Repositorio Institucional de Documentos

MATERNAL LEARNING AND CREEPFEEDING: TWO STRATEGIES DETERMINING THE POST-WEANING PERFORMANCE OF PIGLETS Laia Blavi Josa October 2013 José Francisco Pérez Hernández y David Solà Oriol Als meus pares, el Josep Mª i la Francesca. Per ser un exemple a seguir, i pels valors que m’heu transmès. COM EL BATEC DE L’AIGUA EN ELS NIELLS M’ho deia la mare: tu i jo, un dia, Pujarem a una petita barca i navegarem mar endins; serà tan petita la barca que només allargant una mica el braç podrem acaronar l’aigua, només allargant una mica el braç. Jo era petit, tanmateix petit, i la mare semblava tenir totes les forces per a ben menar la barca. M’ho va dir el vent, molt més tard: no sempre trobaràs un port o un amic; ni tan sols totes les promeses es poden complir. Jo havia crescut, de sobte, i els rems ja doblegaven les tremoloses mans de la mare. Sol i amb el llast de la urgència vaig fer-me a la mar. Pirata, un temps, vaig ser, em bategava el cor com l’aigua en els niells sempre a proa i amb vents favorables un tresor d’algues em floria a cada cala. Mal oratge, altre temps, s’esdevingué, i no tenia ni sabia llibants on amarrar la barca, nàufrag de mi mateix a la deriva fressava mars feréstegues sense ports ni albades. D’una bonança vaig aprendre que tot tal com arriba passa i ara m’he tornat pescador, pescador del temps que em resta i treballo ferm totes les jornades i cada vesprada, a atzur morent, deleixo sempre més aquella petita barca, una mà acaronant l’aigua, l’altra encaixada a la de la mare. Josep Civit i Mateu Montblanc, 2006 ACKNOWLEDGEMENTS ACKNOWLEDGEMENTS Primer de tot, vull donar les gràcies als meus directors. Francisco, mil grácias por abrirme la puerta de esta pequeña família, llamada nutrición, donde además de aprender he hecho unas buenas amistades. Pero sobretodo por ayudarme durante todo este año, por estar siempre disponible. Valoro muchísimo tus consejos, sin ellos este trabajo no sería el mismo. Al David, per tenir tanta paciència i per la seva dedicació en aquest treball, des dels temes pràctics de la granja fins a la redacció. Moltíssimes gràcies! A l’SNiBA haver-me donat l’oportunitat de treballar amb vosaltres, durant tot aquest any. Per ampliar els meus coneixements. Al Josep, per dedicar part del teu temps en aconsellar-me i en orientar-me. A la Lorena i a l’Olga. Al IAMZ i a Armando por haberme aceptado en este fantástico máster y por ayudarme siempre que lo he necesitado. A NOREL Animal Nutrition, S.A Spain. Por financiar y permitir que llevara a la práctica este proyecto. A tots els meus companys de Nutri! Amb els que he compartit despatx, amb els que he compartit estones de cafès, dinars, sopars, estones de riure, de sofriment, vaja de tot. Mil gràcies a tots vosaltres perquè sou un dels principals motius perquè vingués cada dia amb un somriure a treballar i un agraïment encara més especial aquells que vau dedicar part del vostre temps en ajudar-me. Rosa (mil gràcies per tot, ets un sol), Piero, Sergi, Cinta, Alexei, Jaime, Emili, Gemma, Ramon, Ester, Victor F, Roger, Clara, Sergio, Manel, Victor, Aline i Anna. Als professors del departament Susana, Roser, Ana Cris i Alfred. A la colla dels Hippie’s Dream, als amics de tota la vida i a les Vetes ArrRr per escoltar les meves preocupacions i distreure’m d’elles. Sou genials!! A tota la meva família, al meu germà Oriol, a les meves àvies Lola i Rosita, a la Padrineta i al Padrinet. Per estimar-me i estar al meu costat. Al Sisco i la Mª Serra per acollir-me com una més a casa vostra, a l’Oriol i a la Nuri. Al Marc (tutu!) perquè ets com un germà. Al Francesc, per entendre’m, escoltar-me, fer-me costat i sobretot per estimar-me. Desitjo que aquest camí que estem construint plegats, tingui un llarg recorregut. Als meus pares, la Francesca i el Josep Mª. Perquè sense vosaltres no seria la persona que sóc, pel vostre sacrifici, temps, consell, amor, per la vida que m’heu regalat. Us ESTIMO molt. Glossary Glossary CF Creep-Feed CS Conditioned Stimulus CTA Conditioned Taste Aversion ETEC Enterotoxigenic Escherichia coli G Gestation GIT Gastrointestinal Rract L Lactation SDBM Spray-Dried Blood Meal SDP Srpay-Dried Plasma TUTA Triple-U-Testing Arena UAB Universitat Autònoma de Barcelona US Unconditioned Stimulus W Weaning CONTENTS CONTENTS TABLES AND FIGURES ..................................................................................................................... I Index of Tables ........................................................................................................................... I The previous study. ................................................................................................................ I Review Literature ................................................................................................................... I Material and Methods............................................................................................................ I Results .................................................................................................................................... I Annex..................................................................................................................................... II Index of Figures ......................................................................................................................... II Review Literature .................................................................................................................. II Material and Methods.......................................................................................................... III Results .................................................................................................................................. III SUMMARY / RESUMEN / RESUMÉ ................................................................................................ V 1. INTRODUCTION ......................................................................................................................... 1 2. THE PREVIOUS STUDY ............................................................................................................... 5 3. LITERATURE REVIEW ................................................................................................................. 9 3.1 PIGLET AT WEANING: the low feed intake ........................................................................ 11 3.2 CREEP-FEEDING: effects of pre-weaning experience ........................................................ 15 3.2.1 Creep-feed intake ....................................................................................................... 16 3.2.2 Creep-feeding behaviour ............................................................................................ 17 3.2.3 When is creep-feed consumed? ................................................................................. 19 3.2.4 Does Creep-feed affects body weight at weaning? ................................................... 20 3.2.5 Post-weaning performance related to creep-feed .................................................... 20 3.2.6 Strategies to increase creep-feed consumption ........................................................ 21 3.3 LEARNING STRATEGIES: how to increase feed consumption ............................................ 23 3.3.1 Trial-and-error learning .............................................................................................. 23 3.3.2 Social learning ............................................................................................................ 25 3.3.3 Maternal learning ....................................................................................................... 26 4. HYPOTESIS AND OBJECTIVES ................................................................................................... 33 5. MATERIAL AND METHODS ...................................................................................................... 37 5.1 PRODUCTIVE AND PERFORMANCE OUTCOME (First Trial) ............................................... 39 5.1.1 Pre-weaning performance ......................................................................................... 39 SUMMARY VII In the commercial pig industry piglets are abruptly weaned at 21 or 28 days of age becoming the most critical period for pigs. During this period piglets suffer stress, neophobia to new food and anorexia in response to the sudden separation of piglets from their mothers, mixing with others litters, reallocation to a different environment and pass from a liquid diet (milk) to a solid food. There are some strategies to reduce the negative impact of weaning: (i) promoting the early creep-feed intake and number of eaters, (ii) utilising good quality, high digestibility and palatable feedstuff, (iii) including flavours and taste enhancers in the diet, but not always increases fed intake. Recent studies support the use of flavours in the sow and piglets diets as a means of promoting maternal learning. It has been suggested that flavour cues from the maternal diet can reach the foetus through the amniotic fluid and/or the placental blood stream. The early exposure of foetus to certain volatiles may result in a further preference for these compounds later in life and consequently can positively affect the acceptance of food containing a similar flavour before and after weaning. The present study is based on maternal learning to increase feed intake and performance at weaning. The flavour studied is Fluidarom 1003®, a commercial flavour based on >25 % of anethole and cinnamaldehyde and more than 10 % of eugenol as major volatile compounds. The study is divided in three trials, the first one previously carried on at SNiBA (Servei de Nutrició i Benestar Animal), what it was the basis for the hypothesis and objectives of our second and third trials. In this work we describe all the three trials to easily understand the global results. 1-First trial: the aim of this trial was to determine the effects on piglet performance of Fluidarom 1003® supplementation in the sows diets, during late gestation (28 days before farrowing) and lactation, and in the weaning diets. We measured sow’s performance and piglet’s productive parameters during pre-starter and starter phase (0-35 d). There was not difference in sow’s performance related to flavour. Higher feed intake and growth were observed along the pre-starter and starter phase (0 to 35 days post-weaning) mainly due to the maternal exposure to the flavour. Inclusion of Fluidarom 1003® in the sows diets was linked with the presence of flavour in the creep-feed, so piglets probably consumed a higher quantity of creep-feed (uncertain). However, we did not register creep-feed intake so were not able to distinguish the maternal effect from creep-feed effect. 2-Consequently, the second trial was based on the uncertain of the first. We aimed again to find out at what extent Fluidarom 1003® affect sow’s and piglet’s productive and performance parameters, including the determination of creep-feed consumption and evaluating the effect SUMMARY VIII of maternal learning. In order to evaluate the effect of maternal learning on Day 2 after weaning we conducted a double choice test (DCHT) for thirty minutes between two diets supplemented either with Fluidarom 1003® or Lacto-Vanilla (negative control) based on >25 % vanillin and more than 10% of butyric acid and diacetyl. Sow productive parameters did not show difference between treatments. Piglets consumed more unflavoured creep-feed, and those piglets perinatally exposed to flavour with unflavoured creep-feed showed the highest ADFI and BW. DCHT in control animals (without any previous exposure to flavours) showed a high preference for the Lacto-Vanilla. Those animals previously exposed to flavour by maternal diet and creep feed showed no preference for any aroma. But those piglets exposed perinatally to Fluidarom 1003® but not in the creep-feed showed a high preference for the Fluidarom 1003®. 3-In the third trial we go into detail about if each one, pre-natal (foetal environment) and postnatal (maternal milk) contact by their own is able to contribute to reinforce the positive reward and preference in piglets after weaning. We performed an experimental design to discern the independent role of Fluidarom 1003® inclusion (with or without) in late gestation, lactation and in creep-feed. We performed the double choice test (DCHT) like in the Trial 2. The results showed again that control pigs preferred the Lacto-Vanilla flavour. The other treatments did not prefer any flavour. The piglets prenatally exposed to the aroma but with unflavoured creep-feed consumed more feed with Fluidarom 1003® than Lacto-Vanilla, however it was not statistically significant. It is concluded that the inclusion of Fluidarom 1003® in the sow diets for late gestation and lactation improved piglets’ appetite when the same flavour is included at weaning. The positive reward associated to Fluidarom 1003® included in the sow diet is stronger if piglets are offered a non-flavoured creep feed. It was also observed that only exposure to the end of gestation would be enough to condition the piglets. Key words: creep-feed, maternal learning, flavour, piglet, preferences RESUMEN IX El destete de los lechones a edades tempranas (21-28 días) es la etapa más crítica en la producción intensiva de ganado porcino. Durante este periodo es frecuente que los lechones experimenten estrés, neofobia al pienso y anorexia provocada por la separación de la madre, el cambio de lugar y la dieta (de leche a pienso sólido). Para suavizar el destete se utilizan dietas palatables con alimentos frescos y de alta calidad, incluyendo aromasy potenciadores del sabor, que no siempre funcionan. También se da el pienso “creep-feed” durante la lactación para que los lechones se acostumbren al pienso sólido. Recientemente se ha vuelto a poner sobre la mesa la posibilidad de utilizar estrategias de aprendizaje temprano basado en la incorporación de aromas en las dietas maternas. Las claves aromáticas llegan hasta el feto mediante el líquido amniótico y/o el flujo sanguíneo placental. Esta previa exposición puede resultar en una preferencia por este aroma que permanece en el tiempo y en consecuencia, podría facilitar la aceptación del alimento con un aroma similar antes y después del destete. Este estudio se basa en el aprendizaje materno para mejorar la adaptación del lechón durante el destete. El aroma estudiado es el Fluidarom 1003®, un aroma comercial basado en >25 % de anetol y cinamaldehido y más de un 10 % de eugenol, como compuestos principales. El estudio está dividido en tres ensayos, el primero realizado previamente en el SNiBA (Servei de Nutrició i Benestar Animal) ha servido de punto de inicio para el planteamiento de las hipótesis y objetivos de nuestro segundo y tercer ensayo. En la presente memoria describiremos los 3 ensayos para que el lector pueda comprender mejor el conjunto de resultados. 