The Journal of Heredity 2001:92(5)
© 2001 The American Genetic Association 92:409-414
The Genetic Basis of Floral Variation in Senecio jacobaea (Asteraceae)
From the Department of Systematic Botany, Lund University, Ö. Vallgatan 18-20, S-22361 Lund, Sweden.
Address correspondence to S. Andersson at the address above or e-mail: stefan.andersson@sysbot.lu.se.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The self-incompatible composite Senecio jacobaea (ragwort) exhibits geographic variation in the frequency of rayed and discoid (rayless) individuals. Hybrid progenies from within- and between-morph crosses were established in a seminatural (garden) environment to determine whether patterns of segregation conform to single-gene predictions (as found in other Senecio species), whether the direction of dominance is conducive to rapid evolutionary change in ray morphology, and whether geographically distant populations of the discoid morph utilize the same or different genes to suppress ray development. Data from segregating F2 and BC families were consistent with a genetic model involving one major locus and an unknown number of modifiers. Analysis of F1 progenies from different intermorph crosses using the same rayed plant as a seed parent revealed a variable and incomplete pattern of dominance, with a trend toward partial dominance in some crosses. Hybridizations between discoid populations produced a few rayed progeny (4%), but there was no tendency for the frequency of rayed progeny to increase with the geographic distance separating the parent populations. Results of this study indicate that major mutations have been important for the evolution of discoid populations of ragwort, that ray-suppressing mutations should be directly available to selection in most populations, and that the suppression of rays is conditioned by the same or similar gene(s) in Atlantic and Baltic populations of the discoid taxon.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Contrary to the neo-Darwinian view of adaptive evolution, with its emphasis on slow, gradual accumulations of allelic substitutions (Coyne and Lande 1985; Fisher 1958; Lande 1981), there is growing evidence that mutations of large effect can be selectively advantageous (Orr and Coyne 1992) and that extreme differences in morphology can be governed by a small number of major mutations (Bradshaw et al. 1995; Gottlieb 1984). However, the detection of major genes does not necessarily mean that the difference between two species or populations evolved rapidly (Coyne and Lande 1985). For example, the response to selection is expected to be slow when the novel, advantageous mutation is recessive (Haldane 1924), when the mutation has a low penetrance in the natural habitat (Andersson 1995), or when the mutation has deleterious pleiotropic effects on other aspects of fitness (Coyne and Lande 1985; Lande 1983). In addition, there are ample indications that a species can achieve the same adaptation by different genetic mechanisms (Andersson 1995; Fenster and Barrett 1994; Ford 1975; Mayer and Charlesworth 1992), stressing the importance of comparative genetic studies before any broad generalizations are made regarding the heritable basis of adaptation.
Specialized ray florets (rays), a distinctive feature of many genera in the sunflower family (Asteraceae), probably evolved under pollinator-mediated selection for more conspicuous flower heads, as evidenced by the drastic reductions in pollination efficiency following ray removal (Andersson 1991; Stuessy et al. 1986) and the high outcrossing rate of rayed plants in populations polymorphic for rayed and discoid individuals (Abbott and Irwin 1988; Marshall and Abbott 1982). However, if resource allocation to rays is costly (Andersson 1999) and rays become redundant as pollinator attractants (Andersson 1996), one would expect selection to favor a reduction or loss of the rays. Such reversal has occurred throughout the family (e.g., Bremer and Humphries 1993) and appears to have a simple genetic basis (Clausen et al. 1947; Fick 1967; Ford and Gottlieb 1990; Jackson and Dimas 1981). For example, the suppression of rays has been shown to be controlled by a single codominant locus in Senecio vulgaris and S. squalidus (Comes 1998; Ingram and Taylor 1982; Trow 1912). It remains to be seen whether this hypothesis can be generalized to other Senecio species that are polymorphic for rayed and discoid individuals.
