Abstract
• Background and Aims Seven related species of Antirrhinum (A. siculum, A. majus, A. latifolium, A. linkianum, A. litigiosum, A. cirrhigherum and A. tortuosum) were studied in order to compare levels of genetic variation and its partitioning in them, and to check relationships between genetic patterns and the reproductive system.
• Methods Eight hundred and fifty-one plants were screened for variability at 13 allozyme loci by means of horizontal starch gel electrophoresis. Parameters of genetic diversity and its partitioning, the inbreeding coefficient as well as an indirect estimate of gene flow based on the equation: Nm = (1 − GST)/4GST, were calculated.
• Key Results Genetic variability in A. siculum was found to be the lowest known in the genus. Mean values of FIT and FIS were mostly positive and not significantly different from zero. Population differentiation (FST) ranged between 6·1 in A. tortuosum and 17·6 in A. linkianum. The inbreeding coefficient within populations ranged between FIS = −0·5 in A. tortuosum and FIS = 1 in A. siculum. Estimates of gene flow ranged between Nm = 15 in A. majus (considered as very high) to Nm = 0·42 in A. siculum (considered as low).
• Conclusions Correlation was found between levels of diversity and differentiation on one hand, and the reproductive system of the studied taxa on the other. Striking differences among species in the inbreeding coefficient (FIS) show different reproductive systems, which mostly support previous reports. Strategies for the conservation of A. siculum are recommended, such as preservation of natural populations as well as ex situ preservation of seeds from different populations.
Keywords: Antirrhinum, reproductive system, inbreeding, out-breeding, self-incompatibility, self-compatibility, populations, genetic variability, genetic structure, gene flow
INTRODUCTION
Plant species show a wide variety of breeding systems (Holsinger, 2000) ranging from obligate inbreeders to obligate outbreeders. In angiosperms, self-incompatibility is a widespread genetic mechanism to prevent self-fertilization due to the evolutionary advantages of out-crossing, and it is widely accepted that self-compatibility is repeatedly derived from self-incompatibility (Barrett and Harder, 1996). Recent data show that the gametophytic self-incompatibility system of Scrophulariaceae has a common phylogenetic origin to that of Solanaceae and Rosaceae (Uenoyama, 1995). Moreover, the independent loss of self-incompatibility has been shown in different groups of species of Lycopersicon (Kondo et al., 2002), supporting former theories on the derived origin of self-compatibility from self-incompatibility.
The reproductive system determines the way genes are transmitted between generations and, consequently, levels of genetic variability are closely related to this. It has long since been recognized that inbreeders show lower levels of genetic variability than outbreeders (Baker, 1959; Hamrick et al., 1979), and that such variability is mainly found among populations in the former while it remains within populations in the latter.
The genus Antirrhinum comprises species with different reproductive systems, as documented by Baur (1919, 1924, 1932), Gruber (1932) and Sherman (1939). Although most species are strictly self-sterile (Baur, 1932; East, 1940), A. siculum is completely self-compatible, A. cirrhigerum and A. linkianum are reported as partially self-compatible (Vieira and Charlesworth, 2002), and A. latifolium, A. litigiosum, A. tortuosum and wild populations of A. majus are self-incompatible. Despite all this, the breakdown of self-incompatibility is documented in A. latifolium and A. majus. (Baur, 1919, 1924, 1932) as well as in A. valentinum (Mateu-Andrés and Segarra-Moragues, 2003b).
Understanding population genetic concepts such as diversity and linkage disequilibrium as well as within-species patterns of variability are important to all biologists (Charlesworth, 2003). Molecular markers are powerful tools to measure selfing rates within populations, and enable the species reproductive system to infer when extended to the species range, and to detect differences among populations. In this paper, data on levels of genetic variability and its partition within and among populations are presented, as well as linkage disequilibrium between pairs of loci, estimates of inbreeding coefficient and gene flow in seven related species showing different reproductive systems: A. siculum, A. majus, A. latifolium, A. linkianum, A. litigiosum, A. cirrhigherum and A. tortuosum, whose delimitation has been studied previously (Mateu and de Paco, 2005).
The aim was to check to what extent the levels of genetic variation of species are affected by differences in the reproductive system in these related taxa. The results will contribute to the knowledge of the reproductive system of these plant species and will help in the management of rare endemic taxa of the genus.
MATERIALS AND METHODS
Taxa studied
All seven species of Antirrhinum L. studied are perennials, and most of these species are unspecialised with populations growing on limestone crevices, roadsides, walls and roofs, while A. cirrhigherum grows on sands.
Antirrhinum siculum Miller and A. tortuosum Bosc. are widespread taxa, distributed throughout a vast area in the Mediterranean basin, while A. latifolium Miller, A. litigiosum Pau, A. cirrhigherum Welw ex Rothm., A. majus L. and A. linkianum Boiss. et Reut. have small ranges.
Plant material
A total of 851 individuals from 49 natural populations were studied. Individuals were grown in a greenhouse from seeds sampled in different plants. Sampled individuals were between 10 % and 85·7 % of the total number of mature plants growing in each population. Sampled populations were distributed through the species range. Data on sampled populations are provided in Table 1, and the species distribution used is that of Mateu-Andrés and de Paco (2005).
Table 1.
