“Puku are gregarious in small to large herds” ANSELL (1960a). This common opinion about the puku’s social organisation can be supported by the findings of this study. Most of all the data about the puku’s social organisation were collected during the dry seasons in the different study areas: generally in Tanzania (RODGERS 1984), from Kilombero Controlled area, Tanzania, (CORTI et al. 2002), from Chobe NP, Botswana, (CHILD & VON RICHTER 1969) and from Zambia: from Luambe NP (RDUCH 2008), from Kafue NP and Luangwa Valley (DE VOS &
DOWSETT 1964) and from Kasanka NP (GOLDSPINK et al. 1998). The two latter studies can be compared directly with the results of this study, also based on observations in the dry season, to evaluate possible changes over the passage of time.
Many of the groups in Kasanka NP could not be classified for group composition. The same applies to the evaluation of sex and age ratio and body condition. This is certainly be related to the fact that the amount of sightings during each survey in Kasanka NP was much higher than in Kafue Region. Although overall sighting distance is larger in Kafue Region, fewer sightings occurred above 200 m in contrast to Kasanka NP.
The non-territorial adult and subadult males assemble at specific places more or less apart from the territorial males and females (own obs.). The transect lines in Kasanka NP lead along some of the ‘male-only’-sites situated along Puku-Loop (southern and northern end) or on Fibwe Plain. Others of these sites were not covered e.g. at the northern border of Wasa Lakes (own obs.). In Kafue NP, the transect lines Airstrip or Dambo covered areas with males. In PPKR only Between covered a patch occupied by some males. But another site with more adult and subadult males was situated about halfway between Puku Pan-Lodge and the airstrip of the lodge and not covered by the transect lines. This might have influenced the results of the overall group composition amount as well as of the sex ratio. This small- scale influence of study sites within the study regions might also explain the differences between the studies reported from Chobe NP by CHILD & VON RICHTER (1969) and DIPOTSO &
SKARPE (2006). Furthermore, as the study regions were sampled different years, inter-annual differences might affect the results of the population characteristics and thus influence the comparison.
Average values of group sizes are difficult to compare as not always given as a mean for all observed groups in the respective studies. In Luambe NP, a mean group size of 3.56 puku was found for July and 4.22 for September (RDUCH 2008) which is lower. DE VOS &
DOWSETT (1964) gave a mean group size of 6 which corresponds almost exactly to the values of Kafue Region from this study, and is lower than the mean group sizes of Kasanka NP.
Although large groups occurred, puku appear to prefer small groups. Especially single animals were encountered often, the modal group size during all surveys. The modal group size of one puku was described for other areas, too, as in Kilombero Valley (CORTI et al.
2002), in Chobe NP (DIPOTSO & SKARPE 2006) or Luangwa Valley (RDUCH 2008). Small groups of up to 5 puku were sighted most frequently in the study regions. This agrees with the findings of other studies in Chobe NP (DIPOTSO & SKARPE 2006), Tanzania (RODGERS 1984) or Luambe NP (RDUCH 2008). Group sizes of puku in Kasanka NP reached 104 in the cool season and 169 the hot dry season, respectively. They were larger than those observed in Kafue NP which were ranging up to 37 puku. Other studies give a maximum group size of 28 (CHILD & VON RICHTER
1969) to 46 (CORTI et al. 2002); these are in the range of the maximum group sizes of Kafue Region. In Kasanka NP, maximum observed group sizes in the cool dry season were large and increased dramatically from 39 in 1996 (GOLDSPINK et al. 1998) to 104 in 2009. These larger groups are rather uncommon; although groups of up to 100 individuals were recorded in Zambia (DE VOS & DOWSETT 1964). But in Kasanka NP they made up 14% of the sightings in
July and comprised about 40% of the observed individuals. This was less in November but still more than in other regions. While selective browsers, in particular if they are small species, are likely to occur in smaller and very small groups, unselective grazers have a potential for larger groups (JARMAN 1974, WIRTZ & LệRSCHER 1983, BRASHARES et al. 2000).
Relating group size to feeding style, JARMAN (1974) classified puku in one group with lechwe, waterbuck, springbok, most gazelles, impala, greater kudu: “although adult males may be found singly, young males and females very rarely occur singly except when sick or giving birth. Most individuals find themselves in groups between 6 and 60 animals, sometimes as many as 200.” This was supported by the findings of this study: although single animals were encountered and small groups made up high amounts of the sightings, most of the individuals were encountered in groups between 6 and 25 individuals in Kafue region and between 6 and 169 individuals in Kasanka NP, respectively. Thus, for puku, it makes a difference if one looks at the occurrence of group sizes or individuals encountered.
Puku in open habitat tended to occur in larger groups than in dense habitat. Variation in social behaviour is observed for other bovids: e.g. buffalo (Syncerus caffer), Grant’s gazelle (Gazella granti), impala or common reedbuck; each form smaller groups in mixed woodlands, scrub or tall grass habitats than they do in short grasslands (BRASHARES et al.