Ragwort (Senecio jacobaea L.) exhibits extensive ecogeographic variation with a discoid, coastal morph, usually referred to as ssp. dunensis (Dumort.) Kadereit and Sell, replacing the type variety (ssp. jacobaea) in parts of Ireland, Scotland, Belgium, and the Netherlands, and having a more sparse and scattered occurrence in the Baltic region (Andersson 2001; Harper and Wood 1957; van der Meijden 1976). Although historical factors cannot be excluded, it seems reasonable to invoke geographically varying selection as one possible mechanism underlying this pattern (Andersson 2001). In the present study, I have collected data from a series of crossing experiments to determine whether the pattern of segregation in ragwort conforms to the single-gene system observed in other Senecio species, whether the pattern of dominance is conducive to rapid adaptive change in ray morphology, and whether geographically distant populations of the discoid phenotype utilize the same or different genes to suppress ray development.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Plant Material
Ragwort is a self-incompatible, insect-pollinated, biennial or perennial plant of open, grassy places, both in its native Eurasian range and in areas where the species occurs as a naturalized weed, for example, North America and Australia. Plants of ragwort are single-stemmed or branched from the base, the stems branching above the middle to give a loose corymbiform inflorescence with 20–300 heads. Each flower head is 12–25 mm in diameter and consists of 50–60 bisexual disk florets surrounded by 12–15 pistillate ray florets. Disk florets have a tubular corolla with five small radially symmetrical lobes and five connate anthers forming a cylinder around the style. Ray florets are much larger, with an 8–12 mm long, yellow, three-lobed, strap-shaped ligule surmounting a short corolla tube (Harper and Wood 1957). Heads on discoid plants have an extra whorl of disc florets instead of ray florets (see Results). Due to self-sterility, insect visitation is essential for full fruit set (Andersson 1996). Fruits from ray florets lack dispersal structures and require more time to germinate than disc fruits (McEvoy 1984). Chromosome counts indicate 2n = 32–40 for populations in the western and northern parts of Europe (Bain 1991; Harper and Wood 1957).
The plants used in the crossing experiments (henceforth parent plants) represent one rayed population (Lund) and five discoid populations (Visby, Sundsvall, Veluwe, Meijendel, Strathy Bay; Table 1). Plants from the Lund and Visby populations were transplanted from the field in 1993, while the other populations were represented by natural sib families derived from seeds collected in 1997. All parent plants were kept in an unheated, insect-proof greenhouse and enclosed in transparent perforated plastic bags before cross-pollination.
|
Analyses of Segregation and Morph Fitness
In 1993–1994 I established a series of F2 and backcross (BC) families from a cross between the Lund and Visby population. The F2s represent six crosses involving 12 different F1 hybrids, derived from a single pair of parent plants, whereas the BC progenies represent a randomly chosen F1 plant crossed with each of the parents. In the autumn of 1996 I planted 100 seedlings per cross in separate pots with sandy soil and randomized the pots in a dry, sunny part of an experimental garden. In 1997 the first flowering head on each F2 or BC plant was scored for the number of ray and disk florets, followed by the dissection of a randomly chosen outer floret to obtain data on floral morphology: length of the corolla (including the ray, if present), length of the ray (score 0 if absent), maximum width of the ray (score 0 if absent), distance between the two outermost lobes at the apex of the ray (score 0 if absent), number of lobes, lobe length, pistil length, and length of the anthers (score 0 if absent). The presence or absence of well-formed pollen in the anthers was also recorded. A plot of the first two principal components, based on the product-moment correlations among the quantitative characters, was used to display the variation among the F2 and BC plants. There was little variation in ray morphology among outer florets within the same head (Figure 1) or on different heads of the same plant, at least for plants flowering in the 1997 season (Andersson S, unpublished data).
|
To evaluate the single-gene hypothesis, I classified each F2 and BC individual as a "rayed homozygote" if the outer florets produced well-developed rays but lacked anthers, as a "discoid homozygote" if the outer florets were identical to the tubular, hermaphroditic disk florets, or as a "heterozygote" if the outer florets produced short, stubby rays or other types of intermediate or "novel" trait combinations, for example, tubular or deeply lobed rays, or rudimentary anthers without pollen. Chi-squared analyses were used to test for differences in the proportion of morphs among F2 families and to compare the observed morph frequencies with segregation ratios expected under different versions of the single-locus model (see Results). The overall fit of a particular model was examined by pooling the chi-squared values across the F2 and BC families and testing this statistic against a chi-squared distribution with degrees of freedom equal to the sum of all degrees of freedom.