Populations of the taxa studied
| Species | Population code | Country | Province | Locality | Population size | % of population studied |
|---|---|---|---|---|---|---|
| A. siculum (AS) | AS1 | It | Sicily | Palermo | >100 | 17 |
| AD2 | It | Sicily | Siracusa | 100–150 | 11·6–16 | |
| AS3 | It | Sicily | Nicolosi | 37 | 14 | |
| AS4 | It | Sicily | Taormina | 12 | 41·6 | |
| AD5 | It | Sicily | Messina | 12 | 33·3 | |
| AS6 | It | Napoles | Napoles | – | – | |
| AS7 | Is | Haifa | Haifa | – | – | |
| AS8 | It | Grosseto | Santo Stefano | – | – | |
| A. majus (AM) | AM2 | Hs | Gerona | Queralbs | 11 | 72·7 |
| AM3 | Fr | Pyréneés Orientales | Colliure | 50–60 | 11·6–14 | |
| AM4 | Hs | Gerona | Banyolas | <100 | 20 | |
| AM5 | Hs | Gerona | Blanes | <100 | 15 | |
| A. tortuosum (AT) | AT1 | It | Sicily | Agrigento | <40 | 17·5 |
| AT2 | Hs | Sevilla | Alcalá de Guadaira | <100? | 33 | |
| AT4 | Hs | Córdoba | Cerro Muriano | 30 | 76·6 | |
| AT5 | Hs | Córdoba | de Posadas a Espiel | <50? | 58 | |
| AT8 | Hs | Córdoba | Lucena | <60? | 45 | |
| AT9 | It | Lacio | Roma | – | – | |
| AT10 | It | Latina | Norma | – | – | |
| AA1 | Hs | Cádiz | Zahara de la Sierra | <60 | 50 | |
| AA2 | Hs | Cádiz | Peñón de la Parra | <50 | 60 | |
| AA3 | Hs | Cádiz | Grazalema | <40 | 45 | |
| AA4 | Hs | Cádiz | Benaocaz | <50 | 50 | |
| AA6 | Hs | Málaga | Torcal de Antequera | <100 | 30 | |
| AA7 | Hs | Córdoba | de Lucena a Cabra | <100 | 19 | |
| AA8 | Hs | Córdoba | Carcabuey | >60 | 50 | |
| AA9 | Hs | Jaén | Alcaudete | 100–150 | 20–30 | |
| A. latifolium (AL) | ALI2 | Fr | Pyréneés Orientales | Villefranche de Conflent | 44 | 15·9 |
| ALI3 | Fr | Pyréneés Orientales | Usson-les-Bains | <20 | 15 | |
| ALL2 | Hs | Gerona | Tossas | <50 | 14 | |
| ALL3 | Fr | Provence | La Fontaine de Vaucluse | 8 | 37·5 | |
| ALL4 | Fr | Massif Central | Beaulieu | <30 | 36·6 | |
| ALL5 | Fr | Alpes Maritimes | Auron | 7 | 85·7 | |
| ALL6 | Fr | Alpes Maritimes | Valdeblore-Rimples | 27 | 29·6 | |
| ALL7 | Fr | Alpes Maritimes | Menton | <40 | 22·5 | |
| A. cirrhigerum (AC) | AC5 | Lu | Beira Litoral | San Pedro de Muel | >100 | 30 |
| AC6 | Lu | Beira Litoral | Figueira da Foz | 100–150 | 18–27 | |
| AC7 | Lu | Douro Litoral | Aveiro | <35 | 71·4 | |
| A. linkianum (ALK) | ALK1 | Lu | Estremadura | Serra da Arrabida | 120–150 | 20·6 |
| ALK2 | Lu | Estremadura | Trafaria | 100–120 | 10–12 | |
| ALK3 | Lu | Ribatejo | Almada | 150 | 15 | |
| ALK6 | Lu | Estremadura | Livramento | 32 | 50 | |
| ALK7 | Lu | Beira Litoral | Coimbra | 23 | 52·2 | |
| ALK8 | Lu | Estremadura | Pernes | 42 | 33·3 | |
| A. litigiosum (ALI) | ALIT6 | Hs | Castellón | Begis | <30 | 80 |
| ALIT9 | Hs | Tarragona | Mas de Barberans | <30 | 60 | |
| ALIT17 | Hs | Tarragona | Sant Vicenç de Calders | <30 | 63·3 | |
| ALIT19 | Hs | Valencia | Lliria | <80 | 37·5 | |
| ALIT20 | Hs | Valencia | Bugarra | <40 | 75 |
Fr, France; Hs, Spain; Is, Israel; It, Italy; Lu, Portugal.
Population codes are the same as in Mateu and de Paco (2005).
The sign ‘–’ in population size and % populations studied indicates that data were not available for these populations.
Electrophoresis and analysis
Electrophoresis was carried out on horizontal 10 % starch gels. The extracting buffer consisted of 0·2 m Tris–HCl. pH 7·5, 2 mm EDTA, 0·12 m Na2S2O5, 1 m Cl2Mg, 40 mg mL–1 (w/v) PVP, 4 µm mL–1 mercaptoethanol. The material employed for the extracts consisted of young leaves from plants grown in the greenhouse, which were absorbed onto 3-mm wicks of Whatman chromatography paper.
Nine enzyme systems were assayed: aconitase (ACO; EC 4.2.1.3), aspartate aminotransferase (AAT; EC 2.6.1.1), diaphorase (DIA; EC 1.6.99), isocitrate dehydrogenase (IDH; EC 1.1.1.42), malate dehydrogenase (MDH; EC 1.1.1.37), menadione reductase (MNR; EC 1.6.99), phosphoglucoisomerase (PGI; EC 5.3.1.9), phosphoglucomutase (PGM; EC 5.4.2.2) and triosephosphate isomerase (TPI; EC 5.3.1.1). All of them presented banding patterns. AAT and IDH could not be scored due to inconsistent banding patterns. The electrophoretic buffer system II of Wendel and Weeden (1989) was employed to resolve IDH and MDH; system VI for DIA, PGI, PGM, and TPI; and system VII was used for AAT, ACO and MNR. All staining methods followed Wendel and Weeden (1989), and AAT was modified following H. C. Prentice (pers. comm.). As reported in previous studies in Antirrhineae, both in other genera (Elisens and Crawford, 1988; Elisens, 1992; Elisens and Nelson, 1993) and in Antirrhinum (Mateu-Andrés, 1999; Mateu-Andrés and Segarra-Moragues, 2000, 2003a,b), PGI1 and TPI1 showed differences in band thickness, which were interpreted as duplicated comigrant loci. Thus, they were not scored.
Analysis of data
Diversity parameters (A, mean number of alleles per locus; P95 P99, proportion of polymorphic loci at 95 % and 99 % criteria, respectively; Ho, observed heterozygosity; He, expected heterozygosity), tests for significant deviations from Hardy–Weinberg expectations and Nei's genetic distance measures within taxa among populations were calculated for each population using Biosys-1 (Swofford and Selander, 1989). Outcrossing rates within populations were estimated through FIS. Partitioning of genetic variation was performed with F statistics (Wright, 1965, 1978), estimated according to the equations of Weir and Cockerham (1984). The FST statistics measure the divergence in allele frequencies among populations, whereas FIS and FIT measure heterozygote excess (<0) and deficit (>0), in relation to the Hardy–Weinberg expectation in local populations and the total set of populations, respectively. The population genetic structure and FIS were calculated using FSTAT 2.9.3 (Goudet, 2001). Sequential Bonferroni correction (Rice, 1989) was applied to the results of tests on the Hardy–Weinberg expectations. Total genetic diversity HT was computed for each species to make comparisons easier with others studied previously.
