We recorded underwater vocalizations of captive and wild Florida manatees (Trichechus manatus latirostris) to assess variability in acoustic structure of their sounds and to test hypotheses regarding the importance of specific acoustic traits in individual distinctiveness and in certain behavioral contexts. Manatees use vocalizations to maintain contact when in groups. The highest rates of vocalizing occur during antiphonal calling between females and calves. Vocalizations are complex, single-note calls with multiple harmonics, frequency modulations, nonharmonically related overtones, and other nonlinear elements. We measured 6 acoustic variables and found that individuals varied significantly in fundamental frequency, emphasized band, frequency range, and call contour (the overall pattern of complexity in frequency modulation). These traits did not vary within individuals on different dates or when manatees were alarmed and fleeing. Individual fundamental frequencies ranged from 1.75 to 3.90 kHz, and were negatively correlated with body size. Little sound energy occurred above 18 kHz in 502 call notes of 6 captive manatees sampled with a recording oscilloscope. Presence of harmonics and call duration differed by date and manatees emitted longer calls when fleeing disturbance. Call duration varied from 118 to 643 ms (geometric mean = 271 ms, 95% confidence limits = 264, 279 ms) in a sample of 479 vocalizations we recorded from 14 individuals. The maximum call duration recorded over the entire study was 900 ms. Females and calves responded only to each others' vocalizations when rejoining a group after brief separations, strongly suggesting individual recognition by sound. Structural complexity in the calls of manatees is similar to that in other sirenians, and may reflect their auditory capabilities and the unique physical properties of sound in shallow water.

Bugle calls of male North American elk (Cervus elaphus) are common sounds during fall in the Canadian and United States Rocky Mountains. In contrast, bugle calls of female elk are rarely heard. We quantified the acoustic structure of elk bugle calls, which is an essential 1st step to understanding of the function of the call. We also investigated whether motivation–structural rules apply to these long-distance calls. We measured male elk bugle calls in Rocky Mountain National Park, Colorado, during autumn of 1998 and 1999 and we measured female elk bugle calls on 2 Colorado elk ranches (private establishments that raise elk for commercial purposes) during spring of 2001 and 2002. All bugle calls had 3 segments: on-glide, whistle, and off-glide. Male bugle calls were longer in duration than female bugle calls (P < 0.01). Bugle calls emitted in aggressive interactions had 4 or 5 low-frequency formants, resulting in harsher, wider bandwidth bugles (P < 0.001) compared to the tonal calls emitted in nonaggressive contexts, which lacked formants. Thus, elk bugle calls appear to conform to motivation–structural rules.

The Cape dune mole-rat (Bathyergus suillus) is a solitary, seasonally breeding, rodent mole. Body mass, reproductive tract morphometrics, ovarian histology, and plasma estrogen and progesterone concentrations were measured in females during 1 calendar year to elucidate if this species had an inclination toward seasonal breeding. Qualitative analyses of ovarian histology revealed that females retain the potential for ovulation and subsequent production of corpora lutea from early austral winter to spring (June–October). Seasonal differences were found in ovarian morphometrics and hormone concentrations that are associated with follicular activation from April and subsequent conceptions from June to November. Body mass, reproductive tract morphometrics, testicular histology, and plasma testosterone concentrations were obtained for males over the period of 1 calendar year. There was a gradual increase in seminiferous tubule diameter from May to October, which was mirrored by fluctuations in testosterone concentration. The current data support a strong seasonality to reproduction in the Cape dune mole-rat from April to October with a peak that is linked to the period of maximal rainfall within the distributional range of this species.

The striking patterns in seasonal activity and mass change exhibited by hibernating species of ground squirrels (Spermophilus) are thought to be shaped by climatic conditions and seasonal peaks of environmental productivity. I examined the annual cycle of Mexican ground squirrels (Spermophilus mexicanus) in southeastern New Mexico to test the prediction that because of the moderate winters and summer peaks in precipitation typical of the Chihuahuan Desert, the low-latitude, desert-dwelling Mexican ground squirrel would have a longer active season, a delayed reproductive season, and reduced seasonal changes in body mass compared to species that live at higher latitudes and elevations. I found that adult Mexican ground squirrels are active for roughly 5 months of the year, which is similar to other long-season ground squirrels. Litter emergence and prehibernation fattening in adults and juveniles coincide with the summer peak in precipitation and plant productivity, occurring later in the year than in most Spermophilus. The mating season is delayed after spring emergence and is long relative to other Spermophilus, lasting 30–71 days. Similar to other ground squirrels, mass change in adults is characterized by gains before hibernation, and loss in males during the mating season. However, adult males emerge from hibernation at a very low mass and exhibit an extended period of mass gain after emergence. These findings provide support for the hypothesis that the environment is important for the evolution of annual patterns in Spermophilus.

