Read an Excerpt
Coevolution of Life on Hosts
Integrating Ecology and History
By Dale H. Clayton, Sarah E. Bush, Kevin P. Johnson, John N. Thompson The University of Chicago Press
Copyright © 2016 The University of Chicago
All rights reserved.
ISBN: 978-0-226-30230-0
CHAPTER 1
INTRODUCTION TO COEVOLUTION
Coevolution occurs on a very broad scale and may comprise most of evolution.
— Van Valen 1983
Life began with simple interactions between simple molecules. Over vast periods of time, these interactions increased in complexity, yielding macromolecules capable of transmitting heritable information. More intricate levels of biological organization followed, leading to the evolution of cells, tissues, and whole organisms. The main mechanism underlying these evolutionary changes involved random genetic mutations that coded for phenotypic changes, most of which had negative or neutral effects on fitness. However, a tiny fraction of these changes were adaptive, in that they improved the ability of organisms to deal with environmental pressures. Individuals with these adaptations left more offspring and thus greater genetic representation. In this way, the adaptive changes quickly became more common. This is the process Darwin (1859) called evolution by natural selection.
Populations of individuals evolve in heterogeneous natural environments that exert selection for a multitude of solutions to a maze of challenges. Environmental heterogeneity subdivides populations, further increasing diversity through restricted gene flow and random genetic drift. Over time, evolution by natural selection has generated a dizzying array of life forms of mind-boggling complexity and beauty. It is these species that make up the immense diversity of life on earth.
Species do not live in isolation; they live and interact with other species. The history of life has been profoundly influenced by these interactions. Interactions between species are so pervasive that they are a central feature of the environments of nearly all organisms. Many interactions are selective in nature, with one species influencing the fitness of other species. These selective effects can lead to evolutionary changes in the affected species. In some cases, the selective effects are reciprocal. When reciprocal selective effects lead to evolutionary changes in two or more interacting species, the process is known as coadaptation, or coevolution in the strict sense (box 1.1).
Darwin (1859) did not use the term coevolution, but he did refer to "coadaptations of organic beings to one another" (Thompson 1982, 1989). A century later, in a paper published in the journal Evolution, Mode (1958) used coevolution in reference to reciprocal microevolutionary changes between agricultural crops and fungal pathogens in ecological time. Soon thereafter, in another paper published in Evolution, Ehrlich and Raven (1964) used coevolution to describe reciprocal adaptive radiations between butterflies and their food plants (plate 1a). Although both papers were concerned with selective interactions between species, they took very different approaches, one focusing on microevolutionary time, and the other on macroevolutionary time.
The divide between micro- and macroevolutionary approaches to coevolution continues. Microevolutionary studies test one or more assumptions underlying the process of coadaptation, such as reciprocal selection, heritability of traits, or the extent of evolutionary responses to selection. By contrast, macroevolutionary studies compare the phylogenetic histories of interacting groups, such as parasitic lice and their hosts (plate 2), in order to understand patterns of codiversification. Microevolutionary studies thus tend to focus on evolutionary ecological processes, while macroevolutionary studies tend to focus on historical patterns. Regardless of the time scale, both approaches test the null hypothesis that coevolution has not taken place, until proven otherwise.
Coevolution has been pervasive in the history of life. Complex organisms depend on interspecific interactions to survive and reproduce (Thompson 2009). Eukaryotic organelles such as mitochondria and chloroplasts are descendants of ancient, independent lineages that evolved mutualistic relationships with primitive host cells, and ultimately became part of those cells. In short, coevolutionary dynamics have played a major role in generating the diverse catalog of ecological interactions that make up Darwin's (1859) "tangled bank." These relationships include a large variety of antagonistic interactions, like those between hosts and parasites, predators and prey, and competing species. Coevolved relationships also include a diversity of mutually beneficial interactions, like those between pollinators and flowering plants, dinoflagellates and corals, endosymbiotic bacteria and insects, and the fungi and algae that make up lichens.
Coadaptation
Despite this incredible diversity of coadapted systems, effects between species boil down to one of two possibilities: Species A can have a positive effect on Species B, or it can have a negative effect. The reciprocal also holds: Species B can have a positive effect on Species A, or it can have a negative effect. Interactions between species thus generate four broad outcomes: mutualism, parasitism, predation, or competition (fig 1.1). Mutualism is an interaction in which both parties benefit. The remaining three combinations are antagonistic interactions. In two of these cases, predation and parasitism, one party benefits, while the other is harmed. Predatory interactions are brief, lasting minutes or hours. Parasitic interactions are more durable, lasting days, weeks, or even years (Combes 2001). In the fourth type of interaction — competition — both parties are harmed. Interactions in which one species benefits, but the other is unaffected, are called commensalism. Examples of commensalism include bryophytes in the canopies of tropical trees, or remora fish dining on the leftovers of foraging sharks. Commensalistic interactions are not examples of coadaptation because they are unidirectional, by nature.
