Some abstracts do not have video files because ASAS was denied recording rights.
64
Identification of lameness using lying time, rumination time, neck activity, reticulorumen temperature, and milk yield
Early identification and treatment of lameness can reduce pain and negative performance related to the disease. However, producers frequently misidentify lame cows; therefore, automatic identification of lame cows is needed. The objective of this study was to identify lame cows using precision dairy monitoring technologies. Cows (n = 98) were housed at the University of Kentucky Coldstream Dairy from January 11, 2012 to May 3, 2013. Cows were equipped with an IceQube (IceRobotics, Edinburgh, Scotland) on their left rear leg, which measured daily lying time (LT), an HR tag (SCR Engineers ltd., Netanya, Israel) around their neck, which recorded daily rumination time (RUM) and neck activity (ACT), and a DVM bolus (DVM Systems, LLC Boulder, CO) that measured reticulorumen temperature (RETT). Milkline Milpro P4C (Gariga di Podenzano, Italy) milking system recorded daily milk yield (MY). Seasons were categorized as season one (December, January, and February), season two (March, April, and May), season three (June, July, and August), and season four (September, October, and November). Cow gait was assessed weekly using a one (sound cow) to five (severely lame cow) scale. General symmetry, tracking, spine curvature, head bob, speed, and abduction/adduction were scored individually. Final gait score was calculated as a weighted average of general symmetry (24.92%), tracking (20.38%), spine curvature (19.81%), head bob (13.77%), speed (13.12%), and abduction/adduction (8%). Cows that scored ≥ 2 overall were classified as lame. A generalized linear model using the GENMOD procedure in SAS (SAS Institute, Inc., Cary, NC) was used to evaluate the effects of parity, season, LT, RUM, ACT, RETT, and MY on overall gait scores and their two-way interactions. Stepwise backward elimination was used to remove non-significant interactions (P ≥ 0.05). Variables associated with a probability of being lame were an increase in LT by one hour, a decrease in RUM by 30 minutes, a decrease in ACT by 150 units, and an increase in RETT by one degree (Table 1). Milk yield was not associated with a probability of being lame (Table 1). When identifying lameness, using precision dairy monitoring technologies that determine LT, RUM, ACT, or RETT may be useful.
Keywords:
lameness, accelerometer, precision dairy farming, technology