Food for rumination : developing novel feeding strategies to improve the welfare of veal calves

Summary of thesis entitled: “Food for Rumination – Developing novel feeding strategies to improve the welfare of veal calves”, Laura Webb Veal calves are typically fed high levels of milk replacer supplemented with solid feed, which tends to contain a relatively small roughage component. Feeding strategies used in veal production have been associated with welfare issues, including the development of abnormal oral behaviours (AOB) and poor gastrointestinal health. AOB include tongue playing, excessive oral manipulation of the environment, grazing of the coat of other calves, and sham chewing, and are thought to develop in calves when chewing activity (i.e. eating and rumination) is not adequately stimulated. Common gastrointestinal health issues include poor rumen development and lesions in the abomasum. The aim of this thesis was to develop novel feeding strategies to improve the welfare of veal calves, i.e. to minimise the development of AOB and gastrointestinal health disorders as well as maximise chewing activity. The EU legislation stipulates a minimum of 250 g of `fibrous feed' for 20 week-old calves, but this amount does not seem supported by previous research in terms of it optimising calf welfare. In addition, it does not specify what fibrous feed refers to in terms of source and particle length of roughage. Developing novel feeding strategies for calves necessitates a better understanding of how different roughage characteristics might affect behaviour and gastrointestinal health, and this is what was investigated in Chapter 2. Because none of the single roughage sources investigated were able to improve both behaviour and health, it is likely that a combination of roughage sources would be optimal. For example, an appropriate diet choice may include a combination of roughage sources that facilitate good ruminal papillae development (e.g. maize silage), minimise plaque formation, and encourage both rumen muscularisation and rumination (e.g. straw). This chapter also suggested that hay, as a roughage source with both high levels of structure and high levels of fermentable fibre, could achieve both objectives of encouraging rumination and rumen development. Hay, however, is not used in veal production due to its high iron content that would lead to darker meat colour, which is less preferred by consumers. In Chapter 3, different amounts of a solid feed mixture were fed to calves and behaviour was monitored. The results showed that calves fed no solid feed on top of their milk replacer still displayed a rumination-like behaviour, which was in previous literature referred to as `sham chewing'. This result gives an indication as to the importance of rumination in calves. Moreover, this chapter failed to find a straightforward linear relationship between amount of solid feed provided and level of AOB displayed. Certain amounts of solid feed were found to initially stimulate chewing activity to a high level, but later, as calves grew older and more experienced with roughage, failed to stimulate chewing above the level displayed by calves fed no solid feed. Providing such an amount of roughage seemed to be more detrimental in terms of behaviour than providing an amount that results in a constant level of chewing activity throughout the fattening period. In order to develop animal-friendly feeding strategies, it is important to know what the animals would choose when given free choice. Therefore, in Chapter 4, the feed preferences of calves for milk replacer, concentrate, hay, straw and maize silage were investigated. This study showed that at 6 months, calves selected on average approximately 1250 g dry matter (DM) milk replacer, 1000 g DM roughage and 2000 g DM concentrate. Although all calves with free choice showed high levels of chewing activity and subsequently low levels of AOB, large individual differences existed in intake levels and feed preferences. Moreover, outcomes were dependent on the variable used to assess preferences: i.e. intake (in g DM relative to metabolic body weight), duration of feeding, or number of visits to each diet component. On average, however, calves showed a preference for milk replacer, concentrate and hay, over straw and maize silage. In contrast to free choice testing, as was used in Chapter 4, double demand operant conditioning gives an indication as to the strength of a preference. In Chapter 5, different methods to analyse data collected from double demand operant conditioning studies were investigated. Due to the dependence level between the two resources presented simultaneously, i.e. at any given time the test animal can only work for one resource, it would seem that proportions of rewards achieved for one resource over the total number of rewards achieved for both resources would be an adequate dependent variable in this type of analysis. In Chapter 6 the statistical method developed in Chapter 5 was used to assess the preference of calves for long and chopped hay and straw, and their preference for hay versus straw. Two to five month-old calves learned the double demand operant task and were motivated to work for roughage on top of a high energy diet of milk replacer and concentrate. They showed a preference for long over chopped hay, but not for long over chopped straw, and showed a strong preference for hay over straw. In Chapter 7 it was investigated whether temperament might affect learning of a double demand operant task in calves. Studies in horses and voles previously found that certain individuals