All-offspring dispersal in a tropical mammal with resource defense polygyny

In polygynous mammals, males are usually responsible for gene flow while females are predominantly philopatric. However, there is evidence that in a few mammalian species female offspring may disperse to avoid breeding with their father when male tenure exceeds female age at maturity. We investigated offspring dispersal and local population structure in the Neotropical bat Lophostoma silvicolum. The mating system of this species is resource defense polygyny, with the resource being active termite nests, excavated by single males, which are then joined by females. We combined field observations of 14 harems during 3 years and data about the genetic structure within and between these groups, calculated with one mitochondrial locus and nine nuclear microsatellite loci. The results show that both male and female offspring disperse before maturity. In addition, we estimated life span of excavated termite nests and the duration they were occupied by the same male. Our findings suggest that long male tenure of up to 30 months is indeed a likely cause for the observed dispersal by female offspring that can reach maturity at a low age of 6 months. We suggest that dispersal by offspring of both sexes may occur quite frequently in polygynous tropical bats and thus generally may be more common in mammals than previously assumed.     

Mean colony relatedness is a poor predictor of colony structure and female philopatry in the communally breeding Bechstein's bat (Myotis bechsteinii)

In order to understand why animals are social and how group members interact with each other it is important to know their relatedness. However, few studies have investigated the genealogy in complete social groups of free-living animals with low reproductive skew. This holds particularly true for bats. Although almost all bat species are social, their sociobiology is not well understood. Because they are volant, nocturnal and have a rather cryptic life-style, bats are difficult to observe in the wild. Furthermore females are generally gregarious making genetic parent-offspring assignment a challenging task. We used genetic markers in combination with knowledge about age and colony membership of individually marked bats to construct pedigrees in completely sampled maternity colonies of Bechstein's bats (Myotis bechsteinii). Despite considerable fluctuations in population size, no immigration occurred over 5 years in four colonies living in close proximity. Additionally, confrontation tests showed that females of one maternity colony were able to detect and attempted to prevent the intrusion of foreign females into a roost they occupy. Although colonies were absolutely closed, and 75% of the colony members lived together with close relatives (rS=0.25), mean colony relatedness was nearly zero (0.02). Average relatedness therefore is a poor estimator for the potential of kin selection in Bechstein's bat colonies and may be misleading when attempting to understand the social structure of animals living in groups where many members breed. Based on our results we discuss the potential adaptive value of living in closed societies with low reproductive skew.     

Storage and display of odour by male Saccopteryx bilineata (Chiroptera, Emballonuridae)

Males of the sac-winged bat, Saccopteryx bilineata, actively fill their propatagial sacs with secretions from the genital region, the gular gland, urine and saliva. From our observations and those of Starck we deduce that propatagial sacs in S. bilineata do not have a glandular function, but are instead organs for the storage and display of odours. In addition to the already known “salting” and hovering behaviour of male S. bilineata, we describe in detail how odour is fanned to roosting individuals during the complex, stereotypic hovering displays. S. bilineata males also coat the fur of their backs with saliva using the wing tip and might scent-mark territory boundaries. “Yawning” may represent a visual as well as an olfactory cue. Odour seems to play an important role in the social communication of S. bilineata and in other emballonurids, as revealed by the broad distribution of wing sacs in this family. S. bilineata males display odour during energetically costly hovering flights in front of females. We demonstrate that the number of hovering displays increases with harem size. The mating effort of S. bilineata males comprises a multimodal signalling behaviour. Although males defend harem territories in which females gather, females seem to be able to choose the father of their progeny freely among the males of a colony. This may have led to the evolution of the complex mating displays by male S. bilineata. Received: 9 December 1998 / Received in revised form: 6 May 1999 / Accepted: 13 June 1999     

