Abstract

The pygmy hippopotamus (Choeropsis liberiensis) is endangered in the wild and we have limited information about its reproductive biology and well-being under managed care. We therefore developed non-invasive methods for characterizing reproductive hormone profiles and adrenocortical activity in the pygmy hippo using enzyme immunoassays (EIAs). Our objectives were to 1) identify a biologically relevant EIA for measuring metabolites of androgens, glucocorticoids (cortisol), estrogens, and progestogens; 2) characterize androgen profiles in juvenile and adult males; 3) characterize the reproductive cycle and pregnancy in females.

Fecal samples were collected twice weekly for 1–3 years from 37 female and 12 male pygmy hippos housed at 27 zoological institutions. Fecal hormone metabolites were assayed in three separate laboratories using separate EIAs and established methanol- or ethanol-based extraction methods. Results were compiled for analysis.

Hormone metabolite concentrations generated by a corticosterone assay^a were significantly correlated with testosterone^b metabolite concentrations for both males and females, so the corticosterone assay could not be used to assess glucocorticoid metabolites in this species. However, a group-specific EIA exhibiting cross-reactivity with 11,17-dioxoandrostane (DOA) metabolites of cortisol clearly reflected adrenocortical activity (response to ACTH^c challenge) in both males and females. The testosterone metabolite assay also produced biologically coherent data: adult males exhibited the highest androgen metabolite concentrations, followed by adult females and juvenile males, and proven breeding males had higher concentrations than unproven males. There were significant differences in mean concentrations among seasons for adult males overall, with higher values in spring and summer than in fall and winter. Adult males housed outdoors year-round in subtropical climates exhibited higher mean androgen concentrations than males in temperate climates that were housed indoors year-round or in colder weather.

Data from three progestogen metabolite EIAs^d and three estrogen metabolite EIAs^e accurately reflected reproductive events. Average estrous cycle length was 31.8±7.4 days based on estrogen metabolite peak concentrations and 30.9±7.3 days based on nadir to nadir progestogen metabolite concentrations. Cyclical patterns were detected throughout the year, indicating a lack of seasonality. Peaks in estrogen metabolite concentrations observed during pregnancy and lactation suggested follicles may develop during both reproductive states. Pregnancy was most reliably demonstrated by an average 3- to 5-fold elevation in progestogen metabolite concentrations in the second half of gestation. Average gestation length was 203±4 days for 15 pregnancies based on breeding to calving date.

Our results provide valuable data for reproductive management and serve as a baseline for future studies monitoring gonadal and adrenocortical activity in both sexes. Additionally, we demonstrate the importance of biological validation of EIAs used for non-invasive assessment of hormone metabolites and reveal the potential for cross-reactivity between assays to confound results.

Endnotes

a. CJM006, Coralie Munro, University of California, Davis, CA, USA
b. C196, Arbor Assays, Ann Arbor, MI, USA
c. Adrenocorticotropic hormone; Cosyntropin 0.25 IU/mL, Sandoz Inc., Princeton, NJ, USA
d. Pg-diol (5β-pregnane-3α,20 αdiol 3HS:BSA), PdG (pregnanediol-3-glucuronide R13904), mono-P4 (Quidel clone 425, Coralie Munro, University of California, Davis, CA, USA)
e. E2a (estradiol-17b-OH 17-HS:BSA), E2b (estradiol 17b R0008, Coralie Munro, University of California, Davis, CA, USA), E2c (estradiol 17b R4972; Coralie Munro, University of California, Davis, CA, USA).