Background: Koalas are a vulnerable marsupial species with unique reproductive traits. Efforts to develop assisted breeding technologies have been hindered by a limited understanding of maternal recognition of pregnancy and physiological changes induced by the foeto-placental unit. Differences in the reproductive physiology of pregnant and non-pregnant koalas were examined to investigate the possibility of maternal recognition and identify potential pregnancy and/or embryonic loss biomarkers. Methods: Koalas were separated into three groups: pregnant (n = 4 cycles from three females), mated but non-parturient (n = 4 cycles from three females), and gonadotropin-releasing hormone (GnRH) agonist-treated females (n = 7). Plasma was collected on day of mating/GnRH injection (D0) and on multiple subsequent days. Progesterone concentrations were measured by enzyme immunoassay, and plasma proteomes were analysed using filter-aided sample preparation followed by liquid chromatography-tandem mass spectrometry (LC-MS/MS), employing sequential window acquisition of all theoretical fragment ion spectra. Results: Ovulation induction mechanisms influenced peri-ovulatory progesterone secretion (pregnant 40.7 ± 3.3 ng/mL vs. GnRH-treated 15.7 ± 1.8 ng/mL), with no significant differences in progesterone that occurred later in the luteal phase. LC-MS/MS identified 158 proteins, representing the first koala plasma proteome. Leucine-rich alpha-2-glycoprotein (LRG1) was significantly elevated at D2 in pregnant females compared to GnRH-treated females, and pregnant D9 and D19. In pregnant females, fibronectin (FN1) was significantly more abundant at D19 compared to D9 but not significantly different between treatments. Conclusions: These preliminary findings provide foundational data for further investigation into maternal recognition and pregnancy/embryonic loss in koalas.

Background Captive koala breeding programmes are essential for long-term species management. However, breeding efficacy is frequently impacted by high neonatal mortality rates in otherwise healthy females. Loss of pouch young typically occurs during early lactation without prior complications during parturition and is often attributed to bacterial infection. While these infections are thought to originate from the maternal pouch, little is known about the microbial composition of koala pouches. As such, we characterised the koala pouch microbiome across the reproductive cycle and identified bacteria associated with mortality in a cohort of 39 captive animals housed at two facilities. Results Using 16S rRNA gene amplicon sequencing, we observed significant changes in pouch bacterial composition and diversity between reproductive time points, with the lowest diversity observed following parturition (Shannon entropy - 2.46). Of the 39 koalas initially sampled, 17 were successfully bred, after which seven animals lost pouch young (overall mortality rate - 41.18%). Compared to successful breeder pouches, which were largely dominated by Muribaculaceae (phylum - Bacteroidetes), unsuccessful breeder pouches exhibited persistent Enterobacteriaceae (phylum - Proteobacteria) dominance from early lactation until mortality occurred. We identified two species, Pluralibacter gergoviae and Klebsiella pneumoniae, which were associated with poor reproductive outcomes. In vitro antibiotic susceptibility testing identified resistance in both isolates to several antibiotics commonly used in koalas, with the former being multidrug resistant. Conclusions This study represents the first cultivation-independent characterisation of the koala pouch microbiota, and the first such investigation in marsupials associated with reproductive outcomes. Overall, our findings provide evidence that overgrowth of pathogenic organisms in the pouch during early development is associated with neonatal mortality in captive koalas. Our identification of previously unreported, multidrug resistant P. gergoviae strains linked to mortality also underscores the need for improved screening and monitoring procedures aimed at minimising neonatal mortality in future.

