Pastoralists perceptions and rangeland evaluation for livestock production in South Eastern Ethiopia

Teshome Abate, Abule Ebro* and Lisanework Nigatu**

Hawassa University, Department of Animal and Range Science, Hawassa, Ethiopia, P. O. Box, 05,
teshome_abate@yahoo.com
* Adami Tulu Agriculture Research Center, Zeway, Ethiopia, P.O.Box 35
** Haromaya University, Dire Dawa, Ethiopia, P.O.Box, 138

Abstract

This paper reports the results of a study carried out to examine the traditional rangeland utilization practices, and evaluation of the rangelands in Rayitu district of south-Eastern Ethiopia. Traditional pastoralists’ rangeland utilization practices were assessed using structured questionnaires and group discussions. The rangelands were evaluated by stratifying the district in altitude and grazing types using grass, soil and woody layers parameters.  

About 91% of the pastoralists replied that the condition of their rangelands was poor and the main problems were drought, overgrazing and bush encroachment. Over 86% reported that compared to past, their grazing lands are now covered with bushes and shrubs. Annual and less desirable grass species dominated the communal grazing sites while perennial and highly desirable grass species were most frequent in the benchmark sites. Mean total range condition score in the communal, enclosure and benchmark sites were 23.40 +1.35 (poor), 39 + 1.35 (fair) and 50.42 +0.9 (good), respectively. The perception and rangeland evaluation studies confirmed that the communal grazing sites have deteriorated in condition and need improvement interventions. The enclosure sites were in transitional state from poor to fair condition and this suggests a need for intervention to improve their condition while the benchmark sites need maintenance of their present condition.

Key words: altitude, biomass, grazing types, species composition, traditional rangeland utilization

Introduction

In Ethiopia, rangelands cover about 61 to 65% of the total area of the country and are characterized by arid and semi-arid agro-ecologies; experience a relatively harsh climate with low, unreliable, and erratic rainfall, and are home to 12%-15% of the human population, and 26% of the total livestock population. Pastoralism and agro-pastoralism are the dominant types of land use systems in these areas. The lowland livestock play a significant role in the national economy. Lowland breeds of cattle and sheep make up over 90% of the legal export of live animal and lowland cattle provide around 20% of the draught animals for the highlands. The rangelands are not only known for livestock rearing, but there are many wildlife parks, sanctuaries, and reserves.  Many of these rangelands and rangeland based lifestyle at present are shrinking and degrading due to natural and human induced factors (PADP 2004).

Pastoral communities usually have a detailed knowledge of the environment of the grazing lands. This knowledge is acquired through extensive observation and continuous herding  practice (Mapinduzi et al 2003). Despite the existence of such knowledge, researchers and development policy experts previously overlooked community-based knowledge when evaluating the rangelands. Development interventions that did not integrate traditional range management strategies have not been successful. Combining community-based knowledge with scientific knowledge may provide a more complete understanding of environment from the perspective of those utilizing the resources (Ayana and Gufu 2008).The need for incorporating community-based knowledge in assessing rangelands has been widely acknowledged (Fernandez-Gimenez 2000). Studies on indigenous knowledge of rangeland management have shown that pastoralists have management strategies and own traditional ecological knowledge to classify rangelands and assess range condition and trend (Fernandez-Gimenez 2000; Ayana and Fekadu 2003; Mapinduzi et al 2003, Ayana and Gufu 2008).These traditional practices provide a useful source of information for the development and sustainable use and conservation of natural resources..

Rayitu is one of the districts found in the lowlands of Bale (South eastern Ethiopia) where pastoralism and agro- pastoralism are the main land use systems and livestock are the main assets of the community. The researchers are unable to trace any documented research efforts on rangeland evaluation in the lowlands of Bale, south eastern Ethiopia, in general, and in the study district, in particular. The significance of this study is partly to fill the gap in the knowledge of the conditions of rangelands as influenced by altitude and grazing types. Therefore, this study was initiated to assess the traditional grazing land utilization practices and evaluate the condition of the rangeland in Rayitu district of Bale zone, south east Ethiopia.

