| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
What's this? |
Ecology |
School of Biological Sciences, The University of Sydney, NSW, 2006, Australia
Received for publication 11 April 2007. Accepted for publication 8 November 2007.
ABSTRACT
Up to 22% of plant species are the result of breeding among species—hybridization—directly conflicting with the prediction that hybrids, compared to parental species, are intermediate in character and of low fitness and little consequence. Few studies, however, have compared the fitness of hybrids and parental species under field conditions. This study evaluates components of fitness in the field for naturally occurring hybrids of the shrub Kunzea, relative to the parental speciesKunzea rupestris. Hybrid plants did not differ from the parental species in the level of effective pollination. Thus, we found no support for Grant’s model (Evolution: International Journal of Organic Evolution3: 82–97) of reduced fitness of hybrids via reduced pollination level (the intermediate hypothesis). Hybrids displayed variable fitness across the measured fitness components. Seed set levels for hybrids were structured among populations, suggesting genetic structuring for this fitness component at this scale. The response of hybrids to fire (a major selective force in the study system) was partly consistent with a resource trade-off model. Hybrids were large robust plants but most did not resprout after fire. Hence, the fitness of hybrids was complex. We developed a model for relative fitness to estimate fitness for species and hybrids with complex life histories.
Key Words: disturbance • fitness • hybrid • Myrtaceae • pollination • resprouting • seed set
Hybrids are individuals that have arisen from the interbreeding of two species or from the subsequent reproduction of those individuals. Since the work of Grant (1949) in the 1940s, hybrids have been dismissed as of minor evolutionary importance. Grant proposed that hybrids will have characteristics intermediate of parental species and that such plants would be of low relative fitness in the parental habitat. Hence hybrids, it was assumed, would contribute little to the evolution of species. However, it is now known that up to 22% of flowering plant species may be of hybrid origin (Ellstrand et al., 1996
), and hybridization is being recognized as a creative force in generating variation (see Rieseberg and Wendel, 2004 and papers cited therein; Hersch and Roy, 2007) and contributing to the colonization of novel habitats and invasions (Rieseberg et al., 2004). These observations and a number of important new hypotheses conflict with Grant’s widely accepted model. Therefore, the fitness of natural hybrids compared to their parents remains a critical debate, but there is limited empirical evidence to settle the matter (reviewed in Arnold and Hodges, 1995). Thus, further theoretical advances in evolutionary theory now depend on a better understanding of the process of hybridization. In particular, the fitness of hybrids relative to parental species needs to be accurately assessed so that a sound empirical basis is established to test new concepts and models of ecological speciation.
Campbell (2003) reviewed the data on hybrid fitness and concluded that there are few case studies that demonstrate relative levels of hybrid and parental fitness under field conditions. She investigated the fitness of Ipomopsis hybrids relative to parental species and partitioned fitness among a number of intrinsic fitness components (pollination and seed set) as well as an extrinsic fitness component (the field environment). The hybrids could be of lower, equal, or greater fitness than parental species, depending on the genotype class and the environment. Importantly, the work highlighted the need for fitness to be assessed in the relevant field environment. The other case studies reviewed by Campbell (in particular Wesselingh and Arnold, 2000; Johnston et al., 2001) and evidence from other sources (e.g., Lester and Kang, 1998; Sornsathapornkul and Owens, 1998), including more recent studies (e.g., Pelabon et al., 2005), support the concept that the relative fitness of hybrids could result from variations across pre- and postzygotic reproductive mechanisms or from physiological differences that interact with the field environment. Because Ipomopsis is a relatively short-lived monocarpic plant, it was possible to derive a measure of lifetime fitness. However, measuring lifetime fitness under field conditions is often more complicated because many parental species and hybrids are long-lived, with variable seed set patterns across their lifetime; produce a large numbers of seeds and have persistent seed banks from which seeds recruit intermittently; and potentially produce seeds from reciprocal crosses and backcrosses with parental species. Lifetime fitness is therefore difficult to assess for many plants. If we are to make progress and understand hybrid fitness in these complex systems, a first step is to investigate the variation in fitness of hybrids relative to parental species across a number of fitness components. This approach will allow researchers to focus on those fitness components that determine lifetime fitness.
The aim of the study was to estimate fitness components for hybrid plants relative to the parental species in the habitat of the parental species. We measured pollination levels, seed set, and survivorship through fire to determine whether these components, either in combination or individually, explain the performance of hybrid plants compared to the parental species in this system. For example, if hybrids survive fire at relatively low levels, this low survivorship may significantly reduce the lifetime fitness of hybrids relative to parental species. Alternatively, if hybrid plants are seed or pollen sterile, then these plants have zero fitness and fire has no impact on the lifetime fitness of such plants. We recognize that the measured fitness components do not capture all variation in survivorship or reproductive success. Similarly, the level of genetic variation among and within populations, the parentage of hybrids, and the hybrid generations present (e.g., F1 or F2) are yet to be determined. Hybrids, however, were all large robust plants, at least 7 years old (this can be determined by the annual growth pattern of shoots), flowered at high levels annually, and are considered likely to be F1 plants (see Tierney, 2003).
