Natural selection of allozyme polymorphisms: a microgeographic climatic differentiation in wild emmer wheat (Triticum dicoccoides)

Summary Allozymic variation in proteins encoded by 48 loci was analyzed electrophoretically in 1984 and 1985 in 137 individual plants of wild emmer wheat, Triticum dicoccoides, from a microsite in Yehudiyya, northeast of the Lake of Galilee, Israel. The test involved two climatic microniches in the open Tabor oak forest (1) sunny between trees and (2) shady under trees' canopies. Significant genetic differentiation at single-, two- and multilocus structures was found between neighboring climatic niches, which were only separated by a few meters. Our results suggest that allozyme polymorphisms in wild emmer wheat are partly adaptive, and differentiate primarily at the multilocus level by climatic factors presumably related to aridity stress.     

Spatiotemporal allozyme divergence caused by aridity stress in a natural population of wild wheat, Triticum dicoccoides, at the Ammiad microsite, Israel

Spatiotemporal diversity at 35 allozyme loci was assayed over 6 years in 1,207 individuals of wild emmer wheat (Triticum dicoccoides)from a microgeographic microsite, Ammiad, north Israel. This analysis used new methods and two additional sample sets (1988 and 1993) and previous allozymic data (1984–1987). This microsite includes four major habitats (North-facing slope, Valley, Ridge, and Karst) that show topographic and ecological heterogeneity. Significant temporal and spatial variations in allele frequencies and levels of genetic diversity were detected in the four subpopulations. Significant associations were observed among allele frequencies and gene diversities at different loci, indicating that many allele frequencies change over time in the same or opposite directions. Multiple regression analysis showed that variation in soil-water content and rainfall distribution in the growing season significantly affected 10 allele frequencies, numbers of alleles at 8 loci, and gene diversity at 4 loci. Random genetic drift and hitchhiking models may not explain such locus-specific spatiotemporal divergence and strong allelic correlation or locus correlation as well as the functional importance of allozymes. Natural ecological selection, presumably through water stress, might be an important force adaptively directing spatiotemporal allozyme diversity and divergence in wild emmer wheat at the Ammiad microsite. Received: 3 July 2000 / Accepted: 1August 2000     

Natural selection of allozyme polymorphisms: a microgeographical differentiation by edaphic,topographical, and temporal factors in wild emmer wheat (Triticum dicoccoides)

Summary Allozymic variation in proteins encoded by 47 loci was analyzed electrophoretically in 1983/4 and 1984/5 in 356 individual plants of wild emmer wheat, Triticum dicoccoides, from a microsite at Tabigha, north of the Sea of Galilee, Israel. Each year the test involved two 100-meter transects, each equally subdivided into basalt and terra rossa soil types, and comparisons were based on 16 common polymorphic loci. Significant genetic differentiation, genetic phase disequilibria, and genome organization according to soil type were found over very short distances. Our results suggest that allozyme polymorphisms in wild emmer wheat are partly adaptive, and that they differentiate at both single and multilocus structures primarily from environmental stress of such ecological factors as soil type, topography, and temporal changes, probably through aridity stress.     

Edaphic microsatellite DNA divergence in wild emmer wheat, Triticum dicoccoides, at a microsite: Tabigha, Israel

Twenty eight microsatellite markers were used to analyze genetic divergence in tandem dinucleotide repeated DNA regions between two edaphic subpopulations of Triticum dicoccoides growing on the contrasting terra rossa and basalt soilsfrom a microsite at Tabigha, north of the Sea of Galilee, Israel. The terra rossa soil niche was drier and more stressful than the basalt throughout the growing season (November to May). Significant microsatellite divergence in allele distribution, repeat length, genetic diversity, and linkage disequilibria were found between emmer wheat from the two soil types over two short transects of 100 m each. Soil-specific and -unique alleles and linkage disequilibria were observed in the terra rossa and basalt subpopulations. A permutation test showed that the effects of random genetic drift were very low for the significant genetic diversity at microsatellite loci between the two subpopulations, suggesting that an adaptive molecular pattern derived by edaphic selection may act upon variation of the microsatellites. Received: 4 February 2000 / Accepted: 31 March 2000<@head-com-p1a.lf>Communicated by H.F. Linskens     

RAPD polymorphism of wild emmer wheat populations, Triticum dicoccoides, in Israel

