Aus dem Institut für Tierzucht und Tierhaltung der Agrar- und Ernährungswissenschaftlichen Fakultät der Christian-Albrechts-Universität zu Kiel

Transcription

1 Aus dem Institut für Tierzucht und Tierhaltung der Agrar- und Ernährungswissenschaftlichen Fakultät der Christian-Albrechts-Universität zu Kiel Effects of an ad libitum milk supply during early life of calves on shortterm stress as well as long-term development of the endocrine pancreas and assessment of milk quality aspects Dissertation zur Erlangung des Doktorgrades der Agrar- und Ernährungswissenschaftlichen Fakultät der Christian-Albrechts-Universität zu Kiel vorgelegt von: M.Sc. agr. Luise Prokop aus Berlin Dekan: Prof. Dr. Eberhard Hartung 1. Berichterstatter: Jun.Prof. Dr. Steffi Wiedemann 2. Berichterstatter: Prof. Dr. Gerald Rimbach Tag der mündlichen Prüfung: 04. November 2015 Die Dissertation wurde mit dankenswerter finanzielle Unterstützung im Rahmen eines Landesgraduiertenstipendiums (Graduiertenzentrum der CAU Kiel) und der Rügenwalder Mühle im Rahmen des Arbeitskreises Tierwohl durchgeführt.

2

3 TABLE OF CONTENTS GENERAL INTRODUCTION... 1 CHAPTER ONE Effect of acidification and storage temperature on the microbiological quality of milk intended for the ad libitum feeding of calves... 5 CHAPTER TWO Effect of an ad libitum milk supply during the first three weeks of life of dairy calves on heart rate and heart rate variability during feeding and rehousing CHAPTER THREE Intensive rearing of male calves during the first three weeks of life has long-term effects on number of islets of Langerhans and insulin stained area in the pancreas GENERAL DISCUSSION GENERAL SUMMARY ALLGEMEINE ZUSAMMENFASSUNG... 99

4

5 GENERAL INTRODUCTION In Germany, dairy calves are typically separated from their mother within the first hours of life and fed limited amounts of milk or milk replacer twice per day until weaning (e.g., 10 % of the calf s body weight). Arguments for this restrictive feeding protocol are based on reports of earlier and higher solid feed intake which was found to result in faster rumen development as well as in reduced costs for rearing (Drackley, 2008; Hill et al., 2010). However, recent calculations on energy requirements of calves reported that the recommended restrictive milk allowance only barely meets the energy requirements for maintenance particularly during winter times, but does not exploit the needs for the high growth potential of the pre-weaned calves (Kunz, 2012). On the other hand, it has been reported that calves left with their dams suckle from their mothers teat 7-10 times a day and ingest approximately 25% of body weight per day in milk (Albright and Arave, 1997). This results in much higher weight gains during the pre-weaning period in suckler calves. Therefore, research on the effects of the restrictive feeding protocol in comparison to higher nutrient intakes during the first weeks of life has intensified over the last decade. So far, several studies have revealed that the provision of free access to milk at all times results in comparable milk intakes between individually housed calves fed via suckling buckets and suckler calves (Appleby et al., 2001; Hammell et al., 1988; Jasper and Weary, 2002; Maccari et al., 2014). As a consequence, the postnatal growth rates in ad libitum fed calves are also considerably higher with up to 1,200 g per day in comparison to restrictively fed calves (Hammon et al., 2002b; Jasper and Weary, 2002; Maccari et al., 2014). Furthermore, it has been reported that a higher milk allowance during the early neonatal period positively affects performance, health, behavior and welfare (summarized in Khan et al., 2011a). Calves with free access to milk have lower disease and mortality rates compared to restrictively fed calves (Appleby et al., 2001; Jasper and Weary, 2002). Furthermore, ad libitum feeding of calves through a nipple has positive effects on behavior such that cross-suckling and other behavioral disorders are reduced because natural sucking behavior is satisfied in these calves (Flower and Weary, 2001; Hammell et al., 1988; von Keyserlingk et al., 2006). Free access to milk at all times also allows calves to suckle several times and small portions of milk respectively, which is said to improve digestion (de Passillé et al., 1992). Increased feeding of milk is also reported to improve a calf s immunity (Nonnecke, 2000). 1

6 In addition to the mentioned positive short-term effects of an increased milk allowance in preweaned calves, numerous long-term effects have also been demonstrated previously. In some studies an intensified feeding during the first weeks of live results in higher body weights compared with restrictive feeding far beyond the time of intervention (Maccari et al., 2014; Moallem et al., 2010). This early weight advantage could lower the age of first calving and therefore the costs of production (Cady and Smith, 1996). Female calves with free access to milk in their first four weeks of life also displayed higher milk yields (+ 0.8 kg per day) during their first lactation in comparison to restrictively fed calves (Wiedemann et al., 2015). Despite the previously published results, some questions remained unanswered about the effects of feeding a higher quantity of milk during the first weeks of life in calves. It was therefore the aim of this thesis to evaluate three different aspects of feeding calves ad libitum in their first weeks of life in comparison to feeding them according to the conventional restrictive protocol. (1) Is ad libitum feeding of calves associated with an increased risk of microbial infection because of the long storage period of milk in the nipple buckets? (2) Are there differences in the stress reaction during feeding and rehousing between ad libitum and restrictively fed calves? (3) Does the feeding protocol during early life have long-term effects on the endocrine pancreatic tissue which could indicate a metabolic programming of the animals? Chapter 1 discusses the analysis of the microbiology of milk intended for the ad libitum feeding of calves. Untreated, pasteurized, acidified and pasteurized acidified milk was stored for different time periods and temperatures. Furthermore, the development of two bacterial pathogens in untreated and acidified milk was investigated at different temperatures of storage. Finally, a recommendation for the preservation of milk is made in regard to the health of calves. Chapter 2 presents the results of measuring heart rate and heart rate variability in ad libitum fed calves during feeding time and rehousing in contrast to restrictively fed calves. These parameters were measured in 56 Holstein Friesian calves at the end of their 3 rd living week before and during rehousing. In conclusion, statements are made about the influence of feeding intensity on stress load in calves. Chapter 3 deals with the impact of ad libitum feeding on the development of pancreatic islets of Langerhans. 42 Holstein bull calves were either reared intensively or according to an established rearing protocol. After the calves were slaughtered at an age of 9 months, pancreatic tissue was histologically analyzed with regard to number and area of islets of 2

7 Langerhans. The results could provide preliminary conclusions on the long-term effects of the feeding protocol during very early life on the later life metabolism. In the general discussion, some additional results are presented which are not included in the main chapters. At the end, general conclusions are given for each treated topic to summarize the important findings and make recommendations. References Albright, J. C., and C. Arave Behavioural responses to management systems. The Behaviour of Cattle: Appleby, M. C., D. M. Weary, and B. Chua Performance and feeding behaviour of calves on ad libitum milk from artificial teats. Applied Animal Behaviour Science 74: Cady, R. A., and R. R. Smith Economics of heifer raising programs. Proceedings from the Calves, Heifers, and Dairy Profitability National Conference. p 7. Northeast Reg. Ag. Eng. Serv. NRAES-74, Harrisburg, PA. de Passillé, A. M. B., J. H. M. Metz, P. Mekking, and P. R. Wiepkema Does drinking milk stimulate sucking in young calves? Applied Animal Behaviour Science 34: Drackley, J. K Calf nutrition from birth to breeding. Veterinary Clinics of North America: Food Animal Practice 24: Flower, F. C., and D. M. Weary Effects of early separation on the dairy cow and calf: 2. Separation at 1 day and 2 weeks after birth. Applied Animal Behaviour Science 70: Hammell, K. L., J. H. M. Metz, and P. Mekking Sucking behaviour of dairy calves fed milk ad libitum by bucket or teat. Applied Animal Behaviour Science 20: Hammon, H. M., G. Schiessler, A. Nussbaum, and J. W. Blum Feed intake patterns, growth performance, and metabolic and endocrine traits in calves fed unlimited amounts of colostrum and milk by automate, starting in the neonatal period. Journal of Dairy Science 85: Hill, T. M., H. G. Bateman, J. M. Aldrich, and R. L. Schlotterbeck Effect of milk replacer program on digestion of nutrients in dairy calves. Journal of Dairy Science 93:

8 Jasper, J., and D. M. Weary Effects of ad libitum milk intake on dairy calves. Journal of Dairy Science 85: Khan, M. A., D. M. Weary, and M. A. G. von Keyserlingk Effects of milk ration on solid feed intake, weaning, and performance in dairy heifers Journal of Dairy Science 94: Kunz, H.-J Kälber Handbuch. Agrar- und Veterinär-Akademie, Laer. Maccari, P., S. Wiedemann, H. J. Kunz, M. Piechotta, P. Sanftleben, and M. Kaske Effects of two different rearing protocols for Holstein bull calves in the first 3 weeks of life on health status, metabolism and subsequent performance. Journal of Animal Physiology and Animal Nutrition 99: Nonnecke, B Effects of dietary energy and protein on the immunological performance of milk replacer-fed Holstein bull calves. In: American Dairy Science Association Abstracts von Keyserlingk, M. A. G., F. Wolf, M. Hötzel, and D. M. Weary Effects of continuous versus periodic milk availability on behavior and performance of dairy calves. Journal of Dairy Science 89: Wiedemann, S., P. Holz, H.-J. Kunz, E. Stamer, and M. Kaske Effect of ad libitum feeding of Holstein-Friesian calves during the first four weeks of life on weight development as well as milk yield and feed intake during first lactation. Züchtungskunde 87(6):

9 CHAPTER ONE Effect of acidification and storage temperature on the microbiological quality of milk intended for the ad libitum feeding of calves Luise Prokop a, Karin Knappstein b, Hans-Georg Walte b, Steffi Wiedemann a, * a University of Kiel, Animal Health, Institute of Animal Breeding and Husbandry, Olshausenstraße 40, Kiel, Germany, lprokop@tierzucht.uni-kiel.de, swiedemann@tierzucht.uni-kiel.de b Max-Rubner-Institute, Department of Safety and Quality of Milk and Fish Products, Hermann-Weigmann-Straße 1, Kiel, Germany, karin.knappstein@mri.bund.de, hans-georg.walte@mri.bund.de * Corresponding author: Steffi Wiedemann, Tel ; swiedemann@tierzucht.uni-kiel.de 5

10 Abstract In enhanced feeding programs for neonatal calves, whole milk typically remains in suckling buckets for long time periods. Therefore, the objective of this study was to compare the total bacterial count (TBC) between acidified milk (ph = 5.5) and untreated milk (trial 1; n = 10 bulk tank milk samples) and between pasteurised milk (mobile on-farm pasteuriser) and untreated milk (trial 2; n = 6 bulk tank milk samples) at different incubation temperatures (10 C, 20 C and 30 C). To analyse the effect on selected pathogenic bacteria, milk with a low total bacterial count was artificially contaminated with two pathogens (Staphylococcus aureus Wood 46, Escherichia (E.) colio45). For both trials the total bacterial count of samples was determined after 0, 6, 9 and 12 h of incubation. In trial 1, no difference in TBC between untreated and acidified milk was observed for incubation at 10 C for 12 h. At 20 C, an increase in TBC was detected after 12 h both in untreated and acidified milk. Incubation temperatures of 30 C resulted in a bacterial growth after 6, 9 and 12 h in untreated milk and after 9 and 12 h in acidified milk, respectively and TBC in untreated milk was higher compared with acidified milk at all incubation periods. In trial 2, higher numbers of S. aureus were detected in untreated milk compared with acidified milk after 9 and 12 h at 20 C and after 6, 9 and 12 h at 30 C, respectively. Acidification resulted in a reduction of E. coli growth for 12 h at 20 C and for 6, 9 and 12 h at 30 C, respectively. The TBC in pasteurised milk increased after 12 h at 20 C and after all periods of incubation at 30 C. No bacterial growth at any temperature and incubation time was found for pasteurised and acidified milk. In conclusion, it is recommended to acidify milk above environmental temperatures of 10 C to reduce the bacterial uptake of calves fed an enhanced milk allowance twice a day. At and above temperatures of 30 C, acidified milk should be replaced every 6 h. Pasteurisation and acidification is advised when milk with a high initial TBC or waste milk is fed to calves. Key words: microbiological quality, ad libitum feeding, total bacterial count, calves, raw milk, pasteurisation 6

11 Introduction During the last decade, numerous studies on enhanced feeding of calves have shown positive effects on animal performance, health, behaviour and welfare (Appleby et al., 2001; Bar- Peled et al., 1997; Jasper & Weary, 2002). Furthermore, it has also been demonstrated that calves with free access to milk in the first weeks of life (ad libitum feeding) produce more milk during their first lactation compared to calves which were fed restrictively in early life (summarised in Kaske et al., 2009; Wiedemann et al., 2015). Under practical conditions, neonatal calves are often housed in individual hutches for hygienic and monitoring reasons and are fed with milk from suckling buckets. To provide milk ad libitum the suckling buckets typically remain at the calves hutches over periods of 12 to 24 h and are only removed for cleaning/disinfection and refilling. It is commonly suggested that refilling the buckets at least twice a day results in a milk storage time of about 12 h at varying outside temperatures. Also, group housing of calves and enhanced feeding of whole milk via automatic calf feeders is increasingly practiced on European farms, which also results in long storage time periods for the milk (Hepola, 2003). However, non-cooled milk is an ideal culture medium for bacteria and other micro-organisms which may compromise calf health and milk acceptance. For example, in colostrum the total bacterial plate count (TBC) increased from 100,000 cfu/ml to over 18,000,000 cfu/ml after incubation at 73 F (22.8 C) for 24 h (Stewart et al., 2005). Additionally, the initial TBC has great impact on the multiplication of germs particularly against the background that often non-saleable milk of infected mammary glands is fed to calves (Muir, 1996). The viability and proliferation of bacteria are also affected by physical and chemical factors, including the availability of nutrients in the growth medium, temperature, oxygen tension and ph (Atlas, 1984; Phillips & Griffiths, 1987). To extend the storage time of milk, different preservation procedures have been established. In this regard, the pasteurisation and acidification of milk before feeding are the most common practical methods to reduce the initial count and/or the further proliferation of micro-organisms. Pasteurisation of colostrum results in a reduction in the total plate and total coliform counts and a decrease in calves risk of being subjected to diarrhoea (Godden et al., 2005). Acidification reduces the ph in milk to ph values of , which may destroy acidintolerant bacteria and inhibit bacterial growth (Booth et al., 1989). For acidification, organic acids (chemical acidification) or yoghurt cultures (natural acidification) can be used (Wouters et al., 2002; Davidson & Taylor, 2007). It was the aim of this study to analyse the effect of preservation and incubation at different temperature on the development of the total bacterial 7

