comparison of voluntary food and water consumption and digestion in kirk's dikdik and suni

E. A f r . Wildl. J . , 1977, Volume 15, pages 41-48

Comparison of voluntary food and water consumption and digestion in Kirk's dikdik and suni

PET E R P. H 0 P P E Department of Animal Physiology, University of Nairobi, P.O. Box 30197, Nairobi, Kenya

Summary Voluntary food and water consumption of Kirk's dikdik (Madogua kirki) and suni (Nesotragus moschatus) was determined under controlled laboratory conditions. Both species consumed large amounts of dried matter (lucerne hay leaves) per day. Dry matter consumption of dikdik accounted for 3.8 &0.5% and of suni for 3.5f 0.2% of body weight. Dry matter intake in one immature dikdik and one female dikdik during late pregnancy and lactation was considerably higher.

Dikdik drank very little water (278 ml/day, equalling 83 ml/kgosz) which is further proof of their adaptation to arid environment. Suni drank about twice as much (401 ml/day, equalling 155 ml/kg0'82), but individual variation of water consumption was high. Thus, the forest-dwelling suni seems to be less adapted to dehydration than the dikdik.

There was no difference in the digestibility of dry matter and gross energy of lucerne hay leaves for dikdik and suni. The fraction of N-free extracts which consists mainly of soluble carbohydrates was digested very well (84% and 81 x, respectively), whereas fibre and cellulose digestibility was poor. Suni digested crude protein, minerals and cellulose significantly better than dikdik.

Introduction Kirk's dikdik (Madogua kirki Gunther, 1880) and suni (Nesotragus moschatus von Dueben, 1846) are among the smallest existing ruminant species, the adult body weight of dikdik ranging between 4 and 6.5 kg and that of the suni between 3.4 and 8.2 kg (Hofmann, 1973). Dikdik inhabit semi-arid areas and dry bush-country (Hen- drichs & Hendrichs, 1971) and are independent of drinking water. Investigations of their water metabolism (Maloiy, 1973; Schoen, 1972) and thermoregulation (Hoppe et al., 1975) have shown that they are physiologically well adapted to high ambient temperatures and aridity. Suni occur in dense mountain and coastal forests. Apart from small areas of overlapping distribution they are not sympatric with dikdik. According to Lydekker (1926) they too are independent of drinking water.

Dikdik and suni were chosen as experimental animals because their body size is at one extreme of the suborder ruminantia which comprises species weighing between 3 and over 1000 kg. Due to their small body size they can be expected to have the

41

42 Pcivr P. Hoppe

highest metabolic rate and accordingly the highest energy requirement per kg of body weight, of all ruminants. Detailed anatomical investigations (Hofmann, 1973) of their forestomachs indicate obvious morphological differences from the ‘typical’ forestomachs of domestic ruminants. Both dikdik and suni have a relatively small rumen with dense papillation, and a large reticulum. Rumen pillars which function as food- retaining barriers are poorly developed. Furthermore, there is no dorsal zone of adhesion of the rumen which suggests that rumen--reticular motility is different from domestic ruminants. Both are highly selective feeders (Hendrichs & Hendrichs, 197 I , Haltenorth, 1963; Tinley, 1969) and have been termed selectors of fruit and dicotyledonous foliage by Hofinann (1973). Hungate (1959) found a very high rate of microbial fermentation in the rumen of one suni. This paper provides quantitative data on food and water consumption and digestibility of food in dikdik and suni as basic information for further work on the function of their forestomachs. Since field methods to establish food consumption are inaccurate the trials were done under controlled laboratory conditions.

