Cheetah (Acinonyx jubatus)- Data, Pictures & Videos

[Acinonyx sp.]

CHAPTER 9: DEMOGRAPHY OF THE NAMIBIAN CHEETAH 
ABSTRACT 
Namibian cheetahs have suffered, and continue to suffer, high levels of removal due to conflict with local farmers, and it is important to understand the demography of this population in order to determine its likely persistence. Examination of cheetahs reported live-trapped or killed by local farmers, combined with subsequent information from radio-telemetry, allowed demographic parameters such as sex ratios, age and social structure, litter size, interbirth intervals and survivorship to be estimated for cheetahs on Namibian farmlands. Cub mortality was relatively low, but adult mortality was high, particularly for males, and peaked at 5-6 years of age. Neither marking nor relocating cheetahs seemed to affect survivorship, and there was no difference in survivorship between the sexes. Time spent in captivity did not appear to affect survival after release. These findings will be useful in formulating recommendations regarding the conservation and sustainable utilization of cheetah populations outside protected areas.


9.1 INTRODUCTION 
Determining vital rates and demographic parameters is fundamentally important to the accurate understanding of any population (Eberhardt 1985, Lebreton et al. 1992, Lebreton et al. 1993). When a population is subject to high offtake, it is essential to establish whether the level of removal threatens its long-term viability. Large mammals are particularly sensitive in light of their long gestation and interbirth intervals, extended parental care and long maturation. Their life history parameters effectively lower the potential rate of population increase (Eisenberg 1981, Harvey et al. 1989), creating a higher extinction risk. Adult survival is a particularly important parameter, documented to exert a substantial impact on population viability for large, long-lived species (Crooks et al. 1998, Doak et al. 1994, Eberhardt 1985, Taylor et al. 1987). The Namibian cheetah is an example of a threatened population which has been subject to a high level of removal, and whose vital rates require more accurate determination in order to assess and manage the impact of such removals. Vital rates of cheetahs have been reported in the Serengeti (Caro 1994, Kelly and Durant 2000, Kelly et al. 1998, Laurenson 1995, Laurenson et al. 1992), but the population in Namibia is subject to strikingly different pressures (Marker-Kraus et al. 1996). The intense conflict with humans results in a high number of adults being removed, which could have a more severe effect on long-term population viability (Crooks et al. 1998) than the lion-perpetrated high cub mortality reported in the Serengeti ecosystem (Laurenson 1995).In order to ascertain the impact of such mortality, this paper reports and examines the vital rates and life history parameters of cheetahs on Namibian farmlands. Assessment of these reproduction and survivorship estimates provides insight into the vulnerability and likely persistence of the cheetah population in Namibia. In 1996, a Population and Habitat Viability Assessment was conducted, and called for, as a priority, more data on demographic parameters such as annual female mortality and reproductive information (Berry et al. 1996). Here, we provide these data.

9.2 METHODS 
Demographic parameters were estimated using data from cheetahs captured opportunistically by farmers between 1991 and 2000. Some of these cheetahs were subsequently radio-collared and released, which provided information regarding reproduction and survivorship in the wild. Our interpretations take note of a potential bias in the sample population arising from the high proportion of cheetahs captured at ‘playtrees’, which are used more by adult males than by adult females (Marker-Kraus and Kraus 1995). 

9.2.1 Trapping, immobilising and marking cheetahs 
During the course of the study, we examined cheetahs that had been trapped on Namibian farms, using the capture and immobilisation protocols described in Chapter 3. Each live cheetah that we examined during the study was marked with a uniquely numbered implantable transponder (Trovan Electronic Identification Systems, Model-ID 100) placed at the base of its tail, and/or an aluminium ear-tag with a unique ID number in the individual’s ear. On designated animals (those released in the core study area), a neoprene radio-telemetry collar with external antenna was also fitted (Advanced Telemetry Systems, Minnesota). The radio-collars were 3.8 cm wide with an adjustable strap from 30-45 cm long, with a 30 cm antenna extending about 18 cm from the collar. The collars were fitted with a "C" cell lithium battery with a life expectancy of over 36 months. Radio-collars weighed 280g and were equivalent to 0.56% of body mass for a 50kg male and 0.76% for a 37kg female, well below the limit suggested by Amlaner and Macdonald in 1980. In line with Caro (1994), we could detect no impact of these collars on cheetah survival. 

