Competition and coexistence of larval ant lions

Ecology, Sept, 1997 by Nicholas J. Gotelli

INTRODUCTION

The coexistence of species remains a central, unsolved problem in community ecology (Case and Diamond 1986, Gotelli and Graves 1996). If a pair of species consumes shared resources, why doesn't one member of the pair go extinct? Traditionally, there have been three kinds of answers. First, factors such as predation or disturbance may suppress densities to chronically low levels, so that shared resources never become limiting (Connell 1975, Sousa 1984). Second, coexistence on a local scale may be transitory, but regional coexistence may be maintained through immigration and patch dynamics (Hanski 1983, Wilson 1992). Third, competitors may partition available resources so that species coexist in a stable equilibrium, but at a lower abundance than they would in the absence of a competitor (MacArthur 1972, Tilman 1982).

Resource use and morphology of adult organisms have provided the traditional framework for studying resource partitioning (Wiens 1982). However, different size and age classes in a population may effectively function as different "species" in resource use (Wilson 1975, Polis 1984), particularly in animals with complex life histories (Wilbur 1988). Moreover, many species adopt different trophic roles depending on their stage or size (Werner and Gilliam 1984). Some species pairs may interact both as competitors and as predators of one another (Polls and McCormick 1987, Wissinger 1989, Moran 1995). This phenomenon of intraguild predation is widespread in nature (Polis et al. 1989) and may enhance or destabilize the coexistence of competing species (Polis and Holt 1992).

In this paper, I describe the results of six manipulative field experiments and one laboratory experiment designed to test for the effects of intraguild predation on the coexistence of two species of larval ant lion (Neuroptera: Myrmeleontidae). My experiments manipulate food, spatial arrangement, and larval density, and assess their effects on different combinations of species and larval instars. The results provide insight into mechanisms that promote species coexistence.

MATERIALS AND METHODS

The study system

Ant lion larvae are ideal for studying species coexistence. Larvae are sit- and-wait predators that capture arthropod prey - including other ant lions - in sand pits (Plate 1) (Wheeler 1930, Topoff 1977). There are three larval instars. Larval development in the field takes 1-2 yr, so overlapping generations of larvae coexist. Larvae pupate in the soil, and adult ant lions emerge in the spring and summer. The adult ant lion is short lived; it ecloses, mates, and oviposits in the soil (Wheeler 1930).

In Caddo County, Oklahoma, Myrmeleon crudelis and M. immaculatus coexist at the sheltered bases of sandstone cliffs, where densities can exceed 100 larvae/[m.sup.2]. Both species are restricted to this microhabitat by rainfall and high temperature in exposed areas (Gotelli 1993). Within the ant lion zone, there is no spatial segregation of the two species, although their numbers are negatively correlated in small-quadrat samples (Gotelli 1993). Species coexist stably in this region, and I have collected both species at a number of cliff ledges in Caddo County (Salyer East, Salyer West, Pugh Canyon, North Canyon) for the past 7 yr.

Third-instar larvae of M. immaculatus are larger than those of M. crudelis (Lucas and Stange 1981), but there is considerable overlap of body size and feeding habits among all instars. Ants are the most common potential prey item in pitfall catches (Marsh 1987, Lucas 1989, Gotelli 1993, 1996), and the probability of capture is inversely related to ant body size (Gotelli 1996). However, field experiments reveal that ants can detect and effectively avoid high-density ant lion aggregations (Gotelli 1996).

Collectively, these observations suggest that both space and food may be limiting resources for ant lion larvae. Space is probably limiting because abiotic factors restrict both species to a specialized microhabitat. Food is probably limiting because the most common prey taxon is effective at avoiding ant lion aggregations. In the following sections, I describe field experiments that test for the effects of space and food limitation and reveal the mechanisms of species coexistence. Because of the potential for complex trophic interactions, different larval instars are treated as different "species" in intra- and interspecific competition experiments.

Overview of the experiments

I conducted six field experiments and one laboratory experiment to test for the effects of competition among larvae for food or space (Table 1). In all experiments, larvae were maintained in enclosures that restricted their movement but allowed them to construct natural feeding pits and intercept ambient prey. The following section describes the methods and procedures that are common to all of the experiments. Then I provide details for each experimental manipulation.

Field procedures

In Experiments 1-5, larvae were collected randomly from the field and returned to the laboratory for identification, weighing, and treatment assignment. Larvae were identified to species by the width: length ratio of the head capsule and by the pigmentation pattern on the head capsule. Both traits vary predictably between species (Lucas and Stange 1981). Larval stages were determined by head capsule width. Each larva was weighed to the nearest 0.1 mg in the laboratory and assigned randomly to one of the experimental treatments. One-way ANOVAs confirmed that there were no significant differences (P [greater than] 0.30) in initial body mass of individuals of the same stage and species that were assigned to different treatments. Within a replicate, each larva was uniquely marked with a tiny spot of colored enamel paint on the dorsal side of the abdomen. Larvae were held in plastic dishes (12 cm diameter, 2 cm depth) filled with 270 g of oven-dried sand collected from the ant lion zone. Larval densities in these experiments bracketed the range of densities typically encountered in the field (Gotelli 1993).