Animals and methods Five adult Kirk’s dikdik and five adult suni were captured near Nairobi and kept singly in metabolism cages of 60 x 40 x 55 cm at thermoneutrality. Urine and faecal pellets were voided on the floor of the cages resulting in some contamination of faeces with urine. At the time of the experiments the animals had been in captivity for 3 months and weighed between 3.9 and 6.5 kg (dikdik) and 2.8 and 3.7 kg (suni). They had free access to lucerne hay (Medicago sativa) of which they selected the leaves only, sweet potato vines (Ipomea hatatas), leaves of the bush Crewia similis and water. A mineral mixture was offered twice a week. One week before the start of an experiment they were only offered lucerne hay and water.

Fifteen experiments were done on dikdik and seven on suni to establish voluntary food and water intake. The experiments lasted from 6 to 20 days. The animals were offered daily weighed amounts of lucerne hay leaves which had been separated from the stems (Table I) , and water in bowls. Food was continuously on offer in amounts well above the quantity which had been eaten the week before the experiment. Once daily, the remaining food and water were weighed to the nearest gram. Dry matter (DM) and proximate nutrient content of the leaves were analysed according to the AOAC (1960) method, and gross energy was determined by bomb calorimetry. The animals were weighed at least once at the beginning and the end of each experiment. Voluntary food consumption was expressed as the amount of DM consumed per day

Table 1. Nutritive value of lucerne hay leaves

Proximate nutrient ”/, of dry matter

Crude protein

Crude fat Crude fibre Ash Nit rogen-free extracts Gross energy

(N x 6.25) 23.7

5 .7 20.9 15.7 34.0

4 .55 Cal/g D.M.

Food consumption in dikdik and suni 43

Table 2. Voluntary daily consumption of dry matter (lucerne hay leaves) and water in Kirk’s dikdik and suni. Means1S.D. of eleven and seven experiments on dikdik and suni, respectively

Kirk’s dikdik Suni

g dry matter 161k20 111k17

Dry matter as percentage of body weight 3 .76h0 .51 3.46 f 0 .24 ml water* 278 f 56 401 f 123 ml water*/kgo.82 8 3 . 0 h 5 . 2 155k49 ml water*/g dry matter 1 . 8 f 0 . 3 3 .62& 1 .O

g dry matter/kg0”5 54.7 & 7 .O 4 5 . 4 1 5 . 3

*Including plant-borne water.

and per kilogram metabolic weight (kg body weight respectively. Water consumption was expressed as ml water (drinking and plant borne, respectively)/kgO M2

according to Macfarlane er al. (1972) in order to facilitate comparisons with other animal species.

Digestibility trials were carried out with three dikdik and three suni in experiments lasting 14 days. Dry matter and water consumption were measured each day. Total faeces were collected every morning, weighed, and the DM content of the faeces determined by drying a subsample of 20 % in a ventilated drying-oven at 110°C for 48 h. Total daily faecal DM was then determined. The difference between total DM consumed and total faecal DM excreted in 14 days was taken as the apparent DM digestibility. 10% of each day’s faecal DM was collected, mixed and finely ground, and the composite faecal samples analysed for proximate nutrient content to calculate digestibility coefficients. No corrections were made for any eventual loss of nitrogen during the drying process. Urine and serum samples were analysed for osmolality using a wide range osmometer working on the principle of freezing point depression. Statistical evaluation of species differences was done by the t-test (Student).

Results Adult Kirk’s dikdik consumed a mean of 161 20 (S.D.) g DM/day which equals 55 & 7 g DM/kg metabolic weight (Table 2). Daily intake was between 2.8 and 4.7 % of body weight. Food was eaten frequently in small portions, during each hour of day and night. The water intake, 90% of which was drinking water, was 278 f 56 ml/day or 1.8 times the DM intake. Dikdik have a rather tedious way of drinking by inserting the tongue into the water and licking it in a cat’s manner. The highest amount of the water drunk by any of the non-pregnant, non-lactating adults was 394 ml in a day.

