Unique belowground ecological strategies of subtropical and tropical plant species expand the root trait space

biogeography; functional diversity; phylogeny; root economics space; root 67 functional traits; subtropics; tropics 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 4

84 Root trait variation may reflect the ecological and evolutionary processes shaping biodiversity, 85 but remains poorly quantified in the (sub)tropics. Here, we aim to further complete our 86 knowledge of belowground functional strategies by assessing the contributions of subtropical 87 and tropical species to global root trait diversity. We gathered root data for 1618 temperate, 341 88 subtropical, and 775 tropical species. We compared functional diversity among biomes and 89 calculated the unique contribution of each biome to the global root economics space. Further, we 90 determined if the within-variation of subtropical and tropical biomes is shaped by species’ niches 91 and/or differences in evolutionary history. Root trait expressions differed among biomes, but root 92 functional diversity did not. Furthermore, subtropical and tropical biomes accounted for 40% of 93 the unique root functional space within the global traits space. Species’ climate niches and 94 phylogenetic turnover explained variation in root traits (e.g., denser root tissue was associated 95 with drier sites) among subtropical but not tropical species. Through their unique root traits, 96 sub(tropical) species strongly expand the current ‘global’ root trait space. This work underwrites 97 their importance in conceptual models for more complete insights into how various belowground 98 strategies drive plant functional biogeography and biodiversity globally. 99 100 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 5

101 Subtropical and tropical biomes are hotspots of biodiversity across taxonomic groups. Recent 102 estimates suggest that tropical regions in America, Africa, and Asia jointly harbor ~224,000 103 species amounting to ~65 % of the known terrestrial vascular plant species (Raven et al. 2020, 104 Govaerts et al. 2021). The large contribution of tropical ecosystems to global biodiversity may be 105 associated with the high functional diversity in these systems, owing to e.g., biogeographical 106 transition zones in the subtropics where species with temperate and tropical strategies can coexist 107 (Morrone 2024); their long evolutionary time and strong competition and niche partitioning; and 108 their abundance of rare species with distinct plant ecological strategies that may minimize niche 109 overlap, enabling the coexistence of a large number of functionally distinct species (Leitão et al. 110 2016, Umaña et al. 2017). Trait diversification may be more pronounced belowground than 111 aboveground because root traits are more flexible in their coordination, giving rise to great 112 variability in root trait combinations compared to leaf traits (Kramer-Walter et al. 2016). As data 113 on root traits lag behind data on leaves in general but especially in the tropics (Cusack et al. 114 2024), our current insights on functional diversity and how it underlies species diversity and 115 coexistence in (sub)tropical biomes may only represent the tip of the iceberg and remain largely 116 incomplete. Here, we explore the emergent root trait variation among species across subtropical 117 and tropical biomes from around the globe, how much it contributes to global belowground 118 functional diversity, and to what extent it is driven by biogeographic and evolutionary histories 119 at the continental scale. 120 Interspecific variation in functional trait expressions generally reflects species’ variation in 121 ecological strategies (Grime 1977). Ecological frameworks approach this relationship from an 122 ‘economics’ perspective based on resource returns on carbon (C) investments into plant tissue 123 (Bloom et al. 1985). Species invest either in the construction of robust, durable tissue with slow 124 rates of resource return or in producing short-lived tissue with fast resource returns on relatively 125 low biomass investment (Reich et al. 1997, Westoby et al. 2002, Wright et al. 2004). This 126 tradeoff is evidenced by the global ‘leaf economics spectrum’ that spans a continuum of species 127 that exhibit a suite of ‘slow’ versus ‘fast’ leaf traits (Reich et al. 1997, Wright et al. 2004) and is 128 associated with high survival rates in adverse environments versus high growth rates on 129 productive sites, respectively (Reich et al. 1998, Aerts and Chapin III 2000, Poorter and Bongers 130 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 6 2006). The traits of absorptive roots (i.e., fine roots responsible for nutrient and water acquisition 131 and therefore, functional analogs of leaves; hereafter: roots) have been assumed to be similarly 132 coordinated along a fast-slow axis in a ‘root economics spectrum’ (Mommer and Weemstra 133 2012, Reich 2014), but more recent work shows that root trait covariation is multidimensional 134 (Kramer-Walter et al. 2016, Weemstra et al. 2016, Bergmann et al. 2020). A global analysis of 135 root trait data showed that interspecific covariation in root traits was best captured by two 136 independent axes in a ‘root economics space’ (Bergmann et al. 2020). The ‘collaboration axis’ is 137 assumed to reflect the extent to which species rely on (arbuscular) mycorrhizal symbiosis for 138 nutrient and water acquisition. It ranges from species with a high specific root length (SRL, root 139 length per unit root dry mass) that enhances direct water and nutrient uptake by roots (i.e., the 140 do-it-yourself strategy), to species with thick roots that rely more heavily on mutualistic 141 associations for resource capture (i.e., the outsourcing strategy). The ‘conservation axis’ 142 resembles the traditional fast-slow tradeoff, separating species that have tough, long-lived roots 143 (e.g., by having high root tissue density, RTD) with slow resource returns on C investments from 144 those with resource-acquisitive roots with low biosynthesis costs and fast returns on C 145 investments (e.g., by having high N concentrations and thus, assumingly, fast metabolism) 146 (Bergmann et al. 2020). This global multidimensional belowground framework is increasingly 147 serving as a baseline for comparative approaches in functional ecology. 148 The root economics space was tested using 748 woody and non-woody species sampled around 149 the globe but came with the major caveat that tropical and subtropical species are vastly 150 underrepresented (Bergmann et al. 2020). As such, this lack of data and knowledge on 151 (sub)tropical root trait (co-)variation hinders our understanding of overall ecological processes 152 and mechanisms (Freschet et al. 2017, Bergmann et al. 2020, Carmona et al. 2021). Taking this 153 bias towards the temperate biome into account is particularly relevant since (co)variation in root 154 traits may well deviate among biomes (Gu et al. 2014, Weemstra et al. 2023) owing to different 155 environmental conditions and unique evolutionary processes that select for different trait 156 expressions (Comas et al. 2012, Eiserhardt et al. 2017, Valverde-Barrantes et al. 2021). Biotic 157 and abiotic differences among biomes, such as nitrogen (N) versus phosphorus (P) limitation or 158 co-limitation (Vitousek 2004), nutrient cycling rates (Marklein et al. 2016), mycorrhizal 159 associations for the dominant canopy species (Steidinger et al. 2019, Soudzilovskaia et al. 2019), 160 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 7 and the length of the growing season suggests that trait syndromes found in temperate areas may 161 not represent belowground functional diversity in lower latitudes. 162 Improved data availability already demonstrates overall differences in root traits across biomes 163 associated with the collaboration axis. Subtropical and tropical species generally have thicker 164 roots with lower SRL than temperate species (Freschet et al. 2017, Ma et al. 2018), possibly 165 reflecting the dominance of arbuscular-mycorrhizal (AM) canopy trees of tropical forests 166 compared to ectomycorrhizal (EcM) dominance in temperate forests (Tedersoo 2017). Since AM 167 fungi colonize the root cortex, their hosts tend to have thicker root tips with a larger cortex than 168 EcM-associated species, where the fungal mantle develops outside the root (Brundrett 2002, 169 Kong et al. 2014). Biomes may also exhibit root trait differences associated with the 170 conservation axis. Temperate ecosystems with lower temperatures, slower nutrient cycles, and 171 shorter growing seasons, may in turn filter for species with ‘slow’, robust root traits (e.g., high 172 RTD) to conserve resources over the long term (Grime 1977, Reich et al. 1998, Aerts and Chapin 173 III 2000, Reich 2014). Tropical forests, in turn, may provide overall more favorable abiotic 174 environments with longer growing seasons, higher temperatures, and faster microbial-driven 175 nutrient cycling that would select for ‘fast’ roots (high root N) with a strong competitive 176 advantage for resource uptake. Alternatively, tropical plant species may be expected to have 177 ‘slower’ roots than temperate species because of higher pressure from natural enemies, e.g., 178 soilborne pathogens (Delgado-Baquerizo et al. 2020), selecting for chemically-protected and less 179 palatable roots with high RTD (Freschet et al. 2017, Xia et al. 2021) and lower root N 180 concentrations (Freschet et al. 2017, Laughlin et al. 2021). Owing to these different biotic and 181 abiotic selective forces, subtropical and tropical tree species may take up distinct positions in the 182 global root economics space, which currently mostly describes variation in root traits in 183 temperate, continental ecosystems. 184 Subtropical and tropical systems are often treated as single homogenous biomes. Overlooking 185 the large variation in evolutionary trajectories and geographical and environmental 186 characteristics they comprise (Townsend et al. 2008, Fujii et al. 2018) constrains our knowledge 187 of root trait variation within the (sub)tropics and its implications for global diversity. As a first 188 example, evolutionary differences may be large as the tropics comprise multiple 189 phytogeographic regions, i.e., phytogeographical delineation based on evolutionary relationships. 190 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 8 This variability leads to greater evolutionary distinctness within the tropics and co-occurrence of 191 deeply diverged lineages within and across continents (Carta et al. 2022). Secondly, in the 192 American tropics, the arrival of boreo-tropical plant species (e.g., Quercus and Salix), large 193 topographic variation resulting from the rise of the Andes (Antonelli 2021), and rapid plant 194 radiations at high elevations (Madriñán et al. 2013) may further enlarge root trait variation within 195 the region. Thirdly, several taxonomic groups evolved unique, contrasting belowground 196 symbiotic traits within and across continents, reflected by the local dominance of canopy trees 197 like Dipterocarpaceae in Southeast Asia or Fagaceae and Juglandaceae in Central and South 198 America highlands breaking the perceived dominance of AM-associated species and N-fixers in 199 the tropics (Steidinger et al. 2019). Fourthly, geological changes in Africa have resulted in 200 ecosystem turnover from the typical moist forests into savannas and dry woodlands (Raven et al. 201 2020), with savannas currently covering 50% of the African continent (Osborne et al. 2018). 202 These examples illustrate how - across the tropics - different continents would have their own 203 root trait expressions because of their unique biogeographic and evolutionary history (Comas et 204 al. 2012, Echeverría-Londoño et al. 2018, Weemstra et al. 2023) to fill their belowground niche 205 in these heterogeneous environmental gradients. 206 Environmental gradients e.g., in nutrient and water availability, may also differ within the 207 (sub)tropical biomes, shaping belowground plant strategies (Fujii et al. 2018, Cusack et al. 2021, 208 Yan et al. 2022b). In Australia, the presence of non-mycorrhizal proteoid roots on severely P-209 impoverished soils is characteristic (Lambers et al. 2006), while adaptations to fires play another 210 important role in root allocation and trait expression. Plants in Africa had higher median RTD 211 and SRL than sites in Asia and the Neotropics, which was partly attributed to root morphological 212 plasticity in response to frequent exposure to water stress in African systems (Addo-Danso et al. 213 2020), and within African tree species, seedlings from dry and moist ecosystems had distinct root 214 trait syndromes (Boonman et al. 2020). In other words, across the “global (sub)tropics'', as a 215 response to contrasting environmental gradients, it is expected that variation in root trait 216 expressions depends on species’ niches, such as the “typical” climatic or soil conditions of the 217 species, i.e., niche position, as well as the size of the species' environmental space, i.e., niche 218 breadth (Vleminckx et al. 2023). 