Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Jun 5.
Published in final edited form as: Am J Bot. 2010 Jul 26;97(8):1328–1333. doi: 10.3732/ajb.1000020

Embryonic inbreeding depression varies among populations and by mating system in Witheringia solanacea (Solanaceae)1

Judy L Stone 1,2, Emily E Wilson 1, Andrew S Kwak 1
PMCID: PMC3673560  NIHMSID: NIHMS469628  PMID: 21616885

Abstract

Premise of the study

Embryonic inbreeding depression is a key influence on mating system evolution and can be difficult to estimate in self-incompatible species. A pollen chase experiment was used to estimate the magnitude of embryonic inbreeding depression in Costa Rican Witheringia solanacea, a species polymorphic for self-incompatibility (SI). In a pollen chase experiment, bud self-pollinations are followed after anthesis by outcross pollinations, with a comparable pair of outcross pollinations used as a control. Lowered seed set for the self-precedence treatment indicates embryonic inbreeding depression.

Methods

Embryonic inbreeding depression was assayed for self-compatible (SC) individuals and for SI plants from two populations that differ quantitatively in the onset and enzymatic activity of their SI response. Microsatellite markers were used to assay the selfing rate of a sample of surviving progeny from the prior self-pollination treatment.

Key results

SC individuals showed no evidence of embryonic inbreeding depression. In SI plants, prior self-pollination reduced seed number by 28–70%, depending on population. Microsatellite genotyping revealed that embryonic inbreeding depression was even more severe than estimated by the phenotypic data: for mature fruits resulting from self-pollination precedence, the majority of the progeny were the result of outcross fertilization.

Conclusions

Lineage-specific purging of recessive lethals has accompanied the evolution of SC in this species. SI populations show contrasting levels of embryonic inbreeding depression, with nearly complete embryonic lethality upon selfing in the Monteverde population. In the face of high embryonic inbreeding depression, an increase in selfing rate can evidently occur only under severe pollen limitation.

Keywords: Costa Rica, embryonic inbreeding depression, evolution, mating system, self-compatibility, self-incompatibility, Solanaceae, Witheringia solanacea

The transition from genetically enforced self-incompatibility (SI) to a selfing mode of life has been traversed by many plant lineages, perhaps more often than any other evolutionary transition in flowering plants (Stebbins, 1974; Igic et al., 2008). Lineages that make this transition from SI to selfing must cope with initially high inbreeding depression. Early-acting inbreeding depression, that is, inbreeding depression operating during embryogenesis, seed maturation, and germination, is typically much higher in predominantly outcrossing than in predominantly selfing species, whereas late-acting inbreeding depression is similar for species with different breeding systems (Husband and Schemske, 1996). The increased homozygosity caused by selfing can eliminate or purge recessive lethals, reducing inbreeding depression and thereby permitting the spread of selfing variants (Lande and Schemske, 1985; Husband and Schemske, 1996; Crnokrak and Barrett, 2002; but see Byers and Waller, 1999). The ability to purge early-acting inbreeding depression is evidently key in the evolution of selfing from out-crossing (Glémin et al., 2001; Porcher and Lande, 2005), and estimates of its magnitude in SI populations are therefore of interest.

For self-compatible (SC) species, early-acting inbreeding depression can be estimated simply by comparing seed production resulting from self- and outcross-pollinations. This method cannot be used for species with intact SI systems. Selfed seed can be produced by bud pollinations or other methods used to circumvent SI (de Nettancourt, 2000), but the cause of reduced seed set in these cases is due to an unknown combination of residual SI and embryonic inbreeding depression (e.g., Kärkkäinen et al., 1999). Inbreeding depression can be estimated by comparing inbreeding coefficients of adults to juveniles, but this method is recommended only for species with selfing rates of at least 10% (Ritland, 1990). Another approach to estimating inbreeding depression of SI species is to construct outcross matings among relatives (e.g., Vogler et al., 1999). This method has the virtue of permitting a regression of genetic load over inbreeding coefficient; however, it is infrequently used, perhaps due to labor and space requirements.

A pollen chase experiment can be used to estimate the extent of embryonic inbreeding depression for some species with suspected or confirmed SI systems (Krebs and Hancock, 1990; Barrett, 2002). In a pollen chase experiment, self-pollen is applied in the bud, with outcross pollen being applied after anthesis (the SO treatment). As a control, outcross pollen is applied on a paired flower, both in the bud and after anthesis (the OO treatment). To the extent that self-pollen is successful in fertilization, but generates progeny of lowered embryonic fitness, seed set will be lowered in fruits with self-pollen precedence. The reduction in fruit and seed production for SO as opposed to OO treatments can be used as an estimate of lethal inbreeding depression acting at the embryonic stage. If the SI system involves degradation of self-pollen tubes in the style, fluorescence microscopy can be used to observe stylar pollen growth to document the success of bud pollinations in overcoming the SI response. Such data can be used to correct potential errors of interpretation that arise when partial SI persists in the bud (Fig. 1).

