W.L. Jansen

Jagran Zoological Research & Development, Eemnesserweg 109B, Hilversum, The Netherlands.

Several experiments are done by the author in which horses or ponies were given extra fat in the form of soybean oil (Jansen, 2001). In all experiments apparent total tract digestibility of crude fibre declined. Similar reductions were seen for the digestibility of neutral and acid detergent fibre. The results of the experiments were pooled. In order to trace out other dietary components that could be related to fibre digestibility, a multiple regression analysis (Wilkinson, 1990) was performed with horses, experiments, periods and dietary crude fibre, crude protein, crude fat content as factors. By adding and deleting the factors the percentage of variance accounted for by the factors to the regression model could be analyzed. The factors fat content and experiment were statistically significant (P<0.001) in the model, when crude fibre or neutral and acid detergent fibre, hemicellulose and cellulose digestibility were chosen as the response variable. The factors horse, periods, crude fibre, crude protein and crude fat content were not statistically significant (P > 0.10). The addition of 10 g fat per kg dry matter at the expense of an iso-energetic amount non-structural carbohydrates lowered the total tract digestibility of crude fibre by 1.0 percent unit (Table 1). Thus, for a high-fat diet, the digestible energy from components rich in crude fibre may be overestimated when calculating the energy content of the diet on the basis of feedstuff tables. An increase of soybean oil by 10 g per kg DM increased apparent fat digestibility by 2.5 percent units (Table 1). An increase in fat intake will raise the amount of faecal fat of dietary origin and thus would lower the proportion of endogenous fat in the faeces. By comparing the low-fat diets without added soybean oil and the high-fat diets with soybean oil the digestibility of soybean oil could be estimated as 74.6 ± 14.9% (mean ± SD, n=42). This digestibility is about one fifth (20 percentage units) lower than used in the Dutch net energy system (CVB 2000). The net energy content of soybean oil would thus be overestimated with 5 MJ/ kg product. Other studies in which crude fat digestibility was measured show wide variation in the outcome. The results probably are related to macronutrients other than fat intake. This was the reason to subject the data on digestibility of crude fat digestibility to multiple regression (Wilkinson, 1990) with horse, experiment, period, crude fat, crude protein and crude fibre content as factors. The multiple regression showed that crude fibre content significantly (P<0.0001) diminished crude fat digestibility.
Refere nces:
Anon (1996) Het definitieve VEP- en VREp-systeem. CVB-documentatierapport nr. 15. CVB (2000) Veevoedertabel. Centraal Veevoederbureau, Lelystad, The Netherlands. Jansen, W.L. (2001) Fat intake and apparent digestibility of fibre in horses and ponies. PhD thesis, University of Utrecht, the Netherlands Wilkinson, L. (1990) Systat: The system for statistics. Systat Inc. Evan


M.J.R. Jordan* and A.M. Chestnutt

Animal Management Section, Sparsholt College Hampshire, Sparsholt, Winchester, Hants. SO21 2NF.#

The science of nutrition is a rapidly expanding field of zoo biology and the formulation of appropriate and scientifically based diets has become an integral part of the husbandry of many species. Yet for most free ranging mammals and birds information on their wild diets still remains unclear and often only the most generalised information exists. Even when information exists it frequently represents an approximation of the ‘average’ diet, ignoring the intricacies of habitat, seasonal, sex or even age specific variations. Ecologically, an accurate knowledge of species’ diets, and variations within them, contributes greatly to our understanding of biodiversity and competition between species. More critically, knowledge of accurate diets contributes to formulating strategies for wildlife conservation and management and allows husbandry to be specifically refined. There are a variety of techniques that can be used to ascertain the diets of wild mammals and birds, from direct observation of items being eaten, or the identification of feeding remains through to digestive tract and faecal analysis. The limitations of such wildlife forensic work depend upon the level of diagnosis possible for a variety of feeding remains. In carnivores the identification of hair, feather, bone and invertebrate fragments can all yield information on items consumed whilst in herbivores plant cell fragments, seeds or pollen can be used to derive similar information. The quantification of dietary information can be confusing and issues and considerations in understanding results will be discussed. For a number of techniques detailed captive studies have the potential to greatly refine and quantify analyses by allowing the correlation of information collected in the field.


