This is the accepted version of the article: Veresoglou, Stavros; Peñuelas, Josep. Variance in biomass-allocation fractions is explained by distribution in European trees. DOI 10.1111/nph.15686 This version is avaible at https://ddd.uab.cat/record/218158 under the terms of the license 1 Variance in biomass-allocation fractions is explained by distribution 2 in European trees 3 4 5 Stavros D Veresoglou1,2*, Josep Peñuelas3,4 6 7 8 9 1 10 Freie Universität Berlin, Institut für Biologie, Plant Ecology, D-14195 Berlin, Germany 2 11 Faculty of Agriculture, Laboratory of Ecology and Environmental Protection, Aristotle 12 University of Thessaloniki, 541 24 Thessaloniki, Greece. 3 13 CSIC, Global Ecology Unit CREAF-CSIC-UAB, Bellaterra, 08193, Catalonia, Spain 4 14 CREAF, Cerdanyola del Vallès, 08193, Catalonia, Spain 15 16 Summary 17  Intraspecific variability in ecological traits confers the ability of a species to adapt to an 18 ever-changing environment. Fractions of biomass allocation in plants (BAFs) represent 19 both ecological traits and direct expressions of investment strategies and so have 20 important implications on plant fitness particularly under current global change. 21  We combined data on BAFs of trees in more than 10000 forest plots with their 22 distributions in Europe. We aimed to test whether plant species with wider distributions 23 have more or less variable intraspecific variance of BAFs foliage-woody biomass and 24 shoot-root ratios than species with limited distribution. 25  Irrespective of corrections for tree age and phylogenetic relatedness, the standard 26 deviation in BAFs was up to three times higher in species with most extensive 27 distributions than in those with least extensive distribution due to a higher genetic 28 diversity. Variance in BAFs also increased with latitude. 29  We show that a combination of 36% tree genetic diversity and 64% environmental 30 variability explains variance in BAFs and imply that changes in genetic diversity occur 31 quickly. Genetic diversity should thus play a key role in regulating species responses to 32 future climate change. Loss of habitat, even if transient, could induce a loss of genetic 33 diversity and hinder species survival. 34 35 Key-words: allometry; effective population size; extinction; genetic diversity; plant allometry; 36 plant morphology; woody plants 37 38 39 Introduction 40 41 The strategies of fitness and growth of sessile organisms are largely determined by biomass 42 allocation (Hodge, 2004; Poorter et al., 2012; Veresoglou et al., 2017), which in turn determines 43 the long-term morphology of the individual. The morphology of sessile organisms has been 44 studied the most in plants (Kokko, 2007). Plant morphology can be summarised in several ways, 45 each with distinct strengths and limitations. Poorter and Sack (2012) reviewed these methods and 46 concluded that fractions of biomass allocation, especially after correcting for confounding 47 allocation parameters such as size, represent a particularly effective measure. Plant morphology 48 has high functional importance, so the influence of environment on morphology has been well 49 studied. Some representative parameters that have been studied extensively for their influence on 50 plant allometry are the size of individual plants (Reich et al., 2005; Poorter et al., 2012), 51 historical environmental abiotic parameters such as stressors (e.g. uniform stress hypothesis) 52 (Mogan & Gannell, 1994; Dean et al., 2002), shading (Lusk et al., 2008; Duursma et al., 2010; 53 Forster et al., 2011), temperature (Reich et al., 2014), precipitation (McCarthy & Enquist, 2007), 54 biotic interactions such as competition and diversity (Forrester et al., 2017a); the growth form of 55 plants (Wyka et al., 2013) and species-specific ontogeny (Forrester et al., 2017b). Plant age, 56 however, is an influential factor that is often not available for integration into such analyses 57 (Nikklas, 1997; Nikklas et al., 2003; Barthélémy & Caraglio, 2007; Duursma et al., 2010). 58 Analysts often use plant size as a proxy of age (Nikklas, 2004; Bowman et al., 2013). 59 Biomass-allocation fractions (BAFs) describe biomass ratios, usually at a logarithmic 60 scale, of plant organs. BAFs are measureable to an individual level and thus constitute plant 61 traits (Müller et al., 2000; Reich et al., 2003). Trait variance is often comparable to or even more 62 important than trait means (Messier et al., 2010; Bolnick et al., 2011; Violle et al., 2012). High 63 variance in trait values could facilitate, for example, the evolution and adaptation of a species to 64 new environmental settings (Bolnick et al., 2011). Quantifying variation in BAFs could improve 65 modelling uncertainty in the standing biomass of woody habitats because most of our estimates 66 of standing biomass are projections of allometric equations (Chave et al., 2005; Muukkonen, 67 2007). 68 Plant morphology has been described as an equilibrium between constraints to plant 69 growth and exogenous environmental stressors (Barthélémy & Caraglio, 2007). Constraints to 70 plant growth and other intrinsic factors can negate the influence of the environment to varying 71 degrees, i.e. environmental conditions tend to shift BAFs against the stabilising influence of 72 intrinsic factors. We would then expect that BAFs would be more variable in plant species that 73 experience extreme environmental conditions more often. Reyer et al. (2013) argued that 74 extreme conditions occur mainly at the edges of the distribution of a species. Source-sink 75 population dynamics describe instances where species only occur in an area because of a 76 constant influx of propagules from surrounding areas where the species grows better. At the 77 edges of the distribution of a species it is more likely to observe source-sink population 78 dynamics (Remeš, 2000) than in the kernel of the distribution which could induce unique 79 phenotypes. Plant species that have a small distribution should more frequently experience such 80 “extreme edge” conditions so plants with smaller distributions may have the most variable BAFs. 81 The reason is that environmental heterogeneity mainly increases the variance in BAFs in these 82 species (Fig. 1a). We thus hypothesize that plant species with smaller distributions have more 83 variable BAFs (hypothesis one). Alternatively, the higher effective population sizes and genetic 84 variability could allow plant species that have extensive distributions to be those showing the 85 highest intraspecific trait variance, including BAFs. The reason could be that environmental 86 variability acts independently of distribution but more populous species show more diverse 87 BAFs due to a higher genetic diversity (Fig. 1b). As a result, plant species with smaller 88 distributions could vary less in their BAFs (hypothesis two). A final expectation is that variance 89 in BAFs should indicate the ability of species to adapt to local environmental conditions, so a 90 high variance suggests rapid evolution. Stapley et al. (2017) reported considerably lower 91 recombination rates in gymnosperms than angiosperms. The variance in BAFs should thus be 92 lower in conifers, compared to angiosperms (hypothesis three). 