Reproduction in male swamp wallabies (Wallabia bicolor): puberty and the effects of season

Abstract

This study describes pubertal changes in testes and epididymides and seasonal changes in the adult male reproductive organs and plasma androgen concentrations of the swamp wallaby (Wallabia bicolor). Pre-pubescent males had testes with solid seminiferous cords and spermatogenesis only to the stage of gonocytes. Their epididymides had empty lumina along their entire length. The testes of three males undergoing puberty had some lumen formation and mitotic activity. Their epididymides were similar in appearance to those of adult males but were entirely devoid of any cells within the lumen of the duct. Three other pubescent males showed full lumen formation in the testes and spermatogenesis up to the elongating spermatid stage. Their epididymides were similar in appearance to those of adult males but with no spermatozoa in the duct. However, cells of testicular origin were found in the lumen of the duct in all regions suggesting that testicular fluids and immature germ cells shed into the rete testes flow through the seminiferous tubules into the epididymis before the release of mature testicular spermatozoa. The weights of testes and epididymides of adult males showed no change throughout the year but prostate weight and plasma androgen concentrations varied significantly with season, with maximums in spring and summer and minimums in winter. The volume fraction of Leydig cells and seminiferous tubules was significantly lower in winter than in summer; but, despite this, maturing spermatozoa were found in the testes throughout the year. Females in the area conceived year-round, suggesting that seasonal changes in the male reproductive tract did not prevent at least some males from breeding throughout the year.

Introduction

The swamp wallaby (Wallabia bicolor, Desmarest 1804) is a common, medium-sized, browsing macropodid marsupial. It possesses a unique suite of behavioural, dental, karyotypic and reproductive features and has been classified as the sole extant member of the genus Wallabia. The only study on seasonal reproductive activity in male swamp wallabies is from two populations, one fenced to exclude dingoes and the other unfenced, in north-eastern New South Wales. Testis weight declined outside the breeding season in a seasonally breeding population but remained constant in a continuously breeding population (Robertshaw & Harden, 1986). However, despite the decline in testis weight (from x¯ = 20.2 g during the breeding season to x¯ = 16.9 g outside the breeding season) in the seasonally breeding population, mature spermatozoa were still present in the testes of all adult males throughout the year in both populations (Robertshaw & Harden, 1986). Neither puberty nor seasonal changes in the male accessory glands and reproductive hormones have been described in the swamp wallaby.

In many species, both males and females are seasonal breeders (Bronson & Heideman, 1994), which allows for the synchronisation of reproduction with a variety of external stimuli so that birth or weaning occurs at the optimum time each year. Seasonal breeding is controlled by proximate and ultimate factors (Baker, 1938 in Sadleir, 1972). Ultimate factors are those which, through evolution, shape species to their environment to maximise reproductive success whereas proximate factors affect the physiology of reproduction in the short term. The most important factors affecting reproduction are food availability, rainfall, ambient temperature and photoperiod and the latter three all influence food availability. The behaviour, morphology and physiology of animals change as they enter puberty and develop towards sexual maturity and, as with seasonal breeding, this process is timed such that the lifetime reproductive output of an individual is maximised. In male mammals, development and maturation of the hypothalamic-pituitary axis, increased testosterone production and the development of accessory gland function and secondary sexual characteristics occur at puberty and the initiation of spermatogenesis and the sperm maturation function of the epididymis results in the production of mature spermatozoa (Bronson & Rissman, 1986).

Very few studies have examined changes during puberty in male marsupials. In male macropodids, studies of the external signs of puberty are limited to observations of a rapid increase in body weight and testicular and scrotal size (Frith & Sharman, 1964; Poole, 1973, 1976; Catt, 1977; Inns, 1982; Williamson et al. 1990). The end of puberty in males is difficult to ascertain. Unlike females, puberty in males is not associated with a single event such as the first oestrus or first mating. Most studies of puberty in male macropodids divide males into immature, maturing and mature classes based on testis histology and the age spread of the males in each class is then recorded (Frith & Sharman, 1964; Sharman & Pilton, 1964; Poole, 1973, 1976; Poole & Catling, 1974; Merchant, 1976; Catt, 1977; Merchant & Calaby, 1981; Bolton et al. 1982; Inns, 1982; Williamson et al. 1990).

The most detailed study, in the tammar wallaby (Macropus eugenii), examined changes in the testes during puberty (Williamson et al. 1990). The seminiferous cords of immature males (at 13 months of age) contained pre-Sertoli cells, pre-spermatogonia and gonocytes. The interstitium mainly consisted of fibroblasts with some mature Leydig cells. By 19 months of age, spermatogenesis had been initiated but not all stages were present in all males and no spermatozoa were found in the epididymides. In mature males, by 25 months of age, all developmental stages of spermatogenesis were present within the testis and spermatozoa were found in all regions of the epididymis. The interstitium contained clumps of large, mature Leydig cells. The hypothalamic-pituitary axis was fully functional at the onset of puberty but the pituitary-testicular axis matured later, although before 25 months of age (Williamson et al. 1990).

Only one study has described structural changes in the developing epididymis of a marsupial, the brown antechinus, Antechinus stuartii (Taggart & Temple-Smith, 1992). During pubertal development in this species, the diameter and epithelial height of the duct increase and the most proximate regions of the epididymis develop before the more distal regions. The presence of spermatocytes in the lumen of the duct before the initial outflow of spermatozoa indicates that testicular fluids, and probably androgens, are passing into the duct from the testis during development as occurs in other mammal species (Setchell, 1970; Orgebin-Crist et al. 1975; Sun & Flickinger, 1979). The asynchronous development and differentiation of the regions of the epididymis occurs when circulating androgens are low, suggesting that epididymal development is influenced by androgens in testicular fluids (Setchell et al. 1969; Cooper & Waites, 1974; Voglmayr et al. 1977; Taggart & Temple-Smith, 1992).