Materials and methods 

Description of the study area

The study was conducted in Rayitu district, which is found in Bale zone, south eastern Ethiopia. It covers an area of about 6 139.39 km² of land and its climate varies from hot to warm sub-moist plains (Sm1-1) sub-agro ecological zone. The rainfall pattern is bimodal with erratic distribution; the main rainy season extends from March to end of June and the short rainy season usually extending from September to end of October. The average annual rainfall is usually about 450 mm (PADS 2004), and the mean annual temperature of the district range between 26-40⁰C. The production system in the district is pastoral (PADS 2004). The district lies within an altitudinal range between 500 to 1,785 m.a.s.l. and the dominant soil types are Solancharks, fluvisols and xerosols. The human population of the district is estimated to be 43,914 (excluding 60% of nomadic people). The livestock population in the area is estimated to be 22,099 to 37,581 cattle, 31,065 goats, 10,869 sheep, 6,705 camels, 4,357 donkeys, 112 horses, 64 mules, 7,662 poultry and 5,874 local beehives (CSA 2003).

Assessment of the traditional grazing land utilization practices

A single-visit formal survey method (ILCA 1990) was followed to gather data.  Prior to the actual survey, visits were made to the district and secondary information relevant to the study were gathered from all possible sources. Informal surveys and group discussions were made to gather information about the district and to get insights from community members who were directly or indirectly involved in the production system. Group discussions were held with elders, key informants, and development agents working on the sites and district administrative officers. The group discussions focused on rangeland condition and management, and bush encroachment. The information gathered through the above process was summarized and used as basis to design other data collection instruments. A structured questionnaire was administered to obtain data on rangeland condition and management and bush encroachment. To undertake the questionnaire enumerators were recruited and training was given. A pre-test of the questionnaire was made before the actual data collection, and appropriate modifications and corrections were made. In the district, there were 19 Pastoral Associations (PAs) which are the lowest administrative units. Out of these, six PAs were selected purposely based on accessibility, representativeness of grazing land and livestock potential. From the selected PAs, a total of 90 households were chosen randomly and interviewed individually. Prior to the interview, the objective of the survey was explained and discussed with the informants in order to insure their cooperation. Furthermore, in each of the selected PA, community leaders who are familiar with the area were used as facilitator for data collection

Assessment of rangeland condition

Field layout and site selection

The grazing areas of the district were stratified into two, based on altitude,.i.e., 765-1070 m.a.s.l and >1070-1350 m.a.s.l. The altitudes of each of the sample sites were measured using altimeters. Within each altitude group, the grazing areas were further stratified into communal, enclosure and benchmark. Communal grazing lands were those grazing sites which were owned communally, the vegetation was being grazed through out the year, and the intensity of grazing was high. Lands were taken as “enclosure”or locally called Hogga when they were privately used as grazing fields, protected and fenced with thorny bushes for dry season grazing and the intensity of grazing was medium. Benchmark sites were those areas which were protected from grazing animals for a long time (10-15 years old), and relatively had low grazing intensity and were used for comparison purposes. A total of 15 grazing sites (namely 8 communal, 5 enclosure and 2 benchmark) and 9 grazing sites (namely 4 communal, 3 enclosure and 2 benchmark) were selected in terms of their altitude group of >1070-1350 and 765 to 1070 m.a.s.l. respectively. On each grazing site, a sampling block, which was 3 km x 1 km, of the land area, representative of the vegetation under consideration, was demarcated either in an unbroken or a broken form. The demarcated area was further sub-divided into 3 equal plots of 1 km x 1 km for the purpose of stratification. In each of the sub-divided plots, a belt transect of 50m x 4m was laid out randomly to accommodate both vegetation layers, i.e., woody and herbaceous (Friedel 1991; Abule et al 2007a,b).

Herbaceous floristic composition and biomass

The species composition of the herbaceous layer at each of the belt transect of 50m x 4m was determined, based on the frequency of occurrence, using a wheel point apparatus where the nearest plants were recorded (Tidmarsh and Havenga 1955). At each of the sample site, 300 point observations were recorded and the readings were undertaken at 3m interval by revolving the wheel point. At each observation point, the nearest herbaceous plant, within a radius of 300mm was recorded by species which included both annuals and perennials. Non-grass herbaceous species were recorded as forbs. If no herbaceous species of the given criteria occurred within the given radius of the point, it was recorded as “bare ground”. Bare ground and species composition was determined based on percentages of occurrence of bare ground area and individual species in relation to the total number of observation points.