The aims of this study were to investigate the proteome of koala spermatozoa and that of the prostatic bodies with which they interact during ejaculation. For this purpose, spermatozoa and prostatic bodies were fractionated from the semen of four male koalas and analysed by HPLC MS/MS. This strategy identified 744 sperm and 1297 prostatic body proteins, which were subsequently attributed to 482 and 776 unique gene products, respectively. Gene ontology curation of the sperm proteome revealed an abundance of proteins mapping to the canonical sirtuin and 14-3-3 signalling pathways. By contrast, protein ubiquitination and unfolded protein response pathways dominated the equivalent analysis of proteins uniquely identified in prostatic bodies. Koala sperm proteins featured an enrichment of those mapping to the functional categories of cellular compromise/inflammatory response, whilst those of the prostatic body revealed an over-representation of molecular chaperone and stress-related proteins. Cross-species comparisons demonstrated that the koala sperm proteome displays greater conservation with that of eutherians (human; 93%) as opposed to reptile (crocodile; 39%) and avian (rooster; 27%) spermatozoa. Together, this work contributes to our overall understanding of the core sperm proteome and has identified biomarkers that may contribute to the exceptional longevity of koala spermatozoa during ex vivo storage.

Although it is well established that chlamydial disease renders female koalas infertile, there has been limited research on its effects on male koala fertility, specifically sperm quality. This study determined whether chlamydial infection adversely affects semen quality of naturally infected koalas and spermatozoa recovered from Chlamydia negative koalas co-incubated in vitro with C. pecorum elementary bodies (EBs). Semen from 102 south-east Queensland sexually mature wild koalas exhibiting varying degrees of chlamydiosis and clinical signs of disease were assessed for semen quality and compared to 11 clinically healthy, Chlamydia-free captive male koalas. For in vitro studies, semen samples were collected from 6 Chlamydia-free captive koalas, and co-incubated over 24 h with high and low concentrations of C. pecorum EBs and sperm quality assessed. Wild koalas displaying severe signs of clinical disease with C. pecorum present in the semen had significantly greater sperm DNA damage (P = 0.0267). The total % of morphologically abnormal spermatozoa was highest in wild koalas that had severe signs of clinical disease but whose semen was negative for C. pecorum (P = 0.0328). This apparent contradiction is possibly associated with wild males having resolved the infection but still possessing underlining reproductive pathology. A higher incidence of loose head spermatozoa occurred in semen of wild koalas not infected with C. pecorum compared to those that were C. pecorum infected (P = 0.026). In vitro incubation of semen with C. pecorum significantly decreased sperm motility and viability over 24 h.

Evaluation of physiological stress in the tiger (Panthera tigris) using faecal cortisol metabolite (FCM) enzyme immunoassays (EIAs) provides a powerful conservation physiology tool for the species. However, it is important to validate non-invasive endocrine sampling techniques in field conditions to ensure that the method provides a reliable parameter of physiological stress in the species. This is because endocrine measurements are highly species specific and FCM concentrations can be influenced by environmental factors. Here, we studied the impact of the decay rate of FCMs and intra-sample variation of FCMs using a previously validated EIA. To determine the decay rate of FCMs, we measured FCMs in freshly deposited tiger faeces (n = 8 tigers and 48 scats) that were randomly exposed to the natural environment (dry conditions with no rainfall) for up to 192 h. To determine intra-sample variation in FCMs, we used 10 scats from 10 tigers, divided each sample into four sections and each section into four sub-sections and measured FCMs in each section and sub-section. The results of this decay-rate experiment showed that FCMs in tiger faeces began to decay after 48 h exposure to the environmental conditions available. Thus, FCMs within freshly deposited tiger faeces are influenced by available environmental conditions. Changes in weather conditions (e.g. increased rainfall and humidity) could influence the stability of FCMs. The results of the intra-sample variation study showed that inter-variation among scats accounted for 52% of the variations in FCMs, while intra-sample variation between sections (32%) was greater than the sub-sample variation (16%). Intra-sample variation can be reduced by homogenizing the entire lyophilized faecal sample prior to the EIA. In conclusion, careful evaluation of decay rate and complete homogenization of faeces prior to EIA analysis will increase the reliability of FCMs as a non-invasive index of physiological stress in the tiger.

