American Journal of Botany 93(5): 739746. 2006.

INBREEDING

EFFECT ON MALE AND FEMALE FERTILITY AND

INHERITANCE OF MALE STERILITY IN

NEMOPHILA (HYDROPHYLLACEAE)1
AND

MENZIESII

NADILIA N. GOMEZ2

RUTH G. SHAW

University of Minnesota, Department of Ecology Evolution and Behavior, 100 Ecology, 1987 Upper Buford Circle, St. Paul, Minnesota 55108 USA Models of the evolution of gynodioecy assume that inbreeding affects male and female fertility equally and ignore quantitative variation in sex expression. The objectives of this study were to assess inbreeding effects, genetic background, and plant maturity on male and female fertility and the mechanism of male sterility inheritance for Nemophila menziesii (Hydrophyllaceae). Frequency of male-sterile flowers, number of anthers and ovules, and percentage of viable pollen were measured on plants from different pedigrees and five inbreeding levels (F 0, 0.0625, 0.25, 0.5, and 0.75). Quantitative variation in male sterility was evident. As inbreeding increased, anther and ovule number decreased; the effect on anther number was greater than on ovule number. Pedigrees varied in number of male-sterile flowers and inbreeding effects. Frequency of male-sterile flowers was greatest among first flowers. No trade-off between male and female fertility was detected. A model attributing male sterility to a cytoplasmic locus and restoration to male fertility to a nuclear locus accounted for the distribution of complete sterility and hermaphroditism over the pedigrees. This study suggests that models of the evolution and maintenance of gynodioecy should allow for quantitative variation in male and female fertility components due to inbreeding, pedigree, and plant maturity. Key words: cytoplasmic male sterility; gynodioecy; Hydrophyllaceae; partial male sterility; plant gender; qualitative variation; sex determination; vestigial anther.

More than 90% of flowering plant species are hermaphroditic. A small proportion of species are gynodioecious. Gynodioecy is a sexual polymorphism in which female and hermaphroditic plants comprise a population. Gynodioecy has been considered a transitional stage in the evolution from hermaphroditism to dioecy (Ross, 1970), but under certain conditions, gynodioecy may be an evolutionarily stable polymorphism (reviewed in Bailey et al., 2003). A complication in evaluating the evolutionary stability of gender polymorphisms is that sexual expression of the two morphs may vary. Plants of gynodioecious species are often characterized as female or hermaphrodites even though variation within individuals in the gender of flowers, or sex expression, within single plants is not unusual (Bawa and Beach, 1981). Numerous phenomena influence sex expression and the maintenance of gynodioecy in populations including the genetic basis of gender inheritance, inbreeding depression, and developmental effects. One of the best-studied factors influencing the gender of plants in a gynodioecious species and the maintenance of gynodioecy in populations is the genetic basis of gender inheritance. Male sterility mutations, causing strictly female plants, can arise in either nuclear or cytoplasmic genes. When male sterility is compensated by an increase in fitness through female function, it can persist in a population particularly if the male sterility mutation is cytoplasmic, because the maternal inheritance of cytoplasmic genomes ensures its transmission even when male function is lost
1

Manuscript received 6 August 2005; revision accepted 7 February 2006.

The authors thank S.-M. Chang, K. Mercer, H. Hangelbroek, and L. Stone for valuable comments on the manuscript and M. Sanders for assistance with digital imaging. This work was supported in part by the Plant Biological Sciences Block Grant Program. 2 Author for correspondence (e-mail: [email protected]), present address: University of St. Thomas, Department of Biology, Mail OWS 390, 2115 Summit Ave., St. Paul, MN 55105 USA

(Taylor et al., 1999). Counteracting the effect of cytoplasmic genes inducing male sterility are nuclear genes that restore male fertility (Frank, 1989). Restorers differ in the degree to which they restore male fertility. In some cases they can completely restore male fertility, whereas in other cases they may only partially restore normal pollen production. Inbreeding can also influence sex expression (Hayes et al., 2005) and contribute to stable sexual polymorphisms (Schultz, 2002). In self-compatible species, inbreeding depression can also favor the spread of male sterility (Lande and Schemske, 1985; Uyenoyama et al., 1993) and the maintenance of gynodioecy because hermaphrodites are more likely than females to undergo inbreeding, and thus inbreeding depression is expected to be more often expressed in the progeny of hermaphrodites than in the progeny of females. Differences in the magnitude of inbreeding depression between male and female fitness components have been reported in the hermaphroditic species Mimulus guttatus (Robertson et al., 1994; Carr and Dudash, 1995, 1997), Phacelia dubia (del Castillo, 1998), Ipomoea purpurea (Chang and Rausher, 1999), and Cucurbita pepo (Hayes et al., 2005). Whereas models of mating system evolution tend to assume implicitly that inbreeding depression is equal in both male and female fertility components (Schultz, 2002), Rausher and Chang (1999) have modeled differences in inbreeding depression in male and female fitness components to explain the stabilization of mixed-mating systems. Sex expression can also vary among flowers of a single plant depending on development, e.g., floral position (Mazer and Dawson, 2001). Preferential resource allocation to developing fruits in the proximal end of the inflorescence is associated with production of male flowers in the distal portion of inflorescences (Brunet and Charlesworth, 1995). These patterns of floral variation among flowers of a plant have a genetic basis as shown by the variability observed among pedigrees and populations (Guitian et al., 2004). A previous study of the gynodioecious species Nemophila 739

