The southwestern Phlox species P. dolichantha, P. superba, P. stansburyi, and P. grayi are a group of morphologically similar species that differ notably in the length of their corolla tubes (approximately 35 to 45 mm, 26 to 33 mm, 19 to 25 mm, and 13 to 16 mm, respectively). We assessed insect visitation to these 4 species over 24-h periods during peak flowering. A diverse array of visitors was documented, including members of the orders Lepidoptera, Diptera, Hymenoptera, and Coleoptera. Although Phlox is generally considered to exhibit a lepidopteran pollination syndrome, all of these species experienced some visitation from non-lepidopterans. Phlox superba and P. grayi received greatest visitation from a hawkmoth (Hyles lineata), and P. dolichantha and P. stansburyi received the most visits from small flies (Anthomyia and Tachypeza, respectively). All Phlox species were visited by lepidopterans, and the hawkmoth species H. lineata was observed visiting all 4 species. Bees were only observed visiting the shorter-tubed P. stansburyi and P. grayi, although bees were present among populations of all 4 species. Our findings suggest that these Phlox species might be more generalist than previously hypothesized and that further investigation into plant-pollinator interactions in the group might contribute to our understanding of their floral evolution.

Since its introduction into the United States, the imported red fire ant (Solenopsis invicta Buren), a freeze-intolerant species, has extended its range northward and now occupies areas where winter temperatures are consistently below freezing. The objective of our study was to determine if more northern S. invicta had greater supercooling capability than more southerly S. invicta. Such information would help to delineate the northern boundary of range expansion. We collected S. invicta colonies in early summer, late summer, and early spring from 3 locations along a latitudinal gradient in Texas. Supercooling point determinations were made immediately after colonies were collected and after colonies were acclimated in the laboratory for 1 and 2 months. Regardless of time or acclimation period, the northernmost colonies (−5.91 ± 0.21°C) had higher mean supercooling points than the southernmost colonies (−16.49 ± 0.11°C). Colonies from central Texas had intermediate mean supercooling points (−10.40 ± 0.19°C). This pattern is difficult to explain. Tests from a broader geographic range might allow us to better assess supercooling ability in S. invicta. Additionally, we tested whether supercooling points were affected by body size. We found no relationships between supercooling points and head capsule widths. However, there was little variation in head capsule widths of the minor workers that we measured. Differences in supercooling points might be evident only on a larger scale of body size.

Africanized honeybees, hybrids between African (Apis mellifera scutellata) and western European (A. m. mellifera) and eastern European (A. m. caucasica, A. m. carnica, and A. m. ligustica) subspecies are widely distributed in urban areas of the southwestern USA. However, little is known about their distribution in rural regions. We collected bees at 54 sites in a 5,350-km² study area in the Sonoran Desert of southwestern Arizona. We used mitochondrial DNA (mtDNA) analysis of individual worker honeybees (10 per site) to assess genetics of colonies within the study area. Among collected bees, 86.9% possessed African mtDNA. Western European, eastern European, and Egyptian (A. m. lamarckii) mtDNA was present in 5.6%, 4.1%, and 3.4% of collected bees, respectively. There was no apparent relationship between the percentage of bees having African mtDNA and distance to agricultural fields or elevation of the collection site. The preponderance of Africanized honeybees confirms previous studies and predictions concerning their distribution in the southwestern USA.

A new species of Gyrodactylus (Monogenoidea, Gyrodactylidae) is described from the skin of the White Sands pupfish, Cyprinodon tularosa Miller and Echelle, 1975 (Cyprinodontiformes, Cyprinodontidae) from Salt Creek, White Sands Missile Range, Tularosa Basin, New Mexico. The new species is compared with the 6 species of Gyrodactylidae known to parasitize pupfish in the southern United States.

Laguna de Sayula, a seasonal freshwater wetland in Jalisco, western Mexico, supports many resident and migratory bird species, particularly during the dry season (October to May). We documented changes in habitat condition (i.e., water vs. mudflat vs. land cover) and occurrences of 12 waterbird species during the dry season. From October to February, much of the lagoon was covered with water and mudflats, and the avifauna was dominated by resident killdeer (Charadrius vociferus) and migratory long-billed curlew (Numenius americanus). Between March and May, there was significantly less area covered by water and mudflats, and killdeer and the migratory spotted sandpiper (Actitis macularia) were most common. Species tended to co-occur in plots, often in large numbers, suggesting that they were tracking patchy resources within the study area.