The study was carried out during the main growing season (May 2006) at the time when most of the grass species were at the flowering stages. Plants with a full flowering head and other vegetative parts were collected and identified at the Herbarium of Haramya University, Ethiopia. All the identified grass species were classified into highly desirable, desirable and less desirable based on the information obtained from the pastoralists and the insights from the literature. For herbaceous species composition (1 to 10) points were considered based on the contribution of grasses only. The maximum score of 10 points was given if the contribution of highly desirable grasses was 91-100%, 9 points if the highly desirable grasses contributed 91-90% etc. score of 3 and 4 points were given if highly desirable grasses made up 10-40%, as well as <30% and equal or > 30% increasers, respectively. Score of 1 and 2 points were given if highly desirable grasses were less than 10%, as well as<50% and equal or >50% intermediate desirable grasses, respectively (Baars et al 1997). The dry matter (DM) biomass of the herbaceous plant was determined in each belt transect. Herbaceous cover was harvested in quadrant of 1m x 1m randomly placed along the belt transect at each sampling site. A total of 4 quadrants per belt transect were harvested at ground level using hand shears. The cut samples were weighed using a simple balance immediately and hand separated into grasses and forbs. The grass species were further separated into highly desirable, intermediate and less desirable species. DM biomass of the samples was determined in an oven dried at 68^oC for 72 hours at the laboratory of Sinana Agriculture Research Center, Ethiopia.

Basal area and litter cover

Within each belt transect, the basal cover of the living plant parts were estimated in a randomly laid out 1mx1m quadrant that was partitioned into eight equal parts, and three measurements were taken at each transect. All the basal cover of plant in the selected square meter was transferred (drawn) to one of the eights for easy visual estimation. The highest score (10 points) was given for the basal cover of tufted species if the eight was completely filled (corresponding to 12% basal cover of the original square meter). Accordingly, classes of < 3%, 3-6%, 6-9%, 9-12% were distinguished. Lower scores (0, 1 and 2 points) were given for basal cover with < 3%. The maximum score was given to creeping grasses such as Cynodon dactylon. The rating for litter cover within the same square meter was given the maximum score (10 points) when it exceeds 40% and the minimum score when the litter cover was less than 3% (Baars et al 1997).

The number of seedling and age distribution

The number of seedlings and age categories of the grasses species were recorded from each belt transect using three areas equal to the size of an A4 sheet of paper (26 x 24cm) chosen at random. The sheet was dropped from a height of 2m above the ground. The numbers of seedlings were counted. The highest score 5 point was given for more than 4 seedlings, for 3, 2 1 and no seedlings, 4, 3, 2,1and 0 points were given respectively. Similarly, when all age categories, young, medium and old plants of the dominant species were present, the maximum score of 5 points were given. If there is only old plant, medium-aged or young plants, the score 3, 2 or 1 point, respectively was assigned. Young and medium-aged plants defined as having approximately 20% and 50%, respectively of the biomass of old and mature plants of the dominant species (Baars et al 1997).

Soil erosion and soil compaction

The extent of soil erosion and compaction in each of the belt transect of the study area was

evaluated from three measurements of 1m x 1m quadrant by visual observations. The score of soil erosion (0 to 5 points) was based on the amount of pedestals (higher parts of soils, kept together by plant roots, with eroded soil around the tuft), and in severs cases, the presence of pavements (terraces of soil, normally without basal cover, with a line of tufts between pavements). The highest score 5 point was given for no sign of erosion, (4 points= slight sand mulch; 3 points= weak pedestals; 2 points = steep sided pedestal; 1= pavement, 0= gullies). Soil compaction (1 to 5 points) was evaluated based on the level of capping or crust formation of the surface soil. The scores given to soil compaction range from 1-5 points. The highest score (5 points) was given for a soil surface with no capping; 4 points=isolated capping; 3 = greater than 50% capping; 2 = greater than 75% capping and 1= almost 100% capping) (Baars et al 1997).

Sampling of woody vegetation layer

In each of the 50 m x 4 m belt transect, the vegetation’s density, hedging, and canopy cover were recorded. All rooted live woody plants regardless of being single stemmed or multi-stemmed were counted and identified to determine wood species density per hectare. Identification and nomenclature of the woody species followed as that indicated for herbaceous species. The lowest score (0 points) was given for a woody density 0= >5000/ha; 1= >4000-5000/ha; 2=>3000-4000/ha; 3=>2000-3000/ha; 4=>1000-2000/ha; 5=>0-1000/ha (Abule et al 2007b). Furthermore, within each belt transect, the cover of the canopy was measured in percent using tape meter. The canopy cover (1-5 points) was rated as described by Kuchar (1995). The highest score (5 points) was given for a woody cover <15% cover, 4=15-25%, 3=26-35%, 2=36-45%, 1=>45% cover. The hedging effect for the woody vegetation was rated as described by Kuchar (1995) with in 50m x 4m belt transects.  For the hedging effect 1 to 5 points was considered. The highest score 5 point was given if the highly palatable and palatable shrubs dominant and most of hedgeable plants were lightly to moderately hedged and few or no decadent plants were present in belt transect.  Three points were given, if palatable plants dominant and hedgeable plants moderately to heavily hedged and some shrubs decadent due to hedging were present Two points were given, if palatable and less palatable plants dominant and hedgeable plants heavily to very heavily hedged and if considerable numbers of decadents' shrubs were present and some may be dead due to hedging. One point was given, if less palatable and unpalatable shrubs dominant and some normally unhedgeable shrubs were hedged  moreover, hedgeable shrubs were very heavily hedged the crowns often reduced to nubbins. Many shrubs decadent and dead from hedging.

