In the early dry season, the grass species Panicum infestum was the main component of the diet of all three test animals. The number of plant species in the diet and the diet breadth were lowest for reedbuck. In a humid savannah, the diet overlap between the three leafy herbivores red duiker, common duiker and suni decreased in the dry season (Prins et al. 2006).
In contrast, goats and sheep in a semi-arid savanna have a less diverse diet in the dry season compared to the wet season (Omphile et al. 2004). Dietary overlap between reed, sedge and wild animals is higher in the early dry season than in the late dry season. All study animals ate higher proportions of dicots in the late dry season than in the early dry season.
Materials and methods
This is presented in Figures 5 and 6, but for further analysis the unidentified fragments and indeterminate monocots were excluded and, where possible, the proportion of the plant species in the diet was not recalculated. In the further analyzes of the diet composition, Panicum infestum and Panicum "indeterminate" were combined. For this index, the proportions of the plant species in the manure were recalculated, as the elimination of the unidentified and undetermined monocot fragments caused major disruptions in the pattern.
For each of the four habitat types existing in the two regions (Mkwaja North and Saadani North), five locations were selected to represent the study area. The proportions of the plant species in the diet were recalculated, as the elimination of unidentified and undetermined monocots caused a large disturbance in the pattern of the data set. Various relationships between nutrients (i.e., nitrogen concentrations, phosphorus concentrations), dry matter digestibility, and frequency of plant species found in experimental animal dung were calculated using Spearman correlations.
Results
Plant species are sorted based on plant species preference observed by livestock in Mkwaja North (Cossack high preference, ** = preference, * = low preference, ne = not eaten. The total number of plant species eaten was very low for all three study animals in both regions and seasons. Especially the reedbuck in Mkwaja North had a very low number of plant species during the late dry season (Table 7).
The highest number of plant species included in the diet was found for the water hyacinth in Mkwaja North during the early dry season. In Mkwaja North, the mean number of plant species in the diets of the reed and waterfowl differed significantly (two-way ANOVA; Table 8). In Saadani North, the number of plant species included in the diet did not differ significantly between animal species or seasons (Table 8).
In Mkwaja North, the diet breadth of the reedbuck was significantly lower than that of the waterbuck. At the end of the dry season, diet overlap between animal species remained high in North Saadani but was very low in North Mkwaja. Differences in diet composition between seasons were greater than between animal species, with overlap particularly low in Northern Mkwaja.
Nitrogen and phosphorus concentrations were significantly higher in the late dry season compared to early dry season values (Table 12). Nitrogen concentrations were significantly higher in plant species in Mkwaja North compared to Saadani North, whereas phosphorus content was significantly higher in plant species in Saadani North (Table 12). The digestibility of plant species did not differ significantly between the two regions Mkwaja North and Saadani North (Figure 13).
Furthermore, the proportion of plant species found in experimental animal manure was not significantly correlated with nitrogen, phosphorus, N/P ratio or dry matter digestibility of these plant species (Table 14).
Discussion
Only three main components (i.e. Panicum infestum, Heteropogon contortus and dicotyledons) were found in the diet of the investigated animals. Both Panicum infestum and Heteropogon contortus are widespread in the study area and occur in dense patches in all habitat types, with the possible exception of Panicum infestum in the late dry season (see above) (Halsdorf 2002). In other study areas, where Panicum and Heteropogon contortus species were abundant, they also represented large portions of the diet of waterbuck, wildebeest, buffalo and zebra (Tomlinson 1980, Ben-Shahar 1991, Bodenstein et al. 2000, Macandza et al. 2004).
However, I found that plant species that were not included in the diet have even higher nutrient values (Figures 11, 12). This could explain why much less Heteropogon contortus was found in the reedbed diet. Reedbuck, which must feed on high-quality food at all times, may therefore have adapted to select more reliable plant species such as dicots during the late dry season.
The reedbuck's diet contained large amounts of dicots, especially in the late dry season. Preliminary data on nitrogen concentrations of grasses, legumes and other plants in the Mkwaja area confirm this pattern (Cech, unpublished). Legumes are a high-quality food source, so a large proportion of legumes in the diet supports my hypothesis that the cane feeds on better quality forage than bread and wildlife.