12 count as well as of selected bacterial pathogens in milk intended for the ad libitum feeding of calves. Materials and Methods Due to the high labour input of the respective series of tests, the microbial proliferation in both acidified and pasteurised milk were compared separately with the proliferation in untreated milk. Bacterial growth in untreated and acidified milk Sample collection and microbiological analysis In order to include different bacterial populations and different levels of contamination, bulk tank milk samples from 5 dairy farms in Northern Germany (3 research farms, 2 practice farms) were collected on two different days, respectively. Practice farms with known high bacterial counts in milk were chosen. Bulk tank milk samples were collected in sterile 1-L screw-cap canisters. Samples were held on ice and analysed within 24 h. On the test day, the milk samples were mixed thoroughly and 200 ml milk was transferred to a separate tube. This aliquot was acidified with a feed acid (Schaumacid Drink C flüssig, Schaumann GmbH, Pinneberg, Germany; 1.5 ml/l milk; Figure 1). Untreated and acidified milk were apportioned in tubes containing 10 ml milk and stored in duplicates at 10 C, 20 C and 30 C for 0, 6, 9 and 12 hours (test scheme; Figure 1), respectively. Immediately after the desired incubation intervals, standard plate counts were performed in duplicates, to determine TBC according to ISO : (International Organization for Standardization, 2013). Bacterial growth of selected bacterial pathogens Sample collection and microbiological analysis In order to obtain milk with a TBC <300 cfu/ml, milk samples of cows with a low somatic cell count (<10,000/ml) were collected with sterilised milking equipment on four different days. The milk was held on ice until analysis within 24 h. For contamination, the Grampositive Staphylococcus (S.) aureus strain Wood 46 and Gram-negative strain Escherichia (E.) coli strain O45 were exemplary chosen because of their relevance as potential causes of diseases in calves (E. coli) and cows (S. aureus) (Jamali et al., 2015; Janke et al., 1990). The inoculum was prepared by incubation in Tryptone Soya Broth (CM0129B, Oxoid GmbH, Wesel, Germany) for 18 h at 37 C. To obtain an initial bacterial count of 10² cfu/ml, 4 8

13 ml of the inoculum was added to 400 ml milk with a low bacterial count. Half of the milk was acidified with a feed acid as described above. Aliquots of milk samples were prepared and incubated according to the test scheme (Figure 1). After the desired incubation intervals decimal dilutions in 0.85% sodium chloride were prepared and 0.1 ml of the dilution was spread on blood agar plates (Columbia blood agar basis, CM0331B, Oxoid GmbH; plus 5% defibrinated sheep blood) in duplicates. After an incubation of 24 h at 37 C only plates containing 10 to 400 colonies were counted. Comparison of pasteurised and pasteurised, acidified bulk tank milk Sample collection and microbiological analysis Bulk tank milk samples (at least 40 L) from 3 dairy farms in Northern Germany (2 research farms, 1 practice farm) were collected on two different days, respectively. The milk samples were cooled prior to the pasteurisation process. An on-farm pasteuriser (Milchtaxi 80 L Pasteurisierer; Holm & Laue, Westerrönfeld, Germany) was used to pasteurise milk in a batch process (64 C, 35 min). The pasteuriser is equipped with a heating element in the bottom of the tank, a cooling coil in the external wall working with cold water and a stirring device at the bottom of the steel tank. The milk circulates through a dispensing device during pasteurisation. Immediately after filling the milk into the pasteuriser, a 50 ml milk sample was collected into a sterile screw-cap vessel to determine TBC before pasteurisation (0 h). The programmed pasteurisation started automatically at a.m.. The milk was cooled to 13 C before and after pasteurisation. One sample of milk was taken in a 1-L screw-cap canister to determine TBC directly after pasteurisation. According to the manufacturer s instruction the remaining milk was removed after each pasteurisation process and the pasteuriser was refilled with warm water (60 C) and an alkaline cleaning agent (1% solution). Manual cleaning was performed with a large brush. Subsequently, the cleaning solution was pumped automatically through the dosage unit for 20 min. Rinsing was performed manually for 5 min. An aliquot of 200 ml of the pasteurised milk sample was acidified with a feed acid as mentioned above. Aliquots of pasteurised and pasteurised, acidified milk samples were prepared and incubated according to the test scheme (Figure 1). Statistical analysis All statistical analyses were performed using SAS (Version 9.3, SAS Institute Inc., Cary, NC 27513, USA). Bacterial counts were logarithmised to base 10 for statistical evaluation. 9

14 The repeatability of microbial testing was calculated according to ISO :2009 IDF 128-1: 2009). The GLIMMIX procedure in SAS was applied to estimate differences between untreated and acidified milk. The treatment of milk, the initial TBC and the interaction between treatment of milk, time point and temperature were included as fixed effects. The initial TBC was classified as low, medium, or high ( 4 log 10 cfu/ml, >4 to 5 log 10 cfu/ml, >5 log 10 cfu/ml). The milk sample was considered as a random effect. If an overall significant effect was found, a subsequent Bonferroni post hoc analysis was performed. Probability values less than 0.05 (p<0.05) were considered statistically significant. Data are presented as LS-means (± standard error). Results Bacterial growth in untreated and acidified milk The repeatability determined between the standard plate counts in duplicates was r = 0.24, which complies with the ISO norm ISO IDF 128-1:2009. The initial bacterial count, the treatment of milk and the interaction of treatment of milk, time and temperature of incubation affected TBC (all P < 0.01; Table 1; Figure 2). At 10 C, no change in TBC was detected neither in untreated nor in acidified milk over 12 h of incubation. At 20 C TBC increased from 0 h to 12 h of incubation by 1.09 log 10 units in untreated and by 0.56 log 10 units in acidified milk, respectively. At 30 C, an increase in TBC by 1.0 log 10 unit was demonstrated after 6 h of incubation of untreated milk, whereas a significant increase in acidified milk only appeared after 9 h. After 12 h of incubation, TBC increased by 3.0 log 10 units in untreated milk and by 2.26 log 10 units in acidified milk, respectively. Incubation times of 12 h at 20 C and all incubation times at 30 C resulted in reduced TBC in acidified milk in comparison to untreated milk. Bacterial growth of selected bacterial pathogens The repeatability of the results of milk inoculated with S. aureus (r = 0.48) was lower than required. The repeatability of milk inoculated with E. coli (r = 0.23) fulfilled the requirements. 10

15 Treatment of milk and the interaction of treatment of milk, time and temperature of incubation affected the bacterial counts of S. aureus strain Wood 46 and E. coli strain O45 (all P < 0.05; Figure 3). At 10 C no increase in S. aureus was detected for all incubation time periods in acidified and untreated milk. At 20 C S. aureus counts increased by 1.04 log 10 units in untreated milk after 9 h of incubation compared to 0 h and remained on the same level up to 12 h of incubation. At 20 C, an acidification resulted in an inhibition of growth of S. aureus at any incubation time period. At 30 C, an increase in S. aureus counts by 1.86 log 10 units was demonstrated after 6 h of incubation in untreated milk. After 9 and 12 h S. aureus counts increased by 4.46 log 10 units and 5.86 log 10 units, respectively. At 30 C S. aureus increased also in acidified milk over all time periods of incubation by up to 2.38 log 10 units. Incubation times of 9 and 12 h at 20 C and all incubation times at 30 C resulted in lower counts of S. aureus in acidified compared to untreated milk. As in S. aureus, no increase in E. coli was detected for all incubation times at 10 C in acidified and untreated milk. At 20 C E. coli counts increased in untreated milk after incubation of 9 h by 1.16 log 10 units and 12 h by 1.59 log 10 units, respectively, whereas no increase in acidified milk could be detected after all incubation time periods at 20 C. At 30 C and 6 h of incubation E. coli counts in untreated milk increased by 2.28 log 10 units, whereas no increase in acidified milk was detected. After 9 and 12 h of incubation a comparable increase of E. coli was found in acidified and untreated milk (by 3.30 log 10 units vs log 10 units and 4.29 log 10 units vs log 10 units in comparison to 0 h values). Only incubation time periods of 6 h at 30 C and of 9 and 12 h at 20 C resulted in significant differences in the count of E. coli between acidified and non-acidified milk, respectively. Comparison of pasteurised and pasteurised, acidified bulk tank milk The repeatability of the analyses of TBC (r = 0.24) was within the requirements. The TBC of pasteurised and pasteurised, acidified milk was affected by the initial bacterial count, the treatment of milk and the interaction of treatment of milk, time and temperature of incubation (all P < 0.05; Figure 4). Pasteurisation reduced the TBC by 0.4 log 10 units on average. The highest reduction rate after pasteurisation was 1.3 log 10 units (initial TBC of 5.19 log 10 cfu/ml). In 2 out of 6 milk samples, an increase in TBC was detected after pasteurisation (by 0.15 and 0.18 log 10 units, respectively). At 10 C and 20 C, no change in TBC was detected neither in pasteurised nor in 11

16 pasteurised and acidified milk over all periods of incubation. At 30 C, TBC in pasteurised milk increased after 6 h by 0.85 log 10 units and after 12 h by 2.68 log 10 units compared to 0 h values. No increase in bacterial growth was observed independent of time and temperature in pasteurised, acidified milk. Pasteurisation and subsequent acidification resulted in a significant lower TBC at 20 C after 12 h and at 30 C after all periods of incubation in comparison to pasteurised milk with no subsequent acidification. Discussion This study aimed to evaluate bacterial proliferation in milk, which is intended to provide ad libitum feeding of pre-weaned calves in dependence on time and temperature of storage and acidification. In order to offer free access to whole milk at all times, it is recommended that suckling buckets are filled with 6 to 8 L milk twice a day and left at the calves hutches. This approach results in longer milk incubation times in comparison to milk which is provided according to the established rearing protocol for only a limited time period. Therefore, the risk of microbial spoilage of ad libitum offered milk is increased (Spreer, 1998). It has to be mentioned that an increase in TBC does not automatically result in an increase in pathogens, but nonetheless this might serve as an indicator of a possible health risk. For bacterial growth, the initial number and the composition of competing micro-organisms as well as their metabolic products play an important role (Claeys et al., 2013). In the milk of healthy cows, the initial milk TBC is about cfu/ml milk, but initial TBC can increase up to 1,000,000 cfu/ml in milk collected from infected mammary glands (Jeffrey & Wilson, 1987). In this context, it is worth noting that waste milk is often fed to calves under practical conditions. Regularly, this non-saleable milk contains a high number of micro-organisms, which reduce the potential storage time and should therefore not be provided to ad libitum fed calves. Several further influences on the initial microbial quality of raw milk have previously been described such as animal and equipment cleanliness, season, feed and cow health (Torkar & Teger, 2008). Hygienic management measures during the milking process, e.g. premilking teat disinfectants or drying teats with paper towels could contribute greatly to a low TBC (Galton et al., 1986). Also, regular cleaning and disinfection of the suckling buckets is an important measure to prevent contamination of the milk. The storage temperature is one of the main influences on microbial proliferation (Spreer, 1998). In this context, our results confirm previous findings that raw milk should not be stored for any time period at temperatures at or above 30 C, whereas storage at temperatures 12

17 below 10 C - 15 C does not result in a TBC increase for h, respectively (Hamosh et al., 1996). At temperatures between 10 C and 40 C, mainly mesophilic bacteria multiply contributing largely to the increase in TBC in untreated milk which was confirmed by our results. However, also during cooling at 3 7 C, some psychrotrophic micro-organisms can grow, e.g. pseudomonads (Sperber et al., 2009). Nonetheless, we recommend cooling down the warm raw milk immediately after the milking process and before filling the suckling buckets of the calves to prevent further multiplication of TBC. Due to the fact that ph of the medium (Spreer, 1998; El-Zubeir & El-Owni, 2009) also has a major influence on microbial growth, different milk preservation methods are commonly recommended. One possibility to extend the storage times and to reduce the risk of the transmission of pathogenic microorganisms to calves is the acidification of milk with feed acids. A reduction of the ph is known to result in an inhibition of bacterial proliferation (Robinson, 2005). Particularly at higher temperatures a reduction and delay in the bacterial growth rate could be demonstrated in our study. Nonetheless, the ph of the acidified milk was between 5.5 and 5.9 and thereby higher than the recommended milk ph of the manufacturer. A lower ph might have fully inhibited the bacterial growth at 20 C and 30 C. However, according to our own subjective experience under practical conditions and according to previous scientific results, a higher concentration of feed acid and thereby a lower ph was commonly accompanied by a lower milk acceptance by the calves (Rindsig & Bodoh, 1977; Otterby et al., 1980). Furthermore, flocculation of milk became visible when higher amounts of the feed acid where used, which was sometimes accompanied by a blockade of the artificial teat and a subsequent stop of the calves regular milk supply. In addition to a reduction in bacterial growth rate at higher temperatures, the acidification of milk or the milk replacer has further advantages such as better fecal consistency scores and fewer days with diarrhoea during the 4 th and 17 th day of life of calves (Jaster et al., 1990). The inoculation of milk with one strain each of S. aureus and E. coli, respectively was performed to provide preliminary information on the proliferation ability of pathogenic bacteria in acidified milk in comparison to raw untreated milk. Both S. aureus and E. coli have relevance as mastitis pathogens and may occur in waste milk fed to calves in high numbers. S. aureus is a Gram-positive coccal bacterium and the cause of a variety of calf and cow diseases which often have to be treated with antimicrobials (Jamali et al., 2015). E. coli as a fast-growing Gram-negative bacterium belongs to the group of coliforms and certain strains are associated with calf diarrhoea (Janke et al., 1990). The two strains used in this study were selected due to their characteristic haemolysis (alpha-haemolysis in E. coli O45, 13

18 alpha-beta-haemolysis in S. aureus Wood 46), which made it possible to easily identify inoculated strains. Jayarao and Henning (2001) demonstrated that 26.7 % of bulk tank milk samples in the US states of South Dakota and Minnesota contained one or more pathogenic bacterial species. The prevalence of S. aureus in North American and European herds ranged from 31% to almost 100% (Oliver et al., 2009; Schlegelova et al., 2002). The major cause of S. aureus in untreated milk is cows displaying mastitis (Henning et al., 2004). Different E. coli strains were detected in approximately 40% of bulk milk samples, whereas E. coli strains pathogenic for humans were identified in % of untreated milk samples depending on geographic location, season, farm size and differences in management and husbandry of dairy cattle (Claeys et al., 2013; Schlegelova et al., 2002). In this study, no bacterial strains with haemolysis were determined in the initial milk sample with a low bacterial count. As expected, high temperatures of >20 C resulted in an increase of the inoculated strains of S. aureus and E. coli after short storage periods. However, the acidification of milk inhibited the multiplication of S. aureus and E. coli at 20 C, which confirms the low acid tolerance of both strains (Millette et al., 2007; Sun et al., 1998). Furthermore, antimicrobial processes such as the lactoperoxidase-thiocyanate-hydrogen peroxide system are activated in raw milk by a low ph value (Reiter et al., 1976). Yet, at temperatures of 30 C the acidification to ph values of approximately 5.5 was not sufficient to fully inhibit growth of both strains. It has to be taken into account that our analyses were performed with only two potential pathogenic bacterial strains. The development of further pathogenic strains in acidified milk warrants additional investigations. The pasteurisation of milk has been described as a further on-farm management measure to reduce initial bacterial count and further proliferation of bacteria (Knappstein et al., 2012; Elizondo-Salazar et al., 2010). In our study, a relatively high TBC after pasteurisation was detected in accordance to previous findings of a large variation in the efficiency of the pasteurisation process (Ruzante et al., 2008). Nonetheless, the decrease in TBC after the pasteurisation process was not as high as reported by the manufacturer and in some samples an increase was detected. Thus, an additional experiment was performed with sterile milk (UHT milk) in order to detect a possible contamination of milk during the heating process: the pasteuriser was first filled with water and the heating process was started. Subsequently, 40 L UHT milk was filled into the pasteuriser and the routine program was started. The microbial analysis of swab samples of various parts of the pasteuriser revealed a high TBC particularly in hose samples, which was confirmed by a clearly visible biofilm in the hose. In this context, it has to be noted that on-farm pasteurisation is typically repeated every day whereas in our 14