To compare whether parameters of genetic variability were significantly different among taxa, a one-way ANOVA test was applied to P95 P99 and A. The spatial pattern of genetic variation within taxa was illustrated by a cluster analysis, and the Mantel test was performed comparing the pairwise Nei's genetic distances against the geographical distances between populations. Both were based on Nei's genetic distance (Nei, 1972) and performed with the Ntsys 2.0 (Rohlf, 1998) program. An indirect estimate of gene flow was taken based on the equation: Nm = (1 − GST)/4GST, where Nm is the number of migrants per generation (Wright, 1931).
A test for genotypic disequilibrium between pairs of loci was performed using FSTAT. Due to the fact that multiple tests were involved, the sequential Bonferroni correction was applied to test for significance (Rice, 1989).
RESULTS
Seven out of the nine enzymatic systems studied were interpreted, giving a total of 13 putative loci and 39 alleles. Only MNR1 out of the 13 scored loci was monomorphic. PGI1 and TPI1 had duplicated commigrant loci, also reported for other species of Antirrhineae (Elisens and Crawford, 1988; Elisens, 1992; Elisens and Nelson, 1993) and Antirrhinum (Mateu-Andrés, 1999; Mateu-Andrés and Segarra-Moragues, 2000), so they were not scored. The number of alleles was 10 in A. siculum, 18 in A. majus, 21 in A. tortuosum, 20 in A. latifolium, 17 in A. cirrhigerum, 20 in A. linkianum and 21 in A. litigiosum. Allele frequencies are shown in Mateu and de Paco (2005). No linkage disequilibrium was found between pairs of loci after sequential Bonferroni correction.
Table 2 shows the diversity parameters, all of which are in the range previously reported for other species of Antirrhinum (Mateu-Andrés, 1999; Mateu-Andrés and Segarra, 2000, 2003a, b), and are similar to those obtained by Hamrick and Godt (1996) for Scrophulariaceae. The proportion of polymorphic loci ranged widely among taxa, varying between 0–77 % and 0–84·6 for 95 % and 99 % criteria, respectively. The mean number of alleles per locus among taxa ranged between 1 and 2·1. Ho ranged between 0 and 0·34, and He ranged similarly between 0 and 0·31, among taxa. Among the populations showing variability in at least one locus, FIS values ranged between −0·004 and 0·904 with different levels of signification. Antirrhinum siculum, showed the lowest levels of variability where all monomorphic loci in most populations were only AD2 and AS7 polymorphic, both at the P95 and P99 levels in the former case, and only at P99 in the second; the mean number of alleles per locus was 1·15 and 1·08, respectively. Correspondingly, FIS = 1 was found in six out of eight populations and FIS = 0·904 in one of the populations showing variability, while it was lower and non-significant in the other. The ANOVA tests performed to compare the genetic variability parameters revealed that A (P = 0·000, F = 12·936), P95 (P = 0·000, F = 15·533) and P99 (P = 0·000, F = 16·248) were significantly different among taxa when all populations were considered in the analysis. The Tukey test performed to check differences between pairs of species revealed that all three parameters of diversity were significantly different when A. siculum was compared with any other species (P = 0·000), and was lower (P = 0·004 for A and P = 0·001 for P95) when A. siculum was compared with A. cirrhigerum. Antirrhinum litigiosum had a significantly higher variation than A. latifolium in P95 (P = 0·010) P99 (P = 0·017), A (P = 0·014) and A. linkianum in P95 and P99 (P = 0·011 and P = 0·047).
Table 2.
Parameters of genetic variability
| Species | Pop. | P95 | P99 | Ae | Ho | He | FIS |
|---|---|---|---|---|---|---|---|
| A. cirrhigerum | AC5 | 46·1 | 61·5 | 1·69 ± 0·17 | 0·20 ± 0·06 | 0·21 ± 0·06 | 0·057** |
| AC6 | 38·5 | 38·5 | 1·38 ± 0·14 | 0·21 ± 0·09 | 0·20 ± 0·07 | –0·054* | |
| AC7 | 30·8 | 38·5 | 1·38 ± 0·14 | 0·07 ± 0·03 | 0·11 ± 0·05 | 0·387*** | |
| A. latifolium | ALI2 | 53·8 | 53·8 | 1·62 ± 0·18 | 0·19 ± 0·06 | 0·20 ± 0·06 | –0·005ns |
