Chromosomal rearrangements can promote reproductive isolation by reducing recombination along a large section of the genome. We model the effects of the genetic barrier to gene flow caused by a chromosomal rearrangement on the rate of accumulation of postzygotic isolation genes in parapatry. We find that, if reproductive isolation is produced by the accumulation in parapatry of sets of alleles compatible within but incompatible across species, chromosomal rearrangements are far more likely to favor it than classical genetic barriers without chromosomal changes. New evidence of the role of chromosomal rearrangements in parapatric speciation suggests that postzygotic isolation is often due to the accumulation of such incompatibilities. The model makes testable qualitative predictions about the genetic signature of speciation.

Variations in the origination and extinction rates of species over geological time often are linked with a range of factors, including the evolution of key innovations, changes in ecosystem structure, and environmental factors such as shifts in climate and physical geography. Before hypothesizing causality of a single factor, it is critical to demonstrate that the observed variation in diversification is significantly greater than one would expect due to natural stochasticity in the evolutionary branching process. Here, we use a likelihood-ratio test to compare taxonomic rate heterogeneity to a neutral birth-death model, using data on well-supported sister pairs of taxa and their species richness. We test the likelihood that the distribution of extant species among angiosperm genera and families could be the result of constant diversification rates. Results strongly support the conclusion that there is significantly more heterogeneity in diversity at the species level within angiosperms than would be expected due to stochastic processes. This result is consistent in datasets of genus pairs and family pairs and is not affected significantly by degrading pairs to simulate inaccuracy in the assumption of simultaneous origin of sister taxa. When we parse taxon pairs among higher groups of angiosperms, results indicate that a constant rates model is not rejected by rosid and basal eudicot pairs but is rejected by asterid and eumagnoliid pairs. These results provide strong support for the hypothesis that species-level rates of origination and/or extinction have varied nonrandomly within angiosperms and that the magnitude of heterogeneity varies among major groups within angiosperms.

Plant traits that show little variation across higher taxa are often used as diagnostic traits, but the reason for the stasis of such traits remains unclear. Wild radish, Raphanus raphanistrum, exhibits tetradynamous stamens (four long and two short, producing a dimorphism in anther height within each flower), as do the vast majority of the more than 3000 species in the Brassicaceae. Here we examine the hypothesis that selection maintains the stasis of dimorphic anther height by investigating the effects of this trait on pollen removal, seed siring success, and seed set in R. raphanistrum using both experimental and observational methods. Observational selection gradient analysis based on lifetime seed siring success provided evidence for an optimum dimorphism that was greater than zero in one of three years. In both experimentally manipulated and unmanipulated flowers, more pollen was removed in single visits from flowers with less dimorphism. There was no significant effect of anther dimorphism on female fitness (seed set). Therefore, there is some evidence to suggest that selection is maintaining anther dimorphism in wild radish, and that higher male fitness might result from restriction of single-visit pollen removal. We discuss these results in light of pollen presentation theory.

Measurements of the genetic variation and covariation underlying quantitative traits are crucial to our understanding of current evolutionary change and the mechanisms causing this evolution. This fact has spurred a large number of studies estimating heritabilities and genetic correlations in a variety of organisms. Most of these studies have been done in laboratory or greenhouse settings, but it is not well known how accurately these measurements estimate genetic variance and covariance expressed in the field. We conducted a quantitative genetic half-sibling analysis on six floral traits in wild radish. Plants were grown from seed in the field and were exposed to natural environmental variation throughout their lives, including herbivory and intra- and interspecific competition. The estimates of heritabilities and the additive genetic variance-covariance matrix (G) obtained from this analysis were then compared to previous greenhouse estimates of the same floral traits from the same natural population. Heritabilities were much lower in the field for all traits, and this was due to both large increases in environmental variance and decreases in additive genetic variance. Additive genetic covariance expressed was also much lower in the field. These differences resulted in highly significant differences in the G matrix between the greenhouse and field environments using two complementary testing methods. Although the G matrices shared some principal components in common, they were not simply proportional to each other. Therefore, the greenhouse results did not accurately depict how the floral traits would respond to natural selection in the field.

