Niche breadth measurements indicated no significant difference between sexes. But focussing on food plants, females and males consumed different plants as shown by the amount of different grasses in the diet and underlined by the results of the PIANKA index and the PCA. In the cool dry season, compared to female puku, male puku were found to feed on Sporobolus spp. to much higher extent and less on Panicum spp. This was supported by results of the JACOBS index that indicated avoidance of Sporobolus ssp. in females.
Additionally, the PIANKA index gave a value of 0.67, the lowest value between sexes in one season for puku, indicating the greatest dietary differences between sexes to happen in the cool dry season. This was also visible in the results of the PCA. In the drier seasons of the year, dietary differences between sexes were small. Still, after the first rains, males fed up to 12% more on Brachiaria sp. than females. These slight differences were neither reflected for the hot dry season by the values of the PIANKA index of 0.99 and 1.00 nor after the onset of the rains, respectively.
In the course of a year, according to the results of the PIANKA index, and also apparent in the scatterplots of the PCA, especially male puku showed a change in their diets more than females did. Some male puku hold territories to attract females for mating purpose (BALMFORD et al. 1992). The territories occupy areas with good habitat quality or reduced predation risk (BALMFORD et al. 1992, ROSSER 1992). Other males remain in bachelor herds (see chapter 2). Both territorial males and bachelor males occurred along the Puku-Loop, sometimes very close to each other (own obs.). In contrast to Luangwa Valley, where the territorial males showed a clear neckpatch (ROSSER 1990, own obs.), this feature was less prominent or not visible in Kasanka NP. This made it difficult to distinguish the males and thus, the origin of dung samples. It cannot be excluded that these marked differences between sexes were a result of differences between males. As territorial males occurred along with females, it could be assumed that differences between these males and females are less prominent but still existed, as males were observed alone (see chapter 2).
The niche breadths differed significantly between the seasons. The niche breadths (males and females pooled together) got narrower with the progression of the dry season.
They were broadest in the cool dry season, with a mean Bstandard of 0.09 and 0.10 for females and males; after the onset of the rains it was Bstandard of 0.06 and 0.07 for females and males, respectively. In the late rainy season, the niche breadth was smallest with Bstandard 0.05 only.
This was supported by the PCA showing less scattered dots from the hot dry season towards
the late rainy season. It should be mentioned that general patterns of food plants and the preferences for food plants remain the same throughout the seasons. The exception is the late rainy season where Brachiaria sp. was consumed to much higher extents, to 66%, but not strongly preferred as shown by the JACOBS index. This pattern is interesting: according to general opinion, the range of food is smallest in the dry season, especially in the late dry season when food is scarce and species use their ‘food refuges’, but largest in the wet season as enough food is available (SINCLAIR 1983). Do the puku concentrate on fewer food plants in the late rainy season as their ‘normal’ food plants are less available due to the inundated floodplain? Here again it would be interesting to get information about puku at other sites. During the dry season, especially in the hot dry season, the diet of puku at Kasanka River might represent the diet of most of puku in Kasanka NP; this is not the case in the late rainy season when puku are widespread over Kasanka NP. Thus, the pooled diet breadth of puku of more than one site might reveal a broader niche breadth, a diet not based on Brachiaria sp. but perhaps on more or on other grasses.
Generally, the diets of puku in Kasanka NP and Kafue Region consisted of the same food plants. Interestingly, the diets of puku in PPKR and in Kafue NP were more alike than were diets from puku at different sites within Kasanka NP. This was underlined by the results of the PCA: the dots of puku showed similar patterns in the sub-regions in Kafue Region. The assessment of grasses was conducted in one cool dry season only and therefore was considered incomplete. The results of this assessment showed that, except Hyparrhenia/Andropogon-grasses, other grasses were observed less frequently in Kafue NP than in PPKR. Maybe they had already been eaten by antelopes and thus could not be determined during the assessment. Generally, vegetation in the sub-regions appeared alike (own obs.), both are situated at the border of Kafue River, which might result in similar grass communities and thus to similar diets.
Although differences between sites in Kasanka NP reached a PIANKA index of up to 0.26, differences between puku females in PPKR and Kafue NP resulted in a PIANKA index of up to 0.65 only. Dots of different sexes overlapped most in PPKR, which was supported by a higher PIANKA index for this region. Similar to the situation in Kasanka NP, the slight changes in overlap might have been be due to the origin of the collected dung of puku bachelor males or territorial males. Additionally, females in both sub-regions exhibited a slightly different diet that partly caused the overlap between sexes as well.
