Investigating the biotic and abiotic drivers of body size disparity in communities of non-volant terrestrial mammals

AimThe species that compose local communities possess unique sets of
functional and ecological traits that can be used as indicators of biotic
and abiotic variation across space and time. Body size is a particularly
relevant trait because species with different body sizes typically have
different life history strategies and occupy distinct niches. Here we used
the body sizes of non-volant (i.e., non-flying) terrestrial mammals to
quantify and compare the body size disparity of mammal communities across
the globe. LocationGlobal Time periodPresent Major taxa studiedNon-volant
terrestrial mammals MethodsWe used IUCN range maps of 3,982 terrestrial
mammals to identify 1,876 communities. We then combined diet data with
data on climate, elevation, and anthropogenic pressures to evaluate these
variables’ relative importance on the observed body size dispersion of
these communities and its deviation from a null model. ResultsDispersion
is significantly greater than expected in 54% of communities and
significantly less than expected in 30% of communities. The number of very
large species, continent, range sizes, diet disparity, and annual
temperature collectively explain >50% of the variation in observed
dispersion, whereas continent, the number of very large species, and
precipitation collectively explain >30% of the deviation from the
null model. Main conclusionsClimate and elevation have minimal predictive
power, suggesting that biotic factors may be more important for explaining
community body size distributions. Continent is consistently a strong
predictor of dispersion, likely due to it capturing the effects of
climate, human-induced extinctions, and more. Overall, our results are
consistent with several plausible explanations, including, but not limited
to, competitive exclusion, unequal distribution of resources,
within-community environmental heterogeneity, habitat filtering, and
ecosystem engineering. Further work focusing on other confounding
variables, at finer spatial scales, and/or within more causal frameworks
is required to better understand the driver(s) of these patterns.