93 Most of our existing understanding on the way traits of woody species vary with genetic 94 diversity and the environment originates from provenance tests, where plants differing in their 95 origin are grown under common environmental conditions (Thompson et al. 2008; Wang et al. 96 2010). Some of the limitations of provenance test relate to the choice of the common 97 environment (Leites et al. 2012), the time they require for long-lived species to grow and thus the 98 logistics of destructively harvesting them (e.g. to assess BAFs). Here we worked on an 99 observational approach which synthesizes across existing BAF measurements from records of 100 destructive tree harvests in Europe (Fig. 1). Schepaschenko et al. (2017) have recently released 101 two large data sets detailing the biomass fractions of many trees that had been destructively 102 harvested between 1930 and 2014 in Eurasia. We combined this data set with information on the 103 distributions of many of these plant species in Europe from Mauri et al. (Mauri et al., 2017) to 104 address the above mentioned three hypotheses. 105 106 Materials and Methods 107 108 Sources of data 109 We used the two data sets published by Schepaschenko et al. (2017) for our main analysis. The 110 Biomass-tree data set provided information on biomass fractions following destructive 111 harvesting for 9613 trees, mainly in Europe. The Biomass-plot data set provided information on 112 biomass fractions for 10 351 plots distributed throughout Eurasia, each of which provided 113 information for cumulative biomass of two or more trees. The two data sets synthesized across 114 approximately 1200 experiments over the period 1930-2014 and contained information on 115 location and age of the trees as well as biomass information on several different fractions. We 116 used biomass information on foliage vs. woody aboveground biomass for our main analyses. We 117 also used root:shoot ratios as part of our sensitivity analyses. We extracted distributions of tree 118 species in Europe from Mauri et al. (Mauri et al., 2017). Mauri et al. (2017) only described tree 119 distributions in Europe, so we limited our analysis to trees that occur mainly in Europe. We used 120 QGis v 2.12.3 to estimate the size of the polygonal envelopes provided by Mauri et al. (2017). 121 We used Phylocom v 4.2 to reconstruct the phylogenetic relationships of the tree species in our 122 analysis. 123 We worked with an aggregate of 80 species in our analysis of European trees. A list of 124 the species and the phylogenetic reconstruction used to correct for phylogenetic relatedness can 125 be found in Table S1. The three data sets varied considerably in terms of their resolution and 126 their suitability for the different analyses. The cumulative analysis of the Biomass-tree data set 127 used biomass and age information for 4719 trees from 42 species (median observations per 128 species, 10.5; 1st quartile, 7; 3rd quartile, 78.5). The analysis of the Biomass-plot data set used 129 information for 3898 plot entries describing 63 species (median observations of plot per species, 130 15; 1st quartile, 8.5; 3rd quartile, 46) making it particularly suitable for assessing how BAFs are 131 influenced by environmental heterogeneity. The analysis of individual stands in the Biomass-tree 132 data set used an aggregate of 1854 tree observations from 40 species (median observations per 133 species, 8.5; 1st quartile, 7; 3rd quartile, 33.25). Because for the analysis on individual stands (i.e. 134 plots) in the Biomass-tree data set we only used per species information of trees in a single stand 135 (i.e. plot with most tree observations in the Biomass-tree data set), this made it particularly 136 suitable for quantifying the influence upon BAFs of genetic diversity. The Biomass-tree data set 137 contained proportionally more observations located in Europe than Biomass-plot data set did 138 (Fig.2 – inserts). 139 140 Rationale for the analyses 141 We modelled three sources of variance in BAFs, age, environment and genetic diversity. We 142 inferred genetic diversity from genetic variability (i.e. the degree to which the genetic 143 characteristics of populations vary) using a phenomenological approach that determined the 144 variance of morphological characteristics in populations after controlling for sources of variance 145 that were not of genetic origin such as age, environment and latitude. To control for age-related 146 differences in BAFs we used the slopes of linear models with BAFs as response variables and 147 age as predictor. To assess the fraction of BAFs that is explained by genetic diversity we 148 assessed how variance in BAFs scaled with tree species distribution in tree individuals which 149 belonged to the same species and stand after correcting them for tree age (individual stands in the 150 Biomass-tree data set; Fig. 1b). To assess the fraction of BAFs that is explained by 151 environmental variability we assumed that environmental variability is mainly due to 152 latitudinal range (de Frenne et al. 2013; we provide more details at the section poling effect 153 sizes) and quantified how much the variance in BAFs changed with each degree of latitudinal 154 range (i.e. latitudinal breadth – to avoid an overrepresentation of stands we used the Biomass- 155 plot dataset). 156 157 Statistical analyses 158 Our analyses only considered tree species (which were the unit of our analysis) for which a 159 minimum of five observations of age, foliage biomass and aboveground woody biomass were 160 available, because we needed a minimum of three observations to fit a linear model and then two 161 additional residuals to yield meaningful estimates of variance for the fit of the regression line 162 (results from a sensitivity test on the inclusion threshold are presented in Appendix 1 and Fig. 163 S1). To extract the fraction of variance in BAFs that was not due to age differences we fitted 164 linear models with BAFs as a response variable and age as predictor and then quantified the 165 standard deviation of the residuals. We explain the procedure in higher detail below (also 166 Appendix 2; Table S2): 167 1. We calculated the natural log response ratio (logRR) of foliage over aboveground woody 168 biomass ( log RR  ln  m foliage  , where m stands for biomass) for each individual tree in the case of m   woody  169 the Biomass-tree data base or plot in the case of the Biomass-plot data base (Appendix 2). In the 170 form of a sensitivity analysis we also analysed logRR of root over shoot BAFs. LogRR 171 represents a widely used nonstandard effect size in synthesis studies. A large logRR indicates a 172 higher investment in foliage than woody biomass. 173 2. We fitted a linear model with logRR as a response variable and age as the sole predictor. The 174 single most important idiosyncratic cause of variability in BAFs is age (Nikklas et al., 2003; 175 Nikklas, 2004; Barthélémy & Caraglio, 2007; Duursma et al., 2010). To correct our data for this 176 source of variance, we extracted the residuals of the linear model and assessed their standard 177 deviations. We assumed a first-order linear correlation between logRR and age using the 178 standard deviation of the residuals as a measure of the variance, in agreement with preliminary 179 analyses (Appendix 2). 180 3. We correlated our metric of variance of the relationship between BAFs and age with the 181 distribution of the plants. 