This is the accepted version of the article: Peñuelas, Josep; Fernández Martínez, Marcos; Ciais, Philippe; [et al.]. «The bioelements, the elementome and the “biogeochemical niche”». Ecology, Vol. 100, Issue 5 (May 2019), art. e02652. DOI 10.1002/ecy.2652 This version is avaible at https://ddd.uab.cat/record/204025 under the terms of the license 1 The bioelements, the elementome and the “biogeochemical niche”. 2 Josep Peñuelas1,2*, Marcos Fernández-Martínez2,3, Philippe Ciais4, David Jou5, Shilong Piao6, 3 Michael Obersteiner7, Sara Vicca3, Ivan A. Janssens3, Jordi Sardans1,2 4 1 5 CSIC, Global Ecology Unit CREAF-CSIC-UAB, Bellaterra, 08193 Catalonia, Spain. 2 6 CREAF, Cerdanyola del Valles, 08193 Catalonia, Spain. 3 7 Research Group Plants and Ecosystems (PLECO), Department of Biology, University of 8 Antwerp, B-2610 Wilrijk, Belgium. 9 4 Laboratoire des Sciences du Climat et de l’Environnement, IPSL, 91191 Gif-sur-Yvette, 10 France. 5 11 Department of Physics. Universitat Autònoma de Barcelona, Bellaterra, 08193 Catalonia, 12 Spain. 6 13 Sino-French Institute for Earth System Science, College of Urban and Environmental Sciences, 14 Peking University, Beijing 100871, China. 7 15 International Institute for Applied Systems Analysis (IIASA), Ecosystems Services and 16 Management, Schlossplatz 1, A-2361 Laxenburg, Austria. 17 * Corresponding author: 18 Josep Peñuelas, CSIC, Global Ecology Unit CREAF-CSIC-UAB, Bellaterra, 08193 Catalonia, Spain 19 and CREAF, Cerdanyola del Valles, 08193 Catalonia, Spain. E-mail: josep.penuelas@uab.cat 20 21 Running Head: Elementome and biogeochemical niche. 22 23 24 1 25 Keywords: 26 Bio-elements, elementome, biogeochemical niche, ecological niche, phylogenetic distance, 27 sympatry, plasticity, ecosystem function and organization, evolution. 28 29 30 31 32 Abstract 33 Every living creature on Earth is made of atoms of the various bioelements that are harnessed in 34 the construction of molecules, tissues, organisms and communities, as we know them. Organisms 35 need these bioelements in specific quantities and proportions to survive and grow. Distinct 36 species have different functions and life strategies, and have therefore developed distinct 37 structures and adopted a certain combination of metabolic and physiological processes. Each 38 species is thus also expected to have different requirements for each bioelement. We therefore 39 propose that a “biogeochemical niche” can be associated with the classical ecological niche of 40 each species. We show from field data examples that a biogeochemical niche is characterized by 41 a particular elementome defined as the content of all (or at least most) bioelements. The 42 differences in elementome among species are a function of taxonomy and phylogenetic distance, 43 sympatry (the bioelemental compositions should differ more among coexisting than among non- 44 coexisting species to avoid competitive pressure), and homeostasis with a continuum between 45 high homeostasis/low plasticity and low homeostasis/high plasticity. This proposed 46 biogeochemical niche hypothesis has the advantage relative to other associated theoretical niche 2 47 hypotheses that it can be easily characterized by actual quantification of a measurable trait: the 48 elementome of a given organism or a community, being potentially applicable across taxa and 49 habitats. The changes in bioelemental availability can determine genotypic selection and 50 therefore have a feedback on ecosystem function and organization, and, at the end, become 51 another driving factor of the evolution of life and the environment. 52 53 54 55 56 1. The elements of life and the construction of molecules, tissues, organisms and 57 communities 58 All living beings are made of atoms of various bioelements (elements used by living organisms). 59 The most common bioelements are: hydrogen (H) 59%, oxygen (O) 24%, carbon (C) 11%, 60 nitrogen (N) 4%, phosphorus (P) 1% and sulfur (S) 0.1-1% (percentages of total number of 61 atoms in organisms) (Slade 2006). C, H, and O are the basis of organismic structure, i.e. water 62 and primary structural molecules, whereas N, P and S enter in the various biomolecules allowing 63 biochemical reactions. C, H and O from CO2 and H2O are converted into the skeleton of all types 64 of organic molecules by photo- and chemosynthesis. The molecular structures that control life 65 processes (e.g. RUBISCO or chlorophyll), that carry information (e.g. DNA and RNA), and that 66 control the storage and release of energy when and where necessary (e.g. ATP and NAD) all 67 incorporate N, P and in lower amounts S. 3 68 These six bioelements, however, are not the only bioelements that have important roles in 69 organisms (Slade, 2006). Other bioelements, normally present in low concentrations such as 70 potassium (K), magnesium (Mg), iron (Fe), calcium (Ca), molybdenum (Mo), manganese (Mn) 71 and zinc (Zn) have functions that are more specific. They are cofactors in catalytic reactions in 72 several biochemical and physiological processes. For example, in plants, K, Mg, Fe, S, Ca, and 73 Zn are involved in photosynthesis, respiration, regulation of ion balance in chloroplasts and 74 vacuoles, transport of sugar into the phloem, secondary metabolism, maintenance of internal 75 cellular osmotic or pH equilibria and balances, and strategies for stomatal control and avoidance 76 of water stress (Knight et al. 1991, Bourgis et al. 1999, Shaul 2002, Tripler et al. 2006, Peñuelas 77 and Sardans 2009, Baribault et al. 2012, Naples and Fisk 2010, Rivas-Ubach et al. 2012, 78 Peñuelas et al. 2013, Sardans and Peñuelas 2015). All these bioelements thus contribute 79 essentially to organismic functions, albeit to different degrees (Figure 1). 80 The concentrations and ratios of C, N and P have been the basis of ecological 81 stoichiometry (Elser et al. 2000a,b, 2001, Sardans et al. 2015, 2016). However, to link 82 bioelemental composition and stoichiometry with ecological traits in all types of organisms and 83 communities will thus require information about bioelemental composition beyond the C:N:P 84 concentrations and ratios (Peñuelas and Sardans 2009). The other bioelements such as S, K, Ca 85 or Mg are also crucial and can be limiting in some ecosystems (Naples and Fisk 2010, Baribault 86 et al. 2012, Lapenis et al. 2013). These other bioelements have been introduced in ecological 87 stoichiometric studies to link the elemental composition of individuals and species with their 88 ecological traits, such as drought resistance (related to K), light environment (related to Mg) or 89 the levels of N and S deposition (related to K, Ca, Mg and S) (Sardans et al., 2011;2012a; 4 90 Sardans et al., 2015). All these bioelements should thus be considered in ecological 91 stoichiometric studies. 92 93 2. Elementome and biogeochemical niche. 94 Species use these bioelements to different degrees based on their ecological strategy (example 95 for plants in Figure 2). Classic Hutchinson’s (1957) description of species niche has been largely 96 applied to the main taxa from a point-of-view of distinct species in a site position in a “n- 97 dimensional hypervolume” formed by the values of determined environmental variables and/or 98 organisms traits. For instance, plant species have been classified in a hypervolume formed by 99 three axes, i.e. corresponding to Grime’s 3 strategies: competitor, stress-tolerant and ruderal 100 (Grime 1977). In animals, the species-specific niche has been determined and located in 101 function of hypervolumes for n-axis of different values of energy expenditure (Wilson et al., 102 2011), distinct diet sources (Araujo et al., 2007; Bearhop et al., 2004) or different elemental 103 composition (González et al., 2017, 2018). In microbial communities, genotype-niches have also 104 been described in function of the different availabilities of different nutrients (Pereira and Berry, 105 2017) or as a function of the values of diverse functional traits (Lennon et al., 2012) thus 106 describing multivariant spaces with a determined combination of nutrients availability or 107 functional traits where a determined genotype can exist. Any living species should thus tend to 108 optimize its overall function within a determined and space within this volume generated by a 109 combination of these axes describing strategies or biotic and/or abiotic variables while avoiding 110 the overlap with other coexisting species. Following these niche approach, we can hypothesize 111 that distinct species have different functions and life strategies and have therefore developed 112 distinct structures and adopted a certain combination of metabolic and physiological processes. 5 113 Therefore, each species (of all taxa types) also has different requirements for each bioelement, 114 and should thus tend to have its own elemental composition (Yu et al. 2011; Kattge et al. 2011; 115 Sardans and Peñuelas 2013, 2014; Sardans et al. 2015; Pierce et al. 2016; Bartrons et al., 2018). 