Coevolution Of Consumers And Victims

We might expect predators and prey, herbivores and plants, and parasites and their hosts to evolve in an "arms race," whereby the victim evolves ever greater resistance, defense, or evasion, and the consumer evolves ever greater proficiency in finding and attacking the victim. However, the coevolu-tionary dynamics may be more complex than this, because of factors such as costs of adaptation and diffuse coevolution.

Considerable evidence supports the assumption that greater elaboration of a defensive or offensive feature imposes costs resulting from the character's interfering with another function or simply from the energy required for its development.

The population dynamics and the course of character evolution depend on many parameters and are often sensitive to starting conditions. An indefinitely extended arms race or escalation of the two species' characters is unlikely, because the cost of a sufficiently elaborated character eventually exceeds its benefit. Rather, the characters of both the prey and the predator may evolve to an intermediate stable state. Perhaps counterintuitively, species may become less proficient in attack or defense; for instance, a prey species may evolve a lower level of defense if it is so well defended that the predator becomes rare and thus becomes a weaker agent of selection than the energetic cost of defense. In some models, both the population densities and the character means of both species may change indefinitely, either in stable limit cycles or chaotically, and may even result in extinction.

When the consumer feeds on multiple species of victims, or a victim is attacked by multiple consumer species, diffuse coevolution may affect the outcome. For example, if there exists a negative genetic correlation between a host's resistance to different species of parasites, then resistance to each carries a "cost," selection will vary in time and space, depending on the relative abundance of the two parasites, and resistance to each parasite is constrained. Diffuse coevolution can be very difficult to document and might often be sluggish. Because prey species have characteristics (e.g., cryptic coloration, dis-tastefulness, speed of escape) that provide protection against many species of predators, and predators likewise have characteristics that enable them to capture and handle many prey species, changes in the relative abundance of different predators (or prey) may not greatly alter selection. During the "Mesozoic marine revolution," lineages of crustaceans and fishes capable of crushing hard shells evolved, and many groups of molluscs evolved features (e.g., thicker shells, spines) that made them more difficult to crush. Surely these changes reflect diffuse coevolution, but our inability to ascribe changes in any one species to changes in any other one species makes it hard to discern a coevolutionary process.

Predators and Prey

Geographic variation in the identity and strength of interactions among species provides some of the best evidence of coevolution. For example, the shape of the cones of lodgepole pine (Pinus contorta) differs among populations, depending on whether its major seed predator, the red squirrel (Tamiasciurus hudsonicus), is present or absent. In mountain ranges without red squirrels, red crossbills (Loxia curvirostra complex) are abundant seed predators. In these areas, the pine has evolved modifications of the cone that reduce seed extraction by this species of bird, and the shape and size of the crossbill's bill have evolved to enhance seed extraction.

Such evidence of coevolution, however, is rare, compared with evidence of unilateral adaptation. For example, Mediterranean populations of the braconid parasitoid Asobara tabida have higher "virulence" (capacity to survive host defenses) than northern European populations. Although one of its hosts, Drosophila melanogaster, shows a somewhat parallel geographic pattern in defense, the cline in Asobara appears to be most parsimoniously explained not by coevolution, but by the fact that D. melanogaster, its major host in the south, has stronger defenses than the major northern host, D. subobscura.

Parasite—Host Interactions

The evolution of interactions between hosts and parasites (including pathogenic microorganisms) can differ from predator—prey interactions in several respects. Whereas improvement in a predator or prey trait (such as size or fleetness) is likely to enhance fitness regardless of the specific genotype in the opponent species, parasite—host interactions are more likely to be affected by "gene-for-gene" interactions, in which each allele for host resistance is matched by a parasite "virulence allele" that enables the parasite to overcome resistance. Such gene-for-gene relationships have been described for several plant—fungus interactions and for the relationship between the Hessian fly (Cecidomyiidae: Mayetiola destructor) and resistant genotypes of wheat. Selection in gene-for-gene systems may be frequency-dependent: as a parasite allele that matches the most common host allele increases in frequency, rare host alleles acquire a selective advantage by conferring resistance against most of the parasites and so increase in frequency and initiate selection for a currently rare virulence allele. The genetic composition of local populations is likely to differ at any one time, because these oscillatory genetic dynamics may be out of phase unless the populations are connected by high gene flow. Geographic variation in genetic composition has been reported for trematodes and snails, trematodes and fish, microsporidians and Daphnia, and fungal parasites and plants. In most of these parasite—host pairs, populations of the parasite are best adapted to their local host populations, suggesting that the parasites adapt faster than their hosts.

The fitness of a parasite genotype may be approximately measured by the number of potential hosts it infects, compared with other genotypes. Often, the rate of transmission to new hosts is proportional to the parasite's reproductive rate, which in turn often (though not always) determines the parasite's virulence to the host. For example, the probability that progeny of a virus are transmitted by a mosquito is a function of the density of viral particles in the host's blood. However, the probability of transmission is reduced if the host dies too soon, i.e., if the parasites die before transmission. Such conflicting factors result in an evolutionary equilibrium level of virulence that is determined by several factors, especially the mode of transmission. If transmission is "vertical," i.e., only to the offspring of infected individuals, then parasite fitness is proportional to the number of surviving host offspring, and selection favors benign, relatively avirulent parasite genotypes. If transmission is "horizontal," i.e., among hosts of the same generation, the equilibrium level of virulence is likely to be higher, because (a) an individual parasite's fitness does not depend on successful reproduction of its individual host and (b) the likelihood is higher that an individual host will be infected by multiple parasite genotypes that compete for transmission to new hosts. As predicted by this theory, among species of nematodes that parasitize fig wasps (Agaonidae), those that are mostly horizontally transmitted cause a greater reduction of their hosts' fitness than those that are vertically transmitted.