9.2.2 Age classification 
Age classification was based both on experience with captive cheetahs and on information from previous studies (Burney 1980, Caro 1994), and followed the protocol described in Chapter 3.

9.2.3 Determining social structure 
Cheetahs are relatively social felids and often occur in groups: in many cases, farmers left adjoining traps open after catching a cheetah, to ensure that all members of any social group were captured at the same time. In other instances, capture was more random and it was likely that other cheetahs in the same social group remained free. Parameters such as coalition size, litter size and age-specific cub mortality were therefore determined using data from cases where determined attempts had been made to capture the entire group of cheetahs. Cheetahs were classified as to the social group of which they were a part when they were captured, using the following categories. Males over 18 months old were classed either as single males or as members of male coalitions, while females over 18 months old were classed as either being single females or as mothers trapped with cubs. The remaining classes were cubs (18 months old or younger) trapped without a dam, and mixed-sex groups of young independent cheetahs (19-30 months old), which were presumed to be littermate groups. 

9.2.4 Estimating reproductive parameters 
Age at parturition was estimated by examining females trapped with cubs, and by observing new litters of cubs produced by radio-collared females of known age. Long-term monitoring of six radio-collared females that had multiple litters provided information regarding interbirth intervals. Information regarding the distribution of births through the year was gathered from the examination and ageing of cubs trapped, and from observations made of females and cubs during radio-telemetry. Litter size was determined from groups trapped where determined efforts had been made to capture the entire family unit, and from observations of radio-collared females with cubs during aerial tracking. Although we have no data regarding litter size at birth, observations of litters of different ages allowed some estimation of agespecific cub mortality. 

9.2.5 Estimating mortality and survivorship
 Most cheetahs were released at site of capture, but when this was not possible, the cheetahs were relocated. Relocation was classified as being 100km or more away, as this should be well beyond the diameter of a normal home range (Marker 2000). The majority of cheetahs released within the core study area were radiocollared in order to gain information regarding post-release movements and home ranges. In addition, the tracking of cheetahs enabled information to be gathered regarding survivorship. Wild cheetah deaths reported to CCF included cheetahs that had been radio-collared, some that were tagged, and some that had not been marked at all. The deaths of marked cheetahs, whether radio-collared or simply ear-tagged, were often reported and enabled comparisons to be made about the approximate age of death of handled cheetahs versus those of cheetahs that had never been handled. Mortality rates and life expectancy data were calculated following Downing (1980). The age of a cheetah at death was taken to be the midpoint of the age category in which it was recorded at the time of death. By using this midpoint, the formulae used should underestimate and overestimate the age at death for equal numbers of cheetahs and thereby give an approximation that is close to the actual distribution of ages at death. Statistical analyses were performed using SPSS version 10.0 software (SPSS Inc. Chicago, USA). Means significance testing was carried out using the parametric independent samples t-test, using Levene’s test to determine equality of variances, while departures from expected ratios were analysed using a chi-squared test. The non-parametric Spearman’s rank correlation coefficient was used to determine the significance of relationships between variables measured on nominal scales, while Pearson’s correlation coefficient was determined for interval data. Tests performed included one-way analysis of variance, and log rank for the equality of survival distributions following a Kaplan Meier analysis. All tests were two-tailed unless otherwise stated. 

Table 9.1 Annual demographic breakdown into social groups of the cheetahs captured and examined (SM = single males; C = coalition, CM = coalition males, CS = coalition size, SF = single females, LS = litter size, LM = littermate).

*This image is copyright of its original author

One hundred and seventy adult males were reported trapped, of which at least 97 (51.2%) were in coalitions. Coalition size (Figure 9.1) ranged from 2 to 4, with a mean of 2.3 throughout the study (n = 42), and showed no significant change through the course of the study (F = 1.11, p = 0.389).

Figure 9.1 Coalition sizes for cheetahs captured throughout the study (n = 42 coalitions) 

*This image is copyright of its original author

9.3.2 Age and sex structure of sample population 
A summary of the age structure of captured cheetahs through the study is shown in Figure 9.2. There was a highly significant variation in overall capture frequency for each age cohort (F = 2.02, p = 0.030). Assuming that captures reflect trends in the wild population, the age structure of the population was not stable.