740

A MERICAN J OURNAL

OF

B OTANY

[Vol. 93

Fig. 1. Variation in stamen morphology of Nemophila menziesii flowers. Stamen morphology was used to score the first three or five flowers of each plant into one of three flower type categories of sex expression: (A) hermaphroditic flowers with five pollen-bearing anthers, (B) male-sterile flowers with five vestigial anthers, and (C) partially male-sterile flowers with at least one vestigial anther. One vestigial anther and four pollen-bearing anthers can be observed in the partial male-sterile flower, but the number of vestigial anthers observed in other partial male-sterile flowers ranged from 14 anthers per flower. Two stigmas and a bifurcated style are noticeable in all three flowers. Bar 1 mm. Figure abbreviations: b, bifurcated style; p, pollen-bearing anther; s, stigma; v, vestigial anther.

menziesii H. & A. (Hydrophyllaceae) indicated that some plant traits are affected by inbreeding more than others (Shaw et al., 1998). During the execution of that experiment, a wide range of sexual morphotypes was observed. Some plants were completely male sterile, some had flowers with vestigial or abnormal anthers, while others were hermaphroditic. We here present a study of male and female fertility in this study population, based on previously unreported data from a preliminary experiment, which motivated our main experiment. Plants in the main experiment are siblings of the ones in Shaw et al. (1998) and were grown specifically to assess the effects of inbreeding and flower position in more detail. As primary objectives, we assessed effects of genetic background, inbreeding, and developmental stage on male and female fertility. Further, we evaluated whether the patterns of inheritance of male sterility support the nuclearcytoplasmic male sterility model in this population of N. menziesii. MATERIALS AND METHODS
Study speciesNemophila menziesii H. & A. (Hydrophyllaceae) is an annual plant native to California and Oregon, USA (Munz, 1959). Most N. menziesii flowers are hermaphroditic, with five stamens, pollen-bearing anthers, a bifurcated style, and two stigmas (Fig. 1). Flowers are protandrous, produce nectar, and are frequently pollinated by bees but can self-pollinate (Cruden, 1972; Andersson, 1994). Flowers may be male sterile, with anthers and filaments either completely absent or vestigial. Flowers with an intermediate morphology between hermaphroditic flowers and male-sterile flowers are also observed. Partial male-sterile flowers have fewer than five anthers, and anther dehiscence may be incomplete in some or all of the existing anthers. Ganders (1978) reported a population of N. menziesii with a high proportion (2226%) of male-sterile plants. The male-sterile phenotype of the plants found in the wild was retained under greenhouse cultivation (Ganders, 1978). After conducting experiments to identify the inheritance mechanism of male sterility,