1-En el primer ensayo el objetivo era determinar las consecuencias productivas de añadir Fluidarom 1003® en la dieta de las cerdas (final gestación y lactación), así como en la ración que se administraba a los lechones durante el destete. Se registraron los parámetros productivos de las cerdas y los rendimientos de los lechones durante la transición (0-35 d). No se observó ninguna diferencia en los rendimientos de las cerdas asociada a la administración del aroma. Se observó durante toda la transición (0-35 d) un incremento de la ingestión y del peso vivo de los lechones expuestos perinatalmente al aroma. De estos resultados planteamos la incertidumbre de si el aroma en la dieta de las cerdas y en el pienso “creep-feed” se daba conjuntamente, es posible que el efecto se deba en parte a que los lechones consumieron más cantidad de pienso “creep-feed”. Sin embargo, no se registró la ingestión de pienso “creepfeed” y no se pudo discernir el efecto maternal (gestación y lactación) del efecto “creepfeeding”. RESUMEN X 2-Partiendo de esta incertidumbre se realizó el segundo ensayo, donde esta vez se volvió a evaluar los efectos de la inclusión de Fluidarom 1003® pero teniendo en cuenta el factor “creep-feeding”. Se registraron los parámetros productivos de las cerdas y los rendimientos de los lechones durante la transición (0-14 días), así como el consumo del pienso “creep-feed”. Para poder evaluar el efecto del aprendizaje materno se realizó un test de preferencia doble (DCHT) durante treinta minutos a los 2 días post-destete entre el Fluidarom 1003® y el aroma Lácteo-Vainilla (control negativo) basado en más de un 25 % de vanilina y más de un 10 % de ácido butírico y diacetilo. No se observaron diferencias en los parámetros productivos de las cerdas. Los lechones, consumieron más pienso “creep-feed” sin aroma; y en la transición el mayor consumo y crecimiento se registró en los animales perinatalmente expuestos al Fluidarom 1003® pero sin aroma en el pienso “creep-feeding”. En el DCHT los animales control (los que no recibieron nunca aroma) mostraron una alta preferencia hacia el Lácteo-Vainilla. Sin embargo, los animales previamente expuestos al aroma mediante la dieta materna no mostraron preferencia por ningún aroma. Pero aquellos lechones expuestos perinatalmente al aroma pero no en el pienso “creep-feed” mostraron una mayor preferencia por el aroma. 3En el tercer ensayo se quiso profundizar más en conocer la etapa que resulta más importante para configurar el aprendizaje del lechón por el nuevo flavour. Se realizó un diseño experimental que nos permitiera discernir el papel independiente que ejerce la inclusión de Fluidarom 1003® (con o sin) en la dieta materna de final de gestación, de la lactación, y en el pienso “creep-feed” que se administraba a los lechones durante la lactación. Para ello se realizó un DCHT en las mismas condiciones que en el segundo ensayo. Los resultados volvieron a mostrar que los lechones control, preferían el aroma Lácteo-Vainilla. El resto de tratamientos no presentaron diferencias significativas en la preferencia entre aromas. Los lechones prenatalmente expuestos al aroma pero con un pienso creep-feed sin aroma añadido consumieron más pienso con Fluidarom 1003® que de Lácteo-Vainilla, pero sin ser estadísticamente significativo. En conclusión, la incorporación de Fluidarom 1003® en la dieta de las cerdas (final de gestación y lactación) y la administración de un pienso “creep-feed” sin aroma es suficiente para crear en los lechones una mayor apetencia al consumo de pienso postdestete con el aroma incluido. También se aprecia que dar el aroma solamente al final de gestación sería suficiente para condicionar a los lechones al aroma. Palabras clave: “creep-feed”, aprendizaje materno, aroma, lechón, preferencia RESUMÉ XI Le sevrage des porcelets à un âge précoce (21-28 jours) est l'étape la plus critique dans la production porcine intensive. Pendant cette période, les porcelets souvent expriment le stress, la néophobie alimentaire et l'anorexie causée par: la séparation de la mère, le changement d'emplacement et de l'alimentation (lait par l’aliment solide). Pour faciliter le sevrage des aliments appétissants sont utilisés avec des produits frais et de haute qualité, y compris les arômes, qui ne fonctionnent pas toujours. On peut également donner du “creep-feed” au cours de l'allaitement, afin que les porcelets s'habituent aux aliments solides. Récemment, on est revenu sur la possibilité d'utiliser des stratégies d'apprentissage précoce basé sur l'incorporation d'arômes dans l'alimentation maternelle. Des touches aromatiques atteignent le liquide amniotique et le flux de sanguin fœtal et / ou placentaire. Cette exposition prénatal précoce du fœtus aux touches aromatiques peut entraîner une préférence pour cette arôme qui reste au fil du temps et peut donc faciliter l'acceptation de la nourriture avec un arôme similaire avant et après le sevrage. Cette étude est basée sur l'apprentissage maternel pour améliorer l’adaptation du porcelet pendant le sevrage. L'arôme étudiée est Fluidarom 1003®, une arôme commerciale basée en >25 % de l'anéthole et de l'aldéhyde cinnamique et plus de 10 % d'eugénol. L'étude est divisée en trois essais, le premier réalisé précédemment dans SNiBA (Servei de Nutició i Benestar Animal) a servi de point de départ pour planifier les hypothèses et les objectifs de notre deuxième et troisième essais. Dans cette mémoire, nous décrivons les 3 essais pour que le lecteur puisse mieux comprendre l’ensemble de résultats. 1-Dans le premier essai L'objectif était de déterminer les conséquences productives de l'ajout de Fluidarom 1003® dans l'alimentation des truies (en fin de gestation et de l'allaitement) et aussi dans la ration qui a été administré aux porcelets durant le sevrage. Les paramètres productifs des truies et les rendements des porcelets ont été enregistrés au cours de la transition (0-35 d). Il n'y avait aucune différence dans les performances de truies associées à l'administration de l'arôme. On a observée tout au long de la transition (0-35 d) une augmentation de la consommation et le poids vif des porcelets exposés pendant la période périnatale à l’arôme. A partir de ceci se pose l'incertitude de savoir si l'arôme dans l'alimentation des truies et l’aliment "creep-feed" a été donné en même temps, il est possible que l'effet est dû en partie au faite que les porcelets consommaient plus de quantité de l’aliment "creep-feed". Cependant, on n’a pas enregistré une consommation de l’aliment "creep-feed" et on ne pouvait pas distinguer entre l'effet de la mère (gestation et allaitement) et l’effet "creep-feeding". RESUMÉ XII 2-Sur la base de cette incertitude on a réalisé le deuxième essai, où cette fois on a essayez d’évaluer les effets de l'inclusion du Fluidarom 1003® mais en tenant compte le facteur "creepfeeding". Également on a enregistré de nouveau les paramètres productives et rendements des porcelets au cours de la transition (0-14 jours), en plus de la consommation de l'aliment "creep-feed". Afin d'évaluer l'effet de l'apprentissage maternel on a effectué un test de préférence double (DCHT) pendant trente minutes aux cours du 2e jour après le sevrage entre Fluidarom 1003® et l’arôme Lait-Vanille (contrôle négatif) basé sur plus de 25 % de la vanilline et plus de 10% de l'acide butyrique et le diacétyle. Les résultats ont montré qu’il n’y a pas de différence dans les performances des truies. Les porcelets ils ont consommé plus "creep-feed" sans arôme, au cours de la transition la plus grande consommation et croissance se sont produites chez les animaux exposés pendant la période périnatale à Fluidarom 1003® mais sans arôme dans "creep-feed". Au cours du ‘DCHT’ les animaux témoins a montré une forte préférence pour le Lait-Vanille. Toutefois, les porcelets exposés pendant la période périnatale n’ont montré aucune préférence pour arôme. Mais les porcelets exposés pendant la période périnatale à l'arôme mais pas dans "creep-feed" ont montré une plus grande préférence pour le Fluidarom 1003®. 3-La troisième essai était destiné à connaitre l'étape la plus importante pour configuré l'apprentissage du porcelet pour la nouvelle arôme. On a réalisé un modèle expérimental qui a permis de discerner le rôle indépendant que exerce la inclusion de Fluidarom 1003® (avec ou sans) dans le régime alimentaire de la mère de la fin de gestation, de lactation et dans "creepfeed" qui a été administré aux porcelets pendant la lactation. Pour cela un ‘DCHT’ a été réalisé dans les mêmes conditions que dans la 2e essai. Les résultats ont montré que les des animaux témoins, préféraient l'arôme Lait-Vanille. Le reste des traitements ne présentent pas de différence significative pour la préférence entre les arômes. Les porcelets pré-natalement exposé à l'arôme mais avec "creep-feed" sans arôme ajoutée ont consommé plus d'aliments avec Fluidarom 1003® que du Lait-Vanille, mais sans être statistiquement significative. En conclusion, l'incorporation de Fluidarom 1003® dans l'alimentation des truies (en fin de la gestation et de la lactation) et de "creep-feed" sans arôme suffit pour les porcelets à créer un appétit d’ingestion d'aliments de post sevrage avec l'arôme inclus. Il a été montré que donner l'arôme seulement à la fin de gestation serait suffisante pour crée une préférence pour l’arôme. Mots clés: “creep-feed”, apprentissage de la mère, arôme, porcelet, préférence 1 1. INTRODUCTION INTRODUCTION 3 THE WEANING Commercial weaning is a sudden change in piglet life during which piglets have to be adapted to eat a novel food (usually solid diet) after the separation from their mother. In natural conditions weaning is a gradual process that piglets perform during a long lactation and may be not completed until 17 weeks of age (Jensen, 1988). However, in the modern pig industry it is commonly used the “early weaning” with the objective of reducing and improving sow productive cycle. It means that weaning often occurs abruptly at 21-28 days of age and becomes a stressful process. Weaning for piglets represents a period of adaptation and stress in response to the simultaneous stressors imposed on pigs at weaning. The sudden separation of piglets from their mothers, mixing with others litters and reallocation to a different environment drive piglets to exhibit aggressive behaviour while leading to the formation of a new social group (Varley and Wiseman, 2001). Moreover, piglets at weaning are switched from highly-digestible milk to a less-digestible more-complex solid feed, composed by cereals (starch source) and vegetal proteins (Pluske et al., 2003) and often supplemented with artificial flavours and savouries. As a consequence piglets usually suffer a post-weaning “growth check” which last 7-14 days (Pluske et al., 2003). Piglets respond with low and variable feed intake or directly by stop eating, poor and variable growth rate and increased susceptibility to enteric pathogens. Anorexia brings about gastrointestinal disturbances; with alterations in small intestine architecture and enzyme activities, transiently-increased mucosal permeability, disturbed absorptive-secretory electrolyte balance and altered local inflammatory cytokine patterns (Lallès et al., 2007). The major diseases are diarrhoea, particularly from enterotoxigenic Escherichia coli (ETEC) and Salmonella (Pluske et al., 2003; Lallès et al., 2007). For this reason main strategies to reduce the negative impact of weaning focus into the objective of reducing stress and anorexia (Pluske et al., 1997; Varley and Wiseman, 2001); such as by weaning pigs at older ages, creating eaters of creep-feeding, avoiding environmental challenges, and utilising good quality, high digestibility and palatable feedstuff. Recent studies support the use of flavours in the sow and piglets diets as a means of promoting maternal learning. LITERATURE REVIEW 11 3.1 PIGLET AT WEANING: the low feed intake Weaning in natural conditions represent a gradual transition from a liquid diet to a solid diet, over a period of several weeks. This gradual transition (exposure to solid feed) leads to piglet to stimulate enzyme system and develop gastrointestinal tract (GIT) (Gestin et al., 1997; Makkink et al., 1993) and also provides time for changes in the ecology of the GIT (microbial population adapts to the new feed inputs) (Buddington, 1998). On the other hand the weaning on commercial pig industry occurs abruptly 21 to 28 days of age, and pass from a liquid diet of 20% DM to a compound diet with 85% DM (Varley and Wiseman, 2001). The nutritional, psychological and environmental changes produces stress response, and anorexia on the first days after weaning (Pluske et al., 2007); even some piglets abstain from eating for over 50 hours (Bruininx et al., 2002). There is positive relationship between feed intake and villous height or villus/crypt ratio (Pluske et al,. 