Figure 9.2 Breakdown into age groups of cheetahs captured throughout the study period. 

*This image is copyright of its original author

Overall, the combined increase in the proportion of young animals and decline in the proportion of older animals meant that there was a significant relationship between age groups captured and year of the study (F = 2.02, p = 0.030), with more young animals being captured as time went on (rs = -0.12, p = 0.016). This overall result arises from a combination of trends, although these were not in themselves statistically significant. Thus, the proportion of young cubs captured increased through time, although the trend was not statistically significant (rs = 0.15, p = 0.667), and the same positive trends were seen for both large cubs (rs = 0.22, p = 0.519) and adolescents (rs = 0.55, p = 0.079). By contrast, the proportion of newly independent cheetahs captured showed some decline through time (rs = -0.47, p = 0.145), as did the proportions of young adults (rs = -0.19, p = 0.574) and prime adults (rs = -0.53, p = 0.096). The proportion of old and very old adults (age groups 7 and 8 combined) captured stayed virtually constant through time (rs = 0.003, p = 0.989).

There was a strong bias towards capturing adult males, with 2.9 adult males captured for every adult female. This proved to be a significant deviation from a 1:1 sex ratio with regard to the adult cheetahs trapped (Ȥ 2 = 47.1, p = 0.000). Amongst the adult cheetahs captured, the fraction of males trapped declined through the study, but the trend was not statistically significant (rs = -0.53, p = 0.097).

9.3.3 Reproductive parameters 
Age of breeding females ranged from 19 months to 12 years old (n = 43, mean = 5.3 years) (Figure 9.3). The percentage of adult females that were trapped with cubs each year ranged from 22.2% to 70.0%, with an overall mean of 44.5% (n = 60). 

Figure 9.3 Estimation of age of dam at parturition, for the litters of cubs examined during the study. 

*This image is copyright of its original author

Litters were produced throughout the year, but varied significantly between months (Ȥ 2 = 18.3, p = 0.05), indicating some degree of seasonality. Birth peaks were evident in March and July, with 40% of litters born in these two months, while only 5% of litters were born from October-December (Figure 9.4)

Figure 9.4 Distribution of cheetah cub births throughout the year, using estimated month of birth from examination of cubs captured. 

*This image is copyright of its original author

Litter size obtained through trapping ranged from 1-6 with a mode of 3 (mean = 3.1, n = 27 litters), with no statistically significant variation between years (F = 0.56, p = 0.812). Litters observed during radio-telemetry alone ranged in size from 2 to 5 with equal modes of 3 and 4 cubs (mean = 3.4, n = 13 litters), and also did not vary significantly between years (F = 1.51, p = 0.131). There was no significant difference in mean litter size observed between the two techniques (t = -0.93, p = 0.357). Overall, therefore, the mean litter size observed, using data gathered from both methods, was 3.2 post-emergence (n = 40 litters, Figure 9.5).

Figure 9.5 Litter sizes of cheetahs captured and examined throughout the study.

*This image is copyright of its original author

A regression on litter size and cub age for all the observed litters, including those seen multiple times at different ages (n = 65 observations) implies an estimated litter size at age 0 (birth) as 3.5 (Figure 9.6).

Figure 9.6 Regression of litter size with estimated cub age for cheetahs captured and examined during the study. 

*This image is copyright of its original author

It is very likely, however, that field observation data underestimate mean litter size at birth, as there is likely to be pre-emergence mortality in the den (Laurenson et al., 1992). There was no significant deviation from an expected 1:1 sex ratio regarding cubs aged 12 months old or below (Ȥ 2 = 0.62, p = 0.432). While females were captured with as many as six dependent cubs, average litter size for newly independent littermates ranged from one to three with a mode of 2 and a mean of 2.4 (n = 9 litters). This may be an underestimate, however, if newly independent animals are less likely to stay with a trapped littermate and are therefore less likely to be captured and recorded. Reproductive information was gathered on 19 litters from 10 radio-collared dams (Table 9.2). Interbirth intervals following litters that were raised to independence (n = 6) ranged from 21 to 28 months, with a mean of 24 months. 

