Abba, Guillermo H. Hairy armadillos constitute an ecologically homogeneous and morphologically similar group with currently 5 species classified in the subfamily Euphractinae. Among them, the Andean hairy armadillo Chaetophractus nationi Xenarthra, Cingulata, Dasypodidae is a small, endangered armadillo that has long been suspected to represent a high-altitude variant of Chaetophractus vellerosus. Here, we report the 1st phylogenetic systematics assessment of hairy armadillos using morphological and molecular analyses of all described species with focus on the status of the Andean hairy armadillo.
|Published (Last):||6 September 2014|
|PDF File Size:||14.14 Mb|
|ePub File Size:||18.28 Mb|
|Price:||Free* [*Free Regsitration Required]|
Abba, Guillermo H. Hairy armadillos constitute an ecologically homogeneous and morphologically similar group with currently 5 species classified in the subfamily Euphractinae. Among them, the Andean hairy armadillo Chaetophractus nationi Xenarthra, Cingulata, Dasypodidae is a small, endangered armadillo that has long been suspected to represent a high-altitude variant of Chaetophractus vellerosus.
Here, we report the 1st phylogenetic systematics assessment of hairy armadillos using morphological and molecular analyses of all described species with focus on the status of the Andean hairy armadillo. Multivariate analyses of shape variation based on 3-dimensional landmark coordinates of skulls allowed a clear differentiation of each species with the exception of C.
Identical sequences in C. Based on these data, we propose that C. However, this taxonomic change should not preclude the protection of the high-altitude Bolivian populations that are steadily declining because of their overexploitation for traditional purposes. Finally, phylogenetic analyses of euphractine armadillos based on a combination of 6 non-coding nuclear markers and 2 nuclear exons suggest the paraphyly of the genus Chaetophractus , with C.
A partir de estos resultados proponemos que C. There are currently 21 described species of armadillos Xenarthra, Cingulata, Dasypodidae— Gardner A recent reassessment of their conservation status has shown that virtually all species are affected by hunting and habitat degradation Abba and Superina That study, as well as a more recent literature review by Superina et al.
One clear example is the 5 euphractine armadillos: Euphractus sexcinctus Linnaeus, , Zaedyus pichiy Desmarest, , and 3 species of Chaetophractus , which over time have been considered a subfamily called Euphractinae Frechkop and Yepes , recognized as a tribe Euphractini— Cabrera , lumped into a single genus Euphractus — Moeller , and reclassified as a tribe by Wetzel b. However, based on the phylogenetic distinctiveness of fairy armadillos, Delsuc et al.
As a consequence, the subfamily Euphractinae would now consist of the 5 species originally included by Frechkop and Yepes , namely E. Nonetheless, several taxonomic inconsistencies and phylogenetic relationships within this subfamily still need to be elucidated. The genus Euphractus contains a single species: the 6-banded armadillo E. This relatively common species occurs from southern Suriname and adjacent Brazil to Uruguay and northeastern Argentina Wetzel a ; Abba and Superina It is the largest species of Euphractinae and can be distinguished from the other members of this tribe by its pale yellow or tan carapace and the absence of a movable band at the anterior margin of the scapular shield Redford and Wetzel The genus Zaedyus is also monospecific: the pichi Z.
Unlike other euphractines, the marginal scutes of its carapace bear sharply pointed apices, it has no pelvic glands, and it lacks teeth on the premaxillary Wetzel a , b ; Superina and Abba The genus Chaetophractus is described as comprising 3 species of hairy armadillos Wetzel a , b ; Wetzel et al. The main population of C. In the only comprehensive account on the taxonomy and distribution of armadillos based on a search of museum collections, Wetzel , b questioned the validity of C.
The only difference between 2 putative C. As a consequence, small hairy armadillos are usually identified based on their relative size and capture site, with larger or more hairy individuals, as well as high-altitude specimens or those originating from western Bolivia, northern Chile and northwestern Argentina being assigned to C. Although the skull of C. The ongoing taxonomic uncertainty about C. According to Redford and Eisenberg , C. Although Gardner , limited the range of C.
It also occurs in the southern regions of Tacna and Puno provinces of Peru E. The most recent range maps Abba and Superina suggest that the distribution areas of C. Finally, from the phylogenetic point of view, early molecular studies including all 3 genera of hairy armadillos failed to resolve the relationships within the subfamily Euphractinae Delsuc et al.
The grouping of these 2 genera to the exclusion of Zaedyus would be congruent with the cladistic study of craniodental characters by Gaudin and Wible Such a relationship was strongly supported by the phylogenetic analysis of more than 12kb of non-coding retroposon flanking sequences, and also by the shared presence of a specific DAS-III3b element insertion, plus a nucleotide diagnostic deletion in the An5 genomic locus.
This result is in contradiction with craniodental characters Gaudin and Wible but appears to be compatible with relationships obtained from postcranial evidence Abrantes and Bergqvist One potential explanation for the discrepancy between these 2 molecular studies may lie in the species used as representatives for the genus Chaetophractus.
Delsuc et al. In fact, these 2 species have never been included in the same molecular or morphological phylogenetic data set and may turn out to be only distantly related, in which case such inconsistencies are to be expected. In the present work, we aimed to investigate the phylogenetic systematics of Euphractinae by including for the first time all extant species in an integrative approach using skull geometric morphometrics and molecular phylogenetic analyses based on mitochondrial and nuclear markers.
Within this context, we are particularly interested in testing the putative taxonomic distinctiveness of C. We used qualitative and quantitative approaches to evaluate the morphological differences between species.
