On-line version ISSN 1996-7489
S. Afr. j. sci. vol.111 n.11-12 Pretoria Nov./Dec. 2015
Jamel Ben Mimoun; Saïd Nouira
Laboratory of Animal Ecology, Department of Biology, Faculty of Science, Tunis El Manar University, Tunis, Tunisia
A micro-histological analysis of composite faecal samples was used to determine the food plants consumed by the aoudad Ammotragus lervia (Barbary sheep) in the mountains of Bou Hedma National Park, as well as to study food preferences and seasonal variation of the diet of this ungulate. A total of 19 plant species was identified in the faecal samples: 8 grasses, 6 browse types and 5 forbs. The annual diet of this ungulate is composed of approximately 67% grasses, 17% browse and 16% forbs. Two grass species, Stipa parviflora and Stipa tenacissima, comprised 97% of grasses consumed and 63% of the annual diet, and were eaten in the year with considerably greater frequency than any other plant species. The diet of the aoudad at Bou Hedma National Park showed a seasonal variation. For example, grasses were consumed the entire year with a preference in spring (72.57%) and summer (78.31%), but their occurrence in the diet dropped in winter (38.7%). Browse was eaten preferentially in autumn (21.86%) and forbs were utilised most during winter (43.22%) and least during summer (7.53%). Our results show the plasticity of the Tunisian aoudad's diet and identify it predominantly as a grazer.
Keywords: Barbary sheep; food habits; north Africa; seasonal variation; ungulate
The Barbary sheep or aoudad (Ammotragus lervia) is a bovid within the Tribe Caprini (sheep and goats) and is endemic to North Africa.1 It may be found in any rugged terrain or mountain chain throughout the northern part of the continent.2-6 Its distribution within the Sahara region extends 'south about 14° near the Niger and through N. Tchad and Sudan to the Red Sea'.6,7 The aoudad has become rare and has been locally extirpated from its original range as a result of human encroachment, overgrazing and constant hunting.8,9
In Tunisia, the aoudad is the only representative of wild caprinids and it is considered the flagship species of Tunisian wildlife. According to the Red List of Threatened Species of the International Union for Conservation of Nature (IUCN), the aoudad is classified as a globally 'vulnerable' species.6 It has disappeared from most of its original environment and was considered by Schomber and Kock to be on the verge of extinction in Tunisia as a result of overhunting and a reduction in its habitat. This species is considered in Tunisia by the National Register of Wild Species as 'critically endangered'. Indeed, the Barbary sheep is exposed to threats that may cause its extinction in the wilderness such as overhunting for its meat and overgrazing of its habitat by livestock.11 The National Park of Bou Hedma is one of the country's biotopes that has succeeded in rehabilitating a very small number of aoudad. In a report of the General Directorate of Forests,12 the population density of the aoudad in the park was estimated at about 50 individuals.
In its native habitat in northern Africa, the aoudad is largely unknown apart from the little research done regarding its distribution.3,4,9,11 The social organisation of the aoudad population of Chambi in Tunisia was recently studied by Ben Mimoun et al.13 The study of diet is a necessary step in understanding the eco-biology of herbivorous animals. Indeed, besides its applicability in quantifying the food ingested by animals, it can also contribute to effective management and conservation of endangered herbivores such as the aoudad. No sophisticated food habit studies have been made on this species in North Africa, but general references note that both browse and herbaceous vegetation are eaten.14,15 In Tunisia, the diet of this endangered North African wild caprid has never been studied.
Our objective was to study the feeding behaviour of the aoudad occupying the mountains of the Bou Hedma National Park in central Tunisia. The knowledge of the trophic regime of this ungulate is an opportunity to increase our understanding of the feeding behaviour and the requirements of this species.
