ONE OR TWO DIETS FOR LACTATING SOWS?

ONE OR TWO DIETS FOR LACTATING SOWS? EN ELLER TO FODERBLANDINGER TIL DIEGIVENDE SØER? TRINE FRIIS PEDERSEN MASTER THESIS AGROBIOLOGY FEBRUARY 2015 Main supervisor: Peter Kappel Theil, Department of Animal Science, Aarhus University Co-supervisor: Thomas Sønderby Bruun, SEGES, Danish Pig Research Centre AARHUS UNIVERSITY

PREFACE The current thesis is completed as the final part of the master degree in Agrobiology at Aarhus University. The thesis corresponds to 45 ECTS and includes a practical experiment and a review of published literature. The experiment was carried out during spring 2014 at the facilities of Aarhus University, Faculty of Science and Technology, Foulum. The practical experiment and results are presented in a manuscript, intended to be submitted to the Journal of Animal Science. As the manuscript only presents some results and a discussion of these, a separate section was made to present additional results and discus the results in general. This thesis aims to clarify perspectives for feeding lactating sows with a new dynamic feeding regime composed of two diets. The target audience of this master thesis are students, pig farmers, advisors and scientists within the field of pig nutrition with special interest in feeding of lactating sows. I would like to express my gratitude to my supervisors Peter Kappel Theil, Senior Researcher at the Department of Animal Science, Faculty of Science and Technology, Aarhus University, and Thomas Sønderby Bruun, Project Manager at SEGES, Danish Pig Research Centre. Peter Kappel Theil for support, critical view and comments on the thesis and Thomas Sønderby Bruun for his practical view and comments. A special thanks to team Theil: Peter, Christine, Uffe and Takele from the Department of Animal Science, Aarhus University, for helping with collection of data during the practical experiments. Furthermore, I thank Takele for instructions concerning calculations of nutrient balances and my aunt for proofreading this thesis. Foulum, February 2015 Trine Friis Pedersen

ABSTRACT There has been a long tradition for feeding lactating sows ad libitum (semi ad libitum) with a single diet (1-diet feeding regime). Despite numerous attempts to feed sows ad libitum, sows still lose weight throughout lactation which can have long term effects on the reproduction. It might therefore indicate that the feeding strategy does not provide sows the right amount of nutrients or the right ratio among nutrients. Sows requirement for nutrients has, however, never been greater due to a rapid increase in milk production during the last 20 years. The objective of the present study was to investigate whether sows should be fed a combination of two diets instead of one diet throughout lactation in order to separately take energy and lysine (Lys) (and nitrogen (N)) requirements into account. The 1-diet feeding regime represented the Danish feeding standards and recommendations whereas the new 2-diet regime supplied sows feed according to their individual requirement. The 2-diet regime was composed of a basal diet, formulated to cover the energy requirement for maintenance and a lactation supplement formulated to cover the increasing requirement for milk production. The study showed that feeding lactating sows with a 2-diet feeding regime increased milk yield and mean piglet weight. Sows feed intake could also be reduced by feeding sows with a well-balanced diet in the right amount. It was not possible to supply the 2-diet sows with enough energy to reach a zero energy balance, and sows lost weight throughout lactation. This study has opened for the option to feed sows more dynamically with a 2-diet regime and be able to take into account sows changing requirement for Lys and N relative to energy. The theoretically required ratio of Lys to energy throughout the lactation was found to increase from 0.38 on day 2 to 0.57 on day 28 of lactation, which corresponds to 4.8 and 7.2 g SID Lys/kg (12.56 MJ/kg), respectively. It could therefore be recommended to increase the Danish requirements for Lys from the present recommendation of 6.6 g SID Lys/Fe for sows fed a single lactation diet. Furthermore, sows fed the 2-diet feeding regime were over supplied with N, which was also supported by the increase in plasma urea concentration. Based on this it became evident that the requirement for nonessential amino acids does not follow the same increase in Lys requirement; it would therefore be beneficial to uncouple the dietary supply of Lys from N in diets for lactating sows.

TABLE OF CONTENTS INTRODUCTION... 1 Objectives and hypothesis... 2 Abbreviation key... 2 LITERATURE REVIEW... 3 1 Practical feeding of lactating sows... 3 1.1 Voluntary feed intake... 4 1.1.1 Sow factors... 5 1.1.2 Environmental factors... 6 1.1.3 Diet factors and water supply... 7 2 Milk yield... 8 2.1 Piglet characteristics influence on milk production... 9 2.2 Sow characteristics influence on milk production... 10 2.3 Environment characteristics influence on milk production... 11 2.4 Diet characteristics influence on milk production and push pull effects... 12 2.5 Inadequacies of sows milk production... 14 2.6 Estimation of sows MY... 15 3 Nutrient requirement of the lactating sow... 16 3.1 Energy requirement... 17 3.2 Lysine requirement... 18 3.3 Nitrogen requirement... 18 4 Nutrient balances during lactation... 19 4.1 Relation between plasma parameters and nutrient balances during lactation... 22 5 Long term consequences of suboptimal feeding... 23 6 Final remarks and conclusion of literature review... 25 MANUSCRIPT TO BE SUBMITTED... 26 ADDITIONAL RESULTS AND GENERAL DISCUSSION... 54 Accuracy of sows MY... 54 Milk production and piglet weight gain... 55

Feed intake, changing requirement and BW loss... 57 Nutrient balances and mobilization of body protein and fat... 60 Intermediary metabolism of lactate, TAG and urea... 63 CONCLUSION... 65 PERSPECTIVES AND IMPLEMENTATION... 66 LITERATURE CITED... 67

Section I

INTRODUCTION Productivity of lactating sows has improved significantly during the past decades and high prolific sows today give birth to on average 15.4 live born piglets per litter (Vinther, 2014). On average sows raise 2.26 litter per year (Vinther, 2014) and can produce as much as 1 to 1.5 times their own body weight (BW) in milk in each lactation (Hansen et al., 2012b). Furthermore, modern sows today produce double the amount of milk that sows produced 20 years ago. Breeding programs have also aimed to select for leaner pigs with lower body fat depots, which has unintendedly resulted in leaner sows with a lower voluntary feed intake (VFI) (Smith et al., 1991; Bergsma et al., 2009). The lactation period of sows is an important phase of the reproductive cycle, due to the high requirement for nutrients to support milk synthesis, and because a high milk intake is crucial for survival and performance of piglets. The onset of lactation makes a very sudden and large metabolic demand and sows feed intake is normally insufficient to overcome the requirement. Sows are therefore often in negative energy balance during lactation due to mobilization of body reserves to maintain milk production. To meet the high demand for milk production, the sow will start mobilizing body reserves, and it is common for sows to lose 20-30 kg during lactation (Theil et al., 2012; Christensen and Sørensen, 2013). It has previously been speculated that sows must mobilize nutrients from their own body to reach their potential for milk production. However, more recent scientific evidence suggests that improved sow nutrition is able to both increase MY and greatly decrease sow mobilization (Flummer et al., 2014). Milk production and VFI vary among sows. However today s feeding system for lactating sows is not designed to cover the individual sow s requirement including e.g. parity, litter size and body condition. A fixed energy and protein ratio in the lactation diet does not either match the shift in N, Lys and energy requirements which appears during the lactation period (Feyera, 2014b). Improper feeding of lactating sows can result in excessive weight loss, which is inappropriate for the longevity of sows. Excessive tissue mobilization during lactation is undesirable because it can have long term effects on sow reproduction (Dourmad et al., 1994), sows weaning to estrous interval (WEI) will be extended (Zak et al., 1997). Furthermore, thin sows are more prone to developing shoulder lesions (Kaiser and Petersen, 2014). Extended weight loss is often compensated during the following gestation. 1

INTRODUCTION Objectives and hypothesis Objective of the study: The objective of this master thesis was to investigate whether sows fed a combination of two diets perform better than sows fed a single diet throughout lactation. The hypothesis of the study: Uncoupling of the dietary supply of Lys and energy using a 2-diet feeding regime will improve MY and nutrient balances and reduce sows weight loss. Abbreviation key Amino acid Back fat Body weight Crude protein Dry matter Lysine Milk yield Non-esterified fatty acid Nitrogen Triglycerides Thermoneutral zone Voluntary feed intake Weaning to estrous interval AA BF BW CP DM Lys MY NEFA N TAG TNZ VFI WEI 2

Section II

LITERATURE REVIEW 1 Practical feeding of lactating sows Feed intake during lactation is important to ensure high performance of the sow and to avoid excessive weight loss. During lactation sows are fed to approach ad libitum feeding (semi ad libitum), where sows should have eaten up within a 30 min. range from feeding. This feeding strategy aims to ensure a high VFI to meet the elevated requirement for nutrients during lactation and furthermore to avoid feed waste. The Danish recommendation for feed allowance is to gradually increase the intake within the first and second weeks of lactation and hereafter reach a plateau (Hansen, 2012). Intake of feed and nutrients is illustrated in Figure 1. Sows receive approximately 3.3 kg of feed from parturition until day 3 of lactation. Feed intake around parturition is reduced compared to late gestation to avoid problems with e.g. MMA and birth complications (Dourmad et al., 1994; Danielsen, 2003). If feed allowance is too high in early lactation, sows will reach the peak in feed intake within a few days from parturition, which unfortunately triggers a major drop in feed intake (Sørensen, 2005). After day 3 and until day 10 sows feed supply rapidly increases by about 0.5 kg per day and 0.2 kg from d 10-20. From lactation day 20 feed intake reaches a plateau of approximately 9 kg/day. Feeding of Danish sows is in general limited to one lactation diet, fed from one week before parturition until weaning, and consequently the protein and Lys to energy ratio is constant throughout the lactation period (Figure 1). Feed is normally distributed automatically 3-4 times a day as dry or liquid feed (Sønderby, 2013). The one diet system prevents the possibility to adjust the composition of diet according to the individual sow s requirement. The feeding strength is therefore adapted from a standard feeding curve and does not take into account the differences in requirement for maintenance or milk production. In general there is only one or two feeding curves for all sows in the herd and the feed allowance is adjusted upwards if the sows have eaten up and licked the trough or downgraded in case of leftovers. Weight loss due to inadequacies in feed intake or allowance in the lactation period must therefore be reconstituted in the following gestation period (Dourmad et al., 1996). 3

LITERATURE REVIEW Kg 10 9 8 7 6 5 4 3 2 1 0 1 3 5 7 9 11 13 15 17 19 21 23 25 27 Day g SID 180 160 140 120 100 80 60 40 20 00 1 3 5 7 9 11 13 15 17 19 21 23 25 27 Day 140 120 100 80 60 40 20 00 MJ Figure 1 - Upper figure represents a standard feed intake curve for lactating sows expressed as kg/day (Danish Pig Research Centre). Lower figure represents intake of energy (- - -) in MJ/day, N ( ) in SID g/day and Lys (. ) in SID g/day based on typical contents (Theil, 2015). 1.1 Voluntary feed intake The regulation of sows feed intake is controlled by a multifactorial system (Forbes, 1988). Recent reviews of Eissen et al. (2000) and Hansen (2012) showed that several factors can affect feed intake during lactation. These factors can be grouped into sow, environment and diet related factors and will be discussed below (Williams, 1998). 4

LITERATURE REVIEW 1.1.1 Sow factors Animal factors affecting feed intake during lactation are related to parity/age, BW and condition, length and stage of lactation, litter size and genetics (O'grady et al., 1985; Neil et al., 1996; Hansen, 2012). The feed intake pattern varies individually for sows during lactation, but in general feed intake covaries with milk production and stage of lactation (Koketsu et al., 1996a). After parturition the voluntary feed intake (VFI) is low and starts to accelerate, and the sows will relatively quickly reach a high feed level within 3 days from parturition. After this rapid increase VFI drops and stays low for a few days before it slowly increases again and reaches the highest feed intake around day 11-16 of lactation (Sørensen, 2005). In contrast, milk production will reach a plateau around day 16-19 and feed will therefore not continue to be supplied in increasing amounts. The lower feed intake in early lactation has been reported to be correlated with size of gastrointestinal tract, which will adapt gradually to an increased feed intake (Dourmad, 1991). Increased feed intake induced by milk production corresponds to an additional feed intake of 45 g/day per 1 MJ milk output (Noblet and Etienne, 1989; Dourmad, 1991). This corresponds to an increase in feed intake of 225 g/day if MY is increased by 1 kg (1 kg milk with 5 MJ 0.45 g/day). The feed intake of sows with a greater litter size and milk production will increase due to a greater requirement. Noblet et al. (1998) demonstrated a relationship between feed intake and litter size. Feed intake was increased by 0.6 kg from 4.4 to 5.0 kg/day as litter size increased from 3 to 13 piglets. In contrast, O Grady estimated an increase of 0.96 kg/day and indicated a limit for feed intake when litter size reached 14 piglets. The additional increase in VFI when litter size increases is, however, considered negligible compared with the increased requirement for nutrients to cover milk production (Noblet et al., 1998). Feed intake has shown also to be dependent on sows parity (Noblet et al., 1998). Primiparous sows have a lower feed intake compared to multiparous sows. Koketsu et al. (1996a) calculated the relationship between lactation feed intake and parity. From 1 st to 9 th parity feed intake increased by 0.81 kg/day, from 4.51 to 5.32 kg/day, respectively. Noblet et al. (1998) suggest the increase to be more appreciable from first to second parity, which is confirmed by Koketsu et al. (1996a) with an increase in feed intake of 55% from 1 st to 2 nd parity. The increased feed intake for higher parity sows is suggested by Eissen et al. (2000) to be due to an increased gastrointestinal capacity in older sows 5

LITERATURE REVIEW with higher BW. Sow BW increases with parity and gastrointestinal capacity follows the BW and thereby affects the VFI. Sows body condition at farrowing can affect VFI during lactation (Dourmad, 1991), which is in accordance with O'grady et al. (1985) who suspected that lactation feed intake seems to be more reliant on body condition than BW at farrowing. The reduced feed intake for fat sows is confirmed by (Revell et al., 1998a) who observed a 30% decrease in feed intake during lactation for sows with high body fat at farrowing. The mechanism of why fat sows respond by decreasing feed intake is not clear, but it is suspected to be related to a hormonal feedback system initiated in the adipose tissue. Woods et al. (1990) proposed the possibility that increased insulin concentration in plasma, due to adiposity, provides the signal for CNS to reduce feed intake. In contrast Revell et al. (1998a) proposed that fat sows might have fewer insulin receptors than lean sows and therefore the response would be lower. Explanations have also been suggested to be connected to insulin resistance and glucose intolerance (Eissen et al., 2000) or the increased leptin concentration in fat sows (Williams, 1998). High feed intake during gestation may cause insulin receptors to be insensitive or reduce the number of insulin receptors, which will inhibit the glucose response (Weldon et al., 1994). High feed intake before parturition may also cause a higher glucose intolerance by decreasing glucose receptors. Leptin is produced in the fat tissue and more fat may accumulate more leptin which will act on the CNS (Williams, 1998; Woods et al., 1998). Another reason for the lower feed intake of fat sows has been related to the capacity to secrete milk and the supply of endogenous substrates from body reserves (Revell et al., 1998a; Garst et al., 1999b). If the capacity to use stored energy for milk production is high, the feed intake will decline (Williams, 1998). Lactating sows VFI is a heritable trait and can indirectly be changed by selection for production traits, which may also affect litter performance (Eissen et al., 2000; Bergsma and Hermesch, 2012). Selection for a high daily feed intake or daily gain will increase milk production by affecting VFI during lactation. Selection for increased VFI can be accomplished by direct selection or indirectly by selecting for factors related to daily gain or daily feed intake during the growth phase. Difference in feed intake may also be connected to different breeds, which mainly involves factors related to differences in BW and condition (Eissen et al., 2000). 1.1.2 Environmental factors Environmental factors affecting sows feed intake during lactation are related to ambient temperature, (O'grady et al., 1985; Hansen, 2012). Sows productivity is highly depressed with elevated 6

LITERATURE REVIEW ambient temperature to above the thermoneutral zone (TNZ) (Black et al., 1993; Silva et al., 2009). The TNZ for lactating sows is between 12 and 22 C. When the ambient temperature is above the TNZ, feed intake can be reduced by up to 40% and consequently milk production can be reduced by as much as 25% together with an additional inhibition of piglet growth (Black et al., 1993). When the temperature is above the TNZ, the sow will respond by decreasing feed intake to regulate body temperature. The sow maintains body temperature by heat production. To lower the heat production associated with digestion, absorption and utilization of nutrients sows will decrease their feed intake. In the same context Black et al. (1993) suggested that a temperature above the TNZ increases blood flow to the skin at the expense of the udder, which may contribute to a decrease in milk production. 1.1.3 Diet factors and water supply Feed intake can also be influenced by diet factors related to e.g. digestibility, composition, energy density, availability of water and feeding frequency (O'grady et al., 1985; Hansen, 2012). Increasing energy density of the diet is one way to accomplish that sows feed intake meets the increasing demand for nutrients. Increasing energy density will, however, decrease the feed intake (Noblet et al., 1998). Fiber content of the diet will also affect the digestibility of the diet and heat production of the sow. Diets containing high amounts of fiber could also limit the feed intake by changing the physical capacity of the gastrointestinal tract. Increasing the feeding frequency has been studied, but with confounding results (Williams, 1998). Sørensen (2009) found no significant effect on feed intake when sows were fed 3 times a day compared with 5-8 times. The same author did, however, find an increased feed consumption when sows were fed ad libitum contra restricted (Sørensen, 2005). Ad libitum feeding increased feed intake by 500 g/day and furthermore reduced sows weight loss by 7.4-9.1 kg. High feed intake (and hence fatter animals) during pregnancy has also been demonstrated to depress feed intake during the following lactation (Mullan and Williams, 1989). This link seems to be associated with body reserves of fat and protein at the end of gestation as previously mentioned. Revell et al. (1998a) also demonstrated an increased feed intake of 0.8 kg/day in the 3 rd and 4 th week of lactation for sows fed high protein levels during lactation compared to sows fed the low protein diet. Water supply during lactation is also important for maintaining feed intake. On average Kruse et al. (2011) found the average water intake of sows to be 27.5 kg/day during lactation and the water-to- 7

