CSIRO PUBLISHING Australian Journal of Zoology http://dx.doi.org/10.1071/ZO13009 High incidence of multiple paternity in an Australian snapping turtle (Elseya albagula) Erica V. Todd A,D, David Blair A, Colin J. Limpus B, Duncan J. Limpus B and Dean R. Jerry A,C A School of Marine and Tropical Biology, James Cook University, Townsville, Qld 4810, Australia. Aquatic Threatened Species Unit, Department of Environment and Heritage Protection, Brisbane, Qld 4001, Australia. C Centre for Sustainable Tropical Fisheries and Aquaculture, James Cook University, Townsville, Qld 4810, Australia. D Corresponding author. Email: ericavtodd@gmail.com B Abstract. Genetic parentage studies can provide detailed insights into the mating system dynamics of wild populations, including the prevalence and patterns of multiple paternity. Multiple paternity is assumed to be common among turtles, though its prevalence varies widely between species and populations. Several important groups remain to be investigated, including the family Chelidae, which dominate the freshwater turtle fauna of the Southern Hemisphere. We used seven polymorphic microsatellite markers to investigate the presence of multiple fathers within clutches from the white-throated snapping turtle (Elseya albagula), an Australian species of conservation concern. We uncovered a high incidence of multiple paternity, with 83% of clutches showing evidence of multiple fathers and up to three males contributing to single clutches. We confirm a largely promiscuous mating system for this species in the Burnett River, Queensland, although a lone incidence of single paternity indicates it is not the only strategy employed. These data provide the first example of multiple paternity in the Chelidae and extend our knowledge of the taxonomic breadth of multiple paternity in turtles of the Southern Hemisphere. Additional keywords: Burnett River, freshwater turtle, mating system, paternity assignment, polyandry. Received 16 January 2013, accepted 6 May 2013, published online 23 May 2013 Introduction Animal mating systems reflect a battle of competing interests between male and female optimal mating strategies (Reynolds 1996) and have important implications regarding the intensity of sexual selection (Emlen and Oring 1977), effective population size (Nunney 1993; Sugg and Chesser 1994), and the maintenance of genetic diversity (Baer and Schmid-Hempel 1999). Understanding a species’ mating system is thus not only relevant to behavioural ecology, but is necessary to comprehend how different reproductive strategies affect overall population genetic diversity, which is an important consideration in species conservation and management. Molecular investigations of parentage have been particularly important in elucidating details of the mating system of many species that would otherwise be impossible to obtain through observations of wild populations (Fleischer 1996; Hughes 1998; Jones and Ardren 2003). Multiple paternity describes a scenario where different offspring within a single brood are fathered by different males and molecular parentage studies have revealed it to be an important aspect of the mating system of diverse taxa (Coleman and Jones 2011). The phenomenon is particularly common among reptiles (Uller and Olsson 2008), especially turtles (Pearse and Avise 2001). Journal compilation
CSIRO 2013 Since Harry and Briscoe (1988) used allozymes to provide the first genetic evidence of multiple paternity within clutches of the loggerhead sea turtle (Caretta caretta), studies of chelonian mating systems have contributed important information for conservation programs of many threatened taxa, including terrestrial, freshwater, and marine species (Roques et al. 2006; Zbinden et al. 2007; Davy et al. 2011). Accounts of the prevalence and patterns of multiple paternity in turtles, however, vary widely among species, among populations of the same species, and are sometimes contrary to expectations based on observed promiscuous mating behaviour (FitzSimmons 1998). Published studies are also by no means representative of turtle diversity and several families remain to be investigated. The family Chelidae (Pleurodira) is one such group, significant by virtue of its clear Gondwanan heritage and its dominance of the extant freshwater turtle diversity of Australia, New Guinea, and South America. Limited observations of courtship indicate that elaborate precoital behaviour and female mate choice may be important features of the mating system of Australian chelid turtles (Murphy and Lamoreaux 1978; Hamann et al. 2007). Otherwise, very little is known regarding mating behaviours and reproductive dynamics for this important group. www.publish.csiro.au/journals/ajz B Australian Journal of Zoology Elseya albagula, like other Australian snapping turtles, is a large-bodied riverine specialist with cryptic behaviour and a narrow geographic range. The species is endemic to the Burnett, Mary, and Fitzroy drainages in central and south-east Queensland, recognised as a significant freshwater biodiversity hotspot within eastern Australia and a conservation priority for chelonians worldwide (Buhlmann et al. 