The different measures of optical flow were not equally informative.
Contrary to our predictions and previous findings (Dawkins
et al. 2009), neither the mean flow rate nor the variance was
significantly correlated with gait score.We suggest that this may be
the result of the narrow range of gait scores found in this study.
Aydin et al. (2010) showed that although the rate of movement in
small groups of birds with the most abnormal gaits (Bristol Gait
scores of 4 and 5) was different from that of groups of birds with
scores of 0e3, there was a great deal of overlap among these
groups, so that it was difficult to distinguish between the movement
of birds with gait scores of 0,1, 2 or 3. None of the flocks in the
present study had an average gait score above 2.38. Mean flow rate
may therefore be sufficient to discriminate between healthy
walkers and very lame birds but less able to detect smaller differences
in walking ability such as we observed in this study. Skewness
and kurtosis, on the other hand, were both significantly
correlated with mortality, hockburn and gait score, even though the
24 flocks in this study were selected to be as similar as possible.
This suggests that these measures might be sensitive enough to be
useful on standard commercial flocks. [Direct comparisons with the