Reproductive hormonal patterns in pregnant, pseudopregnant and acyclic captive African wild dogs (Lycaon pictus)

Animal Reproduction Science 156 (2015) 75–82

Contents lists available at ScienceDirect

Animal Reproduction Science journal homepage: www.elsevier.com/locate/anireprosci

Reproductive hormonal patterns in pregnant, pseudopregnant and acyclic captive African wild dogs (Lycaon pictus) L.K. Van der Weyde a,b,c , G.B. Martin a,d , M.A. Blackberry a , V. Gruen e , A. Harland f , M.C.J. Paris a,b,g,∗ a

School of Animal Biology, University of Western Australia, Crawley, WA 6009, Australia Institute for Breeding Rare and Endangered African Mammals (IBREAM), 9 Ainslie Place, Edinburgh, EH3 6AT, UK c Botswana Predator Conservation Trust, Private Bag 13, Maun, Botswana d Nuffield Department of Obstetrics & Gynaecology, University of Oxford, Oxford OX3 9DU, UK e Zoo Duisburg, Mülheimer Straße 273, 47058 Duisburg, Germany f Port Lympne Wild Animal Park, Lympne, Kent, CT21 4PD, UK g Mammal Research Institute, University of Pretoria, Private Bag X20 Hatfield, Pretoria 0028, South Africa b

a r t i c l e

i n f o

Article history: Received 1 July 2014 Received in revised form 29 January 2015 Accepted 5 March 2015 Available online 17 March 2015 Keywords: Oestrous cycle Reproductive suppression Faecal oestradiol Faecal progestagen Pseudopregnancy African wild dog

a b s t r a c t African wild dogs are one of the most endangered canid species, with free-living populations declining as a consequence of habitat loss, disease and human conflict. Captive breeding is considered an important conservation strategy, but is hampered by a poor overall understanding of the reproductive biology of the species. To improve our basic knowledge, we studied hormone patterns in 15 female wild dogs using non-invasive faecal collections. By comparing longitudinal hormone profiles with behavioural and anatomical changes, females could be allocated among three reproductive classes: pregnant (n = 1), pseudopregnant (n = 9) and acyclic (n = 4). We also monitored a single female in which contraception was induced with a deslorelin implant. Comparison of pseudopregnant and acyclic females showed that, in both classes, faecal oestradiol concentrations increased from anoestrus to pro-oestrus then declined into the oestrous and dioestrous phases. Progestagen concentrations rose steadily from anoestrus to the dioestrous phase in both pseudopregnant and acyclic females and, pseudopregnant females had significantly higher concentrations of progestagens than acyclic females in all phases of the oestrous cycle. Most females classed as pseudopregnant were found in female-only groups, suggesting that wild dogs are spontaneous ovulators. Furthermore, only one adult female did not ovulate, so suppression of reproduction in subordinates is likely to be behavioural rather than physiological. © 2015 Published by Elsevier B.V.

1. Introduction With less than 6000 individuals surviving today (Lindsey and Davies-Mostert, 2009), the African wild dog ∗ Corresponding author at: Institute for Breeding Rare and Endangered African Mammals (IBREAM), 9 Ainslie Place, Edinburgh EH3 6AT, UK. Tel.: +61 415 946713. E-mail address: [email protected] (M.C.J. Paris). http://dx.doi.org/10.1016/j.anireprosci.2015.03.003 0378-4320/© 2015 Published by Elsevier B.V.

is currently classified as endangered on the IUCN red list (McNutt et al., 2008). The species is found across Africa south of the Sahara, but the populations are small and fragmented, and face numerous threats such as habitat loss, disease and human conflict (Woodroffe et al., 2007). Viable captive populations and productive breeding programmes are important aspects of wild dog conservation (Frantzen et al., 2001) but, in captivity, there has been mixed success in reproduction (Brand