1996), so it seems that low feeding intake immediately after weaning is responsible for some changes in gut morphology. In the review of Dong and Pluske (2007) demonstrated that level of feed intake was the most important determinant of mucosal function and integrity. Therefore, anorexia produces a reduction of gut functions: a reduced enzyme activity and absorption (Kenworthy and Allen, 1966). In addition in a study of Pluske et al. (1996) suggested that the villous atrophy could be avoided if post-weaning feed consumption is maintained. Moreover, weaning piglets show a reduced low acid secretion (lack of lactose substrate) and consumption of large meals at infrequent intervals results in elevated pH (Kidder and Manners, 1978). At this higher pH, the gastric conditions could allow pathogens to survive and is a greater opportunity to colonise the digestive tract (Yen, 2001). Weaning piglets are less mature of digestive system; they have limited capacity of digestion and absorption of feeds, mainly vegetal proteins (Souza et al., 2012). In addition, the high feed intake after the underfeeding period may saturate the absorptive capacity, which causes undigested material beginning undesirable microbial activity (Makkink et al., 1993). Another factor which influences feed consumption at weaning is neophobia. It is known as rejection or aversion to the new ingredients, flavours or scents. It happens when an animal is LITERATURE REVIEW 12 exposed to novel food and is reinforced by the associated novel environment (Hursti and Sjödén, 1997). Neophobia can be adaptive under natural conditions as it prevents toxicosis while animals are learning from the post-ingestive consequences of eating a novel and potentially toxic feed (Rozin and Vollmecke, 1986), so is a precaution mechanism. It is characterized by a period of low feed intake, followed by increased consumption leading to a relatively stable level of intake (Figure 1). If the consumption of the new feed does not result in a gastro-intestinal disease, then the animals increase the new feed consumption. On the other hand, if eating the new feed results in illness, the animal forms a dislike for the feed called a conditioned taste aversion (CTA) (Launchbaugh, 1995). Figure 1. Intake pattern typical of animals expressing neophobia adapted from (Launchbaugh, 1995). However, neophobia can be reduced through a learning process before or after birth involving contact with flavour cues and associations between those components and positive consequences that induces a food recognition and then its higher consume (Mennella et al., 2001; Wells and Hepper, 2006; Oostindjer et al., 2010; Figueroa et al., 2013a). Methods to increase feed consumption at weaning: a. Weaning pigs according to their size and ability to cope (including the possibility of weaning different pigs in a single litter at different times). b. Accustoming pigs to the consumption of viable quantities of solid feed before they are weaned (creep-feeding). These diets should be highly digestible and include ingredients with high palatability (Solà-Oriol et al., 2012). Creep-feed may also help to LITERATURE REVIEW 13 satisfy requirements of piglets to achieve a good growth potential and weaning weight, in addition to prepare the digestive system of the sucking pigs to cereal-based solid diets after weaning (Pluske et al., 2003). This item will be explained more extensively below. c. Avoiding environmental (especially temperature) challenges and competitive stress. d. Good quality and fresh ingredients: weaned pig is very sensitive to dietary mold growth and rancidity, so addition of for example proper amount of mold inhibitors and antioxidants to the manufactured diet is very important to prevent reduction in palatability during storage. e. Optimum balance among nutrients in the diet: provide adequate amounts of vitamins, trace minerals and limiting amino acids in the diet. f. Using weaning diets that are highly digestible: the digestibility of diets has a positive relationship with the feed intake. Whittemore and Kyriazakis (2006) developed an equation to show this relationship: Maximum voluntary feed intake (kg/d) = 0.013BW (1-DM digestibility coefficient) Where BW is Body Weight (kg) and DM is Dry Matter. g. Palatable feedstuff with palatable ingredients: in the revision of Dong and Pluske (2007) showed that lactose-containing products, spray-dried plasma (SDP), spray-dried blood meal (SDBM), and high quality fish meal are palatable ingredients for the newlyweaned pigs. In addition, the inclusion of whey or lactose in the starter diet ensures continuation of bacterial fermentation and some, though reduced, lactic acid production (Kidder and Manners, 1978). However, improving the palatability of feed offered has variable and mostly only minor effects (Appleby et al., 1991; Pajor et al., 2002). h. In relation with the previous point, inclusion of flavours and taste enhancers in the weaning diets may increase fed intake, but not always (Dong and Pluske, 2007). i. Promoting ways of a maternal learning based on learning from flavours added in the maternal diet. It has been suggested that flavour cues from the maternal diet in uterus can reach the foetus through the amniotic fluid and/or the placental blood stream. This early exposure of foetus to certain cues generally may result in a preference for these flavours later in life and consequently can positively affect the acceptance of food with a similar flavour before and after weaning (Hepper, 1988; Mennella et al., 2001). This flavour preference can be more important when the flavour is also present LITERATURE REVIEW 14 in maternal milk (Galef and Henderson, 1972). It has been demonstrated that maternal learning improves feed intake, maladaptive behaviour, gastro-intestinal problems and poor growth (Figure 2) (Oostindjer et al., 2010). This item will also be explained more extensively below, so is the central hypothesis of our study. Figure 2. Schematic overview showing hypothetical links between neophobia, stress and performance at weaning. At weaning piglets suffered dietary and other changes like mother separation, mixing with other piglets in a different environment, thus create to piglets neophobia and stress and in consequence may negatively affect feed intake, gastrointestinal health, growth rate and behaviour. Dotted lines indicate variables which were found to be affected by prentatal or preand postnatal flavour learning. Adapted from (Oostindjer et al., 2010). Among the previous strategies, we will focus those that involve an early learning of the piglet to the new food. Piglets need to establish a connection of continuity between the suckling period and post-weaning, to reduce the low feed intake and thus the “growth-check”. Therefore, piglets should learn what to eat and how to forage before weaning. The following sections of this bibliographic revision will focus on creep-feeding and maternal learning as likely strategies to smooth the weaning. Stress Neophobia Mother separation Mixing with other piglets Different environment Dietary changes Maladaptive behaviour Poor Growth Gastro-intestinal problems Low-feed intake LITERATURE REVIEW 15 3.2 CREEP-FEEDING: effects of pre-weaning experience Creep-feeding is defined as the milk replacer solid diet provided to the suckling litter during lactation in order to establish an early contact with solid feed and to make easier the transition from milk to solid feed at weaning. Creep-feed formulation may be very complex and varies from one feed mill to another; but milk by-products are always included as a source of lactose. Milk replacers commonly contain cereals (oats, corn, barley…), protein sources of animal origin (milk whey and whey powders, egg meal; animal plasma, fishmeal), protein sources of vegetal origin (soybean meal, soybean meal concentrate, wheat gluten, pea gluten, potato protein…) and lactose as main ingredients. For this period, technological treatments of the feed ingredients are also a common practice so that it is very common the use of heat processed cereal and vegetal origin protein sources (cooked, extruded, thick rolled), decorticated, micronized with the aim to obtain highly digestible ingredients and highly digestible diets. In commercial practice, the creep-feed is introduced from Day 7 to 10 of life, because piglet attraction for dry feed during the first week of life is very low (Pluske et al., 2003). The main reasons that are suggested to supply solid food (creep-feed) during lactation are: 1First, creep-feed will familiarize piglets feeding behaviour and intestinal physiology to the next changes caused by weaning (Pluske et al., 2003). Physiologically, creep-feed stimulates the gastrointestinal tract to produce amylases and proteases enzymes (Aumaitre, 1972) for digestion of complex carbohydrates and proteins. Also introduction of creep-feeding can reduce the enteropathologic changes, as Makinde et al. (1997) observed that only piglets without creep-feed had diarrhoea compared to piglets with creep-feed. According to feeding behaviour, early creep-feed consumption may allow the weaned piglets to focus more on feed intake and less on exploratory behaviour (Bruininx et al., 2002; Sulabo et al., 2010d). 2Second, creep-feed will allow higher daily weight gain (270 g/day) (NRC, 2012). A higher piglet weight at the end of lactation is possible because creep-feed may compensate the decreasing milk production throughout lactation (Klindt, 2003; Bruininx et al., 2004). Whittemore and Morgan (1990) describe the lactation curve for a sow and daily milk yield. It can be observed that at day 21 with 12 piglets per litter, the piglets demand >3 kg/d of milk than the sow production (Figure 3). This result confirms that piglets require extra sources of energy and nutrients. It has been also described that creep-feed consumption occurs at a LITERATURE REVIEW 16 higher extent in smaller piglets within a litter (Sulabo et al., 2010b) and if only these piglets have the necessity to consume creep-feed (Solà-Oriol, 2013). Figure 3. Supply of milk from the sow (solid line) and demand for milk by the litter (broken line), n is the litter size, adapted from (Whittemore and Morgan, 1990). 3Finally, and probably as a consequence of the two previous reasons, creep-feed stimulates feed intake after weaning (Bruininx et al., 2002; Sulabo et al., 2010d). In summary, it is accepted that creep-feed helps to smooth the adaptation of piglets after weaning. However, it has been also suggested that the use in creep-feed of vegetal origin feed ingredients with antinutritional factors or antigenic compounds, like glycinin and β-conglycinin from soybean products, could sensitize piglets to antigens. Exposure of sensitized pigs to an increased intake of the same dietary antigens after weaning may rise to a hypersensitivity response (Pluske et al., 2003). Although, Friesen et al. (2010) observed that all pigs (previously fed to soybean and not) developed a tolerance to soy proteins within 2 weeks post-weaning. 3.2.1 Creep-feed intake Creep-feed consumption is highly variable between litters, within litters (Barnett et al., 1989; Pluske et al., 1995; Bruininx et al., 2002; Sulabo et al., 2010d), and among studies (Table 1). Creep-feed contribution to the total daily energy intake before weaning at 21 to 35 d of age ranged from 1.2 to 17.4% (Pluske et al., 1995). LITERATURE REVIEW 17 Table 1. Creep-feed intake per litter, per piglet and the duration of creep-feeding in different studies. Study Creep-feed Intake per litter Creep-feed Intake per piglet duration of Creep-feed (d) Barnett et al., 1989 107-1550g 13-194 18 Pajor et al., 1991 2-205 g/day 27±16.8 g/day 18 Delumeau and Meuniersalaun, 1995 0-2382g 0-674g 14 Bruininx et al., 2002 445-7840g 377±200 g 17 Bruininx et al., 2004 ----- 301±151 g 17 Pluske et al., 2007 380±332.5 g/day 37±32.7 g/day 19 Sulabo et al. 2010b 263-2349 g ------ 18 Sulabo et al. 2010d 350-1034g ------ 13 Factors that influence the large variation in creep-feed intake are difficult to study. However, it should be noted that most studies considered feed disappearance as feed intake and wastage may have a high impact as a source of variation. There are different theories about causes of the variable creep-feed intake. One is that some piglets may eat more creep-feed because of a more mature gastro-intestinal tract, which allows them to utilize nutrients from solid feed at an earlier age (Aumaitre, 1972). Another theory was that smaller piglets eat more creep-feed to compensate inadequate milk consumption (Algers et al., 1990), but Bruininx et al. (2004) did not observed a relationship between creep-feed consumption and body weight. Finally Sulabo et al. (2010d) speculated that creep diet composition, creep access, and feeder design may strongly contribute to this variability. 