Final Report Project code: Prepared by: B.PDS.1307 Bob Kilgour and Edward Joshua & NSW Department of Primary Industries Date published: September 2015 PUBLISHED BY Meat & Livestock Australia Limited PO BOX 1961 NORTH SYDNEY NSW 2059 The relationship between arena behaviour and lamb rearing ability Meat & Livestock Australia acknowledges the matching funds provided by the Australian Government to support the research and development detailed in this publication. This publication is published by Meat & Livestock Australia Limited ABN 39 081 678 364 (MLA). Care is taken to ensure the accuracy of the information contained in this publication. However MLA cannot accept responsibility for the accuracy or completeness of the information or opinions contained in the publication. You should make your own enquiries before making decisions concerning your interests. Reproduction in whole or in part of this publication is prohibited without prior written consent of MLA.
Abstract Reproductive wastage is a major limiting factor in the Australian sheep industry and lamb mortality has been shown to be a major component of this. While management options are available, they incur an annual cost and genetic selection is a preferred option. However, reproductive rate is made up of many components and these are of low heritability, difficult to measure and limited to one or other of the sexes. However, the behaviour of sheep in an arena test has shown promise as an indirect selection criterion. In a study of sheep from the Trangie Fertility Flock, selected for the ability of its ewes to rear their lambs to weaning and an unselected flock, the Trangie Random Flock, the sheep from the Fertility Flock moved around the arena and bleated less than their Random Flock counterparts. The arena test was devised so as to assess the attraction of the test animal to a group of sheep at the end of the arena and the repulsion of a human sitting in front of this group of sheep. The differences between the two flocks were interpreted as the greater ability of the Fertility Flock ewes to tolerate social isolation, a possibly desirable trait on the birthsite. In this project, we measured the arena behaviour of ewes from several commercial and research flocks with lifetime reproductive performance measurements. We were anticipating that the ewes with the greater reproductive success would move around the arena and bleat less than ewes with lower reproductive success. However, in only one of the flocks was there a significant relationship between arena behaviour and reproductive performance and, in this case, the arena criterion was closest approach to the human in the arena. Page 1 of 13
Executive summary Reproductive rate is a major limitation to profitability in the Australian sheep industry and, despite considerable effort, lambing marking percentages remain at about 80%. When ovulation rate is taken into account, this amounts to about 67 lambs born per lamb conceived or about two-thirds of the potential. Potential means of overcoming this limitation are management and genetic selection. Genetic selection is a preferred option since it is permanent and does not involve the annual cost that management options do. However, reproductive performance has many components and these are distributed among the ewe, the ram and the lamb. Further, these components all have heritabilities of around 0.1 or lower, so the annual rate of progress is low. They are also relatively difficult to measure. One possible way to surmount these difficulties is to use indirect selection. That is, rather than selecting directly for reproductive rate or one of its components, selecting for a character that is highly correlated with reproductive rate, is more highly heritable, is easier to measure and is not restricted to the ewe, the ram or the lamb. One place to look for such a character is in a flock selected for high reproductive rate. This was used in the past when the Trangie Fertility Flock, selected for its ability to rear lambs to weaning was compared to a random-bred control, the Trangie Random Flock. These two flocks differed in their behaviour in an arena test, a test designed to assess the reaction of a sheep to the attraction of a group of sheep and the repulsion of a human. The conclusion was that the test indicated the greater tolerance of the Fertility Flock sheep to isolation from flockmates, possibly a desirable trait on the birthsite. This project was initiated to test this concept in several industry flocks rather than in selection flocks on research stations. Several producers were approached and, as a result of this, approximately 100 sheep from each of the following flocks were arena tested; the Information Nucleus Flock at the Cowra Agricultural Research and Advisory (Merinos and Border Leicester x Merinos), Poll Dorsets (Marocara Dorsets, Dubbo), Dorpers, Burrawang West, Ootha), Border Leicesters (New Armatree, Armatree) and Merinos (Centre Plus, Tullamore). The arena was an enclosure 13m x 3m into which the test sheep is placed. At one end of the arena was a pen containing approximately ten sheep from which the test sheep was taken. In front of this pen and sitting in the arena was a human. Thus the pen of sheep acts as the attraction, the human as the repulsion. The sheep was then assessed for its reaction to this situation