LITERATURE REVIEW feed ratio was 4.9. Inclusion of fiber in the diet can also lead to increased water consumption, leading to a lower feed intake. 2 Milk yield Sows milk yield (MY) has increased significantly in the last 20-30 years (Noblet and Etienne, 1986; Feyera, 2014b) and today the best-performing sows produce 16.5 kg milk at peak lactation (day 17-21) (Hansen et al., 2012b). For a whole lactation period of 28 days sows produce around 350 kg of milk. The secretion of copious milk is initiated on the day after parturition (Theil et al., 2006) and starts to increase, especially during the first 10 days of lactation. The increase in sows milk production after parturition is associated with an ongoing growth of the mammary gland and the MY is to a great extent determined by the number of mammary epithelial cells that produce and secrete milk (Theil et al., 2012). Mammary growth can be affected by lactation stage, nutrition, litter size and gland location (Kim et al., 2000; Hurley, 2001; Nielsen et al., 2001). Piglets nutrient requirement for survival and growth is solely covered by sows milk production during lactation until weaning. Piglets are also often offered creep feed, but the intake is rather low and will not contribute to meeting the requirement significantly. The availability and composition of milk are therefore the major factors contributing to litter weight gain. At the same time milk production is of high priority for the sow, and if feed intake is insufficient to meet the requirement for nutrient, the sow will start mobilizing body tissue to maintain milk production. Sow milk is macrochemically composed of water, protein, fat, lactose, minerals and vitamins (Theil et al., 2012). The chemical composition of sow milk during a 28-day lactation period can be seen in Table 1. The greatest differences in composition occur in the transition period (until day 10 of lactation). After d 10 the milk composition is rather constant (Theil, 2015). Sows ability to produce milk at the expense of other body functions interacts with factors like litter size, parity and ambient temperature. The factors suspected to affect MY and composition will be described in the following. 8

LITERATURE REVIEW Table 1- Development in milk composition during a 28-day lactation period (Klobasa et al., 1987). Time after Total solids [%] Fat [%] Protein [%] Lactose [%] parturition [d] 1* 17.3 5.6 6.4 4.6 2* 18.6 6.5 6.4 4.8 3* 19.0 6.7 6.1 5.2 7 18.3 6.7 5.4 5.6 14 18.2 6.4 5.1 5.9 21 18.7 6.6 5.2 5.8 28 18.1 6.1 5.4 5.8 * 24, 48 and 72 hours after parturition, respectively 2.1 Piglet characteristics influence on milk production Sows milk production is related to the number of functional glands, the suckling intensity and milk removal from individual glands (Auldist et al., 2000). Functional glands are associated with litter size, whereas heavier piglets are associated with increased production per gland. Furthermore, milk production from individual glands may also be influenced by suckling frequency. Litter size has been found to be the most important factor affecting sows MY (Auldist et al., 1998; Revell et al., 1998b; Hansen et al., 2012b). Noblet et al. (1998) calculated an increase in MY of 0.6 kg/day per additional piglet. In the same context, Auldist et al. (1998) found that when litter size increased from six to 14 piglets, the litter gain increased from 1.7 to 2.8 kg/day. However, increasing the litter size and thereby total litter gain will reduce milk intake per piglet, which corresponds to a decrease in average gain from 283 to 202 g/day. This relationship is also supported in a Danish report by Christensen and Sørensen (2013). The decrease in milk intake in litters with more piglets was hypothesized by Kim et al., (1999b) to be ascribed to a decrease in mammary gland size. The relationship between increased MY and increased litter size is mainly suggested to be due to the increased number of functional teats (Auldist et al., 1998). The number of functional teats decides the quantity of milk, and when litter size is increased, more total milk will be removed (Hurley, 2001). In contrast if all the teats of the sow are not utilized, the potential for milk production is reduced and unused teats will undergo involution (Theil et al., 2006). 9

LITERATURE REVIEW A positive relationship between the size of the piglet and the MY has been demonstrated by King et al. (1997). It is suggested that heavier piglets have a greater ability to stimulate the udder and remove more milk resulting in increased mammary growth and higher MY (Kim et al., 2000). A new bout of milk is available for the piglets approximately within 35 min of the preceding suckling bout (Spinka et al., 1997). Piglets will therefore receive a similar amount of milk after 35 min and piglets suckling more frequently can obtain a greater amount than piglets suckling less frequently. In same context Auldist et al. (2000) and Auldist et al. (1995) (cited by King (2000)) were able to demonstrate that shorter suckling intervals result in greater total milk removal leading to an increase in MY. Sow MY could be increased by 14% when the suckling interval was reduced from 44.9 min to 34.9 min. Consequently, the nursing behavior of the sow and piglets may also control the frequency of milk let down (Fraser, 1980). The results of Auldist et al. (2000) furthermore suggest that the frequency of suckling plays a role in mammary gland development and mammary gland size has previously been shown to correlate with MY. Nielsen et al. (2001) also found a correlation between piglet weight gain and the weight of the mammary gland. Heavier and more active piglets are also able to stimulate the mammary gland more, resulting in increased mammary gland growth. The intensity of udder massage was also found by Thodberg and Sørensen (2006) to positively influence mammary gland growth. 2.2 Sow characteristics influence on milk production Sows milk production was previously described according to MY and milk composition at different lactation stages. Differences in milk composition and MY also appear between sows and among breeds (Zou et al., 1992; Mackenzie and Revell, 1998; Farmer et al., 2004). Dourmad et al. (1998) found individual differences among sows from 6.7 to 9.3 kg/d for lower-producing sows and higher-producing sows respectively. Milk production can further be influenced by parity and sow body condition (King, 2000), which will be discussed below. Differences in MY have been found for sows in different parities. In general milk production will be lower for primiparous sows than for multiparous sows. The maximum MY is reached between the second and fourth parity and will hereafter decrease (Salmon-Legagnur (1958), cited by Etienne et al. (1998)). This is supported by Christensen and Sørensen (2013) who found the highest litter gain of piglets for 2-4 parity sows. Beyer et al. (2007) reported that the MY of primiparous sows was 14% lower than that of multiparous sows in second parity. From second to fourth parity MY 10

LITERATURE REVIEW was additionally increased by 10%., from 7.11 to 7.90 kg/day. A greater difference in MY from first to second parity has been found by Elsley (1971; cited by King (2000)), who suggested an increase of 25% from first to second parity. In contrast, it has been demonstrated by Boyce et al. (1997) that primiparous sows were able to produce the same amount of milk as multiparous sows if litter size and weight were identical. Body condition is an important factor influencing sows milk production since sows are able to maintain milk production by mobilizing body reserves when feed intake is insufficient to meet dietary requirement. Sows MY during early lactation seems to be affected more by the body condition of the sow at farrowing than by feeding level during lactation (King, 2000). The intensity of mobilization of body reserves can affect the composition of sow milk and MY (Noblet et al., 1998; Theil et al., 2012). Body condition influence on MY is rather complex because both thin and fat condition of sows will affect MY and should be minimized. Typically the effects of body condition on MY become more evident in the fourth week of lactation when body reserves become depleted. The effect on MY will depend on which nutrient is the limiting factor and whether the sow is able to buffer milk production by catabolism of body reserves (Theil et al. 2012). When sows are too thin, the body reserves will become depleted and MY will be compromised before well-conditioned sows. On the other hand Revell et al. (1998b) reported that fat sows produced 15% less milk than lean sows. In same context Head et al. (1991) and Head and Williams (1991) (cited by Eissen et al. (2000)) reported that fat sows, in comparison with lean sows, had a lower capacity to secrete milk because they had fewer milk secretory cells. 2.3 Environment characteristics influence on milk production Environmental factors such as ambient temperature, construction of crates, noise and photo period may affect sows milk production. The impact of photo periods and noise stimuli from the environment on MY can however be questioned (Fraser, 1980; Etienne et al., 1998; Farmer et al., 2004; Lachance et al., 2010). The key factor in question here will therefore be the effect of increasing temperature. The productivity of sows is depressed when they are exposed to high ambient temperature as previously mentioned. Silva et al. (2009) evaluated the effect of tropical humid climate on sows performance. The authors found a decrease in milk production by 16% from 8.1 to 6.8 kg/d when temperature was increased from 23.7 to 26.1 C. Milk composition content was also affected by tempera- 11

LITERATURE REVIEW ture, milk fat content was higher and milk protein tended to be less when temperature was increased. A temperature about 23.7 C is after all above the TNZ around 12-22 C, and milk production could be depressed even more when comparing with sows within the TNZ. This is confirmed by Black et al. (1993) who observed a reduction of 25% when ambient temperature was increased from 18 to 28 C. Black et al. (1993) developed the hypothesis that when sows are exposed to temperatures above the TNZ, blood flow is redirected from the mammary gland to skin. Decreasing blood flow to the mammary gland will result in reduced nutrient uptake and contribute to the reduced MY. The construction of crates, fixed vs. loose, and space allowance can affect the availability of the udder and the rate of udder massage (Moustsen and Pedersen, 2010). Pedersen et al. (2011) found that piglet weight gain can be increased by 11% when sows are loose housed rather than fixed, indicating an easier access to the udder and higher MY. Space allowance in fixed crates also seems to be important for the availability of the udder and thereby litter gain. 2.4 Diet characteristics influence on milk production and push pull effects Feeding of lactating sows is important for the performance of sows and their litters (Hansen et al., 2012a). Mammary gland growth and thereby MY are influenced by energy and protein intake during lactation (Kim et al., 1999). Although the sow is able to maintain milk production by catabolism of body reserves, milk production still responds to dietary energy and protein intake during lactation. Noblet and Etienne (1986) showed that the energy content in the diet affected milk production. In that experiment energy restricted sows were able to maintain the same MY as normally fed sows. Energy restricted animals were, however, able to increase milk fat and energy output in milk by mobilizing body reserves until day 17. Milk production is therefore reasonably independent of energy intake unless body reserves become depleted. Theil et al. (2004) also demonstrated that sows fed energy above their maintenance requirement prioritize synthesis of lactose higher than covering of maintenance. Increasing the energy density in feed by inclusion of fat was proven by Lauridsen and Danielsen (2004) to increase the daily output of energy and fat in sow milk. The inclusion of different dietary fat sources also reflected the composition of fatty acid in the milk. Overall improvements in MY have not been detected when energy intake was increased (Lauridsen and Danielsen, 2004; Theil et al., 2004). In contrast, energy restricted animals were able to maintain milk production (Noblet and Etienne, 1986). The milk production of sows must however be com- 12

LITERATURE REVIEW promised when body depots become depleted as lactation progresses (Hansen et al., 2012a; Hansen et al., 2012b). The same pattern has been observed when dietary protein and amino acids (AA) content in the diet was increased. Results of Dourmad et al. (1998) indicated that increasing dietary crude protein (CP) from 15.5-17.1% and lysine (Lys) from 0.66-0.87% did not affect MY or the output of energy and nitrogen (N) in milk. Milk composition was not affected by increasing CP and Lys content in diet either. In the experiment of Dourmad et al. (1998) the levels of protein and Lys were above the requirement, and the tested effect was a response of increasing CP and Lys above the requirement. Kusina et al. (1999) determined the effects of insufficient dietary protein intake on the lactation performance of the sow. In the latter experiment sows milk production was impaired by 13% (9.15-7.97 kg/day) when dietary CP content decreased from 19.99 to 8.36% during lactation. Reducing dietary protein also reduced protein and fat content in milk, and the effect seems to be more pronounced on day 18 of lactation than on day 8. In the above section dietary energy and protein intake affecting MY has been illustrated. These factors could contribute to the phenomenon push pull effects. This phenomenon has been explained by Theil et al. (2012) as a relationship between feed intake and MY. When sows are fed insufficiently, the MY may be improved by increasing feed intake, which is known as a push effect. In contrast, MY cannot be further improved by increasing the feed intake when sows are fed optimally. In this situation when sows are fed optimally, the relation between feed intake and MY is the reverse and a high yield will stimulate sows feed intake, which is known as a pull effect. The same picture is evident for energy and protein supplementation. Increasing the supply of energy and protein cannot enhance the MY in well fed sows, but MY will be depressed in highly catabolic sows when body reserves are insufficient to support milk production. Attempts to overfeed primiparous sows, the phenomenon super-alimentation, of (Pluske et al., 1998) clearly illustrate that sows milk production cannot be further enhanced by infusion of additionally 38% energy above requirement. In the mentioned experiment primiparous sows were used, which may prioritize further growth compared to a multiparous sow. The results could then have been different if multiparous sows were being used instead. Addition of 38% energy above the requirement must, however, more than cover the supply for growth, and if milk production could be improved by increasing energy, it should have become evident. Alternatively when energy supply is 13

LITERATURE REVIEW increased, protein and AA should follow too, which was not the case in the experiment of Pluske et al. (1998). 2.5 Inadequacies of sows milk production Sows MY is the most limiting factor for potential piglet growth during lactation (Hartmann et al., 1984; Harrell et al., 1993; Zijlstra et al., 1996). In an experiment with artificial rearing of piglets Harrell et al. (1993) found that sows milk production becomes insufficient to meet piglets energy demands on days 8 to 10 of lactation, which progressively increases through lactation. On day 21 of lactation the estimated milk requirement for all piglets in litters of 10 is 18 kg/day. Zijlstra et al. (1996) also found that piglets weaned and fed milk replacer from day 18-25 weighed 20% more than piglets suckled until day 25. Offering milk replacer ad libitum from day 1 to 21 (Azain et al., 1996) and day 4 to 21 (Park et al., 2014) increased the weaning weight by 16.4% and 11%, respectively. These results demonstrate the insufficiency of sows milk production to meet optimal piglet growth whether weaning occurs during the third or fourth week of lactation. Piglets growth potential during lactation is illustrated in Figure 2. It is estimated that piglets biological potential for growth is 466 g/day from parturition until weaning on day 21 of age, which would require an increase of milk production by 43% each day (Harrell et al., 1993). Figure 2 Piglet growth potential during the lactation period (modified from Le Dividich and Seve (2001)). Although sows milk production seems to be the limiting factor for piglet growth, sows milk production is on the other hand also affected by the nursing demand of the litter (King et al., 1997). 14

LITERATURE REVIEW King et al. (1997) investigated the relationship between piglet BW and milk production in a crossfostering experiment. The authors found that newly farrowed sows fostering 2-week old piglets could increase milk production by 26% in the first week of lactation. Conversely, sows receiving newborn piglets after 2 weeks could decrease MY by 22% in the subsequent week. Auldist et al. (1998) also suggested that sows may be able to produce more milk and particularly in early lactation when litter size increased. The results of the cross-fostering and litter size experiment support the view that sows have the capacity to produce more milk and that the capacity is influenced by the nursing demand. Also it seems more evident that milk production can be altered in early lactation when the sows milk production is not a limiting factor. Attempts have been made by Garst et al. (1999b) and Garst et al. (1999a) to machine milk sows to determine the MY of sows. The above mentioned authors found that litter weight gain was depressed when sows were machine milked and taking care of piglets at same time. The results then indicate that sows are not able to compensate for piglets nursing demand by increasing the MY. Transgenic alteration of sow milk has, however, been proven to improve the lactation performance of sows (Wheeler et al., 2001). Noble et al. (2002) conducted an experiment with first-parity transgenic gilts expressing bovine α-lactalbumin in their milk to improve piglet growth. The authors found that an increased lactose content leading to a higher MY increased piglet growth rate and weaning weight. MY was especially improved in early lactation (d 3-9) by 10-18% resulting in a weaning weight of 5.09 kg compared with 4.85 kg after a 21-day lactation period. 2.6 Estimation of sows MY In practice, MY is predicted indirectly by measuring the milk intake of the piglets. The common way to measure MY has previously been the weigh-suckle-weigh (WSW) or D 2 O method. These methods are laborious and expensive to use so models have been established to estimate MY from the average litter growth (Noblet and Etienne, 1989) and a more recent model from litter growth and litter size (Hansen et al., 2012b). Older models only give vague estimations of the MY or predictions of the average MY. Theil et al. (2002b) also found that milk production when using the WSW method was 12.7% lower than MY found by the D 2 O dilution technique. Hansen et al. (2012b) also found that MY was underestimated by approximately 20% when using the WSW method compared with the D 2 O dilution technique. 15

LITERATURE REVIEW The recent approach to developing a model by Hansen et al. (2012b) was obtained from 18 studies published after 1980. The model estimates the lactation curve of sows by using litter size and average daily gain for the individual sow/ litter as inputs to the model. The influence of litter size and litter gain on MY is illustrated infigure 3. The model can also be used to estimate protein, lactose, fat and energy output in the milk. Figure 3 - Left figure illustrates the effect of changing litter gain from 2.5 kg/day (. ) to 3.1 kg/day (- - -) or 3.7 kg/day ( ) and fixed litter size of 14 on MY. Right figure illustrates the effect of changing litter size from 10 (. ) to 12 (- - -) or 14 ( ) and fixed litter gain of 2.65 kg/day on MY. The figures are obtained from the equations of Hansen et al. (2012b). 3 Nutrient requirement of the lactating sow The improved development in milk production over the years has increased sows requirement for nutrients. The VFI of the sow may become insufficient to meet the increasing demand for nutrients and the sow will compensate by mobilizing body depots to maintain milk production. Feyera (2014a) has also demonstrated that sows requirement for nutrient does not follow the same pattern. This shift in nutrient requirement is however not possible to cover with the one diet system, which will only provide a fixed ratio between energy and protein/aa throughout lactation. During lactation sows nutrient intake should cover the requirement for maintenance and milk production. Requirement for maintenance will depend on age/weight whereas first parity sows will not be full-grown and have a larger nutrient requirement for body growth (Eissen et al., 2000). Older sows have a higher BW than young sows and therefore have a higher maintenance requirement. Furthermore, requirement for milk production and maintenance can be influenced by many factors as previously discussed in sections 1 and 2. 16

LITERATURE REVIEW Lactating sows requirement for nutrients can be determined factorially according to their expenditure for milk production, growth and maintenance and the efficiency of nutrient utilization and/or catabolism of body reserves (Noblet and Etienne, 1987b). In the following section lactating sows requirement for energy, Lys and N will be described. 3.1 Energy requirement Lactating sows requirement for energy (ME) can roughly be ascribed to the sum of energy requirement for maintenance, milk production and heat loss associated with milk production (Theil, 2015). Energy for milk production and maintenance accounts for 69% (54% secreted + 15% heat) and 30%, respectively (Feyera, 2014a). Sows requirement for energy can be determined by the following four equations: Energy for maintenance (HE), MJ: BW 0.75 482 kj Energy secreted in milk (ESM), MJ: MY Energy conc Energy associated with milk production (EMP): (ESM / 0.75) - ESM Total energy requirement (TER), MJ/day = HE + EMS + EMP Energy requirement for maintenance is proportional to metabolic weight and has been estimated by Theil et al. (2004) at 482 kj/kg 0.75 per day. The requirement for milk production can be estimated by calculating the MY and measuring the energy concentration in milk. EMP can be calculated as the efficiency of energy utilization. Efficiency estimates for utilization of ME for milk production were found to be 0.72 by (Noblet and Etienne, 1987b) and 0.78 by (Theil et al., 2004). An average efficiency for energy utilization of milk production can be assumed to be 0.75. TER accounts for at least 95% of the energy required. The supply from regeneration of uterus during the first week and the requirement for mammary growth are left out from the equation because of negligible impact. The efficiency of utilizing body depots to sustain milk production has been calculated at 0.86-0.89 (Noblet and Etienne, 1987b). ME is retained with an efficiency of 0.74 (Strathe et al., 2012). Reconstitution of depots during subsequent pregnancy results in a utilizing efficiency corresponding to the efficiency of direct utilization of dietary ME for milk. The energy efficiency will therefore depend on whether energy is utilized from feed or depots: Energy from feed Energy in milk = 0.75 Energy from feed Body fat depots Energy in milk = 0.74 x 0.875 = 0.65 17

