Abstract

Background and Aims Although hybridization can play a positive role in plant evolution, it has been shown that excessive unidirectional hybridization can result in replacement of a species’ gene pool, and even the extinction of rare species via genetic assimilation. This study examines levels of introgression between the common Saxifraga spathularis and its rarer congener S. hirsuta, which have been observed to hybridize in the wild where they occur sympatrically.

Methods Seven species-specific single nucleotide polymorphisms (SNPs) were analysed in 1025 plants representing both species and their hybrid, S. × polita, from 29 sites across their ranges in Ireland. In addition, species distribution modelling was carried out to determine whether the relative abundance of the two parental species is likely to change under future climate scenarios.

Key ResultsSaxifraga spathularis individuals tended to be genetically pure, exhibiting little or no introgression from S. hirsuta, but significant levels of introgression of S. spathularis alleles into S. hirsuta were observed, indicating that populations exhibiting S. hirsuta morphology are more like a hybrid swarm, consisting of backcrosses and F2s. Populations of the hybrid, S. × polita, were generally comprised of F1s or F2s, with some evidence of backcrossing. Species distribution modelling under projected future climate scenarios indicated an increase in suitable habitats for both parental species.

Conclusions Levels of introgression observed in this study in both S. spathularis and S. hirsuta would appear to be correlated with the relative abundance of the species. Significant introgression of S. spathularis alleles was detected in the majority of the S. hirsuta populations analysed and, consequently, ongoing introgression would appear to represent a threat to the genetic integrity of S. hirsuta, particularly in areas where the species exists sympatrically with its congener and where it is greatly outnumbered.

INTRODUCTION

Hybridization is a common phenomenon in plant taxa, and can result in a wide range of potential evolutionary consequences, both positive and negative (Barton, 2001). The process can give rise to the evolution of new species (Anderson and Stebbins, 1954; Seehausen, 2004), and introgression can act as a source of potential adaptive variation (Lewontin and Birch, 1966), but excessive unidirectional hybridization can result in replacement of a species’ gene pool (Beatty et al., 2010) and has even been implicated in the extinction of rare species via genetic assimilation (Levin et al., 1996). It is thus clear that no single evolutionary trajectory results from the process of hybridization, and the unpredictability of the final outcomes has led to ongoing interest from researchers.

In recent years, global climate change has led to shifts in the distribution ranges of many species (Parmesan and Yohe, 2003; Hickling et al., 2006; Kelly and Goulden, 2008). One consequence of these range shifts is the creation of novel species assemblages. Such ecosystem reorganizations offer new opportunities for hybridization between species that would otherwise remain isolated from each other, both ecologically and reproductively (Garroway et al., 2010; Muhlfeld et al., 2014; Potts et al., 2014). In addition to observational studies on emergent hybridization, species distribution modelling approaches are now being employed to determine the role of future range dynamics in the potential creation of new hybrid zones (Sanchez-Guillen et al., 2014). To date, however, these models have not been applied to assess the possible impacts of climate change on existing known hybrid zones in the context of potential genetic assimilation.

Saxifraga spathularis Brot. and S. hirsuta L. are both members of the Lusitanian flora, whose disjunct distribution between Ireland and northern Spain has been of great intrigue to botanists (Webb, 1983; Beatty and Provan, 2013, 2014). Both species belong to the Saxifraga Section Gymnopera, which also includes two other members, S. umbrosa L. and S. cuneifolia L. Saxifraga umbrosa readily hybridizes with both S. spathularis, resulting in S. × urbium, an artificially induced/propagated hybrid commonly referred to as London Pride, and with S. hirsuta, resulting in S. × geum L., a wild naturally occurring hybrid, found where the ranges of the two parental species overlap in the western Pyrenees (Webb and Gornall, 1989). Saxifraga cuneifolia, however, does not form hybrids with any of its congener species, either in the wild or artificially. The two Lusitanian species exhibit contrasting patterns of occurrence throughout their divided ranges. In the Iberian reaches of their range, S. hirsuta is more commonplace, stretching from the mountains of Galicia and Andalusia in the north-west, across northern Spain to the Pyrenees, compared with the more limited range of S. spathularis, which is restricted to the north-western corner of Iberia. In Ireland, however, levels of abundance are reversed: S. spathularis is significantly more abundant and wide-ranging, being found throughout the south and west of the country, primarily Counties Galway, Kerry, Cork and Waterford, whilst S. hirsuta is limited to the extreme south-west in Counties Cork and Kerry. Both species have similar habitats, although S. spathularis reaches a greater altitude and S. hirsuta is less tolerant of exposed sites. A putative hybrid between the two species, S. × polita (Haw.) Link, has been identified in both the Irish and Iberian parts of their range. Records of S. × polita are rare in Iberia compared with Ireland, most probably due to the relatively limited area of overlap in the parental species’ ranges in Iberia, and the fact that where they do overlap they tend to occupy different environmental niches (McGregor, 2008). In Ireland, however, the hybrids are common where both species co-occur in Counties Cork and Kerry and, interestingly, also in Co. Galway, where only one of its parental species, namely S. spathularis, occurs (Webb, 1951).

