Evaluation of a Linear Type Program in Holsteins

Abstract
Data of 18 traits in the linearized type appraisal program of Midwest Breeders Cooperative were analyzed. Mean scores ranged from 25.1-32.8, with scoring 1-50, except for disposition (mean 2.0 and scoring 1-3) and milkout (mean 2.1 and scoring 1-4). Parity, sire and stage of lactation were sources of variation for most traits. Evaluator by parity interaction, which indicated differential age adjustment by evaluators, was a source of variation for basic form, strength, udder depth and rear udder width. Heritabilities ranged from 0.07 for disposition to 0.67 for basic form, with most greater than 0.25. Heritabilities were similar to but slightly larger than estimates reported for a scoring system based on a small number of nonlinear classes. Basic form with strength, rear udder height with rear udder width, and teats with median suspensory ligament had genetic correlations in excess of 0.90. Phenotypic correlations were often small (< 0.10), possibly because of scoring linearly vs. in relation to an ideal. Type appraisal on a linear basis is recommended over scoring in relation to an ideal because heritabilities of linear traits were slightly larger and correlations between linearly scored traits are interpreted easily.