76

L.K. Van der Weyde et al. / Animal Reproduction Science 156 (2015) 75–82

and Cullen, 1967; Cade, 1967; Dekker, 1968; Fanshawe et al., 1991; Woodroffe et al., 2004), suggesting that we need a deeper understanding of their reproductive biology. Among the canids that have been studied, most have shown similar reproductive attributes, including monooestrous cycles, long pro-oestrous and luteal phases, behavioural suppression and spontaneous ovulation with obligate pseudopregnancy (Asa, 1996). For wild dogs, the findings of previous investigations have been limited by small sample size (van Heerden and Kuhn, 1985; Monfort et al., 1997), male-only studies (Johnston et al., 2007) and the logistical difficulties inherent in frequent sampling in wild populations (Creel et al., 1997). Nevertheless, there is a general consensus that wild dogs are also mono-oestrous seasonal breeders, with breeding mostly restricted to the dominant pair and suppressed in subordinates (van Heerden and Kuhn, 1985; Creel et al., 1997; Monfort et al., 1997; Johnston et al., 2007). The mechanism by which reproduction is inhibited in subordinates is not well understood, but behavioural suppression is considered likely because, in many canids, the physiological ability to breed is retained (Asa, 1996). Behavioural suppression also seems likely in African wild dogs because there are recorded instances of subordinate females becoming pregnant, as well as evidence for multiple sires within litters (Creel et al., 1997; Spiering et al., 2010). On the other hand, physiological suppression, at least in females, has been observed with subordinates not displaying normal oestrous cycles (van Heerden and Kuhn, 1985). It is therefore still not clear whether dominant animals limit reproduction of subordinates by inhibiting behaviour or physiologically suppressing ovulatory cycles. One approach to answering this question is to determine whether ovulation is inhibited or delayed by studying the occurrence of pseudopregnancy, the phenomenon in which non-pregnant females undergo a normal dioestrus, similar to pregnancy, following non-fertile or infertile mating (Concannon et al., 2009). Pseudopregnancy is considered obligatory in canids (Asa, 1996), as demonstrated in domestic dogs (Gudermuth et al., 1998), blue fox Alopex lagopus (Sanson et al., 2005) and maned wolves Chrysocyon brachyurus (Velloso et al., 1998). It has also been observed in two studies of wild dogs (Monfort et al., 1997; Newell-Fulgate et al., 2012) but no difference was detected between pregnant and pseudopregnant cycles in either case. In the present study, therefore, our objective was to observe longitudinal profiles of reproductive hormones during the breeding season in captive female wild dogs under a variety of group-housing conditions, using noninvasive faecal collections. Specifically, we tested whether the endocrine profiles would show whether subordinate individuals were pseudopregnant or acyclic. After individuals were allocated into one or both of these classes, we also tested whether faecal oestrogen concentrations were higher in ovulatory than non-ovulatory individuals, and whether faecal progestagen concentrations were highest in pregnant females, followed by pseudopregnant and then acyclic females.

2. Methods 2.1. Animals and sample collections During the breeding season (June–November) in 2008 and/or 2009, faecal samples were collected from 15 females from four zoological parks: Port Lympne Wild Animal Park (UK), Duisburg Zoo (Germany), West Midland Safari and Leisure Park (UK) and Artis Zoo (Netherlands). Females were aged between 1.7 and 7.3 years and were housed as opposite-sex pairs, or mixed-sex or single-sex groups (Table 1). Opportunistically, faeces were also sampled from a female treated before the onset of the 2009 breeding season with an implant containing deslorelin (Suprelorin® ), a contraceptive known to be effective in domestic dogs (Trigg et al., 2001). The individuals providing the samples were identified by using small, non-toxic, food-grade, coloured plastic beads (Universal Polymer Supplies, Malaga, Australia), or corn or glitter that had been added to food during normal feeding times and days, as appropriate to each zoo. Faeces were collected the following day. The gut transit time averaged 22.7 ± 0.9 h (n = 26, range 17–50 h), similar to the previous estimate of 24 h (Monfort et al., 1997). Samples were collected from most females on a regular basis (generally 1–4 times per week) for up to 3.5 months. Marker feeding time, sample collection times, marker colour and age estimates were recorded for each sample. All samples were collected in sealed plastic ‘ziploc’ bags or plastic tubes and stored at −20 ◦ C before being transported to Australia on dry ice. 2.2. Extraction of faecal samples We used a modification of a previously published method for extraction (Wasser et al., 2000; Palme, 2005). Each sample was thawed and mixed before approximately 0.25 g of wet sample was weighed accurately into a plastic tube, to which 4 mL of 100% methanol was added. After vortex-mixing for 20 min, the tubes were centrifuged for 20 min (4 ◦ C, 3000 × g) and the supernatant was decanted into clean glass tubes and dried under air. Tubes were then capped and stored at −20 ◦ C until assay. Mean recovery during extraction of radiolabeled hormone was 119% for oestradiol and 94% for progesterone. 2.3. Radioimmunoassay Oestradiol concentrations were measured with a single run of an in-house assay developed at UWA. Dried extracts were reconstituted with 1 mL 100% methanol and briefly vortexed. Each sample was diluted 1:10 with gelatine phosphate buffer (GPB) and duplicate 150 ␮L aliquots were removed for assay. Samples and standards were mixed with antiserum (diluted 1:18,000; Bioquest Ltd., NSW) and tracer (2,4,6,7-3 H-E2 , Perkin Elmer, VIC). Samples were shaken, incubated overnight at 4 ◦ C and then treated with 300 ␮L dextran-coated charcoal. After incubation for 20 min at room temperature, the tubes were centrifuged for 10 min (4 ◦ C, 3000 rpm). The supernatant was then added to scintillant and counted. Cross-reactions of the antiserum were listed as oestradiol-17␤ (100%), oestrone

L.K. Van der Weyde et al. / Animal Reproduction Science 156 (2015) 75–82

77

Table 1 Housing conditions of the female African wild dogs from which faecal samples were collected. Individuals are identified with their zoo: P (Port-Lympne); D (Duisburg); A (Artis); W (West Midlands). Location

Individuals sampled

Group housing

Age (year)