116 We postulate that the more functionally separated two species are, the more different the 117 content of bioelements should be. This led us to propose a “biogeochemical niche” (BN) defined 118 as the multidimensional space defined by the concentrations of bioelements in individuals of a 119 given species (Peñuelas et al., 2008). This BN has been used in the recent years by other authors 120 such as Gonzalez et al. (2017) who proposed the “multidimensional stoichiometric niche” as the 121 specific position of different taxonomic/trophic groups in the three-dimensional space 122 determined by C, N, and P, and other elements as additional axes in their application of the BN 123 to animals (Gonzalez et al., 2018), as previously proposed for plants by Peñuelas et al. (2008), 124 and for plants and animals together by Bartrons et al (2018). BN is thus based on the species- 125 specific needs and use of bioelements in different amounts and proportions resulting from the 126 organismic structure and function in the given abiotic and biotic environmental space where the 127 species has evolved (Peñuelas et al. 2008, 2010). Each species should thus have an optimal (at 128 maximum fitness) equilibrium of bioelemental composition, resulting from its functional and 129 structural adaptations, equivalent to the optimum “elemental phenotype” (Jeyasingh et al., 2014). 130 The BN represents thus the species-specific strategies of growth and resource uptake and the 131 differences in soil-space occupation. 132 We propose that the processes underlying this species-specific elementome and the BN 133 are based on three basic complementary rules (Figures 3). First, distinct taxonomic groups 134 exhibit different elementomes, with larger differences as taxonomic distance and evolutionary 135 time increase. Second, at equilibrium, coexisting species tend to have distinct elementomes to 6 136 minimize competitive pressure. Third, trade-offs between adaptation to being competitive in a 137 stable environment versus being successful in more fluctuating environments cause homeostasis 138 and plasticity differences between species in a continuum of strategies. The BN distances among 139 species should thus be a function of taxonomic difference, sympatry (coexistence), and 140 homeostasis/plasticity and sympatry (Figure 3). The BN of taxa i is assumed to be the result of 141 its taxonomical evolutionary determination and its capacity to respond to changes in external 142 conditions, which is also partly taxonomically-determined. External changes include nutrient 143 availability (including the level of atmospheric CO2), changes in growing season duration, 144 disturbance regimes, and, on longer time scales, changes in soil properties. This capacity to 145 respond to environmental changes is, in turn, a function of taxonomy, sympatry, and 146 homeostasis: 147 BNi = BN of taxai = Elementomei1−n = (C: N: P: K: S: Mg: Ca: … )i1−n = BNtaxi + ∂BNtaxi 148 = BNtaxi + f(Taxi , Sympi , Homi ) (Eq. 1) ∂Env 149 where: 150 BNtaxi = Biogeochemical niche explained by taxonomy of taxa i (measured in n 151 individuals of taxa i). A n‐dimensional hypervolume describes the fundamental BN of a species. 152 The domain of this hypervolume function is restricted to a finite region corresponding to the 153 possible values of bioelemental combinations. ∂BNtaxi 154 = Changes in BNtaxi of taxa i explained by environmental changes ∂Env 155 taxi = taxonomy of taxa i; Sympi sympatry of taxa i with other competing taxa (j), and 156 Homi = elemental homeostasis-plasticity of the taxa i 7 157 BNj = BN taxaj = Elementomej1−n = (C: N: P: K: S: Mg: Ca … )j ∂BNtaxj 158 BNj = BNtaxj + = BNtaxj1−n + f(Taxj , Sympj , Homj ) (Eq. 2) ∂Env 159 The BN distance between taxa i and j, ∆BNi−j , will be: ∂BNtaxi −∂BNtaxj 160 ∆BNi−j = (BNtaxi − BNtaxj ) + = (BNtaxi − BNtaxj ) + ∂Env 161 [f(Taxi , Sympi , Homi ) − f(Taxj , Sympj , Homj ) (Eq. 3) 162 The segregation of the BNs of the dominant species of a holm-oak evergreen 163 Mediterranean forest (Figure 4a) and their displacement when exposed to drought in a climatic 164 field experiment (Figure 4b) constitute a consistent confirmation of these two first BN rules. 165 Very similar results for segregation and displacement were found using different calculation 166 approaches, from the additional use of ratios as explanatory variables to the use of discriminant 167 and other multivariate analyses (Appendix S1: Fig. S1). Pairs of species with more overlap in 168 their distributions had greater differences in their BNs as estimated from PCA scores (PCA based 169 on N,P, K, S, Ca and Mg concentrations) in a set of the 12 most representative Mediterranean 170 tree species of South Europe (Figure 4c), confirming the third rule. The BN hypothesis is an 171 extension of the ecological niche concept (Schoener 1986) to the bioelemental components of 172 life. The BN aims to establish the morphological and functional “adaptation” to the abiotic and 173 biotic traits that define the niche of each species-genotype from simple determination of the 174 elemental composition. It is based on the different contribution of each bioelement to the 175 different organismal functions and structures. Its main advantage over other theoretical classical 176 niche concepts is that it can be characterized by the measurable bioelemental composition, the 177 elementome, of the organism, species, community or even biome (Figure 5). This BN 8 178 measurement is not biased by the most abundant bioelements since the analysis is conducted for 179 a multivariate space defined by the standardized elemental concentrations (Figures 4 and 6). 180 The quantitative BN framework can then be used to improve projections of how species 181 will respond to environmental changes, provided elementome homeostasis/plasticity (Hom) on a 182 given time horizon is known. This Hom can be defined for example as the coefficient of variance 183 of the scores for the different individuals of a species in the PCA. It results from both 184 intraspecific variation and phenotypic individual responses. These two responses can be 185 differentiated when the hypervolume occupied by the individuals of one species is compared 186 with the hypervolume generated after changing the environmental conditions, for instance in 187 experiments in field conditions where environmental conditions are changed. 188 BN aims to refocus the use of the hypervolume in ecological niche studies. A n‐ 189 dimensional hypervolume was originally proposed by Hutchinson (1957) and defined as a subset 190 of the space, i.e. an n-dimensional geometrical shape, to describe the fundamental niche of a 191 species. Hutchinson suggested that the axes correspond to requirements of the species (e.g. light, 192 temperature, food size) and that the boundaries of the shape indicate the conditions that permit 193 the growth and reproduction of the individuals of a species. This description of the species niche 194 has been widely used and discussed (Begon et al, 2006 Chase and Leibold 2003, Colwell and 195 Rangel 2009, Holt 2009, Peterson et al 2011). There is now a growing set of statistical methods 196 to operate this hypothesis of species niche with data (Blonder 2017). BN should be a tool to 197 detect the niche shifts at short-, medium- and long-term scales in response to all type of 198 environmental changes (drought, species invasion, pests, new biotic relationships, etc.). The 199 mathematical description of each specific BN position in a multi-space in function of the 200 concentrations of several bioelements can be done by using essential multivariate statistical 9 201 analyses such as principal components analyses (PCA) or functional/general discriminant 202 analysis (F/GDA) (Sardans et al. 2015; Urbina et al. 2017), or other analyses such as kernel 203 density estimation, support vector machine, Bayesian ellipse, etc. (Blonder 2017). BN distances 204 and the level of overlap/separation among different groups can be computed with the score 205 means (centroids) of the main axes (in the PCAs; Figure 4, Appendix S1, Fig. S1) or with 206 squared Mahalanobis distances (in the discriminant analyses; Appendix S1: Fig. S2). These 207 metrics provide numerical evidences of the “distances” between the BN of different species and 208 of the BN displacement (if any) when comparing the BN of different populations of the same 209 species along space or time. Correlations between matrices provide a way of comparing the 210 distances between hypervolumes, for instance the correlation matrix of the genome differences 211 or of the differences in some ecological or ecophysiological variables in a particular set of 212 species with the corresponding correlation matrix of the BN distances of the same set of species. 213 Statistical tests such as Mantel tests (Mantel, 1967) can be used with this aim. All these 214 mathematical tools to estimate the BN allow all levels of comparisons, from species (Sardans et 215 al. 2015) to different taxa and trophic groups (Bartrons et al. 2018), also including the 216 comparison of different populations of the same species growing under distinct environmental 217 conditions (Urbina et al., 2015, 2017). 