Comparison of methods to estimate historic species richness of mammals for tests of faunal relaxation in Canadian parks

Abstract

AimSome recent tests of faunal change in reserves have relied on, but been limited to, estimates of species richness from random samples of historic range maps. We evaluated a different, Geographic Information Systems (GIS)‐based, approach to count species directly, as the latter method might facilitate rapid estimation of historic species richness as well as composition for samples of the same size, shape and exact location as present‐day reserves.LocationNational parks throughout Canada.MethodsGeographic Information System.ResultsThe GIS‐based method tended to count, in exact locations of modern parks, fewer species (on average, seven disturbance intolerant and five disturbance tolerant) present historically than extrapolated from randomly sampled sites, but the differences were not greater than expected by chance. However, correlations between number of species lost and park size were weaker than reported previously, suggesting a greater potential for other factors to influence a change in species richness (and composition) than inferred earlier.Main conclusionsDirect counts of historic range maps using GIS tools provided a quick means to estimate site‐specific historic species richness not statistically different from estimates produced by random sampling. As well, the GIS‐based method could yield data about historic species composition for the specific location and size of modern reserves, which may be more ecologically meaningful in terms of assessing what factors may have contributed to the observed species losses.