[ { "file_name": "Werden et al_2022_Below-ground traits mediate tree survival in a tropical dry forest restoration.pdf.tei.xml", "page_content": "The reassembly of diverse forests is an important component in the fight against biodiversity loss and climate change [1].Moreover, many countries and organizations have committed to develop large-scale forest landscape restoration projects as part of programmes such as the Bonn Challenge [2] and the Trillion Tree initiative (1t.org).It is therefore imperative that appropriate management interventions be applied and targeted to local conditions [3,4].Natural regeneration is often an optimal strategy to promote the re-establishment of native species [5,6], yet in many degraded landscapes, planting trees is a necessary step in the re-establishment of local biodiversity [7].Within such environments, survival rates can be particularly low in the early stages of tree development, and so species selection can be a key determinant of restoration outcomes [8,9].Identifying mixtures of species that can establish and survive within harsh environmental conditions is critical to ensure effective long-term outcomes of active forest landscape restoration efforts.Moreover, increasing plant survival in restoration projects across the globe is important to ensure that resources are efficiently allocated by countries committed to the UN Decade on Ecosystem Restoration (see https://www.decadeonrestoration.org/).\nDespite the importance of appropriate tree species selection for the success of restoration efforts, few studies have characterized the plant traits that are most effective at promoting survival within degraded landscapes.A long history of ecological literature highlights how the performance of plants is contingent upon the expression of traits that can confer a selective advantage under a given set of environmental conditions.A range of morphological, physiological and biochemical functional traits [10] can predict tropical plant growth under different environmental conditions [11,12].Moreover, there have been calls to use functional traits to develop species mixes for tropical forest restoration [9,13].Additionally, integrating local knowledge of species life-history traits (e.g.successional status) when designing restoration interventions can not only improve outcomes, but also ensure that culturally and economically important species are explicitly included in the design process [14].Yet, applying this knowledge to predict the survival of seedlings within a restoration context remains a major challenge as trait expression, survival rates and their interactions can vary significantly at these initial stages of development.In particular, within tropical dry forests-the second largest forested tropical biome [15] and one of the most endangered ecosystems [16]-seedlings are exposed to strong dry seasons within the first year of growth, and survival in the initial years of development is imperative to ensure that restoration is effective in this system [17,18].Globally, tropical dry forests have been extensively deforested [19].It is estimated that 97% of remaining tropical dry forests are threatened by climate change and human activities [20], making them a top-priority for landscape scale restoration efforts.Recent work highlights how physiological leaf traits such as water-use efficiency can be key indicators of seedling survival during initial phases of development in both tropical dry [21] and wet forests [22].As such, the selection of species mixtures with high community-weighted water-use efficiency could potentially improve early survival rates at the community level.Increasing survival at early restoration stages has direct consequences for the development of soil organic matter [23], canopy structure [24] and subsequent recruitment of other native species [25,26].However, by focusing only on the linear correlations with a few plant traits, these studies cannot account for the majority of the variation in species survival rates, or identify the mixtures of trait combinations likely to promote the initial survival of seedlings in the long term.\nGiven the multitude of approaches that plants use to compete for space, light, water and nutrients within heterogeneous environments, it is rare that any single trait can directly predict tree survival rates in a given location [27].However, certain traits can be indicative of general life-history strategies.In particular, the growth rate of trees is a well-described indicator of changes in community assembly over time, with fast-growing resource-demanding species dominating in early successional stages and being gradually replaced by slower-growing species with more efficient resource-use [28].As such, differences between 'acquisitive' (i.e.typically fast-growing species that maximize resource capture and are sensitive to abiotic stress) and 'conservative' (i.e.typically slow-growing, stress-tolerant species with higher resource-use efficiency) functional strategies (sensu [27]) may provide a useful framework for guiding species selection that leads to high establishment rates of planted seedlings within a restoration context.Indeed, in a tropical dry forest field experiment, Gerhardt [29] highlighted that within-species tree seedling survival increased with higher height increment growth rates.However, although this 'acquisitive' growth strategy may promote tree survival within early stages of succession, rapid growth can often come at the expense of more 'conservative' strategies (e.g. that can be critical for species survival in harsh dry conditions).\nIn tropical dry forests, tree growth and survival are strongly limited by seasonal water availability [30], and species survival may be tightly linked to the successful development of deep, robust rooting systems [31].However, while below-ground plant traits have been linked with plant vital rates (e.g.survival and growth) across terrestrial ecosystems, these traits are rarely considered when evaluating plant performance, especially in the context of restoration [32].If investment in fast aboveground growth (i.e.acquisitive strategy) comes at the expense of plant investment in thicker absorptive fine roots and deeper supporting root systems (i.e.conservative strategy), then it may potentially limit seedling survival within tropical dry forest.Yet, until now, no study has explored the relative importance of below-ground traits relative to above-ground traits, or the trade-offs between trait combinations, that may be essential for improving species selection to promote increased survival rates within tropical dry forest restoration.\nIn this study, we examined growth and survival of 14 native tree species (840 seedlings in total) within a tropical dry forest restoration experiment in Costa Rica.These species are common in early successional forests in the region, are inclusive of the most dominant tree families [33] and encompass a wide range of life-history and resource-use traits [34].Across all species, we measured six below-ground and 22 aboveground traits used to place species on leaf, stem and root economic trait spectra so that we could group trait values onto an acquisitive to conservative gradient [35], then examined how interspecific trait variation corresponds to species survival rates over the first 2 years of seedling growth.By examining the interactions between trait complexes, we test the relative importance of above-versus below-ground traits, and examine how interactions between trait combinations mediate seedling survival rates.Last, we investigated if trait syndromes (i.e.fast versus slow [35]) are coordinated across above-and belowground organs, an area of open debate for tree species [36].Ultimately, by taking a broad scope, we aim to identify a clear hierarchy of traits that should be considered in order to promote the survival of planted seedlings within tropical dry forest restoration.", "title": "Below-ground traits mediate tree survival in a tropical dry forest restoration", "id": "0.1", "keywords": [ "ecology", "ecosystems", "plant science biomass allocation", "growth rates", "plant functional traits", "root traits", "trait coordination", "Vertisol" ] }, { "file_name": "Werden et al_2022_Below-ground traits mediate tree survival in a tropical dry forest restoration.pdf.tei.xml", "page_content": "Our study was conducted at Estaci\u00f3n Experimental Forestal Horizontes (10.712N, 85.594W) in \u00c1rea de Conservaci\u00f3n Guanacaste (ACG) in northwest Costa Rica.This tropical dry forest has a mean annual precipitation of 1730 mm, mean annual temperature of 25\u00b0C and a strong five-to six-month dry season from November to May, with little to no rainfall (http://www.investigadoresacg.org/main_eng.html).The study site was previously used for cattle grazing and agricultural crop production for decades, similar to land use in other tropical dry forests across Central America royalsocietypublishing.org/journal/rstb Phil.Trans.R. Soc.B 378: 20210067 [37] and the globe [19].Succession is arrested in the study area, which had been regenerating for approximately 28 years before management intervention, due to intensive use that both compacted the soil and led to nutrient depletion [38].As a result, the study site is dominated by only three tree species, Cochlospermum vitifolium (Bixaceae), Crescentia alata (Bignoniaceae) and Guazuma ulmifolia (Malvaceae) [38], making it particularly species-poor given that more than 60 tree species are observed in nearby forest inventory plots [39].", "title": "Below-ground traits mediate tree survival in a tropical dry forest restoration", "id": "0.2", "keywords": [ "ecology", "ecosystems", "plant science biomass allocation", "growth rates", "plant functional traits", "root traits", "trait coordination", "Vertisol" ] }, { "file_name": "Werden et al_2022_Below-ground traits mediate tree survival in a tropical dry forest restoration.pdf.tei.xml", "page_content": "Here we report new observations of above-and below-ground plant functional traits that complement published data on seedling survival and growth in an experimental species selection trial [38].Our original experiment aimed to determine how above-ground traits and soil amendments can be used to improve restoration outcomes in degraded Vertisols.Vertisols are soils that impede regeneration due to high shrink-swell clay content that leads to cracks in the dry season and flooding in the wet season [40].In brief, in September 2014 we planted approximately 60 individuals of 32 native tropical dry forest tree species (N = 1710 seedlings) common across this landscape [33] and that have a wide range of functional strategies [34].\nSeedlings were grown in an on-site nursery, and we accounted for intraspecific variation by collecting seeds across the ACG from at least three trees per species separated by at least 1 km.\nSeedlings were planted at 1 \u00d7 1 m spacing, with each species planted randomly in rows across four soil amendment (hydrogel, rice hulls, rice hull ash, sand) and two control blocks.Previous results showed that soil amendments did not influence survival or growth of seedlings after 2 years, so we do not consider this factor in the current study.Seedling competition with remnant vegetation (e.g.exotic forage grasses) was minimized by fully clearing around all seedlings with machetes at regular intervals (October and November 2014, and July 2015).We conducted the current study with species (N = 14) that had least three surviving individuals 2 years after planting (table 1).\nWe leveraged data collected in the original experiment to calculate species-level survival and growth rate for 14 species [38].We made an effort to control for the influence of plant size on calculated species-level growth rates by initiating production of all species in the nursery at the same time.Mean planted seedling heights were similar at the species-level both when initially planted (mean: 29.8 cm; s.d.: 13.7) and after 2 years when harvests were made (mean: 72.9 cm; s.d.: 25.6).As such, we effectively compared growth rates between species computed from similarly sized seedlings.We calculated final species-level survival percentages and average relative height growth rate (RGR; ln(cm) d -1 ) after 2 years (two full growing and dry seasons; September 2014-June 2016) on 60 seedlings per species (N = 840).RGR was calculated using the standard equation: (ln[final height]-ln[initial height])/(final day-initial day) [41] and assumptions for this approach were met (i.e.RGR did not slow or approach an asymptote over time).In July and August of 2016, we measured physiological traits and whole-plant structural characteristics in situ, then harvested both above-and below-ground biomass of two-year-old seedlings of each species (electronic supplementary material, figure S1).All measurements were made on three individuals of each species (N = 42 individuals total; see electronic supplementary material, table S1 for above-and below-ground trait list).", "title": "Below-ground traits mediate tree survival in a tropical dry forest restoration", "id": "0.3", "keywords": [ "ecology", "ecosystems", "plant science biomass allocation", "growth rates", "plant functional traits", "root traits", "trait coordination", "Vertisol" ] }, { "file_name": "Werden et al_2022_Below-ground traits mediate tree survival in a tropical dry forest restoration.pdf.tei.xml", "page_content": "Physiological traits related to photosynthesis and water-use were made on individuals in the field prior to destructive harvests.We used an LCi portable photosynthesis system (ADC Bioscientific Ltd, Hoddesdon, UK) to measure maximum photosynthetic capacity at 1200 par (Amax; \u03bcmol CO 2 m -2 s -1 ), transpiration (E; mmol m -2 s -1 ) and instantaneous water-use efficiency (WUE; \u03bcmol CO 2 mmol H 2 O -1 ).Next, we measured the change in leaf water potential (\u03a8) from pre-dawn to midday (\u03a8 diurnal ; MPa) to place species on a drought-tolerator versus avoider spectrum [42].We measured pre-dawn and mid-day \u03a8 on two leaves per seedling with a Scholander pressure chamber (1505D, PMS Instruments, Albany, OR, USA), then calculated \u03a8 diurnal .\nTable 1.The 14 focal tropical dry forest tree species from the restoration experiment, their life-history traits, final survival percentages and relative height growth rates (RGR) over 2 years.Trait abbreviations are as follows: dispersal syndrome (W, wind; A, animal), leaf habit (D, deciduous; EG, evergreen), leaf compoundness (S, simple; C, compound).For all harvested seedlings (N = 3 per species), we quantified a suite of above-and below-ground functional traits related to resource acquisition and stress tolerance.This sample size did not allow for intraspecific comparisons; however, we were able to compare how an extensive suite of above-and belowground traits differed concurrently between many focal species using principal component analysis.In the field, we measured the crown radius (cm) by measuring from the seedling stem to the tip of the furthest leaf, then excavated whole seedlings taking care to harvest the entire root system, including all fine roots (roots \u2264 2 mm diameter).During harvests we measured the maximum lateral rooting extent by measuring from the base of each seedling to the end of the furthest excavated fine root tip (root lateral extent; cm) and maximum depth of the deepest fine root (root depth; cm).No harvested seedlings were observed competing for either above-(i.e.no overlapping canopies) or below-ground resources (i.e.no overlapping root structures) with adjacent seedlings.We washed then scanned all fresh fine roots (absorptive firstand second-order roots \u2264 2 mm diameter; functional classification sensu [43]) to calculate fine root traits for each individual with a transparency scanner (Epson Perfection V800, Suwa, Japan; 2-4 images per individual), placing roots in a clear polycarbonate tray filled with water to ensure no root overlap.We used IJ_Rhizo [44] to calculate mean fine root diameter from scanned images.We then dried roots and calculated specific root length (SRL; m g -1 ) and tissue density (RTD, g cm -3 ) using total fine root length and volume calculated from images.\nOn three fresh leaves per harvested individual (N = 126 leaves total) we measured petiole length (mm) and leaf thickness (mm), then scanned and calculated leaf area (cm 2 ) with IMAGEJ [45].To quantify whole-plant light capture capacity we scanned all fresh leaves of each individual and calculated whole canopy leaf area (total leaf area; cm 2 ) and leaf perimeter (total leaf perimeter; cm).We then dried all leaves at 60\u00b0C to constant weight and calculated specific leaf area (SLA; cm 2 g -1 ) and leaf density (g cm -3 ).We bulked these leaves per individual and transported them to the University of Minnesota, where they were ground and analysed for foliar carbon to nitrogen ratio (C : N) and stable isotopes (\u03b4 13 C; \u2030).\nLast, for each harvested individual, we dried all plant parts (leaves, stems, roots) to constant weight at 60\u00b0C and determined their mass (g), then calculated leaf (LMF), stem (SMF) and root (RMF) mass fractions for each seedling.In final individual-level biomass calculations, we included the mass of all leaves and roots used to quantify all traits above.For all species, we compiled a list of life-history traits from a previous study in the region [34]: dispersal syndrome (wind or animal), leaf habit (deciduous or evergreen), leaf compoundness (simple or compound), nitrogen-fixing status (fixer or non-fixer), seed mass (g) and stem wood density (g cm -3 ).We calculated species-level trait means for statistical analyses (electronic supplementary material, table S2).", "title": "Below-ground traits mediate tree survival in a tropical dry forest restoration", "id": "0.4", "keywords": [ "ecology", "ecosystems", "plant science biomass allocation", "growth rates", "plant functional traits", "root traits", "trait coordination", "Vertisol" ] }, { "file_name": "Werden et al_2022_Below-ground traits mediate tree survival in a tropical dry forest restoration.pdf.tei.xml", "page_content": "In an effort to consider all above-or below-ground traits simultaneously, we performed principal component analysis, using the prcomp function within R statistical software [46].All variables were centred and scaled relative to their means and variances to facilitate interpretation of principal components.We performed two principal component analyses, one using all above-ground traits, and a second using all below-ground traits.This generated two sets of species-level principal components which reflected variation in above-and below-ground traits, respectively.\nWe analysed survival rates at the species level using Bayesian phylogenetic generalized linear models (GLM) to account for species relatedness, via the MCMCglmm function in R [47] and a phylogeny built with V.PhyloMaker [48].Tree species survival rates were logit-transformed prior to analysis, as raw proportional data violate the assumption of homoscedasticity [49].To ensure that results were equivalent to using non-transformed survival values, we compared the logit-transformed Bayesian phylogenetic GLMs to beta GLMs and found that model fits and results were almost identical (electronic supplementary material, figure S2 and table S3).To understand how relative growth rate, above-ground traits, below-ground traits and potential interactions influenced species-level survival rates we fit five statistical models.We focused on predicting species' survival rates rather than growth rates as the goal of our study was to develop a framework for restoration species selection that optimizes initial survival of planted seedlings.The first three models were univariate regressions between survival rate and either relative growth rate, the first principal component (PC1) of aboveground trait variation, or PC1 of below-ground trait variation.Then, we considered potential interactions between relative growth rate and trait principal components by fitting models with relative growth rate, PC1 of above-or below-ground trait variation, and their interaction, respectively.\n(1) Survival rate \u223c above-ground PC1 (2) Survival rate \u223c below-ground PC1 (3) Survival rate \u223c relative growth rate (RGR) (4) Survival rate \u223c RGR + above-ground PC1 + RGR * aboveground PC1 (5) Survival rate \u223c RGR + below-ground PC1 + RGR * belowground PC1\nWe explored how additional principal components of aboveand below-ground trait variation influenced survival rates, but these models performed substantially worse than models fit with the first principal component, and so were excluded from further investigation.Additionally, we performed a 'whole-plant traits' PCA including all above-and below-ground traits, however, PC1 of this PCA only described slightly more variation in survival than above-ground traits PC1, and far less variation than the below-ground traits PC1, so it was excluded from analyses.To determine the extent to which traits were coordinated across plant organs, we calculated pairwise Pearson's correlation coefficients (r) with bootstrapped standard errors and p-values (10 000 samples) between all continuous above-and below-ground functional traits.", "title": "Below-ground traits mediate tree survival in a tropical dry forest restoration", "id": "0.5", "keywords": [ "ecology", "ecosystems", "plant science biomass allocation", "growth rates", "plant functional traits", "root traits", "trait coordination", "Vertisol" ] }, { "file_name": "Werden et al_2022_Below-ground traits mediate tree survival in a tropical dry forest restoration.pdf.tei.xml", "page_content": "Survival to 2 years varied from 7.8% to 90.1% among species.Overall, relative growth rate was the strongest single predictor of tree species survival rate (R 2 adj \u00bc 0:42), indicating that species with faster growth rates typically had higher survival.However, PC1, which captured the greatest amount of variation in either above-or below-ground traits on their own, were poor predictors of tree species survival.However, when relative growth rate and tree species traits were considered in combination (relative growth rate, PC1 of below-ground trait variation and their interaction), variance explained increased by approximately 32% compared to models fit with relative growth rate alone (R 2 adj \u00bc 0:73 versus R 2 adj \u00bc 0:41, respectively).By contrast, models fit with relative growth rate, PC1 of aboveground trait variation and their interaction slightly decreased model R 2 adj relative to a model fit with relative growth rate alone (adjusted R 2 adj \u00bc 0:28, figure 1).To better understand how the PC1s of above-and belowground trait variation are driven by individual traits, we visualized bi-plots of principal component analyses with royalsocietypublishing.org/journal/rstb Phil.Trans.R. Soc.B 378: 20210067 covariate vectors superimposed, as well univariate regressions between four traits that best correlated with PC1 within each analysis (figures 2 and 3; see electronic supplementary material, table S4 for variable contributions to and correlations with PC1).Below-ground trait variation was dominated by a trade-off between either having thick fine roots (high values of fine root diameter) or high biomass allocation to roots (root mass fraction), greater root length per unit root mass (SRL) and higher root tissue density (figure 2).Above-ground trait variation was dominated by a trade-off between low versus high values of total leaf area and perimeter, crown radius and \u03b4 13 C (figure 3).\nTo evaluate the interaction between below-ground traits and relative growth rates, we used the fitted model to develop an interaction plot that allowed us to vary relative growth rate at fixed levels of below-ground PC1.We varied relative growth rate from its 5% to 95% quantile in the dataset, and then set below-ground trait PC1 at either its 5%, 50% or 95% quantile.The 95% quantile of below-ground trait PC1 reflects thick fine roots and deeper maximum rooting depth, but lower root allocation and root tissue density.We refer to the 95%, 50% and 5% quantiles of PC1 as 'conservative', 'average' and 'acquisitive' below-ground trait syndromes, respectively.We observed that the interaction between relative growth rate and below-ground trait variation has a strong effect on estimated tree species survival rates, so much so that trees with otherwise high relative growth rates have much lower survival rates if those trees also have thin fine roots and high overall biomass allocation to roots (figure 4).\nTo examine trait coordination between plant organs, we generated a correlation matrix of pairwise Pearson's correlations between above-and below-ground traits (electronic supplementary material, figure S3).Root traits were largely not correlated with above-ground traits; however, some patterns emerged with respect to biomass allocation.Notably, no above-ground traits were correlated with root diameter, and only one above-ground trait (SMF) was correlated with root depth (i.e.species with deeper roots allocated more resources to stems; r = 0.60).In the few cases where SRL and RTD correlated with above-ground traits, relationships were typically negative; however, species with high RTD generally had high stem wood density (r = 0.69).Species that allocated more biomass below-ground (i.e. higher RMF) had acquisitive leaves (i.e. higher SLA; r = 0.66), narrower crowns (r = -0.74)and allocated less resources to leaves overall (i.e.lower LMF, lower total leaf area and perimeter; r of -0.58, -0.61 and -0.65, respectively).The opposite patterns were observed for species with high root lateral extent (i.e.lower SLA, larger crowns, higher total leaf perimeter and area; r of -0.54, 0.78, 0.75 and 0.76, respectively) and these species also allocated more resources to stem biomass (i.e. higher SMF; r = 0.73).", "title": "Below-ground traits mediate tree survival in a tropical dry forest restoration", "id": "0.6", "keywords": [ "ecology", "ecosystems", "plant science biomass allocation", "growth rates", "plant functional traits", "root traits", "trait coordination", "Vertisol" ] }, { "file_name": "Werden et al_2022_Below-ground traits mediate tree survival in a tropical dry forest restoration.pdf.tei.xml", "page_content": "The selection of tree species remains a major challenge facing restoration projects in tropical dry forest.Given the harsh growing conditions, characterized by prolonged dry seasons, mortality rates can be exceptionally high in the first 2 years following seedling establishment [38,50].Surviving this initial growth phase is key to ensure effective restoration outcomes.In this analysis, only 34% of planted seedlings survived this initial 2-year period.Yet, survival rates of different tree species ranged from 7.8 to 90.1%, highlighting that effective species selection is necessary to maximize the chances of initial seedling establishment when restoring this ecosystem type.\nTo explain this variation in survival rates among our species, we explored a range of above-and below-ground traits.The prominent role of relative growth rate in governing initial survival emerged strongly, explaining over 40% of the variation in survival rates alone (figure 1).Specifically, species able to tolerate conditions in this degraded soil and allocate resources to rapid initial height growth consistently established quickly and tended to survive throughout the initial 2-year growth phase.These findings are consistent with evidence from both tree planting efforts [51] and natural regeneration projects [29] in Neotropical tropical dry forests, highlighting that tropical dry forest seedling growth rates can be tightly and positively linked with survival for the first few years of establishment.Whereas the growthmortality trade-off dictates that fast-growing species typically have the highest mortality rates in closed canopy forest [52], in these high light early successional environments this tradeoff does not appear to hold (also see [53]).Therefore, in early stages of tropical dry forest restoration, acquisitive growth strategies clearly promote rapid establishment and seedling survival.This suggests that integrating fast-growing species that persist naturally at early successional stages into planting mixes is a straightforward approach towards improving restoration outcomes on extensively degraded sites where there is little to no extant canopy cover.Moreover, this pattern was not driven by nitrogen-fixing species included in our study, which can have much higher growth and survival than non-fixers in regenerating tropical wet forest [54] and can have higher growth than non-fixers in tropical wet forest restoration plantings [55].Rather, both fixers and non-fixers were integrated along growth and survival gradients, highlighting that species with both strategies should be considered for species mixes.Indeed, ensuring that species mixes containing a wide range of functional groups with appropriate traits for a given situation can not only restore targeted ecosystem services, but also increase forest resilience to ongoing global environmental change [1].\nDespite the important role of seedling growth rate in governing survival in this system, several species diverged strongly from this linear relationship, and the inclusion of functional traits considerably improved the capacity to predict overall survival rates across species.However, the importance of different traits varied considerably between above-and below-ground characteristics.Although the inclusion of above-ground traits did not significantly ( p > 0.05) affect the model fit, the inclusion of below-ground traits considerably improved the predictive accuracy of our model.Specifically, the final model including relative growth rate, the PC1 for below-ground root traits and their interaction explained 73% of the variation in survival rates across our study species.While our results indicate that below-ground traits are highly predictive of initial survival rates in this system, survival rates over longer timescales may be dictated by different suites of traits, ontogeny [56] and/or cyclical climate cycles (e.g.El Ni\u00f1o and La Ni\u00f1a).This prominent role of belowground traits in predicting initial species survival is consistent with the idea that root investment is critical for plant survival, especially in regions with limited water availability.Indeed, both broad- [57] and fine-scale [58] studies consistently highlight the importance of root investment within arid and seasonally dry regions, as the root : shoot ratio of plants tends to increase in drier conditions.Moreover, in this tropical dry forest, species that allocated more resources below-ground (i.e. higher RMF) allocated less resources to leaf construction (i.e. higher SLA), crown development and total canopy leaf area (electronic supplementary material, figure S3).This tradeoff is consistent with observations from Panama indicating that seedling growth in tropical dry forest was maximized when more resources were invested in roots versus leaves [59].\nThe specific nature of the relationships between root traits and species survival can provide mechanistic insights into their effects on seedling survival.Principal Component 1 for the below-ground traits correlates negatively with root depth, root diameter and root lateral extent and positively with RMF, SRL and root tissue density (figure 2).As such, the negative overall relationship between PC1 and species survival indicates that investing in thicker fine roots, and developing deeper and more laterally extensive root structures can promote species survival within the initial years of seedling growth.These patterns appear to be tightly linked to the optimization of resource capture, and observations demonstrate that early successional tropical dry forest trees in Mexico typically have deep root structures that increase water foraging capacity [60].Moreover, all of the species in this study associate with arbuscular mycorrhizal fungi and species with thicker fine roots typically have high mycorrhizal colonization rates [61], which could directly increase root surface area for nutrient and water uptake.Our findings likely hold across the dry tropics in forests dominated by deciduous species.However, in seasonal tropical forests with communities dominated by evergreen species, which inherently allocate more biomass to roots [62], other suites of traits may better differentiate between species when predicting initial seedling survival rates.Additionally, the low sample size of our seedling harvests may have limited our ability to well resolve trait means for specific species, potentially influencing observed trait syndrome\u00d7growth rate interactions.However, our initial analysis highlights that integrating below-ground traits into restoration design is an important research direction, and further characterizing below-ground trait variation will only help to improve our understanding of this topic.\nWe also found a strong interaction between the effects of root investment and vertical growth rate in predicting overall survival rates.That is, fast-growing species had higher survival if they also had thick fine roots and deep root structures (figure 4).This interaction between root traits and growth rate is counterintuitive as these growth traits are often considered to be associated with distinctly opposing growth strategies.Specifically, fast vertical growth is generally considered to be associated with an 'acquisitive' growth strategy, whereas the investment in thick fine roots, and allocating resources to root structure development are often considered to occur in 'conservative' species.Therefore, to maximize initial survival rates in this tropical dry forest region, and in other tropical seasonal forests with strong dry seasons when little to no rain falls for four months or more, it is essential to select species that exhibit relatively acquisitive above-ground growth, with relatively conservative root structures.Thus, assigning binary resource-use strategies (acquisitive versus conservative) at the whole-plant level may be misleading when attempting to determine the drivers of plant performance in a restoration context.Moreover, below-ground traits were largely uncorrelated with or decoupled from above-ground traits in this system (electronic supplementary material, figure S3).This finding supports growing evidence that coordination across above-and below-ground organs in tree species may be limited [36,63].However, there was some indication that wood and root tissue densities were linked across species, a pattern that has also been observed across a broad range of Neotropical tree species [64].Additionally, the traits tightly coupled with seedling survival in below-ground PC1 (root diameter and RMF; figure 2) were only correlated with above-ground traits in one instance.Thus, our findings suggest that despite the difficulty associated with quantifying below-ground traits, investing time into these efforts and improving our capacity to predict species-level trait values out of sample (e.g.[65]) will most likely improve the design of tropical dry forest restorations.Moreover, large-scale forest biodiversity-ecosystem function experiments have demonstrated that maximizing functional diversity in planted forests promotes higher productivity [66] and increases resilience and ecosystem services [67].Thus, ensuring that planted species mixes are highly diverse, in terms of both above-and below-ground functional attributes, is likely to be important, and using planting mixes with high species richness is a strategy demonstrated to meet this goal [68].", "title": "Below-ground traits mediate tree survival in a tropical dry forest restoration", "id": "0.7", "keywords": [ "ecology", "ecosystems", "plant science biomass allocation", "growth rates", "plant functional traits", "root traits", "trait coordination", "Vertisol" ] }, { "file_name": "Werden et al_2022_Below-ground traits mediate tree survival in a tropical dry forest restoration.pdf.tei.xml", "page_content": "By highlighting the traits that predict tree seedling survival, our analysis provides a path forward for the selection of species to promote survival in tropical dry forest restoration efforts.Selecting species with fast vertical growth rates has the potential to promote rapid seedling establishment when planted into degraded Vertisols, but this effect is only apparent in species that also invest in thicker fine roots, and more extensive root structures.As such, this analysis highlights the importance of considering below-ground root characteristics, relative to above-ground stem and leaf traits, when predicting seedling survival within these seasonally dry ecosystems.Given the typically low seedling survival rates in tropical dry forests, these insights into species selection may potentially be invaluable for promoting the initial establishment of vegetation.By highlighting the key traits that promote seedling survival, our study can serve as a guideline to select species that can tolerate initial planting conditions in tropical dry forest restoration efforts.Data accessibility.All survival and growth data are available from the Dryad Digital Repository: https://doi.org/10.5061/dryad.fd57r[69].All functional trait data and code and data to reproduce analyses and figures are available from Figshare [70].\nSupplementary material is available online [71].", "title": "Below-ground traits mediate tree survival in a tropical dry forest restoration", "id": "0.8", "keywords": [ "ecology", "ecosystems", "plant science biomass allocation", "growth rates", "plant functional traits", "root traits", "trait coordination", "Vertisol" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "The study site is in the inner southern suburbs of Brisbane, Queensland, Australia and is known locally as Oxley Creek Common (27\u00b032\u2032S 152\u00b059\u2032E, Figure S1).It is situated on an alluvial floodplain along a section of Oxley Creek, approximately 2 km from its confluence with the Brisbane River to the north.Rainfall averages 1,058 mm/ year and is typically summer dominant with extended dry periods in late winter and early spring (July-September).Mean daily minimum and daily maximum temperatures are 14.4\u00b0C and 26.4\u00b0C, respectively (station number 40,211;Australian Bureau of Meteorology, 2017).\nOxley Creek Common comprises infrequently inundated pasture and parklands on dermosols.Prior to European settlement, the areas adjacent to Oxley Creek most likely supported gallery notophyll vine forest (Regional Ecosystem 12.3.1;Queensland Herbarium, 2018).While the region receives less rainfall than is typically used to define rainforest in the tropics, this type of vegetation is commonly referred to as riparian subtropical rainforest in Australia.\nEstablished in November 2016, the experiment was embedded within a larger project funded by the Australian Government's National Landcare Programme '20 Million Trees Round II'.Overall, the experiment aimed to test whether the long-term cost-effectiveness of forest restoration could be improved by designing planting mixes based on the functional traits of plants.", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.3", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "A 'trait targets' approach (Laughlin, 2014) was used to generate two experimental assemblages with distinct trait profiles: a fast assemblage comprising more resource-acquisitive species, and a hardy assemblage comprising species with more conservative strategies (see Appendix S1).Each assemblage had 16 species and similar species evenness.Fast and Hardy assemblages each had eight species unique to them and the remaining eight were common to both assemblages, resulting in a total of 24 species.One species (Jagera pseudorhus) could not be sourced in time for planting, leaving 23 species in the present study (Table S1).", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.5", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "The areas available for planting were mainly confined to a 50 m wide strip of alluvial floodplain (Figure S1) that was previously managed as beef cattle pasture.To capture environmental variation across the site, and variation between grazing units (paddocks), a randomized block design that included 10 blocks distributed along the entire extent of the available planting area was implemented.Each block contained two plots to which the two mixes were randomly assigned.\nEach plot was 30 \u00d7 30 m with 400 trees planted in a grid at 1.5 m spacing.Paddocks containing plots were first fenced to exclude livestock, slashed approximately 2 months prior to planting and then sprayed with glyphosate 2-4 weeks prior to planting to suppress excessive grass and forb growth, as is standard restoration practice in the region.\nPlants were sourced in 50 mm tubestock from a number of local nurseries but varied considerably in size within and among species (from 5 to 65 cm shoot heights, mean = 22 cm).The 400 plants per plot were pre-randomized at a separate location in the days before planting.Planting was undertaken between the 31st of October and 18th of November, 2016 at the beginning of the typical springsummer growing season.Holes were pre-dug using motorized augers and a handful of water crystals were placed in the bottom of each hole at the time of planting and mixed with soil.Each plant was watered-in within an hour of planting.Supplementary watering was provided 1-2 times a week until mid-January 2017.Individual plants in each plot were mapped and their heights measured within 2 weeks of planting.", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.6", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "The survival of all plants was assessed 3.5 months after planting.\nPlants were considered dead if they had no visible living leaves, tips or shoots.The height of living plants was measured as the vertical distance from the ground to the live central apex.The exact number of days between planting and monitoring was recorded for each plot and used to calculate daily growth rates.", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.7", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "Maximum temperatures on the day of planting were obtained for each plot (Australian Bureau of Meteorology, 2017) to provide a measure of initial heat stress.Four soil samples were collected from each plot (one in the centre of each plot quarter) at a depth of 2-10 cm, making a total of 80 samples.Samples were passed through a 2-mm sieve and sent to the UQ Soil Analysis Laboratory to measure total ammonium and nitrate, plant-available phosphorus and potassium, as well as soil pH and electrical conductivity.Soil cores were taken immediately adjacent to sample locations to measure bulk density and gravimetric water content (GWC).This was undertaken during a dry period from the 4th of June to the 7th of June when the site had not received rainfall for over 2 weeks.Bulk density rings were 8.5 cm deep, with a radius of 6.5 cm, and were 'trimmed' so that the surface of soil was flush with the ring.Soil cores were weighed immediately in the field.Soil cores were then oven-dried for 48 hr at 105\u00b0C before being re-weighed.Bulk density (g/cm 3 ) was calculated as the mass of oven-dried soil divided by the volume of the soil at field condition, while the GWC was calculated as the weight of water in the core sample at field condition, divided by the dried soil weight (g/g).\nA LiDAR-derived topographical raster was queried to estimate the elevation of individual plants (Source: QLD Government Remote Sensing Centre, Department of Science, Information Technology and Innovation).Plots were located on the raster using GPS points taken at each plot corner.Individual plant elevations were then extracted in using each plant's grid location within plots.From these data, we created two elevation variables.To capture plot-level variation in elevation, we calculated the mean elevation of plants in each plot.We also subtracted the plot-level minimum from each plant's elevation to describe the relative elevation of plants within plots (i.e.whether plants were on mounds or depressions).", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.8", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "Leaf, stem and root traits were measured on well-watered, sunhardened seedlings of each species.Five replicate seedlings were sampled, except for three species that had four replicate seedlings.\nThree new, fully expanded leaves were selected from each seedling for measurement of SLA, leaf dry matter content (LDMC) and lamina area.Leaves with obvious herbivore or pathogen damage were avoided.After being cut from the stem, leaves were immediately weighed and scanned.The one-sided leaf area (mm 2 ) was obtained using ImageJ software, version 2.0.0 (Schneider, Rasband, & Eliceiri, 2012).Area and mass measurements of all leaves included petioles (simple leaves) or petiolules (compound leaves).Leaves were ovendried at 60\u00b0C for 72 hr before being reweighed to obtain the leaf dry mass.SLA was calculated as the one-sided leaf area divided by the leaf dry mass (mm 2 /mg), and LDMC as the oven dry mass divided by fresh leaf mass (mg/g).\nWood density was calculated as dry mass of wood per unit of volume (mg/mm 3 ).One wood sample was cut from the base of each seedling (average stem sample diameter of approximately 4 mm).\nSecondary phloem and bark were removed before volume was determined via water displacement (Perez-Harguindeguy et al., 2013).\nWood samples were oven-dried at 105\u00b0C for 72 hr and then retained at room temperature for 1-2 min before weighing.Total root length and root volume were measured from fresh, washed root material, which was scanned for each seedling and analysed using WinRhizo image analysis software (Regent Instruments Inc.).Following a preliminary root wash, small samples of the finest living roots in each root system were selected and further washed in deionized water.\nAll effort was made to select absorptive roots; however, some samples may have included structural or transport roots.Roots were scanned at 720 dpi using an EPSON Expression 11000XL LA2500 scanner (EPSON).Based on the estimated mean root diameter for each scanned sample, the overall mean (across species) was 0.3 mm, and species' means ranged from 0.193 for Melaleuca bracteata to 0.505 mm for Toona ciliata.Roots were then oven-dried at 60\u00b0C for 48 hr before weighing to obtain their dry mass.Specific root length (SRL) was calculated as root length divided by the root dry mass (m/g) and root tissue density (RTD) was calculated as root dry mass divided by fresh root volume (mg/mm 3 ).\nTraits that were positively skewed were log-transformed (SLA and SRL) or square-root-transformed (lamina area) prior to analyses.", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.9", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "Separate principal components analyses (PCA) were conducted on soil variables and seedling functional traits to assess correlations between variables.The first three components of the soil PCA explained 79% of the measured variation.Soil PC1 (41%) was negatively loaded by pH, and positively loaded by electrical conductance, phosphorus and nitrate, while the soil PC2 (23%) was positively loaded by ammonium and negatively loaded by potassium (Figure S2 and Table S2).Lastly, soil PC3 (15%) was negatively loaded by bulk density and positively loaded by gravimetric water content.The two elevation variables (mean plot elevation and relative elevation) were not included in the soil PCA because they did not directly describe physical or chemical properties of the soil.\nThe first three components of the functional trait PCA explained 83% of the variation in the seven measured functional traits.Trait PC1 (52%) described overall seedling economics and was positively loaded by log-transformed SLA and log-transformed SRL, and negatively loaded by LDMC, RTD and WD (Figure 1; Table S2).Trait PC2 (18%) was negatively loaded by square-root-transformed lamina", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.10", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "Plant performance was measured as the probability of surviving to 3.5 months, and the height growth increments of surviving plants.\nSurvival (1 = alive, 0 = dead) was modelled using a generalized linear mixed-effect model (GLMM) with binomial errors and logit link function.Random effects were included to account for the spatial nesting of plants within plots, plots within blocks and blocks within paddocks.Species was included as an additional random effect to account for the fact that members of the same species are likely to respond more similarly to each other than members of different groups.\nThe survival model included both environmental and plant-related variables as fixed effects.Environmental variables included the maximum temperature on the day of planting (herein maximum temperature), the plot-scale average elevation, the relative elevation of each stem within plots and the first and third soil principal components.We did not include soil PC2 because it was strongly correlated with plot-level mean elevation.We also included twoway interactions between maximum temperature and the soil and elevation variables to allow temperature effects on survival to vary depending on soil characteristics or the elevation of seedlings.The plant-related variables included the initial height of seedlings and the three selected functional traits, log-transformed SLA, squareroot-transformed lamina area and leaf structure (compound = 1, simple = 0).Seedling height was square-root-transformed to reduce the influence of occasional very tall seedlings.We also included a quadratic term for seedling height because exploratory plots indicated that height-survival relationships were not necessarily monotonic.Thus, we included square-root-transformed seedling height and seedling height to fit the quadratic relationship.Two-way interactions between seedling height (linear and quadratic terms) and trait variables were included to allow species' traits to modulate the shape of the height-survival relationships for each species.We also included random slopes for species with respect to seedling height (linear and quadratic terms).\nHeight growth of surviving seedlings was calculated as (height t1i -height t0i )/(t 1i -t 0i ) and expressed as cm/day, where t 0i was the date that seedling i was initially mapped and measured and t 1i is the date that seedling i was re-measured.Plants that were leaning", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.11", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "The overall probability of surviving the first 3.5 months was 0.45.The Hardy mix had a slightly higher overall probability of survival (0.46) than the Fast mix (0.42), but this difference was not significant when assessed in a simple binomial GLMM (p = .063).Individual species survival probabilities ranged from 0.11 (Toona ciliata and Toechima tenax) to 0.87 (Acacia leiocalyx), while growth rates ranged from 0.02 (Toechima tenax) to 0.45 cm/day (Acacia leiocalyx; Table S1).", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.12", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "The survival model had a marginal R 2 of 0.13 (deviance explained by fixed effects alone), indicating that the fixed effects explained (Figure 2a).This interaction indicated that the probability of survival for seedlings planted in compact soils was lower when they were planted on hot days (Figure 3).We found a strong overall hump-shaped (concave quadratic) relationship between the initial height of seedlings and the probability of survival (Figure 2a), and the shape of this relationship was modulated to some extent by functional traits.In particular, interactions between each trait and the quadratic height term were all significant.For species with simple leaves, hump-shaped curves were narrowest for low SLA, small-leaved species and broadest for high SLA, large-leaved species (Figure 4).For species with compound leaves, curves tended to be boarder and have taller optimal heights.To assess how well these fixed-effect relationships matched observed height-survival relationships for each species, we fitted a simple GLMM to each species separately and only included a quadratic height term if it was significant.When modelled separately, only 8 of the 23 species had significant quadratic terms, while most of the remaining species had positive monotonic relationships (Figure 4).\nFurthermore, the conditional R 2 (representing deviance explained by both fixed and random effects) was 0.41, much higher than the marginal R 2 .And as such, height-survival relationships fitted using species-specific random effects matched the observed data much more closely than those using fixed effects (Figure 4).", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.13", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "The marginal R 2 of the growth model was 0.27 (around double that of the survival model despite having fewer terms) and the conditional R 2 was 0.57.The mean elevation of plots was significantly and positively related to height growth (Figure 2b).The interaction between maximum temperature (on planting day) and relative elevation was also significant.This interaction indicated that growth was slower for seedlings planted on hot days, especially if they were planted in high positions within plots, compared to lower positions.\nInitial seedling height and log-transformed SLA also significantly influenced height growth (Figure 2b).Acquisitive species predictably had faster growth than conservative species, but no other trait variable was related to growth.The relationship for seedling height was again quadratic, indicating that seedlings with intermediate initial heights grew faster than smaller and larger seedlings.", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.14", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "Suboptimal outcomes are common in restoration (Hobbs, 2007;Suding, 2011), but are rarely critically evaluated using empirical data or within the context of a randomized experimental design to reveal barriers impeding restoration.We found that survival and growth of planted seedlings were influenced by planting day temperature and soil properties, and depended on initial seedling height and species.