Behavior and demography in an urban colony of Tadarida brasiliensis (Chiroptera: Molossidae) in Rosario, Argentina

Bat colonies were sampled in the city of Rosario to increase the understanding of bat ecology in urban areas of the southem cone of South America. Seven species were recorded, of which three are new records for Rosario. One representative colony was chosen for intensive ecological study. ApproximateJy 64 000 Tadarida brasiJiensis forrned a matemity colony in the attic of an old building. Most of the bats were pregnant or lactat­ ing females and their young .. Adults arrive in the colony in mid-September and ¡eave in February, no bats were present at tbis site from the beginning of March to mid-September. Births occur between mid-November and mid-December. Pups roosted in compact c1usters in the nursery areas, spatially segregated from adults. Densities of these aggregations were 643 ± 76 bats/m2 (p < 0.20). and adult densitíes were 161 ± 21 bats/m2 (p < 0.20). 182 bats were captured. They were sexed and weighed. Mean adult females mass was, 13.26g before parturition and 12.01g after parturition. Recorded data inc1uded daily activity pattems .. Barn Owls (Tyto alba) and domes­ tic cats prey on bats leaving the colony. Rabies virus was no! detected. T he insect biological control exerted by this colony was estimated and it would be of 209 to 385 kg of insects nightly between September and late February, demonstrating the important role they play in the urban ecosistem.

Relatively recent studies have begun to shed light on the roles that bats play in various ecosystems, in aspects such as pollination, dis persal of propagules, control of insect popula tions, etc. (Estrada and   Whitaker et al. 1996).This has begun to change attitudes towards bats, with the result that now many countries have specific legislation for their protection and that there are intemational treaties for the protection ofmigratory species (Anonymous 1981, Racey 1992, Anonymous 1994a, Anonymous 1994b).

MAlERIALS AND METHODS
The study was conducted in the city of Rosario, Argentina, (32° 57' S, 60° 39' W), Sex, weight, and age class were de ter mined for each captured bat, distinguishing "sub-adults" (those born duríng the season under study) from "adults" (those born before the season).FUf color and teeth sharpness were used to determine age class as suggested by Short (1961), with grey••colored individu als being sub-adults and red-brown individu als adults.In addition, 20 bat pups caught within 24 hours of birth were weighed and measured.
Day and night observations were recorded to establish behavior patterns.Times of depar ture and arr ival of bats were recorded continu ously at the attic entrance, and at hourly inter vals within the roost.
A smalJ number of samples for virological studies was taken.Briefly: i) 125 bats were anesthetized by ether, then their brain, sub scapular fat and salivary glands were removed from each of them.A 10 % suspension by mas s was prepared and centrifuged for 5 minutes at 150-200 g, to remove tissue particles.The cJear supernatant of each sample was man tained at 4 oC and then inoculated in adose volume of 0.03 mI by intracerebral route, according lo the standard mouse inoculation test (Koprowski, 1996).The animals were checked daily, the number of mice fouad nor mal, sick or dead were recorded as a perma neot record of the experiment.
ii) One hundred percent brain samples (125) were submited to fluorescent antibody tests, to detect rabies antigens in fresh speci meno Four smears of each brain sample were prepared.After mixing in acetone, they were stained with fluorescent rabies antibody conju gate produced at CePanZo (PHO/WHO), according to the technique described in Dean, et al. (1996).
iii) Simultaneously, rabies neutralizing antibodies were determined in serum samples from 32 individuals from the same colony using the standard procedures in mice blood samples (Delpietro, 1972) that were diluted 1: 4 in PBS sterile solution when blood samples were removed.
The attic of the building chosen was occu pied by T. brasiliensis and about 100 M. levis.Between September and February, most of the T. brasiliensis were pregnant or lactating females and their young.No bats were present al this sÍte from the beginning of March to sometime in September.The nursery area was unoccupied by adults until the onset of births.Parturition began in mid-November (9 and 10 in 1988; 12 and 13 in 1989) and lasted for about one month.Young were born naked and averaged 2.70g, ± O.03g (n = 20), with length of forearms averaging 18.01mm, ± 0.29mm (n = 20).By five weeks of age, the bats began to take theír first flights.The first females that gave birth roosted on the higher parts of the walls and successive bats in the lower areas.Young bats roosted in the nursery areas, but in response to disturbance, females moved their young to other parts of the building.
Roosting T. brasiliensis adjusted their behavior according to weather conditions.On cooler days (temperatures < 23°C), the bats were less active and tended to cluster together, while on very hot days (temperatures > 27°C) the bats were more evenly distributed on the roost surface.On cloudy days the bats began to emerge from the colony 8-10 min earlier than on clear days.
In late October and early November (mid Spring) the nightly pattems �f departure from and retum to the roost were as follows: during the first 10-15 minutes after approximately 2000 h, small groups of five or six bats began to emerge, and over the next two hours the groups gradually increased in size until there was a massive flow of bats.By 2200 h the departure had stopped.The retum started about 0245 h with small groups and increased until there was a massive flow of arriving bats.By 0530 h no bats were observed flying into the building.This pattem prevailed at all five attic entrances used by the bats.
Most bats were away from the roost for about five to six hours.Before the onsel of par turition, all of the bats emerged.Afterwards, sorne adults remained in the nursery are as of the roost after the others had left.
Bam owls (Tyto alba) were always seen around the building, attacking emerging bats.Domestic cats also preyed on bats, typically catching those that had fallen to the ground.
In bats that were sampled for virological studies, neither rabies virus was isolated by the mice inoculation tests nor rabies antigens were detected by the FA tests.Rabies neutralizing anti bodies were found in 5 to 32 bat sera (three adults females, two young males) at 1: 8 dilution.

