The objective of this study was to evaluate ewe and lamb performance in an accelerated lamb production system, and to compare the performance and lamb output between a conventional and an accelerated lamb production system. In the “Conventional” system, ewes were bred in March to lamb in August. The “Accelerated” system was based on the “STAR” system (Lewis et al., 1996), in which there were five breeding periods within each year. In the current experiment these were 14th January, 28th March, 9th June, 21st August and 2nd November. Progesterone was used to synchronise the breeding periods and during the non-breeding season, eCG was used to induce reproductive activity. Lambing began on each of these dates and weaning was 73 days later, coinciding with the next breeding period. The experiment ran over a three-year period beginning with breeding in March 2003 and was complete with the weaning of lambs from the January 2006-bred ewes. This resulted in 15 lambing and breeding periods over the three years in the Accelerated system and three lambing and breeding periods in the Conventional system. Average pregnancy rates were lower in the Accelerated system than in the Conventional system. Lamb growth rates were similar between the two systems, although lamb live weights at weaning were lower in the Accelerated system due to the age of the lambs at weaning (average = 69 vs 96 days). More lambs were born and weaned, resulting in more kilograms of lamb weaned in the Accelerated system relative to the Conventional system over the experimental period (26,200 vs 24,300 kg). Labour input was 35% higher in the Accelerated system, or 13% higher per lamb weaned. Average annual ewe energy requirements were 6% higher in the Accelerated system. Ewe energy requirements per kilogram of lamb weaned was lower (6%) in the Accelerated system due to more breeding and lambing periods per ewe per year. Laparoscopic observation of ewes’ ovaries at each breeding period revealed that most ewes had active ovaries and were therefore capable of successfully producing a viable foetus. In a subsequent experiment, blood samples were collected for analysis of progesterone concentrations from ewes bred during the spring and autumn breeding periods. Observations of data indicated that a small number of ewes conceived and lost their conceptus, or had abnormal corpora lutea. Results suggested that pregnancies were failing due to a lack of an appropriate signal from the embryo to the dam/uterus. Exposing Romney ewes to an artificial lighting regimen was unsuccessful for inducing reproductive activity during spring. In another experiment, melatonin implants administered to Romney ewes in spring and used in conjunction with eCG and progesterone, resulted in 61% more lambs born per ewe treated, compared to eCG and progesterone alone. This result indicated that melatonin implants, used with eCG and progesterone may be a suitable method for improving reproductive performance in sheep bred out of season in New Zealand. Delaying weaning of lambs and breeding lactating ewes can be used to obtain heavier lamb weaning weights in the Accelerated system. Spring-bred ewes had lambs weaned at either 69 days post partum or 90 days post partum. Reproductive performance was similar between the two groups of ewes, and lamb live weights in the later weaned group were heavier when lambs were 90 and 120 days of age. This research has shown that accelerated or out-of-season lamb production is an option for some New Zealand sheep farmers. However, the mechanisms associated with reproductive seasonality and methods of successfully circumventing this seasonality require further attention in order to achieve optimum reproductive performance.

L'étude des mécanismes comportementaux et cérébraux sous-tendant la modulation hédonique de la prise alimentaire chez le porc juvénile présente d'importants enjeux en termes d'applications en production porcine et en recherche biomédicale. L'objectif de cette thèse était de décrire ces déterminants par le biais du développement d'un modèle porcin des préférences et aversions alimentaires conditionnées, en utilisant des méthodes comportementales et des techniques d'imagerie cérébrale fonctionnelle. Sur le plan comportemental, nous avons validé un modèle porcin de l'aversion alimentaire conditionnée, obtenue via des injections duodénales de chlorure de lithium et caractérisée par une modification drastique des choix alimentaires. Nos résultats suggèrent également que les préférences alimentaires sont plus difficiles à induire que les aversions, et qu'un renforcement calorique et sucré, comme le saccharose, représente un stimulus efficace pour l'induction d'une préférence chez le porc. Sur le plan neurobiologique, l'exposition à des flaveurs alimentaires aux valeurs hédoniques contrastées et la perception combinée de saccharose aux niveaux oral et viscéral ont engendré des différences d'activation dans les circuits neuronaux impliqués chez l'Homme et les rongeurs, dans l'évaluation hédonique des stimuli sensoriels, la motivation, le traitement de la récompense (amygdale, cortex préfrontal, noyaux de la base) et la mémoire (hippocampe, cortex parahippocampique). Des études complémentaires sont nécessaires pour déterminer dans quelle mesure des facteurs de variation liés à l'animal ou aux paradigmes expérimentaux ont pu influencer l'expression des conditionnements.

