When plant communities won't stand still: Rapid change detection and causality on a barrier island using the National Vegetation Classification

Background/Question/Methods A beach restoration project was planned by the National Park Service (NPS) and partners for part of Assateague Island, Maryland, USA, in which anthropogenic shoreline alterations had created a sand-starved, eroding section of low elevations, frequent storm disturbance, and sparsely vegetated upper beach habitat. While the unnaturally frequent disturbance was a cause of concern for island managers, it also appeared to provide sparsely vegetated habitat for rare species that require early successional barrier island habitat, including the federally listed piping plover (Charadrius melodus). A project compliance requirement was that the NPS monitor the abundance of sparsely vegetated habitat and consider corrective action if this declined more than 25% after restoration. With little time and funding to develop conventional metrics to define rare species habitat, we employed the newly developed National Vegetation Classification (NVC) treatment for the area. A well-described sparsely vegetation type was employed as the major suitable surrogate metric for early successional habitat. Each year from 1996 through 2007, NVC types were keyed at each of approximately 400 observation plots throughout the study area.

Results/Conclusions Pre-treatment (before restoration) monitoring data (1996 – 2002) showed that the point estimate for the mean frequency (cover) of sparse vegetation stands in the study area was 48.4% of all observations (95% CI: 46.6% – 50.3%) for these years. Post-treatment data (2003-2007) showed that the point estimate for this frequency declined to 32.9%  (95% CI: 30.9% – 34.9%). The monitoring showed a highly significant probability that the threshold of 25% reduction in target vegetation (habitat) abundance was exceeded and triggered the management requirement to assess and consider action. The monitoring was not designed to determine cause of reduction. It is suspected that less storm activity during the post-restoration period may have been partly or mostly cause for the change. A mild declining trend in plover breeding success began shortly after the restoration project. The decline in sparse vegetation abundance and in plover breeding plover success triggered mitigation responses by local managers.

The monitoring demonstrated that a well-defined vegetation classification scheme, thoughtfully implemented and with reasonable understanding of limitations, could be used as a convenient monitoring metric, in the absence of funding to develop and implement more challenging methods. It proved reliable enough that a consortium of land management partners accepted it as both rigorous and germane enough to guide the management of a federally listed species, including mitigation actions.