Effects of different levels of inbreeding on progeny fitness in Plantago coronopus.
There are two main hypotheses for the genetic basis of inbreeding depression. According to the overdominance hypothesis, outbred progeny perform better because they have a larger proportion of heterozygous loci and many of these loci are overdominant. Under the partial dominance (or mutation-selection balance) hypothesis, outbred progeny perform better because their deleterious, mainly recessive alleles tend to be masked in heterozygous condition. In outcrossing populations, mildly and highly deleterious mutations and polymorphisms for overdominant alleles may all contribute to inbreeding depression, while in highly selfing populations, mildly deleterious recessive mutations are expected to be the main cause of inbreeding depression, because of the purging of highly deleterious alleles (D. Charlesworth et al. 1990). Genetic analyses suggest that recessive deleterious alleles, rather than overdominant genes, are the primary cause for inbreeding depression (Wright 1977; D. Charlesworth and Charlesworth 1987; Johnston and Schoen 1995; but see Fu and Ritland 1994). When genetic load is determined primarily by recessive mutations, interactions between the mating system and inbreeding depression should lead to one of two evolutionary stable endpoints: predominant outcrossing associated with strong inbreeding depression or predominant self-fertilization with weak inbreeding depression (Lloyd 1979; Lande and Schemske 1985; D. Charlesworth et al. 1990, 1992). The threshold between these two states depends partly on the magnitude of inbreeding depression. When in-breeding depression is less than one-half selfing should evolve, otherwise outcrossing will be the mode.
The degree to which inbreeding depression determines the evolution of selfing rates, however, has been questioned from different angles. Many plants display intermediate levels of self-fertilization (Schemske and Lande 1985; Aide 1986; Barrett and Eckert 1990; Husband and Schemske 1996). A major issue is what level of inbreeding depression is maintained in partially self-fertilizing populations given the purging effects of self-fertilization on genetic load. Moreover, Campbell (1986), Holsinger (1988, 1991), and Uyenoyama and Waller (1991a,b,c) have demonstrated that population-level estimates of inbreeding depression alone are insufficient to predict whether a selfing variant can invade in a particular population. They showed that associations may develop between viability and mating-system loci that are as important as inbreeding depression itself. One of the mechanisms for this association is the production of highly fit, inbred progeny that are homozygous for the wild-type alleles. These models indicate the importance of determining the variation in inbreeding depression among individuals within a population, as this may better predict whether a selfing variant can invade in an outcrossing population.
Not only the number of deleterious mutations, but also the way these mutations interact between loci (epistasis) influences the level of inbreeding depression and is of importance for the evolution of mating systems (Crow 1970; Kondrashov 1982; B. Charlesworth 1990). Epistasis influences the rate at which the genetic load can be purged with inbreeding. When deleterious mutations act independently, fitness effects of different loci multiply, so there is a linear relationship between the log of individual fitness and the number of mutations in the genome, and the mean fitness of the population is independent of whether reproduction is sexual or asexual. When deleterious mutations act synergistically, however, such that the log of fitness declines at a greater than linear rate with the number of mutations, the mean fitness of sexual populations always exceeds that of asexual populations. The selective advantage is often sufficient to offset the cost of sex (Kondrashov 1988). The effect of synergism on the evolution of self-fertilization is to increase the equilibrium inbreeding depression over what would be expected if mutations act independently (B. Charlesworth et al. 1991). Because it is often impossible to determine directly the relationship between the number of mutations and fitness per individual in natural populations, epistasis among deleterious alleles can only be studied indirectly by monitoring the effect of different levels of inbreeding on fitness components (Willis 1993). Synergism should be manifest in a nonlinear negative relationship between log-transformed fitness and the expected inbreeding coefficient (Crow and Kimura 1970).
Here I present results of a study designed to estimate the mating system, magnitude of inbreeding depression, the relationship between fitness and inbreeding level, and the variation among maternal plants in inbreeding depression in one population of partially self-fertilizing Plantago coronopus. The experimental approach used was to create different inbreeding levels within randomly selected maternal plants and follow the fate of their progeny at the site of origin and in the greenhouse. Thus, the same genetic material was grown in two environments, because the expression of inbreeding depression may depend on the experimental environment (Schemske 1983; Dudash 1990; Schmitt and Ehrhardt 1990; Wolfe 1993; Eckert and Barrett 1994). Hence, where feasible, experiments should preferably be conducted in natural habitats to get proper estimates of mating-system parameters. The design allowed me to determine the variation in inbreeding depression among individuals within a population, to determine whether purging of the genetic load occurs uniformly among families, and to gain insights into whether dominance, overdominance, and/or epistasis underlie the expression of inbreeding depression in fitness related traits.
MATERIALS AND METHODS
Plantago coronopus is a small, short-lived perennial in The Netherlands that mainly grows in meadows along the coast. The species flowers from the beginning of May through September and overwinters as a rosette. It is a wind-pollinated, gynodioecious, and predominantly outcrossing species (mean outcrossing rate 0.77; Wolff et al. 1988). The flowers are protogynous, with an overlap in sexual phase. Flowers are formed on spikes and produce a fixed number of five ovules. Flowering and subsequent maturation occur from the base upward.
In February 1988, 60 nonflowering, adult plants were collected from a salt-meadow (Westplaat population, see Koelewijn and van Damme 1995), transferred to the greenhouse, and raised until full flowering. Eight hermaphrodites were selected randomly and used to create different inbreeding levels and to produce seeds with different inbreeding coefficients. The motherplants were selfed (S1, f = 0.5) and crossed with a pollen mixture of 10 randomly chosen, other plants (0, f = 0). From each mother, five S1-seeds were sown and self-pollinated to give the S2- (f = 0.75) generation, and five O-seeds were sown and backcrossed to the mother to give the B1- (f = 0.25) generation. Subsequently, from each S2-individual, five seeds were sown and selfed to give the S3- (f = 0.875) generation.
