Livestock Research for Rural Development 18 (9) 2006 Guidelines to authors LRRD News

Citation of this paper

Screening of potential indigenous browse species in semi-arid central Tanzania: A case of Gairo division

E J Mtengeti and A Mhelela

Department of Animal Science and Production, Sokoine University of Agriculture,
PO Box 3004, Morogoro, Tanzania

emtengeti@yahoo.co.uk


Abstract

The potential indigenous browse species at Gairo Division in central semi arid Tanzania were screened by using local goats, livestock keeper's indigenous knowledge and laboratory analysis of their nutritive value. Screening covered three grazing sites. The sites were; bush lands, two-year fallow lands and seasonal sand riverbanks. Local goats were followed in the grazing sites to see what they browsed. Samples of the browsed shrubs and trees species were collected in gunny bags and then shown to local livestock keepers and requested to give their indigenous knowledge on these browse species and then rank them according to their importance as livestock feed. The top ten ranked browse species were used for acceptability test with local goats so as to compare the ranking of livestock keepers and livestock preference. Crude protein (CP) and in-vitro dry matter digestibility (IVDMD) of the top ten ranked browse species were also determined.

A total of 26 browse plants were observed being browsed by goats but local livestock keepers identified only 24 as potential browse species. Most of the browse plants had several other uses. Somehow the ranking of the browse species by livestock keepers was close to the preference by local goats. Albizia petersiana had the highest intake rate followed by Grewia bicolor and the least was Lannea tomentosa (13.2, 11.2 and 3.6 g DM/min., respectively). The crude protein content ranged from 12.6 % in Lannea tomentosa to 28.1% in Acacia polyacantha. Maerua trychophylla had the highest digestibility followed by Lannea schweinfurthii (64.4 and 61.3 %, respectively).

It was concluded that in order to identify the right indigenous browse plants for conservation; the browsing animals, local livestock keepers' indigenous knowledge and laboratory analysis of the nutritive value must be included in the screening work.

Key words: Indigenous browse species, indigenous knowledge, screening, semiarid Tanzania


Introduction

In Tanzania, about 70% of the national ruminant herd of 15, 9 and 4 million of cattle, goats and sheep respectively are kept in semi-arid areas mainly under agro-pastoral system (BDAS 2001). The area covers about 80 million ha that is about 40% of the Tanzania landmass (BDAS 2001). The area is heavily overstocked and signs of overgrazing, that include loss of vegetative cover, gullies and bush encroachment, are rather visible in various parts of the region. The main grazing fields are few isolated bushed grassland in abandoned shift cultivated lands and the seasonal river valleys. The crop fields are also grazed after harvest.

Due to overgrazing the area is mainly covered by short-lived grass species such as Urochloa, Dactoctylenum, Aristida, Chloris and Rhynchelytrum. Perennial grass species such as Hyperhenia, Themeda, Heteropogon, Cynodon and Cenchrus are found in isolated sites. These forages are severely overgrazed and nearly depleted within the first half of the 7 to 8 month of dry season. In the dry season therefore, as in other semi-arid areas in Africa (Le Houérou 1980) browse plants becomes an important component of the grazing animal diet.

The factors limiting introduction of exotic browse plant species in semi-arid areas include shortage of rain, communal uncontrolled grazing, wild fires and plant diseases. Indigenous browse species have high persistence at their respective areas (Woodward and Coppock 1995). They are however in danger of extinction due to over use (e.g. browsing by ruminants, charcoal making and medicinal uses), shifting cultivation and other land use and deforestation related activities. Further, not all indigenous browse species are liked by browsing animals (Kuria et al 2005), free from anti-nutritional factors and easy to establish. Moreover, the local livestock keepers may have their own preference on browse plant species to preserve depending on its multiple uses. Alemu et al (1998) found that Afar and Tigrai of Ethiopia prioritize the uses of indigenous shrubs and trees differently. Afars considered livestock fodder to be the best use of shrubs and trees while Tigrians chose construction as their best use. Therefore, any method used to screen potential indigenous browse plant species must take into consideration both the preference of browsing animals and local livestock keepers and thereafter laboratory analysis to verify the nutritional quality of the browse species. By using these three aspects, the screening work may be rapid and a short list of potential indigenous multipurpose shrubs and trees to preserve or domesticate can be provided as early as possible to the farming systems in the respective sites.