218 These studies have shown the facility and consistency in the use of different types of 219 multivariate analyses to detect the BN in spatial and temporal gradients/scales. Moreover, the 220 different levels of homeostasis/plasticity of different studied groups can be calculated by the 221 area/volume in that the individuals of each group are spread in the multidimensional space 222 generated. The temporal shifts of the BN of a species or populations of one species can also be 223 calculated and thus their evolution can be followed. BN also allows the description and 10 224 quantification of niche appearance (i.e. when new species colonize the ecosystem), expansion, 225 contraction, and extinction. Therefore, BN constitutes a quantifiable tool to detect, quantify and 226 thus better comprehend the mechanisms and processes underlying community evolution and 227 species turn-over (Figure 7). Niche construction vs niche expansion or niche contraction vs niche 228 extinction are apparently difficult to differentiate because of their similar outcomes, and niche 229 contraction (and extinction) could be the result of environmental filtering, whereas niche 230 expansion could be due to addition of species to a community. Phenomena within the niche 231 theory such as niche convergence (Pianka et al., 2017) can be followed by monitoring the species 232 BN hyper-dimension along spatial and temporal gradients and focusing on the changes in the 233 level of overlap/distance in their BN. 234 235 3. Evidence from observational data 236 3.1- Taxonomy 237 The BN hypothesis is supported by strong evidence in recent studies. Observational data 238 (Sardans and Peñuelas 2014; Bartrons et al., 2018) provide evidence that plant and animal 239 species have their own elementome, their own BN. For example, different groups of species in 240 distinct climatic biomes have different foliar bioelemental compositions and stoichiometries 241 (Sardans et al. 2014, 2015, 2016). Foliar bioelemental composition of forest tree species, for 242 example, has been positively correlated with phylogenetic distance between taxa at different 243 geographic extents, from local to continental (Sardans et al. 2014, 2015, 2016). Bioelemental 244 composition and stoichiometry thus tend to differ more between phylogenetically distant than 245 between phylogenetically close tree species (Sardans and Peñuelas 2014, Sardans et al. 2015, 246 2016). These positive correlations between elemental composition differences and phylogenetic 11 247 distances have been also observed in animals (Gonzalez et al. 2018) in a further demonstration of 248 the first process underlying the BN hypothesis. The stoichiometry of an individual under a given 249 climatic and edaphic condition is determined in part by the uptake and the efficiency with which 250 the individual uses each bioelement, all of which depend on the trade-offs among the various 251 individual functions that maximize individual fitness under that specific climatic situation. Each 252 species and taxon have its own space in the two-dimensional plot formed by the two first 253 components of the PCA of individual composition (N, P, S, K, Ca and Mg concentrations and 254 their ratios) (See example in Figure 4A and see Appendix S1: Table S1 for the scores of 255 additional PCs). 256 257 3.2- Sympatry 258 Larger differences in the stoichiometry of foliar composition among sympatric than allopatric 259 species have recently been reported both in plant communities (Sardans et al. 2015, 2016) and in 260 animal communities (González et al., 2018) as implied in the third process involved in the BN. 261 The overlapping frequency of all pairwise comparisons among many tree species is correlated 262 with the elementome distances of the corresponding tree pairs (Figure 4c) (Sardans et al. 2015), 263 thus supporting this third process. Different grass and shrub species growing under different 264 climatic conditions in communities with distinct species compositions maintain distinct species- 265 specific foliar compositions, although there may be displacement, expansion or concentration of 266 the biogeochemical space occupied (Figure 6) depending on the coexisting species (Urbina et al. 267 2015, 2017). The species bioelemental composition changes in parallel with the shifts of 268 environmental conditions, in accordance with the new optimal functions that require new 269 proportional use of different bioelements. Contraction, expansion, and displacement at the short- 12 270 term would thus be complemented by construction (allopatric speciation), extinction, 271 segregation, and equalization at the long-term both for estenoic (narrow range of ecological and 272 environmental conditions) and euroic (wide range) niches (Figure 7). 273 In-depth investigation of these processes constitutes a challenge for ecologists. Some 274 studies have already demonstrated different biogeochemical niche space (euroic-estenoic) among 275 different coexisting species-genotypes in plants (Urbina et al., 2015, 2017, Sardans et al., 2015), 276 animals (González et al., 2017, 2018; Bartrons et al., 2018) and microbes (Pereira and Berry, 277 2017). Other studies have observed how the same species displaced their biogeochemical niche 278 depending on the intraspecific and intraspecific competition (Urbina et al., 2015, 2017; Sardans 279 et al., 2016). However, observational field studies along successional stages, field manipulation 280 studies of biotic (introducing new species, changes in density,..), and abiotic (changing climate 281 conditions or nutrient availability) followed along time with similar studies in greenhouse, 282 micro- and mesocosm conditions would be necessary to further test the consistency of the new 283 holistic view of ecosystem processes provided by the BN. 284 285 286 3.3- Homeostasis/Plasticity 287 Changes in environmental supply (for example soil or water bioelemental composition) can 288 affect the bioelemental composition of organisms, as observed in microbes (Frenken et al 2017) 289 and plants (He et al. 2016, Salvagiotti et al. 2017). A trade-off between adaptations to be 290 competitive in stable environments versus being successful in less stable environments should 291 underpin the differences in the continuum of homeostasis-plasticity strategies in bioelemental 292 composition and stoichiometry among individuals, species and communities (Figure 3, Figure 13 293 4B). Higher N and P concentrations and lower N:P ratios coincide with higher stoichiometric 294 plasticity in opportunistic (r-strategy) than in competitor (K-strategy) plant species (Yu et al. 295 2011). Climate can also account for some of the variability in the bioelemental composition of 296 plant tissues at intra- and interspecific levels (Sardans et al. 2015, 2016, Zhang et al. 2012) 297 indirectly though climate controls on bioelemental availability and directly as a forcing of traits 298 linked to the elementome. Environments with fluctuating nutrient availabilities relative to the 299 species’ lifetime would have to favor organisms with greater stoichiometric plasticity. 300 Stoichiometric plasticity tends to decrease in response to fertilization as environmental nutrient 301 richness increases, in both terrestrial and aquatic systems (Sistla et al. 2015), thus suggesting a 302 different position in the homeostatic-plasticity continuum between species in nutrient-rich versus 303 nutrient-poor environments. The BN hypothesis allows to detect the plasticity at two levels: at 304 the individual level (phenotypic plasticity) by showing how its individual elemental composition 305 can vary when environmental conditions shift, and at the intraspecific level as a result of 306 intraspecific variability (individual genotypic differences, different ontogenic stage, distinct 307 sex,..) in elemental composition of a set of individuals of the same species living under the same 308 environmental conditions. This dual source of variability due to natural intra-population 309 variability and to individual capacity to shift BN in response to environmental changes can be at 310 least partially detected experimentally with multivariate analyses working with populations of 311 the same genotype growing in different environmental conditions across space or time. 312 313 3.4 Evolutionary change 314 Species have stoichiometric strategies to adapt to heterogeneous environments (Sun et al. 2017) 315 and are able to adapt their elemental compositions to those of the media after several generations 14 316 (Turner et al. 2017). The stoichiometry of phytoplankton appears to reflect oceanic stoichiometry 317 over geological timescales, providing even stronger support to this link between organisms’ 318 composition and media nutrients availability (Quigg et al. 2003; Reinhard et al. 2017). Other 319 studies of terrestrial ecosystems found that microbial community N:P ratios were correlated with 320 plant-community and soil N:P ratios (Kerkhoff and Enquist 2006, Cleveland and Liptzin 2007). 321 Close relationships between microbial and soil C:N ratios across a soil C:N gradient have been 322 also observed, e.g. in Scandinavian boreal forests (Högberg et al. 2006). Evidence suggests that 323 variation in the environmental supply of P affects the expression of highly conserved genes (e.g. 324 those of the phosphate-transporter system) and that the acquisition, assimilation and allocation of 325 P is genetically variable (Jeyasingh and Weider 2007). Transcriptomic studies of algae 326 (Grossman 2000), microbes (Baek and Lee 2007), terrestrial plants (Morcuende et al. 2007) and 327 more recently herbivores (Jeyasingh et al. 2011) have also reported that the availability of P 328 influences genetic expression. Such effects of the supply of P should impinge on fitness and 329 drive evolutionary change. Jeyasingh et al. (2009) observed that genetic variation in 330 phosphoglucose isomerase (Pg) indicated that Pg heterozygotes of Daphnia pulicaria 331 outcompeted Pg homozygotes under conditions of low C:P ratios, whereas the opposite was 332 observed under conditions of high C:P ratios. 