<![CDATA[Camel production systems are increasingly central to livelihood resilience, climate adaptation, and food security in the arid and semi-arid lands (ASALs) of Kenya, where recurrent droughts and climate variability have undermined conventional cattle-based systems. Consequently, pastoral communities have progressively shifted towards camel-dominant production strategies. Despite the growing economic and ecological significance, systematic empirical forecasting of national camel population trajectories remains limited. Most existing livestock assessments rely on descriptive statistics rather than formal stochastic modelling frameworks. This study developed a comparative univariate time-series forecasting framework. Specifically, it evaluated the forecasting performance of Autoregressive Integrated Moving Average (ARIMA), Holt's linear trend, Exponential Smoothing State Space (ETS), and Linear Trend models using the annual FAOSTAT livestock stock database for Kenya (1961 to 2024). Model performance was evaluated using stationarity diagnostics, residual independence tests, and out-of-sample forecast accuracy metrics, including Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE). Forecast comparisons were further evaluated using the Diebold–Mariano predictive accuracy test. The ARIMA (2,1,1) model achieved the lowest AIC (-46.55) and BIC (-36.25). However, the ETS (M, N, N) model demonstrated the highest predictive accuracy with the lowest forecast errors (RMSE = 622,710; MAE = 429,282; MAPE = 11.36%). The projections suggest that Kenya's camel population is likely to stabilise or moderately increase between 2025 and 2029, reflecting continued adaptation to climate variability. These findings provide valuable evidence for livestock planning, climate adaptation, and sustainable development of Kenya's camel sector.]]>