The outcrossing rate at the study site was estimated by two indirect methods. Fifty-three hermaphroditic plants, originating from the study site, were screened for six isozyme loci by means of horizontal starch gel electroforesis, following the methods described by van Dijk et al. (1988). Five loci (Acon, aconitase; Idh, isocitric dehydrogenase; 6Pgd, 6-phosphogluconate dehydrogenase; Pgm, phosphoglucomutase; and Gpi, glucosephosphate isomerase) were polymorphic. The other locus (Acp, acid phosphatase) was monomorphic. From the polymorphic loci the observed fixation index (F) for the maternal plants, a measure of the deviation in the frequency of heterozygotes from that in a randomly mating population, was calculated by means of the MLT program of Ritland (1989). The expected outcrossing rate was estimated as t = (1 - F)/(1 + F) (Wright 1969; Brown 1979). This method assumes, however, that the population is at an inbreeding equilibrium.
The second method made use of a difference in seed set between male-sterile and hermaphrodite siblings. Progeny of five plants, expected to segregate both male steriles and hermaphrodites in a 1:1 ratio, were grown in the greenhouse until their sex type could be determined. Next, siblings of opposite sex type were paired, and transplanted into the field site at a distance of 25 cm apart. In total, 20 pairs were randomly distributed over the field site.
Male steriles are obligate outcrossers, while hermaphrodites are capable of self-fertilization. Hand pollinations in the greenhouse (Koelewijn, unpubl. data) had shown no inherent differences in seed set between the sex types, and that in hermaphrodites the seed set upon selfing was not different from outcrossing. Because either sex type produces five ovules per flower, a difference in seed set between the two sex morphs in favor of hermaphrodites might be caused by selfing and thus sets an upper limit to the outcrossing rate. Data were analyzed by a paired t-test on the maternal means with the alternative hypothesis (H1): difference in seed set H - MS [greater than] 0 (Sokal and Rohlf 1981). The outcrossing rate was estimated as MS seed set/H seed set.
Moreover, Wolff (unpubl. data) provided results from a survey in 1986 from another site in this population. One ripe spike each was sampled from seven plants. Seedlings were grown in the greenhouse and screened for three loci (6Pgd, Idh, and Pgm). Outcrossing rates per plant, based on 10 progeny and one to three loci, were estimated using the multilocus estimation procedure of Shaw et al. (1981; cf. Wolff et al. 1988).
In August 1989, seeds of each mother-inbreeding level combination (n = 40, eight maternal parents times five inbreeding levels) were placed in petri dishes with moist filter paper and germinated under natural light conditions in the greenhouse. There was no difference in seed weight ([F.sub.4.35] = 0.27, ns, mean seedweight 0.15 [+ or -] SE 0.02 mg) nor in germination (average germination after seven days 72%; [G.sub.4] = 7.4, ns) among inbreeding levels. After 14 days, 40 seedlings were placed in small jiffy pots (2 cm x 2 cm), filled with dune sand poor in humus, and precultured outdoors for one month. Remaining seedlings were used to estimate seedling dry weight. There was a significant difference in seedling weight among the inbreeding levels ([F.sub.4,57] = 3.4, P [less than] 0.05). This difference, however, was only caused by S2-seedlings (seedling weight 0.29, 0.28, 0.30, 0.23, and 0.28 mg for seedlings with inbreeding coefficients, f of 0, 0.25, 0.50, 0.75, and 0.875, respectively) and had no influence on the results when using seedling weight as a covariate. Seedlings were transferred to the field in the beginning of October. The experiment was carried out at the Westplaat, a nature reserve in the southwestern part of The Netherlands, at the site from which the maternal parents originated. A seemingly homogeneous part of the salt meadow of 10 m x 10 m was selected and divided in 16 rectangular plots (2 m x 1 m) laid out in a four (columns) by four (rows) design. Each maternal plant was represented in each plot by two seedlings per inbreeding level, except for the highest inbreeding level (f = 0.875), which was represented by one seedling. Thus, a total of 32, or 16 seedlings per maternal plant per inbreeding level, were used. Seedlings were randomized within a plot and placed at 10-cm intervals in a rectangular fashion (-square-lattice array) and checked monthly to record survival. In spring 1990, one plot was destroyed by rabbits. In May 1990, the plants from three other plots were carefully dug up to estimate shoot and root biomass at the beginning of the flowering season. The remaining 12 plots were followed weekly from the onset of flowering in 1990 (and twice a month in 1991) to record plant presence and spike formation and development. From each flowering plant at least one ripe spike was sampled to estimate seed production. In 1990 detailed observations were made on these spikes (length of the spike, seed set, seed weight, and flowers per mm spike), while in 1991 only the total number of seeds was counted. The experiment was terminated by the end of 1991, due to the low number of surviving plants.
Because of a shortage of seeds, only six of the eight maternal parents and only four of the five inbreeding levels could be used for a greenhouse experiment. For each inbreeding level (f = 0, 0.25, 0.5, and 0.75) and maternal parent, seeds were germinated in petri dishes. After two weeks, 40 seedlings from each maternal parent (10 per inbreeding level) were transplanted into individual pots (9-cm square pots filled with dune sand) and grown in a randomized block design in the greenhouse at a constant temperature of 20 [degrees] C and relative humidity of 70%. Plants were watered daily and received 100 mL of a half-strength Hoagland nutrient solution each month. Each block contained two individuals per mother-inbreeding level combination. The plants were checked daily, and survival and the start of flowering were noted. Plants were harvested after 16 weeks. At harvest, the plants were cut at the soil level and separated in root, shoot, and reproductive biomass. Next, plants were dried at 70 [degrees] C to constant weight, and weighed to the nearest 1.0 mg. Seed production was estimated by multiplying the number of spikes X spike length (mm) X number of flowers per mm spike X seed set per flower.
Although transplant experiments are commonly used to study local adaptation (van Tienderen 1992), some difficulties arise in analysis and interpretation. Because of the death of plants during the experiments, the experimental design can become unbalanced or even worse, incomplete. This is especially true for traits related to reproduction, as only a fraction off all plants produce seeds. Therefore, the field experiment was designed as a randomized plot design with two observations per plot. This increases the probability of getting at least one observation per mother-cross-type combination per plot and allows a more accurate estimate of the experimental error (Steel and Torrie 1980).
