**Calf and Heifer Feeding and management**

Editors

**Zhijun Cao Michael Van Amburgh**

MDPI • Basel • Beijing • Wuhan • Barcelona • Belgrade • Manchester • Tokyo • Cluj • Tianjin

*Editors* Zhijun Cao China Agricultural University China

Michael Van Amburgh Cornell University USA

*Editorial Office* MDPI St. Alban-Anlage 66 4052 Basel, Switzerland

This is a reprint of articles from the Special Issue published online in the open access journal *Animals* (ISSN 2076-2615) (available at: https://www.mdpi.com/journal/animals/special issues/Calf and Heifer).

For citation purposes, cite each article independently as indicated on the article page online and as indicated below:

LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. *Journal Name* **Year**, *Article Number*, Page Range.

**ISBN 978-3-03943-661-3 (Hbk) ISBN 978-3-03943-662-0 (PDF)**

Cover image courtesy of Hoard's Dairyman.

c 2020 by the authors. Articles in this book are Open Access and distributed under the Creative Commons Attribution (CC BY) license, which allows users to download, copy and build upon published articles, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications.

The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons license CC BY-NC-ND.

## **Contents**



## **About the Editors**

**Zhijun Cao** is Professor and Associate Dean of College of Animal Science & Technology, China Agricultural University (CAU). His bachelor's degree was from Heilongjiang Bayi Agirucltural University and his Ph.D. degree was from CAU. He spent 5 months in South Dakota State University and University of Wisconsin–Madison as an exchange student in 2006, and 15 months in Cornell University as Visiting Professor in 2010–2011. His research program over the last 10 years has been focused on the nutrient requirements of dairy calves and heifers, and the control of energy metabolism in transition dairy cows. As the first author or corresponding author, he has published 62 peer-reviewed articles (11 in Journal of Dairy Science) and is the recipient of several awards, including the National Special Support Program for High-Level Personnel Recruitment, the Beijing NOVA Program Award, and the DBN Award for Young Scientists. He serves as the Committee Member of Nutrient Requirements of Dairy Cattle in China, and Deputy General Secretary of China Cattle Science Association, Chinese Association of Animal Science and Veterinary Medicine. He is also Editor-in-Chief of Hoard's Dairyman China and an editorial board member of *Animals*.

**Michael Van Amburgh** (Professor, Cornell University) is Professor at the Department of Animal Science, Cornell University. His research is devoted to the discovery and problem solving of contemporary issues in nutrition, physiology, and management in the dairy and livestock industries, and disseminating new knowledge to students and the industry. He is currently leading the development of Cornell's net carbohydrate and protein system (CNCPS). The development and use of CNCPS has greatly improved the efficiency of ruminant nutrition utilization, with improvements for the environment and animal production performance, which has resulted in significant contributions to ruminant research. With the increasing pressure on the dairy industry to reduce the environmental impact of cattle, Van Amburgh and his team are currently working on developing a basic understanding of whole animal nitrogen metabolism and the efficient use of absorbed amino acids. In support of the CNCPS, new tools are also being developed to describe how NDF digestion occurs in various feeds. Van Amburgh is recipient of the Most-Cited Paper Award in 2018 for the Nutrition, Feeding and Calves section of the Journal of Dairy Science. He has won the American Dairy Science Association Stephen H. Weiss Presidential Fellow Award (2016) and the Cornell University Donald C. Burgett Distinguished Advisor Award (2012).

## **Preface to "Calf and Heifer Feeding and management"**

From birth to first calving, the replacement heifer undergoes tremendous changes anatomically as well as in feeding and management practices. The calf changes from being a pseudo-monogastric to a full ruminant within a period of two months. During the same period, the calf is fed colostrum, milk, or milk replacer, and starter with or without hay. Notably, the lifetime milk production and health of a dairy cow is highly dependent on early life nutrition and management of the calf and, subsequently, the heifer. Hence, animal scientists continue to investigate critical areas such as colostrum feeding, the level of liquid feeding, gut microbial succession, energy and protein levels, housing, health management, and their interactions with the animal in an effort to help dairy producers raise successful and sustainable dairy enterprises. Emerging research techniques have opened new frontiers to better understanding the whole animal and how its diet and environment might influence its microbial, endocrinal, immunity, and metabolic systems. The integration of existing and current knowledge will help refine replacement heifer feeding and management practices. The aim of this Special Issue is to publish current and relevant information related to the nutrition, metabolism, housing, and health of replacement heifers. All papers are open access and can be freely accessed at the Special Issue website: https://www.mdpi.com/journal/animals/special issues/Calf and Heifer.

