Sampling biases in demographic analyses of black rat snakes (Elaphe obsoleta)

Population ecology requires reliable population samples. We assessed sampling reliability for black rat snakes (Elaphe obsoleta) using 1724 captures obtained by two different methods: trapping at communal hibernacula and opportunistic capture of snakes at large. Recapture intervals indicated that opportunistic captures were biased by size (larger snakes were over-represented) but not by sex. Furthermore, opportunistic captures of snakes seen on roads (while observers were driving) had a stronger size bias than other opportunistic captures. Trapping at hibernacula sampled the respective hibernacula populations reliably, but the hibernacula populations themselves were not representative samples of the local population. Among 13 hibernacula, sex ratios ranged from 31-65% females and age structure from 42-86% sexually mature individuals. Because rat snakes can take many years before they join communal hibernacula, young snakes were under-represented in all hibernacula samples. We found highly significant differences in the size and sex composition of our samples from the Ontario population and from samples from a population in Maryland (from published data). Those differences seem more likely to be a consequence of biases associated with how snakes were sampled at each location than a reflection of real population differences. We recommend that future sampling of rat snakes include opportunistic sampling of snakes at large combined with sampling at several hibernacula. Also, researchers sampling snakes should assess biases in their samples, because biases that are not recognized will be more problematic than those of which researchers are aware.