| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
0 Institute of Environmental and Evolutionary Biology, School of Biology, Harold Mitchell Building, University of St Andrews, St Andrews, Fife KY16 9TH, UK
Received for publication October 30, 1998. Accepted for publication October 26, 1999.
ABSTRACT
The possible pathways of origin of two recently arisen introgressant forms of Senecio vulgaris (i.e., var. hibernicus and York radiate groundsel) were investigated in experimental crosses between tetraploid S. vulgaris var. vulgaris and the normally diploid S. squalidus. Comparison of the morphology of synthesized hybrid progeny with established taxa, by discriminant function analysis, revealed that fertile hybrid offspring similar in morphology to S. vulgaris var. hibernicus and York radiate groundsel could be synthesized: (1) following formation of genomically stable diploid gametes by the triploid hybrid; (2) through the production of unreduced gametes by diploid S. squalidus; and (3) when a tetraploid form of S. squalidus acted as one of the parents. It was evident that hybrid offspring similar in morphology to the two introgressant taxa were more often produced in backcrosses to S. vulgaris than in segregating F2 or F3 generations (53% as opposed to 36%), and that fertile hybrid progeny were formed within two generations. Because hybridization between S. vulgaris and S. squalidus occurs regularly, although at very low frequency, in natural mixed populations in the British Isles, there is the potential for multiple origins to occur in the wild of both S. vulgaris var. hibernicus and York radiate groundsel.
Key Words: Asteraceae • hybrid origin • hybridization • introgression • Senecio.
Interspecific hybridization is an important mechanism for generating evolutionary novelty in plants (Arnold, 1997
) and can lead to the origin of stabilized introgressants, and homoploid and allopolyploid hybrid species (Abbott, 1992
; Arnold, 1992
; Rieseberg and Wendel, 1993
; Soltis, Doyle, and Soltis, 1992
; Arnold and Hodges, 1995
; Rieseberg, 1997
; Rieseberg and Carney, 1998
). Case studies of the evolution of stabilized introgressants and/or hybrid species have now been conducted in several angiosperm genera, e.g., Helianthus (Rieseberg, 1991, 1995
), Iris (Arnold, 1992
), Penstemon (Wolfe, Xiang, and Kephart, 1998
), Senecio (Abbott and Lowe, 1996
), Spartina (Raybould et al., 1991
), and Tragopogon (Soltis, Doyle, and Soltis, 1992
), and this has led Rieseberg and Carney (1998)
to suggest that future studies of hybridization "...that likely will prove most useful are those that combine experimental ecological and historical genetic approaches."
The best example of the historical genetic approach in research on plant hybridization is currently a comparison by Rieseberg and co-workers (1996)
of the genomic composition of three artificially synthesized hybrid lineages of Helianthus with that of an ancient, natural homoploid hybrid species that originated from the same two parents. It was established that the three synthetic lineages recovered fertility within 3–5 generations and converged to a nearly identical genomic composition that was concordant with that of the natural hybrid species. From these findings, Rieseberg and Carney (1998)
concluded "that deterministic forces such as selection, rather than stochastic forces, largely govern the formation of ‘recombinational' species." To confirm this, and by extension, to elucidate the factors that determine the genomic composition of other types of stabilized introgressants, similar studies need to be conducted on other hybrid taxa, particularly those resulting from parental taxa of different ploidy level (Ramsay and Schemske, 1998
). As a first step to this objective, we report here an analysis of the routes of origin of two tetraploid stabilized introgressant forms of Senecio that originated in the British Isles in the very recent past. Once their routes of origin are determined, it will be possible to establish whether wild forms of the two introgressant taxa of Senecio are genomically concordant with their artifically sythesized counterparts, in which case deterministic forces such as selection would be considered important in their formation as already demonstrated for recombinational species of Helianthus.
The origin of two stabilized introgressive forms of Senecio vulgaris L. (var. hibernicus Syme and "York radiate groundsel") following hybridization between S. vulgaris (2n = 40) and Senecio squalidus L. (2n = 20) represents one of the best known examples of the recent evolution of such hybrid taxa (Abbott, Ashton, and Forbes, 1992
; Irwin and Abbott, 1992
; Abbott and Lowe, 1996
). Senecio squalidus was introduced to the British Isles ~300 yr ago and did not escape into the wild and begin to spread until the early 1790s (Abbott, 1992
). Since the early part of the 19th century, the sterile triploid hybrid between S. squalidus and S. vulgaris, Senecio x baxteri Druce (2n = 3x = 30), has been reported from various sites in the British Isles that support mixed stands of the two parent species (Crisp, 1972
; Benoit, Crisp, and Jones, 1975
; Marshall and Abbott, 1980
). Subsequently, S. vulgaris var. hibernicus was first collected in 1832 (Crisp, 1972
), while the form of S. vulgaris known as York radiate groundsel (Irwin and Abbott, 1992
) was not collected until 1979 and is considered to have originated not more than 20 yr previously (Lowe, 1996
). The origin of both introgressant forms thus involves crosses between species of different ploidy levels (heteroploids) for which the literature is particularly thin (Ramsay and Schemske, 1998
), although such case studies offer the opportunity for an excellent understanding of the evolution of introgressant taxa (Petit, Bretagnolle, and Felber, 1999
).