Habitat degradation and loss can result in population decline and genetic erosion, limiting the ability of organisms to cope with environmental change, whether this is through evolutionary genetic response (requiring genetic variation) or through phenotypic plasticity (i.e., the ability of a given genotype to express a variable phenotype across environments). Here we address the question whether plants from small populations are less plastic or more susceptible to environmental stress than plants from large populations. We collected seed families from small (<100) versus large natural populations (>1000 flowering plants) of the rare, endemic plant Cochlearia bavarica (Brassicaceae). We exposed the seedlings to a range of environments, created by manipulating water supply and light intensity in a 2 × 2 factorial design in the greenhouse. We monitored plant growth and survival for 300 days. Significant effects of offspring environment on offspring characters demonstrated that there is phenotypic plasticity in the responses to environmental stress in this species. Significant effects of population size group, but mainly of population identity within the population size groups, and of maternal plant identity within populations indicated variation due to genetic (plus potentially maternal) variation for offspring traits. The environment × maternal plant identity interaction was rarely significant, providing little evidence for genetically- (plus potentially maternally-) based variation in plasticity within populations. However, significant environment × population-size-group and environment × population-identity interactions suggested that populations differed in the amount of plasticity, the mean amount being smaller in small populations than in large populations. Whereas on day 210 the differences between small and large populations were largest in the environment in which plants grew biggest (i.e., under benign conditions), on day 270 the difference was largest in stressful environments. These results show that population size and population identity can affect growth and survival differently across environmental stress gradients. Moreover, these effects can themselves be modified by time-dependent variation in the interaction between plants and their environment.

This study partitions selection in a natural metapopulation of a riparian plant species, Silene tatarica, into individual- and patch-level components by using contextual analysis, in which a patch refers to a spatially distinct stand of individual plants. We estimated selection gradients for two morphological characters (plant height and number of stems), their respective patch means, and plant density with respect to reproductive success in a two-year study. The approach was also extended to partition selection separately within habitats with varying degrees of exposure to river disturbances and herbivory. The selection differentials and gradients for plant height were positive at both individual and patch levels, with selection forces highest in the closed habitat with low exposure to disturbance. This pattern suggests that local groups with taller than average plants are more visible to pollinators than to groups that are shorter than average plants; and, within patches, individuals with short stature are visited less often than taller ones. Selection on the number of stems was in opposition at individual and patch levels. At the individual level the character was selected toward higher values, whereas selection at the patch-level favored smaller mean number of stems. The strength of the latter component was associated with the intensity of herbivory in different habitats, suggesting that the patch-level selection against a large number of stems might be due to high attractiveness of such patches to the main herbivore, reindeer. Consequently, direction and strength of selection in spatially structured populations may depend significantly on fitness effects arising at the group level.

Given the enormous number and high mortality of fertilized ovules in plants, it is possible that selection during the earliest stages of the life cycle plays an important role in shaping the genetic composition of plant populations. Previous research involving selection component analyses found strong evidence for viability selection in annual plant species. Yet despite this evidence, few attempts have been made to identify the magnitude and timing of viability selection as well as the mechanisms responsible for mortality among genotypes. Platypodium elegans, a Neotropical tree with high rates of early fruit mortality, represents an opportunity to study viability selection at a level of discernment not previously possible. Microsatellite markers were used to analyze the genetic composition of aborted embryos, as well as mature seeds and seedlings of the same cohort. While selection resulted in an overall decrease in self-fertilized progeny across each life stage, the greatest change in the genetic composition of progeny occurred between mature seeds and established seedlings. This suggested that inbreeding depression, and not late-acting self-incompatibility, was responsible for early selection. An investigation of the mature seed stage revealed that self-fertilized seeds weigh significantly less than outcrossed seeds. The result of this early selection conceals the mixed-mating system and high levels of inbreeding depression in Platypodium elegans, resulting in an apparently outcrossed adult population that does not differ significantly from Hardy-Weinberg expectations.

We provide evidence from comparisons of populations of Drosophila that evolutionary correlations between longevity and stress resistance break down over the course of laboratory evolution. Using 15 distinct evolutionary regimes, we created 75 populations that were differentiated for early fecundity, longevity, starvation resistance, desiccation resistance, and developmental time. In earlier experiments, selection for postponed aging produced increases in stress resistance, whereas selection for increased stress resistance produced increases in longevity. Direct estimates of correlations also indicated an antagonistic relationship between early fecundity on one hand and longevity or stress resistance on the other. Laboratory evolution of extreme values of stress resistance, however, led to a breakdown in these evolutionary relationships. There was no evidence that these significant changes in correlation resulted from genotype-by-environment interactions or inbreeding. These findings suggest that correlations between functional characters are not necessarily durable features of a species, and that short-term evolutionary responses cannot be extrapolated reliably to longer-term evolutionary patterns.