We studied home range, habitat use, and vital rates of radio-marked Mexican spotted owls (Strix occidentalis lucida) in 2 study areas in the Sacramento Mountains, New Mexico. One study area (mesic) was dominated by mixed-conifer forest, the other (xeric) by ponderosa pine (Pinus ponderosa) forest and piñon (P. edulis)-juniper (Juniperus) woodland. Based on existing knowledge of relative use of forest types by Mexican spotted owls, we predicted that the mesic area would provide habitat of higher quality for spotted owls. Results generally supported this prediction. Median home-range size for owls in the mesic area was approximately half that of owls in the xeric area during both the breeding and non-breeding seasons (n = 6 owls in each area). Despite their reduced size, however, mesic-area home ranges contained twice as much mixed-conifer forest as xeric-area ranges. Owls roosted primarily (>80% of roosting locations in both seasons) in mixed-conifer forest in both study areas, and home-range size was inversely related to relative amount of mixed-conifer forest within the home range during both seasons. Both survival and fecundity rates were higher in the mesic than in the xeric area. Estimates of population trend based on observed vital rates suggested that the population in the mesic area was self-sustaining or nearly so during the period of study (1992 through 1994), but the population in the xeric area was not. Collectively, our findings suggest that habitat quality for spotted owls was higher in the mesic area than in the xeric area, and that the xeric area might function as an ecological sink. These results support the need for data linking demographic performance to habitat conditions in development of strategies for recovering threatened and endangered species.

Geomys texensis, the Llano pocket gopher, consists of 3 subspecies, all identified using classical morphological methods. Enzymatic studies have found little support for the existence of one of these subspecies (G. t. llanensis); however, it continues to be recognized regularly today. To clarify the subspecific status of members within this species, we have sequenced the entire 1,140 base pairs of the cytochrome b gene for 25 individuals collected from throughout the range of the species. Our data refute the designation of 1 of the 3 currently recognized subspecies and allude to the existence of a new, as yet unnamed subspecies.

Nucleotide sequences from intron 7 of the beta-fibrinogen gene (Fgb-I7) were used to evaluate phylogenetic relationships among members of the genus Sigmodon. In addition, these sequences were combined with mitochondrial cytochrome-b sequences and analyzed from a total evidence perspective. Results from parsimony and Bayesian analyses indicated support for 3 species groups (alstoni, fulviventer, and hispidus). A sister relationship was depicted for S. alleni and S. hirsutus, followed by the addition of S. toltecus; however, little resolution was provided for relationships between other members of the hispidus species group.

Seventy-six American mink (Mustela vison) from 8 localities were genotyped using 4 polymorphic microsatellite loci. This study was designed to test the hypothesis that genetic structure was partitioned within streams due to the close association of mink with riparian habitats. Six of the localities were streams in Lonoke County in central Arkansas, one was in Independence County, 143 km from Lonoke County, Arkansas, and one was in Tipton County, Tennessee (204 km away from the populations in Lonoke County). F_ST values indicated that the Tipton County site was genetically distinct from those from Arkansas. However, most F_ST values among the 7 localities in Arkansas (6 in Lonoke County and one in Independence County) indicated little genetic difference among these sites. There was no statistically significant relationship between genetic and geographic distances. The straight-line distance between sites was positively correlated with genetic distances based on F_ST, while riverine distances were negatively correlated with genetic distances, indicating that straight-line distance was a better predictor of genetic structure in mink, perhaps due to overland movements among streams. Genetic structure was not partitioned within streams because there was substantial gene flow among streams.

Previous descriptions of growth, development, and composition of deer antlers have not included factors that affect antler breakage. We related morphology and mineral concentration to occurrence of broken antler tines in white-tailed deer (Odocoileus virginianus) using 126 cast antlers collected during 1989 to 1995. Antlers were divided into 2 categories: broken and intact. Density, break strength, spongy bone percentage, and mineral composition were measured for 1-cm cross-sections removed from the base of the main beam, from between the G2 and G3 tines, and from the base of the G3 tine. We compared these variables between broken and intact antlers and estimated correlations between tine density, break strength, spongy bone ratio, and mineral composition. Density, percentage of spongy bone, break strength, and mineral concentration did not differ significantly between broken and intact antlers. Main beam break strength differed significantly among years. Mineral composition differed significantly among antler locations and year. Rainfall during the antler growth period was correlated with concentrations of aluminum and manganese. Although there were significant correlations between morphology and mineral concentrations, morphology and mineral composition did not explain the presence of broken antlers.

