Bidirectional facilitation and interference between shrubs and annuals in the Mojave Desert - Mojave Desert, CA

Ecology, July, 1999 by Claus Holzapfel, Bruce E. Mahall

INTRODUCTION

Understanding the relationships of coexisting species forms a central theme of community ecology (Watt 1947, Roughgarden and Diamond 1986). Composition and structure of a community are shaped by both abiotic factors and interactions among organisms. While the structure of plant communities can be characterized with descriptive methods, the specific types of interactions partly responsible for that structure are usually less evident, and their resolution requires thorough, experimental analysis. Even though positive interactions between plant species have been described in the scientific literature since the turn of the century (Phillips 1909, see also Kropotkin 1902 for a general account), research in community ecology has focused primarily on negative interactions, and interference has often been considered the key factor shaping communities (Clements et al. 1929, Schoener 1983, Keddy 1989, Cornell and Lawton 1992). Only after it became evident that negative interactions alone could not account for all observed vegetation patterns (see Schoener 1982, Goldberg 1990), were positive interactions considered again (e.g., Hunter and Aarssen 1988, Wilson and Agnew 1992, Bertness and Callaway 1994, Bertness and Hacker 1994, Callaway 1995, Callaway and Walker 1997).

That both positive and negative effects can be important simultaneously in interactions between species has been by now demonstrated in a number of studies (e.g., Walker and Chapin 1986, Callaway et al. 1991, Aguiar et al. 1992, Bertness and Shumway 1993, Aguiar and Sala 1994), and the interaction between positive and negative effects has been discussed (Callaway 1994, Callaway and Walker 1997, Holmgren et al. 1997). However, even though it has been recognized that the overall or net effect of one species on another may be the sum of both positive and negative effects (Callaway 1995), net effects have never been experimentally dissected into their positive and negative components. Such a dissection would allow an evaluation of the degree to which positive and negative effects compensate for each other and result in variations in net effects through space and time. Net effects can result in facilitation (positive effects [greater than] negative effects) or interference (positive effects [less than] negative effects) of one neighbor by another. Furthermore, positive and negative interactions between neighbors are very likely to be bidirectional, including positive and negative effects in both directions. This bidirectionality has seldom been assessed. Most studies have addressed only the effects of one neighbor on another. Thus, a truly mechanistic understanding of plant interactions and of the long-term changes or stability in plant associations which such interactions direct, will be possible only after the bidirectional interplay between positive and negative effects of neighboring species on each other is quantitatively assessed.

The apparent structural simplicity and stark, salient abiotic features of deserts, plus the prevalence of close, preferential associations between shrubs and herbaceous plants (often annuals) in many deserts of the world (West and Klemmedson 1978, Noy-Meir 1979) suggest that the desert shrub/annual association provides an ideal model system for a quantitative evaluation of bidirectional, positive and negative interactions among neighboring plants. Facilitative effects of shrubs on annuals have been described in the deserts of the southwestern United States (Went 1942, Muller 1953, Muller and Muller 1956, Halvorson and Patten 1975, Patten 1978, Shmida and Whittaker 1981) and the Middle East (Zohary 1973, Sarig et al. 1994, Tielborger and Kadmon 1995, 1997). In contrast, reports on interference, such as apparent allelopathic suppression of annuals by shrubs (Friedman et al. 1977) or resource competition between herbaceous plants and shrubs, are comparatively rare for deserts (Casper 1996, Tielborger and Kadmon 1997). Facilitation in arid areas often involves water and/or nutrient availability. Microclimates beneath shrubs are characterized by lower radiation exposures and thermal amplitudes, and, therefore, lower evaporative demands compared to conditions in open, inter-shrub areas (e.g., Keeley and Johnson 1977, Nobel 1980, Fuentes et al. 1984, Belsky et al. 1989, Vetaas 1992). Soil surface coverage by herbaceous plants and their litter has the potential of increasing soil water content by increasing infiltration and decreasing evaporation from the soil surface (Evans et al. 1981, Knoop and Walker 1985). "Hydraulic lift" (Richards and Caldwell 1987) is a mechanism by which deep roots of woody plants can transport water to shallow soil layers, where it may be utilized by associated shallow-rooted plants (Caldwell 1990, Dawson 1993). Elevated concentrations of nutrients in the soil under woody plants as compared to that in open interstices between woody plants have been found in numerous studies of arid ecosystems (Radwanski and Wickens 1967, Garcia-Moya and McKell 1970, Bate 1981, Weltzin and Coughenour 1990). These "islands of fertility" (Garcia-Moya and McKell 1970) appear to be due to accumulation of wind-borne organic material and litter deposition (Radwanski and Wickens 1967, Bernhard-Reversat and Poupon 1980, McNaughton 1983, Caldwell and Manwaring 1994).