Evaluation of the condition of the rangeland

The condition of the rangelands in the study district was analyzed focusing on three factors, i.e., grass layer, soil layer, and woody layers (Friedel 1991). Moreover, percentage of bare ground cover was determined to strengthen the result of range condition rating. The factors for grass and soil layers were considered based on the criteria developed for semi-arid rangelands in south and eastern Africa and the analysis of rangeland condition was carried out following Baars et al (1997).The factors for woody layer were considered and adapted based on the criteria developed for Southeast Ethiopia (Kuchar 1995). The assessment factors based on grass, soil and woody parameters sum up to a total of 65 points. The over all rating was interpreted as excellent when the score (53-65 points), good (40-52 points), fair (27-39 points), poor (14- 26 points) and very poor when the rating was less or equal to 13 points (Baars et al 1997).

Statistical analyses

The collected household data were summarized and analyzed using Statistical Package for the Social Sciences, (SPSS, version 10 1996). Descriptive statistics such as mean, percentage and standard deviation were used to present the results. A General Linear Model (GLM) procedure of Statistical Analytical System (SAS) computer software (1987) was used for data analysis of range condition and biomass data.  The least significant difference was used for mean comparison.

Results 

Traditional rangeland management practices

There was no communal rangeland management practice controlled by the clan leaders in Rayitu district. About 88% of the respondents conserved feed, as standing hay by establishing enclosures. Enclosures were privately owned, and this might be due to the need to manage the rangeland privately so as it would help in conserving forage for their animals and for cultivation of more land. Enclosures were usually located around the homestead and farmlands and were mainly used for dry season feeding of lactating cows, calves, draught oxen, weak and sick animals. This rangeland management practice is a recently introduced practice. Drought, bush encroachment, poor productivity and lack of proper management of enclosures were among the major constraints raised by the respondents in the study area. Splitting of livestock species based on livestock type was a common practice and pastoralists separate large ruminants from small ruminants. Cattle and camels were separately herded with independent herders while, small ruminants (goats and sheep) were herded together. Fifty eight percent of the respondents splitted their herd into village based and satellite based herds (Table 1).

Village based herds such as calves and small ruminants are kept around the homestead, while cattle and camels are taken to distance places. Moreover, during the dry season, animal under production (lactating cows), sick animals and calves were kept around the village. Satellites herds (bulls, heifers, dry cows and camels) exploit pasture and water far from homestead.

Out of the total respondents, 71.1% (64) were transhumance; they moved their livestock seasonally in order to exploit areas away from their permanent settlement sites. The entire villages rarely move with the herd and only 5.6% (5) of the respondents were nomadic while the remaining 23.3% (21) were sedentary. The extensions and directions of movements depend on the availability of rainfall, water, feed, and security. In the dry season (December to February), all classes of livestock moved to a distant place where water and feed were abundant and stay as godaantu (migrant). Livestock classes, which are moved as godaantu, were mostly cows, bulls, heifers, goats, sheep and camels. The household heads and boys above the age of 15 years were responsible for migration. According to the opinion of the pastoralists, compared to the past, the frequency of migration has now decreased by 63% (55). This could be due to settlement; decline in livestock ownership per household, human population pressure, conflict and government policy (encourages settlement, education and water points). In contrast, 37% (33) of the sampled households indicated that compared to the past, the intensity of migration has now increased in the study district because of shortage of rainfall, feed and water. Most of the pastoralists replied that migration was undertaken while facing many problems such as wild animal attack 87.8%(65), death of livestock 73%(59), water shortages 45.8%(33), feed shortages 13.9%(10), conflict 12.2%(9), theft 4.1%(3) and human disease 2.4%(2).