\nIn particular, the overall relationship between initial height and survival was hump-shaped, indicating optimal seedling heights between 25 and 35 cm (for the purpose of surviving hot planting conditions).\nSpecies varied in the shape of their height-survival relationships and\nPlot from the survival model showing the interaction between soil PC3 (bulk density) and the maximum temperature on the day of planting.Soil PC3 is plotted against the probability of survival, with separate fitted lines (and associated 95% confidence intervals) for planting days with average maximum temperatures (29.5\u00b0C) and those with above average temperatures (32.5\u00b0C).\nThe underlying points and standard error bars shown for each relationship were calculated from 25 bins of observed binary data F I G U R E 4 Fitted relationships between the probability of survival and the initial seeding height (cm) for each of the 23 species examined in this study.Species are located approximately in trait space formed by log-transformed SLA (x-axis) and square-root-transformed lamina area (y-axis) for species with simple leaves (a) and compound leaves (b).Orange curves are predictions from the full survival model using the trait values of each species and fixed-effect model coefficients.All other explanatory variables were held at their mean values when plotting these 'trait-based' predictions.Red curves are species-specific predictions from the full survival model using the trait values of each species and each species' random-effect coefficients.Black curves are relationships fitted to each species separately in simpler GLMMs with square root (initial seedling height) as the sole fixed effect, unless the quadratic height term was significant in which case the quadratic was fitted (as indicated by *).Grey points are observed binary data that have been jittered to show trends.Black points and standard error bars were calculated from seven bins of observed binary data to indicate how the raw data fit each relationship.The solid portions of curves indicate the limits of observed seedling heights for each study species.The grey envelope shows the 25-35 cm height range.The number of observations used for each species is included in parentheses some of this variation was explained by three easy-to-measure functional traits.Growth was predictably faster for species with acquisitive seedling leaf economics.Environmental variables had significant but weaker effects on survival and growth than seedling heights.", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.15", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "The influence of soil bulk density and micro-topography on early plant performance has been highlighted in previous restoration studies (Cheesman, Preece, Oosterzee, Erskine, & Cernusak, 2018;Douterlungne et al., 2015;Mart\u00ednez-Garza, Campo, Ricker, & Tob\u00f3n, 2016;Mart\u00ednez-Garza et al., 2011).The survival of seedlings in high bulk density soils is likely reduced because compaction can physically restrict root growth (Bassett et al., 2005;Skinner, Lunt, Spooner, & McIntyre, 2009), reduce water infiltration and retention (Chyba, Kroul\u00edk, Kri\u0161tof, Misiewicz, & Chaney, 2014;Ekwue & Harrilal, 2010) and, consequently, limit plant available water (Archer & Smith, 1972).Similarly, seedlings planted on mounds may have reduced water availability as water moves to depressions, a process that is likely exacerbated on high bulk density soils due to reduced infiltration (Ekwue & Harrilal, 2010).It is also well understood that temperature can directly influence the growth and survival of seedlings (Baumber et al., 2017;Close & Davidson, 2003;Dordel, Seely, & Simard, 2011).Importantly, however, effects of these variables might also emerge through interactions.\nIn this study, the effect of hot planting days on survival depended on soil bulk density.While seedlings were watered multiple times immediately following planting, compact soils could have reduced the ability of seedlings to respond to acute heat stress.It is also possible that the combination of heat and compact soils resulted in shallower planting due to planter fatigue, which would have exposed seedling surface roots to extreme temperatures.Previous studies have identified planter effects, often relating reduced growth or survival to a lack of planting experience (Charles et al., 2018;Close & Davidson, 2003;Vogt, Watkins, Mincey, Patterson, & Fischer, 2015).While planting effects may be possible in our study, we did not record information on planter experience nor changes in planting quality and so cannot explore these factors further.\nThe growth of surviving seedlings was significantly slower in lower lying plots, most likely due to temporary waterlogging and associated reductions in water absorption, transpiration and growth rates compared to plants in more elevated plots (Delgado, Z\u00fa\u00f1iga-Feest, & Piper, 2018).", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.16", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "Seedling attributes are well understood to influence initial growth and survival in revegetation settings (Grossnickle, 2012).Many studies have reported that initial seedling height positively correlates with survival and growth of seedlings.Poor performance of small seedlings is generally attributed to the disturbance of root systems during planting (Close & Davidson, 2003;Mart\u00ednez-Garza, Bongers, & Poorter, 2013) or suppression by recolonizing weeds (Cordell et al., 2016;Douterlungne et al., 2015).Instead of the increasing monotonic relationships between seedling height and survival reported in other studies (e.g.Kabrick, Knapp, Dey, & Larsen, 2015), we found strong support for an overall hump-shaped relationship (Figures 2a and4), strongly suggesting intermediate 'optimal' seedling heights for surviving extreme planting temperatures.The reduced survival and growth of larger seedlings may, in part, be due to stock being pot-bound, as 'root circling' in pots reduces anchorage as well as nutrient and water uptake (Allen, Harper, Bayer, & Brazee, 2017;South, Harrisa, Barnett, Hainds, & Gjerstad, 2005).Given that all plants were supplied in 50 mm planting tubes, larger seedlings probably also had lower root:shoot ratios, and thus a reduced capacity to supply water to transpiring foliage under stressful, open pasture situations (Close, Beadle, & Brown, 2005;Grossnickle, 2012).\nIt should be noted that only eight of the 23 species had significant quadratic height-survival relationships when modelled alone.\nThe strong quadratic relationship in the overall survival model occurred because many species with apparently monotonic relationships only included seedlings that spanned the lower half of the height range.When all species were modelled in a single GLMM, information was borrowed from species that spanned the full height range (the effect of partial pooling in multi-level models; Gelman & Hill, 2007), including a number of species with strong hump-shaped relationships.While it would have been ideal for all species to span a similar height range, sharing of information across species during model fitting provided conservative estimates of optimal heights for all species.\nSignificant interactions between trait variables and initial seedling heights explained species' differences in optimal seedling heights to some extent.Low SLA species had a narrower range of predicted optimal heights than those with high SLA, as did species with simple leaves compared to those with compound leaves.However, the result for compound leaves was influenced by two species with very low survival probabilities (Toona ciliata and Toechima tenax; Figure 4) and weakened when these species were removed (not shown).In general, trait effects on optimal heights were small in absolute terms, and most species were predicted to have highest survival with initial shoot heights between 25 and 35 cm tall.\nInterestingly, species with more conservative traits did not consistently have high survival in this study.This is counter to reported trait-vital rate relationships in natural systems (Anderegg & HilleRisLambers, 2016;Markesteijn, Poorter, Bongers, Paz, & Sack, 2011) and a recent trait-based restoration study in the Australian tropics reporting a positive relationship between seedling survival and wood density (Charles et al., 2018).It is possible that seedling mortality due to planting stress, mishandling or a decline in planting quality (as previously discussed) could have obscured trait-survival relationships.More consistent with theoretical and empirical expectations (Shipley, Vile, Garnier, Wright, & Poorter, 2005), the growth of surviving seedlings was strongly and positively related to SLA.Thus, once seedlings overcome stresses associated with planting, commonly measured functional traits such as SLA are effective at predicting species' performance in restoration settings, especially if traits are measured on seedlings instead of adult plants (Gibert, Gray, Westoby, Wright, & Falster, 2016).", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.17", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "This study identified the antagonistic effects of planting in compact soils on hot days, as well as the importance of planting seedlings at optimum sizes (25-35 cm tall) to increase survival during the establishment period.Clearly, planting should be avoided on days with high maximum temperatures.However, it is also recognized that forecasts can be uncertain for the lead-times required to make planning decisions, and the ability to modify decisions upon new information can be limited (Hobday, Spillman, Eveson, & Hartog, 2016).For example, if planting is delayed due to extreme conditions, it may not be possible for nurseries to hold stock for long periods, and it may be difficult to reschedule planting or maintenance teams.Even if stock can be held, seedlings will likely develop taller shoots and thus be more susceptible to transplant shock when eventually planted.Due to the complexity of coordinating projects, and with record-breaking temperatures predicted to occur more frequently in Australia (CSIRO & Bureau of Meteorology, 2015) and elsewhere, it is reasonable to believe that many future restoration projects will experience similar conditions to those presented here.This highlights the need to either make better use of evidence-based strategies that improve planting success when environmental conditions are highly uncertain or develop implementation strategies that enable necessary flexibility to delay restoration projects when adverse environmental conditions arise (Hagger, Dwyer, Shoo, & Wilson, 2018).", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.18", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "We thank the Oxley Creek Catchment Association, and Phil Gunasekara and Chris Jensen in particular, for implementing the project and inviting us to collaborate so enthusiastically.We are also grateful to Carole Bristow, Mary-Lou Simpson and the Friends of Oxley Creek Common for advice and support.Thanks also to the BSU team, Green Army and all those involved in implementing and maintaining the plantings.Finally, thanks to Melissa Fedrigo and John Armston for generously sharing the LiDAR data.AUTH O R S ' CO NTR I B UTI O N S L.S. and J.M.D. designed the experiment, supervised its implementation and collected baseline mapping and plant height data.R.G. conducted all remaining fieldwork with support from L.S. and J.M.D. R.G. and J.M.D. analysed the data.R.G. wrote the first draft of the manuscript with input from L.S. and J.M.D. J.M.D. revised the manuscript for journal submission.", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.22", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "We thank the Oxley Creek Catchment Association, and Phil Gunasekara and Chris Jensen in particular, for implementing the project and inviting us to collaborate so enthusiastically.We are also grateful to Carole Bristow, Mary-Lou Simpson and the Friends of Oxley Creek Common for advice and support.Thanks also to the BSU team, Green Army and all those involved in implementing and maintaining the plantings.Finally, thanks to Melissa Fedrigo and John Armston for generously sharing the LiDAR data.AUTH O R S ' CO NTR I B UTI O N S L.S. and J.M.D. designed the experiment, supervised its implementation and collected baseline mapping and plant height data.R.G. conducted all remaining fieldwork with support from L.S. and J.M.D. R.G. and J.M.D. analysed the data.R.G. wrote the first draft of the manuscript with input from L.S. and J.M.D. J.M.D. revised the manuscript for journal submission.", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.23", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "Data available via the Dryad Digital Repository https ://datad ryad.org/stash/ datas et/doi:10.5061/dryad.3kc37b(Gardiner, Shoo, & Dwyer, 2019).\nJohn M. Dwyer https://orcid.org/0000-0001-7389-5528", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.26", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "Data available via the Dryad Digital Repository https ://datad ryad.org/stash/ datas et/doi:10.5061/dryad.3kc37b(Gardiner, Shoo, & Dwyer, 2019).", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.27", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "John M. Dwyer https://orcid.org/0000-0001-7389-5528", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.28", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "Additional supporting information may be found online in the Supporting Information section at the end of the article.\nHow to cite this article: Gardiner R, Shoo LP, Dwyer JM.\nLook to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration.J Appl Ecol.2019;00: 1-11.https ://doi.org/10.1111/1365-2664.13505", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.29", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Gardiner et al. - 2019 - Look to seedling heights, rather than functional t.pdf.tei.xml", "page_content": "Additional supporting information may be found online in the Supporting Information section at the end of the article.\nHow to cite this article: Gardiner R, Shoo LP, Dwyer JM.\nLook to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration.J Appl Ecol.2019;00: 1-11.https ://doi.org/10.1111/1365-2664.13505", "title": "Look to seedling heights, rather than functional traits, to explain survival during extreme heat stress in the early stages of subtropical rainforest restoration", "id": "1.30", "keywords": [ "bulk density", "land degradation", "pasture", "restoration", "revegetation", "riparian", "tubestock" ] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "Shade tolerant tree species, also known as non-pioneers, are important elements of mature forest structure and composition, representing more than 80% of tree species in tropical forests (Denslow, 1987;Uhl et al., 1988).This guild recovers at a lower rate along succession because abandoned agricultural land in the tropics frequently lacks seed and seedling banks of shade tolerant species, which are typically of late successional status and display low seed dispersal reach (Mart\u00ednez-Garza & Howe, 2003;Mu\u00f1iz-Castro et al., 2012;Suganuma & Durigan, 2015).Since shade tolerant species are more vulnerable to habitat transformation and play a key role in ecosystem functioning, restoration focused on assisting the recovery of this guild in particular is urgently required (Aide et al., 2011;Holl et al., 2000;Mart\u00ednez-Garza & Howe, 2003).\nTropical forest tree species are arranged along a continuum of shade tolerance (Wright et al., 2003), but most studies contrast the attributes and responses of pioneer versus shade tolerant species, with much still to be understood about the complexity that exists within the shade tolerant functional group (Valladares et al., 2016).\nShade tolerance is associated with a conservative-use strategy, featuring biomass and energy conservation traits that maximize survival under the low light conditions of the forest understorey (Kitajima, 1994;Reich et al., 2003;Veneklaas & Poorter, 1998).While shade tolerant species display high survival in the dark conditions of the forest understorey, experimental essays have shown that shade tolerant transplants can also establish under high irradiance in earlysuccessional environments if they arrive via ecological restoration efforts (Cole et al., 2011;Mart\u00ednez-Garza et al., 2013;Pedraza & Williams-Linera, 2003).However, an important variation in survival has been found within the shade tolerant group in disturbed habitats (Ben\u00edtez-Malvido et al., 2005;Camacho-Cruz et al., 2000;Hooper et al., 2002;Mart\u00ednez-Garza et al., 2005;Mu\u00f1iz-Castro et al., 2015).\nDue to the high species diversity and variation in environmental conditions across altered tropical habitats, the development of criteria to select appropriate species for restoration efforts is central to design more effective interventions.\nPlant functional traits have been proposed as a useful tool with which to predict the response of species to environmental conditions in restoration efforts (Mart\u00ednez-Garza et al., 2005, 2013;Pywell et al., 2003).Based on the functional theory that establishes that species traits reflect resource use and life-history trade-offs, plant species can also be selected according to their functional traits with the aim of restoring ecosystem functions (Ostertag et al., 2015;Reich, 2014;Zirbel et al., 2017).Leaf mass area (LMA) is a descriptor of the efficiency of light capture per unit of carbon invested, and is thus a key trait involved in carbon return and balance that influences whole plant performance (Poorter et al., 2009;Reich, 2014).Leaf dry mass content (LDMC), the ratio of leaf dry mass to fresh mass, is a trait associated with drought resistance, presumably because high fibre content increases the hydraulic integrity of leaf veins, improving tissue resistance to the cell damage caused by severe drought (Kursar et al., 2009;Markesteijn et al., 2011).Plants adapted to shaded habitats tend to exhibit a conservative leaf morphological pattern, including high LMA and LDMC, traits that minimize carbon loss through respiration and maximize tissue defence (Reich et al., 2003;Valladares & Niinemets, 2008).These leaf traits favour persistence under low-resource conditions at the expense of low growth rates (Reich et al., 2003).Plant adaptations for coping with low light conditions are generally incompatible with those required for development under high light conditions and vice versa (Wright et al., 2010;Reich, 2014).However, there is limited information about the performance of shade tolerant planted trees in different disturbance settings and the use of LMA and LDMC as criteria for species-site matching to improve restoration effectiveness in highly diverse tropical landscapes.\nIn this study, we examined the performance of eight shade tolerant tree species (six endangered) in restoration plantings in various disturbed environments, including pastures, secondary forests and forests subjected to traditional selective logging, in a tropical montane cloud forest (TMCF) landscape.Restoration interventions are required under a variety of conditions because TMCF landscapes are mosaics of mature, secondary and degraded forest and active and abandoned pastures (Aide et al., 2011;CONABIO, 2010;Mu\u00f1iz-Castro et al., 2012;Pedraza & Williams-Linera, 2003).One of the main causes of TMCF loss has been its transformation to agricultural land (Aide et al., 2011).Such agricultural and livestock production lands are subsequently abandoned, giving rise to secondary forests worldwide (Mulligan, 2011).In remnant forest fragments, unplanned logging, frequently targeting shade tolerant species, contributes to further degradation and the depletion of locally valuable tree species (Ram\u0131\u0155ez-Marcial et al., 2001; R\u00fcger species.Species selection based on LMA could thus improve restoration initiative outcomes: tree species with high LMA present higher survival probability and can be introduced into pastures, secondary forests and selectively logged forests.", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.1", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "enrichment planting, late successional, leaf dry mass content, leaf mass area, pasture, secondary forest, seedling survival, traditional selective logging et al., 2008;Williams-Linera, 2002).Recruitment of many shade tolerant tree species is limited in such disturbed habitats (Mu\u00f1iz-Castro et al., 2012;Ortiz-Colin et al., 2017;Toledo-Aceves et al., 2021).For instance, in Mexico, 60% of TMCF tree species face some degree of threat (Gonz\u00e1lez-Espinosa et al., 2011) and many of these are shade tolerant species for which no information is available regarding their performance under different altered environments.\nWe address two questions across the three altered environments:\n(a) What are the survival and growth rates of seedling transplants of eight shade tolerant tree species?and (b) can LMA and LDMC be used as predictors of seedling survival and growth among shade tolerant species in restoration plantings?Given that some shade tolerant species can experience photoinhibition under high radiation (Agyeman et al., 1999) and that high air temperature and low water availability, as well as high soil compaction, are characteristic of tropical pastures (Williams-Linera et al., 1998;Loik & Holl, 1999), we expected lower seedling survival and growth in pastures than in the secondary and selectively logged forests.Under the high irradiance conditions of pastures where water deficit can be limiting (Mart\u00ednez-Garza et al., 2013;Mu\u00f1iz-Castro et al., 2015), we expected a strong relationship between LDMC and seedling survival.In contrast, relatively high canopy cover is maintained in cloud forests subjected to traditional selective logging (Ortiz-Colin et al., 2017;Toledo-Aceves et al., 2019), as well as in secondary cloud forests (Mu\u00f1iz-Castro et al., 2012, 2015;Trujillo-Miranda et al., 2018).Given that, under low-resource conditions, conservative resource-use seedling traits can be associated with higher survival (Poorter & Bongers, 2006;Wright et al., 2010;Reich, 2014), we hypothesized that higher LMA will be related to higher seedling survival in the shaded understorey of the secondary and selectively logged forests.", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.2", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "The study sites are located in the TMCF of central Veracruz, Mexico, a region of high priority for conservation and restoration due to the high concentration of biodiversity and considerable loss of forest cover (CONABIO, 2010).In the study region, TMCF ranges between 1,200 and 2,200 m a.s.l.The climate is mild and humid throughout the year with three seasons: a wet-warm season (June-October), a humid-cool season (November-March), and a short dry-warm season (April-May).Mean annual temperature in the study sites ranges from 14 to 19\u00b0C, and total annual precipitation between 1,523 and 1,707 mm (based on interpolations calculated by Cuervo-Robayo et al., 2014).TMCF remnant fragments in this region are immersed in a landscape matrix dominated by agricultural land and agroforestry systems (mainly shade coffee plantations; CONABIO, 2010).\nThe study sites comprised four pastures, five secondary forests, and three forests subjected to traditional selective logging, which together represent the most common causes of TMCF transformation (Aide et al., 2011;CONABIO, 2010).Field work permits were secured for all sites.These agreements with the landowners were verbal, as is common in the study region.The study did not require ethical approval.Site locations and characteristics are presented in Appendix ST1.The imbalance in the number of sites was due to the limited number that shared similar management and environmental conditions and in which the landowners were interested in maintaining the restoration plantings.In all pastures, cattle had been recently excluded and the vegetation was dominated by the exotic grass Cynodon plectostachyus (Poaceae).The pastures presented isolated trees and shrubs, some of them remnants of the TMCF.Common tree species include Liquidambar styraciflua (Altingiaceae), Quercus spp.(Fagaceae), Vachellia pennatula (Fabaceae), Trichilia havanensis (Meliaceae) and Turpinia insignis (Staphyleaceae).Scattered trees in pastures are a common element in tropical landscapes, and occur in varying densities and spatial arrangements (Prevedello et al., 2018).\nThe secondary forests, developed in areas previously used as agricultural and agroforestry lands, were between 15 and 40 years of age.Traditional selective logging in the studied sites is of low intensity, focused on the extraction of oaks to produce charcoal and other hard wood species for rural construction needs (Ortiz-Colin et al., 2017;Williams-Linera, 2002), including the species used in this study.\nBased on hemispherical photographs and convex densitometer measurements taken at the study sites (see methodology below), the mean percentage of canopy cover in pastures was 35.5 \u00b1 2.41 (\u00b1 SE; mean values of canopy cover per site are shown in Appendix ST1).In the secondary forests, this value was 90.8 \u00b1 0.3 and in the logged forests it was 95.2 \u00b1 0.2.Important variation in canopy cover occurred within each environment, but particularly in the pastures (minimummaximum: 0%-77%) due to the presence of scattered trees, and in the secondary forests (61%-99%).Lower variation characterized the logged forests (86%-99%).To characterize the soil, six samples were collected across each site and mixed into one compound sample for the following laboratory analysis: pH, soil bulk density, total carbon (C), nitrogen (N) and phosphorus (P) (see Appendix ST1).", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.4", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "Eight native cloud forest tree species were used to establish the plantings, all of which were reported as being shade tolerant sensu Swaine and Whitmore (1988) (Table 1).Seedlings were grown in polythene bags (30 \u00d7 16 cm) with a mixture of forest soil and fine gravel Seedling survival and height were recorded in all the plants a month after planting and again 3 years after.Relative growth rate (RGR) in height was calculated as follows: RGR = (ln\nwhere H is plant height; t is time, and the initial and final measurements are denoted by subscripts 1 and 2, respectively (Hunt, 1982).\nAt each site, 10 seedlings per species were selected at random and one mature undamaged leaf (developed after planting) was collected from each seedling (including the petiole) to quantify LMA and LDMC (following P\u00e9rez-Harguindeguy et al., 2013).Leaves were collected 1 year after planting, which provided information regarding the plant response to the environment.In a previous study, we measured leaf traits under two levels of shade in controlled conditions for the same set of species, with the exception of J. pyriformis Light affects tree seedling performance and leaf traits (R\u00fcger et al., 2009;Poorter et al., 2009); therefore, for each of the 10 seedlings selected, canopy cover was measured using hemispherical photographs at the time of leaf collection.For this purpose, a camera (Canon Eos rebel XSi) equipped with a fisheye lens was set up at height 1.3 m above the ground on a tripod, levelled and orientated northwards.The photographs were analysed using Gap Light Analyzer (Version 2.0.).", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.5", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "A generalized linear model (GLM; binomial family and logit link function) was used to assess the survival of each species as a function of environment.Environment was included as a factor with three categories (pasture, secondary forest, logged forest).The GLMs were performed using the proportion of surviving seedlings per site.A linear mixed effects model (LMM) with a Normal error distribution was used to assess the relative growth rate (RGR) of each species as a function of environment (factor with three categories); site was included as a random factor.\nTo evaluate the power of the functional traits LMA and LDMC as predictors of seedling survival (binomial response variable), a GLM was used including the three environments, irrespective of tree species.In the first full models, environment was included as a factor with three categories (pasture, secondary forest, logged forest), while LMA or LDMC, canopy cover and seedling initial height were included as covariables.The interactions of functional traits with canopy cover or environment and environment with canopy cover were included.Seedling initial height was included because it can affect the establishment of plantings (Ben\u00edtez-Malvido et al., 2005) ).In all cases, the step-wise procedure based on the Akaike information criterion (\u2206AIC > 2) was used for selection of the best fitting model except in the case of tied values, in which case the most parsimonious model was selected (Crawley, 2013).\nTo evaluate relative growth rate (RGR) as function of the leaf traits (LMA or LDMC), LMM including the three environments were used, irrespective of tree species, in a similar manner to the procedure described above.In these models, site was included as a random factor.Since the interactions environment \u00d7 LMA and environment \u00d7 LDMC were significant (p < 0.05; Appendix ST4), a separate model was fitted for each environment in order to better understand the relationships between leaf functional traits and species growth.Models included the leaf traits (LMA or LDMC), canopy cover and their interaction (LMA \u00d7 Canopy or LDMC \u00d7 Canopy) in each environment, with site included as a random factor.The best fitting model was selected based on the criteria described above.\nAll analyses were carried out in R with the MASS and nlme libraries (R Core Team, 2019).", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.6", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "Overall, proportion of survival was highest in logged forests (0.875 \u00b1 0.028), followed by secondary forests (0.798 \u00b1 0.034) and finally pastures (0.624 \u00b1 0.055) after 3 years.Six species presented high survival (>0.70) across all the environments (Figure 1).\nSurvival was significantly lower in the pastures than in the secondary or logged forests in seven species (Figure 1; Appendix ST5).Only\nFraxinus uhdei, Juglans pyriformis and Magnolia vovidesii displayed significantly lower survival in secondary forests than in logged forests (Figure 1).Juglans pyriformis had the poorest performance, with the lowest survival of all species.\nIncluding all species, RGR was highest in the logged forests (0.290 \u00b1 0.01 cm cm -1 year -1 ), intermediate in secondary forests (0.261 \u00b1 0.01 cm cm -1 year -1 ) and lowest in pastures (0.159 \u00b1 0.03 cm cm -1 year -1 ).Significantly lower RGR was displayed in Oreomunnea mexicana in the pastures compared to the secondary and logged forests (Figure 1; Appendix ST5).A tendency for lower RGR in the pastures could be observed for other species but, probably due to high variation within the environments, the differences were not significant.The RGR of J. pyriformis in pastures was not included in the analysis because growth data was available for only three individuals.The average height values reached per species after 3 years are presented in Appendix ST2.", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.8", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "LMA was a strong predictor of seedling survival in the three environments (Table 2).LDMC was also significantly related to survival (Table 2).\nIn the three environments, the species with higher LMA displayed higher probability of survival and, those species with higher LDMC also had a higher probability of survival (Figure 2; Appendix SF6).In addition to the functional trait LMA, canopy cover contributed to a better model fit in the pastures and secondary forests (Table 2).However, canopy cover had no significant effect on the logged forest models.For the models that included LMA as a predictor of seedling survival, initial tree height was not significant in the pastures, but contributed to a better model fit in the secondary and logged forests.In the models that included LDMC, tree height also contributed to predicted survival in the two environments.The magnitude of this contribution is much lower compared to that of the functional traits (see regression coefficients in Table 2).Models with LMA as a predictor of survival explained more variation (Deviance: 27%-68%) than models with LDMC (21%-50%).\nSelection of models based on \u2206AIC can be found in Appendix ST4.\nLMA was also a predictor of RGR in the pastures and secondary forests, but not in the logged forests (Table 3).In the pastures, lower RGR was related to higher LMA.In contrast, the opposite relationship was found in the secondary forests (Figure 3).Lower RGR was related to higher LDMC in the pastures but no significant relationship was found in the secondary and logged forests (Table 3).Canopy cover contributed to a better fit for models that included LMA as a predictor of RGR in pastures and logged forests, but had no significant effect on secondary forests (Table 3; Appendix ST4).For models that included LDMC as a predictor of RGR, canopy cover also contributed to a better model fit in the three environments (Table 3; Appendix ST4).", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.9", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "Our results regarding the relationship between leaf traits and seedling performance within the shade tolerant functional group support a higher probability of survival in species with higher LMA under different disturbed environments.LDMC is also linked to seedling survival among the shade tolerant species, but to a lower extent than LMA.Overall, the high seedling survival found supports the potential of the studied species for use in restoration plantings, in particular under the canopy of secondary and selectively logged forests.\nWhile conservative leaf attributes were linked to higher survival, this occurred at the expense of low growth under the conditions of the pastures, thus potentially reducing the establishment success of species of high LMA and LDMC in these altered environments.", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.10", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "The overall survival found after 3 years (76%) surpassed the average survival of 62% reported in a metanalysis of restoration seedling plantings in varied ecosystems (Palma & Laurance, 2015).\nImportant interspecific variation in transplant performance occurred among environments.However, due to differences among environments in terms of the size of transplants, time of planting and site characteristics (such as orientation, topography, soil conditions and prior land use) comparisons of performance among environments should be treated with some caution.The high variation in survival and growth in the pastures not only denotes the differential responses of the species to the environment, but also highlights the important heterogeneity present within and among the pastures, probably as result of variation in the soil conditions and in grass and canopy cover.Overall lower survival and growth in the pastures, in comparison to the secondary and logged forests, denotes that while most of the shade tolerant species were capable of resistance, they were probably stressed due to the extreme environmental conditions, such as high radiation and temperature accompanied by lower soil water availability, soil compaction, as well as the dominance of exotic grasses as reported for tropical pastures (Ben\u00edtez-Malvido et al., 2005;Loik & Holl, 1999;Williams-Linera et al., 1998).Plants in high irradiance conditions have lower leaf area and greater investment in roots and stem support in order to meet the greater demand for nutrients and water imposed by greater transpiration (Poorter et al., 2019), although compacted soil can limit root expansion (Alameda & Villar, 2008).\nShade tolerant tree seedlings exposed to high solar radiation can also experience photoinhibition, which involves a reduction F I G U R E 1 Survival proportions (top) and relative growth rate (RGR; bottom) of cloud forest tree seedlings in restoration plantings established in pastures (white), secondary forests (black) and forests subjected to selective logging (grey).Values are mean \u00b1 standard error.Different letters denote significant differences within species among environments (p < 0.05).ns, not significant of light use efficiency in photosynthesis (Agyeman et al., 1999;Lovelock et al., 1994).Previous studies also report lower survival of shade tolerant transplants in pastures than in young secondary and mature tropical forests (Alvarez-Aquino et al., 2004;Ben\u00edtez-Malvido et al., 2005;Mu\u00f1iz-Castro et al., 2015).Nevertheless, the overall 62% survival recorded in the pastures after 3 years in our study is an indicator of an acceptable restoration success (Palma & Laurance, 2015) in this harsh environment, thus supporting the use of transplants of shade tolerant species in early successional environments to assist the recovery of this vulnerable group and its associated functions.\nIn the pastures, J. pyriformis in particular displayed very high mortality indicating that this species is unsuitable for these environments.These results confirm the findings of previous experiments in which J. pyriformis attained poor survival in open sites in comparison to seedlings under the canopy of Pinus patula plantations (Avenda\u00f1o-Yanez et al., 2016).However, they contrast with the c. 60% survival of J. pyriformis reported after 20 months in a recently abandoned pasture (Pedraza & Williams-Linera, 2003).Oreomunnea mexicana and S. contrerasii also presented very low RGR in the pastures.This indicates that, although they were able to persist, the conditions were unfavourable for their development.In contrast, the similar performance across environments in M. vovidesii, Q. germana, Q. sartori and U. mexicana support their potential for establishment in varied disturbed habitats.\nThe range of conditions presented by the secondary forests is more favourable than those in pastures, where the developing canopy cover reduces intense radiation and buffers temperature and humidity (Alvarez-Aquino et al., 2004;Mu\u00f1iz-Castro et al., 2015), resulting in higher survival and RGR.However, in comparison to mature forests, secondary forests are highly heterogeneous and have a higher tree density and dominance of the herbaceous strata (Kappelle et al., 1996;Mu\u00f1iz-Castro et al., 2012;Trujillo-Miranda et al., 2018), factors that could be less favourable for tree early establishment.In the traditional selectively logged forests, the highest survival and RGR values were reached for most species.While lower growth rates could be expected in the low light environment of the forest understorey (Poorter & Bongers, 2006), the overall higher survival and growth rates found in the logged forests denote the high potential for successful establishment for all species under the prevalent conditions, which tend to be comparable to those of mature forests.Oscillations in humidity and temperature are also buffered in the understorey, which can have an important influence on plant carbon balance (Valladares et al., 2016).Interestingly, some species that presented high survival in the shaded logged forests also pre-", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.11", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "Our results support that LMA is a strong predictor of seedling performance within the shade tolerant guild in disturbed habitats, ranging from recently abandoned pastures to secondary and selectively logged forests.Despite the smaller range of trait values expressed in the understorey of the selectively logged forests, a strong pattern emerged in which the species of higher LMA displayed higher survival.\nIn a previous study, LMA was also found to be a strong predictor of seedling survival among 13 shade tolerant tree species along an elevation gradient in the TMCF understory (Toledo-Aceves et al., 2019).\nUnder the low light conditions in the forest understorey, conservative resource-use seedling traits can be associated with higher survival (Poorter & Bongers, 2006;Reich, 2014).Additionally, our results denote that this conservative leaf trait provides resistance to the seedlings established in the transformed habitats.Indeed, higher LMA was also linked to higher survival in the pastures.Higher LMA is also associated with increased longevity and better protected leaves with lower nutrient concentrations, which lead to higher plant persistence (Poorter et al., 2009), but this leaf investment strategy occurs at the expense of lower growth in the pastures.Leaves with high LMA constrict potential growth under high irradiance due to lower light interception and less efficient carbon gain (Jantse-ten Klooster et al., 2007;Kitajima, 1994).Thus, a negative relationship between LMA and seedling growth rates has been widely documented (Poorter & Bongers, 2006;Poorter et al., 2009;Wright et al., 2010).Overall, our results are in line with previous studies showing the significant role of LMA in plant performance under controlled and natural forest conditions (Kitajima, 1994;Poorter & Bongers, 2006;Wright et al., 2010).\nHowever, most studies have contrasted pioneer versus shade tolerant species, leaving an important gap in the information regarding differential performance within the shade tolerant functional group and, in particular, our results add to previous findings by providing support for the inclusion of shade tolerant species with high LMA in early successional habitats for restoration purposes.\nPlastic adjustments in LMA could contribute to acclimatization to contrasting environments.Intraspecific LMA values were higher in the pastures and lower in the secondary forests, with the lowest values occurring in the selectively logged forests.Seedlings in the pastures were continuously exposed to high radiation and it is known that plant individuals exposed to high radiation display increased LMA (Poorter et al., 2009).Under high irradiance, a greater photosynthetic biomass per unit of leaf area enhances photosynthetic interception is augmented by increasing the area per unit of leaf biomass (Gratani et al., 2006;Poorter et al., 2009).While shade tolerant species display low plasticity in shaded environments (Toledo-Aceves et al., 2019;Valladares & Niinemets, 2008), the plastic leaf response to the varying levels of solar radiation present across the disturbed habitats may play a role in their establishment success.\nSupporting our prediction, we found a relationship between LDMC and seedling survival under the high irradiance conditions of pastures where water deficit can be limiting, as was reported in non-pioneers planted in a pasture in prior tropical lowland rain forest (Mart\u00ednez-Garza et al., 2013).LDMC also had a significant relationship with survival in the secondary and logged forests.Higher LDMC is associated with thicker, tougher leaves, which can be better protected from the damage caused by drought and herbivory and is also associated with longer leaf lifespans (M\u00e9ndez-Alonzo et al., 2012;P\u00e9rez-Harguindeguy et al., 2013).Well-defended and long-lived leaves allow plants to maintain a positive carbon balance, enhancing plant survival (Coley & Barone, 1996;Poorter & Bongers, 2006;Reich et al., 2003).\nOverall, the models that included LMA accounted for a greater variation in seedling survival, compared to those that included LDMC.LMA can therefore be considered a better predictor of seedling performance for shade tolerant species in cloud forest restoration plantings.\nThe canopy cover had a stronger influence on seedling survival and growth within the pastures than in the secondary forests, but no effect in the logged forests.Canopy cover plays an important role in the establishment of tree seedlings in early successional habitats.The development of canopy cover as forest succession progresses acts to reduce light availability in the understory, directly affecting tree recruitment (Bazzaz & Pickett, 1980).The absence of or low canopy cover in pastures leads to high levels of irradiance, higher variation in temperature, lower humidity and competition with grasses, which can limit establishment of shade tolerant species.However, our results indicate that most of the species introduced as 1-to 2-year-old seedlings could establish.The presence of scattered trees and shrubs contributes to a high heterogeneity in tropical pastures and can provide favourable conditions as they may act as nurse species and regeneration nuclei (Douterlungne et al., 2015;Garc\u00eda-Orth & Mart\u00ednez-Ramos, 2011).The higher establishment success of the introduced seedlings under the canopy of the secondary forests indicates the occurrence of facilitation by the secondary vegetation of forest development into advanced successional stages (Brown & Lugo, 1990).The small variation in canopy cover in the selectively logged forests may have led to the lack of a significant effect of this structural variable on seedling performance.In the shaded understorey, growth is light limited (Valladares et al., 2016), which could explain the small predictive power of the leaf traits studied on seedling growth in this environment.\nInitial seedling height had apparently little influence on survival, which is contrary to that found in previous reports (Charles et al., 2018;Comita et al., 2009).Indeed, height at time of planting had no effect on survival in pastures when considering LMA, and made only a small contribution to explaining survival in secondary and logged forests.The seedlings might have been tall enough to overcome the limiting factors in the studied sites, but we did not further explore the effect of height since the seedlings were all within a small range of this parameter.These results also suggest that other factors are more determinant for the outcome, and support the predictive nature of LMA for tree seedling performance.", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.12", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "An overarching goal in ecological restoration is to re-establish biodiversity and ecological functions, for which the achievement of Valencia.We thank Victor V\u00e1squez, M.A. Garc\u00eda-Hern\u00e1ndez, Claudia Gallardo, Javier Tolome, Marichu Peralta, Mart\u00edn SanGabriel, Carlos Iglesias, Karina Osorio and Carlos Flores for their help in the field, and Graciela S\u00e1nchez for format edition.\nRosario Landgrave calculated temperature and precipitation for the study sites and produced graphics for the figures.K. Macmillan edited the text.We thank the anonymous reviewers for their suggestions to a previous version.", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.13", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "The authors declare that there is no conflict of interest.", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.14", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "Grant/Award Number: CB2014-01/238831; Instituto de Ecolog\u00eda A.C., Grant/Award Number: 20030-11218", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.21", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "Grant/Award Number: CB2014-01/238831; Instituto de Ecolog\u00eda A.C., Grant/Award Number: 20030-11218", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.22", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "Data available via the Dryad Digital repository https://doi.org/10.5061/dryad.q2bvq83mj (Toledo-Aceves et al., 2022).\nTarin Toledo-Aceves", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.23", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "Data available via the Dryad Digital repository https://doi.org/10.5061/dryad.q2bvq83mj (Toledo-Aceves et al., 2022).", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.24", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "Tarin Toledo-Aceves", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.25", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "Additional supporting information may be found in the online version of the article at the publisher's website.", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.26", "keywords": [] }, { "file_name": "Toledo-Aceves et al. - Leaf functional traits predict shade tolerant tree.pdf.tei.xml", "page_content": "Additional supporting information may be found in the online version of the article at the publisher's website.", "title": "Leaf functional traits predict shade tolerant tree performance in cloud forest restoration plantings", "id": "2.27", "keywords": [] }, { "file_name": "Charles - 2018 - Plant Functional Traits and Species Selection in T.pdf.tei.xml", "page_content": "The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The data presented here was from research that was funded by the Australian Research Council (LP0989161).", "title": "Plant Functional Traits and Species Selection in Tropical Forest Restoration", "id": "3.4", "keywords": [ "functional traits", "restoration", "seedling growth", "succession", "tropical landscapes" ] }, { "file_name": "Charles - 2018 - Plant Functional Traits and Species Selection in T.pdf.tei.xml", "page_content": "The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The data presented here was from research that was funded by the Australian Research Council (LP0989161).", "title": "Plant Functional Traits and Species Selection in Tropical Forest Restoration", "id": "3.5", "keywords": [ "functional traits", "restoration", "seedling growth", "succession", "tropical landscapes" ] }, { "file_name": "Charles - 2018 - Plant Functional Traits and Species Selection in T.pdf.tei.xml", "page_content": "The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.\nLachlan S. Charles http://orcid.org/0000-0003-0055-2510", "title": "Plant Functional Traits and Species Selection in Tropical Forest Restoration", "id": "3.6", "keywords": [ "functional traits", "restoration", "seedling growth", "succession", "tropical landscapes" ] }, { "file_name": "Charles - 2018 - Plant Functional Traits and Species Selection in T.pdf.tei.xml", "page_content": "The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.", "title": "Plant Functional Traits and Species Selection in Tropical Forest Restoration", "id": "3.7", "keywords": [ "functional traits", "restoration", "seedling growth", "succession", "tropical landscapes" ] }, { "file_name": "Charles - 2018 - Plant Functional Traits and Species Selection in T.pdf.tei.xml", "page_content": "Lachlan S. Charles http://orcid.org/0000-0003-0055-2510", "title": "Plant Functional Traits and Species Selection in Tropical Forest Restoration", "id": "3.8", "keywords": [ "functional traits", "restoration", "seedling growth", "succession", "tropical landscapes" ] }, { "file_name": "Ostertag et al. - 2015 - Using plant functional traits to restore Hawaiian .pdf.tei.xml", "page_content": "Ecosystem restoration efforts are carried out by a variety of individuals and organizations with an equally varied set of goals, priorities, resources and time-scales.Once restoration of a degraded landscape or community is recognized as necessary, choosing which species to include in a restoration programme can be a difficult and valueladen process (Fry, Power & Manning 2013;Jones 2013).Species choice in restoration is often carried out with limited ecological information, particularly in regard to species interactions, successional processes and resource-use patterns.Selecting species can be particularly problematic in systems where there is no available baseline data on historical communities, or when restoration to a historic state is not feasible for ecological, logistic or economic reasons.In such cases, it may be preferable to focus on restoring site 'functionality' rather than returning to a historic baseline composition.We present a method for species selection in restoration, based on the collection of plant functional trait data.Using this method, managers can develop species mixtures with desired properties, including expected predictions of interspecific interactions and potential changes in biotic and abiotic conditions.\nTo illustrate this approach, we present a case study in Hawaiian lowland wet forests (HLWF) in which plant species for a restoration project were chosen based on their functional traits, in order to help land managers achieve their restoration goals while at the same time allowing researchers to better understand invasion resistance and ecosystem functioning.In our case, our choices led to the development of hybrid ecosystems including both native and introduced species.However, the approach that we present is not limited to novel or hybrid ecosystem creation, because the candidate species exam-ined and functional traits measured are determined by the user.\nWhereas restoration usually implies a return to historic conditions, there is also growing attention to what some authors have called 'intervention ecology' (sensu Hobbs et al. 2011).This view emphasizes maintaining ecosystem services and functions (Hobbs et al. 2011).The contrast between these frameworks has been widely debated in the literature, and we do not intend to advocate the merit of one view over the other here.Rather, we present the logic behind a functional trait approach, describe why its use is feasible in a Hawaiian lowland forest and present a stepby-step approach to the method that can be applied to a wide variety of ecological systems.While we readily acknowledge that in our study system, we cannot return to a pre-human state, we still consider our approach 'restoration' in a broad sense.", "title": "Using plant functional traits to restore Hawaiian rainforest", "id": "4.0", "keywords": [ "community assembly", "functional restoration", "Hawaiian lowland wet forest", "hybrid ecosystem", "invasion", "multivariate trait space", "tropical forest restoration" ] }, { "file_name": "Ostertag et al. - 2015 - Using plant functional traits to restore Hawaiian .pdf.tei.xml", "page_content": "Functional trait theory holds that characteristics or traits of each species reflect their resource use and life-history trade-offs (Reich 2014).A body of evidence shows that plant traits vary continuously, in predictable ways, along resource availability gradientssuggesting that species' functional traits can be linked to ecosystem properties and to ecosystem services (Lavorel 2013).Many studies have suggested that a robust image of a species' functional profile can be obtained by considering traits related to resource acquisition (e.g.foliar nitrogen, leaf area), resource limitation (e.g.midday leaf water potential, d 13 C-integrated water-use efficiency, leaf mass per area), reproductive investment (e.g.height, seed mass, dispersal type) and resource allocation patterns (e.g.leaf mass per area, specific root length, wood density) (Drenovsky & James 2010;Douma et al. 2012;Sonnier et al. 2012;Fry, Power & Manning 2013).For restoration, selecting species with certain functional trait values can influence species interactions (including competition) and ecosystem properties (Suding et al. 2008).For example, if creating a more fire-tolerant community is necessary, practitioners could select species that increase fine-fuel loads or species whose presence will favour succession from forest to grassland.Other considerations relevant to managers include importance to local wildlife, ability to serve as nurse logs, nitrogen-fixing properties, etc.