Assessing researcher impacts from a long-term sampling program of wetland communities in the Everglades National Park, Florida, USA

1. Long-term monitoring requires repeated visits to a study site, greatly increasing the potential for cumulative visitation effects. For ecological studies in general, and for monitoring in particular, data must be evaluated for confounding artefacts from researcher presence. We compared aquatic communities at long-term sampling plots (nine sites, each with three plots, studied continuously from 6 to 22 years) in the Everglades National Park to previously unsampled reference plots adjacent to them to assess the effects of researcher visitation on the flora and fauna. 2. We identified two criteria that are sensitive to local habitat heterogeneity for assessment of visitation impacts. First, the long-term plots must differ from adjacent reference plots by a magnitude that exceeded variation among plots separated by equal or greater distance (i.e. the difference is greater than expected by scaling of community change proportional with distance); and second, multiple reference plots must consistently differ in direction (e.g. greater abundance or less abundance) from adjacent long-term plots. We also tested for increased heterogeneity among samples from long-term plots compared with those not previously visited. 3. We found no evidence of researcher effects on fish or macroinvertebrates, and only weak evidence for alteration of emergent plants and periphyton floating mats. Our failure to document visitor impacts may result from either low visitation rate or the dynamic nature of the wetlands studied.