FEMALE MARSUPIALS

3.1 Physical examination: reproductive tract, external genitalia and associated structures

Female reproductive and associated anatomy is described in Vogelnest and Woods (2008) for most species.

There is considerable variation among the marsupial taxa in the morphology of the normal pouch in both form and the number of teats (Tyndale-Biscoe and Renfree 1987; Tyndale-Biscoe 2005). In species such as macro­pods, wombats and the koala, which typically have single young, the pouch is well developed, fully enclosing, and grows in size, with the developing young. In smaller spe­cies with large litters (e.g. dasyurids), the pouch either lacks defined edges or has only minor folds of skin that develop during the oestrous cycle, pregnancy and lacta­tion to provide some support to the suckling young (Tyndale-Biscoe 2005). Teat number can range from two (wombats and koala) up to 13 in the yellow-footed ante­chinus (Antechinus flavipes) (Tyndale-Biscoe 2005). In species such as the antechinuses, teat number variation (6-13) is normal and is associated with geographical loca­tion, suggesting a link between litter size, altitude, adult survival and environmental conditions (Cockburn et al.

The opening of the urogenital sinus within the common vestibule somewhat restricts examination of the distal reproductive tract. Given the small size of the marsupial neonate at birth and subsequent development within the pouch, the reproductive tract is consequently small, much of it located within the pelvic canal. As a result, internal reproductive structures are difficult to manipulate, observe directly or palpate for reproductive assessment, restricting the primary methods of assessment in larger sized marsu­pials to laparoscopy or ultrasonography.

At any one point in time, follicles in varying stages of development and/or CL may occur on one or both ova­ries (Tyndale-Biscoe and Renfree 1987). In some species, such as wombats and the koala, the ovary is covered with a coelomic bursa, which may obscure direct visualisa­tion of structures on the ovarian surface, even by lapa­roscopy (Hogan et al. 2013; Johnston and Holt 2014; Pagliarani et al. 2023). A uterotubal junction separates the oviduct from the uterus (Tyndale-Biscoe and Ren- free 1987). The paired marsupial uteri vary considerably in size and form between species; for example, some have distinct ellipsoidal uteri with long uterine necks extending from the bilateral cervices, while others have well-defined fusiform bodies, with caudal extremities lying close together (Tyndale-Biscoe and Renfree 1987; Vogelnest and Woods 2008). There is also significant variation in the structure of the vaginal complex, with varying degrees of partitioning and extension of the vaginal cul-de-sac and lateral vaginae, respectively. The two lateral vaginae are separated by a mass of fibrous connective tissue called the urogenital strand. Prior to each parturition, a pseudovaginal canal forms within the urogenital strand, connecting the vaginal cul-de-sac cranially with the urogenital sinus caudally to permit the passage of the fetus. In most species of macropods, and the honey possum, the pseudovaginal canal becomes epithelialised after the first parturition and remains patent as a permanent median vagina.

Ultrasonography has been used to evaluate the koala urogenital tract to detect structural disease associated with chlamydiosis (Marschner et al. 2014). Ultrasonogra­phy may also be useful to detect developmental reproduc­tive tract abnormalities, including the absence of a contralateral uterus and cervix (confirmed in a Tasmanian devil at necropsy; T Keeley and R Hughes, unpublished) and intersex variations. Ultrasound exami­nation of the female reproductive tract can be readily achieved by placing the transducer in the pouch (unless PY are present). This negates the need for clipping fur, which is particularly important if a birth is expected.

3.2 Pregnancy

Gestation length varies among the marsupial species from 10.5 d in the stripe-faced dunnart (Sminthopsis macroura) to ~45 d in the dibbler (Table 5.1) (Selwood and Woolley 1991; Mills et al. 2012). Embryonic and fetal development, placentation and maintenance of pregnancy in marsupials differs significantly to that of eutherians (Johnston and Keeley 2015). Most notably, as placentation is relatively short in marsupials, pregnancy is predominately maintained by the secretion of proges­terone from the CL and the implantation stage is brief.

In marsupial species that have embryonic diapause, it is characterised by either a period of suppressed develop­ment or a variable period of stasis in which the embryo stops developing at the early blastocyst stage for an extended period of time (Renfree and Shaw 2000). Embryonic diapause is usually facultative such that if the female has no current PY, a gestation of normal duration occurs. However, if the female is lactating or affected by environmental or seasonal parameters, the embryo(s) resulting from a pre- or postpartum oestrus undergoes a period of diapause (Renfree and Shaw 2000).