This thesis is concerned with a theoretical simulation model for pig breeding, as part of the ongoing search for the “perfect” genotype. The starting point is an additive model to investigate how accurately the classical, infinitesimal model predicts genetic gain for traits controlled by few loci and few alleles. This initial investigation demonstrates that the infinitesimal model is robust, providing that at least 15 loci are controlling a trait and there is symmetry in the allele distributions. A Genotype-Pig (GE-Pig) model is then developed to apply the additive effects of alleles on sub-phenotypic traits like maximum protein deposition, minimum lipid to protein content in the whole body, ad libitum digestible energy intake, energy for maintenance requirement and water content in the whole body. These parameters are then used in a nutrient partitioning simulation model to growth a pig and calculate traditional breeding traits such as average daily gain, feed conversion ratio, and backfat thickness for any combination of alleles. Three algorithms, Genetic Algorithm, Tabu Search, and Simulated Annealing, are used to investigate the GE-Pig model and find optimal combination of alleles for different dietary and selection objective situations. The two diets investigated were either of a low or high quality, and the three selection objectives used were, maximising average daily gain, minimizing feed conversion ratio, and minimizing back fat. A graphical method is developed for easy comparison of the genotypes. Of the algorithms, the Genetic Algorithm performed the best, followed by Tabu Search and finally Simulated Annealing. It is demonstrated that, in general, there is a different, single, optimum for any given selection objective and diet. However under the back fat selection objective, both diets produce the same optimal genotype. Also there are many similarities between the optima for the average daily gain and feed conversion ratio selection objectives. When the theoretical minimum number of generations of selection to the optima is considered, the feed conversion ratio selection objective is the quickest for a breeding program to achieve the optimal solutions, followed by back fat, then average daily gain. It is demonstrated that diet also has an effect on the theoretical number of generations. A Multiple selection objective, using relative economic values applied to the individual selection objectives, is also investigated. For both diets, the majority of the multiple selection objective solutions are in the vicinity of the feed conversion ratio optima, indicating that feed conversion ratio is the most prominent factor. It is also demonstrated that the optimal solution is most affected by the objective parameter weights under low diet conditions.