Like puberty, the effects of season on reproduction in the male have been reported in few marsupial species. The common brushtail possum (Trichosurus vulpecula, Gilmore, 1969) shows no annual cycle of testicular or epididymal weights but has a restricted mating period and motility of spermatozoa voided in the urine increases during the breeding season (Bolliger, 1942 in Biggers, 1966). Similarly prostate weight and 5α-dihydrotestosterone content of the prostate increase during the breeding season and there is evidence that metabolic activity of the prostate also increases at this time (Gilmore, 1969; Curlewis & Stone, 1985). Similar observations have recently been made in the southern hairy-nosed wombat (Lasiorhinus latifrons) in which seasonal changes in accessory glands and semen quality were linked to reduced levels of circulating androgens and Leydig cell regression but testicular weight and spermatogenesis showed no seasonal change (Taggart et al. 2005).

In contrast, male ringtail possums (Pseudocheirus peregrinus) and greater gliders (Petauroides volans) show an annual regression of the testis, which is associated with cessation of spermatogenic activity, and epididymis after the highly restricted seasonal mating and birthing periods (Hughes et al. 1965; Smith, 1969; Baldwin et al. 1974). Regressed testes in the greater glider resemble those of sexually immature males with solid seminiferous cords consisting of Sertoli cells and spermatogonia, and small, regressed Leydig cells with pyknotic nuclei in the interstitium (Baldwin et al. 1974).

Seasonal regression of the testis is rare in male macropodid marsupials, even in species in which breeding is strictly seasonal or opportunistic (Poole & Pilton, 1964; Poole, 1973, 1975). The tammar wallaby and Bennett's wallaby (M. r. banksianus) have a highly seasonal pattern of births and a breeding strategy which includes a photoperiodically controlled seasonal embryonic diapause (Berger, 1966; Hearn, 1973; Renfree & Tyndale-Biscoe, 1973; Hearn, 1974; Catt, 1977; Merchant & Calaby, 1981; Renfree et al. 1981; Loudon et al. 1985). In both species there is no seasonal change in testicular or epididymal size and males are potentially fertile year-round (Inns, 1982). However, in both these species, prostate weights decline outside the breeding season (Merchant & Calaby, 1981; Inns, 1982). In the tammar wallaby, plasma testosterone and luteinising hormone (LH) also increase during the breeding season but only in the presence of oestrous females (Catling & Sutherland, 1980).

The specific aims of this study were to examine (1) changes in the male reproductive system during puberty, and (2) the effects of season on circulating androgen levels and the morphometry of the testes, epididymides and prostate in a continuously breeding population of the swamp wallaby. This study of reproduction in a mammal population in the wild can be used as a comparison to other studies of wild populations of mammals, living in various environments and at different time-points, to elucidate how environmental factors such as climate and food availability, and the interactions between these factors, influence reproductive physiology.

Methods

Animals

All animal handling and experiments were approved by the Melbourne University Animal Experimentation Ethics Committee (number 00110). Fourteen 2 to 3-day trips were made to a culling site in the Grand Ridge Plantation (Maryvale), a timber (pine and eucalyptus) plantation located in South Gippsland, Australia, approximately 170 km east of Melbourne (38°E 11′S, 146°E 26′E) between November 2000 and April 2003. Animals were culled by the plantation under an Authority to Control Wildlife by two shooters on contract to Grand Ridge Plantations. A mobile laboratory was parked within the plantation. Animals were shot at night throughout the plantation and rapidly transported to the laboratory for processing.

All males were weighed to the nearest 100 g using a 50 kg dial-suspended balance (Salter, Melbourne, Australia). Males weighing more than 11 kg, a conservative estimate of the minimum weight of sexually mature males (Jarman, 1983), were classed as adult and males weighing less than 11 kg were classed as sub-adult. A total of 149 adult males and 19 sub-adult males were sampled.

A 20 mL–25 mL sample of blood was collected by heart puncture or from the dorsal aorta of sampled males. All blood samples were collected via sterile 18 g needles into sterile or washed 20 mL syringes containing approximately 50 µL of a 125 i.u. mL⁻¹ solution of heparin sodium (DBL laboratories, Melbourne, Australia) in 0.9% saline. Blood was centrifuged at ~438 g for 10 min after which plasma was collected and stored in two to three aliquants at –20 °C.

Testes, epididymides and prostate from each male were dissected and weighed individually using an electronic balance. One testis, one epididymis and the prostate from each male was then fixed by immersion in 10% neutral buffered formalin (NBF) for histology. Prostates were bisected and the tunica albuginea of each testis was pierced to aid fixation.

Reproductive organ morphometry

Paired testis and epididymis weights and prostate weight were plotted against body weight in EXCEL (Microsoft Corp., 2002) for all adult males. For each data set the Pearson product-moment correlation coefficient was calculated and the significance of the correlation tested. Because all organ weights correlated significantly with body weight, organ weight as a percentage of body weight was used for all further analysis to remove the effects of body weight on organ weight. Boxplots of each organ weight as a percent of body weight grouped by month and by season were constructed in SYSTAT version 10 (SPSS Inc., Chicago, IL, USA). Kruskal-Wallis analyses (Zar, 1999) were performed in SYSTAT version 10 (SPSS Inc.) to test if organ weight as a percentage of body weight differed in each month and in each season. The non-parametric test was used instead of a parametric test because of a high degree of heteroskedasticity (Zar, 1999). An a posteriori non-parametric multiple comparison test adjusted for unequal sample sizes and tied ranks (which calculates the test statistic Q,Zar, 1999) was carried out if the Kruskal-Wallis analysis rejected the null hypothesis.