MATERIALS AND METHODS
We studied hybridization in the hybrid-prone genus Kunzea (Mrytaceae). We compared fitness components across the life cycle of K. rupestris and naturally occurring hybrids that had established in populations of K. rupestris. Kunzea is considered the genus most prone to hybridization within the Myrtaceae (P. Wilson, Royal Botanic Gardens, Sydney, personal communication), but hybridization also occurs in other Myrtaceae (e.g., Darwinina, Briggs, 1964; Leptospermum, Harris et al., 1992). In Kunzea, the field environment can contribute to the relative fitness of hybrids (de Lange and Norton, 2004). However, both prezygotic mechanisms (incompatibility systems) and postzygotic mechanisms (seed abortion) are recorded in the family and are mechanisms that could regulate hybrid fitness (Beardsell et al., 1993). The genus Eucalyptus is the most widely studied in the Myrtaceae (Pryor, 1976; Moran and Hopper, 1983; Potts and Reid, 1988; Schemske and Morgan, 1990; Passioura and Ash, 1993; Turner et al., 2001; Potts et al., 2003). Closely related species of Eucalyptus hybridize, and some hybrids are considered to have high levels of fitness (Potts et al., 2003). In our study system, hybrids occur in low numbers, which in itself could limit pollinator service and hence lower the relative fitness of hybrids (e.g., Jennersten, 1988).
The study was undertaken in a fire-prone habitat where key life history traits ensure persistence, in particular the capacity to resprout after fire (sensu Noble and Slatyer, 1980). The extent of adaptation to fire by species in this habitat is well documented (e.g., Keith et al., 2002). Hence, it is clear that fire exerts a major selective force in this habitat. Both of the parental species of the hybrids we studied have some capacity to resprout (Tierney, 2003). Resprouting assures survival after fire but possibly at a cost to growth and fecundity (Schwilk and Ackerly, 2005). We observed that the hybrids of these species were large robust plants and estimated flower-set at two to three times that of one parental species (and at least equal to the other). We reasoned that a measure of relative fitness of these hybrids should account for any fecundity difference as well as any trade-off in growth that this might entail.
The study species
We studied flowering phenology, pollination, seed set, and the response to fire in Kunzea rupestris Blakely (Myrtaceae) and in naturally occurring hybrids of this species, which had resulted from interbreeding with K. capitata Reichenbach (Myrtaceae) (Tierney, 2003). Kunzea rupestris and K. capitata are both in the subgenus Kunzea, based on molecular analyses (P. de Lange, Department of Conservation, Newton, Auckland, New Zealand, unpublished manuscript; P. Wilson, Royal Botanic Gardens, Sydney, personal communication), but whether they are sister taxa is not known because the infrageneric relationships have not yet been examined in detail. But it is clear from the molecular work that the relationships within the subgenus are highly reticulated, consistent with hybridization being a significant and ongoing evolutionary force in the genus. Previously, we demonstrated differences in seed germination characteristics among these species and hybrids (Tierney and Wardle, 2005). Where the species co-occur, hybrid plants are frequently found (J. Cohn, National Parks and Wildlife Service, unpublished manuscript; Tierney, 2003). In these locations, disturbance is generally evident, and it is likely that K. capitata has invaded these populations as a result (Tierney, 2003). Both of these species are described in detail by Wilson (1991). Kunzea rupestris is a clonal shrub to 1.5 m high with small oblanceolate leaves. It bears sessile white flowers in headlike clusters on elongated branches in spring. Fruit are small, indehiscent, hairy two-locular capsules. This species is restricted to 20 known populations on rocky sandstone outcrops northwest of Sydney, New South Wales, Australia. Kunzea capitata is generally a smaller shrub, although described by Wilson as growing up to 2 m. Leaves are similar to those of K. rupestris but usually recurved and held appressed to the branch. Flowers are sessile, pink to purple, and occur in headlike clusters on branch ends. Fruit are dehiscent and three-locular. Flowering occurs in late winter to spring (Wilson, 1991). This species has a large distribution in heaths and dry woodlands across New South Wales. Wilson described the known hybrids among Kunzea species as typically bearing pink flowers (which are lighter pink than those of K. capitata; D. A. Tierney, personal observation) and generally of intermediate character to the presumed parent species. Hybrids were identified in the field based on characteristics observed on specimens held in the Royal Botanic Gardens, Sydney (Appendix). These plants were then subjected to independent analyses of flower structure that determined that these plants formed a distinct group from the two species present within the study locations (Tierney, 2003). Although three Kunzea species occur in the vegetation where this study was undertaken, Tierney (2003) also determined that only K. rupestris and K. capitata have flowering periods that overlap, and that the third species (K. ambigua) was absent from the rocky outcrops where K. rupestris is found. Across all study locations, K. rupestris and hybrid plants were on average more than twice the size of K. capitata plants. They also fruit along branches (rather than only terminally, as for K. capitata), and therefore we estimated that they produce two to three times more flowers than K. capitata (D. A. Tierney, personal observation).
Relative fitness: the pollination component
Pollination difference between Kunzea rupestris and the hybrids
We estimated "effective pollinator service" (the germination of pollen applied to stigmas; Vaughton and Ramsey, 1995) by counting pollen tubes in the stigma and style of K. rupestris and hybrid plants. Some hybrid plants failed to produce any seed over our 3-year study, and we considered these to be sterile (see seed set studies later). Hence pollen tubes were examined in a sample of the hybrids identified as fertile or sterile (both open-pollinated and pollinator excluded treatments were applied). Pollen tubes were also examined in open-pollinated and pollinator-excluded flowers of K. rupestris.