19 study the pasteurisation was performed twice per week under laboratory conditions. This circumstance results in longer time periods for bacterial proliferation and may have led to higher rates of recontamination compared to conditions on commercial dairy farms. On the other hand, it can be assumed that in practice the recommendations for cleaning procedures are not always closely followed. Therefore, the recontamination in our study may represent worst case, but realistic conditions. Also, it is well known that thermoduric bacterial species such as Bacillus, Micrococcus, Lactobacillus, Microbacterium, Streptococcus and Enterococcus tolerate the temperatures used in pasteurisation (Richter & Vedamuthu, 2001; Barbano et al., 2006). Particularly psychrotrophic Bacillus species survive the typical pasteurisation conditions in the spore state with subsequent germination and outgrow of the vegetative state and spoilage of milk by degrading enzymes (Meer et al., 1991). However, the pasteurisation of raw milk and subsequent storage at temperatures of 10 C and 20 C resulted in no TBC increase, which might be explained by the optimum growth temperatures of thermoduric species at or above 25 C (Richter & Vedamuthu, 2001). Also, probably only acid-sensitive thermoduric microbial flora developed in the pasteuriser which is explained by the lack of proliferation of bacteria in pasteurized and acidified milk at any temperature and storage time and which warrants further investigations. It can be assumed that the inactivation of vegetative pathogens such as mastitis pathogens is sufficient for practical purposes by proper pasteurisation (Knappstein et al., 2012). Nonetheless, based on our results the manufacturer of the pasteuriser has responded to the problem of a possible re-contamination of milk during the heating process and corrected the cleaning protocol by implementing an acidic cleaning solution. Also, the prevention of re-contamination of the milk after pasteurisation by a correct cleaning and disinfection of the suckling buckets is an important factor for ensuring high milk quality over long storage times (Aust et al., 2012). Under suboptimal practical conditions potential post-pasteurisation pathogen contaminants of milk are coliforms which present a potential risk for foodborne diseases in calves (Aust et al., 2012). This underlines the importance of good hygienic management not only in the milking parlour, but also in the calf barn. Conclusion In conclusion, only milk with a low microbial contamination should be fed to calves. Highquality raw milk can be stored at or below 10 C for up to 12 h without further processing. At temperatures exceeding 10 C, the raw milk provided for ad libitum intake should be treated 15

20 with preservative measures because it represents a potential hazard to the health of calves. Furthermore, heat preventive measures should be considered, such as the avoidance of the buckets exposure to sunlight through the roofing of the calf barn. According to our study, acidification in milk to a ph of approximately 5.5 as the only preservation method is sufficient to reduce the proliferation of bacteria at and above 20 C for up to 9 and at 30 C up to 6 of storage. However, a general conclusion on acidification of milk on the inhibition of pathogenic bacteria cannot be drawn from the results of this study because only two selected strains were tested. Pasteurisation is recommended when milk with a high initial TBC or waste milk is fed to calves, but does not result in the complete inactivation of bacteria. High initial TBC in milk before pasteurisation and post-pasteurisation contamination are risk factors for calf health. Subsequent acidification of pasteurised milk reduces the potential of the further proliferation of bacteria. Because of re-contamination-problems during pasteurisation in our study, further investigations are needed. Acknowledgement The scholarships of Ruegenwalder Muehle and the University of Kiel for the first author are gratefully appreciated. The authors wish to thank the manufacturer of the pasteuriser for providing the equipment. We thank the staff of the Department of Safety and Quality of Milk and Fish Products of the Max-Rubner-Institute, Kiel, Germany, especially Petra Wundram for her support in the laboratory. The laboratory work of the student Lydia Jensen is also gratefully acknowledged. 16

21 Tables & Figures Table 1. Total bacterial count (Log 10 cfu/ml) of untreated and acidified milk depending on time and temperature [LSmean±SE]; abcd significant differences between different periods of storage at the same temperature for untreated and acidified milk, respectively (presented line by line); # significant differences between untreated and acidified milk between the same period of storage at the same temperature Untreated milk Acidified milk 0 h 6 h 9 h 12 h 0 h 6 h 9 h 12 h 10 C 4.75 a 4.53 a 4.54 a 4.71 a 4.75 a 4.53 a 4.44 a 4.63 a 20 C 4.75 a 4.72 a 5.05 a 5.84 b# 4.75 ab 4.62 a 4.74 ab 5.31 b# 30 C 4.75 a 5.75 b# 6.75 c# 7.75 d# 4.75 a 4.98 a# 6.20 b# 7.01 c# 17

22 Figure 1 Graphical presentation of the test scheme used in all 3 test series. 18

23 Figure 2 The course of TBC of untreated and acidified milk depending on storage time and temperature [LSmean]; a significant differences of TBC between 10 C and 20 C, b significant differences of TBC between 10 C and 30 C, c significant differences of TBC between 20 C and 30 C 19

24 Figure 3 The course of pasteurised and pasteurised, acidified milk depending on time and temperature [LSmean]; a significant differences of TBC between 10 C and 20 C, b significant differences of TBC between 10 C and 30 C, c significant differences of TBC between 20 C and 30 [A] [B] 20

25 Figure 4 The course of pasteurised and pasteurised, acidified milk depending on time and temperature [LSmean]; a significant differences of TBC between 10 C and 20 C, b significant differences of TBC between 10 C and 30 C, c significant differences of TBC between 20 C and 30 21

26 References Anonymous (2009) Milk -- Definition and evaluation of the overall accuracy of alternative methods of milk analysis -- Part 1: Analytical attributes of alternative methods. In ISO :2009 (IDF 128-1: 2009). Appleby, M.C., Weary, D.M. & Chua, B. (2001) Performance and feeding behaviour of calves on ad libitum milk from artificial teats. Applied Animal Behaviour Science 74(3), Atlas, R.M. (1984) Microbiology: Fundementals and Applications. New York: MacMillan Publ. Co. Aust, V., Knappstein, K., Kunz, H.J., Kaspar, H., Wallmann, J. & Kaske, M. (2012) Feeding untreated and pasteurized waste milk and bulk milk to calves: effects on calf performance, health status and antibiotic resistance of faecal bacteria. Journal of Animal Physiology and Animal Nutrition 97(6), Bar-Peled, U., Robinzon, B., Maltz, E., Tagari, H., Folman, Y., Bruckental, I., Voet, H., Gacitua, H. & Lehrer, A.R. (1997) Increased Weight Gain and Effects on Production Parameters of Holstein Heifer Calves That Were Allowed to Suckle from Birth to Six Weeks of Age1. Journal of Dairy Science 80(10), Barbano, D.M., Ma, Y. & Santos, M.V. (2006) Influence of Raw Milk Quality on Fluid Milk Shelf Life1,2. Journal of Dairy Science 89, Supplement(0), E15-E19. Booth, I.R., Kroll, R.G. & Gould, G.W. (1989) The preservation of foods by low ph. Mechanisms of action of food preservation procedures., Claeys, W.L., Cardoen, S., Daube, G., De Block, J., Dewettinck, K., Dierick, K., De Zutter, L., Huyghebaert, A., Imberechts, H., Thiange, P., Vandenplas, Y. & Herman, L. (2013) Raw or heated cow milk consumption: Review of risks and benefits. Food Control 31(1), Davidson, P.M. & Taylor, T.M. (2007) Chemical Preservatives and Natural Antimicrobial Compounds. In Food Microbiology: Fundamentals and Frontiers, Third Edition: American Society of Microbiology. El-Zubeir, I.E.M. & El-Owni, O.A.O. (2009) Antimicrobial resistance of bacteria associated with raw milk contaminated by chemical preservatives. World Journal of Dairy & Food Sciences 4(1),

27 Elizondo-Salazar, J.A., Jones, C.M. & Heinrichs, A.J. (2010) Evaluation of calf milk pasteurization systems on 6 Pennsylvania dairy farms. Journal of Dairy Science 93(11), Galton, D.M., Petersson, L.G. & Merrill, W.G. (1986) Effects of Premilking Udder Preparation Practices on Bacterial Counts in Milk and on Teats. Journal of Dairy Science 69(1), Godden, S.M., Fetrow, J.P., Feirtag, J.M., Green, L.R. & Wells, S.J. (2005) Economic analysis of feeding pasteurized nonsaleable milk versus conventional milk replacer to dairy calves. Journal of the American Veterinary Medical Association 226(9), Hamosh, M., Ellis, L.A., Pollock, D.R., Henderson, T.R. & Hamosh, P. (1996) Breastfeeding and the Working Mother: Effect of Time and Temperature of Short-term Storage on Proteolysis, Lipolysis, and Bacterial Growth in Milk. Pediatrics 97(4), Henning, D.R., Flowers, R., Reiser, R., Ryser, E.T. & Frank, J.F. (2004) Pathogens in Milk and Milk Products. In Standard Methods for the Examination of Dairy Products: American Public Health Association. Hepola, H. (2003) Milk feeding systems for dairy calves in groups: effects on feed intake, growth and health. Applied Animal Behaviour Science 80(3), International Organization for Standardization (2013) Microbiology of the food chain - Horizontal method for the enumeration of microorganisms - Part 1: Colony-count at 30 degrees C by the pour plate technique (ISO :2013). Jamali, H., Paydar, M., Radmehr, B., Ismail, S. & Dadrasnia, A. (2015) Prevalence and antimicrobial resistance of Staphylococcus aureus isolated from raw milk and dairy products. Food Control 54(0), Janke, B.H., Francis, D.H., Collins, J.E., Libal, M.C., Zeman, D.H., Johnson, D.D. & Neiger, R.D. (1990) Attaching and effacing Escherichia coli infection as a cause of diarrhea in young calves. Journal of the American Veterinary Medical Association 196(6), Jasper, J. & Weary, D.M. (2002) Effects of Ad Libitum Milk Intake on Dairy Calves. Journal of Dairy Science 85(11),

28 Jaster, E.H., McCoy, G.C., Tomkins, T. & Davis, C.L. (1990) Feeding Acidified or Sweet Milk Replacer to Dairy Calves. Journal of Dairy Science 73(12), Jayarao, B.M. & Henning, D.R. (2001) Prevalence of Foodborne Pathogens in Bulk Tank Milk1. Journal of Dairy Science 84(10), Jeffrey, D.C. & Wilson, J. (1987) Effect of mastitis-related bacteria on total bacterial count of bulk milk supplies. International Journal of Dairy Technology 40(2), Kaske, M., Kunz, H.J. & Koch, A. (2009) Basic essentials for rearing healthy and productive calves. Übersichten zur Tierernährung 37(2/3), Knappstein, K., Aust, V., Kunz, H.-J. & Kaske, M. (2012) Efficiency of two commercial onfarm pasteurizers for inactivation of mastitis pathogens in milk intended for feeding of calves. Berliner und Munchener tierarztliche Wochenschrift 126(1-2), Meer, R.R., Baker, J., Bodyfelt, F.W. & Griffiths, M.W. (1991) Psychrotrophic Bacillus spp. in Fluid Milk Products: A Review. Journal of Food Protection 54(12), Millette, M., Luquet, F.M. & Lacroix, M. (2007) In vitro growth control of selected pathogens by Lactobacillus acidophilus- and Lactobacillus casei-fermented milk. Letters in Applied Microbiology 44(3), Muir, D.D. (1996) The shelf-life of dairy products: 2. Raw milk and fresh products. International Journal of Dairy Technology 49(2), Oliver, S.P., Boor, K.J., Murphy, S.C. & Murinda, S.E. (2009) Food safety hazards associated with consumption of raw milk. Foodborne pathogens and disease 6(7), Otterby, D.E., Johnson, D.G., Foley, J.A., Tomsche, D.S., Lundquist, R.G. & Hanson, P.J. (1980) Fermented or Chemically-Treated Colostrum and Nonsalable Milk in Feeding Programs for Calves1,2. Journal of Dairy Science 63(6), Phillips, J.D. & Griffiths, M.W. (1987) The relation between temperature and growth of bacteria in dairy products. Food Microbiology 4(2), Reiter, B., Marshall, V.M., Björck, L. & Rosén, C.G. (1976) Nonspecific bactericidal activity of the lactoperoxidases-thiocyanate-hydrogen peroxide system of milk against Escherichia coli and some gram-negative pathogens. Infection and Immunity 13(3),

29 Richter, R.L. & Vedamuthu, E.R. (2001) Milk and milk products. Compendium of the methods for the microbiological examination of foods. 4th ed. Washington: APHA, Rindsig, R.B. & Bodoh, G.W. (1977) Growth of Calves Fed Colostrum Naturally Fermented, or Preserved with Propionic Acid or Formaldehyde1,2. Journal of Dairy Science 60(1), Robinson, R.K. (2005) Dairy microbiology handbook: the microbiology of milk and milk products. John Wiley & Sons. Ruzante, J.M., Gardner, I.A., Cullor, J.S., Smith, W.L., Kirk, J.H. & Adaska, J.M. (2008) Isolation of Mycobacterium avium subsp. paratuberculosis from Waste Milk Delivered to California Calf Ranches. Foodborne Pathogens and Disease 5(5), Schlegelova, J., Babak, V., Klimova, E., Lukasova, J., Navratilova, P., Sustackova, A., Sediva, I. & Rysanek, D. (2002) Prevalence of and Resistance to Anti-Microbial Drugs in Selected Microbial Species Isolated from Bulk Milk Samples. Journal of Veterinary Medicine, Series B 49(5), Sperber, W.H., Doyle, M.P., Ledenbach, L.H. & Marshall, R.T. (2009) Microbiological Spoilage of Dairy Products. In Compendium of the Microbiological Spoilage of Foods and Beverages, pp Springer New York. Spreer, E. (1998) Milk and Dairy Product Technology. Taylor & Francis. Stewart, S., Godden, S., Bey, R., Rapnicki, P., Fetrow, J., Farnsworth, R., Scanlon, M., Arnold, Y., Clow, L., Mueller, K. & Ferrouillet, C. (2005) Preventing Bacterial Contamination and Proliferation During the Harvest, Storage, and Feeding of Fresh Bovine Colostrum. Journal of Dairy Science 88(7), Sun, C.Q., O'Connor, C.J., Turner, S.J., Lewis, G.D., Stanley, R.A. & Roberton, A.M. (1998) The effect of ph on the inhibition of bacterial growth by physiological concentrations of butyric acid: Implications for neonates fed on suckled milk. Chemico-Biological Interactions 113(2), Torkar, K.G. & Teger, S.G. (2008) The microbiological quality of raw milk after introducing the two day's milk collecting system. Acta Agri Slovenica 92(1),