 Genetic diversity in random amplified polymorphic DNAs (RAPDs) was studied in 110 genotypes of the tetraploid wild progenitor of wheat, Triticum dicoccoides, from 11 populations sampled in Israel and Turkey. Our results show high level of diversity of RAPD markers in wild wheat populations in Israel. The ten primers used in this study amplified 59 scorable RAPD loci of which 48 (81.4%) were polymorphic and 11 monomorphic. RAPD analysis was found to be highly effective in distinguishing genotypes of T. dicoccoides originating from diverse ecogeographical sites in Israel and Turkey, with 95.5% of the 100 genotypes correctly classified into sites of origin by discriminant analysis based on RAPD genotyping. However, interpopulation genetic distances showed no association with geographic distance between the population sites of origin, negating a simple isolation by distance model. Spatial autocorrelation of RAPD frequencies suggests that migration is not influential. Our present RAPD results are non-random and in agreement with the previously obtained allozyme patterns, although the genetic diversity values obtained with RAPDs are much higher than the allozyme values. Significant correlates of RAPD markers with various climatic and soil factors suggest that, as in the case of allozymes, natural selection causes adaptive RAPD ecogeographical differentiation. The results obtained suggest that RAPD markers are useful for the estimation of genetic diversity in wild material of T. dicoccoides and the identification of suitable parents for the development of mapping populations for the tagging of agronomically important traits derived from T. dicoccoides. Received: 13 July 1998 / Accepted: 13 August 1998     

Genetic diversity and environmental associations of wild wheat,Triticum dicoccoides,in Israel

Summary Allozyme variation in the tetraploid wild progenitor of wheat, Triticum dicoccoides, was studied for the proteins encoded by about 50 gene loci in 457 individuals representing 12 populations from Israel. Six spikelet morphological traits were measured in the same populations. The results indicate that: (a) 16 loci (= 32%) were monomorphic in all 12 populations, 15 loci (= 30%) were locally polymorphic, and 19 loci (= 38%) were regionally polymorphic. All polymorphic loci (but one) displayed high levels of polymorphism ( 10%). In Israel, the proportion of polymorphic loci per population, P, in wild wheat averaged 0.25 (range, 0.16–0.38), and the genetic diversity index, H_e averaged 0.07, (range, 0.03 – 0.12). (b) Altogether there were 110 alleles at the 50 putative loci tested (c) Genetic differentiation of populations included regional and local patterns: (i) The coefficients of genetic distance between populations were high (mean D = 0.10 range, 0.02 – 0.25), and indicated sharp genetic differentiation over short distances, (ii) Common ( 10%) but sporadic and localized alleles were frequent (76%), and (iii) Rare alleles were few (only 5 alleles). (d) The patterns of allozyme and spikelet variation in the wild gene pool were significantly correlated with, and partly predictable by, water factors, including those of precipitation, evaporation, and relative humidity as well as of soil type, (e) All six spikelet characters showed statistically significant variation among localities and (f) Allozymic variation was correlated with spikelet variation.These results suggest in T. dicoccoides: (i) the operation of natural selection in population genetic structure, (ii) local adaptive genetic differentiation caused by diversifying selection through climate and soil, and (iii) the guidelines for sampling these resources for use in wheat breeding programs.     

Microsatellite polymorphism in natural populations of wild emmer wheat,Triticum dicoccoides,in Israel