Zoo 1

P1 P2, P3, P4, P5

Paired with 1 ♂ Single-sex group

3.7 1.7

D1, D2, D3 D3, D4, D5

Single-sex group Single-sex group

3.1 3.1

Zoo 3

A1

Held with 2 ♂

4.0

One male castrated

Zoo 4

W1 W2, W3 W4

Mixed-sex 11 ♀, 3 ♂ Single-sex group Paired with 1 ♂

7.3 6.4 4.6

Subordinate females on contraceptives In adjacent enclosure to mixed groupa On contraceptive treatment, held adjacent to mixed group above

Zoo 2

a

Notes Siblings, no visual but smell and sound possible from pair above Siblings but separated into two groups Used same enclosure alternately D3 joined other females 2 weeks into study

Females present in 2008 but transferred in 2009.

(1%), oestradiol-17␣ (1%), oestriol (1%) and <0.1% for testosterone, progesterone and corticosteroids. Validation of the assay was by parallelism between the standard curve and serially diluted samples. The limit of detection of the standard curve was 62.5 pg/mL and detection range was 0.1–0.3 ng/g. The intra-assay coefficient of variation was 3.9% and 18.7% for high (n = 2) and medium (n = 2) concentrations. Progestagens were assayed with a commercial progesterone kit (Beckman Coulter, Sydney, Australia), as per kit instructions, using 50 ␮L duplicate aliquots of reconstituted sample that had been diluted at 1:10 with GPB. The limit of detection of the standard curve was 0.11 ng/mL and the detection range was 0.2–0.5 ng/g. Serially diluted samples showed parallelism with the standard curve. The kit was shown to cross-react with a number of progestagen metabolites with those above 1% being progesterone (100%), 6␤-hydroxyprogesterone (2.38%), 5␣-pregnanedione (7.31%), 5␤-pregnanedione (20.18%) and corticosterone (2.03%). Although progesterone is not found in wild dog faeces, pregnanes such as 5␣- and 5␤pregnanediones are most likely present (Monfort et al., 1997) and did cross-react with this kit, so our findings represent progestagens. Quality controls provided with the kit were used to estimate the intra-assay coefficient of variation (7.6 ± 3%; n = 4) and the inter-assay coefficient of variation (15.3%). 2.4. Behavioural and anatomical data Reproductive behaviours, agonistic interactions and anatomical changes were recorded. Reproductive behaviours included interest by males (male courtship behaviours) and/or females towards females, such as sniffing or licking the ano-genital region, sniffing urine and faeces, marking behaviour, physical contact, mate guarding, mounting and copulations. Anatomical changes included swelling of the vulva, sanguineous discharge, teat swelling and belly distension. Interactions among two or more individuals were recorded, along with the winner and loser as defined by aggression or submissive actions (e.g., head lowered, ears flattened, showing of belly, direct approach and fighting). Other behaviours

such as digging, begging, whining and “hoo” calling were noted when observed. Individuals from Port Lympne Wild Animal Park were monitored for an average of 2 h (1–3 h) per day for 10 weeks over a 14-week period. Individuals held at Duisburg Zoo were monitored for a minimum of 1 h per day during three 2-week periods during the same season. Observations of wild dogs from the other zoos were recorded by keepers on an opportunistic basis, and were usually limited to dates of mounting, copulation and aggressive encounters. 2.5. Classification of individuals Females were classified as either pregnant, pseudopregnant (ovulatory non-pregnant) or acyclic (non-ovulatory) on the basis of combined observations of oestrous behaviour, anatomical changes, progestagen concentrations and parturition. The female that had been treated with deslorelin was classed as ‘contracepted’. All profiles for each female were aligned to the first oestradiol peak (defined to be Day 0) as it was not possible to confirm the day of ovulation by any other means. We used an iterative method to determine day 0: all values for each female were averaged and then values that were greater than two standard deviations (SD) above or below the mean were removed, and the average recalculated. This procedure was repeated until no values were more than two SD from the mean. The day of the first value that was greater than two SD above the mean was deemed to be day 0. Using behavioural and anatomical observations from this study, as well as previous reports for this species (van Heerden and Kuhn, 1985; Monfort et al., 1997) and endocrine changes for canids (Concannon et al., 1975), we allocated data points to one of four phases: (1) Pro-oestrus – male courtship behaviours, vulva oedema (days −7 to +1); (2) Oestrus – presumed ovulation following oestradiol peak and mating (days +2 to +13); (3) Dioestrus – end of mating, average length of pregnancy (days +14 to +84); (4) Anoestrus – all days outside the above periods.

