Artificially reared herbivores are highly dependent on the stockperson at an early age in order to learn to drink from an artificial milk provider. This period of training may be a determinant for the animals' subsequent responses toward humans. However, long-term responses may also depend on the human contact (e.g., visual, physical interactions, gentling, and handling) provided to the young lambs after this training period. We examined whether different levels of subsequent contact (no visual and physical contact, stroking, and feeding reward) affect long-term attachment responses of lambs to the caretaker that provided the subsequent contact, after a common initial training period for artificial feeding provided by another person. Ewe lambs (n = 45) were artificially reared from multinippled buckets in groups of three. All the lambs were trained by a stockperson (S1) to suck from the bucket (4.4 +/- .3 sessions of 3 min per animal for the first 2 d of life). Subsequently, 15 lambs received no further human contact (T0). Fifteen other lambs received only stroking from a second stockperson (S2) for 6 min three times a day during the first 4 wk (T1). The remaining 15 lambs (T2) were stroked and bottle-fed by S2 during the same posttraining period as for T1. Tests were performed at 4, 6 (just before weaning), 9, and 13 wk of age in an unfamiliar arena marked in a grid pattern. The test procedure included three successive parts: 1) isolation for 1 min; 2) S2 presence for 2 min; and 3) isolation for 1 min. The T0 lambs spent a similar amount of time in the grid square close to S2 regardless of whether he was present. The T1 and T2 lambs spent more time close to S2 than T0 (P < .01), and T2 spent more time close than T1 (P < .05). In the presence of S2, T2 vocalized less (P < .01) than T0, and T1 did not differ from either T2 or T0. The T2 lambs also crossed fewer squares than T0. When S2 left the arena, T2 vocalized more (P < .01) than T0 and more (P < .05) than T1, and T1 had a tendency to vocalize more (P = .08) than T0. Differences persisted with increasing age. Human contact, especially stroking and feeding, during the 4 wk following initial training strongly and durably influenced the lambs' response not only to the appearance but also to the disappearance of a familiar stockperson. This last result supports the idea that lambs could form a social bond with their stockperson.

A 4-year interdisciplinary study was performed to compare two organically managed sheep production systems, using 118 ewes and 24 ha each. The systems differed in the ewes' reproduction rhythm: one lambing per ewe per year (1L/1Y), with the aim of balancing feed self-sufficiency and lamb marketing periods v. three lambings over 2 years (3L/2Y), with the aim of maximising ewes' productivity. The sustainability was evaluated through ewes' reproductive performance, lamb growth rate, carcass characteristics and quality, animal health, forage and feed self-sufficiency, soil mineral balance and gross margin. General animal health was assessed by recording ewe and lamb mortality and putative cause of death. Nematode digestive-tract strongyles parasitism was studied using faecal egg counts and necropsies. Carcass quality was assessed by recording carcass weight, conformation and fatness, and colour and firmness of subcutaneous fat. Thirty-three percent of 3L/2Y ewes lambed twice a year against 4% of 1L/1Y. Mean ewe productivity was 161.3% and 151.0% in 3L/2Y and 1L/1Y, respectively, and it was more variable between years in 3L/2Y. Average concentrate feed consumption and the corresponding cost per ewe were higher in 3L/2Y than in 1L/1Y (156 v. 121 kg, and h49.5 v. h39.3, respectively). Finally, average gross margin was lower in 3L/2Y than in 1L/1Y (h59 v. h65 per ewe, respectively). Even in year 2002, when ewe productivity was highest in 3L/2Y (193%), gross margin was not different between systems (h90 v. h86 per ewe in 3L/2Y and 1L/1Y, respectively), because of higher concentrate costs in 3L/2Y. The 3L/2Y animals presented a lower health status, with a higher lamb mortality ( P , 0.05) and a higher digestive-tract strongyles and coccidia parasitism level ( P , 0.05). Lamb carcass conformation, fatness and fat colour were not different between systems, but carcass weight and subcutaneous dorsal fat firmness were lower in 3L/2Y lambs than in 1L/1Y lambs ( P , 0.05 and ,0.001, respectively). Intensification in an organically managed sheep system, through an increased reproduction rhythm, thus did not lead to better economic results and proved riskier, more variable and more difficult to manage, and so less sustainable. The less intensive system (1L/1Y) was both highly efficient from the animal standpoint and highly feed self-sufficient. The technical and economic results of this system were better than those of organic private farms in the same area and matched those of non-organic farms.

