Genetic and phenotypic parameters and trends of reproduction traits in Kenya Boran cows

A total of 26,233 calving records from three commercial ranches: Mogwooni, Galana and O'ljorai; located in high altitude, humid and low altitude, coastal, and medium altitude rangelands of Kenya, were available for the study over periods between 10 and 30 years. The traits studied were: age at first calving and calving interval. Ranch, year and month of calving, parity, sex previous parous status, coat colour, cow registration class effects were fitted and adjusted for as fixed effects depending on the trait. The mixed models included the same fixed effects plus animal and permanent environment as random effects. Variance components and genetic parameter estimates were obtained from univariate analyses within ranch, using the Derivative Free Restricted Maximum Likelihood (DFREML) program of Meyer (1991). Phenotypic and genetic trends were obtained using DFREML's mixed animal model 1 to estimate the phenotypic, genetic and environmental components which were then averaged and weighted for each animal records within year of entry for each trait within ranch. The respective component values were plotted over the years to give the yearly trends. Age at first calving phenotypically and environmentally deteriorated by 2.0 (+ or -) 2.6 and 5.7 (+ or -) 2.0 days/year in Mogwooni, while the genetic component improved at a rate of -0.35 days/year over the study period. Phenotypic changes in CI were -0.14 (+ or -) 0.08, 3.98 (+ or -) 1.44 and 1.41 (+ or -) 0.90 days/year for Mogwooni, Galana and O'ljorai herds respectively, while the corresponding genetic changes were: -0.14 (+ or-) 1 and -0.02 (+ or -) 0.05 days/year for Mogwooni and O'ljorai respectively, were desirable, while Galana cows registered a significant genetic deterioration (3.3 (+ or -) days/year) in CI over the same period. From the results, it can be concluded that: Little advantage were taken of the existing opportunities (improved environmental trends and substantial additive genetic variability among animals) and therefore insignificant genetic gains were achieved for traits, mainly due to use of inadequate selection criteria as well as inept animal evaluation methods. Use of more directly measurable reproductive traits such as scrotal circumference, coupled with the use of Best Linear Unbiased Predictor (BLUP) procedures for animal evaluation would enhance genetic improvements in overall performance of Kenya Boran cows under rangelands conditions and are therefore recommended.