Strategic test-day recording regimes to estimate lactation yield in tropical dairy animals

Table 2 Comparisons between estimation methods and sampling protocols based on the number of possible sires tested

Sampling protocol	Average peak and persistent tail			High peak and less persistent tail
Sampling protocol	Average bias (±σ_Δ)^†	Number daughters	Possible sires tested	Average bias (±σ_Δ)^†	Number daughters	Possible sires tested
Weekly	−9 (±53)	16.3	11.2	−5 (±56)	15.3	11.9
Monthly	−18 (±110)	19.3	35.4	−29 (±113)	16.8	40.7
Quarterly	−64 (±199)	28.0	66.9	−177 (±201)	20.9	89.6
6 Test-days/lact	5 (±163)	23.9	52.4	61 (±169)	19.2	65.3
5 Test-days/lact	7 (±174)	25.1	59.8	46 (±182)	19.8	75.8
4 Test-days/lact	10 (±188)	26.7	70.2	76 (±202)	21.0	89.3
Weekly	2 (±51)	16.3	11.2	7 (±54)	15.3	12.0
Monthly	14 (±102)	18.7	36.4	21 (±108)	16.6	41.1
Quarterly	33 (±174)	25.1	74.8	31 (±183)	19.9	94.3
6 Test-days/lact	20 (±134)	21.1	59.1	19 (±142)	17.9	70.0
5 Test-days/lact	23 (±145)	22.1	67.9	21 (±156)	18.5	81.1
4 Test-days/lact	27 (±156)	23.1	81.1	25 (±171)	19.2	97.6

The values reported in this table are the median values calculated from the TDSR tested that have at least 80 animals in common in the top 100 compared to the ‘true’ EBV; ^†the ‘Average Bias’ of each method is presented here with its residual standard deviation (σ_Δ, calculated using equation 12); the ‘Number Daughters’ reports the theoretical number of daughters that would be required given the σ_Δ of the estimation method to prove one sire with a reliability of 50%, and a heritability if 0.2 (calculated using equation 11); the ‘Possible Sires Tested’ shows the predicted number of sires according to equation 13 that could theoretically be proven (with a reliability of 50%) given the required number of daughters.

ISSN: 1297-9686