Adaptive capacity of evicted agro-pastoralists from Ihefu Basin in Tanzania

This paper looks at the adaptive capacity of evicted agro-pastoralists who were evicted from Ihefu Basin in Mbarali District and forced to settle in new resettlement areas. Household questionnaires, life stories, and focused group discussions with a sample of 110 resettled agro-pastoralists was used to capture the data. Quantitative and qualitative data were analysed. The adaptive capacity index was used to determine the adaptive capacity of the resettled pastoralists. Findings revealed that the resettled pastoralists adopted different coping behaviors and livelihood strategies to adapt the resettlement areas. It was further observed that different livelihood capitals had different contributions on adaptive capacity among the resettled agro-pastoralists.Variations in the adaptive capacity of resettled agropastoralists were noted. For improving the adaptive capacity of resettled agropastoralists, the Government and other stakeholders should improve infrastructure such as rural roads, clean and safe water supply, schools, extension and veterinary services, reliable markets for agro and livestock products and marketing information.


INTRODUCTION
Pastoralism often refers to extensive husbandry of herds of different animal species (cattle, sheep, goats, camelids, and equines) requiring periodic migration to access pasture.A commonly used definition in various literatures is that pastoralist households are those in which at least 50% of household gross revenue (including income and consumption) comes from livestock or livestock-related activities (Oxfam, 2008).Agropastoralism describes the coexistence of both agricultural and grazing activities, although there may be different degrees of integration of these activities, with specific consequences for land use.An economic definition is that agro-pastoralists derive more than 50% of household gross revenue from livestock and 10-50% from farming (Oxfam, 2008).
Pastoralists in Ihefu Basin in Mbarali District have stayed in the area for more than thirty years from 1972.The Ihefu Basin area was gazetted as a conservation area in 1998 (SMUWC, 2001).Gazettement of Ihefu Basin as a conservation area opened up processes for pastoralists" eviction.As part of the move to safeguard the environment in Ihefu Basin, the Government issued a notice on 9 th March 2006 banning all livestock keepers" activities in Mbarali District.About 1,000 pastoralist households were evicted by the Government from Mbarali District.During the operation, an estimated 218,000 herds were actually relocated to various places (Walsh, 2007).
Various studies have shown that there is impoverishment and a change in livelihoods in pastoral societies following transformation in land use and ownership (Vangen, 2009;Brockington and Igoe, 2006).Cernea et al. (2003) examined how eviction affects people"s livelihoods in terms of major hardship risks, which include landlessness, joblessness, homelessness, marginalization, food insecurity, increased morbidity and mortality, loss of access to common property and social disempowerment and disruption to social institutions.
The eviction of agro-pastoralists from the Ihefu Wetlands and their definitive settlement was incited by a series of factors.The intention of the Government was to safeguard the environment and seen to have deteriorated affecting not only the Usangu plains and the Ihefu wetlands but also the Rufiji Basin ecosystem.Agropastoralist was taken as the major cause of severe water depletion leading to drying up of Mtera and Kidatu dams.During the 2006 dry season, the drying of Great Ruaha River forced Mtera Hydroelectric Plant to close and this measure reduced the Kidatu Hydroelectric Plant"s production of electricity by almost 50% (PINGOS, 2007;Walsh, 2007;Ngailo, 2011).
Eviction of agro-pastoralists from Ihefu to new areas of Tanzania has created some negative socio-economic impact on their livelihoods.However, there is little documentation on these adverse impacts.The exact nature and extent of these effects on the receiving or resettlement areas are not well known and documented, and in particular, the adaptive capacity of evicted pastoralists from Ihefu to new resettlements.The research questions addressed in this research includes; what livelihood strategies were undertaken by agropastoralists to sustain their livelihoods and what were the determining factors for the adaptive capacity of the agropastoralists in the resettlement areas.Therefore, this paper reveals the state of adaptive capacity and examines the livelihood strategies used by agropastoralists in the course of adapting in the resettlement areas.

CONCEPTUAL FRAMEWORK
The conceptual framework of this study was adopted by modifying some elements from the DFIDs Sustainable Livelihood Approach framework (SLA).Livelihood concerns the way people shape their lives by using Msomba et al. 57 material and non-material assets.The earlier livelihood definition was provided by Chambers and Conway (1992) who stated as follows: "A livelihood comprises the capabilities, assets (stores, resources, claims and access) and activities required for a means of living".The livelihood framework is an analytical approach that puts people"s livelihoods, meaning their interaction with their environment at its centre.The framework incorporates five elements of analysis; context and policy analysis, analysis of livelihood resources, institutions and organizations, livelihood strategies and sustainable livelihood outcome.The core livelihood approach lies in analyzing different assets and capital upon which individuals or households draw to produce.
Adaptive capacity is the ability to design and implement effective adaptation strategies, or to react to evolving hazards and stresses to reduce the likelihood of the occurrence and/or the magnitude of harmful outcomes resulting from different hazards (Adger and Brooks, 2007).At the local level, the ability to adapt can be influenced by such factors as managerial ability, access to financial, technological and information resources, the institutional environment within which adaptation occurs, political influence and kinship networks (Kelly and Adger, 2000;Smith and Pilifosova, 2001;IPPC, 2007;Schipper and Burton, 2009).However, it is suggested that humans possess the ability to plan and manage adaptation (Schipper and Burton, 2009).