Combined testis weight, combined epididymis weight and prostate weight were also plotted against body weight for each male weighing less than 11 kg to show trends in organ weight changes over time in pubescent males. However the small sample size precluded statistical analyses.

Stereological analysis

A group of 10 adult males from summer and 10 adult males from winter were chosen at random for stereological analysis of the testis. A t-test was performed to check for a significant difference in body size between the groups. Because no significant difference was found, subsequent measurements were not adjusted for body size. A testis from each male was divided into three 2 mm segments using systematic uniform random sampling (Wreford, 1995). These segments were embedded in glycol methacrylate (Technovit 7100, Heraeus, Grale Scientific, Mulgrave, Vic., Australia) and two 3 µm sections were taken from each for analysis. The sections were mounted on Superfrost Plus glass slides (Menzel-Galser, Braunschweig, Germany), brought to water then stained with haematoxylin and eosin and coverslipped. The slides were analysed under a light microscope fitted with a motorised stage and connected to a computer with a high resolution monitor. A point counting method (Wreford, 1995) with 40 regularly spaced counting frames and 16 regularly spaced points per frame was used in OLYMPUS CAST (Computer Assisted Stereological Toolbox, version 2.1.4, Olympus, Albertslund, Denmark) analysis software to estimate the volume fraction of Leydig cells and seminiferous tubules. Mean volume fractions for Leydig cells and seminiferous tubules in summer and winter were plotted and compared using Mann–Whitney tests (Zar, 1999). The volume fraction of Leydig cells was plotted against the plasma androgen concentration of each male (see next section) and the Pearson product-moment correlation coefficient was calculated. The significance of the correlation was tested (Zar, 1999).

Radioimmunoassay procedure

All plasma collected from the field was thawed once and two 1 mL aliquants for each animal were removed and re-frozen until the extraction stage. Levels of circulating androgen in wild-shot males at different times of the year were determined by radioimmunoassay using the method of Williamson et al. (1990), validated for the swamp wallaby (see next section). Assay results were accepted only if they fell within the central working range of the standard curve. Any samples with concentrations falling outside the working range were either concentrated and re-assayed in a 200 µL aliquot or diluted by using only 100 µL in the assay.

All unknown androgen concentrations were determined in two assay runs. The mean plasma concentration of androgen in each season was plotted and compared by an ANOVA performed in SYSTAT (SPSS Inc.). An analysis of androgen concentrations by month was omitted as the sample sizes in some months were too small (e.g. n = 4 for May and August). An a posteriori one-tailed Dunnett's test (a multiple comparison test which calculates the test statistic Q,Zar, 1999) was carried out on mean plasma androgen concentrations for each pair of seasons.

Radioimmunoassay validation

The antiserum (C-0457, Bioquest Limited, North Ryde, NSW, Australia) cross reacts significantly with testosterone (100%) and 5α-dihydrotestosterone (98%). Parallelism between testosterone standards and endogenous androgen was tested by two serial dilutions (50 µL, 100 µL and 200 µL) of plasma from an adult male. There was no obvious lack of parallelism in the working range of the standard curve to suggest non-linearity caused by interfering substances in the serum. The extraction efficiencies, as determined by the recovery of radioactive testosterone equilibrated with adult male plasma (two replicates per assay), were 83% and 78% for the first assay and 87% and 81% for the second assay. Accuracy was given by the addition and quantitative recovery of 0 pg mL⁻¹, 20 pg mL⁻¹, 200 pg mL⁻¹ and 800 pg mL⁻¹ of testosterone added to female plasma. The female plasma was found to have no detectable androgens when no testosterone was added. The amount added correlated strongly with the amount recovered (r = 0.999). The sensitivity of the assay was calculated as the lowest standard which differed from the zero standard by more than twice the standard deviation multiplied by the procedural losses. For the first assay the sensitivity was 154 pg mL⁻¹ and for the second it was 98 pg mL⁻¹. Precision was evaluated by the calculation of an intra-assay coefficient of variation and an inter-assay coefficient of variation. The intra-assay coefficient of variation was determined by assaying quality control samples and was 2.9% for QC084 (x¯ = 1330 pg mL⁻¹, n = 5) and 2.3% for QCHI (x¯ = 3250 pg mL⁻¹, n = 4). The inter-assay coefficient of variation was 7.0% for QC084 and 5.3% for QCHI. Solvent and buffer blanks were always below the sensitivity of the assay.

Light microscopy

The stage of sexual maturity was assessed in histological sections of the testes and epididymides of males weighing less than 11 kg. A slice of tissue approximately 5 mm orthogonal to the long axis was taken from the centre of each testis collected from all males weighing less than 11 kg. This slice was subdivided into six pie shaped segments and one randomly selected segment was embedded in paraffin wax. Three groups of three 5–10 µm sections from each segment, separated by 100 µm, were cut, mounted, deparaffinised and stained with haematoxylin and eosin. Each section was examined using light microscopy. Epididymides were paraffin embedded lying on their side either in their entirety or, where too large, divided at the midline of the long axis, and were sectioned parasagitally. Two to four 5–10 µm sections were taken from the centre of each epididymis, mounted, deparaffinised, stained with haematoxylin and eosin and examined by light microscopy. The duct of the epididymis was examined in the approximate centre of each of the caput, corpus and cauda epididymides (Fig. 1). A testis and epididymis sample from a large adult male (20 kg) was prepared and examined in the same way for comparison.

Fig. 1.

Epididymis of an adult swamp wallaby indicating the location of the caput, corpus and cauda epididymides. ED = efferent duct, VD = vas deferens.

Results

Reproductive organ morphometry

Nineteen males weighing less than 11 kg were collected from the Maryvale culling site. Combined testis weight, combined epididymis weight and prostate weight all increased with body weight in males weighing more than 8 kg (Fig. 2).

Fig. 2.

