The Evolution of Resistance and Tolerance to Herbivores

Tolerance and resistance are two different plant defense strategies against herbivores. Empirical evidence in natural populations reveals that individual plants allocate resources simultaneously to both strategies, thus plants exhibit a mixed pattern of defense. In this review we examine the conditions that promote the evolutionary stability of mixed defense strategies in the light of available empirical and theoretical evidence. Given that plant tolerance and resistance are heritable and subject to environmentally dependent selection and genetic constraints, the joint evolution of tolerance and resistance is analyzed, with consideration of multiple species interactions and the plant mating system. The existence of mixed defense strategies in plants makes it necessary to re-explore the coevolutionary process between plants and herbivores, which centered historically on resistance as the only defensive mechanism. In addition, we recognize briefly the potential use of plant tolerance for pest management. Finally, we highlight unresolved issues for future development in this field of evolutionary ecology. 541 A nn u. R ev . E co l. E vo l. Sy st . 2 00 7. 38 :5 41 -5 66 . D ow nl oa de d fr om a rj ou rn al s. an nu al re vi ew s. or g by I ns tit ut o de E cl og ia , A .C , C on so rc io C O N A C yT o n 06 /0 5/ 09 . F or p er so na l u se o nl y. ANRV328-ES38-22 ARI 24 September 2007 8:7 Resistance: constitutive or induced response of plants against herbivory to avoid or reduce the amount of damage Tolerance: response of plants induced after consumption to buffer the negative fitness effect of damage Alternative redundant strategies of defense: two or more strategies that confer similar fitness benefits against the same selective pressure INTRODUCTION The evolution of plant defense against natural enemies, such as herbivores and pathogens, has been the motif of study of evolutionary biologists during the past three decades, and has retained a central role because the theory of coevolution emerged directly from the study of interactions between plants and herbivores (Ehrlich & Raven 1964). Specifically, studies aimed to understand the origin and maintenance of plant and animal adaptations that serve as mediators of the interactions (Rausher 1996). These studies encompass an ample range of foci, from cell and molecular responses to herbivore damage, to phylogenetic patterns of relationships (Agrawal & Fishbein 2006, Becerra 1997, de Meaux & Mitchell-Olds 2003). After the recognition arose that many plant attributes function as defenses against herbivores and pathogens (Ehrlich & Raven 1964, Painter 1958), the understanding of trait evolution, and coevolutionary relationships, received considerable attention (Bergelson et al. 2001; Rausher 1996, 2001). Today, resistance and tolerance are well recognized as two components of plant defense against natural enemies, which involve different plant traits and genetic backgrounds, and with different effects on the fitness of both the plant and the enemy (Rausher 1996, 2001; Rosenthal & Kotanen 1994; Stowe et al. 2000; Strauss & Agrawal 1999). Both strategies are considered adaptive, and both are assumed to imply fitness costs to a plant genotype. Thus, all else being equal, the evolution and maintenance of each defensive strategy in plant populations should be affected by the relative fitness costs and benefits that each strategy involves. Seven years ago, Stowe et al. (2000) reviewed the topic of the evolution of plant tolerance to herbivores and raised many relevant questions about the joint evolution of both tolerance and resistance. Despite the many remaining unanswered questions, there have been advances derived from empirical and theoretical work. We here highlight these advances in the study of the evolution of plant defense against herbivores (and other natural enemies). Particularly, we review the models for the joint evolution of tolerance and resistance, the assumptions and predictions of these models, and empirical work related to this issue. We propose lines of research in this field of evolutionary ecology. A cornerstone of the theoretical and empirical development of this field was the proposal that both tolerance and resistance can function as alternative redundant strategies of defense (Simms & Triplett 1994). This hypothesis establishes that natural selection would not act to increase resistance if no fitness reduction is observed in the presence of herbivores because of tolerance. Thus, highly tolerant genotypes would not experience selection on resistance, and