Behavioural determinants of gene flow in the bat Miniopterus schreibersii

Dispersal and migratory behaviours are often important determinants of gene flow in wild species, and we have studied their role using ringing-recapture data in the Portuguese population of Miniopterus schreibersii, a cave-dwelling bat that forms large maternity colonies. Juvenile dispersal, usually a major agent of gene flow, appears to be negligible, as young females never settled to give birth in foreign colonies. Likewise, there was virtually no dispersal of adult females to foreign maternity colonies. This strong philopatry virtually eliminated female-mediated gene flow, but we found a great potential for male-mediated gene flow among colonies, as regional migrations temporarily joined both sexes from different colonies in the same roosts, during the mating season. In fact, females from some colonies were more likely to mate with males from foreign colonies than from their own, thus potentially bringing home genes of foreign males. In spite of this abundant gene flow, we found a pattern of isolation by distance and even strong barriers to gene flow, which is explained by the fact that migrations were usually too short to allow direct flow among distant colonies. We concluded that potential gene flow is influenced by the distance between colonies and the availability of mating roosts between them. In addition, we found this flow to be asymmetrical, with a dominant direction from the largest to the smallest colonies. Our ringing-recapture estimates of potential gene flow based on dispersal and migratory behaviour are compatible with the genetic structure of the population for both nuclear and mitochondrial DNA. Our conclusions have implications for the conservation of bats with a spatial behaviour similar to that of M. schreibersii. Colonies should be managed individually because strict female philopatry not only promotes their isolation but also minimises the contribution of immigration in rescuing declining colonies. Furthermore, the results underline the importance of preserving mating roosts to maintain gene flow among colonies.     

Social organization in the bat Pipistrellus pipistrellus

Summary The social organization of the pipistrelle bat (Pipistrellus pipistrellus) was studied by means of bat boxes in southern Sweden. The males set up territories around a roosting site in the beginning of the summer at the same time as the females formed nursing colonies. After breeding, the females joined the single males in their day roosts establishing transient mating harems. Subsequently, immatures arrived at the mating grounds. The immature females, which probably attained sexual maturity during their first autumn, were admitted to the day roosts of the harem males, in contrast to the immature males. The size of the harem was dependent on the total number of females present on the mating grounds. The size, however, was also restricted by some factor, presumably the quantity of food resources in the surroundings of the specific roost site, or the capability of the harem male for mating. The mating system in the pipistrelle bat is best characterized as a resource defence polygyny. Available data on other related temperate species indicate a similar social organization in Pipistrellus nathusii and Nyctalus noctula.     

The effects of age and relatedness on mating patterns in thornbug treehoppers: inbreeding avoidance or inbreeding tolerance?

The social environment of many species includes synchronous maturation of siblings in family groups, followed by limited dispersal of adults from their natal site. Under these conditions, females may experience high encounter rates with same-age siblings during mate searching, increasing their risk of inbreeding. If inbreeding depression occurs, mating with a sibling is often considered maladaptive; however, in some contexts, the inclusive fitness benefits of inbreeding may outweigh the costs, favoring females that tolerate some level of inbreeding depression. We evaluated mating patterns in the treehopper Umbonia crassicornis, a semelparous species in which females encounter same-age siblings during mate searching. A female U. crassicornis that mates with a brother suffers from inbreeding depression. We used a free-choice mating design that offered females simultaneous mating opportunities with three groups of males: siblings, same-age nonsiblings, and older nonsiblings. These groups represent the types of males typically encountered by females during mate searching. Our goal was to assess whether mating patterns were influenced by inbreeding avoidance by evaluating two hypotheses: kin discrimination and age-based mating (older males cannot be siblings in this species). There was no difference in the proportions of females mating with siblings vs nonsiblings, suggesting an absence of kin discrimination. However, females mated with a greater proportion of older vs younger males. Given that females do not avoid siblings as mates despite a cost to inbreeding, our results provide a possible example of inbreeding tolerance. We also discuss some factors that may have contributed to the mating advantage of older males.     

Social influences on territorial signaling in male greater sac-winged bats

Acoustic territorial displays are common among birds but comparatively rare among mammals. An exceptionally vocal mammal well-known for its elaborate territorial displays is the polygynous greater sac-winged bat, Saccopteryx bilineata. Male S. bilineata are often philopatric and establish small territories in their birth colony in which females can roost during the day. During territorial defense, males produce complex territorial songs that are learned through vocal imitation. Territorial songs are mainly produced at dawn and dusk. We studied social influences on male vocal activity and the occurrence of vocal signatures in territorial songs of 27 male S. bilineata from 12 different-sized colonies in Panama. Males produced significantly more territorial songs when they had more territorial neighbors or when they had females roosting in their territories, indicating that male vocal activity rises with increasing male–male competition. Territorial songs are multisyllabic vocalizations with low-frequency buzz syllables being most prominent. We found statistical evidence for a pronounced individual signature encoded in the buzz syllables of territorial songs that could facilitate individual recognition among rival neighbors. Additionally, we found a vocal group signature in territorial songs, suggesting that young males may learn territorial songs from more than one tutor male. Resident male S. bilineata appear to cooperatively defend their colony against male intruders, making a group signature in territorial songs potentially advantageous.     

