I just ran a Mantel test between Bray-Curtis matrices of plants and bees at a site. The Mantel says that the matrices are "almost" significantly correlated (P=0.06) but a traditional correlation say that they are not (P=0.95 R-sq 0.0). Waht is going on???? all the other alayses make perfect scense - but this one set of relationships is completely screwy. Any insights?
The attached file has a plot attached with Mantel P values for each analysis
John.