Given the previously documented threats of genetic assimilation by a more abundant congener in other plant species, and the relative abundances of S. hirsuta and S. spathularis, along with the occurrence of the hybrid S. × polita in Ireland, the questions addressed in the current study were as follows. (1) Is there evidence of introgression of S. spathularis alleles into S. hirsuta? (2) If so, to what level is the introgression occurring and is it placing the gene pool of the rare S. hirsuta under threat of genetic assimilation? (3) What is the genetic composition of S. × polita hybrid populations? (4) Is the distribution of the two parental species likely to change under future climate scenarios, and thus increase or decrease the likelihood of hybridization?

MATERIALS AND METHODS

Study species

Saxifraga spathularis (St. Patrick’s cabbage) and S. hirsuta (kidney saxifrage) are perennial stoloniferious herbs. The two species and their hybrid (S. × polita) can be distinguished by relative leaf shape, pubescence toothing and the width of the cartilaginous margins of the leaf lamina (Webb and Gornall, 1989). Saxifraga spathularis leaves are extremely distinctive, with smooth, thick, waxy, spathulate leaves, each of which has a sharply toothed margin. The leaves of S. hirsuta are soft and hairy, and are fringed with a crenate toothing, with petioles that are long thin and hairy. The leaves of S. × polita exhibit an intermediate form, in terms of both relative leaf shape and toothing of the leaf margin. The width of the cartilaginous margins of the leaf lamina in S. spathularis is 0·1 mm, for S. hirsuta it ranges between 0·1 and 0·2 mm, while for S. × polita, an intermediate range of between 0·1 and 0·15 mm has been recorded (Webb and Gornall, 1989). The leaves of all three congeners form a basal rosette/turf, from which the erect flowering stem projects. All three taxa flower from May through to July, when the stem bears panicles of star-shaped, five-petalled white/pink flowers. The flowers are insect pollinated and seeds are primarily wind dispersed.

Sampling and DNA extraction

Samples were collected from 29 locations across the ranges of S. spathularis, S. hirsuta and S. × polita in Ireland (Fig. 1; Table 1). These included two sites in Co. Galway, where both S. spathularis and S. × polita were present, but where no verified records of S. hirsuta exist (Stelfox, 1947), as well as samples of S. spathularis from Co. Waterford, where neither S. hirsuta nor S. × polita is present. DNA was extracted using the cetyltrimethylammonium bromide (CTAB) method of Doyle and Doyle (1987).

Fig. 1

Maps showing locations of populations sampled in this study. Numbers refer to populations listed in Table 1.

Fig. 1

Maps showing locations of populations sampled in this study. Numbers refer to populations listed in Table 1.

Table 1.

