Effects of three stressors on vegetation in an oligohaline marsh

Summary 1. When multiple stressors have interactive effects they can lead to important changes in ecosystem function. We examined how three stressors affected the plant community in an oligohaline marsh in southeastern Louisiana, U.S.A. These stressors included herbivory (mostly by the introduced rodent Myocastor coypus), disturbance (herbicide application) and nutrient enrichment (three levels of N–P–K fertilizer). Sampling was conducted six times over 4 years. 2. Recovery after disturbance was slow, such that after 26 months biomass in disturbed plots was 36% that of controls. Slow recovery appeared to be due to herbivory, as exclusion of herbivores for 14 months led to much more biomass compared to non-excluded plots. Exclusion did not, however, aid recovery of species richness; this recovery required 51 months in total. 3. Nutrient enrichment increased biomass by 41% and decreased species richness by c. 20% in later sampling periods. Decreased species richness was due primarily to a reduced ability of dominant species to co-exist (as determined with Hill’s diversity number N1). Nutrient enrichment did not interact with the other treatments. 4. Disturbance favoured two grasses (Echinochloa crus-galli and Leptochloa fascicularis), while lack of disturbance favoured two herbs (Sagittaria lancifolia and Polygonum punctatum) and two vines (Ipomoea sagittata and Cuscuta pentagona). Nutrient enrichment positively affected abundance of two species (C. pentagona and L. fascicularis). Herbivory did not affect species composition. 5. The effect of one stressor (experimental disturbance) on plant biomass depended on the strength of another stressor (herbivory). Nutrient enrichment was also important in affecting the plant community, but only as a single stressor. All effects changed over time, and it was clear that to understand properly the effects of multiple stressors, long-term, manipulative field experiments are necessary.