Perception of pastoralists towards rangeland condition and degradation

The pastoralists stated different opinions on the current condition of the rangeland as compared to the past 30 years. Based on their own judgment, 91% of the respondent indicated that the condition of rangelands was poor. The present study revealed that drought, overgrazing, bush encroachment and human population pressures were among the factors for rangeland degradation in the study district (Table 2).

The communities have been aware of the extent of degradation that has been going on in the rangelands.  Plant, soil and animal parameters are used as indicators of rangeland degradation (Table 3).

Periodic assessment of the condition of the rangelands has been part of the traditional resource management practice of the communities. In the district, pastoralists assess the condition of the rangelands, which was called 'Aburu' and the assessment was carried out individually or on group basis. This assessment of range condition is mainly based on the availability of grasses, water, free of animal and human disease, suitability to the different livestock species and security to the herders. In the past, when grasses were abundant, pastoralists used fire as range management tool. However, recently due to ban on the use of fire no rangeland improvement practice is currently applied.

Perception of pastoralists towards bush encroachment

About eighty six percent of the respondents (77) replied that compared to past, their grazing land was covered with bushes and shrubs. About 83% of the sampled households perceived bush encroachment as an indicator of rangeland degradation.  The present study also revealed that drought 55.6% (50), overgrazing 44.4% (40),  uncontrolled livestock movement 42.2% (38), lack of burning 8.9% (8) to be the triggering factors for bush encroachment in the area as perceived by the pastoralists. However, 17.8% of the respondents failed to give any kind of reason (Table 4).  

According to the respondents, decrease in the production of herbaceous vegetation mainly the grass layer (84.4%) (79), difficulty in herding (54.4%) (49), damage by wildlife (50%) (47) and problem of bloating (3.3%) (3) were the major problems associated with abundance of trees and shrubs in the district in a decreasing order. Acacia bussie and Commiphora species were the major encroaching species in the district as indicated by the pastoralists.

Herbaceous vegetation composition and bare ground cover

A total of 44,2,2,8 and 45 species grasses, legumes, sedges, other herbaceous and woody plants were identified in the study district (Table 4). The grass species contributed about 78.57% of the total herbaceous vegetation composition and categorized into highly desirable (25%), desirable (25%) and less desirable species (50%) grass species. In the study district, Aristida adscension, Sporobolus panicoides, Sporobolus pyramidalis, and Tragus berteronianus were the frequent species in the communal grazing lands, whereas Aristida vestita, Cenchrus ciliaris, Sporobolus panicoides, Sporobolus pyramidalis, and Tetrapogon cencriformis were the frequent and/or most frequent species in the enclosure. In the benchmark grazing sites, Bothriochloa radicans, Cenchrus ciliaris, Cynodon dactylon, Eragrostis tenuifolia, Panicum maximum and Sporobolus pyramidalis were the frequent and/or most frequent species. Based on the frequency of occurrence, the percentage bare ground area varied between 12.25 to 23.44% with a mean of about 18% across the study sites (Table 4).

The effect of grazing type and altitude on range condition and dry matter biomass production

There was a significant difference (P<0.05) among the grazing types in species  composition, basal cover, litter cover, number of grass seedlings and age distribution of  grasses (Table 5).

These parameters were higher in the benchmarks followed by the enclosures and the lowest values were recorded in communal grazing sites. On the other hand, the soil was more eroded and compacted (P<0.05) in the communal grazing sites than in the other two grazing types. The density of woody vegetation (no. /ha) was significantly (P<0.05) affected by grazing types, i.e. a higher density and canopy cover of woody species was observed in communal grazing sites  and enclosure site than in the benchmark sites. Communal grazing sites had higher hedging effect (P<0.05) than enclosures and benchmark sites. The present study also revealed that the overall range condition was significantly (P<0.05) affected by grazing types. Thus, communal, enclosure and benchmark sites were classified as poor, fair and good condition, respectively. Total above ground DM biomass, DM biomass of grass,  DM biomass of highly, DM biomass of intermediate and DM biomass of less desirable  grasses was significantly highest (P<0.05) in the benchmark grazing sites while the  lowest was recorded in the communal grazing sites. On the other hand, significantly (P<0.05) lower DM of forbs biomass was observed in benchmark than the other grazing types.

Species composition rating, and basal cover were significantly (P<0.05) higher in the higher altitude than in the lower altitude ranges (Table 6).