2009). However, areas of intense urban development and agricultural land use surround these catchments, which contain some of the most heavily impounded stretches of river in Australia (Stein et al. 2002). Anthropogenic threats common to many other freshwater turtles worldwide, namely, habitat fragmentation resulting from water infrastructure development and intense egg predation by native and introduced predators (Thomson et al. 2006; Hamann et al. 2007; Limpus et al. 2011), have led to conservation concerns for E. albagula across its range. A management plan (Hamann et al. 2007) has been partly implemented for this species in the Burnett River, but little is known regarding reproductive strategies employed by this species and how these may influence population genetic diversity in the context of the above threats. In order to complement conservation efforts for E. albagula, we conducted a preliminary investigation into the genetic mating system of this species using a multilocus microsatellite approach. Specifically, we sought to document the incidence of multiple paternity and to assess the validity of future paternity-assignment studies in this system. Materials and methods Sample collection Six clutches from separate E. albagula females were investigated for the presence of multiple paternity. Wild gravid females were collected during the 2007–08 breeding season from the lower Burnett River in central Queensland, Australia, within an ~50 km stretch of river below the Paradise Dam to the upper impounded waters of the Ned Churchward Weir. Females were induced into oviposition using synthetic oxytocin (illium synotcin) before being released at their original site of capture. Clutches were incubated in artificially constructed nests at the Paradise Dam turtle hatchery, which operated as part of conservation efforts for this species. Barriers were placed around individual nests to contain hatchlings upon emergence and prevent mixing between clutches. Tissue biopsies were collected from adult females (skin) and hatchlings (scute) before release and were preserved immediately in 70% ethanol to provide DNA for genetic analyses. Tissue samples from seven adult males also caught from this area were included to investigate the potential for paternity assignment in this system. Genotyping A panel of 14 microsatellite markers, developed for E. albagula from the neighbouring Fitzroy River drainage (Todd et al. 2011), was evaluated for its discriminatory power for paternity analyses in the study population. Calculations were based on allele frequency estimates for a sample of 47 unrelated adult turtles from the lower Burnett River (Todd et al. in press). The 13 adult animals sampled for the paternity analysis were included in this sample. This ensured that all alleles appearing in the paternity samples were included in population allele frequency estimates, E. V. Todd et al. used as prior information in paternity analyses. Number of alleles and expected and observed heterozygosities were calculated in GENALEX 6.4 (Peakall and Smouse 2006). Exclusion probabilities and polymorphic information content (PIC) were calculated in CERVUS 3.0 (Kalinowski et al. 2007). Probabilities of detecting a multiple mating (PrDM) (Neff and Pitcher 2002) were evaluated for scenarios of equal or skewed (90 : 10) paternal contribution by two males, assuming an average of 12 hatchlings sampled per clutch (E. albagula average clutch size: Hamann et al. 2007), averaged across 10 replicate runs. Conformity of each locus to Hardy–Weinberg expectations was tested in GENALEX, and potential linkage between pairs of loci was tested using GENEPOP 4.0 (Rousset 2008). For multiple comparisons, significance levels were adjusted using a false discovery rate correction (Benjamini and Hochberg 1995). The presence of null (non-amplifying) alleles and genotyping error was evaluated in MICRO-CHECKER 2.2 (van Oosterhout et al. 2004). Total genomic DNA (gDNA) was extracted from preserved tissues using a DNeasy blood and tissue kit (Qiagen) according to manufacturer’s instructions. Microsatellite loci were subsequently amplified by multiplex PCR in 10 mL reactions using the Type-It system (Qiagen), with 2 mM of each primer and 10–100 ng gDNA as per Todd et al. (2011). PCR products were column purified using Sephadex G-50 resin before analysis on a MegaBACE 1000 capillary sequencer (GE Lifesciences). Allele sizes were scored with reference to a 500-bp size standard in Fragment Profiler