The NZ dairy industry has adopted an encompassing measure of performance, total factor productivity (TFP), as a target measure to guide on-farm improvements. Dairy farmers pay a levy in order to fund agricultural research and extension. Extension services and R&D will continue to be of critical importance to maintain and improve productivity at the farm level. Consequently, it is in the best interest of the dairy industry to adequately target R&D and extension funds and make the best use of resources. To date, the methodology employed to estimate productivity growth has some shortcomings that seriously hamper the ability of potential users to extract useful information from it. First, productivity growth has been reported as an aggregate for the entire dairy industry. Second, it makes no assumption about the efficiency with which resources are being used. Third, it implicitly assumes that all farms face the same technology. Productivity growth can be achieved either through better (more efficient) use of the technology applied, through the adoption of a new technology (technical progress) or a combination of both. Given that the sources of productivity change—technical progress and technical efficiency change—are fundamentally different phenomena, they are, in turn, influenced by different factors. This distinction is important for policy orientation because different instruments/tools may be required to address them. Furthermore, empirical evidence suggests that a variety of farming systems have emerged as a result of dairy farming geographical expansion. Farm-level panel data were used to estimate the Malmquist productivity change index. This index can provide additional insights since it can be decomposed into two additional components, one that measures changes in technical efficiency (i.e., whether firms are getting closer to the production frontier over time), and one that measures changes in technology (i.e., whether the production frontier is moving outwards over time). Hence, it provides individual (farm) estimates of TFP. Moreover, the methodology applied allows to test whether farms in the two regions considered in this study are operating under the sametechnology. These two regions were the long-established dairy areas of Waikato-Taranaki and the newly developed dairy areas of Canterbury-Southland. Results for farms in Waikato-Taranaki indicate that annual TFP change is modest, ranging from 0.29% per annum to 0.59% per annum. Most importantly, technical progress is the only source of TFP change in all four models. Therefore, it is necessary to encourage investments in new R&D targeted to remove the technological constraints that impede the realisation of further productivity gains in the regions. However, important differences in the estimates of TFP, technical progress and change in technical efficiency between models were found for farms in Canterbury-Southland. Estimates of TFP change ranged from 0.7% per annum to 2.8% per annum. Even though technical progress and change in technical efficiency contributed to total factor productivity growth (TFPG), the latter component was the most important contributor in three of the four models. Moreover, in two models the rate of technical progress was negative (i.e., technical regress). The analyses indicate that dairy farms in Canterbury-Southland were on average 10% more productive than farms in Waikato-Taranaki when farms in both regions faced the frontier. These results were consistent for all the input/output set chosen. Furthermore, the null hypothesis that the two regions do not face the same production technology (i.e., each region has it own production frontier) was accepted irrespective of the input/output set chosen. The rejection of the null hypothesis, that farms in traditional and non-traditional dairy regions were operating under the same underlying technology (and hence face the same production frontier), called for a review of the traditional approach to R&D in one central experimental station, strengthening the need for a local approach through the promotion of networks and synergies with universities and other research institutions.

Appendices removed due to copyright restrictions Appendix 1: Beausoleil, N J, Stafford, KJ, Mellor, DJ. 2005. Sheep show more aversion to a dog than to a human in an arena test. Applied Animal Behaviour Science, 91 : 2 1 9-232. Appendix 2: Beausoleil, NJ, Stafford, KJ, MelIor, DJ. Does direct human eye contact function as a warning cue for domestic sheep, Ovis aries? Journal of Comparative Psychology, J 20 (3) : 269-279. Appendix 3:Beausoleil, NJ, Stafford, KJ, MelIor, DJ. 2004. Can we use change in core body temperature to evaluate stress in sheep? Proceedings of the New Zealand Society of Animal Production, 64: 72-76. Appendix 4: Beausoleil, N.J, Mellor, DJ, Stafford, KJ 2004. Methods for marking New Zealand wildlife: amphibians, reptiles and marine mammals. Department of Conservation, Wellington, New Zealand 147p. ISBN 0-478-2263 1-4. Appendix 5: Mellor, DJ, Beausoleil, NJ, Stafford, KJ 2004. Marking amphibians, reptiles and marine mammals: Animal welfare, practicalities and public perceptions in New Zealand. Miscellaneous Publication, Department of Conservation, Wellington, New Zealand 55p. ISBN 0-478-22563-6. / Both humans and dogs are integral in sheep production systems; however, which is more aversive to sheep, or indeed, whether either causes significant stress, has not been shown experimentally. The aim of this thesis was to examine some behavioural and physiological responses of domestic sheep to the presence of humans or dogs. An arena test was