78

L.K. Van der Weyde et al. / Animal Reproduction Science 156 (2015) 75–82

2.6. Statistical analysis All statistical analysis was done using IBM SPSS Statistics v19 (SPSS, IBM, Chicago, IL, USA). Sample age and collection times were not significantly correlated with hormone concentrations so were excluded from all remaining analyses. A linear mixed model that allows for missing data was used to compare hormone concentrations among classes and reproductive phases. These comparisons were calculated by using maximum likelihood with class and phase used as fixed effects. As multiple samples were collected from the same individual, we included individual within class as a random factor in the model. We used variance components as our covariance matrix because it was found to be the best fit using Akaike’s information criteria. We tested intercept and slope as random models and they remained in the model when significant. Means reported are estimated marginal means as they are adjusted for other variables used in the model. Data were checked for normality by plotting residuals after each analysis and nonnormal data were log-transformed where necessary. Post hoc pairwise comparisons were carried out using Bonferroni adjustments and significance was set at p < 0.05. 3. Results Three females were held with males and all were observed mating, but only one successfully conceived and gave birth to pups (in mid-November) so she was considered to be the only pregnant female that was observed. The remaining two females, despite mating, did not show clear signs of pregnancy (e.g., enlarged teats or an extended belly that would typify the later stages of gestation) and were not observed to give birth, so were classed as pseudopregnant. It is possible that these two females were pregnant and experienced spontaneous abortion or embryo reabsorption during early gestation, but we could not determine this from the endocrine profiles since luteal function is not affected by embryo loss in canids. A further 7 of 11 females from single-sex groups showed increased progestagen concentrations suggesting ovulation and thus were classed as pseudopregnant. The remaining 4 females from single-sex groups were classed as acyclic due to consistently low progestagen concentrations throughout the season. Endocrine profiles for these classes, as well as the female undergoing contraceptive treatment, are shown in Fig. 1. All mating observations occurred within a 2-week period beginning in late August and continuing into early September. 3.1. Behavioural pro-oestrus and oestrus For all 3 females held with males, male courtship behaviours began at the onset of the oestradiol peak. These behaviours included sniffing and licking of the ano-genital region, sniffing of urine or faeces, scent marking, mate guarding, physical contact and the male resting his head on the back of the female. Sexual proceptive behaviours by females, such as presentation to males and tail deviation (Beach, 1976), were less obvious or not observed. Vulva oedema was also observed in these females from the onset of the oestradiol peak and lasted throughout most

of gestation in the pregnant female. Female sexual receptivity, identified by mating, commenced 13 days after the first oestradiol peak in the pregnant female and lasted 4 days, with gestation lasting 71 days from the first day of mating. A second female mated with up to 3 males for a period of 6 days, 25 days after an oestradiol peak; the third female mated for at least 3 days but the oestradiol peak was not detected due to sampling limitations. Judging from the pregnant female, sexual proceptivity and receptivity lasted approximately 18 days. Among the females in single-sex groups, behaviours such as sniffing and licking of the ano-genital region, sniffing of urine and faeces, scent marking and mountings were also observed during times of peak and declining oestradiol concentrations. Vulva oedema (lasting 2–3 weeks) and, in at least one female classed as pseudopregnant, a sanguineous discharge, were observed approximately one week after an observed oestradiol peak. 3.2. Faecal oestradiol concentrations Faecal oestradiol concentrations remained low during anoestrus, but when females entered pro-oestrus, oestradiol concentrations increased markedly. The pregnant female presented two peak oestradiol values (36.7 and 41 ng/g) 4 days apart and a third peak (48.3 ng/g) on day 26. Peak values ranged from 14.4 to 65.9 ng/g in the pseudopregnant females and from 16.9 to 23.1 ng/g in the acyclic females. The contracepted female also showed a peak value (15.1 ng/g) during this time, on the same day as the peak value in the dominant female in an adjacent enclosure. Concentrations were lower in the contracepted female than in the cycling females, but not as low as anoestrus values. As we only had one female from each of the pregnant and contracepted classes, we restricted our analysis to pseudopregnant and acyclic classes. Oestradiol concentrations did not differ significantly between these classes, but did differ significantly among the four phases (F3,307 = 5.64, p < 0.01), with values for pro-oestrus and dioestrus being significantly higher than those for anoestrus. However, within each phase, there were no differences between the classes (Fig. 2a). 3.3. Faecal progestagen concentrations Faecal progestagen concentrations were highly variable in all females, but particularly in the pregnant and pseudopregnant females. Concentrations were low during anoestrus and pro-oestrus and then rose during the oestrous and the dioestrous phases of the cycle. In the pregnant female, a pre-ovulatory peak (2045 ng/g) was observed immediately after the first peak in oestradiol, coinciding with the likely time of the LH surge. In the contracepted female, progestagen concentrations rose to a peak value of 1330 ng/g at the expected time of prooestrus, coinciding with the oestradiol peak, but then returned to basal values. A significant difference among phases was observed among the classes (F3,325 = 8.31, p < 0.01) with concentrations during oestrus and dioestrus being higher than those during anoestrus. Overall, pseudopregnant females had significantly higher progestagen

L.K. Van der Weyde et al. / Animal Reproduction Science 156 (2015) 75–82

1500

30 1000

20

500

10 0

0 -50

-30

Acyclic

-10

10

30

50

70

90

110

50

2000

40

1500

30 1000

20

500

10 0 -30

-10

10 30 50 Day of cycle

70

90

1500

30 1000

20

500

10

60

0 -50

2000

40

0

2500

110

Oestradiol (ng/g)