Material and methods
The feeding behaviour of the aoudad in Tunisia was determined for a wild population of this species occupying the mountain steppe of Bou Hedma National Park (16 448 ha) of Sidi Bouzid in the centre of the country (Figure 1). The habitat of the National Park is represented by the Djebel of Bou Hedma, culminating at 790 m, and is characterised by vegetation including Lygeum spartum, Stipa tenacissima, Stipa parviflora, Artemisia herba alba, Pistacia lentiscus, Retama raetam and Acacia tortillis. This habitat is located in a bio-climate of lower semi-arid to lower arid, with an annual average rainfall of about 150 mm.3
Sample collection and analysis
Monthly collections of fresh faecal pellets of the aoudad were made between March 2009 and February 2010 at Bou Hedma National Park. Samples were collected in an opportunistic manner throughout the mountain of the park. A total of 180 piles of droppings was collected, with an average of 15 piles of droppings per month. Within each pile, an average of five fresh droppings was randomly collected. Droppings were harvested and preserved in 95% ethanol in screw-top jars labelled with the month and locality of collection. For each month, a composite sample was prepared by combining the five fresh faeces from the 15 piles.16,17
Composite sample analysis was performed in the Animal Ecology Laboratory at The Arid Land Institute, Medenine Tunisia as described in Simpson et al.18 This method was also applied by Wagner et al.17 for bighorn sheep (Ovis canadensis) in North America. A reference atlas of the epidermis of the harvested plants in the habitat of the aoudad was prepared by applying the experimental protocol followed by Sparks et al.19 Reference material and faecal samples were prepared for analysis following the micro-histological slide preparation technique described by Free et al.20 Examination and identification procedures followed Sparks et al.19 As indicated by Williams21 and Storr22, no staining was performed for microscopic observation. Four slides were prepared for each composite sample. About 275 epidermal fragments were recognised for each composite sample by systematic scanning of the blades.23
where ni is the total number of fragments of the species (i) identified and Ni is the total number of fragments counted.
The average frequencies (AF%) provide a good estimate of the diet by indicating the range of plant species collected by the animal from its habitat and they also show food use within the year. Species with an AF<1.0 were not a major part of the annual diet and were classified as trace foods.18 Non-parametric Kruskal-Wallis tests24 were applied to compare average frequencies of the different forage classes and plant species consumed by the aoudad; p<0.05 was considered significant.
Average diet composition
The food spectrum of wild aoudad in Bou Hedma National Park comprised 19 plant species. The average annual diet of the aoudad sampled in this study was found to consist of 67.12% grasses, 16.70% browse and 15.81% forbs (Table 1). Unidentified species represented only 2.36% of the diet.
The principal forage plants identified in the diet of aoudad in Bou Hedma National Park throughout the year were two grass species (Stipa parviflora and Stipa tenacissima) four browse species (Periploca leavigata, Juniperus phoenicea, Pistacia lentiscus and Retama raetam) and two forbs (Helianthemum kahiricum andGlobularia alypum) (Table 1).
Seasonal diet variation
The diet of the aoudad at Bou Hedma National Park shows a seasonal variation. The spring food spectrum is the largest with 14 plant species in comparison to 13 in autumn, 12 in winter and only 9 species in summer (Table 1).
There was no significant difference in the consumption of each forage class among the four seasons for grasses (K=1.44; df=3; p>0.05), for browse (K=2.43; df=3; p>0.05) and for forbs (K=3.16; df=3; p>0.05). However, interaction between the effects of season and forage class was identified (K=6.57; df=2; p=0.037). Indeed, the aoudad shows a higher use of grasses during the spring (72.57%), summer (78.31%) and autumn (58%), than in winter when the diet is concentrated on forbs (43.22%). Furthermore, a net increase in the consumption of browse was observed in autumn (21.86%) (Table 1).
