Skip Navigation



AOBPreview published online on July 26, 2007
Annals of Botany, doi:10.1093/aob/mcm154
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
100/4/735    most recent
mcm154v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Albaladejo, R. G.
Right arrow Articles by Aparicio, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Albaladejo, R. G.
Right arrow Articles by Aparicio, A.
Agricola
Right arrow Articles by Albaladejo, R. G.
Right arrow Articles by Aparicio, A.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2007. Published by Oxford University Press on behalf of the Annals of Botany Company. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Population Genetic Structure and Hybridization Patterns in the Mediterranean Endemics Phlomis lychnitis and P. crinita (Lamiaceae)

Rafael G. Albaladejo* and Abelardo Aparicio

Departamento de Biología Vegetal y Ecología, Facultad de Farmacia, Universidad de Sevilla, E-41012 Sevilla, Spain

* For Correspondence. E-mail albaladejo{at}us.es

Received: 14 March 2007    Returned for revision: 16 April 2007    Accepted: 8 June 2007   


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 LITERATURE CITED
 

Background and Aims: The historical influence of gene flow and genetic drift after the last glacial phase of the Quaternary Period is reflected in current levels of genetic diversity and population structure of plant species. Moreover, hybridization after secondary contact might also affect population genetic diversity and structure. An assessment was made of the genetic variation and hybrid zone structure in Iberian populations of the Mediterranean Phlomis lychnitis and P. crinita, for which phylogenetic relationships are controversial, and hybridization and introgression are common.

Methods: Allozyme variation at 13 loci was analysed in 1723 individual plants sampled from 35 natural locations of P. lychnitis, P. crinita subsp. malacitana and P. crinita subsp. crinita in southern and eastern Spain. Standard genetic diversity parameters were calculated and patterns of genetic structure in each taxon were tested to fit the equilibrium between gene flow and genetic drift. Individual multilocus genotypes were subjected to Bayesian clustering analysis to estimate hybridization and introgression rates for both geographic regions.

Key Results: Contrasting patterns in the distribution of genetic variation among the three taxa were found. Phlomis lychnitis showed no significant inbreeding, low genetic differentiation among populations and no evidence of isolation by distance. Phlomis crinita subsp. malacitana and P. crinita subsp. crinita showed high levels of genetic structure consistent with a pattern of gene flow–drift equilibrium. Higher instances of hybridization and introgression were detected in locations from southern Spain compared with locations from eastern Spain, matching unimodal and bimodal hybrid zones, respectively.

Conclusions: High instances of historical gene flow, range expansion and altitudinal movement during the Quaternary Period, and lineage sorting can explain the diversity of patterns observed. The results suggest that P. lychnitis is the most differentiated lineage in the group; however, the relationship between the three taxa remains unclear.

Key words: Allozymes, glacial refuges, hybridization, Iberian Peninsula, introgression, isolation by distance, Lamiaceae, lineage sorting, gene flow–drift equilibrium, Phlomis


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 LITERATURE CITED
 
Historical processes play a major role as driving forces shaping the contemporary patterns of genetic architecture in natural populations (Comes and Kadereit, 1998; Taberlet et al., 1998; Hewitt, 2001). Recent studies on patterns of genetic variation in plant species within areas that have historically acted as refuges for biodiversity, such as the peninsulas around the Mediterranean Basin, have revealed complex patterns that mostly result from the Quaternary Period (Gutiérrez Larena et al., 2002; Olalde et al., 2002; Bittkau and Comes, 2005). Although pinpointing the location of glacial refuges and migration routes is a challenging task, the scarce available data indicate that the mountain ranges of southern and eastern Spain (Betic mountains) acted as refuges for many forest plant species, when most lowlands on the Iberian Peninsula had sub-Saharian climatic conditions (Carrion, 2002). The climate changes that occurred during the Quaternary altered the altitudinal range of some species and resulted in a scattered mosaic of fragmented populations that are currently restricted to the massive Betic mountains (e.g. Armeria sp., Gutiérrez-Larena et al., 2002; Anthyllis montana, Kropf et al., 2002). Other plant species quickly expanded their distribution range through horizontal migration from these refuges to the central and western Iberian Peninsula (e.g. Pinus pinaster, Salvador et al., 2000; Quercus ilex, Lumaret et al., 2002). However, in contrast to the knowledge about the recent plant recolonization of central and northern Europe (Hewitt, 2004), few studies have been conducted on plant species in glacial refuges except for forest trees (e.g. Petit et al., 2003), and further knowledge is required to obtain a clearer picture of how historical factors have structured the current genetic variation of these species.

Furthermore, plant range expansion after the last glacial episode allowed the secondary contact of formerly isolated lineages, providing the opportunity for interspecific matings and the formation of hybrid zones (Hewitt, 2001). Hybrid zones might persist if pre-zygotic barriers against hybridization between divergent lineages are lacking or weakly developed, and/or if selection (either endogenous or exogenous) acting upon hybrid progeny is not remarkable. In this case, populations will usually consist of pure parental genotypes and a wide array of hybrids and backcrosses (i.e. a hybrid swarm; Harrison, 1993). In contrast, if nearly complete barriers against hybridization are developed and/or hybrids are overwhelmingly negatively selected, hybrid zones might be ephemeral and populations will comprise parental genotypes and a few hybrids. These two extremes have been referred to as ‘unimodal’ and ‘bimodal’ hybrid zones, respectively (Harrison and Bogdanowicz, 1997). Jiggins and Mallet (2000) have stated that this hybrid zone classification is not discrete and exists as a continuum from unimodality to bimodality towards the complete isolation of lineages on the way to speciation, although it might also take place in the opposite direction when two previously isolated lineages mate and merge to form a single lineage (Rhymer and Symberloff, 1996). Investigating the structure of hybrid zones might shed some light on the processes that maintain isolation between lineages. However, high rates of introgression between lineages also affect patterns of genetic diversity by increasing levels of polymorphism (Rieseberg and Wendel, 1993), and determining whether these polymorphisms are the result of hybridization or acquired through the divergence process is a complicated task.