Oestradiol (ng/g)

60

2500

Pseudo-pregnant

-50

-30

-10

10

30

50

70

90

Contraceptive

Progestagens (ng/g)

2000

40

60 50

0 110 2500

50

2000

40

1500

30 1000

20

500

10 0

Progestagens (ng/g)

2500

Birth

Oestradiol (ng/g)

Copulations

Progestagens (ng/g)

Pregnant

50

Progestagens (ng/g)

Oestradiol (ng/g)

60

79

0 -50

-30

-10

10

30 50 Day of cycle

70

90

110

Fig. 1. Faecal concentrations of oestradiol (closed circles) and progestagen (open circles) in captive African wild dogs that had been allocated among four reproductive classes during the breeding season. For the pregnant (n = 1) and contraceptive (n = 1) classes, the values are for all samples collected from a single female. For pseudopregnant (n = 9) and acyclic (n = 4) classes, weekly means (±s.e.m.) are presented. Individual profiles were aligned to the oestradiol peak, Day 0 of the cycle. Arrows represent first day of copulation and birth in the pregnant female. The hatched bar indicates the duration of anatomical and behavioural signs of pro-oestrus and oestrus.

concentrations than acyclic females (F1,16 = 25.76, p < 0.01). There was a significant interaction between class and phase (F3,325 = 3.97, p < 0.01), with pseudopregnant females having significantly higher concentrations in every phase of the cycle (Fig. 2b).

Log oestradiol (ng/g)

1.10

a)

1.00 0.90 0.80 0.70 0.60 1,400

b)

*

Progestagens (ng/g)

1,200 1,000 800

*

*

*

600 400 200 0 Anoestrus

Pro-oestrus

Oestrus

Dioestrus

Fig. 2. Estimated marginal means (±s.e.m.) for faecal concentrations of oestradiol (a) and progestagen (b) in pseudopregnant (black bars) and acyclic (white bars) captive female African wild dogs, for each phase of the oestrous cycle. *p < 0.05.

4. Discussion By monitoring longitudinal hormonal patterns and aligning them with behavioural and anatomical observations, we were able to classify individuals as pregnant, pseudopregnant or acyclic. In the single pregnant female, a pre-ovulatory surge in progestagens was evident after a peak in oestradiol, coinciding with declining oestrogen concentrations, as seen in domestic dogs and considered the primary means for initiating sexual behaviour (Concannon et al., 1977). Male courtship behaviours began at the onset of the oestradiol peak but female sexual receptivity began considerably later, after the likely time of ovulation, suggesting that oviductal oocytes have long life spans in wild dogs, as they do in the domestic bitch (Concannon et al., 2009). However, another female mated approximately 21 days after the likely timing of ovulation, perhaps explaining why she did not conceive. Our observations of mating span, lengths of pro-oestrus, oestrus and gestation, and behaviours associated with reproduction, were all consistent with previous reports of breeding individuals (Cade, 1967; Dekker, 1968; Reich, 1981; van Heerden and Kuhn, 1985; Creel et al., 1997; Monfort et al., 1997). Interestingly, females in some single-sex female groups displayed male courtship behaviours, including mounting of an oestrous female by a non-oestrous female, considered by Beach (1976) to be unusual. Significant changes in faecal concentrations of oestradiol and progestagen throughout the oestrous cycle were detected in pseudopregnant and acyclic females despite the large amount of variability within and between individuals. We used wet samples for ease of handling, so some of this variability might be removed with the use of dry samples (Wasser et al., 1993; Palme, 2005). Faecal