DISCUSSION
Based on the species recorded in this study, and those previously cited for the region (Barquez el al., 1993), at least ten species had be en recorded for the city of Rosario.The dis tribution areas reported by Barquez et al. (1993) suggest that olher species may also be found in Rosario.A more detailed and specific study would be necessary to clarify local pat tems of distribution.
There are numerous reports of predation on T. Brasiliensis colonies (Twente 1956, Glass 1958, Barbour and Davís 1969, Tuttle 1994), and our study corroborated these find ings as various T. alba were observed living in the same building at the same time as the bats.This relation was studied in more detail (Romano et al. in prep.).
In recent years a growing number of stud ies have sought to deepen our understanding of interspecific ecological interactions, demonstrating the vital roles that many species of bats play in terrestrial ecosystems.Thus, when discussing the "management" of bat populations in urban ecosystems, it is important to balance the benefits provided by these bats against the risks they might repre sent for humans and domestic animals (Constantine 1970, Tuttle 1988).
Based on the relatively low incidence of rabies in the present study, they suggested a low risk of transmition from T. brasiliensis from this colony to the human population of Rosario.
Moreover if each adult T. brasiliensis con sumed 50 % to 100 % (in nursing mothers) of its body mass in insects each night (Fenton 1983, McCracken & Gustin 1987, Tuttle 1994, Kunz, et al. 1995 a), the bats in this colony would eat 209 to 385 kg of insects nightly between September and late February.These figures demonstrate the predation pressute exerted by T. brasiliensis on populati ons of nocturnal insects and theirpossible role in the biological control of sorne insects pests, just as it is documented in a recent report of McCracken (1996), this should help improve their public image.

RESUMEN
Con el objetivo de incrementar el conocimiento de la ecología de los murciélagos en �as urbanas, se mues trearon colonias en la ciudad de Rosario.Fueron registra das siete especies, de las cuales tres son nuevos registros.

Latin
America has not yet reached this same level of protection.By comparison, there has been little systematic research on the biol ogy and ecology of most neo tropical species, and consequently a lack of information and public education campaigns on the crucial eco logical roles played by many species of bats.This is further aggravated by the general belief that bats are dangerous animals, associated with diseases such as rabies (Constantine et al.

1968,
Anonymous 1984, Tuttle 1985, Rosatte 1987, Fenton 1992, Bredt et al. 1996).This sit uation has led to the extermination of impor tant bat populations, a common practice in rural areas throughout Latin America, and par ticularly in cities where both large and small colonies associated with buildings are system atically eliminated.The great majority of such colonies are composed of beneficial insecti vores, fructivores, or polinivores, with the result that their persecution has highly nega tive consequences.The purpose of this study was to sample and detect the presence of urban bat colonies, to chose one sufficiently representativefrom the point of view of size, location, degree of conflict, etc., and to conduct ecological studies of it to generate basic information on the dynamics of these mammals in an urban ecosystem, as a basis for future management decisions.As a secondary aspect of this study, sam pIes of bats were tested for the rabies virus.No cases of rabies have been detected in humans or domestic animal s in Rosario for more than fifteen years.No systematic studies in wild animals have been done.Nevertheless, there is a latent concem about rabies, as it has been found in domestic animals and bats in regions close to the province of Santa Fe (Delpietro et al. 1972, 1987, De Diego and Valotta 1981).The city of Rosario is within the distrib u tion range of seven species of bats, four ves pertilionids: Eptesicus diminutus, Eptesicus furinalis, Lasiurus cinereus, and Lasiurus ega, and three molossids: Tadarida brasiliensis, Eumops bonariensis, and Eumops patagonicus (Barquez et al. 1993).
45m aboye sea leve!.The climate is temperate subtropical with a median annual temperature of 16.7°C (-8.8/41.9°C)and a mean annual precipitation of 1 016 mm, principally in the summer and autumn.A survey was conducted between December of 1987 and February of 1988 in Rosario to locate bat colonies and select one for more detailed study.To locate colonies, people from various neighborhoods were asked whether there were bats in their area.Colonies detected in this way were visited, and the largest and most accessible (and thus pos sibly the one with the greatest potential for conflict) was selected for direct observation.T he selected colony inhabits the attic (650 m 2 ) of th e LawFac uI ty (National University ofRosario),built in 1888 and located down town.This site Was visited 60 times from 1988 to 1990 to observe, photograph, and sample lOe bats.Various counting methods were analysed for determining the populatíon size (Twente 1956, Humphrey 1971, Freeman et al. 1988, Fraze et al. 1990).Finally a photographic count method was designed and used to estí mate the number of bats in the roost.Initial visits showed that the bats distributed them selves in two quite different areas: the "aduIt" area and t h e "nursery" area; with areas• of 206 m 2 and 49 m 2 respectively.Paint was used to delineate a grid with l-m intervals.Two sets of 20 photographs were taken, the fr rst between 30 September and 15 November 1989 (aduIt area) and the second between 30 November and 15 January 1990 (nursery area).The num ber of bats was determined from the pho tographs. .Stratified sampling (Caughley 1978) was U sed to estimate bat populations in the aduIt and nursery areas, strata A and B respectively.The samples included 17 m 2 in stratum A and 6m 2 in stratum B. The mean population size and standard deviation were caIculated for each area, and taking into account the total occupied roosting area, the total number of bats in the roost was calculated.An 80 % reli ability level was used according to the formu la of Mood and GraybilI (1978).A metal box-trap designed for this pur pose was used to capture bats in the roost.This trap had a 20-1 capacity, and a door acti vated b y the weight of the bats.The trap was mounted on a S-m long extensible• handle allowing sampling of bats on the wal1s and ceiling of the attic.