182 In the first of our three analyses (Table S2) we used all observations of individual trees in 183 the Biomass-tree data set, which consisted of multiple trees per stand and multiple stands per 184 species. We then used information for all plots in the Biomass-plot data set, consisting of a single 185 plot per stand, which allowed us to correct for spatial autocorrelation. Finally, we reanalysed the 186 Biomass-tree data set but only using the information for each species in a single stand, with the 187 sole criterion that the stand had provided most observations for that species. This analysis 188 addressed the concern that the cause of higher variance in BAFs was due to a larger distance 189 across observations and thus a higher variability in environmental conditions. We conducted this 190 analysis both with and without phylogenetic correction. We used an analysis of phylogenetically 191 independent contrasts to correct for phylogeny. 192 We further compared our estimates of variance between angiosperms and gymnosperms. 193 We thus used t-tests, assuming unequal variance. 194 195 Sensitivity analyses 196 197 We tested the specificity of our observations to logRR between foliage and aboveground woody 198 biomass by replicating the analysis for logRR between total root biomass vs. total aboveground 199 biomass. The two data sets contained unequal numbers of observations of total root biomass, and 200 the number of species which we could analyse was low. We directly assessed how the variance 201 in biomass fractions scaled with distribution (i.e. independent of age) by directly assessing the 202 standard deviation of biomass-allocation ranges and correlating it with distribution (Appendix 2). 203 We further replicated the analysis with the Legacy Tree Data data set (Radtke et al., 204 2016). To estimate the distribution of North American trees we used information from the United 205 States Department of Agriculture (USDA 2017). This data set did not contain any information on 206 tree age. We used data from the Legacy_Tree table and used the variability of logRR foliage over 207 aboveground woody biomass for single locations. Sufficient information was available for only 208 three species, which were assessed for four locations. All three species belonged to the Pinaceae 209 family, had nested distributions and occurred in southeastern North America. We assessed 210 distribution by extracting information from the USDA for the number of states in which the 211 species occurred and regressed this metric against the variance of logRR. The amount of 212 information we extracted from the Legacy Tree Data base and the resolution, given that the size 213 of US states varies considerably, were much lower than in our main analysis so we only used as 214 a means of supporting the main results with a different dataset. Because we did not have accurate 215 distribution range data, we used here non-parametric statistics. 216 To assess the degree to which our observations were sensitive to the inclusion of invasive 217 tree species we classified tree species in native vs invasive as in Veresoglou et al. (2018; Table 218 S3). We repeated the analysis separately for the subsets of native and invasive species. 219 220 The influence of latitude 221 Latitude is an influential predictor of variance in plant traits (Heibo et al., 2005; Aerts et al., 222 2012), and as such it was important for us to show that it was not driving our results. We 223 assessed the degree to which our patterns were explained by latitude using two analyses. (a) We 224 first extracted the average latitudes of the distributions of the woody species in Europe reported 225 by Mauri et al. (2017). We compared these values with the variances in BAFs corrected for plant 226 age (i.e. the standard deviation of the residuals of the linear model described above). (b) We then 227 divided Europe into northern and southern Europe. We thus used the average of the two 228 latitudinal extremes of Europe: 82°N for the northern region and 35°N for the southern region, 229 averaging 58.5°N. We used the subset of points west of 69°W. We separately calculated the 230 variance of BAFs corrected for age for the woody species for which the Biomass-tree data set 231 contained a minimum of five observations both north and south of the latitudinal average 232 threshold, which we then combined in a new logRR (i.e. northern variance over southern 233 variance). We calculated the variance for these two sites and used them to recalculate logRR. 234 235 Pooling effect sizes to quantify genetic and environmental variance in BAFs 236 We partitioned here variance in BAFs into a fraction due to genetic variability and a fraction due 237 to environmental variability. We used a phenomenological approach for this (Fig. 1; Appendix 238 4). We made the following three assumptions: 239 1. The key factor contributing to environmental variability was latitudinal range of the 240 distribution of an organism. In support to this simplistic statement, see de Frenne et al. (2013) 241 demonstrating that some key environmental parameters such as temperature, precipitation, soil 242 pH and human influence covary with latitude. We assumed that there is a first order linear 243 relationship between latitudinal range and environmental variability in BAFs. 244 2. Genetic variability in BAFs increased with distribution at a slope equivalent to that in Fig. 2b. 245 3. The relative importance of the two fractions was assessed for an imaginary species with an 246 “average” latitudinal range which in this case was 19.4o latitude and a mean distribution range of 247 10,000 km2. 248 We use the additive principle of variances in statistics to add the two fractions on the assumption 249 that the observed variance was independent of our sample sizes. Our variance partitioning was 250 carried out for an idealized species having an average latitudinal range of 58.9o – 39.5o = 19.4o 251 latitudinal range and a distribution of 10.000 km2 which represented averages in our dataset and 252 were also consistent with (Tkach et al., 2008). We estimated the overall form of the model and 253 we calculated relative effect sizes for mean parameters of distribution and latitude in Europe. We 254 provide more information on the analysis in Appendix 4. 255 256 257 258 259 Results 260 261 The main results are on records throughout Eurasia but these were supported by a 262 reanalysis exclusively to the records from Europe. Our main analysis gave consistent results for 263 all three data sets (Fig. 2), namely that variance in BAFs increases with the distribution size of a 264 species. LogRR of foliage vs aboveground woody biomass was significantly positively 265 correlated with distribution in all three tests (Fig. 2). The correlation was strongest (i.e. highest 266 Kendall’s tau) for individual trees in the same stand (tau=0.51, Fig. 2c). We found yet stronger 267 correlations when we narrowed down the analysis on tree records occurring in Europe which we 268 here define as those with a longitude smaller than 69oW (Fig. S2). Correlations would have been 269 considerably weaker if we had not corrected for age of the trees (Fig. S3). No phylogenetic 270 conservatism was found in either the variance of biomass fractions (Blomberg’s K in the 271 cumulative analysis of the Biomass-tree data set was 0.37 with P = 0.95) or the distribution of 272 the species (Blomberg’s K in the same data set was 0.53 with P = 0.12; Appendix 3). The 273 correlations were weaker after correction for phylogeny with phylogenetically independent 274 contrasts, except for the Biomass-tree data set (Fig. 3). The variances of the BAFs did not differ 275 significantly between angiosperms and gymnosperms in any of the three data sets (Fig. 4 – 276 evidence against hypothesis three). 