Sites and numbers of samples collected

Site no. Location Lat. (N) Long. (W) No. of samples
 
S. spathularis S. hirsuta S. × polita 
Glengariff Forest Park 51·7539 9·5649 22/22/22 – 22 
Glengariff Forest Park 51·7531 9·5652 – 
Glengariff Forest Park 51·7533 9·5649 22 16 
Glengariff Forest Park 51·7545 9·5694 11 11 
Glengariff Forest Park 51·7562 9·5713 – 14 22 
Glengariff Forest Park 51·7573 9·5724 22/4 15 15/8 
Glengariff Forest Park 51·7551 9·5674 22 22 22 
Glengariff Forest Park 51·7561 9·5713 
Glengariff Forest Park 51·7566 9·5671 – 
10 Glengariff Forest Park 51·7536 9·5613 19 16 
11 Kenmare Tunnel 51·7560 9·5872 20 17 
12 Beara Way 51·7072 9·6213 22 – 21 
13 Beara Way Bridge 51·7049 9·6523 11 19 9/11 
14 Dromagowlane 51·6886 9·6613 – 22 20 
15 Lauraghbridge 51·7675 9·7705 
16 Glenbeg Lough 51·7177 9·8737 22 12 
17 Cleanderry Wood 51·7377 9·9247 – 22 – 
18 Torc Mountain 51·9959 9·5083 20 – 15 
19 Torc Waterfall 52·0067 9·5083 17 10 12 
20 Meeting of the Waters 52·0048 9·5305 – 21 – 
21 Old Torc Mountain Path 52·0045 9·5305 – 22 
22 Gap of Dunloe 51·9918 9·6440 14/3 12 
23 Ballaghbeama Gap 51·9407 9·8047 16 22 
24 Camp 52·2152 9·9019 – 
25 Owenmore River Bridge 52·2076 10·2084 10 
26 Conor Pass 52·1818 10·2078 22 22 21 
27 Connemara National Park 53·5514 9·9458 – 22 
28 Clare Island 53·7974 10·0431 10 – 
29 Mahon Falls 52·2324 7·5470 22 – – 
 Total   353 310 362 
Site no. Location Lat. (N) Long. (W) No. of samples
 
S. spathularis S. hirsuta S. × polita 
Glengariff Forest Park 51·7539 9·5649 22/22/22 – 22 
Glengariff Forest Park 51·7531 9·5652 – 
Glengariff Forest Park 51·7533 9·5649 22 16 
Glengariff Forest Park 51·7545 9·5694 11 11 
Glengariff Forest Park 51·7562 9·5713 – 14 22 
Glengariff Forest Park 51·7573 9·5724 22/4 15 15/8 
Glengariff Forest Park 51·7551 9·5674 22 22 22 
Glengariff Forest Park 51·7561 9·5713 
Glengariff Forest Park 51·7566 9·5671 – 
10 Glengariff Forest Park 51·7536 9·5613 19 16 
11 Kenmare Tunnel 51·7560 9·5872 20 17 
12 Beara Way 51·7072 9·6213 22 – 21 
13 Beara Way Bridge 51·7049 9·6523 11 19 9/11 
14 Dromagowlane 51·6886 9·6613 – 22 20 
15 Lauraghbridge 51·7675 9·7705 
16 Glenbeg Lough 51·7177 9·8737 22 12 
17 Cleanderry Wood 51·7377 9·9247 – 22 – 
18 Torc Mountain 51·9959 9·5083 20 – 15 
19 Torc Waterfall 52·0067 9·5083 17 10 12 
20 Meeting of the Waters 52·0048 9·5305 – 21 – 
21 Old Torc Mountain Path 52·0045 9·5305 – 22 
22 Gap of Dunloe 51·9918 9·6440 14/3 12 
23 Ballaghbeama Gap 51·9407 9·8047 16 22 
24 Camp 52·2152 9·9019 – 
25 Owenmore River Bridge 52·2076 10·2084 10 
26 Conor Pass 52·1818 10·2078 22 22 21 
27 Connemara National Park 53·5514 9·9458 – 22 
28 Clare Island 53·7974 10·0431 10 – 
29 Mahon Falls 52·2324 7·5470 22 – – 
 Total   353 310 362 

Multiple values indicate multiple patches sampled.

Single nucleotide polymorphism ascertainment and genotyping

Species-specific single nucleotide polymorphisms (SNPs) were developed from an ascertainment set containing four individuals of each of the two parent species using the ISSR (intersimple sequence repeat) cloning method outlined in Beatty et al. (2010). Allele-specific PCR (AS-PCR) primers were designed as described in Provan et al. (2008), and SNP analysis was carried out as described in Beatty et al. (2010). In total, 1025 individuals were genotyped for seven SNP loci (Table 2).