Combined testis weight (a), combined epididymis weight (b) and prostate weight (c) of pre-pubescent, pubescent and post-pubescent swamp wallaby males collected from the Maryvale culling site. All organ weights increase with body weight in males weighing over 8 kg, that is, pubescent and post-pubescent males.

A total of 149 adult males were collected from the Maryvale culling site. Body weight correlated significantly with paired testis weight (r = 0.65, t₁₄₇ = 10.41, P < 0.0001), paired epididymis weight (r = 0.86, t₁₄₇ = 20.93, P < 0.0001) and prostate weight (r = 0.59, t₁₄₇ = 8.92, P < 0.0001). Neither combined testis nor combined epididymal weights differed significantly from month to month (testis: H₁₀ = 15.67, P = 0.109, Fig. 3; epididymis H₁₀ = 15.57, P = 0.11, Fig. 4), however, prostate weights differed significantly (H₁₀ = 26.77, P = 0.003, Fig. 5). Prostate weights were higher in February than in April (Q = 3.2, 0.005 > P > 0.01), June (Q = 2.9, 0.01 > P > 0.025) or July (Q = 2.9, 0.01 > P > 0.025), higher in March than in April (Q = 2.8, 0.025 > P > 0.05) or July (Q = 3.4, 0.0025 > P > 0.005) and higher in September (Q = 2.6, 0.025 > P > 0.01), November (Q = 3.0, 0.01 > P > 0.025) and December than in July (Q = 2.7, 0.025 > P > 0.05).

Fig. 3.

Median monthly combined testis weights as a percentage of body weights (central line of box) for swamp wallaby males from South Gippsland, showing the interquartile range (ends of box) and values which lie within 1.5 times (whiskers) and 3 times (asterisks) of the interquartile range.

Fig. 4.

Median monthly combined epididymis weights as a percentage of body weights (central line of box) for swamp wallaby males from South Gippsland, showing the interquartile range (ends of box) and values which lie within 1.5 times (whiskers) and 3 times (asterisks) of the interquartile range.

Fig. 5.

Median monthly prostate weights as a percentage of body weights (central line of box) for swamp wallaby males from South Gippsland, showing the interquartile range (ends of box) and values which lie within 1.5 times (whiskers) and 3 times (asterisks) of the interquartile range. The circle is an outlier. Significant differences between months are shown by the lines at the top of the graph. See text for details.

Stereological analysis

There was no significant difference in body weight between males in summer and males in winter (t_12.6 = 1.155, P = 0.270). The interstitium of adult testes in summer contained more and larger Leydig cells than in winter (Fig. 6) and stereological analysis showed that the volume fraction of Leydig cells was significantly higher in summer than in winter (summer x¯ = 3.61, s = 1.04; winter x¯ = 2.46, s = 0.96; U = 80.0, P = 0.023). Spermatozoa were found in the seminiferous tubules of summer and winter males (Fig. 7), however, the volume fraction of seminiferous tubule was significantly higher in summer than in winter (summer x¯ = 25.56, s = 4.36; winter x¯ = 22.54, s = 4.05; U = 76.0, P = 0.049). The volume fraction of Leydig cells correlated significantly with androgen concentrations (r = 0.61, t₁₉ = 8.35, P < 0.0001, Fig. 8).

Fig. 6.

Changes in the appearance of the testicular interstitium of adult swamp wallaby males collected from South Gippsland in summer (a) and winter (b). The interstitium contains smaller and fewer Leydig cells (LC) in winter. Seminiferous tubules (ST) can be seen.

Fig. 7.

Appearance of the seminiferous tubules of adult swamp wallaby males from South Gippsland in summer (a) and winter (b). Spermatozoa (Sp) were found in all males examined. Note the difference between the amount of interstitium and Leydig cell size between summer and winter. St = late stage spermatids, LC = Leydig cells.

Fig. 8.

The plasma testosterone concentration of swamp wallaby males collected from the South Gippsland culling site plotted against their volume fraction of Leydig cells, showing the least-squares line of best fit. A significant correlation was found between the volume fraction of Leydig cells and plasma testosterone concentration (r = 0.61, t₁₉ = 8.35, P < 0.0001).

Plasma androgen

Circulating androgen concentrations ranged from 14.1 ng mL⁻¹ in an 18 kg male in spring to 0.04 ng mL⁻¹ in a 23 kg male in winter. Mean circulating androgen concentrations (Table 1) differed significantly between seasons (F₁₁₃ = 3.751, P = 0.013) with concentrations in summer being higher than in winter (q = 2.39, P > 0.05) and autumn (q = 2.42, P > 0.05) and concentrations in spring being higher than in winter (q = 2.32, P > 0.05) and autumn (q = 2.34, P > 0.05). The minimum testosterone concentration was less than 0.2 ng mL⁻¹ in all seasons (Table 1). The maximum testosterone concentrations in spring (14.06 ng mL⁻¹) and summer (11.25 ng mL⁻¹) were 3.3 and 2.6 times greater, respectively, than those in autumn (4.32 ng mL⁻¹) and winter (4.32 ng mL⁻¹).

Table 1.