We discovered a mandible of the flat-headed peccary (Platygonus compressus) in a sand and gravel borrow pit (Pit Stop Quarry) between Taylor and Show Low, Navajo County, Arizona. We also found isolated skeletal elements of a mole salamander (Ambystomatidae), pygmy rabbit (Brachylagus idahoensis), sagebrush vole (Lemmiscus curtatus), and several other small mammals in sediments immediately surrounding the peccary jaw. The presence of P. compressus indicates that the deposit is of late Pleistocene age. Specimens also represent the first Pleistocene record of Ambystomatidae on the Colorado Plateau, an important additional record of P. compressus on the Colorado Plateau, and 1 of 2 records of B. idahoensis in Arizona.

Some characteristics of germination and seedling establishment of Stenocereus stellatus, a columnar cactus that has vegetative propagation and sexual reproduction, were determined through field and laboratory experiments with seeds collected at 2 sites. Germination and seedling establishment varied between sites. Seedling survivorship also was affected by solar radiation. Different intrinsic and extrinsic characteristics, such as genetic variation, seed predation, and rainfall, could affect the successful establishment of this species. Vegetative propagation might be a complementary mechanism to sexual reproduction that contributes to population persistence.

Ipomoea pescaprae (L.) Roth (railroad vine, Convolvulaceae) is a pantropical, perennial beach plant that forms large patches just above the high tide line on coastal beaches and dunes throughout tropical and subtropical areas of the world. In spite of its wide distribution, only rare occurrences of I. pescaprae have been documented in inland habitats. Here we report on an extensive population of I. pescaprae growing on the shores of Lake Nicaragua in the interior of Nicaragua.

We monitored a population of pickerel frogs (Rana palustris) in an Ozark cave over a 2-year period. Frogs were found from August to April with densities peaking from November to December. Although densities were quite high in these peak months, stomach content and stable isotope analyses reveal that R. palustris does not play a significant role as a predator in this subterranean system. These results suggest that the caves are being used as thermal refugia during the coldest months of the year.

We describe the diet of the thornscrub hook-nosed snake (Gyalopion quadrangulare) using prey remains removed from the alimentary tracts of museum specimens. Diet consisted of small arthropods, including 9 (31%) insects, 9 (31%) spiders, 8 (28%) scorpions, and 3 (10%) that could not be identified further. Insects included 7 orthopterans (both crickets and grasshoppers) and 2 that we could not identify further. We identified half of the scorpions as Diplocentrus spitzeri. Prey class frequencies are dependent on geographic distribution, but independent of sample source (stomach vs. intestine/feces). Snakes that ate different prey classes did not differ significantly in snout-vent length.

We examined large series of the plains pocket gopher (Geomys bursarius) and Baird's pocket gopher (G. breviceps) collected in every month of the year in 10 Texas counties. Based on lengths of testes, both species potentially reproduce in every month of the year, although the likelihood of reproduction decreases in summer and autumn. Data from females revealed that breeding actually begins in late December or early January and continues through October (in Baird's pocket gopher) or November (in the plains pocket gopher).

Hispid cotton rats (Sigmodon hispidus) occur primarily in grass-dominated habitats. These habitats abut tidal wetlands along the Gulf and Atlantic coasts of the United States occupied by the marsh rice rat (Oryzomys palustris). Few data document whether cotton rats also occupy wetlands and whether interspecific interactions exist with rice rats. We used livetraps to sample 4 grids that each encompassed wetland and upland habitats near Galveston, Texas. Cotton rats occurred in both upland and wetland habitats. Density of cotton rats was lower in wetlands; their residence time was less in wetlands, which perhaps indicated poorer survival or movement to uplands. Sex ratio, proportion reproductive, and age structure, however, did not differ between upland and wetland habitats. Density of cotton rats and rice rats was negatively correlated. Such density compensation coupled with differences in diel activity and diet minimizes potential for interspecific competition.

We report the first documented use of a building as a winter roost by Euderma maculatum. We suggest that sufficient numbers of E. maculatum have been encountered in urban settings that researchers should consider these areas during the course of investigations of species occurrence. We recommend that urban settings in the vicinity of large cliff features be sampled systematically for the presence of E. maculatum.