The Role of Race Specific Resistance in Natural Plant Populations

Abstract

Disease resistance in natural plant-pathogen interactions can be divided broadly into resistance that is expressed against all isolates of a pathogen (race nonspecific resistance) and resistance that is expressed only against specific pathogen phenotypes (race specific resistance). Variation in race non-specific resistance is usually controlled by many genes and may be expressed as a broad range of morphological (pubescence, floral structure), biochemical (secondary phenolic compounds), phenological (earliness) or ontogenetic traits (Burdon 1987a). Because of its perceived broad genetic base or expression as major morphological characters (e.g. the 'hooded' trait in barley), race non-specific resistance is widely regarded as being 'invulnerable' to a process of reciprocal pathogen co-evolution. In contrast, variation in race specific resistance is usually controlled by single major genes in the host. Particular pathogen isolates may possess corresponding virulence alleles and hence be able to render the resistance ineffective. Major genes for resistance have been identified in a wide range of plants including annual (e.g. Senecio vulgaris, Amphicarpaea bracteata, Hordeum spontaneum) and perennial (Linum marginale, Glycine canescens) herbs, shrubs (Coffea arabica) and even trees (Populus spp., Pinus spp.) (Burdon 1987a, Thompson and Burdon 1992). Such resistance genes may be found in many different populations of a species in many different geographic regions. The presence of race specific resistance in natural plant-pathogen associations is sufficiently established (Table 1) that it can no longer be dismissed as 'an artifact of agriculture' (Barrett 1985). However, its role and importance is the subject of some debate. One school of thought sees such resistance as essentially 'one-off and ephemeral' in effect and that more broadly based race non-specific resistance mechanisms are the essential part of a plant's defensive armoury in natural situations. This viewpoint assumes that a pathogen race with virulence genes matching all the major resistance genes present in a host population will inevitably arise or arrive; with time this race will inevitably dominate the pathogen population and, consequently, any fitness advantage associated with the corresponding host resistance genes will be lost permanently. Superficial circumstantial evidence for this view may be derived from two sources. First is the seemingly inevitable loss of effectiveness that accompanies the use of major gene resistance to counter rust and mildew diseases of agricultural small grain crops. Second is the paradoxical observation that although the frequency of genes for race specific disease resistance may be high in natural host populations occurring in areas where the environment is particularly conducive for pathogen development, disease may frequently still occur (Burdon et al. 1983, Dinoor and Eshed 1990). We believe this view assumes too great a degree of long-term stability in individual plant-pathogen associations at the population level. Indeed, it reflects perceptions about plant-pathogen dynamics that spring from agricultural rather than natural systems; and from the infinite population size assumptions of simple population genetic models that invoke selection but ignore stochastic processes due to the complexities of the subdivided real world. In reality, interactions between plants and their pathogens occur in a highly subdivided world. Individual populations of plants are typically unevenly distributed across the landscape in various degrees of isolation to form part of a broad metapopulation framework. While the metapopulation as a whole may show considerable stability its con-