Phlomis lychnitis L. (Iberian Peninsula and southern France) and P. crinita Cav. (southern and eastern Spain) are two species endemic to the Mediterranean region and constitute the morphological extremes of variation in a hybrid complex where hybridization and lineage sorting have played a major evolutionary role (Albaladejo et al., 2005). An ecogeographic and morphometric analysis of the hybrid complex (Albaladejo et al., 2004) demonstrated the existence of two allopatric sub-specific entities within P. crinita – P. crinita subsp. malacitana (Pau) Cabezudo, Nieto Caldera and Navarro (above 1000 m in the mountains of southern Spain) and P. crinita subsp. crinita (below 800 m in the mountains of eastern Spain). Intermediate morphotypes between each P. crinita sub-species and P. lychnitis are common in southern Spain but rare in eastern Spain. Karyologically these plants conform to a homoploid (diploid; 2n = 20) hybrid complex; however, meiotic aberrations and pollen abnormalities (chromosome clumping, chromatinic bridges, occurrence of multivalents/univalents and polysporads) are common (Aparicio and Albaladejo, 2003). These data were interpreted as evidence for hybridization and introgression between P. crinita subsp. malacitana and P. lychnitis rather than between P. crinita subsp. crinita and P. lychnitis.

Furthermore, genetic support for hybridization and introgression in the hybrid complex was recently gained through a study of nuclear internal transcribed spacer (ITS) and non-coding plastid DNA sequence variation (Albaladejo et al., 2005). This study showed that ITS sequence variation agrees with taxonomic delimitations, although the presence of nucleotide polymorphisms in the ITS sequences is the rule rather than the exception. In contrast, chloroplast DNA (cpDNA) variation follows a geographic trend through the sharing of haplotypes among taxa in different geographic areas (i.e. southern Spain vs. eastern Spain). The authors suggested localized instances of introgression, coupled with the sorting of polymorphisms among lineages through the divergence process, as possible causes for the incongruence between markers (Albaladejo et al., 2005). However, the genetic relationships in this group remain controversial because of the low variability in nuclear and plastid DNA sequences that is inherently linked to the recent divergence of lineages.

In this study, allozyme variability was used to evaluate the relative historical influence of gene flow, genetic drift and hybridization on the genetic diversity and regional population structure of this Mediterranean endemic hybrid complex, and to deepen understanding of their biosystematics.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 LITERATURE CITED
 
Studied plants
All the taxa studied are common suffruticose chamaephytes growing in dry open places in disturbed helm-oak vegetation. The plants have vegetative spread and produce yellow to brown flowers, which are protandrous, nectariferous and visited by bumblebees (Brantjes, 1981). They have a self-compatible breeding system; however, insect visitation is required because of low fruit set through spontaneous selfing (R.G.A. and A.A., unpublished data). After fertilization, one (seldom two) nucule develops within the persistent calyx, supposedly as an adaptive mechanism to promote dispersal by wind (Aparicio, 1997).

Sampling and electrophoresis
To quantify genetic diversity and determine population structure of the three parental taxa, 35 locations were sampled covering the distribution range of the complex (Table 1 and Fig. 1). Phlomis lychnitis was collected from 23 populations (16 in southern Spain and seven in eastern Spain), P. crinita subsp. malacitana was collected from 16 populations and P. crinita subsp. crinita was collected from ten populations. To characterize hybrid zone structure in both geographic regions, morphologically intermediate plants – putative hybrids – that co-occurred with both parental types (i.e. mixed/sympatric zones) were sampled. Twelve such putative hybrid or contact zones were sampled across the studied area, nine in southern Spain (locations 1, 6, 8, 12, 13, 14, 15, 16 and 21; Table 1 and Fig. 1) and three in eastern Spain (locations 27, 32 and 35). The rarity of hybrid plants in eastern Spain limited the sampling. Individual plants were classified in the field as parental or hybrid phenotypes by their morphological traits, as detailed in Albaladejo et al. (2004). In all populations, plants were collected at least 2 m apart to avoid the sampling of different ramets from the same genet. Sampled leaves were refrigerated on ice until proteins were extracted by grinding small pieces of leaves in three drops of extraction buffer (Werth, 1985). Crude extracts were absorbed on chromatography paper wicks and stored at –80 °C until they were electrophoresed.


Figure 1
View larger version (83K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 1. The 35 locations in the Iberian Peninsula from which Phlomis individuals were sampled. Numbers correspond to those in Table 1: 1–22 southern Spain, 23–35 eastern Spain. Black circles correspond to the 12 hybrid or contact zones studied: nine in southern Spain and three in eastern Spain. Dashed lines represent the main clusters of P. crinita in the UPGMA phenogram.

 


View this table:
[in this window]
[in a new window]

  		TABLE 1. Location and population codes of the Phlomis taxa studied in southern and eastern Spain

 
Starch gel electrophoresis was conducted following the general methods of Weeden and Wendel (1989), which have been previously used in this genus by Aparicio et al. (2000). Electrophoreses were carried out in 9 % starch (Sigma) gels containing 2·5 % sucrose. Nine enzyme systems were used to resolve 13 loci, which provided consistent patterns of activity and reproducibility throughout the samples analysed. Malate dehydrogenase (MDH; EC 1·1·1·37), malic enzyme (ME; EC 1·1·1·40), phosphoglucoisomerase (PGI; EC 5·3·1·9), phosphoglucomutase (PGM; EC 2·7·5·1), 6-phosphogluconate dehydrogenase (6-PGD; EC 1·1·1·44) and shikimate dehydrogenase (SKDH; EC 1·1·1·25) were resolved in a morpholine citrate buffer (Clayton and Tretiak, 1972), whereas isocitrate dehydrogenase (IDH; EC 1·1·1·42), colorimetric esterase (EST; EC 3·1·1) and aspartate aminotransferase (AAT; EC 2·6·1·1) were resolved in a lithium borate/Tris citrate buffer (Ashton and Branden, 1961). Staining recipes followed those previously reported by Soltis et al. (1983). Gels were run for 6–7 h at 40 and 50 mA for morpholine citrate and lithium borate/Tris citrate buffers, respectively. Loci and alleles within the gels were numbered and labelled, starting from the most anodally migrating form. Overall, 1723 allozyme multilocus genotypes were generated.