LITERATURE REVIEW The efficiency of utilizing energy is reduced when energy is restored in body tissues before it is being secreted in milk. Theoretically, it is therefore more efficient to supply sows optimally during lactation. 3.2 Lysine requirement The Lys requirement for lactating sows can roughly be ascribed to the sum of requirement for maintenance and milk production. Requirement for Lys corresponds to 5% and 94% of Lys intake to cover the requirements for maintenance and milk production, respectively (Feyera, 2014a). Sows requirement for Lys can be determined by the following three compounds: Lys for maintenance (LM), g SID: 2.46 SID g/day Lys secreted in milk (LSM), g SID: MY Lys conc Total Lys requirement (TLR), g/day = LM + LSM The amount of Lys required for maintenance is a table value from NRC (2012) for a 200 kg lactating sow. The Lys requirement for milk production can be estimated by calculating the MY and measuring the Lys concentration in milk. After parturition regression of the uterus will supply approximately 3.89 g/d from day 2-7, which more than covers the requirement for maintenance. In total Lys supply from the uterus accounts for approximately 13% of the Lys requirement from day 2-7 of lactation and should ideally be taken into consideration in the supply of Lys from day 2-7. The contribution of Lys from the uterus is only valid if the uterine degradation after parturition happens uniformly, which is unlikely (Theil, 2015). The requirement for Lys given here accounts for at least 95% of the requirement. The requirement for mammary growth is a minor part and can be left out from these equations. 3.3 Nitrogen requirement Lactating sows requirement for N is expressed on basis of net energy (feed units) in the Danish feed evaluation system. The dietary supply of N or CP also needs to meet the requirement for AA. Overall the requirement for N is not well defined, mostly because of the uncertainty about the sows requirement for individual essential AAs. Furthermore, urea will be secreted if there is an imbalance in the AA profile and if sows are fed an excess amount of N. It is therefore more appropriate to quantify the amount of N lost in the urine than to evaluate the amount of N required for maintenance (Hansen et al., 2014). 18

LITERATURE REVIEW Roughly N intake should therefore cover the demand for milk production and maintenance corrected for N lost as urea. A factorial approach by Feyera (2014a) estimated N requirement for maintenance based on N lost in urine and N required for milk production, which accounted for 29% and 70%, respectively. Sows requirement for N can be expressed as the following three compounds: N for maintenance (NM), g SID: Loss in urine N secreted in milk (NSM), g SID: MY N conc Total N requirement (TNR), g/day = NM + NSM The amount of N lost as urea can vary from 28-30 % during lactation and is dependent on dietary intake, because protein turnover will increase when the dietary supply of N increases (Theil et al., 2004; Theil, 2015). The N requirement for milk production can be estimated by calculating the MY and measuring the N concentration in milk. The requirement for N given here accounts for at least 95% of the N requirement. In addition an N supply of 10.89 g/d from regeneration of uterus during the first week must be incorporated into the endogenous pool because it contributes with 19-11 % of the N requirement from day 2-7, respectively (Feyera, 2014a). The requirement for mammary growth is a minor part and can be left out from these equations. 4 Nutrient balances during lactation During lactation sows are in negative energy balance in early lactation when feed intake is insufficient to meet the requirement for nutrients. The energy balance for sows can be seen in Figure 4; the upper figure illustrates the energy intake for traditionally fed sows in Denmark and a curve for required energy. From the figure it appears that sows are in negative energy balance in early lactation until day 15 of lactation and become in positive energy balance from day 15 when feed intake is sufficient to meet the requirement. The lower graph in Figure 4 illustrates the required ratio between Lys and energy for lactating sows throughout the lactation period against the ratio sows obtain through feeding with a 1-diet system. The 1-diet system applied in Denmark will therefore only supply sows feed with a fixed ratio throughout lactation although the required ratio increases from approximately days 1 to 16 of lactation. It also appears from the figure that the ratio in a standard lactation diet will only provide sows the right ratio until day 4, after which the diet will be insufficient in Lys relative to energy. 19

LITERATURE REVIEW 140 120 100 ME (MJ) 80 60 40 20 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 Day 0.60 0.55 Lysine:Energy ratio 0.50 0.45 0.40 0.35 0.30 2 4 6 8 10 12 14 16 18 20 22 24 26 28 Day Figure 4 Upper figure: Energy intake (- - -) is obtained from the feeding curves formulated by the Danish Pig Research Centre. Energy requirement ( ) is based on requirement calculation for maintenance, heat energy associated with milk production (Theil et al., 2004) and requirement for milk production calculated from Hansen et al. (2012b) (input litter size = 14 and litter gain = 2.65 kg/day). Lower figure: Lys and energy ratio in a traditional diet for lactating sows (- - -) and the required ratio ( ) throughout the lactating period. The supply of energy and nutrients for maintenance and milk production during lactation comes from feed and mobilized body tissues when feed intake is insufficient (Hansen, 2012). Because of the insufficient feed intake during early lactation high producing sows start mobilizing body re- 20

LITERATURE REVIEW serves to supply energy and nutrients for milk production (Mullan and Williams, 1989) and it is common for older sows to lose 20-30 kg during lactation (Theil et al., 2012). Mullan and Williams (1990) investigated the energy balances of lactating sows in relation to feeding level and found that when sows were fed to appetite, sows will still lose both lipid (520 g/day) and protein (130 g/day); however, the loss was dependent on which feeding strategy was used for the sows during gestation. These results are comparable with the results of Theil et al. (2004) who found that sows fed a high amount of fat (384 g/d) still mobilize the same amount of body fat as sows with normal fat inclusion in the diet (80 g/d), 131 vs. 153 g/day, respectively. In a recent study of Feyera (2014b) sows BW loss in response to four dietary treatments with two levels (high and low) of energy and Lys (HEHL, HELL, LEHL and LELL) was evaluated. LELL, LEHL, HELL and HEHL refer to insufficient amount of energy and Lys, insufficient amount of energy, insufficient amount of Lys, sufficient amount of energy and Lys, respectively. The weight loss, fat and protein loss for each dietary treatment are illustrated in Figure 5. This study underlines the fact that sows mobilize from body reserves to cover the requirement for milk production even though the energy and protein balances are positive. Sufficiently fed sows still lose 12 kg BW during the lactation period. The study also illustrates that the ratio between Lys and energy is crucial for weight loss, because apparently sows are not able to metabolize fat or protein when energy or Lys supplies are insufficient. Flummer et al. (2014) also concluded that in spite of large differences in feed Lys to energy ratio, only minor changes in the ratio between mobilized protein and energy were observed. The optimum ratio between Lys and energy to increase MY was estimated to be 53 g of SID Lys per MJ. Furthermore, it can be seen that LELL sows lose the same amount of protein and fat as sows receiving LEHL; the same pattern was seen for HELL and HEHL fed sows. Overall, it appears that an insufficient supply of energy is most crucial for mobilization from body stores. Furthermore, it is possible for the sows to be in positive energy balance and still mobilize fat because there is an unfavorable ratio between protein and energy. This relation was also observed by Kusina et al. (1999) who found that weight loss and back fat (BF) were reduced when protein and Lys intake was increased. 21

LITERATURE REVIEW 40 a Kg 35 30 25 20 15 10 5 0 a b a a b b a ab b bc c LELL LEHL HELL HEHL Dietary treatment Figure 5 - Weight loss (black column), protein (shaded column) and fat loss (grey column) during lactation in four dietary treatments (modified from Feyera (2014b)). LELL, LEHL, HELL and HEHL refer to low energy and low Lys, low energy and high Lys, high energy and low Lys and high energy and high Lys, respectively. Different letters (a, b, c) within the same color differ significantly (P < 0.05). 4.1 Relation between plasma parameters and nutrient balances during lactation During sows lactation substantial physiological changes occur, going from negative energy balance in early lactation to either zero or positive energy balance later in lactation (around day 15, Figure 4), which will affect the concentration of metabolites in plasma (Mosnier et al., 2010; Hansen et al., 2012a). When milk production is initiated, the requirement for milk synthesis rapidly increases and the feed intake can, however, not keep up with the increasing requirements for nutrients or the sows are not supplied enough. As a consequence sows become catabolic and start mobilizing nutrients to cover the requirement. The catabolism of body depots can to some extent be traced in blood plasma samples, and central plasma metabolites are described in the following: In general plasma concentrations of insulin is closely tied to the circulating concentrations of glucose as insulin acts to lower glucose in plasma and keep glucose constant (Akers and Denbow, 2013). Furthermore, insulin acts to lower plasma concentrations of fatty acid and AAs and to promote the conversion into storage (Cunningham and Klein, 2013). In contrast low plasma insulin will stimulate the catabolism of glycogen and mobilization of muscle and fat tissue, which will cover a part of the nutrient requirement. During the lactation period the general interpretation of the 22

LITERATURE REVIEW role of insulin is, however, too simple to explain the observed changes in plasma insulin and plasma glucose and no clear explanation exists. It has, however, been suggested to be related to an altered distribution of nutrients and lowered insulin sensitivity in peripheral tissues during lactation. This favors the supply of nutrients to the udder (which is insulin independent) at the expense of other body tissues (Theil et al., 2012). The plasma concentration of insulin was found by Mosnier et al. (2010) to increase 4 hours post prandial, after parturition (from app. 24 to 38 µui/ml) and decrease from day 4 of lactation until reaching a plateau at weaning (app. 12 µui/ml). For plasma glucose the concentration slightly increases in early lactation (app. up to 0.93 g/l) and reduces throughout lactation (app. 0.78 g/l at weaning). The decrease in plasma insulin and plasma glucose in lactation is in contrast with the general understanding of insulin, because when insulin levels are high glucose levels should be the reverse. Negative energy balance is associated with increased plasma non-esterified fatty acid (NEFA), which indicates that energy is mobilized from fat tissue. The urea concentration in plasma is indicative of oxidation of AAs and high levels indicate either breakdown of muscle tissues (to cover the requirements for energy, protein or individual AAs) or oxidation of excessive AA (imbalances of AA profile) supplied in the feed. Triglycerides (TAG) and lactate are mainly a response controlled by feed intake as TAG represents the quantity of fatty acids taken up from feed and lactate concentration to some extent reflects the amount of starch converted by lactic acid bacteria. However, lactate can also originate from the conversion of N and propionate in the liver or anaerobic metabolism in the muscles. 5 Long term consequences of suboptimal feeding Commercially housed sows lose large amounts of BW during lactation and excessive weight loss due to negative energy balance during lactation has an unfavorable impact on sows in subsequent reproductive cycles (King and Williams, 1984; Whittemore, 1996). The catabolism of body tissues for maintaining milk production instead of conserving nutrients as reserves for the next reproductive cycle affects the weaning to estrus interval (WEI) and litter size in subsequent reproduction (Zak et al., 1997). Culling of sows in commercial farms often happens because of reproduction failures (Vestergaard et al., 2004). Reproduction problems as a consequence of excessive weight loss have been found to affect reproduction in several ways, including increasing interval from weaning to estrus, an increased inci- 23

LITERATURE REVIEW dence of anestrus, a decreased conception rate, lower ovulation rate and higher embryonic mortality (Zak et al., 1997). In contrast reproductive problems may also arise around farrowing due to increased weight and condition by overfeeding during late pregnancy (Dourmad et al., 1994). These problems may be related to farrowing complication and metabolic disorders and studies have found that increasing the feed intake during late gestation increased the occurrence of agalactia (Persson et al., 1989) and the number of still born piglets (Persson et al., 1989; Cools et al., 2014). The reestablishment of a new reproduction cycle is initiated after weaning when suppression of LH increases (Kemp, 1998). LH stimulates the follicle maturation and is essential for ovulation. The reproductive process is dependent on nutrient availability because depletion of body reserves during lactation establishes a hormonal background which affects follicular growth (Baidoo et al., 1992). When sows feed intake is low during lactation, the concentration of glucose is low as is the secretion of insulin and LH (Koketsu et al., 1996b; Koketsu et al., 1998). As a consequence the low LH secretion will adversely affect the WEI. The WEI seems to be similar in different parities although primiparous sows generally seem to be more sensitive than multiparous sows (Whittemore and Morgan, 1990). Prolonged WEI has also been proven to depend on which body tissue is being mobilized (Reese et al., 1984). Reese et al. (1984) suggest that the WEI is more affected by catabolism of body fat than catabolism of muscle tissue during lactation. In contrast, King (1987) found a critical daily level of energy about 45 MJ DE for which the WEI is affected. A more recent suggestion for lactating sows is to keep weight loss in the range of approximately 15-25 kg (Theil et al., 2012). The longevity and long term performance of sows are best accomplished by avoiding severe body loss and extreme fluctuations in BW and fat reserves (Aherne and Kirkwood, 1985; Eissen et al., 2000). Koketsu and Dial (1997) also found that ensuring a greater feed intake of sows during lactation will improve the reproductive performance and thereby the longevity of sows. Furthermore, thin sows are also more prone to developing shoulder lesions, which has been implemented in Danish culling strategies (Kaiser and Petersen, 2014). Continued selection for lean pigs will further reduce the body fatness of sows and reduce feed intake (Eissen et al., 2000). Selection for higher feed intake or at least average daily weight gain should therefore be considered in further breeding. Reconstitution of body tissue in the subsequent reproductive cycle is not energetically and economically favorable. The efficiency is lower when feed is deposited as body reserves and later mobilized (during lactation) due to oxidation, section 3.1. As a rule of thumb, it takes 4 kg of feed to gain 24

LITERATURE REVIEW 1 kg of BW. Deposition of 1 kg protein is associated with additionally 4.2 kg water and 1 kg fat is associated with additionally 0.17 kg water (Noblet and Etienne, 1987a). The equation for weight and water loss is: Weight loss = 5.20(± 0.12) x protein mobilization + 1.17(± 0.12) x fat mobilization Water loss = 4.20 x body protein loss + 0.17 x body fat loss Recalling the weight loss of 37 kg from Figure 5 for the LELL sows composed by a protein loss of 5.31 kg and a fat loss of 13.9 kg, the following calculations are made: Weight loss = 5.20(± 0.12) x 5.31 + 1.17(± 0.12) x 13.9 Water loss = 4.20 x 5.31 + 0.17 x 13.9 = 44 kg = 24 kg With the equation obtained from Noblet and Etienne (1987a) it is possible to approximately estimate the weight loss from protein and fat loss. Furthermore it is possible to calculate how much of the weight loss can be ascribed to water loss. 6 Final remarks and conclusion of literature review - There may be a potential to further increase sows MY by paying attention to non-nutritional factors (e.g. litter size, litter weight or nursing frequency) - Sows requirement for nutrients is more diverse according to differences in voluntary feed intake and MY - It should be possible to generate a feeding curve more adapted to the energy requirement of the sow to avoid excessive weight loss during lactation - Sows can be fed closer to their requirements in early lactation, but practical experience shows that voluntary feed intake is greatly reduced. - The 1-diet feeding strategy used so far for lactating sows is not well balanced with nutrients required to support milk production during the entire lactation - Negative energy balance should be avoided to improve sows longevity - Avoiding weight loss during lactation improves feed efficiency and feeding costs The literature review indicates the need for a feeding strategy for lactating sows that takes into account the individual sow s requirements to cover maintenance and milk production. This is, however, not possible with a 1- diet feeding strategy. 25

Section III

1 MANUSCRIPT TO BE SUBMITTED 2 3 Running head: Nutrition of lactating sows 4 5 6 7 A 2-diet feeding regime for lactating sows reduces nutrient deficiency in early lactation and improves milk yield 1 8 9 10 T. F. Pedersen* 11 12 13 14 *Department of Animal Science, Faculty of Science and Technology, Aarhus University, DK-8830 Tjele, Denmark. 15 16 17 18 19 20 21 1 The research project was funded by the Ministry of Food, Agriculture and Fisheries of Denmark, grant no. 3405-11-0342 2 Corresponding author: trinef.pedersen@anis.au.dk 22 26

MANUSCRIPT TO BE SUBMITTED 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 ABSTRACT: The objective of the present study was to evaluate whether a new feeding regime composed of two diets throughout lactation could minimize sow weight loss and increase milk yield (MY) and piglet weight gain. In total, 14 sows were included in the experiment from parturition until weaning 28 d later. The sows were fed one of two dietary feeding regimes from lactation d 2 and throughout the lactation period. The 1-diet feeding regime represented the Danish feeding standards and recommendations and the new 2-diet regime supplied sows feed according to their individual requirement. The 2-diet regime was composed of a basal diet, formulated to cover the energy requirement for maintenance and a lactation supplement formulated to cover the increasing requirement for milk production. Sows feed intake and weight loss were affected by an interaction between diet regime and treatment. In lactation wk 4 sows fed the 1-diet feeding regime produced less milk (13.0 kg/d) than the sows fed the 2-diet regime (14.9 kg/d). Piglet weight gain was numerically higher (P = 0.11) throughout the lactation period for sows fed the 2-diet regime. Sows in both dietary regimes were in negative energy balance throughout lactation. Sows fed the 1-diet regime were negative in N and Lys and reached a positive or zero balance in late lactation. For the 2-diet feeding regime sows N and Lys balance was positive throughout lactation, and N loss was higher for sows fed the 2-diet feeding regime. The concentration of urea in plasma was lower for sows fed the 1-diet feeding regime. In conclusion feeding lactating sows with the 2-diet feeding regime throughout lactation improved sows MY and mean piglet weight (as lactation progressed), and the mobilization pattern was altered although the mobilization over the entire lactation period was not affected. Key Words: milk yield, nutrient balances, feeding regime, plasma metabolites. 44 45 INTRODUCTION 27

MANUSCRIPT TO BE SUBMITTED 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 The lactation period is an important part of the sow s reproduction cycle and is crucial for the survival of piglets. Sows milk yield (MY) has increased during the past decades and sows are today able to produce 1.5 times their own weight in milk (Hansen et al., 2012b). Sows MY is, however, relatively insensitive to manipulation by feeding e.g. additional supply of protein or fat (Dourmad et al., 1998; Lauridsen and Danielsen, 2004). During lactation substantial changes occur, and sows mobilize great amounts of nutrients from body reserves (Theil, 2015). Typically, commercial sows lose 10-30 kg of BW during lactation even though they are fed ad libitum (Beyer et al., 2007; Smits et al., 2013; Cools et al., 2014), but weight losses may be even higher (Kim et al., 2009). Excessive BW loss has an unfortunate effect on the subsequent reproductive cycle by e.g. delaying return to estrus (King and Dunkin, 1986; Zak et al., 1997). Furthermore, in the long run, it is not energetically efficient to use body depots to support MY instead of using energy directly from feed (k = 0.72-0.78) (Noblet and Etienne, 1987b; Theil et al., 2004). The efficiency of utilizing body depots (k = 0.86-0.89) and retaining energy (k = 0.74) in the subsequent gestation period reduces the efficiency to approximately 0.65 to sustain MY (Noblet and Etienne, 1987b; Strathe et al., 2012). Lactating sows nutrient requirement is today covered by a single diet fed throughout lactation. The 1-diet regime is not able to take into account the changing requirements of Lys, nitrogen (N) and energy (Feyera and Theil, 2014). Sows requirement for energy is both determined by maintenance and MY, while the Lys requirement is almost exclusively determined by MY (Theil, 2015). An increasing interest in using a dynamic 2-diet regime to cover lactating sows requirement has therefore evolved, but experience is still scarce. The objective of the present study was to evaluate whether a new feeding regime composed of two diets throughout lactation could minimize sow weight loss and increase sow MY and piglet weight gain. 28