30 Wiedemann, S., Holz, P., Kunz, H.-J., Stamer, E. & Kaske, M. (2015) Effect of ad libitum feeding of Holstein-Friesian calves during the first four weeks of life on weight development as well as milk yield and feed intake during first lactation. Züchtungskunde 87(6): Wouters, J.T.M., Ayad, E.H.E., Hugenholtz, J. & Smit, G. (2002) Microbes from raw milk for fermented dairy products. International Dairy Journal 12(2-3),

31 CHAPTER TWO Effect of an ad libitum milk supply during the first three weeks of life of dairy calves on heart rate and heart rate variability during feeding and rehousing L. Prokop a, G. Hoffmann b, M. Kaske c, S. Wiedemann a, * a Animal Health, Institute of Animal Breeding and Husbandry, University of Kiel, Olshausenstraße 40, Kiel, Germany, lprokop@tierzucht.uni-kiel.de, swiedemann@tierzucht.uni-kiel.de b Leibniz Institute for Agricultural Engineering Potsdam-Bornim, Department of Engineering for Livestock Management, Max-Eyth-Allee 100, Potsdam, Germany, ghoffmann@atb-potsdam.de c Department for Farm Animals, Vetsuisse Faculty, University of Zurich, Winterthurer Strasse 260, 8057 Zurich, Switzerland, mkaske@vetclinics.uzh.ch * Corresponding author: Tel ; swiedemann@tierzucht.unikiel.de 27

32 Abstract The objective of this study was to investigate the effect of nutrient supply during early postnatal life on heart rate (HR) and heart rate variability (HRV) as indicators of the cardiac stress response. Male and female calves were fed either a restrictive milk allowance twice per day (6 L/d; RES; N = 28) or an unlimited amount of milk (ad libitum; ADL; N=28) during the first three weeks of life. All calves were housed in individual straw bedded hutches from d 1 to 23 of life and were moved to a group pen on d 23 ± 2 of life. Starting at least one day before rehousing until one hour after the rehousing process HR and HRV variables in the time- and frequency domain were measured continuously using a portable recording system. To study the cardiac response to the feeding process 6 time windows of 5 min each were chosen: resting time at a.m., start of personnel activity in the barn, 15 min before feeding, during feeding, 15 min after feeding and 1 h after feeding. For the evaluation of cardiac response to an unknown stressor such as rehousing 4 time windows of 5 min each were selected: resting time at a.m., during rehousing, 30 min after rehousing and 1 h after rehousing. During resting as well as before feeding and rehousing HR was higher in ADL calves compared with RES calves. During feeding and rehousing HR reached peak values which were comparable in both groups. HRV variables of the time- and frequency domain indicated a shift towards a sympathetic dominance in the balance of the autonomic nervous system during feeding time particularly in RES calves. Differences between resting and feeding values were demonstrated in RES calves at low frequency and high frequency power, whereas no differences were observed in ADL calves. The cardiac response of calves to rehousing was inconsistent in both groups. An increase in RMSSD and SD1 in ADL calves indicated that the vagal component in the vegetative neurological control was increased in these calves during rehousing. In respect to all measured parameters, it can be concluded that the sympathetic and vagal nervous system in calves mainly remained in balance during feeding and rehousing. Keywords: heart rate, heart rate variability, stress, ad libitum feeding, calves 28

33 Introduction During the last decade, numerous studies in calves have investigated the effects of an enhanced nutrient supply during the early postnatal period in comparison to the conventional restrictive allowance of milk or milk replacer (typically 10 % of body weight/d). It has been demonstrated that an ad libitum supply of milk in individual pens resulted in milk consumption up to 25 % of body weight which is comparable to the milk intake of calves left with its dam which suckle their mother 7-10 times a day (Appleby et al., 2001; Jasper and Weary, 2002). The higher milk intakes are accompanied by higher weight gains during the first weeks of life (Maccari et al., 2014). Furthermore, a higher nutrient supply and an unlimited access to milk is associated with positive or with no adverse effects on animals health as well as with positive effects on animals behavior (Khan et al., 2011b; Maccari et al., 2014). At present, it is unclear whether the conventional restrictive feeding of calves results in an increased stress response of the autonomic nervous system particularly before, during and shortly after the time of feeding in which conventionally fed calves could face restlessness caused by hunger and an unsatisfied sucking desire. It has been reported that for pre-weaned calves nutritive sucking of milk is more reinforcing then non-nutritive sucking (Hammell et al., 1988). Furthermore, it is not known if the nutrient supply during the first weeks of life has an effect on the stress reaction to unknown events and to the short-term adaptation of calves to new environments. One of a potential stress inducing event is the rehousing of calves from individual hutches into group pens which is commonly performed after a few weeks of life. During rehousing calves are in close contact to humans and they need to adapt to a new environment and new group mates which may result in social stress. A successful method to measure the autonomic nervous system in calves is the investigation of the heart rate variability (HRV) because detailed and objective conclusions about the level of stress can be drawn immediately (Mohr et al., 2002; von Borell et al., 2007). In comparison with the analysis of the stress reaction of the hypothalamic-pituitary-adrenocortical axis no human contact is needed during the time of measurement. The HRV shows the cardiovascular autonomic regulation which allows to draw conclusions on the balance between the sympathetic and the parasympathetic autonomic nervous system (Jong et al., 2000). The autonomic nervous system dynamically controls the response of the body to a range of external and internal stimuli to provide stability within a physiological range (Pumprla et al., 2002). Therefore, HRV measurements can be applied to determine cardiac stress from physical, pathological and emotional origins (Mohr et al., 2002; von Borell et al., 2007). To 29

34 measure the stress response in calves, HRV i.e. continuously changing temporal distance between succeeding heartbeats (Mohr et al., 2002) was used. It was the objective of the study to investigate HRV before, during and after a) time of feeding and b) rehousing and regrouping to clarify the effect of feeding intensity of calves during the early postnatal period. Material and Methods Calves, Management and Feeding The experiment was set up and performed according to strict federal and international guidelines on animal research (AZ ). The study was carried out at the research farm Karkendamm of the University of Kiel (Bimöhlen, Germany, 190 cows, mean lactation milk yield 11,500 kg). Male (N = 28) and female (N = 28) Holstein Friesian calves which were born between November 2014 and February 2015 were enrolled in the study. All calves were separated from their dam and moved into individual straw-bedded hutches shortly after birth. Within the first hours of life all calves were fed 3 L colostrum from their dam. Thereafter, all calves were randomly assigned to either the ad libitum fed group (ADL, N = 28) or to the restrictively fed group (RES, N = 28). The gender ratio was equal between both groups. After the supply of colostrum ADL calves received 6 to 9 L milk twice a day (07.00 a.m. and p.m.) in nipple buckets. During the morning feeding the rest milk in the nipple bucket was filled up to 6 to 9 L in total. In the evening residual amounts were discarded, buckets were washed and refilled. During the first 4 d of life calves received transition milk. From d 5 to d 23 of life a mixture of pasteurized waste milk (Milk Taxi 150 L Pasteurizer; Holm & Laue, Westerrönfeld, Germany) and surplus transition milk was offered. A supplement (Quota Mix, Sprayfo, Sloten GmbH, Diepholz, Germany; CP 11.4 %, crude fat 1 %, CF 0 %, crude ash 8 %; 100 g/l milk) and an acid (AcidFit, Sprayfo, Sloten GmbH, Diepholz, Germany; 2.5 g/l milk) was added to the milk mixture. The ingested volumes of milk were recorded twice daily for each calf. After the colostrum feeding the RES calves were offered 3 L transition milk twice per day (07.00 a.m. and p.m.) in nipple buckets. From d 5 to d 23 of life 3 L of the 30

35 supplemented and acidified waste and transition milk mixture which was offered to the ADL calves was fed. ADL and RES calves were moved into straw-bedded group pens on d 23 ± 2 of life. All calves were led by ropes from the individual hutch to the group pen by a familiar person (distance approximately 200 m, average duration 6 min). Typically one calf was moved at a time. If a new pen was to be used two calves were rehoused in parallel. Maximum number of animals per group was 5. At all times, all calves had free access to water and calf starter. Body weight The body weight of each calf was recorded after birth and right before rehousing. Calves were guided on a halter over the scale, which was located directly on the way out of the calf barn with individual hutches. Heart rate variability Heart rate (HR) and heart rate variability (HRV) were recorded with a portable recording system (emotion, Mega Electronics, Kuopio, Finland) attached with two self-adhesive skin electrodes (Ambu Blue Sensor VL, Ballerup, Denmark) to the left side of the calves. The electrodes were positioned next to the heart in height of the elbow and approximately one hand under the withers. To ensure a strong attachment of the electrodes to the skin, the skin was firstly sheared and cleaned with a dry towel and an adhesive spray (Henry Schein, Hamburg, Germany) was applied. The two electrodes were connected via cable to the recording device, which stored the data of the individual calf. A belt was fixed around the thorax of each calf to secure the position and to protect the electrodes against chewing by other calves after rehousing. The recordings of data were started at least one day before rehousing to allow the calves to adapt to the electrodes and the belt. One hour after rehousing the electrodes and the belt were removed to avoid later chewing by other calves due to increasing exploratory behavior. Cardiac response to feeding was analyzed during morning feeding on the day of rehousing. Data from the recording devices were transferred via USB connection to a computer (software emotion LAB, version , Mega Electronics, Kuopio, Finland). HR and HRV were analyzed with the Kubios HRV software (version 2.1, Biomedical Analysis and Medical Imaging Group, Department of Applied Physics, University of Eastern Finland, Kuopio, Finland). Repeated short-term 5-min windows were analyzed (Camm et al., 31

36 1996). To evaluate the stress around feeding, six time windows were selected: during resting at a.m., during time when light was switched on in the calf barn at approximately a.m., 15 min before feeding in the morning, during feeding at approximately a.m., 15 min after feeding and one hour after feeding. For the evaluation of the stress during rehousing four time windows were chosen: during resting at a.m., during the rehousing process, 30 min after rehousing and one hour after rehousing. Data were detrended (detrending method: smooth priors) and an artifact correction (level: 300 ms) was conducted. HRV variables in the time- and frequency-domain were analyzed: HR, square root of variance of all R-R intervals (SDNN), root mean square of successive interbeat interval differences (RMSSD) as well as low frequency power (LF; 0.04 to 0.3 Hz), high frequency power (HF; 0.3 to 0.8 Hz) and the LF:HF ratio, whereby LF and HF were calculated by an autoregressive model and in normalized units (nu). The limit of the frequency ranges in calves were adopted from Mohr et al. (2002). Furthermore, short term variability (SD 1) of Poincaré Plot was analyzed. Statistical evaluation All statistical analyses were performed using SAS (Version 9.3, SAS Institute Inc., Cary, NC, USA). To estimate differences between groups the GLIMMIX procedure in SAS was applied. For analyses of milk intake and body weights the group, sex and living day were included as fixed effects and calf was considered as random effect. For HR and HRV analyses the group, sex, time window and the interaction of group x time window were included as fixed effects. The calf was considered as random effect. The adjusted means were compared with the Tukey-Kramer-Test. Probability values less than 0.05 were considered statistically significant. Data are presented as LS-means (± standard error). Results Milk intake and body weight Until the third day of life no difference in colostrum and transition milk intake was observed between both groups (Figure 1). Between d 4 and 7 of life ADL calves ingested more milk in comparison with RES calves, but no differences were observed between d 8 and 10 of life. From d 10 to d 21 of life ADL calves displayed higher daily intakes (P < 0.05 for each day). 32

37 One day before rehousing the milk intake was higher in ADL calves (9.3 ± 0.39 L milk/d) in comparison with RES calves (5.8 ± 0.4 L milk/d; P < 0.05). Total milk consumption during the first three weeks of life was almost 35 % higher in ADL calves (162 vs. 120 L milk/d; P < 0.05). In ADL calves the sex of the animals had an effect on milk intake. Male calves (8.4 ± 0.12 L milk/d) displayed a higher daily milk intake compared with female calves (7.0 ± 0.13 L milk/d; P < 0.05). In RES calves milk intake did not differ between male and female calves (5.7 ± 0.11 L milk/d vs. 5.7 ± 0.11 L milk/d; P = ). Birth weights did not differ between ADL and RES calves (42.3 ± 1.3 kg vs ± 1.2 kg, P = 1.0). Body weight in both groups increased during the first 23 days of life, but tended to be higher in ADL calves in comparison to RES calves (58.2 ± 1.3 kg vs ± 1.2 kg, P = 0.09). In total, the ADL calves displayed a higher average daily gain between birth and d 23 of life (754 ± 55 g vs. 544 ± 52 g; P < 0.05). No differences were observed in average weights and daily weight gains between male and female animals. HRV before, during and after feeding The time window of measurement and the group significantly affected HR (both P < 0.05). The resting HR at a.m. in ADL calves was higher in comparison to RES calves (P < 0.05; Table 1). During times of activity of the stall personnel such as after turning on the light and during preparation of the milk 15 min before feeding the HR increased in both groups, but HR was still higher in the ADL group (each P < 0.05). At feeding, HR did not differ between groups (P = 1.0). The HR increased from resting to feeding 25 % in ADL calves and 59 % in RES calves (both P < 0.05). After feeding HR decreased in both groups and was comparable to resting values. The P-values obtained for SDNN and RMSSD did not differ between the six time windows and both groups. One hour after feeding RES calves displayed higher SDNN and RMSSD values compared with values during feeding (both P < 0.05). LF norm and HF norm were affected by the time window and the interaction of group and time window (P < 0.05). LF norm was higher in ADL calves (76 ± 4.3 nu) compared with RES calves (61 ± 4.3 nu) at a.m. (P < 0.05). Thereafter, LF norm increased during feeding in RES calves in comparison to the resting value by 28 % (78 ± 4.1 nu; P < 0.05). The resting HF norm at a.m. was lower in ADL calves (24 ± 4.3 nu) compared with RES calves (39 ± 4.3 nu; P < 0.05). HF norm decreased until feeding by 43 % in RES calves (22 ± 4.1 nu; P < 0.05). Values after feeding were comparable to resting values. Turning on the light in the calf barn particularly affected the LF:HF ratio in ADL calves which increased in comparison to resting values and which resulted in higher LF:HF ratios in comparison with the values of RES calves 33

38 (P < 0.05). After light was switched on LF:HF decreased in ADL calves continuously until one hour after feeding. The parameter of the nonlinear domain SD1 displayed no differences between groups and time windows. SD1 decreased from a.m. until feeding in RES calves and increased thereafter to the resting value. HRV during rehousing The time window of measurement influenced all analyzed parameters of the rehousing process (each P < 0.05). Furthermore the group affected HR and RMSSD as well as SD1 differed between male and female animals (each P < 0.05). The results of all measured parameters of HRV at a.m. were comparable between rehousing and feeding because it was the same day of measurement (Table 1 and Table 2). HR in ADL calves was higher compared with HR of RES calves before rehousing and 30 min after rehousing (both P < 0.05; Table 2). HR increased in both groups until rehousing and decreased thereafter. One hour after rehousing the HR values in ADL calves were comparable to values one hour before rehousing, but in RES calves values were still higher. The parameter SDNN differed not between groups at any time window. From resting to one hour before rehousing SDNN increased in RES calves (+394%, P < 0.05) and decreased thereafter. RMSSD values were 89% higher in ADL calves during rehousing compared with resting values (P < 0.05). In RES calves rehousing and adaptation to the group pen resulted in lower RMSSD values in comparison to values during the rehousing process. LF norm in RES calves increased by 28 % after a 30 min period in the group pen in comparison to resting values (P < 0.05). The increase in LF norm was reflected by a 28% decrease in HF norm from resting to 30 min after rehousing in RES calves. The LF:HF ratio differed between both groups one hour before rehousing (53 ± 10 vs. 15 ± 9.5 for ADL and RES calves, respectively; P < 0.05). At a.m. LF:HF was lower in ADL calves, but not in RES calves in comparison to values before rehousing. Thereafter, values in ADL calves decreased to levels which were comparable to resting values. The parameter of the nonlinear domain SD1 displayed no group differences at any time window. In ADL calves SD1 increased from resting until rehousing and decreased thereafter. Discussion This study was performed to investigate the effect of a restrictive feeding protocol in comparison to an ad libitum supply of whole milk during the neonatal period from birth to 34