| ALI3 | 38·5 | 38·5 | 1·38 ± 0·14 | 0·08 ± 0·04 | 0·16 ± 0·06 | 0·571*** | |
| ALL2 | 46·1 | 46·1 | 1·54 ± 0·18 | 0·10 ± 0·04 | 0·16 ± 0·05 | 0·393*** | |
| ALL3 | 30·8 | 30·8 | 1·31 ± 0·13 | 0·10 ± 0·06 | 0·13 ± 0·06 | 0·273** | |
| ALL4 | 38·5 | 46·1 | 1·54 ± 0·18 | 0·14 ± 0·06 | 0·14 ± 0·05 | –0·020* | |
| ALL5 | 23·1 | 23·1 | 1·23 ± 0·12 | 0·05 ± 0·03 | 0·05 ± 0·03 | –0·053ns | |
| ALL6 | 30·8 | 30·8 | 1·31 ± 0·13 | 0·07 ± 0·04 | 0·07 ± 0·04 | 0·125** | |
| ALL7 | 46·1 | 46·1 | 1·54 ± 0·18 | 0·13 ± 0·06 | 0·12 ± 0·05 | –0·091* | |
| A. linkianum | ALK1 | 53·8 | 61·5 | 1·85 ± 0·25 | 0·19 ± 0·06 | 0·20 ± 0·06 | 0·071** |
| ALK2 | 30·8 | 30·8 | 1·38 ± 0·18 | 0·09 ± 0·04 | 0·08 ± 0·03 | –0·126** | |
| ALK3 | 38·5 | 38·5 | 1·46 ± 0·18 | 0·15 ± 0·06 | 0·12 ± 0·05 | –0·230*** | |
| ALK6 | 30·8 | 38·5 | 1·46 ± 0·18 | 0·12 ± 0·06 | 0·11 ± 0·05 | –0·029* | |
| ALK7 | 46·1 | 46·1 | 1·46 ± 0·14 | 0·20 ± 0·07 | 0·18 ± 0·06 | –0·108** | |
| ALK8 | 23·1 | 30·8 | 1·46 ± 0·22 | 0·04 ± 0·03 | 0·13 ± 0·06 | 0·669*** | |
| A. litigiosum | ALIT6 | 77 | 77 | 2·00 ± 0·20 | 0·34 ± 0·07 | 0·31 ± 0·06 | –0·102*** |
| ALIT9 | 61·5 | 61·5 | 1·70 ± 0·17 | 0·30 ± 0·07 | 0·24 ± 0·06 | –0·193*** | |
| ALIT17 | 77 | 84·6 | 2·10 ± 0·18 | 0·27 ± 0·08 | 0·27 ± 0·06 | –0·004ns | |
| ALIT19 | 61·5 | 61·5 | 1·62 ± 0·14 | 0·27 ± 0·08 | 0·25 ± 0·06 | –0·140** | |
| ALIT20 | 46·1 | 46·1 | 1·46 ± 0·14 | 0·18 ± 0·06 | 0·18 ± 0·06 | 0·003ns | |
| A. majus | AM2 | 61·5 | 61·5 | 1·62 ± 0·14 | 0·20 ± 0·06 | 0·19 ± 0·05 | –0·103** |
| AM3 | 53·8 | 53·8 | 1·54 ± 0·14 | 0·19 ± 0·07 | 0·19 ± 0·06 | 0·019ns | |
| AM4 | 61·5 | 69·2 | 1·69 ± 0·13 | 0·19 ± 0·05 | 0·20 ± 0·05 | 0·013ns | |
| AM5 | 46·1 | 46·1 | 1·54 ± 0·18 | 0·18 ± 0·07 | 0·18 ± 0·06 | –0·018ns | |
| A. siculum | AS1 | 0 | 0 | 1 | 0 | 0 | 1 |
| AD2 | 15·4 | 15·4 | 1·15 ± 0·10 | 0·005 ± 0·005 | 0·005 ± 0·04 | 0·904*** | |
| AS3 | 0 | 0 | 1 | 0 | 0 | 1 | |
| AS4 | 0 | 0 | 1 | 0 | 0 | 1 | |
| AD5 | 0 | 0 | 1 | 0 | 0 | 1 | |
| AS6 | 0 | 0 | 1 | 0 | 0 | 1 | |
| AS7 | 0 | 7·7 | 1·08 ± 0·08 | 0·005 ± 0·005 | 0·005 ± 0·005 | –0·018ns | |
| AS8 | 0 | 0 | 1 | 0 | 0 | 1 | |
| A. tortuosum | AT1 | 46·1 | 46·1 | 1·46 ± 0·14 | 0·15 ± 0·08 | 0·15 ± 0·05 | –0·050ns |
| AT2 | 38·5 | 46·1 | 1·62 ± 0·21 | 0·14 ± 0·06 | 0·15 ± 0·05 | 0·032* | |
| AT4 | 53·8 | 53·8 | 1·70 ± 0·21 | 0·22 ± 0·07 | 0·20 ± 0·06 | –0·099** | |
| AT5 | 30·5 | 46·1 | 1·54 ± 0·18 | 0·08 ± 0·04 | 0·15 ± 0·06 | 0·455*** | |
| AT8 | 53·8 | 61·5 | 1·69 ± 0·17 | 0·14 ± 0·05 | 0·21 ± 0·06 | 0·354*** | |
| AT9 | 23·1 | 23·1 | 1·31 ± 0·17 | 0·05 ± 0·03 | 0·07 ± 0·04 | 0·205*** | |
| AT10 | 15·4 | 15·4 | 1·15 ± 0·10 | 0·11 ± 0·08 | 0·08 ± 0·05 | –0·500*** | |
| AA1 | 69·2 | 69·2 | 1·69 ± 0·13 | 0·33 ± 0·08 | 0·28 ± 0·06 | –0·192*** | |
| AA2 | 53·8 | 61·5 | 1·77 ± 0·20 | 0·26 ± 0·07 | 0·22 ± 0·06 | –0·149*** | |
| AA3 | 53·8 | 69·2 | 1·77 ± 0·17 | 0·20 ± 0·06 | 0·20 ± 0·06 | 0·005ns | |
| AA4 | 61·5 | 61·5 | 1·69 ± 0·17 | 0·16 ± 0·06 | 0·16 ± 0·05 | 0·045* | |
| AA6 | 69·2 | 76·9 | 1·77 ± 0·12 | 0·23 ± 0·07 | 0·21 ± 0·05 | –0·128*** | |
| AA7 | 69·2 | 69·2 | 1·77 ± 0·17 | 0·25 ± 0·07 | 0·22 ± 0·05 | –0·149*** | |
| AA8 | 46·1 | 53·8 | 1·54 ± 0·14 | 0·20 ± 0·07 | 0·18 ± 0·06 | –0·066** | |
| AA9 | 61·5 | 77 | 2·00 ± 0·20 | 0·22 ± 0·06 | 0·20 ± 0·05 | –0·119*** |
P99, P95, proportion of polymorphic loci at 99 % and 95 % criteria, respectively; Ae, mean number of alleles per locus; Ho, mean observed heterozygosity per locus; He, mean expected heterozygosity per locus; FIS, Wright's inbreeding coefficient.
Significance level of FIS values: *P < 0·05; **P < 0·01; ***P < 0·001; ns, not significant.
Significant deviations from the Hardy–Weinberg expectations were detected for ten loci in different populations (Table 3). In most cases, deviations were due to a deficiency of heterozygotes, but deviations were due to an excess of heterozygotes in MDH1 (population ALIT17), DIA1 (population AA2), MNR2 (population AA4) and MDH3 (populations AA6 and AA8).
Table 3.