In late pregnancy (4 weeks pre parturn*) and the first and second week of lactation, food intake of one female dikdik was increased by 12%, 51 % and 42%, respectively, and water intake by 52%, 170% and 162%, respectively, above non- pregnancy level. One immature dikdik of 60 days of age consumed 21 % more DM per kg metabolic weight than adults, and water consumption was 92 ml/kg0’M2.

5.3 g DM/kg0’75, and 401 of water per day (Table 2). Their pattern

The length of gestation is 170- 175 days (Hendrichs & Hendrichs, 1971)

Adult suni consumed 111 f 17 g DM, equalling 45.4 123 ml equalling 155 f 49 ml/kg

44 Peter P . Hoppe

of food consumption was characterized by frequent ‘meals’ of a few minutes duration. The amount of water drunk varied considerably between individuals and in the same individual in different trials. Unlike the dikdik, suni drink with the mouth submerged i n the water. Quantities in excess of 500 ml/day were not drunk infrequently, and the highest amount drunk in a day was 705 ml which equalled 21 % of body weight.

Digestibility coefficients are given in Table 3. Digestibility of dry matter and gross energy were very similar in both species. Apparent digestibility of minerals (‘ash’), protein and cellulose was significantly higher in suni. The fraction of N-free extracts was remarkably well digested by both species, whereas cell waft constituents like cellulose and lignin were digested less efficiently.

Table 3. Apparent digestibility of lucerne hay leaves for Kirk’s dikdik and suni. Means+S.D. of three animals each

Difference Kirk’s dikdik Suni between species

Dry Matter 6 7 . 5 $ 0 . 4 6 9 , 3 1 1 . 7 NS Gross Energy 69 . I j , 0 .9 6 9 . 9 1 1 . 2 NS Crude Protein 66.5 314. I 7 6 . 6 k 5 . 4 P =: 0 .0 1 Crude Fat 66 .5*3 .5 5 5 . 5 k3.6 P<0.05 N-free Extracts 8 4 . 3 k 1 . 1 8 1 , 3 1 2 . 2 NS Ash 44 .5 * 4 . 8 53.815.9 P<O.Ol Crude Fibre 54 .2: t2 .7 52 .O 1.6 .5 NS Cellulose 3 9 . 7 1 3 . 7 64.3 316.7 P<O.Ot

Discussion If one compares voluntary DM intake of ruminants in terms of percentage of body weight (Table 4), dikdik and suni consumed high amounts of food in spite of the high energy density of lucerne hay leaves. Similarly, in nature dikdik seem to select high energy food: gross energy of rumen contents of five Kirk’s dikdik in the dry season was between 4.0 and 4.9 Cal/g DM (Hoppe, unpublished). On the other hand rumen capacity of dikdik was surprisingly small, Total fresh rumen contents of six dikdik shot after peak feeding activity accounted for only 6.7% of body weight, and total rumen DM content was 57 g. This equals approximately one third of their pre- sumptive daily intake. In four suni shot during the dry season, total rumen fresh contents accounted for 8.5 % of body weight, and total DM content was 48 g (Hoppe, unpublished). Thus, rumen capacity of suni seems to be small as well. Regarding the requirement of suni and dikdik for large amounts of food, i t can therefore be postu- kited that in these species which are typical concentrate selectors, digestion proceeds at a faster rate than in predominantly grazing ruminants. This suggestion has already been brought up earlier by Hofmann (1973) who found that their rumens lack the strong muscular pillars which are typical for domestic ruminants.

In these experiments, the DM intake is likely to be less than in free-living suni and dikdik which have a higher energy requirement due to climatic factors, their greater activity and state of productivity like growth, pregnancy and lactation. Although only one young and one pregnant/lactating dikdik were used in the experiments, it is obvious that animals of this age and state of production, respectively would be most affected by scarcity of quality food.