1.2 (GE Lifesciences), using the allelebinning option to ensure consistent scoring across samples and runs. A subset of samples was retyped from replicate PCRs to screen for genotyping error. Paternity analyses Initially, we evaluated our ability to detect multiple paternity in each of the six clutches by calculating PrDM on a clutch-byclutch basis, including the mother’s genotype and the number of offspring actually sampled in each simulation. Three approaches were used to examine patterns of single or multiple paternity in each of the six clutches. All are based on Mendelian laws of inheritance. Initially, paternal alleles were identified manually by simple inference, given that the maternal genotype was known and her alleles could therefore be accounted for. Multiple paternity was indicated when >2 paternal alleles at any locus were identified among the progeny of a single clutch. The presence of >2 paternal alleles at two or more loci (alternatively two or more offspring) was considered strong evidence of multiple paternity as an extra paternal allele at a single locus or in a single offspring can be explained by mutation (FitzSimmons 1998). Two contrasting analysis packages were implemented to reconstruct paternities in each clutch, GERUD 2.0 (Jones 2005) and COLONY 2.0 (Jones and Wang 2010). GERUD uses an exhaustive algorithm to determine the minimum number of parents explaining a progeny array of full- or half-sibs, when one or neither parent is known. All possible multilocus paternal genotypes are then reconstructed and the relative contribution of each father is calculated. Where multiple genotype solutions are returned, they are ranked by likelihood based on (1) segregation of paternal alleles and deviations from Mendelian expectations and (2) expected frequencies of genotypes in the population. COLONY Australian snapping turtle multiple paternity Australian Journal of Zoology implements a full-pedigree likelihood method (Wang 2004; Wang and Santure 2009) to jointly infer parentage and sibship relationships from multilocus genetic data under a range of biological scenarios, incorporating data on population allele frequencies and candidate parents when available. Likelihood values for inferred relationships are based on Mendelian inheritance rules. We used COLONY to determine the most likely number of males contributing to each clutch by identifying the number of paternal sibships (i.e. clusters of full- and half-sibs given a single known mother), to reconstruct paternal genotypes at each locus, and to calculate paternity assignment likelihoods for the seven candidate fathers. Promiscuous mating by both sexes was assumed and a full likelihood model was implemented with precision set to ‘high’. Maternal relationships and population allele frequencies were used as prior information and the seven sampled males were included as candidate fathers. Probability of a true father being among the candidates was set to 0.05. Consistency of results was confirmed with replicate runs, each employing a different random number seed. Results Samples Incubated clutch size ranged from nine to 15 eggs, with an average sampling effort of eight hatchlings per clutch or 73% of the incubated clutch size (Table 1). Unsampled eggs were either unhatched, undeveloped or unaccounted for. Though highquality gDNA was extracted from most individuals, gDNA extracted from hatchlings H2b and H2e (unhatched), and H5k (dead in nest), was highly degraded, and gDNA extracted from female F6 was of very low concentration despite repeated attempts. Although microsatellite loci can regularly be amplified from degraded (sheared) samples due to their small amplicon size (typically 100–400 bp), DNA degradation increases the risk of allele dropout and scoring error (Pääbo et al. 2004). The four samples with poor DNA quality amplified inconsistently over the seven markers. Therefore, we excluded hatchlings H2b, H2e, and H5k from further analyses. The genotype of female F6 could be reconstructed with certainty from her offspring’s genotypes at four of the seven loci. Clutch C6 was therefore analysed without maternal information, allowing GERUD and COLONY to reconstruct the mother’s most likely multilocus genotype. Analyses were then repeated with F6’s most likely genotype. Statistical power Allelic diversity in the lower Burnett River population was lower than that reported for the Fitzroy River population from which the microsatellite loci were originally developed (Todd et al. 2011). Of the 14 markers tested, seven exhibited at least three alleles and were potentially useful for paternity analyses (Table 2). At these seven markers, between three and 11 alleles (mean 5.3) were identified among the 47 lower Burnett River individuals, with 33 alleles identified in total. Discriminating power of the seven markers combined was reasonably high (Table 2). The probability of excluding a putative parent (e.g. one of the seven adult males) given a known maternal genotype (PE2) was 0.98 for the combined marker set. With neither parent known, this probability was 0.88 (PE1). However, our ability to detect multiple paternity depended on sampling effort and the paternity scenario within Table 1. Details of six Elseya albagula clutches investigated for the presence of multiple paternity Upper case letters identify clutches (C), females (F), and hatchlings (H) Clutch C1 C2 C3 C4 C5 C6 Mother Clutch size Hatchlings sampled (%) F1 F2 F3 F4 F5 F6 11 9 15 12 12 12 11 (100) 6 (67) 5 (33) 9 (75) 11 (92) 8 (67) Details: sampled hatchlings Details: unsampled eggs H2b, H2e unhatched 1 undeveloped, 2 unhatched 2 undeveloped, 8 unhatched 1 undeveloped, 2 unhatched 1 unaccounted for 4 unaccounted for H5k dead in nest Table 2. Characteristics and statistical power of seven microsatellite markers used for paternity analyses in Elseya albagula sampled from the