Search

Morphometric characters are congruent with cytogenetic and genic evidence in dividing Peromyscus boylii into three groups in east-central Mexico. These include a monomorphic, low-FN group (FN = 52); a polymorphic, low-FN group (FN = 52–54); and a polymorphic, high-FN group (FN = 58–60). Univariate and multivariate analyses of morphometric features discriminate the three groups, with the overall pattern more evident in the multivariate than the univariate approach. Some of the qualitative characters show the monomorphic low-FN specimens to be distinct from the karyotypically polymorphic groups, but none of them is diagnostic. Morphometric comparisons of the three karyotypic groups with topotypic and reference samples of the recognized taxa of P. boylii in Mexico support the following taxonomic assignments for these forms: monomorphic, low-FN group, P. boylii rowleyi ; polymorphic, low-FN group, P. beatae ; and the polymorphic, high-FN group, P. levipes . The three chromosomal groups appear to represent a good example of cryptic species.

Fifteen environmental variables of potential biological importance are studied. These variables quantify productivity, altitude, topographic heterogeneity, and a variety of climatic dimensions. Most of these variables are not normally distributed, according to the Kolmogorov‐Smirnov One‐Sample Test. Spearman's coefficient of rank correlation indicates that many of them are highly and significantly correlated. Principal components analysis reveals a gradient of increasing temperature from north to south and an orthogonal gradient of increasing precipitation across the state. Based on these vectors four climatic regions (warm‐moist, warm‐dry, cool‐moist, cool‐dry) are recognized in Texas. An interpretation of the biological meaning of the variables is presented. These data are of potential used as a resource base for studies of relationships between environmental factors and biological variation.

A zone of contact among three chromosomal races of Geomys bursarius , designated E (2N = 74, FN = 72), F (2N = 70, FN = 74), and G (2N = 70, FN = 72), was located in Burleson Co., Texas. Interactions between race E and races F and G were assessed. Differences in diploid and fundamental numbers between these two groups were sufficient to detect many potential backcross and F2 hybrid individuals. Of 42 gophers collected in the contact zone, 31% had an apparent F1 karyotype (some of these may be unrecognized backcross and F2 individuals). No karyotypically detectable backcross or F2 individuals were found except in the fetal state. Race E and races F and G could not be separated morphologically, although two general groups were evident. The F1 hybrids were not morphologically intermediate between race E and races F or G, but were more similar to the latter pair of races than to the former race. No ecological differences with regard to soil preferences were found among the three races and hybrids. The limited size of the contact zone and the apparent lack of backcrossing suggest that race E and races F and G interact as biological species where they contact. The absence of any detectable premating isolating mechanisms may be correlated with the recency of the contact zone. It is suggested that races F and G have invaded the existing range of race E, possibly as a consequence of the effects of man-made disturbances such as agriculture and the building of road rights-of-way.

Nongeographic and geographic variation is assessed in the desmarestianus species group of the genus Heteromys (exclusive of H. gaumeri ). Concordant results of cranial, external, and bacular variation reveal that the desmarestianus -group is represented in northern Middle America by two species, H. desmarestianus and H. goldmani . H. longicaudatus and H. lepturus are indistinguishable from H. desmarestianus . H. temporalis is arranged as a subspecies of H. desmarestianus .

Morphological and habitat variation was studied in four populations of the kangaroo rat ( Dipodomys ordii ) in western Texas. Twenty-six external, cranial, and ratio characters were analyzed in 536 specimens. The number of characters exhibiting secondary sexual dimorphism varied from 14 at Tahoka to five at Canadian and Kermit. In the two northern populations (Canadian and Needmore) females were larger than males, whereas males were larger than females in the southern populations (Tahoka and Kermit). Many of the characters studied had low coefficients of variation (CV) for mammals. Least supra-occipital width, however, showed an abnormally high CV, not only in D. ordii but also in three other species of this genus ( D. merriami, D. spectabilis , and D. deserti ). Despite large sample sizes and precautions against sampling bias, many variates did not fit a normal distribution but were skewed. Large correlation coefficients between certain variates suggest many of the characters studied represent a single adaptive complex and should not be considered independently. Factors that may have influenced the existing evolutionary diversity are discussed.

The karyotype of Peromyscus boylii attwateri is contrasted with that of P. b. boylii, P. b. rowleyi , and P. b. utahensis . The latter three taxa have identical karyotypes. Chromosomal and morphological evidence indicate that attwateri is specifically separated from the boylii-rowleyi-utahensis assemblage. Limited samples from Mexican populations of P. boylii imply that current taxonomic alignments in this species require reevaluation. The karyotype of Peromyscus difficilis comanche is described and it appears to be identical to that of Peromyscus truei truei and unlike those of northern populations of P. difficilis .

Both nongeographic and geographic variation was assessed among populations of Peromyscus boylii from northern Mexico and the southern United States. Geographic variation was analyzed using a numerical phyletic approach developed by Eades (1970). Analysis of morphological features of the skin, skull, and molars reveals that Peromyscus attwateri is a separate species from the other taxa of the boylii complex. Morphologic characteristics presented herein parallel cytological differences reported by Lee et al. (1972).

Morphological variation of the South American lizard genus Kentropyxis discussed, with particular emphasis on the Amazonian distributed calcarata species group. Ecological and zoogeographical data are analyzed. Analysis of geographic variation of meristic characters, in con- junction with color and striping pattern variation, supports the recognition of three Amazonian species of Kentropyx with small dorsal scales. A dramatic morphological change is noted for Ken- tropyx calcarata, and it is suggested that the morphological divergence might be an example of character displacement.

The taxonomic status of Peromyscus comanche Blair 1943 is herein taxonomically reviewed. The karyotype of P. comanche is given and compared with Peromyscus truei truei and Peromyscus difficilis nasutus. Morphological comparisons are made between a sample of P. comanche and one of P. truei from New Mexico. Chromosomal and morphological evidence indicate the affinities of P. comanche are with P. truei and not with P. difficilis, as previously suspected. The difficulties in classifying alloptaric populations are discussed. Peromyscus comanche is considered a subspecies of P. truei (Peromyscus truei comanche). Peromyscus comanche is one of three nominal species of the Peromyscus truei species group occurring in the southwestern United States. Peromyscus difficilis, the juniper mouse, occurs in Colorado, New Mexico, Utah, Arizona, Texas, and extreme western Oklahoma, and thence south into Mexico. Peromyscus truei, the pinon mouse, has a wider distribution extending from central Oregon south to Baja California, eastward to extreme western Oklahoma, and thence south to Oaxaca. The third species, P. comanche, occupies a restricted range along the rocky, cedar covered slopes along the "Break of the Plains" in western Texas (Blair, 1943). P. difficilis and P. truei are ecologically, altitudinally, and geographically sympatric over much of their range in the western United States. P. comanche is isolated by several miles of unsuitable habitat from the nearest known population of other mice of the truei group in New Mexico and Oklahoma. The three species are difficult to distinguish on the basis of morphology and are considered by some authors (notably Blair, 1943 and Tamsitt, 1961a) to be "sibling species." Blair (1943) described P. comanche as a full species closely related to Peromyscus nasutus (now considered a subspecies of difficilis). Since Blair's description of comanche, the taxonomic status of the 1 Contribution No. TA 9988, Texas Agricultural Experiment Station, Texas A&M University.