Ganders (1978) concluded that N. menziesii was a gynodioecious species with nuclearcytoplasmic male sterility. Barr (2004) documented considerable variation among 23 populations of N. menziesii in the proportion of malesterile plants, which ranged from 10% to 52%. Preliminary experimentIn 1990, Shaw et al. (1998) collected seedlings from an uncultivated area of the University of California Riverside Botanic Gardens. Plants were randomly crossed in a nested-crossing design, with three plants used as females mated to a single plant used as male. Three subsequent generations were derived from the progeny of this cross. The offspring of the third generation represented 10 independent pedigrees and five different inbreeding levels (F 0, 0.0625, 0.25, 0.5, and 0.75; see Shaw et al., 1998 for details of the pedigree structure). Plants from the third generation were grown in August 1995 and measured to determine the effects of inbreeding on means and genetic variance of several morphological traits and of flower phenology (Shaw et al., 1998). In addition, observations of flower morphology were made on eight of the 10 pedigrees. The first three flowers from each of 1048 plants (3108 flowers) were scored as male sterile, partial male sterile, or hermaphroditic, depending on stamen morphology (Fig. 1). Pedigree selection for the main experimentBased on analyses of the preliminary experiment, four pedigrees were chosen for a more detailed study of variation in female and male fertility components. They span the variation in magnitude and pattern of inbreeding effect on male fertility; two had a high frequency of male-sterile flowers and two had a low frequency of male-sterile flowers. In three pedigrees male sterility increased with inbreeding, whereas male sterility decreased as inbreeding increased in the other pedigree. Two sibships (full sib groups) from each of the five inbreeding levels within each of the four pedigrees (Table 1) were chosen. Sibships that displayed relatively high levels of male sterility were chosen to obtain adequate numbers of individuals expressing the trait. To maximize information for evaluating the models of male sterility inheritance, the sibships having parents in common were chosen preferentially. When sibships from a particular inbreeding level were not available, additional sibships from other inbreeding levels were included to obtain better estimates of the effect of inbreeding. This process resulted in three, rather than two, sibships for some inbreeding levels (Table 1). From inbreeding levels 0.0625 and 0.25, two pairs of reciprocal sibships were

May 2006]

G OMEZ

AND

S HAW I NBREEDING

AND MALE STERILITY IN

N EMOPHILA

741

TABLE 1. Sibships selected from the preliminary experiment to be used in the main experiment. Four pedigrees (I, II, VI, and VIII) of Nemophila menziesii were selected from 10 pedigrees previously developed for the study on inbreeding depression by Shaw et al. (1998). Each pedigree contained crosses representing five different inbreeding levels (F 0, 0.0625, 0.25, 0.50, 0.75). Numbers within the table are parental plant identification numbers; the first number is the female ID, the second is the male ID. Each pair of parental plants was crossed to produce a sibship. Except for pedigree II, F 0.25. 0.50, and 0.75, all inbreeding and pedigree combinations are represented by two sibships. Three sibships were included for F 0.25 and F 0.75 to account for unavailable seeds in F 0.50.
Pedigree Inbreeding coefficient I II VI VII

0 0.0625a 0.25a 0.50b 0.75

44 41 39 39 44

3 3 3 3 3

56 37 40 39 44

40 40 41 40 38

3 3 3 3 3

6 44 40 40 38

28 3 70 32 3 34 28 3 29 27 3 35 3 35 32 3

31 3 65 31 3 29 28 31 3 32 32 29 3 29

79 74 77 78 74

3 3 3 3 3

61 79 76 78 74

74 77 78 76 77

3 3 3 3 3

75 79 79 76 77

55 59 54 54 59

3 3 3 3 3

20 61 55 54 59

59 55 55 55 56

3 3 3 3 3

4 59 51 55 56

Reciprocal sibships were included for all pedigrees at this inbreeding level. Seeds of pedigree II were not available at this inbreeding level.
a response with two categories (flowers with and without pollen). Nominal logistic regression was used to analyze the data (JMP IN version 4, Statistical Discovery Software, SAS Institute, Cary, North Carolina, USA) with inbreeding level, pedigree, and flower position as main factors. The model also included the two-way interaction of inbreeding level and pedigree. In this procedure, the categorical response (flower type) is transformed into a logit variable and the regression estimates the probability of encountering a certain flower type as a function of a treatment factor (SAS Institute, 2000). Number of ovules was analyzed as a continuous variable and did not require transformation. Number of viable and nonviable pollen per flower was estimated by calculating the mean of the four measurements taken per slide. Box-Cox power estimation was used to select an appropriate transformation (ninth power) for percentage of viable pollen. The ratio of completely male-sterile to male-fertile plants for each sibship was calculated as an index of male sterility per sibship. A difference in this ratio between reciprocal sibships suggests that cytoplasmic genes contribute to inheritance of male sterility, whereas no difference apart from sampling variation is expected if male sterility is inherited via nuclear genes. If cytoplasmic genes are the basis for transmission of male sterility, then the ratio would differ depending on the direction of the cross, with a greater number of male-sterile individuals derived from maternal plants carrying the male-sterile cytoplasmic gene. The pedigree of completely male-sterile plants, full sibs, and half sibs was used as evidence concerning the number of distinct nuclear fertility restorers.