Table 4. Voluntary intake and digestibility of dry matter in some East African bovids

D M intake/day Apparent Type

of body weight of D.M. food* Reference as percentage digestibility of

Boran zebu cattle (Eos indicus)

Eland ( Taurotragus o r w )

Coke's hartebeest (Alcephalus buselaphus. cokei) Oryx (Oryx beisa)

Wildebeest (Conno- chaetes taurinus) Grant's gazelle tCazeIla granti) Thomson's gazelle (Carella thomsoizi)

Grey Duiker (Sylvicapra grimmi )

Turkana goat (Capra hircus) Fat-tailed sheep (Ovis aries)

(Kirk's dikdik (Madoqua kirki) Suni

1 . 6

2 '0-2 3

2 6-3 '4

2'3-2.8

2 . 2

2.1

2 . 2

2.65

3.7-4.7

3.5-4.0

3 . 0

2 . 6

3 '2-4 '4

3 . 8

3 . 5

49

66

52

62

61

64

79

80

55

62

67

65

64

68

69

B Taylor & Maloiy, 1967

A Arman& Hopcraft,

I975 A Arman&

Hopcraft, 1975

A Arman & Hopcraft,

1975 B Taylor &

Maloiy, 1967 B Taylor &



Maloiy, 1967 A Arman&

Hopcraft, 1975

A Arman & Hopcraft,

1975 B Taylor &


Maloiy, 1967 A Arman &

Hopcraft, 1975

C This study

C This study (Nesotragus moschatus)

*Food A: ground mixed grainjcottonseed cake, pelleted; crude protein content between 6 .S and 13 . S o / , of DM. Food B: low quality hay, appr. 5 % crude protein. Food C: lucerne hay leaves.

Water consumption Although both species are reputedly independent of drinking water it was striking to see that suni drank almost twice as much as dikdik on the same food. Dikdik drank comparable quantities as desert-adapted Bedouin goats (84-122 ml/kgo76) kept at 30°C (Shkolnik, Borut & Choshniak, 1972). Since indispensable water loss frorh the skin and respiratory tract can be taken to be similar in both suni and dikdik under the described experimental conditions, the higher water intake of suni must be attributed to a higher urinary and/or faecal water loss. In these experiments, faecal water losses

46 Peter P. Hoppe

were the same in both species, with a mean of 114 ml of water excreted/lOO g faecal dry matter. Therefore, the water lost via the urine must be regarded as the most important single factor determining the water intake. Osmolality of dikdik and suni urine (twelve and twenty-six random samples, respectively) was 1859 f 350 and 1236 + 370 mOsm/kg urine, respectively. This equalled a plasma : urine osmolality ratio of I : 6 and 1 : 4, respectively. Therefore, the suni used about 80% more water per kg metabolic weight to excrete a similar amount of solutes than the dikdik. In spite of the fact that suni had to excrete a slightly higher solute load via the kidney because they digested minerals and protein significantly better than dikdik, urine osmolality was markedly lower in suni than in dikdik. This may be due to a genuine difference in urine concentration capability. From these results it cannot be concluded that free- living suni need a similarly high quantity of water, because nothing is known about the chemical composition of their natural diet. f would expect, however, that their nitrogen intake and, therefore, the amount of water necessary for the excretion of urea is less than when fed a protein concentrate like lucerne hay leaves. On the other hand, I have repeatedly observed suni drink (Thego River, W. Mount Kenya) which suggests that they make use of water as long as it is available. In areas like coastal bush where there is no surface water for several months, the abundance of fruits may contribute considerably to their water intake (G. Rathbun, pers. comm.). It seems likely that, in the absence of drinking water, a day’s diet of dicotyledonous plants and fruits contains between 727; and 82% water to meet their water requirement of 3-6 times the DM intake. Dikdik have an outstanding water economy. Maloiy (1973) found that plasma : urine osmolality ratios range between I : 5 and 1 : 7 in dikdik given lucerne hay and water ad libitum, and can increase up to I : 11 in dehydrated dikdik at high ambient temperature. They void very dry faeces which contain even less water (minimum of 68 ml per 100 g faecal DM during dehydration, Maloiy, 1973) than the faecal pellets of dehydrated camels (Schmid-Nielsen, 1964). When dikdik be- come hyperthermic they dissipate heat by panting. During panting, metabolic rate can be reduced by up to 40% of the resting metabolic rate (Hoppe et al., 1975) thereby minimizing endogenous heat production and evaporative water loss from expired air. It is not surprising, therefore, that their water turnover at thermoneutrality as measured by tritium-labelled water is very low (Hoppe, unpublished ). This water economy permits dikdik to rely on plant-borne water alone and explains why they can even afford to eat a certain amount of dry leaves as has frequently been observed. Their selected food including Solanum incanum fruits and wild figs (Hendrichs & Hendrichs, 197 I ) should provide ample water to make them completely independent of drinking water. Only rarely have they been observed spending some time in the early morning licking off dew drops from plants (E. Beutler, pers. comm.). To satisfy their water demand of 1.8 times the DM intake, the diet should contain about 65% water. According to Macfarlane rt al. (1966) sheep do not need to drink when the pasture carries 60-70x of water.