\nRestoration in a Hawaiian contexttesting ideas in a model system\nThe Hawaiian archipelago offers a unique setting for testing novel approaches to applied ecological questions.On one hand, Hawai'i is considered a model system for ecological research because of the combination of welldefined biotic and abiotic gradients with extreme examples of adaptive radiation (Vitousek 2004).On the other hand, like many isolated oceanic islands, the Hawaiian Islands have been extremely vulnerable to invasion and anthropogenic disturbance.The combination of simplicity and susceptibility presents a distinctive platform for conservation research.\nHawaiian lowland wet forests in particular present serious challenges for restoration and conservation.This forest type only exists as remnant patches, is always populated by invasive plant and animal species, and occurs near human habitation, where non-native propagule pressure is likely to be high (Zimmerman et al. 2008).In such ecosystems, it is important to recognize that not all non-native species are equally problematic.For example, many Polynesian introductions in Hawai'i have persisted in the landscape for up to 1000 years without becoming invasive.In addition to being culturally important, many non-invasive, non-native species have functional trait values, which are not present in the native flora.Breadfruit and coconut are two species that fit both of these criteria: they are not classified as invasive, and they have large leaves and seeds, unlike most native plants.From a restoration perspective, we aimed to test the hypothesis that designing communities with greater diversity of functional trait expression will lead to more invasion-resistant communities (Funk et al. 2008;Drenovsky & James 2010).To do so, we examined the functional traits of species using multivariate analysis; the technique projects each species to a specific x,y location in 'trait space,' which should reflect that species' functional profile.\nOur restoration experiment was carried out at the Keaukaha Military Reservation (KMR) within the municipality of Hilo.This site retains a HLWF, which has a canopy that contains significant numbers of native tree species; however, it also has a long history of disturbance, and invasive trees represent up to half of the basal area (Ostertag et al. 2009).A previous removal experiment in this forest has shown that removal of invasive species alone is not enough to get a native forest back, despite its positive influence on seedling recruitment of native species (Cordell et al. 2009;Ostertag et al. 2009).Given current conditions, the forest is poised to lose most of its native species and become like most of the forest patches remaining in the lowlandsalmost exclusively non-native dominated.\nThe trait-based method we used employs five steps.We illustrate the method's use in HLWF; however, it need not include the use of non-native species, and it is exportable to other ecosystems.\nLand managers at KMR are charged with specific targets developed by the US military.These are as follows: (i) conserving and encouraging the regeneration of native biodiversity at the site, (ii) controlling invasive species and (iii) encouraging carbon storage on the landscape.Because restoring this area to an all-native ecosystem is no longer economically feasible, we elected to create hybrid ecosystems.Our experimental communities were assembled using species whose combined functional trait profiles addressed these three management goals to meet restoration objectives of the main stakeholders.", "title": "Using plant functional traits to restore Hawaiian rainforest", "id": "4.1", "keywords": [ "community assembly", "functional restoration", "Hawaiian lowland wet forest", "hybrid ecosystem", "invasion", "multivariate trait space", "tropical forest restoration" ] }, { "file_name": "Ostertag et al. - 2015 - Using plant functional traits to restore Hawaiian .pdf.tei.xml", "page_content": "We had two main considerations for choosing the traits on which to base our species selection.First, we wanted to estimate the 'functional role' or 'strategy' (sensu Reich 2014) of different species within HLWF environments, and secondly, we wanted to identify specific traits that would address our restoration objectives.We first looked at the overall distribution of HLWF species in trait space and then focused on how different species expressed particular traits that were relevant to our restoration objectives.\nThe functional traits measured (Table 1) were derived from literature and field data.We included traits that are informative in the ecological context of HLWF.For example, knowing that light (rather than nutrient or water availability) is the primary limiting resource for native species in this forest (Ostertag et al. 2009), we included a suite of traits related to light capture (as well as canopy and understorey architecture; including leaf : petiole ratio, plant height and canopy shape).Many of the traits we measured are informative of more than one aspect of a plant species' ecological strategy, and some traits can be used as proxies for others, which are more difficult to measure.For example, we included both maximum plant height and seed mass as proxies for dispersal distance (see Thomson et al. 2011).", "title": "Using plant functional traits to restore Hawaiian rainforest", "id": "4.2", "keywords": [ "community assembly", "functional restoration", "Hawaiian lowland wet forest", "hybrid ecosystem", "invasion", "multivariate trait space", "tropical forest restoration" ] }, { "file_name": "Ostertag et al. - 2015 - Using plant functional traits to restore Hawaiian .pdf.tei.xml", "page_content": "The potential species pool included native species thought to have been present at the site historically, native species currently found at the site and non-native species already found in the region that are believed to pose low invasion risk.To address the latter, we based our decisions on the Hawaiian Weed Risk Assessment score (Daehler et al. 2004; see http://www.botany.hawaii.edu/faculty/daehler/wra/).In a past study at KMR, woody species richness was nine native species and 10 introduced species (Zimmerman et al. 2008).\nDue to a lack of formal records of species occurrence and poor pollen preservation, it is difficult to ascertain which species would have been found historically in any given area of HLWF.Because we suspect that some native species once found at KMR have been locally extirpated, we cast a wide net in order to determine which species could potentially be used for restoring the forest.We included native species with similar environmental requirements that could have occurred in HLWF, based on historical range descriptions by Wagner, Herbst & Sohmer (1999) and previous field experience.\nOnce the list of potential species was compiled (including 17 non-native and 19 native species), we eliminated species which were too difficult to sample or propagate, and species for which sufficient data were not availablereducing our pool to 16 native species and 15 non-native species.Other limiting factors that practitioners should consider include economics (e.g.cost of seeds, plants, labour or time), logistics (e.g.availability of species, project or budget timelines), resilience to climatic change or disturbance regimes, as well as the goals and expectations of stakeholders.", "title": "Using plant functional traits to restore Hawaiian rainforest", "id": "4.3", "keywords": [ "community assembly", "functional restoration", "Hawaiian lowland wet forest", "hybrid ecosystem", "invasion", "multivariate trait space", "tropical forest restoration" ] }, { "file_name": "Ostertag et al. - 2015 - Using plant functional traits to restore Hawaiian .pdf.tei.xml", "page_content": "We sampled plant traits across the full range of conditions in which HLWF is found regionally in order to account for both site and environmental heterogeneity.In total, we sampled traits at 25 sites throughout east Hawai'i Island, in addition to using existing data from the literature.By far, the most time-consuming and effortconsuming steps in making species choices by using traits are creating the potential species pool and collecting trait data.However, compilations of trait data are increasingly common world-wide.A variety of data sets are available either directly from researchers and agencies involved in the region (e.g. the Australian Virtual Herbarium) or from collaborative data base compilations (e.g.TRY (http://www.try-db.org),DRYAD (http://datadryad.org/) or LEDA (http://www.leda-traitbase.org/LEDAportal/).\nOnce the trait data were assembled, we grouped the information for each trait into quartiles in order to account for both analytical constraints and the nature of trait data.Because we considered both multiple traits and multiple species, making meaningful comparisons was often challenging.For example, measures of seed mass spanned many orders of magnitude from <0\u00c1001 to 576 g, while the maximum difference in leaf thickness measured no more than 0\u00c14 mm (Table 1).Furthermore, sometimes it is necessary to compare ratios and categories to numerical data.The use of quartiles places more importance on the relative differences in trait values.This approach emphasizes species which are 'outliers' rather than species that display the 'middle range' (i.e. the most or least effective water use vs. the species whose values are nearest to the group mean).This type of data manipulation is appropriate to the multivariate ordination analysis we later employ.", "title": "Using plant functional traits to restore Hawaiian rainforest", "id": "4.4", "keywords": [ "community assembly", "functional restoration", "Hawaiian lowland wet forest", "hybrid ecosystem", "invasion", "multivariate trait space", "tropical forest restoration" ] }, { "file_name": "Ostertag et al. - 2015 - Using plant functional traits to restore Hawaiian .pdf.tei.xml", "page_content": "We used a principal components analysis (PCA) to show how the selected species are arranged, relative to one another, in trait space and to provide an idea of each spe- cies' functional profile.Other multivariate techniques could be substituted.An important outcome of the PCA was the prominent separation of native and non-native species in distinct areas of the graph, primarily based on leaf mass per area, carbon : nitrogen ratio and foliar nitrogen (Fig. 1a).This result reinforces that species in Hawai'i with different biogeographic origins are functionally divergent.Hawaiian species tend to be conservative in regard to growth and nutrient acquisition, whereas the non-native species tend towards faster life-history strategies.Carbon turnover rates are substantially higher in invasive-dominated communities than they are in native plant communities (Hughes et al. 2014).With this in mind, we ran an additional PCA using only traits associated with carbon cycling (including specific gravity, foliar chemistry and maximum plant height) to find which species from our pool have the greatest likelihood of slow rates of carbon turnover (Fig. 1b).We identified two groups of four 'core species' whose trait profiles indicate either slow or moderate carbon turnover, and then located these core species on the first (general) PCA of all species traits (Fig. 1a).Based on Euclidian distances within the general PCA, we determined a centroid point equidistant from all four species in each group and then selected groups of additional species that were closer (i.e.most similar and potentially most redundant) and furthest (i.e.least similar and potentially most complementary) to each centroid (Fig. 1c).The motivation for this selection was that we are trying to determine whether combinations of species whose traits are more dissimilar (complementary) or similar (redundant) confer greater invasion resistance in the hybrid forests (Funk et al. 2008).\nOnce species are displayed in trait space, a variety of decisions can be made based on their relative functional profiles.While our process included two PCAs, for some restoration objectives, decisions could be made with only one (Fig. 1a).In our case, pragmatic and logistical concerns were addressed at this point.For example, one of our species was removed, because although it was culturally important, and functionally a good fit, it is unable to maintain itself without direct, ongoing human intervention.Alternatively, when several species proved to be functionally similar, final species choice was based on pragmatic considerations such as seedling cost, availability within given time frames or projected time to maturity.In short, the process allows for an unbiased way to let the data dictate a first step, and then, the practical concerns of practitioners can be layered onto the final species choices.", "title": "Using plant functional traits to restore Hawaiian rainforest", "id": "4.5", "keywords": [ "community assembly", "functional restoration", "Hawaiian lowland wet forest", "hybrid ecosystem", "invasion", "multivariate trait space", "tropical forest restoration" ] }, { "file_name": "Ostertag et al. - 2015 - Using plant functional traits to restore Hawaiian .pdf.tei.xml", "page_content": "To date, multiple studies have evaluated the role of functional traits in ecosystems and/or been carried out in controlled settings and relatively simple ecosystems such as grasslands (Drenovsky & James 2010;Fry, Power & Manning 2013;Lavorel 2013).We show that a trait-based approach can also be used in a tropical forest system.Because the metrics are based on site-specific restoration objectives, this approach is generalizable, flexible and transferable across ecosystems and taxa.Once refined, the approach can be applied with relatively little effort by managers.\nWe named our project Liko N a Pilina, which translates loosely to 'the budding new relationships', to emphasize the developing associations among the species, and the intertwining of fundamental science questions and the practical needs of land managers faced with a formidable task made more daunting by lack of information on how to achieve their restoration goals.While it is too early to determine if the treatments met our objectives, results 1 year after planting show >90% survival of outplants and increased seedling recruitment by native species.By creating hybrid ecosystems using both native and nonnative species (many of which are culturally significant), the project melds together traditional and contemporary approaches to forest management.This has proved attractive to the general public, school groups and summer programmes, allowing us to tap into an eager set of volunteers.In our experience, the level of community engagement is unusual for science-based restoration experiments.Thus, this new type of restoration can fulfil many goals, providing a rigorous way to choose species for restoration as well as providing a simple framework that is appealing to local communities.", "title": "Using plant functional traits to restore Hawaiian rainforest", "id": "4.6", "keywords": [ "community assembly", "functional restoration", "Hawaiian lowland wet forest", "hybrid ecosystem", "invasion", "multivariate trait space", "tropical forest restoration" ] }, { "file_name": "Ostertag et al. - 2015 - Using plant functional traits to restore Hawaiian .pdf.tei.xml", "page_content": "We thank the Strategic Environmental Research and Development Program for funding (Project RC-2117).Access to field sites was provided by the County of Hawai'i, Division of Forestry and Wildlife and Hawai'i Army National Guard.Jodie Schulten, Amanda Uowolo, Kai McGuire-Turcotte, Chaunda Tactay and Chris Chu assisted with field data collection and laboratory processing.Analyses were conducted by Lucas Mead and Tara Holitzki at the UH Hilo Analytical Laboratory; they were supported in part by National Science Foundation award number EPS-0903833.We thank our partners in the Hawai'i Army National Guard Environmental Office (Angela Kieran-Vast and Craig Blaisdell) and staff at Keaukaha Military Reservation for facilitating the establishment of the Liko N a Pilina project.Karen Holl provided comments on an earlier draft.", "title": "Using plant functional traits to restore Hawaiian rainforest", "id": "4.11", "keywords": [ "community assembly", "functional restoration", "Hawaiian lowland wet forest", "hybrid ecosystem", "invasion", "multivariate trait space", "tropical forest restoration" ] }, { "file_name": "Ostertag et al. - 2015 - Using plant functional traits to restore Hawaiian .pdf.tei.xml", "page_content": "We thank the Strategic Environmental Research and Development Program for funding (Project RC-2117).Access to field sites was provided by the County of Hawai'i, Division of Forestry and Wildlife and Hawai'i Army National Guard.Jodie Schulten, Amanda Uowolo, Kai McGuire-Turcotte, Chaunda Tactay and Chris Chu assisted with field data collection and laboratory processing.Analyses were conducted by Lucas Mead and Tara Holitzki at the UH Hilo Analytical Laboratory; they were supported in part by National Science Foundation award number EPS-0903833.We thank our partners in the Hawai'i Army National Guard Environmental Office (Angela Kieran-Vast and Craig Blaisdell) and staff at Keaukaha Military Reservation for facilitating the establishment of the Liko N a Pilina project.Karen Holl provided comments on an earlier draft.", "title": "Using plant functional traits to restore Hawaiian rainforest", "id": "4.12", "keywords": [ "community assembly", "functional restoration", "Hawaiian lowland wet forest", "hybrid ecosystem", "invasion", "multivariate trait space", "tropical forest restoration" ] }, { "file_name": "Ostertag et al. - 2015 - Using plant functional traits to restore Hawaiian .pdf.tei.xml", "page_content": "Data have not been archived because this article does not contain data.", "title": "Using plant functional traits to restore Hawaiian rainforest", "id": "4.13", "keywords": [ "community assembly", "functional restoration", "Hawaiian lowland wet forest", "hybrid ecosystem", "invasion", "multivariate trait space", "tropical forest restoration" ] }, { "file_name": "Ostertag et al. - 2015 - Using plant functional traits to restore Hawaiian .pdf.tei.xml", "page_content": "Data have not been archived because this article does not contain data.", "title": "Using plant functional traits to restore Hawaiian rainforest", "id": "4.14", "keywords": [ "community assembly", "functional restoration", "Hawaiian lowland wet forest", "hybrid ecosystem", "invasion", "multivariate trait space", "tropical forest restoration" ] }, { "file_name": "Carlucci et al. - 2020 - Functional traits and ecosystem services in ecolog.pdf.tei.xml", "page_content": "Ecological restoration usually focuses on restoring species composition of pre-disturbance communities (Engst et al. 2016) either by reintroducing targeted native species or by favoring their spontaneous recolonization on degraded sites (Rodrigues et al. 2009;Chazdon & Guariguata 2016).Restoration programs usually rely on the premise that the increased abundance of species typically found in reference ecosystems would support the recovery of key ecosystem services such as nutrient cycling, soil protection, and water supply (SER 2004;Wortley et al. 2013;Laughlin 2014;Kollmann et al. 2016).However, recent metaanalyses on restoration success have demonstrated that either species composition or ecosystem services have not been fully recovered (Crouzeilles et al. 2016;Moreno-Mateos et al. 2017;Shimamoto et al. 2018).Incomplete recovery might be explained by a lack of an explicit approach to restore ecosystem services, which are the benefits people obtain from ecosystems, such as provisioning (e.g.fresh water), supporting (e.g.nutrient cycling), regulating (e.g.climate regulation), and cultural services (e.g.recreation) (MEA 2005).Biodiversity drives ecosystem functions that underlie ecosystem service (MEA 2005) provisioning by influencing ecosystem processes that directly or indirectly affect energy and material flows (D\u00edaz et al. 2015).\nHere we consider that every ecosystem function is an ecosystem service by its direct or indirect effects on ecosystem benefits to humans.\nIf the loss of ecosystem services is a major concern in a world facing severe global changes (MEA 2005;IPBES 2019), services must be explicit targets in restoration programs (Ehrenfeld & Toth 1997;Funk et al. 2008;Montoya et al. 2012;Wortley et al. 2013;Laughlin 2014;Perring et al. 2015;Engst et al. 2016), and not only consequences of interventions guided for achieving other goals.However, a recent literature review found that most (52%) restoration project objectives have not explicitly considered ecosystem services (Kollmann et al. 2016).Although several papers and assessments have emphasized this need (e.g.SER 2004;MEA 2005;Montoya et al. 2012;Brancalion & Holl 2016) and some frameworks linking traits and services have been proposed (Laughlin 2014;Laughlin et al. 2018;Rayome et al. 2019), a clear framework on which services to consider and how to restore them using species traits is still missing (Kollmann et al. 2016), especially for broad-scale restoration.Therefore, the stability of communities over time and the services provided by restored ecosystems to support human well-being, especially under global changes, may be compromised.\nOperationalizing ecosystem services as targets for restoration is still a major challenge (Kollmann et al. 2016), and the selection of species to attain target services is not trivial because multiple species may support several ecosystem services (Lavorel & Garnier 2002;D\u00edaz et al. 2007).Knowing which species combination to reintroduce to restore ecosystem services while maintaining a stable restored community is a challenging task (Laughlin 2014;Laughlin et al. 2018).Traditional approaches to consider ecosystem services in restoration programs involve direct measurement of abiotic and biotic components (SER 2004).However, quantifying abiotic features hardly enables a clear association of ecosystem service and species composition, which has led ecologists to look at more meaningful and predictable parameters such as functional traits (Liebsch et al. 2008;Suganuma & Durigan 2015;Laughlin et al. 2017;Brancalion et al. 2018).\nA common approach has been to classify species into \"functional groups\" a priori, and then trust that reintroducing members of each group would restore the targeted ecosystem services (Perring et al. 2012).Nonetheless, a priori grouping of species into \"functional groups\" has several shortcomings.First, the performance of a species in a given functional group may be context dependent, as the performance of a function may be determined by trait-environment interactions (Brancalion et al. 2019a).Second, a species may belong to more than one functional group, and thereby perform several services in varying degrees (D\u00edaz et al. 2007).Third, the link between species and ecosystem services is mediated by effect traits, that is, those that impact ecosystem processes (Lavorel & Garnier 2002;Violle et al. 2007).Species have multiple effect traits that contribute independently or jointly to ecosystem services (Gamfeldt et al. 2008).Therefore, the effect of multiple species on different ecosystem services is more a multivariate continuum than a scenario with species classified into discrete functional groups.Fourth, by separating species into functional groups a priori, one may at best confirm an arbitrary decision rather than interpret how species relate to functional trait patterns a posteriori, which may provide information on which species contribute more to different ecosystem services.Because of the limitations to operationalize the use of traditional approaches of ecosystem service recovery by ecological restoration, new directions have been proposed in the literature (Funk et al. 2008;Laughlin 2014).\nThe claim for a new paradigm on ecological restoration demands a comparison on the advantages and disadvantages relative to the traditional paradigm (Table 1).Although restoration ecology studies have increasingly considered ecosystem services, on-the-ground restoration programs often experience difficulty applying biodiversity and ecosystem service theory in concert with the traditional restoration approach (Aerts & Honnay 2011;Kollmann et al. 2016;Naeem 2016).An alternative would be to use species effect traits clearly related to ecosystem services (Lavorel & Garnier 2002;Funk et al. 2008).An advantage of trait-based approaches is that they enable the assessment of the relationship between community assembly and ecosystem functioning (Laughlin 2014).This advantage exists because functional traits are two-sided coins in which \"function\" can be related either to response to community assembly processes mediated by abiotic and biotic interactions or to effect on ecosystem services (Lavorel & Garnier 2002;Violle et al. 2007;Suding et al. 2008).\nFunctional traits may be classified into \"soft traits,\" that is, those that are easy and quick to measure, and \"hard traits,\" that is, those that are harder to measure (Hodgson et al. 1999).Hard traits are usually more closely linked to mechanistic processes, but as they are hard to measure for a great number of species, they are usually replaced by soft traits that might be relevant to such processes.For instance, seed mass is a soft trait because it is relatively easy to collect and process seeds to get dry mass values, and at the same time represents correlated functions harder to measure, such as competition versus colonization abilities at initial stages of development (Turnbull et al. 1999).Functional traits may be related to mechanisms of community assembly such as competitive hierarchies (Keddy & Shipley 1989), limiting similarity (MacArthur & Levins 1967) and environmental filtering (Kraft et al. 2015).They can also relate to processes explaining how dominance of traits (i.e.mass ratio hypothesis, Grime 1998;and priority effects, Weidlich et al. 2018) or diversity of traits (i.e.niche complementarity, Tilman et al. 1997) influences ecosystem functioning.While the information on how traits affect ecosystem services relates to the biotic and abiotic components at an ecosystem-level framework, approaching how traits mediate community assembly mechanisms may provide information on community stability and resistance to environmental changes (Laughlin 2014).\nThe advantages of the trait-based approach for restoration rely, however, on the availability of information on functional traits for the species used in restoration projects (Table 1).Such shortcoming makes the use of a trait-based restoration especially challenging for species-rich tropical ecosystems, because the existing knowledge on functional traits is still concentrated in temperate species (Hortal et al. 2015).Moreover, in tropical regions, there are many taxonomic uncertainties due to high species richness, lack of surveys, and overall low scientific development (Karlsson et al. 2007;Hortal et al. 2015;Wilson et al. 2016).Consequently, there is a lack of basic taxonomic, ecological, and physiological information for a great number of native species in the tropics, making it hard to recompose species composition of originally mega-diverse tropical communities.This knowledge gap is an important barrier for the ambitious restoration programs planned for the upcoming years, which have a clear focus on supporting human well-being through the recovery of ecosystem services in degraded landscapes worldwide, but especially in the tropics (Brancalion et al. 2019b).Synthesizing the existing knowledge on functional trait-based ecological restoration is imperative to transform the potential of this conceptual approach into more successful onthe-ground projects, fostering research to new directions that may better match the demands of restoration practitioners.\nThe central goal of this article is to assess the progress made in ecosystem service-and trait-based restoration.For this, we systematically reviewed the literature on restoration ecology (2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017).Although we did this review for all kinds of ecosystems, we had a special focus in tropical ecosystems, which have the most pressing needs to advance with the inclusion of a trait-based approach to guide restoration.Specifically, we answered: (1) is there an increasing trend of published studies on ecological restoration which evaluate ecosystem services and functional traits?(2) how often have ecosystem services and functional traits been used as targets for ecological restoration in general and across continents?(3) are there biases related to target organism, ecosystem type, and geography in the literature towards some ecosystem service type and functional trait?(4) what are the most common approaches of using ecosystem services and functional traits in restoration: a priori or a posteriori?(5) which functional traits are used to evaluate different ecosystem services in ecological restoration?(6) what are the existing trait-based frameworks to target ecosystem services in restoration?After answering these questions, we summarized possible connections between simple-to-measure plant functional traits and ecosystem services and discussed the feasibility of applying trait-based frameworks widely in the tropics, considering both the basic limitations such as the lack of trait information or species in nurseries and the concentration of the >140 million hectares of restoration committed to as part of the Bonn Challenge and the New York Declaration on Forests in tropical countries.", "title": "Functional traits and ecosystem services in ecological restoration", "id": "5.1", "keywords": [ "ecosystem function", "functional diversity", "functional restoration", "reassembly", "restoration target", "trait-based restoration Conceptual Implications" ] }, { "file_name": "Carlucci et al. - 2020 - Functional traits and ecosystem services in ecolog.pdf.tei.xml", "page_content": "We systematically reviewed the ISI Web of Science database for papers published between 2007 and 2017.We chose 2007 as a start because a widely accepted concept of functional traits had been published that year (Violle et al. 2007).We followed the PRISMA protocol (http://www.prisma-statement.org/) for paper search and data collection standardization (Supplement S1).We looked for articles and reviews within the categories \"Ecology,\" \"Biodiversity conservation,\" and \"Forestry.\"\nWe made a general search of papers for obtaining overall patterns within the restoration ecology field.We used the following keywords related to ecological restoration in the title of the papers: restor* OR reforest* OR recover* OR regenerat* OR reintroduc* OR refaunat*.The general search resulted in Table 1.The traditional approach for ecological restoration is based on species composition of the original community, often informed by reference ecosystems, and does not use species traits, while the trait-based approach relies upon theories of community assembly and biodiversity and ecosystem functioning, using functional traits that inform about species coexistence and species contribution to ecosystem services.7,362 papers potentially on restoration.We obtained the information on the number of papers on ecosystem service and functional trait per country by using ISI Web of Science summarizing tools.In order to map the general trend of publication on functional restoration across continents, we refined this general search by using the following sets of keywords related to services or traits in the topic of papers (title, abstract, and keywords): function* OR service* OR guild* and trait* OR attribute*.We used each set of keywords at a time to filter only studies on services or traits, and together to obtain studies that evaluated both subjects simultaneously (by connecting keywords sets with an \"AND\").This last refined search (i.e.resulting from searching service-related AND trait-related keywords within the general search) resulted in 337 papers, which were used in the following steps of the methods for obtaining more detailed information (hereafter, \"specific search\").\nWe obtained the full text of 334 out of the 337 papers from the specific search: 327 in English, 4 in Portuguese, and 3 in Spanish.We evaluated these 334 papers to identify those about restoration, that is, those with the aim of reestablishing, by means of re-introduction or natural recovery of native species, a preexisting community entirely or partially lost due to anthropogenic causes.We consider as anthropogenic causes of a native community entire or partial loss phenomena (e.g.land use change, biological invasions).We did not include papers that studied the reintroduction of only one species with no focus on native species regeneration; assessed the recovery of a community after a natural disturbance (e.g.hurricane); aimed to assess restoration success by interviewing local human populations; aimed to spatially prioritize areas for restoration; or did not evaluate biological parameters.\nWe identified 265 papers on restoration (Supplement S2), which we screened to check: whether the study discussed or analyzed ecosystem services, and if the use of services was a priori (services used as targets in restoration) or a posteriori (services monitored in restored communities); whether the study discussed or analyzed functional traits, and if the use of traits was a priori (traits used to target services in restoration) or a posteriori (traits used to monitor restored communities); what ecosystem services were assessed; what functional traits were assessed; whether the study was carried in the tropics or elsewhere; the country of the study ecosystem; whether the focus ecosystem was terrestrial or aquatic; what was the specific terrestrial ecosystem type (Olson et al. 2001); what organism was used in restoration or was assessed during natural recovery; and whether the study proposed explicit and generalizable trait-based framework to select species for targeting ecosystem services in restoration (i.e. a framework that might be applied in different ecosystems).\nEcosystem services followed the classification by MEA (2005).We considered functional traits sensu Violle et al. (2007), that is, morphological, physiological, or phenological characteristics of organisms that impact species fitness and ecosystem processes and expanded on that definition by also considering life history and performance traits.Plant functional traits were classified into leaf economics spectrum, stem, root (including P and N acquisition strategy), flower (including pollination syndrome), diaspore (including dispersal syndrome, dispersal ability), whole-plant (maximum height, crown architecture), life history (life form, lifespan, carnivory), and performance traits (survival rate, growth rate, reproduction rate, physiological rate, competitive ability, stress tolerance, disturbance resistance/resilience, Grime's competitive/stress-tolerant/ruderal ecological strategies, total/aboveground/ belowground biomass, vegetation strata, shade tolerance).We did not consider as functional traits general grouping of species based on habitat use (e.g.pioneer vs. non-pioneer) or mixtures of life forms with phylogeny (e.g.graminoids).For additional information screened from papers, see Supplement S3.\nIn order to test whether the rate of accumulated number of all papers on restoration (from general search; n = 7,362) differed from the rate of accumulated number of papers on restoration that also evaluated ecosystem services, functional traits, and both (from the specific search) as n Service = 135, n Trait = 169, and n S + T = 76, respectively, along time, we fitted generalized linear models (GLMs) using function \"glm\" of the software R (R Core Team 2019).We considered number of papers in each category as the response and year of publication as the predictor variable.We also fitted generalized nonlinear models (GNMs) with an exponential mathematical function using the \"gnm\" function of R package \"gnm,\" but the linear models presented best fitting as shown by the Akaike information criterion: GLM All = 872.6,GNM All = 1882.4;GLM Service = 92.5,GNM Service = 122.9;GLM Trait = 96.2,GNM Trait = 129.7;GLM S + T = 73.8,GNM S + T = 93.9.We used Poisson distribution (in GLM and GNM) because response variables consisted of count data.\nWe used t tests to compare GLM slopes, considering the slopes are part of a Gaussian population of slopes of every possible hypothetical models.The t tests compared the slope of the GLM on the number of papers per each category of the specific search and the GLM on all papers, as follows: All versus Service; All versus Traits; and All versus Service + Trait.For computing the t tests, we used the estimate and the standard error of the estimate of the slope from the GLM.\nWe used Pearson's chi-square with Monte Carlo randomization tests (R function \"\u03c7 2 test\") to check for dependency of relative proportion of studies about services, traits, or both on ecosystem type; services on ecosystem type and geography (tropical or not); and plant traits on ecosystem service categories.We also used chi-square tests to check for the prevalence of a priori versus a posteriori approaches in studies on services or traits.", "title": "Functional traits and ecosystem services in ecological restoration", "id": "5.2", "keywords": [ "ecosystem function", "functional diversity", "functional restoration", "reassembly", "restoration target", "trait-based restoration Conceptual Implications" ] }, { "file_name": "Carlucci et al. - 2020 - Functional traits and ecosystem services in ecolog.pdf.tei.xml", "page_content": "There was an increasing trend of publication of papers on restoration (specific search; n = 265) that assessed ecosystem services, functional traits, or both across the time period analyzed (n = 228), which accompanied the trend for the whole area of restoration (general search; n = 7,362) (Fig. 1).The rates of increase in these specific subjects ( Most of the 7,362 papers (from the general search) were published by authors in North America and Europe, indicating the major role of developed countries on the ecological restoration literature (Fig. 2).Among the 7,362 studies, the proportion of papers that mentioned services, traits, or both was nearly constant across continents (Fig. 2).\nWe identified 228 out of 265 papers (from the specific search) that evaluated either ecosystem services, functional traits, or both.More than half of these papers focused on ecosystem services, whereas two-thirds evaluated functional traits (Table 2).\nMost studies involved plants (72%), while 23% studied animals and only 3% micro-organisms, fungi, or algae.Most papers focused on non-tropical ecosystems (68%), terrestrial habitats (69%), and had forests or woodlands as the target habitat (43%).Only 26% targeted open vegetation such as grasslands and savannas, 16% freshwater ecosystems (rivers, streams, lakes, or wetlands), and 7% marine ecosystems (oceans zones, coral reefs, estuaries, and salt marshes).\nThe relative proportion of studies (from the specific search) on ecosystem services, functional traits, or both depended on the ecosystem type (Fig. 3).Temperate forests and temperate open vegetation were the most common terrestrial ecosystem types in studies that assessed ecosystem services and functional traits, respectively, while tropical forests (dry and moist) were relatively common in studies using functional traits (Fig. 3).Studies integrating both services and traits were more evenly distributed across ecosystem types, being especially common in temperate open vegetation and tropical rainforests (Fig. 3).\nSupporting and regulating were the most common ecosystem services evaluated (Fig. 4).All categories of services (supporting, regulating, provisioning, and cultural) were found in both aquatic and terrestrial ecosystems in non-tropical ecosystems, but no papers assessing provisioning or cultural services were found in the tropics (Fig. 4).\nMost studies (from the specific search) that focused on ecosystem services assessed them a posteriori (\u03c7 2 = 62.59, p < 0.001), that is, they assessed services after rather than before the beginning of the restoration (Table 2).Similarly, most studies on functional traits assessed them after rather than before the restoration began (\u03c7 2 = 112.51,p < 0.001).Overall, more studies assessed functional traits (64%) than ecosystem services (51%), especially a posteriori (72% and 61%, respectively).One-third of the studies assessed both ecosystem services and functional traits.Only 31 studies (11.7%) evaluated both a priori, thereby using traits to target services before the start of restoration.\nAll categories of plant functional traits were found in papers (from the specific search) that evaluated regulating or supporting ecosystem services, while only performance and diasporerelated traits were evaluated together with cultural or provisioning services (Fig. 5).Whole-plant, diaspore, and performance traits were the most common plant traits studied (Fig. 5).Nutrient cycling and food web and community dynamics were the most common supporting services evaluated, while climate regulation, erosion regulation, and water cycling were the most common regulating services (Fig. S1).The different categories of functional traits were well distributed in terrestrial and aquatic ecosystems outside and in the tropics (Figs.S2 &S3).\nOnly 10 out of 265 screened papers (3.8%) presented a clear trait-based framework to target ecosystem services in restoration (Table 3).\nWe provided a summary of well-established linkages between plant functional traits and target ecosystem services to guide restoration practitioners (Table 4).\nSee the number of papers per country and journal in Tables S1 andS2.", "title": "Functional traits and ecosystem services in ecological restoration", "id": "5.3", "keywords": [ "ecosystem function", "functional diversity", "functional restoration", "reassembly", "restoration target", "trait-based restoration Conceptual Implications" ] }, { "file_name": "Carlucci et al. - 2020 - Functional traits and ecosystem services in ecolog.pdf.tei.xml", "page_content": "General Trends and Biases in Functional Restoration Literature Our review elucidates that restoration studies taking into account ecosystem services and functional traits have been growing in number at a similar pace to the whole field of restoration.However, we found a series of biases in functional restoration research.If someone were to randomly pick a study from a pool of restoration studies, the selected study would be most likely about restoration of plants in forests outside the tropics in a developed country.Furthermore, there would be a good chance that the study evaluated functional traits, but most likely it would merely monitor trait variation along time with no clear relation to ecosystem services or to prior specific restoration targets.Ecosystem services were more rarely evaluated, and mainly in terrestrial, non-tropical ecosystems.\nAlthough many international restoration commitments have been made by tropical developing countries harboring speciesrich ecosystems, these countries lack restoration ecology studies considering functional traits, and, especially, ecosystem services.Our findings corroborate the observation made by recent studies about the lack of biodiversity studies in tropical ecosystems in developing countries (Wilson et al. 2016;Clarke et al. 2017).Our results indicate that the most concerning region is Africa, with a large area, high biodiversity, serious threats to native ecosystems, high demand on ecosystem services, and few studies.\nTable 2. Number of papers on ecological restoration (2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017) that assessed ecosystem services and functional traits (n = 228 out of 265 screened papers; see details on the specific search in the Methods section).Note that all the papers on ecosystem services are under the category \"ecosystem services,\" not only those that studied exclusively services.The same applies for the category \"functional traits,\" and for \"a priori\" versus \"a posteriori\" comparisons.Moreover, the category \"Ecosystem services + functional traits\" represents the intersection (not the union) of studies on services and traits.Therefore, summing up values of the table does not result in 100%.More specifically, our results indicate that, while the number of studies differ substantially, the proportion of studies that consider services, traits, or both is similar between developed (North America, Europe, Australia) and developing regions (South America, Africa).Despite this relative consistency across continents, studies evaluating services were more common in temperate forests, while those assessing traits were more common in temperate grasslands and relatively common in tropical forests.Our results suggest we need more functional restoration studies in the tropics, especially in tropical and subtropical grasslands, savannas, and coniferous forests, but also in tropical forests.These studies should include functional traits, and whenever possible, ecosystem services.As we found no studies evaluating cultural or provisioning services in the tropics, it would be interesting that future studies attempt to assess these kinds of services.\nOur finding that most studies used services and traits a posteriori indicates monitoring of services and traits after restoration started is still much more common than their use before the beginning of restoration projects.The use of traits to target services before the start of restoration should be encouraged as a way to pursue the recovery of services in restored ecosystems (Laughlin 2014).Moreover, several of the studies did not evaluate trait functionality properly; they merely used traits as a separate dimension from taxonomic or phylogenetic dimensions of biodiversity without a clear mention to what the function meant.In order to ensure scientific rigor and higher predictability on the success of ecosystem service recovery, it is important that the relation of traits to ecosystem functioning and services be clear upfront in restoration projects.\nWhat Are the Existing Trait-Based Frameworks to Target Ecosystem Services in Restoration?\nA common approach among the trait-based frameworks for targeting ecosystem services in restoration was to use the classic theories of community assembly (Brudvig & Mabry 2008;Bochet & Garc\u00eda-Fayos 2015), biodiversity and ecosystem functioning (Perring et al. 2012;Mahaney et al. 2015;Ostertag et al. 2015), or both together (Funk et al. 2008;Laughlin 2014).Funk et al. (2008) proposed to select native species with traits similar to traits from invaders for biological control of invaders.Laughlin (2014) proposed a quantitative model to use trait values for targeting ecosystem services in restoration.For instance, species with dense woods and low specific leaf area might be selected to provide the restored community with resistance to future dry conditions; functional trait diversity might also be prioritized.Laughlin's approach was the only to enable the adjustment of species abundances to functional goals of restoration.An important point that only four papers explicitly addressed was the trait-based selection of species from regional and habitat species pool (Brudvig & Mabry 2008;Laughlin 2014;Bochet & Garc\u00eda-Fayos 2015;Ostertag et al. 2015).Assessing the functional structure of the habitat species pool at a regional scale may provide alternative candidate species that are functionally similar.Other frameworks focused on selecting species based on functional traits to restore functional redundancy and complementarity in plant-pollinator networks (Devoto et al. 2012), assure primary succession and phytostabilization on soils degraded by mining (Ilunga et al. 2015), and promote natural regeneration under semiarid conditions using nurse species traits (Navarro-Cano et al. 2016).\nMore recently (beyond our systematic review time frame), novel quantitative trait-based approaches for restoration of ecosystem services have been published.Laughlin et al. (2018) proposed a framework that enables selecting species for restoration to achieve convergent trait value targets and simultaneously maximize functional trait diversity.Rayome et al. (2019) proposed a framework and computer program to select species from regional pools based on the interpretation of multivariate trait patterns, which should enable practitioners selecting functionally redundant or complementary species based on restoration goals.Tsujii et al. (2020) developed a framework that enables selecting species for restoration from species pools to maximize functional richness and redundancy simultaneously as proxies for multiple ecosystem services and resilience with the minimal possible set of species.These recent trait-based frameworks permit the choice between functionally similar species from a previously known (regional or habitat) species pool, which may facilitate practitioners to use species that are available in the market, while assuring that certain ecosystem services will be provided in the restored ecosystem.Nevertheless, we consider we still miss a unified approach that simultaneously enables the trait-based selection of species from regional/habitat species pools and the definition of relative abundances of the selected species to achieve the provisioning of target ecosystem services.\nChallenges to Put Trait-Based Frameworks for Restoration in Practice at Broad Scales A major challenge for a trait-based restoration in the tropics (and elsewhere) is the lack of knowledge on functional traits of native species (Aerts & Honnay 2011).Establishing linkages between traits and services is still a major challenge in the ecological literature as a whole and has important implications to restoration ecology.We showed some patterns on how traits and ecosystem services are related in the restoration literature, which might help prioritize future studies.For instance, food provisioning might be targeted in tropical forest restoration, and the edibility of fruits or seeds might be used as a trait to achieve the delivery of this service in the restored ecosystem.In order to help practitioners put in practice a trait-based restoration to recover ecosystem services, we provided a summary of linkages between plant functional traits and target ecosystem services that are more consolidated in the literature.Ideally, functional traits might be available for the whole regional species pool relative to the sites to be restored.Some functional traits, like performance, whole-plant, and diaspore traits, are relatively well studied in the tropics, while others are less studied.In a major compilation of trait data, Petisco-Souza et al. ( 2020) found major gaps of information for key functional traits of 2,236 tree species of the Brazilian Atlantic Forest after searching in global plant trait databases, specialized books, regional datasets, and digitized plant specimens: 88% for seed mass; 85% for specific leaf area; 53% for wood density; and 16% for maximum height.These findings emphasize the importance of more funding for basic biodiversity research in the tropics both to fill biodiversity knowledge shortfalls and to advance scientific understanding on ecosystem functioning.\nTaxonomic and functional information needs also to be constantly reviewed and updated to accomplish the goals of functional restoration.A good example of broad-scale standardization of taxonomic nomenclature exists for Brazil: the database of the Brazilian Flora 2020 project (Forzza et al. 2012).This database compiles the accepted scientific names and basic biological information of more than 46,000 species of plants, algae, and fungi and is frequently updated by more than 400 experts (http://ipt.jbrj.gov.br/jbrj; accessed 24 Aug 2020).Considerable effort has already been made in research groups working in tropical countries for gathering functional traits as well, and data have been made available in major databases such as TRY Plant Trait Database (Kattge et al. 2020) and Botanical Information and Ecology Network (Maitner et al. 2018), besides several regional databases.\nEcosystem services and functional traits are increasingly being used in ecological restoration, although too few studies have considered them for ecological restoration in the tropics.Trait-based frameworks are useful for broad-scale restoration in the tropics because they may help circumvent common limitations of restoration in developing megadiverse countries (e.g.lack of saplings or trait information) by enabling considering alternative sets of species of the regional pool leading to similar resolutions of target services.While funding for tropical research is impaired and more complete datasets are not available, we advocate the use of the best existing ecological theory and available data in trait-based models to select species sets that will enable the restoration of ecosystem services.