The social organization of the common vampire bat

Summary At a site in Costa Rica, three groups of 8–12 adult female vampire bats, Desmodus rotundus, utilize group-specific sets of hollow trees as day roosts. Long-term nonrandom associations between pairs of females, as measured by the proportion of time one bat spends roosting in the same tree with another bat over a 3 year period, occur even when preferences for particular trees are removed. Significant associations exist between both related and unrelated adult females. Adult male bats, however, show few associations with females or other males. By observing bats within trees and while foraging, and by monitoring feeding flights with radiotelemetry, the following potential benefits of association could be tested. Females roost together to (1) share a suitable microclimate, (2) avoid predators, (3) avoid ectoparasite infestations, (4) minimize travel to mobile prey animals, (5) respond to coercive males, (6) feed simultaneously from a bite, (7) remove ectoparasites by allogrooming, and (8) share food by regurgitating blood to other bats within roosts. The data do not indicate that any of the first five hypotheses provide significant benefits for long-term associations although predators and ectoparasite levels may cause occasional changes in roost sites. Simultaneous feeding was uncommon and apparently confined to females and their recent offspring. Allogrooming, although common, occurred independently of the presence of ectoparasites. Food sharing, however, occurred between both related and unrelated adult females with high levels of association and provides at least one selective advantage for maintaining cohesive female groups.     

Dispersal behaviour and colony structure in a colonial spider

Group living in spiders is characterised by two principle modes, the cooperative social mode and the colonial non-cooperative mode. Kin-relationships due to reduced dispersal determine population genetic structure in social spiders, but the dispersal mechanisms underlying group structure remain poorly understood in colonial spiders. Assuming similar ecological benefits of group living, we address the question whether reduced dispersal shapes population structure in a colonial spider, Cyrtophora citricola (Araneidae). We analysed dispersal by studying settling decisions under semi-natural conditions in experimental trees with and without colonies, and in natural populations, we estimated dispersal and colony structure using population genetic analyses. The propensity to disperse decreased with increasing age in experimental colonies. Adult females did not disperse in the experiment. Sub-adult female spiders preferred trees with a colony to trees without a colony. Dispersal in third instar juveniles was influenced significantly by wind but not by the presence of a colony. Thus, we showed that being in a colony did not inhibit juvenile dispersal, but pre-mating females were philopatric. Genetic differentiation among colonies in natural populations was heterogeneous, colonies being either little or highly differentiated. The heterogeneous structure is likely caused by colony founding by one or a few females followed by dispersal among perennial colonies. Gene flow, however, was slightly male-biased. The experimental and indirect, genetic approaches combined showed that dispersal and the breeding system of C. citricola resemble that of solitary spiders, with juvenile dispersal occurring in both sexes, while the colonial distribution is maintained by female philopatry.     

Copulation duration,but not paternity share,potentially mediates inbreeding avoidance in Drosophila montana

Studying the incidence of inbreeding avoidance is important for understanding the evolution of mating systems, especially in the context of mate choice for genetic compatibility. We investigated whether inbreeding avoidance mechanisms have evolved in the malt fly, Drosophila montana, by measuring mating latency (a measure of male attractiveness), copulation duration, days to remating, offspring production, and the proportion of offspring sired by the first (P₁) and second (P₂) male to mate in full-sibling and unrelated pairs. SNP markers were used for paternity analysis and for calculating pairwise relatedness values (genotype sharing) between mating pairs. We found 18 % inbreeding depression in egg-to-adult viability, suggesting that mating with close relatives is costly. Copulation duration was shorter between previously mated females and their brothers than with unrelated males. Based on an earlier study, shorter copulation is likely to decrease the number of inbred progeny by decreasing female remating time. However, shorter copulations did not lead to lower paternity (P₂) of full-sibling males. Progeny production of double-mated females was lower when the second male was a full-sibling as compared to an unrelated male, but we could not distinguish between inbreeding depression and lower female reproductive effort after mating with a relative. Relatedness estimates based on 34 SNPs did not detect any quantitative effect of relatedness variation on copulation duration and progeny production. We suggest that inbreeding depression has been strong enough to select for inbreeding avoidance mechanisms in our Finnish D. montana population.     