3.2.2 Creep-feeding behaviour In order to account for this variability, it is important to remark that not all piglets consume creep-feed within a litter; approximately only 60% of pigs are creep-feed eaters (Sulabo et al., 2010d; Barnett et al., 1989). For this reason, some authors (Barnett et al., 1989; Bruininx et al., 2002; Kuller et al., 2007; Sulabo et al., 2010b) have tried to discern which piglet eats creep-feed inside a litter, using a dye. Most of them use 1% of chromic oxide. However, Barnett et al. (1989) and Bruininx et al. (2002) reported that a considerable proportion of piglets (33%) were not be categorized with chromic oxide due to indistinguishable colours of the faeces. Pluske et al. (2007) used indigo carmine (5g/Kg) to classify 100% of piglets, because this dye clearly stains faeces (cyan). Females show higher creep-feed intake than males (Bruininx et al., 2001; Kuller et al., 2007) and are more implicated in the creep-feeding activity than castrated males (Delumeau and Meuniersalaun, 1995). As a consequence females with higher activity in creep-feeding may LITERATURE REVIEW 18 become familiarized with the solid diet and they start early to consume after weaning. Then, gilts refrain less from eating after weaning than boars (Bruininx et al., 2002) and shows a greater ADFI and ADG in the first days of post-weaning period (Delumeau and Meuniersalaun, 1995; Bruininx et al., 2001). However, the same author in a posterior study did not observe a gender effect neither on the latency time or post-weaning ADFI and ADG (Bruininx et al., 2004). The position of piglets to suckle (teat order) and their BW may also determine creep-feed consumption. Piglets that nurse from posterior teats that produce less milk are smaller than those nursing from anterior teats with greater milk production (Table 2) (Gill and Thomson, 1956; Fraser and Jones, 1975; Solà-Oriol, 2013). Table 2. Teat Order of suckling period and its relationship with BW (from Solà-Oriol, 2013). Position held on the udder Front Middle Rear Body Weight (kg) 1.72 1.65 1.48 Small piglets have less ability to compete at the udder and extract milk (Sulabo et al., 2010d) and they suckled teats that were less productive (Solà-Oriol, 2013). This could predispose small piglets to consume more creep-feed (Table 3). In contrast, heavier piglets sucking from anterior teats consume more milk during lactation so grew faster and were heavier at weaning (remaining 1.1kg heavier (Pluske et al., 2007) compared to their counterparts sucking from the posterior teats). Heaviest piglets may not find interesting to consume creep-feed, and as a consequence reduces post-weaning feed intake (Pluske et al., 2007). Nevertheless, Bruininx et al. (2004) failed to find a relationship between body weight and the pre-weaning consumption of creep-feed; and Pajor et al. (1991), Delumeau and Meuniersalaun (1995) and Pluske et al. (2007) did not observed any relationship between the teat order and the creep-feeding activity. Table 3. Percentage of piglets that were consumers of creep-feed (offered from the 10th day of age) at 18, 21 and 25 days of age according to the position held on the udder during the lactation (Solà-Oriol, 2013). Position held on the udder Days of age Front Middle Rear 18 7 % 5 % 12 % 21 17 % 13 % 21 % 25 41 % 35 % 42 % LITERATURE REVIEW 19 3.2.3 When is creep-feed consumed? Previously it was believed that creep-feed consumption increases linearly throughout lactation (Barnett et al., 1989) but more recently it has been described that a 60 to 80% of total creepfeed intake is consumed in the last week before weaning, either if pigs are weaned at 3 week (Sulabo et al., 2010d) or 4 weeks of age (Pajor et al., 1991; Fraser et al., 1994; Bruininx et al., 2002; Pluske et al., 2007). Thus, Sulabo et al. (2010b) scored that daily creep-feed intake of litters increased quadratically (R2 =0.22) from day 3 to weaning: Litter creep-feed intake, g/d= (0.8965xAge,d2)-(10.607xAge,d)+37.683 It is clear that piglets consume more creep-feed as weaning date approaches, but why at the end of lactation? Sulabo et al. (2010c) introduced creep-feed 3 days before weaning and observed that this short period was enough to create 70% of eaters. Hence it could be concluded that individual creep-feed consumption is more related to the age of piglets rather than how many days creep-feed is offered to the litter (Kuller et al., 2007; Sulabo et al., 2010c). Moreover, piglets at the end of lactation have more digestive maturity and thus increase the opportunity to consume more creep-feed (Barnett et al., 1989; Pajor et al., 1991). However, in another study the same authors observed that as longer is the duration of creepfeed offer (13 d vs. 2 d) higher is the proportion of piglets that consume creep-feed but, without any effect on the pre-weaning performance (Sulabo et al., 2010d). Therefore, there are still contradictions about if it is better to introduce creep-feed early at the lactation or just few days before weaning. Total creep-feed consumption increases according to litter size (6 to 13 pigs per litter) so larger litters have higher feed consumption despite of less individual creep-feed intake and lower daily gains during lactation (Barnett et al., 1989). In contrast, Klindt (2003) introduced creepfeed in litters with more than 8 pigs and observed an increased lactation ADG and weaning weight. It could be suggested that creep-feed could be beneficial for the piglets especially for those sows with a large litter. Pajor et al. (2002) observed that creep-feeding reduced lactation BW loss and weaning-to estrus interval; but Sulabo et al. (2010b) did not observed any effect, and estimated that creep-feeding provided energy equivalent to only 1.27 kg of sow BW loss. The difference in these results may be due to the different length in the lactation period (28d LITERATURE REVIEW 26 Piglets can also learn by observing their mother. This is called vertical social learning (Oostindjer et al., 2011a). Learning from mother will be the best model for social information transfer and also could be more powerful than learning from other experienced animals (Thorhallsdottir et al., 1990). Furthermore, mother and its offspring are genetically related, so they will respond similarly to the food (Laland et al., 1993). In pigs there are few studies about vertical transfer in pigs. In (Nicol and Pope, 1994) concluded that it would be better to learn from the mother than from piglets of the same age. However, in modern pig husbandry is difficult that piglets could observe and learn the feeding behaviour of their mother, as sows are confined in crates during lactation being the troughs inaccessible for them. 3.3.3 Maternal learning It is known that pigs have a well-developed sense of smell (Morrow-Tesch and McGlone, 1990) and piglets can discriminate among auditory, olfactory, visual and tactile stimuli immediately after birth (Parfet and Gonyou, 1991). Furthermore it has been observed that pigs can learn olfactory cues for faster discrimination than visual discrimination tasks (Croney et al., 2003). It has been hypothesized that piglets have a well-developed olfactory system at birth so neonatal piglets were able to recognize their mother’s faecal and skin odours (Morrow-Tesch and McGlone, 1990). Figueroa et al. (2013a) realized a triple-choice stimulus among maternal amniotic fluid, alien amniotic fluid or water and observed that piglets preferred their own maternal amniotic fluid; but this preference disappeared with age (Figure 6). This also happens in other species (mammals, birds and amphibians) (Bolhuis et al., 2009). For example, Hepper (1987) demonstrated that rat pups, either natural born or caesarean, recognized the flavour previously offered to the mother, indicating that the preference is acquired prenatally. So when a mammal is born they have a rudimentary idea of who are their parents/family. This leads to animals a smoothly transition from the prenatal environment to postnatal life, a maternal recognition, nipple localisation and initiation of suckling (in mammals) (Hepper, 1987). LITERATURE REVIEW 27 Figure 6. Piglet’s permanence time in contact with strips impregnated with maternal amniotic fluid, alien amniotic fluid and water for a 7-min preference test during lactation. Means with different subscripts are different (P<0.05) (Figueroa et al., 2013a). If mothers know what kind of food is good to eat and what kind of food is present to their environment, maternal learning would allow offspring to prefer to eat the same. Then, maternal learning starts before farrowing, when the sow eats some compounds of feed that pass to foetus. It has been described the volatile compounds may cross the placental barrier and enter the foetal blood stream, consequently diffuse out of the nasal blood capillaries and come into contact with foetal olfactory receptors. In addition, flavours come into the amniotic fluid; the foetus inhales or swallows it and then stimulates olfactory receptors or taste buds (Hepper, 1988). As a result this exposure to cues lead to create preference for this flavours later in life and also can positively modulate the acceptance of food with a similar flavour before and after weaning (Mennella et al., 2001). After birth, maternal learning continues through milk, the hedonism and lactic post-ingestive effect with the pleasure of nursering may create an associative learning with cues in milk (Hepper and Wells, 2006). Thus, this flavour preference can be strengthened more when the flavour is also present in the maternal milk (Mennella et al., 2001; Oostindjer et al., 2009). However only milk exposure does not increase flavour preference in all species (Hepper and Wells, 2006). Although in rabbit pups the ingestion of mother’s milk (Bilkó et al., 1994) was sufficient to influence the dietary preference; in piglets postnatal exposure alone did not reduce weaning-associated problems (Oostindjer et al., 2010). LITERATURE REVIEW 28 What kinds of flavours can cross the placental barrier? Table 5, shows some of the flavours that have been described to create postnatal effects. Most of them increases post-natal olfactory preference for these flavours or reduces aversion. Some of them only focussed on the short-term recall by neonates and the others observed that flavour preferences may be long-lasting and persist until weaning and beyond. For example (Bilkó et al., 1994) observed that the ingestion of juniper berries on the day of weaning was five times higher in rabbit pups from does fed juniper berries during gestation compared to control ones. Another example is in humans, where (Mennella et al., 2001) also observed an increased acceptance, enjoyment and intake of a carrot-flavoured cereal meal in infants of 6 month-old from mothers who had a regular consumption of carrot juice during pregnant and lactation (Figure 7). Figure 7. The infants’ relative acceptance of the carrot-flavoured cereal as indicated by display of negative facial expressions (A), mothers’ rating of their infants’ enjoyments (B) and intake (C). Proportional responses were calculated by dividing each infant’s response to the carrot-flavoured cereal by his or her response to the carrot cereal plus plain cereal (carrot/[carrot+plain]). CW group indicates that mothers drank carrot juice during pregnancy and water during lactation; WC did the opposite and finally WW drank water during both exposure periods (Mennella et al., 2001). LITERATURE REVIEW 29 Table 5. Overview of experiments demonstrating that prenatal exposure to flavours from the maternal diets affects postnatal preferences. Species, type of flavour, postnatal effects and references are indicated, adapted from (Bolhuis et al., 2009). Species Flavour1 Postnatal effects5 References Rats Garlic ↑ olfactory preference d12 Hepper, 1988 Rabbits Juniper2 ↑ intake juniper at weaning d28 Bilkó et al., 1994 Cumin ↑ olfactory preference d0 Coureaud et al., 2002 Sheeps Citral2 ↑ olfactory preference d0 Schaal et al., 1995 Oregano3 ↑ intake oregano-flavoured feed at 3,4.5,6,7.5 mo Simitzis et al., 2008 Dogs Anise (transanethole) ↑olfactory preference d0 Wells and Hepper, 2006 ↑ intake anise-flavoured treats w10, only when exposed through milk also Hepper and Wells, 2006 Pig Anise2 Milkycheese2 ↑ intake the same flavour on diet, ↑ body weight first days post-weaning and ↓ diarrhoea Positive effects on behaviour ↑olfactory preference d14, 21 and 26 ↑olfactory preference d14, 21 and 26 Oostindjer et al., 2009 Oostindjer et al., 2009 Figueroa et al., 2013a Figueroa et al., 2013a Humans Anise2 ↑ olfactory preference d0 and d4 Schaal et al., 2000 Garlic ↑ olfactory preference d0 Hepper, 1995 Carrot4 ↓ negative facial responses while eating carrotflavoured cereal at 6 month Mennella et al., 2001 ↑ enjoyment perceived by mother while eating carrot-flavoured cereal Mennella et al., 2001 1Provided in the maternal diet throughout gestation unless indicated otherwise. 2Last two gestational weeks. 3Day 50-130 of gestation. 4Last three gestational weeks. 