Although the onion fly, Delia antiqua, is described as an ovipositional specialist, laying its eggs only on Allium cepa and closely related alliums, it will distribute its eggs across a full range of resources in an artificial 5-choice oviposition test. This distribution across the range is generally conserved across flies, although in a specific sampling period a fly might show great selective variability. The distributed neural math model (DNMM) describes onion fly oviposition by showing a mathematical relationship between internal and external inhibitory and excitatory inputs (EEI*IEI)/(EII*III) in onion fly decision-making. This relationship describes the suitability of the plant to elicit oviposition, as well as the readiness of the fly to oviposit. However, eggs are laid only after a stochastic element is introduced. This stochastic element can either elevate or suppress the probability of a fly laying on a given resource. These stochastic elements could be jittery or oscillating neurons, which have been described in vertebrate systems, but to which no function has been ascribed. This combination of stochasticity and computation of acceptability of resource and the fly's internal status has generated a model highly predictive of onion fly egg laying behavior. Measuring the effects of varying EEI, EII and IEI has tested the computational aspect of this model. We hope to test the stochastic aspect of the model by administering various drugs to vary neural jitteriness.
Species 1: Diptera Anthomyiidae Delia antiqua (onion maggot)
Keywords: host selection, model testing
Back to Display Presentations, Section Cd. Behavior and Ecology
Back to Posters
Back to The 2003 ESA Annual Meeting and Exhibition