Eastern red bats (Lasiurus borealis) will move into leaf litter during cold bouts of winter, and because temperatures fluctuate at these roosts, our 1st goal was to quantify winter arousals in response to ambient temperature (T_a). Additionally, we measured changes in metabolism and body temperature (T_b) during hibernation and arousals at various T_a. Using these data, we estimated winter energy budgets. Bats were captured during autumn of 2003 and 2004 in southwestern Missouri and kept in environmental chambers simulating natural conditions. We assessed torpor bout duration using temperature-sensitive data-loggers within environmental chambers at 15°C, 10°C, 5°C, and 2°C. Metabolic rate during torpor was assessed within metabolic chambers at 15°C, 10°C, 5°C, 1°C, and −5°C. Examination of our data suggests that torpor bout duration was not affected by sex and it was inversely related to T_a down to 2°C. Metabolic rate during torpor was not affected by sex but it did vary with T_a. T_b of hibernating bats approximated T_a and the difference between these was greatest at 1°C. Our studies of thermal physiology provide insight of overwintering behaviors of L. borealis, which are apparently less constrained by thermal parameters than in other temperate bat species.

Bot flies are common parasites of Peromyscus leucopus, although determination of a cost to the host has been elusive. The goal of this study was to further explore the potential costs of bot fly parasites for a population of P. leucopus. We investigated the effects of parasitism on host condition (mass after controlling for parasite mass and host body length) and survivorship (the number of days animals persisted on trapping grids). Parasitism was quantified by prevalence (proportion of the population infected), intensity (the number of parasites per infected host), and dispersion of parasites within hosts (clumped, regular, or random). In addition, we searched for spatial and temporal patterns in infection. Finally, we analyzed the relationship between population demography and parasitism. Contrary to expectations, we found that infected mice persisted longer on trapping grids and were in better condition than uninfected mice. Also, we discovered that when considering overall infection levels, parasites were clumped within hosts, but when considering the number of simultaneous infections, parasites were randomly distributed among hosts. Although most animals had single infections, there was a high incidence of reinfections, leading to bimodal patterns of parasitism. Prevalence was not correlated with host density, sex ratio, or proportion reproductive, but there were significant relationships between intensity and density and sex ratio in 1 year. In addition, prevalence and proportion of reproductively active animals were asynchronous. These results suggest that bot flies do not impose an obvious cost to their hosts, and hosts may express some degree of tolerance for bot fly parasitism.

Seasonal food habits and activity patterns were examined for grizzly bears (Ursus arctos) in west-central Alberta, Canada, to better understand habitat requirements in a threatened population. Food habits were based on an analysis of 665 feces collected from 18 grizzly bears between April and October 2001–2003. Trends in the use of foods were comparable to those of other central Rocky Mountain populations, with minor differences likely reflecting regional habitat and forage availability. Five activities (bedding, sweet vetch digging, insect feeding, frugivory, and ungulate kills) were identified for each of 1,032 field-visited global positioning system radiotelemetry locations from 9 female grizzly bears. We predicted the probability of each activity during relevant periods by time of day (crepuscular, diurnal, and nocturnal) and habitat. Selection ratios were used to assess which habitat and time periods were selected. Activity patterns changed considerably over a 24-h period, with foraging activities occurring mostly during diurnal and crepuscular periods and bedding at night. Habitats were important predictors of activity. Forested areas were selected for bedding areas, whereas digging, insect-foraging, and frugivory activities were associated with herbaceous, recently disturbed forest and open-canopy forests. We suggest that researchers consider behavior and time of day in analyses of habitat selection to improve explanations of habitat use and mechanisms of selection.