Genetic diversity and population structure in P. lychnitis and P. crinita
Allele frequencies were used to compute the percentage of polymorphic loci at the 99 % cut-off level (P99), mean number of alleles per locus (A), expected heterozygosity (He) and observed heterozygosity (Ho), with the software GDA 1·1 (Lewis and Zaykin, 2001). Because differences in sampling intensity can bias comparisons for allelic richness, A was standardized to A[g] on the basis of the smallest number of samples following a rarefaction method adapted for population genetic data by El Mousadik and Petit (1996). At the population level, g was set equal to 20 (the sample size for P. lychnitis at location 30 was ten individuals), and at the taxonomic level g was set equal to 500 as the standard sample size of gene copies (286 sampled plants, i.e. 572 gene copies, in P. crinita subsp. crinita). Allelic richness was computed with the software Rarefac (Petit et al., 1998). The inbreeding coefficient (FIS; Weir and Cockerham, 1984) for each population was calculated with FSTATS 2·9·3 (Goudet, 2001), and significant heterozygote deficiency (HD) or excess (HE) tested by randomizing alleles among individuals.

At the taxonomic level, deviations from Hardy–Weinberg (HW) equilibrium and population structure were estimated by computing the inbreeding coefficient (FIS) and the genetic differentiation (FST) unbiased estimators (Weir and Cockerham, 1984). Statistical significance was tested by constructing 95 % confidence intervals by bootstrapping over loci (2000 replicates) with GDA 1·1 (Lewis and Zaykin, 2001).

Concordance with migration–drift equilibrium (‘isolation-by-distance’) in the three taxa was explored by plotting pairwise FST/(1 – FST) values (Rousset, 1997) against Euclidean geographic distances between all sample sites. Mantel test (1000 permutations) and reduced major axis (RMA) regression (Sokal and Rohlf, 1981) were conducted to assess the significance and strength of the relationship between genetic and geographic distances with the software IBD (Bohonak, 2002). Additionally, the absolute residuals of the regressions were saved and then correlated with the geographic distance (1000 permutations) to test whether the spread of the data increased with geographic distance, following Hutchinson and Templeton (1999). For organisms with restricted dispersal ability, these authors stated that at gene flow–drift equilibrium, the scatter of points in a two-dimensional scatterplot should increase with the geographic distance because of the prominent role of genetic drift over larger distances.

Patterns of hybridization and introgression
The individual multilocus allozyme data in each of the 12 hybrid or contact zones (n = 1020 individuals) were subjected to the Bayesian clustering method implemented in the software Structure 2·0 (Pritchard et al., 2000; Falush et al., 2003) to determine and compare hybridization and introgression patterns between regions. Unlike previous methods that are based on maximum likelihood (e.g. Nason and Ellstrand, 1993), the algorithm behind Pritchard's software does not need to specify the gene frequencies of the parental sources in advance and has been proven to work efficiently, even in complex hybrid swarms where parental frequencies could not be confidently estimated (e.g. Beaumont et al., 2001). In this analysis, individuals were assigned to pre-defined K populations having homogeneous allele frequencies. First, it was checked whether the program was consistent in identifying the two expected parental populations (species in the present case) for each putative hybrid zone computing the ad hoc statistic {Delta}K described by Evanno et al. (2005). Briefly, for each hybrid zone, 20 simulations were run for each group, from K = 1 to K = 5, and then the statistic {Delta}K was computed, which detected the highest rate of change in the log-likelihood between successive Ks (see Evanno et al. 2005 for a detailed graphic explanation about {Delta}K calculations). All simulations were run setting 105 iterations as the burn-in period, and 105 iterations for Markov chain convergence. Preliminary runs using higher burn-in periods and Markov chain iterations did not produce different results. For each run, the admixture model was set, allele frequencies were allowed to be correlated and the previous classification of individuals based on morphological traits was not taken into account. Secondly, after estimating the most likely number of ‘genetic populations’ present in each hybrid zone (K = 2 in all cases; see Results), the individuals were classified as pure types (P. lychnitis, P. crinita subsp. malacitana and P. crinita subsp. crinita) or hybrids considering a somewhat relaxed criterion of 90 % posterior probability, which allowed the classification within pure types of some slightly introgressed individuals (see Valbuena-Carabaña et al., 2007). Individuals with a probability between 40 and 60 % were considered hybrids, and individuals with a probability between 11 and 39 % or 61 and 89 % were considered introgressants.

Genetic differentiation and relationships among taxa
To address the genetic relationship among the taxa studied, genetic differentiation (FST; Weir and Cockerham, 1984) and Nei's (1972) genetic identities were computed. At the population level, genetic relationships were assessed by estimating pairwise genetic distances (Nei, 1972) and plotted through a UPGMA phenogram and multidimensional scaling (MDS) analyses conducted with Statistica 6·0 (StatSoft, 2001). To depict connections among genetically close populations, a minimum spanning tree (MST) was computed with Passage 1·1 (Rosenberg, 2004) and superimposed on the scatterplot of the samples against the first two MDS dimensions.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ACKNOWLEDGEMENTS
 LITERATURE CITED
 
Overall, the allozyme analysis reflected limited variability. Eight of the 13 consistently resolved loci were polymorphic (Pgi, Pgm, 6-Pgd, Skdh-1, Skdh-2, Aat-1, Est-1 and Est-2), comprising 23 alleles. Three alleles were exclusive, i.e. occurring in some populations of only one taxon; however, all of these were present at low frequencies (<0·03) (Table 2). Additionally, 128 multilocus genotypes were detected; none of them was found to be unique, but 33 of them (25 % of the genotypes) were shared by only two individuals.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Allele frequencies, Weir and Cockerham's (1984) unbiased estimators of inbreeding coefficient (FIS) and population genetic differentiation (FST), and overall 95 % confidence intervals (95 % CI) for eight polymorphic loci in the studied taxa of Phlomis