Morphological, biochemical, and molecular analyses have been used to demonstrate the hybrid origins of S. vulgaris var. hibernicus (Abbott et al., 1992
) and York radiate groundsel (Irwin and Abbott, 1992
; Lowe, 1996
); however, in each case the possible routes of origin remain to be clarified. The triploid hybrid, S. x baxteri, has been successfully synthesized several times from crosses between S. vulgaris and S. squalidus (Harland, 1954
; Gibbs, 1971
; Ingram, 1977
; Ingram, Weir, and Abbott, 1980
; Taylor, 1984
), but it remains uncertain whether the formation of this hybrid was an important step in the origin of the two tetraploid taxa. Ingram, Weir, and Abbott (1980)
were successful in backcrossing S. x baxteri to S. vulgaris to produce progeny that were approximately tetraploid and fertile and similar in morphology to S. vulgaris var. hibernicus, in that they bore capitula with ray florets. However, it is feasible that a tetraploid F1 radiate plant could be produced directly from a cross between S. squalidus and S. vulgaris, due to fusion of a normal gamete of tetraploid S. vulgaris with an unreduced gamete of diploid S. squalidus or a reduced gamete of a tetraploid form of S. squalidus. Moreover, tetraploid hybrid derivatives might also originate from crosses between the allohexaploid S. cambrensis and S. squalidus, although all tetraploid offspring produced artificially by this route have proved to be sterile (Ingram and Noltie, 1987
). Finally, it is possible that chromosome reduction from the hexaploid to tetraploid state, either by chromosome mispairing of a genomically unstable near-hexaploid hybrid or following backcrossing of the hexaploid to S. vulgaris, might generate fertile tetraploid derivatives. In this regard, it is of interest that Ingram and Noltie (1987)
produced a semifertile pentaploid hybrid following crosses between S. cambrensis and S. vulgaris, which if backcrossed further to S. vulgaris may have yielded tetraploid offspring.
Thus there are several possible pathways by which the fertile, tetraploid hybrid derivatives S. vulgaris var. hibernicus and York radiate groundsel could have originated from crosses between S. vulgaris and S. squalidus. Of these, the two pathways that involve a contribution from the hexaploid S. cambrensis are unlikely, given the restricted distribution of this species, and more importantly, the fact that hybrids between S. cambrensis and S. squalidus and S. vulgaris have never been found in the wild (Ingram and Noltie, 1995
). Thus, the three most likely routes of origin are: route 1—backcrossing of S. x baxteri to S. vulgaris, or production of a balanced chromosome complement in progeny following selfing of the triploid hybrid; route 2—fusion of a "normal" reduced gamete of tetraploid S. vulgaris with an unreduced gamete of diploid S. squalidus; and route 3—fusion of a reduced gamete of S. vulgaris with a reduced gamete produced by a tetraploid form of S. squalidus.
In this paper we describe attempts to synthesize tetraploid hybrid products by all three of these pathways and use multivariate statistical analysis of morphological characters to define classes of naturally occurring taxa, so as to test the morphological affinity of synthesized hybrids to S. vulgaris var. hibernicus and York radiate groundsel.
MATERIALS AND METHODS
Seed collection and plant propagation
Seeds of S. squalidus, S. vulgaris var. vulgaris, and York radiate groundsel were collected from natural populations in York, UK, between 1989 and 1993, while seeds of S. vulgaris var. vulgaris and hibernicus were collected from populations in Edinburgh, Scotland, in 1991. All seeds from field populations or generated by experimental crosses were sown on damp filter paper. Following germination, seedlings with a root of length 1 cm were transplanted to 11.5 cm pots containing a 1:1 mix of Levington's M2 compost to gravel. Plants were raised at ambient temperature in a glasshouse under 400-W mercury vapor lamps with a photoperiod of 16 h.
Crossing program
Crosses were made between individuals of S. vulgaris (var. vulgaris and var. hibernicus) and S. squalidus using the emasculation technique of Ornduff (1964)
. Shortly before anthesis, the terminal 3–4 mm of an unopened capitulum were cut off with a razor blade to remove the anthers. The capitulum was covered with a small bag made from lens tissue and left to mature for 2–3 d. Developing stigmas were then examined for presence of pollen and, if none was present, cross-pollen was dusted onto the stigmas and the capitulum was rebagged until fruit matured.