We trace the evolutionary correlation between stress resistance and longevity in populations of Drosophila melanogaster selected for stress resistance over many generations. Females selected for desiccation resistance and both females and males selected for increasing starvation resistance initially show concurrent increases in longevity, but then begin to decrease in longevity, even as stress resistance continues to increase. We demonstrate that the correlation between two fitness traits can change and that this change is due to sustained selection rather than a genotype-by-environment interaction or inbreeding depression. The breakdown in evolutionary correlation we report underscores the difficulty of extrapolating the results from short-term selection experiments to predictions of long-term evolution.

Surprisingly little is known about what determines a parasite's host range, which is essential in enabling us to predict the fate of novel infections. In this study, we evaluate the importance of both host and parasite phylogeny in determining the ability of parasites to infect novel host species. Using experimental lab assays, we infected 24 taxonomically diverse species of Drosophila flies (Diptera: Drosophilidae) with five different nematode species (Tylenchida: Allantonematidae: Howardula, Parasitylenchus), and measured parasite infection success, growth, and effects on female host fecundity (i.e., virulence). These nematodes are obligate parasites of mushroom-feeding Drosophila, particularly quinaria and testacea group species, often with severe fitness consequences on their hosts. We show that the potential host ranges of the nematodes are much larger than their actual ranges, even for parasites with only one known host species in nature. Novel hosts that are distantly related from the native host are much less likely to be infected, but among more closely related hosts, there is much variation in susceptibility. Potential host ranges differ greatly between the related parasite species. All nematode species that successfully infected novel hosts produced infective juveniles in these hosts. Most novel infections did not result in significant reductions in the fecundity of female hosts, with one exception: the host specialist Parasitylenchus nearcticus sterilized all quinaria group hosts, only one of which is a host in nature. The large potential host ranges of these parasites, in combination with the high potential for host colonization due to shared mushroom breeding sites, explain the widespread host switching observed in comparisons of nematode and Drosophila phylogenies.

Fecundity is usually considered as a trait closely connected to fitness and is expected to exhibit substantial nonadditive genetic variation and inbreeding depression. However, two independent experiments, using populations of different geographical origin, indicate that early fecundity in Drosophila melanogaster behaves as a typical additive trait of low heritability. The first experiment involved artificial selection in inbred and non-inbred lines, all of them started from a common base population previously maintained in the laboratory for about 35 generations. The realized heritability estimate was 0.151 ± 0.075 and the inbreeding depression was very small and nonsignificant (0.09 ± 0.09% of the non-inbred mean per 1% increase in inbreeding coefficient). With inbreeding, the observed decrease in the within-line additive genetic variance and the corresponding increase of the between-line variance were very close to their expected values for pure additive gene action. This result is at odds with previous studies showing inbreeding depression and, therefore, directional dominance for the same trait and species. All experiments, however, used laboratory populations, and it is possible that the original genetic architecture of the trait in nature was subsequently altered by the joint action of random drift and adaptation to captivity. Thus, we carried out a second experiment, involving inbreeding without artificial selection in a population recently collected from the wild. In this case we obtained, again, a maximum-likelihood heritability estimate of 0.210 ± 0.027 and very little nonsignificant inbreeding depression (0.06 ± 0.12%). The results suggest that, for fitness-component traits, low levels of additive genetic variance are not necessarily associated with large inbreeding depression or high levels of nonadditive genetic variance.

Latitudinal genetic clines in body size occur in many ectotherms including Drosophila species. In the wing of D. melanogaster, these clines are generally based on latitudinal variation in cell number. In contrast, differences in wing area that evolve by thermal selection in the laboratory are in general based on cell size. To investigate possible reasons for the different cellular bases of these two types of evolutionary response, we compared the newly established North and South American wing size clines of Drosophila subobscura. The new clines are based on latitudinal variation in cell area in North America and cell number in South America. The ancestral European cline is also based on latitudinal variation in cell number. The difference in the cellular basis of wing size variation in the American clines, which are roughly the same age, together with the similar cellular basis of the new South American cline and the ancient European one, suggest that the antiquity of a cline does not explain its cellular basis. Furthermore, the results indicate that wing size as a whole, rather than its cellular basis, is under selection. The different cellular bases of different size clines are most likely explained either entirely by chance or by different patterns of genetic variance—or its expression—in founding populations.