 
Population genetic diversity and structure in P. lychnitis and P. crinita
At the population and taxonomic level, P. lychnitis showed the highest diversity for all calculated parameters (Table 3). The two P. crinita sub-species displayed very similar values of diversity, except for heterozygosity. The expected heterozygosity across loci was somewhat higher than the observed heterozygosity in P. lychnitis and P. crinita subsp malacitana. However, in P. crinita subsp. crinita, He was barely 2-fold higher than Ho (Table 3). Most analysed populations conformed to HW proportions. Four out of 23 populations of P. lychnitis showed a significant departure from HW equilibrium (two with heterozygote excess and deficit, respectively). In P. crinita subsp. malacitana, one population displayed a deficit and another one an excess of heterozygotes, while in P. crinita subsp. crinita two populations consistently showed a significant excess of homozygotes (Table 3). Accordingly, multilocus FIS estimates over all populations were not significantly different from zero in P. lychnitis (FIS = –0·073) and P. crinita subsp. malacitana (FIS = 0·003), but showed a significant excess of homozygotes (FIS = 0·144) in P. crinita subsp. crinita (Table 2).


View this table:
[in this window]
[in a new window]

  		TABLE 3. Population genetic statistics analysed for the studied populations of Phlomis; proportion of polymorphic loci (P99), number of alleles per locus (A), allelic richness (A[g]), expected (He) and observed (Ho) heterozygosity, and inbreeding coefficient (FIS). Population code as in Table 1

 
All three taxa showed a significant partitioning of the genetic variation among populations (Table 2). Phlomis lychnitis displayed a moderate FST value (0·108), whereas P. crinita subsp. malacitana and P. crinita subsp. crinita showed a markedly higher degree of genetic subdivision (FST = 0·290 and FST = 0·419, respectively). Mantel test correlation between geographic and genetic distance was not significant in P. lychnitis (r = 0·126, P = 0·103), with <2 % of the variation in genetic differentiation explained by geographic distance (Fig. 2). In contrast, correlation between genetic and geographic distances was significantly positive in P. crinita subsp. malacitana (r = 0·531, P <0·001) and in P. crinita subsp. crinita (r = 0·556, P = 0·011), with geographic distance accounting for a similar proportion of the variance in both subspecies (28 % in P. crinita subsp. malacitana and 31 % in P. crinita subsp. crinita; Fig. 2). Correlation analysis of the absolute residuals of the regression showed a significant increase in the scatter of pairwise points with spatial distance (r = 0·524, P = 0·001 in P. crinita subsp. malacitana and r = 0·405, P = 0·024 in P. crinita subsp. crinita), suggesting the achievement of equilibrium between gene flow and drift among populations in these taxa.


Figure 2
View larger version (11K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 2. Scatterplot of Rousset's (1997) genetic distance against geographic distance for the Phlomis taxa studied. The slope (b) and the explained variance (r2) from reduced major axis (RMA) regressions are also shown for each taxon.

 
Patterns of hybridization and introgression
The ad hoc approach used to infer the most likely number of populations consistently showed that K = 2 was the optimal number of clusters in all the hybrid or contact zones studied (see Table 4), which is consistent with the existence of two well-differentiated species. Based on the criterion of 90 % posterior probability of being a parental type, contrasting patterns of hybridization and introgression were evident between the areas in southern and eastern Spain (Fig. 3). In southern Spain, the number of potential hybrids and introgressed individuals between P. lychnitis and P. crinita subsp. malacitana ranged between 32 % (locations 13 and 14) and 79 % (location 8), with an average value of 48 % for the region. In southern Spain, introgression was apparently bidirectional because it occurred at similar rates towards both parental types. It is noteworthy that only 1·3 % (11/832) of the sampled individuals in these nine hybrid zones that were classified as hybrid phenotypes in the field were determined to be pure types by the allozyme analysis. In contrast, hybridization and introgression rates between P. lychnitis and P. crinita subsp. crinita in eastern Spain were extremely low because most individuals were assigned to parental types and only 4 % (7/188) displayed a multilocus genotype that could be considered hybrid or introgressant.


Figure 3
View larger version (20K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 3. Classes of posterior probabilities for belonging to the parental species P. lychnitis, derived from the software Structure (Pritchard et al. 2000), in 12 hybrid or contact zones across the Phlomis complex. The x-axis shows the identification code for each location (see Table 1 and Fig. 1). Zones between P. crinita subsp. malacitana and P. lychnitis in southern Spain are locations 1–21, and hybrid zones between P. crinita subsp. crinita and P. lychnitis in eastern Spain are locations 27–35.

 


View this table:
[in this window]
[in a new window]

  		TABLE 4. Log-likelihood (LnK) (±s.d.) computed by the software Structure (Pritchard et al. 2000) and {Delta}K values calculated following Evanno et al. (2005) for K groups (from K = 1 to K = 5) in 12 Phlomis hybrid zones located in southern and eastern Spain

 
Genetic differentiation and genetic relationship among taxa
Nei's genetic identities over all localities among the three taxa ranged from 0·822 (for the pair P. crinita subsp. malacitana/P. lychnitis) to 0·922 (for the pair P. crinita subsp. crinita/P. crinita subsp. malacitana). The genetic identity for P. lychnitis/P. crinita subsp. crinita was 0·887· Accordingly, FST values among taxa were very high: 0·652 for P. crinita subsp. malacitana/P. lychnitis, 0·520 for P. crinita subsp. crinita/P. lychnitis and 0·489 for P. crinita subsp. malacitana/subsp. crinita, indicating that despite the low occurrence of exclusive alleles there were marked differences in allelic frequencies among the studied taxa.

