This is the accepted version of the journal article: Fernández-Martínez, Marcos. «From atoms to ecosystems: elementome diversity meets ecosystem functioning». New Phytologist, Vol. 234, issue 1 (April 2022), p. 35-42. DOI 10.1111/nph.17864 This version is available at https://ddd.uab.cat/record/257160 under the terms of the license 1 Tansley Insight 2 From atoms to ecosystems: elementome diversity meets 3 ecosystem functioning 4 Authors: Marcos Fernández-Martínez*1,2,3. 5 Affiliation: 1 6 Research group PLECO (Plants and Ecosystems), Department of Biology, University 7 of Antwerp, 2610 Wilrijk, Belgium. 2 8 CREAF, E08193 Bellaterra (Cerdanyola del Vallès), Catalonia, Spain 3 9 BEECA-UB, Department of Evolutionary Biology, Ecology and Environmental 10 Sciences, University of Barcelona, E08028 Barcelona, Catalonia, Spain 11 12 *Correspondence to: M. Fernández-Martínez, m.fernandez@creaf.uab.cat 13 ORCID: https://orcid.org/0000-0002-5661-3610 14 Twitter: @Burriach 15 Summary word count: 117 16 Manuscript word count: 2516 17 Figures and tables: 3 figures, 1 tables 18 References: 40 19 20 Contents 21 Summary………………………………………………………………………………… 2 22 I. Introduction……………………………………………………………………………. 3 23 II. Elementome ecology: niches, evolution and plasticity………………………...… 3 24 III. Elements, diversity and ecosystem functioning………………………………….. 7 25 IV. Conclusions………………………………………………………………………….. 8 26 Acknowledgements……………………………………………………………………… 9 27 References……………………………………………………………………………….. 9 1 28 Summary 29 The elemental composition of plants (the elementome) is a reliable indicator of their 30 functional traits and the ecological strategies that they follow, and thus represents a good 31 predictor of how ecosystems work. Biodiversity and, especially, functional diversity are 32 also widely recognised as important drivers of ecosystem functioning, mainly because of 33 niche partitioning amongst different species. I here review evidence indicating that plant 34 elementomes relate to their ecological niches, and how plant elemental concentrations 35 may shift in response to abiotic and biotic drivers. I propose the use of ecosystem 36 elementome diversity as a universal metric to compare ecosystems and investigate 37 diversity-ecosystem functioning relationships. Future research using this promising novel 38 approach will bring together elementomes, diversity and ecosystem functioning. 39 Keywords: stoichiometry, ecological niche, biogeochemistry, functional traits, 40 competition, nutrients 2 41 Introduction 42 The elemental composition of matter determines its physical and chemical properties. 43 Living matter, mainly composed of carbon (C), oxygen (O), hydrogen, nitrogen (N) and 44 phosphorus (P), is not an exception to this rule. Like the genome, the elementome, 45 defined as the concentration of elements within organisms (Peñuelas et al., 2019), is the 46 quintessence of biological systems. Not surprisingly, the elemental composition and 47 stoichiometry (i.e., the ratio of elemental concentrations, such as N:P) of organisms have 48 been shown to be excellent predictors of their ancestry, their functional traits and the 49 ecological strategies they follow (Wright et al., 2004; Peñuelas et al., 2019; Sardans et 50 al., 2021; Fernández‐Martínez et al., 2021a). Therefore, two organisms presenting very 51 different elementomes will, most likely, occupy different ecological niches. Assessing 52 ecological niches amongst organisms is an important feature for ecological research, 53 because one of the main mechanisms behind the positive relationship between diversity 54 and ecosystem functioning is niche partitioning (Cardinale, 2011) – this is that different 55 organisms use a different set of resources within the ecosystem. 56 In this review, I compile evidence on how plant elementomes depend on their functional 57 traits and habitats, and how they can be defined. Later, I review the role of plasticity on 58 the elemental composition of plants, ecosystems and their functioning. Finally, I propose 59 the use of elementome diversity and distance as universal tools to assess functional 60 diversity, competition and global change impacts on ecosystem functioning. 