Flowers used for the pollen tube studies were fully opened, but style withering and anther drop had not begun. Unless otherwise specified, flowers were picked in the second week of September (peak flowering). Pollinators were excluded by bagging (15 x 15 cm muslin bags) flower-bearing branches before the onset of flower opening. A procedural control to test for the effect of bagging on flower development consisted of bags identical to those in the pollinator-excluded treatment, except the bags were removed once flowering had commenced.
Flowers were fixed and stained as per Beardsell et al. (1993). After fixation in three parts ethanol to one part glacial acetic acid for 12 h and then storage in 70% ethanol, flowers were washed for 15 min. in 50 g NaSO3/L distilled water and then stained with 0.1% aniline blue in 0.1 M K3PO4 at 40°C for at least 2 d. For viewing pollen tubes, styles were severed from the gynoecium mounted in one drop of aniline blue stain and gently squashed between two slides. These samples were then viewed at 100x magnification using a Zeiss axiophot (Zeiss fluorescent filters 2, 9 and 15 were used, Carl Zeiss, North Ryde, Australia). After the initial inspection, styles were further squashed and reinspected. Observations were made in the top, middle, and bottom third of each style.
Relative fitness: the seed set component
Seed set among Kunzea rupestris, K. capitata, and hybrids among study populations and years
A preliminary study in the 1997 flowering season determined the presence of hybrids in three populations where K. rupestris occurred (Tierney, 2003). Each of these populations was then sampled over three years so that each plant type that occurred in a population was sampled at least once. Fruits were dissected using a stereomicroscope (100x magnification). At least 15 fruit from a given plant were sampled, and seed set was scored as the mean seed set per fruit for the plant. The total numbers of populations, plants, and fruits sampled for each year are shown in Table 1. Analyses were carried out to assess the variability in the relative reproductive output of K. rupestris and the hybrid plants among years and populations. Seed set variability for each of these genotypes was analyzed among the two populations that contained a number of hybrid plants. A factorial ANOVA (factor 1 = genotype; factor 2 = years) was then used to determine the variation in seed set among any genotypes (or if a genotype had been determined to have a distinct seed set level in a population, then the level of variation of seed set of this genotype in this population) among years.
|
Ovule number per flower among Kunzea rupestris, K. capitata, and the hybrids
The number of ovules per flower was estimated for both species and the hybrids to estimate how many ovules could potentially be pollinated and develop into seeds in each of these plant types. A total of 30 buds from 10 K. rupestris plants taken from five populations were dissected using a stereomicroscope (40x magnification). A total of 21 buds (three buds from each of seven plants) were dissected from K. capitata and from hybrid plants. Plants were haphazardly sampled to include at least two plants from each population that contained that plant type (K. rupestris, K. capitata, or hybrid). However, in one population only one hybrid was found and could be sampled.
Seed abortion among Kunzea rupestris and hybrid plants
Seed abortion levels were determined opportunistically on the fruits dissected for seed set studies (see Results for the number counted). A seed was counted as aborted if it was flat in profile (in cross section, 
50% the thickness of a fully developed seed). Aborted seeds also clearly lacked turgor and often differed in color. Undeveloped ovules (i.e., ovules that were putatively sterile) were defined as ovules in mature fruits that were 50% the size of an average ovule for undeveloped fruit. The ratio of aborted seeds to ovules per fruit was calculated for K. rupestris and the fertile Kunzea hybrids.
An assessment of the impact of the field environment on fitness
Persistence of Kunzea rupestris, K. capitata and hybrids
A baseline survey in 1997 to determine the status (presence and abundance of Kunzea species and hybrids) in selected populations confirmed the presence of hybrids. This survey was repeated in 1998 and 1999. Five populations of varied tenure and level of disturbance and with a geographical spread across the known range of K. rupestris were surveyed (sites 1, 2, 3, 6, and 8 in J. Cohn, National Parks and Wildlife Service, unpublished manuscript). These were surveyed at least once every two weeks from mid August to mid September over a 3-year period (1997–99). In initial investigations, K. rupestris plants were restricted to defined rocky outcrops. Thus, these outcrops and the area 20 m from the edges of these outcrops were surveyed. Each survey consisted of walked transects across these areas where Kunzea plants were recognized by general morphology and floral features (with a maximum transect width of 10 m). The number of K. capitata and hybrid plants was recorded for each population. Hybrids were also tagged with an individual code number. Kunzea rupestris plants are clonal; therefore, numbers were not recorded (J. Cohn, National Parks and Wildlife Service, unpublished manuscript provides estimates of percentage cover and numbers for these populations).
The impact of fire
An extensive wildfire in December 2002 burned all five study populations. These were resurveyed 4 mo and 20 mo after this fire to determine the effect of this fire on the status of each species and the hybrids in these populations. The number of burned K. capitata and hybrid plants in each population where they had previously been found (three populations) and the number of these burned plants that resprouted were counted. The impact on K. rupestris in these populations was estimated as the percentage of emergent stems that were burnt and the percentage of these burnt emergent stems that were resprouting. In another survey 45 mo postfire (i.e., in the third postfire flowering period), the number and status of hybrid plants were assessed for each population.