Considering each forage class separately, the Kruskal-Wallis test indicated a significant difference in the consumption of grass species by the aoudad in spring (K=23.59; df=7; p=0.002), summer (K=22.86; df=7; p=0.002), autumn (K=22.69; df=7; p=0.002) and winter (K=21.26; df=7; p=0.003). In fact, eight species of grasses were consumed in the National Park, but only three species - Stipa tenacissima, Stipa parviflora and Ampelodesma mauritanica - had an AF of >1.0 in the annual diet. The genus of Stipa alone represents 97% of grasses consumed and 63.22% of the annual diet of animals, and was eaten throughout the year with considerably greater frequency than that of any other species (Table 1).
No significant seasonal variation was observed in the consumption of browse species (K=10.82; df=5; p>0.05), but during each season the aoudad showed a preference for only a few species: in spring (K=14.07; df=5; p=0.015), summer (K=15.54; df=5; p=0.008), autumn (K=12.73; df=5; p=0.026) and winter (K=13.87; df=5; p=0.016). Six browse species were eaten by the aoudad in Bou Hedma National Park: a single tree species (Olea europea) and five shrubs (Periploca laevigata, Juniperus phoenicea, Pistacia lentiscus, Retama raetam and Salsola longifolia). Among this forage class, only two species - Periploca laevigata and Juniperus phoenicea - were eaten by the aoudad during the year with significant relative frequency. The main browse species used by the aoudad was Periploca leavigata which occurred the most in the samples of droppings analysed in spring (7.37%) (Table 1).
No significant seasonal variation was observed in the consumption of forbs (K=9.36; df=4; p>0.05) (Table 1). A clear seasonal preference for certain species of forbs was shown: spring (K=11.22; df=4; p=0.024); summer (K=11.12; df=4; p=0.025); autumn (K=11.88; df=4; p=0.018); and winter (K=11.44; df=4; p=0.022). Indeed, five forb species were identified in the diet of the aoudad in the Bou Hedma National Park but only two species, Helianthemum kahiricum and Globularia alypum, appeared during all seasons with significant relative frequencies compared to the rest of the plants in this forage class. Helianthemum kahiricum of the Cistaceae family was consumed the most throughout the year, except in summer. The consumption of the second species, Globularia alypum, increased markedly in winter (14.57%) (Table 1).
We have shown that the aoudad in Tunisia consumes different types of plants including grasses, forbs, browse, trees and shrubs. These results reveal that the aoudad is a generalist herbivore with a very flexible diet.25-27 The number of plants used by the aoudad in Bou Hedma is very small (19 plant species) compared to that of Barbary sheep that have been introduced elsewhere. In the Canary Islands, 41 taxa were identified28 and in the Guadalupe Mountains of New Mexico, the annual diet of this ungulate includes 74 plant species.18 However, the limited number of plants constituting the diet of the aoudad at Bou Hedma is more comparable with results obtained in Morocco29 and in Palo Duro Canyon, Texas, in which a total of 25 food plants was identified.30
This variation in the number of plants used by the aoudad between these regions can be related to differences in the specific richness of vegetation available. The scarcity of plant species constituting the diet of the Tunisian aoudad could potentially be associated with plant identification problems from faecal samples and the apparent ease of digestibility of tender grass and succulent forbs.31 In Bou Hedma National Park, the aoudad shows a clear preference for grasses (67% of the annual diet), followed by browse (17%) and forbs (16%), which allows this ungulate to be identified as a grazer.31 This preference for grasses can be explained by the abundance of these plants in the mountain steppe of Bou Hedma National Park.32 Our results concur with those obtained for this ungulate in the Edwards Plateau region of Texas where the annual diet consisted of 61% grasses, 21% browse and 18% forbs31, and in California where grass classes represented 97% of the annual diet.33 Furthermore, a relatively high level of grasses (42%) in the diet of aoudads was observed by Ogren34 in the Canadian River Gorge of northeastern New Mexico. The low frequency of forbs in the diet of the aoudad at Bou Hedma National Park may be the result of the high pressure of overgrazing by domestic animals as demonstrated by Ogren34 in the Canadian River Gorge and by Simpson et al.27 in Palo Duro Canyon. Indeed, Ramsey et al.35 found that in pastures with low forbs availability, the aoudad preferred grasses to browse, which led the authors to surmise that this forage class was the alternative food source in overgrazed areas.