Mean peripheral testosterone concentrations of swamp wallabies collected from Maryvale in each season showing sample size and range. The concentration of testosterone differs significantly in each season (F₁₁₃ = 3.751, P = 0.013). Concentrations in summer are higher than in winter (q = 2.39, P > 0.05) and autumn (q = 2.42, P > 0.05) and concentrations in spring are higher than in winter (q = 2.32, P > 0.05) and autumn (q = 2.34, P > 0.05). Concentrations do not differ significantly between spring and summer and between autumn and winter

Season	Plasma testosterone concentration (ng mL−1, x¯ ± SEM)	Sample size (n)	Range (ng mL−1)
Summer	3.28 ± 0.47	35	0.08–11.25
Autumn	1.74 ± 0.08	25	0.18–4.32
Winter	1.67 ± 0.28	21	0.04–4.32
Spring	3.23 ± 0.50	36	0.10–14.06

Light microscopy

A total of twenty males were examined. The testis of the adult male (body weight 20 kg) showed seminiferous tubules with fully open lumina and all developmental stages of spermatogenesis, including spermatozoa, present (Fig. 9e). Sertoli cells appeared differentiated as did the Leydig cells of the interstitium. The duct of the caput region of the epididymis of the adult male (Fig. 10d) was lined by a pseudostratified columnar epithelium of principal cells with a clearly vacuolated appearance and a layer of basal cells flattened against the basal lamina. External to the basal lamina was a thin (1–2 cells thick) layer of myoepithelial cells enclosing the duct. The duct of the corpus region consisted of the same cell types that looked broadly similar to those of the caput but with a slightly thicker (2–3 cells thick) layer of myoepithelial cells (Fig. 11d). The duct of the cauda region was much larger in diameter than in the preceding regions. It was also lined by a pseudostratified epithelium with an irregular luminal surface when compared to the epithelium of the preceding regions, which had a regular luminal surface. Again, a layer of basal cells occurred along the basal lamina and external to this was a thick layer of true smooth muscle cells which replaced the myoepithelial cell layer in the caput and corpus regions (Fig. 12d).

Fig. 9.

Testes of swamp wallabies collected from the Maryvale culling site before, during and after puberty showing the increase in seminiferous cord/tubule diameter and development of the lumen with age. (a) Pre-pubescent swamp wallaby (6.2 kg body weight). Seminiferous cords are small with no lumen. Gonocytes (G) can be seen to the centre of the cords and basally. (b) Pubescent male (9.0 kg body weight). During early puberty some lumen (L) development can be seen and mitotic figures (M) are apparent. (c) Pubescent male (8.8 kg body weight). During later puberty lumina are fully developed and spermatogenic stages up to the early spermatid (ES) stage can be seen. (d) Young post pubescent male (10.3 kg body weight). All developmental stages of spermatogenesis, including spermatozoa are present. (e) Fully grown adult male (20 kg body weight). All developmental stages of spermatogenesis, including spermatozoa are present. The testes of the young post-pubescent male cannot be distinguished from those of the fully grown adult. LC = Leydig cells, LS = Late spermatids, PS = Pre-Sertoli cells, St = Sertoli cells.

Fig. 10.

The caput epididymidis of swamp wallabies collected from the Maryvale culling site before, during and after puberty showing the increase in duct diameter during development. (a) Pre-pubescent swamp wallaby (6.8 kg body weight). A lumen (L) is evident along the length of the duct. The epithelium (E) lining the duct appears as almost a cuboidal epithelium. The layer of myoepithelial cells (M) surrounding the duct is thicker than at later stages. No cells of any type are found in the lumen of the duct. (b) Pubescent male (9.0 kg body weight). The epithelial layer lining the duct of the caput epididymidis is a pseudostratified columnar epithelium similar in appearance to later stages but thicker. The lumen of the duct contains cellular material (C) of testicular origin (see Fig. 14). (c) Young post pubescent male (10.3 kg body weight). The epithelial layer lining the duct of the caput epididymidis is a pseudostratified columnar epithelium Spermatozoa (Sp) are present in the lumen of the duct. (d) Fully grown adult male (20.0 kg body weight). The epithelial layer lining the duct of the caput epididymidis is a pseudostratified columnar epithelium Spermatozoa (Sp) are present in the lumen of the duct. The caput epididymidis of the young post-pubescent male cannot be distinguished from that of the fully grown adult. S = connective tissue stroma.

Fig. 11.

The corpus epididymidis of swamp wallabies collected from the Maryvale culling site before, during and after puberty showing the increase in duct diameter during development. (a) Pre-pubescent swamp wallaby (6.2 kg body weight). A lumen (L) is evident along the length of the duct. The epithelium (E) lining the duct appears as almost a cuboidal epithelium. The layer of myoepithelial cells (M) surrounding the duct is thicker than at later stages. No cells of any type are found in the lumen of the duct. (b) Pubescent male (9.0 kg body weight). The epithelial layer lining the duct of the corpus epididymidis is a pseudostratified columnar epithelium similar in appearance to later stages but thicker. The lumen of the duct contains cellular material (C) of testicular origin (see Figs 5, 7). (c) Young post pubescent male (10.3 kg body weight). The epithelial layer lining the duct of the corpus epididymidis is a pseudostratified columnar epithelium Spermatozoa (Sp) are present in the lumen of the duct. (d) Fully grown adult male (20 kg body weight). The epithelial layer lining the duct of the corpus epididymidis is a pseudostratified columnar epithelium. Spermatozoa (Sp) are present in the lumen of the duct. The corpus epididymidis of the young post-pubescent male cannot be distinguished from that of the fully grown adult. The epithelium of the corpus epididymidis is taller than that of the corpus in both the post-pubescent male and fully grown adult male. S = connective tissue stroma.

Fig. 12.

The cauda epididymidis of swamp wallabies collected from the Maryvale culling site before, during and after puberty showing the increase in duct diameter and change in lumen shape during development. (a) Pre-pubescent swamp wallaby (6.2 kg body weight). A lumen (L) is evident along the length of the duct. The epithelium (E) lining the duct is thicker than at later stages and the lumen is star shaped prior to the expansion of the duct. As the lumen of the duct becomes distended during development the epithelial layer becomes thinner. No cells of any type are found in the lumen of the duct. (b) Pubescent male (9.0 kg body weight). The epithelial layer lining the duct of the cauda epididymidis is thicker than in the preceding regions and has an irregular inner border. The lumen of the duct contains cellular material (C) of testicular origin (see Fig. 15). (c) Young post pubescent male (10.3 kg body weight). The epithelial layer lining the duct of the cauda epididymidis is thicker than in the preceding regions and has an irregular inner border Spermatozoa (Sp) are present in the lumen of the duct. (d) Fully grown adult male (20 kg body weight). The epithelial layer lining the duct of the cauda epididymidis is thicker than in the preceding regions and has an irregular inner border. Spermatozoa (Sp) are present in the lumen of the duct. The cauda epididymidis of the young post-pubescent male cannot be distinguished from that of the fully grown adult. The myoepithelial layer of the preceding regions is replaced by a true smooth muscle layer. S = connective tissue stroma.

