not to mention that there is a huge bias in bibliography by design, because any predation study of type 'we didn't found any evidence of wild animals preyed by

not to mention that there is a huge bias in bibliography by design, because any predation study of type 'we didn't found any evidence of wild animals preyed by cats in our place' wouldn't be published. If you base your search on bibliography you are pruning all the negative results. This is like removing all negative numbers before calculating an average value. We need to remember also that "Local cats killed minus 1000 birds", would be a perfectly acceptable result [1] that is excluded from the search and the article. [1] (= Cats save 1000 birds by killing egg-eating rats)