Utilisation of LIFE-SIM as a Management Decision Support Tool for Smallholder Farms in Kenya

The objective of the study was to evaluate the accuracy of Livestock Feeding Strategies Simulation (LIFE-SIM) model in prediction of dairy cattle productivity of smallholder dairy farming in Mbeere District of Eastern, Kenya. Longitudinal data using individual animal card were collected from 27 lactating dairy cows in six farms in each division for twelve months. The study covered farms with stall managed lactating dairy cows, where data on breed, age, parity, lactation length, daily milk yield, feed types and their daily intake, calving intervals and fortnightly weights were collected. Aryshire, Friesian and their crosses were reared and fed on roughages and concentrate on as is basis at 20.2 kg and 3.3 kg respectively in Siakago and 16.5 kg and 2 kg respectively in Evurore, with a range of 7-38 kg/day for roughages and 2.0–8.0 kg/day for concentrates. Feeds were sourced from both within and outside the farms. During the dry period, when forage feeds were scarce, banana stems and cereal by-products were used to sustain the animals. The model predicted lactation yield of 2351 ± 380kg equivalent to 7 kg/d (305 days), below potential (would-be) yield of 2649 ± 222kg. There was close correlation (0.97, 0.97 and 0.98) between the actual yield and the predicted yield for the three, early, mid and late, (1, 2, 3), stages of lactation respectively. The model predicted an annual methane and manure yield of 95.44 ± 28.98 litres and 1290 ± 532 Kg respectively while the predicted body weight change for 365 days was 0.44 ± 0.24 Kg per day. The model predicted the milk yield during the three stages of lactation, 1-3, of cows in smallholder dairy farms in the District to a reasonable degree of 96 %. The study concluded that the LIFE-SIM model predicted the milk yield accurately.