80

L.K. Van der Weyde et al. / Animal Reproduction Science 156 (2015) 75–82

oestradiol concentrations did not differ between pseudopregnant and acyclic groups within any phase of the oestrous cycle, but were lowest during anoestrus and highest during pro-oestrus. Faecal progestagen concentrations rose steadily from anoestrus to dioestrus in both pseudopregnant and acyclic females. In addition, pseudopregnant females excreted significantly larger amounts of progestagens than acyclic females during all phases of the oestrous cycle. In the maned wolf, pseudopregnant and unpaired (non-mated) females do not differ in mean oestrogen concentrations but do differ in progestagen concentrations during the luteal phase (Songsasen et al., 2006). These observations contrast with one study of wild dogs where a non-pregnant female showed higher baseline and peak values of faecal progestagens than either of two pregnant females and one suspected pseudopregnant female (Newell-Fulgate et al., 2012). Similarly, a study of the red wolf (Canis rufus) showed that oestrogen and progestagen concentrations in acyclic females did not vary over time and that basal values for both hormones were significantly higher than in pseudopregnant females (Walker et al., 2002). Despite the similarities among canids in the characteristics of their reproductive processes, the species appear to differ. The oestradiol and progestagen profiles in our single pregnant female were similar to those in the pseudopregnant females. Sample size prevented further analysis, but our observations do agree with those of Monfort et al. (1997) who suggested that progestagen concentrations are not reliable indicators of pregnancy for wild dogs – indeed, this seems to be the general case for canids (Concannon et al., 1975; Chakraborty, 1987; Gudermuth et al., 1998). A better alternative might be oestrone concentration which seems to differ between pregnant and pseudopregnant bitches (Chakraborty, 1987). Similarly, relaxin, with its advantage of placental origin (Concannon et al., 2009), has been used to determine pregnancy in other wild canids (Carlson and Gese, 2008) and has been trialled in African wild dogs, although initial results were not compelling (Bauman et al., 2008). Clearly, hormone-based pregnancy detection deserves more attention because the lack of a reliable test is holding up progress in dog management. In the female with a deslorelin implant, faecal concentrations of oestradiol and progestagen were low and similar to the values observed in the acyclic females. Interestingly, this female still showed a brief rise in oestradiol coinciding with a rise in progestagens but, as in the acyclic females, this increase appeared to be insufficient to induce ovulation or behavioural oestrus. Treatment with deslorelin has not always proven effective in the prevention of ovulation and pregnancy in female wild dogs (Boutelle and Bertschinger, 2010) and our data, although obviously limited, do suggest that deslorelin implants do not completely suppress ovarian hormone secretion. Importantly, this study confirmed that female wild dogs can ovulate and become pseudopregnant in the absence of males, supporting the general consensus that most canids are spontaneous ovulators with obligate pseudopregnancy (Asa and Valdespino, 1998; Concannon et al., 2009). The exception is the Island fox (Urocyon littoralis) that exhibits induced oestrus (Asa et al., 2007). Failure to ovulate has

been reported in a few studies of canids, with individuals held in single-sex groups or as singletons, but many of these females were apparently pre-pubertal (Porton et al., 1987; Songsasen et al., 2006). In our female-only groups, age was also probably a contributing factor because four females were peri-pubertal at the beginning of our investigation and three of them remained acyclic. The fourth female ovulated at approximately 23 months of age, previously considered the average time of maturity for this species (van Heerden and Kuhn, 1985). In the three females that remained acyclic, physiological suppression and thus delayed sexual maturation (Wasser and Barash, 1983) seems unlikely because subordinate individuals show high progestagen concentrations (van Heerden and Kuhn, 1985; Creel et al., 1997). Sexual maturation was not synchronised so, if our study had been extended, the other three females might have become sexually mature. Only one of our adult females did not become pseudopregnant. During the early stages of the breeding season, she was seriously harassed by two siblings, leading to several fights, and was then removed to a group with two other siblings. Fighting might have been caused by increased aggression by the dominant female (Creel et al., 1997). Stress can suppress reproduction but this explanation seems unlikely because, in African wild dogs, the concentrations of stress hormones are generally higher in dominant than in subordinate individuals (Creel et al., 1997). Overall, it is far more likely that reproductive suppression in this species is behavioural rather than physiological. This perspective is critical for the management of captive breeding because, in captivity, almost all adults ovulate and become pregnant or pseudopregnant, and many females are implanted with contraceptives or held as single-sex groups to prevent inbreeding or unwanted pregnancies. We postulate that behavioural suppression of reproduction is poorer in zoos than in free-ranging populations where multiple female pregnancies in a single pack are rare (Frame et al., 1979; Malcom and Marten, 1982; Creel et al., 1998). If this is the case, more information is needed because single-sex management groups are not always viable – levels of aggression can be high (van Heerden et al., 1996), space is often limited, and future introductions to opposite single-sex groups are complex and risky (L Van der Weyde, unpublished observation). If we are to continue to improve population management in zoos, further studies of reproductive biology are necessary. Studies of captive individuals are a practical way of improving our understanding of the finer mechanisms of reproduction and, equally importantly, they allow zoos to play a vital role in their own management as well as in wildlife conservation (Wildt and Wemmer, 1999; Tribe and Booth, 2003). Despite limited sample size, the present study supplements the current database of reproductive biology for the African wild dog. This knowledge can in turn be applied to the management of African wild dogs in captivity and to studies of individuals in the threatened free-living populations. In addition, the reproductive biology of the African wild dog is intrinsically interesting because it features reproductive suppression, biased sexratio in litters, and female-biased dispersal.

