**4. Discussion**

Despite limitations in the number of animals enrolled in the study, the catching method was observed to affect several of the parameters recorded. Catching the broilers under the abdomen in an upright position was a faster method, gave slightly fewer wing fractures and gave a more consistent animal density in the drawers, compared to catching broilers by two legs and carrying them inverted

to the drawers. A comparison with one-leg catching was not performed since this method is prohibited in Norway [19].

There was a tendency for fewer wing fractures observed in broilers caught upright under the abdomen. An overall low prevalence of wing fractures was recorded (0.20%). Since the wings were examined for fractures in the lairage, a distinction between fractures due to catching and due to transportation could not be made. However, Jacobs et al. [7] investigated wing fractures post catching and post lairage. They found 1.88% wing fractures after catching and 1.90% in lairage, indicating that the fractures are typically attributable to catching and not to transportation. Available literature on wing fractures registered prior to processing is sparse. Kittelsen et al. investigated wing fractures prior to stunning and found 0.8% wing fractures in the observed broilers [9], a slightly higher number than observed in the current study. Langkabel et al. found no significant differences in the number of wing fractures between one-leg and two-leg catching [4], however, it was reported that the animals appeared to be more restless and performed more wing flapping during two-leg catching. In the current study, no one-leg catching was investigated, but catching by two legs gave more wing flapping than catching upright under the abdomen (anecdotal observation). The results from both studies indicate that the activity of the broilers during carrying did not affect the prevalence of wing fractures. It can be hypothesized that the actual crating process represents a larger risk for wing fractures than the carrying. This may explain the tendency for fewer fractures in broilers caught upright under the abdomen, since this method had more controlled placements of the birds in the drawers. There is a possibility that the catcher's behavior and handling of the birds were subject to observer effect. However, the catchers did not know which indicators would be assessed. The observation may have affected all handling equally, but it does not explain the differences between the catching methods.

Crating time was included in the study since the total time taken to catch an entire flock is important for broiler welfare, as well as the logistics and economy of the industry. Prolonged catching will influence time without feed and water, and therefore cause stress to the birds [20,21]. Crating time was directly and simultaneously influenced by both the hybrid and the catching method. Catching of the commercial fast-growing hybrid was quicker compared to the slower growing hybrid, regardless of method. This may be due to the weight of the slow-growing hybrid. The results also show that catching the broilers UPRIGHT, either one or two birds at a time, was faster than getting a grip of both legs of the animals, even though more birds could be collected in one catch with the two-leg method. In accordance with Langkabel et al. [4], the catchers were observed to fumble more to collect both legs than when they lifted one or two birds upright under the abdomen (anecdotal observation). This study did not consider how the level of practice and experience that the catchers had prior to the study may have influenced the crating time. The catchers perceived the abdominal catching to be most exhausting, since this method allowed fewer birds per catch and therefore more squats. The method's physical load on the catchers may affect crating time negatively if this method was performed in the entire flock, since prolonged depopulation time may increase stress and time without feed and water. However, it must be noted that too rapid handling also may be detrimental to bird welfare, as this may increase the risk of injuries, and as such, rapid handling by itself is not a goal. Few broilers were included in this study, compared with a commercial catching of larger flocks. Therefore, further studies are needed to assess the effect of the catching method on the crating time.

The number of broilers per drawer was higher when the birds were caught by LEGS, compared to UPRIGHT. In addition, the range in the number of birds per drawer was higher, making the crating density inconsistent with LEGS. The animal density is an important parameter during transport: If the density is too high, it gets more difficult for the birds to tackle high or low on-board temperatures and dead-on-arrivals may be the outcome [22–24]. The reason for this is that the mechanisms broilers use for cooling are less effective at higher densities [25]. Therefore, it is important to obtain the optimal density in each drawer. The reason for the significant difference in animal density between the two catching methods may be uneven "catches" and a larger range in the number of animals per catch. Especially with LEGS, the number of birds per catch varied between catches and between the catching personell. This made it difficult to keep track of the number of birds per drawer, especially since two catchers filled one drawer at the same time. The problem with crating density will likely be solved with more experience or a catcher designated to count the broilers.

In this pilot study, the catching method was applied randomly. However, due to the limited size of the study, LEGS was performed prior to UPRIGHT in both flocks. This may have affected our results since the catchers could have been more tired when starting abdominal catching. However, only a small proportion of the flocks were a part of this study, and the depopulation started with these two methods. Approximately 1000 broilers were caught by LEGS, followed by 1000 broilers by UPRIGHT in each flock. This equals approximately 3.33% and 5.95% of the flocks, respectively, which is only a small proportion of the animals that the catchers are used to handling each night. The impact of the catching method used when observers left the broiler house could have been used as a control, but this was not possible since the recording of wing fractures in lairage were very time-consuming. Including additional registration on controls would be negative for animal welfare in this study, as it would prolong the total crating time, and the time without feed and water for the entire flocks. In this study, the flocks were categorized as different hybrids. This may be misleading as there were several other factors that may have affected the results, e.g., live weight, management. This should be further investigated in future studies.

In previous studies on catching, bruising is included as a welfare indicator. In the current study, all registrations were observed in lairage, prior to the slaughter process. Bruising can only be observed after defeathering and was therefore not included as an indicator in this study. Several questions need to be addressed in a follow-up study with a larger sample size; these include the effect of crating time if the methods are applied in large flocks and more welfare indicators (e.g., stress evaluation and bruising).
