The Implications of Intraguild Predation for the Role of Spiders in Biological Control
Total Page:16
File Type:pdf, Size:1020Kb
1999. The Journal of Arachnology 27:351±362 THE IMPLICATIONS OF INTRAGUILD PREDATION FOR THE ROLE OF SPIDERS IN BIOLOGICAL CONTROL Margaret A. Hodge: Department of Biology, The College of Wooster, Wooster, OHIO 44491 USA ABSTRACT. Evidence is growing that spiders can be effective biological control agents, particulary assemblages of several species. Other evidence ®nds that spiders prey on each other and other generalist predators, and as such are of limited value in biological control. Such predatory interactions between species which use similar resources have been dubbed intraguild predation (IGP) due to their potential to modify competition as well as cause direct mortality. IGP interactions can have unexpected effects at other trophic levels, and sometimes result in enhancement of a pest population. In this paper I review the evidence for intraguild predation interactions involving spiders in natural systems, and other generalist predators in agroecosystems. To date not much research has examined whether such interactions in¯uence spider biological control potential. Some suggestions as to how we might begin to address these issues are presented. Given their generalist arthropod diet and that use similar resources (Root 1967; Polis et abundance in most terrestrial habitats, spiders al. 1989; Simberloff & Dayan 1991). Preda- likely in¯ict substantial mortality on insect tory interactions among members of the same populations. While the mechanisms by which guild are termed intraguild predation (IGP). spiders limit insect prey populations have This is distinguished from predation as tradi- been debated (Riechert & Lockley 1984; Wise tionally de®ned because, by eating a guild 1993), it is generally agreed that they are im- member, an individual not only directly gains portant in reducing insect numbers, and as energy and nutrients, but also reduces poten- such are of potential value in biological con- tial competition for food (Polis et al 1989; Po- trol (Riechert & Lockley 1984; Nyf¯er & lis & Holt 1992). Intraguild predation and Benz 1987; Young & Edwards 1990; Wise cannibalism (killing and eating a member of 1993). Several studies have shown that assem- the same species), may have profound effects blages of many predator species may be more on community structure (Polis 1981, 1988; effective at controlling agricultural pests than Polis et al. 1989; Polis & Holt 1992). Given single species augmentation (Chiverton 1986; their ubiquity in terrestrial ecosystems, spiders Riechert & Bishop 1990; Clark et al. 1994; are model organisms to investigate the occur- Provencher & Riechert 1994; Chang 1996; rence and consequences of IGP. Riechert & Lawrence 1997). On the other IGP: predation among potential compet- hand, different species of predators and/or itors.ÐIntraguild predation and cannibalism parasitoids may compete with or prey on each have been shown to directly limit predator other, potentially reducing their biological populations (Polis & McCormick 1986, 1987; control potential (Force 1974; Ehler & Hall Spiller & Schoener 1988; Wissinger 1989; 1982; Spiller 1986; Briggs 1993; Rosenheim Leonardsson 1991; Finke 1994; Wagner & et al. 1995; Chang 1996; Ferguson & Stiling Wise 1996; Wissenger et al. 1996). Since un- 1996; Kester & Jackson 1996; Cisneros & Ro- successful predation attempts represent ex- senheim 1997; Rosenheim 1998). treme forms of interference competition (Polis The nature of the diet of spiders suggests et al. 1989; Elgar & Crespi 1992), IGP can that they can prey on each other and other also lead to behavioral adaptations to reduce arthropod predators (Polis 1981; Jackson mortality and con¯ict, resulting in habitat and 1992; Wise 1993), as well as overlap in prey diet shifts by IG prey (Fox 1975; Turner & taxa consumed, thus potentially competing for Polis 1979; Doncaster 1992; Sih 1982; Eben- resources. Diet overlap is one distinguishing man & Persson 1988; Foster et al. 1988; Polis feature of a guild, a group of sympatric taxa et al. 1989; Polis 1993; Dong & Polis 1992; 351 352 THE JOURNAL OF ARACHNOLOGY Holt & Polis 1997). These changes in foraging Two alternative hypotheses could explain and habitat distribution may in turn have ef- these results: removal of the scorpions could fects at other trophic levels (Polis 1984; Wil- have resulted in competitive release by the bur 1988). spiders in experimental plots, or IGP (by scor- The traditional view of feeding relation- pions) in the control plots may have reduced ships has been to assign species in a com- spider population size. There was no evidence munity to a ``trophic-level'', such as second- of competitive release in that there were no ary consumer (predator), primary consumer