Large quantities of spermatozoa were apparent in all three regions of the epididymis of the adult male. The spermatozoa of the caput region showed the classic marsupial ‘thumbtack’ shape (Temple-Smith & Bedford, 1976) with the sperm head perpendicular to the tail and an obvious cytoplasmic droplet below the head (Fig. 13a). In the corpus region, most sperm did not have the cytoplasmic droplet attached and many shed cytoplasmic droplets were apparent. Spermatozoa in the caudal region appeared mature with the sperm head lying parallel to the tail (Fig. 13b).

Fig. 13.

(a) Spermatozoon (in focus) from the caput epididymidis of an adult (20 kg body weight) swamp wallaby collected from the Maryvale culling site, showing the head (H) lying perpendicular to the sperm tail (ST) and the cytoplasmic droplet (CD) below the head. (b) Spermatozoon from the cauda epididymidis of an adult (20 kg body weight) swamp wallaby collected from the Maryvale culling site, showing the head (H) rotated to lie parallel with the sperm tail (ST). The cytoplasmic droplet has been shed.

Six males weighing 9.25 kg to 10.5 kg were classed as post-pubescent based on testis and epididymis histology. The appearance of the testes of these males was similar to that of the 20 kg adult male (Fig. 9e). The lumina of the seminiferous tubules were open and all developmental stages of spermatogenesis including spermatozoa were present in the seminiferous epithelium. Sertoli and Leydig cells were similar in appearance to those of the 20 kg adult male and were therefore morphologically differentiated. The epididymis of the post-pubescent males showed adult morphology (Figs 10c, 12c, 13c) and spermatozoa were present in all regions of the epididymis.

Seven males weighing between 4 kg and 8 kg were classified as pre-pubescent. The testes of all of these males consisted of spermatogenic cords with no evidence of a lumen (Fig. 9a). Within the cords gonocytes were apparent, some towards the centre of the cords and some basal, and Sertoli cells appeared similar to those in the adult. Little interstitium was evident but there were some Leydig cells although these contained little cytoplasm.

A lumen was evident along the entire length of the epididymis. Overall, the size of the epididymal duct was smaller in all regions than in the corresponding regions of those males classed as post-pubescent. The principal cells of the caput and corpus epididymidis did not resemble those of post-pubescent males, being more similar to a simple cuboidal epithelium without the vacuolated appearance found in the older males. The outermost layer of myoepithelial cells was thicker in these males than the post-pubescent males. The duct of the cauda epididymidis was lined with a pseudostratified columnar epithelium with a highly ridged luminal border. The lumen was small and often star shaped. No cells of any type were found in the lumen of the duct throughout the epididymis (Figs 10a, 12a, 13a).

A further group of six males were classed as pubescent as testicular and epididymal development was intermediate between the pre- and post-pubescent males. In three of these males, two weighing 9 kg and one weighing 7 kg, the lumina of the seminiferous tubules were not fully formed. However, some spermatogenic activity was evident with mitotic figures seen (Fig. 9b). The epithelium of the caput (Fig. 10b) and corpus epididymidis (Fig. 11b) was similar to that of the post-pubescent males; however, the cauda epididymidis appeared similar to that of the pre-pubescent males (Fig. 12b). The other three males, weighing 6 kg, 8 kg and 10 kg showed full lumen formation in the testes and a number of more advanced spermatogenic stages (Fig. 9c). In addition to primary spermatocytes and round spermatids, some shaping and elongating spermatids were seen although no spermatozoa were present. The appearance of the epididymis was similar to that of post-pubescent males but, although no spermatozoa were observed in the duct, aggregates of prematurely shed primary spermatocytes, round spermatids, and some shaping spermatids were seen in the lumen of the duct in all regions of the epididymis. Multinucleate cells, possibly spermatid stages that had been shed and had undergone karyokinesis with no cytokinesis, and cells with pyknotic nuclei were also observed (Fig. 14a–d). These six pubescent males were collected in March, July, September, November and December, that is, in each season of the year.

Fig. 14.

Cells found in the lumen of the epididymis of pubescent males. A variety of cells of testicular origin are found. (a) A shaping spermatid (SS). (b) An elongating spermatid (ES). (c) Multinucleate cell (MN), probably a round spermatid, which has undergone karyokinesis with no cytokinesis. (d) Cell with a pyknotic nucleus (P).

Discussion

Testicular structure shows a number of marked changes during puberty. These changes are similar to those reported for the tammar wallaby (Williamson et al. 1990), the red deer stag (Lincoln, 1971a), the rhesus monkey (Macaca mulatta) (Van Wagenen & Simpson, 1954) and human (Mancini et al. 1960). In the swamp wallaby, as in other mammalian species, prior to puberty the seminiferous cords lack a central lumen and the seminiferous epithelium consists only of Sertoli cells and gonocytes, some of which are basal and others centrally located (Van Wagenen & Simpson, 1954; Mancini et al. 1960; Lincoln, 1971a; Williamson et al. 1990). Initiation of spermatogenesis and secretory activity of the Sertoli cells is accompanied by lumen formation and an increase in height of the seminiferous epithelium (Van Wagenen & Simpson, 1954; Mancini et al. 1960; Gier & Marion, 1969; Lincoln, 1971a; Gondos, 1977; Williamson et al. 1990). Despite the small number of samples, this study confirmed that, as in other mammals, testicular growth is most rapid during the time that the lumen is developing and spermatogenesis begins (Frith & Sharman, 1964; Sharman & Calaby, 1964; Lincoln, 1971a; Poole, 1973; Poole, 1976; Catt, 1977; Inns, 1982; Plant, 1988; Williamson et al. 1990; Taggart & Temple-Smith, 1992) and the pattern of development and growth of the epididymis and prostate followed a similar pattern to that of the testis. This is presumably due to the action of testicular androgens, the production of which increases as the Leydig cells in the testes develop, driving the growth and development of these organs (Bronson & Rissman, 1986).