Valuable conservation research on the African wild dog (Lycaon pictus) has identified that its current endangerment is primarily due to human persecution, although habitat alteration, interference competition with other large predators, and disease also are factors. Numerous studies have thus determined what should be avoided to sustain an African wild dog population, yet in this study we identify what is needed to conserve a wild dog population by using Jacobs' index to determine its preferred prey species. Twenty-four assessments of wild dog prey preference were calculated from 18 studies involving 4,874 kills of 45 species from throughout its distributional range. Wild dogs prefer prey within a bimodal body mass range of 16–32 kg and 120–140 kg, which is abundant and less likely to cause injury when hunted. This bimodal range follows that of optimal wild dog pack sizes based on energetic costs and benefits. Greater kudu (Tragelaphus strepsiceros) and Thomson's gazelle (Gazella thomsonii) are killed by wild dogs wherever they coexist and are significantly preferred. Impala (Aepyceros melampus) and bushbuck (Tragelaphus scriptus) also are significantly preferred. Our results allow wildlife managers to more accurately assess the survival chances of reintroduced or small wild dog populations by determining if sufficient preferred prey are available. These techniques are applicable to all adequately studied large predators.

We examined the numerical and spatial responses of Geoffroy's cats (Oncifelis geoffroyi) to a prey decline in central Argentina between April 2002 and November 2003. The 2nd year of the study coincided with a severe drought. Relative abundance of brown hares (Lepus europaeus) declined from 5.6 individuals/10 km during the predrought period to about 0.6 individuals/10 km during the drought. Small-rodent biomass also showed the lowest level for the study area during the drought of 2002–2003 (134.5 g/ha). During the predrought and drought periods, 3 male and 1 female, and 1 male and 9 female Geoffroy's cats, respectively, were radiotagged and monitored. Home ranges for males of the predrought period averaged 202.8 ha ± 156.8 SD and that of the single female was 27.3 ha. During the drought period, 4 females occupied an average home range of 254.9 ± 254.1 ha, and the home-range size of the single predrought female increased by a factor of 2. No obvious change in mean daily distance traveled between the 2 periods was observed. Geoffroy's cats predominantly used habitats of dense cover during the predrought period, but they became more habitat generalists during the drought. Recruitment of juveniles was only recorded during the predrought period, and all monitored Geoffroy's cats dispersed or died of starvation after the prey decline. Consequently, density of Geoffroy's cats dropped from 2.9 individuals/10 km² before the drought to 0.3 individuals/10 km², probably because of food scarcity. This is the 1st study to examine the spatial ecology of a small wild cat species under nutritional (energetic) stress in South America.

The amount of food left by a forager after feeding in a depletable patch of known volume, the giving-up density (GUD), estimates the quitting harvest rate. I constructed a feeder for measuring GUD in Egyptian fruit bats (Rousettus aegyptiacus). The feeder contained liquid food mixed with pieces of hose that interfered with the drinking behavior of the bats and forced them to work progressively harder to obtain more food as the depth of the liquid in the feeder decreased. Harvest rates in bats using these feeders declined with time. When presented with feeders containing different initial food densities, the bats equalized GUD and consumed proportionately more food from rich patches than from poor ones. Thus, the bats recognized patches of different quality and foraged following a fixed quitting harvest rate patch-use strategy.

We used a combination of capture and acoustic monitoring equipment to examine use of habitat by bats in a desert riparian community in southern Nevada. Each habitat type (riparian marsh, mesquite bosque, riparian woodland, and riparian shrubland) was simultaneously and continuously sampled acoustically in 3- to 5-night increments for 54 nights between June 2000 and January 2001. Fifteen species of bats were detected acoustically, 13 of which were captured using harp traps or mist nets. Five species were not detected frequently enough to be included in statistical analyses. California leaf-nosed bats (Macrotus californicus) and Brazilian free-tailed bats (Tadarida brasiliensis) were generalists, spending equal amounts of time in each habitat. Western yellow bats (Lasiurus xanthinus) and pallid bats (Antrozous pallidus) demonstrated strong biases for riparian woodland over the other habitats sampled. The remaining 6 species spent substantially more time in at least 1 of the 4 habitats. Riparian woodlands accounted for more than 50% of all bat activity, whereas riparian marshes were the least used habitat. High species richness and differences in habitat use by most species emphasizes the importance of a diversity of riparian habitats for bats at the study site. The existence of both native and nonnative habitat may elevate bat species richness and increase the degree of differential habitat use to levels higher than would be expected if only native habitat existed at the study site. Understanding differential riparian habitat use by bats in desert ecosystems may have profound management implications.

