The white rhinoceros (Ceratotherium simum) is experiencing unsustainable poaching losses fuelled by a demand for horn. Increasingly, private and state reserves are dehorning their rhinoceros populations in an attempt to reduce poaching pressure. Rhinoceroses use their horns in social interactions aswell as during resource access and so its partial removal as part of reserve management practices may adversely influence these behaviours. Physiological stress can correlate with animal welfare, reproductive state and health and thus acts as a useful indicator of these parameters. To establish whether dehorning causes a physiological stress response, glucocorticoid and gonadal steroid profiles of free-ranging white rhinoceroseswere determined through the collection and analysis of faecal steroid metabolites before and after dehorning. Faecal corticoid profiles were not influenced by the number of occasions a rhinoceros had been dehorned or by the number of days that had elapsed since dehorning. Furthermore, there was no apparent suppression in the concentrations of testosterone or progesterone metabolites in males and females, respectively, after exposure to multiple dehorning procedures. These findings should increase wildlife managers’ confidence that dehorning does not negatively impact white rhinoceros physiology as measured hormonally.
Faecal samples were collected from 16 white rhinos from a 4932-ha privately owned game reserve in Northwest Province, South Africa (see Penny et al., 2020 for further site details). The rhinos received limited husbandry and veterinary care and had a natural breeding strategy. No supplementary feeding occurred but the rhinos had access to several artificial mineral licks and water sources. Thus, despite being fenced, the population met the African Rhino Specialist Group criteria for a wild population (Leader-Williams et al., 1997). Dehorning began at the site in 2014 (Table 1). Rhinos were classed as subadults from maternal independence until they reached socio-sexual maturity (Table 1). This is when males become solitary and/or territorial at 10–12 years old and at ~7 years old in females after the birth of their first calf (Shrader and Owen-Smith, 2002). Rhinos in Group I were dehorned on 02 June 2016; rhinos in Group II were not dehorned on this date or throughout 2016. Sex classes: F indicates female and M indicates male. Age classes: A indicates adult, SA indicates subadult and C indicates calf. a No samples were collected prior to this date. Project methods were reviewed and approved by the game reserve managers and abide by the ethical guidelines set by the American Society of Mammalogists (Sikes and Animal Care and Use Committee of the American Society of Mammalogists, 2016). Ethical approval was also granted by the Animal Welfare and Ethics Review Board of the University of Brighton (Ref: 2018-1127). Permission was granted by rhino owner D. MacTavish to conduct research at the site and monitor the rhinos in response to dehorning. Prior to each dehorning, the necessary permits were obtained from North West Department of Economic Development, Environment, Conservation and Tourism (DEDECT) by the reserve manager (information available on request). The representatives of DEDECT were present at each dehorning to confirm that all official guidelines were followed throughout the procedure. Dehorning personnel consisted of at least two qualified veterinarians, several assistants and the reserve management staff. Immobilization occurred by shooting a dart into the rump or the gluteal muscles of a rhino from a helicopter. The immobilization cocktail consisted of intramuscular etorphine hydrochloride (M99), azaperone (Stresnil) and hyaluronidase with the dose varied according to animal body size (see Morkel and Nel, 2019). Following immobilization, rhinos were placed in sternal or lateral recumbency and equipped with ear and eye covers to reduce sensory input. Once the rhinos were recumbent, respiratory depressant effects were partially reversed with butorphanol and their breathing rates were monitored (Morkel and Nel, 2019). The anterior and posterior horns were then trimmed with chainsaws and the edges smoothed with a disk sander. Care was taken to cut above the skin-horn interface (9–11 cm) to prevent damage to the growth plate and sinuses. Finally, the effects of the opioids were reversed with intravenous naltrexone. The complete dehorning procedure, from initial darting to full recovery took ~15–20 minutes per rhino. Corticosterone, testosterone and progesterone metabolite concentrations were measured from faecal samples; the data per rhino were then analysed over time to evaluate whether metabolite levels changed. As these hormones are heavily metabolized before