Digestibility of lucerne hay leaves The results of crude protein digestibility must be interpreted with caution because the faeces collected for analysis were contaminated with urine. The faecal pellets of both species are so small that they obstructed the fine floor grid which was used initially for the separation of faeces and urine. The grid was finally removed resulting


in mixing of urine and faeces, although most of the urine usually drained from the floor shortly after micturation. The results of protein digestibility therefore are under- estimates of the actual values.

DM digestibility (dikdik 67 %, suni 69%) and gross energy digestibility (dikdik 69%, suni 70%) were moderate considering the excellent nutritional quality of the food. This finding is in contrast to the contention of Hofmann and Musangi (1973) that digestibility coefficients between 78 % and 85 % are characteristic for browsing ruminants. Their results were obtained by using a regression equation derived from experiments on sheep which permits the calculation of organic matter digestibility from the N- concentration in faecal organic matter. However, marked differences in the anatomy (Hofmann and Stewart, 1972) and function of the forestomachs and intestines between the browsing/grazing sheep and the exclusively browsing dikdik and suni such as food retention time and microbial fermentation time in the rumen, invalidate this experimental approach. As far as these small browsers are concerned, indications are that they meet their high energy demand not by means of an excep- tionally good overall digestibility but the rapid fermentation of easily fermentable food components in combination with a particular pattern of food intake and gastro- intestinal motility (Hoppe and Gwynne, unpublished).

Suni which between experiments ate more lucerne stems than dikdik which strictly selected leaves, digested cellulose significantly (P<O.OI) better than dikdik. However there was no significant difference in fibre digestion, a finding which cannot yet be explained. The relatively low digestibility of cellulose for suni and particularly dikdik compares well with the very low cellulolytic microbial activity found i n the rumen of another browsing antelope, the greater kudu (Tragelaphus strepsicerus Pallas, 1766) in South Africa (van Gylswyk and Giesecke, 1973). In contrast, cellulolytic activity in the rumen of grazing antelope species was up to 10 times higher. It would be interesting to see if by measuring rumen microbial fermentation rate in vitro the suggested strategy of rumen digestion of suni and dikdik can be demon- strated.

Acknowledgments I wish to thank George Waweru, Simon Mungai and John Kanja for their help in carrying out some of the experiments. The food analyses were kindly done by the Nutrition Unit of the Dept. of Animal Production, and the cellulose analyses by R. N. B. Kay. I am grateful to the Kenya Game Dept. for permission to capture the dikdik and suni. The work was financed by a Dean’s Committee Research Grant from the University of Nairobi which I gratefully acknowledge. I wish to thank W. Richards for his helpful criticism of the manuscript.