Human development impacts the landscape by altering the size and shape of natural habitat patches, replacing natural vegetation with other types such as lawns and row crops, or introducing environmental stressors such as increased human activity and pollutants. We investigated the effects of human alterations to the landscape on the distribution of 3 mammalian carnivores (coyote [Canis latrans], raccoon [Procyon lotor], and red fox [Vulpes vulpes]) along an urban–rural gradient in northern Illinois. Distribution of each species was assessed from occurrence at scent stations placed within or along the edges of 47 sites ≥ 4 ha, representing 7 different natural or anthropogenically altered habitats. We averaged presence or absence scores across several seasonal samples over a year, and used an outlying mean index analysis to compare them to environmental variables gathered for each site, including habitat and landscape metrics presumed to reflect varying degrees of anthropogenic influence across the urban–rural gradient. Coyotes used a variety of habitats within the rural part of the gradient. Red foxes were found in forest interiors or shrubland and old fields near forests where coyotes were least detected. Both canids were detected more often in areas of lower human densities but prey abundance was not a strong determinant of their occurrence. Overall occurrence along the gradient was highest for raccoons, which were positively associated with urban areas with relatively high residential land use.

We sampled small mammals of fragmented sagebrush steppe on the Snake River Plain, Idaho, and compared the effects of habitat isolation on their diversity, abundance, and species composition in 2 landscapes, kipukas, which are patches of sagebrush-steppe habitat that were isolated by late-Pleistocene and early-Holocene lava flows, and remnant patches of sagebrush steppe that remain within recently developed agricultural areas. Species richness decreased with increasing isolation in both lava and agricultural landscapes, and Peromyscus maniculatus (deer mouse) was the only species found on kipukas that were isolated by more than 400–800 m, suggesting that many native sagebrush-steppe species may decline or disappear from fragmented sagebrush steppe. Density of small mammals increased with isolation on the agricultural patches, but decreased with isolation on kipukas; however, increased densities were entirely due to P. maniculatus. Diversity of small mammals on kipukas was highly correlated with absolute and proportional abundance of shrubs, probable indicators of sagebrush-steppe vegetation. Additionally, Bromus tectorum (cheatgrass) may play a role in the low diversity and density of mammals on kipukas, because both diversity and density were lower where cheatgrass was present than where it was absent.

We surveyed snowshoe hare (Lepus americanus) and mountain cottontail (Sylvilagus nuttallii) in coniferous forests in northern New Mexico. L. americanus was restricted to the Sangre de Cristo and San Juan ranges in a narrow band of elevations corresponding to subalpine coniferous forest. In contrast, S. nuttallii was widely distributed and virtually cosmopolitan, occupying a wide range of elevations and vegetation zones, including subalpine coniferous forest where it was syntopic with L. americanus. Previous reports about the distribution and ecological associations of these species at their southern limits indicated broader ecological associations for L. americanus and narrower ecological associations for S. nuttallii. Historical reports about the biogeography of these species likely were based on misinterpretation of leporid sign and specimen records. We make recommendations for reducing problems associated with the interpretation of anecdotal natural history information and specimen records.

Subterranean rodents are characterized by limited individual mobility and patchy distribution of local populations, all of which have been related to the great evolutionary diversification of this group. Because of their secretive habits, radiotracking becomes an essential tool to collect data systematically on characteristics of their space use, rate of movement, and daily patterns of activity. Compared to other subterranean species, the ctenomyids exhibit more activity above the surface. Despite this, a previous study showed that only 25% of the captive individuals analyzed presented rhythmic bouts of activity associated with light–dark cycles. In our study, we used radiotelemetry to explore quantitatively the home-range dynamics and daily movements of the South American Talas tuco-tuco (Ctenomys talarum). As expected, the home ranges of males were larger than those of females. However, intraspecific variation in home-range size also was observed between 2 different study sites, possibly reflecting differences in body size and soil characteristics between the study sites. Rhythmicity in activity patterns was confirmed to be polymorphic in this species, as previously observed in captivity, and daily activity was mostly concentrated in the diurnal period. We suggest that the present study will contribute to generating important insights into home-range dynamics and daily patterns of activity of free-living subterranean rodents.