61 II. Elementome ecology: niches, evolution and plasticity 62 Specificity of elementomes: the case of C:N:P stoichiometry 63 Since the early stages of life, organisms have evolved to adapt to new colonisable 64 environments and coexist with other organisms. Thus, the large diversity of forms and 65 functions evolved have generated a large diversity of elementomes. The elemental 66 composition of plants can be determined by different analytical methods, such as 67 isotope-ratio mass spectrometry (IRMS), providing concentrations of C and N, and their 68 isotopic signal (i.e., δ13C and δ15C) and inductively coupled plasma mass spectrometry 69 (ICP-MS), that can provide the concentration of a large number of elements, including 70 trace elements. By comparing the C:N:P ratio across organisms, directly derived from 71 differences in their elementomes, we can get an overview of how species-specific 72 elementomes (Sardans et al., 2021; Fernández‐Martínez et al., 2021b) are related to 73 their traits and habitats (Wright et al., 2004; Peñuelas et al., 2019). In general terms, 74 plant elementomes increase C with respect to N and P as organisms get bigger and 75 develop complex structures (Figure 1, Table 1). 3 76 Analysing the elementomes of morphologically simple plants is methodologically easy, 77 but it can be challenging for complex organisms (e.g., trees) if the aim is to capture the 78 composition of their entire organism and allow a fair comparison with simpler plants. 79 Comparisons amongst complex plants are usually based on the elemental composition 80 of their leaves because they are the photosynthetically active part of the plants, where 81 most of the metabolic processes take place, and are easy to sample. However, plant- 82 leaf elementomes reveal stoichiometric ratios that are far from the average C:N:P ratios 83 of the entire biomass in some species (Balboa-Murias et al., 2006). 84 To estimate the elementome of big organisms it is necessary to sample all different 85 organs of the plant (e.g., roots, stems, leaves), estimate each organ’s biomass using 86 allometric equations or similar methods, and then calculate an overall concentration for 87 each element. To achieve a community-level elementome, the concentration of each 88 element would be the mean of its concentration weighted by the biomass of each species 89 within the community. Once this exercise has been made, estimates clearly reflect the 90 vast investment in C-rich structures (wood) in trees compared to that of phytoplankton, 91 bryophytes, grasses or even shrubs, and denote clear differences between the functional 92 traits and ecological niches of these organisms (Table 1). 93 Moreover, the presence of determined elementomes in an ecosystem could indicate the 94 existence of competitive (i.e., similar elementomes) and/or facilitating interactions 95 amongst species (e.g., epiphytes growing in trees, shade tolerant plants in the 96 understory) and different successional stages. These comparisons based on C:N:P 97 stoichiometry, however, have been made only for simplicity purposes, following (Elser et 98 al., 2000). To achieve a clear picture of elementome differences amongst organisms we 99 must consider the differences in their elemental concentrations because very high N:P 100 values may be the result of high N or low P concentrations. 101 Adaptive value of elementomes 102 Phylogenetic analyses have shown that plant functional traits, such as defence or 103 reproduction, have co-evolved with the elemental composition of organisms following an 104 adaptive rather than a random model of evolution (Fernández-Martínez et al., 2018, 105 2019; Bai et al., 2019). Consequently, elementomes have been repeatedly shown to be 106 phylogenetically conserved (Sardans et al., 2021; Fernández‐Martínez et al., 2021b), 107 further indicating the adaptive value of maintaining a particular elemental composition 108 for each species: their biogeochemical niche (Peñuelas et al., 2019). The concentrations 109 of C, N and P are, by far, the most commonly used to study organism features because 110 of their strong involvement in metabolic processes (Sterner & Elser, 2002). However, a 4 111 three-dimensional elementome may be missing key information that would otherwise 112 have been provided by additional elements (Kaspari, 2021). Recent studies have found 113 that the concentration of elements such as K, Mg, Ca, and S, have also been conserved 114 through evolutionary history and, thus, they should have co-evolved with several 115 functional traits (Bai et al., 2019; Sardans et al., 2021; Fernández‐Martínez et al., 2021b). 116 Elementomes beyond C, N and P 117 Micronutrients and trace elements are essential to regulate vital plant functions such as 118 photosynthesis and chlorophyll production (Mg, Fe), drought tolerance (K), defence (Si) 119 or reproduction (Zn) (Couturier et al., 2013; Sardans & Peñuelas, 2015; Fernández- 120 Martínez et al., 2017; Fernández Honaine et al., 2021) and, hence, different 121 concentrations denote differences in plant traits and functions (Kaspari, 2021). 