American Journal of Botany 91(11): 1802–1808. 2004.
MATERNAL
AND PATERNAL CONTRIBUTIONS TO THE
FITNESS OF HYBRIDS BETWEEN RED AND WHITE MULBERRY
(MORUS, MORACEAE)1
KEVIN S. BURGESS2,4
BRIAN C. HUSBAND3
AND
Biology Department, University of Virginia, Charlottesville, Virginia 22904-4328 USA; 3Department of Botany, University of Guelph, Guelph, Ontario, Canada, N1G 2W1
2
The fitness of hybrids depends on the genetic disparity between parental taxa and the magnitude of their nuclear and non-nuclear contributions. To estimate the role of non-nuclear effects, we crossed red (R), white (W) and hybrid (H) mulberry in all combinations and compared the magnitude of maternal and paternal effects on offspring fitness (seed set, germination, survival and aboveground biomass) in a greenhouse environment. Variation in offspring fitness was determined largely by the identity of the maternal parent; specifically, progeny with white mothers had the highest cumulative fitness. As fathers, red, white, and hybrid mulberry had no effect on fitness, and maternal 3 paternal interactions were significant only for survival. Individual cross-types differed significantly for all fitness components except seed set. Offspring from hybrid crosses (W 3 R, H 3 R, H 3 W) often differed from at least one of the within-parent crosses (W 3 W, R 3 R) as well as from other hybrid crosses, although their fitness values never exceeded the most fit parent. Reciprocal crosses differed in only two of 15 possible parental combinations: W 3 H (cumulative fitness) and W 3 R (aboveground biomass). Overall, the strong asymmetry in magnitude of maternal and paternal effects suggests that fitness of hybrid mulberry is governed largely by non-nuclear, parental effects. Key words: common garden; cytonuclear interactions; cytoplasmic contributions; hybridization; maternal effects; Moraceae; Morus; non-nuclear effects; nuclear contributions; paternal effects.
Matings between genetically distinct populations or closely related species can often result in hybrid offspring (Arnold, 1997). This process can have two evolutionary consequences, which are seemingly opposed (Grant, 1981; Slatkin, 1987; Burke and Arnold, 2001). On the one hand, hybridization may homogenize parental genomes, a process that can oppose adaptation to local environments and lead to the breakdown of species barriers (Mayr, 1963; Dobzhansky et al., 1977; Slatkin, 1985). Alternatively, hybridization can introduce new genes or gene combinations into parental genomes and thereby influence a population’s evolutionary response to selection and possibly species formation (Levin, 1979; Harrison, 1990; Arnold, 1992; Rieseberg and Wendel, 1993; Arnold, 1997; Rieseberg, 1997). The evolutionary consequences of hybridization will depend, in part, on the fitness of hybrids relative to their parents (Rieseberg and Carney, 1998; Burke and Arnold, 2001; Arnold et al., 1999, 2001). Hybrids are generally expected to have lower viability and fertility than their parents as a result of negative genetic interactions between disparate parental genomes (Mayr, 1963; Dobzhansky, 1951; Coyne and Orr, 1998; Turelli et al., 2001). However, recent surveys indicate that hybrids can also be more fit or intermediate in fitness relative to parental taxa (Arnold and Hodges, 1995; Arnold, 1997; Arnold et al., 1999, 2001). This variation in fitness likely reflects the magnitude of genetic Manuscript received 16 December 2003; revision accepted 3 August 2004. The authors thank G. Mouland, V. McKay, and M. Smith for logistical assistance at Point Pelee National Park; A. Crawford, C. Leduc, and S. Spender for assistance with pollinations, fruit collection, plant preparation, and data collection; L. Commandeur, D. Duval, J. Gerrath, B. Kennedy, N. Prigoda, H. Sabara, and J. Tindall for help with the final harvest, and Dr. Laura Galloway and two anonymous reviewers for their helpful comments on the manuscript. Financial assistance was provided by the University of Guelph Botany Department, the Ontario Ministry of Natural Resources and a Natural Sciences and Engineering Research Council of Canada Grant to BCH. 4 E-mail:
[email protected]. 1