Effects of Artificial Roosts for Frugivorous Bats on Seed Dispersal in a Neotropical Forest Pasture Mosaic

Abstract:  In the Neotropics ongoing deforestation is producing open and heavily fragmented landscapes dominated by agriculture, mostly plantations and cattle pastures. After some time agriculture often becomes uneconomical and land is abandoned. Subsequent habitat regeneration may be slow because seed inputs are restricted by a lack of incentives—such as suitable roost sites—for seed dispersers to enter deforested areas. Increasing environmental awareness has fostered growing efforts to promote reforestation. Practical and cost-efficient methods for kick-starting forest regeneration are, however, lacking. We investigated whether artificial bat roosts for frugivorous bat species can attract these key seed dispersers to deforested areas, thereby increasing seed rain. We installed artificial bat roosts in a forest-pasture mosaic in the Costa Rican Atlantic lowlands and monitored bat colonization and seed dispersal. Colonization occurred within a few weeks of installation, and 10 species of bats occupied the artificial roosts. Five species of frugivorous or nectarivorous bats colonized artificial roosts permanently in both primary habitat and in deforested areas, in numbers similar to those found in natural roosts. Seed input around artificial roosts increased significantly. Sixty-nine different seed types, mostly of early-successional plant species, were transported by bats to artificial roosts in disturbed habitats. The installation of artificial bat roosts thus successfully attracted frugivorous bats and increased seed inputs into degraded sites. This method is likely to speed up early-vegetation succession, which in turn will attract additional seed dispersers, such as birds, and provide a microhabitat for seeds of mid- and late-successional plants. As well as supporting natural forest regeneration and bat conservation, this cost-efficient method can also increase environmental awareness among landowners.     

Echolocation beam shape in emballonurid bats, Saccopteryx bilineata and Cormura brevirostris

The shape of the sonar beam plays a crucial role in how echolocating bats perceive their surroundings. Signal design may thus be adapted to optimize beam shape to a given context. Studies suggest that this is indeed true for vespertilionid bats, but little is known from the remaining 16 families of echolocating bats. We investigated the echolocation beam shape of two species of emballonurid bats, Cormura brevirostris and Saccopteryx bilineata, while they navigated a large outdoor flight cage on Barro Colorado Island, Panama. C. brevirostris emitted more directional signals than did S. bilineata. The difference in directionality was due to a markedly different energy distribution in the calls. C. brevirostris emitted two call types, a multiharmonic shallowly frequency-modulated call and a multiharmonic sweep, both with most energy in the fifth harmonic around 68?kHz. S. bilineata emitted only one call type, multiharmonic shallowly frequency-modulated calls with most energy in the second harmonic (~46?kHz). When comparing same harmonic number, the directionality of the calls of the two bat species was nearly identical. However, the difference in energy distribution in the calls made the signals emitted by C. brevirostris more directional overall than those emitted by S. bilineata. We hypothesize that the upward shift in frequency exhibited by C. brevirostris serves to increase directionality, in order to generate a less cluttered auditory scene. The study indicates that emballonurid bats are forced to adjust their relative harmonic energy instead of adjusting the fundamental frequency, as the vespertilionids do, presumably due to a less flexible sound production.     

Inbreeding avoidance in a poeciliid fish (Heterandria formosa)

Polyandry is hypothesized to give females an opportunity to avoid inbreeding through postcopulatory selection mechanisms if precopulatory inbreeding avoidance is not possible, for example, because of forced matings. Here, we report a postcopulatory, prefertilization, inbreeding avoidance mechanism in the least killifish, Heterandria formosa, a species in which males can force matings. Females had 50% less sperm in their oviducts and ovarian cavities after mating with a sibling compared to a mating with a nonsibling male. Neither sex showed inbreeding avoidance in a dichotomous precopulatory mate choice test or during mating trials. Females in this species invest substantially in each offspring after fertilization (matrotrophy), whereas males invest little more than sperm. Based on theory, females should therefore be more likely than males to avoid inbreeding in this species. We suggest that females do this by reducing the amount of sibling sperm in their reproductive system. However, the possibility that males invested more sperm to nonsiblings could not be ruled out. In a fertilization success experiment, the first male to mate with a female sired all the offspring in most cases, even if it was a sibling. However, large females were more likely to carry offspring of multiple males. Possibly female sperm storage sites were filled by the first male, and only large females had space for the second male's sperm.     