> **Zhijun Cao, Michael Van Amburgh** *Editors*

*Review*

## **Review: How Forage Feeding Early in Life Influences the Growth Rate, Ruminal Environment, and the Establishment of Feeding Behavior in Pre-Weaned Calves**

## **Jianxin Xiao 1,**†**, Gibson Maswayi Alugongo 1,**†**, Jinghui Li 2, Yajing Wang 1, Shengli Li <sup>1</sup> and Zhijun Cao 3,\***


Received: 29 November 2019; Accepted: 19 January 2020; Published: 22 January 2020

**Simple Summary:** Under natural grazing systems, calves are likely to consume forage in early life. However, forage inclusion in the diet of pre-weaned calves has long been a controversial issue due to it possibly being associated with negative calf performance. Recent published literature seems to confound previous research. This review aims to understand the factors that may influence forage inclusion in the ration of pre-weaned calves. We have explored research related to the effect of feeding forage on rumen and behavioral development to better understand whether forage should be fed to the young calf. Based on the findings, it is concluded that a small amount of good quality forage is recommended for calves to improve their behavioral expression and rumen environment, which may further improve calf performance.

**Abstract:** The provision of forage to pre-weaned calves has been continuously researched and discussed by scientists, though results associated with calf growth and performance have remained inconsistent. Multiple factors, including forage type, intake level, physical form, and feeding method of both solid and liquid feed, can influence the outcomes of forage inclusion on calf performance. In the current review, we summarized published literature in order to get a comprehensive understanding of how early forage inclusion in diets affects calf growth performance, rumen fermentation, microbiota composition, and the development of feeding behavior. A small amount of good quality forage, such as alfalfa hay, supplemented in the diet, is likely to improve calf feed intake and growth rate. Provision of forage early in life may result in greater chewing (eating and ruminating) activity. Moreover, forage supplementation decreases non-nutritive oral and feed sorting behaviors, which can help to maintain rumen fluid pH and increase the number of cellulolytic bacteria in the rumen. This review argues that forage provision early in life has the potential to affect the rumen environment and the development of feeding behavior in dairy calves. Continued research is required to further understand the long-term effects of forage supplementation in pre-weaned calves, because animal-related factors, such as feed selection and sorting, early in life may persist until later in adult life.

**Keywords:** calves; forage; performance; rumen fermentation; behavior

#### **1. Introduction**

As early as 1897, researchers began to evaluate hay feeding in young calves [1]. Since then, more calf related studies involving various aspects such as genetics, nutrition, health, and welfare have been completed [2]. Likewise, over the last hundred years, the use of forage in pre-weaned calves has remained one of the most key concerns in calf nutrition.

Before the 1950s, forage feeding was generally encouraged in pre-weaned calves, as it was believed to reduce abnormal behavior (e.g., eating bedding material) [3], lower diarrhea [4], and improve rumen development [5,6]. However, new research emerged challenging the fact that forage feeding could improve rumen development to the same degree as calf starter [7]. Volatile fatty acids (VFA) were considered to play a more critical role in stimulating rumen epithelial development rather than the physical form of diets [8]. Specifically, forage ration resulted in a higher proportion of acetate [9], which did not stimulate the growth of rumen papillae to the same extent as butyrate and propionate [10,11]. Concentrates, high in rapidly fermentable carbohydrates, produced more butyrate and propionate [12]. Therefore, higher proportions of concentrates could enhance the development of rumen papillae [13]. Furthermore, as fiber had lower digestibility than starch and sugar, many studies claimed that roughage increased gut fill because of low ruminal fermentation rate, thus curbing the consumption of starter feed which had higher energy density [5,14,15]. Therefore, some dairy farms provided calves with ad libitum access to concentrate feed, with no forage until after weaning [16]. More recently, in the 2000s and 2010s, more studies have investigated the effects of forage feeding in pre-weaned calves, yet results were inconsistent. Some of the studies reported a decrease [15,17], an increase [18–23], or no differences [24–27] in solid dry matter intake (DMI) and average daily gain (ADG) when forage was added in the calf diets. As the solid DMI and growth rate of the pre-weaned calves are important factors that drive rumen development [15] and subsequent milk production in the first lactation [28], it is vital to understand the factors that influence feed consumption and calf growth when forage is added to their diet.