differences between their parents, the frequency of genes with antagonistic effects, and the relative importance of their nuclear and non-nuclear contributions to fitness. The nuclear genome is inherited equally from both parents, and, depending on the genetic basis of species differences, can be influenced by chromosomal rearrangements (Rieseberg et al., 2000; Rieseberg, 2001), polyploidy (Wendel, 2000), interactions between parental genes (nuclear 3 nuclear interactions) (Rieseberg et al., 1996, 2000; Burke and Arnold, 2001), or interactions between the nuclear genotype of the hybrid and its environment (genotype 3 environment interactions) (Arnold, 1997). Non-nuclear contributions or ‘‘parental effects,’’ represent contributions to fitness outside the effects of nuclear genes and may include cytoplasmic effects (Roach and Wulff, 1987; Wade, 1998; Levin, 2003), unequal contributions to the endosperm (including genomic imprinting [Haig and Westoby, 1991]) and the influence of the parental environment on the provisioning or phenotype of offspring (Mazer and Gorchov, 1996; Rossiter, 1998; Lacey, 1998; Wade, 1998). Regardless of the cause, non-nuclear effects are transmitted differentially by the parents and thus manifest themselves as fitness differences in reciprocal crosses between two parental genotypes (Mazer and Gorchov, 1996; Roff, 1998; Shaw and Byers, 1998; Levin, 2003). Non-nuclear effects have been reported for many plant species and for several traits (Roach and Wulff, 1987; Rossiter, 1996; Donohue and Schmitt, 1998). For example, reciprocal crosses in cultivated plants have revealed both maternal and paternal influences on seed set, germination, biomass, survival (Roach and Wulff, 1987) and fertility (Wright, 1977). Similar parental effects on fitness have been found in crosses involving natural populations (Andalo et al., 1999; Lacey and Herr, 2000; Galloway, 2001; Galloway and Fenster, 2001). Non-nuclear effects may also play an important role in determining the phenotype and fitness of hybrids between species. Camp-
1802
November 2004]
BURGESS
AND
HUSBAND—MATERNAL
bell and Waser (2001) showed that survival of hybrids between Ipomopsis aggregata and I. tenuituba depended on the species of the maternal parent, and both nuclear and cytonuclear interactions affected fitness in Louisiana Iris hybrids (Burke et al., 1998b). To date, however, relatively few quantitative studies have explored the relative importance of nuclear vs. non-nuclear effects on phenotypic variation in interspecific hybrids (Levin, 2003). In this study we examined sources of variation in the fitness of hybrids formed between red (Morus rubra L.) and white (Morus alba L.) mulberry in Ontario, Canada. Red mulberry is a wind-pollinated, dioecious, understory tree species that is native to forested habitats in eastern North America. At the northern limit of its range, red mulberry hybridizes with the more abundant and introduced white mulberry, a wind-pollinated, dioecious tree species commonly found in open fields, hedgerows, and disturbed environments. Genetic and morphological studies of natural populations within this region indicate that hybrids are present and are in close proximity to both parental taxa (K. S. Burgess and B. C. Husband, unpublished data, University of Guelph). The two species differ in fertility, microhabitat preference and abundance and may affect the fitness of their hybrids through nuclear and non-nuclear pathways. Here, we cross-pollinated red, white, and hybrid mulberry in all possible directions and compared the fitnesses of their progeny in a greenhouse environment. We asked the following questions: (1) Are hybrid mulberry less fit than their parental genotypes? (2) Do hybrid classes differ in fitness? and (3) Do the maternal and paternal taxa make different contributions to fitness? MATERIAL AND METHODS Crossing design—Experimental pollinations were conducted at Point Pelee National Park, Ontario, Canada (ø UTM [Universal Transverse Mercator] coordinates: 373900 N [Northing]; 464500 E [Easting], NAD [North American Datum] 83). All trees occurred in partially disturbed forests within the Carolinian Forest Zone. Twelve trees (six male and six female) each of red (R), hybrid (H), and white (W) mulberry were selected for the experiment and were cross-pollinated in all possible combinations, including all parental and reciprocal crosses. The cross-types consisted of three different maternal 3 paternal classes: (1) within-parent (R 3 R, W 3 W); (2) F1 hybrid (R 3 W, W 3 R) and (3) later generation hybrid (R 3 H, H 3 R, W 3 H, H 3 W, H 3 H). Each cross-type was replicated using six different male-female pairs of trees. Hybrids were included as a parental class so we could also examine parental effects on later generation hybrids. Finding