Similarly, the soil had been more eroded and compacted (P<0.05) in the higher altitude zone than in the lower altitudinal belts. On the other hand, number of grass seedlings was significantly (P<0.05) higher in lower altitude than higher altitudinal ranges. There was no significant (P>0.05) variation in woody density, hedging and canopy cover among altitude zone. The mean density of woody plants (no. /ha) was 2, 538.5. Total DM biomass, DM of grass, DM of highly desirable grass, and DM of least desirable grass were significantly (P<0.05) higher in altitudinal zones of >1070-1350 m.a.s.l. than in the other altitudinal zone.  

Discussion 

As opposed to other pastoral areas of Ethiopia like the Borana of South Ethiopia (Gemedo et al 2006) and Afar middle rift valley of Ethiopia (Abule et al 2005), there was no communal rangeland management practice controlled by the clan leaders in Rayitu district. The utilization of rangeland areas by establishing enclosures were similar to that practiced by the Borana pastoralists of South Ethiopia (Ayana and Gufu 2008) and part of the middle Awash (Abule et al 2005).  Similarly, splitting of livestock herds based on species, type and productivity, migration and free ranging of communal land were more or less similar to those reported by Gemedo et al (2006). Mobility characterizes the pastoral production system. It is a very important strategy of pastoralists to exploit scarce vegetation and water resources in dry lands and it is in harmony with the harsh environment. The extensions and directions of movements, trend of migration and the problems facing pastoralists in present study were in line with the findings of Abule et al (2005).

The decline in the condition of the rangeland as perceived by the pastoralists was similar to that in the Borana rangeland of South east Ethiopia (Ayana and Fekadu 2003), in middle Awash areas of Ethiopia (Abule et al 2005), and in Hamer and Bena-Tsemay districts South west Ethiopia. Similarly, the major factors for rangeland degradation in study district were in agreement with the observation made by Ayana and Fekadu (2003) in Borana of Southern Ethiopia. The parameters used as indicators of rangeland degradation were corresponding to those reported by Gemedo et al (2006), and the Fulani pastoralists of north Burkina Faso (Mapinduzi et al 2003). The seasonal assessment of range condition and the criteria used by the pastoralists in the study district were similar to those practiced by the Borena pastoralists of South Ethiopia and Massi pastoralist of Kenya (Mapinduzi et al 2003). This type of pastoralists' indigenous knowledge to assess range condition might remain to be important for future grazing management plan. However, this traditional ecological knowledge of the pastoralists currently has been weak and neglected in the study area.

Bush encroachment and the expansion of invasive plant species in the rangelands of Ethiopia is common (Abule et al 2005; Ayana and Gufu 2008), and was also indicated by the pastoralists as a problem in the study area. It is also commonly seen as indicator of rangeland degradation in semi- arid savanna. The causal factors for bush encroachment are not absolutely clear; rather they are diverse and complex (Smit 2002). In the study district, pastoralists responded that drought, overgrazing, livestock movement and absence of fire as the major factors triggering woody encroachment which was in agreement with the reports of Abule et al (2005) for the Awash Rift Valley of Ethiopia, Herrmann and Hutchinson (2005) for the Sahelian belt of Africa, Gemedo et al (2006), and Ayana and Gufu (2008) for the Borana zone of Southern Ethiopia which reported changes in vegetation dynamics due to overstocking,  and drought and absence of fire.

The low spatial distribution of highly desirable and perennial grasses observed in the study area might suggest the deteriorating condition of the rangeland and this finding was in agreement with the reports of Gemedo et al (2006). Increases of undesirable perennial species and decreases of desirable species are important indicators of change in the condition of the rangeland but changes in the abundance of pioneer plants may be misleading in view of their quick response to rainfall and rapid disappearance afterwards. A decrease in palatable perennial grasses is often the result of continued selective and/or heavy grazing by livestock, in which case they are normally replaced by unpalatable grasses, weeds and shrubs. The causes for bare ground could be complex and related with climate, landscape, geology, vegetation cover and pattern of land use. The highest percentage of bare ground in communal grazing land of the study district could be associated with poor range condition such as low herbaceous basal cover, increased soil erosion and compaction. On the other hands, the low percentages of bare ground exhibited in the benchmark sites could probably be attributed to the low grazing pressure.