L.K. Van der Weyde et al. / Animal Reproduction Science 156 (2015) 75–82

5. Conclusions Faecal concentrations of oestradiol and progestagens can be used to identify pseudopregnant and acyclic African wild dogs. High oestradiol levels during pro-oestrus and high progestagen levels during dioestrus lead us to conclude that the females are spontaneous ovulators, similar to other canids. Most subordinate females also ovulate, suggesting that reproductive suppression in this species is likely to be behavioural in nature. The outcomes of this research are valuable for captive breeding of this endangered species. Conflict of interest None declared. Acknowledgements Funding was provided by the School of Animal Biology (University of Western Australia) and the Institute for Breeding Rare and Endangered African Mammals (IBREAM). We would like to thank all the keepers and staff from Port Lympne Wild Animal Park, Zoo Duisburg, West Midland Safari and Leisure Park and Artis Zoo who assisted with faecal collections, with particular mention to Richard Barnes, Rebecca Rust, Nina Honer, Angela Potter and Katie McDonald. We must also thank Helen Johnson and Wendy Jaques who provided additional support during the study. References Asa, C., 1996. Hormonal and experiential factors in the expression of social and parental behavior in canids. In: Solomon, N., French, J. (Eds.), Cooperative Breeding in Mammals. Cambridge University Press, Cambridge, pp. 129–149. Asa, C., Bauman, J., Coonan, T., Gray, M., 2007. Evidence for induced estrus or ovulation in a canid, the island fox (Urocyon Littoralis). J. Mammal. 88, 436–440. Asa, C., Valdespino, C., 1998. Canid reproductive biology: an integration of proximate mechanisms and ultimate causes. Am. Zool. 38, 251–259. Bauman, J., Clifford, D., Asa, C., 2008. Pregnancy diagnosis in wild canids using a commercially available relaxin assay. Zoo Biol. 27, 406–413. Beach, F., 1976. Sexual attractivity, proceptivity and receptivity in female mammals. Horm. Behav. 7, 105–138. Boutelle, S., Bertschinger, H., 2010. Reproductive management in captive and wild canids: contraception challenges. Int. Zoo Yearb. 44, 109–120. Brand, D., Cullen, L., 1967. Breeding the cape hunting dog at Pretoria Zoo. Int. Zoo Yearb. 7, 124–126. Cade, C., 1967. Notes on breeding the cape hunting dog Lycaon pictus at Nairobi Zoo. Int. Zoo Yearb. 7, 122–123. Carlson, D., Gese, E., 2008. Reproductive biology of the coyote (Canis latrans): integration of mating behavior, reproductive hormones, and vaginal cytology. J. Mammal. 89, 654–664. Chakraborty, P., 1987. Reproductive hormone concentrations during estrus, pregnancy and pseudopregnancy in the Labrador bitch. Theriogenology 27, 827–840. Concannon, P., Castracane, V., Temple, M., Montanez, A., 2009. Endocrine control of ovarian function in dogs and other carnivores. Anim. Reprod. Sci. 6, 172–193. Concannon, P., Hansel, W., McEntee, K., 1977. Changes in LH, progesterone and sexual behaviour associated with preovulatory lutenization in the bitch. Biol. Reprod. 17, 604–613. Concannon, P., Hansel, W., Visek, W., 1975. The ovarian cycle of the bitch: plasma estrogen, LH and progesterone. Biol. Reprod. 13, 112–121.

