ANIMAL BEHAVIOUR, 2003, 65, 363–369 doi:10.1006/anbe.2003.2067, available online at http://www.sciencedirect.com

Secretions of the interaural gland contain information about individuality and colony membership in the Bechstein’s bat KAMRAN SAFI & GERALD KERTH

Zoologisches Institut, Universita¨t Zu ¨ rich (Received 30 April 2002; initial acceptance 9 July 2002; final acceptance 2 August 2002; MS. number: 7313)

Mammals use chemical signals for individual and kin recognition, to establish social hierarchies, mark territories and choose mates. The nocturnal and social lifestyle of bats suggests that, besides acoustic signals, they also use scent to communicate. We investigated in the communally breeding Bechstein’s bat, Myotis bechsteinii, whether secretions of the facial interaural gland contain information that can be used for individual and colony recognition. Since female Bechstein’s bats live in closed societies and show cooperative behaviour, we predicted they would recognize colony members. We analysed interaural gland secretions, which we repeatedly sampled from 85 females belonging to four free-ranging colonies. Gas chromatography/mass spectrometry profiles were individually specific and differed between colonies. Comparing odour profiles between colonies we found a relation between chemical similarity and the mitochondrial haplotype of colony members. Within colonies there was no correlation between mass spectrometer profile similarity and genetic relatedness. Our results suggest that female Bechstein’s bats may use interaural gland secretions for individual and colony recognition but not to infer kinship directly. 

2003 The Association for the Study of Animal Behaviour. Published by Elsevier Science Ltd. All rights reserved.

Mammals possess a variety of scent glands to produce substances that are used as chemical signals influencing the behaviour of conspecifics (Macdonald & Brown 1985). Olfactory communication is important for individual and kin recognition, to establish social hierarchies, to mark and defend territories and to choose mates (Gosling 1990; Johnston & Robinson 1993; Wedekind et al. 1995; Porter 1998; Sun & Mu ¨ ller-Schwarze 1998; Clarke & Faulkes 1999; Gosling & Roberts 2001). There may be particular advantages for nocturnal mammals to communicate with scent because odour signals are independent of daylight and, unlike acoustic signals, they often persist even if the signaller is temporarily absent (Macdonald & Brown 1985). Because bats are nocturnal and often colonial, olfactory communication could be important, especially during social interactions. However, although their striking acoustic abilities have received much attention (Neuweiler 1993), few studies have investigated their olfaction (reviewed in Bloss 1999). Histological examination of the cranial integument of European vespertilionid bats has revealed the existence of large glandular complexes with storage chambers as well as structures for storage and application of sebum. The size and position of these glands, which are in a highly Correspondence: K. Safi, Zoologisches Institut, Verhaltensbiologie, Universita¨t Zu¨rich, Winterthurerstr. 190, CH-8057 Zu¨rich, Switzerland (email: [email protected]). 0003–3472/02/$30.00/0



tactile zone between the eye and the nose, suggest that they may be used for communication (Haffner 1998, 2000). The few behavioural field studies dealing with olfactory communication in bats have shown they use odour in social interactions (De Fanis & Jones 1995b; Brooke & Decker 1996; Voigt & von Helversen 1999; Bouchard 2001). Olfaction, in concert with acoustic signals, is important for mother–offspring recognition (Gustin & McCracken 1987; McCracken & Gustin 1991; De Fanis & Jones 1995a). Furthermore, choice experiments in the laboratory have shown that adult females can distinguish roostmates by scent (Vespertilionidae, Pipistrellus pipistrellus: De Fanis & Jones 1995b; Molossidae, Mops condylurus: Bouchard 2001; Vespertilionidae, Eptesicus fuscus: Bloss et al. 2002). In our study of the European Bechstein’s bat, Myotis bechsteinii, we investigated whether secretions of the interaural gland contain information for individual and colony recognition. Bechstein’s bat is a long-lived and medium-sized vespertilionid bat that lives in deciduous forests. Although males are solitary, females, as in all other European bats, form maternity colonies during summer to rear their young communally (Kerth 1998, in press). Females do not switch colonies, which are socially closed. Maternity colonies consist both of closely related and genetically unrelated females, and individual colony composition is stable over years (Kerth et al. 2000, 2002).