MANUSCRIPT TO BE SUBMITTED 70 71 72 73 74 MATERIALS AND METHODS The experiment complied with the Danish Ministry of Justice, Act no. 253 of March 8 2013 concerning experiments with animals and care of experimental animals, and a license issued by the Danish Animal Experiments Inspectorate. 75 76 Animals and Housing 77 78 79 80 81 82 83 84 85 86 87 A total of 14 cross-bred (Danish Landrace x Yorkshire) second parity sows were included in the experiment from parturition until weaning 28 d later. The experiment was carried out at Aarhus University, Foulum, Denmark in the period from March 2014 to May 2014. Sows and their litter were individually housed in fixed farrowing crates (2.7 x 1.8 meter). Pen floor consists of one half concrete floor and one half iron slatted floor. The temperature was kept around 20ᵒC around farrowing and then gradually reduced to 16 ᵒC throughout the experimental period. The piglets were provided with heating lamps in the cave throughout the experimental period and floor heating was turned on the first 14 d after farrowing. Sawdust was provided in the piglet cave before parturition. Until 48 h after farrowing the light was on 24 h and for the rest of the experimental period the light was on from 06-18 and again in connection with feeding from 00.30 to 01.00. 88 89 Diets and Feed Composition 90 91 92 93 Two different feeding regimes were supplied; a standard lactation diet (1-diet feeding re- gime) as a control diet or a basal diet and lactation supplement (2-diet feeding regime). The 1-diet group represented the traditional feed composition and feeding strategy in Denmark, where feed 29

MANUSCRIPT TO BE SUBMITTED 94 95 96 97 98 99 allowance increases in early lactation and is kept high in wk 3 and 4. The 2-diet group received a basal supplement (which covered maintenance requirement) and a lactation supplement (covering requirement for milk production). Dietary formulations of the 1-diet, basal diet and lactation supplement are shown in Table 1. Feed was provided automatically three times a d at 00.30 and 08.30 AM and 04.30 PM. Sows had free access to water. No straw was provided but sows were supplied a rope for stimulation of nursing and rooting behavior. 100 101 Feed Regime Formulation 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 The control diet was formulated according to Danish standards and recommendations (Jørgensen and Tybirk, 2010). Sows fed the 2-diet feeding regime received the energy required for maintenance from the basal diet, and Lys and N content in the basal diet was kept low and formulated to match the requirement for gestating sows (Theil et al., 2004; NRC, 2012). Lactation supplement was formulated to cover the Lys requirement for milk production, and protein was included to obtain the optimal Lys supply based on factorial calculations. And inclusion level of synthetic Lys was the same as in the control diet. The CP content for basal diet and lactation supplement was dictated by the Lys content. The energy content of the basal diet was comparable to that commonly used for gestating sows, whereas the dietary energy in the lactation supplement was comparable to a standard lactation diet. The energy content was elevated in the lactation supplement by addition of soybean oil. The basal diet and lactation supplement was formulated to contain the same amount of barley and wheat. The basal diet was, furthermore, formulated to contain more dietary fibers. Feeding Level. The 1-diet regime sows were fed 3.3 kg from lactation d 1 to 3, from d 4 to 10 feed allowance increased by 0.5 kg per day to 6.6 kg. From lactation d 11-20 sows feed allowance increased to 9 kg and it was kept constant until weaning. The feeding level for sows fed the 2-30

MANUSCRIPT TO BE SUBMITTED 118 119 120 121 122 123 124 125 126 diet feeding regime was based on the individual sow s energy requirement. The requirement for energy was calculated to ensure optimal supply of energy, and supply of basal diet and lactation supplement together ensured optimal supply of energy and Lys for both the sow and the milk production. Energy requirement for maintenance was calculated according to 482 kj/kg 0.75 by Theil et al. (2004). Requirement for milk production was calculated based on a forecast of the MY, the predicted energy content of milk (Hansen et al., 2012b), energy associated with milk production (Theil et al., 2004) and corrected for the energy available from uterus regression (Theil, 2015). The requirement was initially covered with 70% of requirement on d 2, and it increased by 2% units daily based on the results of Feyera and Theil (2014). 127 128 Experimental Procedure 129 130 131 132 133 134 135 136 137 138 139 140 141 Back Fat and BW. The sows were weighed on d 2, 7, 21 and 28 after parturition. Back fat (BF) was measured on d 2, 14 and 28 d after parturition by ultrasound using a 7.5-MHz Linear Intraoperative Probe Aloka (UST-556/7.5, Simonsen & Weel, Vallensbæk Strand, Denmark). The BF was measured 3 times on the right and left side of the sow, 65 mm from the last (12 th ) backbone (conventionally known as P2 measurement). Sows BF is represented as mean values from six measurements. Piglets were weighed individually on d 2, 4, 7, 14, 21 and 28 for forecasting MY during the next week and for determination of MY in the past week. Feces Scoring. Feces were scored every d from farrowing until weaning. The feces were given a score from one to five, score one being dry and pellet-shaped and five being very wet feces, unformed and liquid (Oliviero et al., 2009). Milk Samples. Milk samples from the sows were drawn on d 3, 10, 17 and 24 after farrowing for further analyses. Piglets were removed from the sow during milk sampling and the sow was 31

MANUSCRIPT TO BE SUBMITTED 142 143 injected 2 ml oxytocin intramuscular to induce milk let down. A total of 45-50 ml milk was drawn from 2 to 3 teats and filtered before storing at -80ᵒC until analysis. 144 145 146 147 148 149 150 151 152 153 154 155 156 Blood Samples. Blood samples were drawn 4 h after morning feeding on d 3 and 17 after parturition. Blood samples were collected (2 x 9 ml) by jugular vein puncture into 10 ml heparinized tubes (Greiner BioOne GmbH, Kremsmünster, Austria) and placed on ice. Blood samples were centrifuged at 3,000 rpm at 4ᵒC for 20 min. Plasma was harvested immediately after centrifugation and stored at -80ᵒC until analysis. General Management and Health. Feed leftover from sows was collected once daily. The feed leftover was weighed and registered from every sow throughout the experimental period. Sows with feed leftovers were supplied less the following two days (50 % and then 75%) until the scheduled level was obtained. All litters were standardized 24 h postpartum to 14 piglets (over 850 g). Piglets were further given iron injection, tail docking, and castrated around d 4 of lactation. Sows and piglets were monitored daily for health condition and were treated in compliance with normal procedures. 157 158 Chemical Analyses 159 160 161 162 163 164 The diets were analyzed for GE, DM, CP, crude fat, crude fiber, starch, ash and AA. Feed GE was determined with an adiabatic bomb calorimeter (Parr Instrument Company, Moline, Illinois, USA). Feed DM and ash were analyzed gravimetrically according to the Official Journal of the European Union (152/2009). Feed CP was determined according to the dumas method (Hansen, 1989) and AA was determined based on amino acid analyzer (EU 152/2009). Crude fat was deter- 32

MANUSCRIPT TO BE SUBMITTED 165 166 167 168 169 170 171 172 173 174 175 mined by petroleum ether extraction after hydrochloric acid hydrolysis (EU 152/2009). Dietary content of starch and fiber was analyzed as described by Bach Knudsen (1997). Transient milk was analyzed in triplicate for DM, fat, protein and lactose content with infrared spectroscopy using a Milkoscan 4000 (Foss MilkoScan, Hillerød, Denmark). Plasma was analyzed in duplicate for insulin, glucose, lactate, NEFA, triglycerides and urea. Plasma samples were analyzed for insulin by time-resolved fluoroimmunoassay as described by (Løvendahl and Purup, 2002). Plasma content of NEFA was determined using the Wako, NEFA C ACS-ACOD assay method (Wako Chemicals GmbH, Neuss, Germany), and glucose, lactate, triglycerides and urea were analyzed according to standard procedures (Siemens Diagnostics Clinical Methods for ADVIA 1650) using an auto analyzer (ADVIA 1650 Chemistry System, Siemens Medical Solution, Tarrytown, NY). 176 177 Calculations 178 179 180 181 Dietary ME content was calculated based on analyzed GE content and expected ME/GE ra- tio from the feed formulation. SID N and Lys were calculated from the analyzed content in feed relative to the expected amount of Lys g/kg and SID Lys g/kg by the following equation (Eq. 1): 182 183 Analyzed N/Lys x Expected N/Lys Expected SID N/Lys where expected N and Lys are calculated from the feed evaluation. (Eq. 1) 184 185 186 187 Sows MY was estimated from Hansen et al. 2012, based on litter weight gain and litter size. Energy secreted in milk was calculated as the sum of energy secreted through milk protein, fat and lactose, respectively. The energy values of 23.86, 39.79 and 16.51 kj/g (Weast, 1969) for protein, fat and lactose, respectively, were multiplied by their respective gram content in daily MY. 33

MANUSCRIPT TO BE SUBMITTED 188 189 190 191 192 193 194 Daily protein and fat mobilization and body protein and fat content were estimated from BW and BF measurements on d 2 and 28 according to the model developed by Rozeboom et al. (1994) for Landrace x Yorkshire sows as: Body protein (kg) = 2.5 + 0.146 x BW - 0.11 x BF (Eq. 2) Body fat (kg) = -15.1 + 0.36 x BW + 0.306 x BF (Eq. 3) Nutrient balances were estimated in accordance with the calculations made by Feyera and Theil (2014). 195 196 Statistical Analyses 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 The experiment was regarded as a complete randomized design in which sows were stratified for weight. Weekly recordings of feed and nutrient intake, feces consistency, milk composition, change in body composition and weight, litter size, mean piglet weight, piglet weight gain, nutrient balances, nutrient requirement for maintenance and milk production and plasma metabolites were applied to the following model to analyze data: Y ijk = µ + α i + β j + (αβ) ij +t k + Ɛ ijk, (Eq. 4) where Y ijk is the observed trait, µ is the overall mean of the observations, α i is the main effect of the dietary regime (i = 1-diet, 2-diet), β j is the effect of wk of lactation (j = 1, 2, 3, 4 or 1, 3, for plasma metabolites), (αβ) ij is the interaction between diet and wk, t k is the random effect of sows (k = 1,2,3,, 14) to account for repeated measurements within sow and Ɛ ijk is the residual random components. Plasma NEFA was subjected to a logarithmic transformation to stabilize the residual variance. Initial sow weight, BF, initial body composition, live born piglets, total born piglets, still born piglets, piglet birth weight, weaning weight, weaning weight and number of weaned piglets on 34

MANUSCRIPT TO BE SUBMITTED 212 213 214 215 216 217 218 219 220 d 28 were analyzed using a model where the effect of wk and random effect of sows were omitted, resulting in the following model (notation as described above): Y i = µ + α i + Ɛ i. (Eq. 5) The statistical analysis was performed using the MIXED procedure of SAS (version 9.4), except for still born piglets and piglet mortality where odds ratios were analyzed using GENMOD procedure of SAS (version 9.4) with the model shown in Eq. 5. Data were assumed to be binomially distributed and observations were regarded as dependent observations. All variables were considered significant when P < 0.05 and tendencies were accepted at P 0.1. Mean values are represented as least square mean ± largest standard error of mean (SEM). 221 222 RESULTS 223 Experimental Dietary Regime 224 225 226 227 228 229 230 231 232 The chemical compositions of control diet, basal diet and lactation supplement are listed in Table 2. The control diet and basal diet had almost the same GE and ME content. Lactation supplement is provided with more GE and CP than the control and basal diet. The Lys content varies among the three supplements/diets and was highest for lactation supplement, intermediary for control diet and lowest for basal supplement. The basal diet was formulated to contain 4 g SID Lys/kg, analyzed Lys content was 4.82 g SID Lys/kg. Lactation supplement was formulated to contain 10 g SID Lys/kg, the analyzed content of Lys was 8.60 g SID Lys/kg. The dietary content of N was higher for lactation supplement than for basal diet, and lowest for control diet. 233 234 Sow Performance Characteristics 235 35

MANUSCRIPT TO BE SUBMITTED 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 In total 14 sows were included in the experiment. One sow was excluded from the experiment because half of the mammary glands were not functional. Initial BW, body protein and body fat were similar for the dietary regimes (Table 3). There was a difference in initial BF, 1- diet sows had more BF than 2-diet sows (P = 0.01). No differences in live born, total born, still born, piglet birth weight, weaned piglets or mortality until d 28 were registered. No piglets died from d 2 (after litter equalization) until d 5. At weaning piglets weighed 7.31 and 7.96 kg (P = 0.23) in the dietary regimes receiving 1 diet and 2 diets, respectively. Feed intake was not affected by dietary regime (P = 0.88) and the average feed intake for the 4 wk lactation period was 6.92 kg/d and 6.88 kg/d for sows fed the 1-diet or 2-diet regime, respectively. Feed intake increased, however, from wk 1 to 4 (P < 0.001), and there was an interaction between dietary regime and wk (P < 0.001). Sows weight loss was not affected by dietary regime or lactation wk, but there was an interaction between diet regime and wk (P < 0.01). Sows MY increased throughout lactation (P < 0.001), and an interaction between diet and wk was observed (P = 0.03). In wk 4 sows fed the 2-diet feeding regime produced more milk than the sows fed the 1-diet regime. Milk content of DM, fat, protein and energy was not affected by dietary regime but contents were affected by wk (P = 0.01, 0.02, < 0.001 and 0.01, respectively). Milk lactose content was higher for sows fed the 1-diet feeding regime (P = 0.01) and increased as lactation progressed (P < 0.001). Loss of BF, protein and fat did not differ between treatments, but there was a clear interaction between wk and diet regime (P = 0.03, < 0.001 and 0.001, respectively). Piglet weight gain, mean piglet weight and litter size were not affected by dietary regime. Piglet weight gain increased from wk 1 to wk 3 and decreased in wk 4 and litter size decreased as lactation progressed. There was an interaction between dietary regime and wk for mean piglet weight, where piglets in the 2-diet feeding regime gained more weight. 259 36

MANUSCRIPT TO BE SUBMITTED 260 Nutrient Balances 261 262 263 264 265 266 267 268 Energy, N and Lys intake was different between diet regimes (P < 0.001) and increased as lactation progressed (P < 0.001). N loss was higher for sows in the 2-diet group which also had a positive N and Lys balance throughout lactation. Furthermore, there was an interaction between diet regime and wk for N intake (P = 0.06) and N loss in urine (P = 0.01). Sows in both dietary regimes were in negative energy balance throughout lactation. For sows fed the 1-diet regime energy balance increased from wk 1 to wk 4,for sows fed the 2-diet feeding regime energy balance was constant throughout lactation. 269 270 Plasma Metabolites 271 272 273 274 275 276 277 278 279 280 Plasma concentration of insulin was not affected by diet regime and lactation week. There was a tendency for a higher triglycerides (TAG) concentration in plasma for sows in the 1-diet group (P = 0.10) and a tendency for a higher NEFA concentration in the first wk of lactation (P = 0.06). The plasma concentration of glucose tended to be higher for sows in the 1-diet group (P = 0.07) and tended (P = 0.10) to increase from early lactation to mid lactation. There was interaction between diet regime and wk (P = 0.04) on plasma lactate concentration and an effect of wk (P = 0.02). Sows in the 1-diet feeding regime had a lower plasma concentration of lactate than sows in the 2-diet group in wk 1. Sows in the 1-diet group had a lower plasma concentration of urea in wk 1 and 3 (P < 0.01), and urea increased from wk 1 to 3 of lactation (P < 0.01; Table 6). 281 282 283 DISCUSSION Sows MY, Piglets Weight Gain and Feed Intake 37

MANUSCRIPT TO BE SUBMITTED 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 Lactating sows are today more than ever before challenged. Sows are nursing larger litters, which concomitantly increases MY (Noblet and Etienne, 1989; Vinther, 2014; Vadmand et al., Submitted). Litter size has been improved by selection, and Danish sows today give birth to on average 17.1 piglets per litter, which is approximately 4 piglets more than in 1992 (Pedersen et al., 2010; Vinther, 2014). The MY estimated by Hansen et al. (2012b) based on 21 studies from 1983-2012 also confirms that MY has increased from 9.23 kg at peak lactation to 15.6 kg at peak lactation (observed in the current study for sows fed the 2-diet feeding regime). At the same time the feeding strategy for lactating sows has not changed from the ad libitum supply with a 1-diet feeding strategy (Neil et al., 1996; Cools et al., 2014). Sows requirement for nutrients has therefore never been greater. In the present study an interaction between dietary feeding regime and week, as regards the effect on MY, was found. The interaction was mainly driven by the last wk of lactation when sows fed the 1-diet feeding regime had a lower MY than the 2-diet sows. In wk 4 the 2-diet sows not only produced more milk but did also receive less feed, 8.79 kg and 7.68 kg for sows fed the 1- diet and 2-diet feeding regime, respectively. Piglet weight gain in the present study was numerically higher throughout the lactation period for sows on the 2-diet feeding regime, and at weaning the 2-diet piglets were numerically larger (7.96 kg) than piglets from the 1-diet feeding regime (7.31 kg). The numerically higher weight gain for the 2-diet piglets may also indicate that MY was higher for sows fed the 2-diet regime during the entire lactation period, although the MY assessed by the model developed by Hansen et al. (2012b) estimated the same MY for the 2-diet regime. A higher MY for 2-diet sows is also supported by the lower plasma concentration of glucose for sows fed 2 diets. This might indicate that sows fed the 2-diet feeding regime had a great pull of nutrients by the udder, which probably decreases the glucose concentration in plasma. 38