39 week 3 of life on the stress response of dairy calves. The feeding intensity of the RES calves was based on the still often recommended standard feeding protocol for rearing of dairy and fattening cattle in Germany which is comparable to that in the US (10% of BW per day). The mean uptake of acidified milk by the ADL calves of approximately 8 L per day is in accordance to previously reported amounts (Appleby et al., 2001; Jasper and Weary, 2002). On the other hand, in female and male calves also slightly higher values of approximately 11 L and 10 L per day were measured, respectively (Maccari et al., 2014; von Keyserlingk et al., 2006). In our study milk intake in ADL calves between d 8 and 11 of life decreased to comparable amounts of RES animals. One possible explanation might be the late change from sweet transition milk to acidified milk on d 5 of life which subjectively caused acceptance problems by calves as was also reported by farmers applying the ad libitum feeding protocol. Therefore, it is now frequently recommended to feed acidified milk after the first colostrum supply to minimize the reduction in milk intake. Nonetheless, before the measurement of HR and HRV milk intake was considerably higher in ADL calves for more than 10 d. Hence, results on the HRV measurement can be associated with the feeding intensity. On dairy farms calves are typically exposed to various potential environmental and management related stressors. The autonomous cardiac reaction of the individual animal to such stressors depends on various factors such as on previous positive or negative experiences and on the level of psycho-physiological and mental stress (Rietmann et al., 2004; von Borell et al., 2007). Also, the supply of nutrition has an effect on the autonomic cardiac response of humans and animals. In healthy female young women it has been reported that fasting from 12 h up to 48 h is linked with changes in HRV parameters reflected by a lower HR and a lower LF value as well as by a higher HF value (Mazurak et al., 2013; Ohara et al., 2015). These results are in accordance with our findings of a lower HR and a decreased LF:HF ratio in RES calves during resting. Furthermore, in cattle, young growing goats and immature broiler chicken a higher resting HR in ad libitum fed animals in comparison to restrictively fed animals were reported (Brosh et al., 2002; Puchala et al., 2009; Savory et al., 2006). Because heart rate measurements can also be used to estimate the energy status of farm animals a higher energy expenditure in ADL calves can be assumed (Brosh, 2007). The influence of recording body weight directly before rehousing on stress load during rehousing cannot be excluded. But weighting conditions were equal for all calves. It was a minimal effort for calves, to hold a few seconds on the scale which were located directly on 35

40 the way out of the calf barn. It can be concluded that the influence of weighing on the stress response of a calf is negligible. In the present studies typical potential stress inducing scenarios were exemplified by the time of feeding and the rehousing of animals. Before the time of feeding restrictively fed animals are typically seen to be restless and are heard to vocalize as soon as the farm personnel enter the barn. In agreement with our subjective observation it was reported that calves with limited access to milk usually drink all offered milk within a short period of time without reaching satiety (Hammon et al., 2002a). Non-nutritive sucking of subjects or objects is therefore seen more often in restrictively fed compared with more intensively fed calves (Hammell et al., 1988; Jensen, 2003). In this study ADL calves were also standing shortly before time of feeding, but no vocalization was observed. It is possible that the restlessness of the RES calves transferred to the ADL animals. Rehousing of young animals from individual hutches into a group pen is also very typical in established calf rearing. In our study being haltered and walked by one person was unfamiliar for all calves. After rehousing an increase in the general activity was observed as was previously described (Veissier et al., 2001). Animals also soon started to sniff and explore each other. Therefore, the HRV could only be analyzed for one hour after rehousing in the unobserved animals. During the experimental period the HRV recording system was left at the calves body over several days without any problems in compliance. Because activity of the calves in their hutches caused movement of the electrodes in some cases only 5 min periods with regular recordings were analyzed. The HR and HRV values in our calves were largely comparable to previous reported values (Mohr et al., 2002) and indicated an induction of stress by both tested scenarios in a similar pattern. To evaluate the resting values video recordings of 15 animals of each group were analyzed. At a.m. all animals were lying and activity of the other calves at this time period was not expected, but cannot be ruled out. Interestingly, also in ADL calves an increase in HR was observed during feeding. The experiment was carried out during wintertime and according to our observation with increasing age calves drank increasing amounts of milk within the first hour after feeding. Calves seemed to prefer the warm milk and learned to wait for the next feeding period. It has to be noted, that the large quantities of milk per feeding were not associated with an increased prevalence of diarrhea in ADL calves. However, the relative increase in HR during feeding and rehousing was higher in RES calves in comparison to ADL calves. An increase in HR 36

41 may result from an increased sympathetic activity, a reduced vagal activity or a combination of changes within both branches, but the complex interplay between the branches make an understanding of the underlying control mechanisms difficult (Malik, 1996; McCraty et al., 1996; von Borell et al., 2007). Next to nutritional factors external non-linear effects on HR are the workload as was shown in horses or the level of physical activity (Hagen et al., 2005; Persson, 1983; von Borell et al., 2007). In our study, differences in individual HR and HRV parameters at time windows after resting could therefore also be associated to differences in the posture during measurements, e.g. to standing or lying of the animals particularly before feeding and rehousing (Maros et al., 2008; Palestrini et al., 2005). Also, the speed of the individual calves during the walk of the 200 m distance throughout the rehousing process differed which might be reflected by a variance within the HR. The stress response to feeding activity reflected by HR was only short because one hour after fresh milk supply the HR was comparable to resting values. In contrast to these findings, it can be assumed that the adaptation of the calves to rehousing took more than one hour because HR was still higher at that time in comparison to resting values. Because HR is influenced by various physiological loops and enables assessment only of short-term effects on animals parameters of HRV were also included in the analyses to draw conclusions about long-term effects of stress (Hansen, 2000). The evaluation of HRV variables in the time-domain revealed inconsistent results. Long-term variability (SDNN) was significantly lower in RES calves during feeding than one hour after feeding and showed no significant differences in both groups between initial measurement, during rehousing and one hour after rehousing, respectively. The short-term variability (RMSSD) and long-term variability (SDNN) are negatively correlated with the level of stress (Mohr et al., 2002) which is in agreement with the finding of the lowest values during feeding in RES calves and after rehousing in both groups (Mohr et al., 2002). Low RMSSD and SDNN values are indicators of a reduced parasympathetic activity (Mohr et al., 2002; Visser et al., 2002). During rehousing an unexpected increase of the RMSSD value was observed in ADL calves which reflects alterations in the autonomic nervous system that are predominantly vagally mediated, but values decrease shortly after (Malik and Camm, 1995; Sztajzel, 2004). Parameters of HRV of the frequency-domain are used to assign the power in different bands to underline different physiological functions (von Borell et al., 2007). In the past the LF oscillation has often been regarded as a reflection of sympathetic activity (Pagani et al., 1986). Yet, recent studies have shown that a change in the 37

42 LF band results from the interaction of the sympathetic and vagal nervous system (Houle and Billman, 1999; Malliani et al., 1991; Öri et al., 1992). An increase of the LF component as was seen in RES claves during feeding implies a shift of the sympatho-vagal balance towards the sympathetic nervous system and therefore indicates high mental stress (Sztajzel, 2004). The HF component of the heart rate power spectrum which is identical to the LF with reversed signs represents the vagal activity and is highly correlated with RMSSD (Camm et al., 1996; Friedman and Thayer, 1998; Sztajzel, 2004; von Borell et al., 2007). The vagal tone was reduced in RES calves during feeding as expressed in a decrease of the HF component. The ratio of the LF component and the HF component reflects the sympatho-vagal balance (Sztajzel, 2004). No differences were observed in LF:HF in both groups between resting, during and after rehousing or during feeding, respectively. Remarkable are the increasing values of LF:HF in both groups when light was switched on in the calf barn and one hour before rehousing indicating a shift of the sympatho-vagal balance in the direction of the sympathetic nervous system. The reported influence of thermoregulation on LF:HF (Mohr et al., 2002) has no relevance for the results in this study because all calves were borne in the winter months within a short period of time. The parameter SD1 represents short-term variability of the parasympathetic activity comparable to RMSSD (von Borell et al., 2007). Therefore, the high values of SD1 in ADL calves during rehousing support the surprising results on RMSSD which indicated an increased vagal tone during the rehousing process. The lower values of SD1 after feeding in RES calves imply a decrease in stress load in comparison to the time of feeding. Taken together, HR and the analyzed parameters of time- and frequency-domain indicate an increased stress level during time of feeding evident from a shift towards sympathetic activity in RES calves. This result supports the subjective visual observation of restlessness in these calves during that time. During rehousing the results of all measured parameters is not conclusive. Particularly the unexpected outcomes in some of the HRV parameters which indicated a marked increase in vagal activity in ADL calves during the rehousing process warrant further investigations. It has to be noted that in calves contrary to results in humans the comparison of HRV values between different studies is limited in validity because of missing standard procedures of analysis (Camm et al., 1996). Future research could also include determination of plasma catecholamine concentrations, to receive additional information about the activation of the sympathetic nervous system (Sgoifo et al., 1997). Also non-linear parameters of HRV should 38

43 be analyzed which distinguish between quantitative different levels of stress, whereas linear parameter of HRV are useful to separate qualitative different levels of stress (Mohr et al., 2002). 39

44 Tables & Figures Figure 1 Average daily milk intake [LS-means ± SE] of calves fed ad libitum (ADL, black columns) and restrictively (RES, grey columns) during the first three weeks of life; * P<0.05; ** P<0.001; *** P<

45 Table 1. Changes in heart rate (HR) and heart rate variability (HRV) parameters in ad libitum fed (ADL) and restrictively fed (RES) calves at different time windows of measurement in relation to morning feeding on d 23 of life; abcd significant differences (P < 0.05) of one parameter at different times of sampling within one group (line by line); bold numbers significant differences (P < 0.05) of one parameter and at one time of window between groups; % indicates the increase/decrease of the previous LS mean value to the basal measurement at a.m. resting value activity of personnel time window in relation to feed supply a.m. light on 15 min before during feeding 15 min after 1 h after HRV parameter group LS mean LS mean % LS mean % LS mean % LS mean % LS mean % HR (bpm) ADL 120 b 131 b ab a ab b -1 RES 96 c 108 bc bc a b c 8 SDNN (ms) ADL 41.4 a 46.7 a a a a a -4 RES 35.3 ab 41.4 ab ab b ab a 36 RMSSD (ms) ADL 40.2 a 39.1 a a a a a 3 RES 40.6 ab 37.5 ab ab b ab a 29 LF norm (nu) ADL 76.3 ab 81.9 a ab ab ab b -14 RES 61.1 b 76.6 a ab a ab ab 11 HF norm (nu) ADL 23.7 ab 18.1 b ab ab ab a 44 RES 38.8 a 23.4 b ab b ab ab -17 LF:HF ADL 8.0 b 28.1 a b b b b -69 RES 2.0 a 7.6 a a a a a 190 SD1 ADL 28.5 a 27.7 a a a a a 3 RES 28.7 ab 26.5 ab ab b ab a 29 41

46 Table 2. Changes in heart rate (HR) and heart rate variability (HRV) parameters in ad libitum fed (ADL) and restrictively fed (RES) calves at different time windows of measurement in relation to rehousing on d 23 of life; abcd significant differences (P < 0.05) of one parameter at different times of sampling within one group (line by line); bold numbers significant differences (P < 0.05) of one parameter and at one time of sample between the groups, % indicates the increase/decrease of the previous LS mean value to the basal measurement at a.m. resting value time window in relation to rehousing a.m. 1 h before during 30 min after 1 h after HRV parameter group LS mean LS mean % LS mean % LS mean % LS mean % HR (bpm) ADL 122 d 127 cd a b cb 18 RES 98 c 111 c a b b 32 SDNN (ms) ADL 29.3 a 92.6 a a a a 20 RES 30.6 b a ab b ab 91 RMSSD (ms) ADL 31.3 b 50.7 ab a ab b -16 RES 40.0 abc 56.6 ab a c bc -14 LF norm (nu) ADL 76.0 a 70.5 a a a a 1 RES 59.9 b 71.3 ab ab a ab 26 HF norm (nu) ADL 24.0 a 29.4 a a a a -2 RES 40.1 a 28.7 ab ab b ab -38 LF:HF ADL 7.0 b 53.1 a b b b 83 RES 1.5 a 15.4 a a a a 1093 SD1 ADL 22.1 b 35.9 ab a ab b -16 RES 28.3 ab 40.3 a a b ab

47 References Albright, J. C., and C. Arave Behavioural responses to management systems. The Behaviour of Cattle: Anonymous Milk -- definition and evaluation of the overall accuracy of alternative methods of milk analysis -- part 1: Analytical attributes of alternative methods. Appleby, M. C., D. M. Weary, and B. Chua Performance and feeding behaviour of calves on ad libitum milk from artificial teats. Applied Animal Behaviour Science 74: Atlas, R. M Microbiology: Fundementals and applications. MacMillan Publ. Co., New York. Aust, V. et al Feeding untreated and pasteurized waste milk and bulk milk to calves: Effects on calf performance, health status and antibiotic resistance of faecal bacteria. Journal of Animal Physiology and Animal Nutrition 97: Bar-Peled, U. et al Increased weight gain and effects on production parameters of holstein heifer calves that were allowed to suckle from birth to six weeks of age1. Journal of Dairy Science 80: Barbano, D. M., Y. Ma, and M. V. Santos Influence of raw milk quality on fluid milk shelf life1,2. Journal of Dairy Science 89, Supplement: E15-E19. Booth, I. R., R. G. Kroll, and G. W. Gould The preservation of foods by low ph. Mechanisms of action of food preservation procedures.: Brosh, A Heart rate measurements as an index of energy expenditure and energy balance in ruminants: A review. Journal of Animal Science 85: Brosh, A., Y. Aharoni, and Z. Holzer Energy expenditure estimation from heart rate: Validation by long-term energy balance measurement in cows. Livestock Production Science 77: Cady, R. A., and R. R. Smith Economics of heifer raising programs. Proceedings from the Calves, Heifers, and Dairy Profitability National Conference. p 7. Northeast Reg. Ag. Eng. Serv. NRAES-74, Harrisburg, PA. 43

