Colony structure and parentage in wild colonies of co-operatively breeding Damaraland mole-rats suggest incest avoidance alone may not maintain reproductive skew

Please download to get full document.

View again

of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report



Views: 63 | Pages: 10

Extension: PDF | Download: 0

Colonies of co-operatively breeding African mole-rats have traditionally been thought to be composed of a single breeding female, one or two breeding males, and their offspring. In the naked mole-rat (Heterocephalus glaber), the occurrence of
   Molecular Ecology (2004) 13  , 2371–2379doi: 10.1111/j.1365-294X.2004.02233.x© 2004 Blackwell Publishing Ltd  BlackwellPublishing,Ltd.  Colony structure and parentage in wild colonies of co-operatively breeding Damaraland mole-rats suggest incest avoidance alone may not maintain reproductive skew  TAMSIN M. BURLAND,  *  NIGEL C. BENNETT,  †   JENNIFER U. M. JARVIS  ‡  and CHRISTOPHER G. FAULKES  *  *  School of Biological Sciences, Queen Mary, University of London, London, E1 4NS, UK, †   Mammal Research Institute, Department of Zoology & Entomology, University of Pretoria, Pretoria 0002, South Africa, ‡  Zoology Department, University of Cape Town, Rondebosch, 7701, South Africa Abstract Colonies of co-operatively breeding African mole-rats have traditionally been thought tobe composed of a single breeding female, one or two breeding males, and their offspring.In the naked mole-rat (    Heterocephalus glaber    ), the occurrence of facultative inbreedingmeans incest avoidance cannot prevent reproduction in subordinate group members, andphysiological suppression of reproductive function by the breeding female occurs inboth sexes. In contrast, previous studies of captive colonies of the Damaraland mole-rat(   Cryptomys damarensis   ) suggest that breeding within a colony is restricted to a singlebreeding pair, simply because all other colony members are highly related (first- or second-order relatives) and this species is an obligate outbreeder. Using microsatellite markers, weinvestigated parentage and colony composition in 18 wild Damaraland mole-rat colonies todetermine whether inbreeding avoidance alone can explain the high levels of reproductiveskew in this species. Multiple and unidentified paternity was widespread within coloniesand immigrants of both sexes were regularly identified. Unrelated, opposite-sex nonbreederswere found coexisting in two colonies. These results suggest that, in the wild, conditionsexist where nonreproductive females can come into contact with unrelated males, evenwhen they do not disperse from their natal colony. Inbreeding avoidance alone is thereforeinsufficient to maintain the high levels of reproductive skew identified in this speciessuggesting that the breeding female somehow suppresses the reproductive function innonbreeding females.  Keywords  :African mole-rat, Bathyergidae, colony structure, parentage, reproductive skew, repro-ductive suppression Received 7 February 2004; revision received 7 April 2004; accepted 7 April 2004  Introduction  Co-operative breeding strategies have evolved in a varietyof animal taxa, and are characterized by social groupswhere nonbreeding individuals assist dominant breedinganimals to successfully rear young (either directly orindirectly). The mechanisms underlying the evolution of co-operative breeding and its maintenance at a proximatelevel may be varied, but are dependent on the patterns of genetic relatedness within groups. The role of kin selectionand other factors such as mutualism have received muchdiscussion (for review see Clutton-Brock 2002; Griffin &West 2003). By definition, co-operative breeding requiresthat some members of a social group refrain from breeding,resulting in a clear-cut reproductive division of labour.How this reproductive skew is achieved may be highlyvariable among species, or even populations (for reviewsee Faulkes & Bennett 2001). In the cases of delayeddispersal from monogamous family groups (one breederof each sex plus their offspring), reproduction may belimited simply as a result of incest avoidance. In other cases,  Correspondence: Dr Christopher G. Faulkes. Fax: + 44 20 89830973;E-mail:   2372  T. M. BURLAND ET AL.  © 2004 Blackwell Publishing Ltd,  Molecular Ecology  , 13, 2371–2379  physiological or behavioural suppression of nonbreedingcolony members by breeders, may occur (Keller & Reeve1994; Clutton-Brock 1998; Reeve et al  . 1998).Species of the African mole-rat family (Bathyergidae)display a wide range of co-operative breeding strategies(reviewed in Bennett & Faulkes 2000). Estimates of skew inindividual lifetime reproductive success reach the highestlevels observed in vertebrates, both in the Damaralandmole-rat (  Cryptomys damarensis  , < 8% individuals achiev-ing direct reproductive success) and the naked mole-rat(   Heterocephalus glaber  , < 1% individuals). Wild colonies of  both these species are traditionally thought to be com-posed of a single breeding female (the queen), togetherwith one or two breeding males, and their offspring. In thenaked mole-rat the occurrence of facultative inbreedingmeans that inbreeding avoidance is not sufficient to pre-vent reproduction among subordinate group members,and consequently the queen suppresses reproductivefunction in both males and females. It is thought that thissuppression is mediated through behavioural interactions between the queen and the nonbreeders (Faulkes & Abbott1997). In contrast to the naked mole-rat, field, laboratoryand genetic studies suggest that the Damaraland mole-rat displays inbreeding avoidance (Jarvis & Bennett 1993;Burda 1995; Bennett et al  . 1996; Cooney & Bennett 2000;Burland et al  . 2002). In this instance, ‘inbreeding’ refers to breeding with related colony members, which, in the caseof colonies comprising a single breeding female with breeding male(s) and offspring, would mean first- andsecond-order relatives.Mark–recapture studies of the Damaraland mole-ratshow extremely low rates of natal dispersal and movementamong colonies (Jarvis & Bennett 1993), strongly suggest-ing that inbreeding avoidance within an extended familysituation is sufficient to explain the high levels of reproduc-tive skew associated within the Damaraland mole-rat. Thereare no recorded cases of plural breeding among females of this species, either in captivity or in the wild, where colonysizes may exceed 40 individuals. Furthermore, physiolog-ical studies have demonstrated diminished reproductivefunction in nonbreeding females (reviewed in Faulkes &Bennett 2001). Such findings suggest either that nonbreed-ing individuals show ‘self-restraint’ from reproductiveactivity (which may result from a lack of available mates,and therefore be consistent with inbreeding avoidance,Clarke et al  . 2001), or that a form of external, sociallyinduced suppression of reproductive function (dominantcontrol) is exerted over nonbreeding females, probably bythe breeding female. Two, nonexclusive, hypotheses there-fore exist to explain the manner by which reproductiveskew is maintained in the Damaraland mole-rat: (i) thatnonbreeding females do not breed, and show physiolo-gically reduced reproductive function because of the costs of  breeding with closely related males; (ii) that the breedingfemale suppresses reproductive function in the nonbreed-ing females within her colony (dominant control).Previous studies on captive colonies of the Damaralandmole-rat support the first hypothesis (Cooney & Bennett2000; Clarke et al  . 2001). However, despite the importanceof these studies, their conclusions stem from the assump-tion that a socially and genetically monogamous familyunit is the norm for wild colonies. A genetic investigationinto the mating behaviour and composition of wild colon-ies is therefore essential to resolve this issue fully. Herewe describe the use of microsatellite markers to invest-igate, in detail, parentage and colony structure in Damara-land mole-rat colonies from two geographically separatepopulations. The results of this study provide sufficientresolution to assess, for the first time, the relative import-ance of inbreeding avoidance and dominant control in theevolution and maintenance of reproductive skew in thisspecies in the wild.  