differences between the experimental and con- (herbivore), primary producer (plant), and so trol plots in insect prey abundance or spider forth, with each level feeding on the former reproduction. Release from scorpion predation (Krohne 1998). Thus, in classic biological was the most likely cause of the increased control, insect herbivore populations are re- numbers of spiders. duced by addition of predators, and this in Two independent studies on Anolis lizards turn reduces damage to crop plants (van den examined evidence for intraguild predation on Bosch et al. 1982). In reality, however, ani- spiders cascading to populations of shared in- mals may feed from a variety of trophic lev- sect prey. Pacala & Roughgarden (1984) ma- els, especially generalist predators, which take nipulated anole densities in a Carribean forest prey of whatever size they can handle (Polis and found a direct effect of lizards on forest 1988; Polis et al. 1989; Spence & Carcamo ¯oor arthropods, their primary prey, and an 1991; Dong & Polis 1992; Finke 1994). If indirect effect on ¯ying insects, the prey of these prey include younger conspeci®cs or orb-weaving spiders. Since anoles also prey other predators, then control of the herbivore on orb-weavers, the increase in ¯ying insects population is not guaranteed. Various studies on the high density lizard plots was thought suggest that direct effects of one predator on to be due to intraguild predation by the lizards another can indirectly affect a shared prey on the spiders. On Bahamian islands Spiller species by releasing it from intense predation & Schoener (1990) also found a direct effect or competition (Press et al. 1974; Pacala & of lizards on spiders, but no indirect effect on Roughgarden 1984; Hurd & Eisenburg 1990; ¯ying insects. They did, however, observe that Polis & Holt 1992; Rosenheim et al. 1993; more spiders were feeding on lizard removal Wissinger & McGrady 1993; Wootton 1993; plots than on plots where they co-existed with Cisneros & Rosenheim 1997; Fagan & Hurd lizards. The authors hypothesized that inter- 1994). If shared prey are herbivores then the ference competition or predation by lizards indirect effects could cascade to plants, in¯u- may displace spiders from prime web-sites, encing primary productivity, an issue of ag- resulting in a reduction in prey capture for the ricultural relevance. The purpose of this paper spiders. is to review the theory and empirical evidence Although the generalist diet of most spider relevant to the implications of IGP for the po- species suggests that exploitative competition tential role of spiders in biological control of for food should be important (Marshall & herbivorous pests in agriculture. Rypstra, this volume), experimental tests have IGP between spiders and other generalist found little evidence (Schaefer 1978; Wise predators.ÐSeveral studies of IGP in natural 1981; Horton & Wise 1983; Riechert & Cady communities have uncovered direct and indi- 1983; see Spiller 1984 a, b for an exception). rect interactions involving spiders. Polis & In spider removal experiments to test for ex- McCormick (1986, 1987) investigated a desert ploitative competition among four genera of community of arachnids including spiders, web-building spiders, Riechert & Cady (1983) solpugids and scorpions, all generalist preda- not only found no competitive release, but on tors that use similar prey and prey on each some of their removal plots they observed a other. Scorpions were continually removed negative effect of spider removals on the spe- from experimental plots, but not from control cies remaining. They hypothesized that this plots, and the relative abundances of the spi- may have been due to the fact that by remov- ders and solpugids were tracked over time. At ing the other species of spiders, they may have the end of the experiment (29 months), sig- been removing potential prey. ni®cantly more spiders occurred in the scor- Hodge & Marshall (1996) tested Riechert pion removal plots than in the control plots. & Cady's hypothesis that intraguild predation HODGEÐIGP, SPIDERS & BIOLOGICAL CONTROL 353 masked competitive release in their system of ground caused the crickets to move upward web-building spiders on rock outcrops in Ten- on vegetation (a direct effect of spiders on nessee. After 12 weeks of removing each of crickets); but when present with both mantids three species from experimental plots we and lycosids, crickets were captured by man- found that one of the species, Hypochilus tho- tids hunting in the vegetation. Finally, lyco- relli (Araneae, Hypochilidae) had lower body sids may have consumed other cricket preda- condition indices (indicating lower fecundity, tors (other spiders; an indirect effect of spiders Jakob et al. 1996) on spider removal plots as on crickets); this was supported by