48 Camm, A. J. et al Task force of the european society of cardiology and the north american society of pacing and electrophysiology. Heart rate variability: Standards of measurement, physiological interpretation and clinical use. Circulation 93: Claeys, W. L. et al Raw or heated cow milk consumption: Review of risks and benefits. Food Control 31: Davidson, P. M., and T. M. Taylor Chemical preservatives and natural antimicrobial compounds Food microbiology: Fundamentals and frontiers, third edition. American Society of Microbiology. de Passillé, A. M. B., J. H. M. Metz, P. Mekking, and P. R. Wiepkema Does drinking milk stimulate sucking in young calves? Applied Animal Behaviour Science 34: Drackley, J. K Calf nutrition from birth to breeding. Veterinary Clinics of North America: Food Animal Practice 24: El-Zubeir, I. E. M., and O. A. O. El-Owni Antimicrobial resistance of bacteria associated with raw milk contaminated by chemical preservatives. World Journal of Dairy & Food Sciences 4: Elizondo-Salazar, J. A., C. M. Jones, and A. J. Heinrichs Evaluation of calf milk pasteurization systems on 6 pennsylvania dairy farms. Journal of Dairy Science 93: Flower, F. C., and D. M. Weary Effects of early separation on the dairy cow and calf: 2. Separation at 1 day and 2 weeks after birth. Applied Animal Behaviour Science 70: Friedman, B. H., and J. F. Thayer Autonomic balance revisited: Panic anxiety and heart rate variability. Journal of Psychosomatic Research 44: Galton, D. M., L. G. Petersson, and W. G. Merrill Effects of premilking udder preparation practices on bacterial counts in milk and on teats. Journal of Dairy Science 69: Godden, S. M., J. P. Fetrow, J. M. Feirtag, L. R. Green, and S. J. Wells Economic analysis of feeding pasteurized nonsaleable milk versus conventional milk replacer to dairy calves. Journal of the American Veterinary Medical Association 226:

49 Hagen, K., J. Langbein, C. Schmied, D. Lexer, and S. Waiblinger Heart rate variability in dairy cows - influences of breed and milking system. Physiology & Behavior 85: Hammell, K. L., J. H. M. Metz, and P. Mekking Sucking behaviour of dairy calves fed milk ad libitum by bucket or teat. Applied Animal Behaviour Science 20: Hammon, H. M., G. Schiessler, A. Nussbaum, and J. W. Blum. 2002a. Feed intake patterns, growth performance, and metabolic and endocrine traits in calves fed unlimited amounts of colostrum and milk by automate, starting in the neonatal period. Journal of Dairy Science 85: Hammon, H. M., G. Schiessler, A. Nussbaum, and J. W. Blum. 2002b. Feed intake patterns, growth performance, and metabolic and endocrine traits in calves fed unlimited amounts of colostrum and milk by automate, starting in the neonatal period1. Journal of Dairy Science 85: Hamosh, M., L. A. Ellis, D. R. Pollock, T. R. Henderson, and P. Hamosh Breastfeeding and the working mother: Effect of time and temperature of short-term storage on proteolysis, lipolysis, and bacterial growth in milk. Pediatrics 97: Hansen, S Kurz-und langfristige änderungen von herzschlagvariabilität und herzschlagfrequenz als reaktion auf veränderungen in der sozialen umwelt (gruppierung und grooming-simulation) von hausschweinen. Universitäts-und Landesbibliothek. Henning, D. R., R. Flowers, R. Reiser, E. T. Ryser, and J. F. Frank Pathogens in milk and milk products Standard methods for the examination of dairy products. American Public Health Association. Hepola, H Milk feeding systems for dairy calves in groups: Effects on feed intake, growth and health. Applied Animal Behaviour Science 80: Hill, T. M., H. G. Bateman, J. M. Aldrich, and R. L. Schlotterbeck Effect of milk replacer program on digestion of nutrients in dairy calves. Journal of Dairy Science 93:

50 Houle, M. S., and G. E. Billman Low-frequency component of the heart rate variability spectrum: A poor marker of sympathetic activity. American Journal of Physiology - Heart and Circulatory Physiology 276: H215-H223. International Organization for Standardization Microbiology of the food chain - horizontal method for the enumeration of microorganisms - part 1: Colony-count at 30 degrees c by the pour plate technique (iso :2013). Jamali, H., M. Paydar, B. Radmehr, S. Ismail, and A. Dadrasnia Prevalence and antimicrobial resistance of staphylococcus aureus isolated from raw milk and dairy products. Food Control 54: Janke, B. H. et al Attaching and effacing escherichia coli infection as a cause of diarrhea in young calves. Journal of the American Veterinary Medical Association 196: Jasper, J., and D. M. Weary Effects of ad libitum milk intake on dairy calves. Journal of Dairy Science 85: Jaster, E. H., G. C. McCoy, T. Tomkins, and C. L. Davis Feeding acidified or sweet milk replacer to dairy calves. Journal of Dairy Science 73: Jayarao, B. M., and D. R. Henning Prevalence of foodborne pathogens in bulk tank milk1. Journal of Dairy Science 84: Jeffrey, D. C., and J. Wilson Effect of mastitis-related bacteria on total bacterial count of bulk milk supplies. International Journal of Dairy Technology 40: Jensen, M. B The effects of feeding method, milk allowance and social factors on milk feeding behaviour and cross-sucking in group housed dairy calves. Applied Animal Behaviour Science 80: Jong, I. C. d. et al Effects of social stress on heart rate and heart rate variability in growing pigs. Canadian Journal of Animal Science 80: Kaske, M., H. J. Kunz, and A. Koch Basic essentials for rearing healthy and productive calves. Übersichten zur Tierernährung 37:

51 Khan, M. A., D. M. Weary, and M. A. G. von Keyserlingk. 2011a. Effects of milk ration on solid feed intake, weaning, and performance in dairy heifers Journal of Dairy Science 94: Khan, M. A., D. M. Weary, and M. A. G. von Keyserlingk. 2011b. Invited review: Effects of milk ration on solid feed intake, weaning, and performance in dairy heifers. Journal of Dairy Science 94: Knappstein, K., V. Aust, H.-J. Kunz, and M. Kaske Efficiency of two commercial on-farm pasteurizers for inactivation of mastitis pathogens in milk intended for feeding of calves. Berliner und Munchener tierarztliche Wochenschrift 126: Kunz, H.-J Kälber handbuch. Agrar- und Veterinär-Akademie, Laer. Maccari, P. et al Effects of two different rearing protocols for holstein bull calves in the first 3 weeks of life on health status, metabolism and subsequent performance. Journal of Animal Physiology and Animal Nutrition 99: Malik, M Heart rate variability. Annals of Noninvasive Electrocardiology 1: Malik, M., and A. J. Camm Heart rate variability. Futura Publishing Company Armonk, NY. Malliani, A., M. Pagani, F. Lombardi, and S. Cerutti Cardiovascular neural regulation explored in the frequency domain. Circulation 84: Maros, K., A. Dóka, and Á. Miklósi Behavioural correlation of heart rate changes in family dogs. Applied Animal Behaviour Science 109: Mazurak, N. et al Effects of a 48-h fast on heart rate variability and cortisol levels in healthy female subjects. European Journal of Clinical Nutrition 67: McCraty, R., W. A. Tiller, and M. Atkinson Head-heart entrainment: A preliminary survey. In: Proceedings of the Brain-Mind Applied Neurophysiology EEG Neurofeedback Meeting Meer, R. R., J. Baker, F. W. Bodyfelt, and M. W. Griffiths Psychrotrophic bacillus spp. In fluid milk products: A review. Journal of Food Protection 54:

52 Millette, M., F. M. Luquet, and M. Lacroix In vitro growth control of selected pathogens by lactobacillus acidophilus- and lactobacillus casei-fermented milk. Letters in Applied Microbiology 44: Mohr, E., J. Langbein, and G. Nürnberg Heart rate variability: A noninvasive approach to measure stress in calves and cows. Physiology & Behavior 75: Muir, D. D The shelf-life of dairy products: 2. Raw milk and fresh products. International Journal of Dairy Technology 49: Nonnecke, B Effects of dietary energy and protein on the immunological performance of milk replacer-fed holstein bull calves. In: American Dairy Science Association Abstracts Ohara, K. et al Cardiovascular response to short-term fasting in menstrual phases in young women: An observational study. BMC women's health 15: 67. Oliver, S. P., K. J. Boor, S. C. Murphy, and S. E. Murinda Food safety hazards associated with consumption of raw milk. Foodborne pathogens and disease 6: Öri, Z., G. Monir, J. Weiss, X. Sayhouni, and D. H. Singer Heart rate variability. Frequency domain analysis. Cardiology clinics 10: Otterby, D. E. et al Fermented or chemically-treated colostrum and nonsalable milk in feeding programs for calves1,2. Journal of Dairy Science 63: Pagani, M. et al Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho-vagal interaction in man and conscious dog. Circulation Research 59: Palestrini, C., E. P. Previde, C. Spiezio, and M. Verga Heart rate and behavioural responses of dogs in the ainsworth's strange situation: A pilot study. Applied Animal Behaviour Science 94: Persson, S Analysis of fitness and state of training. In: S. e. al. (ed.) Equine physiology. p Granta editions, Cambridge. Phillips, J. D., and M. W. Griffiths The relation between temperature and growth of bacteria in dairy products. Food Microbiology 4:

53 Puchala, R., I. Tovar-Luna, T. Sahlu, H. C. Freetly, and A. L. Goetsch Technical note: The relationship between heart rate and energy expenditure in growing crossbred boer and spanish wethers. Journal of Animal Science 87: Pumprla, J., K. Howorka, D. Groves, M. Chester, and J. Nolan Functional assessment of heart rate variability: Physiological basis and practical applications. International Journal of Cardiology 84: Reiter, B., V. M. Marshall, L. Björck, and C. G. Rosén Nonspecific bactericidal activity of the lactoperoxidases-thiocyanate-hydrogen peroxide system of milk against escherichia coli and some gram-negative pathogens. Infection and Immunity 13: Richter, R. L., and E. R. Vedamuthu Milk and milk products. Compendium of the methods for the microbiological examination of foods. 4th ed. Washington: APHA: Rietmann, T. R. et al Assessment of mental stress in warmblood horses: Heart rate variability in comparison to heart rate and selected behavioural parameters. Applied Animal Behaviour Science 88: Rindsig, R. B., and G. W. Bodoh Growth of calves fed colostrum naturally fermented, or preserved with propionic acid or formaldehyde1,2. Journal of Dairy Science 60: Robinson, R. K Dairy microbiology handbook: The microbiology of milk and milk products. John Wiley & Sons. Ruzante, J. M. et al Isolation of mycobacterium avium subsp. Paratuberculosis from waste milk delivered to california calf ranches. Foodborne Pathogens and Disease 5: Savory, C. J., L. Kostal, and I. M. Nevison Circadian variation in heart rate, blood pressure, body temperature and eeg of immature broiler breeder chickens in restricted-fed and ad libitum-fed states. British Poultry Science 47: Schlegelova, J. et al Prevalence of and resistance to anti-microbial drugs in selected microbial species isolated from bulk milk samples. Journal of Veterinary Medicine, Series B 49: Sgoifo, A. et al Incidence of arrhythmias and heart rate variability in wild-type rats exposed to social stress. American Journal of Physiology - Heart and Circulatory Physiology 273: H1754-H

54 Sperber, W. H., M. P. Doyle, L. H. Ledenbach, and R. T. Marshall Microbiological spoilage of dairy products Compendium of the microbiological spoilage of foods and beverages. Food microbiology and food safety. p Springer New York. Spreer, E Milk and dairy product technology. Taylor & Francis. Stewart, S. et al Preventing bacterial contamination and proliferation during the harvest, storage, and feeding of fresh bovine colostrum. Journal of Dairy Science 88: Sun, C. Q. et al The effect of ph on the inhibition of bacterial growth by physiological concentrations of butyric acid: Implications for neonates fed on suckled milk. Chemico- Biological Interactions 113: Sztajzel, J Heart rate variability: A noninvasive electrocardiographic method to measure the autonomic nervous system. Swiss medical weekly 134: Torkar, K. G., and S. G. Teger The microbiological quality of raw milk after introducing the two day's milk collecting system. Acta Agri Slovenica 92: Veissier, I. et al Calves' responses to repeated social regrouping and relocation. Journal of Animal Science 79: Visser, E. K. et al Heart rate and heart rate variability during a novel object test and a handling test in young horses. Physiology & Behavior 76: von Borell, E. et al Heart rate variability as a measure of autonomic regulation of cardiac activity for assessing stress and welfare in farm animals - a review. Physiology & Behavior 92: von Keyserlingk, M. A. G., F. Wolf, M. Hötzel, and D. M. Weary Effects of continuous versus periodic milk availability on behavior and performance of dairy calves. Journal of Dairy Science 89: Wiedemann, S., P. Holz, H.-J. Kunz, E. Stamer, and M. Kaske Effect of ad libitum feeding of holstein-friesian calves during the first four weeks of life on weight development as well as milk yield and feed intake during first lactation. Züchtungskunde 87(6): Wouters, J. T. M., E. H. E. Ayad, J. Hugenholtz, and G. Smit Microbes from raw milk for fermented dairy products. International Dairy Journal 12:

55 CHAPTER THREE Intensive rearing of male calves during the first three weeks of life has longterm effects on number of islets of Langerhans and insulin stained area in the pancreas L. Prokop*, M. Kaske, P. Maccari, R. Lucius, H.-J. Kunz # and S. Wiedemann* 2 * Kiel University, Department of Animal Health, Institute of Animal Breeding and Husbandry, Kiel, Germany University of Veterinary Medicine, Clinic for Cattle, Hannover, Germany Department for Farm Animals, Vetsuisse Faculty, University of Zurich, 8057 Zurich, Switzerland Kiel University, Institute of Anatomy, Kiel, Germany # Chamber of Agriculture for Schleswig-Holstein, Department of Animal Housing and Breeding, Blekendorf, Germany 2 Corresponding author: Tel ; fax ; swiedemann@tierzucht.uni-kiel.de Published in Jounal of Animal Science, 2015, Vol. 93,

