Interactions between a hunting spider and a web-builder: consequences of intraguild predation and cannibalism for prey suppression

Abstract. 1. Antagonistic interactions among invertebrate predators such as intraguild predation and cannibalism have the potential to dampen top‐down impacts on shared prey at lower trophic levels. Two abundant spider predators, the large wolf spider Pardosa littoralis and the small sheet‐web builder Grammonota trivitatta co‐occur on the salt marshes of eastern North America where they both attack planthoppers (Prokelisia spp.), the dominant herbivores on the marsh. Experiments both in the laboratory and field were used to assess the incidence of intraguild predation and cannibalism in these spiders and elucidate how such antagonistic interactions influence planthopper suppression. 2. Functional response experiments showed that with an increase in planthopper prey density, Grammonota captured more prey but not a higher proportion of that offered. Pardosa exhibited the same response when Grammonota were offered as intraguild prey. Both functional responses were type I over the range of prey densities offered. 3. Grammonota is moderately cannibalistic, and the presence of planthopper prey reduced the incidence of cannibalism. 4. Factorial experiments in the laboratory showed that Pardosa but not Grammonota reduced planthopper prey populations when prey density was low. By contrast, at high prey densities, both Pardosa and Grammonota had significant adverse effects on planthopper populations. Moreover, there was an interactive effect such that Grammonota reduced planthopper populations relatively more when Pardosa was absent than when it was present. 5. There was direct evidence for the intraguild predation of Grammonota by Pardosa such that fewer Grammonota survived in the presence of Pardosa than when it was absent. This result occurred whether planthopper prey were abundant or not. 6. Field releases of Grammonota in open plots resulted in significant but small decreases in the density of planthopper prey, both nymphs and adults. 7. Enhancing densities of Pardosa in open plots resulted in Grammonota suppression. The intraguild predation of Grammonota at this enhanced Pardosa density, however, did not preclude Pardosa from significantly reducing planthopper populations. 8. Although there was evidence that Grammonota reduced planthopper populations and that the intraguild predation of Grammonota by Pardosa occurred, the strength of these interactions was relatively weak given the low consumption rate of planthoppers by Grammonota (< 3 day^–1) and Grammonota by Pardosa (≈ 2 day^?1). Thus, weak asymmetric intraguild predation among spiders on the marsh likely dampens but does not eliminate the ability of Pardosa to exert significant top‐down control on planthopper populations.