enough naturally occurring male and female hybrids that are reproductive in any one location was difficult. As a consequence, our hybrids included both F1 and later-generation hybrids. Although the hybrid class isn’t homogeneous, previous molecular studies based on Randomly Amplified Polymorphic DNA (RAPD) and chloroplast DNA (cpDNA) sequences indicate that most contain predominantly white nuclear genomes and white cytoplasmic genomes (K. S. Burgess and B. C. Husband, unpublished data, University of Guelph). Moreover, they represent those hybrid genotypes most likely to be involved in natural backcrosses in this species. All parental genotypes were confirmed with speciesspecific RAPD markers (K. S. Burgess and B. C. Husband, unpublished data, University of Guelph). Forty inflorescences were bagged on each female tree prior to anthesis to prevent pollen contamination. At anthesis, pollen was collected from male catkins and applied to 10 female catkins per tree using fine paintbrushes. Fruits were collected in July and stored dry at 48C for 5 mo. Fitness components—Offspring from all nine cross-types were evaluated for four fitness components: seed set, seed germination, survival to 12 wk, and aboveground biomass at 12 wk. A multiplicative fitness function of these
CONTRIBUTIONS TO HYBRID FITNESS
1803
four fitness components was then used to determine cumulative fitness for all cross-combinations. To estimate seed set, the proportion of ovules setting seeds was calculated. We wetted each fruit for 24 h, removed the pulp and counted the number of filled seeds per fruit. To account for differences in ovule number among species, we expressed seed number as a percentage of ovule number. The mean number of ovules per catkin was estimated for 22 red, 27 white, and 10 hybrid trees using two catkins/tree. The mean for each species was 22.5, 15.4, and 17.9 ovules per catkin, respectively. Percentage germination was estimated by sowing seeds onto moist filter paper in petri dishes (maximum of 25 seeds/ dish) and incubated in a growth chamber for two wk (12 h daylight at 268C and 12 night hours at 228C). Seeds were classified as having germinated if the radical had emerged and the cotyledons were beginning to expand. Seedling survival and growth were assessed in a greenhouse environment. All plants that germinated and possessed fully expanded cotyledons were transplanted into 10 cm (4 inch) pots containing perlite, turface, and ProMix in 1 : 1 : 6 proportions. Plants were randomly assigned positions on the bench and grown for 12 wk (16 h daylight and 8 h dark at 218C). To account for mortality due to transplant shock, plants that died within the first 10 d of transplanting were removed from the design. At the end of the 12th wk, survival was measured as the number of plants alive per cross-type, expressed as a proportion of the initial number of seedlings transplanted (after removal). The aboveground portions of all plants were harvested, dried for 144 h at 608C and weighed. A cumulative fitness measure was calculated for each male-female pair separately by multiplying the values obtained for each of the response variables (seed set, percentage germination, survival to 12 wk and aboveground biomass at 12 wk). Analysis—The sources of variation in fitness of red, white, and hybrid mulberry were assessed using a two-way analysis of variance (ANOVA) with maternal parent, paternal parent, and maternal 3 paternal interaction as the sources of variation. The means of the six male-female pairs per cross type served as replicates. To meet the assumptions of normality, percentage germination and survival were arcsine transformed and cumulative fitness was square-root transformed. All values were reported as back-transformed means. Tukey’s Honestly Significant Different (HSD) test was used to compare means of different parental classes. We were also interested in testing for differences between individual parental and hybrid cross-types. Therefore, whenever one of the main effects in the two-way ANOVA was significant, we compared all cross-type means using a Tukey’s HSD test. If there were non-nuclear effects on the fitness of offspring, we expected to find a strong asymmetry between maternal and paternal effects in the twoway ANOVA and differences among reciprocal crosses between red and white, red and hybrid and white and hybrid trees. All statistical analyses were performed using JMP statistical software Version 5.0 (SAS Institute, 2002).
