How big should a mammal be? A macroecological look at mammalian body size over space and time

Abstract

Macroecology was developed as a big picture statistical approach to the study of ecology and evolution. By focusing on broadly occurring patterns and processes operating at large spatial and temporal scales rather than on localized and/or fine-scaled details, macroecology aims to uncover general mechanisms operating at organism, population, and ecosystem levels of organization. Macroecological studies typically involve the statistical analysis of fundamental species-level traits, such as body size, area of geographical range, and average density and/or abundance. Here, we briefly review the history of macroecology and use the body size of mammals as a case study to highlight current developments in the field, including the increasing linkage with biogeography and other disciplines. Characterizing the factors underlying the spatial and temporal patterns of body size variation in mammals is a daunting task and moreover, one not readily amenable to traditional statistical analyses. Our results clearly illustrate remarkable regularities in the distribution and variation of mammalian body size across both geographical space and evolutionary time that are related to ecology and trophic dynamics and that would not be apparent without a broader perspective.

Figure 1.

The number of papers published over the past 20 years as indexed by web of science (WOS) using certain key words. The grey bars represent all papers published in the expanded Science Citation Index (from 1899 to the present), the Social Sciences Citation Index (from 1898 to the present) and the Arts and Humanities Citation Index (from 1975 to the present); open circles, papers in the disciplines of ecology and/or evolutionary biology; open squares with black centres, biogeography; filled squares, macroecology; and filled triangles, phylogeography. Papers were identified by searching for (‘ecolog*’ and/or ‘evol*’), ‘biogeograph*’, ‘macroecolog*’, and (‘biogeograph* and (phylogen* or molecul*’)) as keywords in the title and/or abstract. Note scaling of disciplines as indicated in figure legend. All fields are rising significantly faster than that of WOS as a whole (t-test between each slope and WOS yields p < 0.0001), with both macroecology and PCMs exhibiting the fastest rates of increase (slope (± standard error) of log–log plots 1.888 (0.232) and 1.595 (0.059), respectively, versus 0.234 (0.024) for WOS). While publications in ecology and evolutionary biology and biogeography are each growing rapidly (slopes of 0.866 (0.048) and 0.964 (0.05), respectively), they do not vary significantly from each other, but do differ from macroecology and the comparative phylogenetic method (p < 0.001).

Figure 2.

The limit of body size in mammals as it relates to their life history. The range of mass is shown for volant (pink), aquatic (green–blue) and terrestrial (yellow) mammals; postulated constraints on size are indicated. Redrawn after Smith & Lyons [63].

Figure 3.

The global distribution of mammalian body mass at the Late Quaternary. Patterns are shown separately for volant (dark grey bars, left-hand side of graph), aquatic (grey bars, right-hand side of graph) and terrestrial (blue) mammals. Note log₁₀ scale. Data are prior to the anthropogenic extinction of megafauna in the Americas at the terminal Pleistocene, which significantly depressed the right mode and led earlier authors to characterize the overall distribution as unimodal [71].

Figure 4.

The distribution of mammalian body size across the major continents. Modern species are indicated by the lack of fill; species extirpated during the Late Pleistocene are indicated by hatching. Note log₁₀ scale. The overall pattern is bimodal for the major continents when extinct fauna are included; Australia lacks the mammalian diversity found on other continents. See text for details of statistical analyses. Redrawn after Alexander [59].

Figure 5.

The proportion of each continental fauna composed of the major mammalian orders as a function of body mass. Note that there is very little change in the shape of the ordinal distribution before and after the extinction event. Orange, Rodentia; red, Artiodactyla; dark green, Insectivora; light blue, primates; purple, Proboscidea; black, Lagomorpha; dark blue, Carnivora; light green, Perissodactyla. Panels on the left represent pre-megafaunal extinction composition; right-hand panels are post-extinction. Note log₁₀ scale.

Figure 6.

Turnover of the maximum body mass of mammals on the different continents over the Cenozoic. Body mass is plotted on a logarithmic scale. Note that although the same general pattern is recapitulated on each major continent, the ordinal affiliation of the largest mammal during a particular sub-epoch is often different. Data for South America are lacking because of limited sampling; thus, values for this continent should be considered an underestimate. In all cases, the trajectory of maximum body size is best fit by a Gompertz function, which suggests a saturation of the largest body size niche. Redrawn from Smith et al. [54].