Vegetation
The lack of significant differences in the abundance, richness, and diversity of plants between urban and rural sites (Figure 6D-F) could be due to incomplete sampling, with just 76% of potential plant species identified. Nevertheless, our species accumulation curves exhibited long tails of rare plants, suggesting that any missed species would have contributed very little to the overall percentage cover of the plant communities. Indeed, two of the top three plants by percentage cover in both rural and urban environments were mastic shrub (P. lentiscus ) and slender wild oat (A. barbata ), highlighting the prevalence of native, unmanaged vegetation within urban green spaces on Lipsi. Whilst rural areas had significantly more shrubs and less bare ground than urban sites, there was no difference in the cover of trees, which are scarce on Lipsi, negating the possibility for greater tree cover to promote butterfly species richness (Kurylo et al. 2020). The lack of a marked difference in vegetation structure between urban and rural areas could thus be a key factor in explaining the similarity in butterfly abundance and diversity across habitats. A prevalence of non-native plant species in urban areas may prevent larval development of butterflies (Dylewski et al., 2019), but exotic species were only present at three urban sites and in low abundance, limiting their potentially negative effects on butterflies. Furthermore, butterfly abundance responds negatively to non-native plants in late spring and positively by mid-summer (Kurylo et al., 2020), highlighting the importance of greater temporal resolution of sampling to characterise vegetation effects on butterflies.
Urbanisation altered vegetation community composition (Figure 7C), with cultivated species such as barley (H. vulgare ), castor bean (Ricinus communis L., 1753), and scutch grass (Cynodon dactylon Persoon, 1805) prevalent in urban environments, while natural desert saltgrass (D. spicata ) dominated in rural environments. This may have contributed to the observed differences in butterfly community composition, with cultivated patches hosting different butterfly assemblages than natural forests and scrub (Chong et al., 2014). As noted above, the dominant butterflies in urban environments (geranium bronze, scarce swallowtail, and mallow skipper) are highly associated with cultivated plants, which were prevalent in urban areas (Stefanescu et al., 2006; Tzortzakaki et al., 2019). In contrast, the dominant butterflies in rural environments (Freyer’s grayling and meadow brown) depend on meadows and grasslands, which were largely absent from urban areas (Merckx and Van Dyck, 2002; Grill and Cleary, 2003). Thus, whilst we did not detect any effect of urbanisation on the abundance or diversity of the vegetation and butterfly assemblages, the observed changes in dominance patterns and species composition between rural and urban environments could have major implications for ecosystem functioning, which should be quantified in future studies.