lower Burnett River, Australia NA, no. of alleles; HO, observed heterozygosity; HE, expected heterozygosity; PE1, probability of excluding a potential parent when neither parent is known; PE2, probability of excluding a potential parent when one parent is known; PIC, polymorphic information content; PrDM, probability of detecting multiple paternity given 12 progeny sampled per clutch and two fathers contributing 50 : 50/90 : 10 Locus Repeat NA HO HE PE1 PE2 PIC Ealb07 Ealb09 Ealb10 Ealb15 Ealb18 Ealb20 Ealb24 (ATAG)8 (ATT)10 (AGG)9 (AC)15 (AC)12 (CT)11 (ATT)5G(TAA)10 6 4 3 4 11 6 3 0.74 0.66 0.30 0.64 0.79 0.70 0.50 0.70 0.74 0.32 0.67 0.80 0.72 0.48 0.29 0.32 0.05 0.23 0.45 0.32 0.12 0.47 0.49 0.17 0.38 0.63 0.49 0.22 0.66 0.69 0.30 0.60 0.78 0.68 0.40 37 0.62 0.63 0.88 0.98 0.59 Overall C PrDM 0.98/0.65 D Australian Journal of Zoology E. V. Todd et al. unsurprising given small clutch sizes and moderate marker variability in the study population. COLONY produced a single most likely solution, which could be replicated over multiple runs of the program. Reconstructed paternal genotypes from COLONY are presented in Supplementary Material Table S2. Paternal contributions by individual males ranged from 18% (father C, clutch C1) to 100% (single paternity, clutch C5) and were typically skewed within those clutches with multiple paternity (Table 3). Omitting or including a reconstructed genotype for female F6 from analyses did not alter the conclusion of multiple paternity in clutch C6. An inconsistency was noted in clutch C4, where a single hatchling (H4a) had a genotype incompatible with the mother’s at locus Ealb07 (Table S1). The hatchling is homozygous for the 105 allele, which does not appear in the mother, but is present in all other F4 offspring and must therefore represent a paternal allele. Two scenarios could potentially account for the mismatch: human error or mutation. Mislabelling of either the hatchling or the mother’s sample cannot be ruled out. However, we consider this unlikely, given that genotypes are consistent at the remaining six markers. Scoring error can also be largely ruled out, as both the mother and hatchling genotypes were consistent over multiple replicate PCRs. Mutation at locus Ealb07 may explain the mismatch. For example, mutation of the maternal 104 allele by addition of a single microsatellite repeat could result in a homozygous hatchling genotype given paternal inheritance of a second 105 allele. Alternatively, mutation causing nonamplification of the maternal allele could also produce this result. Exclusion of either the mother or hatchling H4a in all paternity analyses did not alter the conclusion of multiple paternity for this clutch. each clutch. For example, probability of detecting a multiple mating was high (PrDM 0.98) given equal contribution by two males and 12 hatchlings sampled per clutch, but reduced significantly (PrDM 0.65) if two males contributed disproportionately (9 : 1 ratio). Sampling fewer offspring per clutch also has a negative effect on PrDM. On a clutch-by-clutch basis (Table 3), incorporating the mother’s genotype and the number of hatchlings actually sampled per clutch, PrDM was again high assuming equal contribution by two males (PrDM 0.86–0.99), but was significantly reduced if paternal contributions were skewed (PrDM 0.36–0.64). Deviations from Hardy–Weinberg expectations or linkage equilibrium were not detected at any locus, and there was no evidence of null alleles or scoring error. Paternity analysis We uncovered a high incidence of multiple paternity, with up to three fathers contributing to single clutches. Multiple fathers were implicated in five of the six sampled clutches, with evidence for multiple paternity considered strong in four of these (>2 paternal alleles identified at multiple loci, or multiple offspring if support was at a single locus). Genotype information for females and hatchlings is presented in Supplementary Material Table S1 (inferred paternal alleles in bold). Results concerning the presence or absence of multiple paternity were identical between the three approaches and are compared in Table 3. However, the number of fathers and relative paternal contributions differed between the two analytical approaches. GERUD inferred contributions by at least two fathers in four clutches (C2, C3, C4 and C6), and at least three fathers in a further clutch (C1), whereas COLONY suggested three fathers in clutch C1 as well as C2. Clutch C5 showed no evidence of multiple fathers under any approach. In many cases, GERUD reconstructed several equally likely configurations of paternal genotypes and relative paternal contributions. This is Paternity assignment Of the seven males opportunistically sampled from the same local area as the gravid females, all but one could be excluded Table 3. Comparison between methods for detecting multiple paternity in six Elseya albagula clutches, including individual clutch PrDM values and detailed statistics from COLONY PrDM values are calculated for equal and skewed contribution by two males (50 : 50/90 : 10), averaged over 10 replicate simulations. Result frequency is calculated from all plausible configurations