Route 1: formation of tetraploid offspring via the triploid F1 hybrid
Formation of tetraploid progeny via the triploid F1 hybrid is outlined in Fig. 1. Several triploid hybrids (F1a-e) were synthesized from crosses between S. vulgaris (var. vulgaris and var. hibernicus) and S. squalidus. These plants were initially recognized by phenotype among offspring raised from a cross and were later confirmed by chromosome counts. Seed set of these triploid hybrids was very low, but a small amount of open-pollinated F2 seed was collected. Two of these seeds germinated and of the F2 plants raised, one was totally sterile, while the other exhibited partial fertility. This latter plant was backcrossed to S. vulgaris var. vulgaris to produce 17 backcross offspring (eight from backcrosses where S. vulgaris var. vulgaris was used as the maternal parent and nine from backcrosses where it was used as the paternal parent). The same plant also yielded 11 F3 progeny through open-pollination, eight of which were analyzed further.
|
|
Morphometric analysis
A total of 158 plants were raised in a fully randomized design under conditions of cultivation described above. These plants were grown for morphological comparison and included: one plant of S. vulgaris var. vulgaris and two of var. hibernicus; three plants of S. squalidus; 11 plants of York radiate groundsel; eight F3 and 17 B1 hybrid progeny produced by the route-1 crossing program; 15 F2 and 11 B1 progeny produced by the route-2 crossing program; and 24 F2 and 66 B1 progeny produced by the route-3 crossing program. On the day of full anthesis of the apical capitulum, the mid-leaf and apical capitulum were harvested from each individual, after which the plant was returned to the glasshouse. When the capitula of the apical corymb had set seed, a fruiting capitulum was harvested at random from each individual. Sixteen characters were measured on freshly harvested apical capitula, pressed mid-leaves, and fruiting capitula; seven of these characters were descriptors of the capitulum, seven described the mid-leaf, while two described plant fertility. The characters measured were capitulum length (in millimetres), capitulum width (in millimetres), number of outer ray florets, mean outer ray floret length (in millimetres), number of pollen pores, pollen fertility (assessed by determining the proportion of 100 pollen grains that took up aceto-carmine stain), length of mid-leaf (in millimetres), width of mid-leaf (in millimetres), angle of mid-leaf apex (defined as the angle between the apex of the primary vein and the apices of the adjacent marginal tooth sinuses), mid-leaf dissection (defined as the ratio of the perimeter of the mid-leaf divided by the square root of the area measured using a Delta-T area meter; Delta-T Devices, Cambridge, UK), standardized perimeter of mid-leaf (defined as perimeter of the mid-leaf divided by length), standardized square root of mid-leaf area (defined as square root of area of the mid-leaf divided by length), width of basal auricle (in millimetres), number of florets per capitulum, achene length (in millimetres), and proportion of seed set. A full description of each character measured is given in Lowe (1996)
.
The mid-leaf and apical and fruiting capitula were also harvested and measured for the same suite of characters from the following eight plants: one tetraploid S. squalidus; two F2 progeny produced by the route-1 crossing program; one tetraploid F1 produced by the route-2 crossing program; and four tetraploid F1 progeny produced by the route-3 crossing program. Although these plants were grown previous to the 158 raised in the randomized block, the data from them were analyzed together with those collected from plants in the block.
A principal components analysis (PCA) using CLUSTAN (Wishart, 1987
) was conducted on 13 of the 16 characters measured on all 166 individuals. The two fertility measures (pollen and seed) were omitted from this analysis as the high sterility of some hybrids was likely to have an overriding effect. Similarly, ray floret length was excluded, as preliminary analysis showed it to have a very large effect on the first two principal components thus masking the effect of other characters (e.g., leaf shape and achene length) that are important in comparing synthesized hybrids and established taxa. A canonical variate analysis (CVA) was also conducted on the same 13 traits so as to examine differences in mean phenotype of plants of S. vulgaris var. vulgaris, var. hibernicus, S. squalidus, artificially synthesized tetraploid S. squalidus, York radiate groundsel, and the F1 hybrid between tetraploid S. squalidus and S. vulgaris. CVA finds linear combinations of the original variables that maximize the ratio of between-group to within-group sums of squares and products matrices, thereby giving functions of the original variables that can be used to discriminate between the groups. The squared distances between group means are Mahalanobis' D2 statistics when all dimensions are used; otherwise they are approximations (SAS, 1990). To examine the robustness of differences between the six phenotypes as defined by CVA, individuals from each phenotype class were assigned to one of the six phenotype groups blindly using discriminant function analysis (DFA) following comparison of all unit canonical axes to group variances. In all cases, individuals were assigned confidently to their correct phenotype group (P < 0.05), further supporting the delimitation of phenotypic groups. The remaining F2, F3, and B1 hybrid progeny produced by each crossing program were then compared to each of these six phenotypes over all CVA axes, and assigned to a particular type by means of DFA using the DISCRIM routine of SAS (SAS, 1990).