References A.O.A.C. (1960) Oficiul Methods of Analysis (9th ed.) Association of Official Agricultural Chemists,

Washington, D.C. ARMAN, PAMELA AND HOPCRAFT, D. (1975) Nutritional studies on East African herbivores. 1. Digesti-

bilities of dry matter, crude fibre and crude protein in antelope, cattle and sheep. Brit. J . Nutr. 33,255-264.

ARMAN, PAMELA, HOPCRAFT, D. & MCDONALD, 1. (1975). Nutritional studies on East African herbivores. 2. Losses of nitrogen in the faeces. Brit. J. Nurr. 33, 265-276.

HALTENORTH, TH. (1963) Klassifizierung der Saugetiere: Artiodactyla I (18), 1-1 67. In: Kiiken- thuls Hunrlhrcch der Zoologie. 8 Bd, de Gruyter, Berlin.

48 Prtrr P . Hoppe

HENDRICHS, H. & HENDRICHS, U. (1971) Dikdik und Elephanten. Piper, Munchen. HOFMANN, R. (1973) The Ruminant Stomach. Stomach Structure and Feeding Hahits of E.A. Game

HOFMANN, R . & MUSANGI, R.S. (1973) Comparative digestibility coefficients of domestic and game

HOFMANN, R. & STEWART, D.R.M. (1972) Grazer or browser: a classification based on the stomach

HOITE, P.P., JOHANSEN, K., MALOIY, G.M.O. and MUSEWE, V. (1975) Thermal panting reduces oxy-

HUNGATE, R.E. (1959) Microbial termentation in certain mammals. Amner. Ass. fbr the Advancement

LYDEKKER, R. ( 1926) The Gumr Animals of Africu. Rowland Ward, London. MAWARLANE, W.V., DOLLING, C.S.H. & HOWARD, B. (1966) Distribution and turnover of water

in merino sheep selected for high wool production. Ausfr. J . agric. Reg. 17, 491-502. MACFARLANE, W.V., HOWARD, B., MALOIY, G.M.O. & HOPCRAFT, D. (1972) Tritiated water in field

experiments of ruminant metabolism in Africa. Symposium on the use of isotopes in studies on the physiology of’ domestic unimuls, International Atomic EnerKy Agency, Vienna. Athens, Greece.

MALOIY, G.M.O. (1973). The water metabolism of a small East African antelope, the dikdik. Proc. R. Soc. Lond. B. 184, 167-1 78.

SCHMID-NIELSEN, K. (1964) Desert Animals; Physiological Problems of Heat and Wufer. Clarendon Press, Oxford.

SctioEN, A. (1972) Studies on the environmental physiology of a semi-desert antelope, the dikdik. E.A. uxric. forest. J . 40, 325-330.

SHKOLNIK, A, , BORLJT, A. & CIIOSHNIAK, J. (1972). Water economy of the Bedouin goat. Symp. Zoo/.

TAYLOR, C.R. & MALOIY, G.M.O. (1967). The effect of dehydration and heat stress on intake and digestion of food in some East African bovids. Trans. 8th int. Congr. Came Bid . , Helsinki.

TINLEY, K.L. (1969) Dikdik Madoqua kirki in South West Africa. Notes on distribution, ecology and behaviour. Madoqua. 1, 7-33.

V A N GYLSWYK, N.O. & GIESECKE, D. (1973) A summary of preliminary findings in a rumen micro- biological investigation on wild ruminants. Koedoe, 16, 191-194.

Ruminants. East African Literature Bureau, Nairobi.

ruminants from marginal land in East Africa. Bull. epiz. dis. Africa. 21, 385-388.

structure and feeding habits of East African ruminants. Mammalia, 36, 226-240.

gen uptake in the dikdik, Rhynchofragus kirki. Acra physiol. Scand. 95,9 A.

of Science, 130, 1192-1 194.

Sac. Land. 31, 229-242.

(Manuscript received 24 June 1975)

comparison of voluntary food and water consumption and digestion in kirk's dikdik and suni

Documents