In this work we examined the geographic genetic structure of the subterranean herbivorous rodent Ctenomys australis (sand-dune tuco-tuco) using mitochondrial DNA (mtDNA) control region (D-loop) sequences. We found 24 haplotypes in the analysis of 70 individual 403–base pair sequences; most were restricted to single populations, although a few haplotypes were shared broadly across the species' range. Genetic differentiation was not consistent with a simple model of isolation by distance, possibly evidencing a lack of equilibrium between gene flow and local genetic drift. The analysis of mismatch distributions, Fu's F_S-test of neutrality and the “starlike” topology of the gene genealogy showed a pattern consistent with a recent population expansion event. However, we could not rule out an alternative explanation based on departures from strict neutrality of mtDNA.

Kangaroo mice (Microdipodops) inhabiting the Mono Basin and adjoining valley regions of California and Nevada represent a disjunct distributional isolate and have been considered as a distinct species or, more recently, as 2 subspecies of M. megacephalus. Analysis of patterns of geographic variation in 11 populations in the Mono Basin region shows that kangaroo mice inhabiting the northern portion of this peripheral isolate are relatively large and dark (referred to as M. m. nasutus), and those to the southern end are small and pale (termed M. m. polionotus), and a cline exists between the 2. Inasmuch as several morphological characters are correlated positively with environmental measures (e.g., hind-foot length and climatic severity; pelage color and soil color), it appears that variation in morphology is responding to a selection gradient. Mitochondrial DNA sequence data, together with chromosomal and protein information, reveal that the 2 subspecies are nearly identical genetically. Moreover, molecular phylogeographic analysis reveals that the Mono kangaroo mice belong to the southeastern geographic unit of M. megacephalus (a mean of 1.86% sequence divergence) and are genetically most close to animals from the San Antonio locality (more than 100 km to the east). It is hypothesized that distributional shifts in the geographic range of kangaroo mice in response to climatic fluctuation during the late-Pleistocene and Holocene times resulted in the westward expansion and eventual colonization and isolation of kangaroo mice in the Mono Basin region. Our vicariant biogeographical interpretation suggests a historical route through the Lahontan Trough and the physiographic discontinuity east of Mono Lake (i.e., between the Wassuk Range and the White Mountains) that may serve as a biogeographic model for other basin-dwelling organisms. Lastly, the systematic status of the 2 subspecies is evaluated; a single subspecies, polionotus, is recognized, with nasutus placed in synonymy.

The subfamily Heteromyinae (spiny pocket mice) represents a well-defined monophyletic group within the rodent family Heteromyidae. Although 2 genera of spiny pocket mice, Heteromys and Liomys, are recognized in the subfamily, no phylogenetic analysis has demonstrated their reciprocal monophyly. A recent study using DNA-sequence data from the mitochondrial gene cytochrome b suggested that Liomys is paraphyletic but included few species of Heteromys. Here, we conduct phylogenetic analyses of the subfamily with dense taxonomic sampling using allozymic data from a previous study and external and cranial morphological data; our aim is to assess generic monophyly and elucidate phylogenetic structure within the genera, to the degree possible with these data. We also reidentify selected voucher specimens from the allozymic study. Parsimony-based analyses indicate 3 clades in the subfamily: (A) Liomys irroratus, L. pictus, and L. spectabilis; (B) L. adspersus and L. salvini; and (C) all examined species of Heteromys. However, the relationships among these clades are unresolved. The genus Heteromys is characterized by strong support and several unreversed morphological synapomorphies. In contrast, our analyses fail to indicate any synapomorphies for Liomys, but can neither demonstrate nor reject its monophyly. The 3 clades identified here match those recovered from a recent mitochondrial DNA–sequencing study, which found a resolved (B (A C)) topology. Within Heteromys, we recover 5 lineages, but the relationships among them remain unresolved. The examined South American species of Heteromys formed a clade, but 2 species recently described from Ecuador and Venezuela were not included here. Samples referred to as H. desmarestianus crassirostris and H. d. planifrons were quite distinct from other samples of H. desmarestianus, emphasizing the need for alpha-level taxonomic revision of this species complex. Given the current results, future studies can now examine relationships among species of Heteromys using samples of Liomys as outgroups, but studies of Liomys must take into account its likely paraphyletic nature.