RESULTS
Generation of tetraploid hybrid offspring
Triploid F1 hybrids between S. vulgaris and S. squalidus were difficult to generate and crosses were only successful (i.e., produced seed that germinated) using S. vulgaris as the maternal parent (Fig. 1). Five triploid hybrid plants (F1a-e) were produced from 145 crosses. Three of these were formed using S. vulgaris var. vulgaris as a parent, while two were produced when var. hibernicus was a parent. All five hybrids were highly sterile (mean pollen fertility, n = 3, 32.3%, SE 0.318; mean open seed set, n = 5, 0.0088%, SE 0.006), and hybrids produced using var. vulgaris as the maternal parent were of similar fertility to those that used var. hibernicus. All attempts to backcross these F1 hybrids to S. vulgaris or S. squalidus failed; however, two germinable seeds were produced via open-pollination under glass from a total of 385 capitula examined. Of the two F2 plants produced, one was from the cross in which var. hibernicus was the maternal parent (F2e), and was radiate (mean ray floret length = 11.0 mm) with long, narrow leaves and elongated internodes. This plant was male sterile, producing shrivelled, nonfunctional anthers, and no germinable seed when backcrossed to S. vulgaris or when left to open-pollinate. The other F2 plant was derived from the cross in which var. vulgaris was the maternal parent (F2a) and was partially fertile (pollen fertility 71%; open seed set 54.9%), and produced radiate capitula (mean ray floret length 10.0 = mm). Though no chromosome count was made of this plant, it was highly interfertile with S. vulgaris (backcross success = 86%) and was presumed to be near tetraploid.
In addition to the five triploid hybrids, one "spontaneous" tetraploid F1 hybrid (F1f) was produced by crossing diploid S. squalidus to tetraploid S. vulgaris (Fig. 2). This plant was radiate (mean ray floret length = 10.0 mm) and partially fertile (pollen fertility = 68%; open seed set = 10.6%).
Tetraploid F1 hybrids were relatively easy to generate when tetraploid S. squalidus was used in crosses, although only when S. squalidus acted as the maternal parent (Fig. 2). Four tetraploid hybrid plants were produced from 26 such crosses (F1g-j). All four showed high pollen fertility (mean, n = 4, 82.2%, SE 1.8) and reasonable seed set (mean, n = 4, 23.1%, SE 3.1) when left to open-pollinate. They resembled in overall morphology and fertility the spontaneously produced F1 tetraploid hybrid and were partially interfertile with S. vulgaris, exhibiting a mean backcross seed set of 33.7% (SE 0.05).
Morphometric analysis
Means and standard deviations for the 16 morphological and fertility characters measured on all parental material and hybrid progeny are presented in Lowe (1996)
. Principal components analysis on all 166 individuals analyzed for morphology showed that the first two components with eigenvalues of 4.01 and 1.89, described 30.8 and 14.5% of the variance in the data, respectively. Characters that contributed most to the first principal component were leaf characters (eigenvector values shown in parentheses), i.e., mid-leaf length (0.331), mid-leaf width (0.465), mid-leaf dissection (0.401), mid-leaf standardized perimeter (0.417), mid-leaf standardized area (0.402), and mid-leaf basal auricle width (0.319). In contrast, characters contributing most to the second principal component were mainly capitulum characters, i.e., number of ray florets (0.405); pollen pore number (-0.213); capitulum width (0.534); number of florets per capitulum (0.412); and achene length (0.220), although mid-leaf length (0.223) also contributed to this component. The first principal component separated individuals of York radiate groundsel from those of S. squalidus and S. vulgaris (Fig. 3), while the second component separated diploid S. squalidus, tetraploid S. squalidus, S. vulgaris var. vulgaris and var. hibernicus groups of individuals from each other. The "spontaneous" tetraploid F1 hybrid (F1f) was associated with those synthesized using tetraploid S. squalidus as a parent (F1g-h). Thus these tetraploid hybrids exhibit close morphological similarity of a type that is intermediate to that of tetraploid S. squalidus and S. vulgaris. An F2 individual, F2a, produced from the triploid hybrid was positioned close to S. vulgaris var. hibernicus material, indicating a close morphological similarity, while the male-sterile F2 hybrid, F2e, was placed close to S. squalidus. The F2, F3, and backcross hybrid derivatives produced by all crossing programs exhibited considerable morphological variation.
|
|
|
|
|
This study has shown that it is possible to synthesize from crosses between Senecio vulgaris and S. squalidus early-generation, tetraploid, hybrid offspring that are fertile and bear a close morphological resemblance to one or the other of the two recently evolved stabilized introgressant taxa of Senecio vulgaris, i.e., var. hibernicus and York radiate groundsel. Moreover, such hybrid offspring can be formed by at least three different routes of origin: (1) following formation of genomically stable diploid gametes by the triploid hybrid; (2) through the production of unreduced gametes by diploid S. squalidus; and (3) when a tetraploid form of S. squalidus acts as one of the parents.