219 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 9 Here, we aim to expand our understanding of plant biodiversity by assessing the contributions of 220 subtropical and tropical biomes – and the variation within – to the global functional diversity 221 currently known belowground. Specifically: 222 1. We assessed the degree of interspecific covariation in traits related to the belowground 223 collaboration and conservation axes among the three biomes. From temperate to tropical 224 ecosystems, we hypothesized that (H1) along the collaboration axis, roots shift from a 225 “do-it-yourself” to an “outsourcing” strategy because AM associations dominate in 226 tropical biomes. Along the conservation axis, towards the tropics (H1a), roots shift from 227 “slow” to “fast”, as overall environmental conditions become more favorable, or (H1b) 228 roots shift from a “fast” to “slow” strategy towards the tropics due to elevated soil-borne 229 pathogen loads. 230 2. We compared the degree of root functional diversity of temperate, subtropical, and 231 tropical biomes, and hypothesized that (H2) belowground traits in subtropical and 232 tropical biomes are more functionally diverse compared to the temperate biome. 233 3. We determined the contribution of subtropical and tropical biomes to the global root trait 234 space currently dominated by temperate root data. Because different (a)biotic constraints 235 prevail and select for distinct root trait expressions, we hypothesized that (H3) species 236 from subtropical and tropical biomes would occupy distinct spaces in, and their inclusion 237 would thereby expand the global root trait space. 238 4. Finally, we explored drivers of interspecific root trait variation within subtropical and 239 tropical biomes. We hypothesized (H4a) that interspecific root trait variation in 240 subtropical and tropical biomes is associated with different aspects of the species’ niche, 241 i.e., niche and niche breadth, suggesting that specific belowground strategies might offer 242 an overall fitness advantage across climatic and soil conditions in these biomes. Further, 243 we hypothesized (H4b) that interspecific root trait variation is shaped by biogeographic 244 drivers, particularly evolutionary variation among continents. 245

246 Species-level root trait data 247 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 10 We focused on the four key root traits that make up the conservation and collaboration axes in 248 the root economics space (Bergmann et al. 2020): mean root diameter (mm), SRL (m g-1), RTD 249 (g cm-3), and root N concentration (mg g-1). Because trait expressions and covariations generally 250 differ between woody and non-woody species (Roumet et al. 2016, Weemstra et al. 2016), we 251 further included information on plant woodiness. Root trait data and plant woodiness were 252 extracted from the Global Root Trait (GRooT) database ver.2 (Guerrero-Ramírez et al. 2021), 253 which combines root trait observations from the Fine Root Ecology Database ver.3 (FRED; 254 Iversen et al. 2021), the Plant Trait Database ver.5 (TRY; Kattge et al. 2011, 2020), data 255 mobilized by the Tropical Root Trait Initiative (TropiRoot), additional datasets and unpublished 256 data from Ghana (unpublished, S.D.Addo-Danso). Because the root economics space focus on 257 traits associated with belowground resource uptake, we excluded trait data from coarse and 258 transportive roots (identified by the data authors and data contributors as roots with diameter > 2 259 mm) that play no or only a marginal role in acquiring soil resources (McCormack et al. 2015). 260 Individual observations with an absolute error risk value (i.e., the number of mean standard 261 deviations (across all species within a trait) from the respective species means provided by 262 GRooT (Guerrero-Ramírez et al. 2021) higher than 4 were excluded before calculating species-263 level mean values. Data compilation resulted in a dataset consisting of 1035 species for which 264 data on all four core traits were available. For the individual traits, our dataset comprised 2391, 265 2031, 1747, and 1886 species for which we had data on SRL, mean root diameter, root N 266 concentration, and RTD, respectively. 267 Climates and floras across continents 268 Data from GRooT were joined with the new version of the World Checklist of Vascular Plants 269 (WCVP; Govaerts et al. 2021). Due to limited data from the tropics, we categorized the tropical 270 biome by aggregating the “montane tropical”, “seasonally dry tropical”, and “wet tropical” 271 climate categories based on the WCVP climate description. WCVP species-level data for 272 continents were aggregated, i.e., if a subtropical or tropical species occurred in “Northern 273 America” and “Southern America”, it was assigned to “America”, and if the species occurred in 274 “Asia Tropical” and “Asia Temperate”, it was assigned to as “Asia”. When subtropical or 275 tropical species were present in more than one continent, e.g., Dodonaea viscosa Jacq, occurring 276 in “Africa”, “Asia”, “Australasia”, “America”, and “Pacific”, we classified it as “Pansubtropical 277 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 11 or pantropical”, respectively. Because global root trait assessments are mostly based on 278 temperate species (Freschet et al. 2017, Bergmann et al. 2020, Carmona et al. 2021), we used the 279 temperate biome as a root trait space reference. 280 Species occurrences and climatic and soil conditions 281 To ensure that all subtropical and tropical species with available data on the four root traits were 282 included in our analyses, and to calculate their climatic and soil niches (see below), we obtained 283 species’ occurrence data (coordinates) from the Global Information Biodiversity Facility (GIBF; 284 accessed October 2022) using the ‘occ_data’ function in the rgbif package (Chamberlain and 285 Boettiger 2017). Data on (sub)tropical species’ occurrence were restricted between latitudes 286 35°N and 35°S. Data from GIBF were cleaned by (i) removing fossil specimens and managed, 287 introduced, invasive, and naturalized species records; (ii) omitting non-terrestrial coordinates; 288 and (iii) excluding coordinates based on the vicinity of country, country capitals, province 289 centroids, and biodiversity institutions (with a buffer of 2 km) to minimize the chance that 290 coordinates of e.g., planted species outside their natural distribution range would be included 291 using the CoordinateCleaner R package (Zizka et al. 2019). 292 Climatic data were obtained from Chelsa version 1.2 (1 km resolution) and included mean 293 annual temperature (°C), mean annual precipitation (mm), precipitation in wettest and driest 294 months, temperature seasonality (i.e., the standard deviation of the monthly precipitation 295 estimates expressed as a percentage of the mean of those estimates and the standard deviation of 296 the monthly mean temperatures), and maximum and minimum temperatures in warmest and 297 coldest months, respectively (Brun et al. 2022). Soil data were retrieved from the Harmonized 298 World Soil Database (1 km resolution, 0-30 cm, Appendix S1: Figure S2), (Fischer et al. 2008) 299 including variables that generally serve as indicators of soil properties resulting from abiotic and 300 biological processes acting at longer timescales (Garland et al. 2021) (Appendix S1: Table S1). 301 In addition, total soil phosphorus (P) data at a 0.5-degree resolution were retrieved from the 302 Global Gridded Soil Phosphorus Distribution Maps (Yang et al. 2014), but as these data were 303 unavailable for 20 (sub)tropical species and for 12% of the species occurrence observations (i.e., 304 44,688) and were correlated with cation exchange capacity (Pearson r = 0.66, p-value < 0.001; 305 Appendix S1: Figure S2). Thus, we did not include soil P data in our analyses. 306 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 12 Phylogenetic tree 307 To investigate the influence of evolutionary history on trait variation in subtropical and tropical 308 biomes, we built a phylogenetic tree using a dated mega-tree for vascular plants as the backbone 309 (‘GBOTB.extended.tre’; Jin and Qian 2019), which merged trees from Smith and Brown (2018) 310 and Zanne et al., (2014). We resolved missing species with the most commonly used ‘Scenario 3' 311 in the ‘V.Phylo.Maker R’ package (Jin and Qian 2019). The phylogenetic tree was visualized 312 using the ‘ggtree’ package (Yu et al. 2017, 2018, Yu 2020). 313 Data analysis 314 Interspecific (co)variation in root traits among biomes & root functional diversity 315 The multivariate covariation in root traits among species, i.e., the root trait space, within and 316 across temperate, subtropical, and tropical biomes, was visualized using a principal component 317 analysis (PCA) based on logarithmically transformed and scaled species-level mean values of all 318 four core root traits. Differences in the multivariate root trait space among biomes were tested 319 using a Permutation Multivariate Analysis of Variance (Permanova) based on 999 permutations 320 with ‘adonis2’ in the vegan package (Oksanen et al. 2019). In addition, for each of the four core 321 root traits, mean and dispersion (i.e., standard deviation, trait variation) values, as well as their 322 confidence intervals (based on 0.025 and 0.975 quantiles), were calculated across species per 323 biome using the ‘cwm’ and ‘cwd’ functions in the BAT package, assuming equal species 324 abundances (Cardoso et al. 2015) and based on 999 repetitions controlling for species richness 325 by randomly selecting 100 species from each biome. We considered biomes to differ in their root 326 traits if confidence intervals (CI) were not overlapping. 327 We calculated functional diversity by building hypervolumes (i.e., shape and volume of high-328 dimensional objects using a threshold kernel density estimate) of the first two axes of the root 329 trait PCA (Appendix S1: Table S2) (Mammola and Cardoso 2020). Specifically, we randomly 330 selected 100 species from each respective biome and built the hypervolume using the 331 ‘hypervolume_resample’ function in the hypervolume package (Blonder et al. 2018). Based on 332 each hypervolume (i.e., 100 hypervolumes for each of the three biomes), we calculated 333 functional richness (i.e., the total volume of the functional space), evenness (i.e., overlap 334 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 13 between the calculated volume and a simulated volume with even distribution of traits, with 1 as 335 perfect overlap), and dispersion (i.e., the average distance between the centroid and a sample of 336 stochastic points) using the BAT package (Cardoso et al. 2015). 337 Contribution of subtropical and tropical biomes to the global root trait space 338 To assess the proportion of subtropical and tropical functional spaces relative to the global root 339 economics space, we determined the proportion of the root trait hypervolume unique to the 340 respective biome relative to the root trait hypervolume of all other biomes combined (excluding 341 that of the biome of interest, e.g., the unique contribution of tropical species compared to the 342 functional space covered temperate and subtropical species together), using the 343 ‘hypervolume_overlap_statistics’ function (Blonder et al. 2018). Because the number of species 344 differed between each biome and the rest of the space (i.e., excluding the species from the target 345 biome), we standardized by comparing hypervolumes with the same number of species (i.e., 346 using the number of species present in the smaller hypervolume in the target biome or the rest of 347 the space as a baseline). The mean value for uniqueness and its 95% confidence intervals were 348 calculated based on 999 repetitions. 349 Drivers of subtropical and tropical root trait variation 350 We determined if subtropical and tropical root trait variation were associated with species’ 351 climate and soil niche position and niche breadth and/or variation in evolutionary histories 352 among continents. We first built species-specific climatic and soil niches (position and breadth). 353 For species niche positions, first, we carried out two PCAs using logarithmic or square-root 354 transformed (when needed to meet assumptions of normality) and scaled environmental species 355 mean values: one PCA for climatic and one PCA for soil variables in which each data point 356 represents a single species (Appendix S1: Tables S3 and S4 and Figures S3 and S4). For the 357 climatic PCA, the first axis (PC1clim) explained 59% of the total variation and was positively 358 related to MAP, precipitation in the wettest and driest months, and minimum temperature in the 359 coldest month, and is referred to as a precipitation axis (Appendix S1: Table S3 and Figure S2). 