Ultrasonography has been used in macropods and koalas to detect pregnancy, but only in the latter stages of gestation (at expanded vesicle stage; ~14 d before birth) when the mating date is known (Drews et al. 2013).

3.3 Detection of oestrus

In group-housed, mixed-sex marsupials, oestrus detection is rarely required because the males readily detect oestrus in females.

3.3.1 Postpartum oestrus

In macropod species with embryonic diapause, most are facultatively polyoestrous and monovular with a postpar­tum oestrus. The ability for one oestrous cycle to overlap the next and facilitate a postpartum oestrus and embry­onic diapause during lactation in these species is in part due to the functionally independent nature of the ovaries and paired uteri. Essentially, the production of progester­one by the CL in one ovary does not inhibit follicular development in the other. If birth is detected, this can be used to time male-female introductions for mating at the postpartum oestrus.

3.3.2 Urogenital and urinary cytology

For the small dasyurid species commonly housed in managed care for research or breeding programs (e.g. fat­tailed dunnart (Sminthopsis crassicaudata), stripe-faced dunnart, brown antechinus (Antechinus stuartii) and dibbler), urogenital or urinary cytology is commonly used to confirm oestrus to determine the optimal timing of male introductions (Selwood 1980; Woolley 1990; Selwood and Woolley 1991; Millis et al. 1999; Lambert and Mills 2006). Before the onset of the breeding season, urine sampling or vaginal/urogenital swabbing, body­weight measurement and examination of the pouch occurs on a weekly basis (Woolley 1990).

3.3.3 Changes in activity

Changes in non-reproductive behaviours (e.g. activity, play behaviour) associated with oestrus have not been reported for most species. There are anecdotal reports that in some species an increase in activity is associated with oestrus, likely related to the desire to find a mate. This has been confirmed in the Julia Creek dunnart (S. douglasi), a solitary animal in the wild, which displays an increase in wheel-running behaviour during oestrus in

Table 5.1. Reproductive biology of Australian marsupial species, including information on oestrous cycle length, gestation length, number of cycles per year, unique reproductive strategies and features and methods Ofevaluating reproductive physiology

- Assessment and management of reproduction in Australian monotremes and marsupials

Table 5.1. (continued)

¹Merchant 1979; ²PooIe 1974; ³PooIe 1975; ⁴Smith 1992; ⁵McCracken 1990; ⁶BaIIantyne etal. 2009; ⁷Lyne 1976; ⁸GemmeII etal. 1980; ⁹GemmeII 1981; ¹⁰PiIton and Sharman 1962; ¹¹CurIewis etal. 1985; ¹²Oateseta/. 2007; ¹³Bradshaweta/. 2004; ¹⁴Bradshawand Bradshaw 2012; ¹⁵WooIIey 1990; ¹⁶SeIwood and Woolley 1991; ¹⁷Menkhorst etα∕. 2009; ¹⁸Smith and Godfrey 1970; ¹⁹Godfreyand Crowcroft 1971; ²⁰FIetcher 1985; ²¹Hinds 1989; ²²SeIwood 1980; ²³SeIwood 1987; ²⁴Hinds and Selwood 1990; ²⁵MiIIs and Bencini 2000; ²⁶MiIIs etα∕. 2012; ²⁷KeeIey eta∕. 2012c; ²⁸MiIIis eta∕. 1999; ²⁹Hogan eta∕. 2012; ³⁰Johnston eta∕. 2000; ³¹Kusuda eta∕. 2009; ³²Paris eta∕. 2002; ³³FinIayson eta∕. 2006; ³⁴Hogan eta∕. 2010b; ³⁵Swinbourne et al. 2017; ³⁶West eta∕. 2004; and ³⁷North and Harder 2008

*Not an Australian species.

CurrentTherapyin MedicineofAustraIian Mammals

managed care (Pollock et al. 2010), but not for the fat­tailed dunnart (Swinston et al. 2015).

3.3.4 Changes in bodyweight and appetite

A transient increase in bodyweight at oestrus (concurrent with epithelial cells present in the urine) and a secondary increase in bodyweight immediately before parturition have been seen in several dunnart species and in the red­tailed phascogale (Phascogale calura) (Woolley and Valente 1986; Woolley 1990; Woolley 2007; Foster et al. 2008). Although these changes have been used in the laboratory setting to detect oestrus for pairing, it requires daily han­dling and may not be accurate if used as the only method of oestrus detection. Changes in bodyweight have not been confirmed, but food consumption declines or ceases in Tasmanian devils during oestrus and immediately before and during parturition; therefore, if the first decrease in appetite goes undetected (e.g. because of food caching or coincidental starve days), the second decrease may be mis­taken as behavioural oestrus (Keeley et al. 2012c).