K. Leus1*, A. A Macdonald2, G. Goodall3 , D. Veitch3, S. Mitchell2 and L. Bauwens1

1Royal Zoological Society of Antwerp, Koningin Astridplein 26, 2018 Antwerp, Belgium; 2Department of Preclinical Veterinary Sciences, Royal (Dick) School of Veterinary Studies, The University of Edinburgh, Summerhall, Edinburgh EH9 1QH, Scotland, UK; 3Department of Veterinary Pathology, Royal (Dick) School of Veterinary Studies, The University of Edinburgh, Summerhall, Edinburgh H9 1QH, Scotland, UK.

The stomach of the babirusa differs from that of other pigs in several aspects (Leus et al., 1999; Agungpriyono et al., 2000): it is larger and possesses a large diverticulum ventriculi, the gastric glands are confined to a small section at the end of the corpus ventriculi, the cardiac glands occupy a much larger surface area within the stomach (>70% v. ~30% in the domestic pig) and there are some variations in the distribution of endocrine cells. The pH in the lumen of the cardiac gland region was deemed suitable for the survival of the micro-organisms found therein. It was hypothesised that the babirusa is a non-ruminant foregut fermenting frugivore/concentrate selector (Leus et al., 1999). Previous studies of the stomach have concentrated on the gross anatomical and light microscopic structure, largely making use of museum material or specimens not collected immediately after death (Langer, 1973, 1988; Leus et al., 1999). In the mean time, we have been able to obtain stomach tissues from nine babirusa that were being euthanised for veterinary reasons not related to the gastrointestinal tract. The quality of tissue fixation thereby permitted scanning and transmission electron microscopic investigations. Scanning electron microscopy revealed that the surface of the whole of the cardiac gland region was typically characterised by a honeycomb pattern. The entrance to each “honeycomb cell” had a diameter of approximately 0.06 mm. The walls of the honeycomb had a very granular appearance and extended into the stomach lumen a distance of about 0.2 mm in height above the glandular epithelium. At higher magnification the walls appeared to be almost entirely composed of a varied bacterial microflora. Light microscopy and transmission electron microscopy showed that at the luminal border of the cardiac gland epithelium, on top of the ridges between each glandular pit, non-glandular cellular tissue extended ribbon-like into the lumen. Sheets of squamous epithelial-like cells formed thin tube-like structures extending the lumen of the glandular pit. The surfaces of these sheets were covered by a dense layer of mixed gram negative and gram positive bacteria. No histological study of the babirusa stomach has yet drawn attention to anything like this honeycomb structure. No evidence of a similar structure could be found in the published histological studies of the domestic pig, other suids, or indeed the cardiac glands of the forestomach of other mammals. Further studies of fresh tissues of animals such as macropod marsupials, colobine monkeys, peccaries, camelids and sloths, which have larger areas of cardia or mucogenic glands in their stomachs, have been initiated to investigate the uniqueness of this honey comb structure. The latter's function also remains to be explored. Possible hypotheses include surface enlargement in order to increase attachment space and retention time of bacteria in a stomach without strong compartmentalisation and/or to increase the area for absorption of fermentation products. The explanation may have a direct consequence for the feeding requirements of babirusa in zoological gardens, which in turn, may be a key factor for the success of its conservation breeding program.
References: Agunpriyono S., Macdonald A.A., Leus K.Y.G., Kitamura N., Adnyane I.K.M., Goodall G.P., Hondo E., & Yamada J. (2000) Immunohistochemical study on the distribution of endocrine cells in the gastrointestinal tract of the babirusa, Babyrousa babyrussa (Suidae). Anat. Histol. Embryol. 29:173-178. Langer P. (1973) Vergleichend-anatomische Untersuchungen am Magen der Artiodactyla (Owen, 1848). I. Teil: Untersuchungen am Magen der Nonruminantia (Suiformes). egenbaurs morph. Jahrb. 119: 514-561. Langer P. (1988) The mammalian herbivore stomach, comparative anatomy, function and evolution. Gustav Fischer: Stuttgart, New York. Leus K., Goodall G.P. & Macdonald A.A. (1999) Anatomy and histology of the babirusa (Babyrousa babyrussa) stomach. C.R. Acad. Sci. Paris, Sciences de la vie / Life Sciences 322:1081-1092.