The objectives of this study were therefore to use the browsing animal and local livestock preferences and thereafter using the laboratory analysis technique to identify and prioritize the potential indigenous browse plant species in different grazing sites of the semi-arid central Tanzania.


Materials and methods

Study area

The study was conducted at Kilimani, Sanga-njelu, Chigage and Chakwale villages located in Gairo division-Kilosa district. The four villages cover an area of 35 km2. The area lies in between 36o 45' E and 6o 30' S with an altitude of 1200 m above sea level. It receives rainfall ranging from 400 to 500 mm per year falling between December and April with a short dry spell from mid-January to mid-February. The dry season is normally seven months (i.e. between May and November). The rainfall is normally erratic and falling in storms resulting in high run-offs and intensive soil erosion especially in cultivated and overgrazed areas. The area is inhabited mainly by the agro-pastoral Kaguru tribe. The major land use system is subsistence crop farming and keeping indigenous ruminant livestock breeds comprised of cattle, goats and sheep.

Identification of the browse plant species

Three types of grazing sites in each study village were included in this study. These were: bushed grassland, two years fallow land and seasonal sand riverbank. The shrub or tree, that was observed being browsed by goats, was recorded and its sample (a twig) collected and brought to the local livestock keepers to get more information on the potential of the plant as livestock feed and other uses.

Local livestock keeper's knowledge on the browse plant species

A total of 40 local livestock keepers (10 from each village ≥ 50 years old) were requested to assemble in one site and observe the displayed samples of the browsed shrubs and trees that were noted to be browsed by goats in different grazing sites. Their knowledge on these shrubs and trees as potential feed for livestock and other uses was recorded. The displayed samples were then pressed and preserved for scientific identification at Sokoine University of Agriculture (SUA)- Morogoro.

Assessing browse plant species distribution in different grazing sites

Only plants that were considered being important as livestock feed or having other outstanding use were considered for assessment of their distribution in different grazing sites. Twenty-five plots each 10 m x 10 m and 50 m apart, arranged along a transect line, were used to score the frequency of occurrence of the browse plants in each grazing site in each village. The frequency of occurrence of each plant was determined by considering the number of plots it was recorded as a percentage of the total plots in the respective grazing site.

Acceptability test of the browse plant species to the goats

Following the discussion with livestock keepers, ten browse plant species were found to be very important as livestock feed. The ten plants were used for animal acceptability test and analysis of the nutritive value. One-year-old ten male goats were used in this test. The goats were obtained from one livestock keeper. Every morning at 7.00 a.m. for seven days before going to graze the ten goats were locked in a kraal made of branches of thorn bushes within the livestock keeper's homestead. In the kraal the goats were allowed to browse in a cafeteria fashion for 15 minutes, twigs (with stems ≤ 5 mm wood diameters) of the ten browse plant species. Each browse twigs were tied to a bundle of 1 kg and hanged by a twine rope to a horizontal wooden pole supported at both ends by two poles stuck on the ground. The twigs were hanging at 40 cm above the ground surface so as to give easy reach to the browsing goats. The space between the twig bundles was 1.5 m. The position of twigs of each browse plant was changed every day so as to remove bias of browse species position. The goats were not allowed to browse the twigs any more after 15 min. They were allowed to go out of the kraal and lead to the grazing field. The fresh weight eaten from each browse species was obtained by subtracting the weight of the remaining twigs from 1 kg given. Before feeding in each day fresh sample of twigs of each browse species weighing 300 g was collected, air-dried and sent to SUA for the determination of dry matter content, chemical composition analysis and in-vitro dry matter digestibility determination. The crude protein and minerals contents were analysed according to AOAC (1990) procedures and in vitro dry matter digestibility was determined according to Tilley and Terry (1963). The neutral detergent fibre and acid detergent fibre contents were analysed according to Van Soest et al (1991). The reason of extending the test for seven days was to make sure that the animals have enough time to get used to the feed material and thus show their real preference on the browse plant of their choice (Grovum 1988).