used to measure the relative aversion of sheep to the presence of a human or dog, as well as to elucidate differences in the responses of flocks at the University of Western Australia (UWA) which were putatively selected for differences in fearfulness. A Y maze preference test was used to 'ask' sheep whether they preferred a human shaking a rattle or a barking dog. In both tests, adrenocortical responses were measured concurrently to support the interpretation of behaviour. The presence of a human or dog in the arena elicited significantly more avoidance and vigilance behaviour and less exploration than did the presence of a control object. However, the dog elicited significantly more of this fear-related behaviour, and significantly larger adrenocortical responses than did the human. Sheep also expressed a clear preference for a human shaking a rattle over a barking dog in the Y maze test and exhibited larger adrenocortical responses to the dog than to the human in the Y maze facility. The UWA flocks differed in their expression of locomotor and vocal activity; MA sheep were more active/vocal than the other flocks, not only in the presence of the human but also with the box or dog. MA sheep expressed less avoidance and vigilance and more exploration than the other flocks in the presence of the human and exhibited significantly lower plasma cortisol concentrations than LA sheep after exposure to the human (10-min sample). However, there were no inter-flock differences in fear-related behaviour or adrenocortical responses when the flocks were presented with the box or dog. The results do not support the notion that the UWA flocks have been selected for differences in a consistent predisposition to react fearfully. The adrenocortical responses measured in these studies were only moderate in magnitude and duration, with peak plasma cortisol concentrations 2-3 times higher than pre-treatment values, and all concentrations returning to pre-treatment levels within one hour of the start of treatment. If these observations are confirmed in practical situations, the presence of humans and dogs during routine handling should cause little concern on the basis of animal welfare. However, limiting the presence of dogs in certain situations (e.g. before slaughter) may reduce stress in domestic sheep. Significant methodological developments in this research include the use of multivariate statistical techniques to analyze arena behaviour, the concurrent measurement of adrenocortical and behavioural responses in the arena and Y maze tests, and the explicit testing of the effects of individual lateral biases on choice behaviour in a Y maze test. Future studies should measure sheep stress responses to the presence of humans and dogs in practical situations.

Appendices removed due to copyright restrictions Appendix 1: Beausoleil, N J, Stafford, KJ, Mellor, DJ. 2005. Sheep show more aversion to a dog than to a human in an arena test. Applied Animal Behaviour Science, 91 : 2 1 9-232. Appendix 2: Beausoleil, NJ, Stafford, KJ, MelIor, DJ. Does direct human eye contact function as a warning cue for domestic sheep, Ovis aries? Journal of Comparative Psychology, J 20 (3) : 269-279. Appendix 3:Beausoleil, NJ, Stafford, KJ, MelIor, DJ. 2004. Can we use change in core body temperature to evaluate stress in sheep? Proceedings of the New Zealand Society of Animal Production, 64: 72-76. Appendix 4: Beausoleil, N.J, Mellor, DJ, Stafford, KJ 2004. Methods for marking New Zealand wildlife: amphibians, reptiles and marine mammals. Department of Conservation, Wellington, New Zealand 147p. ISBN 0-478-2263 1-4. Appendix 5: Mellor, DJ, Beausoleil, NJ, Stafford, KJ 2004. Marking amphibians, reptiles and marine mammals: Animal welfare, practicalities and public perceptions in New Zealand. Miscellaneous Publication, Department of Conservation, Wellington, New Zealand 55p. ISBN 0-478-22563-6. / Both humans and dogs are integral in sheep production systems; however, which is more aversive to sheep, or indeed, whether either causes significant stress, has not been shown experimentally. The aim of this thesis was to examine some behavioural and physiological responses of domestic sheep to the presence of humans or dogs. An arena test was used to measure the relative aversion of sheep to the presence of a human or dog, as well as to elucidate differences in the responses of flocks at the University of Western Australia (UWA) which were putatively selected for differences in fearfulness. A Y maze preference test was used to 'ask' sheep whether they preferred a human shaking a rattle or a barking dog. In both tests, adrenocortical responses were measured concurrently to support the interpretation of behaviour. The presence of a human or dog in the arena elicited significantly more avoidance and vigilance behaviour and less exploration than did the presence of a control object. However, the dog elicited significantly more of this fear-related behaviour, and significantly larger adrenocortical responses than did the human. Sheep also expressed a clear preference for a human shaking a rattle over a barking dog in the Y maze test and exhibited larger adrenocortical responses to the dog than to the human in the Y maze facility. The UWA flocks differed in their expression of locomotor and vocal activity; MA sheep were more active/vocal than the other flocks, not only in the presence of the human but also with the box or dog. MA sheep expressed less avoidance and vigilance and more exploration than the other flocks in the presence of the human and exhibited significantly lower plasma cortisol concentrations than LA