Nevertheless, our results differ from those obtained in Palo Duro Canyon, where the annual diet of the aoudad is composed of 70% browse, 20% forbs and only 10% grasses,30 and in the Guadalupe Mountains where the aoudad consumes 42% browse, 38% forbs and only 20% grasses.18 According to another study conducted in Texas,27 woody browse is the most important forage class in spring (53%), summer (68%) and autumn (48%), while browse (42%) and grasses (43%) are equally important during winter. Also, in the Eastern High Atlas National Park in Morocco, browse was found to be the major dietary component during winter (52.7%) and autumn (56.6%). Grasses and forbs represented 27.4% and 19.9% of the diet, respectively, in autumn and 21.0% and 23.4% of the diet, respectively, in winter.29 As has been demonstrated in the Guadalupe Mountains by Simpson et al.18, the class of grass in our study has a fairly high rate of consumption in summer and spring, and is at its lowest in winter. Evans36 also recorded low winter use of grasses in Palo Duro Canyon.
The high consumption of grass species during the breeding season, which tends to be focused in spring,25 can be explained by the fact that these plants reach a peak level of their protein content and digestibility during this period of the year.17 Moreover, the Tunisian aoudad tend to prefer open land13 with abundant grass species in the breeding season. The importance of the consumption of grasses with high nutritional value32 in summer can be justified by the increased requirements of the animals during this season, which is characterised by peak lamb drop and lactation.17 In our study, Stipa parviflora and Stipa tenacissima were the only major grass species consumed in all seasons with the highest consumption in spring and summer and the lowest in winter.
The genus Stipa was also among the grasses preferred by the aoudad in Largo Canyon, New Mexico37 and was observed in the diet of this ungulate in Morocco in northern Africa.29 The great affinity to this grass plant is explained by its high pastoral value.32The increased consumption of browse in the autumn can be explained by the reduced availability of herbaceous plants during this season and highlights the plasticity of the diet of this ungulate. Our study showed that the highest use of forbs was in winter and the lowest use was during summer, mirroring the results obtained by Simpson et al.18 in the Guadalupe Mountains, New Mexico.
This observation can be explained by the decrease in the availability of herbaceous plants in summer because of the effects of heat, drought and overgrazing by domestic animals in the mountains of Bou Hedma.32 The variation in the availability of plant resources in Bou Hedma National Park influenced the aoudad's choice of forage class, as was also demonstrated for this ungulate in the Edwards Plateau of Texas.33
We thank the forest directors for their support and for their role in facilitating access to Bou Hedma National Park, the park employees for their voluntary collaboration, and the members of the Animal Ecology Laboratory of Arid Land Institute, Medenine (Tunisia) who gave permission for the micro-histological faecal analyses.
J.B.M. was responsible for the experimental design, for preparing the samples and for writing the manuscript. S.N. performed some of the experiments and made conceptual contributions.