Herbivores and Plants

Most of the many thousands of species of herbivorous insects are fairly or highly specialized, feeding on closely related species—sometimes just a single species—of plants. At a proximal level, this specificity is largely the result of behavioral responses to plant features, especially the many "secondary chemicals" that distinguish plant taxa. Insects often react to compounds in nonhost plants as deterrents to oviposition or feeding and to certain compounds in host plants as stimulants. Phytochemicals may not only deter feeding but also reduce insect fitness by acting as toxins or interfering with digestion. It is generally thought that chemical and other differences among plants select for host-specificity in insects, on the supposition that physiological costs impose trade-offs among adaptations to different plant characteristics; however, only a minority of genetic and physiological studies has supported this hypothesis. Other proposed advantages of host specificity include use of specific plants as rendezvous sites for mating, greater efficiency of finding hosts, and predator escape by several means, such as sequestering defensive plant compounds.

Phylogenetic studies show that associations between some insect clades and plant clades are very old, often dating to the

FIGURE 1 (a) Phylogenies, based on DNA sequence data, of leaf beetles in the genus Ophraella (left) and their host plants (right). Arrows join each beetle species to its host plant. Different shading patterns represent the four tribes of Asteraceae into which the host plants fall; the shading of branches is a parsimonious inference of the tribes with which ancestral Ophraella lineages were associated. Note that most host shifts associated with beetle speciation have been between plants in the same tribe. The incongruence between the phylogeny of the insects and that of their host plants is one of several indications that the beetles and plants did not cospeciate. These plant lineages represent only a few of the tribes of Asteraceae and of the genera within each tribe. [After D. J. Funk et al. (1995). Evolution 49, 1008-1017. The Society for the Study of Evolution.] (b) The leaf beetle O. sexvittata, which feeds on Solidago species, tribe Astereae. (Original illustration by author.)

FIGURE 1 (a) Phylogenies, based on DNA sequence data, of leaf beetles in the genus Ophraella (left) and their host plants (right). Arrows join each beetle species to its host plant. Different shading patterns represent the four tribes of Asteraceae into which the host plants fall; the shading of branches is a parsimonious inference of the tribes with which ancestral Ophraella lineages were associated. Note that most host shifts associated with beetle speciation have been between plants in the same tribe. The incongruence between the phylogeny of the insects and that of their host plants is one of several indications that the beetles and plants did not cospeciate. These plant lineages represent only a few of the tribes of Asteraceae and of the genera within each tribe. [After D. J. Funk et al. (1995). Evolution 49, 1008-1017. The Society for the Study of Evolution.] (b) The leaf beetle O. sexvittata, which feeds on Solidago species, tribe Astereae. (Original illustration by author.)

early Tertiary and in some cases to the Cretaceous or even Jurassic. Nevertheless, only a few instances of cospeciation and phylogenetic concordance have been described. In most cases, much of the speciation within an insect clade has occurred after the host plants diversified, but new species have shifted to plant species closely related to the ancestral host (Fig. 1). That these host shifts have been facilitated by chemical similarity of related plants is supported by instances in which phylogenetic relationships among insect species (e.g., Blepharida flea beetles, melitaeine butterflies) more closely match the hosts' chemical similarities than phylogenetic relationships. Patterns of genetic variation in the ability of host-specific Ophraella leaf beetles to feed and develop on nonhost plants, all within the Asteraceae, indicated greater genetic potential to adapt to those plants that were most closely related to the insect's normal host.

Although physiological, morphological, behavioral, and phenological adaptations of insects to host plants are many and obvious, demonstrating that plant characters have evolved because of selection for their defensive functions has been more difficult. Many chemical and morphological features of plants have the effect of reducing attack or damage by some or many insect species, but some authors have argued that they actually evolved for physiological reasons or as defenses against mammalian herbivores rather than insects. However, both phytochemicals (e.g., furanocoumarins) and morphological features (e.g., trichomes) have been shown to determine fitness differences among genotypes due to their effect on insect herbivores, and the distribution of many plant compounds among tissues conforms to what we should expect if they were adaptively deployed defenses. Still, there have been few demonstrations of adaptive geographic variation in plant defenses in relation to the abundance or identity of particular herbivorous insects. In one of the few examples of probable coevolution at the population level, populations of wild parsnip (Pastinaca sativa) have diverged in their profile of toxic furanocoumarins, and parsnip webworms (Depressaria pastinacella) are adapted to their local host population.

Ehrlich and Raven's escape-and-radiate model of coevolution between plants and herbivorous insects has found some support. Most lineages of plants that have independently evolved latex or resin canals (potent deterrents to most insects) are richer in species than their canalless sister groups, supporting the hypothesis that new plant defenses enhance the rate of diversification. Likewise, herbivorous clades of insects are generally more diverse than their nonherbivorous sister groups. Clades of phytophagous beetles that are thought to be primitively associated with gymnosperms have fewer species than sister taxa that have shifted to angiosperms, perhaps because the latter are so very diverse. The diversity of several moth taxa that feed on Apiaceae with presumably "advanced" chemical defenses is greater than that of those that feed on Apiaceae with "primitive" defenses, paralleling the differences in plant diversity, but phylogenetic analysis is needed to confirm that the diversification rate has been enhanced by novel plant defenses and insect counteradaptations.

0 0

Post a comment