干旱区生态环境知识资源中心

The niche concept is essential to understanding how biotic and abiotic factors regulate the abundance and distribution of living entities, and how these organisms utilize, affect and compete for resources in the environment. However, it has been challenging to determine the number and types of important niche dimensions. By contrast, there is strong mechanistic theory and empirical evidence showing that the elemental composition of living organisms shapes ecological systems, from organismal physiology to food web structure. We propose an approach based on a multidimensional elemental view of the ecological niche. Visualizing the stoichiometric composition of individuals in multivariate space permits quantification of niche dimensions within and across species. This approach expands on previous elemental characterizations of plant niches, and adapts metrics of niche volume, overlap and nestedness previously used to quantify isotopic niches. We demonstrate the applicability of the multidimensional stoichiometric niche using data on carbon, nitrogen, and phosphorus of terrestrial and freshwater communities composed by multiple trophic groups. First, we calculated the stoichiometric niche volumes occupied by terrestrial and freshwater food webs, by trophic groups, by individual species, and by individuals within species, which together give a measure of the extent of stoichiometric diversity within and across levels of organization. Then we evaluated complementarity between these stoichiometric niches, through metrics of overlap and nestedness. Our case study showed that vertebrates, invertebrates, and primary producers do not overlap in their stoichiometric niches, and that large areas of stoichiometric space are unoccupied by organisms. Within invertebrates, niche differences emerged between freshwater and terrestrial food webs, and between herbivores and non-herbivores (detritivores and predators). These niche differences were accompanied by changes in the covariance structure of the three elements, suggesting fundamental shifts in organismal physiology and/or structure. We also demonstrate the sensitivity of results to sample size, and suggest that representative sampling is better than rarefaction in characterizing the stoichiometric niche occupied by food webs. Overall, our approach demonstrates that stoichiometric traits provide a common currency to estimate the dimensionality of stoichiometric niches, and help reduce and rationalize the number of axis required to characterize communities.