\nWe advocate that next steps towards a broad-scale functional trait-based restoration in the tropics are to: (1) define priority target ecosystem services for ecological restoration; (2) focus on measuring key functional traits linked to priority ecosystem services based on up-to-date scientific evidence; (3) test whether current restoration programs are reaching the functional composition and structure of reference ecosystems (Rosenfield & M\u00fcller 2017); and (4) run trait-based quantitative frameworks to select sets of native species and their abundances for areas that will be subject to restoration.", "title": "Functional traits and ecosystem services in ecological restoration", "id": "5.4", "keywords": [ "ecosystem function", "functional diversity", "functional restoration", "reassembly", "restoration target", "trait-based restoration Conceptual Implications" ] }, { "file_name": "Carlucci et al. - 2020 - Functional traits and ecosystem services in ecolog.pdf.tei.xml", "page_content": "This study was financed by Coordena\u00e7\u00e3o de Aperfei\u00e7oamento de Pessoal de N\u00edvel Superior-Brasil (CAPES)-Finance Code 001postdoctoral fellowship to M.B.C.; CNPq-Brazil productivity grants to M.V.C. (307796/2015-9), P.H.S.B. (304817/2015-5), and R.L. (306694/2018-2); and FAPESP grant to R.R.R. (2013/50718-5) and is a contribution of the INCT in Ecology, Evolution, and Biodiversity Conservation founded by MCTIC/CNPq (grant 465610/2014-5) and FAPEG (grant 201810267000023).The authors thank A. C. Petisco-Souza for providing information on trait knowledge gaps in the Atlantic forest.", "title": "Functional traits and ecosystem services in ecological restoration", "id": "5.13", "keywords": [ "ecosystem function", "functional diversity", "functional restoration", "reassembly", "restoration target", "trait-based restoration Conceptual Implications" ] }, { "file_name": "Carlucci et al. - 2020 - Functional traits and ecosystem services in ecolog.pdf.tei.xml", "page_content": "This study was financed by Coordena\u00e7\u00e3o de Aperfei\u00e7oamento de Pessoal de N\u00edvel Superior-Brasil (CAPES)-Finance Code 001postdoctoral fellowship to M.B.C.; CNPq-Brazil productivity grants to M.V.C. (307796/2015-9), P.H.S.B. (304817/2015-5), and R.L. (306694/2018-2); and FAPESP grant to R.R.R. (2013/50718-5) and is a contribution of the INCT in Ecology, Evolution, and Biodiversity Conservation founded by MCTIC/CNPq (grant 465610/2014-5) and FAPEG (grant 201810267000023).The authors thank A. C. Petisco-Souza for providing information on trait knowledge gaps in the Atlantic forest.", "title": "Functional traits and ecosystem services in ecological restoration", "id": "5.14", "keywords": [ "ecosystem function", "functional diversity", "functional restoration", "reassembly", "restoration target", "trait-based restoration Conceptual Implications" ] }, { "file_name": "Carlucci et al. - 2020 - Functional traits and ecosystem services in ecolog.pdf.tei.xml", "page_content": "The following information may be found in the online version of this article: Supplement S1.Steps followed during the systematic review.Supplement S2.Papers on ecological restoration (2007-2017) retrieved from the systematic review.Supplement S3.Additional information screened from papers on ecological restoration (2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017).Table S1.Rank of countries where studies on ecological restoration (2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017) were carried out.Table S2.Rank of journals that published papers on ecological restoration (2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017).", "title": "Functional traits and ecosystem services in ecological restoration", "id": "5.15", "keywords": [ "ecosystem function", "functional diversity", "functional restoration", "reassembly", "restoration target", "trait-based restoration Conceptual Implications" ] }, { "file_name": "Carlucci et al. - 2020 - Functional traits and ecosystem services in ecolog.pdf.tei.xml", "page_content": "The following information may be found in the online version of this article: Supplement S1.Steps followed during the systematic review.Supplement S2.Papers on ecological restoration (2007-2017) retrieved from the systematic review.Supplement S3.Additional information screened from papers on ecological restoration (2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017).Table S1.Rank of countries where studies on ecological restoration (2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017) were carried out.Table S2.Rank of journals that published papers on ecological restoration (2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017).", "title": "Functional traits and ecosystem services in ecological restoration", "id": "5.16", "keywords": [ "ecosystem function", "functional diversity", "functional restoration", "reassembly", "restoration target", "trait-based restoration Conceptual Implications" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "Restoration, in its broadest sense, involves improving conditions at a site to meet desired objectives.Traditionally, improving site conditions has meant an effort to return to a former, less-disturbed state, and much has been learned by examining recovery rates across ecosystem types (e.g., Rey Benayas et al. 2009).Yet in an increasing number of ecosystems, it is not feasible to return to a previous state for reasons that include the lack of reference sites or historical baseline conditions, irreversible climate change, and colonization by highly invasive nonnative species that cannot practically be removed (Hobbs et al. 2014, Zedler et al. 2012).Further, active forms of restoration via planting or encouraging specific species (Holl and Aide 2011) often proceed while information is lacking about species ecology, genetics, physiology, and evolutionary biology (Jones 2013).Choosing plant species for restoration can be a difficult task because it is not always clear which species are the most appropriate to achieve a particular restoration goal.A multivariate approach that allows users to identify a range of species likely to help them meet restoration objectives is one potential solution.Appropriate species chosen based on their life history characteristics can then be combined in a simulated community to see how these species are related to each other in their characteristics.", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.1", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "One major stumbling block in designing restoration plans is deciding which species to use.The motivation for this approach to species choice comes from the desire to merge practical and ecological restoration techniques, as well as the recognition that species choice for restoration can be a difficult and value-laden process.There are practical concerns such as cost, availability of seeds, and ability to propagate that can partially dictate decisions.Yet in many cases, little is known about each species' life history and how each will interact with other species when planted together, particularly if the planting might represent new combinations that are not seen in the field.Situations in which species that do not share an evolutionary history are thrust together provide relevant examples.These new combinations could arise because of invasion by nonnative species, range shifts of species resulting from climate change, or new species distributions resulting from land use activities.\nAlthough \"novel ecosystems\" (Hobbs et al. 2006) are becoming widespread, there is a very limited understanding in the ecological literature about the long-term implications of new species interactions and their effects on ecosystem functioning.", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.2", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "Functional trait-based restoration is based on the principle that ecosystem function depends in part on the expression of various morphological, structural, physiological, or chemical traits of organisms as well as environmental filters and the interaction between traits and environments.Functional traits reflect fundamental life-history and resource-use tradeoffs (Reich 2014).Because these traits vary predictably across environments, it is assumed that they are the products of natural selection.For plants, global datasets show how traits vary continuously along abiotic resource availability gradients and across biomes (Chave et al. 2009, Donovan et al. 2011, Reich 2014, Wright et al. 2005).Evolutionary tradeoffs faced by organisms in resource acquisition (e.g., light, water, and nutrient uptake) and resource processing (e.g., net primary productivity) result in different ways to make a living, termed the \"worldwide 'fast-slow' plant economics spectrum\" (Reich 2014).Plant species on the slow end of the spectrum have low rates of resource acquisition and processing, which requires leaf, stem, and root traits that are more conservative and efficient in resource use than plant species on the fast end of the spectrum.Being a slow species is advantageous under low-resource conditions because resource conservation traits enhance survival, but being a slow species can be a drawback under higher resource conditions.In a given biome, there is selection for trait convergence, but within a more localized community, it is likely that interspecific competition ensures that species differ along the slow-fast continuum (Reich 2014).Thus, at the community and ecosystem levels, consideration of functional trait values can help explain the distribution of species, the assembly of communities, and the rate of ecosystem processes (Reich 2014;Reich et al. 1999Reich et al. , 2003)).\nAt the community level, the functional trait profiles of species can be represented by functional diversity.Simply put, functional diversity is a way to define diversity of species traits within a community or ecosystem, encompassing metrics that focus on the magnitude, variation, and dissimilarity in species' functional traits (Schleuter et al. 2010).Considering functional diversity rather than species diversity may be a more promising approach for addressing questions of how species influence the structure and function of ecosystems (Laureto et al. 2015) or community assembly (Bhaskar et al. 2014).Therefore, selecting species for restoration projects that have a specific set of trait values should influence competitive interactions, resource availability, and ecosystem structure and functioning.Ideally, these functional traits should be easily defined and measured, so that the approach is transportable and flexible, and the predicted successional outcome of restoration can be tested (Ostertag et al. 2015).For example, selecting species with a broad range of functional traits (i.e., low niche overlap or inversely high functional divergence) may preclude exotic species from invading if their functional trait values are already represented in the community (Funk et al. 2008).\nBecause functional traits differ among species and environments in predictable ways, they can be linked to ecosystem properties and used in restoration to achieve specific objectives in ecosystem functioning (Funk et al. 2008).For example, the growth and recruitment of species with certain functional traits could be selected for by choosing species that facilitate plant and animal recruitment.If the objective is to build a community that will be less likely to burn, one could choose species with traits such as high leaf water content and low levels of volatile compounds.\nAlthough most studies attempting to link traits to ecosystem properties have been carried out in relatively simple systems, the field can be expanded to incorporate increasingly complex systems with higher species and life form diversity.The Restoring Ecosystem Services Tool (REST) program allows the user to design new simulated communities to make some assessments about which combinations of species may be best for specific restoration goals.REST has some data incorporated into it, yet allows users to enter their own species list and trait data.This strategy for species selection is generalizable and flexible, allowing users to choose the species and desired functional outcomes, while acknowledging limiting factors such as economics (e.g., cost of seed/plants, labor, time); logistics (e.g., availability of species, project or budget timelines); and predictability of climate or disturbance regimes, as well as the goals and expectations of stakeholders.The choice of species for restoration objectives is not limited to the scores from their traits alone, but could also incorporate other aspects, such as maintaining a diverse and resilient community that fosters the desired environmental outcomes.REST can be used iteratively; e.g., it can be reset and run again after removing species to continually refine choices.", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.3", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "Another difficult decision is the choice of traits.In part, restoration goals determine the traits that should be of interest to consider for a particular restoration project.For example, if the aim is to build an ecosystem that is tolerant of fires, traits such as bark thickness and leaf water content may be of interest.In general, there are six traits that the literature suggests are helpful in the attempt to understand life histories of various species (box 1).\nThese traits appear often in global analyses (Adler et al. 2014, Kunstler et al. 2016, van Bodegom et al. 2014).If you have no prior plant functional trait knowledge, examining these six traits will provide a good foundation.When you chose a restoration goal in the REST program, it will populate with a list of suggested functional traits that might be linked to your restoration needs.", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.4", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "\u2022 Foliar Nitrogen (resource acquisition)\n\u2022 Seed mass (reproductive investment and dispersal)\n\u2022 Specific leaf area (resource allocation)\n\u2022 Wood density (resource allocation)\n\u2022 Leaf lifespan (resource allocation)\n\u2022 Maximum plant height (dispersal)", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.6", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "The Hawaiian archipelago has many extremes of biological invasion, one of which is the remnant Hawaiian lowland wet forest (Zimmerman et al. 2008) that led to the development of REST.Approximately half the flora in Hawaii is not native (Wagner et al. 1999), and a number of invaders have been shown to have strong ecosystemlevel effects on carbon and nitrogen cycling and native biological diversity (e.g., Hughes and Denslow 2005, Litton et al. 2006, Vitousek and Walker 1989).A combination of events has led to systematic alteration of low-elevation lands, including (1) small-scale clearing and burning for agriculture and housing by Hawaiians prior to European contact (Kirch 2002); (2) large-scale clearing for sugarcane agriculture (Cuddihy and Stone 1990); (3) planting and aerial seeding of nonnative trees by territorial foresters, stemming from their lack of understanding about native forest function (Woodcock 2003); and (4) intentional and accidental introduction of many alien plants and animals that benefited from a mild climate, limited interspecific competition, and enemy release (Denslow 2003).The result is a series of communities dominated by mixtures of species that share no evolutionary history, and which contain high proportions of nonnative species classified as invasive.\nREST is a culmination of more than a decade of research conducted in a heav-", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.7", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "In these highly altered habitats, we have no clear historical guide to what species should be planted to achieve traditional restoration goals, and it has become clear that maintaining these forests as all-native species assemblages is unsustainable in terms of labor, logistics, and cost (Cordell et al. 2009, Ostertag et al. 2009).\nFunctional trait-based restoration can involve the use of species not originally found in a given site-including exotic species (Ewel andPutz 2004, Schlaepfer et al. 2011)-guiding the biodiversity toward more favorable species assemblages.The application of functional trait theory in restoration and management is an exciting new approach that can be used to understand the persistence of species and ecosystems as well as build model communities with desired ecosystem functions.", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.8", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "The Liko N\u0101 Pilina project evaluates four different combinations of species to determine the effectiveness of native and nonnative species mixes, or \"hybrid ecosystems,\" for their ability to maintain valuable forest structure and ecosystem services (see Ostertag et al. 2015 for more details).The Hawaiian name reflects the growing relationships that are developing in these new mixtures.The experimental species mixtures were designed using ecological theory related to community assembly rules and functional traits.In each community mixture, four core species were chosen based on functional traits that relate to carbon sequestration (note management goal 1 below), while the six additional species were chosen based on their traits being either redundant or complementary to the core species.Selecting species based on redundancy or complementarity is testing community assembly theory, in relation to invasion resistance, as it has been hypothesized that species with characteristics that complement one another will occupy different niche spaces and lead to a community that resists invasions while allowing native recruitment (note management goals 2 and 3).The particular management goals of this experiment are to develop hybrid ecosystems with the following traits: \u2022 Capable of sequestering substantial amounts of carbon A defining site feature is the substrate-an 'a'\u0101 lava flow dated as being 750 to 1,500 years old.This substrate, challenging for farming or mobility, is why this landscape remains uncleared.Rainfall averages 3347 mm/yr (Giambelluca et al. 2013), and mean annual temperature is 22.7 \u00b0C (Giambelluca et al. 2014).Native trees in the canopy and midstory define the forest, but these species are not regenerating under current conditions (Cordell et al. 2009).Rather, nonnative trees and shrubs comprise approximately 45 percent of the basal area (Ostertag et al. 2009).\nAs methods for selecting and comparing aspects of potential Liko N\u0101 Pilina candidate species informed the development of REST, the trait-based method we used employs five steps:\nStep 1: Articulate objectives and constraints-Because restoring this area to an all-native ecosystem is no longer economically feasible, we elected to create hybrid ecosystems with objectives to increase carbon storage, provide invasion resistance, and enhance native seedling regeneration.\nStep 2: Select appropriate functional traits-We selected a set of traits related to successional facilitation and carbon storage (table 1).Two variables are categorical (stature and canopy architecture) and were given ordinal numbers as a code.\nStep 3: Determine pool of species for trait sampling and restoration potential-Users must define their species pool based on contextually unique knowledge and objectives.To choose species for the experiment, we compiled a list of candidate species capable of surviving in lowland wet forest (LWF) environments in east Hawaii Island.For our purposes, LWF was defined as <700 m elevation and >2500 mm annual rainfall (Price et al. 2007).These climatic conditions are compatible with the study site where the hybrid ecosystem experiment was conducted.In addition, these species were chosen because they were not considered invasive, determined by using Hawaii Weed Risk Assessment scores (Daehler 2009).We examined 29 species for the overall species pool and aimed to use REST to condense to a smaller list that would allow us to simplify the logistics surrounding our experiment (i.e., fewer species to purchase, propagate, and plant).\nStep 4: Collection and preparation of trait data-\nWe sampled plant traits across the full range of conditions in which Hawaiian LWF is found to account for both site and environmental heterogeneity.In total, we sampled traits at 25 sites throughout east Hawaii Island in addition to using existing data from the literature.The most time-and effort-consuming steps in making species choices via trait use involve creating the potential species pool and collecting trait data.However, some shortcuts can be taken for those who do not have the resources to collect original data.REST contains some global trait databases, while other data can be sought out through the literature.There is also a variety of resources that provide background on data collection methods.The Prometheus-Wiki (http://prometheuswiki.publish.csiro.au/tiki-custom_home.php) provides protocols in ecological and environmental plant physiology and is updated by the research community.Another useful reference is Cornelissen et al. (2003) (http:// cef-cfr.ca/uploads/Membres/CornelissenProtocol.pdf), which provides standards for functional trait measurements.\nStep 5: Using REST data analysis for final species choice-", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.9", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "and follow the instructions found in the installation wizard (please note that, as this is an executable file, administrative or other permissions will be required to allow installation).The program will then install itself as well as create shortcuts for easy access.\nProgram on startup and personal database upload-On startup, after loading, the program will appear as in the screenshot shown in figure 1.As a part of the REST loading process, an Internet connection is required to update the database with any species additions or trait changes.However, REST can also start from the most recent archived version.During Step 4, the Liko N\u0101 Pilina project created its own functional trait database, which included 29 species and 15 traits.REST provides users with the option to import a personal database of species and traits in the form of a .csvfile (fig.2) in order to skip the manual process of selecting traits and species.In the event that the user has not created their own database, REST also provides the option to use the species and trait data currently stored within; this manual process is outlined in section 3.0, \"REST User Interface.\"After selecting the \"import database\" option, the screen-captured notification will appear as shown in figure 3.This window's purpose is to warn the user that only .csvfiles will be accepted: select OK.After acknowledging the previous warning, a new window will appear that allows you to browse for the .csvfiles you wish to use (fig.4).All\n.csv files can be imported, but if they are not formatted correctly, the program will notify you that an error has occurred and it will read your .csvfile only partially.\nThe format should be species in the first column followed by traits (fig.5).The units of each trait should be added to the trait name.Species and traits will appear in the program exactly as typed.\nNote that if a species or trait is already in the database but spelled differently, it will appear as a new trait or species.If you would like to follow along with the example in the rest of this section, import the file called LikoNaPilina.csv.REST PCA graph, analysis and interpretation-Once the personal database has been imported successfully, the largest section of the Data tab will show the project data: species, traits, and values (fig.6).Also, in this tab, the upper right corner will list the traits included in the desired analysis.(Note: the \"Restoration Goals,\" \"Species,\"\nand \"Traits\" dropdown menus are not used in this example but will be explained in full in section 3.0, \"REST User Interface.\")The Liko N\u0101 Pilina project's first desired analysis step was to compare all 29 species across the 15 different traits using PCA.\nTo complete this analysis, Check All was selected, followed by the Get Results, both options being provided at the bottom of the Data tab.In figures 7 and 8, species 7 and 10 are very similar to each other along PCA 1 but not along PCA 2. In the subsequent output, you can determine that PCA 1 is the most correlated with the traits Leaf C:N (0.391) and LMA (0.354).Thus species 7 and species 10 are very similar in those traits.A finding such as this example should be evaluated by the user, who will make decisions based on community theory and whether redundancy or complementarity is desired across the chosen species (i.e., how similar the species should be in trait profiles).Positive values indicate that a trait increases its value as that axis increases value, while negative values indicate an inverse relationship.For example, as you move to the right along PCA 1, Leaf Nitrogen (N)\nPercentage values decrease (-0.345), so that species 7 and 10 would have lower leaf N concentrations than all the species positioned to the left of them along PCA 1 (ranging from species 13 to species 4).When examining along PCA 2, it is noted that PCA 2 is positively correlated with specific gravity (0.374) yet negatively with leaf area (-0.387).\nThus, species 7 and species 10 are quite different in those two traits, a factor of their growth habits-species 7 is a tree fern, and species 10 is a slow-growing canopy tree.\nAnother important output to consider is the eigenvalues and variation explained by examining the data along these two dimensions.An eigenvalue reflects the amount of variance in the data in a given axis direction (Quinn and Keogh 2002).\nFor using REST, understanding the percentage of variation explained is sufficient.\nThe highlighted box in figure 8 shows that PCA 1 explains about 36 percent of the variation in the data.Adding PCA 2 explains another 17 percent, for a total of about 53 percent of the variation explained.Principal components analysis will never explain all of the variation in two axes, particularly if there are many traits.\nIn addition, many of the traits examined may be correlated with each other.Low eigenvalues may not be ideal, but the more important consideration is the graph to determine relative distances among species.\nBased on the PCA, we made decisions to eliminate some species, thus simplifying the logistics involved in our experiment:\n\u2022 Species 6 and species 7 are both tree ferns, yet species 6 was more available from growers.We decided to include only species 6, as the two species are close together on the PCA, and thus occupy similar trait space.\n\u2022 Species 4 is similar to species 28.However, species 4 presents propagation challenges, guiding selection toward species 28.\n\u2022 We used a similar logic with species 14 and species 19-we eliminated species 19 because it does not regenerate on its own.\n\u2022 Species 15 was similar to species 13.On site, species 13 would be placed at the lowest elevation of its range, potentially affecting survivorship potential.Thus, we eliminated species 13 in favor of species 15.\n\u2022 Species 26 and species 29 were also similar.We eliminated species 29 as it is less common in the LWF environment than species 26.\n\u2022 Species 9 was deemed unnecessary because it was in a cluster with a large number of species.\n\u2022 We also decided to eliminate the canopy trees already existing on site (species 10, 12, and 22).\nThe analysis can be easily run again by deselecting the nine above-mentioned species within the Data tab.Note the ways in which the output, graph, and PC values of the first analysis (fig.8) differ from the new analysis (fig.9).\nAfter this initial PCA was conducted, resulting in a simplified species list, further PCAs were run to select species mixes based on the Liko N\u0101 Pilina research questions and objectives.To elaborate, a PCA was run using a reduced trait list that included only those traits related to carbon storage/fast-to-slow strategies, in order to select core species that would facilitate slow and moderate carbon turnover.Once the carbon core species were identified, additional species selections per community mix were made based on Euclidean distances (i.e., close = redundant, far = complementary) visualized in trait space.(For a more detailed explanation, see Ostertag et al. 2015).", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.10", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "If desired, users can choose a restoration goal from the displayed dropdown menu in figure 11 by selecting the goal and clicking \"Add.\"Selecting a restoration goal will shorten the list of traits available in the \"Trait\" dropdown menu: displaying only those traits that pertain to that particular goal (fig.12). Figure 12-REST filters functional traits as they pertain to meeting a chosen restoration goal.", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.12", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "To choose a species, select a name from the menu or type the name.Click \"Add\" to include it in the analysis.After adding the species, the species will be added to the selected species window as shown in figure 13.", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.13", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "To choose a trait, select from the dropdown menu or type in the trait name (fig.14).If a species has data for a trait, the value will be visible in the trait column to the right of the species name.If there is more than one value for a species' trait in the databases compiled in the program, REST will take an average.Note that if a species has no value listed, then there are no current data in the program for that species.Values are required, as REST will not complete a PCA unless all species have values for all selected traits (users are encouraged to share their values with the authors as appropriate to augment future program and database updates).At this juncture, if a value is missing from the desired analysis, users should decide whether they would like to remove the species or trait from the analysis.To remove the species, simply uncheck the box to the left of that species.To remove the trait, select the trait where it is listed in the trait box (top right corner) and select Remove.\nThe REST program may be enhanced in the future with updated data in the database or program enhancements.We welcome any suggestions for future features you would like to see.Please send suggestions to ostertag@hawaii.edu.The REST database is still being actively built, and the program will become more user friendly as more of the data gaps are filled.", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.14", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "To edit a trait value, the user can simply double-click on the value and enter the change.Please note that units should be consistent for that trait in the trait list (some units and formatting differ in REST than their real-world counterparts because of programming constraints).Note that the changed value will not be stored in the REST database, but used only for the PCA conducted by the user.", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.15", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "When you are finished selecting your combination of traits and species, click the Get Results button.You will get an error message if you have not selected at least two species and two traits.Note that the graph and analysis output involved at this step were detailed in the previous section.", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.16", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "REST currently has four restoration goals built into the program:\n\u2022 Fire tolerance\n\u2022 Drought tolerance\n\u2022 Successional facilitation", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.17", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "The following are restoration goals for optional use in REST.By definition, included restoration goals filter potential functional trait inputs into those specific to the goal and appropriate for more targeted analysis.Because REST is user-defined, these four goals serve as a basis for popular intervention outcomes, but inputs can be increased, decreased, or otherwise altered as needed.", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.18", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "Fire is a threat to many ecosystems, especially in light of species invasions, greater human development, and climate alteration.Alternatively, the presence of fire events may be a natural part of other ecosystems.Traits related to flammability are included here.", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.19", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "Similar to fire tolerance, the potentials for drought and water use by plants are important concerns for intervention strategies.Traits relating to water storage and use are included here.", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.20", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "One goal of restoration may be to assist ecosystem recovery to another state with increased animal use or plant species more closely aligned with historical observation.\nTraits that can help with modifying conditions such as growth, reproduction, and dispersal are included here.For more information, see Pugnaire and Valladares (1999).", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.21", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "One management goal may be to maximize carbon storage across the landscape.\nTraits that are associated with plant growth and nutrient cycling are included here.", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.22", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "Below is a list of all functional traits as defined in the most recent version of REST.\nEntries include brief definitions, measurement units, useful information for measurement, and references for additional information where appropriate.Connections to restoration goals are denoted using two-letter abbreviations (FT, DT, SF, and CS, respectively).Each trait was assigned to one or multiple restoration goals.This assignment was subjective based on the collective field experiences of the authors.\nUsers of REST have the flexibility to add or subtract traits using the checkboxes next to each trait.Many of these traits are listed here because they are part of global databases.Note that, for some traits, there are very few species with data.For more information about specific functional traits, we recommend a variety of references, including Cornelissen et al. (2003), CSIRO (2013), Fitter and Hay (2002)", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.23", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "Leaf mid-day water potential-This indicator of mid-day water stress is measured in megapascals (MPa).(DT)\nLeaf Mn (percentage, per area, and per mass)-Leaf Mn percentage (%) = proportion of a leaf that is manganese.Leaf Mn per area = total amount of manganese contained within a given leaf area in grams per square meter (g Mn/m 2 ).Leaf Mn per mass = manganese content of overall dry leaf mass in kilograms (g Mn/kg).", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.24", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "Leaf NH 4 + -Ammonium content in a leaf, expressed in grams per kilogram (g/kg).", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.27", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "Leaf NO 3 --Nitrate content in a leaf, expressed in grams per kilogram (g/kg).", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.28", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "Leaf O 18 -Oxygen-18 isotope content in a leaf, expressed in parts per million (\u2030).", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.29", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "Leaf osmolarity-This measure of osmotic potential within leaves is expressed in", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.30", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "REST is best used as an iterative tool to compare different combinations of species.\nHowever, limiting factors exist outside of REST that practitioners should consider.\nThese include economics (e.g., cost of seed/plants, labor, or time), logistics (e.g., availability of species, project, or budget timelines), resilience to climatic change or disturbance regimes, and goals, objectives, or other expectations of stakeholders.\nThe process used by REST allows the data to provide an unbiased and objective first step, followed by practical concerns of limiting factors prior to final species choices.In this regard, REST can serve as a complement to other landscape-management decision tools at the local, regional, or ecosystem level.\nRisks of using REST as a management assistance program are few.As the program is both user-defined and creates simulated outcomes, the majority of risk tends toward the end user.Vetting of sites, trait and final species selection, and enacting conditions necessary for trait expression are all at the discretion of users.\nSuboptimal site conditions, seed germination, transplant survival, and related variables are inherent risks in any intervention attempt.Using REST during planning stages may help to decrease the severity of such circumstances if different species combinations are compared objectively.Yet, the REST user must bring to the table a site-specific understanding of (1) restoration goals, (2) target species, (3) community assembly rules, and (4) desired community densities, as the REST program does not address these areas.\nIn REST, we have accessed publicly available global databases that contain trait data for a variety of the world's species and compiled them into the program.\nHowever, there are many species with limited trait data and many species not in the program.For this reason, REST allows the user to import data on species and traits as a .csvfile.Another caveat for using REST is choosing to input categorical variables; as REST uses principal components analysis (PCA), a statistical technique in which it is proper to statistically analyze only continuous variables.Yet, categorical variables could be added by the user by coding each category with a number.For example, if you wanted to include dispersal type, a file could be inputted with the dispersal type coded so that wind = 1, water = 2, animal = 3, and gravity = 4.In that case, the user would need to import data as a .csvfile.Although it may not be viable to include categorical variables from a strictly statistical sense, the output from including these variables might still be useful in conveying restoration methods to the user, as the intent of the program is to offer useful visuals; we do not suggest using the statistical output beyond making species decisions for restoration planning.Any use of the PCA output for scientific publications should be reviewed by a statistician.Some important categorical variables include carbon pathway, growth habit, reproductive life history, shade tolerance, nitrogen-fixation capacity, and vegetative spread.\nFrom a hardware perspective, REST is relatively compact, an artifact of design with the decreased processing speeds and storage capacities of field computers in mind.When installed, REST requires less than 1 GB of RAM and 512 MB of disk space, allowing for smooth program operation and visual renderings.Because of its size, REST can be run from flash drives or SD cards.REST requires momentary Internet connectivity at startup for program and database update purposes, but can function offline after initial installation, as the master traits database will be copied to the user's local drive.Users can freely edit their local database copy as they please.However, users will not be able to directly edit the master trait database, nor will their local database changes be relayed to the host computer, preserving trait data integrity.Rather, users are encouraged to update their local databases periodically, noting that any local changes will need to be exported and reentered during the update process.", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.31", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "REST is an evolving program with many directions for future development.The program database contains several thousand entries ranging from rare endemics to more cosmopolitan species.More species, functional traits, and restoration goals are currently being drawn from regionally exclusive species traits lists, a comprehensive literature review, new data generated by Liko N\u0101 Pilina and other projects, and information provided by managers familiar with REST.Efforts to expand REST data are ongoing and updated regularly.New versions can be found at https:// hilo.hawaii.edu/faculty/ostertag/.", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.32", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "Hawaii weed risk assessment score-The likelihood of invasion or \"taking over\" of a given plant species as outlined by the Hawaii Weed Risk Assessment Guide and other observational guides regarding plant-ecosystem interactions.For more information, see Daehler (2009).(DT, SF) Heat tolerance-A plant's ability to withstand temperature conditions above its generally accepted upper limit.Heat tolerance is measured in hours (h).(FT, DT,\nLeaf A max -An abbreviation for maximal assimilation, leaf A max is the maximum rate of photosynthesis of a leaf.The unit for A max is micromoles per square meter we would especially like to thank Amanda Uowolo, Taite Winthers-Barcelona, William Buckley, Peter Vitousek, Jodie Schulten, Kaikea Blakemore, Corie Yanger, and many more dedicated undergraduate and post-baccalaureate interns, volunteers, and guest researchers.This report or its contents has not been subject to review by any external agency, university, or otherwise associated entity and therefore does not necessarily reflect their respective organizational views.No official endorsement should be inferred.\nMultiply by: To find: Micrometers (\u00b5m)", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.56", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "Leaf A max -An abbreviation for maximal assimilation, leaf A max is the maximum rate of photosynthesis of a leaf.The unit for A max is micromoles per square meter we would especially like to thank Amanda Uowolo, Taite Winthers-Barcelona, William Buckley, Peter Vitousek, Jodie Schulten, Kaikea Blakemore, Corie Yanger, and many more dedicated undergraduate and post-baccalaureate interns, volunteers, and guest researchers.This report or its contents has not been subject to review by any external agency, university, or otherwise associated entity and therefore does not necessarily reflect their respective organizational views.No official endorsement should be inferred.", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.60", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Rayome et al. - Restoring Ecosystem Services Tool (REST) a Comput.pdf.tei.xml", "page_content": "Multiply by: To find: Micrometers (\u00b5m)", "title": "Restoring Ecosystem Services Tool (REST): a program for selecting species for restoration projects using a functional-trait approach", "id": "6.62", "keywords": [ "Functional traits", "ecosystem services", "land management", "drought tolerance", "fire tolerance", "successional facilitation", "carbon storage" ] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Tropical landscapes have been extensively degraded and deforested, but large-scale passive and active restoration projects have catalysed secondary forest regeneration over the last few decades (Chazdon, 2014).Tropical dry forests (TDFs) have a strong dry season of at least 3-4 months where little to no rain falls (Murphy & Lugo, 1986), which distinguishes TDFs from tropical wet forests, and offers a unique hurdle to restoration projects.Globally, 97% of TDFs are threatened by anthropogenic processes (Miles et al., 2006), and in Central America it is estimated that only 1.7% of the original extent of TDF exists (Griscom & Ashton, 2011;Miles et al., 2006).Despite the fact that TDFs are critically endangered (Janzen, 2002), the restoration of TDFs has been studied minimally compared to wetter tropical forests (Meli, 2003).Following agricultural abandonment in Central America in the 1990s, large tracts of land became available, and active and passive restoration techniques such as plantation establishment, enrichment planting and fire exclusion have been effective in the reestablishment of TDF (Griscom & Ashton, 2011).In north-west (NW) Costa Rica, the conservation of over 160,000 hectares of land in the \u00c1rea de Conservaci\u00f3n Guanacaste created one of the most effective TDF restoration projects in the world (Janzen, 2002).In the \u00c1rea de Conservaci\u00f3n Guanacaste, passive restoration techniques, including removal of unnatural fires and grazers, have been effective in the widespread restoration of TDFs (Janzen, 2002), but active management interventions are necessary (Holl & Aide, 2011) on degraded soils where succession is arrested.\nHighly degraded Vertisols, common in NW Costa Rica, are a barrier to the regeneration of TDF on large tracts of land (Guti\u00e9rrez, pers.obs.).Vertisols have shrink-swell cycles resulting from high expansive clay content (Deckers, Spaargaren, & Nachtergaele, 2001).Degraded Vertisols shrink and crack during the dry season, and swell during the rainy season, resulting in flooding (Figure 1).Consequently, restoration of these soils is particularly challenging due in part to high seedling mortality rates (Guti\u00e9rrez, pers.obs.).To date most studies have focused on developing best practices to use Vertisols for agriculture (Deckers et al., 2001) and not restoration.Most Vertisols in NW Costa Rica were deforested and used for rice cultivation or cattle grazing (Guti\u00e9rrez, pers.obs.), and if livestock density is high compaction resulting from grazing can impede restoration (Nepstad, Uhl, & Serr\u00e3o, 1991).Accordingly, the use of soil amendments during planting may hold promise, as certain amendments facilitate establishment of native TDF seedlings (Fajardo, Rodr\u00edguez, Gonz\u00e1lez, & Brice\u00f1o-Linares, 2013).In Vertisols, amendments that ameliorate microclimatic conditions by improving drainage during seasonal flooding may increase initial survivorship and growth when mixed into the rooting zone of seedlings.Other barriers to TDF restoration include selection of species for plantings, as performance data for native tree species are generally unavailable (Butterfield, 1995).\nPlant functional traits, i.e. morphological, physiological or phenological characteristics linked to plant survival, growth or reproduction (Violle et al., 2007), have been effectively used to select species for tropical wet forest restoration projects (Ostertag, Warman, Cordell, Vitousek, & Lewis, 2015).Furthermore, a restoration experiment in a wet forest in Mexico showed some combinations of resource-acquisition functional traits were correlated with survival and growth rates of pioneer tree species (Mart\u00ednez-Garza, Bongers, & Poorter, 2013).Therefore, using functional traits to understand mechanisms behind species' performance allows results to be extrapolated to other native species with similar traits (Pywell et al., 2003), and ultimately aids the design of restoration plantings that achieve specific outcomes such as rapid carbon sequestration (Ostertag et al., 2015) or restoration of soil fertility (Carpenter, Nichols, Pratt, & Young, 2004).Moreover, defining and identifying the\nactive restoration, Costa Rica, native species, plant functional traits, seedling establishment, soil amendments, stable isotopes, tropical dry forest, Vertisols, water-use efficiency trait syndromes of TDF species, which are hypothesized to be either conservative (efficient water use/drought-tolerant) or acquisitive (higher water use/drought-intolerant; Lohbeck et al., 2013), could further help to design plantings that achieve specific management goals.\nTo develop best practices to restore native tree species on degraded Vertisols, we implemented a large-scale TDF restoration experiment in Estaci\u00f3n Experimental Forestal Horizontes in NW Costa Rica.Our objectives were to: (1) test the restoration potential of many native TDF tree species on Vertisols, (2) determine if plant functional traits explain interspecific variability in species performance, and thus can be used effectively to choose tree species for restorations, and (3) test if affordable and readily available soil amendments increase initial survivorship and growth of planted seedlings.Although not the focus of our study, we also gathered data for a cost-benefit analysis of these management practices.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.1", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Our experiment was carried out from 2014 to 2016 at Estaci\u00f3n Experimental Forestal Horizontes (10.712N, 85.594W) in the \u00c1rea de Conservaci\u00f3n Guanacaste (ACG) in NW Costa Rica.This region has a mean annual precipitation of 1,730 mm with a 5-6 month dry season (December or January-May; www.investigadoresacg.org).During our experiment, sustained wet-season rains did not begin until August for both 2014 and 2015 (Figure S1), leading to the strongest drought on record in this region (IMN Costa Rica, 2015).The site had been abandoned for ~28 years, following decades of rice production and cattle grazing, and received no management since then (Guti\u00e9rrez, pers.obs.).Prior to the experiment, the vegetation was in a state of arrested succession (Figure 1), and the few trees that regenerated, out of 21 species observed growing on these soils in the surrounding area, were dominated by only three species: Cochlospermum vitifolium (Bixaceae), Crescentia alata (Bignoniaceae) and Guazuma ulmifolia (Malvaceae) (L.K. Werden, unpublished).A forest inventory study in the same region found 146 tree species in 84 0.1-ha plots (Becknell & Powers, 2014), corroborating that this is a particularly species-poor site.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.3", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Our goal was to collect native species common in the region (Powers et al., 2009), locally abundant on Vertisols, and/or already proven effective in other local restoration projects.We collected seeds from at least three individuals of 38 species in the ACG and stored them in paper bags.All seed was prepared following Rom\u00e1n, De Liones, Sautu, Deago, and Hall (2012) and local knowledge, and underwent a 12-hr soak in water to soften seed coats.In early April 2014, seeds were planted into a 3:1 locally collected Inceptisol soil:sand mixture in 5 \u00d7 8 cm black polyethylene bags.Seedlings were grown under 90% polyethylene shade cloth for 5 months.Six species with low germination rates were eliminated at this point (Table S1).Seedlings infested with insects were sprayed with diluted insecticide (0.01% w/w Bayer Decis \u00ae , Bayer S.A., San Jos\u00e9, Costa Rica) once each in July and August 2014, otherwise they were not sprayed.Before planting, we removed the shade cloth for 2 weeks to harden seedlings in full sun.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.4", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "In June 2014, we cleared existing vegetation from a 1 ha patch of Vertisols with chainsaws and a tractor.In early September 2014, we planted 1,710 seedlings of 32 species (Table S2) into six randomly applied treatment blocks coupled with four soil amendments (sand, rice hulls, rice hull ash and hydrogel) and two unamended controls (to account for variation in microtopography).Soil amendments were selected based on the criteria that they were inexpensive and readily available locally.We used sand, rice hulls and rice hull ash, all thought to increase drainage during seasonal flooding and hydrogel (Hidrokeeper \u00ae , Qemi International, Inc., Kingwood, TX, USA), an acrylamide and potassium acrylate copolymer which holds 350% its dry weight in water and extends the wet season by maintaining soil moisture in the rooting zone for 2 weeks following a period with no precipitation.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.5", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "After the onset of wet-season rains in late August 2014, we predug holes in a 1 \u00d7 1 m grid for each treatment block.Seedlings were planted over 3 days in early September 2014 at 1 \u00d7 1 density, and 10 seedlings of each species were planted randomly by row into each of the six treatment blocks (Figure S2).Some species had insufficient individuals to include in all treatments, so we included these species in as many treatments as we had multiples of 10, resulting in an unbalanced design (Table S3).We applied amendments in standard quantities to individual seedlings, either around (sand, rice hulls and rice hull ash) or below the rooting zone (1 L of hydrogel pre-mixed with water following manufacturer directions).Amendments applied around the rooting zone were mixed into soil extracted from each hole.Holes were subsequently back-filled with the soil/amendment mixture (3:1 soil to sand; 2:1 soil to rice hulls and rice hull ash).Cost analysis for site preparation and planting appear in Appendix S1.Pasture grasses around seedlings were cleared with machetes to avoid competition and shading, in October and November 2014, and in July 2015.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.6", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "We measured resource-acquisition and allocation traits assumed to be correlated with species' growth and survival.Plant trait data were collected twice during the experiment to examine how traits changed with ontogeny and in response to field conditions.First, we destructively harvested a subset of seedlings from the nursery to quantify leaf, total biomass and leaf chemistry traits.Second, ecophysiological and leaf chemistry traits were collected (non-destructively) from seedlings after they acclimatized to environmental conditions for 3 months.\nIn August 2014, we measured leaf traits on seedlings in the nursery using three leaves of five individual seedlings per species.We measured leaf area (cm 2 ), leaf wet and dry weight (g) dried at 60\u00b0C (>48 hr), leaf thickness (mm) and petiole length (mm) using standard methods (P\u00e9rez-Harguindeguy et al., 2013).These data were used to calculate leaf dry matter content (LDMC; g/g), specific leaf area (SLA; cm 2 /g) and leaf density (g/cm 3 ).Remaining plant material was dried, and roots, shoots and leaves were weighed.We collated additional trait data from previous studies on adult trees in the same region including: leaf habit and compoundness, wood density (Powers & Tiffin, 2010) and maximum height (J.S. Powers, unpublished).