Table 2.

Saxifraga spathularis/S. hirsuta AS-PCR primers

Locus SNP Flanking primers SNP primers 
3A09 G→A AATATGTACTTTACCGTCCTC CAAGTCAGGGAGGGGAG 
TGTGGGAAGTTCAGCATTG TGCACTACGTAAGTACCCT 
3C09 4 bp indel AATCTAAACAAACCCTAGAAAAC CATTCCAGATAAAATATGGCTAC 
AAGTCCAAATATTTAAAAAATATATTTG AATTATGCCTAGACGAACTTG 
3G12 C→T TGTCTACTTTTTTCCCTATGC AATTTTTAATTTACACTAAAAACAGG 
GTATCTATAAACACATATTTATGAAA TGTTAGTATATGAAATTGAGAGTTT 
4C07 4 bp indel CATGCCATATAACTTGATAATAC GGTACGACTAAATCAACAATGG 
 GGTATGGCTAAATCAACATTGA 
4D05 C→T GCACTCTCTCCCTGCACC ATCTCAACGGTCAAAATTTATTC 
 ATCTCAACGGTCAAAACTTATTT 
4F03 A→G GATGTTCTGATTTTGAGAGAAG TTTTTCCCTTTTTTTCTGCCACA 
 CTTCACCTTTTTTCTGCCACG 
4H06 16 bp indel CTAGATCGCCGGCGATC CCTCTGTTATTCCAAATTGCG 
 ACACACAGAACAACTCTCTCTTCTC 
Locus SNP Flanking primers SNP primers 
3A09 G→A AATATGTACTTTACCGTCCTC CAAGTCAGGGAGGGGAG 
TGTGGGAAGTTCAGCATTG TGCACTACGTAAGTACCCT 
3C09 4 bp indel AATCTAAACAAACCCTAGAAAAC CATTCCAGATAAAATATGGCTAC 
AAGTCCAAATATTTAAAAAATATATTTG AATTATGCCTAGACGAACTTG 
3G12 C→T TGTCTACTTTTTTCCCTATGC AATTTTTAATTTACACTAAAAACAGG 
GTATCTATAAACACATATTTATGAAA TGTTAGTATATGAAATTGAGAGTTT 
4C07 4 bp indel CATGCCATATAACTTGATAATAC GGTACGACTAAATCAACAATGG 
 GGTATGGCTAAATCAACATTGA 
4D05 C→T GCACTCTCTCCCTGCACC ATCTCAACGGTCAAAATTTATTC 
 ATCTCAACGGTCAAAACTTATTT 
4F03 A→G GATGTTCTGATTTTGAGAGAAG TTTTTCCCTTTTTTTCTGCCACA 
 CTTCACCTTTTTTCTGCCACG 
4H06 16 bp indel CTAGATCGCCGGCGATC CCTCTGTTATTCCAAATTGCG 
 ACACACAGAACAACTCTCTCTTCTC 

Data analysis

As the populations studied comprised a mixture of parental species and hybrids, and thus were unlikely to be close to Hardy–Weinberg equilibrium, the commonly used software package Structure (Pritchard et al., 2000) was not used. Instead, the percentage of alleles from each parental species found in each individual was represented as a stacked histogram as suggested by Hauser et al. (2012). To investigate further the genetic composition of hybrid individuals, we used the program NewHybrids (V 1·1b; Anderson and Thompson, 2002). As suggested by the authors, the number of generations was restricted to two, which gave six classes of genotypes (both parents, F1, F2, BCS. spathularis and BCS. hirsuta). The program was run with the nuclear SNP data using 50 000 burn-in iterations followed by 500 000 Markov Chain Monte Carlo iterations using default priors for allele frequencies and mixing proportions.