2000). While keeping sensory contact with conspecifics is more difficult in dense habitat, hiding is less feasible in open habitat and grouping becomes the main anti-predator defence (WIRTZ & LệRSCHER 1983). This generally applies for antelopes but as well for the same species in different types of habitat (WIRTZ & LệRSCHER 1983). The groups of puku split up and merge;
nothing indicates a closer relationship between the animals (DE VOS & DOWSETT 1964).
Especially these open groups of puku are likely split up when entering denser habitat (WIRTZ
& LệRSCHER 1983), in contrast to cohesive family groups.
Lastly, group sizes of antelopes are a subject of population density (GERARD et al. 2002, WIRTZ & LệRSCHER 1983), especially if they live in open groups (WIRTZ & LệRSCHER 1983) like puku. If groups are non-permanent units, often flushing and splitting up, any increase of the distance at which animals perceive one another could increase the rate of group fusion and thus group size (GERARD et al. 2002). The puku’s population densities were higher along the transect lines in Kasanka NP than in Kafue Region, reaching the highest population density along Kasanka River, where most of the observations were made, also on group size.
Furthermore, puku reached higher densities in the two study regions in open habitat. These higher population densities, at the scale of the overall study region as well as locally along specific transect lines or in the habitat categories, thus might influence group sizes.
This relation of density and group size as presented by GERARD et al. (2002) might also explain the formation of mixed groups between puku and impala, the kind of mixed group observed most often in Kafue Region. Already DE VOS & DOWSETT (1964) and ANSELL (1960a) stated that puku have more tendencies to mix especially with impala, than waterbuck or lechwe. The two species reached similar densities, which made the formation of mixed groups between these species more probable than any combination with other bovid species. Thus, mixed groups, both in Kafue region and in Kasanka NP, might form rather at random. This was also supported by the variation in number of the two species within mixed groups. However, interspecific association can additionally facilite grazing and reduce predation risk (MCNAUGHTON 1988).
Generally, the share of different group compositions of all classified groups was similar in the study regions and reflected the findings from other regions in southern central Africa.
The amount of male groups, including single males, was higher in Kasanka NP, especially in July 2009 (Figure 2-29 A), and rather low in PPKR. While male groups made up 39% in Chobe NP (DIPOTSO & SKARPE 2006), in July 2009, they reached about 48% of the groups in Kasanka NP. However, in the observation of CHILD & VON RICHTER (1969) these groups made up about 61%, which is even more than in Kasanka NP. With an amount of 35%, 36% and 40%, respectively, the amount of observed male groups was similar in Kafue NP (this study), in Luambe NP (RDUCH 2008) and in Kafue NP and Luangwa Valley reported by
DE VOS & DOWSETT (1964). With 39%, the observed amount of male groups in Kasanka NP in November 2010 fell in this range, too.
The amount of single males is the most often observed group composition in Chobe (DIPOTSO & SKARPE 2006, CHILD & VON RICHTER 1969), in Zambia as observed by DE VOS &
DOWSETT (1964) and RDUCH (2008), as well as generally during this study, except for Kafue NP.
As a compensation, during this study, the amount of breeding groups (Figure 2-29 B) was highest in Kafue NP (26%). Only in Chobe NP, DIPOTSO & SKARPE (2006) observed a higher amount (36%). Thus, apparently the males observed singly in the other regions were accompanied by females and eventually young puku in Kafue NP. The amount of breeding groups in Kasanka NP was comparable to the values given by DE VOS & DOWSETT (1964), CHILD
& VON RICHTER (1969) or RDUCH (2008), but lower in Kafue Region or in Chobe as given by
DIPOTSO & SKARPE (2006). Surprisingly, these low amounts of breeding groups in Kasanka NP were not compensated by higher amounts of groups of females with young puku (Figure 2-29 C).
Nevertheless, in all study region, beside male groups, female groups, possibly accompanied by young puku with or without the presence of an adult male, were very common if not the most important group composition of puku. The differences in the amounts of group composition observed in Kasanka NP might be due to the differences in sex ratio when comparing Kasanka NP to Kafue Region.
A
B C
Figure 2-29: Pictures of group compositions of puku (Kobus vardonii) observed in Kasanka NP (all taken in the cool dry season 2009): a large of adult males, subadult males and juveniles (A), a breeding group of one male with females and juvenile puku (B) and females with juveniles (C).
In Kafue Region as in other areas in southern Africa, the sex ratio of the puku was biased towards the females. The ratio of females to all males of 1.63 in Kafue Region agrees very well with the findings of other studies. For Chobe NP,DIPOTSO & SKARPE (2006) reported a sex ratio of 1.42 and CHILD & VON RICHTER (1969) gave 1.3. In Kilombero Valley, a ratio of 1.21 females per male was found (STARKEY et al. 2002). Even higher are the sex ratios from Luambe NP, with 2.7 females per male in open habitat and 1.9 females per male in closed
habitat (RDUCH 2008). DE VOS & DOWSETT (1964) presented a sex ratio even higher from Kafue NP: 4 females per male. This is high but supporting the very high observed sex ratios in PPKR of 4 females per all male and 6.35 females per adult male.
