Butterfly sampling
The sampling technique implemented was the butterfly census method, which is widely used by the UK Butterfly Monitoring Scheme (UKBMS). Developed in 1973 by Ernest Pollard (Sevilleja et al., 2019), this method uses W- or M-shaped transects to cover heterogeneity within the sampling area. The key variables to standardise with this method are the transect length, walking speed, time of day, and weather (Wheater et al., 2011). The implemented method was adapted from Zografou et al. (2009). The four corners of each site were located using QGIS to create a square shaped plot for one Pollard transect, ranging from 40 – 70 m. The average site size for both urban and rural plots was 583.2 ± 55.1 m2 (mean ± standard error). Butterflies observed 5 m in front and on either side of the transect were recorded and identified. According to Wheater et al. (2011), butterfly surveys should be performed between 10:00 and 16:00, but preliminary surveys indicated that butterflies on Lipsi were very sensitive to changes in temperature during these hours and the highest abundance of butterflies was found before 10:00. Therefore, butterfly surveys were undertaken between 07:00 and 10:00 with temperatures < 27 °C and wind conditions < 25 km h-1. The “Butterflies of Britain and Europe: A photographic guide ” was used for species identification (Haahtela et al., 2019). One transect was conducted at each site in both May and June for a total sample size of n = 36.