There are no previous studies of the development of the pubertal epididymis in any macropodid and only one study of epididymal development in a marsupial, the brown antechinus (Taggart & Temple-Smith, 1992). In the current study, the morphology of the adult swamp wallaby epididymis was similar to that of the adult tammar wallaby (Jones et al. 1984). In both species the duct is lined with pseudostratified columnar epithelium although the stereocilia reported for the tammar wallaby were not obvious at the magnifications used in this study. The smaller diameter and taller epithelium of the caput epididymidis compared to the corpus epididymidis was not noticeable; however, the much greater diameter, and taller and more irregular epithelium of the duct in the cauda epididymidis in the tammar wallaby were also obvious in the swamp wallaby. In the tammar wallaby, as in most other mammals (Martan, 1969; Bedford, 1975), the initial (caput) region of the epididymis appears to be involved in the absorption of testicular fluid and sperm concentration, the middle (corpus) segment in the maturation of spermatozoa and the terminal (cauda) segment in sperm storage (Jones et al. 1984). There was histological evidence of increasing concentration of spermatozoa between the caput and corpus regions of the swamp wallaby epididymis consistent with fluid absorption and sperm concentration in the caput epididymidis of this species. The normal maturation changes in sperm structure during passage through the epididymis (Temple-Smith & Bedford, 1976; Jones et al. 1984; Cummins et al. 1986) were observed in the swamp wallaby. The corpus region contained a mixture of mature and immature spermatozoa and many shed cytoplasmic droplets but no cytoplasmic droplets were seen in the lumen of the duct in the caudal region, suggesting that, as in the tammar wallaby (Jones et al. 1984), brushtail possum (Temple-Smith & Bedford, 1976) and honey possum (Tarsipes rostratus, Cummins et al. 1986), these had been absorbed in the corpus region.

The epididymis of pubescent swamp wallabies was strikingly different to that of mature males. Of most interest was the observation that in pubescent males, the caput and corpus epididymidis resembled that of adult males but the cauda was more similar to that of pre-pubescent males. In other species, not only plasma androgens but also androgens in testicular secretions that pass into the epididymis are thought to influence epididymal development (Orgebin-Crist et al. 1975; Sun & Flickinger, 1979). In the brown antechinus, the development of the more proximal regions of the epididymis precedes the development of the more distal regions of the epididymis primarily under the influence of androgens from testicular fluid rather than plasma androgens (Taggart & Temple-Smith, 1992). This was also observed in the swamp wallaby. Another similarity between antechinus and swamp wallaby epididymal development is the presence of cells that are of clearly testicular origin in the lumen of the epididymis preceding the release of normal maturing spermatozoa (Taggart & Temple-Smith, 1992). This, and the development of the lumina in the seminiferous tubules, indicates that testicular fluids and sloughed germ cells flow through the seminiferous tubules into the epididymis before the release of testicular spermatozoa as has been found in some other eutherian species (Setchell et al. 1969; Sun & Flickinger, 1979).

The lack of a seasonal cycle of change in adult testis weights when females are breeding continuously agrees with the results of the study by Robertshaw & Harden (1986). Female swamp wallabies in the Maryvale population were not seasonal breeders, because individuals with small pouch young were found in each month of the year (Paplinska et al. 2006) and in males no change was observed in testicular or epididymal weights throughout the year. However, in our study, season affected the mass and function of some other parts of the male reproductive system. The volume fractions of Leydig cells in the interstitium of the testis and of the seminiferous tubules were both significantly lower in winter than in summer, and the prostate showed distinct changes in weight with season. Many seasonally breeding male marsupials show no change in testicular or epididymal weights but large changes have been reported in accessory gland weights and plasma testosterone concentrations (Gilmore, 1969; Merchant & Calaby, 1981; Inns, 1982; Curlewis & Stone, 1985; Nogueira, 1988; Hamilton et al. 2000; Taggart et al. 2005) but this is the first report of a species that breeds continuously in the wild in which the males display distinct seasonal changes.

Leydig cells are known to produce testosterone (Christensen, 1975; Huhtaniemi, 1993; Weinbauer & Nieschlag, 1993). In this study, seasonal changes in volume fraction of Leydig cells suggested similar seasonal changes in plasma androgen concentrations. This was confirmed by androgen assays that showed decreased plasma androgen concentrations in winter in male swamp wallabies which correlated positively and significantly. Although the antiserum used in this study cross-reacts significantly with 5α-dihydrotestosterone (98%) as well as testosterone (100%), both of which are potent androgens in marsupials (Wilson et al. 2003), it is likely that the main circulating androgen in the swamp wallaby is testosterone, since in adult male tammars (Catling & Sutherland, 1980) and humans (Mantzoros et al. 1995) the main circulating androgen is testosterone and 5α-dihydrotestosterone concentrations are about 5% of the plasma testosterone concentrations in tammars (Wilson et al. 1999) and about 10% in humans (Mantzoros et al. 1995). In the tammar wallaby, increases in plasma testosterone and luteinising hormone during the breeding season were directly attributed to the presence of oestrous females during the breeding season (Catling & Sutherland, 1980). This may also be the case in the swamp wallaby because males in a population where females bred seasonally showed a marked decrease in testis size outside the breeding season but males in a continuously breeding population showed no change in testis size (Robertshaw & Harden, 1986). In our study, however, it appeared that some male reproductive characteristics of the swamp wallaby, such as testosterone secretion by the Leydig cells and related changes in plasma testosterone concentrations and prostate weights, were directly affected by season rather than the reproductive state of the females, because females in oestrus were present throughout the year (Paplinska et al. 2006). This does not invalidate the results of Catling & Sutherland (1980) because the effect of oestrous females on reproductive hormones and accessory glands outside the normal breeding season was not considered. Neither does it invalidate the results of Robertshaw & Harden (1986) because they did not examine any reproductive organs other than testes in their study.

