Printer Friendly

Genetic and morphometric divergence in ancestral European and descendant NewZealand populations of chaffinches (fringilla coelebs).

Key words. -- Allozymes, drift, Fringilia coelebs, gene flow, morphometrics, population structure, selection.

Received December 4, 1991. Accepted April 16, 1992.

A fundamental postulate of the neo-Darwinian synthesis is that gradual adaptive divergence by natural selection can be extended through time to explain macroevolutionary phenomena (Gould, 1982). The exclusivity of this mode of evolution has been challenged by those who argue that (1) severe reductions in population size are sometimes involved (founder-induced speciation, Carson and Templeton, 1984), or (2) most evolutionary change is concentrated in short periods coincident with speciation events (punctuated equilibrium, Eldredge and Gould, 1972), or (3) that nonselective epigenetic processes in novel environments are responsible for biological diversity (Ho and Saunders, 1979). Proponents of modem neo-Darwinism have replied to these criticisms by stressing that the sudden appearance and prolonged morphological stasis of some species (notably in the fossil record) are compatible with known genetic mechanisms linking microevolution to macroevolution (e.g., Charlesworth et al., 1982; Barton and Charlesworth, 1984). Recently, Turner (1988) has attempted to reconcile these views by arguing that patterns of evolution can be accounted for by cumulative microevolutionary changes as well as differential rates of extinction in different clades.

Comparisons of introduced populations with their ancestral stock are relevant to this controversy because they provide direct estimates of the magnitude and rate of microevolutionary differentiation over different time frames. For example, many species of passerine birds were introduced to New Zealand from England last century, and thus it is possible to compare the amount of population differentiation that has developed in New Zealand in approximately 120 years relative to that in "ancestral" populations which have presumably evolved in several millenia.

A detailed history of the introduction of chaffinches into New Zealand is provided in Thomson (1922), the salient features of which are as follows. Between 1862 and 1877 about 400 birds were imported from England, though it is not clear whether the birds were caught in the breeding season or during winter, in which case overwintering European birds could have been included in shipments. Approximately 100 chaffinches each were released at Auckland, Wellington, and Dunedin, 23 at Nelson, and the rest at Christchurch. Chaffinches have been very successful colonizers, and are now one of the most abundant passerines throughout New Zealand and offshore islands.

A recent analysis of morphometric and genetic differentiation among eight New Zealand populations of chaffinches revealed that the amount of interpopulational divergence that has developed since their introduction is not only small but is also haphazard with respect to geographic proximity of populations and to ecoclimatic gradients (Baker et al., 1990b). Comparisons with populations of chaffinches isolated in the Atlantic Islands (Azores, Madeira, and Canaries) and in the neighboring continental regions of Iberia and Morocco suggested that microevolutionary processes driving divergence among New Zealand populations could be projected through time to account for the larger-scale geographic variation, subspeciation, and speciation in the northern hemisphere populations.

Without some measure of scale of the magnitude of population divergence in the ancestral populations in northern Europe from which the New Zealand founders were drawn, however, this conclusion is rooted firmly in phyletic gradualism. Detailed analyses of variation in Atlantic Island populations relative to their continental counterparts strongly support gradual divergence over the last million to 100,000 years or so in small to moderate-sized populations as the mode of differentiation (Baker et al., 1990a). Nevertheless, we clearly need an assessment of differentiation among connected continental demes in the intermediate term of several millenia to assess more accurately hypotheses of gradual divergence. In this paper, I calibrate the magnitude of population divergence that has developed in New Zealand since colonization last century against that developed in the ancestral" stock, as exemplified in extant European populations. Using this comparative information in the context of different divergence times in ancestral and descendent populations, as well as the pattern of divergence among populations, I attempt to infer processes of microevolution.

MATERIALS AND METHODS

Sample Details

Adult chaffinches were collected from 8 locations in New Zealand (Baker et al., 1990b), and from 10 locations throughout Europe and Great Britain (Fig. 1, Table 1). All samples were collected between 1984 and 1986 to minimize temporal variation among specimens. [TABULAR DATA 1 OMITTED]

Electrophoresis

Heart, liver, and pectoral muscle were removed from each specimen immediately after death and stored in cryogenic tubes in liquid nitrogen or at - 70 [degrees] C until they were electrophoresed. Tissues from the 491 specimens were screened for genetic variation at 42 presumptive loci using running buffers optimized for chaffinches (see Baker et al., 1990a). Gels were run overnight for 16 hr at 4 [degrees] C, and were then stained using the methods detailed in Barrowclough and Corbin (1978), Cole and Parkin (1981), and Harris and Hopkinson (1976). Electromorphs were assumed to be products of different alleles. Alleles from each population were calibrated by comparing them side-by-side on the same gel. Loci were numbered sequentially with integers beginning with one for the most anodal form, and alleles were designated alphabetically (with A for the most common one).