Analyses
Factorial ANOVAs were used to analyze differences in the pollination and seed set experiments, and the experimental designs for these analyses are summarized in Table 2. A Cochran’s test was used to test for homogeneity of variances before the ANOVAs were done. Percentage data were arcsine transformed before the analyses, and log or square-root transformations were used to stabilize variances when necessary (Sokal and Rolf, 1995). ANOVA significant differences were defined as P < 0.05, and a Student-Newman-Kuels (SNK) test was used to determine which means differed significantly. The program GMAV5 (Underwood and Chapman, 2001) was used for all ANOVA, Cochran, and SNK tests. When no significant differences were found, factors were pooled to increase the power of subsequent analyses. When experiments generated different replicate numbers among levels of a factor, replicates were removed randomly from a level to produce a balanced design (see Table 2).
|
RESULTS
Relative fitness: the pollination component
Pollen tubes were observed in both sterile and fertile hybrid styles, indicating that pollen had been transferred to the style and that effective pollination had occurred in both hybrid types (Fig.1). Thus, pollen tubes grew in most styles of both hybrid types (86 ± 7.8% of sterile hybrid styles and 69 ± 15.5% of fertile hybrid styles, with 15 flowers sampled from each of three sterile and three fertile hybrids) when open pollinated; but the level of effective pollination among the hybrid types was not significantly different (F1,5 = 1.13, P = 0.34). The effective pollination level of these hybrids was therefore pooled to increase the power of an analysis comparing pollination among K. rupestris and hybrid plants. This analysis found that K. rupestris and hybrid plants also did not differ significantly in effective pollination level when naturally pollinated or when pollinators were excluded (F1,23 = 8794, P < 0.0001: Fig. 2). The response to pollinator exclusion for K. rupestris and the hybrids was the same—a significantly reduced effective pollination level.
|
|
|
|
|
|
The December 2002 wildfire burned 90% of emergent K. rupestris stems, 95% of all K. capitata plants, and all hybrids. Approximately 60% of the K. rupestris emergent stems had resprouted after 20 mo. Kunzea capitata extensively resprouted in one population (
60% of burnt plants), but in a second population resprouting appeared less pronounced (10% of burnt plants). Hybrid plants were eliminated from two of the three populations (no plants resprouted and no seedlings were observed). In one population (where the hybrids were fertile), one hybrid resprouted. However, while many K. rupestris plants resprouted vigorously at the base of the burnt trunk, the hybrid resprouted from rootstock at some distance from the burnt trunk.
DISCUSSION
The fitness of hybrid plants compared to that of the parental species varied according to the class of hybrid (sterile or fertile) and ultimately depended on the field environment (prefire or survival after fire). Two classes of hybrid plants were identified: sterile hybrids produced fruit but no seeds, and fertile hybrids, which were vigorous plants that produced an equal number of seeds per fruit compared to K. capitata and more seeds per fruit than the other parental species, K. rupsetris. We next examine the details of pre- and postzygotic fitness components and then discuss the ecological situations that ultimately determine the fate of these naturally occurring hybrids.
Pollination
No evidence was found for a reduced fitness in hybrids relative to the studied parental species at the pollination component stage. In this study, Kunzea hybrids occurred in low numbers but were effectively pollinated at levels equivalent to the predominant parent species (K. rupestris) that occurred in these populations. Pollinator-excluded flowers had much lower pollination levels in both the hybrids and in K. rupestris. This strongly indicates that most pollination among these plants is due to a pollen vector. Across three flowering seasons of incidental observations, Apis mellifera and wasps of the genus Bembix were common on these plants (with few other floral visitors noted); we considered these the probable pollinators in this system (although direct observations of pollen transport from male to female flowers by each of these visitors would be necessary to confirm their role as effective pollinators). Grant (1949) predicted that pollination was likely to be a critical point in the life cycle where hybrids (with intermediate characters) would be selected against. This study does not support this prediction. This result may be because pollination is carried out by generalist pollinators including the introduced A. mellifera. These pollinator species are unlikely to have a close association with a particular plant species, and plant–pollinator interactions may be fundamentally altered in this system (e.g., Gross and Mackay, 1998; Kearns et al., 1998). However, low pollination levels relative to parental species have been reported for hybrids in other systems where species have generalist pollinator syndromes (e.g., Sornsathapornkul and Owens, 1998) or specialist pollinator syndromes (e.g., Wolf et al., 2001). In addition, we have not examined the late prezygotic stage (i.e., pollen tube performance in entering the ovule and effecting fertilization) and so have not determined that there are no differences in pollen tube performance at this late point in the prezygotic stage. Recently, context-dependent pollinator behavior has been identified as a cause of differences in patterns of hybridization among populations of three species of Indian paintbrush, Castilleja (Hersch and Roy, 2007). Because both sterile and fertile Kunzea hybrids received pollen and initiated pollen tube growth, we can eliminate pollinator limitation or visitation rates as the most likely explanation for sterile hybrids.
