ANIMAL-BASED INDICATORS FOR ON-FARM WELFARE ASSESSMENT ON CATTLE AND GOATS

Animal welfare is a scientific discipline in continuous development and updating, motivated by the increasing concerns from the European citizens. Recent researches highlighted the importance of the use of animal-based indicators to assess and evaluate the actual welfare status of farmed animals. Moreover, indicators should be feasible (concerning time and money consumption), as well as being valid and reliable. Many welfare assessment protocols have already been developed for various farmed species, such as cattle, pigs and poultry, but they still require refinement and deeper studies, in order to adapt the way of data collection to different situations, giving more appropriate results. In Chapter 1, the aim of the study was to investigate the long-term stability of two behavioural tests (Avoidance Distance – AD; Avoidance Distance at the Feeding rack – ADF) used to evaluate the human–animal relationship, in order to refine on-farm welfare assessment protocols for dairy cattle and to adapt them to alpine husbandry systems. Fifty lactating cows were tested three times (before grazing period – bg; during grazing period – dg; after grazing period – ag) during a year, following their management changes (indoor period vs outdoor summer grazing). Values of AD ag (113.93 ± 21.21 cm) were significantly higher (P < 0.05) than those observed in AD bg (71.07 ± 14.63 cm) and in AD dg (77.68 ± 18.11 cm). A similar trend was recorded also for ADF, which was significantly higher after the grazing period (ADF bg vs ADF ag: 25.00 ± 6.80 vs 47.50 ± 8.72 cm; P < 0.05). For certification purposes, the effect of summer grazing period should be borne in mind, as it may affect the reaction of cows towards humans. In Chapter 2, the aim of the investigation was to compare the prevalence of welfare indicators (integument alterations, lameness and malformations) of poor welfare in 612 dairy cows among five Italian cattle breeds with different milking production and selective pressure (local breeds vs high selected breeds) kept in tie-stalls in the Italian Alps under similar housing and management conditions. The local breeds presented a significantly lower prevalence of all the considered variables compared with the other more productive and selected breeds (P < 0.001). The effect of the breed significantly affected the welfare of dairy cows in tie-stalls in alpine traditional husbandry systems, as the prevalence of the negative welfare indicators studied was lower in local breeds, which were better adapted to local farming conditions. In addition to the refinement required for indicators on already investigated farmed species, the EU commission is presently requiring to develop welfare assessment protocols considering other farmed species in European countries not yet covered by previous research projects, such as small ruminants (sheep and goats), other poultry (turkeys) and equines (horses and donkeys). In order to answer to this request, we concentrated on goats and, in Chapter 3, a literature review on animal-based indicators for this species was carried out to gather information about promising indicators to be included into an on-farm welfare assessment protocol. Indicators were classified into four principles and twelve criteria. Twenty-five promising indicators were found, but in many cases actions are still required in order to assess their validity, reliability and feasibility. As expected, indicators related to health and disease are the more represented. Adult dairy goats are widely considered, as they represent the main category bred in Europe. More studies are needed in order to confirm the already existing results and to improve the development of effective on-farm welfare assessment protocols. In Chapter 4, the study investigated the feasibility in goats of an avoidance distance (AD) behaviour test set-up for cattle, and compared the results in the two species to assess the suitability of the test for on-farm evaluation of human-animal relationship in goats. The tests were performed on 324 lactating cows and 271 lactating goats, housed in free stall farms. Goats exhibited a higher level of confidence with humans, as showed by lower AD (goats: 68.60±4.98 cm; cows: 71.36±4.37 cm; p < 0.10) and higher frequency of contacts with the observer (goats: 45.8%; cows: 31.2%; p < 0.001). The AD test seemed feasible in goats, but its sensitivity needs to be improved, considering the different interactions of goats towards humans, compared to that of cows. In Chapter 5, the aim of the study was to gather information about the possibility of using the hair condition in goats as welfare indicator. This experiment was initially performed in Portugal, and then repeated in Italy, in order to increase the sample size and to test the validity of this indicator in different farming conditions. The results were comparable between the two repetitions and will therefore be presented together. Two homogeneous groups of 24 adult dairy goats with bad hair (RH: rough or scurfy hair) and 24 goats with good hair (NH: shiny and sheen hair) were subjected to general clinical inspection, evaluation of body condition score, hair and blood collection for the analysis of mineral content, inspection for the presence of abscesses, faecal sampling for the analysis of endo-parasites and skin inspection for the presence of ecto-parasites. Nutritional and health conditions were poorer in BH, compared to that of GH. Too thin goats were present with higher prevalence in BH (BCS at sternum area, p < 0.01; BCS at lumbar area, p < 0.001). BH also showed a higher prevalence of abnormal lung sounds (p < 0.001), compared to GH. Preliminary results are encouraging to state that the hair condition can be a valid and feasible indicator of goat welfare. The studies presented show the complexity of the animal welfare discipline. Many results have been achieved, but further efforts need to be accomplished in order to reduce errors in the welfare assessment.