Table 9.2 Information gathered from radio-collared dams (n = 10) regarding observed reproductive rates, interbirth intervals, litter sizes and cub mortality.

*This image is copyright of its original author

Although cheetahs have been known to survive for up to 21 years in captivity (Marker-Kraus, 1997) the maximum age recorded here for an animal that was still reproductively active was 12 years, so this can be regarded as the figure for effective longevity.

9.3.4 MORTALITY AND SURVIVORSHIP 
Mortality rates were calculated from all recorded wild deaths (n = 67), including 45 marked cheetahs and 22 that had never been handled, and these are shown in Table 9.3. These data show that the age specific mortality ranged from 20% to 28% for the first three years of life and then dropped to virtually zero between three and five. There was then a large peak of mortality at age 5-6, but of the few cheetahs that reached six years of age (n = 4), all survived for a further four years.

Table 9.3 Life table showing mortality rates for wild cheetahs throughout the study period.

*This image is copyright of its original author

Survivorship following release could be calculated for the 45 wild marked cheetahs, and ranged from 0.6 months to 48.5 months for males (n = 35) and 0.6 months to 65.4 months for females (n = 10). Survivorship of male and female marked cheetahs is shown in Figure 9.7. 

Figure 9.7 Survivorship of marked male and female cheetahs released during the study period. 

*This image is copyright of its original author

Of the cheetahs shown in Figure 9.7, 71.1% (n = 32) were adults at the time of release, of which 46.9% (n = 15) were of prime breeding age. Mean survival time for tagged and released males was 14.4 months (n = 35), while for females it was 18.5 months (n = 10). Although females lived for slightly longer, the difference in survivorship was not statistically significant (log rank = 0.71, p = 0.401). When analysis was restricted to adult animals, to remove any effect of cub mortality following the death of a female, the mean survival time was 16.2 months for males (n = 23) and 20.3 months for females (n = 9), a difference that again was not significant (log rank = 0.58, p = 0.447). There was no significant difference in estimated age at release between the sexes (t = 0.24, p = 0.812). Cumulative annual survival was calculated for radio-collared male and female cheetahs and is presented in Figures 9.8a and b. 

Figures 9.8a and 9.8b Annual cumulative survivorship for (a) female and (b) male cheetahs. Error bars represent standard errors. 

*This image is copyright of its original author

There was a significant difference between cumulative yearly survival rates between males (mean = 9.4 months) and females (mean = 10.2 months) (t = 2.07, p = 0.009). There was also a significant difference between the frequency of natural mortality and human caused mortality for both males (t = -7.37, p = 0.000) and females (t = - 7.08, p = 0.000), with the majority of deaths attributable to human causes. Marked cheetahs were slightly older at death than unmarked cheetahs, but the difference was not statistically significant (t = -1.94, p = 0.057). Amongst marked cheetahs, survivorship of radio-collared animals was compared to those that were just ear-tagged and released. As only fully-grown (>18 months), healthy cheetahs were radio-collared, whereas the tagged population included subadults, the analysis was restricted to cheetahs that were released at age group 4 or above. Mean post-release survival time for radio-collared cheetahs was 20.2 months (n = 24), whereas cheetahs wearing only a tag survived on average for 9.8 months after release (n = 8): see Figure 9.9. Radio-collared cheetahs survived longer following release than tagged cats, and differed significantly in their survival distribution (log rank = 3.92, p = 0.048). This was not an effect of differing ages at release between the two groups (t = -1.23, p = 0.230). 

Figure 9.9 Kaplan-Meier survival curves for radio-collared versus marked cheetahs. 