Glucocorticoid quantification using non-invasive methods provides a powerful tool for assessing the health and welfare of wild life in zoo-based programmes. In this study, we provide baseline data on faecal-based glucocorticoid (cortisol) monitoring of Sumatran tigers (Panthera tigris ssp. sumatrae) managed at the Melbourne Zoo in Victoria, Australia. We sampled five tigers daily for 60 days. Faecal cortisol metabolites (FCMs) in tiger faecal extracts were quantified using enzyme immunoassays that were successfully validated using parallelism and accuracy recovery checks. Two female tigers had significantly higher mean FCM levels than the two males and another female, suggesting that females may have higher FCM levels. A significant elevation was noted in the FCM levels for one female 2 days after she was darted and anaesthetized; however, the FCM levels returned to baseline levels within 3 days after the event. Comparative analysis of FCM levels of tigers sampled at Melbourne Zoo with tigers sampled earlier at two other Australian Zoos (Dreamworld Themepark and Australia Zoo) showed that FCM levels varied between zoos. Differences in the enclosure characteristics, timing of sampling, size and composition of groupings and training procedures could all contribute to this variation. Overall, we recommend the use of non-invasive sampling for the assessment of adrenocortical activity of felids managed in zoos in Australia and internationally in order to improve the welfare of these charismatic big cats.

Context Captive breeding for release back to the wild is an important component of ex situ conservation but requires genetic diversity that is representative of the wild population and has the ultimate goal of producing ecologically sustainable and resilient populations. However, defining and testing for representativeness of captive populations is difficult. Koalas (Phascolarctos cinereus) are bred for educational and tourism purposes in zoos and wildlife parks in South-East Queensland, but there are drastic declines evident in some wild koala populations in this region. Aim We compared genetic diversity at microsatellite loci and mitochondrial DNA in two captive koala populations with that of the local, wild koalas of South-East Queensland, determining the degree to which genetic diversity of neutral loci had been preserved and was represented in the captive populations. Key results The expected heterozygosity and the allelic richness was significantly greater in one captive colony than one wild South-East Queensland population. There was low but significant differentiation of the captive from wild populations using FST, with greater differentiation described by Jost’s Dest. In contrast, a newly introduced Kullback–Leibler divergence measure, which assesses similarity of allele frequencies, showed no significant divergence of colony and wild populations. The captive koalas lacked many of the mitochondrial haplotypes identified from South-East Queensland koalas and possessed seven other haplotypes. Conclusions Captive colonies of koalas have maintained levels of overall neutral genetic diversity similar to wild populations at microsatellite loci and low but significant differentiation likely resulted from drift and founder effects in small captive colonies or declining wild populations. Mitochondrial DNA suggests that captive founders were from a wider geographic source or that haplotypes have been lost locally. Implications Overall, tested captive koalas maintain sufficient microsatellite diversity to act as an in situ reservoir for neutral genetic diversity of regional populations.

Glucocorticoid quantification using non-invasive methods provides a powerful tool for assessing the health and welfare of wildlife in zoo-based programmes. In this study, we provide baseline data on faecal-based glucocorticoid (cortisol) monitoring of Sumatran tigers (Panthera tigris ssp. sumatrae) managed at the Melbourne Zoo in Victoria, Australia. We sampled five tigers daily for 60 days. Faecal cortisol metabolites (FCMs) in tiger faecal extracts were quantified using enzyme immunoassays that were successfully validated using parallelism and accuracy recovery checks. Two female tigers had significantly higher mean FCM levels than the two males and another female, suggesting that females may have higher FCM levels. A significant elevation was noted in the FCM levels for one female 2 days after she was darted and anaesthetized; however, the FCM levels returned to baseline levels within 3 days after the event. Comparative analysis of FCM levels of tigers sampled at Melbourne Zoo with tigers sampled earlier at two other Australian Zoos (Dreamworld Themepark and Australia Zoo) showed that FCM levels varied between zoos. Differences in the enclosure characteristics, timing of sampling, size and composition of groupings and training procedures could all contribute to this variation. Overall, we recommend the use of non-invasive sampling for the assessment of adrenocortical activity of felids managed in zoos in Australia and internationally in order to improve the welfare of these charismatic big cats.