122 Concentrations of trace elements may also indicate hyperaccumulation and/or 123 hypertolerance of heavy metals (e.g., As, Cd, Ni) (Reeves et al., 2018). Consequently, 124 recent literature has shown that analysing additional elements, together with N and P, 125 improves the prediction of ecosystem functioning and the definition of species-specific 126 elementomes compared to using only N and P (Huang et al., 2019; Ågren & Weih, 2020; 127 Hofmann et al., 2021; Fernández‐Martínez et al., 2021b). 128 Moreover, stable isotopes can also provide valuable information regarding plant traits 129 and functions. δ13C can be used to differentiate between sources of CO2 use in 130 freshwater macrophytes (Maberly et al., 2015) and water-use efficiency in terrestrial 131 plants (Ding et al., 2021), while differences in δ15N can be used to infer contrasting 132 sources of N use (Qian et al., 2021). Similarly, δ18O isotopes have been used to assess 133 the hydrological niche of coexisting plants (Ding et al., 2021). A good estimation of the 134 elementome will thus depend on a trade-off between maximising the information included 135 and reducing the potential noise generated by each additional element. One way to deal 136 with this trade-off, while losing the minimum information as possible, is to perform a 137 dimensionality reduction analysis such as a principal components analysis (Peñuelas et 138 al., 2019; Sardans et al., 2021). The suitable dimensionality of an elementome will thus 139 depend on the biological function under study. 140 Elemental plasticity: abiotic and biotic factors 141 For an organism to be and function as a determined organism it has to maintain its 142 elementome within certain thresholds. This homeostatic regulation is evidenced by the 143 larger variability in the concentration of elements amongst species compared to that 144 within individuals of the same species (Sardans et al., 2021) and lies behind the 145 foundation of the biogeochemical niche. Plasticity, however, is also an important attribute 5 146 of plant elementomes, shifting in response to biotic and abiotic changes within the limits 147 of their biogeochemical niche. Hence, although species elemental concentrations are 148 usually well correlated with environmental gradients (Bai et al., 2019; Fernández‐ 149 Martínez et al., 2021b) changes in community-level elementomes are expected to be 150 larger due to changes in the community assembly than due to the elementome plasticity 151 of the species. 152 Altered global biogeochemical cycles are changing the elemental composition of 153 organisms and ecosystems, both directly and indirectly (Figure 2). The alteration of 154 those cycles have increased the availability of nutrients for plants, creating nutrient 155 imbalances (Peñuelas et al., 2020b) that could hinder ecosystem functioning 156 (Fernández‐Martínez et al., 2020). Increasing atmospheric CO2 concentrations have 157 been shown to reduce the concentration of N and P in trees (Peñuelas et al., 2020a) 158 potentially constraining the CO2 fertilisation effect globally (Wang et al., 2020). Grazing, 159 fertilisation and increased atmospheric N deposition are also known to alter plant 160 elementomes (Mao et al., 2018; Du & Gao, 2021; Vallicrosa et al., 2021). These changes 161 in nutrient availability have also been related to biodiversity loss and species 162 replacement, further modifying how ecosystems work (Soons et al., 2017). On top of 163 that, the alteration of the C cycle is increasing global temperatures and precipitation 164 patterns, which can directly alter plant elementomes (Vallicrosa et al., 2021) and 165 ecosystem biodiversity. 166 On the other hand, elementome differences amongst species increase with their 167 likelihood of coexistence, suggesting that communities are assembled in a way that 168 minimises competition and/or that species tend to differentiate their niches in order to 169 reduce interspecific competition (Peñuelas et al., 2019; Fernández‐Martínez et al., 170 2021b). Compared to monocultures, plants growing in diverse communities change their 171 elemental compositions, to either reduce or increase N and P concentrations depending 172 on the species (Guiz et al., 2016, 2018; Zhu et al., 2020; Wu et al., 2020). The suggested 173 mechanisms behind those elemental shifts (e.g., favouring niche partitioning, reduction 174 of soil nutrients, competition for light forcing physiological changes to increase C-rich 175 structures) indicate that competition is an important driver of elementome plasticity. 