333 Availability and stoichiometry of bioelements can determine transcriptional expression 334 and metabolomic structure (Rivas-Ubach et al. 2012) and phenotypic selection (Chowdhury et al. 335 2015), and can affect genomes and thereby become evolutionary drivers for organisms and 336 ecosystems on the longest timescales (Acquisti et al. 2009). For example, the mechanisms 337 controlling P acquisition are genetically variable, and the expression of the genes involved is 338 strongly affected by its environmental supply (Grossman 2000, Jeyasingh and Weider 2007, 15 339 Jeyasingh et al. 2011, 2014). Selection in environments with very low N availability can even 340 favor new genes richer in N-poor nucleotides (Acquist et al. 2009). Evolution could also be 341 controlled by directly affecting the quality of matter, e.g. different N:P ratios available to 342 genomes for the optimal allocation of N and P to DNA/RNA for maximizing fitness (Sardans et 343 al., 2012b). 344 345 3.5 BN of communities 346 Communities should be more capable to adjust their stoichiometry to additional nutrient inputs 347 and varying stoichiometrical conditions than single species (Danger et al. 2008). A simple and 348 comprehensive example, only considering N and P concentrations and stoichiometry, can be 349 observed when P is added to a freshwater community. In such a case, a shift in community 350 composition typically occurs in favor of N-fixing organisms with a posterior cascade of 351 composition changes towards a new equilibrium in community composition (Davidson et al. 352 2007). Environmental changes (anthropogenic or natural) such as perturbations or shifts in 353 climate, N deposition, species invasion or the increase in atmospheric CO2 can probably generate 354 changes in the availability of various nutrients. Species of the target community can then adapt, 355 depending on their phenotypic flexibility. For example, species with a BN better adapted to new 356 environmental conditions can be favored under large long-term loadings of N by N deposition, 357 but species less well adapted can disappear. Species can also turnover, when possible, with the 358 entrance of new species with traits better adapted to life in the newly generated niches (with 359 different optimal elemental compositions). This sequence of events has been observed in several 360 field studies where shifts in the elemental composition of the media (water or soil) due to 361 changing conditions (anthropogenic or natural) have been translated into changes in the species 16 362 composition of species of the community (Sterner and Elser, 2002). The species with more 363 favorable elemental compositions gain then importance and/or there is species turnover 364 (Novotny et al. 2007; Yu et al. 2011, Poxleitner et al. 2016, Du 2017). Changes in N:P have ben 365 related to substantial changes in species composition of plant communities (Peñuelas et al 2013) 366 and changes in P loads to adjusted C:P ratios in zooplankton communities (Teurlink et al., 2017). 367 368 4- Feedbacks on environment 369 The relationships between evolution and the bioelemental composition of ecosystems are 370 bidirectional, because nutrient supply can affect evolutionary processes, while the changes 371 imposed by evolutionary processes can also affect nutrient supply (Durston and El-Sabaawi 372 2017). As a result, ocean and soil stoichiometries may have been also partly controlled by 373 organisms over long time scales (Tyrrell 1999, Lenton and Klausmeier 2007, Hatton et al. 2015, 374 Zederer et al. 2017). 375 The strongest support of Redfield ratio comes from the fact that the N:P ratio in the open 376 sea areas of all major oceans is remarkably similar to the N:P ratio of plankton (16:1). This is 377 due to the residence time of N and P in the ocean (104 years) relative to the ocean’s circulation 378 time (103 years) (Falkowski and Davis 2004). However, in addition to the geochemical causes, 379 biological processes are also underlying Redfield ratios. N2 fixation can play an important role in 380 ocean N:P ratio homeostasis (Tyrrell 1999, Lenton and Watson 2000, Ganeshram et al. 2002, 381 Kenesi et al. 2009). When ocean N:P ratios fall too low, nitrogen fixation increases by 382 augmenting the biomass of N2-fixers and when these N2-fixers decompose or are eaten, their N 383 rich organic matter turns to dissolved ammonium and nitrate in the water increasing water N:P 17 384 ratios. Conversely, when the ratio rises too high, N2 fixation inputs decrease by P limitation of 385 N2-fixation, thus restoring lower water N:P ratio. This process would confer to P primary control 386 of ocean production (Tyrrell 1999, Lenton and Klausmeier 2007). 387 Soil elemental composition is also controlled by the species of plants (Zederer et al. 388 2017), the soil microbiota (Hartman et al. 2017), and the interactions between plants and soil 389 fungal communities. For example, Quercus species control nutrient dynamics by determining the 390 composition and activity of the forest floor fungal community (Chávez-Vergara et al. 2016) and 391 the litter type (Hatton et al. 2015). Conifers acidifying soil to reduce soil nutrient availabilities 392 (base cations and P) to outcompete species with a different BN provide another consistent 393 example. 394 We should thus expect a constant feedback between organisms and their environment. 395 Understanding the processes underlying species shifts in bioelemental composition and their 396 genetic basis will help us understand the responses of organisms to environmental changes 397 (Yamamichi et al. 2015, Leal et al. 2017) and in turn the effects of organisms on ecosystem 398 functioning and services (Frisch et al. 2014, Leal et al. 2017). We are currently still moving 399 towards understanding and identifying the feedbacks between species evolution and shifts in the 400 bioelemental composition of ecosystems and how these feedbacks interact and determine 401 ecosystem and evolutionary processes. Available data seems to indicate that the long-term 402 availability and use of bioelements can determine individual selection and therefore ecosystem 403 function and organization and evolution of life. 404 405 5- The biogeochemical niche hypothesis and related approaches 18 406 Functional traits such as leaf mass area, seed size or plant height have been also used to build 407 hypervolumes to identify species-specific niche (Violle and Jiang, 2009; Violle et al., 2014; 408 Lamanna et al., 2014). However, taxonomical groups (e.g. animals, plants, fungi,..) have 409 different functional traits, preventing the use of these approaches to different taxonomy clades. 410 Instead, all the organisms, independently of their taxonomical groups, are constituted by the 411 same basic bio-elements, thus providing a general and easy to measure variable, the elementome, 412 with which to establish the species-specific niche in plants (Sardans et al., 2015), in animals 413 (Gonzalez et al., 2017) and in sets of plants and animals analyzed altogether (Bartrons et al., 414 2018). The use of organismal elemental stoichiometry has been proven to be related to the 415 species functional traits in aquatic and terrestrial ecosystems (Meunier et al., 2017, Sardans et al. 416 2012c, Peñuelas et al. 2013). This also allows us to establish relationships between observed 417 shifts in ecosystem N:P medium (water or soil) with the shift in species dominance or 418 community and/or in trophic web composition and structure. 419 We first defined the BN in 2008 for plant communities (Peñuelas et al. 2008) and we 420 have since then developed it in several papers (Peñuelas et al. 2008, Sardans and Peñuelas 2014, 421 Frisch et al. 2014, Sardans et al. 2015, 2016, Urbina et al. 2017). Other groups have also used 422 this approach (De and Bhattacharyya 2009, Menge and Hedin 2009, Chimphango et al. 2015, 423 Zamora et al. 2017), obtaining results that have provided solid experimental consistency to BN 424 hypothesis. The BN that we define as species elemental composition is wide in scope and simple 425 in determination, and reflects the structure and function of each species-genotype resulting from 426 its unique evolutionary history (Sardans et al. 2008, 2015, Urbina et al. 2017). 427 BN aimed to progress beyond the classical ecological stoichiometric studies (Elser et al. 428 2000a,b, Sterner and Elser 2002, Elser and Kyle 2009, Elser et al. 2009, 2010, Rivas-Ubach et al. 19 429 2012, Sardans et al. 2012b). These ecological stoichiometric studies have linked several 430 important ecosystem processes with ecosystem and organismal C:N:P ratios. Several organismal 431 functions and responses to environmental shifts, though, cannot be explained only by the 432 changes in these three main bioelements (Sardans et al. 2012b,c). As stated above, other 433 bioelements (such as K, S, Ca, Mg and Fe) are needed in ecological stoichiometric studies to 434 provide more complete and global information correlating shifts in organismal elemental 435 composition with changes in organismal morphology and function in response to environmental 436 conditions (Peñuelas et al. 2008, Peñuelas and Sardans 2009, Sardans and Peñuelas 2014, 437 Sardans et al. 2015). The BN hypothesis is thus based on the entire elemental composition or at 438 least on the composition of as many elements as possible. The BN hypothesis has afterwards also 439 been used to refer to the concentrations and availabilities of nutrients in soil or water that favor 440 or determine the corresponding spatial occupation by a species or taxon (De and Bhattacharyya 441 2009, Menge and Hedin 2009, Chimphango et al. 2015, Zamora et al. 2017). 