CarniDIET 1.0: A database of terrestrial carnivorous mammal diets

Abstract

Motivation A species’ diet is central to understanding many aspects of its biology, including its behaviour, movement, and ecological niche. The diets of terrestrial carnivorous mammals, defined here as species primarily consuming other mammals (hereafter, mammal‐consumers), have been extensively studied and can vary in the proportion of different food types, and species, consumed across their geographic ranges. Accessibility to data capturing such variation in diets of mammal‐consumers across the variety of ecosystems they occur in would provide valuable information for conservation, and open research avenues for macroevolution and macroecology. However, data on mammal‐consumer diets across their geographic ranges have not been systematically collated. Here, we present CarniDIET (version 1.0), an open‐access database containing quantitative data on the diets of terrestrial mammal‐consumers collated from the literature. Main types of variable contained Diet records capturing the percentage of mammalian prey, to the highest taxonomic resolution available, and non‐mammalian food types (e.g., birds, invertebrates) in the diets of mammal‐consumers at specific sites and times. Associated data with each diet record include, where available, age and sex of mammal‐consumer, sample size, sample origin, and quantification method as well as spatial and temporal variables including dates, season, study site, altitude and coordinates. Spatial location and grain Global, terrestrial. The spatial grain varies among sites from 0.03 to 100,000 km2, with a median of 170 km2. Study centroids are provided as latitude‐longitude coordinates. Time period and grain Original diet samples were collected between 1933 and 2017, with half of studies collected between 1994 and 2008. Studies summarize diets from 1 month to 66 years, with a median of 1 year. Major taxa and level of measurement Terrestrial carnivorous mammals that primarily consume other mammals (103 species). Studies generally represent species’ population averages, although can include demographic breakdowns.

Population density estimates for terrestrial mammal species

“…For example, expressing diet as percentage of item consumption is definitely an improvement compared with coarse diet categories, but still ignores that in some species diet can vary substantially across habitats. For example, the CarniDIET database (Middleton et al., 2021) contains several locally recorded diets for >100 species of carnivores, therefore allowing assessment of how local prey availability and diversity can influence predator population densities. Furthermore, life‐history traits (e.g., litter size and sexual maturity age) and social (e.g., group size), reproductive (e.g., monogamy vs. polygamy) and territorial behaviour are unexplored factors that might also explain why some taxa tend to deviate substantially from model predictions.…”

“…For example, expressing diet as percentage of item consumption is definitely an improvement compared with coarse diet categories, but still ignores that in some species diet can vary substantially across habitats. For example, the CarniDIET database (Middleton et al, 2021) This study provides a comparative overview of population densities in terrestrial mammals that can find applications in many ecology and conservation studies. The quality of our estimates is contingent on the data available and, as more data are collected, we will be able to provide more accurate and reliable predictions of species population density.…”

Echoes of the late Pleistocene in a novel trophic cascade between cougars and feral donkeys

“…Importance of observed prey items for cougars and gray wolves were derived from CarniDIET [33]. We only included studies that reported the frequency of a prey item across predator scats in a study.…”

MOSAIC: A Unified Trait Database to Complement Structured Population Models

“…Despite major improvements in the consolidation and accessibility of trait data, there is not yet a single-source centralised database spanning behaviour, physiology, habitat, and other trait data for a wide range of species. Existing databases are often linked by taxonomy ( e.g., FishBase (Froese & Pauly, 2010), CoralTraits (Madin et al, 2016), MammalBase (Lintulaakso, 2013), AmphiBio (Oliveira, São-Pedro, Santos-Barrera, Penone, & Costa, 2017)); trait type ( e.g., TreeOfSex (The Tree of Sex Consortium 2014), TreeBase (Boettiger & Temple Lang, 2012), Xylum Functional Traits (Borghetti, Gentilesca, Colangelo, Ripullone, & Rita, 2020)); data type ( e.g., GBIF (https://www.gbif.org/), MOL (Jetz, Thomas, Joy, Hartmann, & Mooers, 2012), TetraDensity (Santini, Isaac, & Ficetola, 2018)); or a combination of the taxonomies and traits (WooDiv (Monnet et al, 2021), CarniDiet (Middleton, Svensson, Scharlemann, Faurby, & Sandom, 2021)). A number of other databases take a more general approach in their thematic scope, but are still constrained to a limited set of traits and taxonomy ( e.g., Amniote (Myhrvold et al, 2015), Pantheria (Jones et al, 2009), BIEN (Enquist, Condit, Peet, Schildhauer, & Thiers, 2016; Maitner et al, 2018), TRY (Kattge et al, 2020)).…”