Classification of F2 and BC hybrids into "homozygotes" and "heterozygotes" also provided a possibility to test for physiologically adverse effects of the allele causing discoid heads. To this end I used data from a phenotypic selection analysis (Andersson 2001) to estimate the mean performance for each of the three genotype classes. Particular attention was given to "yield components," that is, characters that should be positively correlated with fitness irrespective of growth conditions: the total number of heads produced, the number of flowers per head, the percent fruit set per head (after open pollination), and the total number of heads produced in the second flowering season. Head number was estimated from the number of side branches on each stem (for details, see Andersson 2001). To simplify comparison, the means of the "discoid homozygotes" and the "heterozygotes" were divided by the mean of the "rayed homozygotes."
Tests for Dominance and Allelism
In 1998–1999 I carried out a test for allelism using two (nonsib) plants from each of the discoid populations. Crosses were made in all possible combinations (including the five reciprocal within-population crosses) and the resulting progenies were grown in the garden plot. Besides classifying each F1 hybrid as discoid or rayed, I measured the length of the rays for each of the plants in the latter category. The presence of rayed offspring in a cross between two discoid populations would indicate that in these populations a discoid head is determined by different genes (complementary gene action). To assess patterns of dominance I crossed each of the 10 discoid parents (used as pollen donors) to a common rayed plant from the Lund population (used as a seed parent) and scored the resulting progeny for ray length. Differences in average ray length between F1 families were tested for significance using nested analyses of variance (ANOVA) with "donor population" and "plant within donor population" as (random) factors.
Assumptions
Inspection of nonpollinated heads on a sample of rayed and discoid parent plants (Andersson S, unpublished data) confirmed that all plants were self-sterile and that no pollen contamination had occurred in the greenhouse. Germination was high (approximately 95%) and seedling mortality was negligible, both in the between- and within-morph crosses, suggesting a limited potential for selection to affect patterns of segregation after seedling emergence. The parents were probably homozygous for genes with large phenotypic effects, as indicated by the lack of nonparental phenotypes in the field-collected progeny families (Andersson S, unpublished data) and in all within-population crosses (Table 2). Hence the presence of rayed offspring in crosses between discoid populations was assumed to reflect complementary gene action rather than segregation at a major locus for which one or both parents were heterozygous. However, given the obligately outbreeding nature of ragwort, there is no certainty that the parents were homozygous for all loci controlling flower development.
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Analyses of Segregation and Morph Fitness
More than 90% of the F2 and BC plants (740 of 800 plants) flowered and could be scored for flower morphology. The first principal component extracted 62% of the variance and was primarily a size component (Table 3), distinguishing discoid plants (low values, Figure 2) from plants with morphologically distinct ray florets (high values). The second axis accounted for a supplementary 17% of the variance and was primarily a shape component (Table 3), contrasting plants with parental phenotypes (high values) from plants with intermediate or "novel" flower types, for example, short stubby rays, deeply lobed rays, or ray corollas with a trumpet-like tube instead of a ligule (Figure 1). Tubular rays usually had a bilaterally symmetrical corolla with three outer lobes and one or two small inner ones, but lacked the bulged region characterizing the "gibbous" ray florets described in Layia (Ford and Gottlieb 1990). Whereas discoid plants formed a dense cluster in the PCA plot (Figure 2), there was no clear separation between plants with normal rays and plants with intermediate or novel ray morphologies.
|
|
The development of discoid heads showed no evidence of aborted ray florets, and the total number of florets per head (mean ± 1 standard deviation) was the same (P = .98; ANOVA), regardless of whether the outer florets were classified as "rays" (74.4 ± 10.4, n = 434) or as disk florets (73.0 ± 8.6, n = 306). Hence the lack of rays was not associated with a reduction in flower number.
