Indirect facilitation: evidence and predictions from a riparian community

Ecology, July, 1999 by Jonathan M. Levine

INTRODUCTION

While the influence of biotic interactions on species diversity is well appreciated (Huston 1994, Rosenzweig 1995), the converse, how diversity influences biotic interactions is only beginning to be explored. Understanding these community-level consequences of diversity is required for differentiating the ecological forces that structure diverse vs. species-poor communities and predicting the impact of species invasion or loss. Models of diffuse competition, or "competition by a constellation of species" (sensu MacArthur 1972), provide rare insights into how the intensity of species interactions changes as the number of species in a community varies (MacArthur 1972, Pianka 1974, Lawlor 1979, Bender et al. 1984, Case 1991, Stone and Roberts 1991). While the original models suggested that increased species number led to more intense competition (MacArthur 1972, Pianka 1974), these studies did not incorporate the numerous indirect interactions that emerge when species are added. Later modeling efforts, accounting for indirect interactions, concluded that increased species number could reduce the intensity of competition, or even lead to facilitation (Levins 1975, Levine 1976, Lawlor 1979, Davidson 1980, Yodzis 1988, Vandermeer 1990, Case 1991, Stone and Roberts 1991).

One of the most influential examinations of how community processes are modified by species addition or loss was provided by the three-species model of Stephen Levine (1976). Levine explored how the addition of a third competitor to a system of two competing species could alter the interaction between these species from competition to facilitation. The model suggested that three species not only compete, but also indirectly facilitate one another via suppressing a shared competitor [ILLUSTRATION FOR FIGURE 1a OMITTED]. Levine pointed out that, when the direct competitive effect of one species on another was weaker than the indirect positive effect, a net facilitative interaction resulted from a system composed entirely of competitors. Other indirect facilitations can be mediated by shared consumers or pathogens or by competing prey (Levine 1976, Vandermeer et al. 1985). I will use "apparent facilitation" (sensu Davidson 1980) to refer specifically to interactions mediated by a shared competitor. Obtaining empirical support for apparent facilitation has proven challenging because of the difficulty in separating out direct and indirect effects that oppose one another (Brown et al. 1986). Probably for this reason, few studies have explicitly examined these interactions in the field (Davidson 1980, 1985, Culver 1982, Miller 1994).

Empirical studies of indirect effects among plants are particularly rare (Miller 1994, Callaway and Pennings 1998). This may be attributed to the prevalence of strong direct interactions among plants (Goldberg and Barton 1992), yet it is precisely these potent direct interactions that lead to strong indirect effects among competitors (Davidson 1985). Furthermore, plants undoubtedly influence one another via their effects on herbivores, parasites, pathogens, and decomposers; and, while, models of these trophically mediated interactions (particularly, apparent competition) are well recognized, empirical studies are few (Clay 1990, Connell 1990, Goldberg 1990). For this reason, our understanding of when indirect interactions between plants can or cannot be ignored is largely undeveloped, though the importance of these interactions in communities is potentially large.

Here, I report the results of a field experiment designed to examine the strength of indirect facilitation among plant species in a riparian community. In particular, I explore the hypothesis that the torrent sedge Carex nudata modifies the competitive effects of the common monkey-flower Mimulus guttatus on three other common plant species along a northern California river [ILLUSTRATION FOR FIGURE 1b OMITTED]. I evaluate indirect facilitation in a simple field experiment where I vary the abundance of Carex and M. guttatus in a factorial design, thus allowing me to compare the effects of Carex on three target species, with and without M. guttatus.

METHODS

Study system

Large, discrete, tussocks formed by the torrent sedge Carex nudata W. Boott line the channels of many northern California streams. Along the South Fork Eel river in northern California, USA (39 [degrees] 44 [minutes] N, 123 [degrees] 39 [minutes] W), Carex tussocks average 30 cm in diameter, are densely covered with stems from late spring to fall, and occupy nearly one-third of the stream channel exposed by midsummer (J. Levine, unpublished data). Carex appears to have direct positive effects on other plant species. Of the [greater than] 60 plant species occurring in this habitat, most are found growing almost exclusively on Carex tussocks, rooting into the accumulated sediment. By growing on tussocks, these small perennials obtain a stable refuge from scouring winter floods (J. Levine, unpublished data). This direct positive effect of Carex is clearly important, but here I examine the indirect facilitative effects of Carex, mediated by its competitive interactions. These competitive effects arise because the associated species grow within the dense Carex root mat and thick vegetation.