56 Abstract Permanent effects of early postnatal nutrition on the development and function of tissues and organs have been previously demonstrated primarily in humans and rodents. The objective of this study in calves was to analyze the impact of rearing conditions during the first 3 wk of life on morphology of insulin producing pancreatic β-cells. Forty-two male Holstein calves were raised during the first 3 wk of life either intensively (intensively reared [INT]; ad libitum milk feeding and individual hutches; n = 21) or according to an established restrictive rearing protocol (4 L milk/ d) during wk 1 in hutches, 720 g/d milk replacer (MR) from d 8 to 21 in group pens (restrictively reared [CON]; n = 21). Thereafter, all calves were housed and fed under comparable conditions. Birth weight and weekly BW up to wk 10 were recorded. Plasma glucose, insulin, IGF-1, and GH levels were assessed in wk 1, 2, 3 and 10 of life. Slaughtering took place after 8 mo and pancreatic tissue from the medium body (corpus pancreatic) was removed. The number of islets of Langerhans and the insulin stained area were examined histologically. Total milk intake of INT calves was nearly double the intake in CON calves in the first 3 wk of life (P < 0.01). Daily starter intake during wk 4 to 10 of life did not differ between groups (P = 0.24). During the first 3 wk, the ADG were up to 9 times higher in INT calves compared to CON calves (P < 0.01), yet BW at time of slaughter did not differ (P = 0.18). Intensive rearing led to increased plasma glucose, insulin and IGF-1 concentrations after 3 wk of life compared with rearing to the established standard protocol (all P < 0.05), whereas GH was lower in INT calves during the second week of life. At time of slaughter, the mean number of islets of Langerhans was higher in INT calves compared to CON calves (9.1 ± 0.3 vs. 7.8 ± 0.3; P < 0.01). Also, the total insulin stained area per photograph was higher in INT calves compared to CON calves (107,180 ± 4,987 µm 2 vs. 84,249 ± 4,962 µm 2 ; P < 0.01). Number of islets of Langerhans was negatively associated with birth weight but positively correlated with insulin and in trend with IGF-1 plasma levels during the second week of life. Insulin stained area tended to be linked with IGF-1 concentration during the third week of life. In conclusion, differences in the morphology of pancreatic islets of Langerhans indicate that calves can be programmed metabolically by an altered postnatal rearing intensity. Keywords: ad libitum intake, calves, insulin, islets of Langerhans, metabolic programming, rearing intensity 52

57 Introduction During the last decade, numerous studies revealed an influence of perinatal environmental and nutritional conditions on long-term health and performance in many species (Srinivasan et al., 2003; Guilloteau et al., 2009; Moallem et al., 2010). During critical periods in fetal and neonatal life, the amount and the composition of nutrients permanently program biological switches such as the hypothalamic neuropeptide regulatory system (Langley-Evans et al., 2005; Waterland, 2005; Taylor and Poston, 2007). In this regard it, was observed that rat pups fed a high carbohydrate formula diet in their first weeks of life had an altered number and size of the islets of Langerhans, a clear shift on the level of insulin secretion, and a higher gene expression of preproinsulin (Srinivasan et al., 2003). These effects were still found during adulthood far after the period of dietary intervention, indicating that pancreatic islet function can be programmed during critical phases of early postnatal development in rats. Insulin produced in the pancreatic β- cells is a key anabolic hormone for glucose homeostasis. In humans, a combination of insufficient pancreatic β-cell secretion of insulin and peripheral insulin resistance results in the development of type II diabetes. In cattle, an impaired metabolism of insulin is also characteristic particularly for dairy cows who display a massive glucose drain toward the udder (De Koster and Opsomer, 2013). It is not known if the morphology of the insulin producing β-cells of the pancreas can be programmed by early postnatal rearing conditions in cattle. It was, therefore, the aim of this study to analyze the long-term effects of intensified rearing of Holstein bull calves during the first 3 wk of life compared to an established rearing protocol on development of the number and area of pancreatic islets of Langerhans. Materials and Methods Calves, Management and Feeding The experiment was set up and performed according to strict federal and international guidelines on animal research (accepted by the State of Lower Saxony, Germany; file number ). The study was conducted at the Research Center Futterkamp (Schleswig-Holstein, Germany, 190 cows and 10,300 kg mean lactation milk yield). The experimental setup has been previously described in detail (Maccari et al., 2014). Male Holstein calves (n = 42) born between May 2010 and September 2010 were separated from their dam after birth and moved into individual straw- 53

58 bedded hutches. The calves were randomly assigned to either the experimental (intensively reared [INT], n = 21) or control group (restrictively reared [CON], n = 21). All calves were fed 3 L colostrum from the dam within the first 3 h of life. At the age of 2 d, each calf received injections of 1 g iron (as Fe³+-III-hydroxyl-dextran; s.c.; Belfer, bela-pharm, Vechta, Germany) and vitamins ( IU vitamin A, 250 mg α-tocopherolacetat, IU cholecalciferol; s.c.; Vitamin ADE, animedica, Senden-Bösensell, Germany). Between the second and eighth day of age, each calf received halofuginon (100 µg/kg; os.; Halocur, Intervet Deutschland GmbH, Unterschleißheim, Germany) once a day for prevention of diarrhea caused by Cryptosporidium parvum. In the second or third week, all calves were vaccinated with an attenuated live vaccine against bovine respiratory syncytial virus and parainfluenza virus 3 (Ripsoval RS PI3 Intranasal; Pfizer Deutschland GmbH, Berlin, Germany). In the second and third month of age, every calf was vaccinated twice with an attenuated live vaccine against ringworm (Trichovac LTF 130; IDT Biologika GmbH, Dessau-Roßlau, Germany). The INT calves were housed in individual hutches during the first 3 wk of life and were offered 6 to 9 L milk twice a d in nipple buckets (0600 h and 1700 h; average energy and composition: 17.3 MJ/kg DM ME; 259 g/kg DM CP, 308 g/kg DM crude fat, 0 g/kg crude fiber [CF], 56 g/kg crude ash, and 377 g/kg DM lactose). Transition milk and bunk milk were mixed and acidified (Schaumacid Drink C flüssig, Schaumann GmbH, Pinneberg, Germany; 1.5 ml/l and ph approximately 5.5) for feeding. Each calf received 50 g of a supplement in the milk (HaGe Vollmilch Aufwerter, HaGe Nord AG, Rendsburg, Germany; 12.2 MJ/kg DM ME; 122 g/kg DM per kg CP, 19 g/kg DM per kg crude fat, 1 g/kg DM per kg CF, 86 g/kg per kg crude ash, 772 g/kg DM per kg lactose, 250,000 IU vitamin A; 25,000 IU vitamin D 3 ; 1,500 mg vitamin E; 2,000 mg vitamin C; 2,000 mg iron) once a day. The ingested volume was recorded and the residual amounts were discarded. On the 22nd and 23rd d of life, each calf received a mixture of 2 L acidified milk and 2 L milk replacer (MR; 120 g/l; BRIO Kälbermilch, Brio BV, Zeegse, Holland; whey powder, whey protein concentrate, vegetable fat [coconut oil and palm oil]; 15.9 MJ/kg DM ME, 214 g/kg DM CP, 179 g/kg DM crude fat, 2 g/kg DM CF, 63 g/kg DM crude ash, and 542 g/kg DM lactose) twice daily. On d 24 the calves moved into group pens and were fed 6 L MR daily by an automatic feeder (SA 2000; Förster Technik, Engen, Germany) up to d

59 The CON calves received during their first week of life, 2 L acidified milk twice daily (0600 and 1700 h) using nipple buckets. They were moved into group pens on their eighth day of life and were fed 6 L MR daily by an automatic feeder up to d 28. After the 25th day of life, INT and CON calves were similarly housed and fed. Weaning took place between d 29 and 70 by reducing the amount of MR from 6 to 2 L per calf gradually. The intake of MR and calf starter after the third week of life was recorded by the automatic feeding system (KalbManagerWIN, version and ; Förster Technik). At all times, all calves had free access to water, hay and calf starter (30.1% wheat, 20.0% wheat gluten feed, 15.0% linseed extraction meal, 12.5% soy extraction meal, 11.5% dried pulp, 5.0% rapeseed expeller, and 3.0% molasses). After weaning, all calves moved to fatteners at an average age of 85 ± 11 d. Two fatteners were involved for technical reasons. Thirty-two calves (INT, n = 16, and CON, n = 16) were housed in an open straw-bedded stable and were fed concentrates ad libitum. Ten calves (INT, n = 5, and CON, n = 5) were housed in a closed stable with a slatted floor and were fed corn silage ad libitum and 3 kg concentrates per day and calf. Animals of both groups were slaughtered at an age of 238 ± 1 d. Body Weight and Blood Data Acquisition The BW of each calf was recorded after birth and weekly up to wk 10 and on day of slaughter. Daily weight gain was calculated from birth to slaughter. At the slaughterhouse, carcass weight was recorded for each animal. Blood samples were collected by puncture of the jugular vein (1.20 x 40 mm; Sterican; B. Braun Melsungen AG, Melsungen, Germany) approximately 2 to 3 h after fresh feed supply on d 2 to 3 (wk 1 of life), d 10 to 12 (wk 2 of life), d 18 to 21 (wk 3 of life), and d 65 to 70 (wk 10 of live). Due to the high intraday variation of blood metabolites in cattle (Wiedemann et al., 2013) and the unknown period of prior starving, no blood analyses were performed during time of slaughter. The blood was conserved in tubes containing K 2 EDTA or sodium fluoride. Centrifugation (3,000 x g for 10 min at 4 C) took place within 1 h after blood sampling. Plasma was stored in Eppendorf cups at -20 C until analysis. The concentrations of glucose in fluoride plasma were measured using a photometric automatic clinical chemistry analyzer (A11A01667; ABX Diagnostics, Montpellier, France) with a hexokinase method (Passey et al., 1977). Quality control 55

60 and calibrations were performed daily. The CV of 20 measurements of 1 sample was 3.1%. A RIA validated for bovine insulin (IM3210; Immunotech, Pasadena, CA) was performed to determine plasma insulin concentrations. The intra-assay CV was 7.6%, and the interassay CV was 10.7%. For total plasma IGF-I determination, IGF-I was separated from its binding proteins by an acidethanol extraction procedure. Thereafter, IGF-I concentrations were determined by an ELISA according to the standard operation manual (DSL ; Diagnostic Systems Laboratories, Inc., Webster, TX). The analytic sensitivity was 0.03 ng/ml. The intra- and interassay CV were 1.5 to 3.5 and 1.5 to 8.5%, respectively. Plasma GH concentration was measured by an ELISA according to Roh et al. (1997) with slight modifications (Piechotta et al., 2013). The GH levels were calculated using Magellan software with a cubic spline modus (Magellan 3.11; Tecan, Crailsheim, Germany). The intra- and interassay CV were 9.8 and 12.6%, respectively. The lowest detection limit was 1.0 ng/ml. At the slaughterhouse, 2 pieces of the central part of the pancreas (Corpus pancreatic) with an edge length of about 2 cm were removed, respectively. The pancreatic tissue was fixed in 4% formaldehyde and transported to the lab. There, 2 pieces of pancreatic tissue with an edge length of 5 mm were dissected and stored at 4 C until analyses in tubes containing 10% paraformaldehyde. Tissue was embedded in paraffin and cut in 7-µm-thick slices according to routine procedures. Histological Analyses A hematoxylin eosin overview stain (Fischer et al., 2008) was performed to determine the quality of the pancreatic tissue samples. Immunohistochemistry was performed to analyze the number of islets of Langerhans and the insulin stained area. Firstly, a rabbit polyclonal antibody to insulin (BioLogo, Kronshagen, Germany) was diluted with 1% BSA (0.75 to 1 g BSA on 100 ml PBS; dilution ratio 10) and incubated overnight. Subsequently, a biotinylated goat anti-rabbit antibody (Vectastain ABC Kit; Vector Laboratories Inc., Burlingame, CA) diluted with PBS (dilution ratio 300; Merck, Darmstadt, Germany) was applied and incubated for 45 min. An avidin-biotincomplex (Vectastain ABC Kit; Vector Laboratories Inc.) was added and incubated for 45 min. The pancreatic preparations were stained for 10 min with a 3,3 -diaminobenzidine kit to visualize peroxidase activity (Vector Laboratories Inc.). The area containing insulin was stained in brown. 56

61 The specificity of staining was determined by preabsorbing the primary antibody with purified bovine insulin (>27 units/mg; Sigma-Aldrich, St. Louis, MO). In brief, the anti-insulin antibody (0.05 mg/ml) was preincubated (molar ratio 1:10) with bovine insulin (0.5 mg/ml) at 4 C overnight and the mixture was used for immunohistochemistry. Nonspecific staining was determined by omission of the secondary antibody. Number of Islets of Langerhans and Insulin Stained Area For the determination of the number of islets of Langerhans, the immunostained preparations were used. The Axiophot photo microscope (4-mm lens; Carl Zeiss Inc., Thornwood, NY) and the associated camera (Axio Cam; Carl Zeiss Inc.) were applied to take pictures of the histological sections in a standardized field (Taniyama et al., 1993). The exposure time was adjusted at 16 ms. All pictures were filled with tissue. Particular care was given to avoid double photographing of 1 tissue section. The Axio Vision 4.5 computer program (Carl Zeiss Inc.) was used to take and process the pictures. As many as possible tissue sections of all calves (6-18, depending on cross-section size of preparation) were photographed. An islet of Langerhans was defined as a unit of at least 6 insulin-positive immunoreactive cells to avoid counting single and newly formed nonislet cells (Petrik et al., 2001). The number of islets of Langerhans was randomly counted twice for each photograph and an arithmetic mean per photograph was calculated for each animal. At first, a validation of the method with samples of 6 randomly selected calves was performed. The mean CV of all calves between the 6 to 18 cross-sections was 31.2%. For 6 randomly selected calves, 2 different sections of pancreatic tissue were stained and the mean values of the cross-sections differed by 16.1%. The determination of the immunohistochemically stained area (insulin stained area) was performed with the Eclipse E600 photo microscope (Nikon, Düsseldorf, Germany) and the associated camera (DS-Fi1; Nikon). A 4 mm lens was used. The NIS-Elements Basic Research 3.2 software program (Nikon) was used to take and process the pictures. The distribution of the 5 pictures on the section was chosen as follows to prevent an influence of the photographer: top left, top right, central, bottom left and bottom right. All pictures were filled with tissue. The quality of resolution was defined at 2,560 x 1,920 pixels and the focus at 1,280 x 960 pixels. The exposure time was adjusted at 1 ms. Quantitative threshold values of the color intensity were individually defined for each picture to mark the brown insulin stained area by the program. Different correcting functions were used to best assess the brown stained area. On the basis of a 57

62 field measurement, the computer program calculated the area of all marked districts with a minimum size range of 20 µm at 1 picture. The mean CV of all calves between the 5 photographs of 1 section was 25.8%. For 6 randomly selected calves, 2 different sections of pancreatic tissue were stained and the mean values of the sections differed by 11.2%. Statistical Analyses All data were analyzed using SAS (Version 9.3; SAS Inst. Inc., Cary, NC). For analyses of total feed intake, birth weight, and weight at slaughter, generalized linear models were applied (GLM procedure). The group was included in all models as fixed effect. Birth weight and age at slaughter were further included as covariates in the model to calculate weight at slaughter. Simple correlations between number of islets of Langerhans and insulin stained area with plasma metabolites, ADG, birth weight, and weight at slaughter were calculated (CORR procedure in SAS). For statistical analyses of data on feed intake and BW generalized linear mixed models (MIXED procedure in SAS) were used. Models for milk and concentrate intake as well as BW between wk 1 and 10 included group, week of life, and group week of life as fixed effects and animal as the random effect. The analyses of number of islets of Langerhans and insulin stained area per photograph (MIXED procedure in SAS) included group as fixed effect. Birth weight and weight at slaughter were highly correlated (P = ). Because birth weight resulted in a better fit of the model, it was chosen as covariate. The animal was considered a random effect. For plasma metabolite concentrations the fixed effects were group, age at sampling, and group age at sampling (MIXED procedure in SAS). The random error term was animal. If an overall significant effect was found, a subsequent Bonferroni post hoc analysis was performed. A P-value of <0.05 was considered significant and a P-value of <0.1 was considered a trend for all models. Data are presented as Least square means ± SE. 58

