Analyzing patterns of change in the orchid flora of Britain and Ireland with correlated environmental variability

Introduction Fifty-seven orchid species have been so far reported to occur naturally in Great Britain and Ireland, some of which are known to have declined over the last century, primarily due to land use changes. In this paper, we examine changes in distribution at the hectad scale (10 × 10 km) from 1930 to 2019. Methods We use hypothesis tests for trends in occupancy for each orchid species, asking if there is significant evidence of change that cannot be attributed to natural processes. We use 1/ f -noise as a background variability to allow for the effect of temporal autocorrelation that models the effects of natural long-term changes. In our data analysis, we also apply Frequency Scaling Local Occupancy (Frescalo) to correct for variation in recording effort. We also investigate distribution centroid shifts and changes in orchid hotspots. Results Analysing a subset of 51 species in Great Britain, 13 show significant (at the 5% level) trends in occupancy, five positive and eight negative. In Ireland, five (out of 31) show significant trends, of which four are positive. There is no prevailing directional shift of orchid distributions. When patterns are averaged across the family, orchids appear to have moved northwards in Great Britain but south-westwards in Ireland. Since recording began, the number of high diversity hotspots (SR ≥16 species) has decreased in Great Britain (from 107 to 41) and increased in Ireland (up to 6 from none), though many hotspots remain stable. Discussion Although there is not a prevailing decline, the absence of significant change in hectad occupancy overall cannot be accepted as proof that all is well with orchids in the British Isles, since examination on finer scales could reveal significant declines at the population level. There have also been redistributions of diversity on small and large spatial scales, emphasizing a major role for processes operating across a range of scales in spatial as well as temporal variability.