Increasing temporal turnover in an East African tropical forest challenges the disturbance-recovery hypothesis

Background/Question/Methods: The disturbance-recovery hypothesis posits that current structural changes in tropical forests, such as increasing stem density and biomass, are due to successional recovery from past disturbances. This hypothesis has typically been tested using taxon analyses to assess increases in abundance of shade-tolerant or late-successional species. Here, we focus on testing the prediction that community change (turnover) decelerates during succession as the community recovers from disturbance and stability increases. Further, we expect that the rate of deceleration in turnover will be greater for a recently disturbed forest relative to an undisturbed forest. To test these predictions, we compare magnitude of turnover and rate of change in turnover in stem number, species composition, and functional composition of forest that was selectively logged 46 years ago to an adjoining undisturbed old-growth forest. Our analysis is based on functional trait data related to carbon storage capacity (wood density, maximum height and maximum diameter at breast height), as well as abundance data from four censuses spaced over 24 years (1989 – 2013) for 26 plots totaling 5.2 ha in Kibale National Park, Uganda.

Results/Conclusions: Rates of change in turnover did not decelerate through time in any of the three community attributes irrespective of plot logging status. Instead, rates of change in turnover for stem number and species composition accelerated while turnover in functional composition showed no temporal changes. However, selectively logged plots showed greater turnover in stem number with slower acceleration, as well as greater turnover in species composition than unlogged plots. Our results do not support the disturbance-recovery hypothesis as an explanation of change in Kibale forest as we do not find the reduced change through time predicted for communities recovering from disturbance. Although disturbance (selective logging) seems to have affected the magnitude of turnover in stem number and species composition, it had little effect on temporal patterns of change. Since, both selectively logged forest and undisturbed forest showed increasing rates of change, we suggest that there is a forest-wide synchronous driver of change unrelated to the disturbance event of selective logging 46 years ago. While our results are consistent with reports of global-scale temporal changes in tropical forests, they raise further questions about the direction and drivers of turnover in Kibale forest.