HoPoMo: A model of honeybee intracolonial population dynamics and resource management

The population dynamics of eusocial insects differ significantly from those of non-eusocial animals. With eusocial insects, one has to distinguish between a population of colonies and the population of individuals inside each colony. These two levels are closely related, because the decision whether a colony reproduces or not is mainly determined by its intracolonial population and resource status. In addition, a population's rate of colony mortality is strongly dependent on each colony's intracolonial status. Honeybees collect their food in the environment, their most important ecological aspect is the pollinating of plants. In recent empirical studies, we demonstrated the importance of food supply on the population dynamics of honeybee colonies. In addition, honeybees show division of labor, which occurs in the form of age polyethism, meaning that morphologically almost identical worker bees choose their tasks mainly dependent on age. This task selection system is very flexible and is significantly affected by changes in task-specific workloads or in the age structure of the colony. Recent studies have shown that environmental factors and/or natural or experimental changes in resource supply heavily affect the age structure, resulting in changes in task allocation within the colony. These changes further affect the intracolony population dynamics and age structure. This situation can be described as a set of delayed feedback loops. We created a mathematically simple honeybee population model (HoPoMo) using difference equations to model the population dynamics and the resource dynamics of a honeybee colony. Our model emphasizes the importance of pollen supply and of brood cannibalism, an aspect that was neglected by other honeybee population models so far. HoPoMo includes simple models of task selection and of nutrient allocation. It allows us to simulate a variety of colony conditions (colony size, intrinsic bee characteristics, resource status) and a variety of environmental conditions (rain, temperature, botanical resource availability). The model can be easily parameterized with real weather data and with experimental colony treatments, so that we can use the model to interpret empirical experimental data. We successfully compared the predictions of our model with data gained from the literature and from own experiments. Extensive sensitivity analysis revealed that the model's predictions of population dynamics are very stable, except at very low mortality rates. And finally, we demonstrate, that the model can be used for several environmental conditions of honeybee living: controlled bee hives, scientifically used colonies and wild-life honeybees.