We investigated variation in environmental conditions and fish assemblages over space and time in the upper Red River drainage (Oklahoma–Texas) using samples from eight sites (three tributaries and five mainstems) collected monthly in 1989–1990 and again in 1998–1999. Fish assemblages varied spatially, with geographically ubiquitous and ecologically tolerant taxa dominating tributaries and obligate species of prairie minnows common in mainstems. At the regional scale, fish assemblages were predictable between time periods. This resulted from strong longitudinal structuring in fishes across sites, which likely was influenced by spatial variation in environmental conditions across the drainage. These patterns suggested that at the regional scale strong forcing of fish by their environment influenced predictability of fishes between time periods. At the local scale, predictability in fish assemblages between time periods differed across type of habitat. Specifically, assemblages in tributary sites varied randomly over time, whereas mainstem assemblages were predictable. We believe that local environmental conditions favored different suites of taxa in tributaries between time periods. When we compared variation in fish assemblages to variation in environmental conditions at both regional and local scales, we detected that structure of fish assemblages tracked environmental variability at the regional scale. However, at the local scale, fish assemblages did not track environmental conditions. This suggests that environmental variation may affect fishes in a relative sense across the drainage, whereas, within sites, fishes may be decoupled from environmental structuring-forces. Based on the predictable patterns in fishes within the Red River drainage, we propose that monitoring efforts, aimed at detecting human-induced changes in this system, should include a range of mainstem sites that capture the strong structuring forces that vary longitudinally within this river system.

In central Texas, western cottonmouths (Agkistrodon piscivorus) inhabit small, low-productivity limestone streams characterized by variability in seasonal flows. An 11-year study of western cottonmouths was conducted at Honey Creek, a spring-fed stream flowing 3.2 km to its confluence with the Guadalupe River, Comal County, to determine how this predator used space in this narrow, linear, and dynamic system. During 57 searches along the 1,564-m study area, 39 sexually mature, 14 subadults, and 4 juvenile snakes were marked. Rates of recapture did not differ between sexes, but females outnumbered males (2.3∶1) and adults were recaptured more frequently than juveniles. Distances between captures were less than predicted if distributions were random, and distances did not vary with number of times captured or time between captures.

The black-tailed prairie dog (Cynomys ludovicianus) is a colonial, burrowing ground squirrel. Its historic range has been reduced during the past 200 years mostly due to agricultural practices and related disturbance. Currently, urbanization is one of the biggest threats to colonies of prairie dogs in eastern New Mexico. As cities expand, colonies contract and become increasingly fragmented. Impacts of broad-scale landscape factors and disturbance by humans on behavior of black-tailed prairie dogs are unknown. To address these impacts, we chose two landscapes in eastern New Mexico, urban and rural. In spring 2006 and autumn–winter 2006–2007, we randomly selected six sites within each type of landscape and conducted 20 5-min behavioral observations at each site. We compared behavioral time budgets using the Mann-Whitney U-test. Black-tailed prairie dogs spent significantly more time being vigilant and gave more warning calls in urban areas, but they spent significantly more time foraging in rural areas. These results suggest that prairie dogs respond behaviorally to differences in type of landscape. Costs associated with higher levels of vigilance might translate as lower overall fitness due to decreased foraging in these populations. Landscape-level factors should be considered in future management decisions for this species.

Data on activity patterns of the desert nightsnake (Hypsiglena chlorophaea) were obtained from 121 specimens. Hypsiglena chlorophaea was active on roads in central Washington during April–October. Specimens were collected at temperatures of 9.1–40.8°C. Males were collected more frequently than females during May, but in equal numbers during June and July, while more females were found in August–September. Hatchling snakes were collected in July–September. Hypsiglena chlorophaea was active 2100-0500 h. Snakes were collected more often during periods of low moon-light (e.g., new moon) than high moon-light (e.g., full moon).

Entomofauna in monospecific stands of the introduced Chinese tallow tree (Sapium sebiferum) and native mixed woodlands was sampled in 1982 along the Texas coast and compared to samples of arthropods from an earlier study of native coastal prairie and from a study of arthropods in S. sebiferum in 2004. Species diversity, richness, and abundance were highest in prairie, and were higher in mixed woodland than in S. sebiferum. Nonmetric multidimensional scaling distinguished orders and families of arthropods, and families of herbivores in S. sebiferum from mixed woodland and coastal prairie. Taxonomic similarity between S. sebiferum and mixed woodland was 51%. Fauna from S. sebiferum in 2001 was more similar to mixed woodland than to samples from S. sebiferum collected in 1982. These results indicate that the entomofauna in S. sebiferum originated from mixed prairie and that, with time, these faunas became more similar. Species richness and abundance of herbivores was lower in S. sebiferum, but proportion of total species in all trophic groups, except herbivores, was higher in S. sebiferum than mixed woodland. Low concentration of tannin in leaves of S. sebiferum did not explain low loss of leaves to herbivores. Lower abundance of herbivores on introduced species of plants fits the enemy release hypothesis, and low concentration of defense compounds in the face of low number of herbivores fits the evolution of increased competitive ability hypothesis.

