João Alves de Oliveira

EVALUATION OF THE PRINCIPAL-COMPONENT AND EXPECTATION-MAXIMIZATION METHODS FOR ESTIMATING MISSING DATA IN MORPHOMETRIC STUDIES

Abstract Vertebrate skeletons, particularly fossils, commonly have damaged, distorted, or missing structures. Because multivariate morphometric methods require complete data matrices, there are two possible solutions: to omit the specimens or characters having missing values, or to estimate missing values from the remainder of the data. Omission of specimens or characters reduces the data available for analysis, and thus the power to detect patterns or differences. Univariate and bivariate-regression methods are known to reduce the total variance of the data, and thus are not considered here. We compared the two most common multivariate methods: expectation-maximization (EM), which uses the covariance matrix directly, and principal-component (PC) estimation, based on regression of characters on principal components. Performance was evaluated by computer simulation of randomly introduced missing data in constructed data sets of known structure, and in several complete fossil (Pterodactylus skeleton) and recent (Alligator skeleton, Canis skull) data sets. The EM and PC methods displayed consistent and similar patterns of behavior for varying combinations of specimens and characters and across a broad range of amounts of missing data. Reliability was greatest for moderate numbers of characters (6–12) and larger sample sizes. For fewer characters the maximum amount of missing data that can be predicted increases substantially, but with a decrease in reliability. Both methods produce accurate estimates of missing values, but EM estimates are more precise. EM also outperforms the PC method in the maximum proportion of missing values that can be reliably estimated (almost 50% for small numbers of characters).

A new species of sigmodontine rodent from the Atlantic forest of eastern Brazil

We describe a new sigmodontine species on the basis of three specimens obtained from a high-altitude locality in the Atlantic forest of eastern Brazil. This new form, a small-bodied pentalophodont with tail longer than head and body, long soft fur, and a brownish ochraceous dorsum, is diagnosed by the presence of an open slit in the suture between the frontal bones in prepared skulls, and by a reduced diploid number of 20 coupled with a relatively high fundamental number of 34. Although the low diploid number suggests a derived sigmodontine, analyses of morphological characters and DNA sequence data (720 bp of the cytochrome-b gene) point to its placement within the recently described genusJuliomys González, 2000, a taxon regarded as belonging to an old and independent sigmodontine lineage. This finding reinforces current hypotheses of the Atlantic forest domain as an important center of diversification for a primitive sigmodontine stock. It also suggests that at least some surviving lineages, often considered rather ancient and unspeciose relicts, were subject to relatively more recent speciation events.

Morphological and Genetic Variation between Two Sympatric Forms of Oxymycterus (Rodentia: Sigmodontinae): An Evaluation of Hypotheses of Differentiation within the Genus

Abstract Species limits in the genus Oxymycterus have been controversial because of difficulties in detecting discrete morphological and cytogenetic variation. Several authors have considered many of the species described from Brazil to be conspecific with O. rufus, including O. dasytrichus described from Bahia. This tendency represents a null hypothesis for the species-level diversity within the genus. A recently proposed alternative hypothesis based on analyses of allopatric samples recognizes O. rufus and O. dasytrichus as valid species within rufus and dasytrichus species complexes, respectively. Two sympatric forms of Oxymycterus from Viçosa, Minas Gerais State, southeastern Brazil, are phenotypically similar to members of these 2 species-groups. Investigation on the status of these populations and their relationships to the rufus and dasytrichus complexes provided a test of the alternative hypothesis regarding diversity within the genus. Morphometric, morphological, and molecular (cytochrome-b DNA sequences) comparisons of samples representing the rufus, dasytrichus, and judex species-groups provided evidence that the 2 sympatric forms from Viçosa are distinct evolutionary units, one related to the dasytrichus species-group and the other to the rufus species-group. The structures of morphological and molecular variation within each species-group supported an isolation-by-distance model of phenotypic differentiation in the rufus group, and a smooth pattern of phenotypic differentiation along a north–south axis in the dasytrichus group. The relevance of these patterns for species delimitation within these groups and the validity of O. dasytrichus and O. rufus are discussed.