METHODOLOGY
This study was conducted in Mbarali, Chunya and Kilwa districts of Tanzania Mainland.Mbarali District was purposively included in this study because it was the area where agro-pastoralists were evicted from.Chunya and Kilwa were among eight districts where agropastoralists were eventually directed for resettlement after eviction from Ihefu Basin in Mbarali District.The study involved 110 respondents who were mainly affected by the eviction process.
Both primary and secondary data were collected.This involved qualitative and quantitative data collection methods.Ouantitative data were gathered using questionnaire, while qualitative data were collected through key informant interviews, life histories and Focused Group Discussions.The quantitative data were analyzed using the Statistical Package for Social Sciences (SPSS) Version 16.0 computer software, which provided descriptive and inferential statistics.In descriptive statistical analysis, frequencies, percentages, means and cross-tabulation were used to measure associations, while, inferential statistics, chi-square and t-test were used to measure variations in some variables in adaptive capacity.Qualitative data were analyzed by using content analysis.To examine the adaptive capacity of the agro-pastoralists resettled in Kilwa, Chunya and Mbarali districts an adaptive capacity index was developed.
An adaptive capacity index was used to determine the adaptive capacity of pastoralists resettled in Kilwa, Chunya and Mbarali districts.The index was developed using 25 indicators (Table 1).Based on the variables in the conceptual framework of this study, the measure of indicators for adaptive capacity was obtained by asking the respondents several questions and their answers were scored.The questions were based on ability of pastoralists to formulate their adaptive strategies, to access various forms of

Total score 31
capital and to follow institutional processes.The respondents" answers were scored based on the interval scale of a particular question.In this study, the social aspect of adaptive capacity (indicators 1 to 4) referred to respondents maintaining a kinship network."Yes/No" questions were used to ask if they had maintained kinship networks.This implies that a respondent who reported "Yes" had lost the kinship network and so was given 0 but if "No" was given 1.Economic aspects referred to their financial status measured by three indicators: price of farm inputs, access to credit, and the sale of crop and animal products (indicators 5 to 7).Physical aspects (indicators 8 to 11) referred to access to dips, market, veterinary clinics and charcoal dam.Natural aspects (indicators 12 to 17) referred to respondents" access to good pasture, water, areas free from disease, land for cultivation, the forest and forest products and grazing land for animals.
Human aspects referred to the ability of the respondent to acquire a residence.If the respondent had a permanent residence he/she was given a score of 2, if it was semi-permanent he/she was given the score of 1 and if it was a temporary residence he/she was given the score of 0 (indicator 18), which implies that this person may not have decided to settle in that area, and so can decide to leave the area and move to another area.Regarding the acquisition of land, if the answer was by inheriting the score given was 1, if by renting the score was 2, if allocated by the government the score was 3, and if the land had been acquired by purchasing it the score For all questions with a response of "Yes"/"No", if the response was "Yes" they were given a score of 1 and if "No" they were given no score (0).The maximum score was 31.The total score for each respondent was divided by 31 so as to have a cumulative index ranging from 0 -1 (0-100%).The adaptive capacity indices of respondents were put into three categories, such as poor adaptive capacity (adaptive capacity index score of 0.00 and less than 0.44), moderate adaptive capacity (adaptive capacity index score of 0.44 and less than 0.55) and good adaptive capacity (adaptive capacity index score of 0.55 and less than 1).The cut-off point for one standard deviation increase was zero + 1 std dev= poor; = 2 std dev = moderate; above + 2 std dev = good adaptive capacity.

Adaptive capacity of resettled pastoralists
Table 2 shows the respondents adaptive capacity categories, which were poor adaptive capacity (adaptive capacity index score between 0.00 and less than 0.44), moderate adaptive capacity (adaptive capacity index score of 0.44 and less than 0.55) and good adaptive capacity (adaptive capacity index of 0.55 and less than 1).It was evident that respondents" adaptive capacity was poor in Chunya District (82.2%).Good adaptive capacity was noted in pastoralists who resettled in Kilwa (45.2%) and Mbarali (53.2%).
A t-test was carried out to determine whether there was a significant difference in the adaptive capacity of respondents in Kilwa, Chunya and Mbarali districts.The test showed that there were highly (t=33.889,p=0.012) statistically significant differences between the adaptive capacity of pastoralists who had resettled in Kilwa and Chunya districts.
The results suggested that the observed difference could be attributed to the ability of respondents to apply different adaptive strategies by accessing several capitals and institutional processes e.g.village governments.For example, in the case of "poor adaptive category", it means a maximum of 44% were capable of applying different adaptive strategies and accessing several capitals.
Those with "good adaptive capacity" of 55% and more were able to use different adaptive strategies and access several capitals.