The relative fitness consequences of interspecific interactions can vary, depending on broader environmental contexts. For example, the effect of brood parasitic cuckoos on carrion crows fluctuates from parasitic to mutualistic, depending on the amount of crow nest predation by mammals and birds in a given year. When predators are abundant, the cost of rearing cuckoo nestlings is offset by the fact that cuckoo nestlings produce a repellent secretion that deters nest predators, thus increasing the fledging success of carrion crows (Canestrari et al. 2014). Hence, interspecific interactions are not static; they can transition between categories (Bronstein 1994).
It is important to bear in mind that, for fitness effects to generate adaptation, those effects must be selective. An effect of one species on another species is selective only if it generates a correlation between phenotypic variation and fitness. If the foreheads of hummingbird pollinators brush up mainly against stamens of flowers with deep corollas, then only flowers with deep corollas will get pollinated. In contrast, if hummingbirds brush up against flower stamens at random with respect to corolla depth, then even though these pollinators have a clear effect on plant fitness, they will not select for a change in corolla depth.
One reason why coadaptation is fascinating is that it has the power to explain remarkable examples of matching traits in nature. One of the most famous cases of matching traits is the Malagasy star orchid (Angraecum sesquipedale) and its pollinator, the Morgan hawkmoth (Xanthopan morgani) (fig 1.2). In his book on orchid pollination, Darwin (1862) wrote about this orchid and its 30-cm-long nectar spur. After experimentally removing pollinia from specimens of A. sesquipedale flowers in different ways, Darwin predicted that the pollinator had to be an undescribed hawkmoth with a proboscis "capable of extension to a length of between 10 and 11 inches [25 centimeters]!" Darwin's prediction was ridiculed by some of his colleagues (van der Cingel 2001; Rodríguez-Gironés and Santamaría 2007). However, 21 years after his death, Darwin was vindicated when Rothschild and Jordan (1903) described Xanthopan morgani praedicta, a Malagasy hawkmoth with a four-inch-long (10 cm) body and 12-inch-long (30 cm) proboscis (fig 1.2b). One cannot help but marvel at the scale, which is equivalent to a six-foot-tall man with an 18-foot-long tongue.
Matching traits need not be morphological; they can involve physiological, behavioral, and other traits. For example, parsnip webworm moth larvae (Depressaria pastinacella) feed exclusively on wild parsnips (Pastinaca sativa) over much of their range (plate 1b) (Zangerl and Berenbaum 2003). Wild parsnip produces furanocoumarins, which are highly toxic defensive compounds (Berenbaum 1991). The webworms detoxify these compounds using P450 proteins. The activity of the proteins is variable, and a target of selection by host plants (Berenbaum and Zangerl 1992). Across most populations, there is a close match between the mean furanocoumarin concentration in wild parsnip seeds and the rate of detoxification of furanocoumarins by webworms (Zangerl and Berenbaum 2003). This system is a classic example of chemical coadaptation.
Despite their intuitive appeal, it is important to keep in mind that matching traits, by themselves, are not evidence of coadaptation. Some matching traits have evolved due to historical interactions with entirely different species. Traits may have evolved in allopatric species that come into sympatry at a later date and happen to fit one another, a phenomenon known as ecological fitting (Janzen 1985a). In an important paper with a citation count that exceeds its word count, Janzen (1980) "call[ed] for more careful attention to the use of 'coevolution' as a word and concept." He argued that casual application of the concept makes it synonymous with the general notion of interspecific interactions.
A tongue in cheek (or snout in ear) example of matching traits is the close fit between the snouts of coati mundis (Nasua narica) and the ears of tapirs (Tapirus bairdii). Coatis service tapirs by removing ticks from their ears (fig 1.3). However, nobody in his right mind would argue that coati snouts and tapir ears are coadaptations. Indeed, the case in figure 1.3 turned out to be a short-lived interaction between a small number of tapirs and coatis looking for handouts near the dining hall on Barro Colorado Island, Panama, in the 1970s. Once visitors to the island stopped feeding the tapirs, the cleaning mutualism between coatis and tapirs disappeared (McClearn 1992).