RESULTS In total, pollinations involved 52 male-female pairs, distributed across the nine cross-types with 2 to 6 replications per cross-type (X¯ 5 5.4). On average, 4.6 female catkins were pollinated per pair, resulting in a total of 242 fruits. The number of plants per cross-type ranged from 21 to 409 individuals (X¯ 5 158.8), and a total of 1429 seedlings were used for the greenhouse comparison. Seed set ranged from 34 to 85% among cross-types. In the two-way ANOVA, only the identity of the maternal parent had a significant effect on seed set (Table 1, Fig. 1A). Crosses with red mulberry as the maternal parent had significantly higher seed set than those with hybrid mothers but did not differ from white mothers. Paternal effects and maternal 3 paternal interactions on seed set were not significant. No significant differences among individual cross-types were found based on the Tukey’s means comparison (Fig. 2A). Maternal parent also had a significant effect on germination,
1804
AMERICAN JOURNAL
TABLE 1. Effect of maternal and paternal parent on the fitness of offspring from crosses among red, white and hybrid mulberry. Results for (A) seed set, (B) germination, (C) survival, (D) above ground biomass, and (E) cumulative fitness were based on a 2-factor ANOVA. Variable
A) Seed set B) Germinationa C) Survivala D) Biomass E) Cumulative fitnessb
Source of variation
Maternal Paternal Maternal Maternal Paternal Maternal Maternal Paternal Maternal Maternal Paternal Maternal Maternal Paternal Maternal
3 Paternal 3 Paternal 3 Paternal 3 Paternal 3 Paternal
* P , 0.05; ** P , 0.01; *** P , 0.001. a Data were arcsine transformed. b Data were square root transformed.
df
2 2 4 2 2 4 2 2 4 2 2 4 2 2 4
MS
F statistics
0.59 5.58** 0.13 1.22 0.04 0.35 0.30 11.25*** 1.78 1.92 0.10 0.66 0.46 3.80* 0.0002 0.002 0.43 3.63* 15.30 16.10*** 0.57 0.60 0.67 0.70 372.92 17.86*** 59.78 2.86 52.34 2.51
OF
BOTANY
[Vol. 91
with progeny from white mothers having higher germination rates than any other parent (Table 1, Fig. 1B). The identity of the father did not affect germination, nor was there a maternal 3 paternal interaction. The means comparison revealed some significant differences among individual cross-types (Fig. 2B). Hybrid cross-types differed significantly from each other, and one hybrid class (W 3 H) had higher germination than the R 3 R within-parent cross. None of the reciprocal crosses involving red and white, red and hybrid, or hybrid and white parents differed from each other (Fig. 2B). The two-way ANOVA indicated a significant effect of maternal parent on survival (Table 1, Fig. 1C). Offspring from hybrid mothers had higher survival than those from white mothers but were not different from offspring of red mothers. Survival was the only fitness component to show a significant maternal 3 paternal interaction. Survival ranged from 60 to 98% among individual cross-types (Fig. 2C). Tukey’s means comparison revealed no differences between the two withinparent cross-types (R 3 R, W 3 W). Hybrid crosses did not differ significantly among each other, but some hybrid crosses (W 3 R) had significantly lower survival than some parental crosses (R 3 R). Later generation hybrids did not differ significantly from their parents but some (H 3 H, H 3 W, W 3 H) had higher survival than one of the first-generation hybrid
Fig. 1. Mean fitness of offspring derived from red (R), white (W) and hybrid (H) maternal plants in a greenhouse environment: (A) Seed set, (B) proportion that germinated, (C) proportion surviving to 12 wk, and (D) aboveground biomass at 12 wk. Means with different letters are significantly different based on a Tukey’s post hoc comparison of means.
November 2004]
BURGESS
AND
HUSBAND—MATERNAL
CONTRIBUTIONS TO HYBRID FITNESS
1805
Fig. 2. Fitness differences among nine maternal 3 paternal cross-combinations between red (R), white (W), and hybrid (H): (A) seed set, (B) proportion that germinated, (C) proportion surviving to 12 wk, and (D) aboveground biomass at 12 wk. Means with different letters are significantly different based on a Tukey’s means comparison.
