DISTRIBUTION AND MORPHOMETRICS OF NATALUS STRAMINEUS FROM SOUTH AMERICA ( CHIROPTERA , NATALIDAE )

Morphometric and distributional data and some observations on the biology of Natalus stramineus Gray, 1838 collected in eastern Bolivia and in northern, northeastern, central, and southeastern Brazil are presented. All new records, combined with the records of the species from Paraguay and Mato Grosso, significantly change the known distribution of N. stramineus in South America. The specimens from northeastern Brazil (Rio Grande do Norte, Ceará, Bahia) are smaller than those found in the northern (Pará), eastern (Espírito Santo, São Paulo) and central regions of the country (Distrito Federal, Goiás, Mato Grosso do Sul). Natalus stramineus specimens from the three latter regions are about the same size, but are larger than those from Santa Cruz, Bolivia. Their size is intermediate between those of central samples and northeastern Brazil samples. The type locality of this species is discussed.


INTRODUCTION
At present, the monotypic family, Natalidae, includes three subgenera and five recognized species (KOOPMAN, 1993), but the specific or subspecific status of some taxa remains unclear.Recent detailed reviews of the taxonomic history of this family are presented by OTTENWALDER & GENOWAYS (1982) and by ARROYO-C. et al. (1997).The bats are most frequently found in the West Indies, including some islands of the coast of Venezuela.Two species of the subgenus Natalus (Natalus stramineus Gray, 1838 andN. tumidirostris Miller, 1900) are also found on the mainland.Natalus stramineus is known to be geographically distributed from northern Mexico to Venezuela, eastern Paraguay and eastern Brazil and throughout the Antilles (KOOPMAN, 1993;ARROYO-C. et al., 1997;LÓPEZ-GONZÁLEZ et al., 1998).There are few data available on the measurements of Brazilian specimens.RUSCHI (1951) provided external and cranial measurements of the holotype of Myotis espiritosantensis Ruschi, 1951(= Natalus stramineus, according to PINE & RUSCHI, 1976) from the State of Espírito Santo (ES), Conceição da Barra, Gruta do Rio Itaúnas.GOODWIN (1959) reported one external and eight cranial measurements of three specimens from Natal, Rio Grande do Norte (RN) and one specimen from Lagoa Santa, Minas Gerais (MG).CARTER & DOLAN (1978) presented cranial measurements of the syntype (The Natural History Museum, London, not numbered), actually the BMNH 70.2324 holotype, an adult male kept in spirits with its skull removed, according to HANDLEY & GARDNER (1990).WILLIG (1983) reported the measurements of two males from Chapada do Araripe, Ceará (CE).HANDLEY & GARDNER (1990) presented some external measurements of the specimens kept in the British Museum, including the BMNH 70.2324, considered to be the holotype.
Natalus stramineus is a species poorly represented in collections probably due to the difficulty in collecting this bat with mist nets set in open areas.Most individuals were captured in caves, their natural day roost.We had the opportunity to examine N. stramineus material collected in Bolivia and in northern, northeastern, southeastern and central Brazil.The aims were to present new records, morphometric and distributional data on this species, including some observations about its biology.Dial calipers were used to obtain 9 external and 12 cranial selected measurements, which were rounded up to the closest 0.1 mm.Measurements of the forearm and metacarpals were taken from the dorsal side and included the carpals.The following cranial measurements were taken: greatest skull length, from the posteriormost projection of the occipital to the anterior edge of the inner incisors; condyloincisive length, from the posteriormost projection of the exoccipital condyles to the anterior edges of the inner incisors; condylocanine length, from the posteriormost projection of the exoccipital condyles to the anterior edge of the canine; length of the maxillary and mandibular toothrows, from the anterior edge of the canine to the posteriormost projection of the last molar; length of the mandible, from the posteriormost projection of the condyloid processes to the anterior edges of the incisors; width across cingula canines, across the labial cingulum of canines; width across molars, maximum width across of the vestibular faces of the molars; posorbital constriction, smaller distance obtained at the postorbital constriction; breadth of the braincase, greatest width across the braincase; zygomatic width and mastoid breadth, greatest width obtained across zygomatic arches and mastoid processes, respectively.

Sixty
Standard statistics (mean, range, and standard deviation) were computed and, for a large sample, the mean measurements of males and females were compared by a t-test (SOKAL & ROHLF, 1969), significance level 0.05, in order to ascertain secondary sexual variation in size.

