Animal stoichiometry affects fundamental processes ranging from organismal physiology to global element cycles. However, it is unknown whether animal stoichiometry follows predictable scaling relationships with body mass and whether adaptation to life on land or water constrains patterns of elemental allocation. To test both interspecific and intraspecific body-size scaling relationships of the nitrogen (N), phosphorus (P), and N:P content of animals, we used a subset of the StoichLife database encompassing 9,933 individual animals (vertebrates and invertebrates) belonging to 1,543 species spanning 10 orders of magnitude of body size from terrestrial, freshwater, and marine realms. Across species, body mass did not explain much variation in %N and %P composition, although the %P of invertebrates decreased with size. The effects of body size on species elemental content were small in comparison to the effects of taxonomy. Body size was a better predictor of intraspecific than interspecific elemental patterns. Between 42 to 45% in intraspecific stoichiometric variation was explained by body size for 27% of vertebrate species and 35% of invertebrate species. Further, differences between organisms inhabiting aquatic and terrestrial realms were observed only in invertebrate interspecific %N, suggesting that the realm does not play an important role in determining elemental allocation of animals. Based on our analysis of the most comprehensive animal stoichiometry database, we conclude that (i) both body size and realm are relatively weak predictors of animal stoichiometry across taxa, and (ii) body size is a good predictor of intraspecific variation in animal elemental content, which is consistent with tissue-scaling relationships that hold broadly across large groups of animals. This research reveals a lack of general scaling patterns in the elemental content across animals and instead points to a large variation in scaling relationships within and among lineages.### Competing Interest StatementThe authors have declared no competing interest.

At the end of the Pleistocene, tropical forests in South America underwent three drastic changes that are analogous to today: the climate warmed, people populated the region and started to use and domesticate of part of the native forest plants, and the abundance of seed-dispersing megafauna collapsed. Here, we evaluate the impact of these three changes on tree populations in South America, using pollen time series of >80 tree and palm genera across 10 sites collected over a 100,000 yr period spanning the end of the Pleistocene and Holocene, and linking this to genus-specific traits that are indicators of climate responses, usefulness to humans, and dispersal by megafauna. We find no support for consistent effects of climate warming, and find that megafauna-dispersed tree genera increased – instead of decreased – in abundance after megafauna decline, indicating that climate warming and megafauna decline were no major drivers of changes in tree populations. However, we find that edible tree and palm species increased in abundance. These results are consistent with the idea that people favoured tree taxa that were useful for consumption of their large seeds and fruits, and in this way took over the role of lost seed dispersers and artificially increased the abundance of previously megafauna-dispersed tree taxa.