Diversity in 20 microsatellite loci of wild emmer wheat, Triticum dicoccoides, was examined in 15 populations (135 genotypes) representing a wide range of ecological conditions of soil, temperature, and water availability, in Israel and Turkey. An extensive amount of diversity at microsatellite loci was observed despite the predominantly selfing nature of this plant species. The 20 Gatersleben wheat microsatellites (GWM), representing 13 chromosomes of genomes A and B of wheat, revealed a total of 364 alleles, with an average of 18 alleles per GWM marker (range: 5–26). The proportion of polymorphic loci per population averaged 0.90 (range: 0.45– 1.00); genic diversity, He, averaged 0.50 (range 0.094– 0.736); and Shannon’s information index averaged 0.84 (range 0.166–1.307). The coefficients of genetic distance between populations were high and averaged D=1.862 (range 0.876–3.320), an indication of sharp genetic divergence over short distances. Interpopulation genetic distances showed no association with geographic distance between the population sites of origin, which ruled out a simple isolation by distance model. Genetic dissimilarity values between genotypes were used to produce a dendrogram of the relationships among wild wheat populations by the unweighted pair-group method with arithmetic averages (UPGMA). The results showed that all the wild emmer wheat populations could be distinguished. Microsatellite analysis was found to be highly effective in distinguishing genotypes of T. dicoccoides, originating from diverse ecogeographical sites in Israel and Turkey, with 88% of the 135 genotypes correctly classified into sites of origin by discriminant analysis. Our present microsatellite results are non-random and in agreement with the previously obtained allozyme and RAPD patterns, although the genetic-diversity values obtained with microsatellites are much higher. Significant correlates of microsatellite markers with various climatic and soil factors suggest that, as in allozymes and RAPDs, natural selection causes adaptive microsatellite ecogeographical differentiation, not only in coding, but most importantly in non-coding genomic regions. Hence, the concept of ”junk DNA” needs to be replaced by at least partly regulatory DNA. The obtained results suggest that microsatellite markers are useful for the estimation of genetic diversity in natural populations of T. dicoccoides and for the tagging of agronomically important traits derived from wild emmer wheat. Received: 27 February 2001 / Accepted: 22 March 2001     

Outcrossing rates and their relationship to phenology in Triticum dicoccoides

Summary Outcrossing rates (t) were estimated in natural and garden populations of wild emmer wheat, Triticum dicoccoides. The estimated t from the natural populations was 0 with the 0.95 upper confidence limit ranging from 0.102 to 0.925 depending on the genetic variation within the population. Outcrossing rates for two genotype classes of this species grown in a common garden were 0.0077 and 0.0018. The difference in Outcrossing rates between the two genotypes is ascribed to phenological differences and hence different available pollen pools. Phenological displacement is discussed as a possible bias in the estimation of outcrossing rates in both predominantly selfing and outcrossing plants.     

Natural selection causing microsatellite divergence in wild emmer wheat at the ecologically variable microsite at Ammiad, Israel

Genetic diversity at 28 microsatellite loci was studied in a natural population of Triticum dicoccoides at the Ammiad microsite, north of the Sea of Galilee, Israel. This microsite was subdivided into four major habitats, North, Valley, Ridge and Karst, and further subdivided into nine subhabitats. The units thus defined showed strong and highly significant differentiation in ecological factors; in particular with respect to cover, proximity and height of rocks, and surface soil moisture after early rains. The results showed that allele distributions at microsatellite loci were nonrandom and associated with habitats. Significant genetic differentiation and variation in repeat number were found among subpopulations in the four major habitats and nine subhabitats. Habitat-specific and -unique alleles and linkage disequilibria were observed in the Karst subpopulation. The subpopulations dwelling in drier habitats and subhabitats showed higher genetic diversities at microsatellite loci. These results suggest that natural selection, presumably through aridity stress, acts upon microsatellite divergence predominantly on noncoding sequences, thereby contributing to differences in fitness. Received: 9 September 1999 / Accepted: 16 September 1999     

High-density molecular map of chromosome region harboring stripe-rust resistance genes YrH52 and Yr15 derived from wild emmer wheat, Triticum dicoccoides

Two stripe-rust resistance genes, YrH52 and Yr15, derived from the Israeli wild emmer wheat, Triticum dicoccoides, have been located on chromosome 1B. The main objectives of the present study were to increase marker density in the vicinity of YrH52 gene by means of AFLP, RAPD and microsatellite markers, to improve the map of another T. dicoccoides-derived stripe-rust resistance gene Yr15 using microsatellite markers, and to preliminarily discriminate these two genes. Additional 26 marker loci comprising 20 AFLPs, three RAPDs, and three microsatellites were found to be linked to YrH52 gene. An updated genetic map consisting of 45 marker loci, in the region of YrH52 gene, was constructed with a total map length of 107.7 cm. The mean interval length was 0.96 cm in the region Xgwm359b–P55M53b carrying YrH52 gene. YrH52 was bracketed by Xgwm413 (Nor1 and UBC212a) and Xgwm273a (Xgwm273d) with map distance of 1.3 and 2.7 cm from either side, respectively. Eight additional microsatellite markers were found to be linked with Yr15, and the linkage map of Yr15 gene was thus obviously improved. In the YrH52-mapping population, no crossover was detected in the interval UBC212a (Xgwm413)–Yr15–Nor1, and YrH52 was located distally outside this interval. It may suggest that YrH52 is different from Yr15 even though both of them are derived from T. dicoccoides and are mapped on chromosome 1BS. The large number of molecular makers revealed in the present study would be helpful for the marker-assisted introgression of the T. dicoccoides-derived YrH52 and Yr15 stripe-rust resistance genes into elite cultivars of wheat, and the high-density map would accelerate the map-based cloning of the two genes. This revised version was published online in July 2006 with corrections to the Cover Date.     