MANUSCRIPT TO BE SUBMITTED 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 The average MY in the present study is 12.5 and 13.3 kg/d for sows fed the 1-diet and 2-diet feeding regime, respectively. A more recent study conducted at the same location also confirms an improved MY for sows fed two diets during lactation (Flummer et al., 2014). Several studies, however, imply that sows MY can hardly be changed through feeding, and sows are able to maintain MY even though the feed intake is lowered considerably, whereby sows respond by mobilizing body depots to sustain MY (King and Dunkin, 1986; Noblet and Etienne, 1986). In the same context, studies have been conducted to feed sows ad libitum, to minimize the weight loss and, in turn, improve sows reproduction. In spite of these attempts, sows still lose weight throughout lactation (Neil et al., 1996; Cools et al., 2014). These findings suggest that sows either are not fed a well balanced diets or sufficient amounts of nutrients. This is further confirmed by the studies of Feyera and Theil (2014) and Flummer et al. (2014). Based on MY for sows fed four different feeding strategies, Flummer et al. (2014) found the optimal ratio of Lys to energy to be 0.53 in feed. However, factorial calculations revealed that the ratio between required Lys and energy increases from 0.38 to 0.57 and emphasized that a dynamic 2-diet feeding regime should be superior to a single-diet regime (Feyera and Theil, 2014). 323 324 Feed Intake and BW Loss 325 326 327 328 329 330 331 Sows feed consumption was not affected by dietary regime, but was different when compared among weeks and dietary regime. In the 1-diet regime, sows were supplied with less feed in early lactation and more in late lactation compared with 2-diet sows which were fed more in accordance with their energy requirement. As a consequence sows mainly lost weight in wk 1 on the 1-diet regime but did not lose weight during wk 3 and 4 of lactation. For sows fed the 2-diet regime the main part of the weight loss happened from wk 2 to 4 of lactation, whereas no weight loss was 39

MANUSCRIPT TO BE SUBMITTED 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 observed in early lactation when the initial supply was 70% of the required energy. It was, however, the intention to minimize weight loss or prevent sows from losing weight using the 2-diet regime. The weight loss could be ascribed to the calculations of feed requirement, which is relying on the model developed by Hansen et al. (2012b). The factorial approach by Feyera and Theil (2014) estimated that 69, 70 and 94% of energy, N and Lys requirement, respectively, are associated with milk production. Estimation of MY is therefore a central player for calculating nutrient requirement and nutrient balances, and the accuracy and precision of MY estimation are therefore crucial. The model made by Hansen et al. (2012b) is relatively insensitive to inputs (litter size and litter weight gain) in early lactation and predicted MY was therefore almost identical in early lactation. The benefit of feeding sows the 2-diet feeding regime would probably become more evident if the experiment was carried out with sows of different parities, and especially the weight loss would expectedly be greater for older sows on the 1-diet regime than observed in the present study. Firstly, because sows in different parities produce different amounts of milk (Beyer et al., 2007), and secondly, sows requirement for maintenance is higher for older sows because of a higher BW (Theil et al., 2004; NRC, 2012; Theil, 2015). Thus, the inclusion of only second parity sows in the present study highly likely favored the 1-diet regime. 348 349 Nutrient Balances and Mobilization of Body Protein and Fat 350 351 352 353 354 355 Sows in both dietary regimes were in negative energy balance throughout lactation. The energy balance of sows fed 1 diet decreased throughout lactation and nearly reached a zero energy balance in late lactation. The energy balance for sows in the 2-diet feeding regime was rather constant, but negative, throughout lactation. Sows feed intake was supplied to cover the requirement for energy and the intention was therefore to reach a zero energy balance in the 2-diet regime. The 40

MANUSCRIPT TO BE SUBMITTED 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 energy balance was, however, planned to be negative in lactation wk 1 because the sows were initially fed 70% of the requirement increasing to 100% in wk 3 and 4 when the sows should have obtained a zero energy balance. The 70% of requirement was found by Flummer et al. (2014) to have a positive effect on MY in the first wk of lactation, after which sows produced more milk when fed 100% of requirement. This is mainly suggested to be related to the high supply of N and Lys relative to energy from the uterus in early lactation, which will cause an unfavorable balance between energy and N. The unfavorable balance between N and energy could possibly cause the reduction in voluntary feed intake observed for sows fed ad libitum in early lactation. The reason for not obtaining the zero energy balance could be ascribed to the theoretical calculation of sows requirement for energy. Feed intake is to a great extent calculated based on the predicted MY, and sows will be fed insufficiently if the forecasted MY turns out to be lower than the actual production. Measuring energy and other nutrient balances based on the factorial approach of Feyera and Theil (2014) could have been supported by respiration and metabolic trials to estimate sows energy and N metabolism as done by Theil et al. (2004), but this is a comprehensive approach. Excessive oxidation of N also contributes to a negative energy balance since energy in protein is lost as urea (Chwalibog et al., 1992), to be discussed below. Sows weight loss and body fat loss support the calculated energy balances, and sows start mobilizing from body depots when the energy intake is insufficient to support the requirement for maintenance and milk production. In line with that, plasma concentration of NEFA also indicates that sows fed the 1-diet feeding regime mobilized more energy than sows fed the 2-diet feeding regime during lactation wk 1. On the other hand, sows fed the 2-diet regime had a more stable plasma concentration of NEFA in lactation wk 1 and 3, which is in line with the constant energy balance for these sows. Theil et al. (2013) and Hansen et al. (2012a) observed a plasma NEFA concentration of 213 and 242 µeq/l on lactation d 1, which is comparable with the 239 µeq/l for 41

MANUSCRIPT TO BE SUBMITTED 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 sows in the present study fed the 1-diet feeding regime (lactation d 3). Plasma NEFA for sows in the 1-diet regime decreased as mobilization decreased and seemed to reach a level in lactation wk 3 (92 µeq/l) comparable with Theil et al. (2013). Sows on the 2-diet feeding regime, however, had a constant low plasma concentration of NEFA (98-86 µeq/l ), which is comparable with the level observed by Theil et al. (2013) from lactation d 10 to 28 (119-90 µeq/l). Sows N and Lys intake was rather different in the two feeding regimes. Sows fed the 1-diet regime were negative in N and Lys and reached a positive or zero balance in late lactation. For the 2-diet feeding regime sows N and Lys balance was positive throughout lactation. The basal diet and lactation supplements in the 2-diet feeding regime were formulated to cover the requirement for energy, and Lys was formulated based on optimally calculated values of Lys to energy ratio at peak lactation. The N to energy ratio was determined by the protein included in the diets to obtain the optimal Lys supply based on factorial calculations. Thus, no attempt was done to optimize N supply. As a consequence the supply of N was compromised because Lys was supplied by increasing the dietary protein content and not by inclusion of crystalline Lys. Despite the higher intake of Lys and N for sows fed the 2-diet feeding regime, sows still mobilized from muscle depots. This is in agreement with the findings of Flummer et al. (2014), who found that sows in negative energy balance would mobilize both fat and muscle depots no matter whether the N balances were positive or negative as long as sows were in negative energy balance. Loss of body protein and fat was, however, lower for both feeding regimes than those observed by Flummer et al. (2014). The estimated loss of protein and fat was calculated from Rozeboom et al. (1994) prediction equation based on sow LW and BF and could have been slightly more accurate by using the deuterium oxide (D 2 O) dilution technique. 402 403 Intermediary Metabolism of Plasma Metabolites 42

MANUSCRIPT TO BE SUBMITTED 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 Plasma urea concentration was substantially higher for sows fed the 2-diet regime mainly because sows were provided with N above the requirement. The excess amount of N was probably used partly as fuel, whereas a part of the carbon skeleton was probably converted into lactate by the liver. The lactate could then be exported to the udder where it may be oxidized and used for lactose synthesis, or used for de novo fat synthesis. Thus, excessive supply of dietary N may both explain the elevated plasma urea (3.94 and 5.61 mm for wk 1 and 3, respectively) and the elevated plasma lactate (2.04 and 1.62 mm for wk 1 and 3, respectively) observed in sows fed the 2-diet feeding regime as compared to the plasma levels obtained for 1-diet sows. The plasma urea concentration of sows supplied the 1-diet feeding regime seems to be comparable with other studies. Valros et al. (2003) observed a plasma urea concentration in the range of 4.6-4.8 mm on lactation d 21, also comparable with the findings of Neil (1996) who observed an average urea level of 4.5-4.9 mm throughout lactation. The plasma concentration of TAG increased in lactation wk 3 for sows fed the 1-diet feeding regime and was reduced for sows fed the 2-diet regime, compared with lactation wk 1. Mosnier et al. (2010) also found a reduced concentration of plasma TAG as lactation proceeds. The lower TAG observed for sows fed the 2-diet regime could possibly be due to the higher MY and greater uptake of nutrients by the udder in lactation wk 3. In conclusion feeding of lactating sows with the 2-diet feeding regime throughout lactation has improved sows MY and increased the mean piglet weight. Feeding of sows with the 2-diet regime should also make it possible to reduce or eliminate sows weight loss during lactation, which will further improve sow reproduction and sow longevity. Expenses for feed can probably also be reduced by feeding sows well-balanced diets throughout the reproductive cycle. 427 43

MANUSCRIPT TO BE SUBMITTED 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 LITERATURE CITED Bach Knudsen, K. E. (1997). Carbohydrate and lignin contents of plant materials used in animal feeding. Animal Feed Science and Technology 67, 319-338. Beyer, M., Jentsch, W., Kuhla, S., Wittenburg, H., Kreienbring, F., Scholze, H., Rudolph, P. E., and Metges, C. C. (2007). Effects of dietary energy intake during gestation and lactation on milk yield and composition of first, second and fourth parity sows. Archives of Animal Nutrition 61, 452-468. Chwalibog, A., Jakobsen, K., Henckel, S., and Thorbek, G. (1992). Estimation of quantitative oxidation and fat retention from carbohydrate, protein and fat in growing pigs. Journal of Animal Physiology and Animal Nutrition-Zeitschrift Fur Tierphysiologie Tierernahrung Und Futtermittelkunde 68, 123-135. Cools, A., Maes, D., Decaluwe, R., Buyse, J., van Kempen, T., Liesegang, A., and Janssens, G. P. J. (2014). Ad libitum feeding during the peripartal period affects body condition, reproduction results and metabolism of sows. Anim Reprod Sci 145, 130-140. Dourmad, J. Y., Noblet, J., and Etienne, M. (1998). Effect of protein and lysine supply on performance, nitrogen balance, and body composition changes of sows during lactation. Journal of Animal Science 76, 542-550. Feyera, T., and Theil, P. K. (2014). Nutrient balances of energy, lysine and nitrogen in late gestating and early lactating sows Proceedings (EAAP Denmark 2014). Flummer, C., Feyera, T., Krogh, U., Hinrichsen, T., Bruun, T. S., and Theil, P. K. (2014). The ratio between energy and lysine is essential to improve milk yield in sows. Proceedings (EAAP Denmark 2014). 44

MANUSCRIPT TO BE SUBMITTED 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 Hansen, A. V., Lauridsen, C., Sorensen, M. T., Knudsen, K. E. B., and Theil, P. K. (2012a). Effects of nutrient supply, plasma metabolites, and nutritional status of sows during transition on performance in the next lactation. J Anim Sci 90, 466-480. Hansen, A. V., Strathe, A. B., Kebreab, E., France, J., and Theil, P. K. (2012b). Predicting milk yield and composition in lactating sows: a Bayesian approach. J Anim Sci 90, 2285-98. Hansen, B. (1989). Determination of nitrogen as elementary-n, an alternative to kjeldahl. Acta Agri Scand 39, 113-118. Jørgensen, L., and Tybirk, P. (2010). "fnutrient recommendations (Normer for næringsstoffer)," 16/Ed. Kim, S. W., Hurley, W. L., Wu, G., and Ji, F. (2009). Ideal amino acid balance for sows during gestation and lactation. Journal of Animal Science 87, E123-E132. King, R. H., and Dunkin, A. C. (1986). The effect of nutrition on the reproductive-performance of 1st-litter sows.3. The response to graded increases in food-intake during lactation. Animal Production 42, 119-125. Lauridsen, C., and Danielsen, V. (2004). Lactational dietary fat levels and sources influence milk composition and performance of sows and their progeny. Livestock Production Science 91, 95-105. Løvendahl, P., and Purup, H. M. (2002). Technical note: Time-resolved fluoro-immunometric assay for intact insulin in livestock species. J Anim Sci 80, 191-195. Mosnier, E., Etienne, M., Ramaekers, P., and Père, M. C. (2010). The metabolic status during the peri partum period affects the voluntary feed intake and the metabolism of the lactating multiparous sow. Livest Sci 127, 127-136. Neil, M. (1996). Ad libitum lactation feeding of sows introduced immediately before, at, or after farrowing. Animal Science 63, 497-505. 45

MANUSCRIPT TO BE SUBMITTED 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 Neil, M., Ogle, B., and Anner, K. (1996). A two-diet system and ad libitum lactation feeding of the sow.1. Sow performance. Animal Science 62, 337-347. Noblet, J., and Etienne, M. (1986). Effect of energy-level in lactating sows on yield and composition of milk and nutrient balance of piglets. Journal of Animal Science 63, 1888-1896. Noblet, J., and Etienne, M. (1987). Metabolic utilization of energy and maintenance requirements in lactating sows. Journal of Animal Science 64, 774-781. Noblet, J., and Etienne, M. (1989). Estimation of sow milk nutrient output. Journal of Animal Science 67, 3352-3359. NRC (2012). "Nutrient Requirements of Swine," 11/Ed. National Academy Press, Washington, DC, USA. Oliviero, C., Kokkonen, T., Heinonen, M., Sankari, S., and Peltoniemi, O. (2009). Feeding sows with high fibre diet around farrowing and early lactation: Impact on intestinal activity, energy balance related parameters and litter performance. Research in Veterinary Science 86, 314-319. Pedersen, L. J., Berg, P., Jørgensen, E., Bonde, M. K., Herskin, M. S., Knage-Rasmussen, K. M., Kongsted, A. G., Lauridsen, C., Oksbjerg, N., Poulsen, H. D., Sorensen, D. A., Su, G., Sørensen, M. T., Theil, P. K., Thodbjerg, K., and Jensen, K. H. (2010). Pattegrisedødelighed i DK: Muligheder for reduktion af pattegrisedødelighed i Danmark. Husdyrbrug nr. 86. Rozeboom, D. W., Pettigrew, J. E., Moser, R. L., Cornelius, S. G., and Elkandelgy, S. M. (1994). In vivo estimation of body composition of mature gilts using live weight, backfat thickness, and deuterium oxide. J Anim Sci 72, 355-366. 46

MANUSCRIPT TO BE SUBMITTED 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 Smits, R. J., Henman, D. J., and King, R. H. (2013). Increasing the energy content of lactation diets fed to first-litter sows reduces weight loss and improves productivity over two parities. Animal Production Science 53, 23-29. Strathe, A. B., Jorgensen, H., Kebreab, E., and Danfaer, A. (2012). Bayesian simultaneous equation models for the analysis of energy intake and partitioning in growing pigs. Journal of Agricultural Science 150, 764-774. Theil, P. K. (2015). Transition feeding of sows. In "The gestating and lactating sow" (C. Farmer, ed.), pp. 147-172, Wageningen Academic Publishers. Theil, P. K., Jorgensen, H., and Jakobsen, K. (2004). Energy and protein metabolism in lactating sows fed two levels of dietary fat. Livest Prod Sci 89, 265-276. Theil, P. K., Olesen, A. K., Flummer, C., Sorensen, G., and Kristensen, N. B. (2013). Impact of feeding and post prandial time on plasma ketone bodies in sows during transition and lactation. Journal of Animal Science 91, 772-782. Vadmand, C. N., Krogh, U., Hansen, C. F., and Theil, P. K. (Submitted). Impact of sow and litter characteristics on colostrum yield, time for onset of lactation, and milk yield of sows. Valros, A., Rundgren, M., Spinka, M., Saloniemi, H., Rydhmer, L., Hulten, F., Uvnas-Moberg, K., Tomanek, M., Krejci, P., and Algers, B. (2003). Metabolic state of the sow, nursing behaviour and milk production. Livest Prod Sci 79, 155-167. Vinther, J. (2014). Landsgennemsnit for produktivitet i svineproduktionen 2014. Danish Pig Research Centre. Copenhagen, Denmark. Weast, R. C. (1969). "Handbook of chemistry and physics,," 49th/Ed. Zak, L. J., Cosgrove, J. R., Aherne, F. X., and Foxcroft, G. R. (1997). Pattern of feed intake and associated metabolic and endocrine changes differentially affect postweaning fertility in primiparous lactating sows. Journal of Animal Science 75, 208-216. 47

MANUSCRIPT TO BE SUBMITTED 520 521 522 523 524 525 526 Table 1 Dietary ingredients of control diet (1-diet regime), basal diet and lactation supplement (2- diet regime). 1-diet 2-diet Ingredient (%) Control Diet Basal Diet Lactation Supplement Barley 30 36.16 16.13 Wheat 43.43 36.16 16.13 Corn 30 Soybean meal 2.1 16.98 Sunflower meal 11 9.96 10.32 Rapeseed meal 5 Oat 3 Sugar beet pulp 3 Alfalfa meal 3 PFAD 1 1.6 0.59 Soyabean oil 1 5 Calcium carbonate 1.35 1.41 1.42 Wheat bran 6.67 1.49 Sodium chloride 0.45 0.42 0.51 Monocalcium phosphate 0.55 0.07 0.94 L-Lysine sulfate, 65% 0.59 0.1 0.55 Molasses 0.5 0.5 L-Threonine, 98% 0.02 0.18 0.01 Valine, 40% 0.35 0.01 Vitamin E 2 0.19 0.19 0.24 Phytase 3 0.02 0.05 0.05 Mineral and vitamin mix 4 0.2 0.2 0.2 1: Palm fatty acid distillate. 2: mg α-tocopherol. 3: 2500 U/kg 4: Supplied per kilogram of diet: retinol 2 IU, α-tocopherol 63mg, phylloquinone 4.4 mg, thiamin 2.10 mg, cyanocobalamin 0.022 mg, riboflavin 5.25 mg, pyridoxine 3.15 mg, biotin 0.44 mg, D-pantothenic acid 15.8 mg, folic acid 1.58 mg, niacin 21 mg, CuSO 4 15 mg, ZnO 90 mg, Se 0.32 mg, FeSO 4 84 mg, CaI 0.84 mg, MnO 42 mg. 48

MANUSCRIPT TO BE SUBMITTED 527 528 529 Table 2 Chemical composition and calculated ME, SID N and SID Lys for control diet (1-diet regime) and basal diet and lactation supplement (2-diet regime). Item Control Diet Basal Diet Lactation Supplement Gross energy (MJ/kg) 16.47 16.45 17.36 Chemical composition (g/kg) DM 868 872 881 CP (N*6,25) 144 147 172 Crude Fat 39 51 102 Starch 439 375 348 Dietary fiber 234 236 240 Ash 46 47 55 Lysine 7.2 5.97 9.52 Methionine 2.5 2.59 2.78 Threonine 5.96 6.51 6.65 Valine 6.72 6.96 7.74 Calculated ME (MJ/kg) 12.56 12.47 14.47 SID N (g/kg) 18.66 18.68 23.55 SID Lys (g/kg) 6.14 4.82 8.60 49