Sex ratios biased towards the males were given only by GOLDSPINK et al. (1998) while presenting the range of 0.4 to 6.6 females per male from different sites over Kasanka NP in the cool dry season 1994 with an average of 3.4. The average sex ratio in the cool dry season 2009 was 0.66 (females to all males) and 1.06 (females to adult males), respectively. This is a shift towards a male biased or less female biased population of puku in Kasanka NP. The high amount of males was generally visible through the survey in July, August and October along Puku-Loop. If including the subadult males, the ratio got biased toward the males. In November, this shift was less visible concerning both the entire Kasanka NP and Puku-Loop only. This might be the reason for more groups of adult and subadult males in Kasanka NP, and vice versa of the very occasionally observed adult and/or subadult male groups in PPRK.
The sex ratios in Kasanka NP concerning adult males were in the range of the values cited above from other areas. This difference in sex ratio concerning all males or adult males only was less exhibited in Kafue NP; it was rather exhibited in PPKR, but the sex ratio was still very much biased toward the females in PPKR. Apparently, especially the higher amount of subadult males in Kasanka NP is unusual for populations of puku.
The general statement that births can take place at any time of the year (SKINNER &
CHIMIMBA 2005, HUFFMAN 2011) can be supported by the observations in Kasanka NP. When speaking about neonate puku, it has to be mentioned that they hide and already ROSSER
(1989) noted that calves were rather difficult to find. Nevertheless, during all surveys, neonate puku were observed, yet in different amounts; and numbers are rather tendencies than facts. Highest numbers and also amounts of all classified puku of neonates were found in July and November. This might indicate that puku in Kasanka NP avoid the time of the year, which provides less food for the lactating females, e.g. , August and October, as also reported from South Luangwa NP (ROSSER 1989). In November, the close rainy season and first rains induce new growth of grasses and thus births. For Kafue Region, as only a total of two neonate puku was observed, the cool dry season is not the time of most births, and that births are more likely to happen in other periods of the year, unlike in Kasanka NP. However, it might be that the puku born during the cool dry season are more likely to be killed by predators (see also Chapter 4).
In Kasanka NP, the amount of juvenile puku was generally decreasing from July to November. At this time of the year, most former juveniles probably had a shoulder height high enough to be classified as females or subadult male, respectively. Despite when comparing the amount of juveniles in the study regions during the cool dry season, the amounts were smaller in Kafue Region, especially in Kafue NP. This result points towards a higher survival rate of young puku in Kasanka NP. The survival of young antelopes as also of zebra affects the development of their respective populations (SINCLAIR 1979b,GRANGE et al.
2004). This fact might explain the very high population densities of puku in Kasanka NP, generally along the transect lines, pooled or when considering Puku-Loop local densities only, and in open habitat compared to Kafue Region. This high survival rate of young puku, due to conservation measures of Kasanka Trust Ltd and probably also due to low densities of large predators (see also Chapter 4), might also be the cause of the increase of the puku’s population size in Kasanka NP from 1994 to 2009/2010.
The body condition of the puku can provide information on the puku’s populations in the study regions. Body condition of puku in Kasanka NP was generally poorer than in Kafue Region; on each survey, 5% of puku showed a poor body condition. Poor body condition was rarely observed in both PPKR and Kafue NP, respectively. Thus, it appears that puku in Kafue Region cope better with the present conditions; resources seem to be available sufficiently for the amount of puku. The surveys in 2009 and 2010 in Kasanka NP have to be considered separately in order to find out about trends throughout a year. Differences in rainfall in the respective preceding rainy seasons might have caused differences in food availability. Indeed the rainy season 2009/2010 resulted in very high precipitations in Kasanka NP (FRANK
WILLEMS, pers. comm.). In 2009, it appears that less food (of good quality) was available for the puku than in 2010. In August 2009 the general aspect of the body condition in the population of puku was poorest indicating a general ‘bad’ year for puku. In 2010 apparently more or better food was available. Nevertheless, body condition of puku decreased with the progression of the dry season in the two years, although body condition was generally poorest in August 2009.
However the data on body condition were not collected within the same dry season, though. Thus, the question arises if in Kafue Region, as in Kasanka NP, rainfall in 2008/2009 was lower than in 2009/2010 and if so how puku were affected in Kafue Region.
Furthermore, only in Kasanka P, puku with damaged limbs (Figure 2-30 A), slow or not- healing wounds (Figure 2-30 B) or other injuries were observed to a noticeable amount (own obs.).
A B
Figure 2-30: Physically disabled puku (Kobus vardonii) in Kasanka NP (pictures taken in the cool dry season 2009): a puku with a stiff limb (A) and a puku with an abscess-like injury (B).