363 2003 The Association for the Study of Animal Behaviour. Published by Elsevier Science Ltd. All rights reserved.

364

ANIMAL BEHAVIOUR, 65, 2

Within colonies, females display fission–fusion behaviour. Colonies often split up into subgroups that are flexible in size and composition and that roost in different tree holes or bat boxes. Day roost associations among colony members are influenced by reproductive status, but not by relatedness (Kerth & Ko ¨ nig 1999). Long-term stability and seclusion of colonies suggest that cooperation among colony members is important in the social life of Bechstein’s bats (Kerth et al. 2000, 2002; Kerth, in press). To stabilize cooperation, interacting animals must detect and avoid possible cheaters, which could take advantage of the benefits of cooperation without sharing the costs. Thus, individual recognition that allows the detection of cheaters is assumed to facilitate the evolution and maintenance of cooperative behaviour (Axelrod & Dion 1988; Crowley 1996; Crowley et al. 1996). In one experiment, Bechstein’s bats belonging to one colony attacked foreign females that entered their roost (Kerth et al. 2002). Thus, females appear to discriminate between conspecifics based on their colony membership. Furthermore, colony members should recognize each other individually to select appropriate cooperation partners. Infrared video monitoring in night roosts of a wild maternity colony revealed that females regularly rub their faces with other colony members shortly after they enter a roost (Kerth 1998). The context of this behaviour suggests that it might be used to identify the presence of colony members in a night roost. Because the part of the face that is rubbed includes the area of the interaural gland, its secretions may be used for individual and/or colony recognition. We used gas chromatography and mass spectrometry to investigate whether interaural gland secretions, sampled from wild female Bechstein’s bats belonging to four colonies, contain information about the carrier. Based on our knowledge of the social behaviour and colony structure, we predicted that secretions would code for individuality, kinship, reproductive status and colony membership. METHODS

Study Animals and Sample Selection In August and September 2000 we sampled interaural gland secretions of females belonging to four maternity colonies (BS, GB2, HB and UH) in deciduous forests near Wu ¨ rzburg, Bavaria, Germany. All adult colony members (individuals born before 2000) had previously been individually marked with a subcutaneously implanted PIT tag (EuroID, Weilerswist, Germany; Kerth & Ko ¨ nig 1996). The colonies roosted mainly in bat boxes, which could be opened. Bats were caught during the day and processed singly. After handling, bats were returned to their day roost. Because we did not always find a colony, recapture success of individuals differed. Members of colony UH were sampled twice (17 August and 1 September), those of GB2 and HB up to three times (GB2: 6 and 26 August and 11 September; HB: 8 and 26 August and 11 September) and colony members of BS up to four times (4 and 18 August, 1 and 12 September). Altogether, we obtained

204 samples from 85 out of 88 females present in the four colonies in 2000. For further analyses, we selected all samples (N=128) from the 41 members of colonies BS, HB and GB2 that had been sampled at least three times, and 24 samples of 12 UH colony members sampled twice. We also used 10 samples of solitary males living close to the maternity colonies.