442 A similar hypothesis of a “multidimensional stoichiometric niche” was recently 443 introduced by Gonzalez et al. (2017). These authors defined a “multidimensional stoichiometric 444 niche” for taxonomic/trophic groups as their position in the three-dimensional space determined 445 by C, N, and P following Peñuelas et al. (2008). Gonzalez et al (2017) also interestingly opened 446 the possibility to incorporate more elements, and also element fluxes and transformation rates as 447 additional axes. In their studies, Gonzalez et al. (2017, 2018) proved the suitability of this 448 approach for animals. This “multidimensional stoichiometric niche” also tried to incorporate 449 rates of nutrient cycling, but these are much more difficult to quantify than elemental 450 concentrations and ratios. Thus, according with the observations in plants, animals, microbes, 451 and also in combination of different taxa altogether (animals and plants) (Peñuelas et al., 2008; 20 452 Sardans et al., 2015; González et al., 2017; Bartrons et al., 2008), the BN multidimensional space 453 proposed here should allow the establishment and quantification of the links between the species 454 BN distances and their taxonomic/phylogenetic distances and between homeostatic capacities 455 and sympatries, and thus to provide information on species/taxa evolutionary processes. The BN 456 thus also aims to refocus the use of the hypervolume in ecological and evolutionary niche studies 457 as discussed above in section 2. 458 Another similar approach to BN, ionomics, i.e. the profiles of mineral elements in plants, 459 has been proposed mostly focused on the genetics of plant mineral nutrition. Ionomics analyzes 460 the elemental composition of organisms in relation to soil concentrations of nutrients and trace 461 elements and also plant functional traits (Salt et al., 2008; Buescher et al., 2010; Baxter et al., 462 2012; Huang and Salt, 2016). Ionomics is mostly used to study the content of bioelements with 463 electrical charge, mostly absorbed from soils, and mostly within a context of the impact on plant 464 health, food quality and/or identification of hyper-accumulation species. It comes from the 465 confluence of metabolomics and plant nutrition (Salt et al., 2008) that, with the incorporation of 466 modern analytical platforms such as ICP techniques, allows to determine most elements 467 altogether. The fusion of this information with bioinformatics and genetic tools, such as genomic 468 sequencing, allows the identification of the genes that control uptake, storage, and use of soil 469 nutrients and trace elements in plants (Baxter et al., 2012; Lowry et al., 2012). This further 470 allows to analyze the effects of genes and environment on plant nutrition (Buescher, et al., 2010; 471 Lowry et al., 2012; Neugebauer et al., 2018) and plant physiological status (Baxter et al., 2008). 472 BN instead considers all, or as many as possible, bioelements, i.e. also including those such as C 473 or N that are initially from atmospheric, not ionic, forms (e.g. N comes mainly from atmospheric 474 N2; nitrates are rare in bedrock). BN focuses on organisms and on the total elemental 21 475 composition and stoichiometry as a proxy of their optimal function and morphology, whereas 476 ionomics focuses on the composition of the medium (soil) to describe different ionomic niches in 477 the abiotic dimension of an ecosystem. Ionomics has aimed to explain shifts in ionic composition 478 by identifying specific genes and their control in changing environmental conditions, thus 479 providing knowledge and information to improve the nutrition of crop species. BN is instead 480 based on the idea that the entire genome of each genotype/species determines species-specific 481 functionality and morphology (more distinct as genomic differences increase) and thus a distinct 482 use of different bioelements, which in turn contribute asymmetrically to different functions and 483 morphological structures. BN can thus be used without knowing about the control of gene 484 expression or about gene function. Moreover, whereas ionomic approach has been developed for 485 plants, BN is developed to be equally applicable to all type of taxa. 486 Even though these ionomic studies are mostly agriculturally oriented, and thus the link 487 with the classical theory of ecological niche is not specifically considered, they are consistent 488 with the BN hypothesis, at least at the level to correlate elemental composition with phylogenetic 489 distance. They constitute strong evidence in favor of BN consistency and constitute a very 490 interesting contribution to the BN. 491 7. Concluding summary and perspectives 492 Here we propose a BN of genotypes/species/communities that should provide an improved and 493 quantifiable proxy of the general ecological niche. The BN is characterized by a particular 494 hypervolume of their individual elementomes based on their specific needs and use of 495 bioelements in different amounts and proportions for their structure and functioning. Several 496 organismal structures and functions and the responses to environmental shifts cannot be 497 explained only by the changes in the three main bioelements, C, N and P. We propose that other 22 498 bioelements such as K, S, Ca, Mg, Fe, Mn and Mo, among others, must be incorporated to fully 499 describe the BN. The BN extends the classical ecological niche concept but has the advantage 500 that it can be easily, directly and quantitatively analyzed. 501 The biogeochemical distances, i.e. the differences in elementome, among species, are a function 502 of taxonomy (phylogenetic distance), sympatry (the bioelemental compositions should differ 503 more among coexisting than among non-coexisting species to avoid competitive pressure), and 504 homeostasis (within the continuum between high homeostasis/low plasticity and low 505 homeostasis/high plasticity). Further studies are warranted to discern the ecological and 506 evolutionary processes involved in the BN of all types of individuals, taxa and ecosystems. The 507 changes in bioelemental availability and use at long timescales should determine phenotypic 508 selection and therefore also ecosystem function and organization, and, at the end, the evolution 509 of life and the environment. 510 511 Acknowledgements 512 The authors would like to acknowledge the financial support from the European Research 513 Council 514 Synergy grant ERC-SyG-2013-610028 IMBALANCE-P, the Spanish Government grant 515 CGL2016-79835-P and the Catalan Government grant SGR 2017-1005. 516 517 References 518 Acquisti, C., J. J. Elser, and S. Kumar. 2009. Ecological nitrogen limitation shapes the DNA 519 composition of plant genomes. Molecular Biology and Evolution 26:953–956. 23 520 Araujo, M. S., D. I. Bolnick, G. Machado, A. A. Giaretta, and S. F. dos Reis. 2007. Using d13C 521 stable isotopes to quantify individual-level diet variation. Oecologia 152:643–654. 522 Baek, J. H. and S. Lee. 2007. Transcriptome Analysis of Phosphate Starvation Response in. 523 Journal of Microbiological Biotechnology 17:244–252. 524 Baribault, T., R. Kobe, and A. Finley. 2012. Tropical tree growth is correlated with soil 525 phosphorus, potassium, and calcium, though not for legumes. Ecological Monograph 526 82:189–203. 527 Bartrons, M., J. Sardans, D. Hoekman, and J. Peñuelas. 2018. Trophic transfer from aquatic to 528 terrestrial ecosystems: a test of the biogeochemical niche hypothesis. Ecosphere 9:e02338. 529 Baxter, I.R., O. Vitek, B. Lahner, B. Muthukumar, M. Borghi, J. Morrissey, M. L.Guerinot, and 530 D. E. Salt. 2008. The leaf ionome as a multivariable system to detect a plant’s 531 physiological status. Proceedings of the National Academy of Sciences, USA 105:12081- 532 12086. 533 Baxter, I., C. Hermans, B. Lahner, E. Yakubova, M. Tikhonova, N. Verbruggen, D. Y. Chao, 534 and D. E. Salt. 2012. Biodiversity of mineral nutrient and trace element accumulation in 535 Arabidopsis thaliana. PlosOne 7:e35121. 536 Bearhop, S., C. E. Adams, S. Waldron, R. A. Fuller, and H. MacLeod. 2004. Determining 537 trophic niche width: a novel approach using stable isotope analysis. Journal of Animal 538 Ecology 73:1007–1012. 539 Begon, M. et al. 2006. Ecology: from individuals to ecosystems. – Blackwell. 540 Blonder, B. 2017. Hypervolume concepts in niche- and trait-based ecology. Ecography 41, 1441- 541 1445. 542 Bourgis, F. et al. 1999. S-methylmethionine plays a major role in phloem sulfur transport and is 24 543 synthesized by a novel type of methyltransferase. Plant Cell 11:1485–1498. 544 Buescher, E., T. Achberger, I. Amusan, A. Giannini, C. Oschsenfeld, A. Rus, B. Lahner, O. 545 Hoekenga, E. Yakubova, J. F. Harper, JM. L. Guerinot, M. Zhang, D. E. Salt, and I. R. 546 Baxter, I.R. 2010. Natural genetic variation in selected populations of Arabidopsis thaliana 547 is associated with ionomic differences. PlosOne:art 211081. 548 Chase, J. M. and M. A. Leibold. 2003. Ecological niches: linking classical and contemporary 549 approaches. – University of Chicago Press. 550 Chávez-Vergara, B. et al. 2016. Quercus species control nutrients dynamics by determining the 551 composition and activity of the forest floor fungal community. Soil Biology and 552 Biochemistry 98:186–195. 