with relatively high likelihood values and is presented for the best configuration. Prob(Inc.) is the probability that all individuals of a given full-sib paternal family are full-sibs and Prob(Exc.) is the probability that no other individuals are full-sibs within this family Clutch N PrDM Inferred fathers across methods Manual GERUD COLONY inference C1 11 0.99/0.64 2 3 3 C2 4 0.95/0.48 2 2 3 C3 5 0.86/0.36 2 2 2 C4 9 0.98/0.57 2 2 2 C5 C6 10 8 0.98/0.61 0.97/0.53A 1 2 1 2 1 2 A PrDM values for clutch C6 are averaged across all possible reconstructed maternal genotypes. Paternal sibship 1A 1B 1C 2A 2B 2C 3A 3B 4A 4B 5A 6A 6B COLONY statistics N full Result Prob sibs frequency (Inc.) 6 3 2 2 1 1 3 2 6 3 10 6 2 0.579 0.858 0.709 0.699 0.955 0.354 0.50 0.97 0.61 0.84 1.00 1.00 0.38 0.93 0.80 0.55 1.00 0.57 0.68 Prob (Exc.) 0.50 0.51 0.54 0.81 0.90 0.91 0.37 0.57 0.60 0.52 1.00 0.57 0.67 Australian snapping turtle multiple paternity as potential fathers of the sampled hatchlings. Analyses in COLONY identified male M1 as a potential father of all 10 hatchlings in clutch C5, the only family with inferred single paternity. Assignment probabilities were extremely high for all 10 offspring and ranged from 0.971 to 1.0 (mean 0.992) (Table S1). Two further assignments were made by COLONY (male M6 was assigned to two hatchlings from clutch C1), although with extremely low probabilities (<0.015) and we consider these spurious matches. Of the reconstructed multilocus paternal genotypes estimated by COLONY (Table S2), no single multilocus genotype was recorded in more than one clutch. Therefore, there was no evidence of shared paternity across the sampled clutches. Discussion Mating system Our genetic parentage analysis revealed a high incidence of multiple paternity in the white-throated snapping turtle (Elseya albagula). We confirm a promiscuous mating system for this species, with evidence of multiple fathers in five of the six sampled clutches and contribution by at least three males to two of these. Results are consistent with high levels of multiple paternity reported previously in turtles (see table 1 in Davy et al. 2011). The lone example of single paternity also indicates that although multiple paternity may be the dominant condition of the mating system in this species, it is not the only strategy employed. It is important to note, however, that apparent single paternity within this clutch may also result from a failure to detect multiple (>2) paternal alleles at the examined loci, and it is difficult to confirm single paternity in such cases. These data provide the first example of multiple paternity in the family Chelidae and extend our knowledge of the taxonomic breadth of multiple paternity in turtles of the Southern Hemisphere. Multiple paternity in E. albagula could have arisen through either (or a combination) of two processes: promiscuous mating by females within a single season, or sperm storage from copulations with separate males across years. We were unable to address the question of whether sperm storage plays a role in encouraging multiple paternity in E. albagula as consecutive clutches from the same female across breeding seasons were not available and because female E. albagula produce a single clutch per season. However, sperm storage across years and within seasons has been shown to be a significant contributor to the incidence of multiple paternity in marine, terrestrial, and other freshwater turtles (Galbraith 1993; Roques et al. 2004, 2006). Long-term sperm storage (up to seven years) is common in reptiles (Gist and Jones 1987; Sever and Hamlett 2002), and is particularly important for turtles as male and female breeding seasons are typically asynchronous (Gist and Jones 1989). Although this is true for E. albagula (mating and oviposition are separated by several months: Hamann et al. 2007), small breeding population size in the study population may make it difficult to differentiate between sperm storage and consecutive matings with the same male, as the same marked individuals have been observed in courtship aggregations in successive years (D. J. Limpus, unpubl. data). Our results are encouraging for the success of future paternity-assignment studies in this system. A single adult male Australian Journal of Zoology E sampled from the same local area as the gravid females was assigned, with high confidence, paternity of all offspring from clutch C6, the only clutch with single paternity. This result may also indicate that females are mating with local males, as suggested by observed mating aggregations. Future studies employing more comprehensive sampling regimes could address questions regarding male and female mate choice, the mating versus reproductive success of males in mating aggregations, and patterns of paternity across years as well as within seasons across the population. Although we found no evidence of multiple mating by males, as none of the reconstructed multilocus paternal genotypes were observed more than once across the six examined clutches, more extensive sampling is necessary to address questions regarding male mating frequency and female mate choice (e.g. McTaggart 