FREQUENCY WEIGHTINGS FOR THE AVERSION OF BROILER-CHICKENS TO HORIZONTAL AND VERTICAL VIBRATION

During transport, broiler chickens experience motion and vibration at frequencies from 0.5 to 25 Hz, some of which they are likely to find a versive. The aversion of chickens to different frequencies (0.5-10 Hz) and root-mean-square accelerations (0-5 m/s(2)) of vertical and horiz ontal vibration was tested using a passive-avoidance, operant-conditio ning procedure. Chickens were trained to receive their daily food rati on by making a pre-selected, random number of pecks (between 1 and 39) at a computer-controlled, operant panel for repetitive small food rew ards during a fixed 2 h period each day. Additionally, every 20th peck initiated 30 s of motion with pre-determined characteristics. Thus, t hey were in conflict in their motivations to acquire their daily food ration and to avoid initiating the motion. The pecking rate in the exp erimental period measured the degree of aversion of the chickens to th e imposed vibration. Each chicken was studied for zero acceleration an d one other value. Four chickens were used in a completely randomized design for each of the 15 horizontal and 16 vertical motions. Models t o predict pecking rate from vibration frequency and root-mean-square a cceleration were obtained using non-linear regression analysis for ver tical and horizontal motions independently with more than 98.5% of the variance accounted for. These models were used to determine families of contours of equal pecking rates. Each contour joins co-ordinates of frequency and acceleration which the chickens find equally aversive. The higher root-mean-square accelerations and the lower frequencies we re most aversive. The chosen models give a linear response with the ef fect of frequency independent of acceleration magnitude. Thus, the eff ects of different vibration frequencies on the chickens are accounted for by frequency weightings derived from the inverse of any one of the contours. The frequency weighting gives the relative value at each fr equency by which the acceleration magnitude must be multiplied to weig ht it according to the level of chicken aversion at that frequency. At frequencies of f Hz, the weightings are 1.66/f(0.518) for vertical mo tion and 1.303/f(0.802) for horizontal motion. Root-mean-square weight ed accelerations derived from these frequency weightings may be used t o predict the likely level of broiler chicken aversion within the scal e of 0.16-0.315 m/s(2) indicating ''not aversive'', 0.5-1.00 m/s(2) '' aversive'' and 2.0-4.0 m/s(2) ''extremely aversive''. (C) 1997 Silsoe Research Institute.