\nIn November 2014, we measured additional traits for all species that still had leaves in the control plots (N = 27).We measured \u03a8 diurnal (\u03a8 mid-day -\u03a8 pre-dawn ; MPa), a metric used to group species into drought-tolerator or avoider categories (Martinez-Vilalta, Poyatos, Aguade, Retana, & Mencuccini, 2014), with a pressure chamber (PMS Instrument Co., Albany, OR, USA) on two leaves of two individuals per species.We also measured photosynthetic light curves before 11 a.m., following Guzman and Cordero (2013), with a LCi portable photosynthesis system (ADC Bioscientific Ltd., Hoddesdon, UK) on three individuals per species.Following K\u00fcppers and Schulze (1985), for each light curve, we calculated: stomatal sensitivity (SS; % per s), stomatal conductance (g s ; mmol m -2 s -1 ), photosynthetic capacity (A max ;\n\u03bcmol CO 2 m -2 s -1 ), quantum yield (\u03a6; \u03bcmol CO 2 \u03bcmol \u03b3 -1 ), instantaneous water-use efficiency (WUE; \u03bcmol CO 2 mmol H 2 O -1 ), dark respiration rate (R dark ; \u03bcmol m -2 s -1 ) and light compensation point (LCP; \u03bcmol m -2 s -1 ).These parameters were included in analyses as they are associated with how plants acquire carbon (A max , R dark ; Poorter, 1999) and cope with water stress (WUE, g s , SS; Galmes, Medrano, & Flexas, 2007), while some directly underlie photosynthetic strategies (LCP, \u03a6; Poorter, 1999).\nLastly, field and nursery collected leaves (2 leaves per individual \u00d7 2 individuals = 4 leaves/species) were dried and transported to Minnesota, USA where they were ground and analysed to determine carbon and nitrogen concentrations, and their stable isotopes, \u03b4 13 C and \u03b4 15 N.The stable isotope of carbon, \u03b4 13 C, is often interpreted as an integrated measure of WUE (Farquhar, Ehleringer, & Hubick, 1989).We quantified leaf chemical traits at two time points (in the nursery and 3 months after planting) to determine whether whole-plant processes contributing to the regulation of these bulk leaf chemical traits varied with ontogenetic stage, and if patterns of up-or downregulation of these traits after exposure to field conditions correlated with species' performance.From foliar isotope data, we calculated the change in foliar \u03b4 13 C values after 3 months in the field as \u0394\u03b4 13 C (\u2030) = \u03b4 13 C field\u03b4 13 C nursery , where positive values indicate upregulation of integrated WUE.\u0394 \u03b4 15 N (\u2030) was calculated in the same way, and positive values represent upregulation of N-fixation for legume taxa.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.7", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Survivorship and growth was monitored for every seedling for two wet and dry seasons.We collected survival data every 1-2 weeks between September 2014 and December 2014 to calculate mortality rates related to transplant shock, and subsequent stresses such as flooding.We measured seedling survival, height and diameter at the base and top (at the apical meristem) monthly from September to December 2014 (end of first wet season).The same survey was conducted in July 2015 (end of first dry season), December 2015 (end of second wet season) and again in July 2016 (end of second dry season).", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.8", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "In November 2014, we collected 10-cm deep volumetric soil samples (442 cm 3 ) every 5 m on a transect down each block centre using a steel ring.Soil samples were air-dried and sieved (2 mm), soil particle size distribution (sand, silt and clay percentages) determined using the hydrometer method (Bouyoucos, 1962), and soil pH was measured on a 1:2.5 soil:water solution using a Oakton pH electrode.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.9", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "We used seedling survivorship and growth surveys to quantify seedling performance over the course of the experiment.Unfortunately, 265 planted seedlings were accidentally destroyed during the first grass clearing in October 2014.These individuals were omitted from statistical analyses; thus, the total number of seedlings was N = 1445 (Table S3).We defined time periods representing distinct stages that had a specific hypothesized influence on seedling growth and survival (Table 1), and for each time period, we computed survivorship percentages and relative growth rates (RGR).Due to our experimental design, some parameters were calculated at the species-level (survivorship percentages) and some were calculated for individual seedlings (RGR).We calculated RGR for both height (RGR ht ; ln(cm) per day) and stem volume (RGR vol calculated as the volume of a conical frustum; ln(cm 3 ) per day) for each seedling using the classic RGR equation:\n(ln[final height or volume] -ln[initial height or volume])/(final dayinitial day) (Hoffmann, 2002).RGR values were calculated for each time period to evaluate changes in RGR over time.We quantified RGR during the dry season because some species we planted are evergreen (Table S2) and may grow year-round.Over the course of our experiment, RGR did not slow or approach an asymptote, thus the assumptions required of the classic RGR model were met (Paine et al., 2012).\nSeedling survival and RGR for each time period were the main response variables of interest.We used visual assessments of survivorship among species, and performed type-III ANOVAs (for unbalanced", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.10", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Soil clay concentrations ranged from 41.6% to 71.2% and pH values from 5.48 to 6.09.Nonetheless, we found no significant pairwise differences (Tukey's HSD; p < .05) in soil particle size distribution or pH among the six treatment blocks (Appendix S2).", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.12", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Over all treatments, survivorship to 2 years had a large range, from 0% to 92.5% among species (Figure 2, Table S2).Of the 32 species, 20 had survivorship of <10% and 15 of those 20 species had no surviving individuals.The remaining 12 species were considered restorationviable, all of which had survival rates \u226510.9%, and 7 of which had >30% survival overall.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.13", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Values for most functional traits had large interspecific ranges, with coefficients of variation for individual traits ranging from 3.02% for foliar \u03b4 13 C nursery to 208.63% for seed mass (Table S4).The only trait correlated with survival when the experiment concluded was \u0394 \u03b4 13 C (r = .53,p < .01,Table 2), indicating that the trait most predictive of long-term seedling survival was the ability to upregulate integrated WUE (\u0394\u03b4 13 C).Distinct trait groups were correlated with survivorship at different ontogenetic stages (Table 2), but no relationships were found between survival and life-history traits (leaf habit, leaf compoundness or seed dispersal syndrome) (Tukey's HSD; data not shown).During transplant shock and the first wet season, wood density and leaf chemistry traits (C and N %) were correlated with species survival, but following the first water limitation (first dry season), all trait-survivorship relationships were driven by photosynthetic (g s ,\nA max , R dark , LCP) and/or water-use traits (g s , WUE, \u0394\u03b4 13 C).\nWood density was the best predictor of survivorship during both transplant shock (r = .58,p < .001),and the first wet season (r = .55,p < .001),as species with higher adult wood densities had higher rates of early survival.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.14", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "A similar set of photosynthetic and water-use traits were correlated with seedling RGR, and these relationships differed slightly by time period when RGR was expressed for height (RGR ht ; Table S5) or volume (RGR vol ; Table S6).RGR ht and RGR vol were highly correlated (p < .001)during transplant shock, the first dry season, and second wet season, but not for the first wet season and second dry season (Table S7).We therefore address relationships between RGR ht and RGR vol and traits separately for the first wet season and second dry season.\nDuring transplant shock, shoot dry mass was predictive of RGR ht (r = -.47,p < .01),and R dark predictive of RGR vol (r = -.65,p < .001).S5), indicating that trees of short stature as adults had higher overall RGR ht .\nT A B L E 2 Pearson correlations for species-level traits and survivorship percentages at specific time periods (see Table 1)", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.15", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Transplant shock", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.17", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Leaf, stem and biomass Seed mass (g) 0.19 0.15 -0.12 -0.12 -0.12 Leaf area (cm 2 ) 0.01 -0.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.21", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "LDMC was the only \"soft\" (easily measured), trait correlated with any \"hard\" traits (requiring specialized equipment to quantify).LDMC was significantly correlated with A max (r = -.6, p < .001),and instantaneous WUE (r = .54,p < .01;data not shown).No \"soft\" traits were correlated with \u0394 \u03b4 13 C.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.22", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "There were large differences in seedling survivorship among soil amendment treatments during the first wet season, but survivorship converged following the first dry season (Figure 3).The rank order of survivorship after the first wet season was: hydrogel = sand > rice hulls > rice hull ash = controls (Tukey's HSD; Figure 4a), and this pattern was essentially the same for restoration-viable species survivorship (Figure 4b).While by rank seedlings planted with hydrogel consistently had the highest survivorship (Figure 4), survivorship did not differ statistically among soil amendments after 2 years (Figure 4e,f).", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.23", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Soil amendments did not alter mean RGR ht for any period except the final dry season (F = 3.03; p = .019;Table S8).During this period, seedlings planted with hydrogel had <1% lower RGR ht (p = .003;\nTukey's HSD) than control seedlings.RGR vol was influenced by soil amendments for several time periods.During transplant shock, the rank order of increases in mean RGR vol relative to the controls was: sand = rice hull ash > controls (sand: 32% higher, p < .001;rice hull ash: 25% higher, p = .018;Tukey's HSD).Finally, during the first wet season, RGR vol in the rice hull ash treatment was 33% higher than the controls (p < .001;Tukey's HSD).", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.24", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "We have shown it is possible, albeit challenging, to establish native TDF species on degraded Vertisols.In our species screening trial, overall seedling survivorship at 2 years was very low, and only three species attained survival rates >50%.These results suggest that continued active management will be essential when restoring TDF on Vertisols.Species with high performance in this stressful environment had some overlapping functional traits, namely the ability to upregulate integrated WUE after planting and to maintain high rates of photosynthesis and instantaneous WUE.Explicitly considering water-use and photosynthetic traits when selecting species for plantings could greatly increase initial effectiveness of TDF restoration projects on degraded soils.While certain soil amendments increased short-term survivorship, no amendment influenced survivorship after 2 years.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.25", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "NW Costa Rica experienced two of the most extreme drought years on record during our experiment, and droughts such as these are becoming more common (IMN Costa Rica, 2015).A natural drought event was therefore imposed on our experiment, and species with the highest survival rates (Figure 2) are not only suitable for Vertisols plantings, but are also likely to tolerate future rainfall deficits.For comparison, an experiment in the same conservation area found some of the species we used (Hymenaea courbaril and Swietenia macrophylla) had 50% higher survival when planted in non-drought conditions on less-degraded soils (Gerhardt, 1993).Much lower survivorship should therefore be expected in Vertisol restorations and/or during drought, but could be partially remedied with management strategies such as replanting seedlings in subsequent years or initially planting multiple seedlings at each hole, and subsequently thinning.Despite these challenges, we identified a diverse group of species tolerant of these harsh conditions.\nThe species we found to be restoration-viable exhibit diverse life-history traits, and could be used to design planting mixes (Table S2).In terms of leaf habit and dispersal syndrome, almost all restoration-viable species were deciduous and wind dispersed,", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.26", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Transplant shock soil fertility in other Costa Rican restorations (Carpenter et al., 2004;Gei & Powers, 2013), provide further utility for Vertisol restoration projects.Not surprisingly, species with the highest (Crescentia alata;\n92.5%) and seventh highest (Guazuma ulmifolia; 32.6%) survival rates are two of the most common tree species on Vertisols at this site, but most other species we found to be restoration-viable are rarely found on Vertisols, or at all (L.K. Werden, unpublished), and seed dispersal limitation could be a barrier to their re-establishment (Janzen, 2002).", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.27", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Water limitation during early TDF succession may influence species composition by favouring drought-tolerant species with conservative trait syndromes (Lohbeck et al., 2013).Consistent with this framework, the only trait we found to predict survivorship through the conclusion of our experiment was the ability to upregulate integrated WUE (\u0394\u03b4 13 C; Table 2).From the traits, we measured, \u0394 \u03b4 13 C may be the most important to consider when the management goal is to select drought-tolerant species for TDF restorations.None of the so called \"soft\" leaf traits we quantified predicted survivorship, and only one leaf trait was predictive of RGR (petiole length), and only during the first dry season (Table S5).This is consistent with findings from a Mexican wet forest, where leaf traits only predicted survivorship and growth of a limited group of pioneer tree species (Mart\u00ednez-Garza et al., 2013).Notably, LDMC, a \"soft\" trait, was correlated with both A max and instantaneous WUE, and LDMC could be an acceptable proxy for traits requiring specialized equipment to measure.\u0394\u03b4 13 C, was not correlated with any \"soft\" leaf traits, but is inexpensive to quantify.\nAt other ontogenetic stages, but not through the conclusion of the experiment, species with conservative trait values (high instantaneous WUE; Lohbeck et al., 2013;high wood density;Lohbeck et al., 2013;high C and low N%;Ostertag et al., 2015) were most likely to survive (Table 2).However, we also found species with high survivorship had photosynthetic traits on the acquisitive end of the spectrum (high A max and g s , low R dark ; Bazzaz & Pickett, 1980; 2).In this region, community weighted mean wood densities were typically highest during early succession (Becknell & Powers, 2014), perhaps because legumes, abundant in young regenerating forests (M.G.Gei, pers.\ncomm.), have high wood densities (Powers & Tiffin, 2010).Notably, we found four species with the highest survivorship were legumes.The last group of traits with conservative values, while only predictive of survival during the first wet season, showed species with high foliar C and low N % had higher overall survival (Table 2).By contrast, high-performing species had acquisitive strategies for some photosynthetic parameters.\nAfter the first dry season, species with acquisitive values (high) of A max and g s in the field were more likely to survive.This could be attributed to the fact that these species are better adapted to drought conditions in Vertisols as they were simultaneously able to maintain high levels of photosynthesis and integrated WUE (Table 2, Figure 5).\nThe same photosynthetic parameters (A max , g s and WUE) were also positively correlated with RGR ht during the second wet season (Table S5), consistent with the idea that species best adapted to Vertisols were those with the highest RGR during the first full wet season after planting.Trees with higher maximum adult heights appear to have lower RGR ht in every period except for the second wet season (p < .05,NS with Bonferroni correction; Table S5), consistent with our observation that short stature trees are dominant at our site (Werden, pers. obs.).Lastly, R dark was negatively correlated with survival from the first dry season forward, and with RGR vol during the first wet season, and we are not aware of a mechanism that explains this result.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.28", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "No soil amendment significantly increased overall survivorship to 2 years (Figure 4e), although the exceptional drought during our experiment may have decreased amendment effectiveness.A caveat to our study is that, for logistic reasons, we did not have sufficient seedlings to plant replicate treatments in multiple blocks.Additionally, while there were significant differences in clay and sand percentages among blocks, Tukey's HSD tests suggest pairwise differences in soil properties among blocks were slight, as only three marginally significant differences were found when comparing between blocks (Appendix S2).Therefore, these uncontrolled variables may have affected our results, but they do not invalidate our main conclusion that no differences among amendments were found after 2 years.We did find seedlings planted with hydrogel always had the highest survival by rank (Figures 3 and4), consistent with an experiment that found hydrogel to increase seedling survivorship in TDF (Fajardo et al., 2013).Both sand and hydrogel increased survivorship over the controls through the first wet season (Figure 3).We therefore recommend testing the effectiveness of applying sand and hydrogel simultaneously, as the two amendments serve different functions.Hydrogel also improved survivorship through the first dry season, and could be providing water to seedlings after dry season onset (Figure 4c).While sand and rice hulls increased RGR over the controls during transplant shock, and rice hull ash increased RGR during the first wet season, no soil amendment predictably influenced seedling growth (Table S8).Lastly, we estimate costs to apply the tested amendments are minimal, from ~$3 to $100 per hectare at standard planting densities (Appendix S1).", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.29", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Our results suggest functional trait-based screenings of tree species for restoration projects may help restoration practitioners efficiently select species for restoration projects.Furthermore, consistent with other studies, we found leaf traits are weak predictors of species' survival and growth, and the traits most predictive of survivorship and growth in TDF correspond to how species capture carbon and tolerate drought.In particular, species that upregulated integrated water-use efficiency (\u0394\u03b4 13 C) had the highest long-term survivorship and restoration potential.Lastly, our experiment helped to determine which native tree species and functional traits are important to consider when re-establishing TDF on degraded Vertisols.\nF I G U R E 5 Species survival percentages after the first dry season plotted against foliar \u0394\u03b4 13 C calculated using leaves collected from seedlings before and after planted R 2 = 0.4018 p = .000881", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.30", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Comments from Rakan A. Zahawi, Susan M. Galatowitsch, Adam Martin, and three anonymous reviewers greatly improved this manuscript.This experiment was supported by a NSF GRFP 11-582, GCA Restoration Fellowship, UMN Carolyn Crosby and Dayton grants (to L.K.W), and a NSF CAREER DEB-1053237 (to J.S.P.).Thanks to Daniel P\u00e9rez-Avil\u00e9s, G\u00e9raldine Derroire, Beatriz G. Exceed, Christina M. Smith, Ronald Castro and many volunteers for excellent field help, to Roberto Cordero (UNA Costa Rica) for use of the ADC instrument, and to Roger Blanco (ACG) for facilitating this work.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.38", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Comments from Rakan A. Zahawi, Susan M. Galatowitsch, Adam Martin, and three anonymous reviewers greatly improved this manuscript.This experiment was supported by a NSF GRFP 11-582, GCA Restoration Fellowship, UMN Carolyn Crosby and Dayton grants (to L.K.W), and a NSF CAREER DEB-1053237 (to J.S.P.).Thanks to Daniel P\u00e9rez-Avil\u00e9s, G\u00e9raldine Derroire, Beatriz G. Exceed, Christina M. Smith, Ronald Castro and many volunteers for excellent field help, to Roberto Cordero (UNA Costa Rica) for use of the ADC instrument, and to Roger Blanco (ACG) for facilitating this work.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.39", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "NSF, Grant/Award Number: GRFP 11-582 and CAREER DEB-1053237; GCA Restoration Fellowship; UMN Carolyn Crosby and Dayton grants clays that impede regeneration following degradation.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.40", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "NSF, Grant/Award Number: GRFP 11-582 and CAREER DEB-1053237; GCA Restoration Fellowship; UMN Carolyn Crosby and Dayton grants clays that impede regeneration following degradation.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.41", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Data available from the Dryad Digital Repository https://doi.org/10.5061/dryad.fd57r(Werden et al., 2017).", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.42", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Data available from the Dryad Digital Repository https://doi.org/10.5061/dryad.fd57r(Werden et al., 2017).", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.43", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "L.K.W., M.G.L. and J.S.P. designed the experiment.L.K.W., P.A.J., S.Z., E.C.M. and E.M.S. implemented the experiment and collected data.L.K.W. performed statistical analyses.L.K.W. and J.S.P. interpreted results and wrote the manuscript, and all others contributed to revisions.\nAdditional Supporting Information may be found online in the supporting information tab for this article.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.44", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "L.K.W., M.G.L. and J.S.P. designed the experiment.L.K.W., P.A.J., S.Z., E.C.M. and E.M.S. implemented the experiment and collected data.L.K.W. performed statistical analyses.L.K.W. and J.S.P. interpreted results and wrote the manuscript, and all others contributed to revisions.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.45", "keywords": [] }, { "file_name": "Werden et al. - 2018 - Using soil amendments and plant functional traits .pdf.tei.xml", "page_content": "Additional Supporting Information may be found online in the supporting information tab for this article.", "title": "Using soil amendments and plant functional traits to select native tropical dry forest species for the restoration of degraded Vertisols", "id": "7.46", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "Our study highlights the importance of wood density and landscape structure to the initial survival of rainforest plantings.Factors influencing seedling survival shifted over time but, most importantly, our results highlight that, when planting into abandoned pastures, it may be preferable to select species with higher wood densities to maximize survival during the crucial early stages of establishment and growth.", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.1", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "With the extensive loss of tropical forest over the last century and the current global increase in land abandonment (Cramer, Hobbs, & Standish, 2008;Munroe, van Berkel, Verburg, & Olson, 2013), ecological restoration to aid the recovery of degraded ecosystems is becoming increasingly important (Society for Ecological Restoration Science & Policy Working Group, 2002).Restoring tropical forested ecosystems commonly involves planting species to accelerate forest recovery (Chazdon, 2008).Despite the abundance of restoration theories and techniques, many restoration projects starting from abandoned pastures still fail to achieve desired goals (Lamb, Erskine, & Parrotta, 2005).\nThe large variability in outcomes among tropical forest restoration efforts is largely due to the many biotic and abiotic factors across spatial scales that impede early-stage seedling establishment and growth.\nSeedling establishment can be influenced by landscape-scale factors such as distance of plantings to established forest patches (Catterall, Kanowski, & Wardell-Johnson, 2008), slope, aspect (Nagamatsu, Seiwa, & Sakai, 2002) and soil fertility (Guariguata & Ostertag, 2001), but these factors have generally not been examined for early-stage seedling survival (but see Mart\u00ednez-Garza, Tobon, Campo, and Howe, 2013).Local-scale factors of the planting site, such as competition with resident grass and weed species, may also impede seedling establishment (Holl & Kappelle, 1999;Hooper, Condit, & Legendre, 2002).\nDespite evidence of their importance, restoration efforts that focus on these factors still vary considerably in whether they are successful (Bechara et al., 2016;Dudley, Mansourian, & Vallauri, 2005).\nTo mitigate these factors, much attention has focused on species selection.Projects differ markedly in how species are selected, particularly in their consideration of species' phenologies, functional traits and responses to local environments.Species selection based on successional status (e.g.early, late secondary or mature species) is recommended under some restoration frameworks (Goosem & Tucker, 1995;Rodrigues et al., 2011), but has largely been untested in replicated planting experiments (but see Mart\u00ednez-Garza, Bongers, and Poorter 2013).Other approaches, including species selection based on plant functional traits (Laughlin, 2014) have not yet been widely incorporated into restoration practice.These trait-based approaches are motivated in part by previous studies linking traits with increased survival and growth in restoration plantings and mature forests.For instance, Mart\u00ednez-Garza, Pe\u00f1a, Ricker, Campos, and Howe (2005) recorded increased growth and survival of planted seedlings with increased variation in specific leaf mass, while Poorter et al. (2008) and Wright et al., (2010) found greater seedling survival for species with higher wood density and seed volume in mature-phase tropical forests.Additionally, comparisons of seedling establishment between monoculture and mixed diversity plantings have also been examined (Piotto, 2008;Plath, Mody, Potvin, & Dorn, 2011), yet comparisons between low and high levels of diversity within mixed species plots in restoration plantings are still lacking.\nThe early stage of seedling establishment (typically between 1 and 3 years) is a strong predictor of mid-term performance of restoration plantings (Mart\u00ednez-Garza, Bongers, et al., 2013;Montagnini, Gonzalez, Porras, & Rheingans, 1995).While seedlings during the initial stages of restoration plantings are highly susceptible to many biotic and abiotic factors, these impacts can be exacerbated by transplanting shock (Burdett, 1990) and poor planting conditions, such as inappropriate seedling handling and planting techniques (Grossnickle, 2005).\nConsequently, seedling mortality during the establishment phase can alter early successional trajectories and potentially lead to arrested forest recovery (Lamb, 2011).\nThe paucity of research into early-stage seedling establishment under different diversity planting regimes (Lamb & Lawrence, 1993), coupled with the distinct lack of monitoring of seedling survival and growth in many restoration projects (Kanowski, Catterall, Freebody, Freeman, & Harrison, 2010), underscores the need for more research into the factors that influence early-stage seedling establishment and future forest recovery.\nThe aim of this study was to increase understanding of the factors influencing seedling survival, and how these factors change over time.\nWe present survival data recorded during the first two and a half years of a rainforest restoration experiment in the Wet Tropics of Australia.\nThe specific questions we ask are: (1) How important are landscape, site and planting conditions to seedling survival?(2) Do species functional traits explain variation in seedling survival?and (3) Do these factors change during the first two and a half years of seedling establishment?\nGiven the wide range of factors that can influence early-stage seedling survival, we predicted a similarly large diversity of factors responsible for explaining survival patterns.In particular, we hypothesis that seedling survival will be highest on shallow slopes near forest, based on previous findings in the literature (Catterall et al., 2008;Nagamatsu et al., 2002).Due to the commonly observed relationship between the functional traits of species and seedling survival, we also hypothesized that species with high wood density and larger seed masses will have higher survival rates than species with low wood densities and small seeds (Baraloto, Forget, & Goldberg, 2005;Poorter et al., 2008).", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.2", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "This study was undertaken on the Thiaki Creek Nature Reserve (Figure 1), located on the southern Atherton Tablelands of far North Queensland, Australia (145\u00b051\u2032E 17\u00b043\u2032S), between 900 and 1,000 m\nfunctional traits, plantings, rainforest, restoration, revegetation, seedling establishment, succession, wood density elevation, with an average annual rainfall of 1,400 mm (Bell, Winter, Pahl, & Atherton, 1987).The underlying substrate is basaltic (Tracey, 1982) with a varied topography of narrow valleys surrounded by 15-45\u00b0 slopes.The 181-ha reserve comprises 130 ha of primary and mature secondary rainforest, classified as Endangered Regional Ecosystem 7.8.4,Upper Barron complex notophyll vine forest (Bell et al., 1987) and 51 ha of abandoned pasture which was the focus of the restoration experiment established in January 2011.The pasture area had a consistent land use history (remnant forest cleared for grazing c. 50 years ago) and the pasture itself was relatively homogeneous in composition prior to planting.", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.4", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "In January 2011, approximately 28,000 rainforest seedlings were row planted in a randomized complete block design consisting of eight blocks each containing eight 50 m \u00d7 50 m plots separated by a 10-m buffer zone.The experiment included a two-level planting density treatment: low-density (seedlings spaced 3 m apart) and high-density (seedlings spaced 1.75 m apart) and a three-level species diversity treatment (1, 6 and 24 species).The diversity treatment was phylogenetically nested with 24 local rainforest tree species from six common families: Lauraceae, Moraceae, Myrtaceae, Proteaceae, Rutaceae and Sapindaceae (Table S1).All monoculture plots were composed of Flindersia brayleyana (Rutaceae).We recognize that this decision prevented any general conclusions about seedling mortality in monoculture and thus deal with data on this species (and thus monoculture plots) separately from the higher diversity treatment plots.The multispecies \"low-diversity\" treatment included one species from each of the six focal families.The high diversity treatment included all 24 species, four species from each family and included all those species present in the low-diversity treatment.Flindersia brayleyana was used in all three treatments.Within each experimental block, we assigned two control treatment plots, with no planted seedlings.Plots were randomly assigned treatments within each block (Figure 1).Species were selected based on their commonality in local intact rainforest, availability in local nurseries and a suit of functional traits selected to match common successional stages.Specifically, we selected species in each focal family to span a broad range of trait values for the following traits: maximum tree height, wood density, dispersal vector, seed size and successional stage.Classifications of successional stages included early, late secondary and mature stage species.Early successional species are typically short-lived, fast growing, shade intolerant species and display both low wood density and small seed mass (Hopkins, Kikkawa, Graham, Tracey, & Webb, 1979).Mature stage species are shade tolerant, longer lived, slow growing species with high wood densities and large seeds (Hopkins et al., 1979).Early-and mature-stage species are functionally analogous with pioneer and non-pioneer species, respectively, as described by Whitmore (1989).Late secondary species display moderate growth rates and increasing shade intolerance with age (Hopkins et al., 1979) and are functionally equivalent to longer lived pioneer species (Whitmore, 1989).\nSpecies from both 6-and 24-species diversity treatments were randomly assigned within rows in each plot.After cattle removal and prior to planting, the pasture was dominated by the exotic grass Urochloa decumbens.We applied a monocot-specific herbicide (Fusillade) along designated planting rows of all treatment plots (excluding one control plot in each block), leaving alternate rows of unsprayed grass to reduce the potential for run-off and erosion.Herbicide application within rows was continued every 6 months until August 2012.To avoid potential desiccation of seedlings, planting was conducted during the regional", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.5", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "Slope and aspect of the plots were measured using a clinometer and compass respectively.We measured the distance from the centre point of plots to the nearest surrounding forest fragments using ArcGIS (ESRI, 2006).The identity of the person responsible for planting individual seedlings (hereafter referred to as planter identity) and the day until first rain were also recorded (plots differed in the time between planting and rainfall).Wood density estimates for all planted species were obtained from Chave et al. (2009) and Zanne et al. (2009).Dry seed mass values were obtained from the literature and the Royal Botanic Gardens Kew Seed information database (Royal Botanic Gardens Kew, 2017).Maximum adult height values and dispersal vectors were obtained from Cooper and Cooper (2004).The complete list of explanatory variables examined in this study can be found in Table S2.", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.6", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "To minimize edge effects within experimental plots, the survival of seedlings was recorded only in a central 25 m \u00d7 25 m subplot within each treatment plot.Seedling status was recorded by visual inspection, with seedlings deemed to be dead if there was pronounced stem desiccation, no leaves and/or the seedling could not be located.\nSeedling survival was assessed at 6-month intervals starting 4 months after planting in April 2011 until July 2013.", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.7", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "Seedling survival (binary response, 1 = living, 0 = dead) was modelled as a function of planting regime, landscape-and site-scale variables, and species' functional traits (Table S2) using generalized linear mixedeffect models with logit link function and binomial error distribution.\nBlock and plot (nested within block) and family and species (nested within family) were included as random effects in all models.Twoway interactions between functional traits (wood density, seed mass and maximum tree height) and plot slope, distance to forest and day until first rain were included in all models to test for specieslevel differences in their responses to different factors.Plot aspect was not included in any interactions due to the uneven distribution of south-facing (31%) compared to north-facing plots (69%).While wood density, seed mass and maximum tree height were included as species-level variables, they likely also vary by family.As such, we first ran a variance component analysis which indicated that c. 35% of variance in both wood density and seed mass occurred between families and 65% within families (Table S3), though much of the within-family variation was driven by two families-Lauraceae and Proteaceae for wood density and Lauraceae and Moraceae for seed mass (Figure S1).\nThus, our wood density and seed mass variables captured both species and family level differences.Variation in maximum tree height occurred entirely within families (Table S3).Before fitting models, there was no evidence of correlation among explanatory variables (Pearson's product-moment correlation, Table S4).\nSeedling survival was analysed separately for three time periods representing early-stage survival (0-4, 4-9 and 9-31 months post planting).Due to low seedling survival within the first 4 months, the density of planted seedlings within plots changed considerably, thus density was excluded as an explanatory variable from all mixed-effects models.Monoculture treatment plots (containing F. brayleyana) were excluded from analyses due to lack of variation in family and functional traits.Separate analyses were conducted for F. brayleyana survival across all plots as it was the only species included in a monoculture treatment.Maximal models containing all relevant explanatory variables for each time period were simplified by removing non-significant terms (determine by Wald Z tests) one at a time.Data analysis was conducted using the r statistical software package version 3.3.3,using the lme4 package (Bates, Maechler, & Bolker, 2011;R Development Core Team, 2010).Post hoc tests of pairwise treatment differences were conducted for F. brayleyana seedling survival using the glht function in the multcomp package (Hothorn, Bretz, & Westfall, 2008).", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.8", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "Of the 6,657 seedlings for which data were collected in this experiment, only 3,544 survived (53.2%) to 31 months post planting.\nOverall probability of seedling survival varied by family, with species from Moraceae having the highest probability of survival (0.7) and Lauraceae (0.33) the lowest (Table S5).Survival probability differed considerably between species within families as well (Table S5) with Stenocarpus sinuatus (0.87), Rhodamnia sessiliflora (0.8) and Guioa lasioneura (0.8), the top surviving species, and Cryptocarya oblata (0.09) and Melicope jonesii (0.09) experiencing the lowest survival.", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.10", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "Seedling survival was affected by multiple factors, and the relative importance of these factors changed through time.At 0-4 months, seedling survival was marginally higher in south-facing than northfacing slopes, with significantly higher survival in plots close to forest fragments than those further from forests.This distance effect was most pronounced for low wood density species (Figure 2, Table S6).\nSurvival of F. brayleyana in monoculture, low-and high-diversity plots over the same period was best explained by slope, with greater seedling survival on steeper slopes (Figure S2 and Table S7).\nFor the 4-to 9-month period, seedling survival was best explained by species' wood density, slope, distance to forest and planter identity (Table S8).Like the preceding time period, seedling survival decreased with distance from forest fragments (Figure S3).Seedling survival declined with increasing slope, and was lower for species with low wood density (Figure 3a), though there were no significant interactions between wood density and other variables during this time period.Survival also varied significantly among planters (Figure 3b).No explanatory variables proved to be good predictors of F. brayleyana survival during this time period.\nDuring the final 9-31 months, survival was best explained by distance to forest, planter identity and the interaction between slope and wood density (Table S9).The significant interaction between wood density and slope (Figure 4a) indicated that high wood density species survived better on shallow slopes, but on steep slopes all species experienced similarly low survival, regardless of wood density.\nSimilar to the previous time periods, survival varied among planter identity, but the order of planters changed through time (planter 5 had lowest survival during 4-9 months, whereas planter 1 had lowest survival during 9-31 months, Figure 4b).Flindersia brayleyana survival during the final period was best explained by the diversity treatment (Table S10), with survival significantly lower in monocultures than in the 6-and 24-species treatments (Figure 5, Table S11).\nApart from wood density, no other functional trait explained variation in seedling survival over the 31-month observation period.", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.11", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "The probability of seedling survival during early-stage restoration was influenced by multiple interacting landscape and biological factors, F I G U R E 2 Probability of seedling survival for the 0-to 4-month time period in relation to distance to nearest forest fragment (m) with high and low levels of species' wood density (g/cm 3 ).Fitted line values for low and high wood density are 0.38 and 0.84 g/cm 3 respectively.Points for high and low wood density were calculated from the upper and lower third of wood density values respectively.Bars are associated SEs on the probability scale.Shaded bands represent 95% confidence intervals F I G U R E 3 Probability of seedling survival 4-9 months post planting in relation to slope (\u00b0) and (a) wood density (g/cm 3 ) and (b) planter ID, with separate lines fitted for each planter ID.Fitted line values for low and high wood density are 0.38 and 0.84 g/cm 3 respectively.Points for high and low wood density were calculated from the upper and lower third of wood density values respectively.Shaded bands represent 95% confidence intervals.Points for planter ID are mean probabilities calculated from seven bins of ordered binary values.All bars are associated standard errors on the probability scale with the strength of some factors changing over time.Only wood density, a species-level predictor, and the distance of plantings to forest fragments consistently influenced seedling survival throughout the 31-month observation period.Overall, seedlings with high wood densities planted near intact forest patches had the greatest probability of survival.This positive wood density effect suggests that when planting into abandoned pastures it may be preferable to include species with higher wood densities to maximize survival during the crucial early stages of establishment and growth.", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.12", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "Throughout this study, wood density had a consistently positive effect on seedling survival.Our results are consistent with seedling survival-growth trade-offs identified in other tropical rainforest species (Chave et al., 2009;Poorter et al., 2008), with higher wood density species displaying higher survival than species with lower wood densities, at the expense of slower growth (Kraft, Metz, Condit, & Chave, 2010;Nascimento et al., 2005).The very high levels of light and heat exposure typically experienced in open tropical pasture may favour species with higher wood densities that typically have reduced risk of xylem implosion under water stress than low wood density species (Hacke, Sperry, Pockman, Davis, & McCulloh, 2001;Sperry, Meinzer, & McCulloh, 2008).Contrary to our expectations, seed mass did not explain variation in early-stage seedling survival.Surprisingly, we observed a negative relationship between seed mass and survival (a relationship also found by Mart\u00ednez-Garza, Bongers, et al., 2013), although this relationship was non-significant.This result is counter to survival patterns observed in plantings (Baraloto et al., 2005) and mature forests (King, Davies, Tan, & Noor, 2006), whereby larger seed mass is correlated with higher survival rates in seedlings.\nDespite strong evidence of a survival-growth trade-off in tropical trees, early successional species with low wood densities are often selected for restoration projects under the assumption that restoration S8) plantings follow the same recovery trajectory as naturally regenerating forests, notably that early successional species start the regeneration process and later successional species move in to replace them once fast growing early successional species have created a closed canopy and start to die out (Palmer, Ambrose, & Poff, 1997).However, there is increasing evidence that restoration projects do not always follow this trajectory (Griscom & Ashton, 2011;Lamb, 2011).Though early successional species are clearly important in the natural regeneration of many forest systems, our results (along with others; Bonilla-Moheno & Holl, 2010;Hooper et al., 2002;Montagnini et al., 1995) suggest that they may not be the best choice for planting-based forest restoration.\nRarely are funds sufficient for supplementary plantings in restoration planting projects (Chazdon, 2008;Holl & Aide, 2011) and natural recruitment can be slow in plantings undertaken in open pasture (Florentine & Westbrooke, 2004).If plantings rely heavily on early successional species with fast growth and high mortality rates, canopies may never actually form, or may reopen over time as plants senesce and recruitment fails to fill gaps.Our results suggest that when restoration funding is limited, planting high wood dense species with slower growth rates may help to maximize mid-term restoration success.", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.13", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "Seedling survival decreased with distance from forest fragments throughout the observation period.Air temperature, vapour pressure deficit and soil temperature can increase with distance from forest (Pareliussen, Olsson, & Armbruster, 2006).Forest edges can provide intermittent shading throughout the day and this reduction in sun exposure and evaporation may provide a more suitable microhabitat for survival and/or growth of young seedlings (Duncan & Duncan, 2000).This is likely important for both early and late successional species in reducing stress levels as discussed above.The impact of distance from forest was stronger for species with low wood densities, further supporting our conclusion that reliance on early successional species with lower wood densities may reduce the likelihood of restoration success in exposed restoration sites.\nThere is some evidence that herbivore pressure can increase with distance to forest as well, though this appears somewhat systemspecific (Myster & McCarthy, 1989;Ostfeld, Manson, & Canham, 1997).While herbivore damage was not directly measured in our experiment, herbivory was not observed to be substantial in any of our experimental plots (L.Charles, T. Smith pers.observ.).Given that many global forest restoration efforts are conducted in abandoned pastures, the effect of planting distance from adjacent forests on seedling mortality is relevant regardless of the root cause.The lower probability of seedling survival in exposed pastures highlights the importance of targeted management schemes, such as enrichment plantings expanding out from existing fragments, or selecting species with appropriate physiological adaptations for use further away from forest fragments.", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.14", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "Consistent with previous studies (Daws, Pearson, Burslem, Mullins, & Dalling, 2005;Nagamatsu et al., 2002), during the 4-to 9-month time period, seedlings planted on steep slopes experienced lower survival compared to seedlings planted on shallow slopes (Figure 3a).This trend became more pronounced with time for all species (Figure 4a).Typical factors associated with steeper slopes include increased water run-off (Snyman & Van Rensburg, 1986), soil erosion and reduced water retention (Meyer & Wischmeier, 1969).At this site, grass strips were intentionally retained to minimize erosion and run-off.Observed increases in seedling mortality on steeper slopes, however, suggest that either these strips were ineffective or that other abiotic factors caused negative associations with slope.To further investigate why mortality was higher on steep slopes at this site, we conducted post hoc linear regression analyses of soil samples collected from each experimental plot.In this analysis, we found no relationship between slope and soil temperature, pH, electrical conductivity or available nitrogen (Figure S4).It is, or course, possible that unmeasured soil factors created the observed slope effect, such as historical nutrient run-off, erosion and compaction associated with initial forest clearing and decades of cattle grazing.In addition, planting may have been more difficult on steep slopes, resulting in suboptimal planting techniques, which likely explains some of the variable performance of individual planters.The planter-mediated effect could also have been exacerbated by drier conditions on steep slopes where drainage and soil moisture holding capacity are probably lower, factors not measured here but worth assessing in the future.As we did not score \"planting quality\" per individual tree, it is hard, however, to disentangle such effects.