Koalas (Phascolarctos cinereus) are the only extant representatives of Australia's unique marsupial family Phascolarctidae and were listed as nationally Vulnerable in 2012. Causes of mortality are diverse, although the disease chlamydiosis, dog attacks, collisions with cars, and loss of habitat represent the principal reasons for the continued species decline. Koala breeding facilities in Queensland and New South Wales, Australia have been established for conservation and tourism. Non-invasive monitoring of physiological stress is important for determining the sub-lethal effects of environmental stressors on the well-being, reproduction and survival of Koalas in Zoos and also in the wild. In this study, we developed a faecal cortisol metabolite (FCM) enzyme-immunoassay (EIA) for monitoring physiological stress in Koalas from two established Zoos in Australia and also within a free-living sub-population from Queensland. Biological validation of the FCM EIA was done using an adrenocorticotropic hormone (ACTH) challenge. We discovered excretory lag-times of FCM of 24 h in females (n = 2) and 48 h in male (n = 2) Koalas in response to the ACTH challenge. FCM levels showed an episodic and delayed peak response lasting up to 9 days post ACTH challenge. This finding should be taken into consideration when designing future experiments to study the impacts of short-term (acute) and chronic stressors on the Koalas. Laboratory validations were done using parallelism and recovery checks (extraction efficiency) of the cortisol standard against pooled Koala faecal extracts. Greater than 99% recovery, of the cortisol standard was obtained as well as a parallel displacement curve against Koala faecal extracts. FCM levels of the captive Koalas (n = 10 males and 13 females) significantly differed by sex, reproductive condition (lactating versus non-lactating Koalas) and the handling groups. Handled male Koalas had 200% higher FCM levels than their non-handled counterparts, while females were not affected by handling as long they were not undergoing lactation. There was no significant difference in FCM levels between the captive and wild Koalas (n = 9 males and 7 females). Overall, these results provide foundation knowledge on non-invasive FCM analysis in this iconic Australian marsupial. Non-invasive stress endocrinology opens up opportunities for evaluating the sub-lethal physiological effects of management activities (including caging, translocation) on the nutritional status, reproductive behaviors and disease status of captive and managed in situ Koala populations. (c) 2013 Elsevier Inc. All rights reserved.

ESR Endangered Species Research Contact the journal Facebook Twitter RSS Mailing List Subscribe to our mailing list via Mailchimp HomeLatest VolumeAbout the JournalEditorsSpecials ESR 18:279-289 (2012) - DOI: https://doi.org/10.3354/esr00454 Non-invasive evaluation of physiological stress hormone responses in a captive population of the greater bilby Macrotis lagotis Edward Narayan1,*, Jean-Marc Hero1, Nicole Evans1, Vere Nicolson2, Al Mucci2 1Environmental Futures Centre, School of Environment, Griffith University, Gold Coast campus, Queensland 4222, Australia 2Dreamworld, Coomera, 4209 Queensland, Australia *Email: e.narayan@ga.griffith.edu.au ABSTRACT: Captive breeding programs are increasingly being used as a management option for threatened mammals. The greater bilby Macrotis lagotis, for example, is a vulnerable species which is maintained in captivity at several facilities in Australia. Non-invasive evaluation of stress hormones (cortisol in mammals) via excretory metabolites can be used to monitor physiological stress responses of captive individuals. In this study, we validated an enzyme-immunoassay (EIA) to measure cortisol metabolites in fresh faecal samples of adult male and female bilbies (n = 7) held in captivity at the Dreamworld Theme Park, Queensland, Australia. The faecal cortisol EIA was validated via parallelism and the recovery of exogenous cortisol added to pooled faecal extracts (>99% recovery). Female bilbies had higher average faecal cortisol metabolite concentrations and higher day-to-day variation than male bilbies; however, there was no relationship with bilby age. Cortisol metabolites for most individuals varied widely through time, with numerous peaks and troughs in response to long-term stressors (illnesses, injury and reproductive issues) and short-term stressors, such as use in shows at Dreamworld or public displays in local schools, manual restraint and short-term veterinary procedures (e.g. general anaesthesia). Overall, the higher mean cortisol metabolite concentrations of individuals suffering long-term stress was related to a greater response to short-term stressors. This suggests an interaction between responses to short-term and long-term stressors which is perhaps due to habituation and/or facilitation of long-term stressors. Non-invasive faecal monitoring of stress hormones could provide further information on the implications of captive breeding programs and the release of animals reared in captivity. KEY WORDS: Macrotis lagotis · Captive breeding · Capture · Reintroduction · Stressor · Health Full text in pdf format PreviousCite this article as: Narayan E, Hero JM, Evans N, Nicolson V, Mucci A (2012) Non-invasive evaluation of physiological stress hormone responses in a captive population of the greater bilby Macrotis lagotis. Endang Species Res 18:279-289. https://doi.org/10.3354/esr00454 Export citation RSS - Facebook - Tweet - linkedIn Cited by Published in ESR Vol. 18, No. 3. Online publication date: October 19, 2012 Print ISSN: 1863-5407; Online ISSN: 1613-4796 Copyright © 2012 Inter-Research.