excretion, the assays rely on the cross-reactivity of their metabolites (Palme, 2019). This makes it more accurate to refer to the measured concentrations as faecal corticosterone metabolites (FCMs), faecal androgen metabolites (FAMs) and faecal progestogen metabolites (FPMs) rather than as native hormones. To investigate whether rhinos exhibited a short-term physiological stress response to dehorning, FCM concentrations were measured in eight rhinos up to 1 week before a dehorning procedure on 02 June 16 (range, 0–117 hours before; mean, 15 hours; n = 8) and compared with the FCMs of samples collected in the week after dehorning (range, 24–166 hours after; mean, 80 hours; n = 8). No samples were collected in the first 24 hours after the procedure to account for the time lag between stressor and response (Riato, 2007) and changes in FAMs and FPMs were not monitored. This was the second time that the rhinos had undergone a dehorning (Group 1, Table 1), as the rhinos had undergone a prior procedure between 584 and 591 days earlier. To investigate whether rhinos experienced a longer-term physiological stress response to dehorning, the FCM profiles of eight rhinos (Group 1) monitored between 1 and 142 days after the procedure (on 02 June 2016) were contrasted against the FCM profiles of the eight rhinos that had not recently been dehorned (Group II) over the same time period. In total, 89 samples were analysed (2–8 per rhino). Changes in FAMs and FPMs were not monitored. Data for both groups were collected over the same contiguous period from 03 June 2016 until 22 October 2016, with all rhinos exposed to similar environmental conditions and comparable levels of forage availability and disturbance (e.g. from anti-poaching units, vehicles and people). Finally, to investigate whether rhinos exhibited a physiological stress response to multiple dehorning events, FCM, FPM and FAM profiles were related to the number of dehorning procedures that each rhino had been subjected to and the number of days that had passed since their first dehorning. FCMs were analysed from 15 rhinos (IDs 1–15, Table 1) collected over a 516-day monitoring period from 24 May 2016 to 22 October 2017. The monitored period did not begin until 110 days or greater after any rhino’s first dehorning. Rhinos had been dehorned between one and three times each by the end of the monitored period. In total, 143 samples were analysed (1–10 per dehorning event per rhino). FPMs were analysed from five female rhinos (IDs 1, 2, 9–11, Table 1) collected over a 482-day monitoring period from 24 May 2016 to 17 September 2017. The monitored period did not begin until 580 days or greater after any female rhino’s first dehorning. The female rhinos had been dehorned twice by the end of the monitored period. FPMs were not analysed from samples collected during pregnancy, which was calculated by back-counting the average white rhino gestation length of 495 days from parturition (Linklater, 2007). In total, 35 samples were analysed (1–9 per dehorning event per rhino). FAMs were analysed from seven male adult and subadult rhinos (IDs 3–8, 12, Table 1) collected over a 516-day monitoring period from 28 May 2016 to 22 October 2017. The monitored period did not begin until 258 days or greater after any rhino’s first dehorning. Rhinos had been dehorned between two and three times each by the end of the monitored period. In total, 89 samples were analysed (1–11 per dehorning event per rhino). Samples were only collected if they could be matched to a specific individual. Faecal samples were collected from the centre of a dung pile or rectally from immobilized rhinos. Samples were sealed in airtight screwcap bijou bottles within 1 hour of defecation, kept cool and frozen within 4 hours of collection. An aliquot (0.150 ± 0.005 g) of each sample was weighed on a balance (VWR LPC-213) and then added to a solution of one-part ethanol (750 μl, 90%) to one-part distilled water (750 μl). Next, the mixture was agitated in a vortex shaker for 5 minutes (Fisherbrand Mini Vortex Mixer; 2.8 K rpm) to create a faecal slurry. The slurry was centrifuged for 15 minutes (SciSpin Mini Microfuge; 7000 rpm) and then 500 μl of the resulting supernatant was pipetted into a screwcap tube (Eppendorf; 2 ml) and stored at 4°C. To establish the dry weight of each sample, a second aliquot of faecal matter was air-dried and weighed until it stabilized. Endocrine analyses were conducted in the UK at the University of Brighton with permission to import rhino faecal extract from South Africa granted from DEFRA (authorization no: ITIMP16/1052). Commercial enzyme immunoassay kits for corticosterone, progesterone and testosterone were purchased from Enzo Life Sciences (lot numbers: ADI-900-097, ADI-901-097, ADI-900-011, ADI-900-065). Samples were vortexed before analysis and then diluted into sample assay buffer to eliminate matrix interference. To determine optimal detection rates, serial dilutions of samples were made for each hormone. Dilutions were selected that fell within the most linear part of the standard curve and did not generate absorbance values beyond the highest standards (corticosterone: 20 000 pg/ml; testosterone: 2000 pg/ml; progesterone: 500 pg/ml). Samples were run in duplicate in 96-well microtiter plates in parallel with a series of known concentrations of target analyte. Once determined, initial concentrations were adjusted by the water content of their initial weight, to equal mass hormone per mass faecal dry weight (see Ezenwa et al., 2012). To calculate the percentage recovery of each hormone, undiluted faecal extracts were spiked with known concentrations of each hormone; these were then diluted in buffer and assayed alongside unspiked samples. A significant recovery of exogenous hormone was demonstrated for corticosterone (91%), testosterone (75%) and progesterone (78%). The dilutions that resulted in the greatest recovery of hormone were selected for further use. Additionally, the manufacturer had previously tested the corticosterone assay against white rhino faeces and reported a sample recovery of 93.9% (Enzo Life Sciences, 2015). Serial dilutions of faecal extracts were checked for and yielded parallelism between their binding inhibition curves and the hormone standards. Assay sensitivity for corticosterone was 7.90 pg/ml (n = 7 replicates), 2.41 pg/ml for testosterone (n = 4 replicates) and 10.79 pg/ml for progesterone (n = 2 replicates). Average intra- and inter-assay coefficients of variability were <16% for FCMs and <10% for FAMs and FPMs. The degree of cross-reactivity between each target metabolite and potential cross-reactants were conducted by the assay manufacturer (Enzo Life Sciences, 2015, 2016a,b). All analyses were two-tailed, and all alpha levels were set at 0.05. All statistical analyses were performed in R (version 3.5.1; R Core Team, 2018). For the analysis of short-term physiological stress, a Wilcoxon signed-rank test was run using the base R function to account for the small sample size and paired dataset (n = 8, obs. = 16). Generalized linear mixed models (GLMMs) with Laplace approximation were conducted for all other analyses using the ‘lme4’ package (Bates et al., 2015). The GLMMs were fit with gamma distributions with log link functions to account for the distribution of the residuals and continuous predictor variables were centred and scaled. To account for the repeated measures design and the pooling of different ages and sexes, rhino identity was included as a random effect with each subject given the freedom to vary by the inclusion of a random intercept and random slope. Thus, directional physiological changes could still be detected even if subjects’ faecal metabolite concentrations started from different baselines or varied in their rate of change. Type-3 Wald χ2 tests were used to extract the significance of fixed effects using the ‘car’ package following checks of over-dispersion (see Thomas et al., 2015). The conditional R2 value, which represents the variance explained by the entire model, was calculated using the delta method in the ‘MuMIn’ package (Johnson, 2014). To investigate changes in FCM concentration over a longer 142-day period (n = 16, obs. = 89), two fixed effects were included in a GLMM: ‘Horn Change’ (whether or not a rhino was dehorned on Day 0) and ‘Sample Collection Date’ (days passed since Day 0). An interaction term was included to indicate whether ‘Horn Change’ had a differential impact over time. To investigate the changes in FCM (n = 15, obs. = 143), FPM (n = 5, obs. = 35) and FAM (n = 7, obs. = 89) concentrations following multiple dehorning procedures, two fixed effects were included in three GLMMS—‘Days Since First Dehorning’ and ‘Number of Dehorning Procedures’—that each rhino had been subject to at the time of sample collection. An interaction term between the two fixed effects was included to indicate whether the ‘Number of Dehorning Procedures’ had a differential impact over time. Rainfall varied over the 516-day monitored period, so season (with two levels: high rainfall and low rainfall) was included as a random effect and fit with a random intercept. However, it was excluded from the final models as its variance was estimated to be zero (FCMs and FAMs) or its inclusion resulted in a singular fit (FPMs) indicating the model had been over-fit. Additionally, the random slope for rhino identity was excluded from the final model of FAMs due to a further ‘singular fit’ warning.
N/A
DIMA AI Care