277 The analysis using root:shoot ratios produced comparable results but with lower 278 statistical power (Fig. 5a,b). The results of our analysis of the biomass data from America agreed 279 with those for European trees (Fig. 5c; tau=0.91, P=0.07). Most of the trees in the datasets were 280 classified as native and there were minimal differences in the additional analyses targeting 281 natives (Fig. S4). There was insufficient statistical power to reach robust conclusions for 282 invasives (Fig. S4). 283 284 The influence of latitude 285 There was no relationship between mean latitude and the variance of BAFs between foliage over 286 woody aboveground biomass in the Biomass-tree data set (Fig. 6a). The five tree species that 287 were observed both North and South of the latitudinal threshold of 58.5°N did not display any 288 patterns with regards to their BAFs (Fig. 6a – insert). Picea obovata showed higher BAFs in the 289 southern extent of its distribution, Pinus sylvestris and Betula alba in the northern extent of their 290 distribution whereas Larix sukaczewii and Pinus sibirica showed relatively balanced BAFs in 291 both extents. The lack of a relationship was even more apparent when we narrowed down our 292 analysis to individual stands in the Biomass-tree data set (data not shown). By contrast we could 293 observe a strong relationship between mean latitude and the variance of BAFs in the Biomass- 294 plot data set (Fig. 6b). 295 296 Relative effect sizes of genetic and environmental variability 297 We had 40 observations and a mean slope of 3.324.10-5 per thousand km2 of distribution in the 298 case of the individual stands in the Biomass-tree data set assessing genetic variability and 18 299 observations and a mean slope of 3.05. 10-2 per degree latitudinal range in the case of the 300 Biomass-plot data set measuring environmental variability. The resulting expression of variance 301 (Appendix 4) was s p  10 4 (0.11D 2  93,025 2 ) with D representing distribution range and φ 302 degrees latitudinal range. For average European settings of D ≈ 10,000 thousands km2 (also 303 consistent with (Tkach et al., 2008)) and φ differences of 58.9o – 39.5o latitudinal range the two 304 factors inside the parenthesis take the values 11.106 and 35.106 suggesting that latitudinal range 305 (i.e. here used as a proxy of environmental variability) exerts on average a 48% stronger (i.e. the 306 resulting variances for genetic and environmental variability, if that the other factor was zero, 307 would be 0.33 and 0.59, respectively, giving relative proportions of 36% and 64%) influence on 308 variance in logRR of BAFs than distribution (i.e. here genetic variability) does. Genetic diversity 309 and latitudinal range, both induce strong changes on the variance in BAFs and the respective 310 standard deviations were multifold higher for species with extensive distributions (Fig. 2c) and 311 high latitudes (Fig. 6b). 312 313 314 Discussion 315 316 Living organisms are constantly challenged to optimally allocate their finite resources to 317 maximise fitness. This challenge leads to multiple investment trade-offs, many of which have 318 been extensively studied (Poorter et al., 2006; Cadotte, 2007; Huot et al., 2014). BAFs represent 319 direct expressions of some of these trade-offs and the phenotypes of the organisms. We 320 experimented with two such BAFs, foliage over woody aboveground biomass and the root:shoot 321 ratio, to show that both these fractions were more variable in tree species with extensive 322 distributions, even after limiting our analysis to neighbouring trees (Fig. 2c), in agreement with 323 hypothesis two. Our results were also valid after correcting for phylogenetic relatedness 324 (Appendix 3). A higher ability to adapt to a changing environment is one of the implications of 325 higher trait variance (Bolnick et al., 2011). The causality of this relationship is unclear and could 326 imply either that tree species that experience a higher phenotypic variability tend to have larger 327 distributions (e.g. van Vallen, 1973) or that some species are phenotypically more diverse in 328 response to a larger distribution and most likely a larger effective population size. Environmental 329 heterogeneity has been identified as a mechanism that facilitates genetic variation in plants 330 (Delph & Kelly, 2013), and a larger distribution implies a higher environmental heterogeneity. 331 Many tree species have extensive distributions, but individual trees can also disperse over very 332 large distances (Bacles et al., 2006; Petit & Hampe, 2006; Kremer et al., 2012) and cross-fertilise 333 with individuals experiencing differing environmental conditions. We thus favour the 334 interpretation that a larger distribution most likely induces a higher variance in BAFs. 335 BAFs in plants represent expressions of an equilibrium between stabilising 336 intrinsic/genetic factors such as ontogeny and the destabilising influence of the environment 337 (Barthélémy & Caraglio, 2007), i.e. any biomass-allocation fraction is an aggregate of these two 338 mechanisms. We thus present an argument detailing why our results were likely due to a higher 339 genetic variability of trees with more extensive distributions and not because of the 340 environmental conditions that the trees experienced (in support of hypothesis two and against 341 hypothesis one). The influence of the environment is expected to be only moderately important 342 at sites close to the centres of species distributions (compared to the edges). Most of the trees in 343 our data sets had been harvested near the centres of their distributions, because moderate 344 environmental conditions facilitate silvicultural practices. Most importantly, our results were 345 valid when using individual sites per species in which case the differences were most likely due 346 to rapid evolution (i.e. the absence of systematic differences between angiosperms and 347 gymnosperms provided evidence against hypothesis three). Co-occurring tree species experience 348 comparable environmental conditions, so the resulting phenotypic variance should best represent 349 the genetic diversity (Fig. 2c). 350 Does latitude influence BAFs? We found no relationship in the Biomass-tree data set but 351 there was a strong positive relationship in the Biomass-plot data set (Fig. 6). A reason why the 352 results across the two data sets were incongruent probably relates to the way these were 353 standardized. The Biomass-tree data set contained multiple tree records per sampling site and 354 there were few replicate sites per tree species. Co-occurring woody species are likely to 355 experience comparable environmental conditions and the resulting BAFs should mainly reflect 356 genetic diversity. Because of the few replicate sites per tree species any influences of latitude 357 were masked in the data set by those of genetic diversity. This was even more pronounced when 358 we narrowed the analysis to a single site per tree species. By contrast in the Biomass-plot data set 359 via being limited to a single record per sampling site, we could effectively investigate differences 360 arising from environmental variability. Based on this analysis, trees that typically occur closer to 361 the poles exhibit higher variances of BAFs. Even though it has been disputed in the past 362 (Vázquez & Stevens, 2004), it is widely appreciated that environmental variability increases with 363 latitudinal range (MacArthur, 1972). This result is in accordance with our first hypothesis that 364 environmental variability should increase BAFs variance (Fig. 1a). The effect size we estimated 365 for environmental variability exceeded that for genetic diversity. 