360 The second axis (PC2clim) explained an additional 25% of the variation, with positive values 361 strongly associated with higher MAT, and mean temperatures in the warmest month, i.e., a 362 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 14 temperature axis. For the soil PCA, PC1soil explained 41% of the variation, with positive values 363 predominantly associated with higher silt content and cation exchange capacity (CEC). The 364 PC2soil explained 34% of the variation, with positive values being mostly related to lower clay 365 content (Appendix S1: Table S4 and Figure S4). Species’ climatic and soil niche positions were 366 then extracted as the species’ scores on PC1 and PC2 of the climate and soil PCAs, respectively 367 (Appendix S1: Figures S3 and S4). For species-specific niche breadths, we first carried out 368 PCAs in which every point represents the environmental conditions associated with species’ 369 occurrences, i.e., multiple points for each species (Appendix S1: Figure S1), using logarithmic or 370 square-root transformed (when needed to meet assumptions of normality) data. Subsequently, we 371 built hypervolumes for each species based on the first two axes from the respective climatic and 372 soil PCAs. For niche breadth, we used the ‘hypervolume_resample’ function in the 373 ‘hypervolume’ package (Blonder et al. 2018). 374 Secondly, we calculated the phylogenetic relatedness of species between continents, i.e., the 375 evolutionary distance between the flora represented in our dataset from different continents. To 376 do that, we followed the methodology described in Taylor et al., (2023): first, we calculated 377 standardized effects sizes of phylogenetic turnover (PhyloSES) between continents using the 378 ‘phylobeta_ses’ function (Procheş et al. 2006) in the phyloregion package (Daru et al. 2020). We 379 used PhyloSES to correct for taxonomic turnover. Specifically, phylobeta_ses determines the 380 observed phylogenetic turnover (PhyloOBS) between continents based on their corresponding 381 phylogenies. It shuffles tip names in the phylogeny of each continent (1000 iterations) to create 382 null assemblages to compute the mean phylogenetic turnover (PhyloNULL_mean) and its standard 383 deviation (PhyloNULL_sd ) (Taylor et al. 2023). PhyloSES is then calculated as: 384 (1) PhyloSES = (PhyloOBS – PhyloNULL_mean) / PhyloNULL_sd. 385 Next, we created a distance matrix using PhyloSES values to carry out a Principal Coordinates 386 Analysis (PCoA; Gower 1966) and extracted the first two PCoA axes representing the 387 phylogenetic relatedness among continents, which accounts for most of the variation (Appendix 388 S1: Figures S4 and S5). For the subtropical biome, differences in phylogenetic relatedness were 389 overall associated with phylogenetic differences between Asia and Australasia, captured by the 390 first principal coordinate axis (PCoA1phyl), and phylogenetic differences between pansubtropical 391 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 15 (i.e., species with a pansubtropical distribution were considered as from another “continent”) and 392 Asia, Africa, America, and Australasia were captured by the second principal coordinate axis 393 (PCoA2phyl, Appendix S1: Figure S5). For the tropical biome, differences in phylogenetic 394 relatedness were overall associated with phylogenetic differences between Asia and America, 395 captured by the PCoA1phyl, and between pantropical and Asia, Australasia and America captured 396 by the PCoA2phyl (Appendix S1: Figure S6). 397 Finally, for each biome, we used linear models to determine if subtropical and tropical root trait 398 variation (i.e, SRL, RTD, mean root diameter, and root N concentration) are associated with 399 species’ climate and soil niche position (PC1clim + PC2clim + PC1soil + PC2soil), niche breadth 400 (Breadthclim + Breadthsoil), and species’ evolutionary histories (PCoA1phyl + PCoA2phyl). To 401 distinguish the effects of evolutionary history from other biogeographic aspects, we compared 402 models including either phylogeny (PCoA1phyl and PCoA2phyl) or continent using the Akaike 403 Information Criterion (AIC). Evidence of the potential role of evolutionary history on root trait 404 variation was associated with either the best model including phylogeny or when models 405 including either phylogeny or continents had ΔAIC < 2. All data were analyzed in R Statistical 406 Software ver. 4.1.1 (R Core Team 2023). 407

408 Species-level root trait data came from 1618 temperate, 341 subtropical, and 775 tropical 409 species, representing 1418 woody, 1167 non-woody, and 41 woody/non-woody species, i.e., 410 plants whose woodiness is mixed or convoluted (Iversen et al. 2021). Of the woody species, 528 411 (37% of the total number of woody species) occurred in the temperate biome, 243 (17%) species 412 in the subtropics, and 647 (46%) in tropical biomes. We first found that among the non-woody 413 species, the vast majority (1007 species; 86% of all non-woody species) were temperate species, 414 and only 73 (6%) and 87 (7%) species occurred in the subtropical and tropical biomes, 415 respectively (Appendix S1: Figure 8). Owing to these data limitations on non-woody subtropical 416 and tropical species (which largely represent the global coverage) and the potential of 417 confounding effects when evaluating belowground functional biogeographic patterns, we 418 restricted our results exclusively to the woody species, hereafter: ‘species’. 419 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 16 Interspecific (co)variation in root traits among biomes and root functional diversity 420 The covariation in root traits among species, i.e., multivariate root trait space, was captured by 421 two main axes. The first component (PC1), recognized as the collaboration axis, explained 45% 422 of the total interspecific variation in root traits, with SRL negatively and mean root diameter 423 positively loading on PC1 (Figure 1). The second component (PC2), recognized as the 424 conservation axis, explained an additional 36 % of the overall variation, with RTD negatively 425 and root N concentration positively loading on PC2 (Figure 1). There was evidence for a shift in 426 the collaboration axis (SRL and diameter) from do-it-yourself (high SRL) to an outsourcing 427 strategy (high RD) from temperate to tropical biomes (Permanova: F= 33.3, p<0.001; Figure 1). 428 Examining root trait variation for traits separately, we found evidence that temperate species had 429 higher SRL and lower mean root diameter than subtropical and tropical species, but subtropical 430 and tropical species were indistinguishable from each other (Figure 2, Table 1). Root N 431 concentration was higher for tropical than subtropical species, with intermediate values for 432 temperate species. Temperate species had lower RTD than subtropical and tropical species, while 433 subtropical and tropical species had similar RTD. Overall, dispersion in the four individual traits 434 did not vary significantly among biomes, except for RTD, which was less dispersed in tropical 435 species than in temperate and subtropical species (Table 1). Further, there were no significant 436 differences in functional richness, evenness, and dispersion among biomes when considering the 437 conservation and collaboration axes together and controlling by species richness (Table 2). 438 Contribution of subtropical and tropical biomes to the global root trait space 439 We determined the unique root trait functional space of subtropical and tropical species relative 440 to the global root trait space (Figure 3). Together, subtropical and tropical biomes accounted for 441 at least 40% of the unique functional space. In other words, of the multidimensional root trait 442 space across all three biomes, ~37% is shared among biomes, 23% (CI: 15 – 30%) is unique to 443 temperate, 23% to tropical (CI: 17 – 30%), and 17% (CI: 10 – 23%) to subtropical biomes. 444 Drivers of subtropical and tropical root trait variation 445 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 17 The inclusion of evolutionary history either improved or performed similarly to subtropical 446 models including continents, suggesting a role of evolutionary history in root trait variation 447 (Appendix S1: Table S5). Subtropical models explained 27, 9, 4, and 56% variance for SRL, 448 diameter, root N, and RTD respectively (Table 3). For instance, higher SRL values were 449 associated with higher climatic niche breadth (Breadthclim) and species linked to environments 450 with lower precipitation. Further, species from phylogenetic-related regions were similar in their 451 SRL (e.g., moving from higher values of SRL in Australasia species and lower species 452 distributed across continents, i.e., pansubtropical species, Figure 4). Higher RTD values were 453 associated with species linked to environments with higher precipitation and were also similar 454 among phylogenetic-related regions, with higher values for pansubtropical species and lower 455 values in Australasia species (Figure 4). For the tropics, we found evidence of a potential role of 456 evolutionary history for diameter and root N (Appendix S1: Figure S9), with differences for SRL 457 and RTD associated with other biogeographic aspects captured by continents and not to 458 evolutionary history per se (Appendix S1: Figure S10). Yet, the full models, including species 459 niches and biogeographic/evolutionary drivers explained 10% or less variation (Table 3). 460

461 Subtropical and tropical biomes are major contributors to global plant biodiversity (Cai et al. 462 2023). This extensive taxonomic diversity is associated with widespread functional diversity 463 aboveground but also likely largely belowground, owing to the vast variability in belowground 464 strategies of plants to acquire a multitude of resources from a heterogeneous and complex soil 465 matrix (Weemstra et al. 2016). Our study shows how tropical woody species have different root 466 traits compared to woody species in the temperate biome (e.g., roots are overall thicker, less 467 dense, and higher in N concentrations). Accounting for their unique combinations of root traits, 468 subtropical and tropical species substantially expanded the global belowground functional space. 469 Even though biomes have similar functional diversity and partially overlap in root functional 470 space, (sub)tropical species seem to employ distinct belowground strategies that have so far not 471 been observed or are rare in the temperate biome. 472 Root trait expressions differ between (sub)tropical and temperate species 473 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 18 Tropical and subtropical species both had significantly lower SRL and larger root diameters than 474 temperate species, supporting the expected shift along the collaboration axis (from do-it-yourself 475 to outsourcing) between biomes (Freschet et al. 2017, Ma et al. 2018). Investing in thick roots 476 that typically have large cortices (Kong et al. 2014) and that permit high AM colonization rates 477 (Brundrett 2002, Ma et al. 2018) may be a particularly relevant foraging strategy in tropical soils 478 (Valverde-Barrantes et al. 2021) where soil P is generally limiting (Reich and Oleksyn 2004, 479 Vitousek 2004) and can be more efficiently acquired by AM fungi than soil N (Smith and Read 480 2008), generally limiting in temperate soils. The overall thicker roots in the (sub)tropics may 481 further be explained by the greater abundance of more basal superorders in the tropics e.g., 482 Magnoliids, that rely heavily on AM fungi (Valverde-Barrantes et al. 2015, 2016) than recently 483 derived superorders at higher latitudes, such as Rosids and Asterids, that have thinner roots and 484 are less dependent on AM symbiosis (Comas et al. 2012, Ma et al. 2018). However, tropical 485 forests often are still dominated by Rosid (Euphorbiaceae, Moraceae, Fabaceae, etc) and Asterid 486 (Sapotaceae, Rubiaceae, Lecythidaceae) rather than Magnoliid (except Lauraceae) families. 487 Therefore, greater dominance of basal clades may explain only a small part of the shift towards 488 the outsourcing strategy in (sub)tropical species. Further, in tropical forests, large root trait 489 diversification still occurs within families, while traits (e.g., root diameter) converge across 490 families (Weemstra et al. 2023). 491 Tropical species had, overall, both higher root N concentration and higher RTD than temperate 492 tree species, which contrasts previous work (e.g., Gu et al., (2014), who found lower RTD in 493 tropical than temperate tree species) and corroborates others reporting higher root N in the 494 tropics, but including non-woody species (Freschet et al. 2017, Ma et al. 2018). These results call 495 for careful interpretation of shifts in individual traits. For example, we observed a tradeoff 496 between RTD and root N corresponding to the conservation gradient across all species combined 497 and when considering biomes separately. However, a higher root N in tropical species does not 498 necessarily imply a lower RTD, nor can it be interpreted as a “faster” belowground strategy 499 relative to temperate species if RTD simultaneously increases too. Improving our understanding 500 of plant belowground functioning requires the establishment of stronger linkages among traits 501 (e.g., RTD and root N), between traits and functions (e.g., root N and resource uptake) and their 502 mechanistic underpinnings. 