Publications

The objective of this study was to define breeding objectives and consequently determine optimum crossbreeding levels for goats in the smallholder production systems. Profits and economic values (EVs) were estimated for four genotypes namely (a) original stock or local goat breeds with 0% German Alpine blood level (OS), (b) F1 with 50% German Alpine blood level (F1), (c) first backcross with 75% German Alpine blood level (B1) and (d) second backcross with 87.5% German Alpine blood level (B2). The EVs were estimated for average daily milk yield (DMY, kg), average post-weaning daily gain (ADG, kg), number of kids weaned (NKW), mature weight (MW, kg) and 12-month live weight (LW, kg). Profitability in Kenyan Shilling (KES) without risk was optimal (KES 6038.02) for the B1 genotype. Economic values without risk for most traits were highest for the F1 genotype, i.e., KES 64.85 (ADG), 24.02 (NKW), −27.55 (MW) and 84.51 (LW). There was, however, a 23% reduction in profits in the F1 genotype. It was evident that crossbreeding would improve the profitability of the smallholder farms, but not beyond the 75% grade level. A similar trend was observed when risk was incorporated. Differences in profitability with and without risk were less than 0.005% for all the genotypes. However, differences in EVs were large, ranging from −28% to +19%; DMY had the largest differences. Therefore, incorporation of risk in estimation of EVs for traits of importance is necessary. This study has also demonstrated that crossbreeding to a higher grade level is not necessarily compensated for by a high performance in most traits. Therefore, a crossbreeding program targeting B1 (75%) crossbreds would be desirable for implementation in the smallholder production systems.

Bulk properties and stability of the entire series of group 4d transition metal carbides and nitrides are reported in this work. The theoretical calculations were carried out within Local Density Approximation and Generalized Gradient Approximation using the Perdew, Burke and Ernzerhof exchange correlation functional. The generalized gradient approximation predictions were found to be closer to experimental values than the local density approximation predictions. In particular, LDA predictions were found to overestimate bulk moduli properties by as much as 5.6–11.5% while equilibrium lattice constants were found to be underestimated by as much as 0.2–5% compared to experimental values. On the other hand, GGA calculations were found to overestimate the lattice parameters by 0.2–6.9%, while underestimating the bulk moduli by as much as 0.07–5%. Out of the carbides considered, TcC and RuC were found to have the highest values of bulk moduli while YC and CdC had the lowest. Similarly, out of the nitrides, MoN and TcN were found to exhibit the largest bulk moduli, indicating that they were the hardest, while CdN had the lowest value and hence relatively softer. Overall, the nitrides presented higher values of bulk moduli than the carbides, an observation that is well supported by their correspondingly shorter bondlengths. The cohesive and structural properties of the 4d transition metal carbides and nitrides are also reported.