Broad portfolios, constrained transformation: Climate adaptation among agro-pastoral households in semi-arid Tanzania

Abstract

Climate adaptation in dryland agriculture is often assessed by counting individual practices, although the distribution of practices across households may conceal whether adaptation is sustained, effective or institutionally supported. This study examines the adaptation portfolio reported by agro-pastoral households in Bahi and Mpwapwa districts of semi-arid central Tanzania and identifies the institutional conditions that shape its depth. This study is guided by the Pressure and Release (PAR) model and Open Systems Theory. A convergent mixed-methods design combined a household survey of 305 randomly selected agro-pastoral households with 24 key informant interviews and four focus group discussions involving 40 participants. Descriptive statistics and village-level chi-square tests were integrated with thematic analysis. Most respondents perceived declining rainfall (91.1%) and increasing temperature (85.9%). Reported adaptation was extensive but concentrated in practices that could be implemented through household labour, existing assets and local knowledge: crop rotation (95.0%), adjustment of planting dates (88.6%), conservation agriculture (87.8%), intercropping (80.8%), use of climate information (80.3%), crop-residue feeding (79.0%) and seasonal livestock migration (76.9%). Practices requiring finance, infrastructure, secure land or specialized inputs were markedly less common, including small-scale irrigation (34.2%), introduction of new crops (30.1%) and access to credit (19.2%). Extension officers were the main reported source of adaptation knowledge (59.5%), followed by peer learning (27.6%) and local experience (12.9%). Qualitative evidence linked uneven adoption to irregular input supply, weak markets and storage, limited extension coverage, financial exclusion and competition over land and water. The findings indicate a broad but predominantly incremental adaptation portfolio, reflecting constrained transformation: households responded actively to climatic stress, yet persistent resource and institutional constraints limited their ability to sustain, combine and scale responses. This suggests that portfolio breadth alone is an insufficient indicator of adaptive depth. Strengthening adaptation therefore requires more than promoting additional practices: it requires localized climate services, reliable extension, finance linked to viable technologies, market and storage support, and negotiated land and water governance. Because the study is cross-sectional and practice adoption was self-reported, the results establish patterns of adaptation rather than causal evidence of reduced climate risk.