Significant deviations from Hardy–Weinberg expectations
| Species | Population | Locus | P | Heterozygotes |
|---|---|---|---|---|
| A. cirrhigerum | AC5 | DIA1 | 0·000 | D |
| AC7 | ACO2 | 0·006 | D | |
| DIA1 | 0·000 | D | ||
| A. latifolium | ALI3 | MDH3 | 0·021 | D |
| ALL3 | DIA1 | 0·021 | D | |
| ALL4 | DIA1 | 0·012 | D | |
| ALL6 | DIA1 | 0·000 | D | |
| ALL7 | DIA1 | 0·001 | D | |
| A. linkianum | ALK1 | PGI2 | 0·004 | D |
| ALK6 | PGI2 | 0·004 | D | |
| ALK8 | DIA3 | 0·01 | D | |
| PGI2 | 0·000 | D | ||
| A. litigiosum | ALIT6 | DIA3 | 0·000 | D |
| ALIT17 | MDH1 | 0·000 | E | |
| A. majus | AM5 | PGI2 | 0·000 | D |
| A. siculum | AS2 | PGI2 | 0·000 | D |
| PGM2 | 0·000 | D | ||
| A. tortuosum | AT1 | PGI2 | 0·000 | D |
| AT2 | PGI2 | 0·000 | D | |
| AT4 | DIA1 | 0·000 | D | |
| PGI2 | 0·000 | D | ||
| AT5 | DIA1 | 0·000 | D | |
| AT6 | ACO1 | 0·001 | D | |
| DIA1 | 0·007 | D | ||
| AT9 | PGI2 | 0·000 | D | |
| AA2 | DIA1 | 0·000 | E | |
| AA3 | PGM2 | 0·000 | D | |
| AA4 | ACO1 | 0·000 | D | |
| MNR2 | 0·007 | E | ||
| AA6 | MDH3 | 0·001 | E | |
| AA8 | DIA1 | 0·005 | D | |
| MDH3 | 0·000 | E | ||
| AA9 | ACO2 | 0·001 | D |
D, deficiency; E, excess.
FIS values were seen to be high (over 0·250) and positive in eight populations, positive with moderate or low values in another four populations, high and negative in only one population, and moderate or low and negative in nine. FIS values were highly significant (P < 0·001) in 17 populations. Throughout species, FIS was very high (0·904) in one of the two populations of A. siculum, showing genetic variability. Antirrhinum litigiosum and A. majus were in equilibrium or showed low values of FIS, one (AC7) out of three populations of A. cirrhigerum showed high and positive FIS. In A. linkianum, two populations showed positive FIS, while this was negative in the other four. FIS values were high and positive in three (ALI2, ALL2, ALL3) out of eight populations of A. latifolium, and they were moderate and positive in one (ALL6). In A. tortuosum, FIS estimates were positive in six out of eight populations of A. tortuosum, they were high in two of them (AT5 and AT6), while one population (AT10) showed high negative values, and moderate or low ones in another 5.
Total genetic diversity (i.e. diversity averaged across all the loci studied, HT) ranged between 0·03 in A. siculum and 0·28 in A. litigiosum (Table 4). Mean values of FIT and FIS for the overall populations of each species were mostly positive and not significantly different from zero, despite departures from zero for some individual loci, indicating that both total and local populations were in Hardy–Weinberg equilibrium. All mean values of FST per species were positive and not significantly different from zero. FST values ranged between 0·061 and 0·270, indicating a low proportion of differentiation among populations in A. tortuosum (6·1 %), A. latifolium (9·3 %) and A. majus, a moderate proportion in A. cirrhigerum (13·2 %), A. litigiosum (14·8 %) and A. linkianum (17·6 %), and a high proportion in A. siculum (27 %).
Table 4.
Partitioning of the total genetic variability
| A. cirrhigerum | A. latifolium | A. linkianum | A. litigiosum | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Loci | HT | FIT | FST | FIS | HT | FIT | FST | FIS | HT | FIT | FST | FIS | HT | FIT | FST | FIS |
| Aco1 | 0·25 | 0·507 | 0·383* | 0·201 | 0·00 | 0·463 | −0·057* | 0·492 | 0·25 | 0·229** | 0·054* | 0·185 | 0·37 | 0·456 | 0·431** | 0·044 |
| Aco2 | 0·48 | 0·154* | 0·073* | 0·088 | 0·06 | −0·020** | 0·015** | −0·035 | 0·14 | 0·421 | 0·471* | −0·096 | 0·38 | 0·129 | 0·164* | −0·043 |
| Dia1 | 0·52 | 0·521* | 0·041 | 0·500 | 0·36 | 0·213* | 0·137* | 0·088 | 0·48 | 0·190* | 0·095* | 0·104 | 0·09 | 0·011* | 0·227* | −0·279 |
| Dia3 | 0·00 | 1 | – | – | 0·07 | 0·466 | 0·295* | 0·242 | 0·15 | 0·038 | 0·221 | −0·236 | 0·07 | 0·330** | 0·049* | 0·296* |
| Mdh1 | 0·49 | −0·239 | 0·120 | −0·408 | 0·15 | 0·280* | 0·056 | 0·237 | 0·25 | −0·023* | 0·191* | −0·264 | 0·48 | −0·397* | 0·023 | −0·429 |
| Mdh2 | 0·13 | 0·493* | 0·163 | 0·394 | 0·23 | 1 | – | – | 0·13 | 1 | – | – | 0·50 | 0·216* | 0·057 | 0·168 |
| Mdh3 | 0·02 | −0·002*** | −0·007*** | 0·005* | 0·26 | −0·104* | 0·051* | −0·164 | 0·00 | 1 | – | – | 0·23 | −0·072* | 0·169 | −0·290 |
| Mnr1 | 0·00 | 1 | – | – | 0·00 | −0·065** | 0·046* | −0·117 | 0·00 | −0·015* | 0·248* | −0·349 | 0·00 | 1 | – | – |
| Mnr2 | 0·39 | 0·225 | 0·281* | −0·078 | 0·21 | 1 | – | – | 0·12 | 0·348* | 0·086 | 0·287 | 0·37 | 0·234 | 0·129* | 0·121 |
| Pgi2 | 0·19 | 0·066* | 0·145* | −0·093 | 0·15 | −0·045* | 0·191* | −0·291 | 0·49 | 1 | – | – | 0·16 | 0·003** | 0·006*** | −0·003 |
| Pgm1 | 0·00 | 1 | – | – | 0·00 | −0·027* | 0·104 | −0·147 | 0·00 | −0·014** | 0·045** | −0·061 | 0·00 | 1 | – | – |
| Pgm2 | 0·01 | −0·002*** | −0·005*** | 0·003* | 0·11 | 1 | – | – | 0·03 | 0·407* | 0·179* | 0·278 | 0·58 | −0·022* | 0·078 | 0·16 |
