Global patterns in coronavirus diversity

Anthony, Simon J.; Johnson, Christine K.; Greig, Denise J.; Kramer, Sarah; Che, Xiaoyu; Wells, Heather; Hicks, Allison L.; Joly, Damien O.; Wolfe, Nathan D.; Daszak, Peter; Karesh, William; Lipkin, W. I.; Morse, Stephen S.; Mazet, Jonna A. K.; Goldstein, Tracey

Due to the small numbers of positive individuals obtained for rodents, non-human primates, and humans, only bat data were analyzed for factors associated with positivity. A total of 12,333 bats were tested. This included 5,624 males and 5,767 females (922 were not sexed). Among aged bats 7,385 were adults, 1,029 were sub-adults, and 8 were neonates (3,911 were not aged). Binary logistic regression models were used to evaluate significant variables, restricting analyses to species for which !50 individuals were tested. Model selection was first performed in R (version 3.2.5) (R Foundation for Statistical Computing, 2008) using the R package 'gmulti' and the step function, and the best model determined by AIC. A robust estimation of standard error was calculated by clustering by individual Animal Identification number to account for non-independence of multiple tests from the same animal using STATA 13.1, SE (College Station, TX, USA). The independent variables included in analyses were Specimen Type (blood, feces/rectal swabs, guano, oral/rectal sample, oral/nasal swabs, tissue, urine/urogenital swabs), Host Taxon (family, sub-family, genus), Host Age (adult, sub-adult, neonate), Season (wet, dry), and Interface (animal use, human activity, land use, pristine area). Season was designated as 'wet' or 'dry' according to month and proximity to the equator for each country. Interfaces were broadly grouped according to the type and intensity of animal contact with people, specifically (1) Land Use (animals sampled in areas with crops, extractive industries, livestock activities), (2) Animal Use (hunting, markets, restaurants, trade, wild animal farms, wildlife managements, zoos/sanctuaries, handling by veterinarian/researchers), (3) Human Activities (ecotourism, in and around human dwellings), and (4) Pristine (where animal human contact was not likely). The model fit for each region was evaluated using the Hosmer-Lemeshow goodness-of-fit test and pseudo R 2 values.