Attack strategies of predatory wasps (Hymenoptera: Pompilidae; Sphecidae) on colonial orb web-building spiders (Araneidae: Metepeira incrassata)

Abstract

Pompilid and sphecid wasps differ significantly in the dynamics of their behavior and their attack strategies on colonial orb-web building spiders (Ara neidae: Metepeira incrassata). In response to wasp attacks, the most common antipredator defense of the spiders is to drop from their webs attached to a dragline. This mode of escape is successful against many wasps, but is exploited by one species. The general attack strategy of pompilid wasps is to attack a large number of spiders per bout, follow large spiders which drop from their webs to the ground, and hunt throughout the large three-dimensional spider colonies. One pompilid, Poecilopompilus mixtus, exploits the escape behavior of the spiders and actively induces spiders to jump from their webs, as it is able to track the spiders as they drop and most successfully captures spiders on the ground. In contrast, sphecid wasps attack relatively few spiders per bout, attempt to sting spiders before they drop from their webs as they never follow the spiders to the ground, and hunt only on the periphery of colonies. Differences in the behavior may be due to multiple versus single prey provisioning of nests by the wasps. To reduce the risk of predation, prey species evolve an array of antipredatory behaviors. Successful predators, in turn, evolve behaviors which overcome the prey's antipredator defenses. Predatory wasps from two families of wasps (Pom pilidae, Sphecidae) capture spiders which they paralyze and provide as a living food source for their offspring to develop upon. Because it is relatively easy to locate pompilid and sphecid nests and remove the paralysed spiders for identifi cation, there is a fairly detailed record of the spider type (cursorial, aerial web builder), species, and size of spiders killed by these wasps (e.g., Evans and Yosh imoto, 1962; Krombein, 1967; Kurczewski and Kurczewski, 1968; Coville, 1982, 1987). Evidence from these nests demonstrates that the wasps selectively capture or shift between prey species as their abundance changes in the field (Field, 1992; Endo and Endo, 1994). But there is surprisingly little information on the dynamics of interactions between the wasp and the spider, other than for the attack behavior of the tarantula hawk (Pompilidae: Pep sis) on tarantulas (Theraphosidae) (Petrunkevitch, 1926; Punzo and Garman, 1989). Although orb-web building spi ders (Araneidae, Uloboridae) are frequent prey of pompilid and sphecid wasps, there are few descriptions of techniques used by the wasps to attack these spiders (Soyer, 1950; Eberhard, 1970; Endo, 1976). Yet, it is the dynamics of the con frontation between predator and prey that determine the ultimate choice of prey by the wasps, and the unsuccessful predation events that lead to the evolution of the spiders' antipredator defenses. The colonial orb-web building spider, Metepeira incrassata (Araneidae) is a frequent target for wasps (Rayor and Uetz, 1990, 1993; Uetz and Hie ber, 1994). The three-dimensional colonies, containing hundreds to thousands of 1 Present address: Department of Entomology, Cornell University, Ithaca, New York 14853. This content downloaded from 207.46.13.156 on Sat, 10 Sep 2016 05:26:06 UTC All use subject to http://about.jstor.org/terms 68 JOURNAL OF THE KANSAS ENTOMOLOGICAL SOCIETY spiders, present a super-abundant resource for wasps able to overcome the intricacies of the webs making up the colony structure and the antipredator defenses of the spiders. The purpose of this paper is to describe the success of different attack strategies used by pompilid and sphecid wasps which prey on M. incrassata. Major differences in the spider size, location, and number attacked by wasps are compared between the two groups. Field Sites and Methods The study was done in Fortin de las Flores, Veracruz, Mexico (19?N Lat.; 97?W Long.; 1000 m elev.), in a region of tropical agriculture. The interactions between wasps and spiders were observed as part of a larger study on trade-offs between foraging success and predation risk relative to spatial position in the spider colony (Rayor and Uetz, 1990, 1993). Colonies of the spider, Metepeira incrassata, were composed of hundreds to thousands of orb-webs connected by sturdy framelines. Spider colonies were suspended in open, sunny areas on a small coffee plantation, hotel grounds, and a small flower nursery. The three-dimensional colonies had a characteristic size structure in which larger, more mature spiders occurred pri marily in the colony core, while smaller, immature spiders were more abundant on the periphery. Webs were defined as being in the Core of the colony if they were blocked from the exterior of the colony by at least 2 other webs. Webs on the 'Periphery' had free access to the exterior. Predator activity was primarily observed in 12 moderate sized colonies of 250-400 individual spiders. A small number of the predation events by unidentified sphecid wasps and Trypoxylon sp. (Sphecidae: Larrinae) occurred in a colony of 1000+ spiders. Data from encounters at different colonies for each wasp species were combined in order to have sufficient sample size. See Ray or and Uetz (1990, 1993) for further details on the habitat and spider colonies. The observations in this paper were collected during dry seasons between Sep tember 1987-January 1988, September-December 1988, September-November 1989 and are based on a total of 500 hours of observation over 12 months by the author. Because both the spiders and wasps were diurnally active, observations were made throughout the day from 600-1830 hours. Whenever wasps entered the spider colony, I recorded detailed information on the success or failure of wasp attacks, the location of the attacks, and the size and response of the spider. Attacks were defined as oriented approaches and attempts to capture individual spiders. A predatory bout was composed of the total attacks which occurred sequentially in a discrete time period, separated from other bouts of activity by longer than ten minutes. I attempted to capture new wasp species for identification after they concluded a bout. Four wasp species which were seen attacking spiders once or twice were not successfully captured for identification. The sphecid, Trypoxylon {Trypoxylon) sp., was identified to subgenus by Dr. H. E. Evans, but the specimens were lost in the mail on the way to a museum for further identification.