Sampling Procedure We used commercially available medicinal cotton to sample the secretions of the interaural gland. Before its use in the field, the cotton was washed in the laboratory, 6 h with methanol and then 6 h with pentane, to prevent possible contamination. The dried cotton was stored for the whole sampling period in a glass flask. In the field we used two forceps to take a small piece of cotton out of the flask. Afterwards we wrapped the cotton around one of the forceps to make a ball of about 3 mm in diameter. By gently rubbing the skin on both sides of the face with the cotton ball, we pressed as much secretion as possible out of the interaural glands. During sampling, latex gloves prevented possible contamination of the cotton through contact with the skin of the (same) experimenter. We also avoided touching the skin of the bat with the forceps. If this nevertheless happened, the forceps were cleaned immediately with paper tissues. Each sample, containing secretions of one bat, was put into an individually labelled glass vial (neoLab 2-7201, neoLab Migge, Heidelberg, Germany) that was tightly closed with a butyl-teflon septum (neoLab 2-7203). We documented the date as well as the identity, reproductive status (lactating or nonlactating) and colony membership of each individual sampled. In the field and during transport to the laboratory, samples were kept on ice. In the laboratory, we stored the samples in a freezer at
25C until they were analysed in November 2000. To test for contamination, each time a colony was sampled, we prepared four pieces of cotton according to our standard procedure, two at the beginning and two at the end of each sampling event. As a control we put them into an empty vial without taking a sample from a bat. Handling, marking and sampling of Bechstein’s bats were carried out under licence from the nature conservancy department of the government of lower Frankonia (obere Naturschutzbeho ¨ rde der Regierung Unterfranken).

Gas Chromatography and Mass Spectrometry Samples were prepared immediately before injection. We put the cotton ball in a fresh test-tube (Fiolax) and added 0.5 ml of methyl t-butyl ether (99.8% GC-Quality, Fluka Chemie, Buchs, Switzerland) and 0.5 ml of water (H2O G-Chromasolv, Fluka). We mixed the contents for about 45 s, with a Vortex mixer, and centrifuged the test-tube for 5 min at 3000 rpm. We removed the emergent nonpolar phase (methyl t-butyl ether) and stored it on ice in a new test-tube. We repeated the extraction of the water–cotton mixture twice with 0.25 ml of methyl t-butyl ether. After uniting the three emergent phases we

SAFI & KERTH: GLAND SECRETIONS OF BECHSTEIN’S BATS

evaporated the methyl t-butyl ether under N2-flow on ice then rediluted with 20
l of methyl t-butyl ether. We manually injected 1
l of the solution into a HewlettPackard gas chromatograph (HP-5890) to which an HP-5971 mass spectrometer was attached. For the analyses we used an HP-5 column (25 m0.2 mm0.33
m). The starting temperature was 80C and after a solvent delay of 3 min the temperature was raised by 20C/min to 280C. After 5 min at 280C the temperature was further raised in 5C/min steps to a final temperature of 300C. The whole run took 32 min. Previous tests have shown that after this time no more secretion components eluted from the column. We avoided analysis of samples from the same individual in successive runs to avoid artificial resemblance of individual GC profiles in the unlikely event of an accumulation of secretion components in the column.

Data Analysis Because we were primarily interested in the individual profiles of the secretion components, we analysed individual mass spectrometry (MS) profiles over the whole run (Software HP G1030 MS ChemStation V.B.00.01.). We could not use the absolute abundance of a peak since the total amount of eluted components may have differed between samples based on our method of sampling and analysis. Therefore, all analyses were carried out on log-transformed data (Aitchinson 1982) with the abundance of each peak expressed as the proportion of the total abundance of a sample. To reduce the dimensionality of our data set we performed a principal component analysis. The principal components were then analysed with a multivariate analysis of variance (MANOVA) to test for individuality, colony membership, sex, reproductive status and relatedness. Relatedness among individuals was determined in two ways. First, individuals were assigned to different matrilines based on their mitochondrial DNA haplotype, which was characterized by the repeat number of a mitochondrial dinucleotide microsatellite (Kerth et al. 2000). Second, degrees of pairwise genetic relatedness among individuals were calculated using the software Relatedness 5.0 (Queller & Goodnight 1989). The genetic data were taken from a previous study by Kerth et al. (2002) which analysed 11 nuclear microsatellite loci to determine pedigrees of our four study colonies. Based on the individual degrees of pairwise relatedness we created a genetic distance matrix among bats within colonies. We used a discriminant function analysis to produce a generalized squared distance matrix of individual MS-profiles. This matrix quantifies difference between individual MS-profiles, with smaller distance numbers indicating more similar profiles. Using a Mantel test we tested whether the genetic distance matrix and the generalized squared distance matrix of individual MS-profiles were significantly correlated. Finally, we used a discriminant function analysis for reassignment tests to investigate whether several samples of the same individual were more similar to each other than to samples of other individuals. All statistical tests were programmed on the