In a large behavioral experiment we reconstructed the evolution of behavioral responses to predators to explore how interactions with predators have shaped the evolution of their prey's behavior. All Enallagma damselfly species reduced both movement and feeding in the presence of coexisting predators. Some Enallagma species inhabit water bodies with both fish and dragonflies, and these species responded to the presence of both predators, whereas other Enallagma species inhabit water bodies that have only large dragonflies as predators, and these species only responded to the presence of dragonflies. Lineages that shifted to live with large dragonflies showed no evolution in behaviors expressed in the presence of dragonflies, but they evolved greater movement in the absence of predators and greater movement and feeding in the presence of fish. These results suggest that Enallagma species have evolutionarily lost the ability to recognize fish as a predator. Because species coexisting with only dragonfly predators have also evolved the ability to escape attacking dragonfly predators by swimming, the decreased predation risk associated with foraging appears to have shifted the balance of the foraging/predation risk trade-off to allow increased activity in the absence of mortality threats to evolve in these lineages. Our results suggest that evolution in response to changes in predation regime may have greater consequences for characters expressed in the absence of mortality threats because of how the balance between the conflicting demands of growth and predation risk are altered.

Although a considerable amount of information is available on the ecology, genetics, and physiology of life-history traits, much more limited data are available on the biochemical and genetic correlates of life-history variation within species. Specific activities of five enzymes of lipid biosynthesis and two enzymes of amino acid catabolism were compared among lines selected for flight-capable (LW[f]) versus flightless (SW) morphs of the cricket Gryllus firmus. These morphs, which exist in natural populations, differ genetically in ovarian growth (100–400% higher in SW) and aspects of flight capability including the size of wings and flight muscles, and the concentration of triglyceride flight fuel (40% greater in LW[f]). Consistently higher activity of each enzyme in LW(f) versus SW-selected lines, and strong co-segregation between morph and enzyme activity, demonstrated genetically based covariance between wing morph and enzyme activity. Developmental profiles of enzyme activities strongly paralleled profiles of triglyceride accumulation during adulthood and previous measures of in vivo lipid biosynthesis. These data strongly imply that genetically based elevation in activities of lipogenic enzymes, and enzymes controlling the conversion of amino acids into lipids, is an important cause underlying the elevated accumulation of triglyceride in the LW(f) morph, a key biochemical component of the trade-off between elevated early fecundity and flight capability. Global changes in lipid and amino-acid metabolism appear to have resulted from microevolutionary alteration of regulators of metabolism. Finally, strong genotype × environment (diet) interactions were observed for most enzyme activities. Future progress in understanding the functional causes of life-history evolution requires a more detailed synthesis of the fields of life-history evolution and metabolic biochemistry. Wing polymorphism is a powerful experimental model in such integrative studies.

Inbreeding, the mating of close relatives, is known to have deleterious effects on fitness traits in organisms. Developmental stability (DS) and canalization may represent two processes that allow an organism to maintain a stable development that will produce the fittest phenotype. Inbreeding is thus expected to affect either DS or canalization. We tested if inbreeding affects DS and canalization using an inbreeding experiment on the cricket Gryllus firmus. We compared mean length, fluctuating asymmetry (as an index of DS), and morphological variation (as an index of canalization) of four limb traits between seven highly inbred lines, their F₁ crosses, and outbred lines originated from the same stock population and maintained in the same environmental conditions. We show evidence for moderate inbreeding depression on the four measures of leg length. The nonsystematic difference in fluctuating asymmetry indices between breed types indicates that inbreeding or heterozygosity did not affect DS, or that fluctuating asymmetry is not a reliable index of DS. In contrast, inbreeding appears to affect canalization, as shown by the significantly higher variation in inbred lines compared to other lines. Identical low variation values in the crossbred and outbred lines indicate that heterozygosity could affect canalization. High variation in morphological variation and fluctuating asymmetry within crossbred or inbred lines, however, suggest the effect of recessive deleterious alleles on both canalization and DS. Although the strong correlation in morphological variation among traits suggests that identical genetic mechanisms govern canalization for all the limb traits, the absence of significant correlation in fluctuating asymmetry among traits causes us to reject this hypothesis for DS. For most of the traits, morphological variation and fluctuating asymmetry were not significantly correlated, which support the hypothesis that canalization and DS consist in two distinct mechanisms.

