Can volunteers provide reliable data for forest vegetation surveys?

Biologists, land managers, and policy makers increasingly utilize volunteer-generated data to supplement their own data collection efforts and to identify habitat degradation or land management strategies. However, many professionals who could potentially benefit from volunteer data are concerned about the quality of such data. This study examined forest data from one of the largest volunteer monitoring programs in the Midwestern United States. Illinois EcoWatch Network recruits volunteers to collect statewide trend data on the habitat quality of forests by monitoring structural dynamics (changes in the dominant tree species) and the density and spread of seven invasive shrub species known to adversely impact native flora. In order to assess the accuracy of volunteer data collection efforts, botanists collected parallel data at 14 sites for comparison to volunteer sampling of tree and shrub identification, abundance, and size class placements. We found no significant difference for 12 out of 20 species identified by volunteers at these sites (P> 0.05). Volunteer accuracy rates for 12 (out of 15) tree genera were 80% or higher. However, species within the Ulmus and Quercus genera presented greater difficulty to the volunteers. We also detected large discrepancies in the shrub survey counts of highly abundant species such as Ribes missouriense Nutt. In spite of these discrepancies, we suggest the data can provide valuable information for measuring long-term changes in forest habitat quality. Improvements in data collection methods and training in combination with strategies to inform users of potential data limitations should enhance the usefulness of volunteer-collected data for most natural resource decision-making processes.