Review: A systematic review of dairy cow health, welfare, and behaviour in year-round loose range housing.

This systematic review compares the health, welfare, and behaviour of dairy cows in year-round loose housing systems against those kept in other housing systems in temperate regions. Year-round loose housing systems comprised housing where dairy cows had no access to the outdoors or only had access to a yard, pen or run. The comparator housing systems also comprised housing with and without outdoor access (including grazing). To contribute to evidence-informed policy, a systematic evidence evaluation was undertaken to assess the scientific evidence base for this question, and determine whether the evidence base is robust enough to determine any association between housing systems and health, welfare and natural behaviour in dairy cows. We assessed 11 181 references and reviewed 53 articles in detail following best practice guidance for systematic review. Seven different types of housing systems were compared and a total of 120 different Health, Welfare and Behaviour (HWB) outcomes were assessed, comprising 839 measurements for HWB. Results indicate both advantages and disadvantages of year-round loose-housing systems. These differences were not just between studies; there were also differences within-studies for individual HWB indicators. There was substantial heterogeneity in methods of collecting and measuring HWB outcomes across the studies; therefore, a robust statistical test (such as meta-analysis) of correlation between potential explanatory variables and HWB outcomes was not possible for any housing comparison or any individual HWB measurement. Assessing the evidence base systematically as a whole, there is only weak evidence that year-round loose-housing is either better or worse than housing systems with grazing for the health and welfare of dairy cows. There is also only weak evidence that year-round loose-housing is either better or worse than housing systems with any outdoor access, including but not limited to grazing, for the health and welfare of dairy cows. Variation in data reporting across studies is too great to allow robust statistical analysis of the direct effects of loose-housing systems and/or grazing on the health and welfare of dairy cows. Data are also often presented in an aggregated form that limits meaningful comparisons. For future research, data collected should be made freely available in a disaggregated form to enable robust meta-analysis to be conducted. In order to change policies and practices, based on evidence, more standardised primary research studies, measuring welfare indicators, including behaviour, are necessary.