81

Creel, S., Creel, N., Mills, M., Monfort, S., 1997. Rank and reproduction in cooperatively breeding African wild dogs: behavioral and endocrine correlates. Behav. Ecol. 8, 298–306. Creel, S., Creel, N., Monfort, S., 1998. Birth order, estrogens and sex-ratio adaptation in African wild dogs (Lycaon pictus). Anim. Reprod. Sci. 53, 315–320. Dekker, D., 1968. Breeding the cape hunting dog at the Amsterdam Zoo. Int. Zoo Yearb. 8, 27–30. Fanshawe, J., Frame, L., Ginsberg, J., 1991. The wild dog – Africa’s vanishing carnivore. Oryx 25, 137–146. Frame, L., Malcom, J., Frame, G., van Lawick, H., 1979. Social organization of African wild dogs (Lycaon pictus) on the Serengeti plains, Tanzania 1967–1978. Z. Tierpsychol. 50, 225–249. Frantzen, M., Ferguson, J., de Villiers, M., 2001. The conservation role of captive African wild dogs (Lycaon pictus). Biol. Conserv. 100, 253–260. Gudermuth, D., Concannon, P., Daels, P., Lasley, B., 1998. Pregnancyspecific elevations in fecal concentrations of estradiol, testosterone and progesterone in the domestic dog (Canis familiaris). Theriogenology 50, 237–248. Johnston, S., Ward, D., Lemon, J., Gunn, I., MacCallum, C., Keeley, T., Blyde, D., 2007. Studies of male reproduction in captive African wild dogs (Lycaon pictus). Anim. Reprod. Sci. 100, 338–355. Lindsey, P., Davies-Mostert, H., 2009. South African action plan for conservation of cheetahs and wild dogs. In: Report from a National Conservation Action Planning Workshop, Bela Bela, Limpopo Province, South Africa, 17–19 June 2009. Malcom, J., Marten, K., 1982. Natural selection and the communal rearing of pups in African wild dogs (Lycaon pictus). Behav. Ecol. Sociobiol. 10, 1–13. McNutt, J., Mills, M., McCreery, K., Rasmussen, G., Robbins, R., Woodroffe, R., 2008. Lycaon pictus, IUCN 2011. IUCN Red List of Threatened Species. Version 2011.2, www.iucnredlist.org (downloaded 21.03.2012). Monfort, S., Wasser, S., Mashburn, K., Burke, M., Brewer, B., Creel, S., 1997. Steroid metabolism and validation of noninvasive endocrine monitoring in the African wild dog (Lycaon pictus). Zoo Biol. 16, 533–548. Newell-Fulgate, A.E., Nothling, J.O., Bertschinger, H.J., 2012. Seasonal changes in steroid hormoneprofiles, body weight, semen quality, and the reproductive tract in captive African wild dogs (Lycaonn pictus) in South Africa. Gen. Comp. Endocrinol. 178, 272–281. Palme, R., 2005. Measuring fecal steroids. Guidelines for practical application. Ann. N. Y. Acad. Sci. 1046, 75–80. Porton, I., Kleiman, D., Rodden, M., 1987. Aseasonality of bush dog reproduction and the influence of social factors on the estrous cycle. J. Mammal. 68, 867–871. Reich, A., 1981. The Behavior and Ecology of the African Wild Dog (Lycaon pictus) in the Kruger National Park, Yale University. Yale University, New Haven, CT, pp. 1–557. Sanson, G., Brown, J., Farstad, W., 2005. Non-invasive faecal steroid monitoring of ovarian and adrenal activity in farmed blue fox (Alopex lagopus) females during late pregnancy, parturition and lactation onset. Anim. Reprod. Sci. 87, 309–319. Songsasen, N., Rodden, M., Brown, J., Wildt, D., 2006. Patterns of fecal gonadal hormone metabolites in the maned wolf (Chrysocyon brachyurus). Theriogenology 66, 1743–1750. Spiering, P., Somers, M., Maldonado, J., Wildt, D., Szykman Gunther, M., 2010. Reproductive sharing and proximate factors mediating cooperative breeding in the African wild dog (Lycaon pictus). Behav. Ecol. Sociobiol. 64, 583–592. Tribe, A., Booth, R., 2003. Assessing the role of zoos in wildlife conservation. Hum. Dimens. Wildl. 8, 65–74. Trigg, T.E., Wright, P.J., Armour, A.F., Williamson, P.E., Junaidi, A., Martin, G.B., Doyle, A.G., Walsh, J., 2001. Long term reversible desexing of male dogs and oestrus postponement of bitches, using a GnRH analogue implant. In: Concannon, P.W., England, G.C.W., Farstad, W., Linde-Forsberg, C., Verstegen, J.P., Doberska, C. (Eds.), Advances in Reproduction in Dogs, Cats and Exotic Carnivores. Journal of Reproduction and Fertility, Supplement 57, pp. 255–261. van Heerden, J., Kuhn, F., 1985. Reproduction in captive hunting dogs Lycaon pictus. S. Afr. J. Wildl. Res. 15, 80–84. van Heerden, J., Verster, R., Penrith, M., Espie, I., 1996. Disease and mortality in captive wild dogs (Lycaon pictus). JSA Vet. Assoc. 67, 141–145. Velloso, A., Wasser, S., Monfort, S., Dietz, J., 1998. Longitudinal fecal steroid excetion in maned wolves (Chrysocyon brachyurus). Gen. Comp. Endocrinol. 112, 96–107. Walker, S., Waddell, W., Goodrowe, K., 2002. Reproductive endocrine patterns in captive female and male red wolves (Canis rufus) assessed by fecal and serum hormone analysis. Zoo Biol. 21, 321–335.

82

L.K. Van der Weyde et al. / Animal Reproduction Science 156 (2015) 75–82

Wasser, S., Barash, D., 1983. Reproductive suppression among female mammals: implications for biomedicine and sexual selection theory. Q. Rev. Biol. 58, 513–538. Wasser, S., Hunt, K., Brown, J., Cooper, K., Crockett, C., Bechert, U., Millspaugh, J., Larson, S., Monfort, S., 2000. A generalized fecal glucocorticoid assay for use in a diverse array of nondomestic mammalian and avian species. Gen. Comp. Endocrinol. 120, 260–275. Wasser, S., Thomas, R., Nair, P., Guidry, C., Southers, J., Lucas, J., Wildt, D., Monfort, S., 1993. Effects of dietary fibre on faecal steroid measurements in baboons (Papio cynocephalus cynocephalus). J. Reprod. Fertil. 97, 569–574.

Wildt, D., Wemmer, C., 1999. Sex and wildlife: the role of reproductive science in conservation. Biodiv. Conserv. 8, 965–976. Woodroffe, R., Davies-Mostert, H., Ginsberg, J., Graf, J., Leigh, K., McCreery, K., Mills, G., Pole, A., Rasmussen, G., Robbins, R., Somers, M., Szykman, M., 2007. Rates and causes of mortality in endangered African wild dogs Lycaon pictus: lessons for management and monitoring. Oryx 41, 215–223. Woodroffe, R., McNutt, J., Mills, M., 2004. African wild dog (Lycaon pictus). In: Sillero-Zubiri, C., Hoffman, M., Macdonald, D. (Eds.), Canids: Foxes, Wolves, Jackals and Dogs. Status Survey and Conservation Action Plan. IUCN, Gland, Switzerland, pp. 174–183.