The low value of grass species composition, basal cover, litter cover, number of grass seedlings and age distribution in communal grazing lands vis-a-vis the enclosure and benchmark could be attributed to the poor range condition associated with recurrent drought in communal grazing lands.  In arid and semi-arid communal rangelands several studies have revealed that species composition, basal cover and age distribution depended on a number of factors such as types of gazing management, drought frequency, rainfall, human and livestock populations (e.g., Van der Westhizen et al 2001). Overgrazing due to high livestock population and prolonged drought might lead to a reduction in herbaceous species composition and diversity, which could aggravate the rangeland deterioration. As reported by other investigators (Amsalu and Baars 2002, Abule et al 2005), heavy grazing pressure may cause a reduction of plant species composition and basal cover. On the other hand, the highest species composition, basal cover, age distribution, number of grass seedlings exhibited in benchmark site could be attributed to the relatively lower impact of livestock grazing and trampling pressures, and high management influence of pastoralists and relatively good range condition.

The variation observed in woody density value in the study district could be attributed to the management practice of the pastoralists such as selective thinning of unwanted woody plants from the benchmark grazing areas. A shrub cover of 40% and/or 2,400 woody plants per hectare has been considered as a borderline between non-encroached and encroached condition (Roquest et al 2001) and a 2,500 tree equivalent per hectare  has been considered as a highly encroached condition (Richter et al 2001; Abule et al 2007b). The overall woody species density of the different grazing types revealed 2,454.9 plants/ha, which was on the border between encroached and non- encroached condition. Consistent with the view of Roquest et al (2001), Richter et al (2001) and Gemedo et al (2006), communal grazing and enclosure areas of the study district were under bush encroachment while the benchmark grazing sites are not bush encroached. The similar value of canopy cover in communal grazing and enclosure sites could be  associated with woody plant density. This result conforms with the findings of Gemedo et al (2006) who indicated that canopy cover of woody vegetation in Borena rangeland was 52% with the range of 27% to 73% and argued that canopy cover and the density of woody plants were positively correlated.

The findings on total DM biomass, DM of grass, DM of highly desirable grass, intermediate and less desirable in communal grazing lands as compared to the other grazing type was in accordance with the reports of Amsalu and Baars (2002) for the rift valley of Ethiopia, Gemedo et al (2006) for the Southern Ethiopia who found that poor range condition had low forage production with less desirable forage than good range condition. The variation observed in biomass could be associated with poor, fair and good range condition class. On the other hand, the increase in the DM biomass of forbs and DM of less desirable grasses in the communal grazing lands might be an evidence for poor range condition. Mean total DM biomass in the communal grazing sites of the study district was observed to be (491.30 kg/ha) and is low in comparison to similar condition of the semi-arid rangelands of Borena (Gemedo et al 2006), Middle Awash (Abule et al 2007a, 2007c). This might point out that such low DM herbaceous biomass in the communal grazing area of study district could directly affect livestock production and sustainability of the rangeland of the study areas over time.

Summary and conclusions 

• The present findings indicated that both the household and the field vegetation studies confirmed that the condition of the communal grazing land has deteriorated and need improvement and rehabilitation.

  
  

• The enclosure sites of the study district were in transitional state from poor to fair condition and this suggests a need for intervention to improve the condition while the benchmark sites need maintenance of their present condition.

  
  

• With the current condition of the communal grazing lands; the sustainable utilization of the rangeland ecosystems is not possible.

  
  

• To sustain the pastoral production system in the district; the current condition of communal rangeland should be reversed through rangeland rehabilitation, conservation and proper management. Therefore, selective thinning of unwanted woody plants, resting of the communal grazing areas, strengthening of privately owned enclosures as well as establishment of community based enclosures in some key sites might be very essential. Furthermore, strengthening of the traditional resource management system, technical and technological supports to the pastoralists could be recommended. To this end, full participation from the pastoralists, government and non- governmental organizations that might be directly or indirectly involved in rangeland resources utilization, management, and conservation are imperative.

Acknowledgements 

The authors acknowledge, the Oromiya Agriculture Research Institute for financing the project, and Sinana Agriculture Research Centre for their logistical assistance during data collection and intensive use of laboratory facilities and Haromaya university for the use of their Herbarium.

References  

Abule E, Snyman H A and Smit G N 2005 Comparisons of pastoralists’ perceptions about rangeland resource utilization in the Middle Awash Valley of Ethiopia. Journal of Environmental Management 75: 21-35

Abule E, Snyman H A and Smit G N 2007a Rangeland evaluation in the middle Awash Valley of Ethiopia: Herbaceous vegetation cover. Journal of Arid Environments 70: 253-271

Abule E, Snyman H A and Smit G N 2007b Rangeland evaluation in the middle Awash Valley of Ethiopia: Woody vegetation: Journal of Arid Environments 70: 272-292

Abule E, Snyman HA and Smit GN 2007c Rangeland evaluation in the middle Awash Valley of Ethiopia: Relationships among soil and vegetation variables. Journal of Arid environments. 70:293-303.