The southern hairy-nosed wombat (Lasiorhinus latifrons) appears to use scent marking, including defaecation, for social communication in the wild. This premise assumes that the receiver wombat is able to distinguish between faeces from different sources. To examine this theory, four types of faeces (male wombat, female wombat, dingo and a plastic control) were placed into the enclosures of 12 captive wombats. Behaviour, inter-individual distance and enclosure use were recorded during the period of placement, as well as the period before and the period after. When faeces were present, the wombats used concealed locations more often than other periods (mean%: pre-treatment: 71.3, treatment: 75.6, post-treatment: 72.7; P<0.05). During the same period they also reduced grazing (mean min/period: pre-treatment: 15.8, treatment: 6.9, post-treatment: 13.1; P=0.0002) and walking activity (mean min/period: pre-treatment: 85.2, treatment: 66.9, post-treatment: 78.2; P=0.01), indicating an increased perception of risk. Wombats approached the dingo faeces 5.6 times per treatment period, which was greater than for the control (3.0; P=0.004) or female wombat faeces (3.7; P=0.049). They also avoided other wombats most when male wombat faeces were present (8.3 retreats/period) compared to the control (4.5; P=0.02), or female wombat (4.3; P=0.01). There was a residual effect of increased wombat avoidance the period after presentation of dingo faeces (9.6; P≤0.05). It is concluded that the southern hairy-nosed wombat can differentiate between faeces from different species and sex of conspecifics, and that predator faeces and those from male conspecifics increase wombat avoidance behaviour either during or after presentation.

The response of animals to handling by humans has been extensively evaluated in domesticated livestock, but rarely examined in wildlife species. Twelve captive wombats (Lasiorhinus latifrons) were subjected to two treatments in a replicated design: (1) daily handling, involving 15min of tactile contact 5 d/wk for 12 wk and (2) no-handling, involving no contact apart from that received during routine husbandry. The effect of handling was assessed via overt responses to human approach and touch, a stressor, and a novel stimulus. Daily handling reduced the wombat's flight distance in response to human approach; more in the first handling replicate (−0.16±0.02m/wk) than in the second (−0.06±0.02m/wk). A behavioural reactivity score also declined faster in the first than second handling replicate. Synthetic ACTH was used to validate the measurement of faecal cortisol metabolites in L. latifrons by EIA. Faecal cortisol metabolite secretion consistently increased in reaction to a handling procedure involving forced human contact (indicating a lack of habituation) but the magnitude of this response was not reduced by regular handling. Regular handling therefore changed the human–wombat relationship by lowering reactivity to and avoidance of the human handler, but did reduce the stress response, suggesting that the wombats entered into a state of learned helplessness.

