TECHNICAL BULLETIN 194 ISSN 0070-2315 GENOTYPE BY ENVIRONMENT INTERACTIONS IN PRODUCTION TRAITS OF SHEEP A.P. Mavrogenis, Chr. Papachristoforou and I. Papastylianou AGRICULTURAL RESEARCH INSTITUTE MINISTRY OF AGRICULTURE, NATURAL RESOURCES AND THE ENVIRONMENT NICOSIA CYPRUS OCTOBER 1998
Editor - in Chief Dr A.P. Mavrogenis, Agricultural Research Institute, Nicosia, Cyprus. All responsibility for the information in this publication remains with the author(s). The use of trade names does not imply endorsement of or discrimination against any product by the Agricultural Research Institute. 2
GENOTYPE BY ENVIRONMENT INTERACTIONS IN PRODUCTION TRAITS OF SHEEP A.P. Mavrogenis, Chr. Papachristoforou and I. Papastylianou SUMMARY Three purebred and two crossbred groups of ewes were evaluated for production and reproductive traits under two environments (semi-intensive and extensive production systems). The genetic groups were the Cyprus Fat-tailed (L), the Chios (C) and the Awassi (A) breeds and the C X L and A X L crossbreds. Litter size and litter weight at birth and at weaning and milk production after weaning adjusted for lactation length, were the traits evaluated. Lambing season significantly affected litter size at birth (P<0.05) and milk yield (P<0.01). Systems had a significant effect on litter weight at weaning (P<0.05) and milk yield (P<0.01). Breed group and ewe lactation number significantly affected all traits studied. Genotype by environment interactions were important for litter size and litter weight at weaning and for milk yield. Poor management, and especially poor feeding conditions, affect more adversely the more productive genotypes. ΠEPIΛHΨH Mελετήθηκε η παραγωγική και αναπαραγωγική συµπεριφορά τριών καθαρών φυλών και δύο διασταυρώσεων µέσα σε δύο διαφορετικά περιβάλλοντα (ηµιεντατικό και εντατικό σύστηµα διατροφής/διαχείρισης). Oι γενετικές οµάδες ήταν το Nτόπιο πρόβατο, το πρόβατο Xίου και το πρόβατο Aβάσσι, και η πρώτη διασταύρωση µεταξύ Xίου και Nτόπιου, και Aβάσσι και Nτόπιου. Eξετάστηκαν οι χαρακτήρες µέγεθος και βάρος τοκετο-οµάδας στη γέννα και τον απογαλακτισµό και η ολική εµπορεύσιµη παραγωγή γάλακτος (µετά τον απογαλακτισµό). H εποχή γέννας επηρέασε σηµαντικά µόνο το µέγεθος της τοκετο-οµάδας στη γέννα (P<0.05) και την παραγωγή γάλακτος (P<0.01). Tο σύστηµα διαχείρισης/διατροφής είχε ση- µαντική επίδραση στο βάρος της τοκετο-οµάδας (P<0.05) και στη γαλακτοπαραγωγή (P<0.01), αλλά δεν είχε επίδραση στους άλλους χαρακτήρες που µελετήθηκαν. Όλοι οι χαρακτήρες επηρεάστηκαν σηµαντικά από τη γενετική οµάδα όσο και από τη γαλακτική περίοδο του ζώου. Σηµαντικές αλληλεπιδράσεις µεταξύ γενότυπου (γενετικής οµάδας) και περιβάλλοντος (σύστηµα διαχείρισης/διατροφής) ση- µειώθηκαν στο µέγεθος και το βάρος της τοκετο-οµάδας στον απογαλακτισµό και στην παραγωγή γάλακτος. Συνθήκες στέρησης και ιδιαίτερα συνθήκες κακής διατροφής, επηρεάζουν περισσότερο τους πιο παραγωγικούς γενότυπους. INTRODUCTION In the Mediterranean area, sheep and goats are called to perform under a wide range of production systems. Natural selection and adaptation have produced breeds capable of performing in unfavorable as well as more favorable environments. Native breeds should, therefore, be utilized as a valuable genetic material in breeding programs. Introduction of new breeds to the diverse environments of the Mediterranean basin should always be associated with performance evaluation at least under particular production environments. Woolaston (1987) suggested that sheep breeds in Australia are still far from being properly adapted to the environment. Hence, we have to make improvements of the environment to accommodate the needs of new introduced breeds. Whatever is the case the phenomenon we are faced with is a genotype x environment interaction which should be thoroughly investigated. Woolaston (1987) concluded that genetic gains from selection on source traits, and specifically wool, are transferable to other environments, implying that there are no serious genotype x environment interactions. Such environments were nutritional regimes, age classes and sexes. On the other hand, Sanna et al. (1994) suggested that the trans- 3
