Residual effects of land management on wetland communities in an agricultural landscape

Background/Question/Methods Wetlands comprise a substantial portion of agricultural landscapes in central and south Florida.  Understanding the effects of agricultural management on these wetlands is a crucial step to sustaining biological diversity and ecosystem function while maintaining agricultural productivity.  We examined effects of pasture management on 40 wetland communities embedded in two pasture types (intensively managed and semi-natural) at the MacArthur Agro-Ecology Research Center, Buck Island Ranch (MAERC), Florida.  Intensively managed pastures are dominated by exotic bahia grass (Paspalum notatum), are fertilized with N and lime, and were fertilized with P from 1960s - 1986.  Semi-natural pastures dominated by native grasses have never been fertilized.  We evaluated differences in wetland species richness (S), diversity (Jost's D), composition, and heterogeneity by pasture type, and compared pasture-type effects between taxonomic groups (plants, insects, vertebrates).  We hypothesized that wetlands in intensively managed pastures would have fewer species, higher abundance, and increased homogeneity between communities. We also hypothesized pasture-type responses would be comparable between insects and plants because habitat structure can constrain insect assemblages.  Conversely, we hypothesized vertebrate patterns would be independent of plant and insect assemblages because vertebrates (e.g. amphibians) are more constrained by hydroperiod and access to terrestrial vegetation than structure within wetlands.

Results/Conclusions Pasture type significantly affected overall wetland composition (MRPP: A=0.04, p<0.01), but taxonomic groups responded differently to pasture management. Plant diversity (D) was higher in semi-natural pastures than in intensively managed pastures (F=13.6, p<0.001), and composition was significantly different (perMANOVA: F=9.3, p=0.02). Insect diversity (D) did not differ by pasture type, but composition was significantly different between pasture types (MRPP: A=0.007, p=0.01). Vertebrate assemblages were not different between pasture types. Overall, wetlands were more heterogeneous in semi-natural than in intensively managed pastures (ANOVA on pair-wise Sorenson distance by pasture type, F=21.5, p<<0.001).  Mantel tests on pair-wise Sorenson distances between taxonomic groups revealed composition changes between wetlands were significantly correlated between plant and insect assemblages in intensively managed pasture (r=0.2, p=0.04), but not in semi-natural pasture (r=0.12, p=0.11). These results reveal that intense management of the landscape surrounding wetlands can alter species composition and increase homogeneity between wetlands embedded within them. Further, effects of previous fertilizer application (P addition) can remain more than 20 years after fertilizing ceases.