Genetic Analysis

Genetic data for all populations were analyzed with the computer package BIOSYS-1 (Swofford and Selander, 1981). Observed (direct count) and expected (based on Hardy-Weinberg equilibrium) heterozygosities were calculated and averaged across all loci for each population. As there were no significant differences between average observed and expected heterozygosities (Kruskal-Wallis tests, P > 0.05) in any of the populations, only the theoretically preferable latter estimate (Nei, 1978) is reported beyond. To test for departures from expected Hardy-Weinberg proportions of genotypes within samples, I employed chis-quare tests with pooling of all uncommon alleles to guard against inflation of the chis-quare values when cells have expected frequencies of less than one (Swofford and Selander, 1981). Exact probabilities for small samples were used in evaluating the chis-quare statistics (Vithayasai, 1973).

Geographic heterogeneity in allele frequencies among all populations was tested at each locus with contingency chi-square analysis, using the method of Workman and Niswander (1970). The extent of population structuring and genetic differentiation were investigated with F statistics (Wright, 1965, 1978). Differentiation in different regions was assessed by averaging F statistics over all polymorphic loci for (1) New Zealand, (2) northern Europe (including Great Britain), (3) Europe (including Iberia), and (4) all 18 populations. In BIOSYS-1 [F.sub.st] is weighted by allele frequencies, and thus is equivalent to [G.sub.st] (Swofford and Selander, 1981). A hierarchical [F.sub.st] for Europe was computed to partition genetic variance among regions (Iberia and northern Europe) and constituent populations.

Multilocus genetic comparisons were made among populations by computing Rogers' (1972) genetic distances and clustering them with UPGMA cluster analysis (Sneath and Sokal, 1973). As a check on possible distortion of relationships of populations represented in the original distance matrix I also ordinated the samples in 3-D space with principal coordinates analysis (Gower, 1966). A minimum spanning tree based on the full dimensional genetic distance matrix was then superimposed between the sample projections to indicate relationships and to reveal any distortions in the reduced 3-D space.

Morphometric Variation

Morphometric variation in the 10 European populations was assessed by taking 12 skeletal measurements with dial calipers to the nearest 0.05 mm on each specimen, as follows: (1) premaxilla length (PREW); (2) cranium depth (CRAD); (3) cranium length (CRAL); (4) mandible length (MAND); (5) humerus length (HUML); (6) ulna length (ULNA); (7) sternum length (STER); (8) coracoid length (CORA); (9) femur length (FEMA); (10) tarsometatarsus length (TARS); (11) synsacrum length (SACL); and (12) synsacrum width (SACW). The same measurements have previously been made for the samples from eight New Zealand localities (see Baker et al., 1990b for details of how the measurements were made), and the two datasets were combined for the present analysis. Only skeletons of males were measured because they predominated in all collected samples.

To quantify the level of divergence between New Zealand and European populations of chaffinches, I conducted single classification analysis of variance among all 18 populations. The pattern of morphometric differentiation among population means was investigated using Student-Newman-Keuls (SNK) tests. Multivariate morphometric divergence of the populations was also conducted with principal component analysis, based on a variance-covariance matrix calculated from population means of In-transformed measurements. To indicate any distortions in the 2-D principal component ordination, a minimum spanning tree computed from a matrix of average taxonomic distances (Sneath and Sokal, 1973) was superimposed on the 2-D plot to indicate full-dimensional relationships.

Environmental Variation in Europe

A detailed analysis of the association of environmental variation and morphometric differentiation among European populations will be presented elsewhere. Here I extend the analysis to the frequencies of common alleles at the five loci showing significant geographic variation in Europe. Separate linear regressions of allele frequencies at Ada(A), Es-1(A), Np(A), PepB(A), and Pgd(A) were computed on the five environmental variables used in the morphometric analysis, as follows: (1-2) mean January and July temperatures, (3) latitude, (4) actual evapotranspiration in the breeding season (April-August), and (5) annual rainfall. Climatic data were taken from Wernstadt (1972), and data for soil water capacity and day length used in calculating actual evapotranspiration were from Thornwaite and Mather (1957).

RESULTS

Genetic Variation

Genetic assays of 42 protein-encoding loci revealed that 20 were monomorphic and fixed for the same allele in all 18 New Zealand and European populations: Acp-2, Acp-3 (Eap), Ak-1, Ak-2, Ak-3, Es-4, Es-5, Got-2, Gpd-2, G6pd, Ldh-1, Ldh-2, Mdh-2, Pt-1, Pt-2, Pt-3, Sdh, Sod-1, Sod-2, and Sod-3. The remaining 22 loci were polymorphic in one or more populations; allele frequencies at these loci are shown in Table 2. For 20 of the polymorphic loci, the New Zealand and European populations have the same common allele. In the other two polymorphic loci (Ada and Np), the three Iberian populations are distinguished from all others by the presence of an alternative common allele (B), which approaches fixation in the Segovia population. Another clear feature of the allele frequency data is that the "ancestral" European populations collectively possess a broader suite of uncommon or rare alleles than do descendent New Zealand populations. Sixteen such alleles occur exclusively in Europe, whereas only five alleles are restricted to New Zealand. Nevertheless, there are no significant differences (t-tests, P > 0.05) in genetic variability between these populations in terms of number of alleles/locus, percentage of polymorphic loci, and average heterozygosity (Table 2). [TABULAR DATA 2 OMITTED]