We conducted prescribed fires on Matagorda Island, Calhoun County, Texas, in late summers (August) and early winters (January) of 1992–1993 and 1993–1994 to determine responses of vegetation and litter. There were reduced litter loads 1 year after burning, but they approached pre-burn levels 2 years after burning. There was no difference in production of grasses, forbs, or the Juncus and other-Cyperaceae complex 1 year after burning. There was a significant difference in production of Juncus and other Cyperaceae when years were compared. We presume this was caused by differences in precipitation between years. There was no difference in percentage frequency between burned and control plots 1 year after burning for most species examined. However, burning in summer significantly increased occurrences of coastal mistflower (Conoclinium betonicifolium), Fabaceae, and Cyperaceae (other than Juncus) over control plots 1 year after treatment. Beach groundcherry (Physalis cinerascens) and Fabaceae also were increased by burning in summer 2 years following treatment. Prescribed burning did not affect density of vertical structure. Because coastal barrier islands serve as important wintering grounds and stopovers for many migratory birds, we recommend conducting burns in late summer every 2–5 years to increase incidence of legumes and other forage for wildlife.

Livetrapping small mammals is a common tool for surveying diversity and determining size of populations, but studies in the southern United States often are complicated by the red imported fire ant (Solenopsis invicta). We tested effectiveness of four chemical repellents applied around Sherman live traps to deter fire ants from removing bait and mutilating captured mammals (Borax, granules of pyrethrum, Raid Ant and Roach Spray, and carbaryl flea powder). A significant difference among repellents was detected, with carbaryl flea powder and Raid Ant and Roach Spray performing best. Carbaryl flea powder significantly reduced occurrence of fire ants in baited traps, while diversity and rates of capture for small mammals were unaffected.

Mist-netting and acoustic monitoring was used to determine occurrence, activity, and use of habitats by bats at Montezuma Castle (i.e., Montezuma Castle Unit and Montezuma Well Unit), Tonto, and Tuzigoot national monuments, Arizona. All calls were visualized by generating sonograms and identified to species except for Myotis, because of similarity between calls within this genus. We used discriminate analysis to identify calls of Myotis to species. Riparian and marsh vegetational associations had the highest overall activity, except at Montezuma Castle Unit where desert, riparian, and transition vegetational associations had similar overall activity. Use of acoustic monitoring and mist-netting enabled us to detect one to eight more species, depending on site, than if only one technique had been used.

We studied demography and behavior of endangered Mount Graham red squirrels (Tamiasciurus hudsonicus grahamensis) inhabiting areas within and outside low-intensity burns following the Nuttall Fire in the Pinaleño Mountains, Arizona. Body mass, proportion of individuals in reproductive condition, and distance squirrels traveled to nest did not differ between squirrels within and outside the perimeter of the fire. Within the perimeter, red squirrels had smaller home ranges and shifted territories less frequently and had shorter distances from their previously held territories than squirrels outside the perimeter. Mount Graham red squirrels evolved with patchy, low-intensity fires like those that burned in mixed-conifer forests in the Pinaleño Mountains and may be able to persist in areas affected by this level of disturbance.

Previous studies have shown that determining sex of adult white-winged doves (Zenaida asiatica) by cloacal characteristics is highly reliable (90%), while using plumage characteristics and body weights are less so. The drawback of the cloacal technique is that the procedure requires use of a modified nasal speculum to probe the cloacal area, which could possibly cause injury to the bird. We hypothesized that females have a wider pelvis than males for the passage of eggs, which should be quantifiable, thereby providing a means of separating adult males and females. We measured pelvic spread (distance from tip to tip of pelvis) of adults (n  =  249). Additionally, body masses were taken for a subset (n  =  215) of the doves. Sex was positively determined by either cloacal examination of live birds or gonad examination of dead birds. Mean pelvic spread of males was significantly less than that of females. Whole body mass was not correlated with pelvic spread for either sex. Accuracy of the pelvic spread technique was 70%. We concluded that the resolution obtainable through determination of sex by pelvic spread in live adult white-winged doves was inadequate for the needs of most field researchers.