Formation of hybrids
No triploid hybrids were produced from crosses between S. vulgaris and diploid S. squalidus, when S. squalidus acted as the maternal parent. This finding has been observed in most previous attempts to resynthesize S. x baxteri (Harland, 1954
; Gibbs, 1971
; Ingram, 1977
, Ingram, Weir, and Abbott, 1980
; Taylor, 1984
), although Ingram (1977)
reported an exception to this rule. Such unidirectional hybridization between a diploid and tetraploid species is common and is thought to be caused by massive abortion of triploid seeds in the diploid due to partial or total failure of the endosperm tissue associated with embryos (Bretagnolle and Thompson, 1995
). In this respect, it is also relevant to note that S. vulgaris, not S. squalidus, was the maternal parent of the spontaneous F1 tetraploid hybrid. More difficult to explain is the observation that other F1 tetraploid hybrids were generated only when tetraploid S. squalidus acted as the maternal parent in crosses with S. vulgaris, for which there should, in theory, be no directionality (Bretagnolle and Thompson, 1995
). This result, however, may simply reflect the low number of crosses involved.
Offspring of triploid hybrids
The five triploid F1 hybrids, produced by crossing diploid S. squalidus and S. vulgaris, were highly sterile and only two germinable seeds were collected from open-pollinated capitula. One F2 plant, derived from a cross in which var. hibernicus was the maternal parent, produced shrivelled, nonfunctional anthers and set no seed even under conditions of open-pollination. This individual was of interest, however, because it exhibited an "elongated" phenotype, with long internodes and long, narrow ray florets and leaves, and resembled the male-sterile dwarf strap plants previously described by Harland (1954)
. The other F2 plant, derived from a cross in which var. vulgaris was the maternal parent, was a partially fertile, radiate plant very similar in overall phenotype to a robust example of S. vulgaris var. hibernicus. It was presumed to be near tetraploid as it was highly interfertile with S. vulgaris. F3 and B1 progeny produced from this partially fertile plant exhibited great morphological variation, and three F3 progeny and four B1 plants were confidently (P < 0.05) assigned to the York radiate groundsel class, while seven other B1 offspring were assigned to var. hibernicus by DFA. The B1 offspring assigned to these taxonomic groups exhibited mean pollen and seed fertilities that exceeded 50%.
Offspring of tetraploid hybrids
Tetraploid F1 hybrids, produced by crossing S. vulgaris with diploid and tetraploid S. squalidus, were very similar to each other in overall morphology and were partially fertile. These plants also exhibited a number of "gigas" and hybrid characters, including long rays (10 mm), long achenes (~3 mm), and highly dissected leaves. The segregating F2 generation and backcrosses to S. vulgaris again yielded plants that exhibited a wide range of morphological phenotypes, and the results of the DFA assigned 14 (out of 116) offspring to York radiate groundsel and 39 to the S. vulgaris var. hibernicus class. These F2 and B1 offspring also exhibited relatively high mean pollen and seed fertilities.
Pathway of origin of tetraploid hybrid taxa
In summary, fertile hybrid progeny similar in morphology to S. vulgaris var. hibernicus and York radiate groundsel were recorded among the offspring of the triploid and tetraploid F1 hybrids produced by all three crossing programs. Unfortunately, it was not possible to continue the analysis over future generations to the point where it could be said with confidence that stabilized introgressants (i.e., nonsegregating) had been produced that were identical in form to either var. hibernicus or York radiate groundsel. However, we might assume that such stable products would be formed within a few generations, if indeed they were not present among some of the hybrid offspring assigned to the var. hibernicus and York radiate groundsel groupings by discriminant function analysis. Rieseberg et al. (1996)
were able to generate stabilized synthetic hybrid lineages of Helianthus that closely resembled a known homoploid hybrid Helianthus species within five generations of the initial cross between the two known parent species. What is clear from the present analysis is that stabilized introgressants of Senecio can be produced in several different ways, which raises the question as to which of these pathways is (or are) most likely to have led to the formation of var. hibernicus and York radiate groundsel in the wild.