Seed set
Relative to parental species, hybrids had variable fitness as measured by seed set. This variability was expressed among the discrete populations (restricted to the rocky outcrops). In one population, hybrids produced more than twice as many seed per fruit as did K. rupestris and similar numbers of seed per fruit as did the second parent species, K. capitata. However, because an average hybrid bears at least twice the number of fruit of a K. capitata plant, these hybrids are more fecund. At two other populations, hybrid plants set no seeds. Ovules in these hybrids appeared normal in bud but did not enlarge postflowering. No aborted seeds were found in these hybrids (but aborted seeds were found in fertile hybrids and in K. rupestris), suggesting that a distinct mechanism imparts sterility (potentially acting late in the prezygotic or early in the postzygotic stage).
The cause of the variation in fitness among hybrid populations is unknown. However, genetic differences among study populations (spatial genetic structure) may play a role together with environmental differences. Kunzea rupestris plants are restricted to discrete rocky outcrops that are mostly more than 1 km apart. Genetic exchange among these populations is limited. Pollen can disperse up to several hundred meters for some Eucalyptus species but typically is shorter than this (Potts et al., 2003). Seed dispersal distances are generally even shorter for nonfleshy-fruited Myrtaceae species and are usually measured in meters (e.g., Auld, 1987). Indehiscent fruit may increase seed dispersal distances in the family (Myerscough, 1998), and K. rupestris fruit are light, hairy, and indehiscent. These can disperse many meters, particularly after fire, which opens up the ground layer (Tierney, 2003). However, the contribution that these dispersing fruits make to dispersal at scales of a kilometer is unknown.
Fine-scale spatial genetic structure in Eucalytus globulus (Myrtaceae) is reported to produce distinct relatedness groups (families) at a scale of about 30 m (Jones et al., 2003). In other restricted taxa, the genetic differences among populations are known to be large enough to result in limited breeding potential within a population (due to mate limitation, e.g., DeMauro, 1993; Rossetto et al., 2004) or dysgenic progeny when plants from different populations interbreed (Young and Murray, 2000). In contrast to E. globulus, K. rupestris is distributed in disjunct locations, mostly more than 1 km apart. In such isolated locations, theoretical models (Rieseberg et al., 2004) and empirical studies (see studies reviewed in Orr, 1997) support that hybrid sterility is likely to be structured among populations, and this is a likely cause of variable postzygotic fitness. However, other mechanisms may also produce the observed pattern of sterile and fertile hybrids. First, sterility may be expressed in F2 hybrids but not in F1 hybrids (Fenster et al., 1997). Second, the direction of cross (i.e., whether K. rupestris or K. capitata was the pollen donor) might determine whether hybrid plants are sterile or fertile. For example, differences in pollen performance have been found in Ipomopsis species that are asymmetric and dependant on the pollen donor species (Wolf et al., 2001). It seems plausible that such a genetic mechanism could operate at the late prezygotic stage (i.e., pollen tubes may fail to fertilize the ovule) or postzygotic stage. In the Myrtaceae, genetically determined seed abortion does occur (Beardsell et al., 1993). However, this indicates either that F1 and F2 hybrids were restricted to different populations (because hybrid fertility is structured among populations) or that the direction of pollen donation was different among populations. Mechanisms can be proposed that could produce such patterns (i.e., F1 plants might not resprout when burnt and have been eliminated by fire at one population, leaving only F2 plants that do resprout, while the plants at the other site are all F1 plants, and F2 plants will not be observed until a recruitment event is triggered by fire). However, these mechanisms seem less plausible than a genetic structuring among populations (i.e., the observed fire event eliminated most hybrid plants from all sites, so it seems unlikely that these were F1 and F2 plants with different resprouting capacities). Genetic markers for maternal inheritance could be used to clarify the pattern of hybrid sterility among populations (Hatfield et al., 1985). If, for example, hybrids in a population are all sterile and this corresponds with a particular species being the maternal species in this population, one would infer that sterility is determined by maternity. Irrespective of the cause, the seed set of hybrids varied greatly among populations.
The effect of the field environment
Crone (2001) demonstrated that survivorship is an important contributing factor to plant fitness (proportionally more so than fecundity for longer-lived plants). In this study (and in the study system in general), survivorship is largely determined by a capacity to survive fire. Hence, a trade-off of increased fecundity for decreased fire survival assurance (e.g., Schwilk and Ackerly, 2005) may represent a loss of fitness. In two populations, hybrids were large, vigorous, sterile, and eliminated by a single fire event. These hybrids had equivalent levels of pollination to the studied parental species, but low levels of fitness as measured by seed set and a low survivorship of fire. At the third population, all hybrid plants were large and fecund. In this population, hybrids also had levels of pollination and seed set equivalent (at least) to the parental species. However, it needs to be acknowledged that the proportion of seeds maturing from ovules was lower in these hybrids than in the parental species. The proportion of seeds maturing per ovule needs to be carefully considered because if this component were adopted as a fitness estimate (rather than seed set per flower), a different estimate of fitness would result. Ultimately then, fitness needs to be assessed at the level of the plant and then measured over the lifetime (which would include estimating survivorship). Survivorship of these hybrids was variable in response to fire.
The observed fire was typical of the fire events in the study area. The fire was an intense summer fire under dry conditions, strong winds, and low humidity (D. A. Tierney, personal observation). This fire eliminated hybrids from two of the three populations, and only one hybrid plant survived at the other population. There is an extensive data set on the effect of fire on plant species in this habitat type, as well as models that predict the impact of the fire regime on plant species with differing survivorships in response to fire (see Keith et al., 2002). Survivorship is critical when fire frequencies are high, unless a plant can establish quickly from seed postfire and then set seed within a few years. It will be important to monitor the establishment of hybrids and the time to first flowering among populations to more accurately predict the impact of fire on hybrid fitness. We have determined that hybrid seed is viable and germinates in response to fire cues (Tierney and Wardle, 2005).