sheep after exposure to the human (10-min sample). However, there were no inter-flock differences in fear-related behaviour or adrenocortical responses when the flocks were presented with the box or dog. The results do not support the notion that the UWA flocks have been selected for differences in a consistent predisposition to react fearfully. The adrenocortical responses measured in these studies were only moderate in magnitude and duration, with peak plasma cortisol concentrations 2-3 times higher than pre-treatment values, and all concentrations returning to pre-treatment levels within one hour of the start of treatment. If these observations are confirmed in practical situations, the presence of humans and dogs during routine handling should cause little concern on the basis of animal welfare. However, limiting the presence of dogs in certain situations (e.g. before slaughter) may reduce stress in domestic sheep. Significant methodological developments in this research include the use of multivariate statistical techniques to analyze arena behaviour, the concurrent measurement of adrenocortical and behavioural responses in the arena and Y maze tests, and the explicit testing of the effects of individual lateral biases on choice behaviour in a Y maze test. Future studies should measure sheep stress responses to the presence of humans and dogs in practical situations.

Appendices removed due to copyright restrictions Appendix 1: Beausoleil, N J, Stafford, KJ, Mellor, DJ. 2005. Sheep show more aversion to a dog than to a human in an arena test. Applied Animal Behaviour Science, 91 : 2 1 9-232. Appendix 2: Beausoleil, NJ, Stafford, KJ, MelIor, DJ. Does direct human eye contact function as a warning cue for domestic sheep, Ovis aries? Journal of Comparative Psychology, J 20 (3) : 269-279. Appendix 3:Beausoleil, NJ, Stafford, KJ, MelIor, DJ. 2004. Can we use change in core body temperature to evaluate stress in sheep? Proceedings of the New Zealand Society of Animal Production, 64: 72-76. Appendix 4: Beausoleil, N.J, Mellor, DJ, Stafford, KJ 2004. Methods for marking New Zealand wildlife: amphibians, reptiles and marine mammals. Department of Conservation, Wellington, New Zealand 147p. ISBN 0-478-2263 1-4. Appendix 5: Mellor, DJ, Beausoleil, NJ, Stafford, KJ 2004. Marking amphibians, reptiles and marine mammals: Animal welfare, practicalities and public perceptions in New Zealand. Miscellaneous Publication, Department of Conservation, Wellington, New Zealand 55p. ISBN 0-478-22563-6. / Both humans and dogs are integral in sheep production systems; however, which is more aversive to sheep, or indeed, whether either causes significant stress, has not been shown experimentally. The aim of this thesis was to examine some behavioural and physiological responses of domestic sheep to the presence of humans or dogs. An arena test was used to measure the relative aversion of sheep to the presence of a human or dog, as well as to elucidate differences in the responses of flocks at the University of Western Australia (UWA) which were putatively selected for differences in fearfulness. A Y maze preference test was used to 'ask' sheep whether they preferred a human shaking a rattle or a barking dog. In both tests, adrenocortical responses were measured concurrently to support the interpretation of behaviour. The presence of a human or dog in the arena elicited significantly more avoidance and vigilance behaviour and less exploration than did the presence of a control object. However, the dog elicited significantly more of this fear-related behaviour, and significantly larger adrenocortical responses than did the human. Sheep also expressed a clear preference for a human shaking a rattle over a barking dog in the Y maze test and exhibited larger adrenocortical responses to the dog than to the human in the Y maze facility. The UWA flocks differed in their expression of locomotor and vocal activity; MA sheep were more active/vocal than the other flocks, not only in the presence of the human but also with the box or dog. MA sheep expressed less avoidance and vigilance and more exploration than the other flocks in the presence of the human and exhibited significantly lower plasma cortisol concentrations than LA sheep after exposure to the human (10-min sample). However, there were no inter-flock differences in fear-related behaviour or adrenocortical responses when the flocks were presented with the box or dog. The results do not support the notion that the UWA flocks have been selected for differences in a consistent predisposition to react fearfully. The adrenocortical responses measured in these studies were only moderate in magnitude and duration, with peak plasma cortisol concentrations 2-3 times higher than pre-treatment values, and all concentrations returning to pre-treatment levels within one hour of the start of treatment. If these observations are confirmed in practical situations, the presence of humans and dogs during routine handling should cause little concern on the basis of animal welfare. However, limiting the presence of dogs in certain situations (e.g. before slaughter) may reduce stress in domestic sheep. Significant methodological developments in this research include the use of multivariate statistical techniques to analyze arena behaviour, the concurrent measurement of adrenocortical and behavioural responses in the arena and Y maze tests, and the explicit testing of the effects of individual lateral biases on choice behaviour in a Y maze test. Future studies should measure sheep stress responses to the presence of humans and dogs in practical situations.

