The community matrix was square root transformed to stabilize
variability followed by Wisconsin double standardization (first,
species divided by their maxima; second, samples divided by
sample totals). From the standardized community matrix a
BrayeCurtis dissimilarity matrix was calculated and subjected to
nonmetric multidimensional scaling. Nonmetric multidimensional
scaling (NMDS) is commonly considered as the most robust unconstrained
ordination method in community ecology (Legendre
and Legendre, 1998). NMDS projects observed community dissimilarities
onto 2-dimensional ordination space and it can handle
nonlinear species responses. NMDS is an iterative process, which
gets easily trapped into local optima instead of finding a global
solution. Therefore, we used a procedure with several random