MANUSCRIPT TO BE SUBMITTED 530 531 532 533 534 Table 3 Initial parameters for sows fed the 1-diet and 2-diet regime. Item 1-diet 2-diet SEM 1 P-value No. Sows 6 7 - - Sow weight d 2 (kg) 240 231 8 0.42 Back fat (mm) 15.3 a 12.8 b 0.6 0.01 Body protein, kg 35.5 34.8 1.1 0.65 Body fat, kg 75.1 72.0 2.8 0.44 Live born piglets 18.3 19.4 1.0 0.44 Total born 20.0 20.4 1.0 0.77 Still born piglets (%) 8.3 4.9 0.51 0.26 Piglet birth weight (kg) 1.16 1.15 0.06 0.95 Weaning weight (kg) 7.31 7.96 0.37 0.23 Weaned piglets 2 13.2 13.1 0.4 0.97 Mortality until d 28 (%) 3 6 6.1 0.62 0.96 Different letters (a, b) within a row differ (P < 0.05). 1: SEM = Standard error of the mean. 2: Litter size standardized to 14 piglets on d 2. 3: From litter equalization on d 2 until weaning on d 28. 50

535 MANUSCRIPT TO BE SUBMITTED Table 4 Mean sow performance in wk 1, 2, 3, and 4 for sows fed the 1-diet and 2-diet regime. 536 537 1-diet 2-diet P-value Item Wk 1 Wk 2 Wk 3 Wk 4 Wk 1 Wk 2 Wk 3 Wk 4 SEM Diet Wk D x W Feed intake (kg/d) 4.04 e 6.65 c 8.20 ab 8.79 a 5.29 d 6.93 c 7.63 bc 7.68 b 0.24 0.88 <.001 <.001 Feces score 2.9 2.9 3.0 3.0 3.0 3.1 3.1 3.0 0.05 0.06 0.31 0.34 Sow weight change (kg) -8.8 b -2.5 ab 0.3 a -0.8 ab -1.7 ab -5.9 b -4.3 ab -3.4 ab 1.9 0.60 0.13 <.01 Milk yield, kg/d 8.9 c 13.6 b 14.6 ab 13.0 b 8.9 c 13.9 b 15.5 a 14.9 a 0.5 0.20 <.001 0.03 Milk yield/ kg feed 2.2 a 2.1 ab 1.8 b 1.5 b 1.7 b 2.0 ab 2.0 ab 2.0 ab 0.1 0.53 <.001 <.001 Milk DM (%) 18.7 ab 18.0 ab 17.6 ab 17.2 b 18.9 a 17.2 b 18.0 ab 17.2 b 0.5 0.99 0.01 0.61 Milk fat (%) 8.2 a 7.6 ab 7.0 ab 6.5 b 8.1 a 6.9 ab 7.7 ab 6.8 ab 0.5 0.82 0.02 0.43 Milk protein (%) 5.2 a 4.5 b 4.5 b 4.7 b 5.5 a 4.4 b 4.5 b 4.5 b 0.14 0.94 <.001 0.19 Milk lactose (%) 4.9 b 5.0 ab 5.2 a 5.2 a 4.8 b 5.0 ab 5.0 ab 5.1 ab 0.06 0.01 <.001 0.44 Milk energy (MJ/100 g) 0.51 a 0.48 ab 0.46 b 0.46 b 0.50 ab 0.48 ab 0.47 ab 0.48 ab 0.02 0.81 0.01 0.77 Back fat change (mm/wk) -0.7 ab -1.1 ab -0.2 a -1.5 b 0.3 0.88 <.001 0.03 Body protein change, kg -0.54 b 0.22 a -0.49 b -0.37 b 0.17 0.18 <.001 <.001 Body fat change, kg -1.72 b 0.48 a -1.36 b -1.69 b 0.5 0.27 <.001 <.001 Piglet weight gain (g) 197 b 229 ab 223 b 196 b 205 b 248 ab 281 a 231 ab 18 0.11 <.01 0.30 Mean piglet weight (kg) 1 2.10 d 3.40 c 4.99 b 6.54 a 2.04 d 3.41 c 5.30 b 7.15 a 0.23 0.44 <.001 0.04 Litter size 14.0 a 14.0 ab 13.7 ab 13.2 ab 14.0 a 13.3 ab 13.3 ab 13.1 b 0.3 0.41 <.01 0.22 Different letters (a, b, c) within a row differ (P < 0.05). 1: Average weight in wk 1, 2, 3 and 4. 51

MANUSCRIPT TO BE SUBMITTED 538 539 541 Table 5 Mean nutrient balances in wk 1, 2, 3, and 4 for sows fed the 1-diet and 2-diet feeding regime. 1-diet 2-diet P-value 540 Item Wk 1 Wk 2 Wk 3 Wk 4 Wk 1 Wk 2 Wk 3 Wk 4 SEM Diet Wk D x W Energy metabolism, MJ/d Intake 50.7 f 83.5 d 103.0 b 110.3 a 67.0 e 96.0 c 106.4 ab 111.4 a 2.28 <.001 <.001 <.01 Maintenance 28.9 28.5 28.2 28.3 28.4 28.0 27.6 27.2 0.66 0.46 <.001 0.16 Secreted in milk 45.0 c 65.6 ab 67.5 ab 59.2 b 44.8 c 65.9 a 72.1 a 71.1 a 3.1 0.24 <.001 0.02 Associated with milk production 17.5 c 25.5 ab 26.3 ab 23.0 b 17.4 c 25.6 ab 28.1 a 27.6 a 1.2 0.24 <.001 0.02 Balance -38.4 d -36.0 cd -19.1 b -0.2-21.3 b -23.6 bc -21.5 b -18.1 b 4.6 0.64 <.001 <.001 Nitrogen metabolism, g/d Intake, SID 3 75.4 f 124.1 d 153.0 e 163.8 b 105.3 e 152.8 c 169.8 ab 178.1 a 3.6 <.001 <.001 0.06 Urine nitrogen loss 26.7 e 44.0 c 54.2 b 58.0 a 36.4 d 52.2 b 57.8 ab 60.5 a 1.2 <.001 <.001 0.01 Secreted in milk 67.0 d 97.3 b 106.7 ab 106.4 ab 69.9 d 96.0 b 105.3 ab 113.4 a 4.8 0.74 <.001 0.49 Balance -7.4 bc -17.1 c -7.9 bc -0.6 ab 9.9 a 4.6 ab 6.7 ab 0.4 ab 5.1 <.01 0.30 0.09 Lys utilization, g/d Intake, SID 3 24.8 e 40.8 c 50.3 b 53.9 b 33.8 d 51.5 b 57.9 a 61.1 a 1.2 <.001 <.001 0.46 Maintenance 1.4 f 2.2 d 2.7 ab 2.8 a 1.7 e 2.3 d 2.5 c 2.6 bc 0.05 0.61 <.001 <.001 Secreted in milk 29.9 c 43.5 b 47.7 ab 47.6 ab 31.3 c 42.9 b 47.1 ab 52.2 a 2 0.58 <.001 0.21 Balance -2.7 cd -4.9 d 0.0 bc 3.5 abc 4.8 ab 6.3 ab 8.3 a 5.9 ab 2.3 <.01 0.09 0.16 Different letters (a, b, c, d, e, f) within a row differ (P < 0.05). 542 52

MANUSCRIPT TO BE SUBMITTED Table 6 Mean concentrations of plasma metabolites in wk 1 and 3 for sows fed the 1-diet and 2- diet feeding regime. 1-diet 2-diet P-value Metabolite Wk 1 Wk 3 Wk 1 Wk 3 SEM Diet Wk D x W Insulin (pmol/l) 85.4 104.2 96.6 62.7 21.70 0.51 0.66 0.14 Glucose (mm) 5.93 5.96 5.16 5.81 0.21 0.07 0.10 0.12 Lactate (mm) 1.34 1.31 2.04 1.62 0.30 0.23 0.02 0.04 NEFA 1 (µeq/l) 239 92 98 86 0.15 0.06 0.15 [135;422] [52;162] [58;166] [49;152] TAG 2 (mm) 0.19 0.21 0.19 0.15 0.02 0.10 0.65 0.16 Urea (mm) 2.84 d 4.77 b 3.94 c 5.61 a 0.24 <.01 <.001 0.55 Measured 4 h postfeeding on lactation d 3 and 17. Different letters (a, b, c, d) within a row differ (P < 0.05). 1: Plasma NEFA was log-transformed before statistical analysis, and hence confidence limits are given in brackets instead of SEM values. 2: Triglycerides. 53

Section IV

ADDITIONAL RESULTS AND GENERAL DISCUSSION Accuracy of sows MY The factorial approach by Feyera (2014a) estimated that 69, 70 and 94% of energy, N and Lys requirement, respectively, are associated with milk production. Estimating MY reliably is therefore central for calculations of nutrient requirements and nutrient balances and a high accuracy and high precision when estimating MY is crucial. The model developed by Hansen et al. (2012b) was used for estimation of MY in the current study. The model estimates the lactation curve of sows by using litter size and average daily gain of the litter as inputs. The influence of litter size and litter gain on MY is illustrated in Figure 3 in literature review. Due to the mathematical model sows MY is relatively insensitive to inputs during the first 2-3 days of lactation for litter size and insensitive during the first 6-7 days of lactation for litter gain, because very few data on MY were available when the model was developed. Thus, the model is regarded reliable for predicting MY of sows, although the precision is probably compromised during the first days of lactation. The model is, however, still a reliable measure for prediction of MY. The feed requirement for milk production is forecasted one week ahead based on predicted MY on an individual basis. This forecast of MY is therefore crucial when calculating sows requirements for nutrients and in this study for calculating feed supply. The measured MY and forecasted MY for sows fed the 1-diet and 2-diet regime are illustrated in Figure 6. Forecasted MY is lower than measured MY in week 2 of lactation and above in week 4 for both dietary regimes. Forecasted MY is, however, more comparable with measured MY for sows fed the 2-diet feeding regime. 54

ADDITIONAL RESULTS AND GENERAL DISCUSSION Milk yield (kg) 15 13 11 9 7 Milk yield (kg) 15 13 11 9 7 5 5 2 4 6 8 10 12 14 16 18 20 22 24 26 28 2 4 6 8 10 12 14 16 18 20 22 24 26 28 Day Day Figure 6 - Left figure: Measured average daily MY ( ) and forecasted MY (- - -) for sows fed the 1-diet feeding regime. Right figure: Measured average daily MY ( ) and forecasted MY (- - -) for sows fed the 2-diet feeding regime. Milk production and piglet weight gain Lactating sows are today more than ever before challenged. Sows are nursing larger litters, which concomitantly increases the MY (Noblet and Etienne, 1989; Vinther, 2014; Vadmand et al., Submitted). Litter size has been improved by selection, and Danish sows today give birth to on average 17.1 piglets per litter, which is approximately 4 piglets more than in 1992 (Pedersen et al., 2010; Vinther, 2014). The MY estimated by Hansen et al. (2012b), based on 21 studies from 1983-2012, also confirms that the MY has increased from 9.23 kg at peak lactation to 15.6 kg at peak lactation (observed in this study in sows fed the 2-diet feeding regime). At the same time the feeding strategy for lactating sows has not changed from the ad libitum supply with a 1-diet feeding strategy (Neil et al., 1996; Cools et al., 2014). Sows requirement for nutrients has therefore never been greater. Several feeding studies suggest that sows MY can hardly be changed through feeding, and sows are able to maintain MY even though the feed intake is lowered significantly (King and Dunkin, 1986; Noblet and Etienne, 1986). There was an interaction between dietary regime and week on sows MY in the present study. The interaction between week and feeding regime was mainly driven by the last half of the lactation period when sows fed the 2-diet feeding regime produced more milk. Furthermore, sows in the 1- diet feeding regime reached peak lactation before the sows in the 2-diet regime and produced approximately 1-2 kg milk less from day 18 until weaning on day 28 (see Figure 7). 55

ADDITIONAL RESULTS AND GENERAL DISCUSSION Litter weight gain, or the mean piglet weight, is indirectly an indicator of MY. Piglet weight gain in the present study was numerically higher and the mean piglet weight was higher throughout the lactation period for the 2-diet feeding regime, which suggests that these sows produced more milk throughout lactation, and not only during the last two weeks as suggested by the MY obtained by the prediction model. Consequently, the higher MY for sows fed the 2-diet regime will lead to the numerically higher weaning weight observed for piglets (7.96 kg) compared to piglets from the 1- diet feeding regime (7.31 kg). With an average piglet weight gain of 196-231 g per day in lactation week four, piglets in the 2-diet regime are 3-4 days ahead compared with piglets from the 1-diet regime. 18 16 14 Milk yield (kg) 12 10 8 6 4 2 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 Day Figure 7 - Average MY for sows fed the 1-diet feeding regime ( ) and the 2-diet feeding regime (- - -). Plasma concentration of glucose was numerically lower in weeks 1 and 3 of lactation when sows were fed the 2-diet regime, most prominently in week 1. This also suggests that sows fed the 2-diet regime had a higher MY due to a great pull of nutrients by the udder. Indeed, if MY is greater, more glucose is then allocated to the udder and glucose concentration in plasma decreases. Plasma concentrations of glucose in the current study are generally lower in the first lactation week for sows fed the 2-diet regime (5.16 mm) and above for sows fed the 1-diet regime (5.93 mm) compared with studies of Hansen et al. (2012a) (5.31 mm on day 1) and Theil et al. (2013) (5.47 mm on day 1 and 5.37 mm on day 17). 56

ADDITIONAL RESULTS AND GENERAL DISCUSSION Average MY of sows fed the 2-diet feeding regime, in the present study, was higher than that observed by Feyera (2014b), whose study was conducted at the same facilities and with sows obtained from the same conventional herd. The study of Feyera (2014b) aimed to feed sows according to energy and LYS requirements (HEHL - high energy and high Lys), 70% of energy requirement (LEHL - low energy and high Lys), 70% of Lys requirement (HELL high energy and low Lys), 70% of LYS requirement and 70% of energy and LYS requirement (LELL low energy and low Lys). The sows receiving the LELL treatment obtained the highest average daily MY of 13.2 kg of all dietary strategies and the HEHL sows fed according to requirement produced slightly less (12.6) kg milk. However, LELL sows lost 37 kg whereas HEHL sows lost 12 kg. In the current study, a combination of the best strategies was applied in order to maximize MY and avoid excessive sow weight loss. It appeared that the 2-diet feeding regime resulted in a higher MY (13.3) and a weight loss comparable to that found for HEHL treated sows (Feyera, 2014b). In general, MY obtained in this study was high compared to another study by Krogh (Unpublished) carried out at the same location; MY was furthermore calculated under the same method as in the current study. Krogh (Unpublished) found an average daily MY of 10.8-11.7 kg, which on average was around 2.3-1.4 kg milk less per day compared with the current study for sows on the 2-diet feeding regime. The lactose content was highest when sows were fed the 1-diet regime. The lactose content in milk is, however, usually rather constant and should not be altered by feeding (Hurley, 2015). Lactose is the component which is responsible for drawing water into the secretory vesicles and is the key element for controlling MY (Theil et al., 2012). No clear explanation for the higher lactose content in milk for sows fed the 1-diet is evident. The variation in milk lactose content is, however, also rather negligible and is of much less importance than the change in MY in terms of daily secreted nutrients. Feed intake, changing requirement and BW loss Several studies have been conducted, partly to improve reproduction and partly to avoid that sows lose weight during lactation by increasing energy supply, either by provision of feed ad libitum (King and Dunkin, 1986; Neil et al., 1996; Beyer et al., 2007; Cools et al., 2014) or by increasing the energy content in the diet (Smits et al., 2013). Despite the numerous attempts to feed sows ad libitum, sows still lose weight throughout lactation. It might therefore indicate that the feeding strategy does not provide sows the right amount of nutrients or well-balanced ratio of nutrients. Feyera (2014a) supports the hypothesis that the current feeding strategy for lactating sows does not opti- 57

ADDITIONAL RESULTS AND GENERAL DISCUSSION mally provide sows with required nutrients. Feyera (2014b) also found that it is not possible for the sow to be in positive energy balance and metabolize from muscle tissue at the same time and that the ratio between metabolized fat and protein is rather stable. Furthermore, Flummer et al. (2014) found that the ratio between Lys and energy is essential to maximized MY and the ratio in feed should be 0.53 SID g Lys per MJ ME, if fed a single diet. Providing sows with additional energy or protein/aa is therefore not beneficial for the sow if the supplied amount of nutrients is not balanced correctly. Lys is the first limiting AA and the supply with dietary N was therefore in excess because Lys content in the diets was obtained by increasing the protein source until Lys supply was optimal. It will be more accurate to consider Lys and N requirement separately. No attempts were made to optimize N supply when formulating the diets. Instead, the N content was determined by the optimal ratio between Lys and energy in the basal diet and lactation supplement. The required ratio of Lys to energy throughout the lactation period is outlined in Figure 8, and it is evident that the ratio is not constant during lactation. The current study aimed to evaluate the effect of providing sows a more dynamic amount of Lys and energy instead of the fixed ratio, which is the case when using a single diet. The dotted line in Figure 8 illustrates the fixed Lys to energy ratio for sows in the 1-diet regime, which received a standard lactation diet. The stippled line shows the ratio of Lys to energy obtained in the 2-diet regime, which is rather comparable with the required ratio (solid line). The required ratio of Lys to energy increases from 0.38 on day 2 to 0.57 on day 28 of lactation corresponding to 4.8 and 7.2 g SID Lys/kg (12.56 MJ/kg) on days 2 and 28, respectively. The required ratio of Lys to energy in Figure 8 also reveals that the ratio is rather constant from day 12 of lactation. It was not possible in the current study to completely reach the required ratio of Lys to energy using the 2-diet approach. The main reason was a lower Lys content than intended in the lactation supplement and a higher Lys content than intended in the basal diet. Consequently, in the first week of lactation sows were fed more Lys relative to energy required and less Lys in subsequent weeks. 58

ADDITIONAL RESULTS AND GENERAL DISCUSSION 0.60 0.55 Lys:Energy ratio 0.50 0.45 0.40 0.35 0.30 2 4 6 8 10 12 14 16 18 20 22 24 26 28 Day Figure 8 - Theoretically required Lys to ME ratio for sows during lactation ( ) and the obtained Lys to energy ratio for the 1-diet feeding regime (... ) and 2-diet feeding regime (- - -). Sows total consumption of feed was not affected by dietary regime, the feed intake was, however, different when compared among weeks. In the 1-diet regime sows were supplied with less feed in early lactation and over supplied in late lactation. Sows in the 2-diet regime were provided with feed according to their requirement and feed intake was higher in early lactation and lower in late lactation compared to the 1-diet regime. As a consequence sows lost weight in the 1-diet regime mainly in the first week of lactation and reached a zero weight loss during the third and fourth weeks of lactation. For sows on the 2-diet regime the main part of the weight loss happened during weeks 2 and 4 of lactation. It was, however, the intention to minimize the weight loss or prevent sows from losing weight. The benefit of feeding sows the 2-diet feeding regime would probably become more evident if the experiment was carried out with sows in different parities. Firstly, because sows in different parities produce different amounts of milk (Etienne et al., 1998; King, 2000; Beyer et al., 2007; Christensen and Sørensen, 2013). In general, MY is lower for primiparous sows than for multiparous sows and the maximum MY is reached between the second and fourth parities. For primiparous sows Beyer et al. (2007) reported a 14% lower MY than for second parity sows and an increase of 10% from second to fourth parities. And secondly, sows requirement for maintenance is higher for older sows because of a higher body weight (Theil et al., 2004; NRC, 2012; Theil, 2015). Thus, the selection of second parity sows in the present study favored the 1-diet regime. 59