fer of genetic progress from the nucleus to the commercial flocks is difficult, because of the lower adaptation of improved stock to traditional production systems (from an improved environment to a harsh one). It is probable, therefore, that some traits are more sensitive than others to changes in the production environment. With this in mind Bouix (1992) examined breeding objectives and selection criteria, with regard to adaptation of sheep in harsh environments. The objectives of the present work were to examine the performance of different genotypes in two different production systems. MATERIALS AND METHODS Data on three purebred and two crossbred groups of ewes were evaluated for two successive lambing seasons under two production systems. The groups were: Cyprus Fattailed (L), Chios (C), Awassi (A), C x L and A x L. Reciprocal crosses were not available. Ewes from all groups were allocated at random to the two production systems. Semi-intensive system (system 1). During the maintenance period, ewes received 0.5 to 0.75 kg of concentrates per head daily and wheat straw ad libitum. Over the last 6 weeks of pregnancy they received 1.0 kg of concentrates and wheat straw ad libitum. After lambing they were fed wheat straw ad libitum and concentrates according to requirements for lactation. Extensive system (system 2). The ewes were kept on stubble grazing and/or fed wheat straw throughout the maintenance and gestation periods. After lambing they received 1.2 to 1.4 kg of concentrates and wheat straw ad libitum. Allocation of ewes to the systems was made at random within breed and irrespective of age. Crossbred ewes were mostly in their first or second lactation. The ewes were housed by system and breed in adjacent pens. Weaning was at approximately 37 days post-partum regardless of system or breed. Following weaning, the ewes were hand milked twice daily and individual milk yields, fat and protein tests were recorded at monthly intervals. All lambs were fed the same diet (concentrates and cereal hay) from weaning until the age of 20 weeks. Data were analyzed using generalized least squares procedures (SAS, 1989) with unequal subclass numbers. The model (Table 1) included the effects of lambing season, system, lactation number of the ewe, breed group and the interaction of system by breed group on production traits. Traits examined included litter size and litter weight at birth and at weaning, and total milk production adjusted for lactation length. Milk yield refers to the total milk production after weaning. RESULTS Lambing season had a significant effect (P<0.01) only on litter size at birth and milk production (Table 1). This can probably be attributed to the fact that ewe performance in the extensive system was more dependent on climatic conditions and vegetation than in the semi-intensive system. Both years were extremely dry.production system significantly affected litter weight at weaning and milk production (Table 1). Ewes in the semiintensive production system had a higher litter weight at birth and at weaning, and higher milk production than those on the exten- Table 1. Mean squares and tests of significance for ewe production and reproductive traits Source df LSB LWTB LSWN LWTWN MILK Lambing season 1 1.98** 6.2 0.28 4 22031* System (S) 1 0.01 0.5 0.06 2118** 113027** Breed group (B) 4 12.48** 40.6** 4.25** 1125** 135050** Lactation number 5 1.76** 22.4** 0.61* 159* 10123* S x B Error 4 2203 0.11 0.24 2.6 4.1 0.84* 0.28 407** 62 27004** 5650 LSB=litter size at birth; LWTB=litter weight at birth; LSWN=litter size at weaning; LWTWN=litter weight at weaning; MILK=milk production after weaning. 4
Table 2. Production and reproductive characteristics of ewes by system of production and breed group Effect Subclass N LSB LWTB LSWN LWTWN MILK System Semi-intensive 1209 1.30 5.5 1.09 13.8 108 Extensive 1012 1.28 5.5 1.03 10.2 93 Breed group Local Fat-tailed (L) 437 1.22 5.5 1.07 10.0 74 Chios (C) 787 1.49 5.9 1.18 13.6 107 Awassi (A) 679 1.13 5.2 0.94 10.0 112 C x L 141 1.23 5.3 1.06 14.4 115 A x L 178 1.19 5.1 1.01 13.5 95 For explanation of abbreviations see Table 1. sive system of production (Table 2). Lactation number had a significant effect on all traits studied (Table 1). Milk production increased linearly with lactation length. The performance of purebred and crossbred ewes in both systems of production is given in Table 2. Chios ewes had the highest prolificacy at birth and at weaning than any other breed group. Litter weight at birth was also higher, whereas at weaning the crossbred groups (C x L and A x L) had a similar litter weight to that of Chios. Chios, Awassi and C x L genotypes had similar milk production, which was higher than that of the other breed groups. Interactions among production systems and breed groups were significant for litter size and litter weight at weaning and milk production (Table 1). The most probable explanation for the interaction appears to be the similar performance of genotypes in the extensive production system. It is evident that under this system genotypes with lower prolificacy are less seriously