Analysis of categorical variables was performed by fitting linear models with GLIM (Baker 1987). Binary response variables were analyzed with a logit-link model. [[Chi].sup.2]-tests were used to assess whether dropping terms from the model significantly reduced the explained variance (McCullagh and Nelder 1983).
A full factorial design including plot (P, random effect), inbreeding level (I, fixed effect), maternal plant (M, random effect), and inbreeding level-by-maternal plant (I X M, random effect) was employed for the reproductive traits measured in 1990 and in the combined dataset, because the design was sufficiently balanced. The residual error consisted of a sampling error (within-plot variance) and experimental error, the latter being the one used for testing the significance of effects (Steel and Torrie 1980). To prevent missing cells, the two highest inbreeding levels were grouped together in the ANOVA. In 1991 the design had become severely unbalanced due to the death of plants and low percentage of flowering plants that year. Consequently, only one-way ANOVAs for each main effect separately were employed. Sex phenotype (hermaphrodite, H; partially male sterile, PMS; or male sterile, MS, defined according to the criteria in Koelewijn and van Damme 1996) was used as a covariate, because of consistent differences among these sex types in reproductive characteristics (Koelewijn 1996). When appropriate, variables were transformed prior to analysis to improve normality or homogeneity of variance.
Epistasis among (partially) deleterious alleles can affect the shape of the relationship between fitness components and inbreeding level. If the effect of homozygous loci acts in a multiplicative fashion, a linear relationship is expected between the log of fitness and the expected inbreeding coefficient. Synergistic or reinforcing epistasis can be detected as a significant nonlinear negative relationship when log-transformed data is regressed on the inbreeding coefficient. Significant nonlinear positive relationships can be due to either diminishing epistasis and/or purging of partially deleterious alleles (Crow and Kimura 1970; Willis 1993). Epistasis at the population level can therefore be detected by calculating the linear (f) and quadratic ([f.sup.2]) coefficients of the quadratic regression of (log)-fitness on inbreeding level. At the family level, epistasis can be detected by a homogeneity of slopes ANOVA. To examine the effect of inbreeding on the two major components of fitness (probability of flowering, seed production per flowering plant, and their product), I therefore performed an ANCOVA with maternal parent as class variable and f and [f.sup.2] as continuous variables. This ANCOVA approach allowed me to examine for main effects of maternal parent, f, [f.sup.2], and interactions between maternal parent and either f or [f.sup.2]. A significant effect of [f.sup.2] indicates a nonlinear relationship at the population level and significant interaction terms indicate differences among maternal parents in their response to increasing homozygosity.
For the greenhouse experiment, a mixed-model ANOVA was used to analyze the influence of inbreeding level (fixed effect) and maternal parent (random effect) on the number of days to first flower, number of spikes at harvest, and the biomass estimates. Because two observations per block were made the residual error was subdivided into an experimental (i.e., variance due to block by treatment interactions) and sampling error (i.e., within-plot variance) (Steel and Torrie 1980).
Cumulative fitnesses for inbred and outcross progeny from each maternal parent were calculated as the the product of mean fitnesses from survival, flowering, and seed production. Inbreeding depression ([Delta]) per inbreeding level was then estimated as:
[Delta] = 1 - [w.sub.s]/[w.sub.o],
where [w.sub.s] and [w.sub.o] are the mean fitnesses of inbred and outcrossed progeny calculated from the eight maternal sibmeans of each crosstype, respectively (cf. Johnston and Schoen 1994). To estimate inbreeding depression per maternal parent, the fitness of outcrossed progeny was compared with the mean fitnes of the four nonoutcrossed treatments.
Seed set per flower in hermaphrodites was consistently higher than in male steriles (H: 3.95 [+ or -] 0.20; MS: 3.35 [+ or -] 0.18; difference H - MS: 0.60 [+ or -] 0.20, paired t-test = 6.44, P = 0.0015; mean [+ or -] SD, n = 5) leading to an estimated outcrossing rate ([t.sub.m] [+ or -] SD) of 0.85 [+ or -] 0.12 (range: 0.78-0.91). The observed fixation index (F) for maternal plants was 0.099 [+ or -] 0.125 SD, indicating a slight excess of homozygotes and leading to an expected outcrossing rate of 0.82. Wolff (unpubl. data) estimated the multilocus outcrossing rate in this same population as 0.76 [+ or -] 0.21 SD (range: 0.37-1.00).
Growth and Survival
Mortality varied among inbreeding levels, and in time, with the highest mortality occurring during the end of the summer [ILLUSTRATION FOR FIGURE 1A OMITTED]. Mortality was positively correlated with inbreeding level and differences among inbreeding levels gradually increased with time, because in every season the highest inbreeding levels performed worst, mounting up to 7% (f = 0.25), 26% (f = 0.50), 59% (f = 0.75), and 48% (f = 0.875) inbreeding depression for survival at the end of the second flowering season ([ILLUSTRATION FOR FIGURE 1B OMITTED], Table 1C). Log-linear analysis detected significant differences in mortality among plots ([[[Chi].sup.2].sub.11] = 79.1, P [less than] 0.001), indicating spatial variation, among maternal plants ([[[Chi].sup.2].sub.7] = 54.5, P [less than] 0.001) and among inbreeding levels ([[[Chi].sup.2].sub.4] = 118.9, P [less than] 0.001). Moreover, a significant inbreeding-by-maternal plant interaction ([[[Chi].sup.2].sub.28] = 78.7, P [less than] 0.001) was observed, indicating differences in inbreeding depression for survival among maternal plants.
Growth was negatively related to the inbreeding level, with plants from the two highest inbreeding levels (f = 0.75 and 0.875) having the lowest shoot and root biomass (Table 2).