The UPGMA phenogram grouped all populations of P. lychnitis together regardless of their geographic origin (Fig. 4). For instance, the population LY12 from southern Spain clustered together with populations LY28 and LY30 from eastern Spain, located >400 km away. In contrast, clustering of P. crinita subsp. malacitana populations was consistent with the mountain ranges of origin of the samples (population MA18 was the only exception). Populations of P. crinita subsp. crinita also grouped following a geographic pattern; however, they did not cluster together: populations CR25 and CR26 (inland populations in Fig. 1) clustered close to P. lychnitis, whereas populations CR27–CR35 (coastal populations) clustered close to P. crinita subsp. malacitana.


Figure 4
View larger version (11K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 4. UPGMA phenogram, based on Nei's genetic distance, for the populations of Phlomis studied. Population code is as in Table 1. The main clusters of P. crinita are mapped in Fig. 1.

 
All the populations studied were consistently separated along the first two dimensions of the MDS scatterplot (Fig. 5). The first dimension clearly separated P. lychnitis and P. crinita subsp. malacitana populations, whereas the second dimension enhanced the separation of P. crinita subsp. crinita populations. The MST showed the role of P. crinita subsp. crinita connecting to P. lychnitis and P. crinita subsp. malacitana through coastal and inland populations, respectively.


Figure 5
View larger version (15K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 5. Minimum spanning tree (MST) superimposed over the scatterplot of the first two dimensions extracted from multidimensional scaling (MDS) analysis. Symbols represent Phlomis crinita subsp. malacitana (squares), P. crinita subsp. crinita (circles) and P. lychnitis (triangles). Dashed lines connect different taxa and ellipses encompass populations of the same taxon. The shaded area in P. crinita subsp. crinita highlights the inland populations (see text).

 


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ACKNOWLEDGEMENTS
 LITERATURE CITED
 
Population genetic diversity and structure in P. lychnitis and P. crinita
Phlomis lychnitis populations consistently showed the highest levels of genetic diversity among the studied taxa. Over all loci, no significant inbreeding was apparent in this taxon, and most populations (83 %) did not depart from HW proportions, revealing the prevalence of outcrossed matings. This taxon had a moderate level of genetic differentiation among populations, and there was no evidence of isolation by distance, and, therefore, a lack of equilibrium between gene flow and drift. Accordingly, populations clustered together in the UPGMA irrespective of their geographic origin. This lack of equilibrium, together with the low degree of variance of the pairwise points seen in the scatterplot (see Fig. 2), suggests high levels of historical gene flow among populations or a rapid range of expansion from a genetically homogenous ancestral source (Templeton et al., 1995; Hutchison and Templeton, 1999). However, previous results of cpDNA haplotype variation in the complex (including P. lychnitis) (Albaladejo et al., 2005) have shown that a clear geographic trend exists (southern Spain vs. eastern Spain), which stresses that nuclear allozymes and plastid markers convey different historical information in this species and argues against the existence of recurrent gene flow, at least via seeds, between regions. It is suggested that sorting of plastid lineages coupled with high levels of gene flow among populations within refuges and range expansion are not mutually exclusive explanations leading to the contrasting pattern of genetic vs. geographic distribution of nuclear and cpDNA variation observed in P. lychnitis. This taxon is not restricted to mountain environments and it seems likely that during the interglacial phases of the Quaternary Period the species expanded its distribution range connecting populations within refuges and colonizing new areas (e.g. Comes and Abbott, 1998). However, localized instances of introgression after secondary contact with P. crinita should also be taken into account to explain the sharing of haplotypes among taxa at some locations (see Albaladejo et al., 2005).

The two subspecies of P. crinita showed similarly low levels of genetic diversity; however, slightly higher allelic richness was detected and heterozygosity was observed in P. crinita subsp. malacitana compared with P. crinita subsp. crinita. Higher rates of introgression in southern Spain compared with eastern Spain (see below) might allow the transfer of low-frequency allozyme alleles from P. lychnitis to the nuclear genome of P. crinita subsp. malacitana (e.g. Pgic Est–2b), which may result in increased allelic diversity and heterozygosity (Rieseberg and Wendel, 1993).

As for P. lychnitis, FIS values in P. crinita subsp. malacitana were consistent with a predominantly outcrossing mating system but, contrary to P. lychnitis, a significant strong genetic population structure (FST = 0·290) was evident in this taxon. Unlike these two taxa, P. crinita subsp. crinita showed a significant excess of homozygotes which, in the absence of selection or assortative mating and given the predominantly xenogamous mating system (R.G.A. and A.A., unpublished results), can be a consequence of strong local genetic structure within some populations or natural variation in breeding systems. In fact, two out of the ten populations studied showed a consistent heterozygote deficit, which may greatly influence the overall pattern detected. Similar to P. crinita subsp. malacitana, high heterogeneity in allele frequencies among populations of P. crinita subsp. crinita was reflected by the markedly strong value of genetic differentiation (FST = 0·419), which was mainly because of strong genetic differences between the inland (CR25 and CR26) and the coastal set of populations (CR27–CR35).

The increase in pairwise FST/(1 – FST) values with geographic distance in both subspecies of P. crinita matched with an isolation-by-distance model. Furthermore, the spread of residuals also increased with geographic distance, indicating an equilibrium situation where, historically, nearby populations were connected by gene flow whereas genetic drift was more influential as spatial isolation increased (Hutchison and Templeton, 1999). Since the time (generations) to achieve this equilibrium is very long (sometimes even longer than the lifetime of a species; Whitlock and McCauley, 1999), it seems that populations of both sub-species of P. crinita, in contrast to P. lychnitis, might have remained relatively undisturbed during the last glacial periods, probably moving their distribution ranges across an altitudinal gradient, similarly to other Iberian endemic taxa (e.g. Gutiérrez-Larena et al., 2002). Restriction to mountain habitats and the long-standing isolation, which should facilitate the divergence process within the complex, might explain why subspecies of P. crinita did not share cpDNA haplotypes (Albaladejo et al., 2005).