On the other hand, over the last ten years, more research has explored the effects of forage inclusion in diets of young calves not only on calf performance and rumen morphological development but also on ruminal fermentation metabolites, bacterial composition, and feeding behavior [2], giving us a more comprehensive and better understanding of this topic. Therefore, the aims of this review are: (1) to discuss the factors that contribute to the inconsistent results in performance in calves with forage inclusion; (2) to summarize and evaluate the latest literature on the role of forage in rumen fermentation and the establishment of feeding behavior of calves.

#### **2. Factors that A**ff**ect Calf Performance with Forage Inclusion**

Generally, under natural grazing systems, adult dairy cows spend 7 to 13 h eating grass every day [29]. Young calves acquire nutrients from both milk and fresh grasses and begin to graze as early as week 2 of age [30,31]. The grazing time usually lasts for a short period, around 20 min at 10 days, which increases rapidly to 360 min at 100 days of age, equivalent to 70 percent of the grazing time in adult cattle [30,32].

On some commercial dairy farms, calves are offered free access to starter feed before weaning without forage [33], which is contrary to natural grazing. Forage inclusion in the pre-weaned calf diet has long been discouraged due to its negative effect on the growth rate [15,17]. However, recent research has shown that several factors need to be taken into account when evaluating the impact of forage provision on calf performance [8,9,34]. These factors include the source, amount, particle size, physical form, offering time, and feeding method of forage and concentrate, as well as the amount of milk offered and milk feeding method. A summary of these studies is presented in Tables 1–6.

#### *2.1. Forage Factors*

Forage provision in pre-weaned calf remains a controversial topic, hence the proliferation of related research. In earlier studies, it was believed that forage was the main component in calf diets that played an essential role in rumen development [5,6]. In two different studies offering a high proportion of roughage (80% vs. 60% vs. 40% and 67% vs. 50%) to calves as hay to grain respectively, Hibbs et al. [5] and Conard and Hibbs [35] found that DMI and ADG increased as the proportion of concentrate in the ration increased. Stobo et al. [36] obtained similar results with calves provided a maximum daily allowance of concentrate at 0.45, 0.91, 1.36, 1.81, or 2.27 kg/d along with ad libitum access to grass hay (crude protein (CP) = 7.5%, crude fiber (CF) = 28.7%). These authors reported that as the concentrate intake increased, the hay intake decreased linearly [5,36], while live and empty body weight (BW) and rumen development were improved [36]. This is probably because concentrates result in more VFA production, especially propionate and butyrate [12], which enhances papillae development [13,36]. Collectively, these early studies suggested that in a high roughage feeding system, the addition of a bulky forage in the diet might decrease the consumption of energy-dense concentrates, leading to less rumen fermentation and lower degradation rates, and subsequently lower total nutrient intake and calf growth.

In the early 2000s, several studies began to investigate the effect of lower levels of forage inclusion in the diet on pre-weaned calf performance. Most of them included a proportion of forage ranging between 5 and 25% of total solid feed intake [15,19,22,27,37–40]. In contrast to previous studies, most studies either reported an increase [19,22,38,40] or a lack of differences [27,37] in DMI, ADG, and BW, indicative of multifactorial effects in these studies.



#### *Animals* **2020**, *10*, 188

negative effect, *p* indicates positive effect, and NS shows no significant effect. DMI and ADG were evaluated by comparing calves fed with or without forage.