\nRegardless of the cause, this finding is particularly important for tropical restoration given that many passive and active forest recovery projects are located on steep terrain (Asner, Rudel, Aide, Defries, & Emerson, 2009).Again identifying local species that survive and grow on steep slopes may be key to improving success rates of rainforest restoration.", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.15", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "Consistent with previous studies (Armesto & Martinez, 1978), we found a significant (though weak) relationship between aspect and seedling survival during the first 4 months post planting, with the probability of seedling survival marginally higher on south-facing slopes (Table S6).North-facing slopes in the southern hemisphere consistently experience more direct solar radiation than south-facing slopes (Tian, Davies-Colley, Gong, & Thorrold, 2001) which can increase surface temperature and moisture evaporation (Monteith, 1965) and lead to less favourable environments for planted seedlings (Turton & Freiberger, 1997).However, in the absence of both westand east-facing plots in our experimental landscape, we advise that these results do not give a full picture of seedling performance across more topographically heterogeneous landscapes.The aspect result may also reflect differences in light tolerance among our study species, a trait we did not assess.While there is evidence that seedling light tolerance can impact survival rates within plantings (Augspurger, 1984), many species are able to acclimate to high-light conditions (Loik & Holl, 1999).Given that these species-specific responses have largely been untested in restoration experiments, and given our strong wood density results, future studies of species-specific responses to highlight conditions are clearly warranted.", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.16", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "Following the first 4 months post planting, the probability of seedling survival was somewhat dependent on who planted the seedling (planter identity; Figures 3b and4b).The absence of a planter effect in the first time period is surprising given past research quantifying planter effects (Rietveld, 1989).The cyclone-induced dry period during planting may have masked short-term planter effects, by inducing immediate stress responses in all seedlings following planting, regardless of planter identity.Past studies have found that poor planter technique can damage roots and reduce root contact with the soil, which causes a reduction in a plant's ability to obtain water many months after planting (Burdett, 1990;Grossnickle, 2012).The professional planters used in this study included a mix of experienced planters (5+ years planting experience) and novices (mostly backpackers with <1 season of planting experience).Though we did not quantitatively evaluate planter background, our qualitative observations were that experienced planters were noticeably better tree planters.Our results suggest that if using untrained planters, training and quality control measures are worthwhile to improve seedling survival.", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.17", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "There was no observed effect of planting diversity on seedling mortality.When analysed separately, however, the probability of F. brayleyana seedling survival across all three diversity treatments was significantly lower in the monoculture plots than in the six species diversity plots (Figure 5), at least during the latest time period.This result contrasts with past studies comparing survival between monoculture and mixed species plantings within the neotropics (Piotto, V\u0131quez, Montagnini, & Kanninen, 2004;Plath et al., 2011;Potvin & Gotelli, 2008).Seedling mortality in monoculture stands can result from increased herbivore or pathogen susceptibility (Jactel, Brockerhoff, & Duelli, 2005), though no obvious herbivores and pathogens were evident in our plots.Past studies have suggested that mortality in monocultures is species-specific and thus our monoculture treatment cannot be generalized beyond F. brayleyana (Forrester, Bauhus, & Khanna, 2004).Results for this species are important, however, as F. brayleyana is commonly used in lowdiversity timber plantations and more diverse restoration plantings.", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.18", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "Restoration projects can be expensive and labour-intensive, with no real guarantee of success.This situation is further impeded by the lack of large-scale, controlled studies of seedling performance in diverse types of restoration projects.Our study demonstrates that seedling survival is influenced by both biological and landscape factors, some of which are persistent, while others are transient.In particular, we show that planted species and sites (context, slope, aspect) can be carefully selected to maximize early-stage seedling establishment, and to reduce the probability of delayed forest recovery.We found no evidence that planting large numbers of species ( 24) improved seedling survival rates early in planting establishment, despite higher diversity being advantageous in later successional stages and being an increasingly common approach to restoration plantings in Australia's tropics (and further afield).Our results also suggest that to maximize seedling survival in the first few years post planting, it is important to use experienced planters, use a high proportion of species with high not low wood densities (or species known to be robust to local planting conditions), and where possible, select restoration sites adjacent to existing forest patches.Though this experiment took place in Australia's wet tropics, many of our findings are likely applicable to other tropical systems, given that our results were robust across a diversity of tropical tree families.Understanding the temporal shift in the influence of species and site selection on seeding establishment can aid in the development of targeted management strategies both before and post planting to maximize restoration success.", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.19", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "We thank Noel Preece and Penny van Oosterzee for access to the Thiaki Creek property and logistical support during this project.Thanks to Alana Burley, David Chittleborough, Corey Bradshaw, Peter Erskine, Michael Lawes, Noel Preece and Penny van Oosterzee for contributions to the design and implementation of the Thiaki experimental plots and to Monica Radovski, Loy Xingwen, John Park and Alex Lindsey for assistance with field work.We also thank the reviewers and editor of Axios for helpful feedback on this manuscript.This work was made possible by the Australian Research Council (LP0989161).", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.24", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "We thank Noel Preece and Penny van Oosterzee for access to the Thiaki Creek property and logistical support during this project.Thanks to Alana Burley, David Chittleborough, Corey Bradshaw, Peter Erskine, Michael Lawes, Noel Preece and Penny van Oosterzee for contributions to the design and implementation of the Thiaki experimental plots and to Monica Radovski, Loy Xingwen, John Park and Alex Lindsey for assistance with field work.We also thank the reviewers and editor of Axios for helpful feedback on this manuscript.This work was made possible by the Australian Research Council (LP0989161).", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.25", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "Data available from the Dryad Digital Repository https://doi.org/10.5061/dryad.t1rv6(Charles et al., 2017).\nhttp://orcid.org/0000-0003-0055-2510", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.28", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "Data available from the Dryad Digital Repository https://doi.org/10.5061/dryad.t1rv6(Charles et al., 2017).", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.29", "keywords": [] }, { "file_name": "Charles et al. - 2018 - Species wood density and the location of planted s.pdf.tei.xml", "page_content": "http://orcid.org/0000-0003-0055-2510", "title": "Species wood density and the location of planted seedlings drive early\u2010stage seedling survival during tropical forest restoration", "id": "8.31", "keywords": [] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "Regenerating secondary forests are now dominant globally (FAO 2015) and much effort has been devoted to understanding the drivers of forest regeneration across the tropics (Chazdon 2014).A complex interplay between predictable and unpredictable factors leads to shifts in vegetation structure during tropical forest succession (Chazdon 2008).Deterministic factors that affect the rate and dynamics of succession include abiotic resources such as light and water availability, in addition to microclimatic conditions such as air and soil temperature (Loik and Holl 2001).These abiotic environmental factors change over the course of secondary Manuscript received 8 July 2019; revised 16 January 2020; accepted 6 February 2020.Corresponding Editor: Carolyn H. Sieg. 6Present address: Lyon Arboretum University of Hawai'i at M anoa 96822 HI, USA 7 E-mail: lwerden@gmail.com\nArticle 2116; page 1 forest regeneration (Lebrija-Trejos et al. 2011), in turn affecting biogeochemical cycles and ecosystem processes such as carbon storage and fluxes (Powers and Mar \u0131n-Spiotta 2017).Because microclimatic conditions (e.g., light and water availability) can change dramatically as a forest ages and its canopy develops (Lebrija-Trejos et al. 2011), these conditions are strong filters of species assembly during tropical forest succession (Lebrija-Trejos et al. 2011, Lohbeck et al. 2013, Buzzard et al. 2016).Furthermore, morphological and physiological strategies of tree species follow predictable patterns as microclimate changes during succession; consequently, as forests age, trees face shifting limitations in resources such as light and water (Bazzaz andPickett 1980, Montgomery andChazdon 2002).\nIt is well established that light availability is the most limiting environmental factor in tropical wet forest succession, contributing to a transition from fast-growing, light-demanding species to slow-growing, shade-tolerant species (Bazzaz and Pickett 1980, Guariguata and Ostertag 2001, Montgomery and Chazdon 2002, Boukili and Chazdon 2017).By contrast, hypothesized amelioration of harsh microclimatic conditions as TDFs mature has led to the prediction that water availability may more strongly influence TDF successional processes than gradients in light availability (Lebrija-Trejos et al. 2010, Lebrija-Trejos et al. 2011, Lohbeck et al. 2013, Pineda-Garcia et al. 2013, Derroire et al. 2018).In theory, TDF species that are able to cope with low soil moisture and/ or high temperatures and vapor pressure deficits should be dominant at earlier TDF successional stages (Lohbeck et al. 2013).Therefore, changes in microclimatic conditions during succession may play a major role in structuring TDF plant communities (Lebrija-Trejos et al. 2011, Jackson et al. 2018).Determining how species' resource-use traits mediate responses to shifts in abiotic gradients may clarify the mechanisms underlying TDF plant community assembly, which are currently not well understood (Quesada et al. 2009).\nIt has been argued that all plant species fall somewhere on a \"fast-slow\" plant economics spectrum, and species tolerant of drought and high temperatures fall on the slow, or conservative, end of the spectrum (Reich 2014).In this framework, acquisitive species maximize resource capture and are more sensitive to abiotic stress, while conservative species use resources more efficiently and are more stress-tolerant.Aligning with this expectation, some observational studies have shown that during succession TDF communities transition from being dominated by species with conservative leaf traits (e.g., smaller, thicker leaves) to species with acquisitive traits (e.g., larger, thinner leaves), both for adult trees (Becknell andPowers 2014, Buzzard et al. 2016) and saplings (Derroire et al. 2018), but some show the opposite trend (Lohbeck et al. 2013, Subedi et al. 2019).Similar patterns have been observed for some stem traits, with TDF tree communities shifting from having conservative (dense) to acquisitive (soft) wood density values over time (Lohbeck et al. 2013, Poorter et al. 2019, but see Subedi et al. 2019).Generally, these studies argue that TDF community assembly patterns are driven by water limitation at earlysuccessional stages, leading to the dominance of species with conservative resource-use strategies (drought and temperature tolerant) in early-successional environments such as completely cleared pastures, eventually transitioning to dominance by species with acquisitive strategies (more profligate resource-use, less stress-tolerant; Lebrija-Trejos et al. 2010, Lohbeck et al. 2013).While studies have demonstrated how TDF functional composition shifts over time, the mechanisms underlying how species-level resource-use strategies dictate performance (i.e., survival and growth) during succession have not been explicitly investigated.Active restoration experiments provide excellent opportunities to test these theories experimentally (Howe and Mart \u0131nez-Garza 2014), while simultaneously promoting regrowth of the critically endangered TDF biome (Janzen 1988).\nWe implemented a 6-ha TDF restoration experiment in northwestern Costa Rica to test how tree species with differing resource-use strategies (acquisitive or conservative) respond when planted into TDF successional stages with differing microclimatic conditions.We used two management treatments to simulate two distinct successional stages including: clearing all remnant vegetation to mimic early-successional environments (cleared treatment), or interplanting seedlings into plots with minimal site preparation to reflect a later, mid-successional environment (interplanted treatment).The goals of our experiment were to determine (1) how TDF tree species with different resource-use strategies perform in the presence or absence of remnant vegetation and (2) how ecophysiological (water-use and photosynthetic) functional traits correlate with tree performance in these simulated successional stages.\nFor our first goal, based on recent findings in secondary TDFs (Lohbeck et al. 2013, Buzzard et al. 2016, Derroire et al. 2018), we hypothesized that species with conservative trait syndromes would have higher performance (survival and growth rates) than acquisitive species when planted into an early-successional stage (lower water availability).This expectation was based on the understanding that conservative species are generally more tolerant of drought and temperature stress, enabled by strategies such as low specific leaf area and higher water-use efficiency (Niinemets 2001).By contrast, we hypothesized species with acquisitive trait syndromes would have higher performance than conservative species in a mid-successional stage (higher water availability) because they would be released from water limitation.Thus, we expected an interaction between trait syndromes and successional stage.Furthermore, there is growing evidence that water-use and photosynthetic traits, hereafter referred to as ecophysiological traits, are better predictors of species performance in regenerating tropical wet (Guimar\u00e3es et al. 2018) and dry forests (Werden et al. 2018b) than easy to measure \"soft\" leaf and stem traits.Consequentially, for our second goal we hypothesized that ecophysiological functional traits related to species' abilities to tolerate drought conditions in Vertisols (Werden et al. 2018b), soils with characteristic shrink-swell cycles (Deckers et al. 2001), would be stronger predictors of seedling performance than resource-use strategies defined with leaf and stem traits commonly used to examine shifts in functional composition during tropical forest succession.Additionally, we expected the strength of the relationships between species performance and ecophysiological traits would be stronger in the cleared than the interplanted treatment, due to amelioration of abiotic conditions by remnant vegetation as succession progresses (Lebrija-Trejos et al. 2011).", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.1", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "We conducted our experiment from 2015 to 2017 at Estaci on Experimental Forestal Horizontes (10.712\u00b0N, 85.594\u00b0W), part of Area de Conservaci on Guanacaste in northwestern Costa Rica.The mean annual precipitation in this region is 1,730 mm, typically with a 5-6 month dry season with little to no rain (December or January-May; data available online).8In 2015, this region of Costa Rica experienced the strongest drought on record (Instituto Meterol ogico Nacional de Costa Rica 2015) and a meteorological station 1 km from our site recorded only 807 mm of precipitation for the year (53% lower than annual mean).The timing of precipitation in 2015 was irregular as well; wet season rains began in June but little rain fell in July and August, before consistent precipitation began in early September.Total precipitation in 2016 was 1754 mm (very close to the mean) and followed typical timing patterns (consistent precipitation June-November).The experiment was conducted on degraded Vertisols, soils that impede regeneration of large areas of TDF and are particularly difficult to restore because of low seedling survival rates (Werden et al. 2018b).The study site was previously used for cattle grazing, likely for decades, similar to other TDFs in this region (Janzen 1988) and across Central America (Griscom and Ashton 2011).Naturally regenerating vegetation at the site was in a state of arrested succession following abandonment for ~30 yr, which is typical for TDF on Vertisols in this region (M.Guti errez, personal observation).", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.3", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "We chose 12 focal species for plantings (Table 1), which commonly occur in TDFs in this region (Powers et al. 2009) and are able to persist on Vertisols (Werden et al. 2018b), with a range of resource-use and life history functional traits (Powers andTiffin 2010, Werden et al. 2018b).For these focal species, we used leaf and stem functional traits to define species' resource-use syndromes and to determine how ecophysiological traits explain species-level performance, thus allowing us to address our two primary study objectives (Table 2).To accomplish this we used a database of functional trait data collected in 2014 using standard methods (P erez-Harguindeguy et al. 2013) for a previous study implemented at an adjacent site on the same soil type (Werden et al. 2018a, b).The functional trait data we used (Werden et al. 2018a, b) were collected 1 yr prior to the implementation of our restoration experiment in 2015, however, the environmental conditions experienced by seedlings were similar in both years due to an ongoing drought in the region (total precipitation 535 and 923 mm lower than average in 2014 and 2015, respectively).Additionally, all trait data were collected from seedlings grown in the same shade house where seedlings were produced for this experiment, or on seedlings planted <200 m from where this experiment was implemented.Because trait data from Werden et al. (2018a, b) were collected under similar environmental conditions, and in a location adjacent to our experiment, using these data in our experiment is unlikely to have influenced our results or conclusions.\nFrom Werden et al. (2018b), we used the following traits commonly used to define resource-use strategies of tropical tree species (Goal 1): leaf thickness (mm), LDMC (leaf dry matter content; g/g), petiole length (mm), SLA (specific leaf area; cm 2 /g), wood density (g/ cm 3 ), foliar nitrogen concentration (%; foliar N).We used the following ecophysiological traits to address Goal 2: g s (stomatal conductance; mmol\u00c1m \u00c02 \u00c1s \u00c01 ), photosynthetic capacity (A max ; lmol CO 2 \u00c1m \u00c02 \u00c1s \u00c01 ), iWUE (instantaneous water-use efficiency; lmol CO 2 /mmol H 2 O), D d 13 C (upregulation of integrated water-use efficiency; &).These ecophysiological traits were used to indicate how the focal species acquire carbon (A max ) and cope with water stress (iWUE, g s , D d 13 C).For additional details on how functional trait data were collected, see Appendix S1: Section S1.\nWe defined acquisitive and conservative trait syndromes for the 12 focal species using six traits demonstrated to shift from conservative to acquisitive values during TDF succession (Lohbeck et al. 2013, Buzzard et al. 2016, Derroire et al. 2018; Goal 1 in Table 2).The 12 focal species were assigned to acquisitive or conservative groups using hierarchical cluster analysis (Appendix S1: Section S2, Figs.S1-S3).In the cluster analysis, we considered the six resource-use strategy leaf and stem traits (Goal 1; Table 2).This resulted in one resource acquisitive and one resource conservative group, with six species per group (Table 1).Consistent with general expectations of species with conservative resource-use strategies, species in the conservative group had lower SLA, thicker leaves, and lower foliar nitrogen concentrations.Petiole length, LDMC, and wood density did not differ between the acquisitive and conservative groups (Appendix S1: Section S2).For more details on the methods used to define resource-use strategy groups, including functional traits means for each group, see Appendix S1: Section S2.", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.4", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "We crossed two management treatments (cleared/interplanted) with two resource-use strategy species mixes (acquisitive/conservative) to test for their interaction.We randomly assigned treatments to six 100 9 100 m plots, for a total of three interplanted plots and three cleared plots (Fig. 1).In June 2015, for the cleared treatment, we obtained necessary permits and fully cleared remnant vegetation with chainsaws and a tractor (Fig. 2).For the interplanted treatment we cleared narrow (~1 m wide) 100 m long planting rows 4 9 4 m apart with machetes.This enabled us to plant seedlings in orderly rows with minimal disturbance to remnant vegetation.Following site preparation, each 1-ha plot was divided into four 50 9 50 m split plots, and two replicates of each species mix (acquisitive/conservative) were randomly assigned to be planted in two split plots per plot (two replicate split plots per species mix 9 two split plots per plot 9 6 plots = 24 split plots; 10 m gap between split plots; Fig. 1).", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.5", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "In July 2016, we characterized remnant vegetation in the three interplanted treatment plots by measuring the DBH (diameter at breast height) of all trees \u226510 cm and identifying all trees to species.For multi-stemmed trees, we measured all stems \u226510 cm in DBH for each individual.We quantified fine root stocks in the six treatment plots in February 2018 as a proxy for belowground competition.We collected six root samples with a root corer (8 cm diameter, 15 cm depth) along two parallel transects (50 m apart) in each plot (6 plots 9 2 transects per plot 9 6 samples per transect = 72 root samples).We washed soil from roots, dried roots to constant mass at 60\u00b0C, and determined root dry mass (g) for each sample.\nTo characterize differences in microclimatic conditions between the cleared and interplanted treatments, we measured air/soil temperature and soil moisture during seasonal drought conditions and after the onset of the 2015 wet season.At the center of each plot, we used iButtons Notes: The six traits used to assign acquisitive or conservative resource-use strategies for each focal species using cluster analysis for Goal 1 (Appendix S1: Section S2) are at the top of the table.Ecophysiological traits used as predictors of seedling performance in Goal 2 are at the bottom of the table.LDMC, leaf dry matter content; SLA, specific leaf area; g s , stomatal conductance; A max , photosynthetic capacity; iWUE, instantaneous water-use efficiency; see Werden et al. (2018b) for more details on D d 13 C measurements (upregulation of integrated water-use efficiency).(DS1921G-F5# Thermochron; Maxim Integrated, San Jose, California, USA) to simultaneously measure air temperature (0.5 m off the ground), and soil temperature (5 cm depth), for 10 d in August and 10 d in September 2015 to account for day-to-day variation in midday temperatures.Measuring over this period captured the transition from dry to wet conditions for this year.Before deployed in the field, all iButtons were checked to ensure measurements were internally consistent.We measured volumetric soil moisture (%) in the top 0-5 cm of the mineral layer with a soil moisture sensor (DeltaSM150, Delta-T Devices, Burwell, UK) three times: before we planted the experiment (during significant drought when almost no rain had fallen for two months; August 2015), after the onset of wet season rains (September 2015), and at the start of the wet season during the second year of the experiment (July 2016).We measured soil moisture every 5 m (n = 20 per transect) along three 100-m transects in each plot (center of plot and two transects equidistant from plot center in opposite directions).No precipitation fell during soil moisture measurements; thus, plot-to-plot measurements were comparable.Finally, in November 2017, when the canopy was fully leafed out, we quantified full-sun photosynthetically active radiation (PAR) at 1.5 m above the ground with an AccuPAR LP-80 Ceptometer (Decagon Devices, Washington, DC, USA), every 5 m along a transect down each plot center on a cloudless day (n = 20 per plot).", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.6", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "We collected seeds and produced seedlings using local best-practice methods (Appendix S1: Section S3).In mid-September 2015, after the start of consistent wet season rains, we pre-dug holes in 4 9 4 m spacing in all six 1-ha plots.Two replicates each of the species mixes (acquisitive/conservative) were planted into the four 0.25-ha split plots at 4 9 4 m planting density (576 seedlings/ha; 3,456 seedlings total) over 10 d in September 2015.We planted seedlings (height: mean, 22.6 cm, minimum, 2.4 cm, maximum, 87 cm; r = 14.8) in each split plot using a multiple Latin-square design following Potvin and Dutilleul (2009), where species were planted in the same sequence with each species occurring twice in each row, to have the ability to examine neighborhood effects in the future.All seedlings were planted with soil amendments that aided first-year establishment on Vertisols in this region (Appendix S1: Section S3; Werden et al. 2018b).To minimize competition with non-native pasture grasses, grasses were cleared from the base of seedlings with machetes three times in both the interplanted and cleared treatments (December 2015, June 2016, May 2017).", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.7", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "We measured seedling survival and growth for two years (two full wet and dry seasons).During each survey, we determined survival and measured the height (at the apical meristem) of all planted seedlings.Seedlings were assumed to be dead if they lacked leaves and had a completely dry stem.Surveys were conducted after planting (September 2015) to obtain initial height and survival measurements (post-transplant shock), at the end of the first wet (growing) season (December 2015) and at the end of both dry seasons (July 2016 and May 2017).A previous study at this site found minimal seedling mortality during the second wet season, and minimal growth during the second dry season (Werden et al. 2018b), therefore no survey was conducted after the second wet season.", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.8", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "We tested for seasonal differences in soil moisture, soil temperature, and air temperature between management treatments using Tukey's HSD post hoc tests performed on one-way ANOVAs, with microclimatic variables as the responses and management treatment as the predictor.We used the same method to test for differences in PAR between the management treatments in November 2017.To characterize remnant vegetation in the interplanted treatment we calculated the total number of stems and basal area (m 2 /ha) for each treatment plot.We estimated \"effective\" stand age of the three interplanted plots by comparing the total basal area of trees in these plots to 84 chronosequence plots of naturally regenerating TDF in the same region (Powers et al. 2009, Becknell andPowers 2014).We compared the microclimatic data we collected to air temperature (Derroire et al. 2018) and PAR data (J.S. Powers, unpublished data) collected in 12 of the chronosequence plots.This allowed us to determine the extent to which the interplanted treatment resembled early-or mid-successional stages from the perspective of basal area and microclimatic conditions.\nTo quantify the performance of TDF seedlings with differing resource-use strategies at early and mid-successional stages (management treatments; Goal 1) we used individual seedling growth and survival measurements over two years.First, we calculated absolute height growth rates (AGR; cm/d) for each year for individual seedlings with the standard AGR equation (height final \u00c0 height initial ) \u00c0 (day final \u00c0 day initial ).Following these calculations, we fit mixed effects models to quantify the effects of management treatments and species' resource-use strategies on seedling survival and AGR.For survival, we fit a mixed effects multiple logistic regression model with seedling survival at the end of the experiment as the response, and management treatment (treatment), resource-use strategy (strategy), and the interaction between treatment and strategy (treatment 9 strategy) as fixed effect predictors (Appendix S1: Table S1; lme4 package in R).The natural logarithm (ln) of initial seedling height was included as a covariate.Variation in micro-topography and other environmental factors was controlled for by including the 0.25-ha split plots nested under the 1-ha whole plots as a random effect.Species was included as a random effect to account for species-level variation.\nFor AGR, we fit a mixed-effects multiple linear regression model with seedling AGR as the response (square root-transformed to meet normality assumptions).The same fixed effect predictors (strategy and treatment) were included with the addition of year as a predictor, to consider the effects of the drought (2015; 807 mm) and average precipitation (2016; 1,754 mm) years on seedling AGR.All two-way interactions and the three-way interaction of the predictors were also included.The same covariate and random effects as the survival model were used in the AGR model (Appendix S1: Table S2).We removed seedlings with negative growth rates, resulting from the top of the stem dying, from the model (n = 104).This allowed us to obtain better estimates of mean AGR, but did not affect the overall significance of model terms.The marginal (fixed effects; R 2 m ) and conditional (fixed and random effects; R 2 c ) R 2 values for all models were obtained (Nakagawa et al. 2017; MuMIn package in R).Interclass correlation coefficients were computed for the survival and AGR models to determine the proportion of variance explained by the random effects (Hox 2002; sjstats package in R).For the survival and AGR models, we used likelihood ratio (LR) chi-squared (v 2 ) tests to determine the significance of fixed effect model terms.For significant model terms we used pairwise comparisons (Tukey corrected) of estimated least-square means to test for differences in seedling AGR between management treatments, and for differences in AGR between the resource-use strategy groups within management treatments (emmeans package in R).\nLast, to determine if ecophysiological traits explained species performance across simulated successional stages (Goal 2), we fit the same mixed effects models described above, this time with four individual ecophysiological traits instead of resource-use strategy as predictors of survival and AGR, again using individual seedlings as an observation (Table 3).We fit the same models using the six individual traits used to define resource-use strategies as predictors to have the ability to directly compare the predictive power of individual ecophysiological traits to individual leaf and stem traits (Table 3).Before fitting models, we performed a z-transformation to standardize all traits values, so the effect sizes from each model (standardized regression coefficients; b std ) could be directly compared, by subtracting the trait mean from each species-level trait value then dividing by the standard deviation.\nWe then fit 10 models total for each response variable (four ecophysiological traits and six leaf and stem traits; Table 3), one for each trait in place of resource-use strategy as a predictor (e.g., abbreviated survival model: survival ~treatment 3 trait + ln(initial height) + random effects; abbreviated AGR model: \u221aAGR ~treatment 3 trait 9 year + ln(initial height) + random effects).For each response variable, we also fit a null model including all predictors except the individual traits (e.g., abbreviated survival null model: survival ~treatment + ln(initial height) + random effects; abbreviated AGR null model:\n\u221aAGR ~treatment 3 year + ln(initial height) + random effects) to determine if including the traits improved fits.We computed variance inflation factors for predictors and verified that collinearity did not affect regression estimates (car package in R).The null model, the individual trait models, and the resource-use strategy model created for Goal 1, were ranked using the Akaike Information Criterion (AIC) for each response variable.Model averaging is not appropriate for this analysis because it is ineffective when evaluating models with interaction terms (Cade 2015).For each model, marginal and conditional R 2 values were determined and the b std for each predictor extracted.\nWe tested for differences in the strength of relationships (slope) between traits and seedling performance (survival or AGR) in each management treatment for models with significant higher order interaction terms (i.e., treatment 9 trait for survival; treatment 9 trait 9 year for AGR) ranked higher than the resource-use strategy model (indicated by lower AIC scores).If the slope of performance 9 trait relationships differed between management treatments, we interpreted this as evidence that this trait had a differential effect on overall seedling performance between the cleared and interplanted treatments.A steeper performance 9 trait slope signified a trait had a stronger effect on seedling performance in the management treatment, or vice versa.No difference between slopes, i.e., approximately parallel lines, indicates the relationship between a trait and seedling performance did not differ by management treatment.We visualized those relationships with interaction plots produced by plotting the full models including all interaction terms (emmeans package in R).Intercepts of performance by trait relationships in these plots represent differences in mean AGR or survival probability between management treatments.All analyses were conducted in R version 3.5.1 (R Development Core Team 2018).", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.9", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "The interplanted treatment was dominated by three tree species that accounted for 85% and 86% of stems and basal area, respectively (see Appendix S1: Section S4 for more on remnant vegetation).The basal area of trees in the interplanted treatment plots ranged from 2.43 to 4.91 m 2 /ha (AE0.73SE).Stands with basal area in this range generally have ~5-10 yr of regrowth (Powers et al. 2009, Becknell andPowers 2014).Thus, while the interplanted stands have ~30 yr of regeneration, they resemble early-successional forest on more amenable soils in Notes: These models were fit to demonstrate how these traits affect (a) tropical dry forest seedling survival and (b) absolute growth rate (AGR).Results from a null model including all predictors except an individual trait is labeled \"no trait.\"(a) Results of logistic regression results with seedling survival as the response and treatment, trait, and interaction of the two as predictors and (b) linear regression results with AGR as the response and treatment, trait, year, and two-way/three-way interactions as predictors.All models include the natural logarithm of initial seedling height as a covariate, and species and split plots nested under whole plots as random effects.Models were ranked with Akaike Information Criterion (AIC).Standardized effect size (b std ) for each predictor and the variance explained by fixed (R 2 m ) and fixed and random effects (R 2 c ) are reported.Likelihood ratio and chi-squared tests were used determine the significance of fixed effect model terms.Dashes indicate terms not present in the model, asterisks indicate significance of predictors (*P < 0.05, **P < 0.01, ***P < 0.001), and ns indicates not significant.the region.Fine root stocks were 60.6% lower in the cleared than the interplanted treatment on average, indicating that management existing vegetation reduced fine root stocks after two years (n = 72, t = \u00c05.63,P < 0.001; Fig. 3a).\nMicroclimatic conditions differed significantly between the management treatments; compared to the cleared treatment, the interplanted treatment received only 17.6% of total incoming PAR (n = 120, t = \u00c08.82,P < 0.001; Fig. 3b), and had 2.1\u00b0C lower midday air (n = 60, t = \u00c05.7,P < 0.001; Fig. 3c) and 3.7\u00b0C lower soil temperatures during the wet season (n = 60, t = \u00c08.0,P < 0.001; Fig. 3d).Differences in soil moisture between the management treatments were very modest (~2-5%), and the relative rank of differences switched from the dry to the wet season.During dry conditions (August 2015), soil moisture was 3.3% higher in the interplanted treatment (n = 360, t = 16.0,P < 0.001), but in the wet-season (September 2015), soil moisture was 5.0% higher in the cleared treatment (n = 360, t = \u00c05.2,P < 0.001; Fig. 3e).Differences in wet-season soil moisture between the treatments remained consistent over time as soil moisture in the wet season of the second year of the experiment (July 2016) was 2.0% higher in the cleared treatment (n = 360, \u00c0t = 6.27,P < 0.001; from nursery to field; see Werden et al. 2018b) interacted significantly with management treatment, indicating that these traits had a stronger effect on survival (steeper slope) in the cleared treatment (iWUE, Fig. 5a, 62% larger effect, z = 3.04, P = 0.002; d 13 C, Fig. 5b, 83% larger effect, z = 2.46, P = 0.01).For seedling AGR, only three models were ranked higher than the resource-use strategy model (two ecophysiological traits: D d 13 C and g s and one leaf trait: foliar nitrogen concentration; Table 3b).For the D d 13 C, g s , and foliar nitrogen models, there was a significant three-way interaction among management treatment, trait, and year (Fig. 6).Interestingly, in year 2, seedlings with the highest D d 13 C (Fig. 6a) and foliar nitrogen concentrations (Fig. 6c) had lower AGR in the cleared treatment.Furthermore, species with higher foliar nitrogen concentrations had higher AGR in the interplanted treatment (Fig. 6c).Last, similar to findings for seedling survival, g s was positively associated with seedling AGR (b std = 0.05, P = 0.02), and the effect was stronger in the cleared treatment (Fig. 6a; year 1, 50% larger effect, t = 2.34, P = 0.02; year 2, 105% larger effect, t = 4.22, P < 0.001).", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.11", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "Our goals were to determine experimentally how TDF tree species with contrasting resource-use strategies defined with leaf and stem traits responded to changes in abiotic environmental conditions along a simulated successional gradient.Our results suggest large shifts in microclimate from early-successional (cleared) to midsuccessional (interplanted) stages likely influenced seedling growth rates, but not survival, and that ecophysiological functional traits differentially influenced responses between treatments.Binning species into resource-use strategies did not explain patterns in seedling survival; but, consistent with recent hypotheses about TDF successional theory (Lohbeck et al. 2013), acquisitive species had higher growth rates than conservative in the interplanted treatment.However, ecophysiological traits were always better predictors of survival than binary resourceuse strategies, and two out of five ecophysiological traits were better predictors of seedling growth rates.As predicted, ecophysiological traits relating to water-use had stronger effects on survival and growth in the cleared treatment, indicating that the influence of water-use traits on plant performance decreases from early-to mid-successional stages in TDF.Our data allow us to quantify the extent to which management treatments affected seedling performance, how functional traits help explain responses, and how we can leverage this knowledge to design more effective TDF restorations.", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.12", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "After ~30 yr of regeneration, the basal area of remnant vegetation in the interplanted treatment was comparable to 5-10 yr of forest regrowth on other soil types in this region, underscoring that Vertisols soils present significant challenges for forest regeneration (Powers et al. 2009).However, from the perspective of microclimate, the abiotic conditions in the interplanted treatment resembled those in nearby mid-successional forest (Appendix S1: Section S4; Derroire et al. 2018) in terms of both light availability (Fig. 3b), and air and soil temperatures (Fig. 3c,d).Thus, from a structural perspective, while the interplanted treatment had low forest cover, what trees and shrubs were there clearly modified the abiotic conditions experienced by seedlings.\nInterestingly, the relative ranks of soil moisture in the management treatments switched seasonally and the interplanted treatment had lower wet season soil moisture (Fig. 3e).While this contrasts strongly with patterns in other successional TDF (Lebrija-Trejos et al 2010), this result is largely consistent with additional observations from this site in Costa Rica and from Puerto Rico (J.Tijerin, personal communication).There are several potential mechanisms that may explain this pattern: first, higher vegetation density (e.g., basal area, fine root stocks) in the interplanted sites may result in higher transpiration rates (Ellison et al. 2017) and/or increased canopy interception (Jim enez-Rodr \u0131guez and Calvo-Alvarado 2014).Regardless of the mechanisms, contrary to our expectation that light would be higher in the cleared sites but wet season soil moisture would be higher in the interplanted sites, these data suggest that both soil moisture and light availability were higher in the cleared sites during the wet season (Figs.3b,e).Because microclimate conditions play a central role in our understanding of how shifts in abiotic conditions during succession influence TDF community assembly, this is a high priority area for additional studies.\nWater limitation during the dry season typically has the largest impact on seedling survival on Vertisols (Werden et al. 2018b) and the cleared treatment exposed seedlings to the lowest mean soil moisture during the dry season (Fig. 3e).However, we did not observe differences in overall survival rates between the management treatments (Fig. 4a), implying that water availability at earlyand mid-successional TDF stages may not differentially influence seedling survival in our system.However, we did observe that overall seedling growth rates were dramatically higher in the cleared treatment (Fig. 4b).This suggests that shifts in species dominance during TDF succession on Vertisols could instead be mediated by decreases in light availability (Fig. 3b) and/or increases in competition for belowground resources by fine roots (Fig. 3a), which can strongly limit seedling growth in this region (Gerhardt 1996), though we cannot tease apart the influence of these factors.Furthermore, our results may be influenced by the selection of species with high performance in drought conditions in a previous study (Werden et al. 2018b), and we may have selected 12 species, from a pool of 32, best adapted to tolerate drought in early-successional conditions.However, one-third of our focal species had no surviving individuals at the end of the experiment, suggesting that the species we planted were not all tolerant of severe drought conditions.Collectively, our results indicate that microclimate and competition may interact in a complex manner to influence TDF seedling growth.Nonetheless, while all species had higher growth in a simulated early-successional environment (cleared treatment), where there was high light and low belowground competition, these factors did not lead to differential survival between our management treatments.", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.13", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "Although plants fall along a continuum of resourceuse strategies (Wright et al. 2004), we were able to bin species into acquisitive and conservative groups (Table 1; Appendix S1: Fig. S1) using functional traits previously demonstrated to shift predictably during TDF succession (Lohbeck et al. 2013, Buzzard et al. 2016).Furthermore, our cluster analysis was robust to which species were included in it, as groupings were consistent when adding 15 additional commonly occurring tree species from the region (Appendix S1: Fig. S2).While these groupings followed general patterns expected of species with conservative or acquisitive strategies (Lohbeck et al. 2013, Buzzard et al. 2016), the acquisitive group appeared to have higher wood density than the conservative group (Appendix S1: Fig. S3), though the groups did not differ statistically (P = 0.10, Appendix S1: Section S2).This was due to the high wood densities of nitrogen-fixing legumes in the acquisitive group (Appendix S1: Section S2, Fig. S3), a finding noted by other studies (Powers and Tiffin 2010).This highlights a potential pitfall of relying on the handful of quantitative leaf and stem traits typically used to assign resource-use strategies across wide tropical precipitation gradients, as the acquisitive group was dominated by nitrogen-fixing legumes (Table 1) usually favored in early-to mid-successional tropical forests (Batterman et al. 2013, Powers andMar \u0131n-Spiotta 2017), particularly in Neotropical TDFs (Gei et al. 2018).However, while higher seedling stem densities can confer increased drought tolerance (Poorter and Markesteijn 2008), we did not find evidence that this was the case for the acquisitive species group.\nAcquisitive-conservative resource-use strategy groupings had some capacity to explain seedling responses, and the acquisitive group had higher mean AGR than the conservative in the interplanted treatment during a year with average rainfall (year 2; Fig. 4b).This result aligns with the recent hypothesis that acquisitive species should dominate in later-successional TDF (Lohbeck et al. 2013); however, this was the only instance where resource-use strategy groupings described patterns of seedling performance.The acquisitive-conservative groupings did not explain seedling survival (Fig. 4a), and a majority of variation in seedling responses in the management treatments was explained by interspecific differences in survival and growth (survival: 70% of variation; growth: 98% of variation).By contrast, ecophysiological traits explained both seedling survival and growth responses and were almost always stronger predictors of seedling performance than species' resourceuse strategies.", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.14", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "Ecophysiological traits always improved predictions of seedling survival when compared to resource-use strategies defined with leaf and stem traits (Table 3a), and two out of five traits improved predictions of seedling growth (D d 13 C and g s ; Table 3b).Furthermore, while iWUE was not predictive of seedling growth in a disturbed tropical wet forest (Guimar\u00e3es et al. 2018), we found iWUE to be positively associated with both seedling survival and growth, consistent with the understanding that water availability is a more important driver of species performance in systems with strong annual water limitation such as TDF (Pineda-Garcia et al. 2013).Individual leaf and stem traits used to define resource-use strategies were generally poor predictors of seedling performance, though one leaf trait was associated with survival (petiole length) and one was associated with growth (foliar nitrogen).Overall, these results confirm our hypothesis that ecophysiological traits are strong predictors of TDF seedling performance (Goal 2).We argue this is strong evidence that leaf and stem traits generally used to define resource-use strategies for tropical trees have limited capacity to categorize tree performance in early-to mid-successional TDF, consistent with recent work highlighting that commonly measured functional traits are poor predictors of tree demographic rates (Yang et al. 2018).We also emphasize the importance of considering ecophysiological traits when explaining patterns of early-successional TDF community assembly, although they are considerably more difficult to measure than \"soft\" traits such as SLA.Because \"soft\" leaf and stem traits are generally not correlated with ecophysiological traits of seedlings in this TDF (Werden et al. 2018b), it is especially important to focus future research on determining how community-level ecophysiological traits shift over TDF successional trajectories.Although ecophysiological traits were almost always better predictors of TDF seedling performance in our restoration experiment, some leaf traits were helpful in understanding how light availability influenced seedling responses, namely petiole length and foliar nitrogen.\nSpecies with long petioles maximize light capture (Niinemets et al. 2004), placing them on the acquisitive end of the resource-use spectrum.We found that petiole length was negatively associated with survival rates (Table 3a), signifying that conservative values of this trait conferred a survival advantage across both clearing treatments.Foliar nitrogen, which is positively associated with photosynthetic rates (Poorter and Bongers 2006), was the only leaf trait predictive of seedling growth.This trait had a complex relationship with seedling performance: species with conservative (low) values of leaf nitrogen had higher growth rates in the cleared treatment and species with acquisitive (high values) of leaf nitrogen had higher performance in the interplanted treatment (Fig. 6c).This finding is consistent with the expectation that TDF communities transition from having conservative to acquisitive trait values.However, our observation that two leaf traits associated with species' light requirements (petiole length and foliar nitrogen) helped to explain seedling performance deserves further investigation especially because TDF community dynamics are generally assumed to be associated more with water availability.It will be important to tease apart how these abiotic environmental factors influence the magnitude of relationships between species' traits and performance.", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.15", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "Many studies have demonstrated that the functional composition of tropical forests can be driven by shifts in environmental conditions during succession (Lebrija-Trejos et al. 2010, Lohbeck et al. 2013, Boukili and Chazdon 2017), and that the relative influence of abiotic and biotic factors on plant performance can shift as succession progresses (Lohbeck et al. 2014, Craven et al. 2018).To our knowledge, our study is the first to build on these findings and demonstrate experimentally that ecophysiological functional traits are the best predictors of TDF tree survival and growth at early-and mid-successional stages (Table 3).Furthermore, we found evidence that the influence of water-use ecophysiological traits on seedling performance decreases during succession in this system, based on our finding that relationships between seedling performance and water-use traits (iWUE, D d 13 C, and g s ) were not as strong in the midsuccessional interplanted treatment (Figs. 5,6).\nWhile seedlings had consistently higher growth rates in the cleared treatment (Fig. 4b), these findings may indicate that decreases in air temperature and vapor pressure deficit during the wet season after vegetation is established (Lebrija-Trejos et al. 2011) may lead to decreases in the influence of water-use traits on seedling performance in mid-successional TDF.Ecophysiological functional traits were positively associated with seedling survival and growth in all but one instance (Table 3): in the second year, Dd 13 C was negatively associated with cleared treatment growth rates (Fig. 6a).This could indicate integrated WUE (measured with d 13 C) for our focal species shifts over time, or that it differs in drought (year 1) versus average precipitation (year 2) years.Because we used values for Dd 13 C that were calculated in the first year after planting seedlings in a previous experiment (Werden et al. 2018b), determining how this dynamic trait varies over the lifetime of a seedling is necessary to fully interpret this result.Additionally, to our knowledge no studies have investigated how resourceuse strategies shift with ontogeny for all TDF tree life stages.Overlapping traits have now been collected for seedlings, saplings (Derroire et al. 2018), and adults of many TDF tree species (Powers and Tiffin 2010) and future efforts should be dedicated to determining how these trait syndromes shift as TDF communities assemble.In addition to furthering the understanding of how species traits interact with shifting environmental conditions during TDF succession, our results can be used to improve TDF restoration strategies.", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.16", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "Recent functional trait-driven approaches have demonstrated how resource-use strategies dictate species performance in active restorations of both tropical wet (Mart \u0131nez-Garza et al. 2013, Ostertag et al. 2015, Giannini et al. 2017, Charles et al. 2018) and dry forest (Werden et al. 2018b).These studies and others have shown the utility of using functional traits as tools for restoration species selection (Charles 2018) and when designing restorations that have targeted functional outcomes (Laughlin 2014).For these approaches to be effective, it is important to determine which traits are associated with plant performance, which can be highly ecosystemspecific.For TDF in particular, our results demonstrate that considering ecophysiological traits when designing and implementing TDF restorations greatly improves the prediction of project outcomes.Specifically, we found that TDF restoration effectiveness can be improved by explicitly examining traits related to species-level responses to water limitation, but that binary resource-use strategy categories and leaf and stem traits did not help to improve TDF restoration approaches (Table 3).Moreover, our results highlight the importance of species selection in determining initial TDF restoration establishment, as species identity was highly predictive of seedling survival but there were no differences in overall survival between the two management treatments (Fig. 4a).Additionally, if carbon accumulation is a major project goal, it is important to consider how decreases in the influence of water-use traits on seedling survival and growth (Figs. 5,6) and increases in belowground resource competition (Fig. 3a) contributed to lower carbon accumulation in the interplanted treatment (Fig. 4b).To increase the probability of restoration success, these findings demonstrate the importance of considering the existing vegetation, or successional status, at a site when designing species mixes.\nInitial site conditions such as the level of disturbance or state of remnant vegetation at a site can have lasting effects on the rates of vegetation recovery in tropical forest restoration (Holl et al. 2018).Our results demonstrate that rates of TDF vegetation recovery can also be predictably influenced by planting species mixes with specific functional characteristics.For example, the differences in the strength of TDF tree performance 9 water-use efficiency relationships between our management treatments (Figs. 5,6) highlight the importance of planting tree TDF species with high water-use efficiency in sites with high levels of disturbance (cleared).However, water-use efficiency is not as strong of an indicator of TDF species performance at intermediate disturbance levels (interplanted) and it may be more pertinent to consider the importance of biotic factors such as competition when designing species mixes for sites with existing vegetation (Fig. 3a).This finding is constant with the observation from tropical wet forest that species dominance is increasingly influenced by biotic factors as succession proceeds (Lohbeck et al. 2014).It also emphasizes the importance of creating species mixes that are functionally representative of the tree communities that dominate at distinct successional stages to ensure restoration outplantings can establish and persist.