503 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 19 From a resource economics perspective, an increase in RTD towards the tropics suggests that 504 low root turnover rates would be advantageous in environments with tropical climates to 505 maximize the return on C investment over the long-life span of the roots. A higher RTD in 506 warmer climates was also observed, mainly in temperate forests, by Laughlin et al., (2021), 507 suggesting that this pattern is consistent within and across biomes and it may be associated with 508 species’ tolerance to freezing conditions. Species that successfully migrated to higher latitudes 509 where freezing events occurred are - among others - characterized by producing cheap 510 aboveground tissue with annual turnover (Zanne et al. 2014), and a similar evolutionary 511 trajectory may have occurred belowground where temperate (and boreal) species produce cheap, 512 low-RTD roots to minimize resource losses as roots are shed during winter. Currently, the 513 assumed resource economics tradeoff between RTD and root lifespan is not or weakly supported 514 by (only few) empirical data across temperate tree species (Withington et al. 2006, McCormack 515 et al. 2012), but it may be more pronounced in (sub)tropical biomes where the lack of freezing 516 events may permit longer root lifespans and greater returns of investing in more expensive, dense 517 root tissue. Biotic drivers may further explain the denser roots of tropical species, as high RTD 518 offer structural and/or chemical protection to roots against herbivores and soilborne pathogen 519 loads (Xia et al. 2021) that are generally higher in tropical soils than at higher latitudes 520 (Delgado-Baquerizo et al. 2020). 521 (Sub)tropical biomes add unique belowground strategies to the global space 522 Mapping the observed multivariate root trait combinations in a global functional space firstly 523 revealed substantial overlap (37%) between species across biomes. These results suggest that at a 524 broad scale, common belowground trait syndromes - made up of these four traits - occur 525 regardless of climate and bioregion, owing to physiological (e.g., resource economics) and 526 evolutionary constraints that plants across bioregions are subjected to (Reich et al. 2003). 527 Despite these common pressures, however, adding the root traits from species in subtropical and 528 tropical biomes extended the global root trait space, so that (sub)tropical species occupied at 529 least 40% of the unique multivariate trait combinations within this space. Our analyses accounted 530 for different numbers of species across biomes, so that the unique belowground traits space of 531 (sub)tropical species cannot be attributed to larger species numbers. Furthermore, the degree of 532 functional diversity was similar across biomes, suggesting that different but not more 533 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 20 functionally rich trait combinations were observed among (sub)tropical species. Possibly, 534 (sub)tropical species are subjected to largely different abiotic and biotic environmental 535 conditions selecting for different root trait combinations unadaptive in other environments, that 536 are further discussed in the following section. The unique contributions of (sub)tropical species 537 to the global belowground functional space, nonetheless, illustrates that a large part of our 538 understanding of belowground resource strategies has so far remained elusive, and that 539 accounting for (sub)tropical species (of which our study still covers only a fraction) is pertinent 540 to obtaining a more complete picture of how trees function belowground in a variety of 541 ecosystems. 542 Evolutionary history and species’ niches explain root trait variation in subtropical biome 543 Evolutionary history and species niches explained root trait variation among subtropical species 544 for SRL and RTD, suggesting that biogeographic imprints on SRL and RTD in the subtropics are 545 associated with evolutionary history. In other words, subtropical species from continents that are 546 more similar phylogenetically, tend to have similar SRL and RTD values, suggesting both the 547 role of evolutionary history and phylogenetic conservatism in root trait variation. In contrast, 548 there was only a weak or no detectable role of evolutionary history for the other traits in the 549 subtropics and for any of the four traits in the tropics. These findings support weak phylogenetic 550 effects on root trait variation previously reported across 218 neotropical tree species that may be 551 ascribed to large trait diversification and niche differentiation within tropical plant families (but 552 see: Pierick et al. 2021, 2023, Weemstra et al. 2023). Together, our results point toward 553 phylogenetic conservatism being context-dependent, varying across biogeographical, 554 phylogenetic scales (Weemstra et al. 2023) and across traits describing both the collaboration 555 and conservation axes. For example, we found evidence of the influence of evolutionary history 556 for RTD but not for root N concentration in the subtropics (and similarly for SRL and not for 557 diameter). Interestingly, lower values of SRL and higher values of RTD were associated with 558 species with cosmopolitan biogeographic distributions (i.e., pansubtropical species), likely 559 conferring the ability to explore the soil matrix via mycorrhiza and exhibit dense roots, with 560 these strategies needed for complex mutualistic- and antagonistic-biosynthesis costs (Xia et al. 561 2021). 562 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 21 Determining whether the lower explanatory power of the models explaining root trait variation in 563 the tropics is related to ecological and evolutionary reasons or data availability remains 564 challenging. Regional studies have highlighted that the role of soil niches in shaping root 565 strategies is nutrient-dependent, with weak significant relationships between SRL and species 566 niches based on soil calcium but not for other soil N, P and potassium (K) in the Amazon 567 (Vleminckx et al. 2023). In addition, biotic factors like mycorrhizal interactions and soilborne 568 pathogen loads and herbivory may have a bigger effect on root selection in the (sub)tropics, 569 while environmental conditions such as freezing temperatures seem more important selecting 570 traits in the temperate biome. While our findings support the growing evidence suggesting a 571 weak role of overall soil conditions (at least at larger scales) and phylogeny explaining root trait 572 variation in the tropics, this needs to be interpreted with caution because of data shortfalls. 573 Moving forward belowground functional biogeography 574 Based on our empirical cross-continental findings, we propose to broaden our view of 575 (belowground) biogeography by including different ecosystems, traits, and species, specifically: 576 • Underrepresented (sub)tropical ecosystems: within the tropics, root data from certain, 577 adverse ecosystems (like savannas, dry forests, and high montane ecosystems) are even 578 more scarce than from e.g., moist tropical forests. Under their unfavorable environmental 579 conditions, even more distinct belowground adaptation may be observed and the global 580 belowground traits space further expanded. 581 ● Additional traits: (sub)tropical species may diverge in traits outside the four core traits 582 proposed (Bergmann et al. 2020), depending on their environment. For example, in 583 tropical, P-limited soils, root exudation rates (Dallstream et al. 2022), phosphatase 584 activity (Guilbeault-Mayers and Laliberté n.d.), branching (Yan et al. 2022a, Weemstra et 585 al. 2023), and mycorrhizal traits may be more relevant than e.g., or SRL. Broadening our 586 scope beyond the temperate zone might require measuring different traits functionally 587 relevant in other biomes. 588 ● Non-woody species: herbaceous species are well represented in tropical and subtropical 589 biomes (24-34% herbaceous species; Taylor et al. 2023), but are largely absent in root 590 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 22 trait databases and, subsequently, global analyses and resulting conceptual frameworks. 591 Their poor coverage likely biases our knowledge of ecological and evolutionary drivers 592 behind species-level root trait variation at global scales (Ma et al. 2018), with potentially 593 misleading assessments on the role of families shaping root trait variation across biomes 594 (Carmona et al. 2021). For example, some families, such as Euphorbiaceae or Fabaceae, 595 are largely sampled as trees in the tropics but as herbs in temperate areas, making it 596 difficult to distinguish the effects of plant growth habits from ecological adaptations to 597 different biomes on belowground trait expressions and variability (Valverde‐Barrantes et 598 al. 2020). These biases need to be considered when interpreting other global analyses and 599 addressed in sampling and data mobilization efforts. 600 ● Environmental information at a relevant spatial scale: Owing to the large heterogeneity 601 of soils at small spatial scales (Ettema and Wardle 2002), root traits are likely to respond 602 to small-scale soil environmental variation (Weemstra and Valverde-Barrantes 2022, 603 Pierick et al. n.d.). The use of global soil dataset with a low spatial resolution in global 604 analyses, as ours, may therefore incur a mismatch between the scale at which soil 605 variables vary and roots respond and lead to weak relationships between edaphic factors 606 and root trait variation. Accurately linking the soil environment to root trait expressions 607 and predicting how plants respond and adapt to soil environmental variation may 608 therefore require more small-scale (in situ) data collection prior when, for instance, 609 aiming to disentangle drivers behind larger scale patterns. 610

611 Our findings illustrate the unique contribution of subtropical and tropical species to the global 612 root economics space, expanding it with belowground strategies that are either rare or absent in 613 the temperate biome and strengthening the foundations of functional belowground biogeography. 614 This study significantly improves our understanding of global belowground strategies using a 615 cross-continental, cross-biome representation of underrepresented species and ecosystems and by 616 elucidating key shortfalls that need to be considered in global analyses. As such, our work 617 highlights the importance of including (sub)tropical species in conceptual models of root 618 functional diversity to develop a more complete view of the various belowground strategies that 619 underlie plant functional biogeography and biodiversity globally. 620 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 23 Author contributions 621 All authors conceived and developed the idea as part of the Belowground Functional 622 Biogeography theme of the Tropical Root Trait Initiative. N G-R compiled the data and 623 developed the analyses with the input of MW, MNU, and CF. N G-R and MW led the 624 development and writing of the manuscript, with substantial input from all authors. 625 Competing interests 626 The authors declare no conflicts of interest. 627 Acknowledgments 628 This study emerged as a collaborative effort of scientists interested in addressing questions about 629 tropical root ecology and evolution: the Tropical Root Trait Initiative (TropiRoot). We thank New 630 Phytologist for supporting this Initiative and our work through its 28 th Workshop: Coordinating 631 and Synthesizing Tropical Forest Root Trait Studies: Understanding Belowground NPP, Root 632 Responses to Global Change, and Nutrient Acquisition Dynamics across tropical forests, held at 633 the Smithsonian Tropical Research Institute in Panama. We thank Sarah Sophie Weil and Kevin 634 Mganga as friendly reviewers and Amanda Taylor for the code to calculate the phylogenetic 635 turnover. NR.G-R thanks the Dorothea Schlözer Postdoctoral Programme of the Georg -August-636 Universität Göttingen and the DFG, grant number 316045089/GRK 2300 for their support. DF.C 637 participation was supported by US Department of Energy (DOE) Office of Science Early Career 638 Award DE – SC0015898, and US National Science Foundation (NSF) Division of Environmental 639 Biology (DEB) Long Term Research in Environmental Biology (LTREB) award #2332006. LFL 640 acknowledges the Bavarian State Chancellery (Project Amazon-FLUX). 641

642 Addo-Danso, S. D., C. E. Defrenne, M. L. McCormack, I. Ostonen, A. Addo-Danso, E. G. Foli, 643 K. A. Borden, M. E. Isaac, and C. E. Prescott. 2020. Fine-root morphological trait 644 variation in tropical forest ecosystems: an evidence synthesis. Plant Ecology 221:1–13. 645 Aerts, R., and F. S. Chapin III. 2000. The mineral nutrition of wild plants revisited: A re-646 evaluation of processes and patterns. Advances in Ecological Research 30:1–67. 647 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 24 Antonelli, A. 2021. The rise and fall of Neotropical biodiversity. Botanical Journal of the 648 Linnean Society. 649 Bergmann, J., A. Weigelt, F. van der Plas, D. C. Laughlin, T. W. Kuyper, N. Guerrero-Ramirez, 650 O. J. Valverde-Barrantes, H. Bruelheide, G. T. Freschet, C. M. Iversen, J. Kattge, M. L. 651 McCormack, I. C. Meier, M. C. Rillig, C. Roumet, M. Semchenko, C. J. Sweeney, J. van 652 Ruijven, L. M. York, and L. Mommer. 2020. The fungal collaboration gradient 653 dominates the root economics space in plants. Science Advances 6:eaba3756. 654 Blonder, B., C. B. Morrow, B. Maitner, D. J. Harris, C. Lamanna, C. Violle, B. J. Enquist, and 655 A. J. Kerkhoff. 2018. New approaches for delineating n-dimensional hypervolumes. 656

