Maximization of lactation milk production without decreasing persistency

Area of Science:

• Animal Breeding and Genetics
• Dairy Science
• Quantitative Genetics

Background:

• Optimizing dairy cow lactation curves is crucial for milk production efficiency.
• Balancing genetic gains for milk yield and lactation persistency presents a significant challenge in breeding programs.
• Estimated Breeding Values (EBVs) are key metrics for genetic selection in livestock.

Purpose of the Study:

• To evaluate six distinct selection strategies for enhancing daily lactation curves in dairy animals.
• To identify strategies that improve lactation milk yield without compromising lactation persistency.
• To develop a genetic method for modifying lactation curves by controlling differential genetic gains across lactation stages.

Main Methods:

• Daily Estimated Breeding Values (EBVs) for lactation were treated as separate traits.
• Six selection strategies were compared: index I(R1) (equal gains at DIM 60 & 280), I(R2) (zero gain at DIM 60), I(d) (desired gains), I(u) (unstandardized EBV & persistency), I(w) (standardized EBV & persistency), and conventional EBV(L) selection.
• Lactation milk yield and persistency responses were assessed for each strategy.

Main Results:

• Index I(R2) maximized persistency but minimized milk yield response, indicating impracticality in inhibiting peak yield.
• Index I(u) yielded similar milk response to conventional selection but decreased persistency.
• Indices I(R1) and I(d) maintained persistency, with I(R1) showing a superior lactation response (669 kg EBV) compared to I(d) (560 kg EBV). Index I(w) showed potential for improving both traits, dependent on economic weighting.

Conclusions:

• Index I(R1) is a practical strategy for enhancing milk yield while preserving lactation persistency.
• No single strategy excelled in both milk yield and persistency; I(w) offers a balanced approach contingent on trait economic values.
• Standardizing EBVs by their standard deviations is essential in index construction when variances differ significantly, as demonstrated by I(u) vs. I(w) responses.