Non-linear Effects of Climate Variability on Tea Green Leaf Productivity in South Rift Kenya: A Generalised Additive Mixed Model Approach
Abstract
Tea is one of Kenya's leading agricultural exports and a major source of rural livelihoods and foreign exchange earnings, yet increasing climate variability threatens tea green leaf productivity in the forest-buffer zones managed by the Nyayo Tea Zones Development Corporation (NTZDC), and localised empirical evidence on how specific climate variables affect this productivity remains limited. This study evaluated the effects of temperature, rainfall and surface net solar radiation on tea green leaf productivity in the South Rift NTZDC-managed tea zones of Kericho, Nyangores and Olenguruone. Anchored on the Climate-Augmented Production Theory, the study drew on a longitudinal panel covering all 37 NTZDC management blocks over 2009-2024, yielding 516 block-year observations, with productivity averaging 9,875 kg/ha/year across a highly variable production environment. Panel data were drawn from NTZDC farm records, the Kenya Meteorological Department and the ERA5 reanalysis dataset. Generalised Additive Mixed Models (GAMMs) with zone- and block-level random effects were used to quantify the non-linear effects of temperature, wet- and dry-season rainfall, solar radiation and fertiliser on productivity. All smooth climatic predictors were statistically significant (p < 0.001), with solar radiation emerging as the strongest predictor; the model explained 74.6% of the deviance (adjusted R² = 0.713). Each climatic variable displayed a productivity-maximising range rather than a simple linear relationship: temperature peaked near 18.3°C, solar radiation peaked at approximately 14 MJ/m²/year, and fertiliser returns diminished beyond 550-570 kg/ha/year. The block random effect was highly significant while the zone random effect was not, indicating that climate exposure rather than zone identity drives the productivity differences observed between zones. These findings imply that climate adaptation interventions should be targeted at the block level and tailored to zone-specific climatic thresholds, particularly rainfall, solar radiation and fertiliser management, to safeguard tea green leaf productivity under a changing climate.