in Ecology and Evolution 9:305–319. 657 Bloom, A. J., F. S. Chapin, III, and H. A. Mooney. 1985. Resource Limitation in Plants--An 658 Economic Analogy. Annual Review of Ecology and Systematics 16:363–392. 659 Boonman, C. C. F., F. van Langevelde, I. Oliveras, J. Couédon, N. Luijken, D. Martini, and E. 660 M. Veenendaal. 2020. On the importance of root traits in seedlings of tropical tree 661 species. New Phytologist 227:156–167. 662 Brun, P., N. E. Zimmermann, C. Hari, L. Pellissier, and D. N. Karger. 2022. Global climate-663 related predictors at kilometre resolution for the past and future. Earth System Science 664 Data Discussions:1–44. 665 Brundrett, M. C. 2002. Coevolution of roots and mycorrhizas of land plants. New Phytologist 666 154:275–304. 667 Cai, L., H. Kreft, A. Taylor, P. Denelle, J. Schrader, F. Essl, M. van Kleunen, J. Pergl, P. Pyšek, 668 A. Stein, M. Winter, J. F. Barcelona, N. Fuentes, Inderjit, D. N. Karger, J. Kartesz, A. 669 Kuprijanov, M. Nishino, D. Nickrent, A. Nowak, A. Patzelt, P. B. Pelser, P. Singh, J. J. 670 Wieringa, and P. Weigelt. 2023. Global models and predictions of plant diversity based 671 on advanced machine learning techniques. New Phytologist 237:1432–1445. 672 Cardoso, P., F. Rigal, and J. C. Carvalho. 2015. BAT – Biodiversity Assessment Tools, an R 673 package for the measurement and estimation of alpha and beta taxon, phylogenetic and 674 functional diversity. Methods in Ecology and Evolution 6:232–236. 675 Carmona, C. P., C. G. Bueno, A. Toussaint, S. Träger, S. Díaz, M. Moora, A. D. Munson, M. 676 Pärtel, M. Zobel, and R. Tamme. 2021. Fine-root traits in the global spectrum of plant 677 form and function. Nature 597:683–687. 678 Carta, A., L. Peruzzi, and S. Ramírez-Barahona. 2022. A global phylogenetic regionalization of 679 vascular plants reveals a deep split between Gondwanan and Laurasian biotas. New 680 Phytologist 233:1494–1504. 681 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 25 Chamberlain, S. A., and C. Boettiger. 2017. R Python, and Ruby clients for GBIF species 682 occurrence data. PeerJ Inc. 683 Comas, L. H., K. E. Mueller, L. L. Taylor, P. E. Midford, H. S. Callahan, and D. J. Beerling. 684 2012. Evolutionary Patterns and Biogeochemical Significance of Angiosperm Root 685 Traits. International Journal of Plant Sciences 173:584–595. 686 Cusack, D. F., S. D. Addo-Danso, E. A. Agee, K. M. Andersen, M. Arnaud, S. A. Batterman, F. 687 Q. Brearley, M. I. Ciochina, A. L. Cordeiro, C. Dallstream, M. H. Diaz-Toribio, L. H. 688 Dietterich, J. B. Fisher, K. Fleischer, C. Fortunel, L. Fuchslueger, N. R. Guerrero-689 Ramírez, M. M. Kotowska, L. F. Lugli, C. Marín, L. A. McCulloch, J.-L. Maeght, D. 690 Metcalfe, R. J. Norby, R. S. Oliveira, J. S. Powers, T. Reichert, S. W. Smith, C. M. 691 Smith-Martin, F. M. Soper, L. Toro, M. N. Umaña, O. Valverde-Barrantes, M. 692 Weemstra, L. K. Werden, M. Wong, C. L. Wright, S. J. Wright, and D. Yaffar. 2021. 693 Tradeoffs and synergies in tropical forest root traits and dynamics for nutrient and water 694 acquisition: field and modeling advances. Frontiers in Forests and Global Change 695 4:704469. 696 Cusack, D. F., B. Christoffersen, C. M. Smith-Martin, K. M. Andersen, A. L. Cordeiro, K. 697 Fleischer, S. J. Wright, N. R. Guerrero-Ramírez, L. F. Lugli, L. A. McCulloch, M. 698 Sanchez-Julia, S. A. Batterman, C. Dallstream, C. Fortunel, L. Toro, L. Fuchslueger, M. 699 Y. Wong, D. Yaffar, J. B. Fisher, M. Arnaud, L. H. Dietterich, S. D. Addo-Danso, O. J. 700 Valverde-Barrantes, M. Weemstra, J. C. Ng, and R. J. Norby. 2024. Toward a 701 coordinated understanding of hydro-biogeochemical root functions in tropical forests for 702 application in vegetation models. New Phytologist 242:351–371. 703 Dallstream, C., M. Weemstra, and F. M. Soper. 2022. A framework for fine‐root trait syndromes: 704 syndrome coexistence may support phosphorus partitioning in tropical forests. 705 Oikos:e08908. 706 Daru, B. H., P. Karunarathne, and K. Schliep. 2020. phyloregion: R package for biogeographical 707 regionalization and macroecology. Methods in Ecology and Evolution 11:1483–1491. 708 Delgado-Baquerizo, M., C. A. Guerra, C. Cano-Díaz, E. Egidi, J.-T. Wang, N. Eisenhauer, B. K. 709 Singh, and F. T. Maestre. 2020. The proportion of soil-borne pathogens increases with 710 warming at the global scale. Nature Climate Change 10:550–554. 711 Echeverría-Londoño, S., B. J. Enquist, D. M. Neves, C. Violle, B. Boyle, N. J. B. Kraft, B. S. 712 Maitner, B. McGill, R. K. Peet, B. Sandel, S. A. Smith, J.-C. Svenning, S. K. Wiser, and 713 A. J. Kerkhoff. 2018. Plant Functional Diversity and the Biogeography of Biomes in 714 North and South America. Frontiers in Ecology and Evolution 6. 715 Eiserhardt, W. L., T. L. P. Couvreur, and W. J. Baker. 2017. Plant phylogeny as a window on the 716 evolution of hyperdiversity in the tropical rainforest biome. New Phytologist 214:1408–717 1422. 718 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 26 Ettema, C. H., and D. A. Wardle. 2002. Spatial soil ecology. Trend in Ecology & Evolution 719 17:177–183. 720 Fischer, G., F. Nachtergaele, S. Prieler, H. T. van Velthuizen, L. Verelst, and D. Wiberg. 2008. 721 Global Agro-ecological Zones Assessment for Agriculture (GAEZ 2008). IIASA, 722 Laxenburg, Austria and FAO, Rome, Italy. 723 Freschet, G. T., O. J. Valverde-Barrantes, C. M. Tucker, J. M. Craine, M. L. McCormack, C. 724 Violle, F. Fort, C. B. Blackwood, K. R. Urban-Mead, C. M. Iversen, A. Bonis, L. H. 725 Comas, J. H. C. Cornelissen, M. Dong, D. Guo, S. E. Hobbie, R. J. Holdaway, S. W. 726 Kembel, N. Makita, V. G. Onipchenko, C. Pico-Cochard, P. B. Reich, E. G. de la Riva, S. 727 W. Smith, N. A. Soudzilovskaia, M. G. Tjoelker, D. A. Wardle, and C. Roumet. 2017. 728 Climate, soil and plant functional types as drivers of global fine-root trait variation. 729 Journal of Ecology 105:1182–1196. 730 Fujii, K., M. Shibata, K. Kitajima, T. Ichie, K. Kitayama, and B. L. Turner. 2018. Plant–soil 731 interactions maintain biodiversity and functions of tropical forest ecosystems. Ecological 732 Research 33:149–160. 733 Garland, G., S. Banerjee, A. Edlinger, E. M. Oliveira, C. Herzog, R. Wittwer, L. Philippot, F. T. 734 Maestre, and M. G. A. van der Heijden. 2021. A closer look at the functions behind 735 ecosystem multifunctionality: A review. Journal of Ecology 109:600–613. 736 Govaerts, R., E. Nic Lughadha, N. Black, R. Turner, and A. Paton. 2021. The World Checklist of 737 Vascular Plants, a continuously updated resource for exploring global plant diversity. 738 Scientific Data 8:215. 739 Gower, J. C. 1966. Some distance properties of latent root and vector methods used in 740 multivariate analysis. Biometrika 53:325–338. 741 Grime, J. P. 1977. Evidence for the existence of three primary strategies in plants and its 742 relevance to ecological and evolutionary theory. The American Naturalist 111:1169–743 1194. 744 Gu, J., Y. Xu, X. Dong, H. Wang, and Z. Wang. 2014. Root diameter variations explained by 745 anatomy and phylogeny of 50 tropical and temperate tree species. Tree Physiology 746 34:415–425. 747 Guerrero-Ramírez, N. R., L. Mommer, G. T. Freschet, C. M. Iversen, M. L. McCormack, J. 748 Kattge, H. Poorter, F. van der Plas, J. Bergmann, T. W. Kuyper, L. M. York, H. 749 Bruelheide, D. C. Laughlin, I. C. Meier, C. Roumet, M. Semchenko, C. J. Sweeney, J. 750 van Ruijven, O. J. Valverde-Barrantes, I. Aubin, J. A. Catford, P. Manning, A. Martin, R. 751 Milla, V. Minden, J. G. Pausas, S. W. Smith, N. A. Soudzilovskaia, C. Ammer, B. 752 Butterfield, J. Craine, J. H. C. Cornelissen, F. T. de Vries, M. E. Isaac, K. Kramer, C. 753 König, E. G. Lamb, V. G. Onipchenko, J. Peñuelas, P. B. Reich, M. C. Rillig, L. Sack, B. 754 Shipley, L. Tedersoo, F. Valladares, P. van Bodegom, P. Weigelt, J. P. Wright, and A. 755 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 27 Weigelt. 2021. Global root traits (GRooT) database. Global Ecology and Biogeography 756 30:25–37. 757 Guilbeault-Mayers, X., and E. Laliberté. (n.d.). Root phosphatase activity is coordinated with the 758 root conservation gradient across a phosphorus gradient in a lowland tropical forest. New 759 Phytologist n/a. 760 Iversen, C. M., A. S. Powell, M. L. McCormack, C. B. Blackwood, G. T. Freschet, J. Kattge, C. 761 Roumet, D. B. Stover, N. A. Soudzilovskaia, O. Valverde-Barrantes, P. M. van 762 Bodegom, and C. Violle. 2021. Fine-Root Ecology Database (FRED): A Global 763 Collection of Root Trait Data with Coincident Site, Vegetation, Edaphic, and Climatic 764 Data. Oak Ridge National Laboratory, TES SFA, U.S. Department of Energy, Oak Ridge, 765 Tennessee, U.S.A. 766 Jin, Y., and H. Qian. 2019. V.PhyloMaker: an R package that can generate very large 767 phylogenies for vascular plants. Ecography 42:1353–1359. 768 Kattge, J., G. Bönisch, S. Díaz, S. Lavorel, I. C. Prentice, P. Leadley, S. Tautenhahn, G. D. A. 769 Werner, T. Aakala, M. Abedi, A. T. R. Acosta, G. C. Adamidis, K. Adamson, M. Aiba, 770 C. H. Albert, J. M. Alcántara, C. A. C, I. Aleixo, H. Ali, B. Amiaud, C. Ammer, M. M. 771 Amoroso, M. Anand, C. Anderson, N. Anten, J. Antos, D. M. G. Apgaua, T.-L. Ashman, 772 D. H. Asmara, G. P. Asner, M. Aspinwall, O. Atkin, I. Aubin, L. Baastrup‐Spohr, K. 773 Bahalkeh, M. Bahn, T. Baker, W. J. Baker, J. P. Bakker, D. Baldocchi, J. Baltzer, A. 774 Banerjee, A. Baranger, J. Barlow, D. R. Barneche, Z. Baruch, D. Bastianelli, J. Battles, 775 W. Bauerle, M. Bauters, E. Bazzato, M. Beckmann, H. Beeckman, C. Beierkuhnlein, R. 776 Bekker, G. Belfry, M. Belluau, M. Beloiu, R. Benavides, L. Benomar, M. L. Berdugo‐777 Lattke, E. Berenguer, R. Bergamin, J. Bergmann, M. B. Carlucci, L. Berner, M. 778 Bernhardt‐Römermann, C. Bigler, A. D. Bjorkman, C. Blackman, C. Blanco, B. Blonder, 779 D. Blumenthal, K. T. Bocanegra‐González, P. Boeckx, S. Bohlman, K. Böhning‐Gaese, 780 L. Boisvert‐Marsh, W. Bond, B. Bond‐Lamberty, A. Boom, C. C. F. Boonman, K. 781 Bordin, E. H. Boughton, V. Boukili, D. M. J. S. Bowman, S. Bravo, M. R. Brendel, M. R. 782 Broadley, K. A. Brown, H. Bruelheide, F. Brumnich, H. H. Bruun, D. Bruy, S. W. 783 Buchanan, S. F. Bucher, N. Buchmann, R. Buitenwerf, D. E. Bunker, J. Bürger, S. 784 Burrascano, D. F. R. P. Burslem, B. J. Butterfield, C. Byun, M. Marques, M. C. Scalon, 785 M. Caccianiga, M. Cadotte, M. Cailleret, J. Camac, J. J. Camarero, C. Campany, G. 786 Campetella, J. A. Campos, L. Cano‐Arboleda, R. Canullo, M. Carbognani, F. Carvalho, 787 F. Casanoves, B. Castagneyrol, J. A. Catford, J. Cavender‐Bares, B. E. L. Cerabolini, M. 788 Cervellini, E. Chacón‐Madrigal, K. Chapin, F. S. Chapin, S. Chelli, S.-C. Chen, A. Chen, 789 P. Cherubini, F. Chianucci, B. Choat, K.-S. Chung, M. Chytrý, D. Ciccarelli, L. Coll, C. 790 G. Collins, L. Conti, D. Coomes, J. H. C. Cornelissen, W. K. Cornwell, P. Corona, M. 791 Coyea, J. Craine, D. Craven, J. P. G. M. Cromsigt, A. Csecserits, K. Cufar, M. Cuntz, A. 792 C. da Silva, K. M. Dahlin, M. Dainese, I. Dalke, M. D. Fratte, A. T. Dang‐Le, J. 793 Danihelka, M. Dannoura, S. Dawson, A. J. de Beer, A. D. Frutos, J. R. D. Long, B. 794 Dechant, S. Delagrange, N. Delpierre, G. Derroire, A. S. Dias, M. H. Diaz‐Toribio, P. G. 795 Dimitrakopoulos, M. Dobrowolski, D. Doktor, P. Dřevojan, N. Dong, J. Dransfield, S. 796 Dressler, L. Duarte, E. Ducouret, S. Dullinger, W. Durka, R. Duursma, O. Dymova, A. E‐797 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 28 Vojtkó, R. L. Eckstein, H. Ejtehadi, J. Elser, T. Emilio, K. Engemann, M. B. Erfanian, A. 798 Erfmeier, A. Esquivel‐Muelbert, G. Esser, M. Estiarte, T. F. Domingues, W. F. Fagan, J. 799 Fagúndez, D. S. Falster, Y. Fan, J. Fang, E. Farris, F. Fazlioglu, Y. Feng, F. Fernandez‐800 Mendez, C. Ferrara, J. Ferreira, A. Fidelis, B. Finegan, J. Firn, T. J. Flowers, D. F. B. 801 Flynn, V. Fontana, E. Forey, C. Forgiarini, L. François, M. Frangipani, D. Frank, C. 802 Frenette‐Dussault, G. T. Freschet, E. L. Fry, N. M. Fyllas, G. G. Mazzochini, S. Gachet, 803 R. Gallagher, G. Ganade, F. Ganga, P. García‐Palacios, V. Gargaglione, E. Garnier, J. L. 804 Garrido, A. L. de Gasper, G. Gea‐Izquierdo, D. Gibson, A. N. Gillison, A. Giroldo, M.-C. 805 Glasenhardt, S. Gleason, M. Gliesch, E. Goldberg, B. Göldel, E. Gonzalez‐Akre, J. L. 806 Gonzalez‐Andujar, A. González‐Melo, A. González‐Robles, B. J. Graae, E. Granda, S. 807 Graves, W. A. Green, T. Gregor, N. Gross, G. R. Guerin, A. Günther, A. G. Gutiérrez, L. 808 Haddock, A. Haines, J. Hall, A. Hambuckers, W. Han, S. P. Harrison, W. Hattingh, J. E. 809 Hawes, T. He, P. He, J. M. Heberling, A. Helm, S. Hempel, J. Hentschel, B. Hérault, A.-810 M. Hereş, K. Herz, M. Heuertz, T. Hickler, P. Hietz, P. Higuchi, A. L. Hipp, A. Hirons, 811 M. Hock, J. A. Hogan, K. Holl, O. Honnay, D. Hornstein, E. Hou, N. Hough‐Snee, K. A. 812 Hovstad, T. Ichie, B. Igić, E. Illa, M. Isaac, M. Ishihara, L. Ivanov, L. Ivanova, C. M. 813 Iversen, J. Izquierdo, R. B. Jackson, B. Jackson, H. Jactel, A. M. Jagodzinski, U. Jandt, S. 814 Jansen, T. Jenkins, A. Jentsch, J. R. P. Jespersen, G.