3.3.5 Overt behavioural signs

With the exception of the koala, which is currently the only marsupial known to be a coitally and/or semen- induced ovulator rather than spontaneous, behavioural signs of oestrus are either subtle or have yet to be docu­mented (Johnston et al. 2000; Feige et al. 2007). Signs of oestrus in the koala include bellowing, restlessness, jerk­ing movements, urination and mounting by a ‘teaser’ male or another oestrous female (Smith 1980; Johnston et al. 2000; Feige et al. 2007). The frequency and pre­dominant behavioural signs of oestrus can vary between individuals. Accurate documentation of any changes in behaviour for an individual over several cycles is impor­tant to confirm normal cycling and to establish what are the most useful behavioural signs of oestrus for each individual female.

3.3.6 Pouch changes

In some dasyurid species, development and stretching of the proximal edges of the pouch and a change in pouch secretions causing reddening of the pouch area have been observed leading up to oestrus (Godfrey and Crowcroft 1971; Woolley and Gilfillan 1991; Hesterman et al. 2008). For example, in the fat-tailed dunnart, the pouch enlarges and reddens with the approach of oestrus, a feature that is more marked in nulliparous than in parous females (Godfrey and Crowcroft 1971). Prior to birth, the pouch becomes more granular in appearance, with clear secre­tions and further enlargement of the pouch cavity (God­frey and Crowcroft 1971; Woolley and Gilfillan 1991; Lambert and Mills 2006; Hesterman et al. 2008). Although changes in pouch morphology and secretions have been observed in cycling female wombats (Finlayson et al. 2006), the association between pouch changes and oestrus has not been well documented in species other than dasyurids.

3.4 Causes of reduced fecundity and weaning success in female marsupials

Once general health, age-related, physical, behavioural, incompatibility and oestrus detection-related causes of infertility have been ruled out, it becomes progressively more difficult to elucidate the cause of reproductive fail­ure. Other than chlamydiosis in koalas, reproductive dis­eases are uncommon or yet to be described. The primary causes of female infertility or lack of reproductive success may be associated with external factors, such as: tempera­ture; day length (artificial or natural); environment (enclosure dimensions, design, complexity, den number and structure, overcrowding); proximity to human activ­ity or conspecifics or other species (including pest spe­cies); nutrition (type, diversity, frequency and timing of feeding); timing and method of mate introductions; mate choice and compatibility; inbreeding; and social struc­ture. Given the broad diversity in reproductive parame­ters and strategies among marsupials (Table 5.1), identifying perturbations in female reproduction can be extremely challenging (Johnston and Keeley 2015).

In marsupials in managed care, it is common for a small number of individuals to never breed. It is difficult to identify a specific reason for this; however, many of the factors listed above may play a role. For example, non­breeding stripe-faced dunnarts, either in permanent anoestrus or cycling regularly without conceiving, are common in low numbers in laboratory colonies. In some cases, when transferred from small cages to large pens they commence breeding (Godfrey and Crowcroft 1971). With successful breeding of stripe-faced dunnarts, epi­thelial cells are found in the urine for 2-9 d per cycle (mean 5.2 ± 0.8 d); however, second-generation females commonly have oestrous cycles that are typically charac­terised by persistent cornification (average 11-20 d), sug­gesting that the physiology of females born in managed care may be altered, potentially by husbandry or environ­mental factors (Godfrey 1969). This could affect the longevity of the managed care population, as reproduc­tive output may decline with each successive generation.

3.4.1 Cannibalism of pouch young

Cannibalism of PY is seen in dasyurids held in managed care whose physical and psychological needs are not being met. Environmental and behavioural enrichment (e.g. pro­vision of live invertebrates, branches/logs for climbing and exploration) are particularly important stimuli for dasyu- rids, to ensure wellbeing and reproductive success. Noise and other disturbances and inadequate environments are known contributors towards PY cannibalism in dasyurids (Aslin 1980; Lambert and Mills 2006). Females born in managed care appear to be more prone to stress-related cannibalism than their wild-caught counterparts (Aslin 1980; Lambert and Mills 2006). Cannibalism was elimi­nated in dibblers born in managed care by reducing human traffic around enclosures and adopting a ‘hands-off’ approach during gestation and lactation, adding visual bar­riers between animals, increasing the live invertebrate prey and doubling the enclosure size (Lambert and Mills 2006).