Overall fitness estimates
Lifetime reproductive fitness could not be directly measured for plants in this study. However, it is useful to derive estimates of the relative fitness of these plants as a basis for understanding which of the measured components might be most important in determining overall fitness, and how these components interact. If seed set (measured over 3 years) is used as the fitness estimate (ignoring pollen flow from plants), the hybrids from two populations have zero fitness and can be eliminated from further analysis. The hybrids from the third population have the highest relative fitness (1.0) compared to parental species (0.3–0.5). However, survivorship estimates based on fire events and plant responses to fire alter these estimates. Assuming that flowering correlates with plant size (which is a reasonable assumption because flowers are borne in a similar configuration on hybrids and on K. rupestris plants) and that there is a threshold size for flowering (say 1.0 m), then relative fitness can be estimated by summing the estimated fruit set across the lifespan of a plant and multiplying this by the seed set per fruit. The area under the curves on Fig. 6 can therefore be used as an estimate of relative fitness, and accurately estimating these areas may provide a way forward in understanding the relative fitness of hybrid plants and parental species. A clear advantage of such an approach is that it allows researchers to integrate fitness estimates across the lifetime of the plant. For example, a large hybrid plant that lacks a capacity to resprout after fire may have lower fitness compared to a small resprouting K. rupestris plant that survives many fire events (i.e., the sum of the areas under the curve for the resprouting plant would be larger than the single area for the nonresprouting plant). Alternatively, the growth rate (vigor) of a resprouting plant may decrease after repeated fires, (i.e., the area under the curve would decrease in size through time, and this could be integrated into a fitness estimate for this plant). Clearly, estimating these areas (i.e., accurate measures of growth rates, time to first flowering, and the probability of resprouting) is critical for estimating relative fitness. Other important variables will be the capacity of a plant to resprout epicormically rather than from a lignotuber (assuming this leads to earlier postfire flower set, and therefore, greater seed production), the growth rate after fire (which is likely to decline with a high fire frequency; Knox and Morrison, 2005), or the size/age threshold for flowering. Therefore, the ecological context is critical for assessing the fitness outcomes in this system, and estimates of fitness from a greenhouse study, also difficult for a shrub, are not sufficient. Indeed, as a further complication, in disturbed systems fitness may vary because of a range of factors. For example, in our study, K. capitata resprouted vigorously in one population that had undergone postmining restoration and that appeared to have artificially deep soils, which would provide extra protection for rootstock from fire events (Tierney, 2003). Although limited, this single observation suggests a long-term change to the habitat of this population and to the level of genetic introgression that would promote recurrent hybridization. Where species are not completely isolated, it may be the level of introgression, relative to the amount of ongoing selection, that determines the fate of a population (Martin and Willis, 2007). Further, under conditions of reduced pollinator service (e.g., for a small population establishing in novel habitat) and an increased reliance on selfing, the Kunzea hybrids could produce about twice the seed per fruit as the K. rupestris plants (i.e., the hybrids may be more fit under these conditions).
|
; Arnold, 1997; Allendorf et al., 2001) that simplistic assessments of the relative fitness of hybrids compared to parental species may be misleading. This situation is problematic because only a few cases of hybridization among particular species have been studied in detail (but see Potts et al., 2003; Rieseberg and Wendel, 2004; Martin and Willis, 2007). Further research on the causes of differences in hybrid fitness should incorporate both extrinsic and intrinsic factors and adequate temporal and spatial sampling, and should document the role of divergent ecological selection across habitats and among populations. It is also important to realize that field-based observations will usually represent only a subset of the plausible variations in extrinsic factors that are important for predicting fitness. Fire regimes, for example, may have variable impacts on fitness estimates because survivorship will be influenced by particular site conditions (e.g., the position, the aspect, the degree of protection of plants, or specific soil conditions). Experimental approaches that determine the effect of varied fire intensities or length of exposure time to fire could be used to better understand survival in relation to varied fire conditions. Additionally, we propose that simulations (based on our initial model for seed production) using varied survivorships and observed growth rates and flowering patterns will be useful in estimating fitness outcomes under a range of scenarios. Such simulations are now a research priority for this study system. Appendix 1. Specimens used for identification of species restricted to Australia. The following specimens were used for the identification of species and hybrids. These specimens are held at the Royal Botanic Gardens, Sydney, Australia.
|
FOOTNOTES
1 P. Myerscough, T. Auld, P. Wilson and S. Hopper are thanked for advice to D.T. that helped guide this study. P. de Lange kindly provided advice on his genetic research on Kunzea. The New South Wales Department of Environment and Climate Change provided research licenses. 
2 Author for correspondence (e-mail: David.Tierney{at}newcastle.edu.au), current address: Environmental and Life Sciences, Ourimbah Campus, University of Newcastle, Ourimah, NSW, 2258 Australia
LITERATURE CITED
Allendorf, F. W., R. F. Leary, P. Spruell, AND J. K. Wenburg. 2001. The problems with hybrids: setting conservation guidelines. Trends in Ecology &Evolution 16: 613–622.[CrossRef][Web of Science]
Arnold, M. L. 1997. Natural hybridization and evolution. Oxford University Press, New York, New York.