-F. Jiang, J. L. Johansen, D. Johnson, 815 E. J. Jokela, C. A. Joly, G. J. Jordan, G. S. Joseph, D. Junaedi, R. R. Junker, E. Justes, R. 816 Kabzems, J. Kane, Z. Kaplan, T. Kattenborn, L. Kavelenova, E. Kearsley, A. Kempel, T. 817 Kenzo, A. Kerkhoff, M. I. Khalil, N. L. Kinlock, W. D. Kissling, K. Kitajima, T. 818 Kitzberger, R. Kjøller, T. Klein, M. Kleyer, J. Klimešová, J. Klipel, B. Kloeppel, S. 819 Klotz, J. M. H. Knops, T. Kohyama, F. Koike, J. Kollmann, B. Komac, K. Komatsu, C. 820 König, N. J. B. Kraft, K. Kramer, H. Kreft, I. Kühn, D. Kumarathunge, J. Kuppler, H. 821 Kurokawa, Y. Kurosawa, S. Kuyah, J.-P. Laclau, B. Lafleur, E. Lallai, E. Lamb, A. 822 Lamprecht, D. J. Larkin, D. Laughlin, Y. L. Bagousse‐Pinguet, G. le Maire, P. C. le 823 Roux, E. le Roux, T. Lee, F. Lens, S. L. Lewis, B. Lhotsky, Y. Li, X. Li, J. W. Lichstein, 824 M. Liebergesell, J. Y. Lim, Y.-S. Lin, J. C. Linares, C. Liu, D. Liu, U. Liu, S. 825 Livingstone, J. Llusià, M. Lohbeck, Á. López‐García, G. Lopez‐Gonzalez, Z. Lososová, 826 F. Louault, B. A. Lukács, P. Lukeš, Y. Luo, M. Lussu, S. Ma, C. M. R. Pereira, M. Mack, 827 V. Maire, A. Mäkelä, H. Mäkinen, A. C. M. Malhado, A. Mallik, P. Manning, S. 828 Manzoni, Z. Marchetti, L. Marchino, V. Marcilio‐Silva, E. Marcon, M. Marignani, L. 829 Markesteijn, A. Martin, C. Martínez‐Garza, J. Martínez‐Vilalta, T. Mašková, K. Mason, 830 N. Mason, T. J. Massad, J. Masse, I. Mayrose, J. McCarthy, M. L. McCormack, K. 831 McCulloh, I. R. McFadden, B. J. McGill, M. Y. McPartland, J. S. Medeiros, B. Medlyn, 832 P. Meerts, Z. Mehrabi, P. Meir, F. P. L. Melo, M. Mencuccini, C. Meredieu, J. Messier, I. 833 Mészáros, J. Metsaranta, S. T. Michaletz, C. Michelaki, S. Migalina, R. Milla, J. E. D. 834 Miller, V. Minden, R. Ming, K. Mokany, A. T. Moles, A. Molnár, J. Molofsky, M. Molz, 835 R. A. Montgomery, A. Monty, L. Moravcová, A. Moreno‐Martínez, M. Moretti, A. S. 836 Mori, S. Mori, D. Morris, J. Morrison, L. Mucina, S. Mueller, C. D. Muir, S. C. Müller, 837 F. Munoz, I. H. Myers‐Smith, R. W. Myster, M. Nagano, S. Naidu, A. Narayanan, B. 838 Natesan, L. Negoita, A. S. Nelson, E. L. Neuschulz, J. Ni, G. Niedrist, J. Nieto, Ü. 839 Niinemets, R. Nolan, H. Nottebrock, Y. Nouvellon, A. Novakovskiy, K. O. Nystuen, A. 840 O’Grady, K. O’Hara, A. O’Reilly‐Nugent, S. Oakley, W. Oberhuber, T. Ohtsuka, R. 841 Oliveira, K. Öllerer, M. E. Olson, V. Onipchenko, Y. Onoda, R. E. Onstein, J. C. 842 Ordonez, N. Osada, I. Ostonen, G. Ottaviani, S. Otto, G. E. Overbeck, W. A. Ozinga, A. 843 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 29 T. Pahl, C. E. T. Paine, R. J. Pakeman, A. C. Papageorgiou, E. Parfionova, M. Pärtel, M. 844 Patacca, S. Paula, J. Paule, H. Pauli, J. G. Pausas, B. Peco, J. Penuelas, A. Perea, P. L. 845 Peri, A. C. Petisco‐Souza, A. Petraglia, A. M. Petritan, O. L. Phillips, S. Pierce, V. D. 846 Pillar, J. Pisek, A. Pomogaybin, H. Poorter, A. Portsmuth, P. Poschlod, C. Potvin, D. 847 Pounds, A. S. Powell, S. A. Power, A. Prinzing, G. Puglielli, P. Pyšek, V. Raevel, A. 848 Rammig, J. Ransijn, C. A. Ray, P. B. Reich, M. Reichstein, D. E. B. Reid, M. Réjou‐849 Méchain, V. R. de Dios, S. Ribeiro, S. Richardson, K. Riibak, M. C. Rillig, F. Riviera, E. 850 M. R. Robert, S. Roberts, B. Robroek, A. Roddy, A. V. Rodrigues, A. Rogers, E. 851 Rollinson, V. Rolo, C. Römermann, D. Ronzhina, C. Roscher, J. A. Rosell, M. F. 852 Rosenfield, C. Rossi, D. B. Roy, S. Royer‐Tardif, N. Rüger, R. Ruiz‐Peinado, S. B. 853 Rumpf, G. M. Rusch, M. Ryo, L. Sack, A. Saldaña, B. Salgado‐Negret, R. Salguero‐854 Gomez, I. Santa‐Regina, A. C. Santacruz‐García, J. Santos, J. Sardans, B. Schamp, M. 855 Scherer‐Lorenzen, M. Schleuning, B. Schmid, M. Schmidt, S. Schmitt, J. V. Schneider, 856 S. D. Schowanek, J. Schrader, F. Schrodt, B. Schuldt, F. Schurr, G. S. Garvizu, M. 857 Semchenko, C. Seymour, J. C. Sfair, J. M. Sharpe, C. S. Sheppard, S. Sheremetiev, S. 858 Shiodera, B. Shipley, T. A. Shovon, A. Siebenkäs, C. Sierra, V. Silva, M. Silva, T. Sitzia, 859 H. Sjöman, M. Slot, N. G. Smith, D. Sodhi, P. Soltis, D. Soltis, B. Somers, G. Sonnier, 860 M. V. Sørensen, E. E. Sosinski, N. A. Soudzilovskaia, A. F. Souza, M. Spasojevic, M. G. 861 Sperandii, A. B. Stan, J. Stegen, K. Steinbauer, J. G. Stephan, F. Sterck, D. B. Stojanovic, 862 T. Strydom, M. L. Suarez, J.-C. Svenning, I. Svitková, M. Svitok, M. Svoboda, E. 863 Swaine, N. Swenson, M. Tabarelli, K. Takagi, U. Tappeiner, R. Tarifa, S. Tauugourdeau, 864 C. Tavsanoglu, M. te Beest, L. Tedersoo, N. Thiffault, D. Thom, E. Thomas, K. 865 Thompson, P. E. Thornton, W. Thuiller, L. Tichý, D. Tissue, M. G. Tjoelker, D. Y. P. 866 Tng, J. Tobias, P. Török, T. Tarin, J. M. Torres‐Ruiz, B. Tóthmérész, M. Treurnicht, V. 867 Trivellone, F. Trolliet, V. Trotsiuk, J. L. Tsakalos, I. Tsiripidis, N. Tysklind, T. Umehara, 868 V. Usoltsev, M. Vadeboncoeur, J. Vaezi, F. Valladares, J. Vamosi, P. M. van Bodegom, 869 M. van Breugel, E. V. Cleemput, M. van de Weg, S. van der Merwe, F. van der Plas, M. 870 T. van der Sande, M. van Kleunen, K. V. Meerbeek, M. Vanderwel, K. A. Vanselow, A. 871 Vårhammar, L. Varone, M. Y. V. Valderrama, K. Vassilev, M. Vellend, E. J. Veneklaas, 872 H. Verbeeck, K. Verheyen, A. Vibrans, I. Vieira, J. Villacís, C. Violle, P. Vivek, K. 873 Wagner, M. Waldram, A. Waldron, A. P. Walker, M. Waller, G. Walther, H. Wang, F. 874 Wang, W. Wang, H. Watkins, J. Watkins, U. Weber, J. T. Weedon, L. Wei, P. Weigelt, 875 E. Weiher, A. W. Wells, C. Wellstein, E. Wenk, M. Westoby, A. Westwood, P. J. White, 876 M. Whitten, M. Williams, D. E. Winkler, K. Winter, C. Womack, I. J. Wright, S. J. 877 Wright, J. Wright, B. X. Pinho, F. Ximenes, T. Yamada, K. Yamaji, R. Yanai, N. 878 Yankov, B. Yguel, K. J. Zanini, A. E. Zanne, D. Zelený, Y.-P. Zhao, J. Zheng, J. Zheng, 879 K. Ziemińska, C. R. Zirbel, G. Zizka, I. C. Zo‐Bi, G. Zotz, and C. Wirth. 2020. TRY 880 plant trait database – enhanced coverage and open access. Global Change Biology 881 26:119–188. 882 Kattge, J., S. Díaz, S. Lavorel, I. C. Prentice, P. Leadley, G. Bönisch, E. Garnier, M. Westoby, P. 883 B. Reich, I. J. Wright, J. H. C. Cornelissen, C. Violle, S. P. Harrison, P. M. V. Bodegom, 884 M. Reichstein, B. J. Enquist, N. A. Soudzilovskaia, D. D. Ackerly, M. Anand, O. Atkin, 885 M. Bahn, T. R. Baker, D. Baldocchi, R. Bekker, C. C. Blanco, B. Blonder, W. J. Bond, R. 886 Bradstock, D. E. Bunker, F. Casanoves, J. Cavender‐Bares, J. Q. Chambers, F. S. C. Iii, J. 887 Chave, D. Coomes, W. K. Cornwell, J. M. Craine, B. H. Dobrin, L. Duarte, W. Durka, J. 888 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 30 Elser, G. Esser, M. Estiarte, W. F. Fagan, J. Fang, F. Fernández‐Méndez, A. Fidelis, B. 889 Finegan, O. Flores, H. Ford, D. Frank, G. T. Freschet, N. M. Fyllas, R. V. Gallagher, W. 890 A. Green, A. G. Gutierrez, T. Hickler, S. I. Higgins, J. G. Hodgson, A. Jalili, S. Jansen, 891 C. A. Joly, A. J. Kerkhoff, D. Kirkup, K. Kitajima, M. Kleyer, S. Klotz, J. M. H. Knops, 892 K. Kramer, I. Kühn, H. Kurokawa, D. Laughlin, T. D. Lee, M. Leishman, F. Lens, T. 893 Lenz, S. L. Lewis, J. Lloyd, J. Llusià, F. Louault, S. Ma, M. D. Mahecha, P. Manning, T. 894 Massad, B. E. Medlyn, J. Messier, A. T. Moles, S. C. Müller, K. Nadrowski, S. Naeem, 895 Ü. Niinemets, S. Nöllert, A. Nüske, R. Ogaya, J. Oleksyn, V. G. Onipchenko, Y. Onoda, 896 J. Ordoñez, G. Overbeck, W. A. Ozinga, S. Patiño, S. Paula, J. G. Pausas, J. Peñuelas, O. 897 L. Phillips, V. Pillar, H. Poorter, L. Poorter, P. Poschlod, A. Prinzing, R. Proulx, A. 898 Rammig, S. Reinsch, B. Reu, L. Sack, B. Salgado‐Negret, J. Sardans, S. Shiodera, B. 899 Shipley, A. Siefert, E. Sosinski, J.-F. Soussana, E. Swaine, N. Swenson, K. Thompson, P. 900 Thornton, M. Waldram, E. Weiher, M. White, S. White, S. J. Wright, B. Yguel, S. 901 Zaehle, A. E. Zanne, and C. Wirth. 2011. TRY – a global database of plant traits. Global 902 Change Biology 17:2905–2935. 903 Kong, D., C. Ma, Q. Zhang, L. Li, X. Chen, H. Zeng, and D. Guo. 2014. Leading dimensions in 904 absorptive root trait variation across 96 subtropical forest species. New Phytologist. 905 Kramer-Walter, K. R., P. J. Bellingham, T. R. Millar, R. D. Smissen, S. J. Richardson, and D. C. 906 Laughlin. 2016. Root traits are multidimensional: specific root length is independent 907 from root tissue density and the plant economic spectrum. Journal of Ecology 104:1299–908 1310. 909 Lambers, H., M. W. Shane, M. D. Cramer, S. J. Pearse, and E. J. Veneklaas. 2006. Root 910 Structure and Functioning for Efficient Acquisition of Phosphorus: Matching 911 Morphological and Physiological Traits. Annals of Botany 98:693–713. 912 Laughlin, D. C., L. Mommer, F. M. Sabatini, H. Bruelheide, T. W. Kuyper, M. L. McCormack, 913 J. Bergmann, G. T. Freschet, N. R. Guerrero-Ramírez, C. M. Iversen, J. Kattge, I. C. 914 Meier, H. Poorter, C. Roumet, M. Semchenko, C. J. Sweeney, O. J. Valverde-Barrantes, 915 F. van der Plas, J. van Ruijven, L. M. York, I. Aubin, O. R. Burge, C. Byun, R. 916 Ćušterevska, J. Dengler, E. Forey, G. R. Guerin, B. Hérault, R. B. Jackson, D. N. Karger, 917 J. Lenoir, T. Lysenko, P. Meir, Ü. Niinemets, W. A. Ozinga, J. Peñuelas, P. B. Reich, M. 918 Schmidt, F. Schrodt, E. Velázquez, and A. Weigelt. 2021. Root traits explain plant 919 species distributions along climatic gradients yet challenge the nature of ecological trade-920 offs. Nature Ecology & Evolution 5:1123–1134. 921 Leitão, R. P., J. Zuanon, S. Villéger, S. E. Williams, C. Baraloto, C. Fortunel, F. P. Mendonça, 922 and D. Mouillot. 2016. Rare species contribute disproportionately to the functional 923 structure of species assemblages. Proceedings of the Royal Society B: Biological 924 Sciences 283:20160084. 925 Ma, Z., D. Guo, X. Xu, M. Lu, R. D. Bardgett, D. M. Eissenstat, M. L. McCormack, and L. O. 926 Hedin. 2018. Evolutionary history resolves global organization of root functional traits. 927 Nature 555:94–97. 928 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 31 Madriñán, S., A. J. Cortés, and J. E. Richardson. 2013. Páramo is the world’s fastest evolving 929 and coolest biodiversity hotspot. Frontiers in Genetics 4:1–7. 930 Mammola, S., and P. Cardoso. 2020. Functional diversity metrics using kernel density n-931 dimensional hypervolumes. Methods in Ecology and Evolution 11:986–995. 932 Marklein, A. R., J. B. Winbourne, S. K. Enders, D. J. X. Gonzalez, T. L. Van Huysen, J. E. 933 Izquierdo, D. R. Light, D. Liptzin, K. E. Miller, S. L. Morford, R. A. Norton, and B. Z. 934 Houlton. 2016. Mineralization ratios of nitrogen and phosphorus from decomposing litter 935 in temperate versus tropical forests. Global Ecology and Biogeography 25:335–346. 936 McCormack, M. L., T. S. Adams, E. A. H. Smithwick, and D. M. Eissenstat. 2012. Predigting 937 fine root lifespan from plant functional traits in temperate trees. New Phytologist 938 195:823–831. 939 McCormack, M. L., I. A. Dickie, D. M. Eissenstat, T. J. Fahey, C. W. Fernandez, D. Guo, H.-S. 940 Helmisaari, E. A. Hobbie, C. M. Iversen, R. B. Jackson, J. Leppälammi‐Kujansuu, R. J. 941 Norby, R. P. Phillips, K. S. Pregitzer, S. G. Pritchard, B. Rewald, and M. Zadworny. 942 2015. Redefining fine roots improves understanding of below-ground contributions to 943 terrestrial biosphere processes. New Phytologist 207:505–518. 944 Mommer, L., and M. Weemstra. 2012. The role of roots in the resource economics spectrum: 945 Commentary. New Phytologist 195:725–727. 946 Morrone, J. J. 2024. Why biogeographical transition zones matter. Journal of Biogeography 947 51:544–549. 948 Oksanen, J., F. G. Blanchet, M. Friendly, R. Kindt, P. Legendre, D. McGlinn, P. R. Minchin, R. 949 B. O’Hara, G. L. Simpson, P. Solymos, H. H. Stevens, E. Szoecs, and H. Wagner. 2019. 950 vegan: Community Ecology Package. 951 Osborne, C. P., T. Charles-Dominique, N. Stevens, W. J. Bond, G. Midgley, and C. E. R. 952 Lehmann. 2018. Human impacts in African savannas are mediated by plant functional 953 traits. New Phytologist 220:10–24. 954 Pierick, K., C. Leuschner, and J. Homeier. 2021. Topography as a factor driving small-scale 955 variation in tree fine root traits and root functional diversity in a species-rich tropical 956 montane forest. New Phytologist 230:129–138. 957 Pierick, K., C. Leuschner, R. M. Link, S. Báez, A. Velescu, W. Wilcke, and J. Homeier. (n.d.). 958 Above- and belowground strategies of tropical montane tree species are coordinated and 959 driven by small-scale nitrogen availability. Functional Ecology n/a. 960 Pierick, K., R. M. Link, C. Leuschner, and J. Homeier. 2023. Elevational trends of tree fine root 961 traits in species-rich tropical Andean forests. Oikos 2023:e08975. 962 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 32 Poorter, L., and F. Bongers. 2006. Leaf traits are good predictors of plant performance across 53 963 rain forest species. Ecology 87:1733–1743. 964 Procheş, Ş., J. R. U. Wilson, and R. M. Cowling. 2006. How much evolutionary history in a 965 10×10 m plot? Proceedings of the Royal Society B: Biological Sciences 273:1143–1148. 966 R Core Team. 2023. R: A Language and Environment for Statistical Computing. R Foundation 967 for Statistical Computing, Vienna, Austria. 