5Relative to (non-exposed) control groups. d=day; w=weeks; mo=months. In pigs the first description of maternal learning was from (Campbell, 1976), who demonstrated that piglets are able to choose diets related on their mothers consumption during lactation. They observed a 12% increase on feed intake of weaning diets with the same flavour that was added in sow’s diet and gained 24% more weight. However, it has been long later when new studies on maternal learning in piglets have been published. Oostindjer et al. (2009, 2010) observed positive effects of perinatal flavour learning on feed intake, growth and behaviour. They saw that during first days of weaning the piglets previously exposed to flavour increased feed intake of the same flavour in the diet and growth (Figure 8). Piglets showed also fewer vocalizations, escape, fighting and aggression, but no increased preference for the flavour. In our department Figueroa et al. (2013a) found that LITERATURE REVIEW 30 prenatal flavour exposure via maternal diet influenced a piglet’s preferences of the pure flavours during lactation and post-weaning (Day 26). Figure 8. Body weight during the first 14 days post-weaning for animals that had been exposed to anise both preand post-natally (FF), prenatally (FC), postnatally (CF), or never (CC) *P<0.05, # P<0.1, ns not significant (Oostindjer et al., 2010). In another study, Oostindjer et al. (2011b) did not observe that pre-weaning (anetholflavoured diet during gestation and lactation) or post-weaning (anethol either in food or in air) treatment affected food intake, preference or growth. But flavour piglets played more and showed less damaging behaviours than control piglets, indicating a reduced stress around weaning when the familiar flavour is present. So it can be concluded that early experience with flavours (prenatal learning), plays a role in postnatal feeding behaviour and development of dietary habits, increasing later acceptance and improving adaptation to weaning. Focusing to post-natal learning, volatiles can be also transferred to the mammary glands via vascular routes from the digestive or respiratory system (Dougherty et al., 1962). It has been observed that pre-weaning exposure to feed flavour (during lactation) from the maternal diet can reduce food neophobia and promote the intake of an otherwise unfamiliar feed after weaning (Bolhuis et al., 2009). For example in Table 6, the studies used twoor three-way food LITERATURE REVIEW 31 choice test to assess the influence of the maternal diet on feed preferences of the offspring in a wide range of species. Table 6. Overview of experiments demonstrating that exposure to flavours from the diet of the lactating mother affects post-weaning preferences. Species, type of flavour, post-weaning and references are indicated. Adapted from Bolhuis et al., (2009). Species Flavour1 Testing Age8 Post-weaning effects9 References Rats Lab chow2 d21 (w) ↑ intake lab chow maternal diet Galef and Sherry, 1973 d 19-23 (w) ↑ intake lab chow maternal diet Galef and Henderson, 1972 Lab chow3 d23 (w) ↑ intake in 30-min test if weanling diet matched with maternal diet Bornstein et al., 1975 Garlic4 d29-40 ↑ intake of garlic-flavoured water Capretta and Rawls, 1974 Onion d 21, d23 (w) d25 ↑ intake of onion-flavoured diet Wuensch, 1978 Mice Fennel d22 (w) ↑ intake of fennel Mainardi et al., 1989 Rabbits Juniper d28 (w) ↑ intake of juniper Bilkó et al., 1994 Goats Rice straw5 6 mo ↓ latency to first ingestion Van Tien, 2002 ↑ intake of rice straw Dogs Anise6 10 w ↑ intake anise-flavoured treats, only when exposed prenatally also Hepper and Wells, 2006 Pigs Firanor nº3 3 w ↑ intake firanor nº3 and ↑growth rate Campbell, 1976 Humans Carrot7 6 mo (w) ↓ negative facial responses while eating carrot-flavoured cereal Mennella et al., 2001 ↓ intake of carrot-flavoured cereal shortly after exposure to carrotflavoured mother's milk Mennella and Beauchamp, 1999 1Provided in the maternal diet throughout the lactation period unless indicated otherwise. 2Day 5-weaning. 3Day 2-weaning. 4Three days before parturition-weaning at day 22. 5Last three weeks before weaning at 3 months. 6First 20 days of lactation 7First two months of lactation, for four days per week. 8(w) indicates that animals were tested at weaning or were not fully weaned at the time of testing. d=days; w=weeks; mo=months. 9Relative to (non-exposed) control groups. In pig, Campbell (1976) demonstrated that flavour can be transferred by mother’s milk and affect flavour preferences after weaning, but Oostindjer et al. (2010) was not able to demonstrate it. Moreover, we have to keep in mind that during lactation, piglets have also contact with the smell of the mother’s body, breath, feed particles clinging to her skin, fur or teeth or faecal LITERATURE REVIEW 32 droppings (Bolhuis et al., 2009). Thus, there are reinforcing conditions like nursing which may further strengthen their preference and acceptance of these flavours. As a result, the postnatal learning can also facilitate the acceptance of a similarly flavoured feed at weaning and thus increases the consumption after weaning. In other words, breast milk may bridge the experiences of flavours in utero to those in solid foods. As seen in this review many studies have studied maternal learning in different species. However, there is little practical knowledge in the pig industry about the mechanisms underlying flavour learning and how perinatal flavour learning can increase acceptance of diets after weaning. 33 4. HYPOTESIS AND OBJECTIVES HYPOTHESIS AND OBJECTIVES 35 The hypothesis of this study is that maternal learning, adding Fluidarom 1003® on sow diets, may create a flavour habit on the piglets, either in utero or via the milk suckling. Fluidarom 1003® could represent a familiar volatile cue for piglets that may facilitate the initiation to feed intake and improve the post-weaning performance when the same flavour is reintroduced in weanling diets. The reason could be that: 1.- the flavour in creep-feed will stimulate the consumption, creating more creep-feed eaters and/or 2.- reintroducing the same flavour in weaning diets will reduce feed neophobia and therefore piglets will consume earlier the solid diet. To confirm or deny these hypotheses, we performed the following trials with the following corresponding objectives: 1) First trial: Productive and Performance Outcome The objective was to determine the sow’s and piglet’s productive and performance parameters associated to the incorporation of Fluidarom 1003® in the sow (late gestation and lactation) and weaning diets. 2) Second trial: Study of the Preference for Fluidarom 1003® in Suckling and Weaned piglets. In the second trial according to the literature it was hypothesised that the positive results previously observed in Trial 1, highest ADFI and BW for piglets perinatally exposed to flavour, could be mainly explained by an early creep-feed contact and intake following the familiar link with volatile compounds exposed via amniotic fluid and/or milk. Thus, the main objective was to determine the preference for Fluidarom 1003® throughout a test of preference during lactation and at weaning. The trial also aimed to find out at what extent Fluidarom 1003® affect sow’s and piglet’s productive and performance parameters, including the determination of creep-feed consumption. 3) Third trial: is the Fluidarom 1003® Habit mainly formed in Uterus or via the Milk Suckling? The results of the first and second trial indicated that diet supplementation with Fluidarom 1003® throughout gestation and lactation has a positive impact on growth performance when the same flavour is included in weanling diets. However, we do not know if each one, pre-natal (foetal environment) and post-natal (maternal milk) contact by their own are able to contribute to reinforce the positive reward and preference in piglets after weaning. MATERIAL AND METHODS 42 Pre-starter diets were offered ad libitum for fourteen consecutive days and starter diets were offered from Day 14 to 35 post-weaning. All diets were presented in mash form. Zinc oxide (at 2500 ppm (2000 ppm of Zinc)), Colistin sulphate (120 ppm) and Amoxicillin (300 ppm) were used as common antimicrobials for weaning diets. The composition and the estimated nutrient content are presented in Tables 4 and 5. Body weight and Feed disappearance was weekly monitored on Days 0, 7, 14, 21, 28 and 35 post-weaning. Piglets were weighted by group. Average daily feed intake (ADFI), average daily gain (ADG) and feed conversion ratio (FCR) was calculated by pen. Table 4. Composition (%) of the experimental diets during the pre-starter and starter phase (0 to 14 and 14 to 35 days post-weaning). Ingredients (*) Pre-starter Starter Maize 28 35 Barley 15 17.07 Wheat 15 18 Sweet Milk Whey 10 - Extruded Soybeans 9.44 11.4 Soybean meal 44%CP 5 6 Plasma animal 80% PB 5 - Whey powder (50% Fat) 5 2.5 Fishmeal LT 2.5 5 Soybean oil 1.85 - Calcium carbonate 1.13 0.25 Monocalcium Phosphate 0.84 2.84 Salt 0.08 0.45 L-Lisina HCL 0.47 0.63 Vit-Min Premix 0.4a 0.4a (*) Antibiotics in feed: Amoxicilin: 300 ppm; Colistin sulphate: 120ppm; zinc oxide: 2500ppm. b Pre-starter and starter diet Premix Supplied (g/kg): 7000 IU of vitamin A (acetate), 500 IU of vitamin D3 (cholecalciferol), 250IU of vitamin D (25-hydroxicholecalciferol), 45 mg of vitamin E, 1 mg of vitamin K3, 1.5 mg of vitamin B1, 3.5 mg of vitamin B2, 1.75 mg of vitamin B6, 0.03 mg of vitamin B12, 8.5 mg of D-pantothenic acid, 22.5 mg of niacin, 0.1 mg of biotin, 0.75 mg of folacin, 20 mg of Fe (chelate of amino acids, 2.5 mg of Cu (sulphate), 7.5 mg of Cu (chelate glycine), 0.05 mg of Co (sulphate), 40 mg of Zn (chelate of amino acids), 12.5 mg of Mn (oxide), 7.5 mg of Mn (chelate of glycine), 0.35 mg of I, 0.5 mg of Se (organic) and 0.1 mg of Se (sodium). MATERIAL AND METHODS 43 Table 5. Estimated nutrient content of the basal experimental diets during the pre-starter phase (0 to 14 days post-weaning (%). Nutrients Pre-starter Starter Moisture 10.2 11.2 Net Energy (kcal/kg) 2584.2 2447.4 Ash 5.46 6.22 Etheric Extract 8.18 5.95 Linoleic Acid 1.93 2.20 Crude Fibre 2.62 3.15 Neutro Detergent Fibre 8.14 9.74 Starch 34.24 41.50 Sugars 10.79 3.12 Crude Protein 19.00 18.00 Lysine 1.496 1.417 Dig. Lysine 1.355 1.292 Methionine 0.446 0.504 Dig. Methionine 0.412 0.468 Met+Cys 0.823 0.779 Dig Met+Cys 0.741 0.703 Threonine 0.897 0.850 Dig. Threonine 0.722 0.648 Tryptophan 0.269 0.255 Dig. Tryptophan 0.231 0.221 Valine 0.943 0.810 Dig. Valine 0.813 0.690 Isoleucine 0.741 0.693 Dig. Isoleucine 0.644 0.601 Calcium 0.830 0.830 Total Phosphorous 0.622 1.008 Dig. Phosphorous 0.400 0.660 Magnesium (0.12 and 0.13 %), Sodium (0.31 and 0.24 %), Calcium (0.49 and 0.54 %), Potassium (0.72 and 0.63 %), Copper (6.16 and 6.36 %), Iron (91.50 and 176.61 %), Vitamin E (15.7 and 19.3 %), Biotin (0.2 and 0.1 %), Coline (1153.6 and 1233.3 %), Sulphur (0.2 and 0.2 %), Zinc (28.5 and 34.5 %) and Manganese (10.9 and 12.6 %) for prestarter and starter diets, respectively. 5.2 STUDY OF THE PREFERENCE FOR FLUIDAROM 1003® IN SUCKLING AND WEANED PIGLETS (Second Trial) 5.2.1 Pre-weaning performance and creep-feed intake A total of 32 (Landrace x Large White) sows were used. An overview of all experimental treatments and procedures is given in Table 6 and Figure 1. Four groups of sows were arranged according to parity, expected farrowing date and body condition scores achieved 28 days before farrowing and allotted to four experimental treatments (Table 6). Control treatment (T4) was a commercial gestation, lactation, creep-feed and pre-starter diets, T1 consisted on MATERIAL AND METHODS 44 the same diets than T4 but containing Fluidarom 1003® (375g/Tm), T2 the same that T1 but with a non-flavoured pre-starter diet and finally T3 was the same that T1 but with a nonflavoured creep-feed. Sows were exposed to a flavoured feed from Day 87 to farrowing and during 28 days of lactation. Composition and the estimated nutrient content were the same as trial 1 and are presented in Tables 1, 2, 4 and 5. Feeding programme was the same than for the first trial. Table 6. Experimental treatments of Trial 2. +: inclusion of Fluidarom 1003® (375g/Tm) -: No inclusion of Fluidarom 1003® Figure 1. Overview of all controls made in second trial. Sow’s body weight was measured at Days 110 of gestation and at weaning; body condition scores was measured at Days 87 and 110 of gestation and at weaning; and feed intake was registered along the entire experimental period. After birth, the number of piglets (born alive and stillbirth), weight of piglets (born alive and stillbirth) and number of weaned piglets were registered. Litters were standardized by cross-fostering after birth. Piglets were weighed after cross-fostering and at weaning following the same procedures as those described in the first trial. On Day 10 after birth a preference test was performed following a Tripe-U-Testing Arena (TUTA) previously used by (Figueroa et al., 2013a). The preference test was conducted among three different diets: a control diet (non-flavoured) and two flavoured diets in order to study a possible effect of flavour pre-exposure on the maternal diet on piglet preference during lactation. Flavoured diets were obtained by supplementing the control diet either with Fluidarom 1003® (375 g/Tm) or Lacto-Vanilla (500 g/Tm; a commercial flavoured feed additive based on >25 % of vanillin, >10 % of butyric acid and >10 % of diacetyl; Norel S.A., Spain). The cumulated time in contact with each diet by pair of littermates was measured for 5 minutes. Experimental treatment Gestation Lactation Creep-Feed Pre-starter T1 + + + + T2 + + + - T3 + + - + T4 (control) - - - - End Birth Day 0 10 TUTA Creep-feed + CR2O3 12 17 23 28 CF intake + eaters Weaning 2 7 14 BW + Feed Intake DCHT Start MATERIAL AND METHODS 45 From Day 12 of age until weaning (Day 28), all litters were offered free access to creep-feed (Tables 1 and 2). Creep-feed was presented in mash form containing 1 % of chromium III oxide (Cr2O3) to determinate the number of piglets consuming creep-feed (eaters). Creep-feed was provided in commercial pan-feeders with hopper in order to ensure ad libitum access to feed. Creep-feed consumption was estimated by weighing the disappearance of creep-feed from the pan-feeder in three different periods (Days 17, 23 and 28). Marker appearance in faeces (Cr2O3) was individually monitored on days 17, 23 and 28 by sampling faeces with swaps, a green colour of the faeces was registered as a creep-feed eater according to the method previously described by (Bruininx et al., 2002). 5.2.2 Post-weaning performance and feed preference A total of 24 complete litters (252 male and female, 10 SD=2 piglets/litter) [Pietrain x (Landrace x Large White)] were weaned at an average of 26 SD=2.5 days old. Piglets were weaned with an average BW of 7.3 SD=1.81 kg (selected by age, number of piglets/litter and litter weight at weaning) and moved to the weaning unit according to the previous exposure to Fluidarom 1003®. Entire litters were weaned and moved at the same pen (10 SD=2 pigs/pen) during 14 days period. Initial body weight was not balanced at weaning in order to keep the body weight differences achieved during the early pre-exposure of piglets to Fluidarom 1003®. The animals had ad-libitum access to feed and drinking water, except 1h before the preference test when piglets had no access to the commercial feeders during the test. Pre-starter diets were offered in mash form for fourteen consecutive days post-weaning. Feed disappearance was recorded on Days 2, 7 and 14 post-weaning. Average daily gain (ADG) was calculated from 0-2 days, 0-7 days and 0-14 days. Moreover, on Day 2 after weaning a preference test was conducted following a Double choice test (DCHT) protocol between two diets supplemented either with Fluidarom 1003® (375g/Tm) or Lacto-Vanilla (500g/Tm) with 4 animals of each pen during 30 minutes. A total of 24 piglets per treatment were used (n=6 replicates per treatment). 