Flowering and Reproduction
Large differences in flowering and reproduction between 1990 and 1991 were observed (Table 1A,B). In 1990 the percentage of flowering plants was 70% and the average seed yield per flowering plant was 136, while for 1991 these figures were 27% and 53. Table 1 shows a detailed analysis of the traits contributing to differences in seed yield. Many statistical tests are shown, not all mutually independent, but the conclusion for the main fitness components, that is, probability of survival, probability of flowering, and seed production per flowering plant, are that inbreeding level, maternal plant, and their interaction have an effect on progeny fitness. Moreover, plot effects were always detectable, indicating small-scale local variation. Inbreeding depression in cumulative seed production after two years mounted up to 34% (f = 0.25), 39% (f = 0.50), 69% (f = 0.75), and 80% (f = 0.875) (Table 1C, [ILLUSTRATION FOR FIGURE 2 OMITTED]). For the two highest inbreeding levels (f = 0.75 and 0.875) the reduction in fitness was the cumulative result of inbreeding effects in survival, probability of flowering, and seed production. For plants from the other two inbreeding levels, the reduction in fitness was mainly caused by a reduction in seed production (Table 1).
At the population level, the quadratic regression analysis of fitness components on expected inbreeding coefficient revealed a significant negative quadratic coefficient for the probability of flowering at least once during the two field seasons (Tables 1C, 3). Seed yield per flowering plant was negatively related to inbreeding level, but no negative quadratic component was observed (Tables 1C, 3). Final fitness, that is, seed yield corrected for survival and flowering, was also negatively related to inbreeding level and had a negative, though not significant, quadratic component (Tables 1C, 3; [ILLUSTRATION FOR FIGURE 2 OMITTED]). [TABULAR DATA FOR TABLE 1 OMITTED] Thus, significant inbreeding depression was detected in these analyses, concurrent with the ANOVA, and synergistic epistasis was detectable for the probability of flowering.
Inbreeding depression among maternal parents, calculated by comparing the fitness of outcrossed progeny versus mean progeny fitness of the four inbred groups together, varied significantly for probability of flowering (range: 0.00-0.44), seed production per flowering plant (range: 0.08-0.67), and seed production corrected for survival and flowering (range: 0.01-0.76; [ILLUSTRATION FOR FIGURE 3 OMITTED]). Responses to increased inbreeding were also variable, with one maternal family exhibiting no significant inbreeding depression (family 4, [ILLUSTRATION FOR FIGURE 3 OMITTED]), one family showing a strong immediate response, and no increase in inbreeding depression afterward (family 7, [ILLUSTRATION FOR FIGURE 3 OMITTED]), while others showed a steady increase in inbreeding depression during the breeding program that was most pronounced in the later stages [ILLUSTRATION FOR FIGURE 3 OMITTED]. Significant interactions between maternal parent and f or [f.sup.2] were detected for the probability of flowering and seed production per flowering plant. These interactions occurred because some families exhibited positive linear and quadratic coefficients, whereas others exhibited negative linear and quadratic coefficients. The magnitudes of these relationships varied among families [ILLUSTRATION FOR FIGURE 3 OMITTED] and suggest that epistasis varies strongly among families.
Results from the greenhouse were in strong contrast to the field. Probability of flowering and survival was 100% and, except for the date of first flowering (35 [+ or -] 1.0, f = 0; 39 [+ or -] 1.8, f = 0.25; 42 [+ or -] 2.3, f = 0.50; and 43 [+ or -] 2.7, f = 0.75, days after transfer to the pots, [F.sub.3,15] = 3.44, P [less than] 0.05; mean [+ or -] 1 SD) no significant differences among inbreeding levels for the other traits (e.g., shoot, root, and reproductive biomass; number of spikes; spike length) were observed. The [TABULAR DATA FOR TABLE 2 OMITTED] number of flowers per mm spike (0.92) and seed set per flower (4.3) were also not different among inbreeding levels. Consequently, no significant inbreeding depression ([Delta] = 0.08, 0.13, and 0.11 for f = 0.25, 0.5, and 0.75, respectively) and no negative relationship between seed production and inbreeding level was observed [ILLUSTRATION FOR FIGURE 3 OMITTED].
The results of these multigenerational inbreeding experiments demonstrate substantial inbreeding depression for important fitness components in the field, but not in the greenhouse. Thus, the expression of genetic load was strongly influenced by the environment. At the population level, probability of flowering, survival, and seed production per flowering plant all showed a negative relationship with increased inbreeding. Though in general also negative, the response of individual plants to extended inbreeding was highly variable, and, consequently, substantial variation in inbreeding depression among these plants was detected. Epistasis was detected at the family level for all traits, but at the population level only for the probability of flowering. Below I discuss these findings in the context of mating-system evolution.
Population-Level Inbreeding Depression
Inbreeding depression is frequently observed in predominantly outcrossing species (e.g., Barrett and Kohn 1991; Husband and Schemske 1996). [TABULAR DATA FOR TABLE 3 OMITTED] Evidence for considerable inbreeding depression also in highly selfing species is now accumulating (e.g., Agren and Schemske 1993; Latta and Ritland 1994; Husband and Schemske 1996). In outcrossing populations, mildly and highly deleterious mutations and polymorphisms for overdominant alleles may all contribute to inbreeding depression, while in highly selfing populations, mildly deleterious recessive mutations are expected to be the main cause of inbreeding depression (B. Charlesworth et al. 1990). Lande and Schemske (1985) suggested that continued inbreeding should result in purging of highly deleterious alleles, while the component of the load due to mildly deleterious mutations or to polygenic mutations is hard to purge. Predominantly self-fertilizing species should therefore display a lower level of inbreeding depression than predominantly outcrossing species. Also, if recessive mutations are the main cause of inbreeding depression, a general negative relationship between inbreeding depression and selfing rate is expected. No such negative relationship is expected under the overdominance model, and inbreeding depression might even increase with increased selfing.
