Aim
Assessing the threat status of declining but yet widespread species poses a challenge to applied ecologists. Previous studies using a common metric to describe the spatial aggregation of occurrences across multiple scales, the fractal dimension Dij, have suggested that species’ distributional trends may be deduced from readily understandable spatial patterns: Expanding species are expected to show more aggregated spatial distributions (higher value of Dij) than declining species (lower value of Dij). Here, we revisited these predictions using a large‐scale empirical dataset on Finnish butterflies.
Location
Finland.
Methods
For each butterfly species (n = 97) and across three spatial scales (grid squares of 10 km, 50 km and 100 km), we calculated the area of occupancy (AOOi) as the sum of occupied grid squares. We employed values of AOOi to derive the Dij for each butterfly species. We then used these metrics to compare the changes in spatial patterns of distribution (∆AOOi and ∆Dij) between two time periods, 2000–2002 and 2009–2011.
Results
Majority of the studied butterfly species showed declining areas of occupancy (at the scale of 10 km, ∆AOO10) and fractal dimensions (across the scales from 10 km to 100 km, ∆D10–100) between the two study periods. In contrast to predictions, AOO10 and D10–100 showed negative impacts on the ∆AOO10, an observation that may be explained by the high proportion of declining species in our data. Butterfly species with the greatest fractal dimensions at regional scales (D10–100) in the years 2000–2002 showed both positive long‐term distributional trends and most notable northern recent range limit shifts.
Main conclusions
Our results were in most cases congruent with the prediction of higher fractal dimension values in expanding compared to declining species. As a novel observation, many butterflies expanded northwards in spite of their occurrences getting simultaneously more scattered, particularly in southern Finland.