selected to assess differences due to the inheritance of cytoplasmic or nuclear genes. Altogether seeds from 56 sibships were used in the main experiment. Main experiment Seed germinationTwelve randomly selected seeds from each sibship were soaked in gibberellic acid (272 mg/L) and incubated in darkness at 48C for 24 h (Platenkamp and Shaw, 1993). Seeds were placed in Petri plates lined with moistened germinator pads and were left in the chamber for 4 d. Seeds were examined weekly for radicle emergence, then transplanted into 10-cm pots with Sunshine Mix Plug No. 5/LP 5 (Sun Gro Horticulture Canada, Bellevue, Washington, USA) and grown in the greenhouse at 258C. Because of the low germination rate in some pedigrees, after 30 d, a second group of seeds was germinated in an effort to obtain five plants per sibship. The number of seeds from each pedigree in this bout of germination depended on the germination success from the first group of seeds. For each sibship, if three or more seeds germinated in the first group, 12 seeds were placed to germinate in the second group. For sibships in which fewer than three seeds germinated in the first group, at least 30 seeds were placed to germinate in the second group. When fewer than 30 seeds were available, all seeds available were used in the second group. The same germination procedure was followed for all seeds in the second group. Anther, pollen, and ovule studiesThe first five flowers from each mature plant were scored as completely male sterile, partially male sterile or hermaphroditic. The number of abnormal and fertile anthers was recorded at dehiscence. Anthers that dehisced fully or partially were considered fertile. Flowers were dissected under a stereomicroscope to count ovules. A total of 415 flowers from 108 individuals were examined. Pollen samples were prepared from a subset of plants in each sibship yielding a total of 307 flowers from 77 plants. Male fertility was estimated as the percentage of fertile pollen. Three anthers from a single flower were placed on a microscope slide in a Petri plate to prevent pollen loss and allowed to dehisce for 3 d. When fewer than three anthers were fertile, all fertile anthers were placed on the microscope slide. Pollen remaining in anthers was tapped onto the slide. Pollen was stained with 10 lL of Alexanders stain to determine pollen viability. Alexanders stain (Kearns and Inouye, 1993) contains malachite green, which stains the walls of abortive pollen green, and acid fuchsin, which stains the protoplasm of viable pollen pink. Four randomly selected areas of the slide were digitally photographed using the 403 objective lens of a Nikon E800 Microscope and a Roper CoolSnap HQ 16-bit monochrome camera. Digital images were then used to score pollen viability using the program Scion Image 1.62c for Mac OS 7.5 to 9.0 (Acquisition and Analysis Software, Scion, Frederick, Maryland, USA). Data analysesHermaphroditic and partial male-sterile flowers were grouped together in the analysis of flower type because of the limited number of partially male-sterile flowers and the wide morphological variation among flowers classified as partially male-sterile flowers. Grouping partially malesterile flowers with hermaphroditic flowers provided comparison of flowers bearing pollen with those altogether lacking pollen. Given the strongly bimodal distribution of anther number, with individuals having mostly five or zero anthers per flower, flower type in the main experiment was also analyzed as

RESULTS Preliminary experimentIn the preliminary experiment, 83% of the 3108 flowers, produced from 1048 individuals, were hermaphroditic. Only 4% were partially male sterile and 13% were male sterile. Of the 1048 plants scored, 68 (6%) plants produced three male-sterile flowers, 48 (4%) plants produced two male-sterile flowers, and 110 (10%) plants produced one male-sterile flower. The remaining individuals produced three hermaphroditic flowers. Twenty-one individuals produced fewer than three flowers. Inbreeding, genetic background, and plant maturity affected production of male-sterile flowers. Inbreeding significantly increased the frequency of male-sterile flowers (P , 0.0001). Pedigrees differed both in the average number of male-sterile flowers produced (P , 0.0001) and in the response of this trait to inbreeding (P 0.0017; Fig. 2). The percentage of malesterile flowers was relatively high in pedigrees I (32%) and II (23.8%) and considerably lower in pedigrees VI (7.1%) and VIII (3.5%). Plant maturity also significantly affected the frequency of male-sterile flowers (P , 0.0001). The earliest flowers were more likely to be male sterile, and the tendency to produce male-sterile flower decreased as the plant developed.

742

A MERICAN J OURNAL

OF

B OTANY

[Vol. 93

Fig. 2. Percentage of male-sterile Nemophila menziessii flowers from eight pedigrees and five inbreeding levels studied in the preliminary experiment to determine the effect of inbreeding on male fertility. The Roman numeral identifying each pedigree reflects its ranking based on percentage of malesterile flowers from highest to lowest, barring family VII, which had the largest SE associated with the mean of mean percentage of male-sterile flowers. These results were used to select four pedigrees for a more detailed study on sex expression (main experiment).