Materials and methods  Study sites and sample collection  Damaraland mole-rats were sampled at two sites, Dordabis(Namibia, 22  °  58  ′  S, 17  °  41  ′  E) and Hotazel (Northern CapeProvince, South Africa, 27  °  17  ′  S, 23  °  0  ′  E). Individuals werelive trapped and sampled using tissue biopsy as previouslydescribed in Burland et al  . (2002). The population at Dordabishas been subject to an extensive mark–recapture studysince 1988. For this study, individuals were sampled from11 colonies between 1993 and 2002. The total number of individuals sampled from each colony (over the samplingperiod) ranged from 10 to 49. Extensive ecological dataincluding sex, reproductive status (as detailed in Jarvis &Bennett 1993), mass and colony membership (includingany movement between colonies) were available for eachindividual. Seven colonies were also sampled once at Hotazelin January 1996, with colony numbers ranging from fiveto 19; no mark–recapture data were available for thispopulation. The composition of all colonies sampled from both populations are detailed in Table 1.  Genotyping and data analysis  DNA was extracted and individuals were genotyped at10 autosomal and one X-linked microsatellite loci: DMR1–5, 7, CH1–3, NCAM and LV25 as previously detailedin Burland et al  . (2001, 2002). Allele frequencies werecalculated separately for each population using the program  relatedness  5.08 ( andthese frequencies were used in all the data analyses. Thisprogram allows for the inclusion of haploid data such asthat generated for males at the X-linked locus DMR1. Inaddition, it allows for frequencies to be weighted by colony   COLONY STRUCTURE AND PARENTAGE IN MOLE-RATS  2373  © 2004 Blackwell Publishing Ltd,  Molecular Ecology  , 13, 2371–2379  size to ensure that they more closely resemble thefrequency of alleles of the breeders within the populationand are not distorted by colonies with a large number of nonbreeding individuals.Parentage was investigated using cervus  2.0 (Marshall  et al  . 1998). This program calculates a likelihood score foreach offspring–candidate parent and determines, by simu-lation, the difference in likelihood score required betweenthe most likely and second most likely parent for parentageto be assigned at a given level of confidence (Delta criteria).The program does not allow the inclusion of haploid data.However, because allele frequencies had already been cal-culated for each population including the haploid data (seeabove), males were designated as diploid homozygotes forthe X-linked haploid allele.For each colony, the offspring of the colony breedingfemale (who was identified easily by the presence of prom-inent teats, Jarvis & Bennett 1993) were identified using theDelta criteria for 95% confidence and specifying 99% of locityped and an error rate of 0.01 (Burland et al  . 2002). Calcu-lations of Delta criteria were performed separately for eachpopulation using the respective weighted allele frequen-cies. All colony members, except those in the Dordabispopulation known from mark–recapture data to srcinatefrom a different colony (all male, Table 1), were regardedas putative offspring of the breeding female. Once mother–offspring pairs had been identified, those males fatheringoffspring within each colony were identified. Any colonymale either known to srcinate from a different colony(Dordabis only) or who had previously been excluded asoffspring of the breeding female was regarded as aputative father. Male offspring of the breeding female werenot considered to be putative fathers, as both behavioural(Jarvis & Bennett 1993; Burda 1995; Bennett et al  . 1996;Cooney & Bennett 2000) and genetic evidence (Burland  et al  . 2002) strongly support the hypothesis that this speciesshows strict inbreeding avoidance. For example, mito-chondrial DNA sequencing revealed that in each of sixcolonies that were studied, the breeding pairs were fromdifferent matrilines (Bennett & Faulkes 2000). Furthermore,average relatedness (  R  ) among breeding pairs in 15 coloniesis estimated at 0.02, indicating that the breeding pair arenot related to one another (Burland et al  . 2002). Simulationsto obtain the Delta criteria were performed separately foreach colony. The number of typed candidate fathers wasassumed to be 90% of the total candidate males (Burland   Location colony name*Colony sampling date(s)Total sampledBreeding femalesMarked males from other colonies†Other femalesOther malesDordabisA01/94–10/95130076B07/93–06/962210813C02/93–07/96161186D01/94–07/9950102722E07/99–09/0127101214F07/98–07/99181197G07/98–07/002011513H07/00–04/01121155I02/98–07/99101036 J02/98–02/992011153K02/98–09/0142121524L01/97–09/0145102024M01/92–06/96140068N11/92–06/96130058HotazelA09/96141—67B09/9651—13C09/9671—33D09/9681—52E09/96101—18F09/96181—413G09/96171—610*Colonies Dordabis A–D and Hotazel A–G are the same colonies as identified using this nomenclature in Burland et al . (2002). In the cases of Dordabis B–D further individuals have  been included for this study.