176 III. Elements, diversity and ecosystem functioning 177 Climate and community-weighted elementome means (Figure 3a) have been proven to 178 be very useful to understand ecosystem functioning (Craven et al., 2018; Fernández‐ 179 Martínez et al., 2020). Similarly, literature shows that biodiversity, and particularly 180 functional diversity, increases productivity and stability of ecosystems under natural and 6 181 experimental conditions (Liang et al., 2016; Craven et al., 2018) through selection and 182 complementarity effects (Cardinale, 2011). The former is based on an increased 183 probability of having species with a strong effect on ecosystem functioning (e.g., very 184 productive) and the latter is based on different species having contrasting ecological 185 niches. As mentioned above, elementomes are an excellent proxy for both plant 186 productivity and their ecological niches. However, despite the potential of elementomes 187 to define traits and niches they have never been used to assess ecosystem diversity nor 188 as a proxy of functional diversity – apart from including N and P within a larger set of 189 traits (Craven et al., 2018). I here propose the use of elementome diversity as a measure 190 of functional diversity and elementome distance as a measure of competition between 191 organisms and temporal changes in plant and ecosystem functioning. 192 Elementome diversity and distance 193 Elementome diversity (ED) can be estimated by means of the elementome dissimilarity 194 index for a given a community formed by S species for which their relative abundance is 195 pi and their concentration of E elements is known, similarly to state-of-the-art functional 196 diversity indices (Villéger et al., 2008). ED is then estimated as the sum of all pair-wise 197 elementome Euclidean distances between species (Di,j) weighted by the minimum 198 relative abundance of the pair of species (min (pi, pj)). Finally, the sum of all elementome 199 distances is divided by the number of elements used (E) so the metric is comparable 200 across metrics calculated with different elementome dimensionality: ∑𝑆−1 𝑆 𝑖=1 ∑𝑗=𝑖+1 𝐷𝑖𝑗 × min⁡(𝑝𝑖 , 𝑝𝑗 ) 201 𝐸𝐷 = 𝐸 202 ED correlates well with other indices of species and functional diversity and so it is 203 expected to present similar relationships with ecosystem functioning to those reported 204 for these other metrics (Figure 3). Depending on the research question, though, ED 205 should be calculated over different compartments of the ecosystem (e.g., autotrophs, 206 decomposers). From a mathematical point of view, ED increases with species diversity 207 and elementome distance between species, as long as species diversity and 208 elementome distance amongst species are both different from 0 (Figure 3e). This 209 property ensures that estimates of ED comprise both ecological features: diversity and 210 elementome distance amongst species. 211 Measures of ED have three main advantages over traditional indices of diversity, such 212 as species, functional or phylogenetic diversity metrics, when investigating how diversity 213 relates to ecosystem functioning: i) given that ED is based on elementomes, it comprises 214 multiple functional traits of the organisms, which in practice makes it a functional diversity 7 215 index. In a way, ED may include hidden functional diversity derived from traits typically 216 overlooked by traditional metrics (e.g., stinging hairs of Urtica dioica, rich in Si and useful 217 to avoid herbivores). Additionally, functional diversity indices are expected to relate to 218 ecosystem functioning better than species diversity indices (Díaz & Cabido, 2001). ii) 219 The comparison of ED is universal across all ecosystem types and can include all living 220 organisms, because they are all composed of the same elements. This is hardly the case 221 when using functional traits: several tree traits are missing in herbs (e.g., wood density), 222 and differences become even larger when comparing more functionally distant 223 organisms (e.g., fungi, animals and plants). Additionally, measuring some frequently 224 used functional traits (e.g., photosynthesis, respiration) can be more resource 225 consuming (time, equipment) than analysing elementomes (with the limitation of big 226 roots). iii) Although two phylogenetically close organisms are expected to be functionally 227 closer, measures of phylogenetic diversity may not properly capture functional diversity 228 under strong cases of convergent evolution whereas ED would reflect those functional 229 differences independently of their ancestry. These advantages highlight the potential of 230 ED to study diversity and ecosystem functioning from a holistic perspective that can 231 integrate a very large array of living organisms. 