\nLast, as we showed previously at this site (Werden et al. 2018b), we highlight that it is particularly difficult to restore TDF on degraded Vertisols given that overall survival was low after two years, ranging from 15.1% to 26.4% in the four resource-use strategy 9 treatment combinations (Appendix S1: Fig. S5).The strongest drought on record in 2015 likely had a large influence on overall survival; however, most of the focal species (eight total) had surviving individuals after two years.Thus, our results demonstrate it is possible to restore TDF on Vertisols in this region, even during extreme drought.Moreover, our results underscore that clearing existing vegetation before planting seedlings may help catalyze regeneration of TDF on Vertisols, because removing vegetation in a state of arrested succession increased seedling growth rates and may help to establish additional species not abundant on these soils.That said, we caution against using clearing as a management strategy without careful consideration of its benefits and drawbacks.", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.17", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "Our results demonstrate that considering how resource-use functional traits describe patterns of TDF tree species' performance not only furthers the understanding of TDF successional dynamics, but also can improve TDF restoration outcomes.In order to do so, it is necessary to focus on traits predictive of survival and growth in this system, namely water-use efficiency and photosynthetic rates.Moreover, our findings suggest that ecophysiological functional traits, specifically those pertaining to water use, had a stronger impact on species performance at early TDF successional stages.Additional focused effort to determine how both environmental conditions and resource competition shift during TDF successional trajectories is necessary to fully understand TDF community assembly.", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.18", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "We thank the many volunteers who helped to plant the experiment, Daniel Perez A. for collecting seeds, and Roger Blanco from Area de Conservaci on Guanacaste for facilitating the study.This work was supported by NSF DDIG (1600710) and GRFP (11-582) and a Garden Club of America Award in Tropical Botany to L. K. Werden, and NSF CAREER (DEB-1053237) to J. S. Powers.Comments from Rebecca Montgomery, Rakan Zahawi, Susan Galatowitsch, Dean Current, and two anonymous reviewers greatly improved this manuscript.", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.27", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "We thank the many volunteers who helped to plant the experiment, Daniel Perez A. for collecting seeds, and Roger Blanco from Area de Conservaci on Guanacaste for facilitating the study.This work was supported by NSF DDIG (1600710) and GRFP (11-582) and a Garden Club of America Award in Tropical Botany to L. K. Werden, and NSF CAREER (DEB-1053237) to J. S. Powers.Comments from Rebecca Montgomery, Rakan Zahawi, Susan Galatowitsch, Dean Current, and two anonymous reviewers greatly improved this manuscript.", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.28", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "Appendix S1: Fig. S4).Furthermore, an ongoing study of TDF microclimate at a nearby site has showed that open field Vertisol sites have lower soil moisture than forested sites at the peak of the dry-season (J.Tijerin, personal communication), consistent with our observation that dry-season soil moisture was lower in the open field cleared treatment than in the interplanted treatment.Overall, these results indicate that the interplanted treatment represents a mid-successional stage (\u226530 yr old) in terms of light availability (PAR) and air temperature when compared to a TDF chronosequence in the same region (see Appendix S1: Section S4).\nOverall seedling survival after two years ranged from 15.1% for acquisitive species in the cleared treatment to 26.4% for conservative species in the cleared treatment (Appendix S1: Fig. S5; see Appendix S1: Table S3 for survival by species).Likely attributed to the severe drought in year 1, three acquisitive (Acosmium panamense, Albizia saman, Thouinidium decandrum) and one conservative species (Pachira quinata) had no surviving individuals in either management treatment by the end of the experiment (Appendix S1: Table S3).There were no significant differences in survival among management treatments or resource-use strategies (Fig. 4a), and management treatments and resource-use strategies explained little variation in seedling survival after two years (R 2 m = 0.04).By contrast, species and split plot level effects explained a majority of variation (R 2 c = 0.42), with species explaining 70% of variation in the random effects.Furthermore, overall species composition of surviving individuals did not differ between the management treatments at the end of the experiment, i.e., results were not driven by intra-specific differences in survival between treatments (Appendix S1: Table S3).The only significant predictor of survival after two years was seedling height at planting (v 2 df\u00bc1 = 22.60, P < 0.001; Appendix S1: Table S1).\nFor seedling growth, management treatment was a significant predictor of AGR (v 2 df=11 = 14.44,P < 0.001; Appendix S1: Table S2).Independent of resource-use strategy, seedlings had dramatically higher AGR in the cleared than the interplanted treatment in both years (Fig. 4b; year 1, cleared 69.1% higher, t = 7.29, P < 0.001; year 2, cleared 143.4% higher, t = 11.34,P < 0.001).A three-way interaction between management treatment, resource-use strategy, and year (v 2 df=1 = 10.28,P < 0.001; Appendix S1: Table S2) was driven by conservative species having 66.8% lower AGR than acquisitive in the interplanted treatment in year 2 (t = 3.90, P < 0.001) under average precipitation.When compared to the survival model, the fixed-effects predictors explained more variation in AGR (R 2 m = 0.17).However, the random effects still explained a majority of variation in seedling AGR (R 2 c = 0.46), with species explaining 98% of variance in the random effects.\nThe series of models we fit demonstrated that ecophysiological functional traits were better predictors of seedling survival and growth than the conservative and acquisitive resource-use strategies defined by \"soft\" leaf and stem traits (Table 3).Here we present results for the models ranked higher with AIC than the resource-use strategy model.For seedling survival, all candidate models were ranked higher than the resource-use strategy model (Table 3a).The best model for survival included A max as a predictor and improved the variance explained considerably over the resource-use strategy model (R 2 m = 0.22 for A max model, R 2 m = 0.04 for strategy model; Table 3).A null model including only a covariate for initial height and the random effect terms as predictors of survival had a modest improvement (AIC = 2,738) over the resource-use strategy model (AIC = 2,739); further evidence that resource-use strategy was a poor predictor of survival.\nHigher rates of A max (photosynthetic capacity; b std = 1.22,P < 0.001), iWUE (instantaneous water-use efficiency; b std = 1.13,P < 0.001), and g s (stomatal conductance; b std = 0.89, P = 0.01) were all positively associated with higher seedling survival.Petiole length was the only leaf trait associated with survival (b std = \u00c00.94,P = 0.007), indicating that species with shorter petioles had higher survival, though all ecophysiological traits were ranked higher (Table 3a).Additionally, specieslevel iWUE and D d 13 C (the capacity of a species to upregulate integrated water-use efficiently after planted", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.29", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "Overall seedling survival after two years ranged from 15.1% for acquisitive species in the cleared treatment to 26.4% for conservative species in the cleared treatment (Appendix S1: Fig. S5; see Appendix S1: Table S3 for survival by species).Likely attributed to the severe drought in year 1, three acquisitive (Acosmium panamense, Albizia saman, Thouinidium decandrum) and one conservative species (Pachira quinata) had no surviving individuals in either management treatment by the end of the experiment (Appendix S1: Table S3).There were no significant differences in survival among management treatments or resource-use strategies (Fig. 4a), and management treatments and resource-use strategies explained little variation in seedling survival after two years (R 2 m = 0.04).By contrast, species and split plot level effects explained a majority of variation (R 2 c = 0.42), with species explaining 70% of variation in the random effects.Furthermore, overall species composition of surviving individuals did not differ between the management treatments at the end of the experiment, i.e., results were not driven by intra-specific differences in survival between treatments (Appendix S1: Table S3).The only significant predictor of survival after two years was seedling height at planting (v 2 df\u00bc1 = 22.60, P < 0.001; Appendix S1: Table S1).\nFor seedling growth, management treatment was a significant predictor of AGR (v 2 df=11 = 14.44,P < 0.001; Appendix S1: Table S2).Independent of resource-use strategy, seedlings had dramatically higher AGR in the cleared than the interplanted treatment in both years (Fig. 4b; year 1, cleared 69.1% higher, t = 7.29, P < 0.001; year 2, cleared 143.4% higher, t = 11.34,P < 0.001).A three-way interaction between management treatment, resource-use strategy, and year (v 2 df=1 = 10.28,P < 0.001; Appendix S1: Table S2) was driven by conservative species having 66.8% lower AGR than acquisitive in the interplanted treatment in year 2 (t = 3.90, P < 0.001) under average precipitation.When compared to the survival model, the fixed-effects predictors explained more variation in AGR (R 2 m = 0.17).However, the random effects still explained a majority of variation in seedling AGR (R 2 c = 0.46), with species explaining 98% of variance in the random effects.", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.31", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Werden et al. - 2020 - Using large\u2010scale tropical dry forest restoration .pdf.tei.xml", "page_content": "The series of models we fit demonstrated that ecophysiological functional traits were better predictors of seedling survival and growth than the conservative and acquisitive resource-use strategies defined by \"soft\" leaf and stem traits (Table 3).Here we present results for the models ranked higher with AIC than the resource-use strategy model.For seedling survival, all candidate models were ranked higher than the resource-use strategy model (Table 3a).The best model for survival included A max as a predictor and improved the variance explained considerably over the resource-use strategy model (R 2 m = 0.22 for A max model, R 2 m = 0.04 for strategy model; Table 3).A null model including only a covariate for initial height and the random effect terms as predictors of survival had a modest improvement (AIC = 2,738) over the resource-use strategy model (AIC = 2,739); further evidence that resource-use strategy was a poor predictor of survival.\nHigher rates of A max (photosynthetic capacity; b std = 1.22,P < 0.001), iWUE (instantaneous water-use efficiency; b std = 1.13,P < 0.001), and g s (stomatal conductance; b std = 0.89, P = 0.01) were all positively associated with higher seedling survival.Petiole length was the only leaf trait associated with survival (b std = \u00c00.94,P = 0.007), indicating that species with shorter petioles had higher survival, though all ecophysiological traits were ranked higher (Table 3a).Additionally, specieslevel iWUE and D d 13 C (the capacity of a species to upregulate integrated water-use efficiently after planted", "title": "Using large\u2010scale tropical dry forest restoration to test successional theory", "id": "9.32", "keywords": [ "abiotic conditions", "active restoration", "community assembly", "Costa Rica", "degraded Vertisol", "ecophysiology", "microclimate", "plant functional traits", "resource-use strategies", "succession" ] }, { "file_name": "Rayome et al. - 2018 - Enhancing Aboveground Carbon Storage and Invasion .pdf.tei.xml", "page_content": "through Restoration: Early Results from a Functional Trait-Based Experiment 1 Donald D. Rayome, 2,5 Rebecca Ostertag, 3 and Susan Cordell 4\nAbstract: One of many ecosystem services essential to land management is carbon regulation, but presence of invasive species can influence carbon (C) in undesirable ways.Here we discuss early results of C accumulation from the Liko N\u00e4 Pilina hybrid wet forest restoration experiment.The focus of our project is to deliberately increase C storage through a functional trait-based approach to restoration.By choosing plant species mixes with specific functional trait values, a novel ecosystem can be assembled that supports desired ecosystem services such as C regulation.We designed species mixtures based on species rate of C turnover (slow or moderate) and their position in trait space (complementary or redundant functional trait values).New species mixes were planted as four treatments (Slow Redundant, Slow Complementary, Moderate Redundant, and Moderate Complementary), with an additional unmanaged Reference treatment.Our objective was to compare C in aboveground woody biomass using allometric equations to determine which mixture had the greatest potential for site restoration, balancing carbon storage with the eventual goal of creating forests better able to resist establishment by invasion species.Initially, we predicted the Moderate Complementary treatment would have increased C storage.However, we found that the Moderate Redundant treatment had the greatest C storage, largely driven by a few fast-growing species during early development.Even though our short-term results did not support our experimental prediction, these data serve as an important benchmark for contrasting with later results when ecological succession might favor complementary species mixes for sustainable biomass productivity and decreased management efforts.\ntions and processes, consequences or potential benefits of novelty can be difficult to ascertain (Mascaro et al. 2008, Kueffer et al. 2010, Hulvey et al. 2013).The challenge of managing novel landscapes and understanding their effects requires new strategies that extend beyond recovering historical compositions and processes ( Hobbs et al. 2011) and instead focus on resulting ecosystem services ( Harborne and Mumby 2011, Hulvey et al. 2013).\nOne of many ecosystem services essential to land management is carbon regulation ( Vitousek, D'Antonio, et al. 1997;Vitousek, Mooney, et al. 1997;Foley et al. 2005).Land managers are likely to be interested in carbon storage from economic and legal standpoints and as a powerful complement to restoration ( Huston and Marland 2003, Lubowski et al. 2005, Torres et al. 2010, Evans et al. 2015, Asner et al. 2016).Humans often diminish the capacity of terrestrial ecosystems to store carbon through utilization, degradation, and biological invasion (Ciccarese et al. 2009).As a result, many ecosystems greatly affected by humans tend to release more carbon than they store (Foley et al. 2005, Fargione et al. 2008, Pongratz et al. 2009).However, there are several examples of management activities such as reforestation and ecosystem service restoration that can help to reduce carbon loss (Albrecht and Kandji 2003, Chazdon 2008, Ciccarese et al. 2009).\nRates of carbon accumulation and cycling are strongly influenced by climate such that tropical forests often contain the highest carbon pools (FAO 2010, Raunikar et al. 2010, Payn et al. 2015).Land managers working in tropical ecosystems can use these high rates to their advantage.Hawaiian forests store substantial amounts of carbon, comparing favorably with their global tropical counterparts (Asner et al. 2011, Ostertag et al. 2014) even with proportionally fewer native tree species, and thus present a unique opportunity to examine carbon cycling in a simplified context.Further, species invasions, disease, and other anthropogenic pressures ( Hughes and Denslow 2005, Asner et al. 2016, Crow et al. 2016) impact forests on all islands.Resulting novel species compositions, structural alter-ations, and functional changes inhibit a return to previously known forest types and support a need for carbon management.\nOne strategy to promote desired levels of carbon accumulation and cycling is to use a functional trait-based approach to assemble plant communities that support desired services such as carbon storage.Functional traits relate to the expression of various morphological, structural, physiological, or chemical traits of organisms.For example, selecting species with a broad range of functional trait expressions may preclude species invasions if the chosen functional traits are already represented in the community (Pokorny et al. 2005, Funk et al. 2008).Because plant functional traits relate to resource capture and processing ability, it is likely that a plant species' position in trait space influences its ability to cycle carbon in ways that affect long-term storage.Complementary assemblages of species in a community are hypothesized to increase ecosystem service variety, species performance, and invasion resistance, and redundant assemblages likely concentrate services ( Hooper 1998, Funk et al. 2008).Traits associated with slower C cycling place many natives in a trait space that is at a disadvantage when competing with invasives, especially where traits that promote faster C cycling overlap with traits that promote weediness or inhibit native species recovery (Cardinale et al. 2011).As succession progresses, functional traits that allow a given species to take advantage of fluctuating environmental conditions likely become less influential.Rather, traits that promote successful competition and resource co-opting become more pertinent (Lohbeck et al. 2014), with highly productive or otherwise influential species often driving ecosystem performance measures (Cardinale et al. 2011).Thus reducing interspecific competition through more complementary functional composition (Suding et al. 2008, Hooper andDukes 2010) becomes important when assessing the trade-offs necessary to meet ecosystem management objectives.However, complementary species mixes may not be advantageous at all stages of restoration because early stages may require species with fast rates of growth that quickly yield suitable microclimatic conditions within a site (Sonnier et al. 2012, Fry et al. 2013, Ostertag et al. 2015).\nFunctional trait-based restoration has rarely been tested in most forested ecosystems (Lavorel 2013, Ostertag et al. 2015).The Liko N\u00e4 Pilina hybrid lowland wet forest restoration experiment addresses functional trait effects in terms of complementary versus redundant experimental plant communities, acknowledging both the need for supporting native species integrity as well as the realities of restoring in areas subject to constant invasion pressure and human use.Carbon is an ecosystem variable readily assessed over a shorter time frame (Ostertag et al. 2009, Cordell et al. 2016).As an experiment, Liko N\u00e4 Pilina consists of a reference (invaded forest) and four restoration treatments in which natives were left in place, all nonnative species were cleared, and different mixtures of 10 redundant or complementary species were planted.Noninvasive nonnative species (exotic) were combined with natives to fill ecological roles and aid in guiding site biodiversity toward species assemblages that promote preferred ecosystem functions and services such as slower carbon cycling rates.In this project, we tested how species assemblages store carbon at an early experimental stage, leading to the following hypothesis: The species mixture with a combination of \"Moderate\" C cycling species and the \"Complementary\" functional trait species will have the highest C storage capacity as measured by aboveground woody biomass.Such early carbon storage results are an important benchmark for comparison with later successional stages when biotic and abiotic limitations may differ.", "title": "Enhancing Aboveground Carbon Storage and Invasion Resistance through Restoration: Early Results from a Functional Trait-Based Experiment", "id": "10.1", "keywords": [] }, { "file_name": "Rayome et al. - 2018 - Enhancing Aboveground Carbon Storage and Invasion .pdf.tei.xml", "page_content": "The process of developing treatments in the Liko N\u00e4 Pilina experiment is described in Ostertag et al. (2015) and is briefly summarized here.To choose species for the experi-ment, we investigated 15 functional traits of native and exotic candidate species that could be found living in the lowland wet forest community in East Hawai'i Island.Species were classified as exotic as opposed to invasive based primarily on the Hawai'i Weed Risk Assessment score [Daehler et al. 2004 (see www.botany.hawaii.edu/faculty/daehler/ wra / )].To focus on C, we ran a Principal Component Analysis (PCA) on a subset of traits related to C allocation ( leaf mass per area, foliar C : N, stem specific gravity, maximum height, stature in the field, and integrated water-use efficiency) to calculate species arrangement in trait space.We identified core species that had sets of traits that can lead to either slow or moderate rates of C turnover (Table 1).Remaining species in each treatment were selected by calculating the centroid of the four core species and then choosing species ( based on Euclidean dis tances) that were either similar (near) in trait expression to the core species (Redundant) or different (far) (Redundant).Four treatment combinations exist: Slow Complementary, Slow Redundant, Moderate Complemen tary, and Moderate Redundant.Intact in vaded reference plots served as a control for comparison.\nLowland [30 m above sea level (a.s.l.)] wet forest portions of the Keaukaha Military Reservation ( KMR, 19\u00b0 42\u2032 15\u2033 N, -155\u00b0 2\u2032 40\u2033 W ) in Hilo, Hawai'i, serve as the test location (Ostertag et al. 2015).The site is located on an 'a'\u00e4 lava flow occurring 750 -1,500 yr ago, with annual temperature average 22.7\u00b0C (Giambelluca et al. 2014) and average annual rainfall 3,347 mm (Giambelluca et al. 2011).Native canopy but limited native tree regeneration (Cordell et al. 2009) and heavy invasion by nonnative, invasive trees and shrubs characterizes forests at KMR [approximately 45% of basal area (Ostertag et al. 2009)].Reference treatment plots received no management, but the four experimental treatments were cleared before planting.Experimental nonnative tree species clearing began in late July 2012 and ended in mid-April 2013, and native tree species were left intact.Introduced trees that were at least 50% rooted in a plot or had a tree canopy that Note: Each experimental treatment had four core species that were chosen for either their slow or moderate rates of C turnover, and six additional species that had functional traits that were either redundant or complementary to the core species.Also noted are existing native species not removed from the experimental plots.\nfell more than halfway into the buffer zone (2.5 m) were removed.Herbicide (30% Garlon 4 Ultra, mixed with 70% crop oil) was sprayed immediately onto cut stumps to prevent resprouting.Planting density was based on data from other Hawaiian lowland wet forests that have maintained a greater abundance of native species (Zimmerman et al. 2008), as well as the mature plant size.Total planted individuals per plot were as follows: 125 for Slow Complementary, 130 for Moderate Complementary, and 120 for the two Redundant treatments.We identified four separate areas ( blocks) with appropriate conditions while using surveying equipment to lay out five plots in each block (Figure 1).Assignment of treatments to plots was random, with 20 plots measuring 20 m by 20 m and a 5 m perimeter buffer.We aimed for a 10 m distance between the buffers for each plot, but actual distances depended on terrain and avoidance of gullies and treefall gaps.\nTo evenly distribute the plants across each plot we set up a grid across each planting area with the number of quadrats depending on the number of large tree species designated for that treatment.These large tree species served as foci, with other species planted around them in a stratified random design (see Figure 1 for spatial configuration in each treatment).Plant spacing was based on adult plant size, such that large plants were placed 2 m away from their nearest neighbor, and medium and small plants were placed 1.5 m and 1 m away, respectively.Planting was done in stages from April 2013 to January 2014 because different species were ready for transfer at different times.All outplants were grown on Hawai'i Island from locally available propagules.When a preexisting native tree was located where an outplant was supposed to be planted we relocated the outplants, making sure that no plant was placed <1 m from any other plant.Plots were weeded before planting because several months had passed after clearing, and new nonnative seedlings had popped up after the disturbance.After planting, plots were weeded at 4-to 6-month intervals; native species plantings and recruits were left intact and invasive species were removed.", "title": "Enhancing Aboveground Carbon Storage and Invasion Resistance through Restoration: Early Results from a Functional Trait-Based Experiment", "id": "10.3", "keywords": [] }, { "file_name": "Rayome et al. - 2018 - Enhancing Aboveground Carbon Storage and Invasion .pdf.tei.xml", "page_content": "Experimental measurements commenced in January 2014 after all individuals had been planted and tagged for long-term monitoring.Carbon values in this article focus on summed aboveground biomass data collected in April and May 2016, which integrate the growth of outplanted individuals and recruits (seedlings of existing or outplant species) during 2 yr of experimental conditions.Data originate from individual on-site woody species but exclude biomass removed from study plots, belowground biomass, and herbaceous species biomass.Data collected for carbon analysis were plant height and diameter at breast height (DBH ) at 1.3 m for all stems \u22651.0 cm diameter.Species were grouped in one of three categories: existing, outplants, and recruits.Existing species were trees left in situ after commencing experimental treatments (also inclusive of initial trees present in reference plots).Outplants refer to those species meant to define experimental treatment mixtures (i.e., all native and nonnative plants installed in cleared experimental plots).Recruits are defined as all new growth originating during the experiment from either existing seed pools or reproducing outplants.\nWe also wanted to determine if abiotic variables ( light, soil nutrients, soil pH ) might contribute to outplant and recruit C storage, regardless of treatment.Before the experiment (2012) and again during outplanting (2014), we took hemispherical photos to estimate percentage canopy cover.The photos were taken with a Canon EOS 5D camera and Canon EF 15 mm fisheye lens before analysis using WinsCanopy software (Regent Instruments, Inc., Quebec City, Canada).In July 2012 soils were sampled no deeper than 10 cm using trowels (n = 4 per plot, with one sample in each subplot).Volumetric soil core data are not expressed on an area basis due to difficulties associated with the extremely rocky terrain.Roots and debris were handpicked out of soil samples to maintain soil aggregates, dried at 60\u00b0C, and ground.Soils were analyzed for carbon (C) and nitrogen ( N ) in a Costech 4010 Elemental Analyzer (Costech Analytical Technologies, Valencia, California), and for", "title": "Enhancing Aboveground Carbon Storage and Invasion Resistance through Restoration: Early Results from a Functional Trait-Based Experiment", "id": "10.4", "keywords": [] }, { "file_name": "Rayome et al. - 2018 - Enhancing Aboveground Carbon Storage and Invasion .pdf.tei.xml", "page_content": "Carbon in aboveground biomass was calculated for outplanted and recruit trees and compared among all treatments.Calculations were completed using allometric equations for individual species or those generally applicable for species in wet tropical forests as described in past studies and the literature (Asner et al. 2011) (Supplemental Table S1).These equations use DBH, wood density, and height as independent variables.For most species, aboveground biomass was determined by diameter-to-biomass equations, supplemented with additional diameter-to-height equations as needed.For all individuals with a measureable DBH, the General Wet Forest equation (Chave et al. 2014), Hawai'i-derived equations (Asner et al. 2011), or more speciesappropriate equations (Donato et al. 2012, Hung et al. 2012, Goodman et al. 2013) were utilized.Tree heights were determined as needed via Asner et al. (2011) or field data when available.Where required, wood density estimates originated from Asner et al. (2011), Chave et al. (2009), the Global Wood Density database (Little and Wadesworth 1964, Anon. 1974, Benthall 1984, Chundoff 1984, Oey 1990, Flynn and Holder 2001, Tree Talk 2005), or previous field measures (R. Ostertag and field assistants, unpubl.data).Calculations include all stems \u22651 cm DBH, but exclude secondary growth such as branches below breast height.Biomass to C-equivalent conversions followed wood production industry standards (Alabama Forestry Commission 2016; D. D. Rayome, R. Ostertag, and S. Cordell, unpubl. data).\nAuthors' Note: Supplemental materials available online at BioOne (http://www .bioone.org/toc /pasc /current) and Project MUSE (http://muse.jhu.edu/journal /166).", "title": "Enhancing Aboveground Carbon Storage and Invasion Resistance through Restoration: Early Results from a Functional Trait-Based Experiment", "id": "10.5", "keywords": [] }, { "file_name": "Rayome et al. - 2018 - Enhancing Aboveground Carbon Storage and Invasion .pdf.tei.xml", "page_content": "Statistics were examined using R 3.1.2(R Core Team 2014) and JMP 11.2.0 (SAS 2013).To examine the effects of experimental planting, we compared the C amount in the outplants plus recruits as the response variable because this biomass represents growth in response to the experimental conditions.We focused on the plot level rather than the individual level because C in the plots integrate survival and growth.C values were ln transformed to achieve normality and equal variances.To examine the effects of the hybrid community treatments, we ran a complete randomized block analysis of variance (ANOVA), based on procedures by Logan (2010), followed by Tukey's tests to compare among the treatments.To examine differences in C across species, we examined the outplants and recruits separately, because it was a fairer comparison across species.Only a few species had recruits, and these were generally small individuals with small amounts of C. We ran one-way ANOVA tests for the outplants and the recruits; we did not test for a treatment effect here because we previously had verified by two-sample t tests that all the species found in multiple treatments (see Table 1) were not significantly different in C amounts.\nWe suspected that environmental variables also influenced C gain and chose several additional analyses to assess this potential connection.First, we utilized one-way ANOVA for pretreatment soil and light variables to test possible treatment differences before experimental manipulation.Among treatments, there was no significant difference in canopy openness, soil pH, or nutrients (C, N, P, Mg, Ca, Na, K).To test whether these abiotic variables influence C storage in outplants (independent of assigned treatment), we ran Pearson correlation tests between outplant C and seven environmental variables: soil pH, C, N, P, Mg, K, and canopy openness in 2014 at the start of the experiment.We omitted some variables from the analysis that were highly correlated (r > 0.8) to other soil variables and would have been redundant (i.e., Ca was omitted because it was correlated with Mg, and Na was omitted because it was correlated with C).", "title": "Enhancing Aboveground Carbon Storage and Invasion Resistance through Restoration: Early Results from a Functional Trait-Based Experiment", "id": "10.6", "keywords": [] }, { "file_name": "Rayome et al. - 2018 - Enhancing Aboveground Carbon Storage and Invasion .pdf.tei.xml", "page_content": "After 2 yr of experimental treatment conditions, our results indicate support for \"Moderate\" C cycling species mixtures but lack of support for \"Complementary\" mixtures.Existing native C totals in treatment plots ranged from 0.10 kg to over 6,200 kg, and reference plots that included invasives ranged from 0.14 kg to 5,652.01 kg.The hybrid community treatment type significantly influenced C storage in outplants (F = 19.8;df = 3, 9; P < .0003),but the block effect was not significant (F = 0.813; df = 3, 9; P = .81).The Moderate Redundant treatment had significantly more C than the other three treatments (Figure 2).\nC measures at the treatment level were driven by a handful of species.Outplants (F = 63.03;df = 15, 715; P < .0001)and recruits (F = 6.09; df = 3, 204; P = .0005)species varied significantly in their C.Total experimental C value ranged greatly (Table 2, Figure 3).For most cases, the differences in treatments can be explained by responses of individual species therein.For outplants, Artocarpus altilis, Cibotium glaucum, Terminalia catappa, Rhus sandwicensis, and Persea americana all had high C contributions (Table 2, Figure 3).Of these, the overwhelming majority of new outplant C originated from the 622 kg contribution of A. altilis.Further, Aleurites moluccana, Morinda citrifolia, Samanea saman, Syzygium malaccense, and Mangifera indica all had notable contributions in at least one treatment type.In contrast, recruit biomass was most heavily influenced by contributions from R. sandwicensis, Pipturus albidus, and C. glaucum (Table 2, Figure 3).Of these, R. sandwicensis contributed the majority of new recruit C, over 30.39 kg.Further, environmental conditions had little influence on the aboveground C values.Before experimental conditions, there were no significant differences in canopy openness, soil pH, or nutrients (C, N, P, Mg, Ca, Na, K).The only significant effect of environmental conditions was that plots with lower soil P had significantly more outplant C (r = -0.5239,P = .0373),but none of the other environmental variables had any significant differences.In addition, none of the environmental variables was related to the existing tree density of basal area before the start of the experiment (data not shown).", "title": "Enhancing Aboveground Carbon Storage and Invasion Resistance through Restoration: Early Results from a Functional Trait-Based Experiment", "id": "10.7", "keywords": [] }, { "file_name": "Rayome et al. - 2018 - Enhancing Aboveground Carbon Storage and Invasion .pdf.tei.xml", "page_content": "We predicted that the treatment composed of species with moderate C cycling traits and more complementary functional trait values would have increased C storage when compared with treatments of species having slower C cycling traits or more redundant functional traits.However, as shown by our results after 2 yr of Liko N\u00e4 Pilina experimental conditions, the Moderate Redundant treatment had higher C storage than the other three treatments (Figure 2).Contrary to our expectations, core species in the two Moderate treatments, although chosen in part to have trait values related to increased rates of C cycling, were in some cases surpassed in growth in the field by \"noncore\" species.Further, Redundant species mixtures composed of fast-growing and high carbonaccumulating species might provide a benefit to a restoration project earlier, whereas Complementary mixes might support long-term carbon accumulation.In our system, the level of weeding required was high; therefore the Moderate Redundant treatment has the advantage of being more likely to obtain quick canopy closure and aid in preventing additional maintenance.Our experimental mixes were selected based on functional trait expression of adult specimens, and the results after 2 yr may not represent mature trees.It is therefore important in restoration planning to recognize the importance of emphasizing growth trajec tories within the greater scope of species life history when prescribing restoration designs.Discarding the value of Moderate Complementary mixtures would be unwise without contextualizing results at the species level.Outplant and recruit species including A. altilis, C. glaucum, T. catappa, R. sandwicensis, and P. americana are indeed influencing C, but they are growing at the same rate regardless of treatment (Figure 3).For example, A. altilis is successfully naturalized or cultivated in almost all suitable terrestrial ecosystems due in part to its ease of management, vigorous growth habit, and versatility in use ( Janick andPaull 2008, Breadfruit Institute 2016).Although the functional trait approach used in this experiment may not be best for all species types or successional stages, it is important to report and understand early results for comparison with later successional stages that may favor complementarity for sustainable productivity and invasion resistance.\nWe found that measuring C storage in outplanted individuals and recruits was an appropriate metric for evaluating effects on aboveground C in restoration approaches.It has been shown in these forests that once invaded, C cycling and storage is altered and no longer benefits C storage in the long term due to decreased longevity of invasives ( Hughes and Denslow 2005, Mascaro et al. 2012, Asner et al. 2016).Indeed, shifting C in a deliberate way toward native and noninvasive exotic species has implications for structural and subsequent ecosystem process changes.Current reference plots as well as that from both preexperimental conditions and nearby lowland sites indicates that increased invasive C detracts from native species recovery (Cordell et al. 2016).These new C cycling regimes may affect overall integrity and long-term resilience, especially when dwindling forests face continued pressure from impacts of continued global change.\nOur experimental work emphasizes the potential of Hawaiian forests as a simple and unique perspective for examining C, with species invasions, human impacts, and management interventions all occurring simultaneously (Friday et al. 2015, Ostertag et al. 2015, Asner et al. 2016).Carbon content varies among tropical tree species (Martin andThomas 2011, Orihuela-Belmonte et al. 2013), yet the relatively few species in historic Hawaiian lowland wet forests store C comparably to that of more biodiverse forests in other regions.Similarly invaded mature Hawaiian wet lowland forests store 72.8 Mg C ha -1 (Asner et al. 2016), and the managed density of our experiment allows up to 1 for species codes).Of contributing species, Artocarpus altilis had the most influence on C measures.Species values in individual treatments are displayed in Table 2. Species included here are outplants only; there was no significant difference between treatments with respect to carbon accumulation.\n291 Mg C ha -1 .This total includes reference conditions as well as existing trees, 4.02 Mg C ha -1 in outplanted individuals, and 0.19 Mg C ha -1 in recruits.Comparable tropical wet forests in Bolivia store approximately 67.5 Mg C ha -1 , and those in Brazil store 90 Mg C ha -1 ( Jespen 2006).\nconclusions Overall, we consider our experiment an important contribution to the growing literature on including C aspects in restoration as well as the importance of increasing stored C in forests globally (FAO 2010, Raunikar et al. 2010, Hurmekoski and Hetema \u00a8ki 2013, Payn et al. 2015).Our experiment supports the finding that mixed species plantings benefit C balances with peripheral benefits for other ecosystem services (Lindenmayer et al. 2012, Hulvey et al. 2013).We have previously shown the approach to be both appropriate and cost-effective for the Hawaiian lowland wet forest context [Cordell et al. 2016;D. D. Rayome, R. Ostertag, and S. Cordell (unpubl. data)].It is important to note that our experimental forest is still developing: canopy closure has not yet occurred, nor have several other factors commonly associated with a mature wet tropical forest.Changing experimental and environmental conditions continue to influence overall forest ecology, and we expect treatments to continue positive C cycling as plots mature and more-complex forest dynamics begin to unfold.We expect C and other services to vary after maturity as well, providing a basis for treatment comparison well into the coming decades.Longerlived species will likely affect C through their growth, and more prolific species or those with traits more useful for survival will affect C dynamics in ways that support competitiondriven C balances (Lusk et al. 2008, Suding et al. 2008).Finally, we understand that a more comprehensive understanding of C would benefit from inclusion of belowground C and not measured aboveground biomass (herbaceous and immature plant matter, snags, and other coarse woody debris; on-site carbon cycling through leaf detritus and connected trophic webs).Focus on more compre-hensive C effects allows for more informed restoration planning and increased likelihood of successful interventions (Pokorny et al. 2005, Cordell et al. 2016).The vital snapshot of early C values in this study will be useful for contrasting with later results when the trade-offs between rate of C accumulation and functional redundancy are realized.", "title": "Enhancing Aboveground Carbon Storage and Invasion Resistance through Restoration: Early Results from a Functional Trait-Based Experiment", "id": "10.8", "keywords": [] }, { "file_name": "Rayome et al. - 2018 - Enhancing Aboveground Carbon Storage and Invasion .pdf.tei.xml", "page_content": "We recognize staff, students, and interns from the University of Hawai'i, the USDA Forest Service Institute of Pacific Islands Forestry, and Stanford University for logistical and technical support.We also express appreciation for work underlying this analysis performed by Amanda Uowolo, Laura Warman, Nicole DiManno, C\u00e9line Jennison, and Palani Akana.We thank our partners in the Hawai'i Army National Guard Environmental Office (Angela Kieran-Vast, Craig Blaisdell, and Kristine Barker) and staff at Keaukaha Military Reservation for facilitating the establishment of the Liko N\u00e4 Pilina project.", "title": "Enhancing Aboveground Carbon Storage and Invasion Resistance through Restoration: Early Results from a Functional Trait-Based Experiment", "id": "10.9", "keywords": [] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "Climate change and human activities (e.g., ore mining and agricultural use) have caused the land ecosystem degraded worldwide (Chapin et al., 2000).These degraded ecosystems are known to seriously affect human economic and social life (Laughlin and Laughlin, 2013;Laughlin, 2014).As such, ecological restoration has rapidly developed to solve the urgent and complex degraded problems (Brown and Amacher, 1999;Cardinale et al., 2012).Among all, identifying the suitable species that can successfully restore the degraded ecosystems is the primary challenge in restoration science (Fry et al., 2013).However, this step requires a comprehensive understanding of the ecological restoration theory, including information on species interactions, successional processes, and resource-use patterns.Because these processes differ greatly across different ecosystems, it remains an enormous challenge.As this type of information is typically lacking, the selection of candidate species for restoration purposes are typically chosen using traditional trial-and-error method.Nevertheless, the trial-and-error method based largely on expert knowledge of the ecosystem (Rosenthal, 2003), and such information requires the involvement of a restoration practitioner who has a wealth of practical experience which is based on years of training (Padilla et al., 2009;Ostertag et al., 2015).\nPlant functional traits (including morphological, physiological and phenological characteristics that are linked to plant life history strategies) are fundamental to understanding plant adaptations and distributions.In theory, plant functional traits can be useful for selecting suitable species to restore degraded ecosystems by computing the similarity index between the target species (e.g., species have been proved to have high survival rates in the degraded ecosystems) and the potential species (e.g., all possible suitable species) (Zhang et al., 2018;Wang et al., 2020).Several studies have successfully used functional traits to select suitable plant species for restoring degraded ecosystems (Bochet and Garc\u00eda-Fayos, 2015;Ostertag et al., 2015;Guimar\u00e3es et al., 2018;Werden et al., 2018;Rayome et al., 2019;Wang et al., 2020).However, each of these studies used different screening methods or different sets of functional traits to select species.A generalized framework and a software platform which can assist this selection process is necessary to people which are new to this field.\nRecently, Laughlin (2014) proposed a quantitative trait-based species selection process for selecting suitable species to restore degraded ecosystems.In a previous study (Wang et al., 2020), we developed a species screening model based on this quantitative trait-based theory (Shipley et al., 2006).Also, we successfully applied this model to select best fit plant species for restoring a tropical coral island which is part of Hainan Island, China (Wang et al., 2020).However, along this way we realized that a software platform which can help automate this modeling process is necessary as it will save time for beginners.\nIn this study, we introduced a newly developed, web-based species selection software platform and named it as \"Restoration Plant Species Selection (RPSS) Platform.\"The platform aims to select suitable plant species for restoration purpose based on plant functional traits.It was developed using the highlevel R programming language1 which is popular for statistical computing and graphics.It is a web-based application that can run on a wide variety of operating systems, like Windows, Linux, and macOS and it can work based on various general browsers (e.g., IE, Chrome, Firefox, and others).The software platform makes use of several external R packages that perform various functions, including computing similarity rankings and drawing multiple graphic functions.For users, there are only two sets of information needed to run the software: (1) the trait values of the target species (species which proved to meet the specific goals) and ( 2) the trait values of the potential species (species to be chosen from).In addition, the RPSS platform also has a graphic user interface that helps users execute each of the functions.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.1", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "In this section, we will introduce the abiotic filtering theory which is the key selection theory for our software platform, and the traitbased plant species selection process in details.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.2", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "Abiotic filtering theory claims that for adapting well to the limiting abiotic environment (e.g., light and soil nutrient), all species have to develop similar traits to optimize fitness and performance (Laughlin, 2014).If we have found at least one or two species that can adapt well to the specific environment in the degraded ecosystem, many species that are appropriate for restoring the degraded ecosystem can easily be selected, as long as we compare the similarities in functional traits among them.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.3", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "Step 1. Determine the Target Species\nThe target species are needed as a prerequisite in this process.Generally, the target species are species that have met certain standards of durability in ecosystems.For restoration purposes, the target species are which we already know have high survival rates in the degraded ecosystems.However, in some cases, the restoration is initiated from bare soil conditions which means that the target species are not easily found based on poor background knowledge.We therefore use the traditional trialand-error method to determine them (Wang et al., 2020).\nStep 2. Prepare the Pool of Potential Species Potential species pool consists all the possible plant species which may be suitable for our specific purposes.For restoration purposes, the potential species pool should include both historical native species and non-native species that have the potential to restore the degraded ecosystems.The non-native species are which found in regions that have similar environments with the study site (Ostertag et al., 2015).However, in some cases, native species or non-native species are not easily found.For example, the study site may be located on a highly degraded island on which almost no native plants can be found.In such situations, only non-native species will be considered for the potential species pool.\nStep 3. Select the Appropriate Functional Traits That Can Capture the Key Characteristics of the Particular Purpose\nThe selected functional traits should capture the key characteristics of the degraded ecosystems' environmental conditions.For example, leaf turgor loss point should be measured to reflect plant resistance to the drought stress, if drought stress is a key characteristic for the degraded ecosystem.Additionally, the appropriate traits should be well measurable for laboratories.If there is not any information on the characteristics already, the users may try to measure all possible functional traits (e.g., commonly measured morphological and physiological traits).We focused on different species that expressed similar trait values that were relevant to our objectives.\nStep 4. Collect and Prepare Functional Trait Dataset All trait information was obtained from field observations and laboratory measurements.Many previous studies have addressed how to collect and measure the functional traits.For the detailed information of how to prepare the functional trait dataset, please refer to some previous studies (Zhang et al., 2019;Wang et al., 2020) and Supplementary Tables 2,3 in the Supplementary Materials.\nStep 5. Use Functional Traits to Select the Best Fit Species\nIn our software platform, the maximum entropy (Maxent) model (Shipley et al., 2006) was used to screen plant species that are most functionally similar to the target species.The model outputs relative abundances to screen suitable species from the potential species pool.We defined relative abundance as the index of similarity between the target species and the potential species.The larger the value of similarity index, the more ecologically similar the two species are, and vice versa.Using this index, we assumed that the most similar species to the target species was also the most suitable species for restoring the degraded ecosystems.\nStep 6. Monitor Seedling Survival of the Selected Plant Species\nAt the final step, the survival rates of the selected species should be monitored to check (1) if they have the high survival rates than the unselected species; and (2) if they have the comparable survival rates to the target species.If both requirements are met, that means our species screening process is succeed.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.4", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "In this study, we developed a web-based software platform named \"Restoration Plant Species Selection (RPSS) Platform\" to aid in the trait-based plant species selection process.The RPSS platform is written in R language, and the main page of the platform is shown in Figure 1.The user only needs to prepare trait values for the target and potential species in text format (see data examples in Supplementary Tables 2,3) when using the software platform.The software outputs both text and figure results of similarity rankings for each potential plant species, ranked from highest to lowest.The returned similarity index can indicate the similarity rank of each candidate plant species to target species.To establish the whole functions of the platform, we make use of numerous external R packages to perform various functions.For example, the FactoMineR package2 is used to perform the PCA analysis, the FD package3 (Lalibert\u00e9 and Legendre, 2010;Lalibert\u00e9 and Shipley, 2010;Legendre, 2010) is used to compute the similarity rankings based on the Maxent model (Shipley et al., 2006), the ggplot2 package4 is used for graphic outputs, and the shiny package5 helps accomplish the web functions.\nThe software is open source.Anyone who needs the software can contact the corresponding author to acquire the whole software platform.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.5", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "In this section, we illustrate how to use our RPSS software platform successfully selects suitable plant species to restore a degraded ecosystem.Our study site is located in a tropical coral island which is part of Hainan Island, China.Because of multiple harsh environmental characteristics (e.g., high temperatures, strong light, drought, no soil, and other harsh conditions) (Zhang et al., 2019;Wang et al., 2020), it is difficult for plants to establish and grow in this area.For human habitation and economic development, restoration of the vegetation on this island is important and urgent.Plantations of some species are considered a quick way to vegetate the island and make the island suitable for the development of plant communities native to the region (Wolfe et al., 2015;Wang et al., 2020).The objective of this section is to use our newly developed software platform to select suitable plant species that could be used to restore the degraded ecosystem of this area.Specifically, this island can be comparable to primary succession, as it is only made of rock and sand and does not have any soil for plants to colonize.If finally, the species selected by our software platform can indeed be suitable for restoring this extremely degraded island, we believe that our software can be widely used for restoring multiple different types of degraded ecosystems.