Incest avoidance,fluctuating asymmetry,and the consequences of inbreeding in Iridomyrmex humilis,an ant with multiple queen colonies

Summary Inbreeding may have important consequences for the genetic structure of social insects and thus for sex ratios and the evolution of sociality and multiple queen (polygynous) colonies. The influence of kinship on mating preferences was investigated in a polygynous ant species, Iridomyrmex humilis, which has within-nest mating. When females were presented simultaneously with a brother that had been reared in the same colony until the pupal stage and an unrelated male produced in another colony, females mated preferentially with the unrelated male. The role of environmental colony-derived cues was tested in a second experiment where females were presented with two unrelated males, one of which had been reared in the same colony until the pupal stage (i.e., as in the previous experiment), while the other had been produced in another colony. In this experiment there was no preferential mating with familiar or unfamiliar males, suggesting that colony-derived cues might not be important in mating preferences. Inbreeding was shown to have no strong effect on the reproductive output of queens as measured by the number of worker and sexual pupae produced. The level of fluctuating asymmetry of workers produced by inbreeding queens was not significantly higher than that of non-inbreeding queens. Finally, colonies headed by inbreeding queens did not produce adult diploid males. Based on the current hypotheses of sex-determination the most plausible explanations for the absence of diploid-male-producing colonies are that (i) workers recognized and eliminated these males early in their development, and/or (ii) there are multiple sex-determining loci in this species. It is suggested that even if inbreeding effects on colony productivity are absent or low, incest avoidance mechanisms may have evolved and been maintained if inbreeding queens produce a higher proportion of unviable offspring. Correspondence to: L. Keller at the present address     

Mating with a kin decreases female remating interval: a possible example of inbreeding avoidance

Inbreeding depression is a relative decline in fitness in offspring of related parents. The magnitude of inbreeding costs varies among taxa and may increase under stressful conditions. Inbreeding tolerance is expected to be low and selection for inbreeding avoidance intense when both sexes invest substantially in shared offspring like in nuptial gift-giving butterflies. This is especially true for increasing mating rate for inbreeding avoidance as nuptial feeding decreases net costs of mating for females. We explored implications of inbreeding in the nuptial gift-giving green-veined white butterfly, Pieris napi. Compared to outbred ones, partially inbred (F = 0.25) eggs and neonate larvae had 25% lower hatching success and 30% lower survival until adult eclosion, respectively. Inbreeding was also associated with small size. Yet, the magnitude of inbreeding depression was independent of larval conditions. A lack of assortative mating and mating durations independent of mating type suggest that neither females nor males discriminate close relatives (r = 0.5) as mates. Indicative of a postcopulatory mechanism to avoid inbreeding, female remating intervals decreased following incestuous matings. Such a plastic response may affect the level of postcopulatory sexual selection as female remating interval (time between successive matings) is necessarily negatively correlated with mating rate (matings per unit time) and mating frequency (lifetime number of matings), and precopulatory mate choice appeared insignificant. Moreover, incest-induced shift in the phenotype towards the adaptive peak may contribute to the evolution of female mating rates, although alternative explanations for polyandry besides material benefits have rarely been invoked when nuptial feeding is involved.     

Mating system of a Neotropical roost-making bat: the white-throated, round-eared bat, Lophostoma silvicolum (Chiroptera: Phyllostomidae)

The vast majority of bats strongly depend on, but do not make, shelters or roosts. We investigated Lophostoma silvicolum, which roosts in active termite nests excavated by the bats themselves, to study the relationship between roost choice and mating systems. Due to the hardness of the termite nests, roost-making is probably costly in terms of time and energy for these bats. Video-observations and capture data showed that single males excavate nests. Only males in good physical condition attracted females to the resulting roosts. Almost all groups captured from excavated nests were single male-multifemale associations, suggesting a harem structure. Paternity assignments based on ten polymorphic microsatellites, revealed a high reproductive success of 46% by nest-holding males. We suggest that the mating system of L. silvicolum is based on a resource-defense polygyny. The temperatures in the excavated nests are warm and stable, and might provide a suitable shelter for reproductive females. Reproductive success achieved by harem males appears to justify the time and effort required to excavate the nests. Reproductive success may thus have selected on an external male phenotype, the excavated nests, and have contributed to the evolution of an otherwise rare behavior in bats.Communicated by G. Wilkinson     