RESULTS AND DISCUSSION
Morphometrics.All the 60 specimens examined and measured (tabs.I, II) had completely fused phalangeal epiphysis and were considered adults.In male samples, mean cranial measurements were generally higher than those in females, but the same did not occur with regards to external measurements.As the sample from Paraíso, Mato Grosso do Sul (MS) was sufficiently large, males (N=15) were tested against females (N=10) in order to detect statistically significant sexual variation.Only one external (tibia length) and nine cranial features (greatest skull length, condylobasal length, condylocanine length, maxillary toothrow length, mandibular toothrow length, mandible length, width across cingula canines, posorbital constriction, and zygomatic width) were larger in males than in females.This fact corroborates the findings of OTTENWALDER & GENOWAYS (1982) who also detected secondary sexual variation in size for N. micropus Dobson, 1880.Thus, Length of maxillary toothrow 5 6.80 ± 0.10 6.7 -6.9 2 6.60 ± 0.14 6.5 -6.7 15 6.94 ± 0.11 ** 6.7 -7.1 10 6.78 ± 0.06 6.7 -6.9 1 6.90 -4 6.77 ± 0.0 5 6.7 -6.As pointed out by PINE & RUSCHI (1976), Myotis espiritosantensis Ruschi, 1951 is a junior synonym of Natalus stramineus.According to these authors, the specimens from Espírito Santo, due to their geographic localization, would be expected to be N. stramineus natalensis.However they are apparently too large for the species so they suggested to denominate them N. stramineus espiritosantensis (Ruschi).Some of the external and cranial measurements reported by RUSCHI (1951) were: forearm length: 40.0; third metacarpal length: 38.0; first phalanx length (digit III): 17.0; tibia length: 22.0; greatest skull length: 18.5; maxillary toothrow length: 8.0; mandible length: 13.0.The accuracy of some cranial measurements should be discussed, because they are too large when compared with the external measurements and the means of other Brazilian samples of the species.Thus, some cranial measurements taken from the M. espiritosantensis (RUSCHI, 1951:4) skull drawing, with bar scale, are more compatible with the external dimensions.For instance, greatest skull length: 16.6; maxillary toothrow length: 6.5 to 7.0; mandible length: 12.6.Based on these data, it is possible to state that the M. espiritosantensis (= N. stramineus) holotype is morphometrically similar to specimens of N. stramineus from central Brazil.Thus, some N. stramineus cranial measurements are smaller in individuals from a northeastern area (RN, CE, and BA); about the same size in the specimens from northern (PA), eastern (ES, SP) and central (DF, GO, MS) Brazil which, on their turn, are larger than those found in the individuals from Santa Cruz, Bolivia, which are intermediate in size between the central Brazil and northeastern samples.GRAY (1838), describing Natalus stramineus, specifically stated that the type locality was unkown.CABRERA (1958) restricted the type locality of this species to Lagoa Santa, MG, Brazil.GOODWIN (1959) disagreed with CABRERA (1958) and, on the basis of measurements and cranial features, stated that the holotype of N. stramineus agreed so exactly with the specimens from Lesser Antilles "that they can be confidently identified as the typical form".He observed that the measurements and cranial features of the specimens from Brazil, Venezuela, Trinidad, Central America, and Mexico definitely ruled out the possibility that Gray's type could come from any of these localities, restricting the type locality to Antigua, British West Indies, Lesser Antilles.GOODWIN (1959) observed that "While the specimen from Lagoa Santa in the Museum Lundi has not been seen by this author, Winge's (1893) drawing of the skull, reproduced natural size, corresponds so closely in measurements and characteristics to the type that it can be confidently referred to as natalensis.Winge's drawing, however, does show the palate as fractionally extended farther posteriorly than in Natal specimens".Thus, the specimens from Lagoa Santa apparently differ morphologically from those from Natal and, regarding cranial measurements, it appears to be a small representative of central and southeastern Brazil populations.Recently, HANDLEY & GARDNER (1990) reviewed the critical specimens of the British Museum and concluded that the specimen examined by GRAY (1838) and described as Natalus stramineus was an adult male in spirits (BMNH 70.2324) and they believed, as did GOODWIN (1959), that it came from the Lesser Antilles (restricted type locality, Antigua).The authors reported that "We are not aware of any South American populations clearly assigned to N. stramineus except for the widespread Brazilian population GOODWIN (1959) named N. stramineus natalensis".The specimens examined in the present study apparently constitute new source of information on morphometrics and distribution of this species in South America and type locality needs to be further discussed.GOODWIN (1959) recognized four subspecies and Varona (1974), apud ARROYO-C. et al., (1997) proposed six subspecies, being N. stramineus natalensis (eastern Brazil from Natal, RN, to Lagoa Santa, MG) the only occurrence in Brazil.PINE & RUSCHI (1976) considered that the specimens from Espírito Santo are apparently large for the species and "it seems appropriate to designate them as Natalus stramineus espiritosantensis (Ruschi)."As formely reports suggest there are at least two subspecies of N. stramineus in South America, further discussions about the type locality of this species are necessary.

Fig. 1 .
Fig. 1.Natalus stramineus from Puerto Suarez, Santa Cruz, Bolívia.(Photo: W. Uieda)canines: 4.0; width across molars: 5.8; posorbital constriction: 3.2; braincase breadth: 8.2; zygomatic width: 8.6; and mastoid breadth: 7.6) presented inCARTER & DOLAN (1978).As pointed out byPINE & RUSCHI (1976), Myotis espiritosantensis Ruschi, 1951 is a junior synonym of Natalus stramineus.According to these authors, the specimens from Espírito Santo, due to their geographic localization, would be expected to be N. stramineus natalensis.However they are apparently too large for the species so they suggested to denominate them N. stramineus espiritosantensis(Ruschi). Some of the external and cranial measurements reported byRUSCHI (1951) were: forearm length: 40.0; third metacarpal length: 38.0; first phalanx length (digit III): 17.0; tibia length: 22.0; greatest skull length: 18.5; maxillary toothrow length: 8.0; mandible length: 13.0.The accuracy of some cranial measurements should be discussed, because they are too large when compared with the external measurements and the means of other Brazilian samples of the species.Thus, some cranial measurements taken from the M. espiritosantensis (RUSCHI, 1951:4) skull drawing, with bar scale, are more compatible with the external dimensions.For instance, greatest skull length: 16.6; maxillary toothrow length: 6.5 to 7.0; mandible length: 12.6.Based on these data, it is possible to state that the M. espiritosantensis (= N. stramineus) holotype is morphometrically similar to specimens of N. stramineus from central Brazil.Thus, some N. stramineus cranial measurements are smaller in individuals from a northeastern area (RN, CE, and BA); about the same size in the specimens from northern (PA), eastern (ES, SP) and central (DF, GO, MS) Brazil which, on their turn, are larger than those found in the individuals from Santa Cruz, Bolivia, which are intermediate in size between the central Brazil and northeastern samples.