| Tpi2 | 0·00 | 1 | – | – | 0·16 | 1 | – | – | 0·08 | 1 | – | – | 0·48 | 0·281 | 0·294** | −0·018 |
| Mean | 0·19 | 0·191 | 0·132 | 0·068 | 0·14 | 0·129 | 0·093 | 0·034 | 0·16 | 0·175 | 0·176 | −0·017 | 0·28 | 0·106 | 0·148 | −0·030 |
| (σ) | 0·06 | 0·27 | 0·13 | 0·27 | 0·03 | 0·23 | 0·10 | 0·25 | 0·05 | 0·18 | 0·13 | 0·24 | 0·05 | 0·24 | 0·12 | 0·23 |
| A. majus | A. siculum | A. tortuosum | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Loci | HT | FIT | FST | FIS | HT | FIT | FST | FIS | HT | FIT | FST | FIS |
| Aco1 | 0·04 | −0·013*** | −0·009*** | −0·004* | 0·00 | 1 | – | – | 0·25 | 1 | – | – |
| Aco2 | 0·14 | 0·241*** | −0·003* | 0·243 | 0·00 | 1 | – | – | 0·41 | 0·378 | 0·205 | 0·219 |
| Dia1 | 0·00 | −0·188* | 0·097 | −0·315 | 0·00 | 1 | – | – | 0·38 | 0·349 | 0·179 | 0·207 |
| Dia3 | 0·41 | 0·268 | 0·120 | 0·169 | 0·00 | 1 | – | – | 0·01 | −0·002*** | 0·024*** | −0·026** |
| Mdh1 | 0·38 | −0·154** | −0·022 | −0·130 | 0·00 | 1 | – | – | 0·27 | −0·083* | 0·098 | −0·201 |
| Mdh2 | 0·45 | 0·400* | 0·132* | 0·309 | 0·00 | 1 | – | – | 0·32 | −0·043* | 0·095* | −0·152 |
| Mdh3 | 0·42 | 0·713 | 0·211 | 0·636 | 0·00 | 1 | – | – | 0·42 | −0·327* | 0·078 | −0·440 |
| Mnr1 | 0·00 | −0·056* | 0·068* | −0·134 | 0·00 | 1 | – | – | 0·00 | 1 | – | – |
| Mnr2 | 0·25 | 0·888* | 0·205 | 0·859* | 0·01 | −0·005*** | −0·006*** | 0·001* | 0·21 | −0·133* | 0·127 | −0·298 |
| Pgi2 | 0·18 | 0·684 | 0·178 | 0·615 | 0·01 | 1*** | −0·008* | 1*** | 0·13 | 0·375** | 0·060* | 0·335 |
| Pgm1 | 0·00 | 0·569* | 0·121 | 0·510 | 0·00 | 1 | – | – | 0·01 | −0·002*** | −0·003*** | 0·001** |
| Pgm2 | 0·109 | 1 | – | – | 0·33 | 0·979 | 0·823 | 0·883** | 0·20 | 0·140* | 0·095** | 0·050 |
| Tpi2 | 0·08 | 1 | – | – | 0·00 | 1 | – | – | 0·02 | 0·008** | 0·133** | −0·144** |
| Mean | 0·193 | 0·666 | 0·100 | 0·250 | 0·03 | 0·658 | 0·270 | 0·628* | 0·20 | 0·227 | 0·061 | 0·235 |
| (σ) | 0·05 | 0·38 | 0·08 | 0·36 | 0·02 | 0·57 | 0·48 | 0·55 | 0·04 | 0·06 | 0·10 | −0·04 |
HT, total genetic diversity and F statistics: FIT and FIS measure of heterozygote excess (<0) or deficit (>0) relative to Hardy–Weinberg expectations for the total and local population, respectively; FST, coefficient of differentiation among-populations.
Significance levels: *P < 0·05; **P < 0·01; ***P < 0·001.
Estimates of the number of migrants (Nm) gave high values for A. majus and relatively high values for A. cirrhigerum, A. latifolium, A. linkianum, A. litigiosum and A. tortuosum (Table 5), indicating high levels of gene flow among populations. However, low values were estimated for A. siculum. It is noted that the estimation of Nm from FST rests on numerous assumptions regarding meta-population dynamics which may not be met in all of the species studied, thus Nm values must be interpreted with caution (Whitlock, 1992). Mantel tests detected a relationship among the geographic and genetic distance in A. litigiosum (r = 0·76, P = 0·99), although it was non-significant. It was weak and non-significant for A. latifolium (r = 0·44, P = 0·99), A. majus (r = 0·44, P = 0·80) and A. tortuosum (r = 16, P = 0·83), weak and negative (r = −10, P = 0·38) for A. siculum, and negligible for A. cirrhigerum (r = 0·05) and A. linkianum (r = 0·004, P = 0·50). Thus, no isolation by distance was detected in these six species. These results show that a low number of migrants is sufficient to genetically homogenize populations, according to Ellstrand and Elam (1993).
Table 5.
Estimates of gene flow (Nm) and correlation coefficients obtained in Mantel tests, within taxa
| Species | Nm | Mantel test |
|---|---|---|
| A. cirrhigerum | 2·2 | r = 0·05 |
| A. latifolium | 4·2 | r = 0·44; P = 0·99 |
| A. linkianum | 1·3 | r = 0·004; P = 0·50 |
| A. litigiosum | 1·8 | r = 0·76; P = 0·99 |
| A. majus | 15 | r = 0·44; P = 0·80 |
| A. siculum | 0·42 | r = −0·10; P = 0·38 |
| A. tortuosum | 1·7 | r = 0·16; P = 0·83 |
There is no estimate of P in A. cirrhigerum due to the low number of populations studied.
DISCUSSION
In agreement with previously reported results for other species of Antirrhinum, the species studied show striking differences in parameters of genetic diversity, inbreeding coefficient (Table 2), total diversity levels (HT) and its partition within and among populations (Table 4).
The total genetic diversity for A. siculum is 0·03, the lowest among the studied species of Antirrhinum (Mateu-Andrés, 1999; Mateu-Andrés and Segarra, 2000, 2003a,b; Mateu-Andrés, 2004) as well as among species of Antirrhineae (Elisens and Crawford, 1988; Elisens, 1992; Elisens and Nelson, 1993). Those for A. litigiosum (0·28) and A. barrelieri (0·25), can be considered as high, and are among those known for Antirrhinum, a genus for which HT = 0·50 levels had been reported in A. microphyllum (Mateu-Andrés, 1999) and HT = 0·30 in A. pulverulenthum (Mateu-Andrés, 2004), two species with similar ranges to those of A. majus and A. linkianum.