vice versa. The hypothesis predicts a negative genetic correlation between tolerance and resistance if they are redundant in terms of fitness and both imply similar costs. Fitness redundancy can occur when a similar reproductive output can be attained through either tolerance or resistance in the presence of herbivores (Valverde et al. 2003). Both the absence of redundancy and/or differences in the magnitude of the cost that tolerance and resistance involve may explain the absence of a negative genetic correlation between the two strategies of defense. Although two studies have demonstrated that costs of tolerance and resistance differ within populations (Fornoni et al. 2004b, Pilson 2000), we do not know 542 Núñez-Farfán · Fornoni · Valverde A nn u. R ev . E co l. E vo l. Sy st . 2 00 7. 38 :5 41 -5 66 . D ow nl oa de d fr om a rj ou rn al s. an nu al re vi ew s. or g by I ns tit ut o de E cl og ia , A .C , C on so rc io C O N A C yT o n 06 /0 5/ 09 . F or p er so na l u se o nl y. ANRV328-ES38-22 ARI 24 September 2007 8:7 much about the environmental conditions that determine the extent of redundancy between tolerance and resistance (Mauricio 2000, Valverde et al. 2003). In contrast, an increasing accumulation of empirical evidence shows that plants allocate resources simultaneously to both tolerance and resistance. This pattern was gathered from the observation that in a few instances, individuals (genotypes) with relatively high levels of resistance have low levels of tolerance, and vice versa (Leimu & Koricheva 2006a). Thus, the theoretical expectation of a negative genetic correlation between tolerance and resistance seems to have rather low empirical support (Fineblum & Rausher 1995, Fornoni et al. 2003a, Pilson 2000, Stowe 1998) and, at the same time, suggests that a pattern of simultaneous allocation to both tolerance and resistance should be common in natural populations. In this review we define a mixed pattern of defense allocation of an individual plant (genotype) as the simultaneous allocation of resources to component traits of resistance and tolerance to herbivory. THE SIMULTANEOUS EVOLUTION OF RESISTANCE AND TOLERANCE Theoretical studies predicted that a mixed pattern of defense allocation may or may not constitute an evolutionarily stable phenotype within populations (Abrahamson & Weis 1997, Fineblum & Rausher 1995, Fornoni et al. 2004a, Mauricio et al. 1997, Restiff & Koella 2004, Roy & Kirchner 2000, Tiffin 2000a). The hypotheses provided by these models can be grouped according to whether they propose that a mixed pattern of allocation to tolerance and resistance does or does not constitute an evolutionary stable equilibrium. Below, each hypothesis is explained and discussed in the context of the available evidence. Mixed Patterns of Defense Allocation as Evolutionary Stable Strategy Intermediate levels of tolerance and resistance are favored by natural selection. A mixed pattern of defense allocation could be selected if maximum fitness is attained at intermediate levels of both strategies, or if a combination of both mechanisms of defense pays higher fitness benefits than either of the strategies alone. This hypothesis predicts that stabilizing selection or positive correlational selection acts on tolerance and resistance to herbivory. Theoretical analyses indicate stabilizing selection may be expected when the costs or benefits of tolerance and resistance are nonlinear functions of allocation to defense (Fornoni et al. 2004a, Tiffin & Rausher 1999) or the relative magnitude of the costs of tolerance and resistance differ within populations (Fornoni et al. 2004a, Tiffin 2000a). In addition, correlational selection is expected when a combination of both tolerance and resistance pays higher fitness benefits than either strategy alone. That is, correlational selection is expected when the benefits of allocating resources simultaneously to tolerance and resistance are more than additive (Fornoni et al. 2004a). Several reviews demonstrated the generalized existence of costs of resistance characters (Bergelson & Purrington 1996, Bergelson et al. 2001, Purrington 2000, Strauss www.annualreviews.org • Resistance and Tolerance to Herbivores 543 A nn u. R ev . E co l. E vo l. Sy st . 2 00 7. 38 :5 41 -5 66 . D ow nl oa de d fr om a rj ou rn al s. an nu al re vi ew s. or g by I ns tit ut o de E cl og ia , A .C , C on so rc io C O N A C yT o n 06 /0 5/ 09 . F or p er so na l u se o nl y. ANRV328-ES38-22 ARI 24 September 2007 8:7 Table 1 Review of papers that estimated direct allocational costs of tolerance to herbivorya Plant species Source of damage Magnitude of