Assessing the impact of climate variability and change on maize performance in semi-arid eastern Kenya

Abstract

Global climate change is elevating the agricultural challenge of feeding the growing world population by causing significant modifications to crop production environments. Effects of climate change and variability in semi-arid Kenya were studied through analysis of long term climate data of two locations, Katumani and Kambi ya Mawe, in the eastern province of the country. Kambi ya Mawe was chosen as an analogue site of Katumani. This analysis was accompanied by field . trials to determine the effect of climate change on performance of three maize varieties with different maturity periods thus, DH 04 (early maturing), DK 8031 (medium) and H 513 (late maturing), under two plant population densities of 37,000 (PI) and 24,700 (P2) plants ha-1 • In Katumani analysis of 57 years of rainfall data showed a decline in annual rainfall totals at a rate of 1.2 mm per annum (p.a) while in Kambi ya Mawe 44 years' record showed reductions at 2.1 mm p.a. Seasonal rainfall reduced at a rate of 0.15 and 0.69 mm pa for October-November-December (short rains) and March-April-May (long rains) periods at Katumani and 1.04 mm and 1.35 mm per annum at Kambi ya Mawe for the two seasons, respectively. The two sites had high chances (> 90% probability) of dry spells in the month of October and high percentages of seasons having less than 250mm rainfall. Mean maximum temperatures increased by 0.04 "C in 23 years at Katumani and 0.03 -c in 19 years at Kambi ya Mawe. Mean minimum temperatures were on the increase in Katumani at a rate of 0.03 "C p.a but on a declining rate of 0.18 -c p.a at Kambi yaMawe. During the experimental season rainfall was low and poorly distributed at both locations. Average temperatures were 4.8°C higher in Kambi ya Mawe than Katumani. There was significance difference (P ::s 0.05) between the varieties in biomass yield at 15 days after emergence (DAE) and at harvest at Katumani. Variety H 513 gave the highest biomass yield at harvest with a grand mean of 2314 kg ha-1 followed by DH 04 and DK 8031 with means of2160 and 1782 kg ha-1 , respectively (l.s.d 273.6). Significant difference in biomass yield was also found between populations at 15 and 30 DAE with population 1 yielding higher than population 2 at both stages. No significant difference was found between variety and population interactions. In Kambi ya Mawe, population effect was significant (P ::s 0.05) at 30 DAE and at harvest with PI yielding higher than P2. Effects of varieties or the interactions between varieties and populations on biomass were not significantly different at P ::s 0.05 throughout the period of growth in this location. Rainfall amounts and distribution as well as temperatures affected crop performance in both sites. Both biomass and grain yields were low partly as a result of moisture stress experienced during anthesis, silking and grain filling stages. Although there were no significant differences, higher biomass and grain yields were achieved at Kambi ya Mawe than Katumani. Higher biomass yields were also realized from the long duration variety at harvest than from the early maturing variety. Thus the long maturing variety with deeper penetrating roots gave higher yields in the hotter environment with higher surface evaporation and higher moisture contents in the lower than surface soil profiles.