As in the New Zealand populations (Baker et al., 1990b), the European populations show departures from Hardy-Weinberg proportions at a few loci. Departures at Acp-1 in Setubal, Ck-1 in Hourtin, Es-1 in St. Quentin, Pgd in Segovia and St. Quentin, and Pep-B in Hourlin are all caused by small deficiencies of heterozygotes involving rare or uncommon alleles, and thus are most likely attributable to sampling error. Small but statistically significant (P < 0.05) heterozygote deficiencies that do not appear to emanate from sampling error are apparent for Es-1 in Wareham and St. Quentin, and for Es-2 in St. Marie.

Genetic Differentiation among Populations

Significant heterogeneity in allele frequencies was detected at 12 loci (Ada, Ck-2, Es-1, Es-2, Es-3, Gda, Gpd-1, Icd-1, Mpi, Np, Pep-B, and Pgd, when all 18 New Zealand and European populations were tested with contingency chi-square analysis (P < 0.0 1, see Table 3). Ancestral populations in northern Europe are essentially panmictic, with only one locus (Es-1) showing significant geographic variation. However, the addition of the three populations from the Iberian Peninsula in southern Europe inflates this number to five loci (Ada, Es-1, Np, Pep-B, and Pgd), pointing to genetic subdivision coincident with the Pyrenees mountains that form a geographic barrier between the Iberian Peninsula and the rest of Europe. Significant geographic variation in allele frequencies is apparent at five loci (Ck-2, Es-2, Gda, Gpd-1, and Np) in the descendent New Zealand populations, and only one of these loci (Np) is similarly variable in Europe. [TABULAR DATA 3 OMITTED]

The same pattern of genetic differentiation is evident for Rogers' (1972) genetic distance, DR (Table 3). The New Zealand populations are less differentiated than populations across Europe, but the northern European subset of populations are the least differentiated of all. Genetic subdivision is much more pronounced in Europe than elsewhere, however, with an approximately seven-fold increase in [F.sub.st] values over those in New Zealand. Individual locus [F.sub.st] values reveal that this subdivision is primarily due to variation at two loci (Ada and Np). Both New Zealand and northern European populations are only very weakly subdivided (Table 3). Hierarchical analysis of variance with regions and populations in Europe as levels (Wright, 1978) revealed that 79.4% of the gene diversity is distributed among European populations, 19.4% is apportioned between Iberia and northern Europe, and the remaining 1.2% is within these two regions.

UPGMA cluster analysis graphically depicts the divergence of the Iberian populations from other European populations, as well as the smaller scale of divergence between the ancestral northern European and descendent New Zealand populations (Fig. 2). The scale of differentiation within New Zealand is small but is of the same order as that within northern Europe. The degree of divergence of the descendent New Zealand populations from the ancestral northern European populations is much less than that between the latter and Iberian populations, and thus the New Zealand and northern Europian populations cluster together, with the Iberian populations as an outlier. The matrix correlation between the original genetic distance matrix and the cophenetic values in the phenogram is high ([r.sub.dd*] = 0.959), indicating that the phenogram accurately represents the full-dimensional matrix. This was confirmed with three-dimensional principal coordinates analysis (not shown), which generated an ordination of populations with identical relationships to those in

the phenogram.

Morphometric Differentiation among Populations

Means and standard errors of the 12 skeletal characters for the 10 European populations are presented in Table 4; for a comparable summary for New Zealand populations see Table 6 in Baker et al. (1990b). Analysis of variance revealed statistically significant (Bonferroni adjustment for multiple comparisons, P = 0.05/216 = 0.0002) variation among the 18 New Zealand and European population means for 11 of 12 morphometric characters (all except tarsometatarsus length, TARS). For four characters (HUML, ULNA, SACL, and SACW), SNK tests indicated that some or most of the New Zealand populations have significantly smaller means than their largest European counterparts (KALO and CHAR). The reverse holds for only one character (CRAD), where the northernmost New Zealand population at Woodhill (WOOD) has the largest mean, but it differs significantly only from the European population in Hourtin (HOUR) in France. For the remaining seven characters, the New Zealand population means are not significantly different from those of central Europe (Great Britain and France), but for skull characters the New Zealand populations have larger means than those of the three Iberian populations. [TABULAR DATA 4 OMITTED]
TABLE 5. Percent variance components for 12 morphometric
characters of chaffinches in Europe and New
Zealand.
Variance component
Northern
All Europe      Europe      New Zealand
Character       (N= 10)        (N= 7)       (N= 8)
PREW             34.3          19.9            0.0
CRAD               7.7           8.3          13.2
CRAL             38.8          23.6            4.0
MAND             39.9          20.5            2.9
HUML             21.7          27.9            3.0
ULNA             22.5          28.2            2.4
STER             31.7          26.5            0.8
CORA             24.1          18.5            0.0
FEML             20.3          19.5            0.0
TARS             12.1            9.8           0.0
SACL             25.4          16.2            2.1
SACW             17.6            5.6           5.2