Available data suggest that partial dominance accounts for most of the observed inbreeding depression (Wright 1977; Charlesworth and Charlesworth 1987; Johnston and Schoen 1995; but see Fu and Ritland 1994). In a review of the literature, Husband and Schemske (1996) observed the magnitude of inbreeding depression in selfing species to be less than outcrossing species. They attributed this result to more efficient purging of genetic load in the selfing taxa. Studies comparing populations of related species that differ in their selfing rate or inbreeding coefficient find, in general, negative associations between inbreeding depression in fitness components and the selfing rate or inbreeding coefficient (Holtsford and Ellstrand 1990; Latta and Ritland 1994; Johnston and Schoen 1997) and relate this result to purging of deleterious alleles. Multigenerational studies of controlled inbreeding provide a less clear answer. Schoen (1983), Dudash (1990), and Latta and Ritland (1994) found no differences between one and two consecutive generations of selfing in Gilia angustifolica, Sabattia angularis, and Mimulus nasutus, respectively. In contrast, fitness declines over consecutive generations of inbreeding have been found in maize (Jones 1939), faba beans (Monti and Fruscante 1984), rapeseed (Schuster and Michael 1976), Mimulus guttatus (Dudash et al. 1996), and Tribolium (Pray and Goodnight 1995), and have been attributed to the continuous expression of recessive deleterious alleles.
The population-level analysis of my data is congruent with the view that mutation to mildly deleterious alleles is the primary cause of inbreeding depression. First, the magnitude of inbreeding depression in fitness for P. coronopus in the field after one generation of selfing ([Delta] = 0.39, f = 0.5) is in between the observed mean values for selfing ([Delta] = 0.23) and outcrossing species ([Delta] = 0.53) (Husband and Schemske 1996). Second, a strong negative correlation (r = -0.96, n = 4, P = 0.045) between relative fitness and inbreeding coefficient [ILLUSTRATION FOR FIGURE 2 OMITTED] and no evidence for strong purging in the first generations of inbreeding was observed in the field. These results might be expected for a partially selfing species with a selfing rate of about 0.25. This selfing rate should be sufficient for purging the load due to highly deleterious alleles (cf. Lande and Schemske 1985), while the load due to mildly recessive deleterious alleles might still be around and expressed under serial inbreeding.
Family-Level Inbreeding Depression
Although P. coronopus suffers from considerable inbreeding depression, it is still a partially selfing species with a mixed mating system. Most models that assume a constant value of inbreeding depression predict for the evolution of mating systems either selection for complete selfing or complete outcrossing, depending on whether the inbreeding depression exceeds one-half (e.g., Lande and Schemske 1985, B. Charlesworth et al. 1991). My results indicate that with low levels of inbreeding (f [less than or equal to] 0.5) the inbreeding depression is below the threshold of one-half, while with high levels of inbreeding it is above the threshold. Thus, the population would be capable of evolving in either direction.
Modeling work by Holsinger (1988, 1991), Uyenoyama and Waller (1991a,b) and Uyenoyama et al. (1993), however, has demonstrated that population-level estimates of inbreeding depression alone are insufficient to predict whether a selfing variant can invade in a particular population. They indicated the importance of determining family-level variation in inbreeding depression. If inbreeding depression estimates within maternal lines and mating system are linked, this may be more relevant in predicting the evolution of mating systems within a population than overall population inbreeding depression and mating-system estimates (Holsinger 1991). Large variation among families in inbreeding depression could maintain mixed mating systems within populations by favoring outcrossing in a frequency-dependent manner even if selfed progeny from outbred parents suffer a fitness reduction less than one-half (Uyenoyama et al. 1993).
My results indicate significant differences among maternal lines in performance, magnitude of inbreeding depression for probability of flowering and seed production, and response to increased inbreeding [ILLUSTRATION FOR FIGURE 3 OMITTED], thereby supporting the theoretical work mentioned above. Both, plants with either high ([Delta] = 0.76) or low ([Delta] = 0.01) inbreeding depression, were found within a 50-[m.sup.2] area. The expression of genetic load during the breeding program varied strongly among maternal plants, but the general response (except for maternal family 4, [ILLUSTRATION FOR FIGURE 3 OMITTED]) was a decline in performance, suggesting that dominance through purging of partially recessive deleterious alleles underlies the expression of inbreeding depression. The difference among maternal plants might reflect the unknown genetic history of the plants and indicate local genetic variation. Variation among maternal plants in inbreeding depression is expected if plants differ in the number of recessive deleterious alleles that they carry. This could be influenced by either their history of inbreeding or by differences in mutation accumulation among individuals (Schultz and Willis 1995). The parental fixation index (F = 0.09), however, did not indicate that inbreeding had a strong effect on the population structure.
Significant variation in inbreeding depression among maternal plants after one generation of selfing has also been detected in Costus allenii (Schemske 1983), Schiedea salicaria (Sakai et al. 1989), Begonia hirsuta and B. semiovata (Agren and Schemske 1993) and Epilobium angustifolium (Husband and Schemske 1994), but was not found in C. laevis (Schemske 1983) and Lobelia species (Johnston 1992). My results are in close agreement with two other multigenerational inbreeding studies (Pray and Goodnight 1995; Dudash et al. 1997). Both found variation among maternal families in inbreeding depression and response to inbreeding. However, my results do differ from theirs in that I did not find any maternal line improvement across the breeding program, though in some families highly inbred individuals (f = 0.875) outperformed their outcrossed relatives. Maternal line improvement would provide a mechanism for the invasion of a selfing variant into the population through any maternal family exhibiting purging of its genetic load.
Maynard Smith (1977, 1978) has emphasized that only looking at the consequences of one generation of selfing might not be sufficient for a proper evaluation of the consequences of inbreeding. He developed a model for the evolution of sex or outcrossing in which multigenerational inbreeding is considered. Applying his model to the P. coronopus data predicts that selection would favor a population with either genotypes that permit some proportion of selfing or a population with a genetic polymorphism for selfing. The first scenario might well apply to the current study because variation in the outcrossing rate among individual plants was found. However, more accurate estimates of the selfing rate of individual plants, in connection with inbreeding depression estimates on the same plants, are necessary for a justified interpretation (cf. Carr et al. 1997).
At the population level, I detected evidence of synergistic epistasis for the probability of flowering, but not for the seed production per flowering plant. The combined fitness estimate did have a negative, though not significant, quadratic regression component, suggesting weak synergism (Table 3). At the family level, I observed both negative and positive curvilinear responses as revealed by the significant maternal parent by [f.sup.2] interaction, suggesting that both synergistic and diminishing epistasis could occur. Weak interactions, while difficult to detect experimentally, can have important influences on evolutionary processes. Epistasis influences the rate at which a population purges its load with inbreeding. Theoretical studies of the effect of synergism on the evolution of sexual systems have shown that even very slight epistasis between harmful mutations can have considerable effects on the maintenance of sexual reproduction and outcrossing (B. Charlesworth 1990; B. Charlesworth et al. 1991). The decrease in the probability of flowering for plants with a f [greater than] 0.5 was large (Table 1C). It implies that the majority of the plants with high inbreeding coefficients, though present in the population in a vegetative state, do not take part in reproduction. These plants are likely to have more homozygous loci and a higher genetic load. The presence of synergistic epistasis therefore suggests that the rate of purging might be higher than if homozygous loci acted independently.
Experimental detection of epistasis at the population level has been difficult. Willis (1993) detected inbreeding depression for many traits in M. guttatus, but found evidence for synergistic epistasis in pollen viability only. Pray and Goodnight (1995) and Dudash et al. (1997) observed no overall population effect of epistasis, but detected both positive and negative curvilinear responses among lineages. They suggested that the effects of epistasis may not be evident unless one examines epistasis at the family level within a population. The response of individual plants to increased inbreeding for final seed production [ILLUSTRATION FOR FIGURE 3 OMITTED] was indeed variable. One plant suggested diminishing epistasis (M7, [ILLUSTRATION FOR FIGURE 3 OMITTED], while four others represented synergistic epistasis (M2, M3, M5, and M8, [ILLUSTRATION FOR FIGURE 3 OMITTED]). Consequently, at the population level only a weak negative response was detected. In contrast, for probability of flowering, seven of the eight maternal parents showed a negative curvilinear response, resulting in a strong negative response at the population level too.
Detection of synergistic epistasis might also be difficult because the expression of genetic load is influenced by the environment. The studies by Willis (1993), Pray and Goodnight (1995), and Dudash et al. (1997) were done in greenhouses and not under competitive conditions, where synergism between deleterious mutations is more likely to be expressed (Hamilton et al. 1990). The reason for this would be that limitation of space and nutrients causes truncation-like selection under such conditions. My results seem to confirm to this pattern; only under field conditions was there evidence for synergism.
The results showed that the expression of genetic load was influenced by the environment. No significant inbreeding depression and response to increased inbreeding was observed in the greenhouse, either at the population or family level. In the field, however, pronounced differences among plants with different inbreeding coefficients and a general decline in fitness with increased inbreeding was observed. Several authors have noted that inbreeding depression is more severe in stressful or competitive environments (e.g., Dudash 1990; Barrett and Kohn 1991; Wolfe 1993; Eckert and Barrett 1994; Latta and Ritland 1994). Greenhouse studies, with their benign environment, may underestimate the component of [Delta] expressed during severe episodes of selection that occur in the field, like overwintering or drought resistance. Johnston (1992), however, found larger differences in flower number between selfs and outcrosses of Lobelia siphilatica and L. cardinalis in the greenhouse compared to the field. He argued that there might be more opportunity for selfing and outcrossing to differ in the greenhouse because outcrossed progeny will exhibit a proportionally greater response to increasing environmental quality. Nevertheless, most studies conducted in both environments found higher inbreeding depression in the field. Therefore, measurements should preferably be taken at the site of origin.
The contrast between field and greenhouse has an important drawback for the interpretation of results with respect to mating-system evolution. based on the greenhouse results, P. coronopus would be expected to evolve toward more selfing, because no inbreeding depression was detected. In contrast, the field results predict a much more dynamic picture. The observed inbreeding depression at the population level is such that P. coronopus could evolve either toward selfing or outcrossing. However, the large variation in inbreeding depression among maternal families together with differences in epistasis make it much more likely that the species will keep its mixed mating system. Under these conditions, associations between inbreeding depression and mating system loci are more likely to occur, making it easier for a selfing variant to invade or be maintained in the population (Holsinger 1991).
Thanks to J. van Damme, W. van Delden, K. Holsinger, M. Johnston, and an anonymous referee for comments on the manuscript, and to K. Wolff for providing unpublished data. This research was supported by a BION-NWO (Netherlands Organisation for Scientific Research) graduate scholarship. This is publication number 2376, Netherlands Institute of Ecology, Centre for Terrestrial Ecology, Heteren.
AGREN, J., AND D. W. SCHEMSKE. 1993. Outcrossing rate and inbreeding depression in two annual monoecious herbs, Begonia hirsuta and B. semiovata. Evolution 47:125-135.
AIDE, T. M. 1986. The influence of wind and animal pollination on variation in outcrossing rates. Evolution 40:434-435.
BAKER, R. J. 1987. GLIM 3.77. Reference manual. Numerical Algorithms Group, Oxford.
BARRETT, S. C. H., AND C. G. ECKERT. 1990. Variation and evolution of mating systems in seed plants. Pp. 229-254 in S. Kawano, ed. Biological approaches and evolutionary trends in plants. Academic Press, Tokyo.
BARRETT, S. C. H., AND J. R. KOHN. 1991. Genetic and evolutionary consequences of small population size in plants: implications for conservation. Pp. 3-30 in D. A. Falk and K. E. Holsinger, eds. Genetics and conservation of rare plants. Oxford Univ. Press, Oxford.
BROWN, A. H. D. 1979. Enzyme polymorphism in plant populations. Theor. Popul. Biol. 15:1-42.
CAMPBELL, R. B. 1986. The interdependence of mating structure and inbreeding depression. Theor. Popul. Biol. 30:232-244.
CARR, D. E., C. B. FENSTER, AND M. R. DUDASH. 1997. The relationship between mating-system characters and inbreeding depression in Mimulus guttatus. Evolution 51:363-372.
CHARLESWORTH, B. 1990. Mutation-selection balance and the evolutionary advantage of sex and recombination. Genet. Res. 55: 199-221.
CHARLESWORTH, B., D. CHARLESWORTH, AND M. T MORGAN. 1990. Genetic loads and estimates of mutation rates in highly inbred populations. Nature 347:380-382.
CHARLESWORTH, B., M. T. MORGAN, AND D. CHARLESWORTH. 1991. Multilocus models of inbreeding depression with synergistic selection and partial self-fertilization. Genet. Res. 57:177-194.
CHARLESWORTH, D. 1988. A method for estimating outcrossing rates in natural populations of plants. Heredity 61:469-472.
CHARLESWORTH, D., AND B. CHARLESWORTH. 1979. The evolutionary genetics of sexual systems in flowering plants. Proc. R. Soc. Lond. B 205:513-530.
-----. 1987. Inbreeding depression and its evolutionary consequences. Annu. Rev. Ecol. Syst. 18:237-268.
CHARLESWORTH, D., M. T. MORGAN, AND B. CHARLESWORTH. 1990, Inbreeding depression, genetic load, and the evolution of outcrossing rates in a multi-locus system with no linkage. Evolution 44:1469-1489.
-----. 1992. The effect of linkage and population size on inbreeding depression due to mutational load. Genet. Res. 59:4961.
CROW, J. F. 1970. Genetic loads and the cost of natural selection. Pp. 128-177 in K. I. Kojima, ed. Mathematical models in population genetics. Springer, Berlin.
CROW, J. F., AND M. KIMURA. 1970. An introduction to population genetic theory. Harper and Row, New York.
DARWIN, C. 1876. The effects of cross and self-fertilisation in the vegetable kingdom. John Murray, London.
DUDASH, M. R. 1990. Relative fitness of selfed and outcrossed progeny in a self-compatible, protandrous species, Sabatia angularis L (Gentianaceae): a comparison in three environments. Evolution 44:1129-1139.
DUDASH, M. R., D. E. CARR, AND C. B. FENSTER. 1997. Five generations of enforced selfing and outcrossing in Mimulus guttatus: inbreeding depression variation at the population and family level. Evolution 51:54-65.
ECKERT, C. G., AND S. H. BARRETT. 1994. Inbreeding depression in partially self-fertilizing Decodon verticillatus: population-genetic and experimental analysis. Evolution 48:952-964.
FISHER, R. A. 1941. Average excess and average effect of a gene substitution. Ann. Eugen. 11:53-63.
FRANKEL, R. 1983. Heterosis: reappraisal of theory and practice. Springer Verlag, Berlin.
FU, Y. B., AND K. RITLAND. 1994. Evidence for the partial dominance of viability genes contributing to inbreeding depression in Mimulus guttatus. Genetics 136:323-331.
GRANT, V. 1975. Genetics of flowering plants. Columbia Univ. Press, New York.
HAMILTON, W. D., R. AXELROD, AND R. TANESE. 1990. Sexual reproduction as an adaptation to resist parasites (a review). Proc. Nat. Acad. Sci. USA 87:3566-3573.
HOLSINGER, K. E. 1988. Inbreeding depression doesn't matter: the genetic basis of mating-system evolution. Evolution 42:12351244.
-----. 1991. Inbreeding depression and the evolution of plant mating systems. Trends Ecol. Evol. 6:307-308.
HOLTSFORD, T. P., AND N. C. ELLSTRAND. 1990. Inbreeding effects in Clarkia tembloriensis (Onagraceae) populations with different natural outcrossing rates. Evolution 44:2031-2046.
HUSBAND, B. C., AND D. W. SCHEMSKE. 1994. Magnitude and timing of inbreeding depression in a diploid population of Epilobium angustifolium (Onagracea). Heredity 75:206-215.
-----. 1996. Evolution of the magnitude and timing of inbreeding depression in plants. Evolution 50:54-70.
JAIN, S. K. 1976. Evolution of inbreeding in plants. Annu. Rev. Ecol. Syst. 7:469-495.
JOHNSTON, M. O. 1992. Effects of cross and self-fertilization on progeny fitness in Lobelia cardinalis and L. siphilitica. Evolution 46:688-702.
JOHNSTON, M. O., AND D. J. SCHOEN. 1994. On the measurement of inbreeding depression. Evolution 48:1735-1741.
-----. 1995. Mutation rates and dominance levels of genes affecting total fitness in two angiosperm species. Science 267:226229.
-----. 1996. Correlated evolution of self-fertilization and in breeding depression: an experimental study of nine populations of Amsinckia (Boraginaceae). Evolution 50:1478-1491.
JONES, D. F. 1939. Continued inbreeding in maize. Genetics 24: 462-473.
KOELEWIJN, H. P. 1996. Sexual differences in reproductive characters in gynodioecious Plantago coronopus. Oikos 75:443-452.
KOELEWIJN, H. P., AND J. M. M. VAN DAMME. 1995. Genetics of male sterility in Plantago coronopus. I. Cytoplasmic variation. Genetics 139:1749-1758.
-----. 1996. Gender variation, partial male sterility and labile sex expression in gynodioecious Plantago coronopus. New Phytol. 132:67-76.
KONDRASHOV, A. S. 1982. Selection against harmful mutations in large sexual and asexual populations. Genet. Res. 40:325-332.
-----. 1988. Deleterious mutations and the evolution of sexual reproduction. Nature 336:435-440.
LANDE, R., AND D. W. SCHEMSKE. 1985. The evolution of self-fertilization and inbreeding depression in plants. I. Genetic models. Evolution 39:24-40.
LATTA, R., AND K. RITLAND. 1994. The relationship between inbreeding depression and prior inbreeding among populations of four Mimulus taxa. Evolution 48:806-817.
LLOYD, D. G. 1979. Some reproductive factors affecting the selection of self-fertilization in plants. Am. Nat. 113:67-97.
-----. 1980. Benefits and handicaps of sexual reproduction. Pp. 69-112 in M. K. Hecht, W. C. Steere, and B. Wallace, eds. Evolutionary biology. Vol. 13. North-Holland Publishing Company, Amsterdam.
MAYNARD SMITH, J. 1977. The sex habit in plants and animals. Pp. 315-331 in F. B. Christensen and T. M. Fenchel, eds. Measuring selection in natural populations. Springer Verlag, Berlin.
-----. 1978. Evolution of sex. Cambridge Univ. Press, Cambridge, U.K.
MCCULLAGH, P., AND J. A. NELDER. 1983. Generalized linear models. Chapman and Hall, London.
MONTI, L. M., AND L. FRUSCIANTE. 1984. Selection methods for yield improvement in faba beans. Pp. 193-200 in P. D. Hebblethwaite, T. C. K. Dawkins, and M. C. Heath, eds. Vicia faba: agronomy, physiology and breeding. Junk Publishers, The Hague.
NAGYLAKI, T. 1976. A model for the evolution of self-fertilization and vegetative reproduction. J. Theor. Biol. 58:55-58.
PRAY, L. A., AND C. J. GOODNIGHT. 1995. Genetic variation in inbreeding depression in the red flour beetle Tribolium castaneum. Evolution 49:176-188.
RICHARDS, A. J. 1986. Plant breeding systems. Unwin and Allen, London.
RITLAND, K. 1989. A series of FORTRAN computer programs for estimating plant mating systems. J. Hered. 81:235-237.
SAKAI, A. K., K. KAROLY, AND S. G. WELLER. 1989. Inbreeding depression in Schiedea globosa and S. salicaria (Caryophyllaceae), subdioecious and gynodioecious Hawaiian species. Am. J. Bot. 76:437-444.
SCHEMSKE, D. W. 1983. Breeding system and habitat effects on fitness components in three neotropical Costus (Zingiberaceae). Evolution 37:523-539.
SCHEMSKE, D. W., AND R. LANDE. 1985. The evolution of self-fertilization and inbreeding depression in plants. II. Empirical observations. Evolution 39:41-52.
SCHMITT, J., AND D. W. EHRHARDT. 1990. Enhancement of inbreeding depression by dominance and suppression in Impatiens capensis. Evolution 44:269-278.
SCHOEN, D. J. 1983. Relative fitnesses of selfed and outcrossed progeny in Gilia achilleifolia (Polemoniceae). Evolution 37:292301.
SCHULZ, S. T., AND J. H. WILLIS. 1995. Individual variation in inbreeding depression: the roles of inbreeding history and mutation. Genetics 141:1209-1223.
SCHUSTER, W., AND J. MICHAEL. 1976. Untersuchungen uber Inzuchtdepression und Heterosiseffekte bei Raps (Brassica napus oleifera). Z. Pflanzenzuecht. 77:56-66.
SHAW, D. V., A. L. KAHLER, AND R. W. ALLARD. 1981. A multilocus estimator of mating system parameters in plant populations. Proc. Nat. Acad. Sci. USA 78:1298-1302.
SHIELDS, W. M. 1982. Philopatry, inbreeding and the evolution of sex. State Univ. of New York Press, Albany, NY.
SOKAL, R. R., AND F. J. ROHLF. 1981. Biometry: the principle and practice of statistics in biological research. W.H. Freeman, New York.
STEEL, R. D. D., AND J. H. TORRIE. 1980. Principles and procedures of statistics. McGraw Hill, Singapore.
UYENOYAMA, M. K., AND D. M. WALLER. 1991a. Coevolution of self-fertilization and inbreeding depression. I. Mutation-selection balance at one and two loci. Theor. Popul. Biol. 40:14-46.
-----. 1991b. Coevolution of self-fertilization and inbreeding depression. II. Symetric overdominance in viability. Theor. Popul. Biol. 40:47-77.
-----. 1991c. Coevolution of self-fertilization and inbreeding depression. III. Homozygous lethal mutations at multiple loci. Theor. Popul. Biol. 40:173-210.
UYENOYAMA, M. K., K. E. HOLSINGER, AND D. M. WALLER. 1993. Ecological and genetic factors directing the evolution of self-fertilization. Oxf. Surv. Evol. Biol. 9:327-338.
VAN DIJK, H. J., K. WOLFF, AND A. DE VRIES. 1988. Genetic variability in Plantago species in relation to their ecology. 3. Genetic structure of populations of P. major, P. lanceolata and P. coronopus. Theor. Appl. Genet. 75:518-528.
VAN TIENDEREN, P. H. 1992. Variation in a population of Plantago lanceolata along a topographical gradient. Oikos 64:560-572.
WILLIS, J. H. 1993. Effects of different levels of inbreeding on fitness components in Mimulus guttatus. Evolution 47:864-876.
WOLFE, L. M. 1993. Inbreeding depression in Hydrophyllum appendiculatum: role of maternal effects, crowding, and parental mating history. Evolution 47:374-386.
WOLFF, K., B. FRISO, AND J. M. M. VAN DAMME. 1988. Outcrossing rates and male sterility in natural populations of Plantago coronopus. Theor. Appl. Genet. 76:190-196.
WRIGHT, S. W. 1969. Evolution and the genetics of populations. Vol. 2. The theory of gene frequencies. Univ. of Chicago Press, Chicago.
-----. 1977. Evolution and the genetics of populations. Vol. 3. Experimental results and evolutionary deductions. Univ. of Chicago Press, Chicago.
|Printer friendly Cite/link Email Feedback|
|Author:||Koelewijn, Hans Peter|
|Date:||Jun 1, 1998|
|Previous Article:||Fine-grained spatial and temporal variation in selection does not maintain genetic variation in Erigeron annuus.|
|Next Article:||Experimental evolution of resistance in Brassica rapa: correlated response of tolerance in lines selected for glucosinolate content.|