Main experimentIn the main experiment, 340 (82%) of the flowers had five fully dehiscent anthers. Eleven individuals produced at least one completely male-sterile flower and three produced only male-sterile flowers. Effects of inbreeding (P 0.3960), pedigree (P 0.1511), and flower position (P 0.2159) on number of fertile anthers were not detected (Fig. 3). When flowers were classified into two groups (flowers without pollen and flowers with pollen), the differences among pedigrees were significant (P 0.0281). Although the effects of plant maturity remained nonsignificant, a slight increase in the tendency of later flowers to be male sterile was observed, contrary to the trend in the preliminary experiment. When the data from both preliminary and main experiments were analyzed together, adding a variable indicating the experiment, inbreeding level (P 0.0245), pedigree (P , 0.0001), and plant maturity (P , 0.0001) significantly affected flower type. A significant interaction between inbreeding level and pedigree (P , 0.0001) indicated that the inbreeding effects differ among pedigrees. Percentage of viable pollen declined with inbreeding, but the trend was not significant. Effects of pedigree (P , 0.001) and flower position (P 0.001) were significant. Pedigree II had the lowest percentage of viable pollen (84%), and pedigree VI had the highest (94%). Both pedigrees I and VIII had 92% viable pollen. Across all pedigrees, as plants matured, they produced flowers with greater pollen viability. The number of ovules per flower ranged from 5 to 17. The modal ovule count was 12 ovules per flower; 34% of the flowers had 12 ovules. Pedigree (P , 0.0001) and flower position (P 0.0094) significantly affected the number of ovules produced. The effect of inbreeding on ovule production also varied among pedigrees (P 0.001). Pedigrees I and II produced fewer ovules as inbreeding increased, whereas the

responses of pedigrees VI and VIII to inbreeding were not monotonic for this trait; these pedigrees produced the greatest number of ovules at an inbreeding level of 0.25 or 0.75, respectively (Fig. 4). The number of ovules produced decreased gradually as the plant matured. On average, malesterile flowers produced fewer ovules (10) than hermaphroditic flowers (12; P 0.0001). Even after accounting for inbreeding, a significant positive correlation between number of ovules and the amount of viable pollen produced was observed. Considering the three plants that were strictly male sterile, as inbreeding increased the average number of ovules per flower decreased (11.4, 11.2, and 10 ovules, for F 0, 0.25, and 0.5, respectively). DISCUSSION Inbreeding reduces both male and female fertility in this population of Nemophila menziesii. These effects, which are expressed quantitatively as increases in the incidence of sterile anthers and reductions in the number of ovules per flower, vary with genetic background, suggesting that a polygenic system contributes to male sterility, as well as to fertility via ovule production. Of the 1156 plants studied, few were completely male sterile, but partial male sterility was relatively common (14% of plants in the larger, initial sample of the preliminary experiment). Schultz (2002), noting growing evidence of partial male sterility, has extended models of evolution of gynodioecy to allow for intermediate fertility of males. Whereas the preliminary experiment unambiguously demonstrated that inbreeding increased the number of male-sterile flowers in most of our pedigrees (Fig. 2), the trends with inbreeding were less clear and less consistent among pedigrees in the main experiment, designed for intensive assessment of

May 2006]

G OMEZ

AND

S HAW I NBREEDING

AND MALE STERILITY IN

N EMOPHILA

743

Fig. 3. Number of pollen-bearing, fertile anthers (mean and SE) in Nemophila menziesii flowers in the four pedigrees (I, II, VI, VIII) and five inbreeding levels (F 0, 0.0625, 0.25, 0.50, 0.75) studied in the main experiment. Seeds were not available for pedigree II F 0.5, and data is missing for plants in pedigree VI F 0.75. All flowers in pedigree VI F 0 and pedigree VIII F 0.75 produced five pollen-bearing, fertile anthers.

male and female fertility and, hence, including considerably fewer plants. In the main experiment, only pedigree I had a clear reduction in the number of fertile anthers per flower with inbreeding (Fig. 3) as expected from the preliminary experiment, and this pedigree also had the clearest decline in ovule numbers with inbreeding (Fig. 4). Both findings support

an interpretation that (partially) recessive alleles that reduce either or both components of fertility segregate in this pedigree. The lack of evidence for sterility in pedigrees II and VI (Fig. 3) is surprising in comparison with the results from the preliminary experiment (Fig. 2). For pedigree VIII, the low expression of male sterility irrespective of degree of inbreeding

Fig. 4. Number of ovules (mean and SE) in Nemophila menziesii flowers in four pedigrees (I, II, VI, VIII) and five inbreeding levels (F 0, 0.0625, 0.25, 0.5, 0.75) studied in the main experiment. Seeds were unavailable from sibships in pedigree II F 0.5, and data is missing for plants from pedigree VI F 0.75.