I did expect an overall increase in the number of dicotyledons in the diet of all three test animals during the late dry season, especially in the small, selective reedbuck. In the following I suggest that reedbuck, waterbuck and wildebeest may follow different feeding strategies, even though they all feed on the same early dry season grass species (i.e. Panicum infestum). Furthermore, they may select higher quality Panicum infestum than waterbuck and wildebeest by feeding exclusively in specific habitat types or at specific locations (e.g. termite mounds) (Melton 1987, Mobaek et al. 2005).
Nitrogen and phosphorus concentrations or the digestibility of green leaves do not sufficiently explain feeding preferences, especially for waterbuck and wildebeest in the Saadani region.
Acknowledgements
Diet selection by roe deer Capreolus capreolus in Kielder Forest in relation to plant cover. Chemical quality of the burnt and unburned grass layer in the Nylsvlei Nature Reserve, South Africa. Seasonal variation in the dung of African grazing mammals, and its consequences for coprophagous insects.
Some aspects of cattle rearing under trypanosomiasis prophylactic treatment at Mkwaja Farm, Tanzania. Ecology of terrestrial grazing ecosystems: Profound functional similarities exist between the Serengeti and Yellowstone. Grazer or Browser: A Classification Based on Stomach Structure and Feeding Habits of East African Ruminants.
The effect of pepsin pretreatment of herbs on the prediction of dry matter digestibility from solubility in fungal cellulase solutions. How a Mega-grazer copes with the dry season: Intake of food and nutrients from White Rhinoceros in the Wild. Elimination of plant toxins by herbivorous woodrats: Revisiting an explanation for dietary specialization in mammalian plants.
Analysis of plant epidermis in faeces: a technique to study the food preferences of grazing herbivores. Spatial variation of summer diet of Red Deer Cervus elaphus in the Eastern Swiss Alps. Wildlife habitat use and dietary preferences of the warthog (Phacochoerus africanus) on a former cattle ranch in a Tanzanian savanna.
Sex differences and seasonal variation in habitat selection in a high-density population of the water goat, Kobus ellipsiprymnus (Bovidae).
Appendix A: Microhistological Dung Analysis
Contents
Introduction
Botanical-taxonomical aspects
Plant micromorphology
The shape and relationship of both cells to each other and to other epidermal cell shapes are important features for recognition. The second cell that makes up the microhair tip is generally poorly visible, and in this paper statements about microhair length refer only to the first cell. They appear only intercostally and have a circular round base and short curved tip.
Large barbed hairs are sometimes present in the costa and most species show a row of large hairs in the basal costa. The ratio of hair length to hair width is an important characteristic that distinguishes macro hair from barbed hair. Lips consist of one pair of lips, poruses, and two concurrent cells; these are round or triangular in shape.
Such patterns are the width (number of cell rows) of intercostal and costal fields or the distribution of trichome, short cells, stomata and homogeneous structure of long cells.
Distinguishing features of analysed plant species
Short cells shapeless or larval with large round cells followed by strong wavy short cells resembling long cells. Short cells are larval in shape with large round cells followed by strong undulating short cells that resemble long cells. Single or paired rod-shaped short cells present in intercostal between each cell and often in costa.
Few short cells are present in intercostal in pairs with one dumbbell-shaped and one square cell. Costal consists of rows of many pairs of round and rod-shaped short cells between long cells. Many rod-shaped short cells are found in the intercostals and sparsely short cells in pairs with a dumbbell-shaped and a rod-shaped cell.
A hooked hair that is larger than the mouth is often present in the intercostals and micro hairs that are longer than the mouth. Small hook-shaped hairs often present in intercostal and micro-hairs shorter than stomata. Hooked drop-shaped hair is often present in the intercostal area, and large barbed hair is present in the costal area.
The costal consists of rows of numerous pairs of short, square-shaped cells between the long cells. Short butterfly-shaped cells, large hair with spines in the margin……….…….Setaria incrassata coastal……….…….Setaria incrassata – Micro hair longer than stingers, hair broad and smaller than stingers. Digitaria milanjiana – Hooked hair drop-shaped, stomachs less wide…..………..…Paspalum dilatatum – Hooked hair sparse, many short cells in intercostal area, long cell walls almost linear.
Short cells caterpillar-shaped, long cells with thick cell walls, square short cells in intercostals………..Chloris mosambicensis cells in intercostals………..Chloris mosambicensis – Short cells dumbbell bell-shaped or reniform. Very large prickles in costal……….………..Aristidia adscensionis costal……….………..Aristidia adscensionis Short cells reniform, single rod-shaped short cells present in intercostals.