crosses (W 3 R) (Fig. 2C). There were no reciprocal differences in survival among first or later generation hybrids. With respect to biomass, only the identity of the maternal parent had a significant impact; offspring with white mothers had higher biomass than those with hybrid or red maternal backgrounds (Table 1, Fig. 1D). No differences were observed between the within-parent crosses, but the W 3 W offspring had higher biomass than two hybrid crosses (R 3 H and R 3 W) (Fig. 2D). There were no differences among hybrid crosses except R 3 H offspring, which had significantly lower biomass than W 3 R. Biomass differed among reciprocal F1 crosses, with W 3 R progeny having more mass than R 3 W (Fig. 2D). The two-way ANOVA indicated that the identity of the maternal parent had significant effects on cumulative fitness (Table 1). In particular, offspring from white mothers had the highest mean fitness, while offspring from red and hybrid mothers did not differ significantly from each other (Fig. 3A). Cumulative fitness was the only fitness component for which within-parent crosses differed, with R 3 R offspring having the lowest value (Fig. 3B). Differences in cumulative fitness were also found between parental cross-types and hybrids. Specifically, R 3 R crosses had lower cumulative fitness than W 3 H crosses. W 3 W progeny had higher cumulative fitness than all other hybrid cross-types except the W 3 H cross. Reciprocal differences in fitness were also found; H 3 W
crosses had lower cumulative fitness than the reciprocal (Fig. 3B). DISCUSSION Our results indicate substantial variation in the fitness of offspring from red, white, and hybrid mulberry parents. We found differences among cross-types for germination, survival, aboveground biomass, and cumulative fitness, but no variation in seed set. Depending on the fitness component, there were differences between parental crosses (W 3 W, R 3 R), between parental and hybrid (W 3 H, R 3 H, R 3 W, H 3 H) crosses, as well as among hybrid crosses. Across all fitness components, reciprocal differences were found for only two crosses: R 3 W (for biomass) and H 3 W (for cumulative fitness). In both cases, fitness was highest when white mulberry was the maternal parent. Importantly, most if not all variation observed among cross types was attributable to the genotype of the maternal parent. For most fitness measures, offspring with white mothers had higher mean fitness than those with red or hybrid mothers. In contrast, the genotype of the paternal parent had no effect on fitness except for survival, in which case the paternal effect varied across maternal genotypes. Below we begin by discussing the differences among parental and hybrid classes and then evaluate the evidence for non-nuclear effects.
1806
AMERICAN JOURNAL
Fig. 3. Mean cumulative fitness of (A) offspring from red, white and hybrid mulberry mothers, and (B) all nine maternal 3 paternal cross-combinations between red (R), white (W), and hybrid (H). Means with different letters indicate significance based on a Tukey’s post hoc comparison of means.
While the fitnesses of hybrid crosses in mulberry often differed from at least one parental cross, their fitness values never exceeded the most fit parent. This result is consistent with that of Campbell and Waser (2001) who showed that growth and survival of F1 hybrids between Ipomopsis aggregata and I. tenuituba were as high as but never exceeded the fitness of within-parent crosses. In contrast, Burke et al. (1998a) and Emms and Arnold (1997) found that Iris hexagona 3 I. fulva hybrids had fitnesses equal to or higher than their parents. These and numerous other studies (Burke and Arnold, 2001; Arnold et al., 2001; Johnston et al., 2001a, b) suggest that the fitnesses of hybrids relative to their parents are highly variable among species pairs. Increased hybrid fitness relative to that of their parents has been attributed to favorable combinations of genetic factors segregating from two disparate parental genomes (Rieseberg et al., 1996, 1999; Burke and Arnold, 2001). Based on our results, there is no evidence for such transgressive segregation in hybrid mulberry. However, the fitness relationships between hybrid mulberry and its red and white parents are based on a greenhouse environment and need to be confirmed with transplant experiments in natural environments. We observed significant differences among certain hybrid crosses for all but one fitness-component. In most cases, some of the later generation hybrid crosses were more fit than other
OF
BOTANY
[Vol. 91
hybrid crosses, including F1s. It is widely recognized that hybrid classes can differ in fitness, depending on the magnitude of epistatic or additive genetic effects (Arnold and Hodges, 1995; Arnold, 1997; Burke and Arnold, 2001). Although F1 hybrids can experience heterosis and hybrid inferiority can be restricted to later generation hybrids (Arnold et al., 1999; Burke and Arnold, 2001), later generation crosses have also resulted in relatively fit hybrids due to favorable allelic combinations or positive epistasis (Rieseberg et al., 1996; Arnold et al., 1999; Burke and Arnold, 2001). These positive interactions appear to be operating in mulberry; however, other explanations related to our experimental design can not be excluded. In particular, the variation in fitness among the hybrid crosses may also be a result of the heterogeneous nature of hybrid parents used in our study. Most of the hybrid parents consisted of predominantly white mulberry nuclear (based on RAPD profiles) and cytoplasmic (cpDNA) genome. This may explain why the fitness of W 3 H crosses was similar to that in W 3 W crosses. In addition, the later-generation hybrid parents have already been exposed to selection and may represent those combinations of nuclear genomes with relatively high fitness. One of the main goals of this study was to investigate the importance of non-nuclear, or