seemed unable to learn certain tasks. If one could find out why, individual training programs could be designed and non-learners would not be removed from studies, potentially avoiding biases in data due to only certain temperament profiles making it through the learning criteria. Chapter 7 gave some indication that temperament may affect learning in calves, and it is the first study in calves to do so. However, due to the low number of animals used, further research is necessary to confirm which temperamental traits affect learning ability in calves. Relationships between tongue playing and: 1) hypothesised measures of chronic stress, and 2) hypothesised temperamental traits were investigated in Chapter 8. Large individual differences in the performance of tongue playing in calves subjected to similar husbandry conditions were found. This suggests that although tongue playing might well be a warning sign for chronic stress, and hence poor welfare, individual variation in the propensity to tongue play in response to stressful conditions exists. This could be due to differences in temperament. In contrast to what theoretical papers suggest, calves that showed more tongue playing showed characteristics of a reactive coping style. This result is, however, consistent with previous experimental papers on calves and other species. Results from Chapters 2 to 8 were combined into the design of the experiment described in Chapter 9. In this chapter, various feeding strategies (i.e. different amounts of solid feed combined with different concentrate to roughage ratios, different types of ad libitum choice diets, and feeding milk replacer via an open bucket or automated milk dispenser[AMD]) were applied and the effect on behaviour was recorded. Rumination was mainly affected by roughage provision, regardless of concentrate provision. Therefore, increasing solid feed provision without increasing the roughage content would most likely have little effect on rumination, although it would probably increase eating time to a certain extent. Because of the timing of tongue playing and oral manipulation of the environment (found in both Chapters 3 and 9), we suggest that the first of these two AOB is related to chewing activity in general, whereas the second may be more related to anticipation of an upcoming meal and positive reinforcement of feeding behaviours following an unsatisfactory meal. Calves provided ad libitum access to long straw in racks showed high levels of chewing activity and low levels of AOB relative to calves that did not have access to a straw rack but otherwise received the same diet. Six-month-old calves with ad libitum access to straw, maize silage and concentrate (but a restricted milk replacer allowance of 1050 g DM/d) consumed on average approximately 900 g DM/d roughage and 2300 g DM concentrate at 6 months of age. Feeding milk replacer via an AMD seemed to have little impact on behaviour, although it led to lower levels of tongue playing at 15 wk relative to bucket-fed calves. In Chapter 10, I first reflect on possible underlying mechanisms of AOB and on the best methods to assess animal preferences. AOB seem to develop in veal calves due to a number of factors, starting with the thwarting of chewing activity, of which rumination at least is most likely a behavioural need. Other factors involved in the development of AOB include chronic stress resulting from the thwarting of chewing activity, anticipation of an upcoming meal, and positive reinforcement of feeding behaviours following a meal that was unsatisfactory. Of great importance is the understanding of individual variation in the propensity to develop AOB, because stereotypic behaviours in sub-optimal environments have been linked to improvements in welfare (relative to non-stereotyping animals). Ruminants seem to be able to select a diet that maximises their comfort. Developing feeding strategies to improve veal calf welfare, therefore, requires the assessment of calf feed preferences. Choice tests and cross point analysis of double demand functions are two possible methods for the assessment of animal preferences, and both these methods include drawbacks and benefits. In contrast to choice tests, double demand offers a setting that closer mimics the complexity of natural environments by imposing a cost on access to resources and enables quantification of the strength of preferences. However, this procedure requires appropriate statistical methods, which take into account the dependence structure between the two simultaneously available resources. Finally, practical implications of the research presented in this thesis are described in Chapter 10. The development of novel feeding strategies to improve the welfare of veal calves is challenged by individual differences in feed preferences, chewing efficiency, and behavioural response to chronic stress caused by inadequate feeding. The latter is demonstrated by only certain calves developing AOB when chewing activity is not stimulated enough by the feeding strategy, whilst others do not develop such behaviours. This complicates the evaluation of the effects of feeding strategy on veal calf behaviour. However, based on the results of this thesis and previous research it seems that young calves should first receive a diet that optimises rumen development, before receiving coarser roughages that stimulate chewing activity, rumen muscularisation, and minimise plaque and hairball prevalence in the rumen. Adequate amounts of roughage and concentrate at 6 months of age seem to be 1000 and 2000-3000 g DM, based on voluntary intake.