全球变化知识资源中心

Aim Climate change is driving many species towards higher latitudes and higher elevations. However, empirical studies documenting these changes have largely focused on presence/absence based range shifts in temperate regions. Studies in lowland tropical ecosystems that control for detection probabilities are especially lacking. Location Hong Kong SAR, China. Methods By analysing a 15-year trapping dataset of geometrid moths along a lowland elevation gradient (0-600 m) in Hong Kong, we used multiple metrics and approaches to estimate occurrence shifts with elevation, changes to range limits, and community-wide shifts in elevation. Our approaches used Bayesian occupancy models to account for false absences, which may bias naive measures of range shifts over time. Results Over the study period, we detected only subtle changes in forest cover but a notable increase in maximum temperatures (similar to 0.5 degrees C per decade) and extreme weather events. Of geometrid moths, one out of the 123 examined species exhibited increased occurrence probabilities across all elevations after accounting for uncertainty in detection; all other species exhibited no significant change in occupancy with elevation. However, at least two species became newly established in Hong Kong over the decade, and 32 species showed significant elevational shifts in lower or upper range limits. At the community level, geometrid moths showed a noticeable upslope shift at all but one of the examined sites. Main conclusions The complex patterns observed highlight the difficulties and limitations in detecting climate change impacts on diverse tropical communities. Our conservative results indicate early responses of tropical species over a relatively short timespan to a decade of environmental change, and the necessity of long-term monitoring for providing insights into the management and conservation of vulnerable species.