2000). Finally, although DNA degradation and patchy sampling of unhatched eggs prevented an analysis of whether paternity influenced hatching success, this would also be a fruitful area for future work. The production of multiply sired clutches seems a common mating strategy across reptilian taxa (reviewed in Uller and Olsson 2008). Patterns in turtles vary widely between species and populations. Variability in sampling strategies and resolving power of molecular markers makes it difficult to compare directly between published accounts and to elucidate overall patterns. Those studies with comprehensive sampling report multiple paternity in between 20 and 90% of clutches (see table 1 in Davy et al. 2011). Up to six males have been reported fathering a single clutch and paternal contributions range from equal (Zbinden et al. 2007) to highly skewed (FitzSimmons 1998). Several factors are thought to influence the extent and variation of multiple paternity within a given population. These include factors influencing the frequency of mate encounter such as the operational sex ratio (Emlen and Oring 1977), size, and density (Jensen et al. 2006) of a population. Costs of mating to females (Lee and Hays 2004), availability of required resources (Garant et al. 2001), and patterns of sperm precedence, competition, and storage (Pearse et al. 2002) are also likely to play a role. Indeed, levels of multiple paternity have been shown to vary significantly between sea turtle populations despite a lack of genetic differentiation between them, supporting the idea that multiple paternity is situation-dependant and also positively correlated with breeding population size (Jensen et al. 2006). Such factors may also be expected to cause geographic and temporal variation in patterns of multiple paternity in E. albagula across its range. Power and limitations of paternity studies Sampling strategy, including numbers and proportions of clutches sampled, and the resolving power of molecular marker systems, are the major factors influencing the statistical power and therefore ability of genetic studies to describe patterns of paternity. Our values for paternity exclusion (PE2 = 0.98) and PrDM (0.98) are consistent with those reported by other studies of multiple paternity in turtles and other groups, which typically quote PE2 and PrDM values >0.9. Our ability to detect multiple paternity, when present, in the current study was reasonably high given equal contribution by multiple sires, but F Australian Journal of Zoology was significantly reduced if paternal contributions were highly skewed (Tables 2 and 3). Overall power to detect multiple matings was limited by low allelic diversity in the study population at typed loci as well as incomplete sampling of clutches. We are therefore likely to have underestimated true levels of multiple paternity in this system. The predictive value of any one locus depends upon the number of alleles as well as their relative frequencies. The more alleles present at a given locus and the closer their frequencies are to equal in a population, the fewer loci are required to achieve high statistical power (Neff and Pitcher 2002). The naturally small clutch sizes of most freshwater and terrestrial turtles limit overall power for detecting multiple paternity but are out of the investigator’s control. We were able to identify multiple fathers among as few as four hatchlings from a single clutch (C2) and when sampling just 33% of a single clutch (C3). However, efforts should be made to analyse entire clutches when clutch sizes are naturally small to allow accurate inferences regarding paternal skew and reproductive success. Future directions In a conservation context, multiple paternity is generally considered as a favourable reproductive strategy, increasing genetic diversity among related offspring (Murray 1964) and, presumably, the effective population size (Ne) of cohorts (Sugg and Chesser 1994; Pearse and Anderson 2009; but see Lee and Hays 2004; Karl 2008; Lotterhos 2011). Elseya albagula is an exemplar species as a riverine specialist highly sensitive to habitat changes associated with serial water impoundment construction throughout a restricted geographic range. We present a snapshot look into the mating system of this species and lay the foundation for future paternity studies in this system. Further research should focus on how anthropogenic activities, particularly those influencing population connectivity and reproductive success, may influence mating system dynamics and, also, what role multiple paternity may play in conferring any long-term adaptive advantage regarding population genetic diversity in context of such threats. Acknowledgements We are grateful to staff of the Paradise Dam Turtle Hatchery and Turtle Research Project (Queensland Department of Environment and Heritage Protection), who assisted with sample collection and field work. Research was conducted under Animal Ethics Approval no. EOA/2007/12/18-22. Comments made by anonymous reviewers greatly improved the manuscript. This work was supported by the Queensland Environmental Protection Agency, James Cook University, and an Australian Postgraduate Award and Queensland Smart Futures Ph.D. Scholarship to EVT. References Baer, B., and Schmid-Hempel, P. (1999). Experimental variation in polyandry affects parasite loads and fitness in a bumble-bee. Nature 397, 151–154. doi:10.1038/16451 Benjamini, Y., and Hochberg, Y. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society. Series B. Methodological 57, 289–300. E. V. Todd et al. Buhlmann, K. A., Akre, T. S. B., Iverson, J. B., Karapatakis, D., Mittermeier, R. A., Georges, A., Rhodin, A. G. J., van Dijk, P. P., and Gibbons, J. W. (2009). A global analysis of tortoise and freshwater turtle distributions with identification of priority conservation areas. Chelonian Conservation and Biology 8, 116–149. doi:10.2744/CCB-0774.1 Coleman, S. W., and Jones, A. G. (2011). Patterns of multiple paternity and maternity in fishes. Biological Journal of the Linnean Society. Linnean Society of London 103, 735–760. doi:10.1111/j.1095-8312.2011.01673.x Davy, C. M., Edwards, T., Lathrop, A., Bratton, M., Hagan, M., Henen, B., Nagy, K. A., Stone, J., Hillard, L. S., and Murphy, R. W. (2011). Polyandry and multiple paternities in the threatened Agassiz’s desert tortoise, Gopherus agassizii. Conservation Genetics 12, 1313–1322. doi:10.1007/s10592-011-0232-y Emlen, S. T., and Oring, L. W. (1977). Ecology, sexual selection, and the evolution of mating systems. Science 197, 215–223. doi:10.1126/ science.327542 FitzSimmons, N. N. (1998). Single paternity of clutches and sperm storage in the promiscuous green turtle (Chelonia mydas). Molecular Ecology 7, 575–584. doi:10.1046/j.1365-294x.1998.00355.x Fleischer, R. C. (1996). Application of molecular methods to the assessment of genetic mating systems in vertebrates. In ‘Molecular Zoology: Advances, Strategies, and Protocols’. (Eds J. D. Ferris and S. R. Palumbi.) pp. 133–161. (John Wiley and Sons: New York.) Galbraith, D. A. (1993). Multiple paternity and sperm storage in turtles. The Herpetological Journal 3, 117–123. Garant, D., Dodson, J. J., and Bernatchez, L. (2001). A genetic evaluation of mating system and determinants of individual reproductive success in Atlantic salmon (Salmo salar L.). The Journal of Heredity 92, 137–145. doi:10.1093/jhered/92.2.137 Gist, D. H., and Jones, J. M. (1987). Storage of sperm in the reptilian oviduct. Scanning Microscopy 1, 1839–1849. Gist, D. H., and Jones, J. M. (1989). Sperm storage within the oviduct of turtles. Journal of Morphology 199, 379–384. doi:10.1002/jmor.1051 990311 Hamann, M., Schauble, C. S., Limpus, D. J., Emerick, S. P., and Limpus, C. J. (2007). Management plan for the conservation of Elseya sp. [Burnett River] in the Burnett River Catchment, Queensland, Australia. Queensland Environmental Protection Agency: Brisbane. Harry, J. L., and Briscoe, D. A. (1988). Multiple paternity in the loggerhead turtle (Caretta caretta). The Journal of Heredity 79, 96–99. Hughes, C. (1998). Integrating molecular techniques with field methods in studies of social behavior: a revolution results. Ecology 79, 383–399. doi:10.1890/0012-9658(1998)079[0383:IMTWFM]2.0.CO;2 Jensen, M. P., Abreu-Grobois, F. A., Frydenberg, J., and Loeschcke, V. (2006). Microsatellites provide insight into contrasting mating patterns in arribada vs. non-arribada Olive Ridley sea turtle rookeries. Molecular Ecology 15, 2567–2575. doi:10.1111/j.1365-294X.2006.02951.x Jones, A. G. (2005). GERUD 2.0: a computer program for the reconstruction of parental genotypes from half-sib progeny arrays with known or unknown parents. Molecular Ecology Notes 5, 708–711. doi:10.1111/ j.1471-8286.2005.01029.x Jones, A. G., and Ardren, W. R. (2003). Methods of parentage analysis in natural populations. Molecular Ecology 12, 2511–2523. doi:10.1046/ j.1365-294X.2003.01928.x Jones, O. R., and Wang, J. L. (2010). COLONY: a program for parentage and sibship inference from multilocus genotype data. Molecular Ecology Resources 10, 551–555. doi:10.1111/j.1755-0998.2009.02 787.x Kalinowski, S. T., Taper, M. L., and Marshall, T. C. (2007). Revising how the computer program CERVUS accommodates genotyping error increases success in paternity assignment. Molecular Ecology 16, 1099–1106. doi:10.1111/j.1365-294X.2007.03089.x Karl, S. A. (2008). The effect of multiple paternity on the genetically effective size of a population. Molecular Ecology 17, 3973–3977. Australian snapping turtle multiple paternity Australian Journal of Zoology Lee, P. L. M., and Hays, G. C. (2004). Polyandry in a marine turtle: females make the best of a bad job. Proceedings of the National Academy of Sciences of the United States of America 101, 6530–6535. doi:10.1073/ pnas.0307982101 Limpus, C. J., Limpus, D. J., Parmenter, C. J., Hodge, J., Forrest, M. J., and McLachlan, J. (2011). The biology and management strategies for freshwater turtles in the Fitzroy River Catchment, with particular emphasis on Elseya albagula and Rheodytes leukops. A study initiated in response to the proposed construction of Rookwood Weir and the raising of Eden Bann Weir. Department of Environment and Heritage Protection, Brisbane. Lotterhos, K. E. (2011). The context-dependent effect of multiple paternity on effective population