Composition and seasonality of flower visitors
Percentages of dipteran insects (63%) and hymenopteran insects (30%) in the flower-visiting insect communities in the present study site (24.2º N, 2950–3230 m a.s.l.) were similar to the Japanese alpine ecosystems in the cool-temperate zone, where dipteran insects occupied 64 and 61%, and hymenopteran insects (mostly bumble bees) occupied 31 and 36% of all flower visitors in the Tateyama Mts. of central Japan (36.6º N, 2400–2800 m a.s.l.) and the Taisetsu Mts. of northern Japan (43.5º N, 1700–1900 m a.s.l.), respectively (Kudo 2016). Thus, bumble bees, syrphid flies, and non-syrphid flies are all dominant pollinators of alpine plants from subtropical to temperate zones in East Asia.
As expected, seasonal trends in foraging activity were different between dipteran and hymenopteran insects. The abundance of bumble bees was low early in the season, and increased in the middle to late season, reflecting the lifecycle of colony development (Pyke et al., 2011; Amsalem et al., 2015; Mizunaga & Kudo, 2017). Thus, the availability of pollination service by bumble bees commonly increased with seasonal progress in alpine ecosystems (Kudo, 2022). In contrast, less clear seasonality in dipteran insects (including unpredictable outbreaks) is reported in the previous studies (Totland, 1994; Mizunaga & Kudo, 2017). Also in the present study, seasonal dynamics of dipteran insects highly varied between years. The abundance of syrphid flies was larger in the early season in 2017, while it increased toward the later season in 2018. The outbreaks of non-syrphid flies were observed in the early season in 2017, but they occurred in the late season in 2018. These unpredictable patterns might reflect the species-specific lifecycle of dipteran insects (Larson et al., 2001; Rotheray & Gilbert, 2011). Dipteran insects are known as important pollinators early in the season in European Alps (Bonelli et al., 2022) and high Arctic regions (Kevan, 1972; Tiusanen et al., 2016), indicating the importance of fly pollinators under harsh environmental conditions (Inouye et al., 2015).
Responses to weather conditions varied between bumble bees and dipteran insects. Foraging activity of bumble bees was positively correlated to ambient temperature as reported in previous studies (Corbet et al., 1992; Bergman et al., 1996). In contrast, the foraging activity of dipteran insects was weakly related to the ambient temperature. This unpredictable pattern in fly visitors may reflect the short lifecycle and diverse species composition in fly communities (Larson et al., 2001) rather than the temporal fluctuation of weather conditions. However, previous studies reported that the visitation frequency of dipteran insects often depended on the ambient temperature in alpine environments (McCall & Primack, 1992; Totland, 1994). For instance, a field survey of flower visitors in the alpine site of northern Japan reported that syrphid flies were most sensitive to ambient temperature, while foraging activity of non-syrphid flies was independent of temperatures (Mizunaga & Kudo, 2017), suggesting that syrphid flies are more sensitive to cool conditions than non-syrphid flies (Inouye et al., 2015). Thus, the temperature dependency of foraging activity of dipteran insects may vary among regions and between taxonomic groups. In the alpine site of this study, at least, sensitivity to weather conditions seemed to be small in dipteran insects. This is partly due to a low seasonal fluctuation of ambient temperature in the low latitudinal location. Mean monthly temperature during the major flowering season (June–September) was maintained between 10.7 and 11.5 ºC, indicating a stable thermal condition (Fig. 1). In contrast, relative humidity was negatively related to the visitation frequencies of major insect groups. The alpine zone in Taiwan was characterized by wet conditions, and wet weather may restrict flight activity of small insects (Inouye et al., 2015).