We identify and describe the distribution of 12 genetically distinct malaria parasite lineages over islands and hosts in four common passerine birds in the Lesser Antilles. Combined parasite prevalence demonstrates strong host effects, little or no island effect, and a significant host-times-island interaction, indicating independent outcomes of host-parasite infections among island populations of the same host species. Host- and/or island-specific parasite lineages do not explain these host-parasite associations; rather, individual lineages themselves demonstrate the same type of independent interactions. Unlike overall prevalence, individual parasite lineages show considerable geographic structure (i.e., island effects) as well as species effects indicating that parasite lineages are constrained in their ability to move between hosts and locations. Together, our results suggest an upper limit to the number of host individuals that malaria parasites, as a community, can infect. Within this limit, however, the relative frequency of the different lineages varies reflecting fine scale interactions between host and parasite populations. Patterns of host-parasite associations within this system suggest both historical co-evolution and ecologically dynamic and independent host-parasite interactions.

Territorial song structures are often the most prominent characters for distinguishing closely related taxa among songbirds. Learning processes may cause convergent evolution of passerine songs, but phylogenetic information of acoustic traits can be investigated with the help of molecular phylogenies, which are not affected by cultural evolutionary processes. We used a phylogeny based on cytochrome b sequences to trace the evolution of territorial song within the genus Regulus. Five discrete song units are defined as basic components of regulid song via sonagraphic measurements. Traits of each unit are traced on a molecular tree and a mean acoustic character difference between taxon pairs is calculated. Acoustic divergence between regulid taxa correlates strongly with genetic distances. Syntax features of complete songs and of single units are most consistent with the molecular data, whereas the abundance of certain element types is not. Whether song characters are innate or learned was interpreted using hand-reared birds in aviary experiments. We found that convergent character evolution seems to be most probable for learned acoustic traits. We conclude that syntax traits of whole verses or subunits of territorial song, especially innate song structures, are the most reliable acoustic traits for phylogenetic reconstructions in Regulus.

The contributions of genetic and environmental factors to differential reproductive success across hybrid zones have rarely been tested. Here, we report a manipulative experiment that simultaneously tested endogenous (genetic-based) and exogenous (environmental-based) selection within a hybrid zone. We transplanted mated pairs of two chickadee species (Poecile atricapilla and P. carolinensis) and their hybrids into isolated woodlots within their hybrid zone and monitored their reproductive success. Although clutch sizes were similar, based on an estimate of the genetic compatibility of a pair, hybrid pairs produced fewer nestlings and fledglings than did pairs of either parental species. According to a linear model generated from the data, a pure pair of either parental species would be expected to produce 1.91–2.48 times more fledglings per nesting attempt, respectively, than the average or least compatible pair in the experiment. Our result of decreased reproduction for hybrid pairs relative to parental species pairs within same environment (the hybrid zone in this experiment) support the endogenous selection hypothesis for maintenance of this hybrid zone. Because the experiment was conducted entirely within the hybrid zone (i.e., the same environment for parental and hybrid pairings), our data do not support the exogenous selection hypothesis as it predicts either all pairings doing poorly or the hybrid pairs more successful than the parental pairs.

We combined mark-and-recapture studies with genetic techniques of parentage assignment to evaluate the interactions between mating, dispersal, and inbreeding, in a free-ranging population of Crocidura russula. We found a pattern of limited and female-biased dispersal, followed by random mating within individual neighborhoods. This results in significant inbreeding at the population level: mating among relatives occurs more often than random, and F_IT analyses reveal significant deficits in heterozygotes. However, related mating partners were not less fecund, and inbred offspring had no lower lifetime reproductive output. Power analyses show these negative results to be quite robust. Absence of phenotypic evidence of inbreeding depression might result from a history of purging: local populations are small and undergo disequilibrium gene dynamics. Dispersal is likely caused by local saturation and (re)colonization of empty breeding sites, rather than inbreeding avoidance.

