Because the response of bee visitors to land use change depends on species specific requirements and these pollinators also have variable effects on plants, understanding the effect of land use change on pollination services requires knowledge not only of which pollinator groups shift to the human-altered landscapes, but also the rate of pollination that those groups have on the plant species in those landscapes. Future research will benefit from looking at a wider range of plants with a different range of target pollinators and that flower earlier in the year to better tease out these hypotheses. If the patterns of bee visitation and seed set that we observed are indeed consistent across other plant species, the novel plant communities created in these human-altered landscapes and the generalist bee species that are favored in such landscapes will lead to a reduction in overall pollination services. Under the Biological Species Concept, speciation is the evolution of barriers to gene flow between sister lineages . Comparative tests of the geography and evolutionary forces driving speciation suggest that reproductive isolation evolves primarily via divergent natural or sexual selection accompanied by some degree of spatial isolation between populations .
Within this framework, plastic plant pot the process of speciation can be conceptualized as comprising three stages: niche expansion through the colonization of a new selective environment, ecological divergence, and the evolution of reproductive isolating barriers between lineages. Within species, populations in unique selective environments reflect all stages along this continuum, from maladapted sink populations to strongly isolated and locally adapted ecotypes . Such populations may become extinct, homogenized through gene flow, or remain partially isolated ecotypes; intraspecific divergence only rarely results in the evolution of a new, persistent biological species. Understanding the ecological and evolutionary processes that determine progress toward adaptive divergence and speciation remains a major goal of evolutionary biology . In this dissertation, I tested specific hypotheses about the processes operating at each of these stages, using acombination of stochastic simulations and field and greenhouse experiments with life history ecotypes of common monkeyflower Mimulus guttatus DC .The initial evolution and persistence of small populations is critical for understanding the process of speciation, particularly in sessile organisms such as plants. Gene flow has historically been viewed as a powerful force maintaining genetic cohesion throughout a species range, necessitating large-scale geographic isolation for speciation to occur .
However, botanists have argued that gene flow is sufficiently limited in many sessile organisms to allow smaller populations to evolve more or less independently, particularly at the periphery of a species range or under strong divergent selection . Contemporary evidence suggests that the spatial scale of speciation is in fact related to the scale of gene flow , and that new plant species are often formed in small, ecologically divergent populations . Speciation in initially small populations is fundamentally different from vicariant speciation in widely-distributed races because of the complex interplay of selection, genetic drift, inbreeding, and gene flow . Small populations face unique Demographic and genetic processes may limit or promote adaptation to novel environmental conditions. Such processes are fundamental to determining when and why adaptation to novel environments may fail, as at species range limits , or when successful adaptation promotes the invasion of non-native species or the evolution of reproductive barriers . Properties that influence the evolvability of species have been central to theory regarding the maintenance of sexual reproduction and outcrossing , as well as recent interest in predicting species’ responses to global environmental change . Historically, models of adaptation to novel environments focused on the genetic consequences of selection, gene flow, and genetic drift as well as demographic processes, such as immigration . However, recent work suggests that phenotypic plasticity may commonly play an important role in allowing population persistence in novel environments . Unlike adaptive evolution, phenotypic plasticity can immediately increase local fitness following colonization of a new environment .
Phenotypic plasticity may be particularly likely in harsh or stressful environments due to developmental instability or the expression of crypticgenetic variation , and can shift a population to within the “realm of attraction” of a new fitness peak . Phenotypic plasticity can evolve as an adaptive response to the range of environments typically encountered by an organism over space and time, or can be a passive response to stress; in either case it can be adaptive or maladaptive in a novel selective environment . Levin suggested that plasticity in traits affecting the mating system may have particularly important consequences for niche evolution in plants because the mating system has direct impacts on both the demographic and genetic properties of a population. Specifically, Levin suggested that niche evolution in plants may be facilitated if colonization of a novel environment is associated with increased self-fertilization via plasticity in floral traits or self-incompatibility systems. Diverse floral traits affect self-fertilization rate in plants, including the spatial and temporal separation of stigma and anthers, the proportion of cleistogamous flowers, and the expression of self-incompatibility mechanisms . Plasticity in one or more of these traits resulting in increased self-fertilization has been widely documented in response to environmental stress, including herbivory , pollen limitation , drought , heat , salt , and shade . Conversely, several studies have found either no plasticity in mating system or increased outcrossing in response to environmental stress . Thus, the magnitude, underlying traits, and environmental drivers of mating system plasticity appear to vary greatly among taxa. From a genetic perspective, increased self-fertilization may promote adaptation by acting as a partial reproductive barrier to maladaptive gene flow or by temporarily increasing genetic variation in traits under selection. Although self fertilization is associated with reduced genetic variation over long timescales , rapid increases in the self-fertilization rate are predicted to increase genetic variation temporarily by generating positive correlations between additive allelic effects within loci among offspring . Such allelic correlations may increase the rate at which genetic variance recovers following a bottleneck and the response to selection in populations with mixed mating . Self-fertilization may further increase genetic variation in traits with more complex genetic architectures by converting epistatic or dominance variance to additive genetic variance . Partial self-fertilization may also result in more rapid fixation of new, beneficial mutations and essentially produces assortative mating for traits under selection . From a demographic perspective, self-fertilization provides reproductive assurance by ensuring at least some reproduction when mates are limiting . Mate limitation may be severe in novel environments because of small population sizes and/or isolation. Plants with abiotic pollination mechanisms often exhibit density-dependent pollen limitation . In animal-pollinated systems, scarce ancestral pollinators or ineffective novel pollinators may limit outcross pollen availability in new environments . The importance of reproductive assurance during colonization is supported by the observation that isolated or peripheral plant populations often exhibit increased self-fertilization . Correspondingly, plastic planter pot self-fertilization is associated with invasiveness in annual weeds and larger range sizes in Collinsia . Despite these potential benefits of mating system plasticity for the persistence and adaptation of colonizing populations, sudden increases in self-fertilization rate are commonly associated with reduced fitness due to the expression and fixation of deleterious alleles that accumulate in previously outcrossing populations . If the segregating genetic load of an outcrossing population is high, a plastic increase in self fertilization rate could decrease the efficiency of selection on quantitative traits and increase the probability of extinction following colonization . In addition, inbreeding depression is often environmentally-dependent, and may be most severe in stressful or novel environments . Alternatively, the process of colonization itself may result in the fixation or purging of deleterious alleles if the number of colonists is small . Population bottlenecks increase the frequency of rare deleterious alleles, and may result in reduced fitness regardless of self-fertilization rate.