affected than those with high prolificacy (Table 3). A more serious interaction was observed for milk production. The milk yield of Chios ewes was higher in the semi-intensive and lower in the extensive system than that of Awassi ewes. A similar case was observed for breed groups C x L and A x L (Table 3), due to a change in scale of the two genetic groups in the two systems (semi-intensive and extensive). Although only two crossbred groups were represented in the systems, their performance may be indicative of what can be expected under such systems. Performance was better in the extensive environment and at least as good, if not better, as that of the Cyprus Fat-tailed sheep in the semiintensive system. DISCUSSION The response of purebred and crossbred sheep in the two production systems provides useful information on their future performance under such conditions. It was evident that productive and in some cases reproductive traits of ewes were seriously affected by the system of production. Milk production was the more seriously affected trait. Chios performed poorly in system 2 Table 3. Interaction effects of production system by breed group for traits studied System Breed N LSB LWTB LSWN LWTWN MILK Semi-intensive L 345 1.24 5.6 1.08 9.2 82 C 372 1.49 6.0 1.26 13.0 124 A 367 1.13 5.3 0.97 7.6 117 CXL 46 1.17 4.9 0.92 13.6 135 AXL 79 1.20 4.9 1.02 12.3 88 Extensive L 92 1.15 5.2 1.02 13.1 45 C 415 1.49 5.8 1.11 14.2 90 A 311 1.13 5.2 0.91 12.9 106 CXL 95 1.26 5.4 1.08 14.8 106 AXL 99 1.17 5.3 1.07 14.4 100 For explanation of abbreviations see Table 1. 5
and its production level deteriorated with deteriorating environment (feeding level), and the rank of the genotypes changed. The present findings are in agreement with Sanna et al. (1994) who suggested that genetic progress made in one environment is not easily transferable to an inferior environment because of the lower adaptation of the improved stock to this new environment. Similar findings have been reported by Mavrogenis and Louca (1980) who concluded that the milk production of different breeds was differentially affected in different environments. Conversely, Woolaston (1987) concluded that genetic gains from selection on some traits are transferable to other environments, implying that there are no genotype x environment interactions. No change in rank, however, was then reported. Since milk production is dependent on feed supply during the lactation period, it is not surprising that the more productive breed (Chios) was the one affected most. The results of the present study suggest that genotype x environment interactions on production traits of ewes are important. The evidence is that litter weight and milk production are more seriously affected. This leads to philosophical questions. Should we breed genotypes for different environments, or should we improve the environment in which genotypes are called to perform? Should we leave the environment (whatever we mean by this term) unchanged or improve it, without destroying it, so that better, more productive genotypes can perform to their potential? REFERENCES Bouix, J. 1992. Breeding objectives and criteria. Adaptation of sheep to harsh environmental conditions. INRA 179-184. Bunge, R., D.L. Thomas, and T.G. Nash. 1995. Performance of hair breeds and prolific wool breeds of sheep in Southern Illinois: lamb production of F1 adult ewes. Journal of Animal Science 75:1602-1608. Gabina, D., and F. Barillet. 1991. Actual tendencies for dairy sheep selection within the EEC. Information Technica Economica Agraria. Produccion Animal 87A: 227-234. Mavrogenis, A.P., and A. Louca. 1980. Effects of different husbandry systems on milk production of purebred and crossbred sheep. Animal Production 31: 171-176. Mavrogenis, A.P., and A. Constantinou. 1986. Performance of purebred and crossbred lambs. Technical Bulletin 77. Agricultural Research Institute, Nicosia. 5p. Sanna, S.R., A. Carta, E. Ugarte, F. Barillet, D. Gabina, B. Portolano, and S. Casu. 1994. Implementation of breeding schemes for dairy sheep in Mediterranean areas. Proceedings of the 45th Annual Meeting of the European Association for Animal Production. Edinburgh, September 5-8, 1994. SAS/STAT. 1989. Users guide, Version 6, fourth edition. SAS Institute inc., Cary, NC. Woolaston, R.R. 1987. Genotype x environment interactions and their possible impact on breeding programs. Proceedings of the National Symposium: Merino Improvement programs in Australia, pp. 421-437. Australian Wool Corporation, Melbourne, Vic. 6
P.I.O. 195/1998-400 Issued by the Press and Information Office, Nicosia Printed by Konos Ltd, tel. 465910, Nicosia