Source-sink dynamics for a generalist insect predator in habitats with strong higher-order predation

The functional importance of higher-order predators in terrestrial ecosyste ms is currently an area of active inquiry. In particular, an understanding of the relative influences of prey availability and higher-order predation on predator populations is of immediate relevance to the theory of biologic al control of herbivorous arthropods. Biological control workers have repea tedly speculated that one cause of failure to establish predators and paras itoids in novel environments is the strong mortality imposed on released ag ents by higher-order predators. Nevertheless, the ability of higher-order p redators to create a habitat where mortality exceeds natality (a "sink" hab itat) has never been tested experimentally with a biological control agent in nature. Although in isolation the predatory lacewing Chrysoperla carnea can consist ently produce strong Suppression of populations of the aphid Aphis gossypii , the full community of predators when tested together exerts minimal aphid control. The age structure of Chrysoperla ln spp. populations in cotton fi elds harboring low to intermediate densities of aphid prey is characterized by a sharp drop in densities from the egg to the first larval instar; this observation is consistent with heavy mortality during either the egg or fi rst larval stage. Egg cohorts followed under unmanipulated field conditions showed relatively high rates of successful hatch, suggesting that the vuln erable developmental stage is the young larva. Larval survival is relativel y high in the absence of hemipteran predators, suggesting that prey availab ility is not the primary limiting factor. Depressed survival is observed in the presence of Geocoris spp., Nabis spp., and Zelus renardii, all common hemipteran predators in cotton. Predation on lacewing larvae appears to dis rupt the strong top-down control of aphid populations in cotton. Chrysoperla spp. densities declined in fields harboring intermediate aphid densities when lacewing subpopulations were experimentally caged to block i mmigration and emigration. In one year (1993) Chrysoperla spp. densities fe ll to very low levels, suggesting that the field was either a true sink hab itat or a pseudosink with a very low equilibrium density. In a second year (1994), densities declined to what appeared to be a lower but stable densit y, suggesting that the habitat was a pseudosink. Thus, in both years, decli nes in Chrysoperla spp. densities were observed following caging, suggestin g that Chrysoperla spp. populations are spatially subsidized. Aphid prey av ailability and higher-order predation interacted strongly in their influenc e on C. carnea survival: larval survival in the presence of higher-order pr edators was 5.6% when prey availability was intermediate and 40.5% when pre y were superabundant. Spatial heterogeneity in aphid prey densities modulat es the intensity of higher-order predation and thereby appears to produce s ource-sink dynamics of Chrysoperla spp. in cotton fields.