Page 7 - Malou van der Sluis
P. 7

Abstract


                    Animal  agriculture,  including  broiler  farming,  has  intensified  to  meet  the
                    increasing  demand  for  livestock  products.  The  resulting  large  numbers  of
                    animals per farm make keeping track of individual broiler activity challenging.
                    However, individual activity records are of great relevance for assessing broiler
                    welfare and for implementation of activity in broiler breeding programs. In
                    this  thesis,  it  was  studied  whether,  and  how,  detailed  activity  data  on
                    individual group-housed broilers could be collected in an automated manner
                    throughout life. To this end, both an ultra-wideband (UWB) and a passive
                    radio frequency identification (RFID) tracking system were implemented to
                    collect data on distances moved by individual broilers. The recorded distances
                    were  compared  to  distances  recorded  on  video.  Both  systems  showed
                    moderate to good agreement with video and with each other. However, RFID
                    worked  with  smaller,  more  lightweight  tags  and  could  therefore  be
                    implemented earlier in life than UWB. Using the collected activity data, the
                    relationships between activity, leg health and body weight were studied, with
                    the ultimate goal of examining whether selecting on activity to improve health
                    and welfare of broilers would be feasible. Using UWB data, general indications
                    for relationships between gait classification and activity were observed, with
                    lower activity levels for birds with a suboptimal gait, but an interaction with
                    body  weight  was  also  observed.  It  remained  difficult  to  distinguish  gait
                    classifications  based  solely  on  distance  moved.  Using  RFID  data,  the
                    relationship  between  activity  patterns  early  in  life  and  average  daily  gain
                    (ADG) was further looked into. A negative correlation between ADG and the
                    root mean square error of activity was observed, indicating that broilers with
                    more deviations, in both directions, from the expected linear trend in activity
                    had a lower ADG. RFID data were also used to estimate the heritability of
                    activity.  An  estimated  heritability  of  0.31  was  observed  across  the  full
                    production  period.  Overall,  the  results  of  this  thesis  improve  our
                    understanding  of  the  relationships  between  activity,  leg  health  and  body
                    weight in broilers and could in the future potentially help to improve broiler
                    (leg) health and welfare, through selection on activity. Potential directions for
                    future  implementation  of  activity  tracking  in  larger  scale  broiler  systems
                    include less detailed RFID tracking and a sensor-fusion approach of RFID and
                    computer vision.
   2   3   4   5   6   7   8   9   10   11   12