Bombus bees species distribution across the UK

Introduction

Bumble bees and other pollinators face mounting threats due to human-induced disturbances, including climate change, habitat loss and fragmentation, pollution, and reductions in the abundance of essential plant species1 . Pollinators provide critical ecosystem services by supporting plant communities and contributing to agricultural productivity, directly impacting food security and biodiversity (Murray, Kuhlmann, and Potts, 2009). Among the primary global drivers of insect and pollinator declines, rising temperatures and climate change have emerged as significant threats to bumblebees (White and Dillon, 2023). For bumblebees, exposure to extreme temperatures—even those below their critical thermal tolerance—has been shown to increase worker mortality, disrupt brood development, and alter behaviour (White & Dillon, 2023). These impacts extend to cognitive functions such as colour-sucrose association and scent recognition, which are essential for efficient foraging (Goulson & Nicholls, 2016). While the immediate effects of heat stress on individual workers are concerning, these impacts reverberate through the colony, affecting its overall health and reproductive success. On a larger scale, climate-induced temperature stress has contributed to disrupted bumblebee distributions and colony declines across continents, underscoring the urgency of conservation efforts (Kerr et al., 2015).

To better understand and mitigate these pressures, species distribution models (SDMs) offer valuable insights. SDMs are powerful tools for predicting species’ geographic distributions, both in current and future climates, which helps in identifying areas of potential habitat loss or resilience (Miller, 2010). By mapping suitable habitats under various climate scenarios, SDMs can guide conservation actions, prioritize areas for protection, and highlight potential refuges that might support bumblebee populations despite climate change.

In this study, we aim to assess the distribution of bumblebee species across predicted habitat suitability maps generated with the Maxent model, a widely used tool for species distribution modelling that utilizes presence-only data to estimate potential habitats. Using occurrence data collected over a five-year period, we compare habitat suitability between 2018 and 2023, focusing on changes in environmental factors that may affect bumblebee populations. This approach allows us to analyse shifts in suitable habitats and identify areas of potential loss or stability within the species’ range. By examining these changes, we hope to gain insights into how recent environmental shifts, driven by climate change and other anthropogenic factors, may impact bumblebee populations, informing future conservation efforts and habitat management strategies.

Methods

Environmental raster layers for this analysis were derived from the TerraClimate dataset, which provides monthly climatic variables for terrestrial surfaces. Elevation data were sourced from the ALOS World 3D global digital surface model (DSM) dataset. The following predictors were included in the model: elevation (elev), slope, hillshade, Palmer Drought Severity Index (PDSI), precipitation accumulation (pr), soil moisture (soil), minimum temperature (tmmn), maximum temperature (tmmx), and wind speed (vs).

Species distribution models were developed using the Maxent (Maximum Entropy) model, a widely used approach in species distribution modelling, particularly effective for presence-only data. Occurrence data for Bombus species were obtained from the Global Biodiversity Information Facility (GBIF).

Together, these environmental variables enable the assessment of habitat suitability for Bombus species and facilitate the comparison of environmental changes between 2018 and 2023. This approach allows us to model how habitat suitability for Bombus species may shift under changing climate conditions.

Results and Discussion

It is important to note that the Maxent species distribution model does not estimate the probability of finding a species in a specific environment but instead predicts the likelihood that the habitat is suitable for the species. The model predicted a total suitable habitat area of 136,080.2 square kilometres in 2018, compared to 127,910 square kilometres of unsuitable habitat. By 2023, the predicted suitable habitat increased to 148,258 square kilometres, while unsuitable habitat decreased to 115,733 square kilometres. This shift highlights the projected changes in habitat suitability over the study period.

The predicted habitat suitability map of the UK is shown below. Areas with higher suitability are represented in green, indicating a higher probability of favourable habitat conditions, while areas in pink denote lower suitability, where habitat conditions are less likely to support the species.

The model revealed shifts in the environmental variables influencing habitat suitability between 2018 and 2023. In 2018, maximum temperature was the primary factor driving habitat suitability. By 2023, however, soil moisture emerged as the most significant factor, followed by maximum temperature, drought index, and precipitation accumulation. This change highlights a potential shift in the ecological drivers of habitat suitability under evolving climate conditions.

The change in habitat suitability between 2018 and 2023 was found to be statistically significant (Chi-square test, p < 0.001), suggesting that the habitat conditions for Bombus species have indeed shifted within this timeframe. While this change may initially appear linked to fluctuations in environmental variables—such as temperature, drought severity, soil moisture, and precipitation—further detailed investigations are required to confirm the causal relationships involved. The significance of these environmental shifts is particularly relevant in light of climate change, which continues to impact habitat suitability for many species globally, often leading to altered distributions and, in some cases, population declines.

Our findings indicate that habitat suitability for Bombus species has not only changed but has increased between 2018 and 2023. However, this increase does not conclusively imply improved long-term habitat stability for these species. Understanding the specific environmental pressures and their seasonal or annual variability is critical for accurately forecasting future suitability trends. With further research, it would be possible to pinpoint which changes are directly influencing habitat quality, distinguishing short-term fluctuations from long-term trends.

Despite focusing on a relatively short timescale, this study underscores the powerful utility of species distribution models, like Maxent, in revealing critical drivers of species’ habitat shifts. Such modelling approaches provide a window into the underlying dynamics of population changes, revealing how shifts in environmental factors influence habitat suitability over time. They also identify at-risk areas, enabling us to track potential habitats that may become unsustainable for Bombus species without targeted conservation interventions.

By highlighting the drivers of habitat suitability change and pinpointing specific areas affected by these shifts, our findings could significantly aid conservation planning. For instance, targeted conservation efforts in areas where habitat suitability has declined could help mitigate future risks of species loss. These models, when coupled with robust field data and longer time series, can help refine conservation strategies and resource allocation, ultimately contributing to the preservation of vulnerable species. Expanding on these findings and incorporating more extensive environmental data will enable a deeper understanding of the complex interplay between species distributions and the changing climate.

References:

Goulson, D. & Nicholls, E. The Canary in the Coalmine; Bee Declines as an Indicator of Environmental Health. Sci. Prog. 99, 312–326 (2016).

Kerr, J. T. et al. Climate change impacts on bumblebees converge across continents. Science 349, 177–180 (2015).

Miller, J. Species Distribution Modeling. Geogr. Compass 4, 490–509 (2010).

Murray, T. E., Kuhlmann, M. & Potts, S. G. Conservation ecology of bees: populations, species and communities. Apidologie 40, 211–236 (2009).

Potts, S. G. et al. Global pollinator declines: trends, impacts and drivers. Trends Ecol. Evol. 25, 345–353 (2010).

White, S. A. & Dillon, M. E. Climate warming and bumble bee declines: the need to consider sub-lethal heat, carry-over effects, and colony compensation. Front. Physiol. 14, 1251235 (2023).