Photosynthetic performance in wild emmer wheat,Triticum dicoccoides: ecological and genetic predictability

Summary In a twin study, we have shown that wild emmer wheat, Triticum dicoccoides, the progenitor of all cultivated wheats, harbours important genetic variation (Vg) in photosynthetic characteristics. This Vg resides within and between populations and ecogeographical regions in Israel, which is the center of origin and diversity of wild emmer wheat. Here we analyzed, by univariate and multivariate methods, the significant differentiation of variation in photosynthetic characteristics of 107 genotypes from 27 populations of wild emmer in Israel, distributed in three ecogeographical regions including central, xeric (northern cold and eastern warm) marginal, and mesic (western) marginal populations. The highest photosynthetic efficiency was displayed by populations of the xeric marginal region, but most variation for photosynthetic capacity occurs between accessions within ecogeographical regions and populations. Genotypes and populations of T. dicoccoides having high photosynthetic capacity can be identified by climatic factors and isozyme markers. The identification by genetic markers, if substantiated by testcrosses, can facilitate the maximization of conservation, in situ or ex situ, and utilization of these photosynthetic genetic resources for improvement of hexaploid wheat (T. aestivum).     

普通小麦与野生二粒小麦A、B基因组的遗传进化关系的SSR分析

以29份普通小麦、野生二粒小麦及野生二粒小麦与节节麦杂交的双二倍体为材料,选取A、B染色体组特异的SSR引物,通过微卫星多态性检测方法探讨上述两物种之间A、B染色体组的遗传进化关系。经NTSYS系统软件UPGMA聚类分析表明,在遗传相似系数为0．18时可以将普通小麦与野生二粒小麦大致分为两大类群,说明野生二粒小麦的A、B染色体组在进化过程中其遗传物质发生了较大变化,但同时野生二粒小麦的D25与D13与普通小麦分为一组,尤其是D13与新疆的红冬麦遗传距离较近,也证明了其两物种间有一定的同源性,并对这些遗传物质分化发生的机制及SSR引物的保守性进行了讨论。     

Use of RAPD markers to determine the genetic diversity of diploid,wheat genotypes

Summary The genetic diversity of two diploid wheat species, Triticum monococcum and Triticum urartu (2n=2x=14), was assessed using random primers and the polymerase chain reaction (PCR). Electrophoretic analysis of the amplification products revealed a higher incidence of polymorphism in T. urartu than T. monococcum. Pair-wise comparisons of unique and shared polymorphic amplification products, were used to generate Jaccard's similarity coefficients. These were employed to construct phenograms using an unweighted pair-group method with arithmetical averages (UPGMA). The UPGMA analysis indicated a higher similarity among T. monococcum than T. urartu. Analysis of RAPD data appears to be helpful in determining the genetic relationships among genotypes.Contribution of the College of Agricultural Sciences, Texas Tech University Journal No. T-4-334. This work was supported by a grant from the National Science Foundation (BSR-8552915)     

Patterns of genetic diversity and candidate genes for ecological divergence in a homoploid hybrid sunflower, Helianthus anomalus