Patterns of hybridization and introgression
Results of Bayesian clustering analysis based on multilocus allozyme genotypes underpin the existence of differential patterns of hybridization and introgression in the areas concerned in this study, southern Spain vs. eastern Spain, involving different taxa. Even considering a conservative criterion for detecting hybridization, high rates of hybridization and bidirectional introgression have been detected between P. crinita subsp. malacitana and P. lychnitis in southern Spain (average frequency = 48 %), in accordance with the high morphological variability within populations (Albaladejo et al., 2004) and the free flowing of chromosome rearrangements between taxa (Aparicio and Albaladejo, 2003). However, this average estimate of hybridization in populations in southern Spain is far higher than previously estimated based on morphometry alone (21·8 %; Albaladejo et al., 2004), which is probably because of the high occurrence (134/832) of slightly introgressed individuals having morphology that is indistinguishable from that of pure types. It is noted that multiallelic co-dominant DNA markers such as microsatellites would allow a more accurate classification of genotypic classes in hybrid or contact zones of Phlomis species.

In contrast, only a few hybrid genotypes (average frequency = 4 %) were detected between P. crinita subsp. crinita and P. lychnitis, which agrees with previous estimates of hybridization based on morphology (2·1 %; Albaladejo et al., 2004). Hybrid zones in southern Spain, with a predominance of hybrid genotypes in most populations, fitted to a unimodal structure (Harrison, 1993), whereas a bimodal structure was evident in populations from eastern Spain where most genotypes corresponded to pure types (Harrison and Bogdanowicz, 1997). Given the current distribution range of the studied Phlomis taxa, the scarcity of hybrid plants in eastern Spain reflects the existence of stronger barriers against hybridization between P. crinita subsp. crinita and P. lychnitis, rather than a lack of historical opportunities to hybridize. In a review dealing with the evolutionary consequences of hybridization, Jiggins and Mallet (2000) stated that bimodal hybrid zones, unlike unimodal hybrid zones, are invariably linked to strong assortative mating [e.g. flowering asynchrony (Cruzan and Arnold, 1993) and ethological isolation (Fulton and Hodges, 1999)] or assortative fertilization [e.g. interspecific pollen competition (Rieseberg et al., 1998)], which avoids the formation of hybrid offspring. However, it does not exclude the existence of post-zygotic barriers because selection, endogenous (Barton and Hewitt, 1985) or exogenous (Arnold, 1997), against hybrids in the early stages of development might help discount hybrid plants from the populations, resulting in the overwhelming predominance of parental genotypes.

Genetic differentiation and relationships among taxa
Systematic and evolutionary relationships within the P. lychnitis/P. crinita complex are controversial because analyses of nuclear ribosomal DNA (nrDNA) and cpDNA sequences (Albaladejo et al., 2005) suggest the existence of two sister lineages where lineage sorting and hybridization are central as evolutionary mechanisms. The MST superimposed over the MDS scatterplot and the UPGMA phenogram underlined the role of P. crinita subsp. crinita as a genetic link between P. lychnitis and P. crinita subsp. malacitana. Interestingly, the inland populations of P. crinita subsp. crinita (shaded populations in Fig. 5) were genetically closer to P. lychnitis, whereas the coastal set of P. crinita subsp. crinita populations were closer to P. crinita subsp. malacitana. These results were paralleled by nuclear ITS sequence variation where samples from the inland population CR25 (Elche de la Sierra) shared more parsimony-informative sites with the pure sequence of P. lychnitis (Albaladejo et al., 2005).

This intermediate situation can be explained by the pattern of allelic additivity displayed by P. crinita subsp. crinita (see Table 2). As previously suggested for the patterns of nuclear and plastid DNA variation (Albaladejo et al., 2005), sorting and gene transfer through introgression are likely to be responsible for the sharing of allozyme alleles among taxa during their evolutionary history.

Alternatively, an evolutionary scenario of homoploid hybrid speciation (Gross and Rieseberg, 2005) involving P. crinita subsp. malacitana and P. lychnitis could also be taken into account. It has been shown that these two taxa can hybridize easily, and species with a well-documented diploid hybrid origin (Gallez and Gottlieb, 1982; Wang et al., 1990) are known to exhibit similar patterns of allelic additivity. However, our data do not allow us to test whether or not homoploid hybrid speciation has occurred in this group of taxa.

Concluding remarks
The taxa involved in this study showed contrasting population genetic diversity, population structure and hybridization patterns despite their phylogenetic relatedness. It has been claimed that the climate changes that took place during the Quaternary Period in southern Europe have been of key importance in the evolution of these Mediterranean endemic Phlomis lineages, providing new information about how these changes shaped the genetic structure of plants within glacial refuges. With the present data, it could be envisaged that P. lychnitis was the ancestral lineage in the group; however, whether P. crinita subsp. crinita is a divergent taxon or the result of a homoploid hybrid speciation event is not possible to discern. Consequently, whether P. crinita subsp. crinita has quickly developed strong interspecific prezygotic barriers or whether P. crinita subsp. malacitana has suffered the breakdown of mating barriers through long-term contact with P. lychnitis remains unanswered and will require further research.


   ACKNOWLEDGEMENTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
LITERATURE CITED
 
The authors thank Dr J. Caujapé-Castells, Dr A. Hampe, Dr J. Fuertes Aguilar, Dr G. Nieto Feliner, the handling editor Dr C. Lexer, and the anonymous referees for helpful reviews of the manuscript. We thank the Andalusian Regional Government for permission to collect material in protected areas, and the University of Seville for a predoctoral grant to R.G.A. This study was financed by the Spanish Ministerio de Ciencia y Cultura project BOS 2000-0450 to A.A.