Biogeographic Determinants of Genetic Diversification in the Mouse Opossum Gracilinanus agilis (Didelphimorphia: Didelphidae)

The genetic variation of Brazilian populations of the mouse opossum Gracilinanus agilis was analyzed on the basis of the mitochondrial Cytochrome b gene (mt-Cytb) and the exon 28 of the nuclear Von Willenbrand factor (e28-vWF). The radiation of Gracilinanus was dated at 4.80 Ma, with the appearance of G. agilis around 1.93 Ma. Gracilinanus aceramarcae appeared as the first offshoot of the genus, followed by Gracilinanus emiliae and Gracilinanus microtarsus, which composed a sister clade of G. agilis. Phylogeographic analyses and genetic distance estimates indicate G. agilis as a single species, with haplotypes grouping in three well-supported clades, one from midwestern Brazil, a second one from northeastern Brazil, and a third one from eastern Brazil. Phylogeographic patterns in G. agilis were interpreted in search for congruence between genetic breaks and historic geomorphologic events documented for the region stretching northeastern to central-western of the Brazilian shield. The Rio São Francisco and the Serra Geral de Goiás were found to represent relevant geographic barriers to gene flow for G. agilis populations as well as for populations of several other widespread taxa.

Divergence in Zygodontomys (Rodentia: Sigmodontinae) and Distribution of Amazonian Savannas

Northern South America presents a diverse array of nonforest or savanna-like ecosystems that are patchily distributed. The distribution of these open habitats has been quite dynamic during Quaternary glacial-interglacial cycles; yet, the relevance of climatically driven vicariance events to the diversification of nonforest Amazonian vertebrates remains poorly known. We analyzed karyologic and mitochondrial DNA sequence data of the genus Zygodontomys, a small cricetid rodent distributed throughout nonforest habitats of northern Amazonia. Samples analyzed represented 4 Brazilian Amazonian localities and 2 French Guiana localities. Karyologic variation among Amazonian Brazilian Zygodontomys populations is high, with, at least, 3 karyomorphotypes. Molecular phylogenetic analyses recovered 3 major clades congruent with known karyotypes, a finding that suggests the existence of 3 species, 2 of which currently undescribed. The French Guiana and Surumú clade, identified as Zygodontomys brevicauda microtinus, is characterized by 2n = 86 and is sister to the clade formed by the 2 nondescribed forms. The Rio Negro-Rio Branco form is characterized by 2n = 82, and the Ferreira Gomes-Itapoá form is characterized by 2n = 84. The distribution of the 3 Zygodontomys lineages identified is in accordance with the geography of the open vegetation patches in Northern Amazonia, and divergence time estimates relate speciation events to the middle-upper Pleistocene, supporting the prominent role of Quaternary climatically driven vicariance events in the diversification of the genus.

Current and potential distribution of Myotis simus (Chiroptera, Vespertilionidae)

Abstract Myotis simus is apparently restricted to tropical and subtropical South American lowlands, with a possible disjunction isolating northern and southern populations. Twenty-eight museum and literature records were assembled and analysed in the context of a taxonomic review of South American species of Myotis. In order to model the distribution of M. simus, to reveal putative areas of occurrence and environmental constraints to its distribution, as well as to test the previously proposed hypothesis of disjunct distribution, Maximum entropy algorithm (MaxEnt) was implemented on the information retrieved from the sampling localities, using nine environmental variables. Two regions with increased probability values were revealed in the Amazon and Paraná basins, connected by a bottleneck in southeastern Bolivia, which provides further support for the previously proposed hypothesis of disjunctive distribution. The predicted distribution for M. simus was strongly associated with the drainage basins, precipitations of the driest quarter, mean temperatures of the warmest quarter and altitude. The Andean eastern slopes and the Guyana, Paraná and Central Brazilian plateaus delimit the geographical distribution of M. simus, and the confirmed records document its presence in both terra firme and floodplain areas in lowland forest and savanna formations across South America.

