To what extent can management variables explain species assemblages? A study of carabid beetles in forests

Studies concerning the influence of forest management on invertebrate communities often focus on a limited set of chosen variables and rarely quantify the importance of management as opposed to other influences. We aimed at: 1) comparing the importance for species assemblages of habitat variables defined by management with those independent of it 2) understanding the ecological significance of the variation remaining when both management and non-management variables are used. We caught carabid beetles according to a stratified pitfall sampling based on forest structure, tree composition and stand age. Forty-nine habitat variables were measured using three spatial scales. We decomposed the variation of species assemblages with successive constrained ordinations based on sets of variables, and Studied the life traits of the species least and best explained by the model including all of the variables. Forest structure, composition and stand age showed important effects but explained a relatively small part of the overall variation in species assemblages. Management accounted for ca 30% of the variation, but non-management variables had a significant impact and the interaction between management and non-management sets resulted in significant influences. Most species for which the variation was highly explained by the model were generally large and with inefficient wings, while the least explained species were small. Our study suggests that: 1) even with highly controlled samples, the influence of management on species assemblages should not be Studied by a limited set of categorical variables; 2) management variables may interact with factors Outside of the manager's control: 3) a significant part of the variation cannot be explained by habitat variables and needs taking ecological processes into account, 4) rules to optimise constrained ordination techniques applied to species-habitat studies call be proposed