Do mothers really know best? Complexities in testing the preference-performance hypothesis in polyphagous frugivorous...

The preference-performance hypothesis (PPH) has widely been used to explain host exploitation patterns by phytophagous insects. However, this hypothesis often fails in the case of polyphagous species when compared with specialists. One explanation, validated by the information-processing hypothesis (IPH), considers that polyphagous insects are unable to process a large array of cues, which hinders females from distinguishing between high- and low- quality hosts. Here we analyzed Anastrepha ludens female host preference and offspring performance, and tested if neuronal limitations could possibly play a role in the incapacity of the polyphagous A. ludens to make 'accurate decisions' and therefore partially explain mismatches related to PPH. Results testing the PPH by correlating female preference to six naturally occurring hosts and its offspring outcomes show that A. ludens females oviposited greater proportions of eggs on fruit according to hierarchical preferences. Infestation level was low in white sapote, the preferential and seemingly putative ancestral host, likely due to sapote defence mechanisms. Pupal weight and adult size were lower when A. ludens larvae developed in guava (conditional host that was artificially infested) and peach, a lower ranked host compared with 'Marsh' grapefruit, white sapote, and 'Manila' mango (three preferred hosts). Larvae reared in 'Manzano' pepper, a low-ranked host, performed better than in peach and guava. Results testing the IPH, show that polyphagous A. ludens females were less accurate when discerning between a non natural host (guava) when compared with a preferred, natural host (grapefruit): error rate was significantly higher, number of oviposited fruit in a 6-h period was extremely low, time searching and ovipositing took longer, and pupae recovery was extremely low. Our findings indicate that both hypotheses tested are complementary and help better understand host use by A. ludens. However, we also discuss the complexity of polyphagy considering other factors such as plant resistance/defence mechanisms which are not fully addressed in both theories tested.