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.6", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "Our study site is located in a tropical coral island which lays to the southern of Hainan Island (lying between 108 \u2022 37 -111 \u2022 03 E and 18 \u2022 10 -20 \u2022 10 N), China.The study site has an area of approximately 1 km 2 and the mean altitude is about 5 meters (Wang et al., 2020).\nThe coral island our study site located has a tropical monsoon oceanic climate with a mean annual temperature of 28 \u2022 C. The average annual precipitation on the island is about 2,800 mm, and most precipitation occurs between April and September.The adverse environments in this study site are characterized by high temperatures, intense light, drought and high salinity and alkalinity soil, in which it is difficult for plants to colonize and grow (Zhang et al., 2019).For human habitation and economic development, developing successful restoration management strategies for this area is exceedingly important.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.7", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "As described earlier, the trait-based species screening process requires target species as a baseline.However, our restoration is initiated from bare soil condition which no native plants can be found.Only the traditional trial-and-error method can be used to acquire the target species.So, we have identified 20 species on a nearby island with similar environmental conditions to the study site.These species were cultivated (watering and fertilizing) in Wenchang City, Hainan Island for 1-3 months and then transplanted to the study site.Three years later (from the year 2014 to 2017), target species were defined as those with survival rates >90% based on trial-and-error method.Ultimately, three species Scaevola sericea, Ipomoea pes-caprae, and Cynodon dactylon'Yangjiang' were selected as target species for trees, vines, and herbs (Li et al., 2016;Luo et al., 2018;Wang et al., 2020).\nIn order to prepare the potential species pool for restoring the degraded coral island, we reviewed literatures and studied surveys of plants in four tropical regions which the climatic and environmental conditions are similar to our study site in global.The four regions are the South China Sea Islands, the South Pacific Islands/Hawaii, the Indian Ocean Islands, and the Caribbean Sea Islands.At last, 66 species were selected to form the potential species pool.These species included a wide range of plant groups containing trees, shrubs, herbs, vines, legumes, semi-mangrove plants, and some medicinal and edible plants.The list of the 66 candidate species is shown in Supplementary Table 1 (Wang et al., 2020).The seedlings of these potential species were cultivated mainly including watering and fertilizing in Wenchang City, Hainan Island.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.8", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "We want to find out key traits that can assist plant species adapt well to the specific environments of the degraded ecosystem.\nThus, the traits will help to select suitable species to restore this place well.Based on our previous studies (Luo et al., 2018;Wang et al., 2020), we identified 28 traits that associated with the harsh environmental conditions on the tropical coral island.The selected functional traits are relate to drought resistance, resource allocation, antioxidant and photosynthetic capacity: specific leaf area (g/cm 2 ), leaf water conductance (mmol m-2 s-1 MPa-1), leaf dry matter content (%), stomatal conductance (mmol m -2 s -1 ), stomatal density (numbers mm -2 ), upper epidermic thickness (\u00b5m), palisade tissue (\u00b5m); spongy tissue (\u00b5m), lower epidermic thickness (\u00b5m), maximum photosynthetic rate (mol m -2 s -1 ), stomatal conductance (mmol m -2 s -1 ), instantaneous water use efficiency (\u00b5mol mol -1 ), transpiration rate (mol m -2 s -1 ) and others.The 28 selected key traits are listed in Table 1.Because we don't have enough fresh herb leaf samples, some leaf related traits were not able to measure for herb species.As a result, we only measured 19 traits for herbs, but 28 traits for wood species.The omitted nine leaf structure-related traits are leaf/palisade/spongy tissue thickness, palisade tissue width, upper epidermis thickness, guard cell length, stomatal density, and stomatal area index.For three target species, only Cynodon dactylon'Yangjiang' had measured 19 traits while others all had 28 traits.Among all 66 potential species, 6 species had only 19 traits, others had 28 traits.For the selected target and potential species, traits on mature and healthy leaves of ten individuals for each species in a growing season were measured.\nThe measurement methods and trait dataset can be available in the supporting files.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.9", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "We input the trait values of target and potential species into the RPSS platform, the input file format can be found in Supplementary Tables 2,3.The software returned the final screening results in both text and graphic formats.In this case study, we used the Windows operating system and Google Chrome browser to accomplish the whole analysis process.The example graphic results are shown in Figures 234.For the target species Scaevola sericea, four plant species (Artocarpus heterophyllus, Pluchea indica, Hibiscus tiliaceus, and Ficus microcarpa) were selected as the potential species.For target species Ipomoea pes-caprae, six plant species (Ipomoea tuba, Pluchea indica, Pandanus tectorius, Hibiscus tiliaceus, Medicago sativa, and Sesuvium portulacastrum) were screened to be more similar to the target species.For target species Cynodon dactylon'Yangjiang, ' four plant species (Spinifex littoreus, Lepturus repens, Miscanthus sinensis, and Cerbera manghas) were outperformed than others.After removing duplicate species, we selected a total of 12 species for vegetation restoration in this study.\nFinally, we monitored seedling survival for all species in order to test (1) whether the selected species had the comparable Leaf water utilization Sources and physiological functions are also shown.survival rates with the target species; and (2) whether the selected species had higher survival rates than the unselected species.We cultivated 1,000 seedlings for each species in a nursery in Wenchang city, Hainan Island, China.In July 2017, we transplanted seedlings from our nursery to the tropical island.Each plant was planted on the same substrate type, and each row was divided into three repeat planting areas.The planting density was maintained at 80-100 plants per hectare.We recorded each plant's annual growth rates.After 2 years' follow-up, we recorded the final survival rate for each species using the following formula:\nsurvival rate = currently remaining seedings original seedlings \u00d7 100%\nIn final results, comparably high survival rates were found among the three target species and the twelve selected species (about 86 to 94%).However, the survival rates of the non-selected species were significantly lower than the target species in all cases (see Supplementary Figure 1).The results showed that the species selected by our software platform had significantly better survival rates than the non-selected species.This means that the software platform successfully performed the species selection for the purpose of restoration in our study site.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.10", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "In this study, we introduce a newly developed web-based species screening software platform -the RPSS platform.It uses a traitbased framework to select suitable plant species for particular purposes (e.g., restore the degraded ecosystem and others).The software is implemented in the high-level R language and uses some external R packages (e.g., FD package, ggplot2 package, and shiny package) to realize the model calculations, the picture productions, and the web functions.The GUI equipped in the software allows an easy execution of purposed analyses.Because it is a web-based platform, it can run across different operating systems including Windows, Linux, and macOS.Compare with other similar work (e.g., Restoring Ecosystem Services Tool, REST, Rayome et al., 2019), our software is a web-based platform which don't need to install any programs on the local computer.\nIn addition, our RPSS platform not only equipped with a webbased GUI to facilitate execution of various functions, it can also run as a script program in R language batch mode.Thus, users can clearly see which commands are executed, so as to better understand the science nature behind the plant species screening process.We demonstrated a successful application of this software platform for selecting suitable species to restore a highly degraded coral island in Hainan Island, China.The example shown in the paper is designed to demonstrate the capabilities of the software platform.We have tested our software platform in different operating systems (e.g., Windows, Linux, and macOS) and various browsers (e.g., IE, Firefox, Google Chrome), all the test samples ran normally and the results were the same which proved that our software platform was stable across different systems.The RPSS platform is an evolving program with many directions for future development (e.g., add new features, output multiple types of figures).We believe that our software platform will have broad applications in the future, especially for selecting many appropriate plant species to restore degraded ecosystems.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.11", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "This study was supported by the National Natural Science Foundation of China (Grant No. 41905094), a start-up fund from Hainan University, China [Grant No. KYQD (ZR) 1876], NSFC-Guangdong Joint Fund, China (Grant No. U1701246), and Youth Innovation Promotion Association, Chinese Academy of Sciences.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.17", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "This study was supported by the National Natural Science Foundation of China (Grant No. 41905094), a start-up fund from Hainan University, China [Grant No. KYQD (ZR) 1876], NSFC-Guangdong Joint Fund, China (Grant No. U1701246), and Youth Innovation Promotion Association, Chinese Academy of Sciences.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.18", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.19", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.20", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "CW conceived the ideas and designed the software platform, analyzed the data, and wrote the manuscript.NL collected the data.All authors contributed critically to the drafts and have final approval for publication.\nThe Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fevo.2021.570454/full#supplementary-material\nThe authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.21", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "CW conceived the ideas and designed the software platform, analyzed the data, and wrote the manuscript.NL collected the data.All authors contributed critically to the drafts and have final approval for publication.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.22", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fevo.2021.570454/full#supplementary-material", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.23", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Wang et al. - 2021 - A Web-Based Software Platform for Restoration-Orie.pdf.tei.xml", "page_content": "The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.", "title": "A Web-Based Software Platform for Restoration-Oriented Species Selection Based on Plant Functional Traits", "id": "11.24", "keywords": [ "vegetation restoration", "plant functional traits", "maximum entropy model", "restoration plant species selection platform", "R language" ] }, { "file_name": "Sandel et al. - 2011 - Using plant functional traits to guide restoration.pdf.tei.xml", "page_content": "Attempts to restore degraded communities often have mixed success across different systems, even when similar treatments are applied.This is in part because the plant community contexts for these treatments exert important controls over the restoration outcome (Hendrickson and Lund 2010).For example, in systems where native and exotic species are functionally similar, it may be difficult to design restoration treatments that benefit one group over the other (Levine andD'Antonio 1999, Corbin andD'Antonio 2010).Further, even if native and exotic species do differ in particular traits, restoration treatments may not benefit all native species equally.\nFunctional traits and trait-based community ecology theory can provide a basis for predicting the success of a restoration treatment in a particular community (Lavorel and Garnier 2002, Pywell et al. 2003, da Silveira Pontes et al. 2010, Roberts et al. 2010), helping to ensure that treatments are applied only where they will be most beneficial.Some research to date has used functional traits to predict species' success in restoration treatments (Pywell et al. 2003, Roberts et al. 2010), with considerable success.For example, Pywell et al. (2003) have shown that seed germination rates and seedling growth rates are both positively associated with species success in sown restorations.These studies have suggested which species are most likely to successfully establish when creating restored communities and indicate, in a general sense, that the relative competitive and colonizing abilities of species may be predicted from their traits.However, little research has addressed how different restoration schemes, mediated by plant traits, can alter these relative abilities.For example, traits that might be advantageous under high-nutrient conditions, such as high seedling growth rates, may be disadvantageous under restored low-nutrient conditions.Such dependencies are predicted based on trait-environment relationships and can have potentially important consequences for restoration success, but have received little testing in a restoration context.\nNitrogen enrichment of terrestrial habitats via atmospheric N deposition (Galloway et al. 1995, Fenn et al. 2003) and invasion of N-fixing plants (Daehler 1998) is a significant component of anthropogenic global change (Vitousek et al. 1997a).Elevated soil N can facilitate the invasion of non-native species and complicate efforts to restore native biodiversity (Maron and Connors 1996, Vitousek et al. 1997b, Perry et al. 2010).In this study, we focus on two strategies designed to restore habitats where species invasion is associated with elevated N inputs (Perry et al. 2010): carbon (C) addition and the mowing and removal of above-ground biomass.Both of these techniques have been applied in a variety of systems to control non-native species and to favor native biodiversity (Maron and Jefferies 2001, Van Dyke et al. 2004, Blumenthal 2009, Alpert 2010, Perry et al. 2010).Carbon addition is intended to stimulate bacterial growth, leading to immobilization of N in microbial biomass and a temporary decrease in plant-available N (Baer et al. 2003, Averett et al. 2004, Corbin and D'Antonio 2004).Mowing is meant to remove aboveground biomass, thereby reducing the competitive advantage of early-germinating or highly productive species, while removing mowed clippings exports N from the system (Zobel et al. 1996, Wilson andClark 2001).\nExperimental tests of these methods have produced mixed results.While both treatments have been shown to decrease the exotic-to-native ratio in some ecosystems, in others native species have experienced no benefit or those benefits have been limited to a few species (e.g., Maron and Jefferies 2001, Cione et al. 2002, Van Dyke et al. 2004, Blumenthal 2009).A likely explanation for this is that the simple classification of species into ''native'' and ''exotic'' categories may provide relatively little functional information.Direct measurement and description of plant strategies may enable better predictions for species responses to these, and other, restoration treatments (Pywell et al. 2003, Eschen et al. 2006).\nOne useful system for describing plant strategies is based on three major axes quantifying leaf, growth, and reproductive strategies (Westoby 1998).Plant species are arrayed along the leaf axis according to the speed at which they obtain returns on leaf investments (Chapin 1980, Reich et al. 1997, Westoby et al. 2002, Wright et al. 2004).Species with fast returns exhibit relatively rapid turnover of leaf biomass and tend to produce thin, low density leaves with high specific leaf area (SLA) (Wright and Westoby 2000, Westoby et al. 2002, Wright et al. 2004).The growth axis describes the structural characteristics of a plant, of which height is a key indicator.Tall species invest in structural biomass that improves their access to light, while short species can allocate a larger portion of their biomass towards photosynthetically active tissue, which is advantageous when light is readily available (Falster and Westoby 2003).The reproductive strategy axis can be related to seed mass, as larger-seeded species typically have a higher probability of germinating and can outcompete smaller-seeded species.However, plants that produce large seeds typically produce smaller numbers of seeds, limiting their ability to colonize available microsites (e.g., Turnbull et al. 1999, Henery andWestoby 2001).\nWe used these three strategy axes to make specific predictions regarding how plant community composition should respond to C addition and mowing.Because C addition reduces N availability, plants that maximize their long-term return on each unit of N obtained should benefit (Poorter and De Jong 1999).This should favor species with slow returns on investments (low SLA, high leaf density, Craine et al. 2001).The ability to fix N is also likely to become particularly valuable under C addition, so Nfixing species are likely to increase in abundance (Table 1).As soil resources become more limiting, competition for light should become less important (Tilman 1984, Knops andReinhart 2000).Under these conditions, large height and high leaf area, both traits related to light capture ability, should be less advantageous (Knops and Reinhart 2000).Finally, large seed masses contribute to improved seedling success, particularly in nutrient-poor environments (Milberg et al. 1998).With reduced soil N availability, the additional resources provided by a large seed should become more useful (Lee and Fenner 1989, Table 1).\nIn contrast, mowing and biomass removal in a restoration context should be expected to favor species that display rapid returns on leaf investments, with high SLA and low leaf density (Craine et al. 2001, Cruz et al. 2010).This is because long-term investments in leaves become less advantageous when those leaves are periodically destroyed by mowing.Mowing, like grazing, should also favor annual over perennial species (Hayes andHoll 2003a, Diaz et al. 2007).Short-statured species, which may avoid the mowing treatment altogether, should also benefit (Table 1, Louault et al. 2005, DiTomaso et al. 2007, da Silveira Pontes et al. 2010).Thus, though mowing treatments may be intended to benefit late-germinating species over early-germinating species in many cases, there are other modes by which the treatment can affect species differentially.\nA final way that we might predict changes in community composition with restoration treatments is in terms of the flexibility of species' traits in the face of environmental variation.Highly plastic species might be better able to take advantage of a wide range of conditions (Funk 2008, Berg and Ellers 2010, da Silveira Pontes et al. 2010).For example, though a species might have low SLA under ambient conditions, it could respond to mowing by producing leaves with a higher SLA, thereby achieving a more appropriate phenotype for the local environmental conditions.Such a species might obtain higher than expected abundances in mowed plots by virtue of its plastic responses.Though such responses are theoretically well-grounded and potentially important, they have rarely been investigated empirically (Berg andEllers 2010, da Silveira Pontes et al. 2010).\nWe tested these predictions in two coastal California grasslands.Exotic species, particularly exotic annual grasses, often enjoy competitive advantages over native California grassland species, and are widespread and abundant (for review see Corbin et al. 2007).Among the mechanisms that have been proposed to explain this competitive dominance are the exotics species' rapid germination and growth that allow them to co-opt light and soil resources (Jackson and Roy 1986, Dyer and Rice 1999, Abraham et al. 2009).Exotic species may also benefit from increases in N availability from nitrogen deposition (Weiss 1999, Fenn et al. 2003) and the invasion of nitrogen-fixing shrubs (Maron andConnors 1996, Haubensak et al. 2004).Both C addition and mowing have been applied in these systems, but without consistent benefits to native species (Alpert and Maron 2000, Maron and Jefferies 2001, Hayes and Holl 2003b, Corbin and D'Antonio 2004).We propose that the success or failure of these measures may depend on the presence or absence of significant functional differences between native and exotic species groups that could drive differential responses to these treatments.\nWe ask (1) whether community composition, summarized using plant functional traits and trait plasticity, responds as predicted to C addition and mowing and (2) whether native species possess suites of traits that allow them to benefit from C addition and mowing.Together, these two components allow a prediction for native species responses, which we test by examining actual responses of native species.This provides an assessment of whether functional traits and trait-based theory can contribute significantly to restoration ecology.Hickman 1993).The two sites have been free from livestock grazing for at least 35 years, though the Point Reyes site is visited frequently by native tule elk (Cervus elaphus nannodes).Soils at both sites may be N-enriched for two reasons.Each site has scattered individuals of Lupinus arboreus, a native shrub that has been shown to increase soil N in coastal California grasslands (Maron and Connors 1996).In addition, coastal California sites in this region receive approximately 5-7 kg/ha/year wet and dry N deposition (Fenn et al. 2003, Weiss 2006).", "title": "Using plant functional traits to guide restoration: A case study in California coastal grassland", "id": "12.1", "keywords": [ "California", "carbon addition", "exotic", "grassland", "height", "mowing", "native", "nitrogen", "seed mass", "specific leaf area" ] }, { "file_name": "Sandel et al. - 2011 - Using plant functional traits to guide restoration.pdf.tei.xml", "page_content": "In 2005, we established 16 replicate mowing plots, 16 carbon addition and 16 control plots at each site.Plots were 5 3 5 m at Point Reyes and 4 3 4 m at Tom's Point, because of space limitations.Beginning in October 2005 and every six months thereafter, we applied the carbon addition treatment to the appropriate plots.Carbon was added as sucrose at 450 g/m 2 (189.5 g C/m 2 ), except the original application, which was 170 g sucrose/m 2 and 360 g sawdust/ m 2 (approximately 240 g total C/m 2 ).Sucrose is the most widely used C source in such experi-\nGeneration times, population response times\nVariation in suites of other traits\nNotes: SLA stands for specific leaf area, and NS indicates a non-significant treatment effect.For categorical traits (lifespan and growth form) hypotheses and treatment effect directions are given for the named trait state (e.g., Annual).Cases where a community mean trait value were expected or observed to increase are indicated with a ''\u00fe'', expected or observed decreases with a ''\u00c0'' and no prediction with a ''0''.Where measured treatment effects differed between sites, the treatment effects are given for Point Reyes first, then Tom's Point.ments because it is readily available and enters the soil quickly, and this application rate is within the range of commonly used rates, based on a review of 55 studies of C addition (Alpert 2010).Plots in the mowing treatment group were mowed to ;10-cm height each spring in late March or early April and all clippings were removed from the plots.Half of the plots of each treatment type also received a mix of native grass seed each fall.Germination rates of these seeds were extremely low, and seed addition plots were lumped with unseeded plots in the following analyses.\nWe sampled the plant community at the end of the growing seasons of 2006-2009 (primarily in June, including sampling in late May or early July in some years).We assessed the presence or absence of all plant species within 16 50 3 50 cm subplots in the central 4 m 2 of each treatment plot.In addition, in the summers of 2008 and 2009, we estimated the percent cover of each plant species occurring in the middle four subplots.This allows two measures of a species' abundance within a plot: 1) a spatial prevalence measure, the occurrence frequency of the species in the 16 subplots, and 2) the relative abundance of the species, based on percent cover estimates.We also estimated the percentage of bare ground in the middle 1 3 1 m of each treatment and control plot in 2009.Finally, in 2008 and 2009, we counted the number of flowering stalks of each native grass encountered in the survey.\nOn March 31, 2006, shortly after the spring C addition, we collected three 15-cm deep, 2-cm diameter soil cores from carbon-treated and control plots, to assess N pools and net N mineralization rates.We bulked the samples, extracted half of each soil sample immediately with 2 M KCl, and incubated the other half in the lab for 14 days.Then, we extracted the remaining soil samples with 2M KCl.All extracts were analyzed for NO 3 and NH 4 concentrations by the UC Davis Agriculture and Natural Resources Analytical Lab using the flow injection analyzer method.Net N mineralization rate was calculated as the amount of NH 4 \u00fe NO 3 in incubated soils minus the amount in soils extracted immediately.\nIn July 2006, we clipped all standing biomass from two 50 3 50 cm squares within half of the treatment plots at each site (n \u00bc 8).We repeated this sampling in July of 2008 and 2009, except that clipping was done in two 25 3 25 cm squares in all plots (n \u00bc 16).We then dried and weighed these samples.\nWe eliminated one plot from all analyses (a control plot from Point Reyes) that was, by the end of the study, nearly completely covered by a very large L. arboreus shrub.", "title": "Using plant functional traits to guide restoration: A case study in California coastal grassland", "id": "12.3", "keywords": [ "California", "carbon addition", "exotic", "grassland", "height", "mowing", "native", "nitrogen", "seed mass", "specific leaf area" ] }, { "file_name": "Sandel et al. - 2011 - Using plant functional traits to guide restoration.pdf.tei.xml", "page_content": "Between April and June, 2009, we collected data on plant functional traits.For species that were sufficiently abundant, we sampled one individual from each C addition treatment and control plot at each site.Mowed plots were not sampled, because the destruction of leaves by mowing made measuring leaf traits problematic.We sampled the first mature individual of each species encountered in a search beginning from a random corner of each plot.In this ideal case, we obtained 16 representatives of a species from each site for two treatments, for a total of 64 samples.However, many species were not present in all plots.In such cases, we took samples from up to two individuals from each plot where the species could be found, in order to come as close as possible to the goal of obtaining 64 samples per species.\nWe measured the height of the highest photosynthetic surface from each individual, and collected one fully-expanded, green, leaf.We measured the thickness of the leaf lamina using a high-precision caliper, avoiding any major leaf veins.Each leaf was stored, refrigerated, for at most 24 hours.We photographed each leaf flattened against a white background and determined the surface area of each leaf using ImageJ (Rasband 2009).Leaves were ovendried at 508C for at least 48 hours, then weighed individually to a precision of 0.01 mg.In addition to directly measuring height, leaf thickness, leaf area and leaf mass values, we calculated specific leaf area (leaf area per unit mass) and leaf density (leaf mass per unit area, divided by leaf thickness).For each species and each trait, we then calculated a species mean trait value from all control plot samples, all C addition plot samples, and all samples combined.Trait values based on all samples were used for calculating plot mean trait values (below), while control-and C addition-specific values were used to assess trait plasticity.All quantitative trait values were log-transformed prior to analysis.\nSpecies were also classified according to several categorical traits, including growth form (graminoid or not), lifespan (annual or not), nitrogen fixing ability and origin (native or exotic).Finally, we supplemented these data with data on species' mean seed sizes from the Kew Gardens Seed Information Database (Liu et al. 2008); most seed size data for the Californian species in this database are originally from Baker (1972).\nTo summarize shifts in community composition with the restoration treatments, we used weighted community trait means (Garnier et al. 2004).These community trait means were calculated by taking the mean trait value of all species occurring within a plot, weighted by each species' abundance within the plot.Two kinds of abundances were used.The first was the spatial prevalence measure, the proportional occupancy of a species within 0.25 m 2 subplots within the sampling plot.The second was the relative abundance measure based on visual estimates of percent cover for all species, taken in 2008 and 2009.", "title": "Using plant functional traits to guide restoration: A case study in California coastal grassland", "id": "12.4", "keywords": [ "California", "carbon addition", "exotic", "grassland", "height", "mowing", "native", "nitrogen", "seed mass", "specific leaf area" ] }, { "file_name": "Sandel et al. - 2011 - Using plant functional traits to guide restoration.pdf.tei.xml", "page_content": "Plot-level responses were assessed using repeated-measures ANOVA, with site and treatment as fixed factors.Three measures (percent bare ground, inorganic N pool and net N mineralization rate) were available only for a single year, so a simple ANOVA was used in these cases.In this manner, we assessed how the restoration treatments affected N pools, net N mineralization, bare ground, aboveground biomass, the abundance of native species (using both spatial prevalence and relative abundance measures) and community mean trait values for each trait.For categorical traits, plot mean trait values were square root-arcsine transformed prior to analysis.For all of the following tests, C addition-control comparisons were conducted independently of mow-control comparisons, because the comparison of C addition to mowing treatments was not of particular interest to us.\nWe tested for overall differences between species mean trait values of native and exotic species.Considering one trait at a time, we performed t-tests comparing the trait means of native species to exotic species.For categorical traits, we instead performed a chi-squared test.In addition, we tested for multivariate differences between native and exotic trait means using the multi-response permutation procedure (MRPP, Mielke and Berry 2001).\nWe also examined how species' traits differed between control and C addition plots.For each species and trait, we calculated a treatment effect size (plasticity treatment effect size, PTES) by taking the difference between that species' log mean trait value in carbon plots and the log mean value in control plots.We used only species with measurements from at least two individuals in each plot type.For each trait, we then performed a one-sample t-test to determine whether these PTES differed collectively from zero.For example, a significantly negative PTES for a trait indicates that species generally had reduced values for a particular trait in C addition plots compared to control plots.\nWe investigated whether species with higher trait plasticity (PTES) performed differently from those with low plasticity.These tests were analogous to the repeated-measures ANOVAs described above, but used PTES for each trait and each species (rather than trait means) to calculate plot means.If no species within a plot showed trait shifts with C addition, or a plot contained equal abundances of species whose traits shifted with C addition in each direction, the plot mean plasticity would be zero.On the other hand, if plots contained mostly species whose trait values (for a particular trait) decreased with C addition, the plot mean would be negative.\nResponses of individual species to C addition and mowing were assessed using a Monte Carlo permutation test.For each species, we took the mean occurrence frequency of that species in 50 3 50 cm subplots within a particular site and treatment.We then randomly shuffled treatment labels among plots within each site and recalculated these mean occurrence rates.We ranked the observed rate for the species against rates derived from 5000 random shuffles.This provides a distribution-free test of the effect of the treatments on each species' abundance.", "title": "Using plant functional traits to guide restoration: A case study in California coastal grassland", "id": "12.5", "keywords": [ "California", "carbon addition", "exotic", "grassland", "height", "mowing", "native", "nitrogen", "seed mass", "specific leaf area" ] }, { "file_name": "Sandel et al. - 2011 - Using plant functional traits to guide restoration.pdf.tei.xml", "page_content": "Carbon addition reduced standing pools of v www.esajournals.orginorganic N and net N mineralization rates in soils at both sites by approximately 90% (F 1,27 \u00bc 81.5, P , 0.0001, F 1,27 \u00bc 147.6, P , 0.0001, Fig. 1A).This reduction led to a decrease in productivity in these plots, with C addition plots typically showing 20-30% reductions in aboveground biomass (F 1,27 \u00bc 13.0, P \u00bc 0.0012, Fig. 1C,D).In 2009, there was 2.5 fold increase in bare ground in C addition plots relative to control plots (F 1,59 \u00bc 25.0, P , 0.0001, Fig. 1B).Mowing decreased aboveground biomass at the end of the growing season by 5 to 40% relative to controls (F 1,27 \u00bc 30.9, P , 0.0001, Fig. 1C,D) but had no significant effect on the percentage of bare ground (F 1,60 \u00bc 1.0, P \u00bc 0.33, Fig. 1B).Productivity in 2009 was unusually low, and both v www.esajournals.orgtreatments produced very small reductions in productivity in that year (treatment-by-year interactions P , 0.05).Here and throughout, interactions of site or year with treatment are not noted unless they were significant.", "title": "Using plant functional traits to guide restoration: A case study in California coastal grassland", "id": "12.6", "keywords": [ "California", "carbon addition", "exotic", "grassland", "height", "mowing", "native", "nitrogen", "seed mass", "specific leaf area" ] }, { "file_name": "Sandel et al. - 2011 - Using plant functional traits to guide restoration.pdf.tei.xml", "page_content": "The restoration treatments produced directional shifts in community composition along all trait axes.Carbon addition led to an increase in plot mean seed mass relative to control plots (F 1,59 \u00bc 9.1, P \u00bc 0.0038, Fig. 2).This was especially true at Point Reyes, with increasingly strong effects through time (treatment-by-year interaction, F 3, 185 \u00bc 9.5, P , 0.0001).At Point Reyes, C addition favored thick-and large-leaved species, while at Tom's Point, thin-, small-leaved species benefitted (treatment-by-site interaction: thickness, F 1,59 \u00bc 4.4, P \u00bc 0.0400, area, F 1,59 \u00bc 4.7, P \u00bc 0.0344, Fig. 2).In some years, C addition produced a shift towards shorter species (treatment-by-year interaction, F 3, 177 \u00bc 4.3, P \u00bc 0.0056).\nMowing produced a decrease in plot mean height (F 1,59 \u00bc 6.3, P \u00bc 0.0145) and an increase in v www.esajournals.orgmean SLA (F 1,59 \u00bc 5.1, P \u00bc 0.0272) that tended to become more pronounced through time (yearby-treatment interaction F 3, 177 \u00bc 2.7, P \u00bc 0.0464, Fig. 2).This change in SLA was caused by an increase in species with low-density leaves (F 1,59 \u00bc 4.8, P \u00bc 0.0325), rather than thinner leaves (P .0.6).\nNeither the total abundance of annual species nor grasses changed with C addition for either occurrence-based (P .0.45) or cover-based analyses (P .0.30).Nitrogen-fixing plants responded positively to C addition (F 1,59 \u00bc 20.1, P , 0.0001).This effect became increasingly apparent through time (treatment-by-year interaction F 3, 177 \u00bc 6.1, P \u00bc 0.0005), and was particularly strong at Point Reyes (treatment-bysite interaction F 1,59 \u00bc 5.7, P \u00bc 0.0206).Annual abundance increased in mowed plots relative to controls, particularly in the third and fourth year of the experiment (year-by-treatment interaction F 3, 177 \u00bc 3.3, P \u00bc 0.0225).\nTraits of individual species were plastic in response to C addition.Plants growing in C addition plots tended to have smaller individual leaf area (t 38 \u00bc \u00c03.6, P \u00bc 0.0010), leaf mass (t 38 \u00bc \u00c02.1, P \u00bc 0.0461) and SLA (t 38 \u00bc\u00c02.2, P \u00bc 0.0349), reduced height (t 38 \u00bc \u00c04.8, P , 0.0001), and higher leaf density (t 38 \u00bc 2.9, P \u00bc 0.0066, Fig. 3).However, the magnitude of species' trait plasticity did not predict their response to C addition for any of the quantitative traits we considered.\nNative and exotic species did not differ significantly in mean values for any of the quantitative species traits (P .0.05), nor were they separable in multivariate space (P .0.49).Native species were not more or less likely than exotic species to be grasses or N fixers (P .0.1), but were more likely to be perennial (v 2 \u00bc 10.7, P \u00bc 0.0011).\nConsistent with the absence of trait differences between native and exotic species, neither C addition (P .0.71) nor mowing (P .0.33, Fig. 4A,B) changed the relative abundance of native species as a group, calculated from either occurrence frequencies or percent cover.However, detailed stem counts of native grasses at Tom's Point did reveal a positive response of Bromus carinatus (F 1,30 \u00bc 4.0, P \u00bc 0.054) and Danthonia californica (F 1,30 \u00bc 6.8, P \u00bc 0.014) to C addition, and a tendency for B. carinatus (F 1,30 \u00bc 3.8, P \u00bc 0.060) to increase with mowing (Fig. 4C).There were no effects of treatments on native grass stem counts at Point Reyes (P .0.1), largely because native grasses were very rare at this site.", "title": "Using plant functional traits to guide restoration: A case study in California coastal grassland", "id": "12.7", "keywords": [ "California", "carbon addition", "exotic", "grassland", "height", "mowing", "native", "nitrogen", "seed mass", "specific leaf area" ] }, { "file_name": "Sandel et al. - 2011 - Using plant functional traits to guide restoration.pdf.tei.xml", "page_content": "Species responses to the C addition ranged from strongly positive to strongly negative.At Point Reyes, species that increased include the native N-fixer Lupinus bicolor and exotic N-fixers Trifolium subterraneum and Vicia sativa Species that strongly decreased in abundance included Bromus hordeaceus and Geranium dissectum, an exotic annual grass and forb, respectively.At Tom's Point, increasing species included the native forb Eschscholzia californica, native perennial grass Danthonia californica and native Nfixers Lupinus bicolor and Trifolium microdon.Some exotic species benefitted from C addition as well, including the exotic annual grass Bromus diandrus, exotic perennial grass Festuca arundinacea and exotic forb Erodium botrys.Pteridium aquilinum, a native fern, and Rumex acetosella, an exotic forb, decreased notably with C addition at Fig. 3. Intraspecific trait shifts induced by carbon addition.For each species and each trait, the treatment effect size is the species' log mean trait in control plots minus the log mean trait in carbon addition plots.For a particular trait, the heavy bar shows the median treatment effect size across 39 species, while the box shows the 25% and 75% quantiles and the whiskers show the extreme values.Traits for which treatment effects differed collectively from zero are shaded grey.\nv www.esajournals.orgTom's Point (see the Appendix).\nSeveral species showed strong responses to mowing.Rumex acetosella increased with mowing at both sites, as did the exotic forb Hypochaeris radicata and exotic annual grass Vulpia myuros.Geranium dissectum increased strongly with mowing at Point Reyes.At Point Reyes, the native forb Oxalis albicans tended to decline with mowing, while the native forb Achillea millefolium L. declined at Tom's Point (Appendix).", "title": "Using plant functional traits to guide restoration: A case study in California coastal grassland", "id": "12.8", "keywords": [ "California", "carbon addition", "exotic", "grassland", "height", "mowing", "native", "nitrogen", "seed mass", "specific leaf area" ] }, { "file_name": "Sandel et al. - 2011 - Using plant functional traits to guide restoration.pdf.tei.xml", "page_content": "Carbon addition successfully decreased soil N cycling rates, and both treatments reduced above-ground biomass, as intended.Previous tests of these strategies in California grasslands and in other systems have yielded similar effects (reviewed in Alpert 2010, Perry et al. 2010).", "title": "Using plant functional traits to guide restoration: A case study in California coastal grassland", "id": "12.9", "keywords": [ "California", "carbon addition", "exotic", "grassland", "height", "mowing", "native", "nitrogen", "seed mass", "specific leaf area" ] }, { "file_name": "Sandel et al. - 2011 - Using plant functional traits to guide restoration.pdf.tei.xml", "page_content": "Our hypotheses for the directions of trait shifts were largely upheld, supporting the utility of functional traits in predictive screening for restoration treatments.Consistent with an in-creased importance of seed provisioning in lownutrient environments, C addition favored largeseeded species.Nitrogen-fixers were also able to benefit from C addition, as their ability to obtain atmospheric nitrogen directly became more valuable as soil nitrogen was reduced.The increase in bare ground in C addition plots suggests that light competition may have become less important in these plots, as hypothesized, leading to the observed increases in shortstatured species.Mowing, on the other hand, favored short, annual species, which are better able to either avoid or rapidly recover from destruction (Diaz et al. 2007, DiTomaso et al. 2007).Thin, low density leaves are cheaper to produce, reducing the cost to the plant when those leaves are lost to mowing (Craine et al. 2001).Thus, we detected an increase in species with high SLA.Our ability to correctly predict these responses indicates that our understanding of plant functional strategies is sufficiently mature to facilitate improved restoration decisions.\nHowever, not all traits responded as predicted.The effects of C addition on certain leaf charac-Fig.4. Responses (mean 6 s.e.) of native species to restoration treatments.Neither mowing nor carbon addition increased native abundance at either site, measured as either occurrence rates (A) or percent cover (B).However, detailed counts of stems of native grass species did reveal positive responses of some species to the treatments (C).Bromus carinatus increased with both mowing and C addition, while Danthonia californica benefitted only from C addition.v www.esajournals.orgteristics that we had predicted to be important were weak or inconsistent across sites.We hypothesized that low-SLA species should benefit from C addition (Table 1), but observed no change in community composition in SLA.Leaf thickness, a trait that is closely related to SLA, showed the expected increase at Point Reyes but decreased at Tom's Point.In contrast, leaf area, a trait we expected to decrease with C addition, showed the expected response at Tom's Point, but the opposite response at Point Reyes.One possible explanation for this is that leaf thickness and area are well-correlated among species at our sites (r \u00bc 0.702).Thus, while C addition would be expected to favor small, thick leaves, the absence of species with this trait combination may have constrained the observed response.At Tom's Point, the decrease in leaf thickness with C addition may have been driven by an advantage of species with small leaf sizes despite a disadvantage of thin leaves, while the reverse occurred at Point Reyes.Our results suggest that height, seed mass and N-fixing ability, not leaf traits, are the most consistent predictors of response to C addition.Predictions of response to mowing, however, are improved when leaf traits are included.\nSurprisingly, intraspecific trait variation and trait plasticity were not predictive of species' responses to C addition (McLendon and Redente 1992).Previous work (Berg andEllers 2010, da Silveira Pontes et al. 2010) has suggested that species with more plastic traits should be able to occupy a wider range of conditions along abiotic gradients.While we did find large responses of species traits to C addition, variation among species in this response appeared not to drive community-level responses to treatments.", "title": "Using plant functional traits to guide restoration: A case study in California coastal grassland", "id": "12.10", "keywords": [ "California", "carbon addition", "exotic", "grassland", "height", "mowing", "native", "nitrogen", "seed mass", "specific leaf area" ] }, { "file_name": "Sandel et al. - 2011 - Using plant functional traits to guide restoration.pdf.tei.xml", "page_content": "Neither mowing nor carbon addition caused a consistent increase in native plant abundance.Though particular native species did benefit from the treatments, their gains were offset by reductions in the abundance of other native species.Thus, we do not recommend the widespread use of these restoration treatments without careful assessment of whether the treatments are appropriate to the target community.Our failure to detect general benefits for native species contrasts with some previous results (McLendon and Redente 1992, Zink and Allen 1998, Alpert and Maron 2000, Averett et al. 2004, Prober et al. 2005, Blumenthal 2009, Perry et al. 2010), but may be expected in systems like coastal Californian grasslands, where functional differences between some native and exotic species are small (Corbin and D'Antonio 2010).The utility of C addition and mowing as restoration measures depends on the existence of substantial functional differences between native and exotic species that could drive differential responses of the two groups.These conditions might be met in Midwestern prairies where the native prairie species are often slowgrowing relative to exotic species (Averett et al. 2004), or when the native and exotic species represent clearly distinct functional groups, such as shrubs and grasses (Zink and Allen 1998).At our sites, native and exotic species were not well separated along functional trait axes, so it is not surprising that neither treatment benefited native species as a group.\nThose native species that did benefit, however, tended to possess traits that we predicted would be advantageous.For example, Danthonia californica (which increased with C addition) is short, with low SLA, high leaf density and small leaves.Lupinus bicolor, one of the species with the strongest positive responses to C addition, is a large-seeded N-fixer.Considering the responses of particular exotic species to the mowing treatment, Geranium dissectum, a short, annual forb with high SLA and Vulpia myuros, a short, annual grass with small, thin leaves and small seeds, showed the strongest positive responses.Thus, it is possible to use functional trait-based predictions not only for species groups, but to predict the responses of particular species within groups.This should be useful to managers who wish to make more specific predictions regarding, for example, how particular rare native species or particular noxious exotic species might respond to restoration.\nInterannual variability in plot mean trait values was substantial, even in control plots.This is unsurprising, given California's highly variable climate and the potential for high turnover in the annual portions of these communities (Reever Morghan et al. 2007). 2008-2009 stands out as a particularly unusual year, with a v www.esajournals.orgrelatively late start to the growing season and low productivity.Species with low average height and high SLA may have benefitted from the climatic conditions in that year in a way that they did not in seasons with other climatic conditions, because of their ability to rapidly obtain returns on growth investments.Though little studied, efforts to restore habitats with high inter-annual climatic variability may be further complicated by variation in the effects of specific restoration methods depending on climate (Vaughn and Young 2010;G. F. Hayes and K. D. Holl, unpublished manuscript).\nThese results contribute to our understanding of the basic mechanisms by which mowing and C addition alter community composition, thereby increasing their utility as restoration methods.Though we detected no increase in native species abundance in response to either treatment, our results indicate the conditions under which one would expect to observe such increases.There are, however, other problems with the large-scale implementation of these restoration strategies that may make their application difficult, even in suitable communities.Sucrose is expensive, and while sawdust may provide a cheaper C source, its low density makes it difficult to spread the necessary quantities over a larger area.Further, the reduction of soil N by C addition is temporary (Sandel 2010), and it remains to be seen how long community changes induced by C addition last after treatments cease.Despite these challenges, though, C addition does hold promise in small-scale restoration projects if the community trait context is suitable (Alpert 2010, Perry et al. 2010).Notes: The abundance shown is the proportion of 0.25 m 2 plots occupied by each species, across all years within the site and treatment.The permutation rank shows the results of the Monte Carlo permutation test, in which treatment effects were assessed by randomly scrambling treatment labels among plots at each site.The value shows the proportion of random permutations in which the observed abundances in the named treatment exceeded the randomly generated abundance.Values greater than 0.975 indicate significant increases in the treatment relative to the control, while values less than 0.025 indicate significant decreases, at a \u00bc 0.05.For clarity, species with no occurrences within a site and treatment are left blank.v www.esajournals.org", "title": "Using plant functional traits to guide restoration: A case study in California coastal grassland", "id": "12.11", "keywords": [ "California", "carbon addition", "exotic", "grassland", "height", "mowing", "native", "nitrogen", "seed mass", "specific leaf area" ] }, { "file_name": "Sandel et al. - 2011 - Using plant functional traits to guide restoration.pdf.tei.xml", "page_content": "We are grateful to many field assistants who helped with the treatments and data collection.Audubon Canyon Ranch and the Point Reyes National Seashore provided exceptional support and we are particularly indebted to John DiGregoria.Wayne Sousa, David Ackerly, John Harte and the members of the Sousa Lab at UC Berkeley provided helpful comments throughout the project.Karen Holl and an anonymous reviewer offered useful suggestions on a previous version of this manuscript.B.S. was supported by an NSF predoctoral fellowship and by grants from the Robert and Nancy Beim Research Fund and the Gray Endowment.", "title": "Using plant functional traits to guide restoration: A case study in California coastal grassland", "id": "12.19", "keywords": [ "California", "carbon addition", "exotic", "grassland", "height", "mowing", "native", "nitrogen", "seed mass", "specific leaf area" ] }, { "file_name": "Sandel et al. - 2011 - Using plant functional traits to guide restoration.pdf.tei.xml", "page_content": "We are grateful to many field assistants who helped with the treatments and data collection.Audubon Canyon Ranch and the Point Reyes National Seashore provided exceptional support and we are particularly indebted to John DiGregoria.Wayne Sousa, David Ackerly, John Harte and the members of the Sousa Lab at UC Berkeley provided helpful comments throughout the project.Karen Holl and an anonymous reviewer offered useful suggestions on a previous version of this manuscript.B.S. was supported by an NSF predoctoral fellowship and by grants from the Robert and Nancy Beim Research Fund and the Gray Endowment.", "title": "Using plant functional traits to guide restoration: A case study in California coastal grassland", "id": "12.20", "keywords": [ "California", "carbon addition", "exotic", "grassland", "height", "mowing", "native", "nitrogen", "seed mass", "specific leaf area" ] } ]