parental, effects on fitness. We achieved this by testing for fitness differences between paternal and maternal effects in a two-way ANOVA of all crosstypes and between reciprocal crosses for individual crosstypes. Significant differences were found for only two of 15 possible reciprocal comparisons among the five fitness components. In contrast, the ANOVA of maternal and paternal effects showed consistently strong impacts of the maternal parent and a lack of paternal influences on offspring fitness. The low frequency of reciprocal fitness differences may have been the result of limited sample size and hence less statistical power in those tests. Nevertheless, the strong maternal influence, and a general lack of paternal contribution to fitness, are indicative of strong non-nuclear (maternal) effects. Similarly, Burke et al. (1998a) found strong cytoplasmic effects on the fitnesses of Iris hexagona 3 I. fulva hybrids, as did Campbell and Waser (2001) for reciprocal crosses between Ipomopsis aggregata and I. tenuituba. Collectively, these results support Levin’s suggestion that non-nuclear effects may have been underestimated in genetic interactions between species (Levin, 2003). The maternal effects influencing the fitness of hybrid mulberry may be caused by one of three different mechanisms (1) maternal nuclear contributions to the endosperm; (2) influence of the maternal environment on offspring; and (3) maternally inherited cytoplasm (chloroplast or mitochondrial DNA) and cytonuclear interactions (Roach and Wulff, 1987). While our experiment was not designed to disentangle these specific causes, we feel that cytoplasmic effects are most likely. Endosperm effects are expected to weaken with age, as the offspring becomes increasingly reliant on its external environment for nutrition. In contrast, in mulberry maternal effects are strong throughout five different life stages, and, based on ongoing research, this effect persists into the second and third years. Alternatively, the effect of maternal environment seems unlikely because the parental individuals used in this study were located in a common natural habitat and the offspring were grown in a common greenhouse environment. For these reasons, we feel it most likely that maternal effects in mulberry are the result of cytoplasmic contributions by the ma-
November 2004]
BURGESS
AND
HUSBAND—MATERNAL
ternal parent. Molecular and quantitative genetic analyses of F1 and later generation hybrids will be needed to disentangle the sources of maternal effects in mulberry with more certainty (Burke et al., 1998b; Levin, 2003). Evolutionary consequences of maternal effects—The maternal effects observed in our study are a significant source of variation in fitness with important implications for adaptation in mulberry. Maternal effects generate a strong fitness advantage for offspring of white mulberry by enhancing their germination, survival and growth relative to offspring from red and hybrid mulberry. Therefore, genes controlling fitness in mulberry are contained in the maternal generation yet influence selection during the offspring generation. Numerous theoretical models have shown that this time lag in the expression of fitness can affect the rate and direction of the response of hybrids to selection (Kirkpatrick and Lande, 1989; Wolf and Brodie, 1998; Thiede, 1998; Wolf, 2000). For example, if selection in the maternal mulberry environment matches that in the offspring generation, then maternal effects can reinforce adaptive selection for white mulberry or, conversely, maladaptation of red mulberry. At this point, information regarding the selective environments of juvenile and adult trees is insufficient to evaluate this outcome. Maternal effects may also influence the dynamics of species hybrid zones. Specifically, hybrids of a single nuclear composition will, on average, experience differential success depending on their maternal parent’s cytoplasmic background. In this sense, maternal effects can mask or counter the effects of maladapted nuclear genotypes and thereby slow their elimination from a population. This process has two consequences for mulberry. First, by elevating the fitness of hybrids to that of their white parents, maternal effects may weaken the strength of post-zygotic reproductive isolation between these species. Secondly, this process will result in the transfer of cytoplasm across nuclear genotypes. As a result, nuclear genotypes, including those with predominantly red genomes, will increasingly carry white cytoplasms. This is consistent with preliminary data on the sequencing studies of hybrid mulberry, which indicate there are more hybrids with white chloroplast DNA than one would predict from the frequencies of the parent species. In summary, the fitness of hybrid mulberry in a common garden setting is determined predominately by its maternal background. Progeny from white mulberry mothers had the highest cumulative fitness. In contrast, paternal contributions to fitness were not significant and maternal 3 paternal interactions were significant only for survival. These results suggest that non-nuclear, maternal effects are a dominant influence on the fitness of hybrid mulberry. These uniparental genetic effects may influence the genetic composition and fate of parental genotypes in the red-white mulberry contact zone. LITERATURE CITED ANDALO, C., S. J. MAZER, B. GODELLE, AND N. MACHON. 1999. Parental environmental effects on the life history traits in Arabidopsis thaliana (Brassicaceae). New Phytologist 142: 173–184. ARNOLD, M. L. 1992. Natural hybridization as an evolutionary process. Annual Review of Ecology and Systematics 23: 237–261. ARNOLD, M. L. 1997. Natural hybridization and evolution. Oxford University Press, New York, New York, USA. ARNOLD, M. L., M. R. BULGER, AND J. M. BURKE. 1999. Natural hybridization: how low can you go and still be important? Ecology 80: 371–381.