553 Chimphango, S. B. M., G. Potgieter, and M. D. 2015. Cramer, M. D. Differentiation of the 554 biogeochemical niches of legumes and non-legumes in the Cape Floristic Region of South 555 Africa. Plant Ecology 216:1583–1595. 556 Chowdhury, P. R. et al. 2015. Differential transcriptomic responses of ancient and modern 557 Daphnia genotypes to phosphorus supply. Molecular Ecology 24:123–135 (2015). 558 Cleveland, C. C. and D. Liptzin. 2007. C:N:P stoichiometry in soil: Is there a ‘Redfield ratio’ for 559 the microbial biomass? Biogeochemistry 85:235–252. 560 Colwell, R. K. and T. F. Rangel. 2009. Hutchinson’s duality: the once and future niche. 561 Proceedings of the National Academy of Sciences USA 106:19651–19658. 562 Danger, M., T. Daufresne, F. Lucas, S, Pissard, and G. Lacroix. 2008. Does Liebig’s law of the 563 minimum scale up from species to communities? Oikos 117:1741–1751. 564 Davidson, E. A. W., and R. Howarth. (2007) Nutrients in synergy. Nature 449:1000–1003. 565 De, C., and S. Bhattacharyya. 2009. First record of the Recent – subRecent diatomaceous earth 25 566 from the Gangetic alluvial floodplains of. Current Science 96:967–973. 567 De Bolos, O., and J. Vigo. 2001. In: Barcino (Ed.), Flora dels Països Catalans (I-IV), Barcelona. 568 Du, E. 2017. Integrating species composition and leaf nitrogen content to indicate effects of 569 nitrogen deposition. Environmental Pollution 221:392–397. 570 Durston, D., El-Sabaawi, R. (2017) Bony traits and genetics drive intraspecific variation in 571 vertebrate elemental composition. Functional Ecology 31, 2128-2137. 572 Elser, J. J., K. Hayakawa, and J. Urabe. 2001. Nutrient limitation reduces food quality for 573 zooplankton: Daphnia response to seston Phosphorus enrichment. Ecology 82: 898–903. 574 Elser, J. J. et al. 2000a. Nutritional constraints in terretrial and freshwater food webs. Nature 575 408:578–580. 576 Elser, J. J. et al. 2000b. Pelagic C:N:P stoichiometry in a eutrophied lake: Responses to a whole- 577 lake food-web manipulation. Ecosystems 3:293–307. 578 Elser, J. J. et al. 2009. Shifts in Lake N:P Stoichiometry and Nutrient Limitation Driven by 579 Atmospheric Nitrogen Deposition. Science 326:835–837. 580 Elser, J. J., and M. Kyle, M. 2009. Nutrient availability and phytoplankton nutrient limitation 581 across\ra gradient of atmospheric nitrogen deposition. Ecology 90:3062–3073 . 582 Elser, J. J. et al. 2010. Atmospheric nitrogen deposition is associated with elevated phosphorus 583 limitation of lake zooplankton. Ecological Letters 13: 1256–1261. 584 Falkowski, P. G. and C. S. Davis. 2004. Natural proportions. Redfield ratios: the uniformity of 585 elemental ratios in the oceans and the life they contain underpins our understanding of 586 marine biogeochemistry. Nature 431: 131. 587 Folk, R., T. Franquesa, and J. M. Camarasa. 1984. Vegetació. In: Història Natural 588 PaïsosCatalans (VII). Fundació Enciclopèdia Catalana, Barcelona. 26 589 Frenken, T. et al. 2017. Changes in N:P supply ratios affect the ecological stoichiometry of a 590 toxic cyanobacterium and its fungal parasite. Frontieres in Microbiology 8:1–11. 591 Frisch, D. et al. 2014. A millennial-scale chronicle of evolutionary responses to cultural 592 eutrophication in Daphnia. Ecology Letters 17: 360–368. 593 Ganeshram, R. S., T. F. Pedersen, S. E. Calvert, and R. Francois.2002. Reduced nitrogen fixation 594 in the glacial ocean inferred\rfrom changes in marine nitrogen and 595 phosphorus\rinventories. Nature 415:156–159. 596 González, A. L., O. Dézerald, P. A. Marquet, G. Q. Romero, and D. S. Srivastava. 2017. The 597 Multidimensional Stoichiometric Niche. Frontiers in Ecology and Evolution 5:1–17. 598 Gonzalez, A.L., Céréghino, R., Dézerald, O., Farjalla, V.F., Leroy, C., Richardson, B.A., 599 Richardson, M.J., Romero, G.Q., Srivastava, D.S. 2018. Ecological mechanisms and 600 phylogeny shape invertebrate stoichiometry: A test using detritus-based communities 601 across Central and South America. Functional Ecology, 32, 2448-2463. 602 Grime, J. P. 1977. Evidence for the existence of three primary strategies in plants and its 603 relevance to ecological and evolutionary theory. American Naturalist 111:1169–1194. 604 Grossman, A. 2000. Acclimation of Chlamydomonas reinhardtii to its Nutrient Environment. 605 Protist 151:201–224. 606 Hartman, W. H., R. Ye, W. R. Horwath, and S. G. A. Tringe 2017. genomic perspective on 607 stoichiometric regulation of soil carbon cycling. ISME Journal:2652–2665. 608 Hatton, P. J., C. Castanha, M. S. Torn, and J. A. Bird. 2015. Litter type control on soil C and N 609 stabilization dynamics in a temperate forest. Global Change Biology 21:1358–1367. 610 He, M. et al. 2016. Influence of life form, taxonomy, climate, and soil properties on shoot and 611 root concentrations of 11 elements in herbaceous plants in a temperate desert. Plant Soil 27 612 398:339–350. 613 Högberg, M. N., D. D. Myrold, R. Giesler, and P. Högberg. 2006. Contrasting patterns of soil 614 N-cycling in model ecosystems of Fennoscandian boreal forests. Oecologia 147:96–107. 615 Holt, R. D. 2009. Bringing the Hutchinsonian niche into the 21st century: ecological and 616 evolutionary perspectives. Proceedings of the National Academy of Sciences USA 106: 617 19659–19665. 618 Huang, X. Y., and D. E. Salt. 2016. Plant Ionomics: From elemental profiling to environmental 619 adaptation. Molecular Plant 9:787-797. 620 Hutchinson, G. E. 1957. Concluding remarks. Cold Spring Harbor Symposia on Quantitative 621 Biology 22:415-427. 622 Jeyasingh, P. D., and L. J. Weider. 2007. Fundamental links between genes and elements: 623 Evolutionary implications of ecological stoichiometry. Moecular Ecology 16:4649–4661. 624 Jeyasingh, P. D. et al. 2011. How do consumers deal with stoichiometric constraints? Lessons 625 from functional genomics using Daphnia pulex. Molecular Ecology 20:2341–2352. 626 Jeyasingh, P. D., L. J. Weider, and R. W. Sterner. 2009. Genetically-based trade-offs in response 627 to stoichiometric food quality influence competition in a keystone aquatic herbivore. 628 Ecological Letters 12:1229–1237. 629 Jeyasingh, P. D., R. D. Cothran, and M. Tobler. 2014. Testing the ecological consequences of 630 evolutionary change using elements. Ecology and Evolution 4:528-538. 631 Kattge J., Díaz S., Lavorel S., Prentice I.C., Leadley P., Bönish G., Garnier E., Westoby M., et 632 al., . 2011. TRY - a global database of plant traits. Global Change Biology 17: 2905-2935. 633 Kenesi, G., H. M. Shafik, A. W. Kovács, S. Herodek, and m. Présing. 2009. Effect of nitrogen 634 forms on growth, cell composition and N2 fixation of Cylindrospermopsis raciborskii in 28 635 phosphorus-limited chemostat cultures. Hydrobiologia 623:191–202. 636 Kerkhoff, A. J. and B. J. Enquist. 2006.Ecosystem allometry: The scaling of nutrient stocks and 637 primary productivity across plant communities. Ecological Letters 9:419–427. 638 Knight, M. R., A. K. Campbell, S. M. Smith, A. J. Trewasvas. 1991. Transgenic plant aequorin 639 reports the effeccts of touch and cold-shock and elicitors on cytoplasmic calcium. Nature 640 352:524–526. 641 Lapenis, A. G., G. B. Lawrence, A. Heim, C. Zheng, and W. Shortle (2013) Climate warming 642 shifts carbon allocation from stemwood to roots in calcium-depleted spruce forests. Global 643 Biogeochemical Cycles 27:101–107. 644 Lamanna, C., B. Blonder, C., Violle, N.J.B., Kraft, B., Sandel, I., Simova, J.C.., Donoghue, J.C., 645 Svenning, J.C., McGill, B.J., Boyle, B., Buzzard, V., Dollins, P.M.., Jorgensen, A., 646 Marcuse-Kubitza, N., Moureta-Holme, P.K.., Peet, W.H., Piel, J., Regetz, M., Schlidhauer, 647 N., Spencer, B., Thiers, S.K.., Wilser, S.K., and B.J. Enquist. 2014. Fungtional trait species 648 and the latitudinal diversity gradient. Proceedings of the National Academy of Sciences 649 USA 111:13745-13750. 650 Leal, M. C., O. Seehausen, O., and B. Matthews. 2017. The Ecology and Evolution of 651 Stoichiometric Phenotypes. Trends in Ecology and Evolution 32:108–117. 652 Lennon, J.T., Aanderud, Z.T., Lehmkuhl, B.K., Schoolmaster, D.R.Jr. 2012. Mapping the niche 653 space of soil microorganisms using taxonomy and traits. Ecology 93, 1867-1879. 654 Lenton, T. M., and A. J. Watson. 2000. Redfield revisited: 1. Regulation of nitrate, phosphate, 655 and oxygen in the ocean. Global Biogeochemical Cycles 14:225–248. 656 Lenton, T. M., and C. Klausmeier. 2007. Biotic stoichiometric controls on the deep ocean N : P 657 ratio. Biogeosciences 4:353–367. 29 658 Lowry, D. B., C. C. Sheng, Z. Zhu, T. E. Juenger, B. Lahner, D. E. Salt, and J. H. Willis. 2012. 659 Mapping of ionomic traits in Mimulus guttatus reveals Mo and Cd QTLs that colocalize 660 with MOT1 homologues. PlosOne 7:art e30730. 661 Mantel, N. 1967. The detection of disease clustering and a generalized regression approach. 662 Cancer Research 27, 209-220. 663 Menge, D. N. L. and L. O. Hedin. 2009. Nitrogen fixation in different biogeochemical niches 664 along a 120 000-year chronosequence in New Zealand. Ecology 90:2190–2201. 665 Meunier, C. L., M. Boersma, R. El-Sabaawi, H. M. Halvorson, E. M. Herstoff, D. B. Van de 666 Waal, and R. J. Vogt, 2017. From elements to function: toward unifying ecological 667 stoichiometr and trait-based Ecology. Frontieres in Environmental Science 5:Art 18. 668 Morcuende, R. et al. 2007. Genome-wide reprogramming of metabolism and regulatory 669 networks of Arabidopsis in response to phosphorus. Plant, Cell Environment 30:85–112. 