Amsalu S and Baars R M T 2002 Grass composition and rangeland condition of the major grazing areas in the mid rift valley of Ethiopia. Africa Journal of Range and Forage Science 9: 161-166

Ayana A and Fekadu B 2003 Current range condition in southern Ethiopia in relation to traditional management strategies: The perceptions of Borana pastoralists. Tropical Grasslands 37(1); 53-59 http://www.tropicalgrasslands.asn.au/Tropical%20Grasslands%20Journal%20archive/PDFs/Vol_37_2003/Vol_37_01_03_pp53_59.pdf

Ayana A and Gufu O 2008 Herder perceptions on impacts of range enclosures, crop farming, fire ban and bush encroachment on the rangelands of Borana, Southern Ethiopia. Human Ecology (In press).

Baars R M T, Chileshe E C and Kalokoni D M 1997 Technical note: range condition in cattle density areas in the western province of Zambia. Tropical Grassland 31: 569-573 http://www.tropicalgrasslands.asn.au/Tropical%20Grasslands%20Journal%20archive/PDFs/Vol_31_1997/Vol_31_06_97_pp569_573.pdf

CSA (Central Statistical Authority) 2003 Ethiopian Statistical Abstract, Central Statistical Authority, Addis Ababa Ethiopia          

Fernández-Giménez M E 2000 The role of Mongolian nomadic pastoralists’ ecological knowledge in rangeland management. Ecological Applications 10: 1318-1326.

Friedel M H 1991 Range condition assessment and the concept of thresholds: A view point. Journal of Range Management 44: 422-426 https://digitalcommons.library.arizona.edu/objectviewer?o=http%3A%2F%2Fjrm.library.arizona.edu%2FVolume44%2FNumber5%2Fazu_jrm_v44_n5_422_426_m.pdf

Gemedo D, Maass B L and Isselstein J  2006 Rangeland condition and trend in the semi-arid Borana lowlands, Ethiopia. African Journal of Range and Forage Science 23: 49-58

Herrmann S M and Hutchinson C F 2005 The changing contexts of the desertification debate. Journal of Arid Environments 63: 538-555 http://www.infra.kth.se/courses/1H1142/attachments/paper2.pdf

ILCA (International Livestock Center for Africa) 1990 Livestock Research Manual. ILCA, Addis Ababa, Ethiopia, No 2: 31-54

Kuchar P 1995 Range monitoring, evaluation and range survey methods, South east range project technical report. Addis Ababa, Ethiopia. 92p.

Mapinduzi A L, Gufu O, Weladji R B and Colman J 2003 Use of indigenous ecological knowledge of the Maasai pastoralists for assessing rangeland biodiversity in Tanzania. African Journal of Ecology 41: 329-336

PADP (Pastoral Areas Development Plan) 2004 General executive summary. Ministry of Agriculture and Rural Development, Addis Ababa, Ethiopia. 58p.

Richter C G F, Snyman H A and Smit G N 2001 The influence of tree density on the grass layer of three sem-arid savanna types of southern Africa. Africa Journal of Range and Forage Science 18: 103-109

Roquest K G, O`Connor T G and Watkinson A R 2001 Dynamics of shrub encroachment in an African savanna: relative influences of fire, herbivory, rainfall and density dependence. Applied Ecology 38(2): 268-280

Smit  G N 2002 The importance of ecosystem dynamics in managing the bush encroachment problem in Southern Africa. P.122. Inaugural Lecture. University of the Free State, Bloemfontein, South Africa

SPSS (Statistical Packages for the Social Sciences) 1996 Cary, North Carolina

SAS (Statistical Analysis System Institute Inc.) 1987 Applied statistics and the SAS programming language 2^nd edition Cary, North Carolina

Tidmarsh C and Havenga 1955 The wheel points methods of survey and measurement of  semi-open grasslands and Karoo vegetation in South Africa. Memories of the botanical survey. Journal of Grassland Society of South Africa 7 (1):52-61

Van der Westhuizen, Snyman H A, Van Rensburg H A and Potgieter W L J 2001 The quantification of grazing capacity from grazing and production values for species in the semi-arid grassland biome of Southern Africa. African Journal of Range and Forage science 18: 43-52

Received 27 February 2009; Accepted 14 March 2009; Published 1 July 2009

Go to top