Creeks in the Pinaleno Mountains of Arizona were renovated in the 1960s to remove nonnative salmonids, and were stocked with native Apache trout (Oncorhynchus apache). Cursory observations of distribution of fish following attempts to transplant have been reported, but following the Clarks Peak Fire of 1996 more thorough investigations of the aquatic communities were needed. In 1997, we sampled communities of fish with a backpack electrofisher and we collected aquatic invertebrates with a dip-net. Lengths of fish were recorded and scales were removed for assessment of age. We collected 398 putative Apache trout 60–257 mm in length and 0–3 years of age from four creeks, while four others contained no fish. Densities of invertebrates were variable and did not appear to be correlated with presence of fish or drainage. Translocation of Apache trout to Grant and Ash creeks has resulted in reproducing populations of trout with adequate densities and rates of growth to serve as potential future sources of fish for stocking. Marijilda Canyon and Big creeks also supported reproducing populations, but at a much lower density and may not be suitable as sources of fish for stocking. However, genetic information from these populations is needed to determine their suitability for service as replicate stocks.

We estimated species richness, distribution, and relative abundance of herpetofauna across a 1,800-m elevation gradient in the Rincon Mountains and compared patterns of occurrence of species with other mountain ranges in southeastern Arizona. We detected 2,378 individuals of 40 species during 596 h of visual-encounter surveys, and 1,793 individuals including six additional species incidentally during 2001 and 2002. Based on probabilistic methods, we estimate that as many as 57 species likely are present in the Rincon Mountains; past observations and expert opinion also suggest as many as 57 species likely are present. Species richness declined with increasing elevation. Relative abundance declined with increasing elevation for seven species, increased with elevation for three species, and was greatest at middle elevations for three species. The Rincon Mountains lack some species that occur in nearby mountain ranges of similar size and elevation, possibly due to climatic or historical factors. Our study was the first detailed inventory of herpetofauna in the Rincon Mountains at a time when these communities may be undergoing significant change.

Cowleech Prairie Preserve, located on the Cowleech Fork of the Sabine River, Hunt County, northeastern Texas, is a wet-to-mesic blackland prairie-wetland complex that represents a plant community currently undescribed by the National Vegetation Classification. However, it bears similarities with those globally imperiled communities having eastern gamagrass (Tripsacum dactyloides) as a first-order dominant. Objectives of this study were to document floristic composition and examine changes in composition and dominance of species from late spring to early autumn and to better define classification of this community as a recognized association. We collected data on species richness, composition, ground cover, and abundance of forbs during early autumn 2003 and late spring 2004. Total number of species (species richness) recorded was 56 in autumn and 61 in spring. However, these seasonal differences in species richness were not significant. Poaceae (grasses) was the most dominant family in autumn, with gamagrass forming the most predominant ground cover. In contrast, sedges and forbs were dominant in spring. Many species of forbs (47.0% in spring and 62.0% in autumn) had a limited distribution among transects, occurring in only one of five transects during the two seasons. In addition, 21.4 and 29.5% of species of forbs were unique to autumn and spring, respectively. Beautiful false dragon-head (Physostegia pulchella), a species endemic to the blackland prairies of Texas, was recorded in spring. This study represents the first comprehensive floristic documentation of the Cowleech Prairie Preserve.

In an ongoing experiment, we investigated how mating behaviors of male western mosquitofish Gambusia affinis change across a range of temperatures. During behavioral trials, we observed high levels of homosexual behavior between males when in a competitive situation. We observed that when two males are in competition for matings with females, the smaller male performed significantly more homosexual acts than the larger. These results suggest that a dominance relationship develops with males in competition, where large males may dominate small males by not allowing them to copulate with the female, thereby causing an increase in same-sex mating by smaller males. Although preliminary data also indicate a possible relationship between density of males and number of homosexual copulations, we were not able to statistically analyze the relationship because these observations occurred during an ongoing experiment. In a future study, we will seek to understand underlying causes of homosexual behavior in the western mosquitofish.

We studied size and fidelity of home ranges of adult (≥3.5 years) male white-tailed deer (Odocoileus virginianus) during 2 years in southern Texas. Size of core areas, determined using the 50% fixed-kernel method, averaged 86 ± 14 ha, whereas size of home ranges, determined using the 95% fixed-kernel method, averaged 403 ± 61 ha. Deer shifted centroids of home ranges 363 ± 45 m from year 1 to year 2. Overlap in home ranges averaged 66 ± 2% between years. These data provide additional support that home ranges of adult male white-tailed deer in southern Texas are relatively small and stable. However, the majority of home ranges are larger than most landholdings. Therefore, development of cooperatives among neighboring landowners may be necessary to effectively manage age structure of males.