#### 2.1.1. Forage Level and Source

Coverdale et al. [19] conducted two experiments in which starter supplemented with relatively low level (7.5 and 15%) of bromegrass hay appeared to improve DMI, ADG, and feed efficiency (FE). In experiment 1, limited amounts of mixed feed (concentrate and forage) were offered before weaning, followed by ad libitum feeding post-weaning. Calves receiving coarse starter with either 7.5 and 15% of bromegrass hay (8 to 19 mm) were heavier and had greater ADG and FE than calves receiving only coarse starter, while calves fed 7.5% of hay tended to have the highest ADG and FE [19]. In experiment 2, calves were offered diets ad libitum and weaned according to intake. The concentrate and total DMI tended to be higher in calves fed 7.5 and 15% of bromegrass hay when compared with the non-forage group [19]. Similarly, Hosseini et al. [40] recently reported that compared to non or 15% straw, feeding 7.5% of chopped wheat straw tended to improve the overall total solid feed intake (659, 685, and 826 g/d, respectively) and ADG (519, 553, and 620 g/d, respectively) when calves were offered 4 L of whole milk per day. Feeding alfalfa hay at 10% of total solid feed increased the overall DMI, ADG, and final BW, and thereby shortened the time to weaning at a target DMI of starter (1 kg for 3 consecutive days), compared with feeding 0 or 5% of alfalfa hay [22]. Nemati et al. [38] also observed a linear increase in total DMI and ADG during the postweaning (d 52 to 70) and overall periods (d 3 to 70) of calves supplemented with chopped alfalfa hay at 0, 12.5, and 25% on dry matter (DM) basis. However, gut fill could be a confounding factor when evaluating the effect of forage feeding on improving ADG in dairy calves. It is commonly believed that an increase in ADG and BW in calves fed forage could be due to greater gut fill [15,20,27]. Therefore, there is a need for further investigation of the relationship between gut fill and ADG.

Poor performance has also been commonly observed when including forage in the diet [15,17]. Hill et al. [15] reported that feeding either 2.5 and 5% of chopped timothy hay linearly reduced starter intake, ADG, empty body weight ADG, and FE. The quality of forage significantly influences the digestibility and the palatability of the diet [46]. Ülger et al. [47] compared two calf total mixed ration (TMR) diets with either 20% of a high-quality alfalfa hay (CP = 18.1%, acid detergent fiber (ADF) = 36.1% and neutral detergent fiber (NDF) = 44.4% on DM basis, relative feed value (RFV) = 127.2) or lower quality wheat straw (CP = 3.7%, ADF = 52.4% and NDF = 80.1% on DM basis, RFV = 55.9) and found that the high-quality roughage improved FE and numerically increased ADG during the preweaning period. In a more recent study, Hill et al. [48] found that moderate to low-quality grass hay (5.4% CP and 62.8% NDF on DM basis) reduced the digestibility of DM, OM, and CP in young calves consuming a textured starter. It is noteworthy that the type of hay was not specified in this study, and we speculated that timothy or mixed hay was included based on other studies at the same period by the authors [49]. On the contrary, Castells et al. [21] and Hosseini et al. [40] reported that a low-quality straw (CP = 4.2%, and NDF = 74.0% on DM basis) could also improve DMI and ADG. The inconsistency in results on calf performance when providing low-quality hay was likely due to the different amount of milk offered, which may affect the solid feed consumption and forage preference as discussed below.



*Animals* **2020**, *10*, 188



#### *Animals* **2020**, *10*, 188

total amount fed during the whole trial period, while milk feeding rate equals the average daily milk

indicates a significant effect (*<sup>p</sup>* < 0.05), \* indicates a tendency (*<sup>p</sup>* < 0.1), N indicates a negative effect, *p* indicates positive effect, and NS shows no significant effect. DMI and ADG were

evaluated by comparing calves fed with or without forage.

allowance/birth

 body weight (%). 4 Outcomes = effect of forage inclusion, \*\*

Castells et al. [21] evaluated ad libitum provision of different types of forages in the diets from 2 weeks of age and found that the inclusion of alfalfa hay (CP = 16.6%, ADF = 30.2%, and NDF = 40.2% on DM basis), rye-grass hay (CP = 6.8%, ADF = 35.1%, and NDF = 59.3% on DM basis), oat hay (CP = 8.4%, ADF = 31.8%, and NDF = 59.6% on DM basis), barley straw (CP = 4.2%, ADF = 42.5%, and NDF = 74.0% on DM basis), corn silage (CP = 8.6%, ADF = 25.2%, and NDF = 41.9% on DM basis), or triticale silage (CP = 7.5%, ADF = 42.3%, and NDF = 64.7% on DM basis) resulted in similar or increased intake and gains without impairing FE and nutrient digestibility. Increased DMI and ADG were observed when oat hay, barley straw, or triticale silage were offered, and the inclusion of alfalfa hay did not exhibit similar benefits, probably due to the preference for this high-quality and palatable forage [21]. Indeed, forage intake was highest in calves fed alfalfa hay (14% of total solid DM) compared with the other treatments (oat hay: 8% rye-grass hay: 4%, barley straw: 5%, corn silage: 5%, and triticale silage: 4%). The higher proportion of forage to concentrate ratio may limit the DM digestibility and hence restrict the DMI and ADG, as we have already discussed [5,35]. More recently, a meta-analysis by Imani et al. [34] evaluating the effect of forage provision on growth performance of dairy calves using 27 published studies from 1998 to 2016 revealed that concentrate DMI was higher in calves offered alfalfa hay compared with those offered other types of forages.

