Rhetoric of "community-based conservation" has gained prominence among development specialists and environmentalists, yet such projects are often implemented from the top-down in Africa. This dissertation contends that only a bottom-up approach can foster resilient livelihoods and environmental stewardship. This study focused on determinants of household resilience within a poverty-stricken agricultural community near Gombe Stream National Park (GSNP) in western Tanzania. The research purpose was to explore: 1) relationships between villagers and GSNP management; 2) how groups and individuals view priority livelihood problems and solutions; 3) various attributes of households; and 4) perceived trends for household resilience and how these are related to natural, social, human, and financial capital as per the Sustainable Livelihoods Framework (SLF). A mixed-methods approach provided qualitative and quantitative assessments. Data collection consisted of Rapid Rural Appraisal (RRA), key informant interviews, focus group discussions, and household surveys. The RRA was conducted adjacent to GSNP while other work was implemented over a larger area. Survey data were analyzed using descriptive statistics, chi square, and logistic regression. Regression examined relationships between trends in resilience variables (quality of life or ability to solve problems) versus trends in capital. Results from the RRA indicated high polarization and problems between villagers and GSNP management. A more complex picture, however, emerged from subsequent investigations revealing that the most important issues facing local communities were inadequate public services, ineffective leadership, and development isolation. This situation was exacerbated by population growth, poverty, and environmental decline. Regression results identified lack of income, manual labor, and skills and knowledge as factors undermining household resilience. Other data indicated a need for improving farming systems. In conclusion, while all forms of capital mattered to resilience, human and financial were most lacking. Knowledge of such variation strengthens future applications of the SLF. Practical implications include how an indigenous educational institution, the Gombe School of Environment and Society (GOSESO), could operate in the area. The GOSESO needs to adopt a bottom-up, participatory approach that emphasizes capacity building for poverty reduction and conservation. This could allow for broader goals of economic and cultural vitality, as well as environmental stewardship, to be achieved.

Natural resource conservation is important for human well-being, especially in fragile environments of developing countries. This study occurred in 2006 among 6,500 smallhold farmers residing along a 25-km segment of a heavily utilized river. Research objectives were to determine use and adoption constraints for 14 soil and water conservation practices (SWCPs). Farms were reportedly contributing to a decline in river water quality via soil erosion. Recent occupation of the upper watershed by immigrants magnified concerns that resource degradation could escalate. A multi-method approach incorporating quantitative surveys, qualitative interviews, and participant observation was used to interpret constraining factors within the biophysical and historical context of the watershed. Adoption rates for SWCPs were expected to be low (less than 20 percent). Increased formal education, income, access to information, and security of land tenure and soil characteristics, were expected to positively influence adoption. Data analysis included descriptive statistics and use of classification and regression trees. Results indicated that all sampled farms had adopted at least two SWCPs, with an average of six per farm. Favored practices were those that were easier to implement and more effective for resource protection and food production. Years in residence (tenure security) and income emerged as primary explanatory variables for adoption of SWCPs, while soil quality and formal education were secondary. Only 27 percent of surveyed farmers held title deeds, but the others perceived that land occupation conferred "ownership" and hence implemented SWCPs. A follow-up visit in 2009, after the region had endured a year of highly publicized ethnic conflict, immigration and farm expansion continued with SWCPs being adopted. Njoro communities mostly remained intact and appeared resilient. While small farms likely contribute to watershed-scale problems and declines in quality and quantity of water in the River Njoro, farmers have made remarkable strides--largely on their own--to conserve natural resources. Future research should examine how a general lack of infrastructure off-farm and study-site context contributes to reduced watershed-resource quality. Further protection of soil and water is best served by a more aggressive policy and extension education framework that links food security, household well-being, and natural resource management.