The fraction of "discoid homozygotes," "heterozygotes," and "rayed homozygotes" showed significant heterogeneity among F2 families (contingency 
2 = 42.7, df = 10, P < .001). This variation was almost entirely due to differences in the relative frequency of discoid plants and "heterozygotes" (Table 4), as demonstrated by the drastic reduction in the chi-squared value after pooling these groups (contingency 2 = 6.0, df = 5, P > .05). A single-locus model involving all three classes (codominance) had to be rejected (pooled 2 = 195.8, P < .001). Most of the F2 and BC progenies showed an excess of discoid plants and a deficiency of "heterozygotes," whereas the frequency of "rayed homozygotes" always conformed to single-gene predictions (Table 4). As expected, there was a much closer agreement between the observed and predicted frequencies after pooling the "heterozygotes" with the discoid individuals (pooled 2 = 6.2, P > .05) than after pooling the "heterozygotes" with the "rayed homozygotes" (pooled 2 = 177.6, P < .001).
|
Discoid plants had a larger number of flowers per head and produced fewer heads in the second flowering season than the "rayed homozygotes," the mean of the "heterozygotes" being intermediate (head number in year 2) or close to the mean of the discoid individuals (flowers per head) (Table 5). Morph-specific differences in other fitness components were too small to be considered significant.
|
Analyses of Dominance and Allelism
The ray length of plants from between-morph crosses was significantly affected by the identity of the discoid donor population (F4,5 = 7.6, P < .05, nested ANOVA), whereas the "plant-within-population" component failed to reach significance (F5,83 = 2.0, P > .05). Inspection of means and standard deviations (Table 2), using the parental range as a reference (0 to approximately 10 mm), indicated a variable and incomplete pattern of dominance, with a trend toward partial dominance in crosses involving the Visby and Strathy Bay populations. Crosses between discoid plants from different localities yielded 17 rayed progeny (out of 389 plants; Table 2), the average ray length being 5.1 mm (range 2.5–8.0 mm). Almost all rayed plants appeared in the crosses Veluwe x Strathy Bay and Veluwe x Meijendel. Crosses between plants from the same discoid population always resulted in discoid progeny.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Patterns of segregation in a cross between a rayed and discoid population of ragwort (Visby x Lund) conformed to Mendelian segregation of two alleles, provided that plants with novel or intermediate ray types were subsumed into the discoid category. Hence results of the present study support the hypothesis that large differences in morphology may be a consequence of small genetic changes in developmental parameters (Gottlieb 1984; Orr and Coyne 1992). When combined with the discrete inheritance of ray floret morphology in crosses within S. vulgaris (Comes 1998; Trow 1912), S. squalidus (Ingram and Taylor 1982), and other composites (Fick 1967; Jackson and Dimas 1981), available data strongly suggest that a single allelic substitution could be involved in the transition from rayed to discoid heads. Obviously this interpretation assumes that the genetic differences between rayed and discoid morphs—as inferred from between-morph crosses—evolved during the process of differentiation and not subsequent to it. However, the fact that several parallel reversals to discoid heads have occurred in Senecio and other Asteraceae genera (Bremer and Humphries 1993; Tutin et al. 1976) provides further support for the evolutionary lability of ray morphology.
The suggestion that major mutations have been important for the evolution of discoid species or populations does not exclude the influence of genes with smaller and/or interactive effects on the ray phenotype (e.g., Ford and Gottlieb 1990). For example, comparison of F2 progenies from the cross Visby x Lund revealed significant among-family variation in the proportion of discoid individuals and "heterozygotes," implying that one or both parent plants were heterozygous for genetic factors that modify the expression of the major gene (Wade et al. 1997). Consequently the genetic control of "raylessness" probably involves a series of (unidentified) modifiers superimposed on a one-locus system. Some of the parental genes may have interacted nonadditively in the F2 and BC generation, as evidenced by the appearance of nonparental phenotypes, for example, ray florets with unusually large lobes or a tubular or bilabiate flower structure. Similar phenotypes have been recorded in crosses between rayed and discoid individuals of S. squalidus (Ingram and Taylor 1982), between Layia glandulosa and L. discoidea (Ford and Gottlieb 1990) and between North American tarweeds and Hawaiian silversword allies (Carr et al. 1996).