232 On the other hand, ED could also be used as a measure of the strength of competition 233 between plants if all species are randomly distributed in space (i.e., the lower ED, the 234 stronger the competition because species occupy the same niche). Otherwise, ED could 235 be multiplied by a coefficient of spatial aggregation (<1 when aggregation occurs) or, 236 ideally, elementome distances could be calculated for neighbouring plants within a 237 determined radius of influence to assess competition amongst individuals. Additionally, 238 elementome distances could be used to monitor shifts in elementomes related to 239 environmental changes (e.g., altered nutrient availability, climate or community 240 composition) at the individual, population or community scales. Increasing elementome 241 distances in plants or communities could be related to changes in plant performance or 242 ecosystem functioning (e.g., increased growth for individual plants or gross primary 243 production for ecosystems), and even changes in the community assemblage as 244 discussed in section II. Obtaining long time series of elementomes could thus be used 245 as a tool for monitoring ecosystem functioning and the impacts of global change on 246 Earth’s biosphere. 247 IV. Conclusions 248 A large body of evidence suggests that the elemental composition of organisms define 249 their ecological niches and how their ecosystems work, and that elements other than C, 8 250 N and P should be used in order capture the diversity of traits between them. Changes 251 in plant elementomes due to changes in biotic and abiotic conditions can also help us 252 monitor how global change impacts ecosystems and their functioning. A focus on 253 elementome diversity and distance looks like a promising approach to study ecosystem 254 functioning under global change. 255 256 Acknowledgements 257 This research was supported by a postdoctoral fellowship of the Research Foundation – 258 Flanders (FWO) and by a fellowship from ”la Caixa” Foundation (ID 100010434), code: 259 LCF/BQ/PI21/11830010. I thank Drs. Catherine Preece, Josep Peñuelas, and Jordi 260 Sardans for their invaluable help in discussing issues regarding the subject of this review. 261 I thank Dr. Roberto Molowny for his advice on sigma notation. I also thank the three 262 anonymous referees and the editor that, with their insightful comments, helped me 263 improve this manuscript. 264 265 Data availability 266 The code supporting the findings of this study is freely available at Figshare: 267 https://doi.org/10.6084/m9.figshare.14748261.v2. 268 269 References 270 Ågren GI, Weih M. 2020. Multi-Dimensional Plant Element Stoichiometry—Looking 271 Beyond Carbon, Nitrogen, and Phosphorus. Frontiers in Plant Science 11: 23. 272 Baek G, Bae EJ, Kim C. 2020. Nutrient stocks of Japanese blue oak (Quercus glauca 273 Thunb.) stands on different soil parent materials. Forest Science and Technology 16: 274 180–187. 275 Bai K, Lv S, Ning S, Zeng D, Guo Y, Wang B. 2019. Leaf nutrient concentrations 276 associated with phylogeny, leaf habit and soil chemistry in tropical karst seasonal 277 rainforest tree species. Plant and Soil 434: 305–326. 278 Balboa-Murias MA, Rojo A, Álvarez JG, Merino A. 2006. Carbon and nutrient stocks 279 in mature Quercus robur L. stands in NW Spain. Annals of Forest Science 63: 557– 280 565. 9 281 Cardinale BJ. 2011. Biodiversity improves water quality through niche partitioning. 282 Nature 472: 86–91. 283 Couturier J, Touraine B, Briat JF, Gaymard F, Rouhier N. 2013. The iron-sulfur 284 cluster assembly machineries in plants: Current knowledge and open questions. 285 Frontiers in Plant Science 4: 1–22. 286 Craven D, Eisenhauer N, Pearse WD, Hautier Y, Isbell F, Roscher C, Bahn M, 287 Beierkuhnlein C, Bönisch G, Buchmann N, et al. 2018. Multiple facets of biodiversity 288 drive the diversity–stability relationship. Nature Ecology and Evolution 2: 1579–1587. 289 Díaz S, Cabido M. 2001. Vive la différence: Plant functional diversity matters to 290 ecosystem processes. Trends in Ecology and Evolution 16: 646–655. 291 Ding Y, Nie Y, Chen H, Wang K, Querejeta JI. 2021. Water uptake depth is 292 coordinated with leaf water potential, water-use efficiency and drought vulnerability in 293 karst vegetation. New Phytologist 229: 1339–1353. 294 Du C, Gao Y. 2021. Grazing exclusion alters ecological stoichiometry of plant and soil 295 in degraded alpine grassland. Agriculture, Ecosystems and Environment 308: 107256. 296 Duarte CM. 1992. Nutrient concentration of aquatic plants: Patterns across species. 297 Limnology and Oceanography 37: 882–889. 298 Elser JJ, Fagan WF, Denno RF, Dobberfuhl DR, Folarin A, Huberty A, Interlandi S, 299 Kilham SS, McCauley E, Schulz KL, et al. 2000. Nutritional constraints in terrestrial 300 and freshwater food webs. Nature 408: 578–580. 301 Fernández-Martínez M, Llusià J, Filella I, Niinemets Ü, Arneth A, Wright IJ, Loreto 302 F, Peñuelas J. 2018. Nutrient-rich plants emit a less intense blend of volatile 303 isoprenoids. New Phytologist 220: 773–784. 304 Fernández-Martínez M, Pearse I, Sardans J, Sayol F, Koenig WD, LaMontagne JM, 305 Bogdziewicz M, Collalti A, Hacket-Pain A, Vacchiano G, et al. 2019. Nutrient 306 scarcity as a selective pressure for mast seeding. Nature Plants 5: 1222–1228. 