The first pathway examined (route 1) relies on the ability of the triploid F1 hybrid to act as a bridge between ploidy levels through the production of balanced diploid gametes (Ratter, 1972, 1973a, b
; Ingram, 1978
; Bretagnolle and Thompson, 1995
). Tetraploid offspring are formed following backcrossing to the tetraploid parent or by fusion with other such gametes. Ingram, Weir, and Abbott (1980)
suggested that S. vulgaris var. hibernicus most likely arose in this way, after synthesizing plants similar in morphology to S. vulgaris var. hibernicus by backcrossing S. x baxteri to var. vulgaris. However, the production of fertile progeny via this route is a two-step process, involving first the generation of S. x baxteri, and second, the production of later generation, fertile progeny either via backcrossing or segregation. In regard to step 1, the best estimate of the frequency of S. x baxteri production in natural populations comes from Marshall and Abbott (1980)
, who noted that the hybrid occurs regularly, but at low frequency in mixed stands of the two parent species. As regards step 2, several studies have examined the potential of experimentally synthesized S. x baxteri to produce fertile progeny, although no field estimates are available. Ingram (1977)
was not able to obtain germinable seed from triploid F1 hybrids backcrossed to S. vulgaris or left to open-pollinate. However, Ingram, Weir, and Abbott (1980)
produced two near tetraploids (two others died before flowering) from a single S. x baxteri plant that had been backcrossed to S. vulgaris var. vulgaris, although no seed was set when the F1 hybrid was selfed. Ingram (1978)
also managed to produce eight partially fertile (seed set 30–55%), near tetraploid (2n = 40–44) progeny when she backcrossed a triploid F1 hybrid to S. vulgaris var. vulgaris. Moreover, on selfing the F1 hybrid, one near diploid (2n = 22) plant was obtained, while eleven partially fertile F2 progeny (ranging from tetraploid to hexaploid) were produced with open-pollination. In the present study, one partially fertile, presumed tetraploid, hybrid was raised from ~25 000 open-pollinated achenes from five triploid plants. Taken overall, therefore, it might be concluded that the production of fertile, near tetraploid progeny from triploid F1 hybrids is likely to be a very rare event in the wild.
The second pathway of origin examined (route 2) depends on the ability of S. squalidus to produce unreduced gametes, which then fuse with normal, reduced gametes of S. vulgaris. The ability of diploid species to produce unreduced gametes is widely acknowledged (Harlan and deWet, 1975
; Bretagnolle and Thompson, 1995
). However, only two tetraploid F1 hybrids are known to have been generated artificially following fusion of an unreduced gamete from diploid S. squalidus with a reduced gamete from S. vulgaris. One of these was produced by Taylor (1984)
and the other during work reported here. Taylor (1984)
reported that from an unspecified number of crosses attempted, one triploid and one tetraploid F1 hybrid were produced, while in the present study, out of six F1 hybrids generated, five were triploid and one was tetraploid. In all other studies where S. squalidus was crossed with S. vulgaris, only triploid F1 hybrids have reportedly been generated (Gibbs, 1971
; Ingram, 1977, 1978
; Ingram, Weir, and Abbott, 1980
). Tetraploid F1 hybrids are partially fertile and produce far higher amounts of germinable seed than triploid hybrids (>2000 times as much, based on the open seed set values observed in this study). This, therefore, should potentially increase the probability with which hybrid progeny produced by this pathway become established.
It is important to note that spontaneous production of unreduced gametes in S. squalidus was not observed in this study, but was inferred from the generation of a hybrid product that must have involved an unreduced gamete. Crisp (1972)
noted that S. squalidus occasionally produces large pollen grains (0.18%; from a count of 9 in 4959 grains), which were presumed to contain an unreduced chromosome complement. Although Crisp's estimate is low, there is some evidence that unreduced pollen may perform better under competitive conditions in the stigma than reduced pollen (Mulinix and Lezzoni, 1988
), and so frequency of occurrence may not reflect fertilization success. Also, production of unreduced gametes may increase markedly under some conditions in the wild. For example, Bretagnolle and Thompson (1995)
noted that low temperature regularly induced production of unreduced gametes.
Recent reviews have shown that the production of unreduced gametes by diploid taxa has played an important role in the evolution of many auto- and allopolyploid species (Harlan and deWet, 1975
; Thompson and Lumaret, 1992
; Bretagnolle and Thompson, 1995
), and the current work suggests that the production of unreduced gametes by S. squalidus could have been instrumental in the origin of both S. vulgaris var. hibernicus and York radiate groundsel. However, both the present and previous studies (Ingram, Weir, and Abbott, 1980
) have also shown that the early stages of the generation of both introgressant taxa are possible via the production of a balanced chromosome complement from a genomically unstable triploid hybrid, which others have recognized as a potentially important process in the establishment of hybrid taxa (Ratter, 1972, 1973a, b
; Ingram, 1978
; Bretagnolle and Thompson, 1995
). Consequently, it is not easy to decide whether either pathway investigated (route 1 or route 2) is more likely than the other to have led to the production of var. hibernicus and York radiate groundsel in the wild. However, given the high fertility of the tetraploid F1 hybrid, it is feasible that the pathway of origin involving unreduced gametes in the diploid parent might be more likely to have led to the origin of the two Senecio introgressant taxa investigated.