The geographic pattern of morphometric differentiation among European populations, as well as their divergence from the New Zealand populations, is depicted in the principal components ordination (Fig. 3). Full-dimensional relationships among populations are adequately represented in two dimensions judging by the magnitudes of their respective eigenvalues; the first two dimensions cumulatively explain 83.0% of the total morphometric variance. Principal component I (PC I) is a general size factor because all characters load positively and highly (with the exception of MAND at 0.385), whereas PC II is a shape factor involving high positive loadings for wing bone characters (HUML and ULNA) and moderate negative loadings for all skull characters (PREW, CRAD, CRAL, and MAND). Thus the New Zealand chaffinches are intermediate in size between the larger northern European birds and the small Iberian birds, as shown by their intermediate position along PC I. The New Zealand populations are morphometrically very similar in size and shape to the southern English population from Wareham (WARE). The divergence of the New Zealand populations from the remaining European samples is also apparent on PC II, mainly reflecting a contrast between the larger mean skull size and shorter wings of the New Zealand populations relative to the Iberian population means. Although the Iberian populations are differentiated from most other European populations in the principal component plot, they are not morphometrically disjunct because the French population in Hourtin (HOUR) to the north of the Pyrenees mountains is intermediate in size and shape between the two regions.

The minimum spanning tree emphasizes the morphometric intermediacy of the New Zealand populations relative to the broader range of variation present among European populations. Both the southern English population (WARE) and the Scottish population (PITL) are most closely allied with New Zealand populations, and the French populations from Hourtin (HOUR) and St. Quentin (QUEN) link the New Zealand chaffinches to the smaller Iberian and larger northern European birds respectively.

The amount of morphometric differentiation that has developed among the New Zealand populations relative to that in northern Europe and all Europe is summarized as percent added variance components in Table 5. With the exception of two characters (CRAD and SACW), the New Zealand populations are only weakly differentiated compared to the northern European populations; the average added variance component over all 12 skeletal characters is only 2.8% for New Zealand whereas for northern Europe it is 18.7%. When Iberian populations are included in the comparison, the average added variance component in the European populations is increased to 24.7%. The restricted amount of differentiation in the New Zealand populations relative to that in Europe is clearly depicted multivariately in Figure 3 where the New Zealand populations are tightly clumped together compared to the broader scatter among their European counterparts.

Rate Tests of Morphometric Divergence of New Zealand Populations

If divergence of the means of morphometric characters of descendent New Zealand populations is genetically based, it is possible to employ a test of evolutionary rates expected under simple neutral models. Because the introduction of chaffinches into New Zealand was recent, it is inappropriate to use mutation-drift models that require the initial genetic variance to have reached mutation-drift equilibrium. I have instead used Lande's (1976) constant heritability analysis, which is recommended when the number of generations t < [N.sub.e]/5 (Turelli et al., 1988). Population sizes of chaffinches in New Zealand over the 120 generations since introduction are not known, but for noncolonial birds [N.sub.e] is typically [10.sup.2] to [10.sup.3] (Barrowclough, 1980), and thus the model seems appropriate.

Lande's (1976) model tests the hypothesis that change in the mean phenotype for a particular character in a single population is caused by random drift. The use of Wareham as the "ancestral" or t(o) population is reasonable because the principal component analysis (Fig. 3) demonstrated that the New Zealand populations are all closely clustered morphometrically near the Wareham population centroid in multivariate space, and historical records indicate that at least some of the birds taken to New Zealand were captured in southern England (Sussex) (Ince et al., 1980).

With Lande's (1976) rate test, drift can be rejected at the 5% level as the cause of divergence of mean phenotypes over 120 generations if [Mathematical Expression Omitted] where [h.sup.2] is the heritability of the trait under study, z is the difference in mean phenotypes, and a is the standard deviation of the trait. For morphometric characters of birds, heritabilities typically range between 60% and 70% (Boag and Van Noordwijk, 1987), and I have used the lower figure to produce a conservative estimate of [Mathematical Expression Omitted]. To obtain a more reliable estimate of [Sigma], pooled standard deviations were computed over all New Zealand populations. Estimates of [Mathematical Expression Omitted] range from 660 for cranium length (CRAL) to 118,520 for ulna length (ULNA). [Mathematical Expression Omitted] > [N.sub.e] for all skeletal characters with the possible exception of CRAL, and random drift cannot be rejected as the cause of the morphometric divergence of the descendent New Zealand birds from the "ancestral" southern English birds. A necessary caveat here is that Lande's test is approximate and is best considered as an exploratory tool, and failure to reject the null hypothesis does not definitively rule out selection as the cause of divergence of ancestral and descendent populations. However, such selection would be very weak and indistinguishable from the effects of drift.