Table 1. Reassignment of individual interaural gland secretion samples to individual female Bechstein’s bats within four maternity colonies using a discriminant function analysis Colony

BS GB2 HB UH

No. of samples

% Correctly assigned

No. of individuals

56 40 32 24

91 100 100 100

14 13 14 12

SAS software package version 6.12. (SAS Institute Inc. 1993). RESULTS Using the mass spectrometer we detected 414 masses with the same integral numbers in 162 samples. The range of masses’ abundance was from near zero to 61 824. Analysis of pure solvent/cotton ball samples (controls) showed that 169 of the 414 masses originated totally or in part from components other than interaural gland secretions and we excluded these from further analysis. We excluded 83 other masses from the statistical analysis because they had an abundance of less than 10 over all samples. The principal component analysis grouped the remaining 162 masses into 25 components, containing 74% of the basic variation, with the first component containing 34% (maximum ‘eigenvalue=1’). First we tested whether several samples taken from the same individual were more similar to each other than comparable samples from different individuals. We included only individuals that had been sampled at least three times. Based on the 25 principal components we tested in a MANOVA model for the influence of individuals nested in colonies. MS-profiles of the interaural gland secretions were individual specific (Nsamples =128, Nindividuals =41, F600,950.15 =1.15, P=0.023). Based on the 25 principal components we could reassign MS-profiles to individual females with high accuracy, using a discriminant function analysis. Within a colony, a meanSD of 97.84.5% of samples were reassigned to the correct individuals (Table 1). We used a second MANOVA model to test for the influence of colony, sex, reproductive status and mitochondrial haplotype. When the influence of individuals was used as the error term, only colony membership had a significant effect on the MS-profiles (Table 2). Despite the overall significant influence of colony membership, colonies differed in their MS-profile similarity from each other (Table 3: generalized square distance matrix). Thus, not all pairwise comparisons between the four colonies resulted in significant differences. Colony HB was statistically not distinguishable from colonies BS and GB2 (Fig. 1). Comparing the generalized square distance matrix of individual MS-profiles with the pairwise relatedness matrix of the analysed bats resulted in a nonsignificant correlation (Mantel test: r1 =0.01, P=0.97). Even

365

ANIMAL BEHAVIOUR, 65, 2

Table 2. Results of a MANOVA model testing for different influences on the MS-profiles of interaural gland secretions of adult Bechstein’s bats (N=53 females and 10 males)

Colony Sex Reproductive status mtDNA haplotype

No. of samples

Wilk’s λ

F

df

P

101 162 101 138

0.039 0.54 0.01 0.0006

2.74 1.20 7.81 1.23

75,105.5 25,35 25,2 150,72.4

<0.0001 0.31 0.12 0.17

Colony and reproductive status were tested using only samples of female bats; mtDNA haplotype includes samples of male bats whose haplotype was known. Table 3. Generalized square distances of chemical similarity between four Bechstein’s bat maternity colonies and overlap in repeat numbers of a mitochondrial dinucleotide microsatellite

BS (10,11,12) GB2 (7,9) HB (7,11,12,13) UH (5)

0.6

BS (10,11,12)

GB2 (7,9)

HB (7,11,12,13)

UH (5)

0.4

— 2.2 1.6 12.5

No — 2.2 8.7

Yes Yes — 10.3

No No No —

0.0

The number of AT-repeats found among colony members is given in parentheses. Overlap of mitochondrial haplotypes is indicated by ‘yes’ or ‘no’.

