No abstract

Generalist predators in terrestrial arthropod communities have traditionally been viewed as predators whose dynamics are less tightly coupled to any particular prey species, but whose ecological roles are in other respects analogous to those of specialist predators. Biological-control theory for predator-prey interactions has been based upon a model of communities composed of three discrete trophic levels-plants, herbivores, and predators-in which biological control agents are top consumers and in which different species of predators interact only through competition for shared prey. Experiments employing single-plant field enclosures have suggested, however, that some generalist predators in the cotton agroecosystem function as higher-order predators, releasing populations of an herbivore, the cotton aphid Aphis gossypii, from control by another predator, the lacewing Chrysoperla carnea. Here we demonstrate through focal observations of neonate C. carnea foraging freely in the field that the high levels of mortality observed experimentally are not an artifact of cage confinement. Five generalist predators in the order Hemiptera were observed preying on neonate C. carnea. Neither cannibalism nor predation by heterospecific chrysopids was observed. The only other potential source of lacewing mortality observed was dislodgment from the plant, which occurred primarily on trichome-rich plant structures. A model of terrestrial arthropod communities incorporating higher-order predators may provide valuable insights into the regulation of herbivore populations and suggest useful avenues for biological-control research.

Generalist predators in terrestrial arthropod communities have traditionally been viewed as predators whose dynamics are less tightly coupled to any particular prey species, but whose ecological roles are in other respects analogous to those of specialist predators. Biological‐control theory for predator–prey interactions has been based upon a model of communities composed of three discrete trophic levels—plants, herbivores, and predators—in which biological control agents are top consumers and in which different species of predators interact only through competition for shared prey. Experiments employing single‐plant field enclosures have suggested, however, that some generalist predators in the cotton agroecosystem function as higher‐order predators, releasing populations of an herbivore, the cotton aphid Aphis gossypii, from control by another predator, the lacewing Chrysoperla carnea. Here we demonstrate through focal observations of neonate C. carnea foraging freely in the field that the high levels of mortality observed experimentally are not an artifact of cage confinement. Five generalist predators in the order Hemiptera were observed preying on neonate C. carnea. Neither cannibalism nor predation by heterospecific chrysopids was observed. The only other potential source of lacewing mortality observed was dislodgment from the plant, which occurred primarily on trichome‐rich plant structures. A model of terrestrial arthropod communities incorporating higher‐order predators may provide valuable insights into the regulation of herbivore populations and suggest useful avenues for biological‐control research.

Recent work in terrestrial communities has highlighted a new question: what makes a predator act as a consumer of herbivores versus acting as a consumer of other predators? Here we test three predictions from a model (Rosenheim and Corbett in Ecology 84:2538-2548) that links predator foraging behavior with predator ecology: (1) widely foraging predators have the potential to suppress populations of sedentary herbivores; (2) sit and wait predators are unlikely to suppress populations of sedentary herbivores; and (3) sit and wait predators may act as top predators, suppressing populations of widely foraging intermediate predators and thereby releasing sedentary herbivore populations from control. Manipulative field experiments conducted with the arthropod community found on papaya, Carica papaya, provided support for the first two predictions: (1) the widely foraging predatory mite Phytoseiulus macropilis strongly suppressed populations of a sedentary herbivore, the spider mite Tetranychus cinnabarinus, whereas (2) the tangle-web spider Nesticodes rufipes, a classic sit and wait predator, failed to suppress Tetranychus population growth rates. However, our experiments provided no support for the third hypothesis; the sit and wait predator Nesticodes did not disrupt the suppression of Tetranychus populations by Phytoseiulus. This contrasts with an earlier study that demonstrated that Nesticodes can disrupt control of Tetranychus generated by another widely foraging predator, Stethorus siphonulus. Behavioral observations suggested a simple explanation for the differing sensitivity of Phytoseiulus and Stethorus to Nesticodes predation. Phytoseiulus is a much smaller predator than Stethorus, has a lower rate of prey consumption, and thus has a much smaller requirement to forage across the leaf surface for prey, thereby reducing its probability of encountering Nesticodes webs. Small body size may be a general means by which widely foraging intermediate predators can ameliorate their risk of predation by sit and wait top predators. This effect may partially or fully offset the general expectation from size-structured trophic interactions that smaller predators are subject to more intense intraguild predation.