SEMICE: An unbiased and powerful monitoring protocol for small mammals in the Mediterranean Region

Schemes to monitor biodiversity change should detect properly target species without harmful effects on individuals and populations, and be powerful enough to detect expected population trends in the face of global change. Targeting is a key aspect of monitoring schemes since there is no single method able to detect unbiasedly all species of any given community, especially the rarest ones. Here we test whether SEMICE (SEguimiento de MIcromamíferos Comunes de España), a monitoring protocol for small mammal biodiversity in the Mediterranean Region, fulfil these requirements. The protocol aims at monitoring common species easy to catch with the two most widely used commercial live traps (18 Sherman and 18 Longworth traps alternated in position across 6 × 6 trapping grids spaced 15 m, brought into operation for three consecutive nights in spring and fall). We used pilot data from twenty-two plots distributed along wide environmental gradients in Catalonia (NE Spain), sampled from 2008 to 2015. The wood mouse (Apodemus sylvaticus) was dominant throughout the study period (992 individuals, 39.0%), followed by the white-toothed shrew (Crocidura russula, 598 individuals, 23.5%) and the Algerian mouse (Mus spretus, 269 individuals, 10.6%). The two most common rodent species experienced strong population declines during the eight-year period (91% for A. sylvaticus and 83% for M. spretus). Regional community data obtained from diet studies of small mammal predators showed that common keystone prey and seed dispersers were sampled properly. No differences among trap types regarding community parameters and similarity indexes, sampling efficiency, detectability, trap-induced mortality, mean size and sex-ratio were detected, confirming previous results for a smaller pilot study. The method was sensitive enough for detecting expected population changes. We recommended extending the SEMICE protocol to sample common keystone small mammals along wide Mediterranean environmental gradients, since the method was sensitive enough to detect, and even test, expected population trends associated to global change for all them.