MS-profiles of mother–daughter pairs on average were not more similar to each other than to the mean distance within a colony (Table 4). DISCUSSION

Individual Recognition Interaural gland secretion samples taken from the same individual were significantly more similar to each other than to samples taken from other females. Furthermore, samples could be reassigned with high accuracy to the individual from which they were taken. Our results therefore suggest that Bechstein’s bats produce individualspecific secretions, whose MS-profile remained relatively constant over several weeks. If female Bechstein’s bats can detect individual differences in the chemical composition of the interaural gland secretions, they could use them to recognize individual conspecifics. The MS-profiles contained no detectable information about kinship. On average, the profiles of closely related individuals were as distinct from each other as they were from that of unrelated colony members. Furthermore, we found no significant correlation between relatedness and chemical similarity. Bechstein’s bat maternity colonies consist of both closely related and genetically unrelated females (Kerth et al. 2002). The genetic structuring between colonies, the high degree of female colony fidelity, and cooperative behaviours such as allogrooming and information transfer about roosts suggest that sociality in female Bechstein’s bats is stabilized by cooperation (Kerth 1998, in press; Kerth & Ko ¨ nig 1999; Kerth et al. 2000). Owing to low average colony relatedness (r=0.02;

BS

UH

0.2 Can 2

366

HB

–0.2 –0.4 –0.6 –0.8 –1.0 –2

–1

GB2 0 Can 1

1

2

3

Figure 1. Two-dimensional (can1, can2) presentation of chemical distances between the MS-profiles of members of four Bechstein’s bat maternity colonies using canonical discriminant function analysis. Arrows indicate significant generalized square distances (P<0.05, with Bonferroni correction) between colonies (see text for further explanation). Table 4. The generalized square distances (X±SD) of chemical similarity between adult female Bechstein’s bats belonging to four maternity colonies compared with the average pairwaise relatedness (r±SD)

Sampled individuals Mother–daughter pairs Z* P

N

r

Generalized square distance

53 14

0.03±0.17 0.44±0.10 6.12 <0.001

35.36±25.56 30.64±20.22 0.45 0.65

*Mann–Whitney U test.

Kerth et al. 2002), indiscriminate cooperative behaviour would be directed mostly towards nonrelatives. Thus, kin discrimination would be necessary, if cooperative behaviour in Bechstein’s bats were stabilized via kin selection (Kerth et al. 2002). Therefore, the apparent lack of information about kinship in the secretions may seem surprising. However, because the individual composition of colonies is stable over years and most colony members are not closely related to each other, individual recognition

SAFI & KERTH: GLAND SECRETIONS OF BECHSTEIN’S BATS

may be more important than kin recognition, which is in agreement with our results. Nevertheless the fact that we did not find any information about kinship in the chemical signature of the interaural gland secretions does not necessarily mean that female Bechstein’s bats are unable or do not need to distinguish between colony members on the basis of their relatedness. Even in a society based on reciprocity among long-term associated individuals, there might be situations where it is beneficial to direct cooperative behaviour to related individuals. Theoretically, there are two ways in which kin recognition can work. Individuals may recognize kin directly through phenotype matching or they may learn to infer kinship from the social context (Beecher 1982; Gosling & Roberts 2001). Both mechanisms occur in mammals (Eggert et al. 1998; Hepper & Cleland 1998; Todrank et al. 1998, 1999). Our results seem to indicate that in Bechstein’s bats phenotype matching via odour in the context of kin recognition is unlikely. However, direct kin recognition could be based on chemical components, which we did not detect with our method, for example MHC-derived components (Eggert et al. 1998). Furthermore, we cannot exclude the possibility that the reduction in information of the MS-profiles that resulted from the principal component analysis may have blurred a signature of kinship in the individual scent of bats. Alternatively, other types of signals, such as acoustic ones, could be used to recognize kin. However, even if scent in Bechstein’s bats contains no information about kinship, females might learn the individual scent profile of their offspring and remember it throughout their life (intrinsic meaning versus learned association; Gosling & Roberts 2001). Then, olfactory kin discrimination would work without a correlation between scent similarity and relatedness. It is not clear whether female Bechstein’s bats show kin recognition beyond that used to nurse their own pup (Kerth 1998, in press).