307 Fernández-Martínez M, Vicca S, Janssens IA, Espelta JM, Peñuelas J. 2017. The 308 role of nutrients, productivity and climate in determining tree fruit production in 309 European forests. New Phytologist 213: 669–679. 310 Fernández‐Martínez M, Corbera J, Cano-Rocabayera O, Sabater F, Preece C. 311 2021a. Do Bryophyte Elemental Concentrations Explain Their Morphological Traits? 312 Plants 10: 1581. 10 313 Fernández‐Martínez M, Preece C, Corbera J, Cano O, Garcia‐Porta J, Sardans J, 314 Janssens IA, Sabater F, Peñuelas J. 2021b. Bryophyte C:N:P stoichiometry, 315 biogeochemical niches and elementome plasticity driven by environment and 316 coexistence. Ecology Letters 24: 1375–1386. 317 Fernández‐Martínez M, Sardans J, Musavi T, Migliavacca M, Iturrate‐Garcia M, 318 Scholes RJ, Peñuelas J, Janssens IA. 2020. The role of climate, foliar stoichiometry 319 and plant diversity on ecosystem carbon balance. Global Change Biology 26: 7067– 320 7078. 321 Fernández Honaine M, Benvenuto ML, Montti L, Natal M, Borrelli NL, Alvarez MF, 322 Altamirano SM, De Rito M, Osterrieth ML, Fernández Honaine M, et al. 2021. How 323 Are Systematics and Biological and Ecological Features Related to Silica Content in 324 Plants? A Study of Species from Southern South America. International Journal of 325 Plant Sciences 182: 210–219. 326 Frias A, Quesada A, Vela D. 2013. Ecological relationships and stoichiometry within a 327 Maritime Antarctic watershed. Antarctic Science 25: 191–197. 328 Garrett LG, Smaill SJ, Beets PN, Kimberley MO, Clinton PW. 2021. Impacts of forest 329 harvest removal and fertiliser additions on end of rotation biomass, carbon and nutrient 330 stocks of Pinus radiata. Forest Ecology and Management 493: 119161. 331 Guiz J, Ebeling A, Eisenhauer N, Hacker N, Hertzog L, Oelmann Y, Roscher C, 332 Wagg C, Hillebrand H. 2018. Interspecific competition alters leaf stoichiometry in 20 333 grassland species. Oikos 127: 903–914. 334 Guiz J, Hillebrand H, Borer ET, Abbas M, Ebeling A, Weigelt A, Oelmann Y, 335 Fornara D, Wilcke W, Temperton VM, et al. 2016. Long-term effects of plant diversity 336 and composition on plant stoichiometry. Oikos 125: 613–621. 337 Hofmann P, Clark A, Hoffmann P, Chatzinotas A, Harpole WS, Dunker S. 2021. 338 Beyond nitrogen: phosphorus – estimating the minimum niche dimensionality for 339 resource competition between phytoplankton. Ecology Letters 24: 761-771. 340 Huang J, Liu W, Li S, Song L, Lu H, Shi X, Chen X, Hu T, Liu S, Liu T. 2019. 341 Ecological stoichiometry of the epiphyte community in a subtropical forest canopy. 342 Ecology and Evolution 9: 14394–14406. 343 Kaspari M. 2021. The Invisible Hand of the Periodic Table: How Micronutrients Shape 344 Ecology. Annual Review of Ecology, Evolution, and Systematics 52: 199–219. 11 345 Liang J, Crowther TW, Picard N, Wiser S, Zhou M, Alberti G, Schulze E-D, 346 McGuire AD, Bozzato F, Pretzsch H, et al. 2016. Positive biodiversity-productivity 347 relationship predominant in global forests. Science 354: aaf8957–aaf8957. 348 Maberly SC, Berthelot SA, Stott AW, Gontero B. 2015. Adaptation by macrophytes 349 to inorganic carbon down a river with naturally variable concentrations of CO2. Journal 350 of Plant Physiology 172: 120–127. 351 Mao Q, Lu X, Mo H, Mo J. 2018. Effects of simulated N deposition on foliar nutrient 352 status, N metabolism and photosynthetic capacity of three dominant understory plant 353 species in a mature tropical forest. Science of the Total Environment 610: 555–562. 354 Martiny AC, Vrugt JA, Lomas MW. 2014. Concentrations and ratios of particulate 355 organic carbon, nitrogen, and phosphorus in the global ocean. Scientific Data 1: 1–7. 356 McGroddy ME, Daufresne T, Hedin OL. 2004. Scaling of C :N :P Stoichiometry in 357 Forests Worldwide : Implications of Terrestrial Redfield- Type Ratios. Ecology 85: 2390– 358 2401. 359 Peñuelas J, Fernández-Martínez M, Vallicrosa H, Maspons J, Zuccarini P, 360 Carnicer J, Sanders TGM, Krüger I, Obersteiner M, Janssens IA, et al. 2020a. 361 Increasing atmospheric CO2 concentrations correlate with declining nutritional status of 362 European forests. Communications Biology 3: 125. 363 Peñuelas J, Fernández‐Martínez M, Ciais P, Jou D, Piao S, Obersteiner M, Vicca 364 S, Janssens IA, Sardans J. 2019. The bioelements, the elementome, and the 365 biogeochemical niche. Ecology 100: e02652. 366 Peñuelas J, Janssens IA, Ciais P, Obersteiner M, Sardans J. 2020b. Anthropogenic 367 global shifts in biospheric N and P concentrations and ratios and their impacts on 368 biodiversity, ecosystem productivity, food security, and human health. Global Change 369 Biology 26: 1962-1985. 