Species distribution modelling

Species distribution modelling was carried out to determine suitable future climate envelopes for both species using the maximum entropy approach implemented in the MaxEnt software package (version 3.3.3; Phillips and Dudik, 2008). Species occurrence data in Ireland between 1950 and 2000 (330 and 181 occurrences for S. spathularis and S. hirsuta, respectively) were downloaded from the Global Biodiversity Information Facility data portal (http://www.gbif.org/). Current-day climatic data (1950–2000; Hijmans et al., 2005) at 2·5 minute resolution were clipped to encompass the island of Ireland (i.e. 10·75°W to 5·20°W, and 51·35°N to 55·55°N) to reduce potential problems associated with extrapolation. Models were generated using cross-validation of ten replicate runs under the default MaxEnt parameters. Model performance was assessed based on the area under the receiver operating characteristic curve (AUC). Models were projected onto climate data for the years 2050 and 2080 generated under the UKMO-HADCM3 model based on three different emissions scenarios (A1b, B2a and A2a; www.ccafs-climate.org). Outputs from the three models were averaged to give a single consensus model for each species at each time period.

RESULTS

Genetic analyses

Analysis of 1025 individuals from all three species, including potential hybrid S. × polita populations, indicated a broad spectrum of intraindividual allele frequencies, ranging from 100 % S. spathularis to 100 % S. hirsuta, with all levels of intermediate allelic composition and hybrid classification evident (Fig. 2). Saxifraga spathularis individuals tended to be largely genetically pure, exhibiting little or no introgression from S. hirsuta. Most individuals were either 100 % S. spathularis, and <5 % of individuals exhibited more than two S. hirsuta alleles. This was reflected in the NewHybrids analysis, which assigned the majority of the individuals clearly to S. spathularis, and only four as having a majority chance of falling onto the BCS. spathularis class. In total, S. hirsuta alleles only accounted for 4·7 % of the S. spathularis gene pool of the populations analysed. There was much greater evidence for introgression of S. spathularis alleles into S. hirsuta populations, with <15 % of individuals exhibiting 100 % S. hirsuta alleles. Overall, the level of introgression of S. spathularis alleles into S. hirsuta populations was 19·2 %, significantly higher than the opposite scenario (Mann–Whitney test; z = −14·37, P < 0·0001). Six individuals exhibited S. spathularis alleles at a frequency >0·5. The NewHybrids analysis showed high levels of assignment to the BCS. hirsuta class, indicating that populations exhibiting S. hirsuta morphology are more like a hybrid swarm, also containing some F2s. Populations of the hybrid, S. × polita, were generally comprised of F1s or F2s, with some evidence of backcrossing. All hybrid individuals exhibited a mixture of alleles from both parental species. In total, S. spathularis alleles were slightly predominant, at a frequency of 0·536.

Fig. 2.

Allele frequencies (above) and results of the NewHybrids analysis (below) for each of the three taxa studied. In all diagrams, each column represents a single individual. The length of each coloured segment in the NewHybrids is proportional to the Bayesian posterior probability of assignment to the corresponding genotypic class. Numbers refer to populations in Fig. 1 and Table 1.

Fig. 2.

Allele frequencies (above) and results of the NewHybrids analysis (below) for each of the three taxa studied. In all diagrams, each column represents a single individual. The length of each coloured segment in the NewHybrids is proportional to the Bayesian posterior probability of assignment to the corresponding genotypic class. Numbers refer to populations in Fig. 1 and Table 1.

Species distribution modelling

The mean AUC values [0·881 (s.d. = 0·034) and 0·957 (s.d. = 0·037) for S. spathularis and S. hirsuta, respectively] indicated a prediction that was far better than random. The present-day (2000) models for both species were an accurate representation of their actual distributions, and projected future distributions indicated an increase in suitable habitat for both species (Fig. 3). A comparison of the area in south-west Ireland where both species are currently found sympatrically indicated an increase in suitable cells of approx. 18 % for S. spathularis between 2000 and 2080, and an increase of approx. 91 % for S. hirsuta over the same time period (Fig. 4). The proportion of suitable cells for S. hirsuta that were also suitable for S. spathularis rose from 94 % in 2000 to 100 % in 2050 and 2080.

Fig. 3.

Species distribution models for S. spathularis (A–C) and S. hirsuta (D–F). (A, B) 2000; (C, D) 2050; (E, F) 2080.

Fig. 3.

Species distribution models for S. spathularis (A–C) and S. hirsuta (D–F). (A, B) 2000; (C, D) 2050; (E, F) 2080.

Fig. 4.

Number of climatically suitable cells predicted by the species distribution models for S. spathularis and S. hirsuta in south-western Ireland for each of tthe three time periods.