The evolutionary importance of postnatal ontogenies has long been recognized, but most studies of ontogenetic trajectories have focused exclusively on morphological traits. For animals, this represents a major omission because behavioral traits and their ontogenies often have relatively direct relationships to fitness. Here four replicate lines of house mice artificially selected for high early-age wheel running and their four replicate control lines were used to evaluate the effects of early-age directional selection, genetic drift, and activity environment (presence or absence of a running wheel) on variation in the ontogenies of three traits known to be genetically correlated: voluntary wheel running, body mass, and food consumption. Early-age selection significantly changed both the shape and position of the wheel-running and food-consumption ontogenies while influencing the position, but not the shape, of the body mass ontogeny. Genetic drift (as indicated by variation among replicate lines) produced significant changes in both the position and shape of all three ontogenies; however, its effect differed between the selection and control groups. For wheel running and food consumption, genetic drift only influenced the control ontogenies, whereas for body mass, genetic drift had a significant effect in both selection groups. Both body-mass and food-consumption ontogenies were significantly altered by activity environment, with the environment causing significant changes in the shape and position of both ontogenies. Overall the results demonstrate strong effects of early-age selection, genetic drift, and environmental variation on the evolution and expression of behavioral and morphological ontogenies, with selection changing only the position of the morphological ontogeny but both the position and shape of the behavioral ontogenies.

The glacial-interglacial cycles of the upper Pleistocene have had a major impact on the recent evolutionary history of Arctic species. To assess the effects of these large-scale climatic fluctuations to a large, migratory Arctic mammal, we assessed the phylogeography of reindeer (Rangifer tarandus) as inferred from mitochondrial DNA (mtDNA) sequence variation in the control region. Phylogenetic relationships among haplotypes seem to reflect historical patterns of fragmentation and colonization rather than clear-cut relationships among extant populations and subspecies. Three major haplogroups were detected, presumably representing three separate populations during the last glacial. The most influential one has contributed to the gene pool of all extant subspecies and seems to represent a large and continuous glacial population extending from Beringia and far into Eurasia. A smaller, more localized refugium was most likely isolated in connection with ice expansion in western Eurasia. A third glacial refugium was presumably located south of the ice sheet in North America, possibly comprising several separate refugial populations. Significant demographic population expansion was detected for the two haplogroups representing the western Eurasian and Beringian glacial populations. The former apparently expanded when the ice cap retreated by the end of the last glacial. The large continuous one, in contrast, seems to have expanded by the end of the last interglacial, indicating that the warm interglacial climate accompanied by marine transgression and forest expansion significantly confined population size on the continental mainland. Our data demonstrate that the current subspecies designation does not reflect the mtDNA phylogeography of the species, which in turn may indicate that morphological differences among subspecies have evolved as adaptive responses to postglacial environmental change.

In this article we explore how host survival and fecundity are affected by host-parasite coevolution. We examine a situation in which hosts upon being infected can mount a defensive response to clear the infection, but in which there is a fecundity cost to such immunological up-regulation. We also suppose that the parasite exploits the host and thereby causes an elevated host mortality rate. We determine the coevolutionary stable strategies of the parasite's level of exploitation and the host's level of up-regulation, and illustrate the patterns of reduced host fitness (i.e., virulence) that these produce. We find that counterintuitive patterns of virulence are often expected to arise as a result of the interaction between coevolved host and parasite strategies. In particular, despite the fact that the parasite imposes only a mortality cost on the host, coevolution by the host results in a pattern whereby infected hosts always have the same probability of death from infection, but they vary in the extent to which their fecundity is reduced. This contrasts with previous results and arises from our inclusion of two important factors absent from previous theory: costs of immunological up-regulation and a more suitable measure of parasite-induced mortality.

Tiffin and Inouye (2000) discussed the use of natural and imposed (controlled) damage in experiments of herbivore tolerance. They constructed a statistical model of the effect of herbivory on plant fitness, including damage level and an environmental factor as the independent factors, in which tolerance is defined as a slope of the regression line when damage level is regressed with plant fitness. They claim that while experiments with imposed damage are more accurate (i.e., they give a more correct estimate of tolerance), experiments with natural damage are more precise under a wide range of parameter values (i.e., tolerance estimates explain a larger part of variation in fitness). I show, however, that experiments with imposed damage are less precise only when an experimenter uses an experimental design that has weaker statistical power than in experiments with natural herbivory. The experimenter can nevertheless control the damage levels to optimize the experimental designs. For instance, when half of the experimental plants are left undamaged and the other half treated with maximal relevant damage level, experiments with imposed damage are almost always much more precise than experiments with natural damage.

Analysis of a recent mutation accumulation (MA) experiment has led to the suggestion that as many as one-half of spontaneous mutations in Arabidopsis are advantageous for fitness. We evaluate this in the light of data from other MA experiments, along with molecular evidence, that suggest the vast majority of new mutations are deleterious.