Analysis of variation in the rayed category provides no clear evidence for bimodality in ray length and other quantitative characters, as illustrated by the more or less gradual transition between fully rayed plants and hybrid phenotypes along the "general size axis" (PC1). For this reason it was necessary to use a combination of characters, for example, ray shape and the possession of sterile anthers, when identifying the genotype of rayed F2 and BC individuals. Judging from the close agreement between the observed and predicted frequency of "rayed homozygotes," this procedure resulted in a good separation between different genotypes in the rayed category. In fact, deviations from single-gene predictions were almost entirely due to an excess of "discoid homozygotes" and a deficiency of "heterozygotes." Presumably many F2 and BC plants failed to produce rays despite being heterozygous at the major locus. Although more detailed genetic analyses are needed to confirm this hypothesis, my observations clearly differ from the more strictly codominant expression of the ray floret locus in S. vulgaris (Comes 1998) and S. squalidus (Ingram and Taylor 1982).
Judging from the F1 data and patterns of segregation in the F2 and BC families, the heterozygotes generally expressed the discoid or some intermediate or novel ray phenotype. Hence, to the extent that patterns of dominance—as inferred from interpopulation hybrids—are inherent attributes of alleles rather than a result of natural selection (Bourguet 1999), there should be a great opportunity for local selection pressures to act on a newly risen discoid mutant in most populations (Haldane 1924). Of interest, comparison of F1 families from different between-morph crosses revealed slight, though significant, differences in average ray length, with some crosses showing incomplete dominance toward the rayed or discoid parent (Lund x Strathy Bay, Lund x Visby). Given these differences and the wide range of ray length seen in some F1 progenies (Lund x Meijendel, Lund x Veluwe), it is possible that ancestral populations differed in their response to past selection pressures.
The spread of a favorable mutation may be a slow process if heritable differences among individuals are obscured by microsite heterogeneity (Andersson 1995) or if selection is too weak to outweigh deleterious pleiotropic effects that may be associated with a major mutation (Lande 1983). Even though differences in ray size appear to have a strong genetic basis under near-natural (garden) conditions, as shown by the high between-year consistency documented for this character (r = 0.91; Andersson 2001), there is no certainty that ray-suppressing alleles have a high penetrance under variable and stressful field conditions. As for the role of pleiotropy, it is worth noting that discoid F2 and BC plants produced the lowest number of heads in the second flowering season (see also Andersson 2001) and that the mean for the "heterozygotes" was close to the midparent value. Hence the reproductive vigor after the first flowering season appears to decrease with the number of ray-suppressing alleles at the major locus. To the extent that head number is positively correlated with lifetime fitness, it may be necessary to invoke strong selection to explain the origin of discoid populations of ragwort.
Results of a ray removal experiment provide no evidence that plants of ragwort compensate for the loss of all rays by increasing the current year's production of heads or the percent fruit set after hand-pollination (Andersson 2001). However, if rays become redundant as pollinator attractants, as seems to be the case in some ragwort populations (Andersson 1996), then the conversion of rays (which are pistillate) into an extra whorl of disc florets (which are bisexual) should increase the male fitness of a discoid plant relative to co-occurring plants of the rayed phenotype (Andersson 2001). Moreover, given the low germination speed of fruits from ray florets relative to fruits from disc florets (McEvoy 1984), one would expect discoid individuals of ragwort to have a reproductive advantage at sites where selection favors early seedling emergence (Andersson 2001). When combined with other possible mechanisms, for example, selection for discoid heads to promote wind pollination at sites with low pollinator abundance (Berry and Calvo 1989), there should be a great potential for local selection pressures to outweigh the physiologically adverse effects of "raylessness." Selection for rayless heads may be greatest in the Atlantic region, where the discoid morph occurs in a more natural habitat (sand dunes) than the rayed phenotype (van der Meijden 1976).