We studied the ability of 32 lambs reared artificially in groups of four to discriminate between their shepherd and an unknown shepherd. Half of the lambs were bottle fed in isolation by one shepherd during the first 3 wk. The other half was fed alternately by three shepherds. Lambs had no visual contact with humans for the next 3 wk. Lambs were weaned at 6 wk of age and reared together with the minimum human contact necessary for rearing management. Lambs were tested at 3, 6, and 14 wk of age, investigating the effect of the rearing conditions on the response to isolation and to reunion with the known or an unknown shepherd. During tests, lambs were observed 1) in isolation for 1 min, 2) in the presence of a shepherd who entered and squatted at one end of the pen for 1 min, trying to touch the lambs if they approached, 3) again in isolation for 1 min. Early rearing management (one vs three shepherds) had no significant effect on any criteria studied. Lambs vocalized and moved less when in the presence of the shepherd than when isolated. They vocalized less, moved less, approached more quickly, and interacted more with the known than with an unknown shepherd. The difference persisted after 3 wk spent without visual human contact. However, no difference was evident at 14 wk of age. The effect of shepherd knowledge is clearly demonstrated by this experiment after an intensive early period of contact.

La maîtrise de la reproduction contribue à optimiser la productivité de l’exploitation. La réglementation de la production biologique n’impose pas de conduite de reproduction spécifique, mais limite les pratiques autorisées parmi l’ensemble des stratégies disponibles pour les éleveurs. Les conséquences de ces limitations sont plus ou moins pénalisantes selon l’espèce animale considérée. Ainsi, l’interdiction des traitements hormonaux représente le principal obstacle pour la maîtrise de la reproduction en agriculture biologique (AB) chez les petits ruminants et les porcins. La maîtrise de la saisonnalité de la reproduction est un objectif pour la plupart des éleveurs ovins et caprins. Dans ces deux espèces, la reproduction est possible tout au long de l’anoestrus saisonnier grâce à des traitements lumineux sans recours à la mélatonine, mais des difficultés de mise en œuvre font qu’ils sont aujourd’hui peu utilisés en AB. En élevage conventionnel ovin et caprin, l’insémination artificielle (IA) est réalisée à un moment fixe après synchronisation hormonale des ovulations et sans détection d’oestrus. En AB, l’IA est autorisée sur chaleurs naturelles : cela implique la pratique systématique de la détection des chaleurs sur plusieurs jours pour identifier les femelles à inséminer. Pour cette raison, l’IA est très peu pratiquée en AB chez ces deux espèces. La synchronisation des ovulations par effet mâle est une alternative prometteuse pour simplifier les protocoles d’IA, facilitant l’accès des éleveurs AB aux schémas de sélection et donc la généralisation de cette pratique. La conduite de reproduction la plus courante en élevage de porc conventionnel est dite en 7 bandes. La durée du cycle d’une bande est de 21 semaines dont 4 semaines d’allaitement. En AB, le règlement européen impose l’augmentation du temps d’allaitement des porcelets à 40 j minimum. En pratique, le sevrage a lieu à 42 j avec une conduite en 8 bandes, le cycle de reproduction s’étend donc à 23 semaines. En AB comme en conventionnel, le sevrage est la principale méthode utilisée pour la synchronisation de l’oestrus des truies dans leur bande. L’efficacité de la conduite en bandes est toutefois pénalisée en AB par l’interdiction des traitements hormonaux, utilisés fréquemment en conventionnel pour intégrer les cochettes dans les bandes, pour recycler des femelles décalées ou pour prévenir des oestrus de lactation. Cependant, l’interdiction de l’utilisation d’hormones ne pénalise pas la pratique de l’IA car elle est réalisée sur oestrus naturels synchronisés par le sevrage chez la majorité des producteurs en AB et en élevage conventionnel. Chez les bovins, l’insémination est pratiquée de manière générale sur chaleurs naturelles. Lors de la mise à la reproduction, la mise en place de la semence est réalisée par insémination artificielle (majoritaire en élevage laitier) ou par monte naturelle (majoritaire en élevage allaitant). L’interdiction des traitements hormonaux d’induction et de synchronisation des chaleurs n’est donc pas pénalisante en AB par rapport à l’élevage conventionnel. La mise à la reproduction est possible toute l’année, avec des contraintes qui diffèrent selon les objectifs de production. Les voies de recherche pour le développement de méthodes naturelles de maîtrise de la reproduction utilisables en AB sont très diverses, multidisciplinaires et souvent communes aux différentes espèces de mammifères d’élevage. L’INRA contribue fortement à l’amélioration des méthodes existantes et à leur adaptation aux contraintes d’élevage. Il s’implique également dans la conception et la validation de nouvelles approches de maîtrise de la reproduction sans hormones. Ce développement ne pourra connaître un véritable essor que dans le cadre d’une relation étroite entre la recherche finalisée et les professionnels.

No abstract