Atmospheric nitrogen deposition has been linked with an overall loss of plant species richness and homogenisation of semi-natural habitats both in GB and elsewhere. We expect that nitrogen-induced changes in plant communities will impact invertebrate species through the loss of reproductive habitat, food plants and suitable microclimatic conditions caused by the shifts in composition of plant communities. Prior to this thesis, no quantitative research had been undertaken to assess the potential effects of nitrogen on fauna in GB. Butterflies are often used as indicator species due to their sensitivity to environmental change, our comprehensive understanding of their ecology, and the existence of long-term datasets on their abundance and distribution.
In this study, I analysed butterfly data from the UK Butterfly Monitoring Scheme alongside data on expected driver variables including nitrogen deposition, sulphur deposition, temperature, rainfall, land use intensity, and elevation. I performed a spatio-temporal analysis on the data for each species individually using generalised additive models to understand the complex and expected non-linear relationships between butterfly trends and their drivers. Model results were summarised to provide an overview of the total number of species exhibiting responses to nitrogen. In addition, results were summarised by trait groupings such as voltinism, host plant category, host plant specificity, and breeding habitat to summarise whether any trait groupings may be particularly strongly impacted by nitrogen pollution. In addition, I performed further detailed analysis on Lasiommata megera, the Wall Brown butterfly, which has been shown to be negatively impacted by nitrogen in studies undertaken elsewhere in Europe. I ran a similar spatio-temporal analysis to that mentioned above, but with the addition of two variables I hypothesised would be key drivers of L. megera: temperature in the previous September and elevation. The results for this additional analysis were presented separately.
I demonstrated that individual butterfly species vary in their relationships with nitrogen deposition and highlighted both species-level and potential trait level responses. Nine butterfly species were negatively correlated with historic nitrogen deposition, and nine were negatively correlated with percentage change in nitrogen deposition at the site over time. Two species showed significant negative relationships with both historic nitrogen deposition and percentage change in nitrogen deposition over time: Fabriciana adippe (High Brown Fritillary) and Hipparchia semele (Grayling). These findings suggest that there is a strong correlative relationship between nitrogen deposition and the abundance of many butterfly species in GB. Other key drivers of change identified in this analysis were time, rainfall, and temperature in the current and previous year. I also demonstrated a strong relationship between abundance of L. megera and historic nitrogen deposition using the model with more detailed covariates. Initial summaries based on traits were inconclusive, not highlighting any particular trait groupings as being especially susceptible to the effects of nitrogen pollution.
The results of this study present the first correlative link between nitrogen deposition and negative impacts on terrestrial fauna in GB. It reinforces the importance of continued efforts to reduce emissions to protect the natural environment. It also provides a basis for further field and lab-based work to be undertaken to better understand the causal mechanisms behind the observed relationships.