The objectives of this research were firstly, to investigate if dairy cattle genotypes in NZ exhibit genetic variation in environmental sensitivity and to determine if this genetic variation is statistically significant from a genetic evaluation perspective, and secondly, to use genetic information including environmental sensitivity data to simulate dairy cattle responses to changes in nutritional regime and variation in climate. A comprehensive review identified that simulation models either overlook, or do not represent environmental sensitivity information where genotypes and breeds respond differently when exposed to variations in environment. A large dataset of daily and total lactation records (yields of milk, fat and protein) from herds participating in the progeny testing of sires from 1989 to 2002 was obtained to test for differences in the environmental sensitivity of dairy cattle in New Zealand. Production data was matched with environmental data relating to climate, herd size, altitude and herd average production levels (a proxy for feeding level). The statistical analyses applying univariate and bivariate multibreed models to environmental character states identified minimal sire re-ranking between environmental character states as measured by genetic and rank correlations. However, differences in yields of milk, fat and protein between New Zealand Jersey and overseas Holstein Friesian systematically diverged with production level, in herds expected to use different levels of supplements. These results suggest New Zealand Jersey cattle are best suited to a grassland-type environment, and overseas Holstein Friesian cattle are more suited to an intensive-type environment. A phenotypic analysis identified thermal environment (cold and hot conditions) significantly affected the expression of production traits in Holstein Friesian, New Zealand Jersey and Holstein Friesian x New Zealand Jersey cattle. Holstein Friesian dairy cattle were more susceptible to the effects of heat conditions than New Zealand Jersey cattle with yields of milk, and concentrations of fat and protein of the former compromised at a lower value for temperature humidity index. Dairy cattle performance is likely to be compromised by heat more frequently than cold conditions in New Zealand. A simulation model that considers how dairy cow genotypes respond to different environments, incorporating the results presented above, was then developed. An initial estimate of feed intake is used to define cow genetic potential based on estimated breeding values for total yields milk, fat and protein, and environmental sensitivity information. A mammary gland module then predicts daily yields of milk, fat and protein based on the cow's genetic potential after considering her age, stage of lactation, body condition score, nutritional status and thermal environment. Live weight change is also predicted via a body energy stores module, which considers the effect of age, stage of lactation, current body condition score, nutritional status, and an estimated breeding value for body condition score. Feed intake is predicted from the requirements for maintenance, growth and pregnancy, and the genetic drive for yields of milk, fat and protein and body fat change. The predictive ability of the model was tested using information from a prior study with two Holstein Friesian genotypes managed in a pasture-based system. The model simulated to a high degree of accuracy, mean values for yields of milk, fat and protein, and concentrations of fat and protein of each genotype. Various tests identified the major source of error between simulated and observed values were due to a lack of simulated variation. In conclusion, the extent of genetic variation in environmental sensitivity for total lactation yields of milk, fat and protein within the range of New Zealand environments are not sufficient to warrant the formation of separate breeding schemes for distinct environments. However, New Zealand Jersey cattle are best suited to a grassland-type environment, and overseas Holstein Friesian cattle are more suited to an intensive-type environment. Genetic variation in the suitability of different breeds for specific environments existed within breeds. A simulation model was developed that was able to simulate the effect of genotype, environment and genotypic differences in environmental sensitivity on daily cow performance.