Data analysis

The villages were rather similar in terms of the frequency of occurrence of the browse species in each grazing site. The data of the same browse species in the same grazing site in each village were therefore pooled together and expressed as a percentage of occurrence in all plots observed in the four villages. The results of intake rate and chemical composition of the browse species were subjected to the analysis of variance according to Steel and Torrie (1980). After the analysis of variance, the means, which showed significant difference, were compared using the Least Significance Difference.


Results and discussion

A total of 26 browse plant species were identified browsed by livestock (Table 1).


Table 1.  Indigenous browse plant species, their uses and ranking by local livestock keepers as important  livestock feed

No.

Vernacular name (Kaguru)

Scientific name

Uses*

Ranking as a browse plant

1

Mkungugu

Acacia tortilis

1,2,3,4,5

6

2

Msega

Acacia brevispica

1,2,5,8

2

3

Muwindi

Acacia polyacantha

1,2,5

8

4

Chifulwe

Acalypha ornata

1,2

18

5

Mgutani

Albizia anthelmintica

1,2,9

11

6

Mchalala

Albizia petersiana

1,2,4,5,6

7

7

Mkobola

Catunaregam spinosa

6,7

 

8

Chibobwe

Commiphora africana

1,2

22

9

Muhingo

Dalbergia melanoxylon

1,2,4

24

10

Mtoyo

Diplorhynchus candylocarpus

1,2,6,7

15

11

Mkelegembe

Dichrostachys cinerea

1,2,5

9

12

Mkilikia

Ehretia amoena

1,2,9

23

13

Mgunga

Faitherbia albida

1,2,3,5

12

14

Fuguti

Ficus sp.

1,2,7

13

15

Mkole

Grewia bicolor

1,2,6,9

5

16

Mkole bwambwa

Grewia goetzei

1,2,6,9

14

17

Ngelula

Justicia sp

1,10

4

18

Kilimelag’hembo

Lannea tomentosa

1,6

10

19

Mumbu

Lannea schweinfurthii

1,2,4,5,9

1

20

Mdudu

Maerua trychophylla

1,7,9

3

21

Mtalawanda

Markhamia obtusifolia

1,2,5,6,9

16

22

Kifwasa

Markhamia zanzibarica

1,2

20

23

Kitukuto

Phyllanthus reticulatus

1

19

24

Mkowe

Pteleopsis myrtifolia

2,4,6,7

 

25

Muhunungu

Zanthoxylum chalbeum

1,5,9

17

26

Mnyange

Ziziphus mucronata

1,2

21

*Uses

1.                   Edible to livestock                                            

2.                   Provide fuel wood

3.                   Making beehive

4.                   Provide shade

5.                   Provide building poles

6.                   Making stick for herding livestock and tool handles

7.                   Edible fruits and leaves as vegetable

8.                    Act as lighter as it keeps flame long time

9.                   Medicine

10.                Indicator of good soil fertility


The local livestock keepers however, singled out Catunaregam spenosa and Pteleopsis myrtifolia that livestock does not browse them. Such a disagreement between the observation of browsing livestock and livestock keepers knowledge was some how perplexing because the same livestock keepers indicated that the two plant leaves can be used as vegetables and their fruits are eaten by human. Therefore, the disagreement requires further clarification.