   LITERATURE CITED
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ACKNOWLEDGEMENTS
 LITERATURE CITED
 

    Albaladejo RG, Aparicio A, Silvestre S. Variation patterns in the Phlomis x composita (Lamiaceae) hybrid complex in the Iberian Peninsula. Botanical Journal of the Linnean Society (2004) 145:97–108.[CrossRef][Web of Science]

    Albaladejo RG, Fuertes Aguilar J, Aparicio A, Nieto Feliner G. Contrasting nuclear–plastidial phylogenetic patterns in the recently diverged Iberian Phlomis crinita and P. lychnitis lineages (Lamiaceae). Taxon (2005) 54:987–998.[Web of Science]

    Aparicio A. Fitness components of the hybrid Phlomis x margaritae Aparicio and Silvestre (Lamiaceae). Botanical Journal of the Linnean Society (1997) 124:331–343.[CrossRef][Web of Science]

    Aparicio A, Albaladejo RG. Microsporogenesis and meiotic abnormalities in the hybrid complex of Phlomis composita (Lamiaceae). Botanical Journal of the Linnean Society (2003) 143:79–85.[CrossRef][Web of Science]

    Aparicio A, Albaladejo RG, Porras M, Ceballos G. Isozyme evidence for natural hybridization in Phlomis (Lamiaceae): hybrid origin of the rare P. x margaritae. Annals of Botany (2000) 85:7–12.[Abstract/Free Full Text]

    Arnold ML. Natural hybridization and evolution. (1997) New York: Oxford University Press.

    Ashton GC, Braden AWH. Serum ß-globulin polymorphism in mice. Australian Journal of Biological Sciences (1961) 14:248–254.[Medline]

    Barton NH, Hewitt GM. Analysis of hybrid zones. Annual Review of Ecology and Systematics (1985) 16:113–148.[CrossRef][Web of Science]

    Beaumont M, Barratt EM, Gottelli D, Kitchener AC, Daniels MJ, Pritchard JK, Bruford MW. Genetic diversity and introgression in the Scottish wildcat. Molecular Ecology (2001) 10:319–336.[CrossRef][Medline]

    Bittkau C, Comes HP. Evolutionary processes in a continental island system: molecular phylogeography of the Aegean Nigella arvensis alliance (Ranunculaceae) inferred from chloroplast DNA. Molecular Ecology (2005) 14:4065–4083.[CrossRef][Medline]

    Bohonak AJ. IBD (Isolation By Distance): a program for analyses of isolation by distance. Journal of Heredity (2002) 93:153–154.[Free Full Text]

    Brantjes NBM. Floral mechanics in Phlomis (Lamiaceae). Annals of Botany (1981) 47:279–282.[Free Full Text]

    Carrión JS. Patterns and processes of Late Quaternary environmental change in a montane region of south-western Europe. Quaternary Science Reviews (2002) 21:2047–2066.[CrossRef][Web of Science]

    Clayton JW, Tetriak DN. Amine-citrate buffers for pH control in starch gel electrophoresis. Journal of the Fisheries Research Board of Canada (1972) 29:1169–1172.[Web of Science]

    Comes HP, Abbott RJ. The relative importance of historical events and gene flow on the population structure of a Mediterranean ragwort, Senecio gallicus (Asteraceae). Evolution (1998) 52:355–367.[CrossRef][Web of Science]

    Comes HP, Kadereit JW. The effect of Quaternary climatic changes on plant distribution and evolution. Trends in Plant Sciences (1998) 3:432–438.[CrossRef]

    Cruzan MB, Arnold ML. Ecological and genetic associations in an Iris hybrid zone. Evolution (1993) 47:1432–1445.[CrossRef][Web of Science]

    El Mousadik A, Petit RJ. High level of genetic differentiation for allelic richness among populations of the argan tree [Argania spinosa (L.) Skeels] endemic to Morocco. Theoretical and Applied Genetics (1996) 92:832–839.[CrossRef][Web of Science]

    Evanno G, Regnaut S, Gaudet J. Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Molecular Ecology (2005) 14:2611–2620.[CrossRef][Medline]

    Falush D, Stephens M, Pritchard JK. Inference of population structure using multilocus genotype data: linked loci and correlated allele frequencies. Genetics (2003) 164:1567–1587.[Abstract/Free Full Text]

    Fulton M, Hodges SA. Floral isolation between Aquilegia formosa and. Aquilegia pubescens. Proceedings of the Royal Society B: Biological Sciences (1999) 266:2247–2252.[Abstract/Free Full Text]

    Gallez GP, Gottlieb LD. Genetic evidence for the hybrid origin of the diploid plant. Stephanomeria diegensis. Evolution (1982) 36:1158–1167.

    Goudet J. FSTAT, a program to estimate and test gene diversities and fixation indices. Version 2·9·3. Distributed by the author over the internet from. (2001).

    Gross BL, Rieseberg LH. The ecological genetics of homoploid hybrid speciation. Journal of Heredity (2005) 96:1–12.[Free Full Text]

    Gutiérrez-Larena B, Fuertes Aguilar J, Nieto Feliner G. Glacial-induced altitudinal migrations in Armeria (Plumbaginaceae) inferred from patterns of chloroplast DNA haplotype sharing. Molecular Ecology (2002) 11:1965–1974.[CrossRef][Medline]

    Harrison RG. Hybrids and hybrid zones: historical perspective. In: Hybrid zones and the evolutionary process—Harrison RG, ed. (1993) New York: Oxford University Press, 3–12.