Appendices removed due to copyright restrictions Appendix 1: Beausoleil, N J, Stafford, KJ, Mellor, DJ. 2005. Sheep show more aversion to a dog than to a human in an arena test. Applied Animal Behaviour Science, 91 : 2 1 9-232. Appendix 2: Beausoleil, NJ, Stafford, KJ, MelIor, DJ. Does direct human eye contact function as a warning cue for domestic sheep, Ovis aries? Journal of Comparative Psychology, J 20 (3) : 269-279. Appendix 3:Beausoleil, NJ, Stafford, KJ, MelIor, DJ. 2004. Can we use change in core body temperature to evaluate stress in sheep? Proceedings of the New Zealand Society of Animal Production, 64: 72-76. Appendix 4: Beausoleil, N.J, Mellor, DJ, Stafford, KJ 2004. Methods for marking New Zealand wildlife: amphibians, reptiles and marine mammals. Department of Conservation, Wellington, New Zealand 147p. ISBN 0-478-2263 1-4. Appendix 5: Mellor, DJ, Beausoleil, NJ, Stafford, KJ 2004. Marking amphibians, reptiles and marine mammals: Animal welfare, practicalities and public perceptions in New Zealand. Miscellaneous Publication, Department of Conservation, Wellington, New Zealand 55p. ISBN 0-478-22563-6. / Both humans and dogs are integral in sheep production systems; however, which is more aversive to sheep, or indeed, whether either causes significant stress, has not been shown experimentally. The aim of this thesis was to examine some behavioural and physiological responses of domestic sheep to the presence of humans or dogs. An arena test was used to measure the relative aversion of sheep to the presence of a human or dog, as well as to elucidate differences in the responses of flocks at the University of Western Australia (UWA) which were putatively selected for differences in fearfulness. A Y maze preference test was used to 'ask' sheep whether they preferred a human shaking a rattle or a barking dog. In both tests, adrenocortical responses were measured concurrently to support the interpretation of behaviour. The presence of a human or dog in the arena elicited significantly more avoidance and vigilance behaviour and less exploration than did the presence of a control object. However, the dog elicited significantly more of this fear-related behaviour, and significantly larger adrenocortical responses than did the human. Sheep also expressed a clear preference for a human shaking a rattle over a barking dog in the Y maze test and exhibited larger adrenocortical responses to the dog than to the human in the Y maze facility. The UWA flocks differed in their expression of locomotor and vocal activity; MA sheep were more active/vocal than the other flocks, not only in the presence of the human but also with the box or dog. MA sheep expressed less avoidance and vigilance and more exploration than the other flocks in the presence of the human and exhibited significantly lower plasma cortisol concentrations than LA sheep after exposure to the human (10-min sample). However, there were no inter-flock differences in fear-related behaviour or adrenocortical responses when the flocks were presented with the box or dog. The results do not support the notion that the UWA flocks have been selected for differences in a consistent predisposition to react fearfully. The adrenocortical responses measured in these studies were only moderate in magnitude and duration, with peak plasma cortisol concentrations 2-3 times higher than pre-treatment values, and all concentrations returning to pre-treatment levels within one hour of the start of treatment. If these observations are confirmed in practical situations, the presence of humans and dogs during routine handling should cause little concern on the basis of animal welfare. However, limiting the presence of dogs in certain situations (e.g. before slaughter) may reduce stress in domestic sheep. Significant methodological developments in this research include the use of multivariate statistical techniques to analyze arena behaviour, the concurrent measurement of adrenocortical and behavioural responses in the arena and Y maze tests, and the explicit testing of the effects of individual lateral biases on choice behaviour in a Y maze test. Future studies should measure sheep stress responses to the presence of humans and dogs in practical situations.