Fig. 1.

Fig. 1

Open in a new tab

Unobserved residual self-incompatibility (SI) deflates estimates of embryonic inbreeding depression. Filled circles represent out-crossed ovules and empty circles represent selfed ovules. A line through the circle represents embryonic mortality. In both examples, inbreeding depression for selfed embryos is 0.5. Overall seed production in the self-pollen precedence (SO) treatment is elevated by outcrossed progeny if bud pollination does not completely overcome SI.

Estimates of inbreeding depression in species where variation in SI has been documented can test whether variation in SI corresponds with variation in inbreeding depression (e.g., Vogler et al., 1999; Busch, 2005). We have found both qualitative and quantitative variation in the SI response of Witherin-gia solanacea L’Her, a tropical shrub occurring in Central America. Most individuals are SI, but autogamous SC variants have been found in two very small isolated Costa Rican populations (Stone et al., 2006). An undescribed SC variant occurs sympatrically with a large SI population in the Monteverde region of Costa Rica and in at least one small isolated population farther west (Stone and Jenkins, 2008). In addition, two large SI populations differ quantitatively in the strength of their SI response, with the Monteverde population producing lower levels of stylar RNase and permitting more self-pollen tubes to grow to the base of the style in bud pollinations than the Las Cruces population (Stone et al., 2006).

To gain insight into the mating system evolution of this species, we used a pollen chase experiment to compare the level of early-acting inbreeding depression for plants that varied in SI response. We compared fruit and seed set for SO vs. OO pollinations for SC plants and for plants from the Las Cruces and Monteverde populations that differ quantitatively in the strength of their SI response. For SI plants, we compared germination rates for seeds resulting from the two treatments. We also genotyped seedlings at five microsatellite loci to estimate selfing rate of surviving progeny, which we used in concert with previously published pollen tube data to estimate the relative fitness of selfed and outcrossed embryos. We hypothesized that selfing lineages would show decreased inbreeding depression due to purging of genetic load. We also hypothesized that the two SI populations might show differences in inbreeding depression corresponding to quantitative variation in the SI response. If the less robust SI system in Monteverde has led to a history of selfing, that population might show lower inbreeding depression due to partial purging of genetic load.

MATERIALS AND METHODS

Study species and populations

Witheringia solanacea is a small shrub that extends from southern Mexico and the Caribbean to South America (D’Arcy, 1973). It possesses the gametophytic SI system characteristic of the Solanaceae, where allele-specific S-RNases expressed in the style cause the degradation of RNA, and therefore the death of pollen tubes bearing the cognate S-allele (McClure et al., 1990). The floral biology of W. solanacea is described in detail by Bohs (2000). Its flowers are pollinated by small bees and its fleshy fruits are bird-dispersed. Witheringia solanacea also reproduces prolifically by vegetative reproduction in gardens and along trailsides. Cuttings were collected in the field from Costa Rican populations and maintained in the greenhouse at Colby College. Populations and locations have been described by Stone and Pierce (2005) and Stone et al. (2006). Plants were sampled from populations in three geographic regions. Self-incompatible plants were sampled from large populations in the Las Cruces region, at 1180–1600 m a.s.l. in southwestern Costa Rica, and the Monteverde region in northwestern Costa Rica, extending from 1100 m elevation in the San Luis Valley to 1500 m a.s.l. in the village of Monteverde. The SC individuals came from the very small population at Vara Blanca at 1360 m a.s.l. in central Costa Rica and from the undescribed but morphologically distinct variety at lower elevations in the Monteverde region. Treatments were carried out on seven SI individuals from Las Cruces and 13 from Monteverde, in addition to the two strongly SC individuals available from Vara Blanca and three from Monteverde. The low sample size of SC plants was a consequence of the relative scarcity of the SC phenotype in Monteverde and the tiny overall size of the population in Vara Blanca, which over a decade of monitoring has never been found to contain more than six individuals.