size. Evolution 65, 1693–1706. McTaggart, S. J. (2000). Good genes or sexy sons? Testing the benefits of female mate choice in the painted turtle, Chrysemys picta. Masters Thesis, University of Guelph, Canada. Murphy, J. B., and Lamoreaux, W. E. (1978). Mating behavior in three Australian chelid turtles (Testudines: Pleurodira: Chelidae) Herpetologica 34, 398–405. Murray, J. (1964). Multiple mating and effective population size in Cepaea nemoralis. Evolution 18, 283–291. doi:10.2307/2406402 Neff, B. D., and Pitcher, T. E. (2002). Assessing the statistical power of genetic analyses to detect multiple mating in fishes. Journal of Fish Biology 61, 739–750. doi:10.1111/j.1095-8649.2002.tb00908.x Nunney, L. (1993). The influence of mating system and overlapping generations on effective population size. Evolution 47, 1329–1341. doi:10.2307/2410151 Pääbo, S., Poinar, H., Serre, D., Jaenicke-Despres, V., Hebler, J., Rohland, N., Kuch, M., Krause, J., Vigilant, L., and Hofreiter, M. (2004). Genetic analyses from ancient DNA. Annual Review of Genetics 38, 645–679. doi:10.1146/annurev.genet.37.110801.143214 Peakall, R., and Smouse, P. E. (2006). GENALEX 6: genetic analysis in Excel. Population genetic software for teaching and research. Molecular Ecology Notes 6, 288–295. doi:10.1111/j.1471-8286.2005.01155.x Pearse, D. E., and Anderson, E. C. (2009). Multiple paternity increases effective population size. Molecular Ecology 18, 3124–3127. doi:10.11 11/j.1365-294X.2009.04268.x Pearse, D. E., and Avise, J. C. (2001). Turtle mating systems: behavior, sperm storage, and genetic paternity. The Journal of Heredity 92, 206–211. doi:10.1093/jhered/92.2.206 Pearse, D. E., Janzen, F. J., and Avise, J. C. (2002). Multiple paternity, sperm storage, and reproductive success of female and male painted turtles (Chrysemys picta) in nature. Behavioral Ecology and Sociobiology 51, 164–171. doi:10.1007/s00265-001-0421-7 Reynolds, J. D. (1996). Animal breeding systems. Trends in Ecology & Evolution 11, 68–72. doi:10.1016/0169-5347(96)81045-7 Roques, S., Diaz-Paniagua, C., and Andreu, A. C. (2004). Microsatellite markers reveal multiple paternity and sperm storage in the Mediterranean spurthighed tortoise, Testudo graeca. Canadian Journal of Zoology 82, 153–159. doi:10.1139/z03-228 G Roques, S., Diaz-Paniagua, C., Portheault, A., Perez-Santigosa, N., and Hidalgo-Vila, J. (2006). Sperm storage and low incidence of multiple paternity in the European pond turtle, Emys orbicularis: a secure but costly strategy? Biological Conservation 129, 236–243. doi:10.1016/ j.biocon.2005.10.039 Rousset, F. (2008). GENEPOP’007: a complete re-implementation of the GENEPOP software for Windows and Linux. Molecular Ecology Resources 8, 103–106. doi:10.1111/j.1471-8286.2007.01931.x Sever, D. M., and Hamlett, W. C. (2002). Female sperm storage in reptiles. The Journal of Experimental Zoology 292, 187–199. doi:10.1002/jez.1154 Stein, J. L., Stein, J. A., and Nix, H. A. (2002). Spatial analysis of anthropogenic river disturbance at regional and continental scales: identifying the wild rivers of Australia. Landscape and Urban Planning 60, 1–25. doi:10.1016/S0169-2046(02)00048-8 Sugg, D. W., and Chesser, R. K. (1994). Effective population sizes with multiple paternity. Genetics 137, 1147–1155. Thomson, S., Georges, A., and Limpus, C. J. (2006). A new species of freshwater turtle in the genus Elseya (Testudines: Chelidae) from central coastal Queensland, Australia. Chelonian Conservation and Biology 5, 74–86. doi:10.2744/1071-8443(2006)5[74:ANSOFT]2.0.CO;2 Todd, E., Blair, D., Hamann, M., and Jerry, D. (2011). Twenty-nine microsatellite markers for two Australian freshwater turtles, Elseya albagula and Emydura macquarii krefftii: development from 454sequence data and utility in related taxa. Conservation Genetics Resources 3, 449–456. doi:10.1007/s12686-010-9377-0 Todd, E. V., Blair, D., Farley, S., Farrington, L., FitzSimmons, N. N., Georges, A., Limpus, C. J., and Jerry, D. R. Contemporary genetic structure reflects historical drainage isolation in an Australian snapping turtle, Elseya albagula. Zoological Journal of the Linnean Society, in press. Uller, T., and Olsson, M. (2008). Multiple paternity in reptiles: patterns and processes. Molecular Ecology 17, 2566–2580. doi:10.1111/j.1365-294X. 2008.03772.x van Oosterhout, C., Hutchinson, W. F., Wills, D. P. M., and Shipley, P. (2004). MICRO-CHECKER: software for identifying and correcting genotyping errors in microsatellite data. Molecular Ecology Notes 4, 535–538. doi:10.1111/j.1471-8286.2004.00684.x Wang, J. L. (2004). Sibship reconstruction from genetic data with typing errors. Genetics 166, 1963–1979. doi:10.1534/genetics.166.4.1963 Wang, J. L., and Santure, A. (2009). Parentage and sibship inference from multilocus genotype data under polygamy. Genetics 181, 1579–1594. doi:10.1534/genetics.108.100214 Zbinden, J. A., Largiader, A. R., Leippert, F., Margaritoulis, D., and Arlettaz, R. (2007). High frequency of multiple paternity in the largest rookery of Mediterranean loggerhead sea turtles. Molecular Ecology 16, 3703–3711. doi:10.1111/j.1365-294X.2007.03426.x Handling Editor: Raoul Mulder www.publish.csiro.au/journals/ajz