3.4.2 Pouch eviction and neonatal loss

The incidence of pouch eviction or neonatal loss among species is unknown, but has been reported in both free- ranging and marsupials held in managed care. Free- ranging macropod females may evict PY in extreme environmental conditions or after trauma such as a vehi­cle collision. In managed care, PY eviction or neonatal loss can occur for a variety of reasons and may go unde­tected if it occurs shortly after birth. Causes include pouch infection, mastitis, pouch trauma, disturbances, capture and physical restraint or stress (Johnson-Delaney and Lennox 2017).

Pouch checking is a common tool used to confirm successful breeding of marsupials in managed care. The optimal timing for this varies and must be weighed up against the potential risks and benefits. If pouch checks occur at the estimated time of parturition, they may dis­rupt the journey of the joey from the common vestibule to the pouch. Delaying pouch checks for 1-2 wk after expected parturition may reduce the risk of PY loss, but also risks missing early loss and thus confirmation of pregnancy. Pouch checking can be conducted under physical restraint in tractable animals by experienced personnel. The use of a flexible inspection camera (e.g. endoscope) may facilitate pouch inspection. In koalas that fail to ovulate, the female will typically return to oestrus either around or slightly before the expected time of parturition, confirming non-pregnancy. A return to oestrus ~2 wk after expected parturition will be observed where ovulation occurred but conception did not, allowing for a less invasive approach in koalas by testing the behavioural response to a male for signs of oestrus rather than a pouch check (Johnston et al. 2000).

In a study where stripe-faced dunnarts were handled daily for urine collection (for epithelial cytology to deter­mine oestrus) and to confirm birth, PY loss was high and occurred predominately at birth (Godfrey 1969). In other studies, where daily handling of females ceased at the confirmation of mating and was reduced (not daily) when confirming the presence of PY, much lower levels of litter loss were reported (Godfrey and Crowcroft 1971; Woolley and Valente 1986). Loss of PY in free-ranging northern brown bandicoots (Isoodon macrourus) is not uncommon after capture and handling or in situations of environ­mental stress (FitzGibbon 2015).

3.4.3 Pouch infection and mastitis

Prior to breeding, examination of the pouch for signs of infection or mastitis is recommended (Johnson-Delaney and Lennox 2017). Microbial culture is necessary to establish a diagnosis and to direct treatment. The causes of pouch infections and mastitis are unclear, however, clean environments and ensuring hands are washed before pouch checks are important preventative meas­ures. Stress, poor nutrition and husbandry, and other environmental factors may also contribute to pouch infections. In koalas, pouch infection has been suspected as a cause of PY loss during the first few months of pouch life (Osawa et al. 1992). Although rare, mastitis has been documented in various marsupial species (Johnson- Delaney and Lennox 2017). Recently, Maidment et al. (2023) have identified changes in the microbiota of the koala pouch that appear to be related to pouch young loss.

3.4.4 Effects of nutrition on reproduction

Good nutrition is crucial for successful reproduction (see Chapter 14). Inadequate nutrition was suspected as a con­tributor to poor reproductive success in a managed care population of fat-tailed dunnarts. Supplementation with iodine and vitamin E improved the production of PY (God­frey and Crowcroft 1971). Both low body condition and obesity affect reproductive potential and performance. Although there are few published studies that have exam­ined the relationship between nutrition and reproduction in marsupials, in several species housed in managed care a change in diet during the breeding season is known to improve reproductive success. For example, feeding a pre­dominately termite-based diet to female numbats during the breeding season improves reproductive success and the weaning of young (Power and Monaghan 2003), and increasing the sugar content of nectar mixes for gliders sea­sonally promotes breeding (M Shaw pers. comm.). Growth, the onset of puberty and reproductive performance are also nutrient-dependent (Loudon 1987). For example, Tasma­nian devil precocial breeders have rapid growth and high bodyweights for 1-yr-old females (~6 kg) (Lachish et al. 2009) and females with low bodyweight (salpingitis, metritis and ovarian bursal cysts^5-8)

• Unilateral and/or bilateral cryptorchidism and/or varying degrees of aplasia⁴

• Scrotal trauma (possibility associated with fighting)⁹

• Scrotal bifurcation⁹ Southern hairy-nosed wombat (Lasiorhinus latifrons) • Penile fistula associated with pyuria and bacteria and reluctance to mate¹⁰

• Penile urethral obstruction associated with a dried, partially ejaculated copulatory plug⁹ Spotted-tailed quoll (Dasyurus maculatus) • Preputial and penile prolapse and penile trauma resulting in injury to the corpus cavernosum¹¹ Dasyurids • Cytomegalovirus-related prostatitis¹

• Mammary adenocarcinoma¹²