Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Animal Genomics and Improvement Laboratory » Research » Publications at this Location » Publication #412254
Research Project: Increasing Accuracy of Genomic Prediction, Developing Algorithms, Selecting Markers, and Evaluating New Traits to Improve Dairy Cattle

Location: Animal Genomics and Improvement Laboratory

Title: Shifting paradigms in daily milk yield correction factors: the DeLorenzo-Wiggans method revisited

Author
item WU, XIAO-LIN - Council On Dairy Cattle Breeding
item CAPUTO, MALIA - Council On Dairy Cattle Breeding
item WIGGANS, GEORGE - Council On Dairy Cattle Breeding
item NORMAN, H - Council On Dairy Cattle Breeding
item Miles, Asha
item Van Tassell, Curtis - Curt
item Baldwin, Ransom - Randy
item BURCHARD, JAVIER - Council On Dairy Cattle Breeding
item DURR, JOAO - Council On Dairy Cattle Breeding

Submitted to: Journal of Dairy Science
Publication Type: Abstract Only
Publication Acceptance Date: 5/21/2024
Publication Date: 6/25/2024
Citation: Wu, X., Caputo, M.J., Wiggans, G.R., Norman, H.D., Miles, A.M., Van Tassell, C.P., Baldwin, R.L., Burchard, J., Durr, J. 2024. Shifting paradigms in daily milk yield correction factors: The DeLorenzo-Wiggans method revisited [abstract]. Journal of Dairy Science. 107(Suppl. 1):18-19(abstr. 1142).

Interpretive Summary:

Technical Abstract: Over the past decades, various methods have been proposed to derive discrete correction factors for estimating daily milk yields from partial yields. These methods assume constant correction factors within discretized milking interval classes but varying between classes. The recent introduction of continuous yield factors promises more accurate yield estimates. This study integrated continuous correction factors into existing methods, demonstrated using the DeLorenzo and Wiggans (D-W) method. The original D-W method is a two-step process: first, it calculates discrete multiplicative correction factors (MCF) for specific MIT classes, which are then finalized by a smoothing process. Such an approach often results in computational inefficiency and estimation biases. Our modifications involve replacing separate local regressions with a single global linear regression, circumventing the need for smoothing MCF. The daily yield is estimated by the product of MCF (denoted by F) and a partial yield, adjusted for additional effects or covariates. The modified D-W model equation is: y_i = F_(t_i)x_i + ß(d_i-d_0 ) + epsilon Here, y_i, x_i, t_i, d_i denote daily milk yield, partial yield, milking interval time, and days in milk, respectively, b_0, b_1, b_2, and gamma are regression coefficients, d_0 is a constant value, and epsilon is an error term. For cows milked twice daily, F_(t_i) = b_0 + b_1 t_i; for more frequently milked cows, F_(t_i) = b_0 + b_1 t_i + b_2 t_i^2. We evaluated the original and modified D-W methods using a dataset of 7,544 milking records from Holstein cows. The modified method showed, on average, a 1-2% improvement in accuracy over the original, with more significant gains when individual MIT deviated from the midpoint of each class and as the milking interval became more uneven. The modified method fully leverages the available data, provides more accurate estimates, and better captures the underlying relationship between your variables. Further enhancement could be achieved by treating DIM effects as non-linear. These findings would propose a paradigm shift from discrete to continuous correction factors for yield, enhancing computational efficiency and the accuracy of daily yield estimates.