Predatory interactions between prey affect patch selection by predators

Abstract

Abstract: When predators can use several prey species as food sources, they are known to select prey according to foraging efficiency and food quality. However, interactions between the prey species may also affect prey choice, and this has received limited attention. The effect of one such interaction, intraguild predation between prey, on patch selection by predators was studied here. The predatory mite Neoseiulus californicus preys on young larvae of the western flower thrips Frankliniella occidentalis and on all stages of the two-spotted spider mite Tetranychus urticae. The two prey species co-occur on several plant species, on which they compete for resources, and western flower thrips feed on eggs of the spider mites. A further complicating factor is that the thrips can also feed on the eggs of the predator. We found that performance of the predatory mite was highest on patches with spider mites, intermediate on patches with spider mites plus thrips larvae and lowest on patches with thrips larvae alone. Patch selection and oviposition preference of predators matched performance: predators preferred patches with spider mites over patches with spider mites plus thrips. Patches with thrips only were not significantly more attractive than empty patches. We also investigated the cues involved in patch selection and found that the attractiveness of patches with spider mites was significantly reduced by the presence of cues associated with killed spider mite eggs. This explains the reduced attractiveness of patches with both prey. Our results point at the importance of predatory interactions among prey species for patch selection by predators.

Significance statement: Patch selection by predators is known to be affected by factors such as prey quality, the presence of competitors and predators, but little is known on the effects of interactions among prey species present on patch selection. In this paper, we show that patch selection by a predator is affected by such interactions, specifically by the feeding of one prey species on eggs of the other.

Keywords: Behaviour; Competition; Neoseiulus californicus; Prey-prey interactions; Shared predator; Spider mites; Thrips.

Fig. 1

Interactions among plants, the two-spotted spider mite T. urticae, the western flower thrips F. occidentalis and the predatory mite N. californicus. Direct and indirect interactions are shown as solid and dotted lines respectively

Fig. 2

Reproduction of predators when being supplied with spider mites alone, thrips larvae alone and the two species together, during 2 and 7 days. Different letters above the bars indicate significant differences (P < 0.0001, GLM)

Fig. 3

a Patch selection and b oviposition (+SEM) by predatory mites when offered a choice between the following two leaf discs: a disc with spider mites vs. a disc with first-instar thrips larvae (the first bar), an empty disc vs. a disc with first-instar thrips larvae (the second bar), an empty disc vs. a disc with second-instar thrips larvae (the third bar), a disc with first-instar thrips larvae vs. a disc with second-instar thrips larvae (the fourth bar), an empty disc vs. a disc with spider mites and second-instar thrips larvae (the fifth bar), a disc with spider mites vs. a disc with spider mites and second-instar thrips larvae (the sixth bar) and a disc with second-instar thrips larvae vs. a disc with spider mites and second-instar thrips larvae (the seventh bar). Asterisks indicate significance of the preference: ns not significant; *P < 0.05, **P < 0.01, ***P < 0.001 (binomial test for the adult predators (a) and Wilcoxon matched pairs signed-rank test for the eggs (b)). Different letters next to the bars indicate significant differences in the total numbers of eggs found on both discs among experiments (P < 0.0001, GLM)

Fig. 4

a Patch selection and b oviposition by predatory mites when offered two leaf discs. a Adult female patch selection, showing the proportion of adult female predators found on each of the two discs. b Oviposition preference, showing the numbers of eggs produced on each leaf disc (+SEM). Each disc contained eggs and cues of spider mites. In addition, one of the leaf discs contained cues of 10 second-instar thrips larvae (first bars from above) or cues of 10 first-instar thrips larvae (second bars). The third bars concern a choice between leaf discs (both with spider mite eggs and cues), of which one contained additional cues of 10 second-instar thrips larvae and with the other with additional cues of 27 first-instar thrips larvae. The fourth bars refer to an experiment where spider mites and thrips produced cues each on a separate half of one of the leaf discs so that thrips could not prey on spider mite eggs. This disc was offered together with a disc with only spider mite eggs and cues. The fifth bars concern a choice between a disc with eggs and cues produced by 30 adult female spider mites and a disc with eggs and cues of 15 spider mites. The last bars show the choice of predators between a disc with eggs and cues of 30 spider mites and a similar disc, on which eggs were damaged to mimic egg predation. See legend to Fig. 3 for details of statistics. Asterisks indicate significance of the preference: ns not significant; **P < 0.01, ***P < 0.001