Finally, it becomes meaningful to ask whether discoid populations have evolved repeatedly, that is, whether the suppression of rays is achieved by different genetic pathways in different populations. To evaluate this "multiple-origin hypothesis," I performed a test for allelism by crossing five discoid populations in all possible combinations and searching for rayed progeny in the offspring generation. A few rayed individuals occurred in crosses between different dune populations (Veluwe, Meijendel, Strathy Bay), but there was no evidence for complementary gene action in crosses between the Atlantic dune populations and the two discoid populations in the Baltic region (Visby, Sundsvall). Although these results suggest that some populations utilize different genes to suppress ray development, there is no reason to invoke a multiple-origin scenario to explain the wide and disjunct distribution of the discoid morph. Indeed, a preliminary survey of herbarium material indicates that most of the discoid populations in the Baltic region can be attributed to introduction through ship ballast in the 1800s (Andersson S, unpublished data).
|
Acknowledgments |
|---|
I wish to thank Louise Lundberg and Rune Svensson for technical assistance; Richard Abbott, Klaas Vrieling, Patrik Waldmann, and Thomas Karlsson for providing plant material; TorbjÖrn Säall for statistical advice; and Richard Abbott for critically reading the manuscript. Financial support by the Swedish Natural Science Council is also acknowledged.
|
Footnotes |
|---|
Corresponding Editor: Brandon Gaut
Received February 6, 2001
Accepted June 30, 2001
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-
Abbott RJ and Irwin JA, 1988. Pollinator movements and the polymorphism for outcrossing rate at the ray floret locus in groundsel, Senecio vulgaris. Heredity 60:295–298.
Andersson S, 1991. Floral display and pollination success in Achillea ptarmica (Asteraceae). Holarct Ecol 14:186–191.
Andersson S, 1995. Differences in the genetic basis of leaf dissection between two populations of Crepis tectorum (Asteraceae). Heredity 75:62–69.
Andersson S, 1996. Floral display and pollination success in Senecio jacobaea (Asteraceae): interactive effects of head and corymb size. Am J Bot 83:71–75.
Andersson S, 1999. The cost of floral attractants in Achillea ptarmica (Asteraceae): evidence from a ray removal experiment. Plant Biol 1:569–572.
Andersson S, 2001. Fitness consequences of floral variation in Senecio jacobaea (Asteraceae): evidence from a segregating hybrid population and a resource manipulation experiment. Biol J Linn Soc 74:17–24.
Bain JF, 1991. The biology of Canadian weeds. 96. Senecio jacobaea L. Can J Plant Sci 71:127–140.
Berry PE and Calvo RN, 1989. Wind pollination, self-incompatibility, and altitudinal shifts in pollination systems in the high Andean genus Espeletia (Asteraceae). Am J Bot 76:1602–1614.
Bradshaw HD, Wilbert SM, Otto KG, and Schemske DW, 1995. Genetic mapping of floral traits associated with reproductive isolation in monkeyflowers (Mimulus). Nature 376:762–765.
Bremer K and Humphries C, 1993. Generic monograph of the Asteraceae-Anthemideae. Bull Nat Hist Mus 23:73–179.
Bourguet D, 1999. The evolution of dominance. Heredity 83:1–4.
Carr DC, Baldwin BG, and Kyhos DW, 1996. Cytogenetic implications of artificial hybrids between the Hawaiian silversword alliance and North American tarweeds (Asteraceae: Heliantheae-Madiinae). Am J Bot 83:653–660.
Clausen J, Keck DD, and Hiesey WM, 1947. Heredity of geographically and isolated races. Am Nat 81:114–133.
Comes HP, 1998. Major gene effects during weed evolution: phenotypic characters cosegregate with alleles at the ray floret locus in Senecio vulgaris L. (Asteraceae). J Hered 89:54–61.
