Vegetation response to disturbance in a coastal marsh in Texas

Disturbance is considered one of the main factors influencing plant species composition and diversity. We conducted a field study to address the plant community response in a coastal marsh to a major disturbance. In 1992, muskrats (Ondatra zibethicus L.) completely removed vegetation within a 450 ha area of intermediate coastal marsh in Texas, USA. We used vegetation data collected prior to the disturbance (1989-1991) as a baseline for comparison to that recorded annually for a decade (1992-2002) following the disturbance. We compared species diversity, richness, relative abundance, evenness, and species similarity between pre and postdisturbance periods to evaluate the temporal response of the disturbed plant community. Plant diversity in the study area returned to predisturbance levels after 10 years. Species diversity in the study area had two peaks following the main disturbance. These peaks are associated with fluctuations of the water levels in the area. Our results suggest that it is possible to control the sedge or grass dominance in a community by subjecting the area to a carefully timed willful disturbance (e.g., grazing or fluctuating water levels) to achieve management goals. However, vegetation composition in the area changed from a grass-dominated (predisturbance) to sedge-dominated (postdisturbance) community. At the conclusion of sampling in 2002, plant species abundance, evenness, dominance, and richness conditions reached levels similar to predisturbance. However, the species composition after a decade postdisturbance differed from that during the predisturbance period. Therefore, even though we are able to predict the return of species diversity, evenness, and richness of a community after a period following disturbance, the actual species composition of an intermediate marsh following recovery is difficult to determine accurately, as it is contingent on several biotic and abiotic conditions that prevail while the system recovers from disturbance. Nomenclature follows: Stutzenbaker (1999) for plants and Davis and Schmidly (1994) for animals. termining vegetation response to disturbance. For example, natural cyclical patterns of vegetation change occur in the coastal wetlands of the Carolina Bays in response to rainfall cycles (Kirkman 1992) or during extended periods of drought, entire extant vegetation stands may be removed due to a fire event. Building of levees or other water-control structures affect hydrology or tidal influence, which disrupts the function of such marshes. During dry conditions, other forms of disturbance in wetlands include mechanical discing and tilling or disturbance caused by animals such as extensive rooting by hogs (Kirkman and Sharitz 1994). An important source of disturbance in coastal marsh is damage following a cyclonic or hurricane event, resulting in excessive salt influx and physical destruction of extant plant communities. In most wetland systems, there is a lack of long-term monitoring of plant communities following a disturbance, which screens much of our understanding of community recovery after disturbance. Successful management of coastal marsh requires an understanding of plant community response to disturbance. Evaluating trends in vegetation recovery following a disturbance is of importance for insight into the fundamental processes of vegetation regeneration and development of conceptual models to predict response of vegetation to such disturbances (Keddy et al. 1989). For example, plant community response to recent (2005) hurricane events in coastal marshes of Louisiana and Texas could be predicted based on documented responses to major disturbance events. However, in most instances, longterm vegetation monitoring data are not available, making recovery predictions difficult or impossible. A major source of disturbance in coastal and freshwater wetlands is extensive grazing by muskrat (Ondatra zibethicus L.). Vegetation response to complete removal by muskrats in wetlands has received little attention compared to fire or grazing by terrestrial mammals mainly because these are common techniques in marsh management (e.g., fire as a tool to enhance pastures for the cattle industry, land management through fire or grazing removal to restore or enhance habitats for game species). However, muskrats can severely affect vegetation structure and community composition (Akkermann 1975). At low densities, they consume only minor parts of the vegetation, mostly those that are rich in carbohydrates and proteins (Danell 1977), but in higher densities muskrats can cause massive damage to vegetation structure and community composition (Weller and Fredrickson 1974). Furthermore, muskrats can cause additional disturbance in a vegetation community through an increase in soil aeration and water infiltration (Meadows and Meadows 1991, Butler 1995, Johnston 1995). Nyman et al. (1993) reported a positive correlation between species richness and muskrat activity in a coastal marsh in Louisiana. Similar results were reported by Connors et al. (2000) in a tidal marsh in the Hudson River. However, muskrats also use a considerable amount of vegetation material, mostly emergent, to make lodges for overwintering. In areas where muskrats occur at high densities, entire extant vegetation stands may be removed (Danell 1977). To examine vegetation response to a complete removal by muskrats, we evaluated natural revegetation patterns in an intermediate coastal marsh in Texas. We expected plant diversity to increase following the disturbance, based on the disturbance-diversity hypothesis (Grime 1973a, b; Connell 1978). However, as communities exhibit some degree of resilience, we also expected the diversity to increase following disturbance and then decrease over time as the communities returned to predisturbance structure. Our primary objectives were to: (1) determine the effects of intensive disturbance on vegetation diversity of an intermediate marsh; (2) evaluate vegetation community response following disturbance and changes in species composition over time; and (3) estimate the timeframe for marsh vegetation recovery to predisturbance conditions following a major disturbance. The exclusivity of our study is the collection, monitoring, and use of long-term data, which allowed close evaluation and reporting of vegetation recovery patterns in a coastal marsh following a major disturbance. Thus, results of our study provide land managers with a tool for estimating time and pattern of vegetation recovery a priori, aiding future management decisions in intermediate coastal marsh.