Abstract #W5

# W5
Identifying farm-level factors affecting milking attendance and productivity in automated milking systems.
Meagan T. M. King*1, Ed A. Pajor2, Stephen J. Leblanc3, Trevor J. DeVries1, 1Department of Animal and Poultry Science, University of Guelph, Guelph, ON, Canada, 2Faculty of Veterinary Medicine, University of Calgary, Calgary, AB, Canada, 3Department of Population Medicine, Ontario Veterinary College, University of Guelph, Guelph, ON, Canada.

The objective of this study was to determine how barn design, herd management, and lameness prevalence relate to automated milking system (AMS) attendance and productivity. Data were collected from 26 AMS farms in Eastern Ontario, Canada. Farms averaged 108 ± 13 (mean ± SE) lactating cows and 2.3 ± 0.3 AMS units. The majority of farms used free-flow cow traffic (n = 24), while 2 farms employed a mixture of free-flow and milk-first traffic for different groups. Physical barn characteristics and stocking densities were recorded, and complemented by a questionnaire regarding feeding and bedding management. At each farm, 40 cows were gait scored (or 30% for herds > 130 cows) using a 5-point numerical rating system (1 = sound to 5 = lame). Cows were defined as lame with gait scores ≥ 3 (28.4 ± 2.8%) and severely lame as ≥ 4 (2.9 ± 0.7%). For a 6-d period, milking activity and production parameters for all cows at each farm were extracted from the respective AMS computers. Data were summarized across all cows and the 6-d observation period for each farm and analyzed in multivariable mixed-effect regression models. Milking frequency (2.9 ± 0.1 milkings/cow/d) increased with herds of greater average parity (P = 0.04; mean = 2.2 ± 0.4 lactations), greater frequency of feed push up/d (P = 0.04; mean = 9 ± 1.2 ×/d), and when stalls were raked ≥ 3×/d vs. 2×/d (P = 0.01), while it decreased with a greater prevalence of severe lameness (P = 0.008). Milk yield (32.6 ± 0.5 kg/d) increased with herd size (P = 0.01) and feed space/cow (P < 0.001; mean = 70.4 ± 2.6 cm/cow). Lying time (706 ± 11.8 min/d) was greater with concrete flooring (P = 0.045) as opposed to rubber flooring, and tended to decrease with lying stall neck rails placed further from the back curb (P = 0.07; mean = 172.0 ± 1.6 cm). Lying bouts (8.9 ± 0.3 no./d) were less frequent with a greater prevalence of lameness (P = 0.01), and tended to be less frequent with longer stall lengths (P = 0.07; mean = 248.9 ± 3.4 cm). These results demonstrate that behavior and productivity in AMS herds may be positively affected by modifiable housing and management practices and are negatively affected by a higher prevalence of lameness.

Key Words: automatic milking, dairy cow, behavior