370 Perez-Quezada J, Pérez CA, Brito CE, Fuentes JP, Gaxiola A, Aguilera-Riquelme 371 D, Lopatin J. 2021. Biotic and abiotic drivers of carbon, nitrogen and phosphorus stocks 372 in a temperate rainforest. Forest Ecology and Management 494: 119341. 373 Qian J, Jin W, Hu J, Wang P, Wang C, Lu B, Li K, He X, Tang S. 2021. Stable 374 isotope analyses of nitrogen source and preference for ammonium versus nitrate of 375 riparian plants during the plant growing season in Taihu Lake Basin. Science of the 376 Total Environment 763: 143029. 12 377 Redfield AC. 1934. On the proportions of organic derivatives in sea water and their 378 relation to the composition of plankton. James Johnstone Memorial 176: 176–192. 379 Reeves RD, Baker AJM, Jaffré T, Erskine PD, Echevarria G, van der Ent A. 2018. 380 A global database for plants that hyperaccumulate metal and metalloid trace elements. 381 New Phytologist 218: 407–411. 382 Sardans J, Peñuelas J. 2015. Potassium: a neglected nutrient in global change. 383 Global Ecology and Biogeography 24: 261–275. 384 Sardans J, Vallicrosa H, Zuccarini P, Farré-Armengol G, Fernández-Martínez M, 385 Peguero G, Gargallo-Garriga A, Ciais P, Janssens IA, Obersteiner M, et al. 2021. 386 Empirical support for the biogeochemical niche hypothesis in forest trees. Nature 387 Ecology & Evolution 5: 184–194. 388 Soons MB, Hefting MM, Dorland E, Lamers LPM, Versteeg C, Bobbink R. 2017. 389 Nitrogen effects on plant species richness in herbaceous communities are more 390 widespread and stronger than those of phosphorus. Biological Conservation 212: 390– 391 397. 392 Sterner RW, Elser JJ. 2002. Ecological stoichiometry : the biology of elements from 393 molecules to the biosphere. Princeton University Press. Princeton, NJ, USA. 394 Vallicrosa H, Sardans J, Ogaya R, Fernández PR, Peñuelas J. 2021. Short-Term N- 395 Fertilization Differently Affects the Leaf and Leaf Litter Chemistry of the Dominant 396 Species in a Mediterranean Forest under Drought Conditions. Forests 12: 605. 397 Villéger S, Mason NWH, Mouillot D. 2008. New multidimensional functional diversity 398 indices for a multifaceted framework in functional ecology. Ecology 89: 2290–2301. 399 Wang S, Zhang Y, Ju W, Chen JM, Ciais P, Cescatti A, Sardans J, Janssens IA, 400 Wu M, Berry JA, et al. 2020. Recent global decline of CO2 fertilization effects on 401 vegetation photosynthesis. Science 370: 1295–1300. 402 Wang Z, Pi C, Li X, Bao W. 2019. Elevational patterns of carbon, nitrogen and 403 phosphorus in understory bryophytes on the eastern slope of Gongga Mountain, China. 404 Journal of Plant Ecology 12: 781–786. 405 Wright IJ, Reich PB, Westoby M, Ackerly DD, Baruch Z, Bongers F, Cavender- 406 Bares J, Chapin T, Cornelissen JHC, Diemer M, et al. 2004. The worldwide leaf 407 economics spectrum. Nature 428: 821–7. 408 Wu X, Du X, Fang S, Kang J, Xia Z, Guo Q. 2020. Impacts of competition and 13 409 nitrogen addition on plant stoichiometry and non-structural carbohydrates in two larch 410 species. Journal of Forestry Research 32: 2087–2098. 411 Zhu D, Hui D, Wang M, Yang Q, Yu S. 2020. Light and competition alter leaf 412 stoichiometry of introduced species and native mangrove species. Science of the Total 413 Environment 738: 140301. 414 415 Competing interests 416 The author declares no competing interests. 14 417 Figures 418 Figure 1. Elemental singularities. Several examples of organisms with very 419 characteristic elemental concentrations for particular elements. Arrows indicate 420 anomalously high or low concentrations of a given element relative to other plants. For 421 marine plankton and Quercus robur, numbers indicate their C:N:P atomic ratio. Picture 422 credits: Salicornia europaea by FortBienVert, Urtica dioica by Peter O'Connor and 423 Alyssum murale by Matt Lavin (CC BY-SA 2.0). Quercus robur by Gikü (public domain). 424 Fontinalis antipyretica by Jordi Corbera. 425 15 426 Figure 2. Impacts of altered biogeochemical cycles on plant elementomes. Fossil 427 fuel burning and the use of fertilisers are the main contributing factors to changing global 428 biogeochemical cycles. Altered biogeochemical cycles change plant elementomes 429 directly by altering the environmental availability of nutrients (a) which, in turn, can modify 430 the composition of the community and lead to loss of biodiversity (b) and changes in 431 microclimatic conditions (c). Global warming (c) changes the environmental conditions 432 for plants, leading to physiological changes such as increased sclerophylly, should 433 climate become drier (i.e., in Mediterranean climates). Further changes in community 434 composition and biodiversity loss take place if the species cannot withstand these new 435 environmental conditions (b), hence creating a feedback loop because of altered