We then used one-way ANOVA to identify significant variation among farming systems in the variables characterising land-use livestock husbandry and farming practices, applying the Holm's correction on the p-values to account for multiple testing and ensure conservative results (Legendre and Legendre, 1998). To detect where the differences occurred between groups we used Games-Howell(GH)post hoc tests, which allows for heterogeneous viriance and uneven sample sizes(Schmitzberger et al., 2005). To assess whether farming systems could be discriminated in terms of farming practices, we conducted a linear discriminant analysis(LDA).