63 Results Feed Intake and BW Development of Calves Mean ad libitum intake of milk increased during the first 3 wk of life up to 10.5 L/d in INT calves (Fig. 1). Until the end of wk 3 of life, total milk consumption was almost twice as high in INT calves compared with the consumption of CON calves (196 vs. 102 kg of milk/mr per calf during the first 3 wk), whereas no differences were obtained between wk 4 and 10 of life (4.3 ± 0.05 vs. 4.2 ± 0.05 kg MR/d). Energy and protein intakes did not differ from those reported by Maccari et al. (2014), who studied all animals of this experiment and additional 6 calves. In short, in the first 3 wk of life, intake of ME was more than twice as high in INT calves compared to CON calves and an even more pronounced difference between the groups was found for intake of CP. Average daily intake of calf starter was neglible during the first 3 wk of life. Thereafter, starter intake increased and was comparable between groups (0.94 ± 0.06 vs ± 0.06 kg/d; P = 0.28). Birth weights in INT and CON calves were similar (P = 1.0; Table 1). Successive weekly BW increased in all calves with exceptions during the time periods of the transfer of calves from individual hutches into group pens. Average daily gain during the first week of life was 2.5 times higher in the INT calves compared with the CON calves (P < 0.05, Fig. 1). During the second and third week, an increased rearing intensity resulted in almost 9 times higher average daily gains compared with a standard rearing protocol. During wk 5, a tendency for lower weight gains was observed in INT calves (P < 0.1). However, the INT calves remained heavier than the CON calves until weaning (P < 0.01; Table 1). Average daily gain, BW, and carcass weight at time of slaughter did not differ significantly between both treatment groups (Table 1). Blood Parameters The basal plasma concentrations of glucose, insulin, IGF-1 and GH on d 2 were comparable for both treatment groups (Fig. 2). Ad libitum feeding of milk was accompanied by higher glucose concentrations in the INT calves compared to the CON calves during the third week of life (P < 0.01). During the second and third week, the INT calves exhibited 3- to 5-fold higher plasma insulin and IGF-1 concentrations compared to the CON calves (P < 0.05), whereas GH was 59

64 found to be 2-fold higher in the CON calves compared to the INT calves during wk 2 (P < 0.01). In wk 10 of life glucose, insulin, GH, and IGF-1 concentrations in plasma did not differ. Number of Islets of Langerhans and Insulin Stained Area The preabsorption of the primary anti-insulin antibody performed to ensure the specificity of the insulin-staining resulted in a disappearance in staining intensity in comparison to the immunostaining with the non-preincubated antibody (Fig. 3c and 3d). The omission of the primary antibody yielded no further detectable specific staining. However, both experiments showed a faint brown unspecific background staining in the exocrine part of the pancreas, which was not included in the analysis of the endocrine pancreas. At time of slaughter with an approximate age of 8 mo, INT calves had a higher number of islets of Langerhans per field of view compared to CON calves (+17%; P < 0.01; Table 2; Fig. 3a and 3b). The different feeding regime also influenced the total insulin stained area per photograph (+27% in INT calves; P < 0.01), whereas the number of insulin stained areas per photograph did not differ (+11% in INT calves; P = 0.4). The mean area of continuous insulin stained cells tended to be higher in the INT calves compared to the CON calves (+18 % in INT calves; P = 0.079; Table 2; Fig. 4). Correlations of Histological Findings with Weight Parameters and Blood Metabolites Simple correlation procedures revealed a correlation in trend between number of islets of Langerhans and insulin stained area (r = 0.46, P = 0.06). The number of islets of Langerhans was negatively correlated with birth weight of calves (r = -0.4; P < 0.01) and weight at slaughter (r = -0.3; P < 0.01). However, birth weight and weight at slaughter are not independent of each other (r = 0.6; P < 0.01) and a multicollinearity with the number of islets is assumed. No relationship was observed between mean insulin stained area with weights at birth or slaughter (P = 0.9 and P = 0.8). The ADG during the first 3 wk of life was weakly correlated to the number of islets of Langerhans (r = 0.21, P = 0.01) but not with the insulin stained area (P = 0.6) and with birth weight (P = 0.3). Insulin concentration during the second week of life was associated with the insulin stained area in trend (r = 0.5, P = 0.08). The IGF-1 concentrations during the second week was related to the number of islets of Langerhans (r = 0.4, P < 0.05) and during the third 60

65 week to the insulin stained area in tendency (r = 0.5; P < 0.08). No further relationships between insulin, glucose, IGF-1, and GH concentrations with number of islets of Langerhans and insulin stained area could be established. Discussion In agreement with our findings, intensive rearing previously has been shown to increase postnatal nutrient intake and BW development of calves in comparison with calves reared according to a conventional protocol (Khan et al., 2011). However, the transfer of calves from individual hutches into group pens and the associated change of the feeding system (from nipple buckets to an automatic feeding system) and of the feedstuff (from whole milk to MR) resulted in reduced feed intakes in both groups. As a consequence, the respective change was associated to a depression in weight gains in CON calves and INT calves after the first and fourth week of life, respectively. In this context, it is recommended to avoid a very early change of housing and feeding. Intensively reared calves seemingly adapted to the new housing and feeding system much easier and faster compared with CON calves. The utilization of the great growth potential by unrestricted feeding lasted up to the 70th day of life in our study and to the 90th day of life in a previous study (Khan et al., 2007b). Yet differences in BW at slaughter were not as great as expected (10.6 kg). This was inter alia attributed to diseases during the subsequent rearing period (Maccari et al., 2014). To our knowledge, this is the first study in cattle which demonstrates an effect of early postnatal weight gain on the permanent morphology of the pancreas. The analysis of pancreatic tissue in a standardized field of a light microscope revealed a higher number of pancreatic islets of Langerhans in 8-mo-old bulls that were raised with an increased intensity during the first 3 wk of life in comparison with bulls that were raised according to an established rearing protocol. This finding in cattle is in agreement with previous experimental results on long-term effects of early postnatal nutrition and environment on the histology of the islets of Langerhans which were obtained primarily in rodents (Laychock et al., 1995; Aalinkeel et al., 2001; Petrik et al., 2001; Srinivasan et al., 2001). In rats, the nutrient availability in the late fetal and/or the early postnatal period determines the pancreatic ontogeny at the cellular, biochemical, and molecular level (Waterland and Garza, 1999; Aalinkeel et al., 2001). During this critical time frame, the rate of 61

66 replication and regeneration of β-cells from the ductal pancreatic epithelium is particularly high and feeding intensity possibly influences the balance between β-cell proliferation and β-cell death (Hill and Hogg, 1991; Kaung, 1994; Petrik et al., 1999). The differences in the number of islets were also reflected by differences in the insulin stained area between both feeding strategies. The NIS-Elements Basic Research 3.2 morphometric analysis software was used for the first time to determine the immunohistochemically stained area. Settings had to be made manually for each picture which might cause a subjective influence of the researcher on the values of the insulin stained area. As a result, absolute values of the insulin stained area have to be interpreted with caution. In this study, only 1 person performed the investigation without knowledge of calves belonging to a feeding group. In contrast to the manual determination of the number of pancreatic islets in this study, a differentiation between pancreatic islet cells and other insulin-producing cells such as newly formed single cells and small cell groups was not possible. That could explain the greater percentage difference between INT and CON calves for insulin stained areas than for islets per field of view. Also, cells belonging to 1 islet could be counted as 2 insulin stained areas, which might be the reason for the percentage difference between the manual counting of islets and the automatic counting of insulin stained areas. Previous studies used NIS Elements Basic Research as morphometric analysis software for different tissues (Zhang et al., 2009; Phipps et al., 2012; Walton et al., 2013). For image analysis, individual islets were circled and selected by red, green and blue threshold (Petrik et al., 2001) or by gray-level threshold (Petrik et al., 1999), which was not feasible in this study. Until now, the control mechanisms of regeneration and cell death of the islets of Langerhans are not fully understood. During neonatal life, apoptotic islet cells are replaced via neogenesis of ß- cells from the ductal epithelium (Scaglia et al., 1997; Petrik et al., 1998). In this developmental stage, the number and size of the newly formed islets are influenced by feeding due to the influence of nutrition on the ß-cell neogenesis and replication (Petrik et al., 2001). Neonatal rats fed a high-carbohydrate diet displayed an increased number of smaller islets compared to rats fed normal rations (Petrik et al., 1999; Petrik et al., 2001; Srinivasan et al., 2001). Furthermore, there is evidence that growth, maturation, and function of β-cells is dependent on IGF because IGF-1 and IGF-2 can protect islet cells against cytokine-induced apoptosis (Bryson et al., 1989; Petrik et al., 1998; Petrik et al., 2001). In line with this findings in humans and rats, 2 to 3 times higher 62

67 IGF-1 plasma concentrations during the second and third week of life in INT calves could at least partly explain differences in the morphology of pancreatic islets between both groups in this study. The assumption is further supported by the observed correlation between IGF-1 plasma concentration during the second and third week of life with the number of islets of Langerhans and insulin stained area, respectively. Interestingly, the lower levels of IGF-1 in CON calves were accompanied by higher GH levels. This pattern indicates a decoupling of the basically functioning somatotropic axis in calves (Sauter et al., 2003) and has been previously reported for poor-growing piglets (Saleri et al., 2001). The synthesis and storage of insulin and its precursors in ß-cells of the endocrine pancreas is regulated by nutrient, hormonal, and neuronal stimuli. In mice (age 1 to 18 mo), a close positive relationship was demonstrated between islets area and total content of pancreatic insulin as well as between pancreatic insulin and the insulin secretory capacity in simultaneously performed perfusion and morphological analyses of the endocrine pancreas (Bonnevie-Nielsen, 1986). In contrast to our results, they also revealed a strong relationship between BW and islet area and insulin secretory capacity and total pancreatic insulin but no relationship between BW and pancreatic islet number. Increased plasma glucose concentrations and plasma insulin concentrations in INT calves during the time of different rearing intensity compared with calves fed restrictively were already established in suckling Simmentaler calves and in intensively fed Holstein calves (Egli and Blum, 1998; Smith et al., 2002). Like in monogastric species, glucose derived from intestinal absorption plays the most important role in the regulation of insulin secretion in young calves (Khan et al., 2007a). Therefore, the plasma glucose and insulin levels are associated with the amount of lactose intake during the first week of life (Hugi et al., 1997). In monogastric animals, changes in glucose levels modify the rate of translation of preexisting mrna and thereby the rate of exocytosis of insulin in a matter of minutes. Long-term effects on insulin secretion are modulated through adaptations in the preproinsulin gene transcription rate (Docherty and Clark, 1994). In rats fed high amounts of carbohydrates during the neonatal period, the plasma insulin levels are also increased (Srinivasan et al., 2000). In such rats, the higher demands of insulin lead to compensatory molecular adaptions of the pancreatic islets (Srinivasan et al., 2001; Srinivasan et al., 2003). The higher lactose levels of INT calves during the first 3 wk of life could be considered as a cause of a similar adaptive reaction of the pancreatic islet. Our results indicate 63

68 that this effect continues after the development of a viable rumen fermentation and replacement of glucose by short-chain fatty acids and level of vagotone as the primary trigger for insulin secretion (Bloom and Edwards, 1981; Peters and Elliot, 1984; Baldwin et al., 2004). In conclusion, intensive rearing during the first 3 wk of life results in an enhanced nutrient uptake and body development in comparison with a conventional rearing protocol. Differences in plasma insulin, glucose, IGF-1 and GH concentrations occurred during the intensive rearing period but were not found thereafter. However, the observed alterations in pancreatic tissue imply that the rearing intensity during the neonatal period has long-term consequences. This result supports the theory of a possible metabolic programming in INT calves. Future studies need to address the detailed mechanisms and influencing factors. Consequences for future performance particularly in dairy cows also warrant further investigations. Acknowledgement The scholarship of the University of Kiel for the first author is gratefully appreciated. We thank the staff of the Training and Research Centre Futterkamp, Germany and the staff of the laboratory of the Clinic for Cattle, University of Veterinary Medicine Hannover, Germany. We thankfully acknowledge the staff of the Anatomical Institute at the Christian-Albrechts- University Kiel, Germany, especially M. Koelln and K. Masuhr for their support in the laboratory. 64

69 Tables & Figures Table 1. Body weight and average daily weight gain (ADG; Least square means ± SE) of Holstein bull calves intensively reared (INT; n = 21) or restrictively reared (CON; n = 21) during the first 3 wk of life Parameter INT CON P-value Birth weight, kg 43.9 ± ± Weight at weaning, kg (d 70) 109 ± ± 1.5 < 0.01 Age at slaughter, d 238 ± ± BW at slaughter, kg 319 ± ± Carcass weight, g 162 ± ± ADG birth to slaughter, g/d 1,159 ± 21 1,114 ± Table 2. Effect of rearing intensity of Holstein bull calves during the first 3 wk of life on number of islets of Langerhans as well as on number, total area, and mean area of insulin stained areas per photograph (Least square means ± SE) Parameter INT CON P-value No. of calves / No. of preparations 21 / / 191 No. of islets per field of view 9.1 ± ± 0.3 < 0.01 No. of calves / No. of photographs 21 / / 105 No. of insulin stained areas per photograph 27.3 ± ± Total insulin stained area per photograph, µm² 107,180 ± 4,987 84,249 ± 4, Mean area of insulin stained cells, µm 2 5,425 ± 341 4,574 ±

70 Figure 1 Effect of rearing intensity of Holstein bull calves during the first 3 wk of life on intake of milk/milk replacer and ADG (Least square means ± SE) during the weaning period (INT = intensively reared; CON = restrictively reared [according to a standard protocol]). * P < 0.1; *** P < 0.01) 66

71 Figure 2 Effect of rearing intensity of Holstein bull calves during the first 3 wk of life on plasma concentration of glucose, insulin, IGF1, and GH (Least square means ± SE) measured in wk 1,2,3, and 10 (INT = intensively reared; CON = restrictively reared [according to a standard protocol]). * P < 0.1; ** P < 0.05; *** P < [a] [b] [c] [d] 67

72 Figure 3 Effect of rearing intensity of Holstein bull calves during the first 3 wk of life on immunhistochemical presentation of brown stained β-cells of Langerhans exemplified by two calves fed either (a) restrictively or (b) intensively (circled areas in red represent insulinpositive immunoreactive cells counted as pancreatic islets); negative controls of staining after (c) overnight preabsorption of the polyclonal insulin antibody with insulin and (d) withdrawal of the insulin antibody. a 4x b 4x c 20x d 20x 68

73 Figure 4 Immunhistochemical presentation of brown stained β-cells of Langerhans of 2 Holstein bull calves reared in the first 3 wk of life either restrictively or intensively and the respective red marked districts for the calculation of the area. (a) Immunhistochemical presentation of pancreatic tissue of a restrictively reared calf. (b) Figure 4a and the respective red marked districts for the calculation of the area (area: 39,793 µm²). (c) Immunhistochemical presentation of pancreatic tissue of an intensively reared calf. (d) Figure 4c plus the respective red marked districts for the calculation of the area (area: 100,151 µm²). a 4x b 4x c 4x d 4x 69