nutrient 436 uptake rates (a). Additional changes in elemental composition to those caused by (a) 437 and (c) can then occur by altered competitive interactions amongst the new or fewer 438 species present in the community. The mechanisms responsible for those elemental 439 shifts are highlighted in the yellow box (differences in nutrient uptake, community 440 composition, competition and environmental constraints). Further details and references 441 can be found in section Elemental plasticity: abiotic and biotic factors. The size of the C, 442 N and P circles indicate an arbitrary concentration of elements. Blue arrows indicate 443 direct effects. Pattern vector created by pch.vector - www.freepik.com. 444 445 446 447 16 448 Figure 3. Plant elemental concentration and elementome diversity. The elemental 449 concentration of nutrients generally increases ecosystem functioning (a), following a 450 pattern of diminishing returns on investment (i.e., when nutrient concentrations are high, 451 factors other than nutrients limit ecosystem functioning). Elementome diversity is 452 expected to be similarly related to ecosystem functioning (b) as all available niches 453 become progressively occupied, potentially, with smaller and less productive organisms 454 (e.g., epiphytes). Other relationships (dashed lines) could also emerge when comparing 455 different successional stages due to the appearance of plants with large contributions on 456 ecosystem functioning (e.g., trees). Panel c) shows communities containing three and 457 four different species. Despite having the same number of species, the green community 458 has higher elementome diversity than the blue one because of a larger elemental 459 difference amongst its species. The grey community shows the minimum elementome 460 diversity, because of having only three species and them being very similar. Panel d) 461 shows the relationship between elementome diversity and Shannon’s diversity (H) and 462 two indices of functional diversity, functional dispersion (FDis) and Rao’s Q from a 463 simulation of 100 communities containing a random assemblage of eight species with 464 contrasting elementomes. Panel e) shows how elemental diversity increases as a 465 function of Shannon’s H and interspecific variability in elementomes, measured as the 466 coefficient of variation (sd · mean-1) in elemental concentrations amongst species. Code 467 to perform simulations in panels d) and e) is available at Figshare: 468 https://doi.org/10.6084/m9.figshare.14748261.v2. 17 469 470 471 18 472 Table 1. Atomic C:N:P stoichiometry of different organisms in contrasting 473 habitats. Examples are meant to be a brief overview of the differences between 474 organisms in relation with their traits and habitats rather than an exhaustive literature 475 review. Values indicate means or median estimates for the group, depending on the 476 centrality measure provided by the reference. Organisms are ordered by C:P ratio. C:N:P Organism Structure Habitat Climate/region Reference 106:16:1 Marine plankton Whole organism Ocean Global 1 110:16:1 Phytoplankton Whole organism Diverse Global 2 145:8:1 Bryophytes Whole organism Springs Mediterranean 3 162:22:1 Marine plankton Whole organism Ocean Global 4 215:23:1 Filamentous green algae Whole organism Lake Antarctica 5 256:27:1 Plants Leaf Freshwater Global 6 435:20:1 Angiosperms Leaf Marine Global 2 500:24:1 Angiosperms Leaf Freshwater Global 2 511:13:1 Grasses, weeds (grazed) Whole organism Alpine meadow Tibetan Plateau 7 591:17:1 Bryophytes Whole organism Freshwater Antarctica 5 660:24:1 Bryophytes Whole organism Epiphytic Subtropical 8 799:27:1 Plants Leaf Terrestrial Global 6 800:49:1 Macroalgae Whole organism Ocean Global 2 878:18:1 Bryophytes Whole organism Soil Subalpine 9 964:29:1 Bryophytes Whole organism Soil Antarctica 5 1334:28:1 Trees Leaf Terrestrial Global 10 1396:20:1 Grasses, weeds Whole organism Alpine meadow Tibetan Plateau 7 1802:32:1 Lichens Whole organism Epiphytic Subtropical 8 2340:26:1 Lichens Whole organism Soil Antarctica 5 3686:25:1 Quercus glauca Aboveground biomass Terrestrial Temperate 11 3989:08:1 Pinus radiata Aboveground biomass Terrestrial Temperate 12 4150:32:1 Quercus robur Whole organism Terrestrial Temperate 13 5507:463:1 Cyanobacterial mats Whole organism Freshwater Antarctica 5 8380:40:1 Rainforest Whole trees Terrestrial Temperate 14 477 References: 1. Redfield, 1934; 2. Duarte, 1992; 3. Fernández‐Martínez et al., 2021; 4. Martiny et al., 2014; 478 5. Frias et al., 2013; 6. Elser et al., 2000; 7. Du & Gao 2021; 8. Huang et al., 2019; 9. Wang et al., 2019; 10. 479 McGroddy et al., 2004; 11. Baek et al., 2020; 12. Garrett et al., 2021; 13. Balboa-Murias et al., 2006. 14. 480 Perez-Quezada et al., 2021. 481 19