1. Brentejs B. The Barbary sheep in ancient North Africa. In: Simpson CD, editor. Symposium on ecology and management of Barbary sheep. Lubbock, TX: Texas Tech University Press; 1980. p. 25-26. [ Links ]
2. Gray GG. Status and distribution of Ammotragus lervia: A worldwide review. In: Hoefs M, editor. Wild sheep. Distribution, abundance, management and conservation of the sheep of the world and closely related mountain ungulates. Whitehorse, Canada: Northern Wild Sheep and Goat Council; 1985. p. 95-126. [ Links ]
3. Le Houérou HN. Outline of the biological history of the Sahara. J Arid Environ. 1992;22:3-30. [ Links ]
5. Shackleton DM, editor. Wild sheep and goats and their relatives: Status survey and conservation action plan for Caprinae. Gland: International Union for Conservation of Nature; 1997. [ Links ]
6. Cassinello J, Cuzin F, Jedeidi T, Masseti M, Nader I, De Smet K. Ammotragus lervia. In: International Union for Conservation of Nature (IUCN). IUCN Red List of Threatened Species. Version 2013.2 [document on the Internet]. c2008 [cited 2015 March 13]. http://dx.doi.org/10.2305/IUCN.UK.2008.RLTS.T1151A3288917.en [ Links ]
7. Corbet GB. The mammals of the palaearctic region: A taxonomic review. Ithaca, NY: Cornell University Press; 1978. p. 314. [ Links ]
8. Ansell WFH. Order Artiodactyla. In: Meester J, Setzer HW, editors. The mammals of Africa: An identification manual. Washington DC: Smithsonian Institution Press; 1971. p. 1-84. [ Links ]
9. De Smet K. Tunisia. In: Shackleton DM, editor. Wild sheep and goats and their relatives: Status survey and conservation action plan for Caprinae. Gland: International Union for Conservation of Nature; 1997. p. 45-47. [ Links ]
10. Schomber HW, Kock D. The wild life of Tunisia: Part 2. Some larger animals. Afr Wildlife. 1960;14:277-282. [ Links ]
11. Clarck JL. The great arc of the wild sheep. Norman, OK: University of Oklahoma Press; 1964. p. 247. [ Links ]
12. General Directorate of Forests (DGF). Strategie Nationale Tunisienne pour la conservation et la restauration des antílopes sahélo-sahariennes et leurs habitats (2001-2020) [Tunisian National Strategy for the Conservation and Restoration of Sahelo-Saharan antelopes and their habitats (2001-2020)]. Tunisia: Direction Générale des Forêts, Ministère de l'Agriculture Tunisie; 2001. p. 29. French. [ Links ]
13. Ben Mimoun J, Nouira S. Social organization of Barbary sheep (Ammotragus lervia) population in the Chambi National Park, Tunisia. Int J Biodivers Conserv. 2013;5(1):15-19. http://dx.doi.org/10.5897/IJBC11.104 [ Links ]
14. Brouin G. Contribution à l'étude de l'Aïr: Notes sur les ongulés du cercle d'Agadez et leur chasse. Institut Francais d'Afrique noire [Contribution to the study of the Aïr: Notes on ungulates of the circle of Agadez and their hunting. French Institute of Black Africa]. Mémoires. 1950;10:425-455. French. [ Links ]
15. Malbrant R. Faune du centre Africain Francais (Mammifères et oiseaux) [Fauna of the French African center (mammals and birds)]. In: Lechevalier P editor. Encyclopédie Biologique. 2nd ed. Paris; 1952. p. 59-61. French. [ Links ]
16. Hansen RM, Lucich GC. A field procedure and study design for fecal collections to be used to determine wildlife and livestock food habits. Fort Collins, CO: Composition Analysis Lab; 1978. p. 15. [ Links ]
17. Wagner GD, Peek JM. Bighorn sheep diet selection and forage quality in central Idaho. Northwest Sci. 2006;80(4):246-258. [ Links ]
18. Simpson CD, Krysl LJ, Dickinson TG. Food habits of Barbary sheep in the Guadalupe Mountains, New Mexico. In: Simpson CD, editor. Symposium on ecology and management of Barbary sheep. Lubbock, TX: Texas Tech University Press; 1980. p. 87-91. [ Links ]