5.3 IS FLUIDAROM 1003® HABIT MAINLY FORMED IN UTERUS OR VIA THE MILK SUCKLING? (Third Trial) A total of 24 (Landrace x Large White) sows were used to study which is the main pre-exposure way before weaning for what piglets are familiarized with a flavour supplemented in sow diet MATERIAL AND METHODS 46 and of its reinforcement via creep-feed. An overview of all experimental treatments and procedures is presented in Table 7 and Figure 2. Sows were selected and distributed at Day 86 of gestation into six experimental groups following a 3 x 2 factorial arrangement, taking into account the inclusion of Fluidarom 1003® in sow diets (without (Control), with during late gestation (Gestation) and with during lactation (Lactation)) and Fluidarom 1003® inclusion in the creep-feed (with or without) as main factors. Sow parity number and expected farrowing date were taken into account for the initial distribution. Dietary composition and the estimated nutrient content were the same as trial 1, and are presented in Tables 1, 2, 4 and 5. Diets were offered following the procedures described in the first trial. Table 7. Experimental treatments of trial 3. Treatment Gestation Lactation Creep-feed Post-weaning Control Ø - - - - - Ø + - - + - Gestation G - + - - - G + + - + - Lactation L - - + - - L + - + + - -: Fluidarom 1003® not added; +: Fluidarom 1003® added. Figure 2. Overview of all controls made in third trial. Again, litter size was standardized immediately after birth by cross-fostering within the same experimental treatment. All piglets were weighted at Day 7 and at weaning (Day 28) following the same procedures described for the first trial. From Day 7 of age until weaning (Day 28) all litters were offered free access to a creep-feed diet. Half of the litters received a flavoured creep-feed (Ø +, G + and L +) and the other half received a non-flavoured creep-feed (Ø -, G - and L -). 5.3.1 Post-weaning feed preference A total of 96 piglets [Pietrain x (Landrace x Large White)] were weaned at an average of 28 SD=1.5 days old and with an average BW of 8.35 SD=1.69 kg. Piglets were moved to the End Creepfeed Weaning 2 DCHT Birth Day 0 7 Start 28 MATERIAL AND METHODS 47 weanling unit taking into account the previous exposure to Fluidarom 1003® and distributed into 12 pens (2 pens per treatment, 8 pigs per pen). The weaning unit and pens have the same conditions as first trial. The animals had ad-libitum access to feed and drinking water. After weaning, the animals received a pre-starter specification and feed was offered in mash form. Two days after weaning, piglets were individually weighed and a preference test was conducted with 2 animals (Double choice test (DCHT); n=8 per treatment) after 1h of fasting. Two diets differing in the flavour supplementation, either with or without Fluidarom 1003® (375g/Tm) or Lacto-Vanilla (500g/Tm), were provided at the same time and their intake evaluated after 30 minutes. 5.4 STATISTICAL ANALYSIS 5.4.1 Performance parameters Parameters for sow and litter performance were analysed with ANOVA by using the GLM procedure of the statistical package SAS® (version 9.2, SAS Institute; Cary, USA). The statistical unit was the sow for all individual measurements and litter for piglet performance. The following mathematical model was used: Yij = µ + αi + βj + αβij + εij, Where Yij were the general observations by sow category (i) and experimental treatment (j); μ was the general mean of all observations; αi was the effect of sow category; βj was the effect of experimental treatment; αβij was the interaction between treatment and effect of farrowing cycle and ε ~ N (0,σ2ε) was the unexplained random error. Sow was also included as a random effect. In order to get a balanced distribution amongst treatments for the data analysis, the farrowing number of sows was classified as sow category: 0= Zero parities (nuliparous) and 1; 1= Sows of 2, 3 and 4 farrows; and 2= parity 5 y 6. In first trial average daily feed intake (ADFI), average daily gain (ADG) and feed conversion ratio (FCR) was analysed with ANOVA following a 2 x 2 factorial arrangement where the two main factors were Fluidarom 1003® supplementation in sows (gestation, lactation and creep-feed) diet and after weaning (preand starter diets) by using the GLM procedure of the statistical package SAS®. The statistical unit was the pen of pigs for all production measurements. MATERIAL AND METHODS 48 In second trial, piglet performance parameters before and after weaning were analyzed with a generalized linear model by using the MIXED procedure of the statistical package SAS® taking into account the day as a repeated measure, the creep-feed intake at each period and the experimental treatment (T1 to T4) as main factors, and their interactions. The statistical unit was the individual piglet for all production measurements. The following complete mathematical model was: Yijkl = μ + αi + βj + δk +γl + αβij + αδik + αγil +βδjk +βγjl + δγkl + αβδijk + αδγijl + βδγjkl+ αβyijl + αβδγijkl + εijkl, From our perspective some interactions had no sense and/or we were not interested in. Then, we removed them from the model. So the final mathematical model used was: Yijkl = μ + αi + βj + δk +γl + αδik + αγil +δγkl +αδγikl + εijkl, Where Yijkl were the parameters for all observations at each day (i), according to creep-feed intake in the first period (Day 12 to 17 of age) (j), creep-feed intake during the second period (Day 17 to 22 of age) (k), and for the different experimental treatments (l); μ was the general mean of all observations; αi was the effect of the day; βj was the effect of creep-feed intake during Day 12 to 16 of age; δk was the effect of creep-feed intake during Day 17 to 22 of age; γl was the effect of the experimental treatment; αδik was the interaction between day and the creep-feed intake during Day 17 to 22 of age; αγjl was the interaction between day and the experimental treatment; δγkl was the interaction between creep-feed intake during Day 17 to 22 of age and the experimental treatment; αδγikl was the interaction between day, creep-feed intake during Day 17 to 22 of age and the experimental treatment and ε ~ N (0,σ2ε) was the unexplained random error. All of the results are presented as Least Square Means taking into account a Tukey adjust and the alpha level used for the determination of significance for all of the analysis was 0.05. 5.4.2 Preference test Triple-U-Test-Arena was analysed with General Linear Models using the GENMOD procedure of SAS®. The mathematical model used was: Yij = µ + αi + j + α ij + εij, MATERIAL AND METHODS 49 Where Yij was the time spend for the observations of each experimental treatment; µ was the general mean of all observations; αi was the effect of treatment (feed with Fluidarom 1003®, Lacto-Vanilla and without flavour); j was the prenatal exposure [piglets coming from sows that ate (pre and post-exposed) or not (control) flavours during late gestation and lactation]; αij was the interaction between experimental treatments and the prenatal exposure; and ε ~ N (0,σ2ε) was the unexplained random error. Double Choice Test was analysed with ANOVA using the MIXED procedure of the statistical package SAS®. The statistical unit was the pen of two or four piglets. The mathematical model used was: Yij = µ + αi + εijk, Where Yij is the consumption estimated of each flavour; µ is the general mean of all observations; αi is the effect of treatment; and ε ~ N (0,σ2ε) was the unexplained random error. All of the results are presented as Least Square Means taking into account a Tukey adjust and the alpha level used for the determination of significance for all of the analysis was 0.05. Additionally, the preference values for the Fluidarom 1003® were compared to the neutral value of 50 % by using a Student’s t-test procedure of the statistical package SAS®. 51 6. RESULTS RESULTS 58 Table 5. Piglet body weight (BW, g) at weaning and on days 7 and 14 post-weaning. Treatment Body Weight (g) G L CF W Day 0 Day 7 Day 14 T-1 + + + + 7507 8480 10080 T-2 + + + - 7630 8410 9970 T-3 + + - + 7892 9308 11099 T-4 Control - - - - 7859 8786 10184 Mean 7722 8746 10333 Pooled Standard Error 531.66 531.69 531.72 Day effect (Pr>F) <0.0001 Creep-feed intake (12-16d) effect (Pr>F) 0.675 Creep-feed intake (17-22d) effect (Pr>F) 0.014 Treat effect (Pr>F) 0.673 Treat x Day effect (Pr>F) 0.0006 Treat x Creep-feed intake (17-22d) effect (Pr>F) 0.585 Day x Creep-feed intake (17-22d) effect (Pr>F) 0.241 Treat x Day x Creep-feed intake (17-22d) effect (Pr>F) 0.048 G: gestation; L: lactation; CF: creep-feed; W: weaning; +: Fluidarom 1003® added; -: Fluidarom 1003® not added Creep-feed intake (12-16d) effect: creep-feed intake during the period of 12 to 16 days of age. Creep-feed intake (17-22d) effect: creep-feed intake during the period of 17 to 22 days of age 6.2.3 Feed preference between Fluidarom 1003® and a lacto-vanilla flavour 2 days after weaning The preference for Fluidarom 1003® (375g/Tm of feed) over a lacto-vanilla (500g/Tm of feed) was directly compared. Results are presented as feed intake corrected by the number of piglets and BW. The percentage of preference is reported in Figure 1. Piglets from the control group (T4) showed a higher preference for lacto-vanilla (P<0.05) and no preferences for Fluidarom 1003® or lacto-vanilla were observed for piglets corresponding to treatments T1 and T2 (P>0.1). However, a strong preference (P<0.05) for Fluidarom 1003® was observed in T3 piglets (the ones that received an early contact with Fluidarom 1003® via the sow diet, but were offered an unflavoured creep-feed). RESULTS 59 +: Fluidarom 1003® added; -: Fluidarom 1003® not added Figure 1. Feed consumption pondered by number and body weight of piglets and feed preference about Fluidarom 1003® as total consume percentage (number on top of the bars), statistical comparison against the preference neutral value (50%; indifference). NS: P>0.1; *: P<0.05 (Trial 2). 6.3 IS FLUIDAROM 1003® HABIT MAINLY FORMED IN UTERUS OR VIA THE MILK SUCKLING? (Third Trial) 6.3.1 Feed preference after weaning (Day 2 post-weaning) measured by using a Double Choice Test (DCHT) The preference for Fluidarom 1003® over a lacto-vanilla (negative control) was directly compared. Results are presented as feed intake pondered by the number and BW of piglets in each pen. Additionally, the percentage of preference for Fluidarom 1003® compared to the negative control is also reported in Figure 2. Piglets from the sows of the control group that were offered a creep-feed without Fluidarom 1003® inclusion (Ø-) showed higher preference for lacto-vanilla (P<0.05). In addition piglets previously exposed to Fluidarom 1003® during the pre-natal period (G+; Fluidarom 1003® supplemented on gestation diet and creep-feed) also tended to prefer Lacto-Vanilla (P<0.1). No preferences for Fluidarom 1003® were observed two days after weaning for piglets corresponding to treatments Ø+, G-, Land L+. However piglets pre-exposed during the prenatal period (G-; dietary Fluidarom 1003® supplementation during late gestation but not on creep-feed) were those showing a quantitative preference (56.1%) for the Fluidarom 1003® RESULTS 60 supplemented diet as compared to Lacto-vanilla (negative control) two days post-weaning (P=0.19). Figure 2. Feed consumption pondered by number and body weight of piglets and feed preference about Fluidarom 1003® as total consume percentage (number on top of the bars), statistical comparison against the preference neutral value (50%; indifference). NS: P>0.1; *: P<0.05; †: P<0.1 (Trial 3). 61 7. DISCUSSION DISCUSSION 63 The present work aimed to study the effects on the piglets behaviour of a preand postnatal exposure through the maternal diet to Fluidarom 1003®, a commercial flavour based on >25 % anethol, >25 % cinnamaldehyde and >10 % eugenol. Preference and acceptance of piglets towards the same flavoured feed at weaning were the main studied parameters, as well as the subsequent growth in the post-weaning period. The first study revealed that preand postnatal exposure to Fluidarom 1003® throughout the maternal diet improved growing performance after weaning. Higher feed intake and growth were observed along the pre-starter and starter phase (0 to 35 days post-weaning) mainly due to the maternal exposure to the flavour. We suggest that Fluidarom 1003® supplementation in sow diet may promote an early food habit formation in the piglets that could affect the piglets feed intake and performance after-weaning. The basis of this effect could be related with the transfer of some compounds to the womb and/or the milk. It has been shown that anise (anethol) offered in swine diets (Langendijk et al., 2007) may be transferred to both amniotic fluid and milk (Hepper, 1988). Moreover, higher feed intake after weaning could also be related with a stress-reducing effect by piglet exposure with a familiar flavour acquired before weaning. This effect widely accepted in other mammals like humans (Varendi et al., 1998) and rats (Smotherman, 1982). However, in our first trial we were not able to discriminate between the pre-natal (gestation) and post-natal (lactation + creep-feed) effects, because both were administered together. Then, piglet learning could have been achieved through one of these steps, or by their interconnection and positive reinforcement from pre-natal to weaning experience. To solve and fractionate these mechanisms, we hypothesized in the second trial that an early creepfeed contact and intake may play an important role on feed and growth performance after weaning. Therefore, the use of the same flavour in maternal diets and in the creep-feed may facilitate creep-feed intake and their further positive reward after weaning when the same flavour is used. The consumption of creep feed and the influence of flavours in the creep feed diet Unexpectedly, the highest growth of