Given the diverse potential effects of mating system plasticity on the demography and evolution of colonizing populations, its overall consequences for niche evolution remain unclear. We used individual-based simulations to examine the evolutionary and ecological contexts in which mating system plasticity may promote or inhibit niche evolution. We tested the effect of a constant plastic increase in prior self-fertilization rate on population persistence and local adaptation by examining the evolution of a quantitative trait under stabilizing selection and the probability of extinction in a novel environment. Specifically, we focus on a region of genetic and demographic parameter space in which sink populations are in a ‘race’ to become locally adapted prior to extinction. To distinguish reproductive assurance, increased genetic variation, and reproductive isolation as potential mechanisms, we tested the effects of pollen limitation, strength of selection, genetic architecture, and maladaptive gene flow on both extinction probability and local adaptation. We examined the evolution of inbreeding depression and genetic load in the colonizing population to test whether consideration of deleterious mutations alters the effects of mating system plasticity on niche evolution. Although this model includes several assumptions based on plant reproductive biology , it could also be applied to hermaphroditic animal systems with mate limitation and/or gamete dispersal .We used stochastic simulations that track individual genotypes to model the evolution and demography of a colonizing sink population connected by gene flow with a source population. Our model is similar to that used by Holt et al. in which one-way migration occurs from a locally adapted source population to an initially maladapted sink population. However, we consider a plastic increase in self fertilization rate following colonization of the sink habitat, the potential for pollen limitation, and the evolution of inbreeding depression and genetic load due to deleterious mutations.The source and sink populations are composed of diploid and hermaphroditic individuals with discrete generations. Initially, the source population contains K individuals and the sink habitat is empty. The source population evolves for 1000 generations to reach mutation selection equilibrium before C individuals are randomly selected without replacement to colonize the sink habitat. Both populations evolve for 1000 generations following colonization or until the sink population goes extinct. The order of life-history events within each generation is: fertilization, selection, density dependence, dispersal, reproduction, and death. Prior to reproduction, individuals undergo selection. This selective period encompasses mortality at all life-history stages between fertilization and reproduction, including seed development, germination, and seedling growth. The fitness of an individual is determined by two forms of selection: optimizing selection on a quantitative trait, and purifying selection on deleterious mutations. Optimizing selection can be either directional or stabilizing, depending on the similarity of the average phenotype to the optimum.Immediately following colonization of the sink habitat, populations either begin to evolve towards the optimum phenotype and increase rapidly in size, or decline towards extinction . Population size is strongly associated with the average genotypic value, and extinction occurs only in maladapted populations with sizes far below K . Thus, density dependence only occurs once sink populations have begun to adapt, and the value of K does not affect colonization success. Extinction mainly occurs within the first 10 generations following colonization, and no populations become extinct after 50 generations . Given that the key demographic and genetic dynamics determining colonization success operate within the first 50 generations, we focus our results on this period. However, we also examine longer-term outcomes after 500 or 1000 generations to test whether these patterns change over time.Mating system plasticity has a profound and immediate effect on the genetic variance in colonizing sink populations . In general, an increase in self fertilization rate temporarily increases genetic variation relative to obligate outcrossing, but this effect decreases through time. Mixed mating maintains higher genetic variation than obligate outcrossing for tens to hundreds of generations following colonization, whereas obligate self-fertilization results in an initial spike in genetic variation that declines rapidly to levels below obligate outcrossing populations . This general pattern was observed across a range of parameter values, though the magnitude and duration of mating system effects on genetic variance depend on the genetic architecture of the quantitative trait . The effects of mating system plasticity on genetic variance have consequences for niche evolution when adaptation is limited by low genetic variation or strong selection. Under these conditions, the sustained increase in genetic variance under mixed mating allows a greater response to selection following colonization.