22. Storr GM. Microscopic analysis of faeces, a technique for ascertaining the diet of herbivores mammals. Austral J Biol Sci. 1961;14:157-164. [ Links ]
23. Delaunay G. Contribution à la mise au point de méthodes de suivi des populations d'ongulés de hautes montagnes en milieu protégé: Etude sur le chamois dans le parc national des Ecrins [Contribution to the development of monitoring methods of ungulate populations of high mountains in protected areas: Study on the chamois in the Ecrins National Park] [thesis]. Rennes: Université de Rennes; 1982. p. 280. French. [ Links ]
24. Siegel S. Non-parametric statistics for the behavioral sciences. New York: McGraw-Hill; 1956. p. 312. [ Links ]
25. Cassinello J. Ammotragus lervia: A review on systematics, biology, ecology and distribution. Ann Zool Fenn. 1998;35:149-162. [ Links ]
26. Barrett RH, Beamson SL. Research needs for management of Barbary sheep in the future. In: Simpson CD, editors. Proceedings of the symposium on ecology and management of Barbary sheep; 1979 Nov 19-21; Texas, USA. Lubbock, TX: Texas Tech University Press; 1980. p.106-107. [ Links ]
27. Simpson CD, Krysl LJ, Hampy DB, Gray GG. The Barbary sheep: A threat to desert bighorn survival. Desert Bighorn Council Trans. 1978;22:26-31. [ Links ]
28. Pinero JCR, Luengo JLR. Autumn food habits of the Barbary sheep (Ammotragus lervia Pallas 1777) on La Palma Island (Canary Islands). Mammalia. 1992;56:385-392. [ Links ]
29. Hafidi MEM. Food habits and preferences of Barbary sheep (Ammotragus lervia) in the eastern High Atlas National Park [MSc thesis]. Morocco: International Centre for advanced Mediterranean Agronomic Studies, Mediterranean Agronomic Institute of Chania; 1996. [ Links ]
30. Krysl L, Simpson CD, Gray G. Dietary overlap of sympatric Barbary sheep and mule deer in Palo Duro Canyon, Texas. In: Simpson CD, editors. Symposium on ecology and management of Barbary sheep; 1979 Nov 19-21; Texas, USA. Lubbock, TX: Texas Tech University Press; 1980. p. 97-103. [ Links ]
31. Schaller GB. Mountain monarchs: Wild sheep and goats of the Himalaya. Chicago, IL: University of Chicago Press; 1977. p. 425. [ Links ]
32. Roselt/OSS. Etude de la biodiversité dans l'observatoire pilote de Haddej - Bou Hedma (TUNISIE) [Study of biodiversity in the pilot observatory Haddej - Bou Hedma (TUNISIA)]. CT no. 7. Montpellier: Collection ROSELT/ OSS; 2004. p. 110. French. [ Links ]
33. Ogren HA. Barbary sheep. New Mexico Department of Game and Fish Bulletin 13. Santa Fe, NM: New Mexico Department of Game and Fish; 1965. [ Links ]
34. Ogren HA. Barbary sheep in New Mexico. New Mexico Department of Game and Fish Bulletin 11. Santa Fe, NM: New Mexico Department of Game and Fish; 1962. p. 32. [ Links ]
35. Ramsey CW, Anderegg MJ. Food habits of an aoudad sheep, Ammotragus lervia (Bovidae), in the Edwards Plateau of Texas. Southwestern Nat. 1972;16:267-280. http://dx.doi.org/10.2307/3670063 [ Links ]
36. Evans PK. The aoudad sheep, an exotic introduced in the Palo Duro Canyon of Texas. Proceedings of the Southeastern Association of Game and Fish Commissioners Conference; 1947; New Orleans, USA. New Orleans: South-eastern Association of Game and Fish Commissioners; 1967. p. 183-188. [ Links ]
37. Bird W, Upham LL. Barbary sheep and mule deer food habits of Largo Canyon, New Mexico. In: Simpson CD, editor. Proceedings of the symposium on ecology and management of Barbary sheep; 1979 Nov 19-21; Texas, USA. Lubbock, TX: Texas Tech University Press; 1980. p. 92-96. [ Links ]
Jamel Ben Mimoun
Laboratory of Animal Ecology, Department of Biology
Faculty of ScienceTunis El Manar University
Farhat Hached, Tunis University Campus BP 94
Rommana 1068, Tunisia
Received: 12 Dec. 2014
Revised: 13 Mar. 2015
Accepted: 16 June 2015