Modelling information transfer in honey bees.

CORTI BIELSA, GONZALO; SEGURA, E; FARINA WM

Buenos Aires

Congreso; 42º Congreso Internacional de Apicultura Apimondia 2011; 2011

Apimondia

Modeling information transfer in honey bees Corti Bielsa, G.D.1; Segura, E.C.2; Farina, W.M.1   1 Grupo de Estudio de Insectos Sociales, Departamento de Biodiversidad y Biología Experimental, IFIBYNE-CONICET, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires. moinonplus@gmail.com 2 Grupo de Investigación en Redes Neuronales, Departamento de Computación, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires   Olfactory information provided by floral sources shapes individual and collective foraging behaviour in honey bees. However, how the population structure of this eusocial insect as well as its temporal dynamics (the formation of groups specialized in different tasks according to age) affect the propagation of this relevant resource cue is unknown. We formalized a computational model to address these issues. As empirical data suggest that this information is stored in hive bees as associative memories (the floral scent associated to sugar reward), we used the Rescorla-Wagner learning rule in an agent-based model. We simulated floral nectar distribution on a hierarchically structured population via mouth-to-mouth food transfers (trophallaxis). In this virtual structure, the agents play different roles in specific areas according to their age. With learning and population parameters (such as switching rates between tasks) that fit empirical values, we evaluated associative memories after a pulse of an odour-rewarded input. We found that odour-rewarded memories established at early ages retard largely the forgetting rate of this dynamic network in older individuals (i.e., foragers and nectar processors). While related models usually restricted the analysis to foraging agents at a short term, ours focuses on longer periods including agents from other sub-castes/tasks. We consider that the inclusion of a broader social horizon developing at longer timescales would allow understanding relevant questions such as how resource information affects individual decision and how it biases collective foraging behaviour.