Association of Genetic and Morphometric Divergence

Mantel's test using Rogers' genetic distances and average taxonomic distances computed from the means of morphometric characters was carried out to determine whether genetic and morphometric divergence among European populations were significantly associated, and for comparison with equivalent tests previously conducted on New Zealand populations (see Baker et al., 1990b). Using the normalization procedure in Smouse et al. (1986), genetic and morphometric distances among European populations are significantly though not highly correlated (Z = 0.380, P = 0.034). This correlation derives principally from the Iberian populations because they are the most differentiated in either distance matrix. When the analysis is confined to the northern European populations the association is negative and not significant (Z = - 0.243, P = 0. 149).

Further Mantel's tests were conducted to ascertain whether either of the above matrices was associated with the geographic proximity of populations. To approximate likely routes of gene flow, geographic distances among populations were defined as the edge lengths in a Gabriel connected graph among sampling localities. Neither genetic nor morphometric distances among European populations are significantly correlated with geographic distances between sampling localities (Z = 0.100, P = 0.304, and Z = -0.070, P = 0.327, respectively). Similarly, genetic and morphometric distances among the New Zealand populations are not significantly correlated (Z = -0.206, P = 0.315), and they are also not associated with the geographic proximity of populations (Z = 0.248, P = 0.287, and Z = -0.023, P 0.910, Baker et al., 1990b).

Population Divergence in Europe in Relation to Environmental Variation

Using the Bonferroni method to adjust significance levels for multiple comparisons (P = 0.05/25 comparisons = 0.002), only two of 25 regressions of allele frequencies al geographically variable loci in Europe were significant. Variation at PepB(A) was successfully predicted by mean January and July temperatures (r 2 = 0.74 each). For the morphometric characters, three of 60 regressions were significant at P = 0.0008. Variation in the mean length of the premaxilla (PREW), cranium (CRAL), and mandible (MAND) was predicted by AE ([r.sup.2] = 0.82, 0.76, and 0.79 respectively). Although these regressions could also be attributed to chance, it is noteworthy that regressions involving other characters [tarsometatarsus length (TARS), and synsacrum length and width (SACL and SACW)] on AE and mean January temperature also approach significance (P < 0.01). More compellingly, multivariate body size as represented by PC I was predicted by AE ([r.sup.2] = 0.79, see Fig. 4).

When the analyses were repeated within northern Europe, none of the regressions were significant for either allele frequencies or morphometric characters. However, PC I was significantly related to AE (P = 0.034, [r.sup.2] = 0.63), indicating that interlocality body size differences in northern Europe are also responding to productivity gradients in this region. Most notably, the largest mean body size and highest AE values are found at localities at opposite ends of the sampling range, Chartreuse (CHAR) in southern France and Kalo (KALO) in the Jutland Peninsula of Denmark.

DISCUSSION

Magnitude of Divergence in European and New Zealand Populations

The magnitude of divergence among all 10 European populations of chaffinches sampled in this study is considerably greater than that among New Zealand populations in both protein-encoding genes and morphometrics. Genetic subdivision between northern European and Iberian populations is quite pronounced, as indicated by the [F.sub.st], of 0.222. However, when the three Iberian populations are removed from the analysis, the [F.sub.st] drops to 0.032 for northern European "ancestral" populations, similar to the level of very weak structuring that is evident among extant descendent New Zealand populations approximately 120 years after they were founded. The Pyrenees mountains between Iberia and France are a major barrier to gene flow among European chaffinch populations, as they are for Iberian and northern European populations of the chiff-chaff (Phylloscopus collymbita) (Salomon, 1987, 1989).

The European populations of chaffinch also exhibit about nine-fold greater morphometric divergence than the New Zealand populations. Unlike the genetic analysis, however, the "ancestral" northern European populations also display a six-fold greater level of morphometric divergence than their descendent New Zealand counterparts.

To calibrate the levels of divergence in the ancestral northern European populations relative to those developed in the New Zealand populations in about 120 generations, it is desirable to have an estimate of the time frame over which differentiation in the ancestral populations has developed. The percent sequence divergence in mitochondrial DNA between Iberia and northern Europe (corrected for within-region variation) is 0.01 (Baker, unpubl. data), and if we accept a constant rate of molecular evolution of 2% per million years for mtDNA in birds (Shields and Wilson, 1987; Shields and Helm-Bychowski, 1988), then the time of divergence is estimated to be approximately 5,000 years. This estimate is in accord with colonization following the last retreat of glacial ice about 20,000 years ago (Seret et al., 1990). Historical evidence also indicates that birds such as the house sparrow colonized Britain from Europe at least 2,000 years ago (Parkin and Cole, 1985).