744

A MERICAN J OURNAL

OF

B OTANY

[Vol. 93

Fig. 5. Pedigree I with all three female Nemophila menziesii plants observed in the main experiment. Plants are designated by a plant identification number that bears no significance on the relationship among plants and generations shown in the pedigree. The relationships among individuals are marked with lines: paternal parents are connected with dotted lines, maternal parents are connected with solid lines, and selfed are connected with thick solid lines. Completely male-sterile plants are designated by boldfaced numbers 839, 757, and 625. Full sibs from the same cross and the direction of the reciprocal cross are listed together. Plants with identification numbers that are not in boldface may have had a male-sterile flower, but did not consistently produce all male-sterile flowers.

(Fig. 3) is consistent with the findings of the preliminary experiment (Fig. 2). Because the same seed lots were used for both experiments, the difference in findings cannot be attributed strictly to genetic differences, e.g., purging of deleterious alleles via selection. Failure to germinate and seedling mortality of plants with higher inbreeding levels may have introduced bias in the effect of inbreeding on male and female fertility components. Previous studies show that germination was lower at higher inbreeding levels (Shaw et al., 1998). Therefore, selection at germination and seedling survival may have introduced a bias in samples minimizing the differences across inbreeding level on male and female fertility components. Thus, in addition to demonstrating inbreeding effects on sex expression and partial male sterility, these findings illustrate inherent variability of sex expression, likely due jointly to genetic and environmental variation. Resource allocation trade-offs between male and female fertility componentsWhereas male-sterile plants are assumed to allocate more resources to ovule production (Charlesworth and Ganders, 1979; Charlesworth and Charlesworth, 1981; Charnov, 1982; Frank, 1989), no trade-offs between ovule production and percentage of viable pollen were detected. Instead, a trend in the opposite direction was found. The predicted trade-off between male and female fertility has likewise not been detected in other studies; Williams and Fenster (1998) found that male-sterile flowers produced only slightly more ovules (NS) than hermaphroditic flowers in Chamaecrista fasciculata, and Robertson et al. (1994) found no evidence of trade-offs in production of pollen and ovules in Mimulus guttatus. However, trade-offs may exist between other male and female fertility components not measured in this study. Significant negative correlations between male and female components of reproduction have been reported for several species (Atlan et al., 1992; Ashman, 1999; Olson and Antonovics, 2000; Parachnowitsch and Elle 2004).

Mechanisms of male sterility inheritanceTwo models can account for the inheritance of male sterility in the three fully male-sterile plants (plants 839, 625, and 757) observed in the main experiment. These plants were all from pedigree I, which contained the greatest number of male-sterile flowers in both the preliminary and main experiment (Figs. 2, 3, 4). The simplest model that could explain male sterility in this pedigree invokes a single nuclear recessive male sterility gene. The second model involves a cytoplasmic gene inducing male sterility and a nuclear restorer. Under a model with a single nuclear recessive male sterility gene, plants 236 and 179 were heterozygous at that locus. Selfing of plant 236 and mating between maternal plant 236 and paternal plant 179 could have resulted in heterozygous plants 40, 39, and 41. These, in turn, could have produced the hermaphroditic and female plants in the third generation, though male sterility of 839 requires in addition that plant 6 also carried the recessive allele. Gynodioecy can be stably maintained under this mode of inheritance of male sterility, as shown by substantial modeling (Valdeyron, et al., 1973; Charlesworth and Charlesworth, 1978; Gregorius et al., 1982; Schultz, 2002). Conditions for the evolution of gynodioecy under this model are considered stringent (Charlesworth and Charlesworth, 1978), though they may be less stringent when certain assumptions are relaxed (Schultz, 2002). A model with one male sterility cytotype and a single-locus dominant nuclear male fertility restorer can also account for the occurrence of complete male sterility observed in this pedigree and would be in agreement with previous studies on this species (Ganders, 1978; Barr, 2004). The difference in the occurrence of male sterility in reciprocal sib groups supports the model of cytoplasmic inheritance of male sterility, but is not conclusive due to the small samples. Offspring from reciprocal sibships have nuclear genes in common but can differ in their cytoplasmic genes. Therefore, male sterility in the progeny can vary depending on the direction of the sibship

May 2006]