Arnold, M. L., AND S. A. Hodges. 1995. Are natural hybrids fit or unfit relative to their parents? Trends in Ecology &Evolution 10: 67–71.[CrossRef][Web of Science]
Auld, T. D. 1987. Post-fire demography in the resprouting shrub Angophora hispida Sm. Blaxell: flowering, seed production, dispersal, seedling establishment and survival. Proceedings of the Linnean Society of New South Wales 109: 259–269.
Beardsell, D. V., R. B. Knox, AND E. G. Williams. 1993. Breeding system and reproductive success of Thryptomene calycina Myrtaceae. Australian Journal of Botany 41: 333–353.[CrossRef][Web of Science]
Briggs, B. G. 1964. The control of interspecific hybridisation in Darwinia. Evolution; International Journal of Organic Evolution 18: 292–303.
Campbell, D. R. 2003. Natural selection in Ipomopsis hybrid zones: implications for ecological speciation. The New Phytologist 161: 83–90.
Crone, E. E. 2001. Is survivorship a better fitness surrogate than fecundity? Evolution: International Journal of Organic Evolution 55: 2611–2614.[Medline]
De Lange, P. J., AND D. A. Norton. 2004. The ecology and conservation of Kunzea sinclairii (Myrtaceae), a naturally rare plant of rhyolitic rock outcrops. Biological Conservation 117: 49–59.[CrossRef][Web of Science]
Demauro, M. M. 1993. Relationship of breeding system to rarity in the Lakeside Daisy Hymenoxys acaulis var glabra. Conservation Biology 7: 542–550.[CrossRef][Web of Science]
Ellstrand, N. C., R. Whitkus, AND L. H. Rieseberg. 1996. Distribution of spontaneous plant hybrids. Proceedings of the National Academy of Sciences, USA 93: 5090–5093.
Fenster, C. B., L. F. Galloway, AND L. Chao. 1997. Epistasis and its consequences for the evolution of natural populations. Trends in Ecology &Evolution 12: 282–286.[CrossRef][Web of Science]
Grant, V. 1949. Pollination systems as isolating mechanisms in angiosperms. Evolution; International Journal of Organic Evolution 3: 82–97.[Medline]
Gross, C. L., AND D. Mackay. 1998. Honeybees reduce fitness in the pioneer shrub Melastoma affine (Melastomataceae). Biological Conservation 86: 169–178.[CrossRef][Web of Science]
Harden, G. J. [ed.] 1991. Flora of New South Wales, vol. 2, New South Wales University Press, Kensington, New South Wales, Australia.
Harris, W., N. G. Porter, AND M. I. Dawson. 1992. Observations of biosystematic relationships of Kunzea sinclarii and on an intergenic hybrid Kunzea sinclarii x Leptospermum scoparium. New Zealand Journal of Botany 30: 213–230.[Web of Science]
Hatfield, P. M., R. C. Shoemaker, AND R. G. Palmer. 1985. Maternal inheritance of chloroplast DNA within the genus Glycine, subgenus soja. The Journal of Heredity 76: 373–374.
Hersch, E. I., AND B. A. Roy. 2007. Context-dependent pollinator behaviour: an explanation for patterns of hybridization among three species of Indian paintbrush. Evolution: International Journal of Organic Evolution 61: 111–124.[Medline]
Jennersten, O. 1988. Pollination in Dianthus deltoids (Caryophyllaceae): effects of habitat fragmentation on visitation and seed set. Conservation Biology 2: 359–366.[CrossRef][Web of Science]
Johnston, J. A., D. J. Grise, L. A. Donovan, AND M. L. Arnold. 2001. Environment-dependent performance and fitness of Iris brevicaulis, I. fulva (Iridaceae), and hybrids. American Journal of Botany 88: 933–938.
Jones, T. H., R. E. Vaillancourt, AND B. M. Potts. 2003. Pollen movement and bi-parental breeding in the Tasmanian blue gum. Consequences of habitat fragmentation workshop. Royal Botanic Gardens Trust, Sydney, Australia.
Kearns, C. A., D. W. Inouye, AND N. M. Waser. 1998. Endangered mutualism: the conservation of plant–pollinator interactions. Annual Review of Ecology and Systematics 29: 83–112.[CrossRef][Web of Science]
Keith, D. A., W. L. McCaw, AND R. J. Whelan. 2002. Fire regimes in Australian heathlands and their effects on plants and animals. In R. A. Bradstock, J. E. Williams, and A. M. Gill [eds.], Flammable Australia: the fire regimes and biodiversity of a continent, 199–237, Cambridge Press, Cambridge, UK.
Knox, K. J. E., AND D. A. Morrison. 2005. Effects of inter-fire intervals on the reproductive output of resprouters and obligate seeders in the Proteaceae. Austral Ecology 30: 407–413.[CrossRef][Web of Science]
Lester, R. N., AND J. H. Kang. 1998. Embryo and endosperm function and failure in Solanum species and hybrids. Annals of Botany 82: 445–453.