With the qualitative approach, we analyzed the variation of the squamosal-jugal suture to assess its specific diagnostic value. Quantitative analyses of form were carried out through classical and geometric morphometrics as described below. We used 2 sets of samples. First, a total of 70 prepared skins assigned to C.
The 2nd set of samples comprised a total of skulls of adult euphractine species and was used to conduct the skull morphogeometric analyses. This latter set included a total of 9 specimens of C. The 9 specimens assigned to C. The specimens of C. Specimens assigned to C. We defined cranial landmarks 45 taken on each side and 10 on the midline; Fig. They include landmarks of type I anatomical , II mathematical , and III semilandmarks from the lateral side and midline.
Each specimen was digitized 3 times and a consensus specimen was used to account for measurement error. Landmark and semilandmarks 1—55 on the right side and midline, shown on outline of skull of C. Names and definitions of landmarks are given in Supporting Information S3. Skull is shown in 3 views: lateral above , dorsal middle , and ventral below. Semilandmarks are indicated in brackets. Landmark and semilandmarks 56— on the left side are not shown.
The morphological analyses were carried out in a morphogeometric framework Dryden and Mardia The data were analyzed using MorphoJ Klingenberg and Morpho package Schlager within R statistical software R Development Core Team , which supports the use of 3-dimensional sources. The centroid size was used as a proxy for size Goodall ; Dryden and Mardia According to Hood , centroid size is a geometric measure of size that follows the same mathematical behavior as body mass.
Other authors Frost et al. Differences in centroid size among euphractine species, and those between putative C. The sequential Bonferroni correction Rice was used to address the problem of multiple statistical tests. The 1st step consisted of landmark configuration superimposition using rotation, translation, and reflection transformations in order to remove spatial variation that did not correspond to form. This was achieved using the generalized Procrustes analysis Rohlf in R. The landmark configurations were projected from the curve Kendall to the tangent Euclidean space Bookstein , which allows exploration of shape variation using multivariate analyses e.
The Morpho package in R allows 3-dimensional rendering to visualize the change in shape along the principal component of interest. A scanned surface of a putative C. A color ramp was incorporated to the mesh material to represent the strength of shape change from the consensus. The patterns of shape change were evaluated in relation to the taxonomy, which allowed defining morphospaces.
Canonical variate analysis was used in MorphoJ to find the shape features that best distinguished species. To assess the influence of allometry, the landmark coordinates of aligned specimens were regressed on logtransformed centroid sizes using MorphoJ software. Then a new canonical variate analysis was performed on the regression residuals i. Discriminant analyses were performed using Procrustes superimposed coordinates half of crania to avoid colinearity and repeated using the principal component scores.
Correlation coefficients were calculated against the number of components to select the number of principal components to retain. This information was then used to determine how many variables were needed to summarize most shape variation Fadda and Corti ; Cardini et al. Discriminant analyses were also performed to evaluate species phenotypic distinctiveness Cardini et al.
We used this approach in 2 steps, building discriminant functions using subsamples learning sample consisting of 5 specimens from, 1st, each euphractine species, and, 2nd, the same specimens of C. These discriminant functions were used to classify the remaining specimens test sample that were not included for obtaining the functions.
This approach is equivalent to a hold-out sample cross-validation Hair et al. Over-fitting is avoided by predicting group affiliation using discriminant functions based on samples that do not include the specimens that are being classified. Instead of taking random subsamples, we modified the Cardini et al. The discriminant analysis was repeated times for the euphractine species and 56 times for localities to use all possible combinations of small samples i.
One of all possible combinations of the other categories was also selected randomly with the combinations eliminated after being used. Thus, the same 5 individuals per species or locality were never used twice. This series of cross-validated discriminant analyses allowed us to explore the robustness of predictions in which sampling error is very large, as may happen when poorly studied populations are included in the analysis.
We assessed the phylogenetic relationships of hairy armadillos by collecting both mitochondrial and nuclear markers for all currently described species. Biological samples were collected from 10 Andean hairy armadillos putative C. The animals came from different localities of the department of Oruro in Bolivia; the zoo staff provided information on capture sites and dates.
Two buccal swab samples were taken from each animal to harvest epithelial cells by gently rubbing sterile cotton swabs against the inner cheek. These samples were taken before the morning meal in order to prevent any possible contamination in the subsequent DNA extractions.
Buccal swabs were dried at room temperature and then wrapped individually in sterile covers. Following Poljak , who previously conducted phylogeographic studies of C. We additionally considered 10 individuals of Z. These markers were chosen in order to maximize nucleotide variability among the available sequences for euphractine armadillos: a screaming hairy armadillo C.
Andean hairy armadillo
The Andean hairy armadillo Chaetophractus nationi is an armadillo located in Bolivia , in the region of the Puna ; the departments of Oruro , La Paz , and Cochabamba Gardner, Nowark describes it as distributed in Bolivia and northern Chile. A recent publication of Pacheco also locates the species in Peru , basically in Puno Region. This species is also thought to be present in northern Argentina. The Andean hairy armadillo averages a tail length of three to seven inches and a body length of eight to sixteen inches.
Share this image — Hide sharing options. Armadillos are one of the oldest groups of mammals. Once thought to be closely related to turtles because of their tough protective carapaces , zoologists now classify them in the mammalian order Cingulata. Their closest relatives are anteaters and sloths 4.
Chaetophractus nationi is endemic to Bolivia and northern Chile, in the Andes mountain range. Yensen et al, Chaetophractus nationi lives in grasslands at high altitudes, in an ecosystem called the Puna. Montgomery, Head and body length reaches to mm and the tail length is 90 to mm. The head shield is 60 mm long and 60 mm wide.