G OMEZ

AND

S HAW I NBREEDING

AND MALE STERILITY IN

N EMOPHILA

745

if cytoplasmic genes are involved. Plants 577 and 584, reciprocal sibs of 625, were not completely male sterile. This model would imply that plant 236 was heterozygous for the restorer allele and that nuclear restorers could have been present in plants 6, 39, and 41. Quantitative nature of the variability in sex expression Our study shows that inbreeding, pedigree, and plant maturity may differentially affect male and female functions. These factors have a quantitative, rather than a qualitative, effect on fertility. Even though the number of pedigrees used in the main experiment were few, they were chosen to represent the range observed in the preliminary experiment and, hence, that of the population. It is likely that other factors not considered in this study also contribute to the observed variation. For example, gene penetrance, the frequency with which a phenotype is expressed among individuals with the same genotype, may contribute to variation in the degree of male sterility observed in an individual. Among the factors studied, however, inbreeding, plant maturity, and genetic differences among pedigrees can cause variation in the proportion of male-sterile flowers. Quantitative measures such as pollen viability or anther counts may characterize plant gender more accurately than a dimorphic system by revealing the range of variability in sex expression (Lloyd, 1980; Bawa and Beach, 1981). Other gynodioecious species have shown variability among individuals classified by sexual morph, suggesting a need for more quantitative estimates of fertility (Lloyd and Bawa, 1984; Pfahler et al., 1996). We have shown that fertility of both male and female components varies with inbreeding, floral position, pedigree, and genotype within pedigree and that the labile expression of male sterility in this species supports the use of quantitative models of gender expression in evolutionary studies of plant mating systems. LITERATURE CITED
ANDERSSON, S. 1994. Floral stability, pollination efficiency, and experimental manipulation of the corolla phenotype in Nemophila menziessii (Hydrophyllaceae). American Journal of Botany 81: 13971402. ASHMAN, T.-L. 1999. Determinants of sex allocation in a gynodioecious wild strawberry: implications for the evolution of dioecy and sexual dimorphism. Journal of Evolutionary Biology 12: 648661. ATLAN, A., P. H. GOUYON, T. FOURNIAL, D. POMENTE, AND D. COUVET. 1992. Sex allocation in an hermaphroditic plant: the case of gynodioecy in Thymus vulgaris L. Journal of Evolutionary Biology 5: 423444. BAILEY, M. F., L. F. DELPH, AND C. M. LIVELY. 2003. Modeling gynodioecy: novel scenarios for maintaining polymorphism. American Naturalist 161: 762776. BARR, C. M. 2004. Soil moisture and sex ratio in a plant with nuclearcytoplasmic sex inheritance. Proceedings of the Royal Society of London, B, Biological Sciences 271: 19351939. BAWA, K. S., AND J. H. BEACH. 1981. Evolution of sexual systems in flowering plants. Annals of the Missouri Botanical Garden 68: 254 274. BRUNET, J., AND D. CHARLESWORTH. 1995. Floral sex allocation in sequentially blooming plants. Evolution 49: 7079. CARR, D. E., AND M. R. DUDASH. 1995. Inbreeding depression under a competitive regime in Mimulus guttatus: consequences for potential male and female function. Heredity 74: 437445. CARR, D. E., AND M. R. DUDASH. 1997. The effects of five generations of enforced selfing on potential male and female function in Mimulus guttatus. Evolution 51: 17971807. CHANG, S.-M., AND M. D. RAUSHER. 1999. The role of inbreeding