†Marked individuals were only caught at Dordabis, where a long-term mark–recapture study has taken place. Table 1 Details of colonies and individualsincluded in the study. For Dordabis colon-ies A, M and N tissue from the breedingfemale was not available. As such, individualsfrom these colonies were included onlyfor the estimation of population allelefrequencies   2374  T. M. BURLAND ET AL.  © 2004 Blackwell Publishing Ltd,  Molecular Ecology  , 13, 2371–2379  et al  . 2002)  .  Where more than one typed male was includedas a putative father, the mean pairwise relatedness valueamong these males (see below) was specified to preventan over-estimation of parentage assignment confidence(Marshall et al  . 1998). The critical likelihood of differencescores obtained for assigning parentage with 95% con-fidence ranged between one and two (except for two colo-nies; see Results). For the paternity analysis, male offspringwere regarded as having a missing genotype for the X-linked locus, as fathers and male offspring are not expectedto share an allele at this locus.Pairwise and mean levels of relatedness were estimatedwith the 10 autosomal loci using relatedness  5.08, whichuses the calculation of Queller & Goodnight (1989). A bias-corrected allele frequency value was incorporated into eachcalculation of relatedness (Queller & Goodnight 1989).Standard errors were calculated by jackknifing over loci.  Colony structure  Using the results of the parentage and relatedness analyses,in combination with mark–recapture data (where available),a summary of colony structure was compiled for eachcolony. The total number of pairwise relationships withineach colony was calculated as (  n *  n   −  1)/2, where n  is thenumber sampled in a colony. Each pairwise relationshipwas assigned to one of the following categories: breedingpair; parent–offspring pair; sibling pair (subsets — fullsiblings, half siblings where both fathers were identified,half siblings where only one father was identified, othersiblings where no father was identified); other pairingswith no identified relationship (subsets — same sex pair,opposite sex pair — either breeding male vs. females excludedas his offspring or other opposite sex pairs).  Results  Genotypes were generated for a total of 401 individualsfrom 21 colonies across the two study sites (Table 1),and included the animals used in the study of Burland  et al  . (2002). Morphological examination identified asingle breeding female in each colony. Although previouslycaptured, tissue samples were missing for breeding femalesfrom three colonies sampled in Dordabis (A, M and N;these individuals had disappeared from colonies beforethe tissue-sampling program began), and so these colonieswere included for the estimation of population allelefrequencies only.  Parentage analyses   Just three of 171 nonbreeding females tested were excludedas offspring of their respective colony breeding female(1.8%). Measures of relatedness between these excludedfemales and their respective breeding female suggest thatthey were not closely related (Table 2). The number of males excluded as offspring of the breeding female washigher; 19 of 205 tested (with a further seven in Dordabisalready known to srcinate from outside the colony, givinga total of 12.7% of males excluded). Mean levels of relatedness between the excluded males and the breedingfemale were generally low, although in a few colonies, it ispossible that they were second- or third-order relatives(Table 2).At least one breeding male (defined as a male for whichoffspring could be assigned with 95% confidence) wasidentified in 16 of the 17 colonies investigated (Table 2).For one colony (Dordabis I), all males were assigned as off-spring of the breeding females, and so this colony was notincluded in the paternity analysis. More than one breedingmale was directly identified in just two colonies (DordabisK and Hotazel A). In both