Sex-specific effects of the local social environment on juvenile post-fledging dispersal in great tits

An individual’s decision to disperse from the natal habitat can affect its future fitness prospects. Especially in species with sex-biased dispersal, we expect the cost–benefit balance for dispersal to vary according to the social environment (e.g., local sex ratio and density). However, little is known about the social factors affecting dispersal decisions and about the temporal and spatial patterns of the dispersal process. In our study, we investigated experimentally the effects of the social environment on post-fledging dispersal of juvenile great tits by simultaneously manipulating the density and sex ratio of fledglings within forest plots. We expected young females in the post-fledging period mainly to compete for resources related to food and, as they are subordinate to males, we predicted higher female dispersal from male-biased plots. Juvenile males compete for vacant territories already in late summer and autumn; thus, we predicted increased male dispersal from high density and male-biased plots. We found that juvenile females had a higher probability to leave male-biased plots and had dispersed further from male-biased plots in the later post-fledging phase when juvenile males start to become territorial and more aggressive. Juvenile males were least likely to leave male-biased plots and had smallest dispersal distances from female-biased plots early after fledging. The results suggest that the social environment differentially affected the costs and benefits of philopatry for male and female juveniles. The local sex ratio of individuals is thus an important social trait to be considered for understanding sex-specific dispersal processes.     

Sex-biased dispersal in a long-lived polygynous reptile (Crocodylus johnstoni )

We compared natal dispersals of freshwater crocodiles (Crocodylus johnstoni) against the prediction of male dispersal bias for a polygynous mating system. The crocodiles inhabited a linear series of pools and we calculated the net distances from natal pools to recapture locations some 12–18 years later, at maturity. Philopatry was assessed in terms of adult social distances. A female social distance was 0.46 pools and a male social distance was 1.0 pool. By these criteria, both sexes showed low levels of philopatry (7–12%). However, individuals of both sexes dispersed from the natal site long before they were sexually mature. Divergence in dispersal patterns by sex occurred after the maturity threshold, as males dispersed two to three times farther than females. Intrasexual competition by males is resolved by a size-based hierarchy. The displacement of small males from local mating access is a probable cause of the longer dispersals undertaken by males. Competition, rather than inbreeding avoidance, is driving dispersal in this population of freshwater crocodiles. Received: 8 May 1998 / Accepted after revision: 26 June 1998     

Low roost-site fidelity in pallid bats: associated factors and effect on group stability

Despite potential costs of changing roost or densites, many animals frequently move between roosts or dens. Pallid bats (Antrozous pallidus) change diurnal roost sites frequently and also reportedly have a variety of cooperative social behaviors, many of which are associated with the care of developing offspring. Roost switching is likely to increase the costs of maintaining the group stability expected with social cooperation. Pallid bats roosting in rock crevices in central Oregon were studied with radiotelemetry to (1) examine characteristics of day roosts, (2) determine what ecological factors were correlated with low roost fidelity, and (3) examine the temporal stability of roosting groups of pregnant and lactating bats. Pallid bats changed roosts an average of once every 1.4 days throughout the summer. The bats exhibited seasonal shifts in roost use, occupying roosts behind thin slabs of rock in cool weather and roosts in deep rock crevices in warm weather. Roost switching was not correlated with daily variations in weather conditions or with structural characteristics of the diurnal roosts, although switching may have allowed bats to maintain familiarity with several roosts that vary in microclimate. Roost switching was positively correlated with ectoparasite load. High ectoparasite levels were correlated with lower body weights in lactating females (Fig. 3), suggesting that parasites may be costly to the bats. Roost switching may be a strategy to decrease ectoparasite loads by interrupting the reproductive cycles of those parasites that spend at least part of their life cycle on the walls of the roost. Both pregnant and lactating pallid bats frequently changed their diurnal roost location, but lactating bats tended to travel shorter distances between consecutive roosts. Lactating bats were more likely to continue to associate with particular roostmates despite changes in the location of the diurnal roost (Fig. 4) and were less likely to roost alone. Although the stability of groups of lactating bats was not absolute, evidence supported the prediction that such groups are more cohesive than are those of pregnant bats. Received: 20 June 1995/Accepted after revision: 13 July 1996