*This image is copyright of its original author

It is possible that since all cheetahs radio-collared were in our core research area, more people were aware of the presence of marked cheetahs which was a deterrent to killing cheetahs, whereas ear-tagging of cheetahs took place throughout the country, including places where many people were unaware of our research. There were 21 reported cases of human-caused mortality amongst the marked animals, and if the radio-collar acted as a deterrent then the expectation would be for there to be relatively few collared cheetahs within this sample. In fact, 14 of the 21 cheetahs reported killed by humans were radio-collared, which was not a significant deviation from an equal ratio of collared to tagged cats (Ȥ 2 = 2.33, p = 0.127). However, 36.8% of the released adult cheetahs were radio-collared, while 63.2% were only tagged. Using these proportions as the expected ratios, radio-collared cheetahs comprise a significantly greater percentage of human-caused mortality than would be expected (Ȥ 2 = 8.05, p = 0.005). Often cheetahs with health problems spent longer time in captivity before release than healthy cheetahs. To investigate whether time spent in captivity was detrimental in terms of subsequent survival in the wild, survivorship following release for adult animals was correlated with total days spent captive prior to release. There was a slight negative correlation between the two variables, but it was not significant (r = -0.52, p = 0.778). Cheetahs were either released at sight of capture or distances varying between 50km to 600 km away from the capture location. There was no significant difference in mean survivorship for adult cheetahs released at the site of capture (mean = 18.0 months, n = 19, s = 13.6), those released within 100km of the capture site (mean = 15.4 months, n = 7, s = 9.1) and those relocated over 100 km away (mean = 28.4 months, n = 8, s = 27.3) (F = 1.34, p = 0.275).

9.4 DISCUSSION 
9.4.1 Social structure 
The most common age groups trapped were young adults and prime adults; removal of these age classes is likely to have a particularly detrimental effect on the population (Crooks et al. 1998). If this sample is representative of the nationwide picture, then this finding is of particular concern. Males are often caught at the time of dispersal when they are trying to establish a territory, travelling long distances across many farms, presumably increasing their chances of being trapped. This bias towards young adult and prime adult males is likely to be a sampling bias rather than a true indication of population structure in the wild, due to the aforementioned ‘playtree’ bias. The disproportionate removal of males has been seen in many mammalian species and, although probably less damaging to the viability of the population than a skew towards removing females, can nevertheless have serious impacts in terms of social structure and behaviour (Tuyttens and Macdonald 2000). In areas of fragmented populations and low density where removed males cannot easily be replaced by immigrants, continued removal of adult males could have a severe effect and lead to lower reproductive rates and an accelerated decline (Tuyttens and Macdonald 2000). This scenario is likely to become of greater importance if cheetah populations become more fragmented in the future. The removal of dominant, territorial males can also be counter-productive to farmers insofar as it may lead to the increased survival of subadult and transient animals that would not normally settle in the area (Young and Ruff, 1982, cited in Tuyttens and Macdonald, 2000) and which may be more likely to become ‘problem’ animals (Marker-Kraus et al. 1996). 

9.4.2 Reproductive parameters 
The live capture of females with cubs presented the opportunity to monitor the sex structure of cubs and to estimate reproductive parameters. Litters could not be studied before emergence from the den, and therefore provide no direct information regarding either sex structure or litter size at birth, as infant mortality can be substantial before emergence (Laurenson 1994). The litter size of 3.5 at birth extrapolated from the data in this study is slightly higher than the 3.1 found in previous studies in Namibia (Marker-Kraus et al. 1996), while information from captive studies give a mean litter size of 3.3 (Marker-Kraus 1997), and data compiled by Gittleman (1986) indicated an average of 3.8 cubs per litter. These data are difficult to compare directly as the litter size at various ages is not given, but the litter size at independence of 2.4 found in this study is similar to the 2.1 reported from the Serengeti (Kelly et al. 1998). Although females are capable of breeding at an earlier age (Marker-Kraus 1997, Wildt et al. 1993), reproduction on the Namibian farmlands usually does not occur before 1.5 to 3 years of age (Morsbach 1987). Similarly, whereas males are physiologically capable of breeding at less than 2 years of age (Marker-Kraus 1997, Wildt et al. 1993), social constraints may limit breeding of Namibian male cheetahs to older, territorial animals in the prime age category.