Except one, the rest of the plants had several other uses besides browsed by goats. Beside browsing, 20 plant species were used for fuel wood, 9 for building poles, 8 for medicine, 8 for making herding sticks and tool handles, 5 for shade, 4 for fruits and leaves as vegetable, 2 for making beehives, 1 for providing fuel ticks used as lighter in the household and 1 for an indicator of good soil fertility. Most of these uses may lead to disappearance of some plant species. Therefore, fast screening method to identify the potential indigenous browse plant species for domestication should be developed. The livestock keeper's ranked Lannea schweinfurthii the highest and Dalbergia melanoxylon as the least browse plant. However, the latter browse species is much liked by wood sculptures due to its hard black wood and it is highly used in the famous Makonde curving in Tanzania (Mbuya et al (1994) and thus require to be conserved from extinction. The use of Albizia anthelmintica as a medicinal plant for livestock is in agreement with what has been reported elsewhere among the Maasai in Kajiado, Kenya (Farah et al 1996).

Most browse plants had lower frequency of occurrence along the riverbanks than both in bush and fallow lands (Table 2).


Table 2.   Mean frequency of occurrence (%) of indigenous browse plants at three grazing sites

No.

Scientific name

                    Grazing sites

Bushland

River bank

Fallow land

1

Acacia tortilis

85

60

60

2

Acacia brevispica

70

35

60

3

Acacia polyacantha

60

40

75

4

Acalypha ornate

40

10

10

5

Albizia anthelmintica

60

10

50

6

Albizia petersiana

90

65

45

7

Catunaregam spinosa

40

25

15

8

Commiphora africana

35

25

30

9

Dalbergia melanoxylon

35

10

10

10

Diplorhynchus candylocarpus

25

35

30

11

Dichrostachys cinerea

75

55

60

12

Ehretia amoena

70

45

45

13

Faitherbia albida

35

40

35

14

Ficus sp

50

10

25

15

Grewia bicolor

65

65

60

16

Grewia goetzei

65

60

50

17

Justicia sp

100

80

65

18

Lannea tomentosa

100

15

25

19

Lannea schweinfurthii

45

55

40

20

Maerua trychophylla

75

50

70

21

Markhamia obtusifolia

55

20

45

22

Markhamia zanzibarica

80

50

35

23

Phyllanthus reticulatus

65

20

20

24

Pteleopsis myrtifolia

35

15

50

25

Zanthoxylum chalbeum

25

25

15

26

Ziziphus mucronata

35

10

15


Only 9 out of the 26 plants had mean frequency of > 50% along the riverbanks. This could be due to plant ecological preference (Alemu et al 1998) and/or being over browsed by livestock as they go down to the river for water. Lannea tomentosa and Justicia sp. were spotted in all bush plots possibly because they are neither used as fuel wood nor to make tool handles (Table 1). Also Lannea tomentosa was rather unpalatable to the browsing animals (Table 3) while Justicia sp. was very palatable but its runners protected lots of its growing points from browsing animals. The differences of occurrence between the fallow and the bush lands for most of the identified browse species were narrow. This shows how resting of land in these areas can rejuvenate the vegetation species which were nearly lost through cultivation. This was the strategy of shift cultivation system that at present cannot be practiced in many areas due to shortage of land because of increased human population. Unexpectedly plants with few uses showed lower recovery than those with many uses after two years fallow, possibly due to genetic differences or because the farmers tend to destroy them completely when they open a new farm. Generally, high recovery rates of these shrubs is in agreement with the concept that overgrazing and shift cultivation in dry rangelands deplete herbaceous understorey layer and promote bush encroachment which reduce grazing lands carrying capacity (Baumer 1990 and Strange 1980). The large ruminants are unable to penetrate dense bushes and cannot subsist well on shrubs and trees as the small ruminants especially goats. One reason of an increasing interest of small ruminants production among agro-pastoralists and settled patoralists (Legesse 1984, Wienpahl 1985) could be due to the fact that, goats with their small pointed mouth can utilize efficiently the bushes and moreover as they are smaller in size than cattle and therefore able to pass easily through the dense bushes. The results of an earlier study conducted at the same area are in agreement with this concept because the livestock keeper's were found keeping one and a half times more goats than cattle (Kavia 1997).