in terms of closest approach to and furthest distance from the human, mean approach to the human, total distance moved and number of bleats. Initially, at Cowra, the duration of the test was ten minutes. However, analysis of the data indicated that the test could be considerably shortened so, subsequently, the test was conducted over five minutes. Despite the promise of the earlier work at Trangie, in only one case was there a significant relationship between any of the arena measures and number of lambs reared per lambing. The one exception was in the Merinos at Cowra where the ewes with the highest rearing success were those that approached closest to the human in the arena. This meant that the project did not produce the expected results, so no economic predictions, practice change or extension of results were possible. Page 2 of 13
Table of contents Background 4 Project objectives 5 Methodology 6 Results 7 Discussion 8 Conclusion 8 Acknowledgments 9 Appendix 9 Page 3 of 13
Background One of the major factors limiting profitability in the Australian sheep industry is reproductive performance and there are two options for improving it, namely, management and genetic selection. Management options include nutritional supplementation, shed lambing and reproductive technologies such as multiple ovulation, artificial insemination and embryo transfer. However, these strategies are not permanent and, therefore, incur an annual cost. On the other hand, genetic selection is permanent but also has its own set of limitations. Firstly, reproduction is a complex characteristic and involves serving capacity and semen quality in the ram as well as ovulation rate, expression of oestrus, embryonic survival and maternal ability in the ewe and the vigour of the newborn lamb. Secondly, this is further complicated by the fact that each of these components is of low heritability, often lower than 0.1 and also quite difficult to measure. Finally, phenotypic reproductive characteristics are sex-limited; only rams produce semen, only ewes give birth. One possible way of overcoming these limitations is to use indirect selection. That is, rather than select for, by way of example, number of lambs weaned, select for a characteristic that is correlated with number of lambs weaned but is more highly heritable, easier to measure and measurable in both sexes. One means of discovering such a characteristic is to study selection flocks. This opportunity was available at the Agricultural Research Centre at Trangie, NSW, where sheep from the Fertility Flock and the Random Flock were available for comparison. The Fertility Flock had been selected mainly for the ability of its ewes to rear their lambs to weaning, while the Random Flock was an unselected flock. The fact that the main selection pressure on Fertility Flock was rearing lambs to weaning is important; it is all very well to get lambs on the ground but, if they do not survive, they are wasted. This is all the more important a large studies of commercial flocks in Australia have indicated that lamb mortality is a major source of reproductive wastage. A three-year study of these two flocks unearthed two possible candidates for indirect selection because they differed between the two flocks. The first of these was pelvic size and the second was the behaviour of the ewes in an arena test. The arena test placed the test sheep in an enclosure 13mx3m from which it could not escape. At one end of the arena was a small group of sheep from which the test sheep had been taken. This would normally be expected to attract the test sheep to that end of the arena but, a human being sat in the arena in front of this pen of sheep. The test sheep was then faced with a situation of attraction to the sheep and repulsion by the seated human. The arena test was an adaptation of that devised by Fell and Shutt, who showed differences in the approach behaviour of recently-mulesed and unmulesed lambs to a standing human in the arena, in that the recently-mulesed lambs avoided the human to a much greater extent than the unmulesed lambs. While Fell and Shutt tested their sheep in groups of three or four, the sheep in the Trangie study were tested individually. This was because the maternal behaviour of the ewe is an individual characteristic and it was thought that the individual behaviour of the sheep in the arena might reflect this. This was given some support during the Trangie study because, in Page 4 of 13
an experiment that compared one sheep in the arena compared to four, the differences between the two flocks that manifested with one sheep under test disappeared when four were used. The differences that appeared between the two flocks were that the Fertility Flock ewes moved around the arena less and also bleated less than the Random Flock ewes. This result appeared to make perfect sense since, once her lamb is born, the ewe has to forget about the rest of the flock and attend to the newborn. The results of the arena test indicated that the ewes of the Fertility flock showed a greater tolerance to social isolation and so might be expected to pay greater attention to their lambs. The Trangie study also showed that the behaviour of the ewes in the arena were repeatable and were not influenced by whether the ewe was dry, pregnant or lactating. This indicated that it was a stable component