Fig. 4.

Number of climatically suitable cells predicted by the species distribution models for S. spathularis and S. hirsuta in south-western Ireland for each of tthe three time periods.

DISCUSSION

The present study further highlights the utility of species-specific SNPs in studies into hybridization in plant taxa, indicating the occurrence of cryptic introgression into both S. spathularis and S. hirsuta, but particularly the latter. Nevertheless, despite developing primer pairs from the ISSR libraries to amplify 96 sequence-tagged sites (STS), we only managed to develop seven sets of AS-PCR primers that could consistently be used to genotype SNPs. This was mainly due to difficulties in amplifying the orthologous locus in S. hirsuta using primers developed from the S. spathularis ISSR library, even after relaxing PCR conditions by lowering the annealing temperature. This was surprising, given the fact that they are sister species (Sanna, 2013), and given that cross-genus amplification of orthologous loci was possible using the same ISSR-based approach in Pyrola (Beatty et al., 2010). Furthermore, it should be borne in mind that genetically more similar orthologues might be more likely to undergo recombination, thus facilitating introgression.

The findings of this study would seem to indicate that levels of introgression are primarily density dependent. This is most apparent in Cleanderry Wood (Site 17, Fig. 2), which is the only location that harboured a large proportion (15/22) of genetically pure S. hirsuta, and where no populations of either S. spathularis or S. × polita were found in the vicinity. Similarly, the majority of S. hirsuta collected from Site 10 in Glengariff Forest Park, where only two S. spathularis plants were found, exhibited 100 % S. hirsuta alleles. Further evidence of this is seen at sites where hybrids are found with only one of the parental species. Significantly different frequencies of S. spathularis alleles were found in hybrid populations from Connemara National Park (Site 27, Fig. 2), where S. hirsuta has never been recorded (0·629), compared with Dromagowlane (Site 14, Fig. 2), where S. spathularis is absent (0·347; Mann–Whitney test; z = −5·53, P < 0·0001). Such density-dependent introgression has previously been cited as a threat to the persistence of a rare species sympatrically with a more abundant congener (Ellstrand and Elam, 1993; Burgess et al. 2005; Chan et al., 2006; Lajbner et al., 2009; Beatty et al., 2010).

Individuals identified as S. × polita all contained a mixture of parental alleles, with S. spathularis allele frequencies ranging from 0·214 to 0·928. The majority were assigned in the NewHybrids analysis to F1 or F2 classes, with some backcrosses to both S. spathularis and S. hirsuta. This is consistent with the observation that wild hybrids within Section Gymnopera (S. × polita in Ireland and S. × geum in the western Pyrenees) are fertile, and form hybrid swarms that display a full range of intermediate morphologies between the relevant parental species (Webb, 1951; McGregor, 2008). Given the apparent fertility of hybrids, and indications of density-dependent introgression, there is thus the real chance that, over time, the rarer S. hirsuta could be under threat of genetic assimilation by the more abundant S. spathularis (Levin et al., 1996; Beatty et al., 2010). Indeed, >50 years ago, Webb had already highlighted the fact that S. hirsuta was rarer than the hybrid S. × polita, and that ‘…  S. hirsuta could not be maintained as a distinct species’ (Webb, 1951, p. 204).

The occurrence of populations of S. × polita in Co. Galway, where only one of the parental species – S. spathularis – is present, is unusual. The extremely limited number of historical records of the occurrence of S. hirsuta in the region have been subsequently attributed to misidentification of the hybrid (Stelfox, 1947). An examination of herbarium samples held at the National Botanic Gardens of Ireland, Glasnevin, found a single sheet of S. hirsuta samples from the early 20th Century, labelled as a ‘garden escape’ and collected in the vicinity of Letterfrack, Co. Galway, which is on the periphery of the current distribution of both S. spathularis and S. × polita. It is possible that this – or other – garden plants might have contributed to the formation of populations of the hybrid.

