Skip to main content
ARS Home » Research » Publications at this Location » Publication #63820
Title: VARIOUS PERSISTENCY MEASURES AND RELATIONSHIPS WITH TOTAL, PARTIAL AND PEAK YIELDS

Author
item GENGLER, N. - GEMBLOUX, BELGIUM
item KEOWN, JEFFREY - UNIV. OF NEBRASKA
item Van Vleck, Lloyd

Submitted to: Brazilian Journal of Genetics
Publication Type: Proceedings
Publication Acceptance Date: 12/20/1994
Publication Date: N/A
Citation: N/A

Interpretive Summary: Persistency is a term used to describe the plot of daily lactation yield. A flat curve is generally considered more persistent than a curve that slopes strongly downward after peak yield. Seven mathematical definitions of persistency were studied. The two main types were based on 1) ratios of yields during segments of the lactation and 2) standard deviations of partial yields. Ratio methods were influenced by total yield as were the methods based on standard deviations. A new method based on standard deviations of segments of partial yields for the trimesters of lactation is related to the shape of the lactation curve but is not greatly influenced by total yield and is suggested to be the best measure of persistency. Estimates of breeding value for this measure of persistency could be used in an index with yield.

Technical Abstract: Seven persistency measures and their relationship with total, partial and peak yields were studied. Heritabilities, genetic and phenotypic correla- tions were estimated using mixed model equations with multiple trait restricted maximum likelihood (REML). Data were from 31482 first lactation Belgian Holstein-Friesian cows. Differences in direction and size of the correlations between measures of persistency and total yield were found. Two types of measures were used. Persistency measures based on ratios showed positive correlations to total yield. Correlations with partial yields also showed a similar pattern. Genetic correlations between ratio methods and peak yield were positive. Methods using standard deviations of test day yields showed a similar problem, but the influence of the level of total production was negative. Heritabilities of several persistency measures were above 0.10. After considering these results, a newly defined dmethod, based on the standard deviation of partial yields, emerged as the best choice to describe persistency and for use in future genetic evaluation of persistency.