Review of Myotis (Chiroptera, Vespertilionidae) from Northern South America, Including Description of a New Species

ABSTRACT We describe a new species of bat in the genus Myotis (Vespertilionidae, Myotinae) from the coastal mountains of Venezuela. The new species (Myotis handleyi, sp. nov.) can be distinguished from other South American congeners by the following set of traits: dorsal fur long, silky, and bicolored with burnished tips; skull long; rostrum long and broad; frontals moderately to steeply sloping; sagittal crest absent or very low; plagiopatagium broadly attached to the foot at the level of the base of the toes; fringe of hairs along the trailing edge of uropatagium absent; and fur on uropatagium not reaching knees. We review Colombian and Venezuelan samples of Myotis, covering all of the currently recognized species known from these countries. Based on our analyses, we provide a sketch of the taxonomic diversity of the genus in Colombia and Venezuela, along with a key to their identification. Among other conclusions, we elevate M. nigricans caucensis to the species level; confirm that populations of M. nigricans from the opposite sides of the Andes represent the same taxon; retain J.A. Allens names M. esmeraldae, M. bondae, and M. maripensis in the synonymy of M. nigricans; and document clinal variation in size along an altitudinal gradient for M. nigricans, with larger specimens from higher elevations. This research, based on museum and field collections, is one of a series of studies by the senior author re-evaluating species limits among Neotropical Myotis.

Morphometric and morphological variation in South American populations of Myotis albescens (Chiroptera: Vespertilionidae)

Myotis albescens (E. Geoffroy, 1806) occurs from Mexico to Uruguay and Argentina. Despite a large number of specimens in collections, its variability in South America has been underestimated, potentially leading to errors in identification. In order to clarify the taxonomic limits of M. albescens and to evaluate previous hypotheses of geographic variation in size we analyzed the type material and studied the variability in South American samples using multivariate exploratory and confirmatory procedures, as well as frequency analyses of discrete morphological data. The presence of a fringe of hairs along the trailing edge of the uropatagium, the long and silky pelage with frosted appearance on the dorsum, ear 9 to 14 mm long, broad interorbital and postorbital constrictions, and a globular braincase were identified as the most useful traits to distinguish M. albescens from its South American congeners. In agreement with Bergmans rule, larger specimens were found in the South. Beyond the geographic component, Individual variation is an important factor affecting the variability in the size and shape of the skull and pelage color.

Taxonomic implications of cranial morphometric variation in the genus Clyomys Thomas, 1916 (Rodentia: Echimyidae)

Abstract Spiny rats of the genus Clyomys Thomas, 1916, include 2 recognized species both described on the basis of a small number of specimens, C. laticeps (Thomas, 1909) comprised of the nominotypical form and subspecies C. laticeps whartoni Moojen, 1952, and C. bishopi Avila-Pires and Wutke, 1981. Univariate and multivariate analyses of craniodental measurements obtained from 78 adult specimens from 8 pooled geographic samples representative of the distributional range of Clyomys structured the samples morphometrically into 2 geographic groups, which do not correspond to the taxonomic units postulated in previous studies. These morphometric groups were checked against data on qualitative skull morphology and pelage coloration from 128 specimens in a search for patterns supporting the taxonomic recognition of the morphometric groups revealed. The morphometric and qualitative analyses do not corroborate the current taxonomic framework of the genus Clyomys, and indicate that C. bishopi is a junior synonym of C. laticeps.

A Taxonomic Update of Small Mammal Plague Reservoirs in South America

Plague is a disease of epidemic potential that may emerge with discontinuous outbreaks. In South America, 50 wild rodent species have been identified as plague reservoirs, in addition to one lagomorph and two marsupials. To review the nomenclature of plague reservoirs, we examined specimens collected in plague foci, carried out new surveys in Brazilian plague regions, and re-evaluated the nomenclature of South American reservoirs on the basis of the current literature. Five of the 15 species involved with plague in Argentina, three of 10 species involved with plague in Bolivia, three of the seven species involved with plague in Peru, five of the nine species involved with plague in Ecuador, and six of the nine species involved with plague in Brazil have undergone taxonomic changes. In the last 20 years, plague cases were recorded in Bolivia, Brazil, Ecuador, and Peru. These four countries have a high rodent species richness in plague foci, a fact that may be decisive for the maintenance of plague in the wild.