Colony Recognition When we controlled for the influence of the individual, MS-profiles were influenced by colony membership, whereas sex and female reproductive status had no significant effect. Three of the four colonies differed significantly in the MS-profiles of their members. Because all the colonies lived in forests within 15 km of each other, diet differences that could theoretically affect odour (Liang & Silverman 2000; Silverman & Liang 2001) are unlikely to be the main source of MS-profile differences. Differences in MS-profiles between colonies are reflected by the behaviour of females that encounter a foreign colony member in an experimental set-up. Confrontation tests between members of the colonies BS and UH, whose MS-profiles were very distinct, resulted in agonistic behaviour, whereas members of HB and GB2, whose MS-profiles were more similar to each other, treated each other like colony members (Kerth et al. 2002). These findings suggest that scent matching could be a possible mechanism of colony recognition (Gosling & Roberts 2001). With distinct colony profiles animals could

compare other individuals and match their scent profile with the colony scent to accept or reject intruders. The behaviour of females under natural conditions may explain why MS-profiles that are individual specific also differed between colonies. Female Bechstein’s bats regularly engage in face rubbing when entering a night roost occupied by colony members (Kerth 1998). This behaviour could lead to the homogenization of individual interaural gland secretions within a colony, thereby creating an overall colony smell (cf. Boulay et al. 2000). Alternatively, bats within a colony could exchange bacterial cultures when face rubbing and the volatile breakdown products of these bacterial cultures could lead to a similarity of colony odour (Albone & Shirley 1984). Members of the colony HB had MS-profiles that fell between those of BS and GB2 colony members. They also carried repeat numbers of a mitochondrial microsatellite that were present in both other colonies. In contrast, UH colony members, whose MS-profiles were very different to those found in the other colonies, all had a single repeat number that did not occur in either BS, GB2, or HB. It is unclear how this relation between mtDNA and MS-profile similarity arises. We found no influence of relatedness on MS-profiles within colonies. The lack of a correlation between genetic relatedness and MS-profile, although there seems to be a relation between mitochondrial haplotype and colony MS-profile, indicates that there is no simple genetic mechanism responsible for the individual scent profile. Further studies are necessary to find which factors control the chemical profile of interaural gland secretions in Bechstein’s bats.

Conclusions Mammals living in family groups, such as beavers, Castor canadensis (Sun & Mu ¨ ller-Schwarze 1997, 1998), often show clan-specific odour profiles that help them to maintain the cohesion of their social groups. Our results suggest that female Bechstein’s bats could use secretions of the interaural glands to recognize colony members. However, this cannot be the only proximate mechanism used for maintaining the high segregation of colonies. Members of colony HB had secretions that were chemically not very distinct from that found in BS and GB2. Nevertheless, none of the colonies exchanged members over a period of 6 years (Kerth et al. 2002). Therefore, females probably also use signals other than interaural gland secretions to recognize colony members. Nevertheless, interaural gland secretions contain information about individual identity, and thus potentially allow colony members to recognize each other individually. In conclusion, our study suggests that scent is likely to be an important proximate mechanism for individual and colony recognition in female Bechstein’s bats. Acknowledgments We thank L. M. Gosling, B. Ko ¨ nig, A. McElligott and two anonymous referees for helpful comments on the manuscript. We are very grateful to P. Ru ¨ edi for allowing us to