Processes of Divergence

The haphazard pattern of population differentiation in New Zealand does not fit geographically ordered patterns such as clines or isolation-by-distance, and is not aligned with environmental gradients. This suggests that gene flow is restricted by geographic barriers and discontinuous habitat, and that there has not yet been selection for climatic adaptation or nonselective environmental induction. Thus Baker et al. (1990b) concluded that random drift was the likely cause of this variation. Rate tests of the divergence of means of morphometric characters of New Zealand populations from those of an extant population in southern England representative of the founding stock are also compatible with sampling drift as an explanation for the observed small morphometric shifts.

The greater degree of morphometric evolution but lack of genetic structuring in the "ancestral" northern European populations most probably reflect the different evolutionary forces acting on morphometric traits and protein loci over thousands of years. Electrophoretically detectable variation in protein loci in birds is very likely neutral or nearly so (Barrowclough et al., 1985; Baker and Moeed, 1987), and analyses of genetic variation in chaffinches populations in various parts of the world support this conclusion (Baker et al., 1990a). Thus the small scale genetic differentiation among "ancestral" northern European populations of chaffinches suggests strongly that homogenizing gene flow is preventing divergence of neutral genes in these geographically connected demes.

Using Wright's (1969) formula [F.sub.st] = 1/(1 + 4Nm), gene flow among northern European populations appears to be extensive (Nm = 7.6). Conversely, gene flow between Iberian and northern European populations is restricted (Nm = 0.9). Although these estimates are based on an island model of population structure, they have also been found to be reasonable approximations for the opposite extreme of stepping-stone arrays (Crow and Aoki, 1984; Slatkin, 1987). Population structure in chaffinches lies between these extremes, with only the Iberia-northern Europe split approaching a two-dimensional stepping stone. Additionally, these estimates assume that the elapsed time since colonization is sufficient for an equilibrium to have been attained between migration and drift. [F.sub.st] is expected to approach its equilibrium value rapidly (Crow and Aoki, 1984), unless strong founder effects have caused the initial divergence among demes. The latter seems unlikely, given current high levels of heterozygosity in populations.

The lack of genetic structuring within regions in Europe could be attributable partly to their shared ancestral polymorphisms, with insufficient time elapsing since their separation for detection of differences in slowly evolving allozymes (Larson et al., 1984, Zink, 1991). One way to estimate levels of gene flow in continental demes at equilibrium is to consider only the Iberian Peninsula because this region of Europe was not glaciated and potentially could have supported populations of chaffinches over a much longer period than northern Europe. Based on an [F.sub.st] of 0.020, Iberian populations experience high gene flow (Nm = 12.2). Values of Nm clearly exceed those necessary for independent divergence of demes under either the island model (Nm < 1) or in a stepping-stone array (Nm < 4, Nagylaki, 1983). It also seems reasonable to conclude that northern European populations are subject to homogenizing gene flow that is sufficient to prevent genetic differentiation via random drift.

The haphazard pattern of among-population differentiation in skeletal characters in New Zealand cannot be simply extrapolated through time to account for the pattern of geographic variation in Europe. Although geographic variation in skeletal characters is not clinal in Europe, it is aligned with temperature and actual evapotranspiration. Actual evapotranspiration in the breeding season, a measure of environmental productivity at breeding sites (Rosenzweig, 1968; Ricklefs, 1980), is a good predictor ([r.sup.2] = 0.77) of general body size (as measured multivariately by PC I) in adult male chaffinches in Europe.

The pattern of morphometric differentiation in Europe is consistent with either genetically-based adaptation, or with ecophenotypic variation in which larger birds occur in more productive localities. Distinguishing between these two alternatives requires knowledge of the relative magnitudes of the additive genetic variance and environmental variance of body size in chaffinches. For example, James (1983) has demonstrated with egg transplant experiments that body size differences among geographically separated populations of red-winged blackbirds (Agelaius phoeniceus) have a significant environmental component. Comparative studies have shown that body size differences among populations of pocket gophers (Thonomys bottae) are strongly associated with differences in the quality of forage (Patton and Brylski, 1987). Conversely, Alatalo and Gustafsson (1988) found no among-population environmental component to tarsus length in coal tits (Parus ater).