cases, these males had been cap-tured together within the colony and values of relatedness between the pairs of males were consistent with thesemales being first-order relatives (Table 2). A number of off-spring in these two colonies could only be assigned a fatherwith 80% confidence (Table 2). This is considered to be aresult of the high relatedness among the candidate males,which meant a much higher likelihood of difference value(5.6–5.7) was necessary to distinguish between them with95% confidence. Including these offspring, the ratios of offspring fathered for each pair was 1.3 : 1 for DordabisK (  n  = 39) and 7 : 1 for Hotazel A (  n  = 8). In the case of Dordabis K, both males fathered offspring born over aperiod of more than 3 years, and mixed paternity withina single litter was indicated when two juveniles, caughtweighing just 18 g, were assigned different fathers with 95%confidence — this weight suggests that the juveniles wereless than 1 month old (Bennett et al  . 1994), while only twoor three litters are born each year (Bennett & Faulkes 2000).For nine of the 17 colonies investigated, no father could be assigned to at least one offspring (Table 2). The mini-mum number of males achieving paternities with the col-ony is therefore one more than the number of breedingmales identified within these colonies. This occurred morefrequently in Hotazel, where six of the seven colonies con-tained offspring with no father assigned, in contrast to half of the colonies in Dordabis (excluding colony Dordabis I).It was not possible to determine whether offspring whereno father was assigned were full or half siblings: when alikelihood approach (Goodnight & Queller 1999) was taken,a Type I error rate of 5% was associated with a Type II errorrate of 50%.Not all males excluded as offspring of the breeding femalewere breeders (Table 2). The mean level of relatednessamong the breeding female and males who did not fatherher offspring was −  0.01 ±  0.14, −  0.11 ±  0.1 and −  0.2 ±  0.1(colonies Dordabis J, Dordabis L and Hotazel A, respectively).   C O L  O N Y  S  T  R  U C T  U R  E  A N D P  A R  E  N T  A G E  I   N  M O L  E - R  A T  S   2  3  7  5   © 2  0  0  4  B  l   a c  k  w e l   l   P  u b  l   i   s  h  i   n  g L  t   d  ,  M o  l   e  c  u l   a  r  E  c  o  l   o   g  y  , 1  3  , 2  3  7  1 – 2  3  7  9  Table 2 Summary of parentage analysis for each colony   ColonyCandidate offspringMaternity analysis Paternity analysisFemale offspring of  breeding femaleFemales excluded( R   ± SE)Maleoffspring of  breeding femaleMales excluded( R   ± SE)Candidate males( R   ± SE)Father assignedNo father assigned(m,f)Breeding males identified in colony( R   ± SE)Minimum no. of males fathering offspringDordabisB2180121 ( −−−− 0.02 ± 0.11)120011C148060114011D49270211 (0.16 ± 0.16)147(1,0)12E26120131 (0.21 ± 0.18)124(0,1)12F169070116011G1850130118011H10505019(0,1)12I93060————1 J18132 (0.04 ± 0.19)21 (0.03 ± 0.22)2 (0.34 ± 0.10)12(0,3)12K391502402 (0.37 ± 0.18)39*02 (0.37 ± 0.18)2L44191 ( − 0.02 ± 0.2)204 ( − 0.1 ± 0.09)4 (0.02 ± 0.08)36(0,3)11HotazelA136034 ( − 0.05 ± 0.15)4 (0.51 ± 0.16)8*(0,1)2 (0.5 ± 0.18)3B41021 (0.34 ± 0.13)10(2,1)01C63021 ( − 0.24 ± 0.14)15011D75011 ( − 0.06 ± 0.14)14(0,2)12E91071 (0.3 ± 0.14)16(2,0)12F1740121 (0.23 ± 0.14)12(10,4)12G166091 ( − 0.1 ± 0.16)114(1,0)12For the maternity analysis, the number of individuals assigned and excluded as offspring of the breeding female are detailed separately for each sex, while the relatedness ( R ) values are mean values calculated between the breeding female and the excluded individuals. For the paternity analysis, candidate males are those males considered as putative fathers; where more than one candidate male was present within a colony, the mean R -value among these males is given. Breeding males are defined as those males to whom offspring could be assigned; where more than one was identified within a colony, the mean R -value among breeding males is given. The number of colony offspring for which no father could be identified is given separately for each sex, with the values for males given first.*For colonies Dordabis K and Hotazel A, nine and two offspring, respectively, were assigned fathers with only 80% confidence (see text).
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!