We undertook analyses of mitochondrial DNA gene sequences and echolocation calls to resolve phylogenetic relationships among the related bat taxa Rhinolophus pusillus (sampled across China), R. monoceros (Taiwan), R. cornutus (main islands of Japan), and R. c. pumilus (Okinawa, Japan). Phylogenetic trees and genetic divergence analyses were constructed by combining new complete mitochondrial cytochrome-b gene sequences and partial mitochondrial control region sequences with published sequences. Our work showed that these 4 taxa formed monophyletic groups in the phylogenetic tree. However, low levels of sequence divergence among the taxa, together with similarities in body size and overlapping echolocation call frequencies, point to a lack of taxonomic distinctiveness. We therefore suggest that these taxa are better considered as geographical subspecies rather than distinct species, although this should not diminish the conservation importance of these island populations, which are important evolutionarily significant units. Based on our findings, we suggest that the similarities in body size and echolocation call frequency in these rhinolophids result from their recent common ancestry, whereas similarities in body size and call frequency with R. hipposideros of Europe are the result of convergent evolution.

Wing morphology is crucial for flight performance and foraging ecology in bats. We describe variations in 5 wing parameters within the 5 species of European horseshoe bats (genus Rhinolophus) based on data taken from 3,081 adult individuals. All 5 species belong to a single ecological guild. Measurements were taken from live bats in the field in southeastern Europe (Bulgaria, Greece, and Turkey), where all 5 species occur in sympatry. Examination of our data shows that the species and accordingly their wings differ substantially in size. Albeit grossly similar in form, we additionally found several size-independent differences in wing shape. For example, the smallest species, Rhinolophus hipposideros, and to a lesser extent also R. blasii, have extremely short hand wings, enabling highly maneuverable search flight close to vegetation. The largest species, R. ferrumequinum, and the 2nd largest one, R. mehelyi, have rather long hand wings, allowing fast and economic commuting flight over longer distances. We argue that both size and shape are likely to play a role for niche separation between species. We found both sexual and geographic variation within species. There was sexual dimorphism for most parameters, with females being larger than males. Populations of R. mehelyi in southeastern Europe had significant variation in wing measurements. This was not so for R. ferrumequinum and R. euryale. We give a discriminant function based on only 2 parameters that correctly assigned 98% of the 3,081 individuals to species. This function may prove useful for identification of museum specimens.

Humeral variation associated with digging ability in the subterranean rodent Ctenomys was analyzed through 6 functionally significant indexes. The humerus of some extinct and living species was slightly more specialized than that of fossorial octodontoids †Actenomys and Octodon, whereas it was highly specialized in some living species. The constant occurrence of greater epicondyles suggests a hierarchical pattern in the acquisition of scratch-digging specializations. A possible relationship between humeral morphological diversity and environments is preliminarily discussed.

The distinct ecological requirements of Microtus xanthognathus (yellow-cheeked vole or taiga vole) and M. pennsylvanicus (meadow vole) warrant accurate discrimination of their remains in studies of paleoecology and past biogeographical shifts. An occlusal length of the lower 1st molars (m1) that is >3.2 mm for M. xanthognathus is the method most frequently used to separate these 2 taxa in archaeological and paleontological samples. However, these measurements alone are unreliable because some specimens of M. pennsylvanicus overlap smaller individuals of M. xanthognathus in size. Therefore, I created and tested a morphometric technique that discriminates Recent lower 1st molars (m1s) of M. pennsylvanicus from those of M. xanthognathus, and is applicable to other taxa (both modern and fossil). Despite overlapping occlusal length, my discriminant function based on landmark data correctly classified 100% (n = 53) of Recent m1s from the 2 taxa and 97.7% (43 of 44) of (assumed) m1s of M. pennsylvanicus from an archaeological site from about AD 1200 in central Nebraska. This landmark scheme is applicable to fossil and modern Microtus worldwide.

Sympatric species of vesper mice (Calomys expulsus and C. tener) from the Cerrado biome are often distinguished by their respective sizes. Using geometric morphometrics, we tested if interspecific differences were mainly due to isometric or allometric size variations or allometry-free shape differences. To delimit species groups, we used and compared linear discriminant analysis, calculated on subsets of individuals of known identity, and pattern recognition techniques, needing no prior information on specimens. They both yielded similar results, indicating that patterns of interspecific morphological differences are mainly due to size-free shape differences located at landmarks defined at the suture between the frontals and the parietals and between the latter and the interparietal. Correct specimen identification was obtained with pattern recognition techniques using Gaussian mixture models. Morphological differences also were found between the 2 species analyzed and the newly described C. tocantinsi represented here by its paratypes. The combination of geometric morphometrics and pattern recognition techniques seems suitable for systematic analyses aimed at elucidating interspecific patterns of morphological variation in closely related species in field studies and museum specimens.