Competition and density-dependent fitness in a plant parasitic fungus

Ecology, Sept, 1997 by M.R. Newton, L.L. Kinkel, K.J. Leonard

INTRODUCTION

The potential impacts of intra- or interspecific competition on population dynamics have long been a central topic in the ecological literature (Tansley 1917, Volterra 1926, Clements 1929, Lotka 1932, Gause 1934). A substantial amount of empirical and theoretical work has been devoted to investigating and modeling the effects of competition on population and community dynamics in diverse natural systems (Roughgarden 1979, Connell 1983, Goldberg and Barton 1992). However, research on microbial competition is not well represented in the ecological literature, particularly in the realm of ecological theory, perhaps because population dynamics for many microbes are not easily quantifiable (Seifert 1981, Andrews and Harris 1986, Keddy 1989).

Several plant pathogenic fungi have properties suitable for studying the effects of competitive interactions on individual fitness (per-individual contribution to the next generation's gene pool; Roughgarden 1979). Specifically, they have readily quantifiable components of fitness, such as discrete, sporulating lesions that result from infections by single spores that produce new generations of spores after a distinct latent period (Leonard 1990). In this work we use the wheat stem rust fungus, Puccinia graminis f. sp. tritici (Pgt), on wheat leaves as a model system for studying microbial competitive interactions.

Among coexisting organisms, the effects of competition on reproductive output increase with increasing population density (Arthur 1987). In rust fungi, as the density of propagative units increases, the reproductive output per unit generally decreases, in terms of either the number of sporulating pustules formed relative to the number of spores inoculated (Petersen 1959, Davison and Vaughan 1964, Sun and Zeng 1993), or spore production per pustule (Yarwood 1961, Leonard 1969a, Imhoff et al. 1982, Shaner 1983, Kardin and Groth 1989). The fitness of each organism thus will depend on the absolute densities of all strains or species present in mixed populations.

To date, research on population dynamics in mixtures of foliar microbes has largely failed to consider the importance of density effects on reproductive output. Previous studies have commonly focused on observing changes in the proportions of strains in mixed cultures over time, given specified initial proportions, rather than initial absolute densities of the strains in the mixture (Irish 1950, Loegering 1951, Broyles 1955, Thurston 1961, Katsuya and Green 1967, Ogle and Brown 1970, Yang and TeBeest 1995, Liu et al. 1996). Although efforts have been made to mathematically model changes in proportions of strains over multiple generations (Leonard 1969b, Ogle et al. 1973, Ostergard 1987, 1996, Welz and Leonard 1993, Huang et al. 1994), only the model of Ostergard (1996) considers the influence of the absolute densities of both strains in a mixture on their relative reproductive outputs. Other studies of microbial competition have used replacement series designs (de Wit 1960), in which the outcome of competition for two strains in mixture is assessed over a range of proportions of strains in the mixture, while the initial total population density of the mixture is held constant (Rai and Upadhyay 1983, Widden 1984, da Luz and Bergstrom 1987, Widden and Hsu 1987, Adee et al. 1990, Wilson and Lindow 1995). However, in these designs, changes in the absolute density of one strain are confounded with concurrent changes in the density of the other, and it is therefore impossible to fully describe the dependence of the competitive outcome on the absolute densities of both strains (Firbank and Watkinson 1985, Connolly 1988, Snaydon 1991).

In this work we present a model that predicts the reproductive output of a pathogen strain as a function of its population density and the density of a competitor. Two different components of reproductive output corresponding to different stages of the pathogen's life cycle, formation of uredinia (pustules) on leaves and production of urediniospores by the uredinia, are modeled as functions of the strains' inoculum and uredinial densities, respectively. Thus, the model permits density-independent measurement of the competitive effects of strains on one another's uredinial formation and urediniospore production. Competitive effects are defined as the effect of a propagative unit (a urediniospore infecting a leaf, or a sporulating uredinium on a leaf) of one strain on the reproductive output (formation of uredinia from initial infections in the leaf, or production of urediniospores by uredinia on the leaf) of the other strain, relative to the effect of a unit of that other strain on its own reproductive output.

The model and the experimental approach used in this research offer a way to define and measure competitive effects independently from intrinsic reproduction parameters. Thus, the contributions of competitive effects and intrinsic reproduction parameters to fitness can be distinguished and compared.