of sheep behaviour. The Trangie work also showed that the same differences that appeared between the ewes of the two flocks also appeared in the rams. The arena test results therefore showed considerable promise. The measures were highly variable, not the 0,1,2 data that you get with numbers of lambs reared. They were easily measured in that the test could be conducted indoors over a period of ten minutes rather than out in a cold, wet, windy lambing paddock and that they could be measured in rams as well as ewes. The fact that they could be measured in the rams suggested that sires could be selected to improve the maternal ability of their daughters, resulting in much faster genetic progress. It appeared that the next step was to prove the concept in commercial flocks and, to that end, we approached the producer members. The producers who were approached were known to have pedigree information and the sheep at the Cowra Agricultural Research and Advisory Station were also known to have pedigree information. It seemed that all that had to be done was to measure the arena behaviour of the animals and relate it to their lamb rearing performance over their lifetimes. The flocks chosen also had a wide representation of the major breeds in the Australian sheep industry. As such, the project only just fell within the realm of a producer demonstration site. Because the concept had only been demonstrated in two Merino research flocks, this was more an on-property research project using producers sheep rather than the demonstration of a solidly-proven principle. Project objectives The project had two objectives. The first was to examine the relationship between the behaviour of ewes in the arena and their lifetime rearing performance. Our prediction was that there would be a relationship between the number of bleats in the arena test and the total distance moved in the test with lifetime rearing performance. The expectation was that the ewes that reared the greater number of ewes per lambing would bleat and move around the arena less. Page 5 of 13
The second objective was to assess how much the test could be shortened from the tenminute test used in the original Trangie study without losing the capacity to meaningfully measure behaviour. Methodology Approximately 100 ewes from each of the members flocks (total of 6 flocks) were tested in an arena. These included Merinos and Border Leicesters (approximately 100 of each) from the Agricultural Research and Advisory Station, Cowra, Poll Dorsets from Marocara Dorsets at Dubbo, Dorpers from Burrawang West, Ootha, Border Leicesters at New Armatree, Armatree and Merinos from Centre Plus, Tullamore. The arena was an area measuring 13m x 3m and surrounded by sheep panels so that the animal was enclosed (Fig 1). The floor of the arena was marked out in squares measuring 1m x 1m, giving a total of 39 squares. At one end of the arena was a pen containing approximately ten ewes and it was from this pen that the animal to be tested was taken. At the end of the test, the animal was returned to this pen and another animal removed for testing. Sitting in front of this pen of sheep and inside the arena was a human being. Y X A C Fig 1. Plan of the arena indicating the arena proper (Y), A the pen containing the group of sheep from which the test animal was taken (A), the race along which the animal moved into the arena (B), the point of entry to the arena through a gate (C) and the position of the seated human (X). At Cowra, the test involved placing each of the test ewes in the arena for ten minutes and, at intervals of 15 seconds, recording the number of the square in which the animal s front left foot rested. From this, four behavioural measurements could be made. These were closest approach to the human, maximum distance away from the human, mean approach to the human and total distance moved. The number of bleats were also counted. At the other four properties, the conduct of the test was exactly the same except for the length of test which, based on the analysis of the Cowra data, indicated that shortening the test to five minutes did not decrease the reliability of the test. At Cowra, the tests were conducted under a roof and on concrete. At Burrawang West they were carried out under a roof and on bark chips, at Centre Plus under a roof and on bare earth and at Marocara and New Armatree tests were carried out in the yards. Data were analysed using generalised linear models procedures and tested for the significance of any relationship between closest approach, mean approach, maximum distance, total distance travelled and number of bleats with the number of lambs weaned per lambing. B Page 6 of 13
