Mammalian body mass strongly correlates with life history and population properties at the scale of mouse to elephant. Large body size is thus often associated with elevated extinction risk. I examined the North American fossil record (28–1 million years ago) of 276 terrestrial genera to uncover the relationship between body size and extinction probability above the species level. Phylogenetic comparative analysis revealed no correlation between sampling-adjusted durations and body masses ranging 7 orders of magnitude, an observation that was corroborated by survival analysis. Most of the ecological and temporal groups within the data set showed the same lack of relationship. Size-biased generic extinctions do not constitute a general feature of the Holarctic mammalian faunas in the Neogene. Rather, accelerated loss of large mammals occurred during intervals that experienced combinations of regional aridification and increased biomic heterogeneity within continents. The latter phenomenon is consistent with the macroecological prediction that large geographic ranges are critical to the survival of large mammals in evolutionary time. The frequent lack of size selectivity in generic extinctions can be reconciled with size-biased species loss if extinctions of large and small mammals at the species level are often driven by ecological perturbations of different spatial and temporal scales, while those at the genus level are more synchronized in time as a result of fundamental, multiscale environmental shifts.
Fossil Occurrence Data
Fossil occurrence data compiled from the MIOMAP and FAUNMAP databases (http://www.ucmp.berkeley.edu/neomap/). Each locality is identified by the combination of database name (“MM” for MIOMAP, “FM” for FAUNMAP), Machine Number, and Analysis Unit. Included in the data table are: Max_age, maximum age of locality in million years; Min_age, minimum age of locality in million years; Genus_accepted, standardized genus designation; Species_accepted, standardized species designation; Genus_NEOMAP, genus designation in MIOMAP/FAUNMAP; Species_NEOMAP, species designation in MIOMAP/FAUNMAP; ID_conf_NEOMAP, level of confidence in the original species identification.
Dryad_Tomiya Fossil Occurrence Data.csv
Composite cladogram for 276 North American fossil genera. See article and Online Appendix B for complete sources of information.
Body Mass Data
Skeletal measurements and estimated body masses of 901 taxa. Refer to Online Appendix B for additional information. Included in the data table are: LnMass, natural-log transformed body mass (originally in grams); m1L, anteroposterior length of the lower first molar; m1W, transverse width of the lower first molar; otherL, anteroposterior length of a skeletal element other than m1 (reported under Remarks); otherW, transverse width of a skeletal element other than m1 (reported under Remarks). See Body Mass References for a list of literature sources of measurements. Measurement data in the Paleobiology Database were obtained from: http://paleodb.org/cgi-bin/bridge.pl.
Dryad_Tomiya Body Mass Data.csv
Body Mass References
Literature sources of skeletal measurements reported in Body Mass Data.
Dryad_Tomiya Body Mass References.pdf
Generic data for phylogenetic comparative analyses. Included in the data table are: Group1/2, taxonomic affiliation above the genus level; Occurrences, number of genus-level occurrences in North America; T_FAD, observed first appearance date; T_LAD, observed last appearance date; T_FAD50, sampling-adjusted first appearance date; T_LAD, sampling-adjusted last appearance date; LnMass, natural-log transformed estimated body mass (originally in grams); LnIBIL, natural-log transformed expected inter-birth interval length (originally in days).
Dryad_Tomiya Generic Data.csv