If body size variation in chaffinches has a large environmental component, then geographic variation aligned with productivity gradients would arise rapidly in colonizing populations. This expectation is not met in New Zealand (Baker et al., 1990b). It therefore seems likely that the greater degree of morphometric differentiation in the European populations of chaffinches relative to the descendent New Zealand populations reflects longer term adaptive differentiation via selection for optimal body size. Development of such morphometric differentiation in Europe but not New Zealand indicates that local adaptation evolves slowly in chaffinches because selective forces are opposed by homogenizing gene flow. As noted by Wright (1980), multilocus evolution would be expected to be slow in populations subject to these opposing forces, and adaptive phenotypic divergence may still be evolving in northern Europe.

ACKNOWLEDGMENTS

I am grateful to Dr. G. Hemery, S. Eis, Dr. L. Batten, A. Lopez Lillo, T. de Azcarate y Bang, Dr. A. Teixeria, and B. Bell for permission to collect and export chaffinches from Europe and New Zealand. For assistance in the field or with local arrangements I thank M. Peck, M. Goldsmith, G. Le Grand, K. Bertelsen, M. Dennison, K. Hansen, A. Lynch, E. Masters, T. W. MacMillan, J. Goss-Custard, D. Parkin, and N. Simonsen. I am indebted to M. Dennison for measuring all skeletons and for suggesting the use of productivity measures. I thank R. Rockwell and R. Zink for constructive criticisms of an earlier version of the manuscript, and A. Lynch for assistance with statistical analysis and computer programming. Expert laboratory assistance was provided by M. Peck and S. Kingsley. This research was made possible by funds from the Natural Sciences and Engineering Research Council of Canada (grant A200) and the Royal Ontario Museum, for which I am most grateful.

LITERATURE CITED

Alatalo, R. V., and L. Gustafsson. 1988. Genetic component of morphological differentiation in coal tits under competitive release. Evolution 42:200-203. Baker, A. J., and A. Moeed. 1987. Rapid genetic differentiation and founder effect in colonizing populations of common mynas, (Acridotheres tristis). Evolution 41:525-538. Baker, A. J., M. D. Dennison, A. Lunch, and G. Le Grand. 1990a. Genetic divergence in peripherally isolated populations of chaffinches in the Atlantic islands. Evolution 44:981-999. Baker, A. J., M. K. Peck, and M. A. Goldsmith. 1990b. Genetic and morphometric differentiation in introduced populations of common chaffinches (Fringilla coelebs) in New Zealand. Condor 92:76-88. Barrowclough, G. F. 1980. Gene flow, effective population sizes, and genetic variance components in birds. Evolution 34:789-798. Barrowclough, G. F., and K. W. Corbin. 1978. Genetic variation and differentiation in the Parulidae. Auk 95:691-702. Barrowclough, G. F., N. K. Johnson, and R. M. Zink. 1985. On the nature of genetic variation in birds, pp. 135-154. In R. F. Johnston (ed.), Current Ornithology, Vol. 2. Plenum Press, N.Y., USA. Barton, N. H., and B. Charlesworth. 1984. Genetic revolutions, founder effects and speciation. Annu. Rev. Ecol. Syst. 15:133-164. Boag, P.T., and A. J. Van Noordwijk. 1987. Quantitative genetics, pp. 45-78. In P. A. Buckley and F. Cooke (eds.), Avian Genetics. Academic Press, London, UK. Carson, H. L., and A. R. Templeton. 1984. Genetic revolutions in relation to speciation phenomena: The founding of new populations. Annu. Rev. Ecol. Syst. 15:97-131. Charlesworth, B., R. Lande, and M. Slatkin. 1982. A neo-Darwinian commentary on macroevolution. Evolution 36:474-498. Cole, S. R., and D. T. Parkin. 1981. Enzyme polymorphisms in the house sparrow, Passer domesticus. Biol. J. Linn. Soc. 15:13-22. Crow, J. F., and K. Aoki. 1984. Group selection for a polygenic behavioral trait: Estimating the degree of population subdivision. Proc. Natl. Acad. Sci. USA 81:6073-6077. Eldredge, N., and S. J. Gould. 1972. Punctuated equilibria: An alternative to phyletic gradualism, pp. 82-115. In T. M. Schopf (ed.), Models in Paleobiology. Freeman, San Francisco, CA USA. Gould, S. J. 1982. Darwinism and the expansion of evolutionary theory. Science 216:380-387. Gower, J. C. 1966. Some distance properties of latent root and vector methods in multivariate analysis. Biometrika 53:325-338. Hapius, H., and D. A. Hopkinson. 1976. Handbook of Enzyme Electrophoresis in Human Genetics. North Holland Publishing Co., Amsterdam, The Netherlands. Ho, M. W., and P. T. Saunders. 1979. Beyond neo-Darwinism -- An epigenetic approach to evolution. J. Theor. Biol. 78:573-591. Ince, S. A., P. J. B. Slater, and C. Weisman. 1980. Changes with time in the songs of a population of chaffinches. Condor 82:285-290. James, F. C. 1983. Environmental component of morphological differences in birds. Science 221:184-186. Lande, R. 1976. Natural selection and random genetic drift in phenotypic evolution. Evolution 30: 314-334.