On average Albizia petersiana had the highest intake rate and Lannea tomentosa the least (Table 3).


Table 3.   Intake rate of the indigenous browse plant species by goats

Browse species

Intake rate, g DM/min

Mean± SD

Day 1

Day 2

Day 3

Day 4

Day 5

Day 6

Day 7

Acacia brevispica

7.8

11.7

13.7

10.5

11.1

7.2

3.3

9.3± 1.31c

Acacia polyacantha

10.9

8.1

10.7

12.2

7.5

7.5

6.1

9.0 ± 0.86c

Acacia tortilis

3.2

3.5

4.7

6.3

3.9

3.9

7.1

4.6 ± 0.58d

Albizia petersiana

5.4

6.8

11.5

11.5

16.2

18.9

22.3

13.2± 2.35a

Dschrostchys cinerea

10.5

7.6

8.1

8.1

9.0

12.0

18.0

10.5±1.29bc

Grewia bicolor

8.8

7.3

13.9

16.1

10.3

10.3

11.7

11.2±1.14b

Justicia sp

6.0

11.0

8.0

15.0

15.5

13.0

13.5

11.3±1.21b

Lannea schweinfurthii

11.4

8.6

9.8

12.0

5.7

9.2

17.7

10.6±1.41b

Lannea tomentosa

1.4

1.4

1.4

2.8

5.6

5.6

8.4

3.6±1.01e

Maerua trychophylla

1.6

6.4

8.0

11.2

16.0

16.0

18.1

10.6±2.16d

   Means followed by different superscripts are significantly different at P ≤ 0.05


Lanneaschweinfurthii was the first and Albizia petersiana the seventh choice of the livestock keepers. The goats however, ate faster Albizia petersiana than Lannea schweinfurthii. The differences between the livestock keepers' choice and the livestock preference could be due to the external features of the plant such as the shape and size of thorns and arrangement of the leaves or internal factors such as bad odour and high tannin. For example the rate of intake of Acacia tortilis was nearly 50 % less than the other two Accacia species possibly because Acacia tortilis have long thorns. The smell of Lannea tomentosa leaves lead the goats to sneeze. This was in agreement with Freer (1981) who reported that the intake of the browse species is affected by their physical and chemical composition. However, the animals took some days to get used with other species. For example there was a gradual increase of intake from day one to day seven for Maerua trychophylla and Lannea tomentosa. Such results strengthen the fact that the animals have to be given enough time to get used to the new feed (Grovum 1988). Therefore, it was appropriate in this study to test the acceptability of the browse species for seven days so as to identify the most preferred species without bias.

The dry matter content of the browse species twigs ranged from 34.5 % for Lannea schweinfurthii to 48 % for Maerua trychophlla (Table 4).


Table 4.   Mean chemical composition and in vitro dry matter digestibility of indigenous browse species  of semi arid central Tanzania

