Animal Well-Being and Behavior

Unlike methods that require the corralling, catching and handling of individual birds (IN), the transect walks (TW) method involves only visual inspection of the flock. During TW, observers walk the length of the barn . TW have been applied to broilers and turkeys . We aimed to assess the feasibility of using TW to evaluate the welfare of commercial ducks by determining 1) whether the TW method produces comparable results with the currently used IN method, and 2) whether the number of transects through the barn affect results of assessments . Six commercial 30-d-old Pekin duck flocks (5850–6300 ducks/flock) were evaluated. One hundred 50 ducks (IN) and 7 transects (TW) were evaluated per flock. During TW, ducks within approximately one meter in front of observer and for the width of transect were scored and recorded . Two observers recorded the incidence of ducks in the following welfare indicators: feather quality (featherless areas <5 cm2 = FQ1; > 5 cm2 = FQ2), feather dirtiness (dirty areas <5 cm2 = FD1; > 5 cm2 = FD2), blood on feathers (presence of blood = BF), eye condition (dirt or staining around the eye = E1; eyes sealed shut or blind = E2), nostril condition (nostril blocked from inside = N1; nostril opening blocked = N2), and lethargic birds (“sick” appearance = S). Data were analyzed using Proc Mixed in SAS (v .9 .4) with barn as experimental unit . Obtained data were checked for observer effect then averaged between 2 observers . Lower proportions of ducks per flock were recorded in the FQ1, FQ2, FD1, FD2, BF, E1, and N1 categories using TW than IN (P < 0 .05 for all indicators) . Observed frequencies of indicators were not different across transect locations within a flock (P > 0 .05), except for BF (P < 0 .0001) . Bootstrapping analyses (calculated expected mean and SE for different numbers of transects) of TW data revealed that similar results could be obtained using as little as 20% of information (2 transects per house) . Method based differences in incidences of FQ and BF have also been reported for turkey flocks. There, the TW (but not IN) results corresponded to whole flock evaluations conducted during load out. It is not clear from our data whether TW or IN yield most accurate results for ducks . However, the similarities with previously reported results suggest that further research into the applicability of TW for on-farm duck welfare assessment is warranted .