Appendices removed due to copyright restrictions Appendix 1: Beausoleil, N J, Stafford, KJ, Mellor, DJ. 2005. Sheep show more aversion to a dog than to a human in an arena test. Applied Animal Behaviour Science, 91 : 2 1 9-232. Appendix 2: Beausoleil, NJ, Stafford, KJ, MelIor, DJ. Does direct human eye contact function as a warning cue for domestic sheep, Ovis aries? Journal of Comparative Psychology, J 20 (3) : 269-279. Appendix 3:Beausoleil, NJ, Stafford, KJ, MelIor, DJ. 2004. Can we use change in core body temperature to evaluate stress in sheep? Proceedings of the New Zealand Society of Animal Production, 64: 72-76. Appendix 4: Beausoleil, N.J, Mellor, DJ, Stafford, KJ 2004. Methods for marking New Zealand wildlife: amphibians, reptiles and marine mammals. Department of Conservation, Wellington, New Zealand 147p. ISBN 0-478-2263 1-4. Appendix 5: Mellor, DJ, Beausoleil, NJ, Stafford, KJ 2004. Marking amphibians, reptiles and marine mammals: Animal welfare, practicalities and public perceptions in New Zealand. Miscellaneous Publication, Department of Conservation, Wellington, New Zealand 55p. ISBN 0-478-22563-6. / Both humans and dogs are integral in sheep production systems; however, which is more aversive to sheep, or indeed, whether either causes significant stress, has not been shown experimentally. The aim of this thesis was to examine some behavioural and physiological responses of domestic sheep to the presence of humans or dogs. An arena test was used to measure the relative aversion of sheep to the presence of a human or dog, as well as to elucidate differences in the responses of flocks at the University of Western Australia (UWA) which were putatively selected for differences in fearfulness. A Y maze preference test was used to 'ask' sheep whether they preferred a human shaking a rattle or a barking dog. In both tests, adrenocortical responses were measured concurrently to support the interpretation of behaviour. The presence of a human or dog in the arena elicited significantly more avoidance and vigilance behaviour and less exploration than did the presence of a control object. However, the dog elicited significantly more of this fear-related behaviour, and significantly larger adrenocortical responses than did the human. Sheep also expressed a clear preference for a human shaking a rattle over a barking dog in the Y maze test and exhibited larger adrenocortical responses to the dog than to the human in the Y maze facility. The UWA flocks differed in their expression of locomotor and vocal activity; MA sheep were more active/vocal than the other flocks, not only in the presence of the human but also with the box or dog. MA sheep expressed less avoidance and vigilance and more exploration than the other flocks in the presence of the human and exhibited significantly lower plasma cortisol concentrations than LA sheep after exposure to the human (10-min sample). However, there were no inter-flock differences in fear-related behaviour or adrenocortical responses when the flocks were presented with the box or dog. The results do not support the notion that the UWA flocks have been selected for differences in a consistent predisposition to react fearfully. The adrenocortical responses measured in these studies were only moderate in magnitude and duration, with peak plasma cortisol concentrations 2-3 times higher than pre-treatment values, and all concentrations returning to pre-treatment levels within one hour of the start of treatment. If these observations are confirmed in practical situations, the presence of humans and dogs during routine handling should cause little concern on the basis of animal welfare. However, limiting the presence of dogs in certain situations (e.g. before slaughter) may reduce stress in domestic sheep. Significant methodological developments in this research include the use of multivariate statistical techniques to analyze arena behaviour, the concurrent measurement of adrenocortical and behavioural responses in the arena and Y maze tests, and the explicit testing of the effects of individual lateral biases on choice behaviour in a Y maze test. Future studies should measure sheep stress responses to the presence of humans and dogs in practical situations.