Browse species

DM, %

CP, %

Ash, %

NDF %

ADF %

Ca %

P %

IVDMD,  %

Acacia brevispica

39.2cd

23.9b

5.8ab

51.5ab

13.8c

0.9b

0.1d

56.2c

Acacia polyacantha

45.0b

28.1a

6.7a

48.9b

15.6bc

1.1ab

0.2c

50.2d

Acacia tortilis

47.2a

16.2c

6.6a

50.6ab

14.7bc

1.0ab

0.2c

48.7d

Albizia petersiana

40.5c

23.4b

5.5ab

58.1a

26.4a

0.9ab

0.2c

40.0e

Dichrostachys cinerea

40.7c

19.6c

4.3b

50.4ab

16.9bc

0.9ab

0.1d

40.2e

Grewia bicolor

44.0bc

19.2c

5.8ab

44.2b

19.5b

0.8b

0.3b

55.6c

Justicia sp

30.0d

23.2b

5.8ab

43.9b

12.2c

1.2a

0.5a

69.6a

Lannea schweinfurthii

34.5d

13.3d

5.0ab

47.5b

14.4c

1.1ab

0.3b

61.3a

Lannea tomentosa

42.0c

12.6d

5.8ab

59.6a

25.8a

1.0ab

0.2c

30.0f

Maerua trychophylla

48.0a

23.4b

6.5a

42.3b

12.4c

1.2a

0.3b

64.4a

 SEM

  3.15

   1.57

0.56

3.25

1.68

0.08

0.01

  3.85

Values in the same column followed by different superscripts and significantly different at P ≤ 0.05, DM = Dry matter, CP = Crude protein, NDF = Neutral Detergent Fibre, ADF = Acid Detergent Fibre, IVDMD = In vitro dry matter digestibility.


These results were consistent with those reported by Mecha and Adegbola (1980) in Southern Nigeria and Kuria et al (2005) in Northeastern Kenya. The crude protein content ranged from 12.6 % in Lannea tomentosa to 28 % in Accacia polyacantha. The results were within the range of 17 % to 30 % reported for indigenous browse plants in Semi Arid Central Tanzania by Olson and Welin (1989) and above the adequate range (11 - 13 % CP) for maintenance and growth requirements of goats and sheep (Kearl 1982). The two species of Lannea were not significantly (P> 0.05) different in terms of crude protein content, but the digestibility of Lannea tomentosa was 50 % less than that of Lannea schweinfurthii. It could be Lannea tomentosa had a high tannin content (Kumar and D'Mello 1995). Normally the livestock keepers base their choice of a fodder plant on the performance of the animals eating the plant. Possibly that is why Lannea Schweinfurthii which was shown to have high digestibility was ranked number one by the local livestock keepers.

Neutral detergent fibre (NDF) and acid detergent fibre (ADF) are the major determinants of overall forage quality (Van Soest 1994). The NDF content of the browse plants ranged from 42.3 to 59.6 % while that of ADF ranged from 12.2 to 26.4 % and compared well to the reported values for East African semi arid browse species (Bhalahenda 2001 and Kuria et al 2005). Calcium content was within while phosphorus was lower than the ruminant animal requirements (Kearl 1982). Similar results were reported by Olson and Welin (1989) and Bhalahenda (2001). It is therefore important to supplement minerals to the browsing animals especially when they lactating or young growing animals.


Conclusions and recommendations

From this study it can be concluded that:


Acknowledgements

The study was supported by Sokoine University of Agriculture (Tanzania) and Tuskigee University (USA) joint project (SUA-TU Linkage)


References

AOAC 1990 Association of Official Analytical Chemists. Official Methods of Analysis, 15th edition, pp. 69 -88. Washington, D.C.

Alemu D T, Farah K O and Mbuvi D M 1998 Impact of land use on woody vegetation in semi-arid area of Abala district, north Afar, Ethiopia. The African Pastoral Forum, No. 16. PINEP. Department of Range Management, University of Nairobi.

Baumer M 1990 The potential role of Agro-forestry in combating desertification and environmental degradation with special reference to Africa. Technical Centre for Agriculture. ACP. Netherlands.

BDAS 2001 Basic Data Agricultural Sector 1993/94 - 1999/2000. Statistics Unit. Ministry of Agriculture and Food Security. Dar es Salaam.

Bhalahenda M K 2001 Feed resource assessment, knowledge and practices under a traditional ruminant production system in semi-arid areas of Tanzania: A case study of Gairo Division. M.Sc. Dissertation. Sokoine University of Agriculture

Farah I O, Ngatia T A and Munyua W K 1996 Ethnoveterinary Techniques practiced by the Maasai Pastoralist of Kajiado district, Kenya. The African Pastoral Forum. No. 8 PINEP. Dept of Range Management, University of Nairobi.