piglets during lactation in the second trial was observed in piglets without preand post-natal contact to Fluidarom 1003®. We related this effect to the consumption of unflavoured creep-feed. DISCUSSION 64 Thus, in our second trial we observed that piglets preferred to consume non-flavoured than flavoured creep-feed. Both, litters corresponding to the control sows (non-flavoured diets) and litters corresponding to sows fed a flavoured diet showed higher intake of control creep-feed (non-flavoured) than the creep-feed containing the flavour (277 vs. 153 g/eater piglet, respectively). Therefore, we observed a negative impact of the flavour supplementation on piglet feed intake during lactation, which is different to the results described previously by Langendijk et al. (2007) who observed that creep-feed intake may be stimulated when specific flavours associated with the diet and the milk of the sow are offered. No preference of the suckling piglets for a creep-feed supplemented with a flavour, even if previously was given to the sows, was also reported by (Figueroa et al., 2013). It appears that during the lactating period piglets preferred a simple unflavoured diet (cereal base, without added flavours). Moreover Sulabo et al. (2010a) reported that flavour supplementation to creep-feed diets did not improve daily creep-feed intake, average daily gain or feed efficiency. Our results and those described in the literature suggest that a non-flavoured creep-feed is more consumed than flavoured creep-feed. However, we should remark that what we call “a non-artificial flavoured diet” is in fact a diet containing the intrinsic flavours of their own ingredients. The high content of the creep-feed in dairy by-products, like milk whey, lactose and even skimmed milk, may be one of the main reasons that explain that piglets preferred a plain creep feed over a flavoured modified diet. It could be speculated that some of these intrinsic flavours may be also found in the sow’s diets and their fluids and secretions, favouring the attraction of piglets towards these diets. We could also suggest that the use of creep-feed diets with similar ingredients than those used in maternal diets should be taken into account for creep-feed formulation. To reduce feed neophobia in the piglets, a variety of ingredients should be included in the maternal diets. This idea has been exposed for humans, as higher is the variety of the ingredients consumed by the mother, larger is the number of new volatiles that the baby may be get used to while sucking (Mennella et al., 2001) and lower the negative impact on food acceptance when complex diets are offered to the babies later on. Creep feed intake is one of the other controversial points because often it is difficult to standardise the different condition in which the studies has been performed, and therefore the results obtained in the literature are not comparable. Among the wide range of intakes DISCUSSION 65 reported in the literature, (Pluske et al., 2007) reported creep-feed consumption of 703 g per piglet (SD= 608) in 19 days of lactation, while (Pajor et al., 1991) found an average creep-feed consumption of 468 g (range 13-1385) per piglet in 18 days of lactation and (Bruininx et al., 2002) observed that total creep-feed disappearance was 377 g per piglet (SD=220.5) in 17 days of lactation. In our study we registered a creep feed consumption of 153 g per piglet (SD= 41) with flavoured creep-feed and 277 g per piglet (SD= 41) with non-flavoured creep-feed. The creep-feed consumption of the present study is lower than in others studies which could depend on the different composition of creep-feed, the animal’s genetics, animals per litter and type of creep-feeding. The differences highlight the difficulty of comparing creep-feeding among different studies. Among the main causes that affect creep-feed consumption, Sulabo et al. (2010b) remarked that only 60% of pigs were creep-feed eaters. In our second trial a higher amount of creepfeed eaters was observed for pigs fed non-flavoured (86%) than for pigs fed the flavoured diet (74.5%). Sulabo et al. (2010a) observed that addition of the feed flavour to creep-feed did not affect the proportion of creep-feed eaters. The difference between studies results might be due to the difficulty of marker detection (chromium III oxide (Cr2O3)) detection in faces. Bruininx et al. (2004) reported that a considerable proportion of piglets (33%) could not be designated into any categories due to “indistinguishable colours” of the faces with this dye. Contrarily, Pluske et al. (2007) reported that by using indigo carmine as indigestible marker almost up to 90% of piglets could be classified as eaters due to the presence of with creep feed as their faces staining faeces (cyan). The influence of flavour in the sow diet to guide the post-weaning feed intake of piglets In natural environments, piglets are exposed to a lot of new challenges, and trainand errorlearning is essential in nature. However, in intensive pig production, animals experience sudden changes with little time to choose or learn on the basis of trial and error. One strategy to reduce this psychological impact in piglets is to give them links between one step and another that allow them to feel a familiar connection and reduce stress (Oostindjer et al., 2011b). We observed that pre-natal (gestation) and post-natal (lactation + creep-feed) flavour exposure via maternal diet may reduce the negative impact of weaning when the same flavour was offered in post weaning diet (Trial 1). Moreover, those piglets with pre-weaning flavour DISCUSSION 66 exposure via maternal diet and creep-feed, showed higher BW and ADG along the entire transition period than those with no previous exposure to flavoured diets. The similar results were obtained for those piglets without pre-weaning exposure to flavour but with postweaning flavour exposure. This could be explained because the different experimental treatments were randomized within the same farrowing room. Then sows and their corresponding litters shared the same ambient and a cross-contamination of volatiles within the room could have been occurred. Therefore, we can’t exclude that piglets did not have a previous familiarity with those volatile compounds. In the second trial (Trial 2) we hypothesised that the positive results previously observed in Trial 1 could be explained by the early creep-feed contact and intake following the familiar link with volatile compounds exposed via amniotic fluid and/or milk. However, and contrarily to our hypothesis, the incorporation of the flavour into the creep-feed diet did not improve feed intake during lactation and the piglet preference for Fluidarom 1003® at weaning. On the other hand, a higher effect of the maternal learning was observed when a non-flavoured creep-feed diet was administered to the piglets. A high preference for Lacto-Vanilla, a commercial flavour based on butyric acid, vanillin and diacetyl was observed in those piglets without any previous exposure to flavours (control group). This might be explained because butyric acid is one of main fatty acids in milk; vanillin is the primary component of the extract of the vanilla bean and diacetyl is a natural by-product of fermentation that give butter its characteristic taste. It is possible that butyric acid and diacetyl, both found in sow’s milk, may justify a high preference for this flavour. Having rejected the hypothesis around the positive effect of flavouring creep-feed to bridge the connection between the prenatal learning and the post-weaning period, a third experiment was performed. This experiment aimed to study the most important maternal pathway that could explain a maternal flavour formation, either pre-natal or post-natal. Even though many studies examined perinatal flavour learning in a lot of species during four decades, there is still little information about the mechanisms affecting flavour learning and how perinatal flavour learning may improve feed acceptance and feed intake in piglets (Oostindjer et al., 2010). So, in our third trial we observed that piglets previously exposed to Fluidarom 1003® during the pre-natal period (only gestation) but not in the creep-feed were able to neutralize the high preference previously observed for lacto-vanilla. Those piglets offered a flavoured creep-feed preferred the Lacto-Vanilla flavour after-weaning. It seems DISCUSSION 67 again, that an aversive reaction to the flavour offered in the creep-feed diet turns down the preference for the pre-natal conditioned flavour. This result might be due to flavour supplemented in the creep-feed, and as we suggested previously, the use of a flavoured creepfeed induces to confusion while piglets are still sucking. Moreover, post-natal flavour exposure alone (milk pathway) did not exert any effect on the preference for the pre-weaning flavour exposure either with or without flavour in the creep-fed. As it was previously reported by Oostindjer et al. (2010) piglets only with post-natal flavour exposure (lactation) did not show any preference for the pre-exposed flavour, which suggests that only a pre-natal exposure is the most powerful pathway for maternal flavour conditioning of newborn piglets and it seems enough to create a post-weaning preference for a previously exposed flavour. Figueroa et al. (2013a) also observed that prenatal flavour exposure (anise and milky-cheese) via maternal diet significantly influenced the piglet’s preferences during lactation and until 5 days after weaning (throughout a Triple-U-Testing Arena (TUTA)). However, a preference for Fluidarom 1003® was not observed during lactation in our trial by using the TUTA method in feed. This might be due to the use of solid feed instead of using adsorbent strips impregnated with a high concentration of flavour like the Figueroa et al. (2013a) study. We decided to offer directly a choice between three different diets in order to actually evaluate if piglets were able to discriminate between the two different volatile compounds (commercial flavours; Fluidarom 1003® and Lacto-Vanilla) supplemented on top of the creep-feed diets. Piglets are able to detect volatile compounds in feed at very low levels of inclusion, and then we tried to simulate as much as possible a practical situation. However, young animals did not pay much attention to the different diets likely because it was the first time to have contact with feed. On the other hand, piglets showed an escaping behaviour. Then, we confirmed that young animals can learn about flavours from the maternal diet that appear in the amniotic fluid and also that these preferences acquired before birth for the prenatally exposed flavour were long-lasting. This flavour continuity, may act as a psychological link, remembering to piglets the hedonic remembrance of the maternal environment the post-weaning periods (Arias and Chotro, 2006). ANNEX 75 Table 1. Sow reproductive and productive parameters and litter performance due to Fluidarom 1003® supplementation in gestating and lactating diets (Trial 1). Experimental Treatments P-values CG TG Pooled SEM Treatment n 40 38 - Body Weight (BW; kg) day 73 of gestation 246 248 4.22 0.782 day 110 of gestation 266 266 4.94 0.917 28 days after farrowing 216 215 6.09 0.911 Body Condition Scoring (BCS) day 73 of gestation 2.98 2.93 0.109 0.752 day 110 of gestation 3.10 3.09 0.107 0.945 28d after farrowing 2.67 2.58 0.112 0.541 Back Fat (BFP2; in mm) day 110 of gestation 15.6 15.5 0.83 0.926 28d after farrowing 12.1 12.6 0.83 0.688 Sow farrowing performance ADFI (late gestation; kg) 2.98 2.77 0.049 0.002 Total number of piglets born 16.8 15.9 0.82 0.410 Piglets born alive 13.8 13.3 0.67 0.536 Stillbirth 2.4 2.0 0.44 0.583 Mummies 0.65 0.63 0.251 0.941 BW of piglets born alive, kg 1.35 1.44 0.052 0.194 BW of stillbirth piglets, kg 1.17 1.16 0.087 0.934 Total litter weight at birth, kg 21.0 21.4 0.94 0.754 Sow and litter performance Initial number of piglets by litter (CF) 12.6 13.1 0.39 0.370 Final number of piglets by litter (23d) 11.0 11.6 0.28 0.100 Mortality rate (CF-23d; %) 11.9 10.6 2.91 0.730 ADFI (Cross-Fostering to 23d; kg) 5.43 5.72 0.210 0.319 TFI (Cross-Fostering to 23d; kg) 121.9 129.2 4.68 0.265 Litter weight at CF, kg 17.3 18.7 0.71 0.133 Litter weight at 23d, kg 66.7 73.5 2.48 0.043 Average initial BW (CF; kg) 1.38 1.44 0.060 0.452 Average final BW (23d; kg) 6.06 6.35 0.164 0.184 ADG (CF-23d; kg) 0.208 0.218 0.0065 0.256 The table 1 is simplified; only show the p-value of treatment. But the model included farrowing, batch and their interactions as factors to control the error experimental. CG: Control group (No inclusion Fluidarom 1003® in diets); TG: Treatment group (inclusion Fluidarom 1003® in diets) ANNEX 76 Table 2. Sow reproductive and productive parameters and litter performance due to Fluidarom 1003® supplementation in gestating and lactating diets (Trial 2). Experimental Treatments Pooled SEM P-values T-1 T-2 T-3 T-4 treatment G L CF W G L CF W G L CF W G L CF W + + + + + + - - + + - + - - - - n 7 8 9 9 - Body Weight (BW; kg) day 110 of gestation 262 259 250 258 9.56 0.820 28 days after farrowing 223 251 240 221 9.85 0.090 Body Condition Scoring (BCS) day 87 of gestation 3.72 3.38 3.50 3.75 0.22 0.450 day 110 of gestation 2.86 3.00 2.85 3.02 0.17 0.795 28d after farrowing 2.67 2.29 2.42 2.63 0.14 0.132 Sow farrowing performance ADFI (late gestation; kg) 3.11 3.15 3.10 3.24 0.09 0.637 Total number of piglets born 15.1 14 14.1 13.7 1.82 0.933 Piglets born alive 14.1 13.3 12.9 10.6 1.52 0.297 Stillbirth 0.7 0.6 1.1 2.3 0.60 0.123 Mummies 0.28 0.08 0.23 0.87 0.33 0.241 BW of piglets born alive, kg 1.544 1.518 1.256 1.568 0.10 0.118 Total litter weight at birth, kg 22.354 20.457 16.260 19.395 2.05 0.197 Sow and litter performance Initial number of piglets by litter (CF) 12 12.3 12.4 10.2 0.72 0.071 Final number of piglets by litter (28d) 10 11 11 10 0.75 0.624 Mortality rate (CF-28d; %) 13.8 11.6 13.7 1.7 0.05 0.211 ADFI (Cross-Fostering to 28d; kg) 4.66 4.76 4.83 4.90 0.11 0.435 TFI (Cross-Fostering to 28d; kg) 158.49 162.38 164.23 158.54 4.99 0.762 Litter weight at CF, kg 18.177 18.651 15.423 16.139 1.41 0.247 Litter weight at 28d, kg 71.503 76.923 77.042 78.884 7.10 0.851 The table 2 is simplified; only show the p-value of treatment. But the model included farrowing and their interactions as factors to control the error experimental. G: Gestation; L: Lactation; CF: Creep-Feed; W: weaning. +: inclusion of Fluidarom 1003® (375g/Tm) -: No inclusion of Fluidarom 1003® 77 10. BIBLIOGRAPHY BIBLIOGRAPHY 79 ALGERS, B., JENSEN, P. and STEINWALL, L., 1990. Behaviour and weight changes at weaning and regrouping of pigs in relation to teat quality. Applied Animal Behaviour Science, 26(1–2), pp. 143-155. APPLEBY, M.C., PAJOR, E.A. and FRASER, D., 1992. Individual variation in feeding and growth of piglets: effects of increased access to creep food. Animal Science, 55(01), pp. 147. APPLEBY, M.C., PAJOR, E.A. and FRASER, D., 1991. Effects of management options on creep feeding by piglets. Animal Science, 53(03), pp. 361-366. ARIAS, C. and CHOTRO, M.G., 2006. Interactions between prenatal ethanol exposure and postnatal learning about ethanol in rat pups. Alcohol, 40(1), pp. 51-59. AUMAITRE, A., 1972. Development of enzyme activity in the digestive tract of the suckling pig: nutrition significance and implications for weaning. World Review of Animal Production, 8(3), pp. 54-68. BARNETT, K., KORNEGAY, E., RISLEY, C., LINDEMANN, M. and SCHURIG, G., 1989. Characterization of Creep Feed Consumption and its Subsequent Effects on Immune-Response, Scouring Index and Performance of Weanling Pigs. Journal of animal science, 67(10), pp. 26982708. BILKÓ, Á., ALTBÄCKER, V. and HUDSON, R., 1994. Transmission of food preference in the rabbit: the means of information transfer. Physiology & Behavior, 56(5), pp. 907-912. BOLHUIS, J.E., OOSTINDJER, M., VAN DER BRAND, H., GERRITS, W.J.J. and KEMP, B. Edited by TORRALLARDONA, D. and ROURA, E., 2009. Voluntary feed intake in pigs. Wageningen Academic Publishers. The Netherlands. BORNSTEIN, P.M., LEVINE, M.J. and MARCUS, M., 1975. A rat's first bite: the non-genetic, cross-generational transfer of information. Journal of Comparative and Physiological Psychology, (42), pp. 183-192. BRUININX, E., VAN DER PEET-SCHWERING, C., SCHRAMA, J., VEREIJKEN, P., VESSEUR, P., EVERTS, H., DEN HARTOG, L. and BEYNEN, A., 2001. Individually measured feed intake characteristics and growth performance of group-housed weanling pigs: effects of sex, initial body weight, and body weight distribution within groups. Journal of animal science, 79(2), pp. 301-308. BRUININX, E., BINNENDIJK, G., VAN DER PEET-SCHWERING, C., SCHRAMA, J., DEN HARTOG, L., EVERTS, H. and BEYNEN, A., 2002. Effect of creep feed consumption on individual feed intake characteristics and performance of group-housed weanling pigs. Journal of animal science, 80(6), pp. 1413-1418. BRUININX, E., SCHELLINGERHOUT, A., BINNENDIJK, G., VAN DER PEET-SCHWERING, C., SCHRAMA, J., DEN HARTOG, L., EVERTS, H. and BEYNEN, A., 2004. Individually assessed creep food consumption by suckled piglets: influence on post-weaning food intake characteristics and indicators of gut structure and hind-gut fermentation. Animal Science, 78, pp. 67-75. BIBLIOGRAPHY 80 BUDDINGTON, R., 1998. The influences of dietary inputs on the neonatal gastrointestinal tract: managing the development of a complex ecosystem. Journal of Animal and Feed Sciences, 7, pp. 155-165. CABRERA, R.A., USRY, J.L., ARRELLANO, C., NOGUEIRA, E.T., KUTSCHENKO, M., MOESER, A.J. and ODLE, J., 2013. Effects of creep feeding and supplemental glutamine or glutamine plus glutamate (Aminogut) on pre-and post-weaning growth performance and intestinal health of piglets. Journal of animal science and biotechnology, 4(1), pp. 29-29. CAMPBELL, R.G., 1976. A note on the use of a feed flavour to stimulate the feed intake of weaner pigs. Animal Science, 23(03), pp. 417. CAPRETTA, P.J. and RAWLS, L.H., 1974. Establishment of a flavor preference in rats: Importance of nursing and weaning experience. Journal of comparative and physiological psychology, 86(4), pp. 670-673. CARSTENSEN, L., ERSBOLL, A., JENSEN, K. and NIELSEN, J., 2005. Escherichia coli post-weaning diarrhoea occurrence in piglets with monitored exposure to creep feed. Veterinary microbiology, 110(1-2), pp. 113-123. COUREAUD, G., SCHAAL, B., HUDSON, R., ORGEUR, P. and COUDERT, P., 2002. Transnatal olfactory continuity in the rabbit: Behavioral evidence and short‐term consequence of its disruption. Developmental psychobiology, 40(4), pp. 372-390. CRONEY, C.C., ADAMS, K.M., WASHINGTON, C.G. and STRICKLIN, W.R., 2003. A note on visual, olfactory and spatial cue use in foraging behavior of pigs: indirectly assessing cognitive abilities. Applied Animal Behaviour Science, 83(4), pp. 303-308. DELUMEAU, O. and MEUNIERSALAUN, M., 1995. Effect of Early Trough Familiarity on the Creep Feeding-Behavior in Suckling Piglets and After Weaning. Behavioural processes, 34(2), pp. 185195. DONG, G.Z. and PLUSKE, J.R., 2007. The low feed intake in newly-weaned pigs: Problems and possible solutions. Asian-Australasian Journal of Animal Sciences, 20(3), pp. 440-452. DOUGHERTY, R., SHIPE, W., GUDNASON, G., LEDFORD, R., PETERSON, R. and SCARPELLINO, R., 1962. Physiological mechanisms involved in transmitting flavors and odors to milk. I. Contribution of eructated gases to milk flavor. Journal of dairy science, 45(4), pp. 472-476. DWYER, D.M., PINCHAM, H.L., THEIN, T. and HARRIS, J.A., 2009. A learned flavor preference persists despite the extinction of conditioned hedonic reactions to the cue flavors. Learning & Behavior, 37(4), pp. 305-310. FIGUEROA, J., SOLÀ-ORIOL, D., BORDA, E., SCLAFANI, A. and PÉREZ, J.F., 2012a. Flavour preferences conditioned by protein solutions in post-weaning pigs. Physiology & Behavior, 107, pp. 309-316. FIGUEROA, J., SOLÀ-ORIOL, D., GUZMÁN-PINO, S., BORDA, E. and PÉREZ, J., 2012b. Flavor preferences conditioned by postingestive effect of sucrose and porcine digestive peptides in postweaning pigs. Journal of animal science, 90(Supplement 4), pp. 381-383. BIBLIOGRAPHY 81 FIGUEROA, J., SOLÀ-ORIOL, D., VINOKUROVAS, L., MANTECA, X. and PÉREZ, J., 2013a. Prenatal flavour exposure through maternal diets influences flavour preference in piglets before and after weaning. Animal Feed Science and Technology, 183(3), pp. 160-167. FIGUEROA, J., SOLÀ-ORIOL, D., MANTECA, X. and PÉREZ, J.F., 2013b. Social learning of feeding behaviour in pigs: Effects of neophobia and familiarity with the demonstrator conspecific. Applied Animal Behaviour Science, 148(1-2), pp. 120-127. FRASER, D., FEDDES, J. and PAJOR, E., 1994. The Relationship between Creep Feeding-Behavior of Piglets and Adaptation to Weaning - Effect of Diet Quality. Canadian Journal of Animal Science, 74(1), pp. 1-6. FRASER, D. and JONES, R.M., 1975. The ‘teat order’ of suckling pigs: I. Relation to birth weight and subsequent growth. The Journal of Agricultural Science, 84(03), pp. 387-391. FRIESEN, K., GOODBAND, R., NELSSEN, J., BLECHA, F., REDDY, D., REDDY, P. and RICHERT, B., 2010. The effect of pre-weaning exposure to soybean meal on subsequent post-weaning growth performance in the early-weaned pig. Report of progress (Kansas State University.Agricultural Experiment Station and Cooperative Extension Service), 641, pp. 31-34. GALEF, B.G. and HENDERSON, P.W., 1972. Mother's milk: A determinant of the feeding preferences of weaning rat pups. Journal of comparative and physiological psychology, 78(2), pp. 213-219. GALEF, B.G. and SHERRY, D.F., 1973. Mother's milk: a medium for transmission of cues reflecting the flavor of mother's diet. Journal of comparative and physiological psychology, 83(3), pp. 374-378. GALEF, B.G., 1996a. Food selection: problems in understanding how we choose foods to eat. Neuroscience & Biobehavioral Reviews, 20(1), pp. 67-73. GALEF, B.G., 1996b. Social influences on food preferences and feeding behaviors of vertebrates. Why we eat what we eat, pp. 207-231. GALEF, B.G. and WHISKIN, E.E., 2004. Effects of environmental stability and demonstrator age on social learning of food preferences by young Norway rats. Animal Behaviour, 68(4), pp. 897902. GESTIN, M., LE HUËROU-LURON, I., LE DRÉAN, G., PEINIAU, J., ROME-PHILOUZE, V., AUMAITRE, A. and GUILLOTEAU, P., 1997. Effect of age and feed intake on pancreatic enzyme activities in piglets. Publication-european association for animal production, 88, pp. 127-130. GILL, J.C. and THOMSON, W., 1956. Observations on the behaviour of suckling pigs. The British Journal of Animal Behaviour, 4(2), pp. 46-51. GRAVES, H., 1984. Behavior and ecology of wild and feral swine(Sus scrofa). Journal of animal science, 58(2), pp. 482-492. HEPPER, P.G., 1987. The amniotic fluid: an important priming role in kin recognition. Animal Behaviour, 35(5), pp. 1343-1346. BIBLIOGRAPHY 82 HEPPER, P.G., 1988. Adaptive fetal learning: prenatal exposure to garlic affects postnatal preferences. Animal Behaviour, 36(3), pp. 935-936. HEPPER, P.G, 1995. Human fetal olfactory learning. The International journal of prenatal and perinatal psychology and medicine, 7(2), pp. 147-151. HEPPER, P.G. and WELLS, D.L., 2006. Perinatal olfactory learning in the domestic dog. Chemical senses, 31(3), pp. 207-212. HURSTI, U.K. and SJÖDÉN, P., 1997. Food and general neophobia and their relationship with self-reported food choice: familial resemblance in Swedish families with children of ages 7–17 years. Appetite, 29(1), pp. 89-103. JENSEN, P., 1988. Maternal behaviour and mother—Young interactions during lactation in free-ranging domestic pigs. Applied Animal Behaviour Science, 20(3–4), pp. 297-308. KENWORTHY, R. and ALLEN, W., 1966. Influence of diet and bacteria on small intestinal morphology, with special reference to early weaning and Escherichia coli Studies with germfree and gnotobiotic pigs. Journal of comparative pathology, 76(3), pp. 291-IN18. KIDDER, D.E. and MANNERS, M.J., 1978. Digestion in the pig. Scientechnica, Kingston, Bath, U.K. KLINDT, J., 2003. Influence of litter size and creep feeding on preweaning gain and influence of preweaning growth on growth to slaughter in barrows. Journal of animal science, 81(10), pp. 2434-2439. KULLER, W.I., SOEDE, N.M., VAN BEERS-SCHREURS, H.M.G., LANGENDIJK, P., TAVERNE, M.A.M., KEMP, B. and VERHEIJDEN, J.H.M., 2007. Effects of intermittent suckling and creep feed intake on pig performance from birth to slaughter. Journal of animal science, 85(5), pp. 1295-1301. KULLER, W.I., TOBIAS, T.J. and VAN NES, A., 2010. Creep feed intake in unweaned piglets is increased by exploration stimulating feeder. Livestock Science, 129(1-3), pp. 228-231. LALAND, K., RICHERSON, P. and BOYD, R., 1993. Animal social learning: toward a new theoretical approach. Perspectives in ethology, 10, pp. 249-277. LALLÈS, J.P., BOSI, P., SMIDT, H. and STOKES, C.R., 2007. Nutritional management of gut health in pigs around weaning. Proceedings of the Nutrition Society, 66(02), pp. 260-268. LANGENDIJK, P., BOLHUIS, J.E. and LAURENSSEN, B.F.A., 2007. Effects of preand postnatal exposure to garlic and aniseed flavour on preand postweaning feed intake in pigs. Livestock Science, 108(1–3), pp. 284-287. LAUNCHBAUGH, K., 1995. Effects of neophobia and avertions on feed intake: why feedlot cattle sometimes refuse to eat nutritious feeds, Symposium: Intake by feed cattle. Oklahoma. proceedings 1995, pp. 36-48. BIBLIOGRAPHY 83 MAINARDI, M., POLI, M. and VALSECCHI, P., 1989. Ontogeny of dietary selection in weaning mice: effects of early experience and mother's milk. Biology of behaviour, 14(3), pp. 185-194. MAKINDE, M.O., UMAPATHY, E., AKINGBEMI, B.T., MANDISODZA, K.T. and SKADHAUGE, E., 1997. Differential Response of Legumes and Creep Feeding on Gut Morphology and Faecal Composition in Weanling Pigs. Comparative Biochemistry and Physiology Part A: Physiology, 118(2), pp. 349-354. MAKKINK, C.A., 1993. Of piglets, dietary proteins, and pancreatic proteases. Ph.D. Thesis. Department of Animal Nutrition, MENNELLA, J.A., JAGNOW, C.P. and BEAUCHAMP, G.K., 2001. Prenatal and postnatal flavor learning by human infants. Pediatrics, 107(6), pp. e88. MENNELLA, J.A. and BEAUCHAMP, G.K., 1999. Experience with a flavor in mother's milk modifies the infant's acceptance of flavored cereal. Developmental psychobiology, 35(3), pp. 197-203. MORROW-TESCH, J. and MCGLONE, J.J., 1990. Sources of maternal odors and the development of odor preferences in baby pigs. Journal of animal science, 68(11), pp. 3563-3571. MYERS, K.P. and SCLAFANI, A., 2006. Development of learned flavor preferences. Developmental psychobiology, 48(5), pp. 380-388. NICOL, C.J. and POPE, S.J., 1994. Social learning in sibling pigs. Applied Animal Behaviour Science, 40(1), pp. 31-43. NRC. 2012. Nutrient requirements of swine. 11th revised edition. National Academic Press, Whashington, D.C. OKAI, D., AHERNE, F. and HARDIN, R., 1976. Effects of creep and starter composition on feed intake and performance of young pigs. Canadian Journal of Animal Science, 56(3), pp. 573-586. OOSTINDJER, M., BOLHUIS, J.E., VAN DEN BRAND, H. and KEMP, B., 2009. Prenatal flavor exposure affects flavor recognition and stress-related behavior of piglets. Chemical senses, 34(9), pp. 775-787. OOSTINDJER, M., BOLHUIS, J.E., VAN DEN BRAND, H., ROURA, E. and KEMP, B., 2010. Prenatal flavor exposure affects growth, health and behavior of newly weaned piglets. Physiology & Behavior, 99(5), pp. 579-586. OOSTINDJER, M., BOLHUIS, J.E., MENDL, M., HELD, S., VAN DEN BRAND, H. and KEMP, B., 2011a. Learning how to eat like a pig: effectiveness of mechanisms for vertical social learning in piglets. Animal Behaviour, 82(3), pp. 503-511. OOSTINDJER, M., BOLHUIS, J.E., SIMON, K., VAN DEN BRAND, H. and KEMP, B., 2011b. Perinatal flavour learning and adaptation to being weaned: all the pig needs is smell. PloS one, 6(10), pp. e25318.