Published online by Cambridge University Press: 18 August 2016
In an experiment lasting from 1.5 years before first calving until third parturition, 24 purebred indigenous Bos indicus (Boran) cows and 24 Boran crosses with Holstein (proportionately 0.50 and 0.75 Holstein blood level) were given a diet composed of Bermuda grass hay and wheat bran (0.65: 0.35) offered at low, medium and high level. This level was adapted in amount to actual body weight every 2nd week corresponding to assumed 1.0, 1.2 and 1.4 times maintenance energy requirements. Cows were subjected to working exercise before and after first parturition. Body weight differences (lower with low and medium feeding level) developed mostly before calving for the first time and showed the typical decline at the start of lactation and increase in the dry period. Additionally, Boran cows given the high level increased their body weight from the first to the second lactation cycle. Body condition scores were higher in the Boran cows and, in both genotypes, with high feeding level. Independent of feeding level and genotype, calving intervals in cycles 1 and 2 were 530 and 421 days, respectively. Lactation length was considerably shorter in the Boran cows than in the crossbred cows. Milk yield, calculated over the first 13 weeks of lactation and over the whole lactation in both cycles, was 2.06 and 3.06 times higher in the crossbreds than in the Boran. Milk of Boran cows had 1.30, 1.15 and 1.20 times higher contents of fat, total solids and protein. In the first 13 weeks of lactation, milk yield of crossbreds with high feeding level (8.7 kg/day) was higher (P < 0.05) than that of the crossbreds with medium (6.5 kg/day) and low feeding level (5.4 kg/day), respectively. Boran cows never showed a significant response in milk yield to feeding level. Accordingly, the amount of organic matter intake required per kg milk increased with feeding level in the Boran cows while it remained unaffected in the crossbreds. Organic matter digestibility, as measured in lactation cycle 1, was higher (P < 0.01) in Boran. Response to high feeding level and estimated maintenance requirements of crossbred cows corresponded with current assumptions, but crossbreds subjected to medium feeding level expressed unexpectedly little difference from those receiving the low level. It seems that current recommendations are not generally applicable to indigenous tropical breeds as these responded differently from crossbreds to feeding level.