Abstract

Nearly half of the species in the family Felidae are included on the IUCN Red List of Threatened Species.⁵ As wild populations dwindle, cats housed in zoos are becoming ever more critical for conservation by serving as assurance populations to protect against extinction. Due to a combination of poor breeding success, low founder numbers, and small population sizes, the long-term sustainability of most felid populations in North American zoos cannot be achieved with natural breeding alone. The development of assisted reproductive technologies (ART) may be critical to ensure their future viability.⁷ Non-invasive fecal hormone metabolite monitoring has provided these breeding programs with a pregnancy diagnostic tool; however, for many felid species, increase in progesterone during a non-pregnant luteal phase is indistinguishable from that of pregnancy, and non-pregnant luteal phases can last up to two-thirds the duration of gestation, depending on the species. For example, pregnancy lasts 66–77 days in the Pallas’ cat (Otocolobus manul), while their non-pregnant luteal phase is ∼50 days, restricting accurate pregnancy diagnosis to just 20 days before parturition.¹ This delay in pregnancy diagnosis to the latter part of gestation creates uncertainty in planning for parturition and neonatal care. Moreover, fecal progesterone monitoring cannot distinguish between post-implantation pregnancy loss and conception failure, which is relevant for assessing both natural breeding and AI success. Because of these challenges, recent research has focused on investigating alternative pregnancy markers to improve the efficiency and accuracy of non-invasive pregnancy diagnosis. One promising pregnancy marker is 13,14-dihydro-15-keto-prostaglandin F2 alpha metabolite (PGFM), a stable fecal metabolite generated during corpora lutea lysis.⁶ A pronounced spike in PGFM occurring only in pregnant individuals during the last trimester has been observed reliably in 10 of the 14 Felidae genera (Acinonyx, Caracal, Catopuma, Felis, Leopardus, Leptailurus, Lynx, Neofelis, Prionailurus, Puma), with variation in the presence of a characteristic peak in one genus (Panthera).^2,4 In this study, our objectives were to 1) characterize fecal PGFM concentrations during the non-pregnant luteal phase and pregnancy in Pallas’ cats, one of the last uncharacterized genera, Otocolobus, and 2) further investigate fecal PGFM levels in pregnant and luteal-phase sand cats (Felis margarita), using a commercially available PGFM enzyme linked immunoassay (PGFM EIA Kit, Arbor Assays®, Ann Arbor, MI, USA). A previous study reported initial characterization of PGFM for one pregnant and one non-pregnant luteal phase sand cat, finding definitive differences in PGFM levels.² In our study, fecal PGFM concentrations were assessed in pregnant Pallas’ cats (natural mating, n=3; AI, n=2) from samples collected beginning 85 days before parturition until 0–1 days after parturition. For non-pregnant Pallas’ cats (natural mating, n=3; AI, n=2), samples were assayed in each female for 79–85 days, beginning 9–14 days before ovulation and initial progesterone increase. In sand cats (natural breeding, pregnant, n=3; AI, pregnant, n=1; natural breeding, non-pregnant, n=2; AI, non-pregnant, n=2), a similar fecal sampling schedule was used to assess PGFM concentrations. Data are presented as means±standard error (SE), and mean values were compared using Student’s t-test. Basal PGFM values were calculated using a two standard deviation (SD) iterative process; an elevation of PGFM, in three consecutive samples, that was greater than the baseline value plus two SDs was considered a substantive increase.³ In pregnant Pallas’ cats, fecal PGFM remained basal (1526.4±260.3 ng/g) until 29–35 days before birth. PGFM concentrations subsequently exhibited an initial increase (5439.2±1259.4 ng/g) above (p=0.02) baseline levels, and continued increasing steadily, with maximum values (132068.3±38,105.7 ng/g) observed 0–9 days before parturition. In contrast, PGFM levels in the five non-pregnant Pallas’ cats remained near baseline (1250.5±165.0 ng/g) throughout their luteal phase, with no increase (p=0.99) in values (1247.8±305.8 ng/g) observed in the 39–43 day time period after initial progesterone elevation (i.e., corresponding to the time frame observed for the initial PGFM rise in pregnant Pallas’ cats). Notably, one Pallas’ cat that was artificially inseminated and categorized as non-pregnant showed a sharp increase in PGFM (39619.7 ng/g) at 51 days post-AI; values remained elevated for 4 days before steadily declining to baseline 67 days post-AI. These findings suggest that this female did conceive but lost the pregnancy several weeks before full-term gestation. In pregnant sand cats, the timing of the initial PGFM increase (2–27 days before parturition) and associated PGFM levels (8415.8±3745.8 ng/g) were highly variable, and did not differ (p=0.10) from baseline values (1888.8±950.8). Following the initial PGFM rise, maximum values (30571.0±8471.5 ng/g) that were higher (p=0.04) than baseline occurred 2–9 days before parturition. PGFM concentrations (520.0±62.7 ng/g) in the four non-pregnant sand cats remained (p=0.78) near baseline (552.9±122.7 ng/g) throughout the sampling period. In conclusion, our findings have shown that PGFM analysis is a reliable diagnostic tool for pregnancy in Pallas’ cats, and confirm its utility for pregnancy diagnosis in sand cats, albeit much closer to the time of parturition. In Pallas’ cats, the initial increase in PGFM occurs ∼1 mo before birth (or 9–15 days earlier than the progesterone decline in a non-pregnant luteal phase) and remains elevated until parturition. In sand cats, the timing of this initial diagnostic PGFM increase is much more variable, occurring from 2 to 27 days before birth. In both species, these PGFM increases appear to be highly specific to pregnancy since similar elevations do not occur during non-pregnant luteal phases. This diagnostic tool is valuable to curators and population managers by allowing earlier confirmation of pregnancy, as well as identification of females that possibly conceive but fail to carry offspring to term.

Acknowledgments

We thank the keeper, curatorial, and veterinary staff at each institution for their assistance in collecting fecal samples and Dr. Jason Herrick for kindly supplying many of the sand cat samples.

Literature Cited

1.  Brown JL. Female reproductive cycles of wild female felids. Anim Reprod Sci. 2011;124(3–4):155–162.

2.  Dehnhard M, Finkenwirth C, Crosier A, Penfold L, Ringleb J, Jewgenow K. Using PGFM (13,14-dihydro-15-keto-prostaglandin F2α) as a non-invasive pregnancy marker for felids. Theriogenology. 2012;77(6):1088–1099.

3.  Herrick J, Bond J, Campbell M, Levens G, Moore T, Benson K, D’Agostino J, West G, Okeson D, Coke R, et al. Fecal endocrine profiles and ejaculate traits in black-footed cats (Felis nigripes) and sand cats (Felis margarita). Gen Comp Endocrinol. 2010;165(2):204–214.

4.  Lueders I, Ludwig C, Schroeder M, Mueller K, Zahmel J, Dehnhard M. Successful nonsurgical artificial insemination and hormonal monitoring in an Asiatic golden cat (Catopuma temmincki). J Zoo Wildl Med. 2014;45(2):372–379.

5.  Nowell K. Cats on the 2009 Red List of Threatened Species. Cat News. 2009;51:32–33.

6.  Schwarzenberger F, Brown JL. Hormone monitoring: an important tool for the breeding management of wildlife species. Wien Tierarztl Monatsschr. 2013;100(9–10):209–225.

7.  Swanson WF. Application of assisted reproduction for population management in felids: the potential and reality for conservation of small cats. Theriogenology. 2006;66(1):49–58.