Evolutionary stability of mutualism: interspecific population regulation as an evolutionarily stable strategy

Abstract

Interspecific mutualisms are often vulnerable to instability because low benefit : cost ratios can rapidly lead to extinction or to the conversion of mutualism to parasite-host or predator-prey interactions. We hypothesize that the evolutionary stability of mutualism can depend on how benefits and costs to one mutualist vary with the population density of its partner, and that stability can be maintained if a mutualist can influence demographic rates and regulate the population density of its partner. We test this hypothesis in a model of mutualism with key features of senita cactus (Pachycereus schottii)-senita moth (Upiga virescens) interactions, in which benefits of pollination and costs of larval seed consumption to plant fitness depend on pollinator density. We show that plants can maximize their fitness by allocating resources to the production of excess flowers at the expense of fruit. Fruit abortion resulting from excess flower production reduces pre-adult survival of the pollinating seed-consumer, and maintains its density beneath a threshold that would destabilize the mutualism. Such a strategy of excess flower production and fruit abortion is convergent and evolutionarily stable against invasion by cheater plants that produce few flowers and abort few to no fruit. This novel mechanism of achieving evolutionarily stable mutualism, namely interspecific population regulation, is qualitatively different from other mechanisms invoking partner choice or selective rewards, and may be a general process that helps to preserve mutualistic interactions in nature.

Benefits and costs to pollinating, seed-eating insects: the effect of flower size and fruit abortion on larval performance

“…Indeed, larval growth rate is positively correlated with adult fecundity in several insect species (e.g., Awmack and Leather 2002;Honek 1993;Kause et al 1999). It would thus be very interesting in future studies to quantify how fruit abortion translates into reduced energetic costs and risk of secondary attack for the plant, and into lower moth survival and fecundity, and as a result lower abundance of the seed predator (Holland and DeAngelis 2002;Holland et al 2004b;Westerbergh and Westerbergh 2001).…”

“…These findings suggest that fruit abortion may also be an effective control mechanism in this non-obligate system. However, to be effective as a defense, abortion of infested fruits needs to reduce the costs of predation to the plants, and lower the profitability to the larvae, so as to ultimately limit survival and reproductive rates of the seed predator (Shapiro and Addicott 2004;Holland et al 2004b).…”

“…For the plant, the outcome of this interaction will depend on the balance of pollination benefits versus seed predation costs. Similarly, fitness of the seed predating and pollinating insect will likely be affected by the quantity and quality of food provided by the developing fruit to its larvae, but also crucially by the plant's ability to control damage (Bronstein 1992;Holland et al 2004b).…”

Plant Reproductive Allocation Predicts Herbivore Dynamics across Spatial and Temporal Scales

“…Thus, in addition to resource-based costs, plant reproductive effort may also incur herbivore-mediated ecological costs. This component of reproductive costs is typically neglected (Klinkhamer et al 1997;Holland et al 2004), and the generality of a positive relationship between plant reproduction and susceptibility to herbivores remains unknown (Obeso 2002). Given that herbivores can reduce plant fitness (Crawley 1983(Crawley , 1997Doak 1992;Louda and Potvin 1995;Bigger and Marvier 1998;Maron 1998;Ehrlen 2003), explicit consideration of the ecological costs of reproduction is needed to refine our understanding of optimal life-history strategies in plants.…”

“…For example, plant nutrient content (White 1984), water content (Huberty and Denno 2004), physical and chemical defenses (Rhoades 1983;Edelstein-Keshet and Rausher 1989;Larsson et al 2000), and combinations of the above (Rausher 1981;Louda and Collinge 1992;Agrawal 2004) are known to influence herbivore performance, abundance, and dynamics. In contrast, the influence of host plant resource allocation has not been intensively examined in this context (but see Holland et al 2004).…”