Martin, N. H., AND J. H. Willis. 2007. Ecological divergence associated with mating system causes nearly complete reproductive isolation between sympatric Mimulus species. Evolution: International Journal of Organic Evolution 60: 68–82.
Moran, G. F., AND S. D. Hopper. 1983. Genetic diversity and the insular population structure of the rare granite rock species Eucalyptus caesia Benth. Australian Journal of Botany 31: 161–172.[CrossRef][Web of Science]
Myerscough, P. J. 1998. Ecology of Myrtaceae with special reference to the Sydney region. Cunninghamia 5: 787–808.
Noble, I. R., AND R. O. Slatyer. 1980. The use of vital attributes to predict successional changes in plant communities subject to recurrent disturbance. Vegetatio 43: 5–21.[CrossRef][Web of Science]
Orr, H. A. 1997. Haldane’s rule. Annual Review of Ecology and Systematics 28: 195–218.[CrossRef][Web of Science]
Passioura, J. A., AND J. E. Ash. 1993. Phenotypic, genetic and ecological variation in the Eucalyptus saligna-E. botryoides complex. Australian Journal of Botany 41: 393–412.[CrossRef][Web of Science]
Pélabon, C., M. L. Carlson, T. F. Hansen, AND W. S. Armbruster. 2005. Effects of crossing distance on offspring fitness and developmental stability in Dalechampia scandens (Euphorbiaceae). American Journal of Botany 92: 842–851.
Potts, B. M., R. C. Barbour, A. B. Hingston, AND R. E. Vaillancourt. 2003. Genetic pollution of native eucalypt gene pools—identifying the risks. Australian Journal of Botany 51: 1–25.[CrossRef][Web of Science]
Potts, B. M., AND J. B. Reid. 1988. Hybridization as a dispersal mechanism. Evolution: International Journal of Organic Evolution 42: 1245–1255.
Pryor, L. D. 1976. Biology of eucalypts. Institute of Biology, Studies in biology no. 61. Edward Arnold, London, UK.
Rieseberg, L. H., S. A. Church, AND C. L. Morjan. 2004. Integration of populations and differentiation of species. The New Phytologist 161: 59–69.[CrossRef]
Rieseberg, L. H., AND J. Wendel. 2004. Plant species—rise of the poor cousins. The New Phytologist 161: 3–7.[CrossRef]
Rossetto, M., C. L. Gross, R. Jones, AND J. Hunter. 2004. The impact of clonality on an endangered tree Elaeocarpus williamsianus in a fragmented rainforest. Biological Conservation 117: 33–39.[CrossRef][Web of Science]
Schemske, D. W., AND M. T. Morgan. 1990. The evolutionary significance of hybridization in Eucalyptus. Evolution: International Journal of Organic Evolution 44: 2150–2151.
Schwilk, D. W., AND D. D. Ackerly. 2005. Is there a cost to resprouting? Seedling growth rate and drought tolerance in sprouting and nonsprouting Ceanothus (Rhamnaceae). American Journal of Botany 92: 404–410.
Sokal, R. R., AND F. J. Rolf. 1995. Biometry: the principles and practice of statistics in biological research. Freeman, New York, New York, USA.
Sornsathapornkul, P., AND J. N. Owens. 1998. Pollination biology in a tropical Acacia hybrid A. mangium Wild xA. auriculiformis A. Cunn. ex Benth. Annals of Botany 81: 631–645.
Tierney, D. A. 2003. Hybridization and its management implications for the rare plant Kunzea rupestris Blakely. PhD dissertation, University of Sydney, Sydney, Australia.
Tierney, D. A., AND G. M. Wardle. 2005. Differential seed ecology in the shrubs Kunzea rupestris, Kunzea capitata and associated hybrids (Myrtaceae): the function of thin-walled fruits in a fire-prone vegetation. Australian Journal of Botany 53: 313–321.[CrossRef][Web of Science]
Turner, C., R. J. E. Wiltshire, B. M. Potts, AND R. E. Vaillancourt. 2001. Variation in seedling morphology in the Eucalyptus risdonii-E tenuriamis complex. Australian Journal of Botany 49: 43–54.[CrossRef][Web of Science]
Underwood, A. J., AND M. G. Chapman. 2001. GMAV5. Institute of Marine Ecology, University of Sydney, Australia.
Vaughton, G., AND M. Ramsey. 1995. Pollinators and seed production. In J. Kigel, G. Galili, and M. Dekker [eds.], Seed development and germination, 475–490, Marcel Dekker, New York, New York, USA.
Wesslingh, R. A., AND M. L. Arnold. 2000. Pollinator behavior and the evolution of Louisiana iris hybrid zones. Journal of Evolutionary Biology 13: 171–180.[CrossRef][Web of Science]
Wolf, P. G., D. R. Campbell, M. N. Waser, S. D. Sipes, T. R. Toler, AND J. K. Archibald. 2001. Tests of pre- and postpollination barriers to hybridization between sympatric species of Ipomopsis (Polemoniaceae). American Journal of Botany 88: 213–219.
Young, A. G., AND B. G. Murray. 2000. Genetic bottlenecks and dysgenic gene flow into re-established populations of the grassland daisy Rutidosis leptorrhynchoides. Australian Journal of Botany 48: 409–416.[CrossRef][Web of Science]
CiteULike Complore Connotea Del.icio.us Digg Facebook Reddit Technorati Twitter What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
