Abstract
Water insufficiency due to scarcity, poor distribution and high variability of rainfall in the Central Highlands of Kenya (CHK) and a decline in soil fertility have contributed to a continuous decrease in water productivity. To increase food production to match the growing population, soil management practices that ensure improved water productivity should be embraced. The study, therefore, sought to characterize rainfall and to establish the effects of selected soil management practices on water productivity in Tharaka-Nithi and Murang’a Counties. The field experiments were laid in Meru South sub-County, Tharaka-Nithi County and Gatanga sub-County, Murang’a County. Rainfall characterisation was done in seven Counties in the C; Meru, Embu, Nyeri, Kirinyaga and Kiambu plus the two Counties where the field experiments were laid. Statistical methods and programs used in rainfall characterisation included RAIN software, cumulative departure index (CDI), rainfall anomaly index (RAI), a coefficient of variation (CV), precipitation variability index (PVI), correlation coefficient, root mean square error and scatter plots. The field experiment was laid in a randomised complete block design with tillage and soil inputs as combined treatments. The treatments in each site were: control, sole mineral fertiliser, crop residues plus mineral fertiliser, crop residues plus mineral fertiliser plus animal manure, crop residues plus Tithonia diversifolia plus rock phosphate, crop residues plus animal manure plus legume intercrop, crop residues plus Tithonia diversifolia plus animal manure. The treatment combinations were laid under both conventional and minimum tillage (MT). Test crop was maize (Zea Mays L.), H516. Soil moisture, soil water productivity and soil physical properties (Bulk density, saturated hydraulic conductivity and aggregate stability) under different treatments were subjected to analysis of variance (ANOVA) in SAS 9.3 software. Differences between treatment means were separated using Duncan's Multiple Range Test (DMRT) at p≤0.05. Rainfall analysis showed that rainfall in the CHK was highly variable temporally and spatial though fairly distributed. There was a high frequency of dry spell with high probability (93%) of future occurrence. Satellite and observed rain gauge data showed close agreement at seasonal scale than at daily scale. Satellite estimates can be used to substitute observed rain gauge data. The experimental results showed treatments had significant effect on soil moisture at 0-10 cm, 10-20 cm, 20-30 cm and at 30-40 cm depths (p=0.0001). Treatment under conventional tillage (CT) plus crop residue plus Tithonia diversifolia plus animal manure (CtRTM) had the highest soil moisture. Treatments had a significant effect on soil water productivity at 0-10 cm, 10-20 cm, 20-30 cm and at 30-40 cm depths in Chuka while in Kandara, the significant difference was only at 0-10 cm depth. Treatments under CT plus crop residue plus mineral fertiliser with or without animal manure (CtRFM/ CtRF) had the highest soil water productivity. Treatments had no significant effect on soil bulk density, saturated hydraulic conductivity and aggregate stability in the two sites. The result of the experiment will be used by various stakeholders in agriculture for planning and decision-making purposes regarding water productivity. It will also provide a baseline for further research and development.