During a radio-tracking study of North American beavers (Castor canadensis) in Arizona, we discovered three adult, lactating beavers using the same bank den at the same time. Two adult females 5.2 km downstream also were using the same den at the same time. For the first case, we documented use of communal dens by lactating adults on seven occasions during 68 days. For the second case, we documented communal use of a den on three occasions during 45 days. To test the hypothesis that these communal females were closely related, we used eight autosomal DNA microsatellites. Two females in the first den were first-order relatives. However, the other communal females were unrelated to the females with which they shared the den. We conclude that communal denning, where multiple females raise one or more litters in the same den, may have occurred in our study area and communal-denning partners are, in some cases, close kin.

We collected a mite, Acomatacarus arizonensis, for the first time from the states of Coahuila and Durango, Mexico. It had been reported previously only in Mexico from the Peninsula of Baja California and from Guerrero. Larvae of this mite was parasitizing the endemic lizards Sceloporus cyanostictus and Crotaphytus antiquus in the Sierra San Lorenzo, Matamoros, Coahuila, Mexico, and Sceloporus jarrovii in the Valle de las Piedras Encimadas, Dinamita, Durango, Mexico.

We used microsatellite analysis to investigate possible reproduction by parthenogenesis in the bisexual Sierra garter snake, Thamnophis couchii. The genotypes of four microsatellite loci were determined for each of four individuals, including three T. couchii, two of which were a female and her offspring, and T. ordinoides. The female T. couchii and her offspring were homozygous and identical for all four microsatellite loci; however, Ts3 and Ts4 are the only loci that supported reproduction by parthenogenesis as Ts1 and Ts2 were uninformative. These data along with absence of a unique paternal allele and known absence of an opportunity to mate prior to and during captivity is consistent with reproduction by parthenogenesis.

We report apparent predation by coyotes (Canis latrans) on one of two male pronghorns (Antilocapra americana) that were locked together. When the locked pronghorns first were spotted at 2030 h, both were alive, one standing and the other on the ground wounded by a horn of his opponent. At ca. 0230 h, four coyotes were seen within 150 m of the pronghorns. At 0630 h, the wounded pronghorn had been partially eaten, presumably by the coyotes, and the other pronghorn freed himself from the carcass and walked away.

The genus Cycleptus is widely distributed in the Mississippi, Missouri, and Ohio rivers and in Gulf-slope drainages, including the Rio Grande. Recent genetic evidence suggests that Cycleptus in the Rio Grande is an undescribed species; however, there are few specimens from Mexico. These include two records from the Río Conchos, two from the Rio Grande and another from the Río Salado. In 1954, Cycleptus was reported in two localities in the Río Conchos: 1.6 km N Saucillo and 1 km above the confluence of the Rio Grande and the Río Conchos. After a purported absence of 51 years, it has been rediscovered in three localities in the state of Chihuahua: the mouth of the Río Conchos at Rio Grande, Cuchillo Parado, and Río Conchos at El Alamo.

This study documents the red crossbill (Loxia curvirostra) from the Cerro El Potosí region, Nuevo León, Mexico. Adults were observed at different sites and elevations in December 2006 and again in July and August 2007. A juvenile male was observed in August 2007. These multiple seasonal observations, as well as presence of a juvenile, indicate an established population in the area of Cerro El Potosí. This report adds the red crossbill to the list of permanent avian residents for Nuevo León.

The range of the northern pygmy mouse (Baiomys taylori) continues to expand through Texas and into adjacent states. We report a new record for B. taylori in Hunt County, near its poorly understood northeastern limit in Texas. This record, together with results from other recent surveys, suggests that the northern pygmy mouse might only now be realizing the extent of range projected in the literature >25 years ago.

The Georgetown salamander, Eurycea naufragia, is a permanently neotenic salamander known only from about a dozen surface springs and caves in Williamson County, Texas. Rapid urbanization places all known populations at risk and conservation strategies are hindered by a lack of information on the ecology of the species. To better understand requirements of microhabitat and spatial distribution of E. naufragia within flows of surface springs, we conducted counts of salamanders on the surface at one locality over a 12-month period. Numbers of salamanders and percentage of cover objects occupied by salamanders varied among months, with a general trend of higher abundance in spring and summer. Few juveniles were observed, and there was no strong seasonal trend in distribution of size of salamanders. Within the flow of the spring, abundance of salamanders decreased linearly with distance from origin of the spring. Salamanders were more likely to be under rocks than under other types of cover objects and they selected larger rocks. Larger salamanders occupied larger cover objects; rocks covering multiple salamanders were larger than those covering single salamanders.

A recent survey of the lower Brazos River in Grimes and Washington counties, Texas, resulted in the first record of a living population of Truncilla macrodon since its description in 1859. Truncilla macrodon is endemic to the Brazos and Colorado basins of central Texas. Anthropogenic impacts resulted in a decline in abundance and distribution for this species prior to extensive scientific study. Given the absence of information on natural history for this species, we documented preliminary physical and biological characteristics associated with this population.