366 The survival challenges of tree species to the accelerating pace of global change is a key 367 topic in the biology of global change (Lenoir et al., 2008; Bertrand et al., 2016; Veresoglou & 368 Halley, 2018). Identifying tree species at a high risk of extinction is important. Several traits such 369 as longevity (Morin & Thuiller, 2009) and seed size (Veresoglou & Halley, 2018) might be 370 informative in terms of tree susceptibility to extinction. The loss of habitat is a key factor that 371 drives the eventual risk of extinction, but the relationship between habitat size and genetic 372 variability is poorly understood. The loss of genetic diversity following habitat loss can further 373 limit the ability of a species to cope with environmental conditions (Sexton et al., 2009) and 374 eventually accelerate extinction. The implications of extensive distributions have been debated. 375 A review of 31 studies by Lowe et al. (Lowe et al., 2005) found that habitat loss usually did not 376 significantly affect the genetic variability of tree species. Another meta-analysis by Vranckx et 377 al. (2012), however, found that habitat loss induced losses in the genetic diversity of species. An 378 extensive synthesis by Morueta-Holme et al. (2013) reported that plant species with extensive 379 geographic ranges were more genetically variable, and Kremer et al. (Kremer et al., 2012) 380 argued that long-distance gene flow amongst trees likely conferred an evolutionary advantage. In 381 contrast, the implications of habitat loss in trees may be fully reversible if the former habitat of 382 these species can be restored before extinction (Newmark et al., 2017). 383 Our results suggest that it is tree genetic diversity that induces variance in BAFs (Fig. 1b; 384 we found higher variance in BAFs in species with large distributions). Our analysis makes the 385 assumption that genetic variability is a good proxy of genetic diversity, which despite being a 386 common and well supported assumption in the literature (e.g. Avolio et al., 2012; Jöqvist and 387 Kremp, 2016), remains less robust than direct estimates of genetic diversity. Most importantly, 388 our general models did not discriminate between native and non-native plant species and used 389 real distributions to assess effect sizes. We often observe that the distributions of non-native (i.e. 390 invasive) species are not at equilibrium with their environment and that they possess a lower than 391 expected genetic diversity (Beaumont et al., 2009; Bradley et al., 2010). We would have 392 expected, as a result, relatively weak relationships between genetic diversity and realized 393 distributions for the subset of non-native trees which was not the case (Fig. S4). This finding 394 implies that changes in genetic diversity can occur quickly and develop after a few generations 395 of growth in isolation. Genetic diversity should thus play a key role in regulating the response of 396 species to future climate change, because of the extensive fitness implications of BAFs on the 397 ability of a species to adapt. The loss of habitat, even if transient, could lead to a loss of genetic 398 diversity, which would hinder species survival. 399 Human activity has sped up gene flow in almost all types of ecosystems and this should 400 continue in the near future (Wilson et al. 2009). On the short term, assisted gene flow 401 homogenizes populations (i.e. and their genetic diversity), allowing species to more effectively 402 cope with unfavourable environmental conditions (e.g. through acquiring more favourable 403 BAFs). Assisted gene flow should thus steepen the positive relationship between variance in 404 BAFs and distribution (Fig. 1b), benefiting disproportionally species with a large distribution at 405 the risk of species that maintain a limited genetic diversity. It has been documented, for example, 406 that invasions, which represent an alternative form of introductions of species with a high 407 competitive ability, induce extinctions of native species (Colautti et al. 2017; Catford et al. 408 2018). Even though we did not directly model assisted gene flow (as for example in Adams et al. 409 1998), the strong relationship between distribution and genetic diversity should be instructive for 410 forest management. Silvicultural practices such as the artificial regeneration of stands from 411 commercial genetic material (e.g. Bradshaw 2004; Finkeldey Ziehe 2004) might, therefore, be 412 precarious for the native species diversity. 413 In summary, BAFs were more variable in trees with extensive distributions, i.e. our 414 results supported only the second of our hypotheses. Most notably, we made the point that it was 415 a higher genetic variability that resulted in more variable BAFs for tree species with extensive 416 distributions. We thus present evidence that the loss of habitat for tree species through a quick 417 loss of genetic diversity could lower the ability of a species to modify its architecture (BAFs) in 418 response to changes in the environment (Fig. 1). It is thus likely that any loss of habitat may not 419 be as reversible as many believe. 420 421 Acknowledgements 422 We want to thank two anonymous reviewers and the editor for constructive criticism on our 423 paper. JP was funded by the European Research Council Synergy grant ERC-SyG-2013-610028 424 IMBALANCE-P. 425 426 Contributions 427 Conceived the study and carried out the analysis: SDV. The two authors wrote together the 428 article and approved the final version. 429 430 431 Data availability 432 All the data we analyse here are already in the public domain. We present intermediate data in 433 our analysis in the form of Appendix 8. 434 435 436 Competing interests 437 The authors declare no conflicts of interest. 438 439 Supplementary Information 440 Fig. S1: Sensitivity analysis on the inclusion criterion of five full records 441 Fig. S2:Relationships between the variance in BAFs in records from Europe 442 Fig. S3: Weighted regressions between the variance in BAFs and the distribution of the species 443 without correcting for age 444 Fig. S4: Senstivity analysis on how relatioships differed between native and invasive tree species 445 Table S1: List of the plant species that we used in our analyses. 446 Table S2: Overview of all analyses. 447 Table S3: : Classification of tree species into native vs invasive 448 Appendix 1: Inclusion Criteria 449 Appendix 2: Effect sizes and model specifications 450 Appendix 3: Strength of BAF relationships 451 Appendix 4: Partitioning of BAFs into environmental and genetic variability 452 Appendix 5: Biases due to differences in coverage of the databases 453 Appendix 6: Phylogenetic tree 454 Appendix 7: Supplementary references 455 Appendix 8: Raw data 456 457 458 References 459 460 Adams WT, Zuo J, Shimizu JY, Tappeiner JC. 1998. Impact of alternative regeneration 461 methods on genetic diversity in coastal Douglas-Fir. Forest Science 44: 390-396. 462 Aerts R, van Bodegom PM, Cornelissen JHC. 2012. Litter stoichiometric traits of plant 463 species of high-latitude ecosystems show high responsiveness to global change without 464 causing strong variation in litter decomposition. New Phytologist 196: 181-188. 465 Avolio M, Beaulieu JM, Lo EYY, Smith MD. 2012. Measuring genetic diversity in ecological 466 studies. Plant Ecology 213, 1105-115. 467 Bacles CFE, Lowe AJ, Ennos RA. 2006. Effective seed dispersal across a fragmented 468 landscape. Science 311: 628. 469 Barthélémy D, Caraglio Y. 2007. Plant architecture: A dynamic, multilevel and comprehensive 470 approach to plant form, structure and ontogeny. Annals of Botany 99: 375-407. 471 Beaumont LJ, Gallagher RV, Thuiller W, Downey PO, Leishman MR, Hughes L. 2009. 472 Different climatic envelopes among invasive populations may lead to underestimations of 473 current and future biological invasions. Diversity and Distributions 15: 409-420. 474 Bertrand R, Riofrío-Dillon G, Lenoir J, Drapier J, de Ruffray P, Géqout JC, Loreau M. 475 2016. Ecological constraints increase the climatic debt in forests. Nature 476 Communications 7: 12643. 477 Bolnick DI, Amarasekare P, Araújo MS, Bürger R, Levine JM, Novak M, Rudolf VHW, 478 Schreiber SJ, Urban MC, Vasseur DA. 2011. Why intraspecific trait variation matters 479 in community ecology. Trends in Ecology and Evolution 26: 183-192. 480 Bowman DMJS, Brienen RJW, Gloor E, Phillips OL, Prior LD. 2013. Detecting trends in 481 tree growth: Not so simple. Trends in Plant Science 18: 11-17. 482 Bradley BA, Blumenthal DM, Wilcove DS, Ziska LH. 2010. Predicting plant invasions in an 483 era of global change. Trends in Ecology and Evolution 25: 310-318. 484 Bradshaw RHW. 2004. Past anthropogenic influence on European forests and some possible 485 genetic consequences. Forest Ecology and Management 197: 203-212. 486 Cadotte MW. 2007. Competition-colonization trade-offs and disturbance effects at multiple 487 scales. Ecology 88: 823-829. 488 Catford JA, Bode M, Tilman D. 2018. Introduced species that overcome life history tradeoffs 489 can cause native extinctions. Nature Communications 9: 2131. 490 Chave J, Andalo C, Brown S, Cairns MA, Chambers JQ, Eamus D, Folster H, Fromard F, 491 Higuchi N, Kira T, Lescure JP, Nelson BW, Ogawa H, Puig H, Riera B, Yamakura 492 T. 2005. Tree allometry and improved estimation of carbon stocks and balance in tropical 493 forests. Oecologia 145: 87-99. 494 Colautti RI, Alexander JM, Dlugosch KM, Keller SR, Sultan SE. 2017. Invasions and 495 extinctions through the looking glass of evolutionary ecology. Philosophical 496 Transactions of the Royal Society B 372: 20160031. 497 de Frenne P, Graar BJ, Rodriguez-Sanchez F, Kolb A, Chabrerie O, Decocq G, de Kort H, 498 de Schrijver A, Diekmann M, Eriksson O, Gruwez R, Hermy M, Leonoir J, Plue J, 499 Coomes DA, Verheyen K 2013 Latitudinal gradients as natural laboratories to infer 500 species’ responses to temperature. Journal of Ecology 101: 784-795. 501 Dean TJ, Roberts SD, Gilmore DW, Maguire DA, Long JN, O'Hara KL, Seymour RS. 502 2002. An evaluation of the uniform stress hypothesis based on stem geometry in selected 503 north american conifers. Trees 16: 559-568. 504 Delph LF, Kelly JK. 2013. On the importance of balancing selection in plants. New Phytologist 505 201: 45-56. 506 Duursma R, Mäkelä A, Reid DEB, Jokela EJ, Porté AJ, Roberts SD. 2010. Self-shading 507 affects allometric scaling in trees. Functional Ecology 24: 723-730. 508 Finkeldey R, Ziehe M. 2004. Genetic implications of silvicultural regimes. Forest Ecology and 509 Management 197: 231-244. 510 Forrester DI, Benneter A, Bouriaud O, Bauhus J. 2017a. Diversity and competition influence 511 tree allometric relationships – developing functions for mixed-species forests. Journal of 512 Ecology 105: 761-774. 513 Forrester DI, Tachauer IHH, Annighoefer P, Barbeito I, Pretzsch H, Ruiz-Peinado R, 514 Stark H, Vacchiano G, Zlatanov T, Chakraborty T, Saha S, Sileshi GW. 2017b. 515 Generalized biomass and leaf area allometric equations for european tree species 516 incorporating stand structure, tree age and climate. Forest Ecology and Management 396: 517 160-175. 518 Forster MA, Ladd B, Bonser SP. 2011. Optimal allocation of resources in response to shading 519 and neighbours in the heteroblastic species, acacia implexa. Annals of Botany 107: 219- 520 228. 521 Heibo E, Magnhagen C, Vøllestad LA. 2005. Latitudinal variation in life-history traits in 522 eurasian perch. Ecology 86: 3377-3386. 523 Hodge A. 2004. The plastic plant: Root responses to heterogenous supplies of nutrients. . New 524 Phytologist 162: 9-24. 525 Huot B, Yao J, Montgomery BL, He SY. 2014. Growth–defense tradeoffs in plants: A 526 balancing act to optimize fitness. Molecular Plant 7: 1267-1287. 527 Jöqvist CO, Kremp A. 2016. Genetic diversity affects ecological performance and stress 528 response of marine diatom populations. The ISME Journal 10, 2755-2766. 529 Kokko H. 2007. Modelling for field biologists and other interesting people. London: Cambridge 530 University Press. 531 Kremer A, Ronce O, Robledo-Arnuncio JJ, Guillaume F, Bohrer G, Nathan R, Bridle JR, 532 Gomulkiewicz R, Klein EK, Ritland K, Kuparinen A, Gerber S, Schueler S. 2012. 533 Long-distance gene flow and adaptation of forest trees to rapid climate change. Ecology 534 Letters 15: 378-392. 535 Leites LP, Rehfeldt GE, Robinson AP, Crookston NL, Jaquish B. 2012. Posibilities and 536 limitations of using historic provenance tests to infer forest species growth responses to 537 climate change. Natural Resource Modeling 25: 409-433. 538 Lenoir J, Gégout JC, Marquet PA, de Ruffray P, Brisse H. 2008. A significant upward shift 539 in plant species optimum elevation during the 20th century. Science 320: 1768-1771. 540 Lowe AJ, Boschler D, Ward M, Bacles CFE, Navarro C. 2005. Genetic resource impacts of 541 habitat loss and degradation; reconciling empirical evidence and predicted theory for 542 neotropical trees. Heredity 95: 255-272. 543 Lusk CH, Falser DS, Jara-Vergara CK, Jimenez-Castillo M, Saldana-Meñdoza A. 2008. 544 Ontogenetic variation in light requirements of juvenile rainforest evergreens. Functional 545 Ecology 22: 454-459. 546 MacArthur RH. 1972. Geographical ecology. Princeton, N.J: Princeton University Press. 547 Mauri A, Stronna G, San Miguel Yanz J. 2017. Data descriptor: Eu-forest, a high-resolution 548 tree occurrence dataset for europe. Scientific Data 4: 160123. 549 McCarthy MC, Enquist BJ. 2007. Consistency between an allometric approach and optimal 550 partitioning theory in global patterns of plant biomass allocation. Functional Ecology 21: 551 713-720. 552 Messier J, McGill BJ, Lechowicz MJ. 2010. How do traits vary across ecological scales? A 553 case for trait-based ecology. Ecology Letters 13: 838-848. 554 Mogan J, Gannell MG. 1994. Shape of tree stems-a re-examination of the uniform stress 555 hypothesis. Tree Physiology 14: 49-62. 556 Morin X, Thuiller W. 2009. Comparing niche- and process-based models to reduce prediction 557 uncertainty in species range shifts under climate change. Ecology Letters 90: 1301-1313. 558 Morueta-Holme N, Enquist BJ, Boyle B, Jørgensen PM, Ott JE, Peet RK, Símová I, Sloat 559 LL, Thiers B, Violle C, Wiser SK, Dolins S, Donoghue II JC, Kraft NJB, Regetz J, 560 Schildhauer M, Spencer N, Svenning JC. 2013. Habitat area and climate stability 561 determine geographicalvariation in plant species range sizes. Ecology Letters 16: 1446- 562 1454. 