In short, matching traits have not coevolved unless they have undergone coadaptation in response to reciprocal selection between the species possessing them. Note also that it is entirely possible for different categories of traits to coevolve — for example, a morphological feature in one species can coevolve with a physiological feature in another species. All that matters is that the trait in one species evolves in response to selection imposed by a second species. The evolutionary change then generates reciprocal selection and evolution in a trait of the second species. This criterion of reciprocity is widely accepted by coevolutionary biologists studying coadaptation (Dawkins and Krebs 1979; Kiester et al. 1984; Thompson 1989, 2005; Wade 2007; Nuismer et al. 2010). The key concept, and one that makes coevolution unique, is that coadaptation leads to the partial coordination of non-mixing gene pools over evolutionary time.
The phenotypic interface of coadaptation
Rigorous tests of coadaptation attempt to isolate and quantify reciprocal selective effects at the phenotypic interface (fig 1.4), which is influenced by a range of morphological, behavioral, physiological, and other performance traits. These traits can often be measured directly. The effect of one species on the performance of another species, and vice versa, governs the fitness of each species, which in turn governs how fitness correlates with phenotypic variation in each species. In other words, reciprocal effects on performance dictate reciprocal selection on corresponding traits between coadapting species. The performance component of the phenotypic interface, and the ability to quantify it, are essential ingredients in tests of coadaptation. No matter how perfectly traits seem to match, it is hard to know whether they are coevolving without knowing something about the reciprocal selective effects on those traits. A typical method of isolating and quantifying reciprocal selective effects is to experimentally manipulate a trait in one species, then quantify the fitness consequences in the other species (Clayton et al. 1999; Nash 2008).
Measuring selection is also important in cases where matching traits of interacting species covary across geographic regions. Anderson and Johnson (2008) documented striking correlations among different localities between flower corolla length (Zaluzianskya microsiphon) and proboscis length in a fly pollinator (Prosoeca ganglbaueri). Such patterns are intriguing, and may well represent different coadaptive end points. However, theoretical models suggest that geographic correlations can also arise due to selection on the members of one group to match a preexisting distribution of phenotypes in the other group (Nuismer et al. 2010). Geographic correlations can also be generated by independent bouts of stabilizing selection if the optimal phenotype for each species happens to correlate with some third abiotic or biotic factor.
Reciprocal selection cannot result in coadaptive evolution unless the phenotypic targets of selection have heritable genetic components. Quantitative genetic theory (Falconer 1981) tells us that the evolutionary response (R) of a trait is the product of the intensity of selection on that trait (s) and its narrow sense heritability (h2):
R = h2s
Strong selection can exhaust genetic variation, meaning that traits under selection may have low heritabilities (Mousseau and Roff 1987). In short, reciprocal selection on the traits of interacting species is a necessary but insufficient condition for predicting coadaptive evolutionary responses. One also needs information concerning the genetic basis of the traits.
Other important genetic factors include whether the genome is haploid or diploid, the number of alleles influencing a given trait, and the possible interactions between those alleles, such as dominance, pleiotropy, or epistasis (Woolhouse et al. 2002). Coevolutionary dynamics may further be influenced by 1) the type of reproduction (sexual or asexual), 2) whether mating is random or assortative, 3) whether generations are discreet or continuous, and 4) the relative generation times of interacting species. Random genetic drift can also play an important role in the coevolutionary dynamics of small populations (Kiester et al. 1984).
Yet another important consideration in predicting coadaptive outcomes is that the evolutionary response to selection depends on environmental context. Fitness cannot be described as a function of phenotypes alone; the "value" of the environment encountered by an individual must also be considered (Brodie and Ridenhour 2003). A genotype (G) with high fitness in one environment (E) may not have high fitness in another environment. Thus, genotype-by-environment (G×E) interactions have a fundamental influence on evolutionary responses to selection.
From a coevolutionary viewpoint, a key component of the environment for one species is the other species with which it interacts. This other species has an evolutionary trajectory of its own and, hence, it effectively has the ability to "evolve back." The evolutionary response to selection will be influenced by variation in the selective agent (the other species), as well as variation in the abiotic and other biotic components of the environment. Thus, the outcome of selection on a given species depends partly on the standing genetic variation in populations of the other species. As noted by Wade (2007), "The reciprocal co- evolution of one species in response to the genetic context of another is integral to the process of co-adaptation between species, and is believed to contribute to the functional integration of ecological communities and maintenance of biodiversity." Coadaptive dynamics are complicated, and intriguing, because they depend on genotype-by-genotype-by-environment (G×G×E) interactions (Piculell et al. 2008; Sadd 2011).
(Continues...)
Excerpted from Coevolution of Life on Hosts by Dale H. Clayton, Sarah E. Bush, Kevin P. Johnson, John N. Thompson. Copyright © 2016 The University of Chicago. Excerpted by permission of The University of Chicago Press.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