M. Gore* and J. Cook

Animal Conservation and Research Dept., Royal Zoological Society of Scotland, Murrayfield, Edinburgh EH12 6TS, Scotland

Research on diets of domestic and laboratory animals has provided diets suitable for the requirements of these animals. The diets for wildlife zoo animals are often extrapolated from domestic or laboratory diets of species closest to the wildlife species in question. This situation does not provide a satisfactory diet for wildlife zoo animals, which have a different purpose to, and are often in much more individual settings than, production animals. The resulting diet could affect the wildlife zoo animal’s health and reproduction. For example, red deer and grey seals have been shown to produce more male than female offspring when the breeding females have abundant food resources and are in good bodily condition. Arabian oryx at Edinburgh Zoo produce predominantly male offspring, and their diet has since been estimated to be too energy rich for the species needs. The present study addressed the problem for a range of wildlife zoo mammals by examining current models that predict basal metabolic rate (BMR). The value of an animal’s BMR can be used to calculate the energy requirements of a diet. Body mass is the most important characteristic of an individual. The best measure of the direct influence of body mass on BMR in mammals is found in groups of species that are physiologically and ecologically uniform. That is, BMR can be more accurately estimated when species are grouped by food habits than by order or class. Wildlife zoo animals represent not only a wide variety of foraging strategies, but within these are a number of different food types ingested. For instance, wildlife zoo carnivores include those that eat largely invertebrates. Wildlife zoo mammals in the present study were categorised for both body mass and food types. In the present study, BMRs for a wide variety of wildlife zoo mammals representing a number of specific food habits was calculated. The results were then compared with current models. BMRs were generally lower than expected by the current models. The results indicated that zoo diets provided considerably more energy than required. A more accurate BMR value was achieved using the corrected model. Results showed that large grazers and vertebrate eaters compared well with predictions from current models although corrected values were lower. For other categories of food habits, including arboreal frugivores and folivores, insectivores, terrestrial folivores and fossorials, the values of the current models were much greater in magnitude. From the present study, it was concluded that the corrected method to quantify BMR was proved to be satisfactory through testing and provided a means to estimate energy requirement for wildlife zoo mammal diets. The companion paper will illustrate how these results can be used practically in calculating energy requirements and activity levels in the diets of wildlife zoo mammals.


J. Cook* and M. Gore

Animal Conservation and Research Dept., Royal Zoological Society of Scotland, Murrayfield,
Edinburgh EH12 6TS, Scotland

There are many possible contributing factors to metabolic rate, many of which remain unknown. Emmans (1997) developed a generic equation for domestic mammals, based on basal metabolic rate (BMR) in domestic mammals. The equation expresses the rate at which an animal would produce eat from its own fat store at an arbitrary activity level in a thermoneutral environment. This clearly needs to be adjusted for practical use in wildlife species. To obtain basic energy requirements for BMR, knowledge of the principal food type exploited by the species is necessary and an estimated weight. Physical activity is a major factor causing differences in the energy budget between zoo animals, as activity is variable among individuals and zoo collections. We demonstrate a simple method to provide this information. These three values are then simply entered into the equation, as we show with a range of worked examples. Data were collected on the BMRs of exotic zoo species representating three major food habits, carnivory and large grazers and omnivores. In addition, average weights for a wide variety of species within these two food habits were collected. This allowed us to develop the original equation for each category to allow a more accurate value of BMR to be predicted for captive wildlife rather than domestic species. We have investigated the activity budgets of representatives of the three food habits in our collection at Edinburgh Zoo. We obtained elocities to calculate the corresponding increase in metabolic rate for these individual animals. Observations were made on the individuals and the data were compared with a keeper survey. The results showed a significant correlation between methods. We were then able to categorise velocities into distinct levels for practical use. The results of our study showed that the BMR provided by the Emmans equation can simply be corrected by a certain percentage related to the general activity level of an animal, the food habit and weight. The results provide energy requirement for maintenance at a given level for individual animals. Zoo animals are kept in groups and we discuss how to calculate this. Ultimately, our simple recommendations estimate more precisely than previous methods the energy requirements for zoo animals. This allows feeding regimes to be optimised, thus improving the nutrition of captive species such as are found in zoo collections today. This paper illustrates, through worked examples for a variety of species, that this new method proves accurate and simple to use.