Although S. hirsuta is generally far less common that S. spathularis in south-west Ireland, where both species occur sympatrically, species distribution modelling under future climatic scenarios indicated an increase in suitable habitat for the former relative to its congener. Assuming the asymmetric introgression observed in the present study is density dependent, this would suggest that future introgression of S. spathularis alleles into S. hirsuta may not be as pronounced as at present. However, two potential caveats should be borne in mind. First, it is well documented that such species distribution models only take into account climatic factors and tend not to incorporate other biotic and abiotic variables, such as species–species interactions (Pearson and Dawson, 2003; Araújo and Guisan, 2006). Secondly, it is very possible that introgression of S. spathularis alleles may lead to a shift in the ecological and/or climatic niche of S. hirsuta, thus compromising these modelled potential future range expansions. Nevertheless, it is likely that ongoing environmental change will lead to changes in ecosystem function, and this could also be a factor at the level of the genome.

In conclusion, the levels of introgression observed in this study in both S. spathularis and S. hirsuta would appear to be correlated with the relative abundance of the species. This is reflected in the genetic composition of populations of the hybrid, S. × polita, and its sympatric occurrence with one or both parental species. From a conservation viewpoint, although significant levels of introgression of S. spathularis alleles were detected in the majority of the S. hirsuta populations analysed, there were only a few individuals in which S. hirsuta alleles were in a minority. This suggests that the S. hirsuta morphology is being retained, despite varying degrees of introgression. Nevertheless, ongoing introgression of S. spathularis alleles into S. hirsuta would appear to represent a threat to the genetic integrity of the latter, particularly in areas where the species exist sympatrically and where S. spathularis outnumbers S. hirsuta.

ACKNOWLEDGEMENTS

The authors are grateful to Robert Beatty for assistance with sampling. This project was funded by British Ecological Society Research grant no. 4309-5281 to G.E.B.

LITERATURE CITED

Anderson
E
Stebbins
GL
Jr
.
1954
.
Hybridization as an evolutionary stimulus
.
Evolution
 
8
:
377
388
.
Anderson
EC
Thompson
EA
.
2002
.
A model-based method for identifying species hybrids using multilocus genetic data
.
Genetics
 
160
:
1217
1229
.
Araújo
MB
Guisan
A
.
2006
.
Five (or so) challenges for species distribution modelling
.
Journal of Biogeography
 
33
:
1677
1688
.
Barton
NH
.
2001
.
The role of hybridization in evolution
.
Molecular Ecology
 
10
:
551
568
.
Beatty
GE
Provan
J
.
2013
.
Postglacial dispersal, rather than in situ glacial survival, best explains the disjunct distribution of the Lusitanian plant species Daboecia cantabrica (Ericaceae)
.
Journal of Biogeography
 
40
:
335
344
.
Beatty
GE
Provan
J
.
2014
.
Phylogeographical analysis of two cold-tolerant plants with disjunct Lusitanian distributions does not support in situ survival during the last glaciation
.
Journal of Biogeography
 
(in press)
.
Beatty
GE
Philipp
M
Provan
J
.
2010
.
Unidirectional hybridization at a species’ range boundary: implications for habitat tracking
.
Diversity and Distributions
 
16
:
1
9
.
Burgess
KS
Morgan
M
Devemo
L
Husband
BC
.
2005
.
Asymmetrical introgression between two Morus species (M. alba, M. rubra) that differ in abundance
.
Molecular Ecology
 
14
:
3471
3483
.
Chan
C
Ballantyne
KN
Aikman
H
Fastier
D
Daugherty
CH
Chambers
GK
.
2006
.
Genetic analysis of interspecific hybridization in the world’s only Forbes’ parakeet (Cyanoramphus forbesi) natural population
.
Conservation Genetics
 
7
:
493
506
.
Doyle
JJ
Doyle
JL
.
1987
.
A rapid DNA isolation procedure for small quantities of fresh leaf tissue
.
Phytochemical Bulletin
 
19
:
11
15
.
Ellstrand
NC
Elam
DR
.
1993
.
Population genetic consequences of small population size: implications for plant conservation
.
Annual Review of Ecology and Systematics
 
24
:
217
242
.
Garroway
CJ
Bowman
J
Cascaden
TJ
et al
.
2010
.
Climate change induced hrbridization in flying squirrels
.
Global Change Biology
 