Larson, A., D. B. Waye, and K. P. Yanev. 1984. Measuring gene flow among populations having high levels of genetic fragmentation. Genetics 106:293-308. Nagylaki, T. 1983. The robustness of neutral models of geographical variation. Theor. Popul. Biol. 74:268-294. Nei, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89:583-590. Parkin, D. T., and S. R. Cole. 1985. Genetic differentiation and rates of evolution in some introduced populations of the house sparrow, Passer domesticus in Australia and New Zealand. Heredity 54:15-23. Patton, J. L., and P. V. Brylski. 1987. Pocket gophers in alfalfa fields: Causes and consequences of habitat-related body size variation, Am. Nat. 130: 493-506. Ricklefs, R. E. 1980. Geographical variation in clutch size among passerine birds: Ashmole's hypothesis. Auk 97:38-49. Rogers, J. S. 1972. Measures of genetic similarity and genetic distance. Univ. Tex. Publ. 7213:145-153. Rosenzweig, M. L. 1968. Net primary productivity of terrestrial communities: Prediction from climatological data. Am. Nat. 102:67-74. Salomon, M. 1987. Analyse d'une zone de contact entre deux formes parapatriques: cas des Pouillots veloces Phylloscopus collybita collybita et P. c. brehemii. Rev. Ecol. Terre Vie 42:377-420. --. 1989. Song as a possible reproductive isolating mechanism between two parapatric forms. The case of the chiffchaffs Phylloscopus c. collybita and P. c. brehmii in the western Pyrenees. Behaviour 111:270-290. Seret, G., E. Dricot, and G. Wansard. 1990. Evidence for an early glacial maximum in the French Vosges during the last glacial cycle. Nature 346: 453-456. Shields, G. F., and K. Helm-Bychowski. 1988. Mitochondrial DNA of birds, pp. 273-295. In R. F. Johnston (ed.). Current Omithology, Vol. 5. Plenum Press, N.Y., USA. Shields, G. F., and A. C. Wilson. 1987. Calibration of mitochondrial DNA evolution in geese. J. Mol. Evol. 24:212-217. Slatkin, M. 1987. Gene flow and the geographic structure of natural populations. Science 236:787-792. Smouse, P. E., J. C. Long, and R. R. Sokal. 1986. Multiple regression and correlation extensions of the Mantel test of matrix correspondence. Syst. Zool. 35:627-632. Sneath, P.H.A., and R.R. Sokal. 1973. Numerical Taxonomy. Freeman, San Francisco, CA USA. Swofford, D. L., and R. B. Selander. 1981. BIOSYS-1: A FORTRAN program for the comprehensive analysis of electrophoretic data in population genetics and systematics. J. Hered. 72:281-283. Thomson, G. M. 1922. The naturalisation of animals and plants in New Zealand. Cambridge University Press, London, UK. Thornwaite, C. W., and J. R. Mather. 1957. Instructions and tables for computing potential evapotranspiration and the water balance. Publs. Climat. 10: 181-311. TURNER, J. R. G. 1988. The evolution of mimicry: A solution to the problem of punctuated equilibrium. Am. Nat. 131:S42-S66. Turelli, M., J. H. Gillespie, and R. Lande. 1988. Rate tests for selection on quantitative characters during macroevolution and microevolution. Evolution 42:1085-1089. Vithayasai, C. 1973. Exact critical values of the Hardy-Weinberg test statistic for two alleles. Commun. Stat. 1:229-242. Wernstadt, F. L. 1972. World Climatic Data. Climatic Data Press, Lemont, USA. Workman, P. L., and J. D. Niswander. 1970. Population studies of Southwestern Indian tribes. II. Local genetic differentiation in the Papago. Am. J. Hum. Genet. 22:24-49. Wright, S. 1965. The interpretation of population structure by F-statistics with special regard to systems of mating. Evolution 9:395-420. --. 1969. Evolution and the Genetics of Populations. Vol. 2. University of Chicago Press, Chicago, IL USA. --. 1978. Evolution and the Genetics of Populations. IV. Variability within and among Natural Populations. University of Chicago Press, Chicago, IL USA. --. 1980. Genic and organismal selection. Evolution 34:825-843. Zink, R. M. 1991. The geography of mitochondrial DNA variation in two sympatric sparrows. Evolution 45:329-339.
COPYRIGHT 1992 Society for the Study of Evolution
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Baker, Allan J.
Publication:Evolution
Date:Dec 1, 1992
Words:6464
Previous Article:Species isolation, genital mechanics, and the evolution of species-specific genitalia in three species of Macrodactylus beetles (Coleoptera,...
Next Article:Algae containing chlorophylls a + c are paraphyletic: molecular evolutionaryanalysis of the Chromophyta.
Topics:

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |