Species Interactions In Intertidal Food Webs: Prey Or Predation Regulation Of Intermediate Predators? - Statistical Data Included

Ecology, August, 2000 by Sergio A. Navarrete, Bruce A. Menge, Bryon A. Daley

SERGIO A. NAVARRETE [1,3]

BRUCE A. MENGE [2]

BRYON A. DALEY [2,4]

Abstract. Most natural food webs have more than one predator species, and many have trophic interactions among these predators. When a top predator feeds on an intermediate predator and they both feed on a shared basal resource, a phenomenon labeled intraguild predation (IGP), the potential exists for complex food web dynamics due to predation and competitive effects. Here we investigate the relative importance of direct predation vs. competition by a top predator on an intermediate predator.

The study system is the rocky intertidal interaction web formed by the predatory seastar Pisaster ochraceus, the predatory whelks Nucella emarginata and N. canaliculata, and a shared resource species, the mussel Mytilus trossulus. Previous experiments documented strong negative effects of Pisaster on mussels and whelks, but the mechanisms responsible for the effects on whelks, whether competition or predation, were not identified. Here we report results of a field experiment that manipulated both Mytilus and Pisaster to determine the short- and longer-term changes in whelk populations. Using a simplified dynamic model for changes in abundance over the initial stages of the experiment, we separated and quantified the top-down effect of direct predation by seastars vs. the bottom-up effects of competition and food limitation. Short-term results were in agreement with longer-term responses. Results suggest that direct and indirect bottom-up influences of mussels were far stronger than predation, and thus, whelk increases in the absence of seastars were due to reduced competition with Pisaster. Large differences between the body sizes of seastars and whelks make it difficult to determine the ultimate nature of the resource under competition between predators. Small mussels may constitute only a food resource for seastars, but to some extent, they also represent a microhabitat for whelks. Differences in the magnitude of the response to mussel manipulations between Nucella species might be due to slight differences in the way the species utilize mussel beds. Some of the predictions of theoretical IGP models regarding coexistence and stability of species may not apply to this interaction web because it includes species with both "open" and "closed" populations, rather than just closed populations as assumed by the models.

Key words: bottom-up effects; effect sizes; exploitation competition; intraguild predation, mussels; Mytilus; Nucella; Pisaster; rocky intertidal; seastars; top-down effects; whelks.

INTRODUCTION

Most food webs have more than one predator, and many also have a complex pattern of competitive and trophic interactions among these predators. In the simplest food webs, predators feed on the trophic level immediately below: a "top" predator feeds on an "intermediate" predator, which in turn feeds on herbivores or basal species. When this pattern of trophic interactions dominates food webs, the removal of species at the top trophic level results in a chain of effects producing a classic trophic cascade with alternation of positive and negative biomass responses between trophic levels (e.g., Hairston, Smith, and Slobodkin 1960, Oksanen et al. 1981, Carpenter et al. 1985, Bronmark et al. 1992, Hunter and Price 1992, Moran et al. 1996). Specific predictions about the forces that regulate different trophic levels can then be made and some of them have been experimentally tested (e.g., Bronmark et al. 1992, Wootton and Power 1993, Wootton et al. 1996). However, this food web scenario can be substantially complic ated by the existence of intraguild predation (IGP), in which a top predator feeds on an intermediate predator and both feed on a shared prey (Polis et al. 1989). In this case, predictions about the effects of predation will depend on the relative strengths of exploitation competition for prey, vs. direct predation of one predator on another (Spiller and Schoener 1990, Polis and Holt 1992, Holt and Polis 1997).

The widespread occurrence of intraguild predation and other forms of omnivory in natural food webs (e.g., Bradley 1983, ter Braak 1986, Menge and Sutherland 1987, Polis 1991, 1994, Diehi 1992, 1993) has led Polis and Strong (1996) to question the validity of the trophic level concept and the generality of trophic cascades. If intraguild predation among predators is strong, predation should keep predator densities low and prey abundance is predicted to be high (Fig. 1, model A). Alternatively, if intraguild predation is weak, but predator effects on shared prey are strong, exploitation competition may structure the higher trophic levels and prey should be subject to severe reductions (Fig. 1, model B; see Spiller and Schoener [1990], Diehl [1995]). A plethora of indirect effects, including resource competition, apparent competition, and mutual enhancement are possible in food webs with IGP (Polis et al. 1989). A simple theoretical analysis of a three species community with IGP, leads to the prediction that fo r an intermediate IG predator to coexist with a top IG predator (see Fig. 1), the former must be superior at exploiting shared prey resource (Polis and Holt 1992). Building on these simple models, Holt and Polis (1997) have recently formulated a basic theoretical framework for intraguild predation, leading to the identification of a series of conditions necessary for the stable coexistence of IG predators, IG prey and a basal resource species. Predictions from these models have not been experimentally tested.