968 Raven, P. H., R. E. Gereau, P. B. Phillipson, C. Chatelain, C. N. Jenkins, and C. U. Ulloa. 2020. 969 The distribution of biodiversity richness in the tropics. Science Advances 6:eabc6228. 970 Reich, P. B. 2014. The world-wide ‘fast-slow’ plant economics spectrum: a traits manifesto. 971 Journal of Ecology 102:275–301. 972 Reich, P. B., and J. Oleksyn. 2004. Global patterns of plant leaf N and P in relation to 973 temperature and latitude. PNAS 101:11001–11006. 974 Reich, P. B., M. G. Tjoelker, M. B. Walters, D. W. Vanderklein, and C. Buschena. 1998. Close 975 association of RGR, leaf and root morphology, seed mass and shade tolerance in 976 seedlings of nine boreal tree species grown in high and low light. Functional Ecology 977 12:327–338. 978 Reich, P. B., M. B. Walters, and D. S. Ellsworth. 1997. From tropics to tundra: Global 979 convergence in plant functioning. Proceedings of the National Academy of Sciences 980 94:13730–13734. 981 Reich, P. B., I. J. Wright, J. Cavender‐Bares, J. M. Craine, J. Oleksyn, M. Westoby, and M. B. 982 Walters. 2003. The Evolution of Plant Functional Variation: Traits, Spectra, and 983 Strategies. International Journal of Plant Sciences 164:S143–S164. 984 Roumet, C., M. Birouste, C. Picon-Cochard, M. Ghestem, N. Osman, S. Vrignon-Brenas, K. 985 Cao, and A. Stokes. 2016. Root structure–function relationships in 74 species: evidence 986 of a root economics spectrum related to carbon economy. New Phytologist 210:815–826. 987 Smith, S. A., and J. W. Brown. 2018. Constructing a broadly inclusive seed plant phylogeny. 988 American Journal of Botany 105:302–314. 989 Smith, S. E., and D. J. Read. 2008. Mycorrhizal Symbiosis. Third edition. Academic Press. 990 Soudzilovskaia, N. A., P. M. van Bodegom, C. Terrer, M. van’t Zelfde, I. McCallum, M. L. 991 McCormack, J. B. Fisher, M. C. Brundrett, N. C. de Sá, and L. Tedersoo. 2019. Global 992 mycorrhizal plant distribution linked to terrestrial carbon stocks. Nature Communications 993 10:1–10. 994 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 33 Steidinger, B. S., T. W. Crowther, J. Liang, M. E. V. Nuland, G. D. A. Werner, P. B. Reich, G. 995 Nabuurs, S. de-Miguel, M. Zhou, N. Picard, B. Herault, X. Zhao, C. Zhang, D. Routh, 996 and K. G. Peay. 2019. Climatic controls of decomposition drive the global biogeography 997 of forest-tree symbioses. Nature 569:404. 998 Taylor, A., P. Weigelt, P. Denelle, L. Cai, and H. Kreft. 2023. The contribution of plant life and 999 growth forms to global gradients of vascular plant diversity. New Phytologist 240:1548–1000 1560. 1001 Tedersoo, L., editor. 2017. Biogeography of mycorrhizal symbiosis. Springer International 1002 Publishing AG. 1003 Townsend, A. R., G. P. Asner, and C. C. Cleveland. 2008. The biogeochemical heterogeneity of 1004 tropical forests. Trends in Ecology & Evolution 23:424–431. 1005 Umaña, M. N., X. Mi, M. Cao, B. J. Enquist, Z. Hao, R. Howe, Y. Iida, D. Johnson, L. Lin, X. 1006 Liu, K. Ma, I.-F. Sun, J. Thompson, M. Uriarte, X. Wang, A. Wolf, J. Yang, J. K. 1007 Zimmerman, and N. G. Swenson. 2017. The role of functional uniqueness and spatial 1008 aggregation in explaining rarity in trees. Global Ecology and Biogeography 26:777–786. 1009 Valverde-Barrantes, O. J., L. Authier, H. Schimann, and C. Baraloto. 2021. Root anatomy helps 1010 to reconcile observed root trait syndromes in tropical tree species. American Journal of 1011 Botany 108:744–755. 1012 Valverde-Barrantes, O. J., A. L. Horning, K. A. Smemo, and C. B. Blackwood. 2016. 1013 Phylogenetically structured traits in root systems influence arbuscular mycorrhizal 1014 colonization in woody angiosperms. Plant and Soil 404:1–12. 1015 Valverde‐Barrantes, O. J., H. Maherali, C. Baraloto, and C. B. Blackwood. 2020. Independent 1016 evolutionary changes in fine-root traits among main clades during the diversification of 1017 seed plants. New Phytologist 228:541–553. 1018 Valverde-Barrantes, O. J., K. A. Smemo, and C. B. Blackwood. 2015. Fine root morphology is 1019 phylogenetically structured, but nitrogen is related to the plant economics spectrum in 1020 temperate trees. Functional Ecology 29:796–807. 1021 Vitousek, P. 2004. Nutrient cycling and limitation: Hawai’i as a model system. Princeton 1022 University Press, Princeton (NJ). 1023 Vleminckx, J., O. V. Barrantes, C. Fortunel, C. E. T. Paine, D. Bauman, J. Engel, P. Petronelli, 1024 N. Dávila, M. Rios, E. H. Valderrama Sandoval, I. Mesones, E. Allié, J.-Y. Goret, F. C. 1025 Draper, J. E. Guevara Andino, S. Béroujon, P. V. A. Fine, and C. Baraloto. 2023. Niche 1026 breadth of Amazonian trees increases with niche optimum across broad edaphic 1027 gradients. Ecology 104:e4053. 1028 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 34 Weemstra, M., L. Mommer, E. J. W. Visser, J. Ruijven, T. W. Kuyper, G. M. J. Mohren, and F. 1029 J. Sterck. 2016. Towards a multidimensional root trait framework: a tree root review. 1030 New Phytologist 211:1159–1169. 1031 Weemstra, M., and O. J. Valverde-Barrantes. 2022. Above- and below-ground trait responses to 1032 environmental variation: the need to distinguish inter- and intraspecific variability. A 1033 commentary on ‘Above and below-ground plant traits are not consistent in response to 1034 drought and competition treatments.’ Annals of Botany:mcac135. 1035 Weemstra, M., O. J. Valverde-Barrantes, C. Fortunel, E. M. Oblitas Mendoza, E. M. B. Prata, M. 1036 Vásquez Pilco, A. Vicentini, J. Vleminckx, and C. Baraloto. 2023. Weak phylogenetic 1037 and habitat effects on root trait variation of 218 Neotropical tree species. Frontiers in 1038 Forests and Global Change 6:1187127. 1039 Westoby, M., D. S. Falster, A. T. Moles, P. A. Vesk, and I. J. Wright. 2002. Plant Ecological 1040 Strategies: Some Leading Dimensions of Variation Between Species. Annual Review of 1041 Ecology and Systematics 33:125–159. 1042 Withington, J. M., P. B. Reich, J. Oleksyn, and D. M. Eissenstat. 2006. Comparisons of Structure 1043 and Life Span in Roots and Leaves among Temperate Trees. Ecological Monographs 1044 76:381–397. 1045 Wright, I. J., P. B. Reich, M. Westoby, D. D. Ackerly, Z. Baruch, F. Bongers, J. Cavender-Bares, 1046 T. Chapin, J. H. C. Cornelissen, M. Diemer, J. Flexas, E. Garnier, P. K. Groom, J. Gulias, 1047 K. Hikosaka, B. B. Lamont, T. Lee, W. Lee, C. Lusk, J. J. Midgley, M.-L. Navas, U. 1048 Niinemets, J. Oleksyn, N. Osada, H. Poorter, P. Poot, L. Prior, V. I. Pyankov, C. Roumet, 1049 S. C. Thomas, M. G. Tjoelker, E. J. Veneklaas, and R. Villar. 2004. The worldwide leaf 1050 economics spectrum. Nature 428:821–827. 1051 Xia, M., O. J. Valverde‐Barrantes, V. Suseela, C. B. Blackwood, and N. Tharayil. 2021. 1052 Coordination between compound‐specific chemistry and morphology in plant roots aligns 1053 with ancestral mycorrhizal association in woody angiosperms. New Phytologist 1054 232:1259–1271. 1055 Yan, H., G. T. Freschet, H. Wang, J. A. Hogan, S. Li, O. J. Valverde-Barrantes, X. Fu, R. Wang, 1056 X. Dai, L. Jiang, S. Meng, F. Yang, M. Zhang, and L. Kou. 2022a. Mycorrhizal 1057 symbiosis pathway and edaphic fertility frame root economics space among tree species. 1058 New Phytologist 234:1639–1653. 1059 Yan, H., G. T. Freschet, H. Wang, J. A. Hogan, S. Li, O. J. Valverde-Barrantes, X. Fu, R. Wang, 1060 X. Dai, L. Jiang, S. Meng, F. Yang, M. Zhang, and L. Kou. 2022b. Mycorrhizal 1061 symbiosis pathway and edaphic fertility frame root economics space among tree species. 1062 New Phytologist 234:1639–1653. 1063 Yang, X., W. M. Post, P. E. Thornton, and A. Jain. 2014. Global Gridded Soil Phosphorus 1064 Distribution Maps at 0.5-degree Resolution. Data set. Available on-line 1065 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 35 [http://daac.ornl.gov] from Oak Ridge National Laboratory Distributed Active Archive 1066 Center. Oak Ridge, Tennessee, USA. 1067 Yu, F., Z. Zhang, L. Chen, J. Wang, and Z. Shen. 2018. Spatial distribution characteristics of soil 1068 organic carbon in subtropical forests of mountain Lushan, China. Environmental 1069 Monitoring and Assessment 190:545. 1070 Yu, G. 2020. Using ggtree to Visualize Data on Tree-Like Structures. Current Protocols in 1071 Bioinformatics 69:e96. 1072 Yu, G., D. K. Smith, H. Zhu, Y. Guan, and T. Tsan-Yuk Lam. 2017. ggtree: an R package for 1073 visualization and annotation of phylogenetic trees with their covariates and other 1074 associated data. Methods in Ecology and Evolution 8:28–36. 1075 Zanne, A. E., D. C. Tank, W. K. Cornwell, J. M. Eastman, S. A. Smith, R. G. FitzJohn, D. J. 1076 McGlinn, B. C. O’Meara, A. T. Moles, P. B. Reich, D. L. Royer, D. E. Soltis, P. F. 1077 Stevens, M. Westoby, I. J. Wright, L. Aarssen, R. I. Bertin, A. Calaminus, R. Govaerts, 1078 F. Hemmings, M. R. Leishman, J. Oleksyn, P. S. Soltis, N. G. Swenson, L. Warman, and 1079 J. M. Beaulieu. 2014. Three keys to the radiation of angiosperms into freezing 1080 environments. Nature 506:89–92. 1081 Zizka, A., D. Silvestro, T. Andermann, J. Azevedo, C. Duarte Ritter, D. Edler, H. Farooq, A. 1082 Herdean, M. Ariza, R. Scharn, S. Svantesson, N. Wengström, V. Zizka, and A. Antonelli. 1083 2019. CoordinateCleaner: Standardized cleaning of occurrence records from biological 1084 collection databases. Methods in Ecology and Evolution 10:744–751. 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 36 Table 1. Mean and dispersion values for the four target traits. Based on logarithmically 1096 transformed species-level mean values for specific root length, mean root diameter, root nitrogen 1097 concentration, and root tissue density. Mean and dispersion values were calculated, standardizing 1098 species richness across biomes by randomly selecting 100 species for each biome. Mean and 95 1099 % confidence intervals (CI, in gray) were based on 999 iterations. Different letters indicate 1100 differences among biomes based on overlapping confidence intervals (0.025 and 0.975%). 1101 Values were back-transformed with exponential. 1102 Traits Temperate Subtropical Tropical -------------------Mean (95% CI) -------------------- Specific root length (m g-1) Mean 27.9 (22.8, 33.4)a 17.1 (14.6, 19.8)b 13.8 (11.6, 16.4)b Dispersion 3.0 (2.5, 4.0)a 2.8 (2.5, 3.0)a 2.6 (2.2, 3.0)a Mean root diameter (mm) Mean 0.39 (0.36, 0.41)a 0.47 (0.44, 0.51)b 0.51 (0.47, 0.55)b Dispersion 1.6 (1.5, 1.7)a 1.6 (1.5, 1.7)a 1.6 (1.5, 1.7)a Root nitrogen concentration (mg g-1) Mean 13.2 (12.1, 14.2)a,b 13.0 (12.3, 13.6)a 15.0 (13.9, 16.1)b Dispersion 1.6 (1.4, 1.8)a 1.6 (1.5, 1.7)a 1.5 (1.4, 1.6)a Root tissue density (g cm-3) Mean 0.28 (0.25, 0.30)a 0.33 (0.30, 0.36)b 0.34 (0.32, 0.36)b Dispersion 1.7 (1.6, 1.8)a 1.8 (1.7, 1.9)a 1.5 (1.4, 1.5)b 1103 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 37 Table 2. Functional richness, evenness, and dispersion for temperate, subtropical, and tropical 1104 biomes. Mean and 95 % confidence intervals (CI, in gray) based on 999 iterations. Different 1105 letters indicate significant differences among biomes based on non-overlapping confidence 1106 intervals (0.025 and 0.975%). 1107 Functional Temperate Subtropical Tropical -------------- Mean (95% CI) ------------ Richness 29.5 (24.9, 35.8)a 31.9 (27.3, 36.9)a 30.5 (24.8, 35.8)a Evenness 0.41 (0.24, 0.54)a 0.52 (0.47, 0.59)a 0.44 (0.38, 0.51)a Dispersion 2.10 (1.92, 2.41)a 2.17 (1.98, 2.36)a 2.14 (1.90, 2.39)a 1108 1109 1110 1111 1112 1113 1114 1115 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint Table 3. Best models explaining root trait variation across species in subtropical and tropical 1116 biomes. To maximize the number of species included in the models, models for each trait varied 1117 in terms of species number and composition. Df, degrees of freedom; F, F statistic; and P, 1118 probability value. Bold P values represent significant relationships between root traits and 1119 species climate and soil niches (position and breadth) and biogeographic and evolutionary 1120 variables (ɑ = 0.05). 1121 Specific root length Mean root diameter Root N concentration Root tissue density Df F P Df F P Df F P Df F P Subtropical Breadthclim 1 ↑13.2 <0.001 1 ↓5.68 0.018 1 0.13 0.723 1 0.47 0.493 PC1clim 1 ↓13.7 <0.001 1 2.34 0.128 1 0.27 0.601 1 ↑12.66 <0.001 PC2clim 1 0.06 0.806 1 0.47 0.492 1 2.16 0.144 1 1.08 0.300 Breadthsoil 1 ↓3.04 0.083 1 2.27 0.133 1 0.01 0.926 1 0.30 0.585 PC1soil 1 0.07 0.796 1 0.85 0.356 1 1.08 0.031 1 ↑2.87 0.091 PC2soil 1 0.03 0.864 1 0.26 0.613 1 0.04 0.842 1 0.98 0.320 PCoA1phyl 1 ↑3.90 0.049 1 0.13 0.722 1 0.25 0.619 1 ↓4.32 0.039 PCoA2phyl 1 ↓34.83 <0.001 1 0.18 0.671 1 0.53 0.467 1 ↓88.85 <0.001 Residuals 186 179 124 174 R2 0.27 0.09 0.04 0.56 Tropical Breadthclim 1 ↑8.04 0.004 1 ↓5.86 0.015 1 ↑4.62 0.032 1 2.00 0.158 PC1clim 1 ↓13.36 <0.001 1 1.58 0.209 1 2.13 0.145 1 ↑5.26 0.022 PC2clim 1 0.51 0.477 1 0.002 0.962 1 1.06 0.030 1 0.003 0.960 Breadthsoil 1 1.53 0.217 1 0.35 0.553 1 0.73 0.392 1 0.05 0.820 PC1soil 1 0.05 0.827 1 ↑3.44 0.064 1 0.80 0.373 1 ↓5.65 0.017 PC2soil 1 ↓3.02 0.082 1 ↑3.43 0.064 1 ↓6.16 0.013 1 0.002 0.960 Continent 4 2.44 0.046 4 4.28 0.001 PCoA1phyl 1 ↑3.47 0.063 1 ↑13.47 <0.001 PCoA2phyl 1 0.13 0.715 1 ↑8.80 0.003 Residuals 571 565 289 545 R2 0.09 0.05 0.09 0.09 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint Figure captions 1122 Figure 1. Variation in the root functional space for species among biomes. The root space is 1123 visualized using a principal component analysis (PCA). Traits included are mean root diameter, 1124 specific root length (SRL), root tissue density (RTD), and root nitrogen concentration (Root N). 