563 Müller I, Schmid B, Weiner J. 2000. The effect of nutrient availability on biomass allocation 564 patterns in 27 species of herbaceous plants. Perspectives in Plant Ecology, Evolution and 565 Systematics 3: 115-127. 566 Muukkonen P. 2007. Generalized allometric volume and biomass equations for some tree 567 species in europe. European Journal of Forest Research 126: 157-166. 568 Newmark WD, Jenkins CN, Pimm SL, McNeally PB, Halley JM. 2017. Targeted habitat 569 restoration can reduce extinction rates in fragmented forests. Proceedings of the National 570 Academy of Sciences USA 114: 9635-9640. 571 Nikklas KJ. 1997. Mechanical properties of black locust ( robinia pseudoacacia l.) wood. Size- 572 and age-dependent variations in sap- and heartwood. Annals of Botany 79: 265-272. 573 Nikklas KJ. 2004. Plant allometry: Is there a grand unifying theory? Biological Reviews 79: 574 871-889. 575 Nikklas KJ, Midgley JJ, Rand RH. 2003. Tree size frequency distributions, plant density, age 576 and community disturbance. Ecology Letters 6: 405-411. 577 Petit RJ, Hampe A. 2006. Some evolutionary consequences of being a tree. Annual Review in 578 Plant Ecology, Evolution and Systematics 37: 187-214. 579 Poorter H, Niklas KJ, Reich PB, Oleksyn J, Poot P, Mommer L. 2012. Biomass allocation to 580 leaves, stems and roots: Meta-analyses of interspecific variation and environmental 581 control. New Phytologist 193: 30-50. 582 Poorter H, Sack L. 2012. Pitfalls and possibilities in the analysis of biomass allocation patterns 583 in plants. Frontiers in Plant Science 3: 259. 584 Poorter L, Bongers L, Bongers F. 2006. Architecture of 54 moist-forest tree species: Traits, 585 trade-offs, and functional groups. Ecology 87: 1289-1301. 586 Radtke P, Walker DM, Weiskiitel AR, Frank J, J.W. C, Westfall JA 2016. Legacy tree data: 587 A national database of detailed tree measurements for volume, weight, and physical 588 properties.In S. M. StantonG. Christensen. Pushing Boundaries: New Directions in 589 Inventory Techniques & Applications; Forest Inventory and Analysis Symposium 2015. 590 General Technical Report PNW-GTR-931. . USDA Forest Service, Pacific Northwest 591 Research Station, Portland. 25-30. 592 Reich PB, Luo Y, Bradford JB, Poorter H, Perry CH, Oleksyn J. 2014. Temperature drives 593 global patterns in forest biomass distribution in leaves, stems, and roots. Proceedings of 594 the National Academy of Sciences USA 38: 13721-13726. 595 Reich PB, Tjoelker MG, Machado JL, Oleksyn J. 2005. Universal scaling of respiratory 596 metabolism, size and nitrogen in plants. Nature 439: 457-461. 597 Reich PB, Wright IJ, Bares JC, Craine JM, Oleksyn J, Westoby M, Walters MB. 2003. The 598 evolution of plant functional variation: Traits, spectra, and startegies. International 599 Journal of Plant Sciences 164: S3. 600 Remeš V. 2000. How can maladaptive habitat choice generate source-sink population dynamics? 601 Oikos 91: 579-582. 602 Reyer CPO, Leuzinger S, Rammig A, Wolf A, Bartholomeus RP, Bonfante A, de Lorenzi F, 603 Dury M, Gloning P, Aboujaoudé R, Klein T, Kuster TM, Martins M, Niedrist G, 604 Riccardi M, Wohlfart G, de Angelis P, de Dato G, Francois L, Menzel A, Pereira M. 605 2013. A plant’s perspective of extremes: Terrestrial plant responses to changing climatic 606 variability. Global Change Biology 19: 75-89. 607 Schepaschenko D, Schvidenko A, Usoltsev V, Lakyda P, Luo Y, Vasylyshyn R, Lakyda I, 608 Myklush Y, See L, McCallum I, Fritz S, Kraxner F, Obersteiner M. 2017. Data 609 descriptor: A dataset of forest biomass structure for eurasia. Scientific Data 4: 170070. 610 Sexton JP, McIntyre PJ, Angert AL, Rice KJ. 2009. Evolution and ecology of species range 611 limits. Annual Review in Plant Ecology, Evolution and Systematics 40: 415-436. 612 Stapley J, Feulner PGD, Johnston SE, Santure AW, Smadja CM. 2017. Recombination: The 613 good, the bad and the variable. Philosophical Transactions of the Royal Society B 372: 614 20170279. 615 Thomson AM, Parker WH. 2008. Boreal forest provenance tests used to predict optimal 616 growth and response to climatic change. 1. Jack pine. Canadian Journal of Forest 617 Research 38: 157-170. 618 Tkach NV, Röser M, Hoffman MH. 2008. Range size variation and diversity distribution in the 619 vascular plant flora of the eurasian arctic. Organisms Diversity & Evolution 8: 251-266. 620 USDA 2017. The plants database.In: National Plant Data Team, Greensboro, NC 27401-4901, 621 USA. 622 van Vallen L. 1973. A new evolutionary theory. Evolutionary Theory 1: 1-30. 623 Vázquez DP, Stevens RD. 2004. The latitudinal gradient in niche breadth: Concepts and 624 evidence. The American Naturalist 164: E1-E19. 625 Veresoglou SD, Halley JM. 2018. Seed mass predicts migration lag of european trees. Annals of 626 Forest Science 75: 86. 627 Veresoglou SD, Wang D, Andrade-Linares DR, Hempel S, Rillig MC. 2017. Fungal decision 628 to exploit or explore depends on growth rate. Microbial Ecology 75:289-292. 629 Violle C, Enquist BJ, McGill BJ, Jiang L, Albert CH, Hulshof C, Jung V, Messier J. 2012. 630 The return of the variance: Intraspecific variability in community ecology. Trends in 631 Ecology and Evolution 27: 244-252. 632 Vranckx G, Jacquemyn H, Muys B, Honnay O. 2012. Meta-analysis of susceptibility of 633 woody plants to loss of genetic diversity through habitat fragmentation. Conservation 634 Biology 26: 228-237. 635 Wang T, O'Neill GA, Aitken SN. 2010. Integrating environmental and genetic effects to predict 636 responses of tree populations to climate. Ecological Applications 20: 153-163. 637 Wilson JRU, Dormontt EE, Prentis PJ, Lowe AJ, Richardson DM. 2009. Something in the 638 way you move: dispersal pathways affect invasion success. Trends in Ecology and 639 Evolution 24: 136-144. 640 Wyka TP, Oleksyn J, Karolewski P, Schnitzer SA. 2013. Phenotypic correlates of the 641 lianescent growth form: A review. Annals of Botany 112: 1667-1681. 642 643 Legends to Figures 644 645 Fig. 1 Conceptual diagram illustrating the three hypotheses we address in the manuscript. Four 646 hypothetical species, the broadleaves blue triangle and yellow circle (represented with an 647 elliptical leaf) and the conifers red rhomb and green square (represented by an acicular 648 leaf) are each sampled at four locations (map on top; note that distribution envelopes 649 differ). The ranking of their distributions is as follows: red rhomb, blue triangle, yellow 650 circle and green square. In Hypothesis One (a) we expect that the fraction of variance in 651 biomass allocation fractions that is explained by genetic variability (purple discontinuous 652 lines) is independent of distribution and that environmental factors increase the variance 653 (green arrow) more in those species that have a small distribution (larger arrow) resulting 654 in a negative relationship between observed variance and distribution (green 655 discontinuous lines). In Hypothesis Two (b) we expect that a larger distribution results in 656 a higher genetic diversity which is depicted with a purple line. The environment increases 657 variance irrespective of distribution (green arrows) resulting in a positive relationship 658 between observed variance and distribution (green discontinuous lines). In Hypothesis 659 Three (c) the biomass allocation fractions vary independently of distribution and can be 660 explained by the evolutionary history of the plant (here angiosperms vs gymnosperms). 