16
:
113
121
.
Hauser
TP
Borh
MB
Dienst
DN
Nielsen
KB
Philipp
M
.
2012
.
Hybridization between Pyrola grandiflora and Pyrola minor in West Greenland: a tension zone maintained by clonality
.
Botany
 
90
:
1036
1047
Hickling
R
Roy
DB
Hill
JK
Fox
R
Thomas
CD
.
2006
.
The distributions of a wide range of taxonomic groups are expanding polewards
.
Global Change Biology
 
12
:
450
455
.
Hijmans
RJ
Cameron
SE
Parra
JL
Jones
PG
Jarvis
A
.
2005
.
Very high resolution interpolated climate surfaces for global land areas
.
Journal of Climatology
 
25
:
1965
1978
.
Kelly
AE
Goulden
ML
.
2008
.
Rapid shifts in plant distribution with recent climate change
.
Proceedings of the National Academy of Sciences, USA
 
105
:
11823
11826
.
Lajbner
Z
Šlechtová
V
Šlechta
V
Švátora
M
Berrebi
P
Kotlík
P
.
2009
.
Rare and asymmetrical hybridization of the endemic Barbus carpathicus with its widespread congener Barbus barbus
.
Journal of Fish Biology
 
74
:
418
436
.
Levin
DA
Francisco-Ortega
J
Jansen
RK
.
1996
.
Hybridization and the extinction of rare plant species
.
Conservation Biology
 
10
:
10
16
.
Lewontin
RC
Birch
LC
.
1966
.
Hybridization as a source of variation for adaptation to new environments
.
Evolution
 
20
:
315
336
.
McGregor
M
.
2008
.
Saxifrages: a definitive guide to the 2000 species, hybrids and cultivars
 .
London
:
Timber Press
.
Muhlfeld
CC
Kovach
RP
Jones
LA
et al
.
2014
.
Invasive hybridization in a threatened species is accelerated by climate change
.
Nature Climate Change
 
4
:
620
624
.
Parmesan
C
Yohe
G
.
2003
.
A globally coherent fingerprint of climate change impacts across natural systems
.
Nature
 
421
:
37
42
.
Pearson
RG
Dawson
TP
.
2003
.
Predicting the impacts of climate change on the distribution of species: are bioclimate envelope models useful?
Global Ecology and Biogeography
 
12
:
361
371
.
Phillips
SJ
Dudik
M
.
2008
.
Modelling of species distributions with Maxent: new extensions and a comprehensive evaluation
.
Ecography
 
31
:
161
175
.
Potts
WM
Henriques
R
Santos
CV
et al
.
2014
.
Ocean warming, a rapid distributional shift, and the hybridization of a coastal fish species
.
Global Change Biology
 
20
:
2765
2777
.
Pritchard
JK
Stephens
M
Donnelly
P
.
2000
.
Inference of population structure using multilocus genotype data
.
Genetics
 
155
:
945
959
.
Provan
J
Beatty
GE
Hunter
AM
et al
.
2008
.
Restricted gene flow in fragmented populations of a wind-pollinated tree
.
Conservation Genetics
 
9
:
1521
1532
.
Sanchez-Guillen
RA
Munoz
J
Hafernik
J
Tierney
M
Rodriguez-Tapia
G
Cordoba-Aguilar
A
.
2014
.
Hybridization rate and climate change: are endangered species at risk?
Journal of Insect Conservation
 
18
:
295
305
.
Sanna
M
.
2013
.
Diversidad filogeografía y systemática de Saxifraga hirsuta L
 .
Universidad de Oviedo
.
Seehausen
O
.
2004
.
Hybridization and adaptive radiation
.
Trends in Ecology and Evolution
 
19
:
198
207
.
Stelfox
AW
.
1947
.
C. C. Babington’s record for Saxifraga geum in Connemara
.
Irish Naturalists’ Journal
 
9
:
29
.
Webb
DA
.
1951
.
Hybrid plants in Ireland
.
Irish Naturalists’ Journal
 
10
:
201
205
.
Webb
DA
.
1983
.
The flora of Ireland in its European context
.
Journal of Life Sciences, Royal Dublin Society
 
4
:
143
160
.
Webb
DA
Gornall
RJ
.
1989
.
Saxifrages of Europe
 .
London
:
Christopher Helm Ltd
.

Comments

0 Comments