Among the species studied, A. siculum is self-compatible, A. cirrhigerum and A. linkianum had been reported as showing variable levels of compatibility among populations (Vieira and Charlesworth, 2002; Zwettler et al., 2002), and A. latifolium, A. litigiosum, A. tortuosum and wild populations of A. majus are considered as self-incompatible, although a breakdown of the self-incompatibility system has been reported from the second year of cultivation onwards in the former species (Baur, 1919, 1924, 1932). The present results coincide with previous reports for A. siculum, although the low FIS value indicates out-crossing among individuals in one population (AS7). In A. litigiosum, A. majus and A. tortuosum, FIS values are either in Hardy–Weinberg equilibrium or are low and negative, in accordance with the obligate out-crossing in self-incompatible plants. Meanwhile, high and positive FIS values are found in A. tortuosum (AT5 and AT6), as were high and negative values (AT10), indicating strong deviations from Hardy–Weinberg equilibrium both upwards and downwards. In A. latifolium, high and positive FIS values in three (ALI2, ALL2, ALL3) out of eight populations indicate different levels of inbreeding in some populations, and suggest that Baur's (1919, 1924, 1932) results in cultivated plants could also be extended to wild plants. The inbreeding coefficient in A. linkianum shows that the species is out-crossing, although the selfing rate is high in at least one of the populations studied, in agreement with that reported by Vieira and Charlesworth (2002). In A. cirrhigerum, FIS values are low in two populations, indicating out-crossing among plants into populations, while high and positive values show a high level of inbreeding in a third population (AC7). Interestingly, populations AC6 and AC7 had been reported as being largely self-compatible, while AC5 is considered as largely self-incompatible (Vieira and Charlesworth, 2002; Zwettler et al., 2002). The present results agree with those of Vieira and Charlesworth (2002) and Zwettler et al. (2002) in both AC5 and AC7, while plants behave as out-crossers in AC6 despite being self-compatible.
Following Loveless and Hamrick (1984) and Hamrick et al. (1979), out-crossing reduces the effect of genetic drift and increases the migration rate, thus increasing the specific levels of variation and preventing divergence among populations. In self-pollinated plants on the other hand, genetic drift increases and species present high differences among sub-populations. In agreement with this, genetic differentiation among populations was high in A. siculum, the only self-compatible taxon, whereas it was low or moderate in the other species studied.
As in any other previously studied species of Antirrhinum, these taxa show patchily arranged populations. The low values of genetic diversity among populations in A. majus agree with the high estimates of gene flow, indicating little population divergence. The geographical proximity among populations, continuity of the habitat and the out-crossing breeding system, allow for a genetic exchange among populations through pollen and/or seed dispersal, as reported for A. charidemi (Mateu-Andrés and Segarra-Moragues, 2000).
Population isolation may lead to stochastic differentiation by genetic drift (Ellstrand and Elam, 1993), and to an increase of differences among populations. The low values of gene flow together with the low correlation among genetic and geographic distances in A. siculum, are consistent with the hypothesis that genetic drift has produced a stochastic differentiation of populations. A particular aspect worthy of consideration in that species is that, in the words of Sutton (1988), some authors ‘have interpreted the occurrence of A. siculum in places such as Perpignan and Jerusalem as introductions dating from the time of Crusades’. Although the only putatively introduced population studied (AS7) is genetically close to most of the others from a supposedly natural area of the species, the presence of allele MNR2-2, present in that population in the species but common in several others in the group (Mateu-Andrés and de Paco, 2005), does not support the hypothesis of a relatively recent introduction. In any case, the present data cannot be conclusive in a species with such low levels of allozymic variability. It is considered that a study on the species phylogeography would shed light on this matter.
By way of conclusion, levels of genetic diversity and its partition were correlated with the reproductive system. Levels of genetic diversity were higher and differentiation among populations was lower in self-incompatible species than in self-compatible ones, in accordance with the assumptions of Hamrick and Godt (1989).
Consequences for species conservation
None of these species are included in red lists (A. linkianum is in the Spanish red list but not at the species range), and most of them show high levels of diversity, a common pattern in species of Antirrhinum, even in some with small and fragmented populations (Mateu-Andrés, 1999; Mateu-Andrés and Segarra-Moragues, 2000). Most species have a small range and small-sized populations, and many are located in places under human influence, which is the main cause of threat for plant species in the Mediterranean area (Domínguez-Lozano et al., 1996; Thompson, 1999). Such a scenario possibly leads to drastic reductions in size or even to the extinction of populations, owing to recreational activities, road construction, grazing etc. The small distance and high gene flow among populations makes the natural recovery of populations possible.
Both the extremely low levels of variability, indicating some genetic erosion of the species, and its reproductive system, make A. siculum a different case. Despite the wide range, the species is severely fragmented, most populations are of a small size and all of them are under human influence, which is likely to lead to drastic reductions or even local extinctions. For such reasons, preservation of natural populations is a priority for species conservation. Although low levels of gene flow among populations exist, recolonization of populations is uncertain owing to large distances. Self-compatibility and, probably the selfing reproductive system, together with the high production of seeds, make the regeneration of populations from remaining seeds possible. The high genetic differentiation among populations indicates that the implementation of complementary strategies, such as ex situ preservation of seeds from different populations, is highly recommended.
Acknowledgments
We are deeply grateful to Dr M. Iberite and A. Dafni for seed collections, to J. Martin and H. L. Warburton for language revision, and to two anonymous referees whose comments very much improved the work.
LITERATURE CITED
- Baker HG. (1959) Reproductive methods as factors in speciation in flowering plants. Cold Spring Harbor Symposia on Quantitative Biology 24177–191. [DOI] [PubMed] [Google Scholar]
- Baur E. (1919) Über Selbsterilität und über Kreuzungsversuche einer selbsterilen Art in der Gattung Antirrhinum. Z. Abtsammgslehre 21.
- Baur E. (1924) Untersuchungen über das Wesen, die Entstehung und die Vererbung von Rassenunterschieden bei Antirrhinum majus. Bibl Genetica 4.