Freer M 1981 The control of food intake by growing animals. Morlye, FWD Elserver Amsterdam.

Grovum W L 1988 Appetite, palatability and control of feed intake. In: Ruminant animal digestive physiology and nutrition (Church D C editor) Prentice - Hall Englewood Cliffs, New Jersey.

KaviaV 1997 Traditional livestock husbandry in Chakwale village Gairo -Division. Special project. Sokoine University of Agriculture

Kearl L C 1982 Nutrient Requirements of Ruminants in Developing Countries. International Feedstuffs Institute, Utah Agricultural Experiments Station, Utah State University, Logan, Utah USA, pp. 67-88

Kumar R and D'Mello J P F 1995 Anti-nutritional factors in forage legumes. In: D'Mello J P F and Devendra C (editors) Tropical Legumes in Animal Nutrtion. CAB International. Wallingford. United Kingdom.

Kuria S G, Wanyoike M M, Gachuiri C K and Wahome R G 2005 Nutritive value of important range forage species for camels in Marsabit district, Kenya. Tropical and Subtropical Agroecosystems, 5: 15-24.

Le Houéron H N 1980 Browse in Northern Africa. In: Le Houéron H N (editor) Browse inAfrica. The current state of knowledge. Proceedings of a symposium held at ILCA. Addis Ababa 8 - 12 April 1980. Addis Ababa, Ethiopia. http://www.ilri.cgiar.org/InfoServ/Webpub/Fulldocs/Browse_in_Africa/Chapter5.htm#

Legesse A 1984 Boran - Gabra pastoralism in historical perspective. In: Joss P J, Lynch P W and Williams O B (editors) Rangelands a resource under siege. Proceedings of the 2nd International Rangeland Congress. Aldelaide, Australia.

Mecha I and Adegbola T A 1980 Chemical composition of some southern Nigeria forage eaten by goat. In: H N (editor) Browse in Africa. The current knowledge. Proceedings of a symposium held at ILCA Addis Ababa 8-12 April 1980. Addis Ababa, Ethiopia. http://www.ilri.cgiar.org/InfoServ/Webpub/Fulldocs/Browse_in_Africa/Chapter30.htm

Mbuya L P, Msanga H P, Ruffo C K, Birnie A and Tengnas B O 1994 Useful trees and shrubs for Tanzania. Regional Soil Conservation Unit/SIDA. Nairobi. Technical Handbook No. 6

Olson A and Welin S 1989 The potential of local shrub as feed for livestock and mineral content of some soil licks in central Tanzania. Working paper 125. Swedish University of Agricultural Science. International Rural Development Centre, Uppsala.

Steel R G D and Torrie J H 1980 Principles and procedures of statistics: A biometrical approach. 2nd edition, McGraw-Hill, New York, USA.

Strange L R N 1980 African Pastureland Ecology. FAO Pasture and Fodder Crop Studies No 7

Tilley J M A and Terry R A 1963 A two-stage technique for the in vitro dry matter digestibility of forage crops. Journal of British Grassland Society, 18: 104-111

Van Soest P J, Robertson J B, Lewis B A 1991 Methods for dietary fiber, neutral detergent fiber, and non starch polysaccharides in relation to animal nutrition. Journal of Dairy Science 74, 3583 - 3597 http://jds.fass.org/cgi/reprint/74/10/3583

Van Soest P J 1994 Nutritional Ecology of the Ruminant 2nd Edition. Cornell University Press

Wienpahl J 1985 Turkana herds under environmental stress. Nomadic peoples 17: 59-87.

Woodward A and Coppock D L 1995 Role of plant defensive mechanism in the utilization of native browse plants in Southern Ethiopia. AgroforestrySystems. 32: 147-161


Received 31 December 2005; Accepted 30 June 2006; Published 11 September 2006

Go to top