Coyne JA and Lande R, 1985. The genetic basis of species differences in plants. Am Nat 126:141–145.[ISI]
Fenster CB and Barrett SCH, 1994. Inheritance of mating-system modifier genes in Eichhornia paniculata (Pontederiaceae). Heredity 72:433–445.
Fick GN, 1967. Genetics of floral color and morphology in sunflowers. J Hered 67:227–230.
Fisher RA, 1958. The genetical theory of natural selection. New York: Dover.
Ford EB, 1975. Ecological genetics. New York: Wiley.
Ford VS and Gottlieb LD, 1990. Genetic studies of floral evolution in Layia. Heredity 64:29–44.
Gottlieb LD, 1984. Genetics and morphological evolution in plants. Am Nat 123:681–709.
Haldane JBS, 1924. A mathematical theory of natural and artificial selection. Trans Camb Philos Soc 23:19–41.
Harper JL and Wood WA, 1957. Biological flora of the British Isles. Senecio jacobaea L. J Ecol 45:617–637.
Ingram R and Taylor L, 1982. The genetic control of a non-radiate condition in Senecio squalidus L. and some observations on the role of ray florets in the Compositae. New Phytol 91:749–756.
Jackson RC and Dimas CT, 1981. Experimental evidence for systematic placement of the Haplopappus phyllocephalus complex (Compositae). Syst Bot 6:8–14.
Lande R, 1981. The minimum number of genes contributing to quantitative variation between and within populations. Genetics 99:541–553.
Lande R, 1983. The response to selection on major and minor mutations affecting a metrical trait. Heredity 50:47–65.
Marshall DF and Abbott RJ, 1982. Polymorphism for outcrossing frequency at the ray floret locus in Senecio vulgaris L. Evidence. Heredity 48:227–235.
Mayer SS and Charlesworth D, 1992. Genetic evidence for multiple origins of dioecy in the Hawaiian shrub Wikstroemia (Thymeleaceae). Evolution 46:207–215.
McEvoy PB, 1984. Dormancy and dispersal in dimorphic achenes of tansy ragwort, Senecio jacobaea. Oecologia (Berlin) 61:160–168.
Orr HA and Coyne JA, 1992. The genetics of adaptation: a reassessment. Am Nat 140:725–742.
Stuessy TF, Spooner DM, and Evans KA, 1986. Adaptive significance of ray corollas in Helianthus grosseserratus (Compositae). Am Midl Nat 115:191–197.
Trow AH, 1912. On the inheritance of certain characters in the common groundsel—Senecio vulgaris L.—and its segregates. J Hered 2:239–276.
Tutin TG, Heywood VH, Burges NA, Moore DM, Valentine DH, Walters SM, and Webb DA, 1976. Flora Europaea, vol. 4. Cambridge: Cambridge University Press.
van der Meijden E, 1976. Het verspreidingsgebied van Senecio jacobaea L. var. nudus Weston. Gorteria 8:57–61.
Wade MJ, Johnson NA, Jones R, Siguel V, and McNaughton M, 1997. Genetic variation segregating in natural populations of Tribolium castaneum affecting traits observed in hybrids with T. freemani. Genetics 147:1235–1247.[Abstract]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Andersson Pollinator and nonpollinator selection on ray morphology in Leucanthemum vulgare (oxeye daisy, Asteraceae) Am. J. Botany, September 1, 2008; 95(9): 1072 - 1078. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. K. Broholm, S. Tahtiharju, R. A. E. Laitinen, V. A. Albert, T. H. Teeri, and P. Elomaa A TCP domain transcription factor controls flower type specification along the radial axis of the Gerbera (Asteraceae) inflorescence PNAS, July 1, 2008; 105(26): 9117 - 9122. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Chapman, J. H. Leebens-Mack, and J. M. Burke Positive Selection and Expression Divergence Following Gene Duplication in the Sunflower CYCLOIDEA Gene Family Mol. Biol. Evol., July 1, 2008; 25(7): 1260 - 1273. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||