- Baur E. (1932) Artumgrenzung und Artbildung in der Gattung Antirrhinum, Sektion Antirrhinastrum. Z. Abstammgslehre 63256–302. [Google Scholar]
- Barrett SCH and Harder LD. (1996) Ecology and evolution of plant mating. Trends in Evolutionary Ecology 1173–79. [DOI] [PubMed] [Google Scholar]
- Charlesworth D. (2003) Effects of inbreeding on the genetic diversity of populations. Philosophical Transactions of the Royal Society London Series B 3581051–1070. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Domínguez-Lozano F, Galicia Herbada L, Moreno Rivero L, Moreno Saíz JC, Sainz Ollero H. (1996) Threatened plants in peninsular and balearic Spain: a report based on the EU habitats directive. Biological Conservation 76123–133. [Google Scholar]
- East EM. (1940) The distribution of self-sterility in the flowering plants. Proceedings of the American Philosophical Society 82449–517. [Google Scholar]
- Elisens WJ. (1992) Genetic divergence in Galvezia (Scrophulariaceae): evolutionary and biogeographic relationships among South American and Galapagos species. American Journal of Botany 79198–206. [Google Scholar]
- Elisens WJ and Crawford DJ. (1988) Genetic variation and differentiation in the genus Mabrya (Scrophulariaceae–Antirrhineae): systematic and evolutionary inferences. American Journal of Botany 7585–96. [Google Scholar]
- Elisens WJ and Nelson AD. (1993) Morphological and isozyme divergence in Gambelia (Scrophulariaceae): species delimitation and biogeographic relationships. Systematic Botany 18454–468. [Google Scholar]
- Ellstrand NC and Elam DR. (1993) Population genetic consequences of small population size: implications for plant conservation. Annual Review of Ecology and Systematics 24217–242. [Google Scholar]
- Goudet J. (2001) FSTAT, a program to estimate and test gene diversities and fixation indexes (version 2.9.3). Available from http://www.unil.ch/izea/softwares/fstat.html.
- Gruber F. (1932) Über die Verträglichkeitsverhältnisse bei einigen selbsterilen Wildsippen von Antirrhinum und über eine selbstfertile Mutante. Zeitschrift für Indukt Abstammungs-Verebungsl 62426–462. [Google Scholar]
- Hamrick JL and Godt JW. (1996) Effects of life history traits on genetic diversity in plant species. Philosophical Transactions of the Royal Society of London Series B 3511291–1298. [Google Scholar]
- Hamrick JL, Linhart YB, Mitton JB. (1979) Relationships between life history characteristics and electrophoretically detectable genetic variation in plants. Annual Review of Ecology and Systematics 10173–200. [Google Scholar]
- Holsinger KE. (2000) Reproductive systems and evolution in vascular plants. Proceedings of the National Academy of Sciences of the USA 977037–7042. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kondo K, Yamamoto M, Itahashi R, Sato T, Egashira H, Hattori T, Kowyama Y. (2002) Insights into the evolution of self-compatibility in Lycopersicon from a study of stylar factors. The Plant Journal 30143–153. [DOI] [PubMed] [Google Scholar]
- Loveless MD and Hamrick JL. (1984) Ecological determinants of genetic structure in plant populations. Annual Review of Ecology and Systematics 1565–95. [Google Scholar]
- Mateu-Andrés I. (1999) Allozymic variation in three species of Antirrhinum L. (Scrophulariaceae–Antirrhineae). Botanical Journal of the Linnean Society 131187–199. [Google Scholar]
- Mateu-Andrés I. (2004) Low levels of allozyme variability in the threatened species Antirrhinum subbaeticum and A. pertegasii (Scrophulariaceae): implications for conservation of the species. Annals of Botany 94797–804. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mateu-Andrés I and de Paco L. (2005) Allozymic differentiation of the Antirrhinum majus and A. siculum species groups. Annals of Botany 95465–473. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mateu-Andrés I and Segarra-Moragues JG. (2000) Population subdivision and genetic diversity in two narrow endemics of Antirrhinum L. Molecular Ecology 92081–2087. [DOI] [PubMed] [Google Scholar]
- Mateu-Andrés I and Segarra-Moragues JG. (2003a) Patterns of genetic diversity in related taxa of Antirrhinum L. assessed using allozymes. Biological Journal of the Linnean Society 78299–307. [Google Scholar]
- Mateu-Andrés I and Segarra-Moragues JG. (2003b) Reproductive system in the Iberian endangered endemic Antirrhinum valentinum F.Q. (Antirrhineae, Scrophulariaceae): consequences for species conservation. International Journal of Plant Sciences 165773–778. [Google Scholar]
- Nei M. (1972) Genetic distance between populations. American Naturalist 106283–292. [Google Scholar]
- Rice WR. (1989) Analyzing tables of statistical tests. Evolution 43223–225. [DOI] [PubMed] [Google Scholar]
- Rohlf FJ. (1998) NTSYS-pc. Numerical taxonomy and multivariate analysis system. Setauket. Version 2.0. (Exeter software, New York, NY).
- Sherman M. (1939) The inheritance of self-sterility in certain species of Antirrhinum. Zeitschrift für Indukt Abstammungs-Verebungsl 771–17. [Google Scholar]
- Sutton DA. (1988) A revision of the tribe Antirrhineae. (Oxford University Press, London & Oxford)67–97.
- Swofford DL and Selander RB. (1989) Biosys-1. A computer program for the analysis of genetic variation in population genetic and biochemical systematics. version 1.7. (University of Illinois, Urbana, Illinois).
- Thompson JD. (1999) Population differentiation in Mediterranean plants: insights into colonization history and the evolution and conservation of endemic species. Heredity 82229–236. [DOI] [PubMed] [Google Scholar]
- Uenoyama MK. (1995) A generalizad least-squares estimate for the origin of self-incompatibility. Genetics 139975–992. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Vieira CP and Charlesworth D. (2002) Molecular variation at the self-incompatibility locus in natural populations of the genera Antirrhinum and Misopates. Heredity 88172–181. [DOI] [PubMed] [Google Scholar]
- Weir BS and Cockerham CC. (1984) Estimating F-statistics for the analysis of populations structure. Evolution 381358–1370. [DOI] [PubMed] [Google Scholar]
- Wendel JF and Weeden NF. (1989) Visualization and interpretation of plant isozymes. In Soltis DE and Soltis PS (Eds.). Isozymes in plant biology(Dioscorides Press, Portland, OH) pp. 5–45.
- Whitlock MC. (1992) Temporal fluctuations in demographic parameters and the genetic variance among populations. Evolution 46608–615. [DOI] [PubMed] [Google Scholar]
- Wright S. (1931) Evolution in Mendelian populations. Genetics 2224–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wright S. (1965) The interpretation of population structure by F-statistics with spatial regard to system of mating. Evolution 19355–420. [Google Scholar]
- Wright S. (1978) Evolution and the genetics of populations(University of Chicago Press, Chicago, IL) Vol. 4.
- Zwettler D, Vieira CP, Schlötterer C. (2002) Polymorphic microsatellites in Antirrhinum (Scrophulariaceae), a genus with low levels of nuclear sequence variability. Journal of Heredity 93217–221. [DOI] [PubMed] [Google Scholar]