Natural hybridization accompanied by a shift in niche preference by hybrid genotypes can lead to hybrid speciation. Natural selection may cause the fixation of advantageous alleles in the ecologically diverged hybrids, and the loci experiencing selection should exhibit a reduction in allelic diversity relative to neutral loci. Here, we analyzed patterns of genetic diversity at 59 microsatellite loci associated with expressed sequence tags (ESTs) in a homoploid hybrid sunflower species, Helianthus anomalus. We used two indices, ln RV and ln RH, to compare variation and heterozygosity (respectively) at each locus between the hybrid species and its two parental species, H. annuus and H. petiolaris. Mean values of ln RV and ln RH were significantly lower than zero, which implies that H. anomalus experienced a population bottleneck during its recent evolutionary history. After correcting for the apparent bottleneck, we found six loci with a significant reduction in variation or with heterozygosity in the hybrid species, compared to one or both of the parental species. These loci should be viewed as a ranked list of candidate loci, pending further sequencing and functional analyses. Sequence data were generated for two of the candidate loci, but population genetics tests failed to detect deviations from neutral evolution at either locus. Nonetheless, a greater than eight-fold excess of nonsynonymous substitutions was found near a putative N-myristoylation motif at the second locus (HT998), and likelihood-based models indicated that the protein has been under selection in H. anomalus in the past and, perhaps, in one or both parental species. Finally, our data suggest that selective sweeps may have united populations of H. anomalus isolated by a mountain range, indicating that even low gene-flow species may be held together by the spread of advantageous alleles.     

Storage-protein variation in wild emmer (Triticum turgidum ssp.dicoccoides) from Jordan and Turkey.

Genetic diversity in the seed storage-proteins encoded at theGlu-A1,Glu-B1 andGli-B1/Glu-B3 loci was studied electrophoretically in 315 individuals belonging to nine populations ofT. dicoccoides from Jordan and three from Turkey. The inter- and intra-population distribution of seed storage-protein alleles at the considered loci and its link with geographical factors were investigated. Population differentiation in seed storage-proteins was in some cases very high with very weak correlations with geographic distance. Greater gene differentiation was found within and between populations which were geographically very close in Jordan than between those from Jordan and Turkey. However the distribution of alleles appeared to be non random. Samples collected from populations at locations over 900 m above sea level were less polymorphic than those collected at lower altitudes (500–700 m), whereas the relative genetic differentiation between populations was greater between those collected at higher altitudes. Seed storage-protein differentiation was significantly correlated with the altitude of the collecting sites. Although it is difficult to point out the selective pressure of altitude per se, altitude can reflect an integration of several environmental parameters. The possible adaptive value of seed storage-proteins is discussed.     

Resistance of wild wheat to stripe rust: Predictive method by ecology and allozyme genotypes

From 114 accessions of wild emmer wheat from 11 sites in Israel, known for their allozymic variation (Nevo & al. 1982), individual genotypes were tested for resistance to one isolate of stripe rust both in the seedling stage in a growth chamber and in the adult plant stage in the field. The results indicate that resistance to stripe rust in seedlings and adults are significantly correlated (r_s = 0.40, p < 0.001). Genetic polymorphisms of resistance to stripe rust vary geographically and are predictable by climatic, as well as allozymic markers. Three variable combinations of rainfall, evaporation, and temperature explain significantly 0.40–0.53 of the spatial variance in disease resistance to stripe rust, suggesting the operation of natural selection. Several allozyme genotypes are significantly associated with disease resistance. We conclude that natural populations of wild emmer wheat in Israel contain large amounts of disease resistance genes. These populations could be effectively screened and then utilized by the phytopathologist for identifying resistant genotypes and producing new resistant cultivars.Patterns of Resistance of Wild Wheat to Pathogens in Israel II.     

Wheat storage proteins: glutenin diversity in wild emmer,Triticum dicoccoides,in Israel and Turkey. 2. DNA diversity detected by PCR

Germplasm of Triticum dicoccoides collected from different environments in Israel was evaluated by using the PCR as a molecular marker. Two pairs of primers were used in the PCR in amplifying the DNA tracts coding the high-and low-molecular-weight glutenin subunits. Analyses reveal great variability within and between populations indicating the high values of this germplasm for future breeding programs to improve the protein quality in wheat.     

Storage-protein variation in wild emmer wheat (Triticum turgidum ssp.dicoccoides) from Jordan and Turkey. I. Electrophoretic characterization of genotypes