367

368

ANIMAL BEHAVIOUR, 65, 2

use his GC/MS-system and for his advice and support during our chemical analyses. We thank S. Krackow for helping us with the statistical analyses. We gratefully acknowledge the support of the local nature conservation and forestry departments. K. Widmer assisted us in the field and S. Hein helped to establish the chemical methods. This work was supported by the Swiss National Science Foundation (SNF, 31-59556.99). References Aitchinson, J. 1982. Statistical analysis of compositional data. Journal of the Royal Statistical Society, Series B, 44, 139–177. Albone, E. S. & Shirley, S. G. 1984. Mammalian Semiochemistry: the Investigation of Chemical Signals Between Mammals. Chichester: J. Wiley. Axelrod, R. & Dion, D. 1988. The further evolution of cooperation. Science, 242, 1385–1390. Beecher, M. D. 1982. Signature systems and kin recognition. American Zoologist, 22, 477–490. Bloss, J. 1999. Olfaction and the use of chemical signals in bats. Acta Chiropterologica, 1, 31–45. Bloss, J., Acree, T. E., Bloss, J. M., Hood, W. R. & Kunz, T. H. 2002. Potential use of chemical cues for colony-mate recognition in the big brown bat, Eptesicus fuscus. Journal of Chemical Ecology, 28, 819–834. Bouchard, S. 2001. Sex discrimination and roostmate recognition by olfactory cues in the African bats, Mops condylurus and Chaerephon pumilus (Chiroptera: Molossidae). Journal of Zoology, 254, 109–117. Boulay, R., Hefetz, A., Soroker, V. & Lenoir, A. 2000. Camponotus fellah colony integration: worker individuality necessitates frequent hydrocarbon exchanges. Animal Behaviour, 59, 1127– 1133. Brooke, A. P. & Decker, D. M. 1996. Lipid compounds in secretions of fishing bat, Noctilio leporinus (Chiroptera: Noctilionidae). Journal of Chemical Ecology, 22, 1411–1428. Clarke, F. M. & Faulkes, C. G. 1999. Kin discrimination and female mate choice in the naked mole-rat Heterocephalus glaber. Proceedings of the Royal Society of London, Series B, 266, 1995– 2002. Crowley, P. H. 1996. Evolving cooperation: strategies as hierarchies of rules. Biosystems, 37, 67–80. Crowley, P. H., Provencher, L., Sloane, S., Dugatkin, L. A., Sophn, B., Rogers, L. & Alfieri, M. 1996. Evolving cooperation: the role of individual recognition. Biosystems, 37, 49–66. De Fanis, E. & Jones, G. 1995a. Post-natal growth, mother–infant interactions and development of vocalizations in the vespertilionid bat Plecotus auritus. Journal of Zoology, 235, 85–97. De Fanis, E. & Jones, G. 1995b. The role of odour in the discrimination of conspecifics by pipistrelle bats. Animal Behaviour, 49, 835–839. Eggert, F., Muller-Ruchholtz, W. & Ferstl, R. 1998. Olfactory cues associated with the major histocompatibility complex. Genetica, 104, 191–197. Gosling, L. M. 1990. Scent marking by resource holders: alternative mechanisms for advertising the costs of competition. In: Chemical Signals in Vertebrates (Ed. by D. W. Macdonald, D. Mu¨ller-Schwarze & S. Natynczuk), pp. 315–328. Oxford: Oxford University Press. Gosling, L. M. & Roberts, S. C. 2001. Scent-marking by male mammals: cheat-proof signals to competitors and mates. Advances in the Study of Behavior, 30, 169–217.