depression in maintaining the mixed mating system of the common morning glory, Ipomoea purpurea. Evolution 53: 13661376. CHARLESWORTH, D., AND B. CHARLESWORTH. 1981. Allocation of resources to male and female functions in hermaphrodites. Biological Journal of the Linnean Society 15: 5774. CHARLESWORTH, D., AND B. CHARLESWORTH. 1978. A model for the evolution of dioecy and gynodioecy. American Naturalist 112: 975 997. CHARLESWORTH, D., AND F. R. GANDERS. 1979. The population genetics of gynodioecy with cytoplasmic male sterility. Heredity 43: 213218. CHARNOV, E. 1982. The theory of sex allocation. Princeton University Press, Princeton, New Jersey, USA. CRUDEN, R. W. 1972. Pollination biology of Nemophila menziesii (Hydrophyllaceae) with comments on the evolution of oligolectic bees. Evolution 26: 373389. DEL CASTILLO, R. F. 1998. Fitness consequences of maternal and nonmaternal components of inbreeding in the gynodioecious Phacelia dubia. Evolution 52: 4460. FRANK, S. A. 1989. The evolutionary dynamics of cytoplasmic male sterility. American Naturalist 133: 345376. GANDERS, F. R. 1978. The genetics and evolution of gynodioecy in Nemophila menziesii (Hydrophyllaceae). Canadian Journal of Botany 56: 14001408. GREGORIUS, H. R., M. D. ROSS, AND E. GILLET. 1982. Selection in plant populations of effectively infinite size. III. The maintenance of females among hermaphrodites for a biallelic model. Heredity 48: 329343. GUITIAN, J., M. MEDRANO, AND J. E. OTI. 2004. Variation in floral sex allocation in Polygonatum odoratum (Liliaceae). Annals of Botany 94: 433440. HAYES, C. N., J. A. WINSOR, AND A. G. STEPHENSON. 2005. Environmental variation influences the magnitude of inbreeding depression in Cucurbita pepo ssp. texana (Cucurbitaceae). Journal of Evolutionary Biology 18: 147155. KEARNS, C. A., AND D. W. INOUYE. 1993. Techniques for pollination biologists. University Press of Colorado, Boulder, Colorado, USA. LANDE, R., AND D. W. SCHEMSKE. 1985. The evolution of self fertilization and inbreeding depression in plants. I. Genetic models. Evolution 39: 2440. LLOYD, D. G. 1980. Sexual strategies in plants. III. A quantitative method for describing the gender of plants. New Zealand Journal of Botany 18: 103108. LLOYD, D. G., AND K. S. BAWA. 1984. Modification of the gender of seed plants in varying conditions. Evolutionary Biology 17: 255338. MAZER, S. J., AND K. A. DAWSON. 2001. Size-dependent sex allocation within flowers of the annual herb Clarkia unguiculata (Onagraceae): ontogenetic and among-plant variation. American Journal of Botany 88: 819831. MUNZ, P. A. 1959. A California flora. University of California Press, Berkeley, California, USA. OLSON, M. S., AND J. ANTONOVICS. 2000. Correlation between male and female reproduction in the subdioecious herb Astilbe biternata (Saxifragaceae). American Journal of Botany 87: 837844. PARACHNOWITSCH, A. L., AND E. ELLE. 2004. Variation in sex allocation and malefemale trade-offs in six populations of Collinsia parviflora (Scrophulariaceae s.l.). American Journal of Botany 91: 12001207. PFAHLER, P. L, M. J. PEREIRA, AND R. D. BARNETT. 1996. Genetic and environmental variation in anther, pollen and pistil dimensions in sesame. Sexual Plant Reproduction 9: 228232. PLATENKAMP, G. A. J., AND R. G. SHAW. 1993. Environmental and genetic maternal effects on seed characters in Nemophila menziesii. Evolution 47: 540555. RAUSHER, M. D., AND S.-M. CHANG. 1999. Stabilization of mixed-mating systems by differences in the magnitude of inbreeding depression for male and fitness components. American Naturalist 154: 242248. ROBERTSON, A. W., A. DIAZ, AND M. R. MACNAIR. 1994. The quantitative genetics of floral characters in Mimulus guttatus. Heredity 72: 300 311.

746

A MERICAN J OURNAL

OF

B OTANY

[Vol. 93

ROSS, M. D. 1970. Evolutions of dioecy from gynodioecy. Evolution 24: 827828. SAS INSTITUTE. 2000. JMP statistics and graphics guide. SAS Institute, Cary, North Carolina, USA. SCHULTZ, S. T. 2002. Partial male sterility and the evolution of nuclear gynodioecy in plants. Genetical Research 80: 187195. SHAW, R. G., D. L. BYERS, AND F. H. SHAW. 1998. Genetic components of variation in Nemophila menziesii undergoing inbreeding morphology and flowering time. Genetics 150: 16491661. TAYLOR, D. R., S. TRIMBLE, AND D. E. MCCAULEY. 1999. Ecological genetics of gynodioecy in Silene vulgaris: relative fitness of females

and hermaphrodites during the colonization process. Evolution 53: 745751. UYENOYAMA, M. K., K. E. HOLSINGER, AND D. M. WALLER. 1993. Ecological and genetic factors directing the evolution of selffertilization. Oxford Surveys in Evolutionary Biology 9: 327381. VALDEYRON, G., B. DOMMEE, AND A. VALDEYRON. 1973. Gynodioecy: another computer simulation model. American Naturalist 107: 454 459. WILLIAMS, H. L., AND C. B. FENSTER. 1998. Ecological and genetic factors contributing to the low frequency of male sterility in Chamaecrista fasciculata (Fabaceae). American Journal of Botany 85: 12431250.