Approximating mammal functional trait diversity using body mass

Quantifying biodiversity (i.e., the numbers and types of organisms) enables us to understand the geographic distributions of species, drivers of those distributions, to identify conservation “hot spots,” and, ultimately, to assess community, population, and geographic change resulting from both anthropogenic and non-anthropogenic processes (1). Measuring biodiversity in the past, in particular, allows us to assess the roles of a variety of long-term drivers (e.g., tectonics, climate change) (2) and develop a baseline understanding of how assemblages of species respond to global and regional changes (3, 4).

There are a variety of ways to measure biodiversity, the simplest of which is richness (i.e., the number of species in an assemblage). The various metrics for Functional Diversity (FD) attempt to describe the array of lifestyles occupied by species in the same assemblage (5). They all rely on measuring or scoring ecologically relevant traits to quantify how species in the same assemblage are distributed in niche or trait space (6). FD has advantages over richness in that it can be used to identify assemblages of species or particular species with unique characteristics and thus to inform species-agnostic conservation targets (7, 8). Ideally, FD is calculated from traits that are correlated with teh major niche axes (e.g., diet, locomotion, body size) and incorporates data from all species present in an assemblage.

Palaeontologists and mammalogists have developed a variety of metrics for approximating the niche of mammal species (e.g., tooth shape, calcaneal gear ration), all of which are intended to capture some aspect of the mammalian niche (e.g., diet, locomotion) (9-13). The process of fossilization, however, means that palaeontologists are often faced with fragmentary fossils (14). Much of the taxonomically informative fossil record is comprised of isolated teeth or mandibulae and maxillae with multiple teeth; other isolated elements (e.g., calcanea) may not be reliably identifiable to species. Thus, many of the niche proxies, such as inferences of locomotory behavior derived from post-cranial elements, that are available for modern mammals are unavailable for large numbers of fossil and extinct species. Therefore, estimates of FD theat incorporate large numbers of traits are often intractable in the fossil record.

Fortunately, body mass is a fundamental mammalian trait that is correlated with many other components of the lifestyle, including diet and locomotion (15, 16). Body mass is also easily estimated from both dental and post-cranial remains (often with R2 values ≥ 0.9). The goal of the proposed study is therefore to understand whether body mass alone can be used to quantify mammalian functional diversity . If body mass capture the major axes of mammalian functional diversity, we can potentially estimate FD for mammal assemblages over their entire evolutionary history. Therefore, we will gain an enhanced understanding of how mammal FD changes in response to various biotic and abiotic perturbations of millions of years.

Methods. Using trait data in the FuTRES database (e.g., body mass, post-cranial elements to determine locomotor type, and occurrences to gather habitat data in combination with GIS-based ecotype and climate data) and species-level traits (diet) for species in the Western Hemisphere, we will compare spatial patterns of FD estimates (17) calculated using body mass alone and using suites of other traits (18). We will make the explicit assumption that space can be substituted for time and that the results apply equally to the fossil record of mammals (19). We will test whether different suites of functional traits identify the same diversity “hot spots'' and test for correlations among patterns of FD using a method such as Kendall’s Tao. Finally, if differences in the apparent spatial patterns of FD are found, we will identify which traits may be driving them. D. Fraser and colleagues (2019), showed qualitatively that FD calculated based on body mass alone identifies similar areas of high and low FD in the Western Hemisphere as estimates incorporating other traits (Fig. 1). If these findings are borne out by a larger, more thorough analysis, mammal palaeontologists may confidently estimate FD using body mass alone, aligning palaeoecology and neoecology.

Engagement. The FuTRES datastore includes data for many North American mammal species, particularly in the USA. Furthermore, the database includes individual-scale data, which will enable us to derive new species means (e.g. for body mass) that are tied directly to specimens in museum collections. The project will also benefit greatly from the expertise of participants in the FuTRES workshops.