The third pathway (route 3) examined requires a tetraploid S. squalidus plant. The tetraploid S. squalidus plants generated by colchicine treatment in this study exhibited a number of "gigas" morphological characters associated with polyploidy (also noted by Crisp, 1972
), that would easily distinguish them in a population of diploid S. squalidus individuals. Such plants have never been recorded in natural populations (Crisp, 1972
; A. J. Lowe, personal observations), and assuming, therefore, that tetraploid S. squalidus occurs only at very low frequency in the wild, we may exclude route 3 as a pathway of origin for S. vulgaris var. hibernicus and York radiate groundsel. The pathway was, however, usefully examined here as it involves the generation of diploid S. squalidus pollen and thus provides further indirect support for route 2 as a means of generating the two stabilized introgressants of Senecio.
It was notable that one generation of backcrossing the triploid and tetraploid hybrids to S. vulgaris produced offspring similar to S. vulgaris var. hibernicus or York radiate groundsel at a higher frequency than did segregation within the F2 and F3 generations (53% as compared to 36%). This might indicate that backcrossing has been important in the origin of both such taxa. Morphological, isozyme, and molecular studies also support this assertion (Irwin and Abbott, 1992
; Lowe, 1996
), with York radiate groundsel probably having undergone less backcrossing to S. vulgaris var. vulgaris during its derivation than var. hibernicus. Finally, the fact that interspecific hybridization occurs regularly, though infrequently, in natural populations (Marshall and Abbott, 1980
) and the finding that stabilized hybrid products of later generation are likely to be produced with some ease, would indicate that multiple origins of S. vulgaris var. hibernicus and York radiate groundsel may be common in the British Isles.
FOOTNOTES
1 The authors thank David Forbes for his assistance in the greenhouse, and Francois Bretagnolle, Christophe Thébaud, and Ron Smith for advice with statistical methods and constructive comments on the manuscript. The research was conducted while AJL held a research studentship from the NERC. 
2 Current address, author for correspondence: Institute of Terrestrial Ecology, Bush Estate, Edinburgh Science Park, Midlothian EH26 0QB, UK.
LITERATURE CITED
Abbott, R. J. 1992 Plant invasions, interspecific hybridization and the evolution of new plant taxa. Trends in Ecology and Evolution 7: 401–405.[CrossRef]
———, P. A. Ashton, and D. G. Forbes. 1992 Introgressive origin of the radiate groundsel, Senecio vulgaris L. var. hibernicus Syme: Aat-3 evidence. Heredity 68: 425–435.[ISI]
———, and A. J. Lowe. 1996 A review of hybridization and evolution in British Senecio. In D. J. N. Hind [ed.], International Compositae Conference, Kew, vol. 1, 679–689. Royal Botanic Gardens, Kew, UK.
Arnold, M. L. 1992 Natural hybridization as an evolutionary process. Annual Review of Ecology and Systematics 23: 237–261.[CrossRef][ISI]
———. 1997 Natural hybridization and evolution. Oxford University Press, Oxford, UK.
———, and S. A. Hodges. 1995 Are natural hybrids fit or unfit relative to their parents? Trends in Ecology and Evolution 10: 67–71.
Benoit, P. M., P. C. Crisp, and B. M. G. Jones. 1975 Senecio L. In C. A. Stace [ed.], Hybridization and the Flora of the British Isles, 404–410. Academic Press, London, UK.
Bretagnolle, F., and J. D. Thompson. 1995 Tansley Review Number 78. Gametes with the somatic chromosome number: mechanisms of their formation and role in the evolution of autopolyploid plants. New Phytologist 129: 1–22.[CrossRef][ISI]
Crisp, P. 1972 Cytotaxonomic studies in the Section Annui of Senecio. Ph.D. dissertation, University of London, London, UK.
Gibbs, P. E. 1971 Studies on synthetic hybrids of British species of Senecio:I Senecio viscosis L. x S. vulgaris L. Transactions of the Botanic Society of Edinburgh 41: 213–218.
Harlan, J. R., and J. M. J. de Wet. 1975 On Ö winge and a prayer: the origins of polyploidy. Botany Review 41: 361–390.
Harland, S. C. 1954 The genus Senecio as a subject for cytogenetical investigation. Proceedings of the Botanic Society of Britain I 1: 256.
Ingram, R. 1977 Synthesis of the hybrid Senecio squalidus L. x S. vulgaris L. f. radiatus Hegi. Heredity 39: 171–173.[ISI]
———. 1978 The genomic relationship of Senecio squalidus L. and Senecio vulgaris and the significance of genomic balance in their hybrid S. x baxteri Druce. Heredity 40: 459–462.
