Abstract: Recent studies from Europe have indicated long-term declines in insect populations. However there are few published studies which have investigated insect trends in North America. In this study, I analyzed butterfly surveys for abundance and richness trends across 132 transects in Wisconsin (USA) over the past 33 years. Additionally, I investigated the role of climate change, agricultural intensification, urbanization, vegetation changes and landscape homogenization in driving these trends. To achieve these objectives, I collected and analyzed climate data, as well as land cover and normalized difference vegetation index (NDVI) data derived from Landsat satellite imagery. I examined these trends and drivers separately for three habitat types (grasslands, pine barrens and bogs) and at two scales (100m and 1km).
Analysis revealed that butterfly abundance and species richness declined across all habitats, with an estimated abundance decrease of 2.9%–4.7% per year and a richness decrease of 1.2%–2.5% per year. Strongest declines in abundance and richness occurred in grasslands. This study did not find evidence that variables related to climate change, agriculture intensification or urbanization were associated with this decline. Change in greenness (NDVI mean trend) was positively associated with butterfly abundance and richness trends in pine barrens and grasslands, indicating that transects with higher rates of greening have lower rates of butterfly decline. On the other hand, average greenness (mean NDVI mean) was negatively associated with abundance trends in pine barrens and grasslands (at the 1km scale), implying that greener transects had a greater rate of butterfly decline. Finally, change in landscape heterogeneity (NDVI sd trend) was positively associated with butterfly abundance and richness trends in pine barrens, suggesting that transects with greater homogenization have greater butterfly decline.
Our findings are consistent with previous studies from Europe and North America which have observed declines in butterfly abundance and richness. Evidence from this analysis indicated that these butterfly declines could not be entirely explained by variables related to climate change, agricultural intensification, urbanization, vegetation changes or landscape homogenization. Further investigation into the drivers of these declines is proposed. Finally, these findings add to literature which asserts that NDVI may be a useful metric in predicting butterfly community trends
|