This study documented the extent of individual animal variation with respect to two proven methods of sperm cryopreservation in a captive population of 22 koalas. Semen samples were collected by electroejaculation, diluted in Tris-citrate glucose and equilibrated to 4 degrees C before being further diluted and frozen in media containing a final concentration of either 14% (v/v) glycerol or 12.5% (v/v) dimethylacetamide (DMA). There were significant differences in post-thaw survival of spermatozoa from different animals that were independent of pre-freeze semen quality. Glycerol proved to be a better cryoprotectant than DMA in terms of maintaining motility, plasma membrane integrity and high mitochondrial membrane potential; however, there was no difference between cryoprotectants with regards to their ability to prevent chromatin relaxation. While a positive correlation was observed between motility and mitochondrial membrane potential, both before and after cryopreservation, the slopes of the pre-freeze regression lines of these relationships were significantly altered following cryopreservation, suggesting that the efficiency of energy generation by the mitochondria was lowered by the freeze-thaw process. Based on a cluster analysis of the post-thaw semen viability parameters, the koalas in this study could be divided into two distinct groups; Cluster 1 had significantly higher sperm viability compared to the other cluster, regardless of the cryoprotectant used. The unpredictability of assessing post-thaw survival from pre-freeze koala semen parameters requires further investigation but is likely to be associated with variation in ejaculate composition or inherent genetic differences between animals.

The aim of the present study was to determine whether analogues of gonadotrophin-releasing hormone (GnRH) could be used to both induce an acute testosterone response and suppress anterior pituitary function in male koalas, and induce a luteal phase in female koalas. Experiment 1 characterised the steroidogenic response of male koalas to administration of 30 mu g (4.3 mu g kg(-1)) natural-sequence GnRH. Intra-muscular injection of natural-sequence GnRH induced the release of LH and testosterone with peak concentrations at 30 min (3.7 +/- 1.9 ng mL(-1)) and 2 h (5.4 +/- 0.5 ng mL(-1)), respectively. In Experiment 2, a single injection of the GnRH antagonist acyline (100 mu g (14.3 mu g kg(-1)) or 500 mu g (71.4 mu g kg(-1))) did not influence the testosterone response to subsequent injections of natural-sequence GnRH. In Experiment 3, 4 mu g (similar to 0.67 mu g kg(-1)) of the GnRH agonist buserelin induced a luteal phase in five female koalas based on a LH surge, secretion of progestogen, and a normal-length oestrous cycle. The findings have shown that (1) natural-sequence GnRH can be used to test gonadotroph cell function and determine the testosterone-secreting capacity of male koalas, (2) the GnRH antagonist, acyline, at the dose rates used, does not suppress the pituitary-testis axis in male koalas, and (3) the GnRH agonist, buserelin, induces a normal luteal phase in female koalas.

Swelling of koala sperm chromatin following cryopreservation has largely been attributed to the absence of intermolecular disulfide cross-linkages in the marsupial sperm nucleus. Fish spermatozoa also lack disulfide bonds within their chromatin, but have been successfully cryopreserved. The present study examined the hypothesis that the cryoprotectants used for fish sperm cryopreservation would confer a similar degree of protection on koala spermatozoa. Three concentrations each of five cryoprotectants (dimethyl sulfoxide, methanol, propylene glycol, ethylene glycol and dimethylacetamide (DMA)) were evaluated. Each treatment was compared against an established koala sperm cryopreservation protocol that uses 14% glycerol. Post-thaw assessment of progressive motility, plasma membrane integrity and mitochondrial membrane potential (MMP) revealed that protocols using 15% DMA achieved 62.2 +/- 3.6% (P < 0.05) sperm survival, of which 79% (P < 0.05) had high MMP, an improvement of 32% and 40%, respectively, over sperm frozen in 14% glycerol. The percentage of spermatozoa with swollen nuclei was also lowest when frozen in 15% DMA, both immediately after thawing (18.0 +/- 3.5%; P < 0.05) and after 2 h incubation at 35 degrees C (35.8 +/- 4.4%; P < 0.05). A second study was conducted to determine the optimal concentration of DMA for use in the cryopreservation of koala spermatozoa. High DMA concentrations (17.5% and 20%) resulted in significantly lower proportions of live spermatozoa showing high MMP immediately after thawing compared with spermatozoa frozen in the lower concentrations. The percentage of koala spermatozoa with swollen chromatin following cryopreservation was not affected by DMA concentration.