Seed storage-protein variation at theGlu-A1,Glu-B1 andGli-B1/Glu-B3 loci in the tetraploid wild progenitor of wheat,T. dicoccoides, was studied electrophoretically in 315 individuals representing nine populations from Jordan and three from Turkey. A total of 44 different HMW-glutenin patterns were identified, resulting from the combination of 15 alleles in the A genome and 19 in the B genome. Twenty-seven new allelic variants, 12 at theGlu-A1 locus and 15 at theGlu-B1 locus, were identified by comparing the mobilities of their subunits to those previously found in bread and durum wheats. The novel variants include six alleles at theGlu-A1 locus showing both x and y subunits. The genes coding for the 1Bx and 1By subunits showed no or very little (3%) inactivity, the 1Ax gene showed a moderate degree (6.3%) of inactivity whereas the gene coding for lAy showed the highest degree of inactivity (84.8%). A high level of polymorphism was also present for the omega- and gamma-gliadins and LMW-glutenin subunits encoded by genes at the linkedGli-B1 andGlu-B3 loci (19 alleles). Some Jordanian accessions were found to contain omega-gliadin 35, gamma-gliadin 45, and LMW-2 also present in cultivated durum wheats and related to good gluten viscoelasticity. The newly-discovered alleles enhance the genetic variability available for improving the technological quality of wheats. Additionally some of them may facilitate basic research on the relationship between industrial properties and the number and functionality of HMW- and LMW-glutenin subunits.     

Genetic diversity in European and Mediterranean faba bean germ plasm revealed by RAPD markers

Broadening of the genetic base and systematic exploitation of heterosis in faba bean requires reliable information on the genetic diversity in the germ plasm. Three groups of faba bean inbred lines were examined by means of RAPDs (random amplified polymorphic DNAs) assays: 13 European small-seeded lines, 6 European large-seeded lines, and 9 Mediterranean lines. Out of 59 primers, 35 were informative and yielded 365 bands, 289 of which were polymorphic with a mean of 8.3 bands per primer. Monomorphic bands were omitted from the analyses and genetic distances (GD) were estimated via the coefficient of Jaccard. The mean GD among the European small-seeded lines was significantly greater than those among the lines of the other two groups. Repeatability of GD estimates was high. Cluster (UPGMA) and principal coordinate analyses identified European small-seeded lines and Mediterranean lines as distinct groups with European large-seeded lines located in between. The results are in harmony with published archaeobotanical findings. We conclude that RAPDs are useful for classification of germ plasm and identification of divergent heterotic groups in faba bean.     

Genomic microsatellite adaptive divergence of wild barley by microclimatic stress in 'Evolution Canyon', Israel

We examined diversity levels and patterns of 19 nuclear microsatellites and four chloroplast microsatellites in 275 genotypes of wild barley Hordeum spontaneum, in seven stations at the ‘Evolution Canyon’ (EC) microsite, Lower Nahal Oren, Mt. Carmel, Israel. EC is sharply subdivided ecologically into a tropical savannoid, ‘African’, xeric, south‐facing slope (SFS) abutting the temperate, dense, liveoak, brushwood, ‘European’, mesic, north‐facing slope (NFS). We found the following. (i) 17 of 19 (89.5%) nuDNA simple sequence repeats (SSRs) were polymorphic across all seven subpopulations and three chDNA SSRs were polymorphic. (ii) A total of 216 nuDNA SSR alleles, with a maximum of 23 alleles in a nuclear locus, and ten chDNA SSRs, with a maximum of four alleles in a locus, were registered. (iii) There were striking and significant inter‐ and intraslope diversities, based on the 19 nuDNA SSRs, climaxing with a remarkable genetic distance between the mid‐slope stations on opposite slopes (D_A = 0.481), across a distance of 200 m. This genetic distance is as large as that between the H. spontaneum populations of Jerusalem and Sede Boqer, which are separated by 100 km (× 500 larger in transect length). (iv) Slope‐unique alleles (103 = 45.6%) were higher on the ‘European’ than on the ‘African’ slope. Slope‐specific (predominant) alleles (17) were equal on opposite slopes. (v) nuDNA SSR gene diversity was higher on the ‘European’ slope and the opposite was found for the chDNA SSR. (vi) nuDNA SSR genic differentiation was very high between opposite slopes, with Gst = 0.187; for chDNA SSR this value was 0.127. Our results are inexplicable by stochastic processes and suggest that: (i) microclimatic diversifying selection is the major evolutionary, fast‐acting, interslope force, overriding migration and drift, and (ii) ecological stress can generate local, regional and global adaptive patterns, suggesting that natural selection is a major differentiating force of both coding and noncoding SSRs linking micro‐ and macroevolutionary processes. © 2005 The Linnean Society of London, Biological Journal of the Linnean Society, 2005, 84 , 205–224.