Estimating possible bumblebee range shifts in response to climate and land cover changes

Abstract

Wild bee decline has been reported worldwide. Some bumblebee species (Bombus spp.) have declined in Europe and North America, and their ranges have shrunk due to climate and land cover changes. In countries with limited historical and current occurrence data, it is often difficult to investigate bumblebee range shifts. Here we estimated the past/present distributions of six major bumblebee species in Japan with species distribution modeling using current occurrence data and past/present climate and land cover data. The differences identified between estimated past and present distributions indicate possible range shifts. The estimated ranges of B. diversus, B. hypocrita, B. ignitus, B. honshuensis, and B. beaticola shrank over the past 26 years, but that of B. ardens expanded. The lower altitudinal limits of the estimated ranges became higher as temperature increased. When focusing on the effects of land cover change, the estimated range of B. diversus slightly shrank due to an increase in forest area. Such increase in forest area may result from the abandonment of agricultural lands and the extension of the rotation time of planted coniferous forests and secondary forests. Managing old planted coniferous forests and secondary forests will be key to bumblebee conservation for adaptation to climate change.

Climate change winners and losers among North American bumblebees

“…We found that most species’ trajectories, regardless of whether they have increased or decreased over the past 120 years, were negatively impacted by changes in temperature (figure 2 a ). This finding is consistent with similar work on bumblebees from Japan, where temperature emerged as the major factor driving species’ range shifts [23]. Importantly, our model predicts that nine North American species ( B. appositis , B. bohemicus , B. borealis , B. neoboreus , B. occidentalis , B. perplexus , B. ternarius , B. terricola and B. vagans ), would have benefited from temperature stasis over the last century but, instead, have been negatively impacted by realized changes in temperature.…”

“…[17,21,22]). For bumblebees, in particular, Suzuki-Ohno et al [23] used species distribution models to link past changes in ranges to changes in temperature and land use. While these are important advances, species distribution models do not account for detection bias and, thus, are not ideal for analysing large-scale historical datasets where such detection bias has likely changed through time.…”

Bee occurrence data collected in citizen science program “Hanamaru‐Maruhana national census” in Japan

“…When our published data are used for scientific researches, spatial filtering, and background manipulation (e.g., Kramer-Schadt et al, 2013;Phillips et al, 2009) are required to mitigate the sampling bias (Ohno et al, 2018;Suzuki-Ohno et al, 2017;Suzuki-Ohno et al, 2020).…”

“…Using original bumble bee distribution data at high resolution, we estimated the distributions of major six bumble bee species (B. diversus, B. ardens, B. hypocrita, B. ignitus, B. honshuensis, and B. beaticola) by species distribution model, Maxent (Suzuki-Ohno et al, 2017). We also estimated possible range shifts of the six bumble bee species for past 26 years by Maxent (Suzuki- Ohno et al, 2020). In addition, we introduced our citizen science program and explained its problems and solutions in Japanese (Ohno et al, 2018).…”

“…Among wild bees, bumble bees have been frequently investigated, and some species have been reported to be declining in Europe and North America (Cameron et al, 2011;Kerr et al, 2015;Nieto et al, 2014;Soroye et al, 2020). Bumble bee decline is also related to agricultural intensification (e.g., Goulson et al, 2006), climate change (e.g., Kerr et al, 2015;Soroye et al, 2020;Suzuki-Ohno et al, 2020), pesticide (e.g., Stanley et al, 2016), invasion of alien species (e.g., Inoue et al, 2008), and pathogens (e.g., Cameron et al, 2016).…”