    Harrison RG, Bogdanowicz SM. Patterns of variation and linkage disequilibrium in a field cricket hybrid zone. Evolution (1997) 51:493–505.[CrossRef][Web of Science]

    Hewitt GM. Speciation, hybrid zones and phylogeography – or seeing genes in space and time. Molecular Ecology (2001) 10:537–549.[CrossRef][Medline]

    Hewitt GM. Genetic consequences of climatic oscillations in the Quaternary. Philosophical Transactions of the Royal Society B: Biological Sciences (2004) 359:183–195.[CrossRef]

    Hutchison DW, Templeton AR. Correlation of pairwise genetic and geographic distance measures: inferring the relative influences of gene flow and drift on the distribution of genetic variability. Evolution (1999) 53:1898–1914.[CrossRef][Web of Science]

    Jiggins CD, Mallet J. Bimodal hybrid zones and speciation. Trends in Ecology and Evolution (2000) 15:250–255.[CrossRef]

    Kropf M, Kadereit JW, Comes HP. Late Quaternary distributional stasis in the submediterranean mountain plant Anthyllis montana L. (Fabaceae) inferred from ITS sequences and amplified fragment length polymorphism markers. Molecular Ecology (2002) 11:447–463.[CrossRef][Medline]

    Lewis PO, Zaykin D. Genetic Data Analysis: computer program for the analysis of allelic data. Version 1.0 (d16c). (2001) Free program distributed by the authors over the internet from http://lewis.eeb.uconn.edu/lewishome/software.html.

    Lumaret R, Mir C, Michaud H, Raynal V. Phylogeographical variation of chloroplast DNA in holm oak (Quercus ilex L.). Molecular Ecology (2002) 11:2327–2336.[CrossRef][Medline]

    Nason JD, Ellstrand NC. Estimating the frequencies of genetically distinct classes of individuals in hybridized populations. Journal of Heredity (1993) 84:1–12.[Abstract/Free Full Text]

    Nei M. Genetic distance between populations. American Naturalist (1972) 106:283–292.[CrossRef][Web of Science]

    Olalde M, Herrán A, Espinel S, Goicoechea PG. White oaks phylogeography in the Iberian Peninsula. Forest Ecology and Management (2002) 156:89–102.[CrossRef][Web of Science]

    Petit RJ, El Mousadik A, Pons O. Identifying populations for conservation on the basis of genetics markers. Conservation Biology (1998) 12:844–855.[CrossRef][Web of Science]

    Petit RJ, Aguinagalde I, de Beaulieu JL, Bittkau C, Brewer S, Cheddadi R, et al. Glacial refugia: hotspots but not melting pots of genetic diversity. Science (2003) 300:1563–1565.[CrossRef][Web of Science][Medline]

    Pritchard JK, Stephens M, Donnelly P. Inference of population structure using multilocus genotype data. Genetics (2000) 155:945–959.[Abstract/Free Full Text]

    Rhymer JM, Simberloff D. Extinction by hybridization and introgression. Annual Review of Ecology and Systematics (1996) 27:83–109.[CrossRef][Web of Science]

    Rieseberg LH, Wendel JF. Introgression and its consequences in plants. In: Hybrid zones and the evolutionary process—Harrison RG, ed. (1993) New York: Oxford University Press. 70–109.

    Rieseberg LH, Baird SJE, Desrochers AM. Patterns of mating in wild sunflower hybrid zones. Evolution (1998) 52:713–726.[CrossRef][Web of Science]

    Rosenberg MS. PASSAGE. Pattern Analysis, Spatial Statistics, and Geographic Exegesis. Version 1.1. (2004) Tempe, AZ: Department of Biology, Arizona State University.

    Rousset F. Genetic differentiation and estimation of gene flow from F-statistics under isolation by distance. Genetics (1997) 145:1219–1228.[Abstract]

    Salvador L, Alía R, Agúndez D, Gil L. Genetic variation and migration pathways of maritime pine (Pinus pinaster Ait.) in the Iberian Peninsula. Theoretical and Applied Genetics (2000) 100:89–95.[Medline]

    Sokal RR, Rohlf FJ. Biometry (1981) 2nd edn. New York: Freeman.

    Soltis DE, Haufler CH, Darrow DC, Gastony GE. Starch gel electrophoresis of ferns: a compilation of grinding buffers, gel and electrode buffers, and staining schedules. American Fern Journal (1983) 73:9–27.[CrossRef][Web of Science]

    StatSoft Inc. STATISTICA for Windows release 6.0. (2001) Tulsa: StatSoft Inc.

    Taberlet P, Fumagalli L, Wust-Saucy AG, Cosson JF. Comparative phylogeography and postglacial colonization routes in Europe. Molecular Ecology (1998) 7:453–464.[CrossRef][Medline]

    Templeton AR, Routman E, Phillips CA. Separating population structure from population history: a cladistic analysis of the geographical distribution of mitochondrial DNA haplotypes in the tiger salamander. Ambystoma tigrinum. Genetics (1995) 140:767–782.

    Valbuena-Carabaña M, González-Martínez SC, Hardy OJ, Gil L. Fine-scale spatial genetic structure in mixed oak stands with different levels of hybridization. Molecular Ecology (2007) 16:1207–1219.[CrossRef][Medline]

    Wang XR, Szmidt AE, Lewandowski A, Wang ZR. Evolutionary analysis of Pinus densata (Masters), a putative Tertiary hybrid. 1. Allozyme variation. Theoretical and Applied Genetics (1990) 80:635–640.[Web of Science]

    Weeden NF, Wendel JF. Genetics of plant isozymes. In: Soltis DE, Soltis PS, eds. Isozymes in plant biology. Portland, OR: Dioscorides Press, 46–72 (1989).

    Weir BS, Cockerham CC. Estimating F-statistics for the analysis of population structure. Evolution (1984) 38:1358–1370.[CrossRef][Web of Science]

    Werth CR. Implementing an isozyme laboratory at a field station. Virginia Journal of Sciences (1985) 36:53–76.

    Whitlock MC, McCauley DE. Indirect measures of gene flow and migration: FST != 1/(4Nm + 1). Heredity (1999) 82:117–125.[CrossRef][Web of Science][Medline]


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
100/4/735    most recent
mcm154v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Albaladejo, R. G.
Right arrow Articles by Aparicio, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Albaladejo, R. G.
Right arrow Articles by Aparicio, A.
Agricola
Right arrow Articles by Albaladejo, R. G.
Right arrow Articles by Aparicio, A.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?