CONTRIBUTIONS TO HYBRID FITNESS
1807
ARNOLD, M. L., AND S. A. HODGES. 1995. Are natural hybrids fit or unfit relative to their parents? Trends in Ecology and Evolution 10: 67–71. ARNOLD, M. L., E. K. KENTNER, J. A. JOHNSTON, S. CORNMAN, AND A. C. BOUCK. 2001. Natural hybridisation and fitness. Taxon 50: 93–104. BURKE, J. M., AND M. L. ARNOLD. 2001. Genetics and the fitness of hybrids. Annual Review of Genetics 35: 31–52. BURKE, J. M., S. E. CARNEY, AND M. L. ARNOLD. 1998a. Hybrid fitness in the Louisiana irises: analysis of parental and F1 performance. Evolution 52: 37–43. BURKE, J. M., T. J. VOSS, AND M. L. ARNOLD. 1998b. Genetic interactions and natural selection in Louisiana iris hybrids. Evolution 52: 1304–1310. CAMPBELL, D. R., AND N. M. WASER. 2001. Genotype-by-environment interaction and the fitness of plant hybrids in the wild. Evolution 55: 669– 676. COYNE, J. A., AND H. A. ORR. 1998. The evolutionary genetics of speciation. Philosophical Transactions of the Royal Society of London, B 353: 287– 305. DOBZHANSKY, T., F. J. AYALA, G. L. STEBBINS, AND J. W. VALENTINE. 1977. Evolution. W. H. Freeman, San Francisco, California, USA. DONOHUE, K., AND J. SCHMITT. 1998. Maternal environmental effects in plants: adaptive plasticity? In T. A. Mousseau and C. W. Fox [eds.], Maternal effects as adaptations, 137–158. Oxford University Press, New York, New York, USA. EMMS, S. K., AND M. L. ARNOLD. 1997. The effect of habitat on parental and hybrid fitness: transplant experiments with Louisiana irises. Evolution 51: 1112–1119. GALLOWAY, L. F. 2001. Parental environmental effects on life history in the herbaceous plant Campanula americana. Ecology 82: 2781–2789. GALLOWAY, L. F., AND C. B. FENSTER. 1999. The effect of nuclear and cytoplasmic genes on fitness and local adaptation in an annual legume, Chamaecrista fasciculate. Evolution 53: 1734–1743. GRANT, V. R. 1981. Plant speciation, 2nd ed. Columbia University Press, New York, New York, USA. HAIG, D., AND M. WESTOBY. 1991. Genomic imprinting in the endosperm: its effect on seed development in crosses between species, and differences between ploidies of the same species, and its implications for the evolution of apomixis. Philosophical Transactions of the Society of London, B 333: 1–13. HARRISON, R. G. 1990. Hybrid zones: windows on evolutionary process. In D. Futuyma and J. Antonovics [eds.], Oxford surveys in evolutionary ecology, vol. 7, 70–128. Oxford University Press, Toronto, Ontario, Canada. JOHNSTON, J. A., R. A. WESSELINGH, A. C. BOUCK, L. A. DONOVAN, AND M. L. ARNOLD. 2001a. Intimately linked or hardly speaking? The relationship between genotype and environmental gradients in a Louisiana iris hybrid population. Molecular Ecology 10: 673–681. JOHNSTON, J. A., D. J. GRISE, L. A. DONOVAN, AND M. L. ARNOLD. 2001b. Environment-dependent performance and fitness of Iris brevicaulis, I. fulva (Iridaceae), and hybrids. American Journal of Botany 88: 933–938. KIRKPATRICK, M., AND R. LANDE. 1989. The evolution of maternal characters. Evolution 43: 485–503. LACEY, E. P. 1998. What is an adaptive environmentally induced parental effect? In T. A. Mousseau and C. W. Fox [eds.], Maternal effects as adaptations, 54–66. Oxford University Press, New York, New York, USA. LACEY, E. P., AND D. HERR. 2000. Parental effects in Plantago lanceolata L. III. Measuring parental temperature effects in the field. Evolution 54: 1207–1217. LEVIN, D. A. 1979. Hybridization: an evolutionary perspective. Dowden, Hutchinson, and Ross, Stroudsburg, Pennsylvania, USA. LEVIN, D. A. 2003. The cytoplasmic factor in plant speciation. Systematic Botany 28: 5–11. MAYR, E. 1963. Animal species and evolution. Belknap Press of Harvard University Press, Cambridge, Massachusetts, USA. MAZER, S. J., AND D. L. GORCHOV. 1996. Parental effects on progeny phenotype in plants: distinguishing genetic and environmental causes. Evolution 50: 44–53. RIESEBERG, L. H. 1997. Hybrid origins of plant species. Annual Review of Ecology and Systematics 28: 359–389. RIESEBERG, L. H. 2001. Chromosomal rearrangements and speciation. Trends in Ecology and Evolution 16: 351–359. RIESEBERG, L. H., M. A. ARCHER, AND R. K. WAYNE. 1999. Transgressive segregation, adaptation and speciation. Heredity 83: 363–372.
