Assessing the impact of forage availability variation on spatial patterns of beehive migration using a hybrid modeling approach - B-Agent

The contribution of beekeeping towards achieving sustainable development is increasingly being recognized. Beekeeping is a landscape-scale process requiring an integrated management approach to address multiple sustainability challenges. Accessing quality forage resources is a key social-ecological interaction within a beekeeping system. The state of the forage landscape affects the decision-making by beekeepers to optimize forage locations for beehives placement and, in turn, influences the sustainability of beekeeping systems. To date, beekeeping system models have not taken into account the complex interconnectivities of beehive migration management. As a result, this study presents B-Agent, a spatially explicit hybrid agent-based model that combines multiple stakeholder engagement approaches to derive beekeeper decision rules and combines a machine-learning algorithm to build forage availability scenarios with an agent-based model to simulate beehive migration processes. We use the Western Australian beekeeping sector as an application for method development, testing the model using a pattern-oriented approach for its ability to reproduce patterns in I beekeeper distances traveled, (ii) beehive migration frequency, and (iii) the spatial distribution of forage site locations harvested by beekeepers within a year. Pattern analysis is carried out for the current baseline conditions. An assessment is made as to likely future change using climate change scenarios to vary forage availability. Relative to the baseline, the results show an increase in travel distance and frequency of hive migration for beekeepers for future forage availability, and indicate that the spatial distribution of forage harvest locations used by beekeepers will shift towards the Southwestern regions of Western Australia. This research provides an important novel tool for assessing migratory pressures on the beekeeping system. Through our presented spatial modeling framework an evidence-base can be harnessed for better-informed management decisions to improve the long-term sustainability of beekeeping systems, not only in Western Australia but transferable worldwide.