Does previous experience enhance foraging on a particular host in a polyphagous frugivore?

Prior oviposition or feeding experience with a host plant has been proposed to improve foraging efficiency for polyphagous insect herbivores. Many laboratory-based experiments have attributed certain modified behavioural responses, such as an improved ability to locate a host, as evidence for the benefits of learning. However, few studies have considered the negative ramifications of learning, particularly under field conditions rather than laboratory conditions. In the current study, using the polyphagous fruit fly Bactrocera tryoni (Froggatt) (Diptera: Tephritidae), we explore both the positive and negative consequences of modified host selection resulting from prior oviposition experience. In field cage experiments, prior oviposition experience on a given host increased the selection of that host despite its abundance and its suitability for offspring development. For example, in a field cage that contained significantly more blueberry (poor larval host) than guava (good larval host), prior experience on guava resulted in more flies selecting and accepting guava than did naïve flies (a presumed positive foraging outcome). However, the opposite also held, that prior experience on blueberry led to increased blueberry use, even when guava was more abundant. In unrestrained flies, prior experience dramatically improved recapture rate with flies locating the training fruit. The results highlight the potential benefits of learning, but also identify that learning can be a disadvantage if prior experience on a poor host leads to flies repeatedly using that host over a more suitable host.