Results At Cowra, in the Border Leicester ewes there were no significant relationships between any of the arena data and number of lambs weaned par lambing. However, in the Merinos there was a significant (P<0.05) relationship between closest approach and number of lambs weaned per lambing (Fig 2). We have included data for these animals only since the scatterplots for all of the other flocks were essentially similar. There was also an almostsignificant (P<0.06) relationship between the number of bleats and number of lambs reared per lambing. These relationships indicated that ewes that reared the greater number of lambs per lambing were those that approached closer to the human and tended to bleat more during the arena test. 2.5 No of lambs weaned 2.0 1.5 1.0 0.5 0.0 0 2 4 6 8 10 12 Approach (m) Fig. 2. Simple scatterplot of the relationship between number of lambs weaned per lambing and closest approach. The Cowra data also indicated that the test could be significantly shortened without any loss of information about the behaviour of the sheep. In fact, the analysis showed that the test could be shortened to as little as one minute but, in order to err on the side of caution, we decided to shorten it to five minutes. In fact, the analysis showed that the coefficients of variation for mean approach and total travel, the two most highly variable measurements, were 38% and 74% respectively. These figures didn t change appreciably even when the test was reduced to only 5 records, where the coefficients of variation for mean approach and total travel become 38% and 83% respectively. In the other four flocks, there were no significant relationships between any of the arena data and the number of lambs weaned per lambing. Page 7 of 13
Discussion The project did not achieve the more meaningful objective of demonstrating a relationship between any of the arena behaviour measurements and lambs born per lambing. The project did achieve the less meaningful objective of shortening the duration of the test from ten minutes. As a result, the measurements that we made do not benefit the producers at all. Therefore there are no economic benefits, no changes that we can recommend to the producers, no improvement of their bottom line or any beneficial environmental effects. As a result, there were no meaningful results to extend to industry. The group worked because the collection of data was carried out over only a few days. This meant that the producers were not trying to juggle their operations with a project that was carried out over a protracted period. This meant that the producers were able to spend some time with us during the conduct of the arena tests, with most of them spending some time as the human in the arena. Conclusion While the original research on which the idea of using behavioural measurements as a means of genetically selecting for improved lamb rearing performance certainly showed great promise, this promise was not borne out in the current work. In fact, there was a significant relationship between arena behaviour and lambs weaned per lambing in only one of the six groups of sheep tested and this relationship was almost the opposite of what was expected. The following comments on the project are from the producer members. From Jenny and Craig Bradley, New Armatree, Armatree NSW: We thought the original idea was a good one and was very worthwhile pursuing. Anything that could improve lamb survival and profitability is worth looking at. There definitely needs to be more work done on strengthening the link between NLW (number of lambs weaned) and behavioural characteristics in sheep. In fact the whole NLW thing needs a lot more work a trait with a relatively low heritability has such a big influence on Maternal $ Index and also profitability. What we got out of the project was that while there was a huge variation in animal behaviour within our flock (and others I guess), it didn t reflect in their NLW figures and highlights how difficult it is to get a firm grasp on measureable reproductive traits. There needs to be follow up work to try to define the link. This was an excellent first step in the process and while the results and links between NLW and temperament were disappointing it should not be dismissed and further investigation would be worthwhile. The NLW trait needs to be worked on to firm up its accuracy within the Page 8 of 13
maternal breeds because it is such a big profit driver and the least reliable figure but one which has a big influence on selection. Acknowledgments We acknowledge the cooperation of the property owners, John, Angus and Alistair Kelly from Marocara Dorsets, Dubbo, NSW, Jana and Graham Pickles and Bill Royal, Burrawang West, Ootha, NSW, Jenny and Craig Bradley and Sam Seaman, New Armatree, Armatree, NSW, Robert, Pam and Mark Mortimer, Centre Plus Tullamore, NSW and Gordon Refshauge, Agricultural Research and Advisory Station, Cowra, NSW. We also gratefully acknowledge the technical assistance of Allie Jones and James Prescott, NSW DPI, Orange, NSW, Julie Sawley, Tooraweenah Prime Lamb Marketing Cooperation, Tooraweenah, NSW for administrative expertise and Gavin Melville, NSW DPI, Trangie, NSW for statistical analysis. Appendix The following photographs were taken during the conduct of the work Arena testing at the Cowra Agricultural Research and Advisory Station, with James Prescott sitting in the arena. Page 9 of 13
Arena testing Poll Dorsets at Marocara ; Allie Jones is sitting in the arena and Bob Kilgour is on the scaffolding collecting the data. Arena testing Dorpers at Burrawang West ; Jana Pickles is sitting in the arena and Bob Kilgour is on the scaffolding collecting the data. Page 10 of 13
Sally Bryant, ABC Western Plains NSW in the chair at :New Armatree JB Tancred from the Tooraweenah Prime Lamb Marketing Co-operative, Tooraweenah NSW takes the chair at New Armatree. Page 11 of 13
The project leaders Bob Kilgour (left) and Edward Joshua during a break in testing at New Armatree Page 12 of 13