Gustin, M. K. & McCracken, G. F. 1987. Scent recognition between females and pups in the bat Tadarida brasiliensis mexicana. Animal Behaviour, 35, 13–19. Haffner, M. 1998. The size of sebaceous glands in relation to the size of hair follicles on the heads of some small mammals (Insectivora, Chiroptera, Rodentia). Annals of Anatomy, 180, 165– 171. Haffner, M. 2000. Structure and function of pilosebaceous units on the heads of small mammals (Insectivora, Chiroptera, Rodentia). Acta Zoologica, 81, 195–203. Hepper, P. G. & Cleland, J. 1998. Developmental aspects of kin recognition. Genetica, 104, 199–205. Johnston, R. E. & Robinson, T. A. 1993. Cross-species discrimination of individual odours by Hamsters (Muridae: Mesocricetus auratus; Phodopus campbelli). Ethology, 94, 317–325. Kerth, G. 1998. Sozialverhalten und genetische Populationsstruktur bei der Bechsteinfledermaus Myotis bechsteinii. Berlin: Wissenschaft und Technik Verlag. Kerth, G. In press. Relatedness, life history and social behaviour in the long-lived Bechstein’s bat (Myotis bechsteinii). In: Functional and Evolutionary Ecology of Bats (Ed. by A. Zubaid, G. F. McCracken & T. H. Kunz). Oxford: Oxford University Press. Kerth, G. & Ko ¨ nig, B. 1996. Transponder and an infraredvideocamera as methods in a field study on the social behaviour of Bechstein’s bats (Myotis bechsteini). Myotis, 34, 27–34. Kerth, G. & Ko ¨ nig, B. 1999. Fission, fusion and nonrandom associations in female Bechstein’s bats (Myotis bechsteinii). Behaviour, 136, 1187–1202. Kerth, G., Mayer, F. & Ko ¨ nig, B. 2000. Mitochondrial DNA (mtDNA) reveals that female Bechstein’s bats live in closed societies. Molecular Ecology, 9, 793–800. Kerth, G., Safi, K. & Ko ¨ nig, B. 2002. Mean colony relatedness is a poor predictor of colony structure and female philopatry in the communally breeding Bechstein’s bat (Myotis bechsteinii). Behavioral Ecology and Sociobiology, 52, 203–210, doi:10.1007/ s00265-002-0499-6. Liang, D. & Silverman, J. 2000. You are what you eat: diet modifies cuticular hydrocarbons and nestmate recognition in the Argentine ant, Linepithema humile. Naturwissenschaften, 87, 412–416. McCracken, G. F. & Gustin, M. K. 1991. Nursing behavior in Mexican free-tailed bat maternity colonies. Ethology, 89, 305–321. Macdonald, D. & Brown, R. E. 1985. Introduction: the pheromone concept in mammalian chemical communication. In: Social Odours in Mammals (Ed. by R. E. Brown & D. W. MacDonald), pp. 1–18. Oxford: Clarendon Press. Neuweiler, G. 1993. Biologie der Flederma¨use. Stuttgart: Thieme Verlag. Porter, R. H. 1998. Olfaction and human kin recognition. Genetica, 104, 259–263. Queller, D. C. & Goodnight, K. F. 1989. Estimating relatedness using genetic-markers. Evolution, 43, 258–275. SAS Institute Inc. 1993. SAS Technical Report P-243. Cary, North Carolina: SAS Institute Inc. Silverman, J. & Liang, D. 2001. Colony disassociation following diet partitioning in a unicolonial ant. Naturwissenschaften, 88, 73–77. Sun, L. X. & Mu ¨ ller-Schwarze, D. 1997. Sibling recognition in the beaver: a field test for phenotype matching. Animal Behaviour, 54, 493–502. Sun, L. X. & Mu ¨ ller-Schwarze, D. 1998. Anal gland secretion codes for relatedness in the beaver, Castor canadensis. Ethology, 104, 917–927. Todrank, J., Heth, G. & Johnston, R. E. 1998. Kin recognition in golden hamsters: evidence for kinship odours. Animal Behaviour, 55, 377–386. Todrank, J., Heth, G. & Johnston, R. E. 1999. Social interaction is necessary for discrimination between and memory for odours of close relatives in golden hamsters. Ethology, 105, 771–782.

SAFI & KERTH: GLAND SECRETIONS OF BECHSTEIN’S BATS

Voigt, C. C. & von Helversen, O. 1999. Storage and display of odour by male Saccopteryx bilineata (Chiroptera, Emballonuridae). Behavioral Ecology and Sociobiology, 47, 29–40.

Wedekind, C., Seebeck, T., Bettens, F. & Paepke, A. J. 1995. Mhc-dependent mate preferences in humans. Proceedings of the Royal Society of London, Series B, 260, 245–249.

369