1125 Each point represents a species, different colors refer to different biomes. Density distributions 1126 of the first two axes are shown on the upper and right sides. PCA and density distributions are 1127 based on 235, 143, and 289 species for temperate, subtropical, and tropical biomes, respectively. 1128 Bivariate correlations between traits across and within biomes are shown in Appendix S1: Figure 1129 S8. 1130 Figure 2. Histograms for species mean values for each of the target root traits. Traits included 1131 are specific root length (m g-1; n = 469, 233, and 615 temperate, subtropical, and tropical species, 1132 respectively), mean root diameter (mm; n = 353, 228, and 614 temperate, subtropical, and 1133 tropical species, respectively), root nitrogen (N) concentration (mg g-1; n = 416, 156 and 322 1134 temperate, subtropical and tropical species, respectively), and root tissue density (g cm-3; n= 329, 1135 223, and 598 temperate, subtropical, tropical species, respectively). Horizontal bar graphs show 1136 the medians (black, vertical lines) and boxes extending from the first and third quartile and the 1137 lines representing the min and max values that are not outliers of the trait distributions per 1138 biome. 1139 Figure 3. Principal component analysis acts as a graphical representation of the unique, i.e., 1140 unique multivariate strategies in each biome, and overlapping space, i.e., shared multivariate 1141 strategies between (A) all three biomes, (B), temperate and subtropical, (C) temperate and 1142 tropical, and (D) subtropical and tropical. The mean value for uniqueness and its 95% confidence 1143 intervals were calculated based on 999 repetitions. The traits included are root diameter, specific 1144 root length (SRL), root tissue density (RTD), and root nitrogen concentration (Root N). 1145 Figure 4. Phylogenetic relatedness (PCOA2phyl), i.e., evolutionary distance between the flora 1146 represented in our dataset from different continents, explaining interspecific variation in (A) 1147 specific root length and (B) root tissue density across subtropical species. Species from 1148 continents that are similar phylogenetically share similar specific root length and root tissue 1149 density values. 1150 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint Figure 1 1151 1152 1153 1154 1155 1156 1157 1158 1159 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint Figure 2 1160 1161 1162 1163 1164 1165 1166 1167 1168 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint Figure 3 1169 1170 1171 1172 1173 1174 1175 1176 1177 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint Figure 4 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint Supporting Information 1198 1199 Appendix S1. Supplementary Figures and Tables 1200 1201 Unique belowground ecological strategies of subtropical and tropical plant species expand 1202 the root trait space 1203 Nathaly Guerrero-Ramírez†, Monique Weemstra†, Shalom D. Addo-Danso, Kelly Andersen, 1204 Marie Arnaud, Amanda L. Cordeiro, Daniela F. Cusack, Martyna M. Kotowska, Ming Yang Lee, 1205 Céline Leroy, Laynara F. Lugli, Kerstin Pierick, Chris M. Smith-Martin, Amanda Taylor, Laura 1206 Toro, María Natalia Umaña, Oscar J Valverde-Barrantes, Michelle Wong, Claire Fortunel 1207 † Equal contribution 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint Supplementary Figures 1229 1230 1231 Figure S1. Number of coordinates (i.e., observations) used to quantity species-specific niches. 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 1244 Figure S2. Pearson correlations for species-specific soil niches (position). Soil variables included sand, silt, and clay content, total 1245 organic carbon (TOC), cation exchange capacity (CEC), and total soil phosphorus (P).1246 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 1247 1248 Figure S3. Principal component analysis of climatic variables for quantifying climate niche 1249 positions. Climatic variables include mean annual temperature (MAT, °C), mean annual 1250 precipitation (MAP, mm), precipitation in wettest and driest months (Wet and Dry), temperature 1251 seasonality (T_Seanonality), and maximum and minimum temperatures in warmest and coldest 1252 months (Warm and Cold), respectively. 1253 1254 1255 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 1256 Figure S4. Principal component analysis of soil variables used to quantify soil niche positions. 1257 Soil variables includes topsoil sand, silt, and clay fraction (% wt), topsoil organic carbon (TOC, 1258 % weight), and topsoil cation exchange capacity (CEC, cmol/kg). 1259 1260 1261 1262 1263 1264 1265 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 1266 Figure S5. Principal Coordinate Analysis (PCoA) to determine the evolutionary distance 1267 between the subtropical flora represented in our dataset from different continents. The PCoA is 1268 based on distance matrices using standardized effects sizes of phylogenetic turnover. 1269 1270 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 1271 Figure S6. Principal Coordinate Analysis (PCoA) to determine the evolutionary distance 1272 between the tropical flora represented in our dataset from different continents. The PCoA is 1273 based on distance matrices using standardized effects sizes of phylogenetic turnover. 1274 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 1275 1276 Figure S7. Root functional space including all available species mean values (n= 1035) together 1277 and separated into the three distinct regions (n= 561, 166, 308 for temperate, subtropical, and 1278 tropical species, respectively). Traits included are root diameter (mm), specific root length (SRL; 1279 m g-1), root tissue density (RTD; g cm-3), and root nitrogen concentration (Root N; mg g-1). Root 1280 functional spaces are visualized using a principal component analysis (PCA), with PC1 1281 representing the collaboration gradient (SRL and Diameter) and PC2 the conservation gradient 1282 (Root N and RTD). The root functional space was separated by biomes to improve visualization 1283 but temperate, subtropical, and tropical panels represent the same PCA shown for all species 1284 together. Each point represents a species, with the colors representing woodiness (yellow dots = 1285 667 woody, red dots = 320 non-woody, and blue dots = 10 non-woody/woody species). Contours 1286 are built using kernel density estimation. 1287 1288 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 1289 1290 Figure S8. Pearson correlations between root functional traits for woody species. 1291 1292 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 1293 Figure S9. Phylogenetic relatedness explaining root trait in the tropics. The role of phylogenetic 1294 relatedness (PCOA1phyl , PCOA2phyl), i.e., evolutionary distance between the flora represented in 1295 our dataset from different continents, explaining interspecific variation in mean root diameter 1296 and root nitrogen (N) concentration. Species from continents that are similar phylogenetically 1297 share similar specific root length and root tissue density values. 1298 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint 1299 Figure S10. The role of continents explaining root traits in the tropics. 1300 1301 1302 1303 1304 1305 1306 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint Supplementary Tables 1307 1308 Table S1. Minimum (Min), first and third quartile (1sr Qu. and 3rd Qu., respectively) median, 1309 mean, and maximum (Max), values of climatic and soil data used to calculate species' niches 1310 from subtropical and tropical species. In addition, total soil phosphorus (P) data at a 0.5-degree 1311 resolution were retrieved from the Global Gridded Soil Phosphorus Distribution Maps (Yang et 1312 al., 2014), but as these data were unavailable for 20 (sub)tropical species and for 12% of the 1313 species occurrence observations (i.e., 44,688) and were correlated with cation exchange capacity 1314 (Pearson r = 0.66, p-value < 0.001; Fig S2), we did not include soil P data in our analyses. 1315 Variables Min 1sr Qu. Median Mean 3rd Qu. Max Climatic variables Annual mean temperature (°C) -9.7 19.1 22.9 22.1 25.5 31.2 Annual precipitation (mm) 0 861 1386 1578 2189 10216 Precipitation wettest month (mm) 0 155 235 241.7 336 2519 Precipitation driest month (mm) 0 6 28 44.64 60 653 Temperature seasonality 0.68 6.8 18.3 27 44.4 100.6 Maximum Temperature Warmest month 35 27.6 29.9 29.8 32.0 47.6 Min temperature coldest month -26.8 8.0 15.3 14.1 20.5 26.7 Soil variables Topsoil sand fraction (% wt) 0.75 34.7 45.0 45.6 55.1 98.2 Topsoil silt fraction (% wt) 0 22.1 27.0 27.2 32.2 65.4 Topsoil clay fraction (% wt) 0 19.6 23.6 27.0 34.0 85.0 Topsoil organic carbon (% weight) 0.09 0.86 1.17 2.25 1.85 38.17 Topsoil cation exchange capacity (cmol/kg) 0.5 9.2 12.0 16.2 20.0 80.0 Total soil phosphorus (g m-2) 45.0 230.3 321.9 404.8 485.2 1577 *standard deviation × 100. 1316 1317 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint Table S2. Principal component analyses of root traits using 667 woody species. 1318 PC1 PC2 PC3 PC4 Standard deviation 1.341 1.194 0.817 0.329 Proportion of Variance 0.450 0.356 0.167 0.027 Cumulative Proportion 0.450 0.806 0.973 1.000 Loadings Specific root length -0.670 0.290 -0.191 0.655 Mean root diameter 0.718 0.024 -0.181 0.671 Root tissue density -0.143 -0.703 0.606 0.342 Root nitrogen concentration 0.118 0.648 0.750 0.052 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint Table S3. Principal component analysis of climatic variables for species-specific climatic niches 1336 (position) for 820 subtropical and tropical species. 1337 PC1 PC2 PC3 Standard deviation 2.036 1.318 0.885 Proportion of Variance 0.586 0.248 0.112 Cumulative Proportion 0.586 0.834 0.946 Loadings Mean annual temperature 0.365 0.496 -0.134 Mean annual precipitation 0.456 -0.123 0.363 Precipitation in the wettest month 0.406 0.008 0.507 Precipitation in the driest month 0.378 -0.354 0.220 Temperature seasonality -0.387 0.206 0.604 Maximum temperature in the warmest month 0.017 0.738 0.168 Minimum temperature in the coldest month 0.448 0.162 -0.388 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint Table S4. Principal component analysis of soil variables for species-specific soil niches 1349 (position) for 820 subtropical and tropical species. 1350 PC1 PC2 PC3 Standard deviation 1.438 1.301 0.972 Proportion of Variance 0.414 0.338 0.189 Cumulative Proportion 0.414 0.752 0.941 Loadings Topsoil sand fraction -0.467 0.561 -0.057 Topsoil silt fraction 0.539 0.019 0.626 Topsoil clay fraction 0.061 -0.693 -0.426 Topsoil organic carbon 0.385 0.369 -0.629 Topsoil cation exchange capacity 0.582 0.261 -0.165 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint Table S5. Model selection within biome variation (subtropical and tropical). 1365 Explanatory variables Response variables Specific root length Mean root diameter Root N Root tissue density ----------------------------AIC--------------------------- Subtropical Breadthclim + PC1clim + PC2clim + Breadthsoil + PC1soil + PC2soil + continent 533.7 269.2 199.5 193.8 Breadthclim + PC1clim + PC2clim + Breadthsoil + PC1soil + PC2soil + PCoA1phyl + PCoA2phyl 533.3 268.5 197.0 191.9 Tropical Breadthclim + PC1clim + PC2clim + Breadthsoil + PC1soil + PC2soil + continent 1600.8 761.9 351.0 458.9 Breadthclim + PC1clim + PC2clim + Breadthsoil + PC1soil + PC2soil + PCoA1phyl + PCoA2phyl 1604.4 759.4 349.7 470.6 1366 1367 1368 1369 1370 1371 1372 1373 1374 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint

1375 Brun P, Zimmermann NE, Hari C, Pellissier L, Karger DN. 2022. Global climate-related 1376 predictors at kilometer resolution for the past and future. Earth System Science Data 14: 1377 5573–5603. 1378 Cardoso P, Rigal F, Carvalho JC. 2015. BAT – Biodiversity Assessment Tools, an R package for 1379 the measurement and estimation of alpha and beta taxon, phylogenetic and functional 1380 diversity. Methods in Ecology and Evolution 6: 232–236. 1381 Fischer G, Nachtergaele F, Prieler S, van Velthuizen HT, Verelst L, Wiberg D. 2008. Global 1382 Agro-ecological Zones Assessment for Agriculture (GAEZ 2008). IIASA, Laxenburg, 1383 Austria and FAO, Rome, Italy. 1384 Yang X, Post WM, Thornton PE, Jain A. 2014. Global Gridded Soil Phosphorus Distribution 1385 Maps at 0.5-degree Resolution. Data set. Available on-line [http://daac.ornl.gov] from 1386 Oak Ridge National Laboratory Distributed Active Archive Center. Oak Ridge, 1387 Tennessee, USA. 1388 Zizka A, Silvestro D, Andermann T, Azevedo J, Duarte Ritter C, Edler D, Farooq H, Herdean A, 1389 Ariza M, Scharn R, et al. 2019. CoordinateCleaner: Standardized cleaning of occurrence 1390 records from biological collection databases. Methods in Ecology and Evolution 10: 744–1391 751. 1392 1393 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 10, 2024. ; https://doi.org/10.1101/2024.10.06.616893doi: bioRxiv preprint