ADDITIONAL RESULTS AND GENERAL DISCUSSION Nutrient balances and mobilization of body protein and fat Sows in both dietary regimes were in negative energy balance throughout lactation. The energy balance of sows fed the 1-diet decreased throughout lactation and nearly reached a zero energy balance in late lactation. The energy balance for sows in the 2-diet feeding regime was rather constant, but negative, throughout lactation. Sows feed intake was supplied to cover the requirement for energy and the intention was therefore to reach a zero energy balance in the 2-diet regime. The energy balance was planned to be negative in the first week of lactation because the sows were initially fed 70% of the requirement increasing to 100% in weeks 3 and 4 when the sows should have been at zero energy balance. The reason for not obtaining the zero energy balance could be ascribed to the theoretical calculation of sows requirement for energy. Feed intake is to a great extent calculated based on the predicted MY, and sows will be fed insufficiently if the forecasted MY is less than the actual prediction. Measuring energy and other nutrient balances based on the factorial approach of Feyera (2014a) could have been supported by respiration and metabolic trials to estimate sows energy and N metabolism as done by Theil et al. (2004), but this is a comprehensive approach. Sows weight loss and body fat loss support the calculated energy balances, and it becomes evident that sows start mobilizing from body depots when the energy intake is insufficient to support the requirement for maintenance and milk production. Plasma concentration of NEFA also indicates that sows fed the 1-diet feeding regime which lost 8 kg in the first week of lactation mobilized more energy from depots than sows fed the 2-diet feeding regime (1.7 kg in the first week of lactation). On the other hand, sows fed the 2-diet regime had a more stable plasma concentration of NEFA in the first and third lactation weeks, which is associated with the stable energy balance and weight loss. Theil et al. (2013) and Hansen et al. (2012a) observed a plasma NEFA concentration of 213 and 242 µeq/l on the first day of lactation, which is comparable with the 239 µeq/l for sows in the present study fed the 1-diet feeding regime (day 3 of lactation). Plasma NEFA for sows in the 1- diet regime decreased as mobilization decreased and seemed to reach a level in the third week of lactation (92 µeq/l) comparable with that of Theil et al. (2013). Sows on the 2-diet feeding regime obtained a constant plasma concentration of NEFA (98-86 µeq/l), which is comparable with the level observed by Theil et al. (2013) from lactation days 10 to 28 (119-90 µeq/l) and suggests that sows weight loss was constant throughout lactation. Sows N and Lys intake is relatively different in the two feeding regimes. Sows fed the 1-diet regime were negative in N and Lys and reached a positive or zero balance in late lactation. For the 2-60

ADDITIONAL RESULTS AND GENERAL DISCUSSION diet feeding regime sows N and Lys balance was positive throughout lactation. The basal diet and lactation supplements in the 2-diet feeding regime were formulated to cover the requirement for energy, and Lys and N were formulated based on optimal calculated values of Lys to energy ratio at peak lactation. The N to energy ratio was determined by the protein included to obtain the calculated required Lys content. Thus, no attempt to optimize N supply was made. As a consequence the requirement for N was compromised because Lys was supplied by increasing the dietary protein content and not by including crystalline Lys. Despite the higher intake of Lys and N for sows fed the 2-diet feeding regime, sows still mobilized from muscle depots. This is in agreement with the findings of Feyera (2014b) and Flummer et al. (2014), who found that sows in negative energy balance would mobilize both fat and muscle depots no matter whether the N balances were positive or negative. Loss of body protein and fat was, however, lower for both feeding regimes than those observed by Feyera (2014b). The estimated loss of protein and fat was calculated from Rozeboom et al. (1994) prediction equation based on sow LW and BF and could have been slightly more accurate by using the deuterium oxide (D 2 O) dilution technique. In the present study the D 2 0 dilution technique was carried out on day 2 and 28 of lactation, however, it has not yet been analyzed. The provision of sufficient amounts of nutrients to establish a zero nutrient balance is illustrated in Figure 9. It appears that sows fed the 1-diet feeding regime, reflecting today s feeding strategy, should be fed additionally 3 kg feed each day to cover energy requirement in the first and second weeks of lactation. The additional requirement for N and Lys is, however, not as high as the additional requirement for energy indicating that a standard sow diet is not optimal for sows in early lactation. The intention of the current study was to separate the daily supply of Lys (and N) from the daily supply of energy and thereby break with today s feeding strategy with the fixed ratio of Lys to energy. The true credibility of the dynamic feeding strategy must, however, become clearer when the supply of Lys is uncoupled from the N supply. By providing sows with the required amounts of nutrients, the 2-diet feeding regime was able to straighten the curves so they appear more parallel. The parallel relationship among nutrients for the 2-diet feeding regime makes it possible to adjust according to feed supplied. For sows fed the 1-diet feeding regime it would not be enough to increase the feed intake to fulfill the nutrient requirement. For the 2-diet feeding regime sows could need additionally 1.5 kg to reach a zero energy balance throughout lactation. 61

ADDITIONAL RESULTS AND GENERAL DISCUSSION 4 4 Additional feed required to reach zero nutrient balances kg feed/d 3 2 1 0-1 -2 1 2 3 4 Additional feed required to reach zero nutrient balances kg feed/d -1 Week -2 3 2 1 0 1 2 3 4 Week Figure 9 - Left figure: Additional feed required to obtain zero ME ( ), N (- - -) and Lys (... ) balances for sows fed the 1-diet feeding regime. Right figure: Additional feed required to obtain a zero ME ( ), N (- - -) and Lys (... ) balance for sows fed the 2- diet feeding regime. The current feeding strategy used so far in Denmark, which is presented as the 1-diet feeding regime in the present study, supplies sows insufficiently with feed in early lactation and above the requirement at peak lactation. The intention for feeding lactating sows in early lactation is to slowly increase the feed intake and reach a feed supply of 5.5-6.0 kg per day around days 6 or 7 of lactation. The main reason for this is to prevent the sows from an unintended reduction in voluntary feed intake, which is a prominent problem in practice. Feeding sows to requirement with a 1-diet feeding regime is therefore not possible in early lactation. This is probably caused by an unbalance between energy and N due to the relatively high supply of N and low supply of energy by the uterus. The supply of feed reaches a maximum supply around days 16-20 when sows are fed around 9 kg or more per day. Sows requirement for energy at peak lactation is, however, reached by supplying sows 7.5 kg/day (assuming: 230 kg sow, MY of 13.7 at peak lactation, milk energy content of 4.8 MJ/kg and energy content in feed of 12.56). The reason for feed intake above the requirement is, however, to cover the requirement for AA and energy at the same time. If the ratio between Lys and energy is not right, then the supply of feed will provide the sow with either energy above the requirement or Lys below the requirement. Sows are also fed additional amounts of feed to prepare for reestablishment of a new reproductive cycle. Excessive amount of dietary N In the current study sows receiving the 2-diet feeding regime were fed excessive amounts of N because increased dietary Lys was obtained by increasing CP. Excessive N intake is costly for the sow 62

ADDITIONAL RESULTS AND GENERAL DISCUSSION because energy is lost via urea (Chwalibog et al., 1992). The breakdown (deamination) of proteins into urea is connected to a heat loss, which will increase as dietary protein exceeds the requirement (Theil et al., 2002a). It is therefore reasonable to think that sows receiving the 2-diet feeding regime are challenged in the present study by the excessive supply of N. The energy available for sows fed excessive amounts of N/ protein was also reduced since a considerable amount of energy is wasted in the urine (23% of energy in protein is lost if AAs are oxidized). Thus, instead of contributing with 23.9 kj/g protein to the energy pool (Chwalibog, 2000), the excess protein will be converted into carbon and urea, and after oxidation the protein contributes with 18.42 kj/g. Additionally, more energy is lost for further usage of the carbon skeleton. Alternatively, the increased CP content compromised the starch content and the sows fed the 2-diet feeding regime received more than 1 kg starch less per day in lactation week 4 compared with the sows in the 1-diet feeding regime. If the N supply in the 2-diet feeding regime had been reduced to the requirement, it would have improved the feed efficiency and likely also further improved the MY. If this should be done, it involves an uncoupling between the dietary supply of Lys and N as lactation progresses, and it can only be done using a 2-diet feeding regime. Intermediary metabolism of lactate, TAG and urea Plasma urea concentration was substantially higher for sows fed the 2-diet regime because sows were provided with N and Lys above the requirement. Urea concentration in plasma can both increase due to muscle breakdown to cover the requirements for proteins and individual AA and due to excessive amounts of proteins/aa being supplied in the feed. In the current study the excessive intake of N most likely increased plasma urea concentration. The excess amount of N was probably used as fuel and hence converted into lactate in the liver. The lactate could then be exported to the udder where it may be oxidized and used for lactose synthesis, or used for de novo fat synthesis. Thus, excessive supply of dietary N may both explain the elevated plasma urea (3.94 and 5.61 mm for weeks 1 and 3, respectively) and the elevated plasma lactate (2.04 and 1.62 mm for weeks 1 and 3, respectively) observed in sows fed the 2-diet feeding regime. Plasma urea concentration of sows supplied the 1-diet feeding regime seems to be comparable with other studies. Valros et al. (2003) observed a plasma urea concentration in the range of 4.6-4.8 mm on lactation day 21, also comparable with the findings of Neil (1996) who observed an average urea level of 4.5-4.9 mm throughout lactation. 63

ADDITIONAL RESULTS AND GENERAL DISCUSSION The plasma concentration of TAG is mainly affected by dietary fat intake and milk fat production. Plasma TAG increased in the third week of lactation for sows fed the 1-diet feeding regime and was reduced for sows fed the 2-diet regime, compared with the first week of lactation. Mosnier et al. (2010) also found a reduced concentration of plasma TAG as lactation proceeds. The lower TAG observed for sows fed the 2-diet regime could possibly be due to the higher MY and greater uptake of nutrients by the udder in the third week of lactation. 64

CONCLUSION Feeding of lactating sows with the new 2-diet feeding regime throughout lactation has proven to have several benefits for lactating sows. The 2-diet feeding regime improved sows MY and mean piglet weight. Feeding of sows with the 2-diet regime should also make it possible to reduce or eliminate sows weight loss during lactation, which will further improve sow reproduction and sow longevity. Expenses for feed could also be reduced by feeding sows with a well-balanced diet in the right amount. From a theoretical aspect it could also be concluded that the optimal ratio of SID Lys to energy for lactating sows varies from 0.38 on day 2 to 0.57 on day 28 of lactation, which corresponds to 4.8 and 7.2 g SID Lys/kg (12.56 MJ/kg). Based on this, the Danish recommendations for Lys should therefore be raised from the present recommendation of 6.6 g SID Lys/Fe. To achieve the full potential of the new feeding regime with two diets, it will be beneficial to uncouple the dietary supply of Lys from N, which will put an end to today s normal protein concept. It also appears that the requirement for essential AAs does not follow the same increase in Lys requirement. The 2-diet feeding regime should further be formulated to contain more energy and less CP. This should be done by adding more synthetic Lys and it is probably possible to further reduce today s recommended amount of N for lactating sows. Reaching the expected SID Lys in the lactation supplement and basal diet could possibly further have improved MY. 65

PERSPECTIVES AND IMPLEMENTATION Future feeding of lactating sows should aim to feed sows under a more dynamic feeding strategy. The feeding strategy should consider the individual sow s requirement for nutrients throughout lactation. This study has opened for the option to feed sows with two diets and be able to take into account sows changed requirement for Lys and N relatively to energy. The present study uses a complicated and labor intense approach, which cannot be used directly in commercial sow herds. To implement our knowledge of the improved feeding strategy, new feeding curves need to be developed for lactating sows based on some generalized equations. Inputs to the equation to form the feeding curves should be some simple inputs, for instance LW at farrowing and litter size. Feeding systems for practical use currently do not exist for feeding two diets in most herds today. But another approach may be easier to implement. If diets were composed to fulfill the requirements on a weekly basis, a feeding system with the possible use of one silo per week batch of lactating sows would be easy to install, and costs would be affordable. In a future study it could be recommended to reduce the CP content in the diets and fulfill the requirement for Lys by adding synthetic Lys. The true value of the dynamic feeding strategy will also be more prominent if different parity sows are included. 66

LITERATURE CITED Aherne, F. X., and Kirkwood, R. N. (1985). Nutrition and sow prolificacy. J Reprod Fert Suppl 33, 169-183. Akers, R. M., and Denbow, D. M. (2013). "Anatomy and physiology of domestic animals," Am Vet Med Assoc. Auldist, D. E., Carlson, D., Morrish, L., Wakeford, C. M., and King, R. H. (2000). The influence of suckling interval on milk production of sows. Journal of Animal Science 78, 2026-2031. Auldist, D. E., Morrish, L., Eason, P., and King, R. H. (1998). The influence of litter size on milk production of sows. Animal Science 67, 333-337. Azain, M. J., Tomkins, T., Sowinski, J. S., Arentson, R. A., and Jewell, D. E. (1996). Effect of supplemental pig milk replacer on litter performance: Seasonal variation in response. Journal of Animal Science 74, 2195-2202. Bach Knudsen, K. E. (1997). Carbohydrate and lignin contents of plant materials used in animal feeding. Animal Feed Science and Technology 67, 319-338. Baidoo, S., Aherne, F., Kirkwood, R., and Foxcroft, G. (1992). Effect of feed intake during lactation and after weaning on sow reproductive performance. Canadian Journal of Animal Science 72, 911-917. Bergsma, R., and Hermesch, S. (2012). Exploring breeding opportunities for reduced thermal sensitivity of feed intake in the lactating sow. Journal of Animal Science 90, 85-98. Bergsma, R., Kanis, E., Verstegen, M. W. A., van der Peet-Schwering, C. M. C., and Knol, E. F. (2009). Lactation efficiency as a result of body composition dynamics and feed intake in sows. Livestock Science 125, 208-222. Beyer, M., Jentsch, W., Kuhla, S., Wittenburg, H., Kreienbring, F., Scholze, H., Rudolph, P. E., and Metges, C. C. (2007). Effects of dietary energy intake during gestation and lactation on milk 67

yield and composition of first, second and fourth parity sows. Archives of Animal Nutrition 61, 452-468. Black, J. L., Mullan, B. P., Lorschy, M. L., and Giles, L. R. (1993). Lactation in the sow during heat-stress. Livest Prod Sci 35, 153-170. Christensen, T. B., and Sørensen, G. (2013). Store variationer i søers vægttab og daglig kuldtilvækst. Vol. Erfaring nr. 1316, pp. 20. Danish Pig Research Centre. Copenhagen, Denmark. Chwalibog, A. (2000). "Husdyrernæring: Bestemmelse af næringsværdi og næringsbehov," 5/Ed. Jordbrugsforlaget. Chwalibog, A., Jakobsen, K., Henckel, S., and Thorbek, G. (1992). Estimation of quantitative oxidation and fat retention from carbohydrate, protein and fat in growing pigs. Journal of Animal Physiology and Animal Nutrition-Zeitschrift Fur Tierphysiologie Tierernahrung Und Futtermittelkunde 68, 123-135. Cools, A., Maes, D., Decaluwe, R., Buyse, J., van Kempen, T., Liesegang, A., and Janssens, G. P. J. (2014). Ad libitum feeding during the peripartal period affects body condition, reproduction results and metabolism of sows. Anim Reprod Sci 145, 130-140. Cunningham, J. G., and Klein, B. G. (2013). "Textbook of Veterinary Physiology," Elsevier Health Sciences. Danielsen, V. (2003). Fodringsstrategier for diegivende søer. Grøn viden. Husdyrbrug nr. 33. Maj 2003. Dourmad, J. Y. (1991). Effect of feeding level in the gilt during pregnancy on voluntary feed-intake during lactation and changes in body-composition during gestation and lactation. Livestock Production Science 27, 309-319. 68

Dourmad, J. Y., Etienne, M., and Noblet, J. (1996). Reconstitution of body reserves in multiparous sows during pregnancy: Effect of energy intake during pregnancy and mobilization during the previous lactation. Journal of Animal Science 74, 2211-2219. Dourmad, J. Y., Etienne, M., Prunier, A., and Noblet, J. (1994). The effect of energy and proteinintake of sows on their longevity - a review. Livest Prod Sci 40, 87-97. Dourmad, J. Y., Noblet, J., and Etienne, M. (1998). Effect of protein and lysine supply on performance, nitrogen balance, and body composition changes of sows during lactation. Journal of Animal Science 76, 542-550. Eissen, J. J., Kanis, E., and Kemp, B. (2000). Sow factors affecting voluntary feed intake during lactation. Livest Prod Sci 64, 147-165. Etienne, M., Dourmad, J. Y., and Noblet, J. (1998). The influence of some sow and piglet characteristics and of environmental conditions on milk production. In "The lactating sow" (W. A. Verstegen, P. J. Moughan and J. W. Schrama, eds.), pp. 285-299. Wageningen Academic Publishers. Farmer, C., Fisette, K., Robert, S., Quesnel, H., and Laforest, J. P. (2004). Use of recorded nursing grunts during lactation in two breeds of sows. II. Effects on sow performance and mammary development. Canadian Journal of Animal Science 84, 581-587. Feyera, T., and Theil, P. K. (2014). Nutrient balances of energy, lysine and nitrogen in late gestating and early lactating sows Proceedings (EAAP Denmark 2014). Feyera, T. B. (2014a). Calculation of daily nutrient requirements of sows during the transition periods and formulation of transistion feed Internship Project. Aarhus University, Denmark, 34. Feyera, T. B. (2014b). Daily energy, lysine and nitrogen balances of lactating sows M.Sc Thesis. Aarhus University, Denmark, 1-58. 69