Diet breadth measurements revealed a broader niche for puku in Kafue Region than in Kasanka NP. Similar to the observations in Kasanka NP, the niche breadth was smaller in the late rainy season. Although dots were scattered in a somewhat different alignment for the late rainy season, the PIANKA index gave a high similarity for the female puku’s diet during the cool dry and the late rainy season. Puku in PPKR, especially in the late rainy season, were sampled with a focus on puku at the waterhole at PukuPan-Lodge. During the dry season, this waterhole was regularly refilled with water regularly to attract animals for the tourists.
Additionally, the grass around the waterhole was artificially clipped from time to time to force new sprouting. This indeed attracted puku (own obs.) that came especially at daytime to feed and rest, whereas they left the area during the night (see chapter 5). This artificial maintenance of sprouting grass might influence the puku’s food.
According to measurements, puku had a small diet breadth. This might be true when considering specific small-scale sites as the Kasanka-river puku. But considered as a whole, puku appear flexible concerning their diet. This was underlined by the PIANKA index analyses for the puku-regional samples. Further, the JACOBS index revealed that grasses preferred at one site could be avoided at another site which was visible for Sporobolus spp. However, in Kasanka NP, puku showed a preference for Panicum grasses. Then, puku showed differences between sexes and even more pronounced ones between the seasons, if one has a look at the quantity of food plants. Comparisons with the results from other protected areas show that there appears to be no real ‘common-theme’ in the diet of puku except that they feed on grasses. This leads to the conclusion that puku might be able to adapt to local differences in grass community and availability of grasses at small-scale and large-scale.
The small-scale differences in diet between sites in Kasanka NP were remarkable. For red deer (Cervus elaphus) in the Alps, it has been shown that only little variability exists between diets at different sites. The sites were situated in a way that excluded red deer roaming in between (SUTER et al. 2004). It was concluded that they exhibit selective feeding which results in this similar dietary composition even in areas with different grassland vegetation (SUTER et al. 2004). In contrast, the puku in Kasanka NP showed a relatively high variability at landscape scale between sites that they can easily reach. ILLIUS & GORDON (1992) present a formula that allows calculating mean retention time (MRT) for ruminants.
According to this formula (MRT = 15.3 W0.251; W is the weight of the animal), the puku’s food needs 40.23 hours to 47.47 hours from being eaten to excretion. This is enough time for puku to travel some distance between sites in Kasanka NP, separated by a few kilometres
only. Does this indicate that puku are faithful to one site at least during the dry season?
Otherwise the differences in the puku’s diet at different sites would not be that distinctive.
The opportunistic food choice of puku might enable them to live at various places in Kasanka NP and elsewhere: the minimum requirement appears to be that the vegetation is characterised by grasses. The grass community might be of less concern. Generally, some antelope species like freshly sprouting grass after bushfires, like impala or kob antelopes (Kobus kob) (WRONSKI 2003). The puku’s flexible food enables them to cope with short-term and to long-term changes in the vegetation community that result from bushfires or other events. The response of puku following new food sources is reported by JENKINS (2013).
Additionally, this flexibility might be one factor that leads to their high numbers of individuals, although the areas can be different as Kasanka NP, Kafue NP (this study, chapter 2) and Luangwa Valley (RDUCH 2008). On the other hand it has to be mentioned that in Kasanka NP, there are places like Chikufwe, Nafulwe and the southern end of Kafubashi, where food plants of puku occur, but where puku were absent during the sampling for this study.
Puku are reported to live and to feed along water (DE VOS & DOWSETT 1964), sometimes feeding in water that is up to 15 cm deep (HUFFMAN 2011). Observations of puku and their diet in the drier seasons of the year were concentrated along rivers and near to water where puku occurred (see chapter 2). This dependence on water, also highlighted by SKINNER &
CHIMIMBA (2005), is more or less cancelled in the wet season, when water is everywhere.
JENKINS et al. (2002) report small scale movements of puku within the course of a year: in the wet season, puku leave protected areas and go into rice farms. Where do the Kasanka puku go during the rainy season and what do they feed there? This question could not be answered during this study (see chapter 2). But information on the puku’s occurrence might tell us if, for example, puku use the Nafulwe area of Kafubashi area more extensively in the wet season.
This opportunistic food choice of puku that depends on the available food plants available has implications for the research as well. The results of analyses of the puku’s diet do not only depend on the general vegetation in the study area as a whole, but also on the exact site where observations were conducted or samples were taken. The results of the puku’s diet in the course of a year would certainly be different if I had chosen another focal site than e.g. Kasanka River for the collection of the puku dung.