1808
AMERICAN JOURNAL
RIESEBERG, L. H., S. J. E. BAIRD, AND K. A. GARDNER. 2000. Hybridization, introgression, and linkage evolution. Plant Molecular Biology 42: 205– 224. RIESEBERG, L. H., AND S. E. CARNEY. 1998. Tansley review no. 102: plant hybridization. New Phytologist 140: 599–624. RIESEBERG, L. H., B. SINERVO, C. R. LINDER, M. C. UNGERER, AND D. M. ARIAS. 1996. Role of gene interactions in hybrid speciation: evidence from ancient and experimental hybrids. Science 272: 741–745. RIESEBERG, L. H., AND J. F. WENDEL. 1993. Introgression and its consequences in plants. In R. G. Harrison [ed.], Hybrid zones and the evolutionary process, 70–109. Oxford University Press, New York, New York, USA. ROACH, D. A., AND R. D. WULFF. 1987. Maternal effects in plants. Annual Review of Ecology and Systematics 18: 209–235. ROFF, D. A. 1998. The detection and measurment of maternal effects. In T. A. Mousseau and C. W. Fox [eds.], Maternal effects as adaptations, 83– 111. Oxford University Press, New York, New York, USA. ROSSITER, M. C. 1996. Incidence and consequences of inherited environmental effects. Annual Review of Ecology and Systematics 27: 451–476. ROSSITER, M. C. 1998. The role of environmental variation in parental effects expression. In T. A. Mousseau and C. W. Fox [eds.], Maternal effects as adaptations, 112–134. Oxford University Press, New York, New York, USA. SHAW, R. G., AND D. L. BYERS. 1998. Genetics of maternal and paternal
OF
BOTANY
[Vol. 91
effects. In T. A. Mousseau and C. W. Fox [eds.], Maternal effects as adaptations, 83–96. Oxford University Press, New York, New York, USA. SLATKIN, M. 1985. Gene flow in natural populations. Annual Review of Ecology and Systematics 16: 393–430. SLATKIN, M. 1987. Gene flow and the geographic structure of natural populations. Science 236: 787–792. THIEDE, D. A. 1998. Maternal inheritance and its effects on adaptive evolution: a quantitative genetic analysis of maternal effects in a natural plant population. Evolution 52: 998–1015. TURELLI, M., N. H. BARTON, AND J. A. COYNE. 2001. Theory and speciation. Trends in Ecology and Evolution 16: 330–343. WADE, M. J. 1998. The evolutionary genetics of maternal effects. In T. A. Mousseau and C. W. Fox [eds.], Maternal effects as adaptations, 5–21. Oxford University Press, New York, New York, USA. WENDEL, J. F. 2000. Genome evolution in polyploids. Plant Molecular Biology 42: 225–249. WOLF, J. B. 2000. Gene interactions from maternal effects. Evolution 54: 1882–1898. WOLF, J. B., AND E. D. BRODIE. 1998. The coadaptation of parental and offspring characters. Evolution 52: 299–308. WRIGHT, S. 1977. Evolution and the genetics of populations; experimental results and evolutionary deductions. University of Chicago Press, Chicago, Illinois, USA.