State of adaptive capacity (human, social, economic, physical and natural capitals and transforming structures)
An ANOVA-F test was conducted to compare the contribution of different capitals to the adaptive capacity of resettled pastoralists in Kilwa, Chunya and Mbarali districts (Table 3).The attribute of human capital differed significantly between the three districts (F (2,107) =50.55, p=0.012).Human capital contributed less to adaptive capacity in Chunya District (M=0.23) and more to adaptive capacity in Kilwa and Mbarali districts (Ms= 0.31 and 0.36 respectively).
Post hoc comparisons using the Fisher LSD test revealed that economic capital contributed more adaptive capacity in Mbarali than Kilwa and Chunya districts.During FGDs it was noted that in Mbarali District the majority of resettled pastoralists had access to financial institutions.Physical capital made a greater contribution to adaptive capacity in Mbarali and Kilwa districts but much less so in Chunya District (Table 3).Adaptive capacity was greatly contributed to by natural capital in Kilwa and Chunya districts.There was plenty of land for grazing and cultivating in these two districts, which increased their animal and plant productivity.Resettled pastoralists in Mbarali District had minimal access to cultivating and grazing land.
Regarding the contribution of social capital to adaptive capacity, the results showed that there was no statistical difference between the three districts (F (2,107) =1.08, p=0.172).From this finding it shows that even if agropastoralists were relocated to different areas, most of them were able to maintain their kinship networks.Social networks were very important for resettled pastoralists in

Adaptive strategies undertaken in new resettlement areas
Adaptive strategies are those that seek to spread the risk of failure in response to anticipated adverse trends.This may be by intensifying existing livelihood strategies or diversifying into new activities.Table 4 shows the adaptive strategies adopted by the evicted pastoralists.The majority of respondents (97%) reported that they had increased crop-farming areas, which caused agriculture to become an important economic activity.This enabled them to improve household food security and have a surplus for sale.Seventy-four respondents revealed that another strategy was to send members of the family to settle in different parts of the country.This also was narrated during FGDs.The study also revealed that during the eviction process, some families divided into different groups to minimize the risk of losing the entire herd if they all settled in one area.Such family division strategy enabled family members to engage in new activities, which improved their livelihoods through sharing their outcome.It was also noted that 23.5% of the respondents, especially in Chunya District, used another adaptive strategy that of growing cash crops, mainly tobacco.
It was further noted that 46.4% of the respondents opted to diversify their major livelihood activities.These included income-generating activities, such as petty trade and local transport services locally named "boda boda", selling roasted meat "Nyama choma" and making local brew (especially by women in Mbarali District).A woman in Mbarali once stated that: "depending only on livestock does not ensure that I will be able to meet the needs of my family, especially when you consider the great loss we sustained during the eviction process, and so it was crucial to make use of all alternatives at our disposal".
Other adaptive strategies used to improve livelihoods were reported.Twenty-nine percent of the respondents modified their herd composition and the number of animals kept.This was reported mainly in Mbarali District (15.5 %) (Table 4).The livestock keepers in Mbarali District decided to increase livestock breeds and adjust herd composition by reducing the number of grazing animals (cattle and sheep) and increasing the number of browsers (goats).This strategy was especially adopted in Mbarali District, where grazing land was shrinking and the quality of the pasture deteriorating.However, among the livestock only goats were able to browse and reproduce to generate income to meet family needs.The income generated from goat sales sometimes was used to generate other business activities such as petty trade.Similar observations of having more grazing animals than browsers have been reported.A study by CARE International in Longido District, following support by HPI Project, found that household members opted to raise camels (Richiè et al., 2009).Through raising camels they could get more milk during the dry season unlike other households who refused the option.Camels can easily browse herbs and thorn bushes.

Conclusion
Resettled pastoralists had various ways of adopting different adaptive strategies and accessing several livelihood capitals.Such capabilities were considered to be good adaptive capacity.Availability of good pastures and water, organizations, different skills, culture and tradition were good drivers for pastoralists to adapt in resettlement areas.Government and its institutions were important in enhancing adaptive capacity of resettled pastoralists.For improving the adaptive capacity of resettled agro-pastoralists, the Government and other stakeholders should improve infrastructure such as rural roads, water supply, schools, extension and veterinary services, reliable markets and marketing information tailored to specific needs of pastoralists.

Table 1 .
List of 25 adaptive capacity indicators.