9.4.3 Mortality and survivorship 
The mortality and life expectancy data reveal that for both male and female cheetahs in our sampled population, the highest peak of deaths is between five and six years of age. This is to be expected given that the trapping and removal methods tend to select prime breeding age adults, as discussed above. The mortality figure found here for the first year of life (25%) should be interpreted with caution as it cannot include mortality before emergence from the den, which has been found to be a period of high mortality in other studies (Laurenson 1994). However, it appears that in Namibia, the level of cub mortality is indeed far lower than in game reserves with a high density of intra-guild competitors. Despite this, even without intra-guild competition, fewer than 50% of the cubs reach independence. Data from the Serengeti show that female cheetahs surviving to independence had a good chance of reaching old age (Kelly et al. 1998). This was not the case here: in this study, female cheetahs that reached independence still had an 86% chance of dying before six years of age. This reflects the differing pressures on the two populations: in the Serengeti, the greatest threat to survival is predation by larger carnivores, particularly lions, on dependent cubs (Laurenson 1994). This threat recedes once a cheetah reaches adulthood, whereas the greatest threat to cheetahs in Namibia appears to be humancaused and focuses on cheetahs of prime breeding age. In this study the threshold seemed to be six years of age; the few cheetahs monitored that lived that long managed to survive until old age. The removal of adult cheetahs has been shown to have a far more significant impact on the overall population than the removal of cubs (Crooks et al. 1998). The selection by trapping adult cheetahs is therefore of major concern regarding the ability of the Namibian cheetah population to persist long-term. The majority of these removals are in reaction to a perceived threat to livestock and/or game by commercial farmers (Marker-Kraus et al. 1996). As a result, conservation efforts should be concentrated on educating farmers in alternative game and livestock management techniques to reduce losses and lessen conflict. Much of the information gathered through this study was only possible by directly handling cheetahs, including fitting radio-collars prior to release. However, the invasive handling and monitoring of wildlife, particularly when it involves an endangered species, has been the focus of much debate (Bateson 1991, Burrows et al. 1994, Creel et al. 1997, Cuthill 1991, Driscoll and Bateson 1986, Smith and Boyd 1991). The handling of wild animals is likely to involve some degree of stress (Laurenson and Caro 1994), and it has been argued that this stress may hamper the eventual survival of the animals (Cuthill 1991, Martin and Bateson 1986). This may be particularly important with cheetahs, a species potentially more vulnerable due to its inherent genetic uniformity (Laurenson and Caro 1994, Smith and Boyd 1991). However, this study showed that the survivorship of the wild cheetahs we handled and marked was no lower than that of wild cheetahs that had never been handled. In addition, there was no evidence that putting a radio-collar on a wild cheetah had a negative impact on survival following release, and the radio-collared cheetahs studied here lived longer post-release than their tagged counterparts. Collaring of cheetahs was conducted in a relatively concentrated area, where farmers were well aware of the research being conducted. Therefore, public awareness may have contributed to the longevity of radio-collared cheetahs that are not causing problems with farmers. Marking of cheetahs in other parts of the country, where public awareness was not as extensive, may have caused the differences between the post-release longevity. It could not be ascertained whether the collars acted as a deterrent to farmers who would otherwise kill the cats, instead prompting them to contact CCF or another authority to deal with problem or trapped cheetahs. The higher than expected incidence of radio-collared cheetahs amongst those killed by humans may indicate that people are more likely to report a death if they see a radio-collar on the cheetah concerned. Cheetahs that were relocated far from their capture site provided important survival information. The relocated cheetahs did not have significantly different survival rates from those released in close proximity to their capture site, which suggests that relocating cheetahs into suitable habitat can be an effective conservation strategy without negatively impacting the survivorship of the individual cheetah. Of importance in our monitoring was the possible effect of keeping cheetahs in captivity prior to release. There is a chance that holding an animal in captivity could have a detrimental effect on survivorship after release, either through a reduction in physical fitness and hunting ability, or by animals losing their territories and being forced into marginal areas. Analysis of cheetahs we handled, however, showed that there was no relationship between the length of time spent in captivity and subsequent survivorship following release, despite the fact that some of the cheetahs were held in captivity due to being in poorer physical condition. From this 10-year study we have identified certain areas of concern, e.g. the continued removal of prime adults from the population, the skewed sex ratio resulting from capture methodology and the apparently unstable age distribution. In addition, we have been able to establish parameters that can be used in future modelling efforts. Such modelling exercises will provide a basis for long-term conservation strategies for cheetahs on Namibian farmlands. 




 Page 186-208
http://citeseerx.ist.psu.edu/viewdoc/dow...1&type=pdf Reply