———, and H. J. Noltie. 1987 The control of chiasma frequency within a polyploid series in the genus Senecio (Compositae). Genetica 72: 37–41.[CrossRef][ISI]
———, and ———. 1995 Biological Flora of the British Isles: Senecio cambrensis Rosser. Journal of Ecology 83: 537–546.[CrossRef]
———, J. Weir, and R. J. Abbott. 1980 New evidence concerning the origin of inland radiate groundsel, S. vulgaris var. hibernicus Syme. New Phytologist 84: 543–546.[CrossRef][ISI]
Irwin, J. A., and R. J. Abbott. 1992 Morphometric and isozyme evidence for the hybrid origin of a new tetraploid radiate groundsel in York, England. Heredity 69: 431–439.[ISI]
Lowe, A. J. 1996 The origin and maintenance of a new tetraploid Senecio hybrid in York, England. Ph.D. dissertation, University of St Andrews, St Andrews, UK.
Marshall, D. F., and R. J. Abbott. 1980 On the frequency of introgression of the radiate (Tr) allele from Senecio squalidus L. into Senecio vulgaris. Heredity 45: 133–135.
Mulinix, C. A., and A. F. Lezzoni. 1988 Microgametophytic selection in two alfalfa (Medicago sativa L.) clones. Theoretical and Applied Genetics 75: 917–922.[ISI]
Ornduff, R. 1964 Evolutionary pathways of the Senecio lautus alliance in New Zealand and Australia. Evolution 18: 349–360.[CrossRef][ISI]
Petit, C., F. Bretagnolle, and F. Felber. 1999 Evolutionary consequences of diploid-polyploid hybrid zones in wild species. Trends in Ecology and Evolution 14: 306–311.
Ramsay, J., and D. W. Schemske. 1998 Pathways, mechanisms, and routes of polyploid formation in flowering plants. Annual Review of Ecology and Systematics 29: 467–501.
Ratter, J. A. 1972 Cytogenetic studies in Spergularia VI: The evolution of true breeding, fertile tetraploids from a triploid interspecific hybrid. Notes of the Royal Botanic Gardens Edinburgh 32: 117–125.
———. 1973a Cytogenetic studies in Spergularia VIII: Barriers to the production of viable interspecific hybrids. Notes of the Royal Botanic Gardens Edinburgh 32: 297–301.
———. 1973b Cytogenetic studies in Spergularia IX: Summary and conclusions. Notes of the Royal Botanic Gardens Edinburgh 32: 411–428.
Raybould, A. F., A. J. Gray, M. J. Lawrence, and D. F. Marshall. 1991 The evolution of Spartina anglica C. E. Hubbard (Gramineae): origin and genetic variability. Biological Journal of the Linnean Society 43: 111–126.[CrossRef]
Rieseberg, L. H. 1991 Homoploid reticulate evolution in Helianthus (Asteraceae): evidence from ribosomal genes. American Journal of Botany 78: 1218–1237.[CrossRef][ISI]
———. 1995 The role of hybridization in evolution: old wine in new skins. American Journal of Botany 82: 944–953.[CrossRef][ISI]
———. 1997 Hybrid origins of plant species. Annual Review of Ecology and Systematics 28: 359–389.[CrossRef][ISI]
———, and S. E. Carney. 1998 Tansley Review Number 102. Plant hybridization. New Phytologist 140: 599–624.[CrossRef][ISI]
———, 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.[Abstract]
———, and J. F. Wendel. 1993 Introgression and its consequences in plants. In R. Harrison [ed.], Hybrid zones and the evolutionary process, 70–109. Oxford University Press, New York, New York, USA.
Soltis, P. S., J. J. Doyle, and D. E. Soltis. 1992 Molecular data and polyploid evolution in plants. In P. E. Soltis, D. E. Soltis, and J. J. Doyle [eds.], Molecular systematics of plants, 177–201. Chapman and Hall, New York, New York, USA.
Taylor, L. 1984 The potential for introgression in a British polyploid complex. Ph.D. dissertation, University of St Andrews, St Andrews, UK.
Thompson, J. D., and R. Lumaret. 1992 The evolutionary dynamics of polyploid plants: origins, establishment and persistence. Trends in Ecology and Evolution 7: 302–307.[CrossRef]
Wishart, D. 1987 Clustan user manual. Cluster Analysis Software. Computing Laboratories, University of St Andrews, St Andrews, UK.
Wolfe, A. D., Q. Y. Xiang, and S. R. Kephart. 1998 Diploid hybrid speciation in Penstemon (Scrophulariaceae). Proceedings of the National Academy of Sciences, USA 95: 5112–5115.
This article has been cited by other articles:
|
|
 |
|
  M. Kim, M.-L. Cui, P. Cubas, A. Gillies, K. Lee, M. A. Chapman, R. J. Abbott, and E. Coen Regulatory Genes Control a Key Morphological and Ecological Trait Transferred Between Species Science, November 14, 2008; 322(5904): 1116 - 1119. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Kirk, K. Vrieling, and P. G. L. Klinkhamer Reproductive fitness of hybrids between Senecio jacobaea and S. aquaticus (Asteraceae) Am. J. Botany, September 1, 2005; 92(9): 1467 - 1473. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
