Addressing species turnover and community changes in vegetation resurvey studies

Questions How are species turnover and community change assessed by a plot-specific approach based on quasi-permanent plots? How can community changes be tracked by a community-specific approach based on community surveys? Do the results of a community-specific approach depend on the grain of the community classification? Locations Atlantic wet heaths, southwest France. Methods We studied vegetation change at one site (Cousseau National Nature Reserve, NNR) and within a network of sites on a regional scale using two resurvey approaches: a plot-specific and a community-specific approach based on sets of old (1968–1998) and new (2006–2012) plots. Two correspondence analyses (CAs) were performed and plots were classified into community types (CTs) using hierarchical clustering. Trajectories of quasi-permanent plots were described along identified gradients. Community change within each CT was detected by testing for differences in mean coordinates along CA axes between old and new plots. Several grains of clustering were tested for the community-specific approach at the regional scale. Results At the Cousseau NNR site, the plot-specific approach identified the trajectory of each plot along two major ecological gradients, while the community-specific approach detected a community change for only one out of the three defined CTs. For the regional network of sites, the community-specific approach found several community changes; their strength and direction differed slightly when the CTs were defined more accurately. Conclusions Our study suggests that the resurvey strategy influences our ability to detect the degree and characteristics of vegetation changes in time. The plot-specific approach allows detection of species turnover and real trajectories but may over-emphasize community changes. The community-specific approach offers another way to interpret community changes that could be relevant for habitat conservation issues, especially at a regional level. For the community-specific approach, the detection and interpretation of vegetation change is influenced by the grain of community classification.