Flummer, C., Feyera, T., Krogh, U., Hinrichsen, T., Bruun, T. S., and Theil, P. K. (2014). The ratio between energy and lysine is essential to improve milk yield in sows. Proceedings (EAAP Denmark 2014). Forbes, J. (1988). Metabolic aspects of the regulation of voluntary food intake and appetite. Nutrition research reviews 1, 145-168. Fraser, D. (1980). A review of the behavioral mechanism of milk ejection of the domestic pig. Applied Animal Ethology 6, 247-255. Garst, A. S., Ball, S. F., Williams, B. L., Wood, C. M., Knight, J. W., Moll, H. D., Aardema, C. H., and Gwazdauskas, F. C. (1999a). Influence of pig substitution on milk yield, litter weights, and milk composition of machine milked sows. Journal of Animal Science 77, 1624-1630. Garst, A. S., Ball, S. F., Williams, B. L., Wood, C. M., Knight, J. W., Moll, H. D., Aardena, C. H., and Gwazdauskas, F. C. (1999b). Technical note: Machine milking of sows-lactational milk yield and litter weights. Journal of Animal Science 77, 1620-1623. Hansen, A. V. (2012). Feed intake in reproducing sows. In "Nutritional physiology of pigs" (K. E. Bach Knudsen, N. J. Kjeldsen, H. D. Poulsen and B. B. Jensen, eds.), pp. 1-30. Danish Pig Research Centre, Copenhagen, Denmark Hansen, A. V., Lauridsen, C., Sorensen, M. T., Knudsen, K. E. B., and Theil, P. K. (2012a). Effects of nutrient supply, plasma metabolites, and nutritional status of sows during transition on performance in the next lactation. J Anim Sci 90, 466-480. Hansen, A. V., Strathe, A. B., Kebreab, E., France, J., and Theil, P. K. (2012b). Predicting milk yield and composition in lactating sows: a Bayesian approach. J Anim Sci 90, 2285-98. Hansen, A. V., Strathe, A. B., Theil, P. K., and Kebreab, E. (2014). Energy and nutrient deposition and excretion in the reproducing sow: Model development and evaluation. Journal of Animal Science 92, 2458-2472. 70

Hansen, B. (1989). Determination of nitrogen as elementary-n, an alternative to kjeldahl. Acta Agri Scand 39, 113-118. Harrell, R., Thomas, M., and Boyd, R. (1993). Limitations of sow milk yield on baby pig growth. Proceedings (USA). Hartmann, P., McCauley, I., Gooneratne, A. D., and Whitely, J. (1984). Inadequacies of sow lactation: survival of the fittest. In "Physiological strategies in lactation." (M. Peaker, R. G. Vernon and C. H. Knight, eds.), pp. 301-326. Academic Press, Symp. zool. Soc. Lond. Hurley, W. L. (2001). Mammary gland growth in the lactating sow. Livest Prod Sci 70, 149-157. Hurley, W. L. (2015). Composition of sow colostrum and milk. In "The gestating and lactating sow" (C. Farmer, ed.), pp. 147-172, Wageningen Academic Publishers. Jørgensen, L., and Tybirk, P. (2010). "fnutrient recommendations (Normer for næringsstoffer)," 16/Ed. Kaiser, M., and Petersen, l. B. (2014). Skuldersår. Vol. Viden, pp. 12. Danish Pig Research Centre. Copenhagen, Denmark. Kemp, B. (1998). Lactational effets on the endocrinology of reproduction. In "The Lactating Sow" (W. A. Verstegen, P. J. Moughan and J. W. Schrama, eds.), pp. 241-257. Wageningen Academic Publishers. Kim, S. W., Hurley, W. L., Han, I. K., and Easter, R. A. (2000). Growth of nursing pigs related to the characteristics of nursed mammary glands. Journal of Animal Science 78, 1313-1318. Kim, S. W., Hurley, W. L., Han, I. K., Stein, H. H., and Easter, R. A. (1999). Effect of nutrient intake on mammary gland growth in lactating sows. Journal of Animal Science 77, 3304-3315. Kim, S. W., Hurley, W. L., Wu, G., and Ji, F. (2009). Ideal amino acid balance for sows during gestation and lactation. Journal of Animal Science 87, E123-E132. 71

King, R. (2000). Factors that influence milk production in well-fed sows. Journal of Animal Science 78, 19-25. King, R., and Williams, I. (1984). The effect of nutrition on the reproductive performance of firstlitter sows 1. Feeding level during lactation, and between weaning and mating. Animal Production 38, 241-247. King, R. H. (1987). Nutritional anoestrus in young sows. Pig News and lnfo., 8: 15-22. King, R. H., and Dunkin, A. C. (1986). The effect of nutrition on the reproductive-performance of 1st-litter sows.3. The response to graded increases in food-intake during lactation. Animal Production 42, 119-125. King, R. H., Mullan, B. P., Dunshea, F. R., and Dove, H. (1997). The influence of piglet body weight on milk production of sows. Livest Prod Sci 47, 169-174. Klobasa, F., Werhahn, E., and Butler, J. E. (1987). Composition of sow milk during lactation. Journal of Animal Science 64, 1458-1466. Koketsu, Y., and Dial, G. (1997). Factors influencing the postweaning reproductive performance of sows on commercial farms. Theriogenology 47, 1445-1461. Koketsu, Y., Dial, G. D., Pettigrew, J. E., and Marsh, W. E. (1996a). Characterization of feed intake patterns during lactation in commercial swine herds. Journal of Animal Science 74, 1202-1210. Koketsu, Y., Dial, G. D., Pettigrew, J. E., Marsh, W. E., and King, V. L. (1996b). Influence of imposed feed intake patterns during lactation on reproductive performance and on circulating levels of glucose, insulin, and luteinizing hormone in primiparous sows. Journal of Animal Science 74, 1036-1046. Koketsu, Y., Dial, G. D., Pettigrew, J. E., Xue, J. L., Yang, H., and Lucia, T. (1998). Influence of lactation length and feed intake on reproductive performance and blood concentrations of glucose, insulin and luteinizing hormone in primiparous sows. Anim Reprod Sci 52, 153-163. 72

Krogh, U. (Unpublished). Kruse, S., Traulsen, I., and Krieter, J. (2011). Analysis of water, feed intake and performance of lactating sows. Livestock Science 135, 177-183. Kusina, J., Pettigrew, J. E., Sower, A. F., White, M. E., Crooker, B. A., and Hathaway, M. R. (1999). Effect of protein intake during gestation and lactation on the lactational performance of primiparous sows. J Anim Sci 77, 931-941. Lachance, M. P., Laforest, J. P., Devillers, N., Laperriere, A., and Farmer, C. (2010). Impact of an extended photoperiod in farrowing houses on the performance and behaviour of sows and their litters. Canadian Journal of Animal Science 90, 311-319. Lauridsen, C., and Danielsen, V. (2004). Lactational dietary fat levels and sources influence milk composition and performance of sows and their progeny. Livestock Production Science 91, 95-105. Le Dividich, J., and Seve, B. (2001). Energy requirements of the young pig. In "The weaner pig: nutrition and management." (M. A. Varley and J. Wiseman, eds.), pp. 17-45. CABI Publishing, Wallingford, UK. Løvendahl, P., and Purup, H. M. (2002). Technical note: Time-resolved fluoro-immunometric assay for intact insulin in livestock species. J Anim Sci 80, 191-195. Mackenzie, D. D. S., and Revell, D. K. (1998). Genetic influences on milk quantity. In "The lactating sow" (W. A. Verstegen, P. J. Moughan and J. W. Schrama, eds.), pp. 91-112. Wageningen Academic Publishers. Mosnier, E., Etienne, M., Ramaekers, P., and Père, M. C. (2010). The metabolic status during the peri partum period affects the voluntary feed intake and the metabolism of the lactating multiparous sow. Livest Sci 127, 127-136. 73

Moustsen, V. A., and Pedersen, M. L. (2010). Adfærd under diegivning - effekt af farestitype. Vol. Meddelelse nr. 874, pp. 17. Danish Pig Research Centre. Copenhagen, Denmark. Mullan, B. P., and Williams, I. H. (1989). The effect of body reserves at farrowing on the reproductive-performance of 1st-litter sows. Animal Production 48, 449-457. Mullan, B. P., and Williams, I. H. (1990). The chemical-composition of sows during their 1st lactation. Animal Production 51, 375-387. Neil, M. (1996). Ad libitum lactation feeding of sows introduced immediately before, at, or after farrowing. Animal Science 63, 497-505. Neil, M., Ogle, B., and Anner, K. (1996). A two-diet system and ad libitum lactation feeding of the sow.1. Sow performance. Animal Science 62, 337-347. Nielsen, O. L., Pedersen, A. R., and Sorensen, M. T. (2001). Relationships between piglet growth rate and mammary gland size of the sow. Livest Prod Sci 67, 273-279. Noble, M. S., Rodriguez-Zas, S., Cook, J. B., Bleck, G. T., Hurley, W. L., and Wheeler, M. B. (2002). Lactational performance of first-parity transgenic gilts expressing bovine alphalactalbumin in their milk. Journal of Animal Science 80, 1090-1096. Noblet, J., and Etienne, M. (1986). Effect of energy-level in lactating sows on yield and composition of milk and nutrient balance of piglets. Journal of Animal Science 63, 1888-1896. Noblet, J., and Etienne, M. (1987a). Body composition, metabolic rate and utilization of milk nutrients in suckling piglets. Reproduction Nutrition Development 27, 829-839. Noblet, J., and Etienne, M. (1987b). Metabolic utilization of energy and maintenance requirements in lactating sows. Journal of Animal Science 64, 774-781. Noblet, J., and Etienne, M. (1989). Estimation of sow milk nutrient output. Journal of Animal Science 67, 3352-3359. 74

Noblet, J., Etienne, M., and Dourmad, J. Y. (1998). Energetic efficiency of milk production. In "The lactating sow" (W. A. Verstegen, P. J. Moughan and J. W. Schrama, eds.), pp. 113-130. Wageningen Academic Publishers. NRC (2012). "Nutrient Requirements of Swine," 11/Ed. National Academy Press, Washington, DC, USA. O'grady, J. F., Lynch, P. B., and Kearney, P. A. (1985). Voluntary feed-intake by lactating sows. Livest Prod Sci 12, 355-365. Oliviero, C., Kokkonen, T., Heinonen, M., Sankari, S., and Peltoniemi, O. (2009). Feeding sows with high fibre diet around farrowing and early lactation: Impact on intestinal activity, energy balance related parameters and litter performance. Research in Veterinary Science 86, 314-319. Park, B. C., Ha, D. M., Park, M. J., and Lee, C. Y. (2014). Effects of milk replacer and starter diet provided as creep feed for suckling pigs on pre- and post-weaning growth. Animal Science Journal 85, 872-878. Pedersen, L. J., Berg, P., Jørgensen, E., Bonde, M. K., Herskin, M. S., Knage-Rasmussen, K. M., Kongsted, A. G., Lauridsen, C., Oksbjerg, N., Poulsen, H. D., Sorensen, D. A., Su, G., Sørensen, M. T., Theil, P. K., Thodbjerg, K., and Jensen, K. H. (2010). Pattegrisedødelighed i DK: Muligheder for reduktion af pattegrisedødelighed i Danmark. Husdyrbrug nr. 86. Pedersen, M. L., Moustsen, V. A., Nielsen, M. B. F., and Kristensen, A. R. (2011). Improved udder access prolongs duration of milk letdown and increases piglet weight gain. Livestock Science 140, 253-261. Persson, A., Pedersen, A. E., Goransson, L., and Kuhl, W. (1989). A long-term study on the healthstatus and performance of sows on different feed allowances during late pregnancy.1. Clinical observations, with special reference to agalactia post partum. Acta Veterinaria Scandinavica 30, 9-17. 75

Pluske, J. R., Williams, I. H., Zak, L. J., Clowes, E. J., Cegielski, A. C., and Aherne, F. X. (1998). Feeding lactating primiparous sows to establish three divergent metabolic states: III. Milk production and pig growth. Journal of Animal Science 76, 1165-1171. Reese, D., Peo Jr, E., and Lewis, A. (1984). Relationship of Lactation Energy Intake and Occurence of Postweaning Estrus to Body and Back Fat Composition in Sows. J Anim Sci 58, 1236-1244. Revell, D. K., Williams, I. H., Mullan, B. P., Ranford, J. L., and Smits, R. J. (1998a). Body composition at farrowing and nutrition during lactation affect the performance of primiparous sows: I. Voluntary feed intake, weight loss, and plasma metabolites. Journal of Animal Science 76, 1729-1737. Revell, D. K., Williams, I. H., Mullan, B. P., Ranford, J. L., and Smits, R. J. (1998b). Body composition at farrowing and nutrition during lactation affect the performance of primiparous sows: II. Milk composition, milk yield, and pig growth. Journal of Animal Science 76, 1738-1743. Rozeboom, D. W., Pettigrew, J. E., Moser, R. L., Cornelius, S. G., and Elkandelgy, S. M. (1994). In vivo estimation of body composition of mature gilts using live weight, backfat thickness, and deuterium oxide. J Anim Sci 72, 355-366. Silva, B. A. N., Noblet, J., Donzele, J. L., Oliveira, R. F. M., Primot, Y., Gourdine, J. L., and Renaudeau, D. (2009). Effects of dietary protein level and amino acid supplementation on performance of mixed-parity lactating sows in a tropical humid climate. Journal of Animal Science 87, 4003-4012. Smith, W. C., Ellis, M., Chadwick, J. P., and Laird, R. (1991). The influence of index selection for improved growth and carcass characteristics on appetite in a population of large white-pigs. Animal Production 52, 193-199. Smits, R. J., Henman, D. J., and King, R. H. (2013). Increasing the energy content of lactation diets fed to first-litter sows reduces weight loss and improves productivity over two parities. Animal Production Science 53, 23-29. 76

Spinka, M., Illmann, G., Algers, B., and Stetkova, Z. (1997). The role of nursing frequency in milk production in domestic pigs. Journal of Animal Science 75, 1223-1228. Strathe, A. B., Jorgensen, H., Kebreab, E., and Danfaer, A. (2012). Bayesian simultaneous equation models for the analysis of energy intake and partitioning in growing pigs. Journal of Agricultural Science 150, 764-774. Sønderby, T. B. (2013). Fodring af diegivendesøer. In "Viden". Danish Pig Research Centre. Copenhagen, Denmark. Sørensen, G. (2005). Tørfoder efter ædelyst til diegivende søer. Vol. Meddelelse nr. 686, pp. 22. Danish Pig Research Centre. Copenhagen, Denmark. Sørensen, G. (2009). Flere daglige fodringer i diegivningsperioden nedsætter risikoen for skuldersår med 30 procent. Vol. Meddelelse nr. 847. Danish Pig Research Centre. Copenhagen, Denmark. Theil, P. K. (2015). Transition feeding of sows. In "The gestating and lactating sow" (C. Farmer, ed.), pp. 147-172, Wageningen Academic Publishers. Theil, P. K., Jorgensen, H., and Jakobsen, K. (2002a). Energy and protein metabolism in pregnant sows fed two levels of dietary protein. Journal of Animal Physiology and Animal Nutrition 86, 399-413. Theil, P. K., Jorgensen, H., and Jakobsen, K. (2004). Energy and protein metabolism in lactating sows fed two levels of dietary fat. Livest Prod Sci 89, 265-276. Theil, P. K., Nielsen, M. O., Sørensen, M. T., and Lauridsen, C. (2012). Lactation, milk and suckling. In "Nutritional physiology of pigs" (K. E. Bach Knudsen, N. J. Kjeldsen, H. D. Poulsen and B. B. Jensen, eds.), pp. 1-49. Danish Pig Research Centre, Copenhagen, Denmark Theil, P. K., Nielsen, T. T., Kristensen, N. B., Labouriau, R., Danielsen, V., Lauridsen, C., and Jakobsen, K. (2002b). Estimation of milk production in lactating sows by determination of deuterated water turnover in three piglets per litter. Acta Agri Scand (Sec A) 52, 221-232. 77

Theil, P. K., Olesen, A. K., Flummer, C., Sorensen, G., and Kristensen, N. B. (2013). Impact of feeding and post prandial time on plasma ketone bodies in sows during transition and lactation. Journal of Animal Science 91, 772-782. Theil, P. K., Sejrsen, K., Hurley, W. L., Labouriau, R., Thomsen, B., and Sorensen, M. T. (2006). Role of suckling in regulating cell turnover and onset and maintenance of lactation in individual mammary glands of sows. J Anim Sci 84, 1691-1698. Thodberg, K., and Sørensen, M. T. (2006). Mammary development and milk production in the sow: Effects of udder massage, genotype and feeding in late gestation. Livest Sci 101, 116-125. Vadmand, C. N., Krogh, U., Hansen, C. F., and Theil, P. K. (Submitted). Impact of sow and litter characteristics on colostrum yield, time for onset of lactation, and milk yield of sows. Valros, A., Rundgren, M., Spinka, M., Saloniemi, H., Rydhmer, L., Hulten, F., Uvnas-Moberg, K., Tomanek, M., Krejci, P., and Algers, B. (2003). Metabolic state of the sow, nursing behaviour and milk production. Livest Prod Sci 79, 155-167. Vestergaard, K., Christensen, G., Petersen, L. B., and Wachmann, H. (2004). Afgangsårsager hos søer - samt obduktionsfund hos aflivede og selvdøde søer. Danish Pig Research Centre. Message nr. 656. Copenhagen, Denmark. Vinther, J. (2014). Landsgennemsnit for produktivitet i svineproduktionen 2014. Danish Pig Research Centre. Copenhagen, Denmark. Weast, R. C. (1969). "Handbook of chemistry and physics,," 49th/Ed. Weldon, W., Lewis, A., Louis, G., Kovar, J., and Miller, P. (1994). Postpartum hypophagia in primiparous sows: II. Effects of feeding level during gestation and exogenous insulin on lactation feed intake, glucose tolerance, and epinephrine-stimulated release of nonesterified fatty acids and glucose. Journal of animal science 72, 395-403. 78

Wheeler, M. B., Bleck, G. T., and Donovan, S. M. (2001). Transgenic alteration of sow milk to improve piglet growth and health. In "Control of Pig Reproduction Vi" (R. D. Geisert, H. Niemann and C. Doberska, eds.), pp. 313-324. Whittemore, C. T. (1996). Nutrition reproduction interactions in primiparous sows. Livest Prod Sci 46, 65-83. Whittemore, C. T., and Morgan, C. A. (1990). Model components for the determination of energy and protein-requirements for breeding sows - a review. Livest Prod Sci 26, 1-37. Williams, I. H. (1998). Nutritional strategy effects during lactation and during the interval from weaning to oestrus. In "The lactating sow" (W. A. Verstegen, P. J. Moughan and J. W. Schrama, eds.), pp. 159-181. Wageningen Academic Publishers. Woods, S., Figlewicz, L. D., Schwartz, M., and Porte Jr, D. (1990). A reassessment of the regulation of adiposity and appetite by the brain insulin system. Int. J. Obes. 14, 69-73. Woods, S. C., Seeley, R. J., Porte, D., and Schwartz, M. W. (1998). Signals that regulate food intake and energy homeostasis. Science 280, 1378-1383. Zak, L. J., Cosgrove, J. R., Aherne, F. X., and Foxcroft, G. R. (1997). Pattern of feed intake and associated metabolic and endocrine changes differentially affect postweaning fertility in primiparous lactating sows. Journal of Animal Science 75, 208-216. Zijlstra, R. T., Whang, K.-Y., Easter, R. A., and Odle, J. (1996). Effect of feeding a milk replacer to early-weaned pigs on growth, body composition, and small intestinal morphology, compared with suckled littermates. Journal of Animal Science 74, 2948-2959. Zou, S., McLaren, D. G., and Hurley, W. L. (1992). Pig colostrum and milk-composition - comparisons between chinese meishan and united-states breeds. Livestock Production Science 30, 115-127. 79