Temperature is a critical environmental factor for ectotherms and the ecosystem services provided by them such as insect pollination. With extreme climatic events becoming more rampant under climate change, it is important to understand how the ecosystem services can be maintained under highly variable temperatures. Many studies suggest that communities with high species diversity can provide reliable services under the variable environment. However, these studies did not consider the multitude of the temporal scales or hierarchical levels in which temperature and the services could vary, which is necessary for the predictions.
The temporal scale matters because climate change is known to alter temperature variability in different ways at different scales. Therefore, the diversity–function relationship must be examined by integrating the differences across scales, which could provide insights for effective future management. In this study, we investigate whether diversity can stabilize and elevate the pollination service in two different temporal scales of temperature variations (within day and between days) and identify the scales in which diversity is more important.
We studied a crop pollination system of common buckwheat in the town of Iijima at the central Japan. We observed buckwheat pollinators under different weather conditions for five years and estimated the responses against temperature, diurnal activity, and phenology for nine different pollinator groups using generalized additive models with mixed effects. We also estimated pollination efficiencies for different pollinator groups. We then constructed a model to simulate the amount of pollination service given the real temperature variations provided by three hypothetical but realistic communities with the same level of abundance but different levels of diversity. The pollination service was evaluated based on the mean level and the stability.
Our results suggest that diversity is more important for increasing and stabilizing the service on a within-day scale on days with low temperatures than on a between-day scale and within-day scale on warmer days. The implications of these findings for the management and climate change impact will be discussed.