In this study we aimed to investigate trap efficiency and specificity by three widely used live trapping methods (Sherman, mesh, and pitfall traps) in an agriculture landscape of NE Spain. 

We trapped 243 small mammals of 8 different species. Sherman traps yielded high number of species (6) than mesh (5) and pitfall (3) traps. Frequencies of occurrence of small mammal species differed depending on the sampling methods used, as revealed by a statistical log-linear model for multidimensional contingency tables. Apodemus sylvaticus, Mus spretus and Eliomys quercinus were over-sampled by mesh traps, whereas Crocidura russula was under-sampled by mesh traps. Crocidura russula, Apodemus sylvaticus, Mus domesticus, and Mus spretus, were over-sampled by Sherman traps, whereas Suncus etruscus was under-sampled by Sherman traps. Finally, Suncus etruscus and Microtus duodecimcostatus were over-sampled by pitfall traps. The composition of the small mammal community studied was rather similar when using Sherman and mesh traps, but differed strongly from the community sampled by pitfall traps. So, in order to have a good reflex of the composition of the small mammal community a combination of  trapping techniques is necessary, as pointed out by many authors.