Koala sperm mitochondria were examined by cryomicroscopy using the fluorescent probe JC-1, which distinguishes high (red) and low (green) mitochondrial membrane potential (MMP). At normal body temperature, ~70% of live and untreated spermatozoa exhibited high MMP whereas <3% of live untreated spermatozoa exhibited low potential. A third class, in which single midpieces contained mixed mitochondrial populations, was also detected. Heterogeneity was noted in the level of MMP between individual koalas, individual spermatozoa and even between mitochondrial gyres within single midpieces. MMP of the live sperm population was not significantly affected by glycerol but was suppressed by freezing and thawing treatments. After thawing, MMP declined significantly during rewarming, especially as the temperature increased from 5 to 35°C. The distribution of the ganglioside GM1 was examined using fluorescent-labelled cholera toxin B. In fresh, untreated koala spermatozoa GM1 was detected on the head and midpiece, but not on the principal piece. No significant redistribution of GM1 was observed after chilling and cryotreatment. Phosphatidylserine translocation across the plasma membrane was examined using fluorescent-labelled annexin V. Few fresh spermatozoa exhibited phosphatidylserine translocation (~1%); this was not increased by chilling or cryopreservation, thus implying that cryotreatment had little effect on plasma membrane lipid asymmetry.

Artificial insemination in the koala using chilled, electroejaculated semen provides for a marked improvement in the reproductive and genetic management of captive koala colonies in Australia and internationally, and makes available the option of using semen collected from wild populations to expand restricted gene pools. Dilution of koala semen for artificial insemination is complicated because koalas are induced ovulators, and it is thought that ovulating factors are present in the semen, so that semen extension for preservation purposes might be anticipated to result in a failure to induce ovulation. The first two experiments of this study were designed to determine whether artificial insemination using undiluted, extended, and extended-chilled semen collected by electroejaculation was capable of inducing a luteal phase and/or the production of pouch young. In Experiment 1, 1 ml undiluted electroejaculated semen, 2 ml diluted (1:1) semen, and 1 ml diluted (1:1) semen resulted in seven of nine, six of nine, and six of nine koalas showing a luteal phase, respectively; four pouch young were produced in each treatment. A second artificial insemination experiment was conducted in which 2 ml diluted (1:1) semen was administered in three groups of nine koalas. The first group received semen that had been collected and diluted immediately without chilling, the second group was deposited with semen stored chilled for 24 h, and the final group received semen that had been chilled for 72 h. In the first group, five females had a luteal phase, but none became pregnant. In group 2, two of the five females that had a luteal phase gave birth, whereas in group 3, four of the six females that had a luteal phase produced pouch young. In addition, experiment 3 was conducted to determine whether it was possible to produce pouch young by naturally mating koalas that were in the latter stages of their behavioral estrus; this information is important to the logistics of transporting koala semen for artificial insemination by establishing the maximum time frame in which females might be expected to shed a fertile oocyte. Of the 12 females mated on Day 8 of estrus, 6 gave birth, whereas only 3 of the 10 females naturally mated on Day 10 of estrus produced pouch young. The majority of females (21 of 22) in experiment 3 showed evidence of a luteal phase. Together, these experiments have shown that it is possible to use undiluted, extended, or extended-chilled semen to produce koala offspring up to Day 8 of estrus at conception rates similar to those achieved following natural mating. These findings represent a significant advancement in the use of reproductive technology in marsupials and provide the basis for the shipment of koala semen over long distances. The pouch young produced in this study represent the first marsupials born following artificial insemination of extended-chilled semen and bring the total number of koalas produced by artificial insemination to 31.