Comparing Pixel Changes and Manual Observations for Mapping Broiler Activity during Dried Black Soldier Fly Larvae (Hermetia illucens) Provisioning

SIMPLE SUMMARY: Exercising broilers may reduce health problems as this benefits leg development and limits the time in which broilers’ skin is in contact with dirty litter. Scattering dried Black Soldier Fly Larvae (BSFL) in the whole pen could increase broiler activity. We studied the activity of broilers during dried BSFL feeding periods in more detail to determine how well BSFL provisioning maintains broiler activity. To do so, we assessed broiler activity, at four daily provisioning periods and at different ages. Apart from manual observations, we used a software that automatically measured pixel changes in successive videoframes as a proxy for broiler activity. These pixel changes were positively related to manual observations of activity. Both the automated method and manual observations detected that scattering dried BSFL increased activity. Activity in response to feeding periods was influenced by the time of day and was lowest at 17:00 in manual observations, and the other method indicated that this could be due to reduced activity of broilers without BFSL at the end of the day. Increased activity could enhance broilers’ quality of life, and future research is necessary to determine how benefits of providing dried larvae can maximized. ABSTRACT: Welfare issues in broilers include inactivity and poor leg health. Activity can be stimulated by providing whole dried Black Soldier Fly Larvae (BFSL), but it is unknown whether this effect changes over time. Therefore, effects of BSFL provisioning on broiler activity per minute were assessed in detail. Additionally, the use of pixel changes as a proxy for broiler activity was explored. Broilers were housed in groups of 28 (n = 4 pens, <33 kg/m(2)). Dried BSFL were scattered through the pen of treatment groups at 08:00, 11:00, 14:00, and 17:00. Activity was assessed per minute both manually using scan sampling, and automatically as percent pixel changes for 30 min around BSFL provisioning, and at 14, 21 and 28 days of age. Both methods were moderately positively correlated and showed that BSFL provisioning increased activity. Activity as assessed by manual observations decreased at 17:00, at 21, and 28 days in both groups. The pixel changes indicated that this could be due to low activity levels in the control group. Using pixel changes seems to be a promising, timesaving tool to score broiler activity, but future research is necessary to validate this method and determine how high activity can be maintained over time.