670 Naples, B. K., and M. C. Fisk. 2010. Belowground insights into nutrient limitation in northern 671 hardwood forests. Biogeochemistry 97:109–121. 672 Neugebauer, K., M. R. Broadley, H. A. El-Serehy, T. S. George, J. W. McNicol, M. F. Moraew, 673 and P. J. White. 2008. Variation in the Angiosperm Ionome. Physiologia Plantarum 1–17. 674 Novotny, A. M. et al. 2007. Stoichiometric response of nitrogen-fixing and non-fixing dicots to 675 manipulations of CO2, nitrogen, and diversity. Oecologia 151:687–696. 676 Peñuelas, J. et al. 2013. Evidence of current impact of climate change on life: A walk from genes 677 to the biosphere. Global Change Biology 19:2303–2338. 678 Peñuelas, J., and J. Sardans. 2009. Elementary factors. Nature 460:803–804 (2009). 679 Peñuelas, J. et al. 2010. Faster returns on ‘leaf economics’ and different biogeochemical niche in 680 invasive compared with native plant species. Global Change Biology 16:2171–2185. 30 681 Peñuelas, J., J. Sardans, R. Ogaya, and M. Estiarte. 2008. Nutrient stoichiometric relations and 682 biogeochemical niche in coexisting plant species: Effect of simulated climate change. 683 Polish Journal of Ecology 56:613–622 (2008). 684 Pereira, F.C., Berry, D. 2017. Microbial nutrient niches in the gut. Environmental Microbiology, 685 19, 1366-1378. 686 Peterson, A. T. 2011. Ecological niche conservatism: a time-structured review of evidence. 687 Journal of Biogeography 38:817-827. 688 Pianka E. R., L.J. Vitt, N. Pelegrin, D.B. Fitzgerald, and K.O. Winemiller. 2017. Toward a 689 periodic table on niches, or exploring the lizard niche hypervolume. The American 690 Naturalist 2017. Toward a Periodic Table of Niches, or Exploring the Lizard Niche 691 Hypervolume. The American Naturalist 190:601-616. 692 Pierce, S., D. Negreiros, B. E. L. Cerabolini, J. Kattge, S. Díaz, M. Kleyer, B., Shiply, S. J. 693 Wright, et al. 2016. A global method for calculating plant CSR ecological strategies 694 applied across biomes world-wide. Functional Ecology 31:444-457. 695 Poxleitner, M., G. Trommer, P. Lorenz, and H. Stibor. 2016. The effect of increased nitrogen 696 load on phytoplankton in a phosphorus-limited lake. Freshwater Biology 61:1966–1980. 697 Quigg, A. et al. 2003. The evolutionary inheritance of elemental stoichiometry in marine 698 phytoplankton. Nature 425:291–294. 699 Rivas-Ubach, A., J. Sardans, M. Perez-Trujillo, M. Estiarte, and J. Peñuelas. 2012. Strong 700 relationship between elemental stoichiometry and metabolome in plants. Proceedings of 701 the National Academy of Sciences USA 109:4181–4186. 31 702 Rivas-Martínez, S., F. Fernández-González, J. Loidi, M. Lousa, and A. Penas. 2001. 703 Syntaxonomical Checklist of vascular plant communities of Spain and Portugal to 704 association level. Itinera Geobotany 14, 5–341. 705 Reinhard, C. T. et al. 2017.Evolution of the global phosphorus cycle. Nature 541:386–389. 706 Salt, D. E., I. Baxter, and B. Lahner. 2008. Ionomics and the study of plant ionome. Annual 707 Review of Plant Biology 59:709-733. 708 Salvagiotti, F. et al. 2017. N:P:S stoichiometry in grains and physiological attributes associated 709 with grain yield in maize as affected by phosphorus and sulfur nutrition. Field Crop 710 Research 203:128–138. 711 Sardans J., J., Peñuelas, and M. Estiarte. 2008a. Changes in soil enzymes related to C and N 712 cycle and in soil C and N content under prolonged warming and drought in a 713 Mediterranean shrubland. Applied Soil Ecology 39: 223-235. 714 Sardans J., J. Peñuelas, M. Estiarte and P. Prieto. 2008b. Warming and drought alter C and N 715 concentration, allocation and accumulation in a Mediterranean shrubland. Global Change 716 Biology 14: 2304-2316. 717 Sardans J., J. Peñuelas, and R. Ogaya. 2008c. Drought-Induced Changes in C and N 718 Stoichiometry in a Quercus ilex Mediterranean Forest. Forest Science 54: 513-522. 719 Sardans, J., and J. Peñuels. 2013. Tree growth changes with climate and forest type are 720 associated with relative allocation of nutrients, especially phosphorus, to leaves and wood. 721 Global Ecology and Biogeography 22:494-507. 722 Sardans, J., and J.Peñuelas. 2014. Climate and taxonomy underlie different elemental 723 concentrations and stoichiometries of forest species: The optimum ‘biogeochemical niche’. 724 Plant Ecology 215:441–455. 32 725 Sardans J., Peñuelas J. 2015. Potassium: a neglected nutrient in global change. Global Ecology 726 and Biogeography 24:261-275. 727 Sardans, J., O. Grau, H.Y.H. Chen, I.A. Janssens, P. Ciais, S. Piao, and J. Peñuelas. 2017. 728 Changes in nutrient concentrations of leaves and roots in response to global change 729 factors. Global Change Biology 23:3849-3856. 730 Sardans, J. et al. 2016. Foliar and soil concentrations and stoichiometry of nitrogen and 731 phosphorous across European Pinus sylvestris forests: Relationships with climate, N 732 deposition and tree growth. Functional Ecology 16:676-689. 733 Sardans, J., A. Rivas-Ubach, and J. Peñuelas. 2011. Factors affecting nutrient concentration and 734 stoichiometry of forest trees in Catalonia (NE Spain). Forest Ecology and Management 735 262:2024-2034. 736 Sardans, J., R. Alonso, J. Carnicer, M. Fernandez-Martinez, M. G. Vivanco, and J. Peñuelas. 737 2016. Factors influencing the foliar elemental composition and stoichiometry in forest trees 738 in Spain. Perspectives Plant Ecology, Evolution and Systematics 18:52–69. 739 Sardans, J., I. A. Janssens, R. Alonso, S. D. Veresoglou, M. C. Rillig, T. G. M. Sanders, J. 740 Carnicer, I. Filella, G. Farré-Armengol, and J. Peñuelas. 2015. Foliar elemental 741 composition of European forest tree species results from evolutionary traits and the present 742 environmental and competitive conditions. Global Ecology and Biogeography 24:240–255. 743 Sardans J., Peñuelas J., Coll M., Vayreda J., Rivas-Ubach A. 2012a. Stoichiometry of potassium 744 is largely determined by water availability and growth in Catalonian forests. Functional 745 Ecology 26: 1077-1089. 746 Sardans, J., A. Rivas-Ubach, and J. Penuelas. 2012b. The elemental stoichiometry of aquatic and 747 terrestrial ecosystems and its relationships with organismic lifestyle and ecosystem 33 748 structure and function: A review and perspectives. Biogeochemistry 111:1–39. 749 Sardans, J., A. Rivas-Ubach, and J. Peñuelas. 2012c. The C:N:P stoichiometry of organisms and 750 ecosystems in a changing world: A review and perspectives. Perspectuives in Plant 751 Ecology Evolution and Systematics 14:33–47. 752 Schoener, T. W. 1986. in Ecological concepts: the contribution of ecology to an understanding 753 on the natural word (ed. Cherret, J. M.) (Cambridge, Blackwell Scientific Publications, 754 1986). doi:10.1126/science.232.4746.113-a 755 Shaul, O. 2002. Magnesium transport and function in plants: the tip of the iceberg. BioMetals 756 15:309–323. 757 Sistla, S. A., A. P. Appling, A. M. Lewandowska, B. N. Taylor, and A. A. Wolf. 2015. 758 Stoichiometric flexibility in response to fertilization along gradients of environmental and 759 organismal nutrient richness. Oikos 124: 949–959. 760 Slade, S. 2006. Elements in living organisms. (The Rosen publishing Group, 2006). 761 Sterner, R. W.., and J. J. Elser. 2002. Ecological Stoichiometry: The Biology of Elements from 762 Molecules to the Biosphere. 763 Sun, L. et al. 2017. Leaf elemental stoichiometry of Tamarix Lour. species in relation to 764 geographic, climatic, soil, and genetic components in China. Ecological Enginnering 765 106:448–457. 766 Tripler, C. E. S. S. Kaushal, G. E. Likens, and M. Todd Walter. 2006. Patterns in potassium 767 dynamics in forest ecosystems. Ecological Letters 9:451–466. 768 Turner, C. B., B. D. Wade, J. R. Meyer, B. Sommerfeld, and R. E. Lenski 2017. Evolution of 769 organismal stoichiometry in a long-term experiment with Escherichia coli. Royal Society 770 Open Science 4:170497. 34 771 Tyrrell, T. 1999. The relative in inluences of nitrogen and phosphorus on oceanic primary 772 production. Nature 400:525–530. 773 Urbina I., Sardans J., C. Beierkuhnlein, A. Jentsch, S. Backhaus, K. Grant, J. Kreyling J., and J. 774 Peñuelas. 2015. Shifts in the elemental composition of plants during a very severe drought. 775 Environmental and Experimental Botany 111:63-73. 776 Urbina, I., J. Sardans, O. Grau, C. Beierkuhnlein, A. Jentsch, J. Kreyling, and J. Peñuelas. 2017. 777 Plant community composition affects the species biogeochemical niche. Ecosphere 778 8:e01801.1 779 Violle, C., and L. Jiang. 2009. Towards a trait-based quantification of species niche. Journal of 780 Plant Ecology 2:87-93 781 Violle, C., P. B. Reich, S. W. Pacala, B. Enquist, and J. Kattge. 2014. The emergence and 782 promise of functional biogeography. Proceedings of the National Academy of Sciences 783 USA 11:13690-13696. 784 Wilson, R.P., McMahon, C.R., Quintana, C.R., Frere, E., Scolare, A., Hays, G.C., Bradshow, 785 C.J.A. 2011. N-dimensional animal energetic niches clarify behavioural options in a 786 variable marine Environmental. Journal of Experimental Biology 214, 646-656. 787 Yamamichi, M., and C. L. Meunier, A. Peace, C. Prater, and M. A. Rúa. 2015.. Rapid evolution 788 of a consumer stoichiometric trait destabilizes consumer-producer dynamics. Oikos 789 124:960–969. 790 Yu, Q., J. J. Elser, N. He, H. Wu, Q. Chen, G. Zhang, and X. Han. 2011. Stoichiometry 791 homeostasis of vascular plants in the inner Mongolia grassland. Oecologia 166:1–10. 792 Zamora, P. B., M. B. Cardenas, R. Lloren, and F. P. Siringan. 