Covariation Of Song, Morphological, And Allozyme Frequency Characters In The Rufous-collared Sparrow, Zonotrichia Capensis

In the Rufous-collared Sparrow (Zonotrichia capensis) song dialects, defined by rate of delivery of the terminal trill, are clearly associated with variation in vegetation. The local adaptation hypothesis suggests that these dialects may act as markers for different populations adapted to different local environmental conditions predicting that there will be a correlation between the patterns of vocal and genetic variation. Thus, covariation of allozyme frequencies, and morphological (external and skeletal) measures and song characters was investigated in northwestern Argentine populations of Rufous-collared Sparrows with particular reference to possible correlations between vocal dialects and population structure. Approximately 20 males were collected from each of four sites within each of six different vegetations (and thus, six vocal dialects): lowland chaco thornscrub, transition forest, montane woodland, montane grassland, Monte desert scrub, and puna high altitude scrub.;There was significant variability in both external and skeletal morphology among all 24 sites and among vegetation/dialect populations. Overall Wright\u27s corrected inbreeding coefficient (F{dollar}\sb{lcub}\rm ST{rcub}{dollar}) was 0.119 indicating significant genetic differentiation among sites within the study area. Hierarchical Wright\u27s F statistics indicated that only 50% of among site variability was due to a vegetation/dialect effect.;Puna scrub sites were differentiated from all other sites with respect to both morphology and allozyme frequencies. Heterogeneity at the PGM-1 locus among puna scrub sites was the major cause of the high F{dollar}\sb{lcub}\rm ST{rcub}{dollar} value overall and within puna scrub vegetation (F{dollar}\sb{lcub}\rm ST{rcub}{dollar} = 0.156). Puna scrub populations have traditionally been treated as a subspecies (pulacayensis) separate from other populations within the study area (hypoleuca) and my data support this contention.;Genetic differentiation overall among all non-puna sites (corrected F{dollar}\sb{lcub}\rm ST{rcub}{dollar} = 0.018) was similar to differentiation among sites within each of the five non-puna vegetations (average corrected F{dollar}\sb{lcub}\rm ST{rcub}{dollar} = 0.0132 {dollar}\pm{dollar} 0.0069). Hierarchical Wright\u27s F statistics indicated that none of the among site differentiation in this subset of samples was due to a vegetation/dialect effect. These observations are not consistent with the local adaptation hypothesis.;Overall covariation between morphology and allozyme frequencies was evident in that puna scrub populations were distinct from all other sampled populations. Mantel\u27s non-parametric tests based on non-puna sites indicated that there was no covariation between morphological Mahalanobis and Rogers\u27 genetic distance matrices. Degree of morphological differentiation was associated with both geographic distance and differences between habitats. There was no significant covariation between genetic distance and geographic distance suggesting no simple isolation-by-distance effect. Finally, there was no relationship between song Mahalanobis distances (based on frequency and temporal characters of song) and either morphology or habitat structure.;All significant genetic heterogeneity occurred among sites in mountainous habitats and I conclude that topography and patchiness of habitat were probably major factors involved in population differentiation, rather than vocal dialects