Pollen chase experiment and pollen tube growth

For each trial, we conducted 3–6 pairs of SO and OO treatments over a period of 1 week to 10 days. During the summers of 2004–2007, 48 trials were carried out: seven on SC plants, 14 on Las Cruces plants, and 27 on Monteverde plants. Pollen was collected in clean vials and applied using clean toothpicks to styles of unopened flowers, 24 h before anthesis. Buds at this stage can be identified because they are still tightly closed but have developed the green splotches that decorate the corolla of mature flowers. Pairs of buds were chosen on each recipient, the bud for the SO treatment receiving self-pollen, and the bud for the OO treatment receiving outcross pollen. Twenty-four hours later, outcross pollen from a single compatible donor was applied to both flowers. Previous work in our laboratory has shown that although bud self-pollinations produce fewer pollen tubes at the base of the style than do bud outcross-pollinations, both types of pollinations produce tubes that grow past midstyle at 24 h and reach the base of the style by 48 h. Therefore, we assume that 24 h priority is sufficient to ensure that bud self-pollen tubes reach the base of the style before the subsequent outcross pollen tubes do. In total, 852 pollinations were done, half of which were bud pollinations. Fruits were allowed to ripen, and seeds were counted.

Germination and genotyping

To compare germination rates among treatments for SI populations, we planted seeds from fruits representing all available mothers and cross types. In most instances, 20 seeds were planted per cross, but in some cases all seeds per fruit were planted. We planted five seeds per cell in a 128-cell flat filled with Metro-Mix 200 (Sun-Gro Horticulture, Vancouver, British Columbia, Canada), alternating the position of cells occupied by seeds from SO and OO fruits. Seeds were covered with a fine layer of vermiculite, and the flats were placed in a mist bench under 12-h light/12-h dark in the greenhouse at Colby College. Liquid fertilizer was applied weekly once initial sprouts appeared. Germination was recorded at 6 wk, at which time seedlings were transplanted to individual cells and randomly placed on the bench. Seedlings were harvested by clipping with scissors at ground level at 12 wk and air dried.

To detect the level of self-fertilization resulting from the SO treatment, we used five microsatellite loci to genotype progeny and parental plants. We genotyped 73 progeny from 13 SO pollinations on eight maternal plants. Microsatellite-enriched libraries were developed by Craig Newton of ATG genetics (Vancouver, Washington. USA). We amplified inserts using universal m13 primers, sequenced the inserts using an ABI 3130 genetic analyzer (Applied Biosystems, Foster City, California, USA), and designed primers using the program Primer3 (Rozen and Skaletsky, 2000). Primer sequences appear in Table 1.

Table 1.

Primer sequences for microsatellite genotyping of Witheringia solanacea.

Name Sequence Tm (°C) Product length (bp) Alleles in sample
WsolgataA6L ctgctacccatggctcaact 60.28 185–205 6
WsolgataA6R ttttctcggttgcaatggtt 60.48
WsolcaA3L gcaaaatacagtcaacacaatgaa 59.10 244–246 2
WsolcaA3R gctcacgtggtgtttgtaga 57.30
WsolcaE1L cgatccacatcagctgaaaa 59.80 187–191 3
WsolcaE1R acccaggaattggggtaaga 60.55
WsolcaE2L gcggatccaggacatgaata 60.83 209–215 3
WsolcaE2R ctcctgcgacctgtctgttt 60.44
WsolcaE9L tccctcattggtaagggatg 59.74 154–166 4
WsolcaE9R cgaccagtgctagctgacaa 60.20
Open in a new tab

We extracted genomic DNA using the Qiagen (Germantown, Maryland, USA) DNeasy Plant Mini Kit on 40 mg dried leaf material, using dry ice to keep the tissue brittle for grinding in the Qiagen tissue lyser. PCR was carried out in 25 μL volumes in standard buffer with 1.5 mM MgCl2, 0.8 mM dNTPs, 0.5 μM each primer, 2 U Taq polymerase, and 10 ng genomic DNA. In each reaction, the left primer was fluorescently tagged with 6-FAM. A touchdown PCR program was used, with the annealing temperature decreasing from 60°C to 50°C by 0.5° each cycle for the first 20 cycles and kept at 50°C thereafter. Other PCR conditions were an initial denaturation at 95°C for 2.5 min; and 35 cycles with 95°C for 20 s, annealing for 30 s, and extension at 72°C for 30 s; with a final extension at 72°C for 10 min. PCR products were suspended in deionized formamide with ROX-500 size standard (Applied Biosystems) and run under standard conditions on an ABI 3130 genetic analyzer. Peaks were visualized using GeneScan and scored manually.

Data analysis

For the two SI populations, we used a two-way ANOVA to compare fruit and seed set of SO and OO treatments, with population as a random factor and treatment as a fixed factor. Maternal plant was nested within population. Treatment and the population × treatment interaction were tested over treatment × maternal plant nested within the population. To meet assumptions of ANOVA, proportion fruit set was arcsine-transformed, and seed number per plant per treatment was averaged and square-root transformed. The low number of available SC plants made it impossible to evaluate assumptions of parametric tests. For analyses involving SC plants, we used a Wilcoxon’s matched pairs signed rank test for paired comparisons, a Mann–Whitney test for unpaired comparisons, and Kruskal–Wallis tests for comparisons involving all three populations. Nonparametric tests were also used for germination data. Mann–Whitney tests were done by hand, and the program Stata 10.1 (Stata-Corp, College Station, Texas, USA) was used for all other analyses. The cumulative embryonic fitness decline associated with self-precedence was calculated for each individual according to the formula: δ = 1 − (wsf/wof · wss/wos), where wsf = proportion of SO pollinations producing fruit, wof = proportion of OO pollinations producing fruit, wss = average seed number resulting from SO pollinations, wos = average seed number resulting from OO pollinations.

To estimate the selfing rate for SO fruits, we calculated the probability, at each locus, that each offspring was a product of self-fertilization. If any locus revealed a definitive outcrossing event, the progeny was recorded as an out-cross. If no loci revealed definitive outcrossing, the likelihood of selfing was found by multiplying the probabilities across loci.

Pollen tube data, together with selfing rate estimates, allowed us to account for residual SI in the estimate of embryonic inbreeding depression. Bud self-pollination of SI plants from the Monteverde population yielded an average of 37 pollen tubes at the base of the style, whereas bud self-pollination of SI plants from Las Cruces yielded an average of only 13 pollen tubes at the base of the style (Stone et al., 2006). The number of pollen tubes at the base of the style after bud self-pollinations provides the maximum number of selfed seeds that could be expected under bud self-pollination. Any seeds in excess of that expectation should be due to the outcross pollination done on the second day of the pollen chase experiment. An expected selfing rate can therefore be obtained by comparing the number of self-pollen tubes with the total number of seeds. The observed selfing rate of surviving progeny as obtained by microsatellite genotyping was compared with the expected rate to estimate the relative postzygotic fitness of selfed progeny.

RESULTS

Fruit set for the OO treatment averaged 52% and did not differ across populations (H = 3.020, df = 2, P = 0.22). For the SI plants, mean fruit set for SO was 24%, significantly less than for the OO treatment (Fig. 2A, F1,18 = 34.2, P < 0.001). For the SC plants, fruit set for the SO treatments averaged 67%, which was not significantly different from the OO treatment (z = 0.54, N = 5 pairs, P = 0.60). For those fruits that set seeds, seed set was reduced by self-pollen precedence in SI plants (Fig. 2B, F1,12 = 6.25, P < 0.03) but not in SC plants (U = 18, n1n2 = 24, P = 0.20). There was an interaction between population and treatment, with Monteverde plants having a greater reduction than Las Cruces plants in seeds per fruit with self-pollen precedence (F1,12 = 6.25, P < 0.03). Loss of cumulative fitness at the embryonic stage resulting from self-bud pollination averaged 0.52 across all trials, but populations showed different degrees of fitness loss (H = 6.0, df = 2, P = 0.05). The self-compatible plants showed negligible fitness decline, whereas Las Cruces and especially Monteverde plants showed greatly reduced fitness for SO vs. OO (Table 2). Germination rate was just over 50% and did not differ by treatment or population (Table 3).

Fig. 2.

Fig. 2

Open in a new tab

Box plots showing nontransformed fitness components for pollinations with bud self-pollen precedence (SO treatment) vs. bud outcross-pollen precedence (OO treatment) for SC individuals and plants from two SI populations of Witheringia solanacea. (A) Proportion fruit set. Number of trials, number of flowers shown under each bar. (B) Seeds per fruit. Number of trials, number of fruits shown under each bar.

Table 2.

Cumulative fitness decline through both fruit and seed set of flowers receiving bud self-pollinations (SO) relative to those receiving bud-outcross pollinations (OO).

Population N Embryonic inbreeding depression
Las Cruces 10 0.28 ± 0.86
Monteverde 21 0.70 ± 0.51
SC plants 4 0.02 ± 0.68
Open in a new tab

Notes: Open flowers were outcross-pollinated in both treatments. Values are means (±SD). N = Number of trials for which both fruit and seed set were recorded. A trial consisted of 3–6 pairs of SO and OO treatments.

Table 3.

Mean germination rates (±SD) at 6 wk.

Treatment Fruits Seeds Las Cruces Fruits Seeds Monteverde
SO 4 173 0.48 ± 0.27 11 425 0.53 ± 0.32
OO 9 266 0.51 ± 0.32 15 296 0.59 ± 0.27
Open in a new tab

Notes: SO, self-pollen was applied in the bud the day before outcross pollination; OO, outcross-pollen was applied in the bud the day before outcross pollination.

The great majority of the progeny genotyped from the SO treatment showed definitive evidence of outcrossing by displaying a paternal allele that was not present in the mother (Table 4). For plants from the Las Cruces population, 28% of the surviving progeny from SO pollinations were likely to be self-fertilized. For plants from the Monteverde population, 15% of surviving progeny from SO pollinations were likely to be self-fertilized. The estimated outcrossing rates of surviving progeny did not significantly differ by population (U = 30, n1n2 = 40, P > 0.05). Individual mothers were not always consistent in the outcrossing rates of replicate progeny arrays. For example, the SO treatment of plant MV4 yielded outcross progeny in one trial and predominantly selfed progeny in another.

Table 4.

Probability of outcrossing of surviving progeny from the SO (self-precedence) pollination treatment, as assessed by genotyping at five microsatellite loci.

Family N Proportion definitively outcrossed p(t) for remainder Mean p(t)
LC3a 6 1 . 1.000
LC3b 8 4/8 0.00 0.500
LC6a 3 1 . 1.000
LC6b 7 6/7 0.125 0.875
LC19 5 1/5 0.22 0.375
Mean 0.724
MV2 4 1 . 1.000
MV4a 2 1 . 1.000
MV4b 8 1/8 0.08 0.195
MV11 2 1 . 1.000
MV12 8 1 . 1.000
MV39a 9 1 . 1.000
MV39b 7 1 . 1.000
MV39c 4 1 . 1.000
Mean 0.854
Open in a new tab

Notes: N = progeny genotyped. p(t) = probability of outcrossing. The mean outcrossing rates for each population are weighted by family size.

Estimates of cumulative inbreeding depression increased greatly when controlled for residual SI (Table 5). Bud self-pollinations were not completely successful in overcoming the SI response, so that we could expect selfing rates of only 0.32 and 0.85 for the Las Cruces and Monteverde populations, respectively. The observed selfing rate of surviving embryos at Las Cruces was similar to the expected, but the observed selfing rate of surviving embryos at Monteverde was much lower than expected. Once residual SI was taken into account, embryonic inbreeding depression was estimated to be 0.49 for the Las Cruces population and 0.97 for the Monteverde population. Inbreeding depression for the Monteverde population was significantly greater than for the Las Cruces population (U = 82.5, n1n2 = 91, P < 0.01).

Table 5.

Embryonic inbreeding depression in self-incompatible populations, corrected for residual self-incompatibility in bud pollinations by incorporating pollen tube data and microsatellite results.

Population Expected selfing rate a Observed selfing rate b Relative postzygotic fitness of selfed embryos c Relative fitness of SO for fruit set Cumulative embryonic inbreeding depression d
Las Cruces 0.32 ± 0.06 0.28 1.08 ± 0.25 0.48 ± 0.13 0.49 ± 0.18
Monteverde 0.85 ± 0.11 0.15 0.07 ± 0.03 0.34 ± 0.10 0.97 ± 0.01
Open in a new tab

Notes: Values are means (for selfing rate) or means ± standard error. N = 7 plants from Las Cruces and 13 plants from Monteverde.

a

Estimated proportion of self-pollen tubes reaching ovary in self-pollen precedence (SO) treatment = number of pollen tubes that reach the base of the style in self-bud pollinations/number of seeds produced in bud outcross-pollinations (OO).

b

Population average from microsatellite data.

c

Ratio of observed/expected selfing and outcrossing rates, (O/E)selfed/(O/E)outcrossed.

d

Embryonic inbreeding depression = 1 − (ws/wo), where ws = relative fitness of selfed embryos and wo = relative fitness of outcrossed embryos. Cumulative relative fitness is the product of relative fitness for fruit set and relative fitness for seed set.

DISCUSSION

The pollen chase experiment demonstrated that SC individuals had markedly lower embroyonic inbreeding depression than did SI individuals (Table 2). Such a contrast in inbreeding depression is widely found in interspecific comparisons between SI and SC species (Husband and Schemske, 1996). Correlations between selfing rate and inbreeding depression have also been observed in intraspecific comparisons among populations or lineages that differ in selfing rate (Chang and Rausher, 1999; Vogler et al., 1999; Takebayashi and Delph, 2000; Fishman, 2001; Stone and Motten, 2002; Busch, 2005; Goodwillie and Knight, 2006; Kennedy and Elle, 2008), although the extent to which we should expect to see such lineage-specific inbreeding depression has been controversial (Schultz and Willis, 1995). Key to the distinction between inter- vs. intraspecific purging of genetic load is the degree of reproductive isolation that may occur at the intraspecific level. In the current study, three of the selfing individuals derived from an undescribed variety of W. solanacea. It is not yet known whether the two varieties are interfertile, but the morphological distinctness suggests that reproductive barriers are present. The other two selfing individuals originated from the Vara Blanca population, which segregates for morphologically indistinguishable fully SI and SC individuals (Bohs, 2000; Stone et al., 2006), suggesting the possibility of lineage-specific inbreeding depression in the absence of any reproductive isolation other than mating system.

Plants from the two SI populations differed in the extent to which the SO treatment reduced seed set, with plants from the Monteverde population showing a 70% loss in production, whereas the plants from Las Cruces declined only ~30%. Estimates of inbreeding depression increased for both populations when pollen tubes and genotype data were taken into account, with embryonic inbreeding depression at 97% for Monteverde and 49% for Las Cruces (Table 5). The incorporation of pollen tube and genotype data probably results in an overestimate of inbreeding depression because the expected selfing rate does not account for attrition of pollen tubes from the base of the style to the ovule. The true level of inbreeding depression should lie between the first and second estimates, which are both quite high, as well as distinct by population. The observation of more-severe inbreeding depression in the Monteverde population contradicts our initial hypothesis that the quantitatively weaker SI response might correspond to a history of leaky SI and attendant purging of the genetic load. Perhaps the reduced stylar RNase activity does not cause leaky SI for plants in the Monteverde population, the degree of inbreeding has been insufficient to purge genetic load, or other factors such as population structure have had a predominant influence in shaping contrasting levels of inbreeding depression for the two regions (Glémin et al., 2003; Roze and Rousset, 2004). We plan to address these possibilities by using molecular markers to estimate current and historical levels of inbreeding and to evaluate demographic histories of populations in the two regions.

Inbreeding depression exceeding 50% at the embryonic stage has been documented in numerous studies (e.g., Husband and Schemske, 1995; Vogler et al., 1999; Raspé et al., 2004). According to classical theory, inbreeding depression of 0.5 provides a threshold that counteracts the transmission advantage of selfing such that a variant that permits competing self-fertilization spreads to fixation below that threshold, but is eliminated above it (Lloyd, 1979). A more recent model found that invasion of a nonfunctional S-allele is favored under conditions of pollinator limitation, low numbers of S-alleles, and low genetic load due to nearly additive or S-linked mutations (Porcher and Lande, 2005). Load caused by deleterious recessive alleles is less critical in the model because it can be readily purged by selfing. Therefore, population size—as it affects S-allele number and pollen limitation—appears to be a more critical factor than inbreeding depression caused by lethal recessive mutations in determining the fate of a SC allele. The predictions of the Porcher and Lande model are consistent with our observations that SC variants have been unable to invade large populations with many S-alleles (Stone and Pierce, 2005), whereas an SC mutation has been able to invade the tiny, isolated Vara Blanca population, and SC lineages at that population have purged their genetic load.

To our knowledge, this is the first pollen chase experiment with the goal of estimating embryonic inbreeding depression in a species with stylar-acting SI. Pollen chase experiments have previously been used to investigate interference by self-pollen in SC species (Krebs and Hancock, 1990), in heterostylous species (reviewed by Barrett, 2002), in species with cryptic or late-acting SI (Bertin and Sullivan, 1988; Waser and Price, 1991; Broyles and Wyatt, 1993), and in Polemonim viscosum, which has stigmatic-acting SI (Galen et al., 1989). Because bud pollination, in some instances, only partially overcomes SI, pollen chase experiments can provide only a lower bound for embryonic inbreeding depression. Despite its limitations, the pollen chase method used in concert with pollen tube data and molecular genetic markers was able to reveal contrasting levels of inbreeding depression in two populations, severe embryonic inbreeding depression in the Monteverde population, and evidence for purging of genetic load in SC individuals.

Footnotes

1

The authors thank B. Hummel, M. Sasuclark, and A. Smith for greenhouse work; P. Easton for processing microsatellite samples; and anonymous reviewers for helpful comments. Funding was provided by NSF DEB-0841482 and by NIH Grant Number P20 RR-016463 from the INBRE Program of the National Center for Research Resources.

LITERATURE CITED

  1. Barrett SCH. Sexual interference of the floral kind. Heredity. 2002;88:154–159. doi: 10.1038/sj.hdy.6800020. [DOI] [PubMed] [Google Scholar]
  2. Bertin RI, Sullivan M. Pollen interference and cryptic self-incompatibility inCampsis radicans. American Journal of Botany. 1988;75:1140–1147. [Google Scholar]
  3. Bohs L. Insights into the Witheringia solanacea (Solanaceae) complex in Costa Rica. I. Breeding systems and crossing studies. Biotropica. 2000;32:70–79. [Google Scholar]
  4. Broyles SB, Wyatt R. The consequences of self-pollination in Asclepias exaltata, a self-incompatible milkweed. American Journal of Botany. 1993;80:41–44. [PubMed] [Google Scholar]
  5. Busch JW. Inbreeding depression in self-incompatible and self-compatible populations ofLeavenworthia alabamica. Heredity. 2005;94:159–165. doi: 10.1038/sj.hdy.6800584. [DOI] [PubMed] [Google Scholar]
  6. Byers DL, Waller DM. Do plant populations purge their genetic load? Effects of population size and mating history on inbreeding depression. Annual Review of Ecology and Systematics. 1999;30:479–513. [Google Scholar]
  7. Chang SM, Rausher MD. The role of inbreeding depression in maintaining the mixed mating system of the common morning glory Ipomoea purpurea. Evolution. 1999;53:1366–1376. doi: 10.1111/j.1558-5646.1999.tb05401.x. [DOI] [PubMed] [Google Scholar]
  8. Crnokrak P, Barrett SCH. Purging the genetic load: A review of the experimental evidence. Evolution. 2002;56:2347–2358. doi: 10.1111/j.0014-3820.2002.tb00160.x. [DOI] [PubMed] [Google Scholar]
  9. D’Arcy WG. Flora of Panama. Annals of the Missouri Botanical Garden. 1973;60:573–580. [Google Scholar]
  10. de Nettancourt D. Incompatibility and incongruity in wild and cultivated plants. Springer; Berlin, Germany: 2000. [Google Scholar]
  11. Fishman L. Inbreeding depression in two populations of Arenaria uniflora (Caryophyllaceae) with contrasting mating systems. Heredity. 2001;86:184–194. doi: 10.1046/j.1365-2540.2001.00820.x. [DOI] [PubMed] [Google Scholar]
  12. Galen C, Gregory T, Galloway LF. Costs of self-pollination in a self-incompatible plantPolemonium viscosum. American Journal of Botany. 1989;76:1675–1680. [Google Scholar]
  13. Glémin S, Bataillon T, Ronfort J, Mignot A, Oliveiri I. Inbreeding depression in small populations of self-incompatible plants. Genetics. 2001;159:1217–1229. doi: 10.1093/genetics/159.3.1217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Glémin S, Ronfort J, Bataillon T. Patterns of inbreeding depression and architecture of the load in subdivided populations. Genetics. 2003;165:2193–2212. doi: 10.1093/genetics/165.4.2193. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Goodwillie C, Knight MC. Inbreeding depression and mixed mating in Leptosiphon jepsonii: a comparison of three populations. Annals of Botany. 2006;98:351–360. doi: 10.1093/aob/mcl105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Husband BC, Schemske DW. Magnitude and timing of inbreeding depression in a diploid population of Epilobium angustifolium (Onagraceae) Heredity. 1995;75:206–215. [Google Scholar]
  17. Husband BC, Schemske DW. Evolution of the magnitude and timing of inbreeding depression in plants. Evolution. 1996;50:54–67. doi: 10.1111/j.1558-5646.1996.tb04472.x. [DOI] [PubMed] [Google Scholar]
  18. Igic B, Lande R, Kohn JR. Loss of self-incompatibility and its evolutionary consequences. International Journal of Plant Sciences. 2008;169:93–104. [Google Scholar]
  19. Kärkkäinen K, Kuittinen H, van Treuren R, Vogl C, Oikarinen S, Savolainen O. Genetic basis of inbreeding depression in. Arabis petraea Evolution. 1999;53:1354–1365. doi: 10.1111/j.1558-5646.1999.tb05400.x. [DOI] [PubMed] [Google Scholar]
  20. Kennedy BF, Elle E. The inbreeding depression cost of selfing: Importance of flower size and population size in Collinisa parviflora (Veronicaceae) American Journal of Botany. 2008;95:1596–1605. doi: 10.3732/ajb.0800322. [DOI] [PubMed] [Google Scholar]
  21. Krebs SL, Hancock JF. Early-acting inbreeding depression and reproductive success in highbush blueberryVaccinium corymbosum. Theoretical and Applied Genetics. 1990;79:825–832. doi: 10.1007/BF00224252. [DOI] [PubMed] [Google Scholar]
  22. Lande R, Schemske DW. The evolution of self-fertilization and inbreeding depression in plants. I. Genetic models. Evolution. 1985;39:24–40. doi: 10.1111/j.1558-5646.1985.tb04077.x. [DOI] [PubMed] [Google Scholar]
  23. Lloyd DG. Some reproductive factors affecting the selection of self-fertilization in plants. American Naturalist. 1979;113:67–79. [Google Scholar]
  24. McClure BA, Gray JE, Anderson MA, Clarke AE. Self-incompatibility in Nicotiana alata involves degradation of pollen rRNA. Nature. 1990;347:757–760. [Google Scholar]
  25. Porcher E, Lande R. Loss of gametophytic self-incompatibility with evolution of inbreeding depression. Evolution. 2005;59:46–60. [PubMed] [Google Scholar]
  26. Raspé O, Guillaume P, Jacquemart AL. Inbreeding depression and biased paternity after mixed-pollination in Vaccinium myrtillus L. (Ericaceae) International Journal of Plant Sciences. 2004;165:765–771. [Google Scholar]
  27. Ritland K. Inferences about inbreeding depression based on changes of the inbreeding coefficient. Evolution. 1990;44:1230–1241. doi: 10.1111/j.1558-5646.1990.tb05227.x. [DOI] [PubMed] [Google Scholar]
  28. Roze D, Rousset F. Joint effects of self-fertilization and population structure on mutation load, inbreeding depression, and heterosis. Genetics. 2004;167:1001–1015. doi: 10.1534/genetics.103.025148. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Rozen S, Skaletsky HJ. Primer3 on the WWW for general users and for biologist programmers. In: Krawetz S, Misener S, editors. Bioinformatics methods and protocols: Methods in molecular biology. Humana Press; Totowa, New Jersey, USA: 2000. pp. 365–386. [DOI] [PubMed] [Google Scholar]
  30. Schultz ST, Willis JH. Individual variation in inbreeding depression: The roles of inbreeding history and mutation. Genetics. 1995;141:1209–1223. doi: 10.1093/genetics/141.3.1209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Stebbins GL. Flowering plants: Evolution above the species level. Harvard University Press; Cambridge, Massachusetts, USA: 1974. [Google Scholar]
  32. Stone JL, Jenkins EG. Pollinator abundance and pollen limitation of a solanaceous shrub at premontane and lower montane sites. Biotropica. 2008;40:55–61. [Google Scholar]
  33. Stone JL, Motten AF. Anther–stigma separation is associated with inbreeding depression in Datura stramonium, a predominantly selfing annual. Evolution. 2002;56:2187–2195. doi: 10.1111/j.0014-3820.2002.tb00143.x. [DOI] [PubMed] [Google Scholar]
  34. Stone JL, Pierce SE. Rapid recent radiation of S-RNase lineages in Witheringia solanacea (Solanaceae) Heredity. 2005;94:547–555. doi: 10.1038/sj.hdy.6800657. [DOI] [PubMed] [Google Scholar]
  35. Stone JL, Sasuclark MA, Blomberg CP. Variation in the SI response within and among populations of the tropical shrub Witheringia solanacea (Solanaceae) American Journal of Botany. 2006;93:592–598. doi: 10.3732/ajb.93.4.592. [DOI] [PubMed] [Google Scholar]
  36. Takebayashi N, Delph LF. An association between a floral trait and inbreeding depression. Evolution. 2000;54:840–846. doi: 10.1111/j.0014-3820.2000.tb00084.x. [DOI] [PubMed] [Google Scholar]
  37. Vogler DW, Filmore K, Stephenson AG. Inbreeding depression in Campanula rapunculoides L. I. A comparison of inbreeding depression in plants derived from strong and weak self-incompatibility phenotypes. Journal of Evolutionary Biology. 1999;12:483–494. [Google Scholar]
  38. Waser NM, Price MV. Reproductive costs of self-pollination in Ipomopsis aggregata (Polemoniaceae): Are ovules usurped? American Journal of Botany. 1991;78:1036–1043. [Google Scholar]

RESOURCES