2017. Seawater-groundwater 793 mixing in and fluxes from coastal sediment oberlying discrete fresh seepage zones: a 35 794 modeling study. Journal of Geophysical Research Oceans 122:6565–6582. 795 Zederer, D. P., U. Talkner, M. Spohn, and R. G. Joergensen. 2017. Microbial biomass 796 phosphorus and C/N/P stoichiometry in forest floor and A horizons as affected by tree 797 species. Soil Biology and Biochemistry 111:166–175. 798 Zhang, S. B., J. L. Zhang, J. W. F. Slik, and K. F. Cao. 2012. Leaf element concentrations of 799 terrestrial plants across China are influenced by taxonomy and the environment. Global 800 Ecology and Biogeography 21: 809–818. 801 802 803 36 804 Figure captions N, P, S Organism Information and energy storage Nucleotides, Proteins, Organic molecular Skeletons Functional control C, H skeleton Oxidative power, energy production Catalytic power, physiological processes and cell homeostasis O2 K, Ca, Mg, Fe, Mo, Mn, Zn,… 805 806 Figure 1. Schematic of the use of bioelements by organism in different functions. C, H, and O 807 are the basis of organismic structure, whereas N, P and S enter in the various biomolecules 808 allowing biochemical reactions. The molecules that control life processes (e.g. RUBISCO or 809 chlorophyll), that carry information (e.g. DNA and RNA), and that control the storage and 810 release of energy when and where necessary (e.g. ATP and NAD) all incorporate N, P and in 811 lower amounts S. Other bioelements, normally present in low concentrations such as potassium 812 (K), magnesium (Mg), iron (Fe), calcium (Ca), molybdenum (Mo), manganese (Mn) and zinc 813 (Zn) are cofactors in catalytic reactions in several biochemical and physiological processes 814 including for example photosynthesis, respiration, or cellular ion balance. The arrows indicate 815 the processes in which each bioelement is involved. Green arrows indicate the main organismal 816 functions to which each bioelement contributes. The red arrows indicate the relationships of the 817 different bio-molecules with basic organismal structures and functions. 37 The example of plants Ecological variables Stoichiometry traits Ecological Growth Body Reproductive Stoichiometry strategy Live-span Other [C] [N] [P] [K] rate size strategy flexibility Few seeds but Strong K-strategy Long Medium Large with highly investment Medium (competitor) caloric reserves in structure Strong A-strategy investment Medium (stress- Medium Slow Medium in C-rich Low -small tolerator) secondary compounds General trend to Many seeds but develop r-strategy Short Fast Small with low caloric belowground High (ruderal) reserves vegetative & reproductive structures 818 819 Figure 2. The example of plants. Hypothesized asymmetric uses of bioelements in evolution, 820 depending on the ecological strategy of each species. Vertical arrows indicate low or high 821 concentration values (double arrows indicate especially high values). Dashes indicate not clear 822 patterns. There is a different use of all these elements depending on the ecological strategy of 823 each species (Grime 1977). 824 38 BNj = BN taxaj = Elementomej1−n = C: N: P: K: S: Mg: Ca … j BNtaxi − BNtaxj ∆BNi−j= BNtaxi − BNtaxj + = BNtaxi − BNtaxj + [f Taxi, Sympi , Homi − f Taxj , Sympj, Homj Env 825 826 Figure 3. Theoretical distribution of the biogeochemical differences between species as a 827 function of (1) the taxonomic and phylogenetic distances between them, (2) the level of 828 sympatry (overlap frequency in their corresponding areas of distribution) and of (3) the 829 homeostatic or flexible response to current environmental conditions (e.g. climate, soil traits). 830 The surface represents the biogeochemical distance resulting from phylogenetic distance and 831 level of sympatry, and the brown and blue arrows the positive and negative residuals of the BN 832 distance induced by variable current environmental conditions (e.g. climate, soil traits, 833 neighbors). 834 39 (A) (B) (C) 1.2 1.2 1.2 0.8 P. latifolia 0.8 Ca P. Latifolia control Mg P S Mg Ca P S 1.0 0.4 K K P. Latifolia drought N 0.4 PC2 (20.2 %) N PC2 (20.2%) PC1distance distances A. unedo A. Unedo drought C 0.8 C 0 0 Q. Ilex drought A. Unedo control 0.6 PC1 -0.4 Q. ilex -0.4 -0.8 -0.8 Q. Ilex control 0.4 -1.2 -1.2 -1.2 -0.8 -0.4 0 0.4 0.8 1.2 -1.2 -0.8 -0.4 0 0.4 0.8 1.2 0.2 -1 0 1 2 3 4 PC1 (25.8%) PC1 (25.8%) Speciesoverlapping Species overlap frequency frequence 835 836 Figure 4. (A) Biogeochemical niche segregation among coexisting plant species. We analyzed 837 the foliar elementomes of the three dominant trees (Quercus ilex, Arbutus unedo and Phillyrea 838 latifolia) in a field experiment of climatic manipulation in an evergreen Mediterranean forest in 839 the Prades Mountains (Catalonia). We plotted the plant scores for the first two principal 840 components of the principal component analysis (PCA) conducted with foliar N, P, K, S, Ca, Mg 841 concentrations as variables. (B) Shifts in biogeochemical niche in a drought experiment. The 842 mean ± confidence intervals (95%, represented by the ellipses) of the species × climate 843 interaction groups are depicted in the plot thus comparing control plants to those submitted to a 844 decrease of 20 % of water availability during six years in field conditions. The other PCAs (from 845 PC3 to PC7) and the corresponding other axes are now shown in the Appendix S1: Table S1. 846 (C) Score distances for PC1 of the PCA of the foliar stoichiometry (foliar N, P, K, S, Ca and Mg 847 concentrations and their ratios) in a set of the 12 most representative Mediterranean species of 848 South Europe as a function of the frequency of pairwise species overlap (coincidence in the same 849 community): six broadleaf species, Quercus ilex, Q. faginea, Q. pyrenaica, Q. suber, Castanea 850 sativa and Eucalyptus globulus, and six needleleaf species, Pinus halepensis, P. pinaster, P. 851 nigra, P. pinea, Juniperus oxycedrus and J. thurifera). 0, no overlap; 1, occasional overlap; 2, 40 852 moderate overlap; 3, frequent overlap. Based on Rivas-Martinez et al. (2001) De Bolos and Vigo 853 (2001), CREAF, Catalonia cover map, (2013) and Folk et al. (1984). 854 855 856 Figure 5. Elementome, full spectrum of bioelemental content, of boreal, wet temperate, 857 Mediterranean and tropical forests. (A) Elementomes calculated from web of science data 858 incorporated to our GEU-CREAF dataset (Boreal forest: 473 sites and 1890 trees; Wet temperate 41 859 forests: 2882 sites and 12665 trees; Mediterranean forests: 2001 sites and 2901 trees; Tropical 860 wet forest: 1100 sites and 1821 trees). The values in plot B show the difference of each forest 861 type relative to the global mean (% ± SE). We have used data from the main available data bases 862 (e.g. TRY and ICP forest) and from an exhaustive search for woody plants from non- 863 manipulated areas in WEB of Science 864 N K C P N N K K Fe Ca C P Mg C P S Fe Fe Ca Ca Mg Mg S S N Biogeochemical niche K Biogeochemical Displacement Expansion C P Fe Ca Mg S Biogeochemical niche Concentration 865 866 Figure 6. Three possible biogeochemical niche responses to short-term changes in the 867 environmental conditions: Contraction, Expansion or Displacement. Blue dots represent 868 individuals of a species, grey and blue ellipses represent the initial and final BN space. 869 42 Biogeochemichal niches defined by the different elementomes C:N:P:H:O:K,.. C:N:P:H:O:K,.. C:N:P:H:O:K,.. C:N:P:H:O:K,.. C:N:P:H:O:K,.. C:N:P:H:O:K,.. C:N:P:H:O:K,.. C:N:P:H:O:K,.. C:N:P:H:O:K,.. Time Niche Niche Equalizing expansion Niche extinction mechanisms Eurioic Niche contraction Competitive niche construction Stenoic Displacement segregation Niche extinction (allopatric speciation) Abiotic factors 870 871 Figure 7. Possible responses of species biogeochemical niches to long-term changes in the 872 abiotic and biotic environmental conditions (possible evolutionary changes in the elementome of 873 a species). We hypothesize that each species has an optimal function related with its niche traits 874 and thus an optimal content of bioelements. Species have an optimal function related with its 875 niche traits and thus an optimal content of bioelements. The species content of bioelements 876 changes in parallel with the species niche shifts according with the new optimal functions for the 877 new environmental conditions. Contraction, expansion, and displacement are here complemented 878 by construction (allopatric speciation), extinction, segregation, and equalization at the long-term 879 both for estenoic (narrow range of ecological and environmental conditions) and euroic (wide 880 range) niches. The communities resulting of interactions with changing abiotic and biotic 881 conditions are even more capable to adjust their stoichiometry than single species to new nutrient 882 supply and stoichiometrical conditions. Each particular niche constitutes a singular functional 43 883 and physical position in the ecosystem that can vary with time in response to changes in abiotic 884 variables (climate, perturbations such as volcanic eruptions, etc.) and to changes in the impacts 885 and feed-backs of the other organisms living in it (biotic factors). In this context, each niche 886 should determine a specific and particular content of different bioelements in each moment. This 887 should also determine a singular elementome accompanying the niche characteristics of each 888 genotype and species at both micro- and macro-evolutionary timescales. 44