661 We could assess genetic variance in BAFs by comparing conspecific trees in the same 662 stand after correcting for age-differences and the sum of genetic and environmental 663 variability by comparing across stands, also after correcting for age. 664 665 666 Fig. 2 Relationships between the variance in biomass-allocation fractions (foliage over woody 667 aboveground biomass) and distribution of the tree species. (a) Biomass-tree data set, all 668 possible observations (multiple trees per location) of the specific data set (lower than in 669 Biomass-plot); (b) Biomass-plot data set, one observation per location (at a plot level); (c) 670 Biomass-tree data set, observations per species only describe trees in the stand (plot) that 671 contained the most tree observations. The dashed lines represent the best fits. Numbers 672 next to the data points indicate the number of observations per species used to calculate 673 the variance. Overlaid map shows in red the location of the sites where the data 674 originated from – we only analysed woody species with an extensive distribution (over 2 675 million km2) in Europe. Phylogenetic correction was not applied. Relationships were 676 stronger when we repeated the analyses with the subset of sites located in Europe (Fig. 677 S2). 678 679 680 681 682 Fig. 3 Relationships between variance in biomass-allocation fractions (foliage over woody 683 aboveground biomass) and distribution of the tree species for phylogenetically corrected 684 data using phylogenetically independent contrasts (PIC). (a) Biomass-tree data set, all 685 possible observations; (b) Biomass-plot data set, one observation per location; (c) 686 Biomass-tree data set, observations per species are from the stand that contained most 687 observations. The dashed red lines represent the best fits. The overlaid phylogenetic tree 688 depicts in the form of squares information (blue: first quartile; white middle two 689 quartiles; red 4th quartile) on variance in biomass allocation fractions (left) and 690 distribution (right) of the woody species included in the analysis and their phylogenetic 691 relationships (tree). Note the absence of phylogenetic signal which was confirmed with 692 Blomberg K tests. 693 694 Fig. 4 Differences in variance in biomass-allocation fractions (foliage over woody aboveground 695 biomass) between angiosperms (elliptical leaf in yellow) and gymnosperms (acicular leaf 696 in green). (a) Biomass-tree data set, all possible observations; (b) Biomass-plot data set, 697 one observation per location; (c) Biomass-tree data set, observations per species are from 698 the stand that contained most observations. None of the t-tests were significant. 699 700 701 Fig. 5: (a; b) Relationships between the variance in biomass-allocation fractions (root over shoot 702 fractions) and distribution of the tree species. (a) Biomass-tree data set, all possible 703 observations; (b) Biomass-tree data set, observations per species are from the stand that 704 contained most observations; (c) variance in biomass-allocation ratios for Pinus echinata 705 (observed in 23 states), Pinus taeda (observed in 20 states, from two studies) and Pinus 706 palustris (observed in 10 states) across four studies in the Legacy Tree Data database 707 (USA). The three species have nested distributions in America. The dashed lines 708 represent the best fits. Numbers next to the data points indicate the number of 709 observations per species used to calculate the variance. Overlaid map shows in red the 710 location of the sites where the data originated from. 711 712 713 Fig. 6 Influence of latitude on variance in biomass-allocation fractions (foliage over woody 714 biomass). The x-axis describes the average latitude of the distribution range of the woody 715 species whereas the y-axis the variance in biomass-allocation fractions. Panel (a) 716 describes the Biomass-tree data set whereas panel (b) the Biomass-plot data set. The 717 insert in panel (a) shows how variance differs in the north and south distribution of five 718 tree species in the Biomass-tree data set (opaque triangles). North here describes latitudes 719 in Europe higher than 58.5°N (mean of the latitudinal extremes of European territory) and 720 South lower than this value. These five species were the only ones that met the inclusion 721 criterion, namely being represented with a minimum of five entries both in North and 722 South Europe. The continuous opaque line is an isocline where variance in the North 723 equals that in the South. 724 725 726 Variance in biomass allocation fractions (a) ● Distribution +genetic environment environment ● ● Variance in biomass allocation fractions ● (b) ● ● Distribution Variance in biomass allocation fractions (c) ● ● (a) tau=0.48, P =0.03 (b) tau=0.34, P =0.028 (c) tau=0.51, P =0.047 ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ●●●●● ● ● ● ● ● 20 ● ● ●● ● ● ● ● ● ● ● ●●● ● ● ● ● ● ● ● ● ●● ● ●● 0.8 0.8 0.8 ●●● ● ●● ● ● ●● ● ● ● ● ● ● ●● ●● ●●● ●● ●● ● ●●● ● ●● ● ●● ●● ● ● ● ● ● ● ● ● ● ● ● ●● ● ●●●● ●●● ● ● ●●● ●●●●● ● ● ● ● ● ● ● ● ● ● ●●● ●● ●●● ●●●● ● ●● ● ● ● ● ● ● ● ●●●● ● ● ●● ● ● ● ● ●●●● ● ●● ● ●●●● ● ●● ● ● ●● ● ●● ● ●● ●● ● ● ●● ● ● ● ● ● ●● ●● ●● ●● ● ● ●● ●● ● ● ● ● ●●● ● ● ●●● ● ●● ●● ● ● ●●●●● ● ●●● ● ●● ● ● ●● ● ●●●●●● ● ●● ● ●● ●●●●●● ● ● ● ● ●●● ● ●● ●● ● ● ● ● ●● ●●● ● ● ●● ● Variance in biomass−allocation franctions Variance in biomass−allocation franctions Variance in biomass−allocation franctions ● 42 ● 1372 0.6 0.6 0.6 ● ● 1677 259 ● ● 20 ● ● 356 ● 370 30 263 ● ● 18 53 ● 915 ● ● ● ● 10 9 247 ● 142 59 ● ●310 ● 0.4 0.4 0.4 201 142 ● ● ● 45 14 111 ● ● ● 13 ● ● 35 35 32540 ● ● 39 23 ● ● 18 24 ● ● ● 44 7 6 ● ● ● 17 17 0.2 0.2 0.2 15 ● 13 ● ● 7 ● 7 7 ● 4 6 8 10 4 6 6 8 10 4 6 8 10 2 2 2 Distribution (millions km ) Distribution (millions km ) Distribution (millions km ) (a) (b) (c) r=0.75, P =0.008 r=0.28, P =0.24 r=0.61, P =0.08 ● 0.04 0.04 0.04 ● ● PIC Variance in biomass allocation franctions ● 0.02 0.02 0.02 ● ● ● ● ● ● ● ● ● ● ● ● ● 0.00 0.00 0.00 ● ● ● ● ● ● ● ● ●●● ● ● ● ● ● ● −0.02 −0.02 −0.02 ● ● ● ● ● −0.04 −0.04 −0.04 −0.4 −0.2 0.0 0.2 −0.4 −0.2 0.0 0.2 −0.4 −0.2 0.0 0.2 2 2 PIC Distribution (millions km ) PIC Distribution (millions km ) PIC Distribution (millions km2 ) Variance in biomass−allocation franctions 0.2 0.4 0.6 0.8 (a) ● ● ● ● ● ● ● ● ● ● ● ●● ●● ●●● ●● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ●● ● ● ● Variance in biomass−allocation franctions 0.2 0.4 0.6 0.8 (b) ● ● ● ● ● ● ● ●● ●● ● ● ● ●● ●● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ●● ●● ● ●● ● ● ● ● ●● ●● ● ● ● Variance in biomass−allocation franctions 0.2 0.4 0.6 0.8 (c) ● ● ● ● ● ● ● ● ● ●● ●● ● ● ● ●● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● (a) r=0.8, P =0.11 (b) r=0.87, P =0.05 (c) tau=0.91, P =0.07 0.7 0.8 0.8 Variance in biomass−allocation fractions (not corrected for age) ● ● ●● ●● ● ● 0.6 ● Variance in biomass−allocation franctions Variance in biomass−allocation franctions ● ●● ● ● 111 73 0.6 0.6 ● 162 0.5 48 ● ● ● 414 20 ● 35 0.4 0.4 0.4 ● 27 0.3 ● 198 ● ● 18 18 ● 16 0.2 0.2 ● 16 ● 30 0.2 4 6 8 10 4 6 8 10 10 15 20 25 2 2 Distribution (millions km ) Distribution (millions km ) Number of states Variance in biomass−allocation franctions (a) 0.2 0.3 0.4 0.5 0.6 0.7 0.8 South ● ● 0.0 0.2 0.4 0.6 0.8 1.0 48 50 North 52 ● ●0.0 0.2 0.4 0.6 0.8 1.0 Mean latitude 54 r=0.48, P =0.19 ● 56 ● Variance in biomass−allocation franctions (b) 0.2 0.3 0.4 0.5 0.6 0.7 0.8 ● ● 44 ● ● 46 ● ● ● 48 ● 50 ● Mean latitude ● 52 54 ● r=0.72, P <0.001 ● ● 56 ●