Female swamp wallabies conceived year-round during this study (Paplinska et al. 2006) showing that some, if not all, males are capable of fertilising females at all times of the year, despite the changes we observed in the male reproductive system. Maturing spermatozoa were found in the testes of adult males throughout the year, even during winter when peripheral testosterone levels reached their nadir. Testosterone is necessary for the maintenance of spermatogenesis (Hearn, 1975; Weinbauer & Nieschlag, 1993; McLachlan et al. 1994; Zhengwei et al. 1998). However, the concentration of testosterone within the testis far exceeds circulating testosterone levels (Maddocks & Sharpe, 1989; Maddocks et al. 1993; Weinbauer & Nieschlag, 1993) and qualitatively normal spermatogenesis may proceed even if intratesticular testosterone is as low as 5% of normal concentrations (Cunningham & Huckins, 1979). Our observations suggest that intratesticular testosterone concentrations were sufficient to maintain spermatogenesis and also to enable sperm maturation to continue in the epididymis throughout the year in the swamp wallaby. The decrease in volume fraction of seminiferous tubules during winter suggests a reduction in spermatogenic activity and reduced production of spermatozoa by the testes but the rate of spermatogenesis was still sufficient to maintain fertility. In the human male, whilst sperm counts of 40–250 million spermatozoa mL⁻¹ of ejaculate are considered normal, some men with sperm counts as low as 5 million spermatozoa mL⁻¹ of ejaculate are fertile (Anonymous, 1990) and pregnancies have been reported in men with sperm counts less than 1 million spermatozoa mL⁻¹, providing that sperm morphology and motility are normal (Harper, 1988).

Testosterone is essential for the maintenance of male sexual behaviour (Goy & McEwen, 1977; Rudd, 1994). In male swamp wallabies, the circulating level of plasma testosterone needed to stimulate mating behaviour is not known but the lower winter concentrations of plasma testosterone must have been sufficient, in some males at least, to stimulate mating and other normal sexual behaviour despite the fact that the maximum plasma testosterone concentrations in winter were approximately one third of the concentrations in summer and spring.

Causes of the seasonal decline in androgen secretion and prostate gland mass in the swamp wallaby are not clear. Steroidogenesis is sensitive to food deprivation (Martin et al. 1994; Blache et al. 2000). Browse is readily available all year at the field site; however, nightly temperatures in the area regularly fall below 4°C during winter. Browse is a low quality food source and therefore the swamp wallaby may be unable to compensate for the increased costs of thermoregulation in winter, resulting in, amongst other physiological changes, a decrease in steroid synthesis. Using this argument one may also expect reproduction in females also to be influenced negatively during winter but our observations do not support this (Paplinska et al. 2006), although steroidogenesis in male swamp wallabies may be more sensitive to food restriction than in female swamp wallabies. Seasonal cycles of testis weight in musky rat kangaroos (Dennis, 2002) and prostate weight, 5′-dihydrotestosterone content, and metabolic and secretory activity of the prostate in the brushtail possum (Curlewis & Stone, 1985) persist in captivity when animals are fed ad libitum although in these animals the effects of oestrous females on male reproduction cannot be discounted. A future study is now needed to determine if seasonal cycles of testosterone secretion and accessory gland weight persist in the swamp wallaby in captivity when housed with oestrous females with readily available high quality food.

Photoperiod is one of the most important factors affecting reproduction in a wide variety of male and female mammals (Breed & Clarke, 1970; Lincoln, 1971b; Elliott, 1976; Lincoln & Short, 1980; Tyndale-Biscoe & Renfree, 1987; Turek & Van Cauter, 1988). In the swamp wallaby, the seasonal cycle of androgen secretion may be under photoperiodic control. Steroidogenesis, like any other physiological process, is energetically expensive; however, it is not necessary for the maintenance of life. It is therefore possible to downregulate steroid synthesis and leave more energy available for processes which are required for the maintenance of life, such as thermoregulation. A photoperiodically controlled, annual downregulation of steroid synthesis during periods when thermoregulatory costs are increased may be of advantage in areas where temperatures vary in a predictable manner. Whether pinealectomised or melatonin-implanted swamp wallaby males still show seasonal cycles of testosterone secretion requires investigation.

Conclusion

This study examined seasonal changes in the reproductive tract of swamp wallaby males and described the changes occurring during puberty in the testes and epididymides. Luminal development in the epididymis was followed by luminal development in the testis. The completion of spermatogenesis followed luminal development in the seminiferous cords. More proximal regions of the ductus epididymis developed earlier than more distal regions, possibly due to the actions of androgens in testicular secretions flowing through the duct prior to sperm release. This study has demonstrated a seasonal cycle of testosterone production and accessory gland weight in a continuously breeding population of the swamp wallaby. The factors which control this seasonal cyclicity need to be further investigated because this is the first report of a population in which seasonal changes in male reproductive physiology are independent of the reproductive state of females. This population, therefore, provides a good model system with which to investigate intrinsic seasonal changes in reproduction of male macropodid marsupials in the wild.