#### 2.1.2. Forage Physical Forms and Processing

The focus is not only on the forage source and level of feeding, but also on the physical form and processing of forage. Mirzaei et al. [27] evaluated the effects of particle size (short at 2.92 mm vs. long at 5.04 mm as geometrical means) of alfalfa hay on growth performance of dairy calves at two different inclusion rates (low at 8% vs. high at 16% on DM basis). The authors observed no differences in growth rates between calves fed with or without hay, but greater DMI and weaning BW were found in calves fed low levels of alfalfa with a long particle size (8% and 5.04 mm) and high levels with a short particle size (16% and 2.92 mm) compared with calves fed low levels with a short particle size (8% and 2.92 mm) and high levels with a long particle size (16% and 5.04 mm). The short particle size at a low level of alfalfa might not have the potential to increase the capacity, motility, and development of the rumen [11,60], while the negative effect on performance by the long particle size at a high level might be attributed to the lower digestibility rate of long particles compared with short particles [61]. Montoro et al. [62] also found that when calves were supplemented with 10% of long chopped (3 to 4 cm) ryegrass hay, DMI, ADG, and FE were greater than those fed 10% finely ground (2 mm) grass hay. Longer particle size improved performance [8], probably because of the increased rumination time of calves, which increased saliva production, and consequently improved buffering effect on the ruminal environment [63,64]. However, inconsistent results have been obtained by Omidi-Mirzaei [65] and Suárez-Mena [66]. Omidi-Mirzaei et al. [65] reported that when calves were fed forage with different particle size (alfalfa hay: short = 1.96 mm or long = 3.93 mm; and wheat straw: short = 2.03 mm or long = 4.10 mm as geometric mean), rumination time increased in calves fed forage with long particle size, but concentrate DMI, ADG, and FE were not affected. Suárez-Mena et al. [66] compared four different particle sizes (0.82, 3.04, 7.10, and 12.7 mm as geometric mean) of low-quality forage (5% straw) mixed in the diet and observed no effect on DMI, growth performance, and minimal changes in rumen fermentation and pH among treatments. In summary, these results imply that interactions may exist among forage source, level, and particle size, and the optimal inclusion level of forage should be determined based on the forage source and particle size.

In recent years, attempts have been made on alternative ways to increase solid feed consumption of dairy calves, such as using non-forage fiber [39], silage based feed [55,67], moisturized starter [68,69], and reconstituted hay [58,59,70]. Beet pulp is a common source included in the diet as a non-effective fiber source. Maktabi et al. [39] observed that 10% of beet pulp in the diet improved DMI and ADG compared to a control group (no fiber inclusion), but growth was not enhanced when 20% of beet pulp was used. Inclusion of corn silage early in life of dairy calves has recently gained more interest [55,67]. In an experiment that compared supplementing 15 against 0% of corn silage, DMI, ADG, and BW increased probably because of the higher moisture content of corn silage that contributed to reduced dustiness and increased palatability of the feed [55,67]. However, feeding a high level of corn silage (30 or 60%; 75 or 100%) offered no benefits compared with feeding concentrate alone [43,44]. Suárez et al. [43] reported that substitution of 30 or 60% of the concentrate by corn silage did not affect DMI and ADG but feeding 30% of straw reduced DMI. Kehoe et al. [44] also found no differences in DMI and ADG when including 0, 75, or 100% of corn silage in pre-weaned calf diet but fed solely corn silage diet stunted the growth of rumen papillae and tended to impair intestinal morphology. Hence, it is possible that corn silage can be used to partly replace the concentrate with little harmful effects on the growth and development of the calf.

It has been documented that moisturizing the concentrate starter feed by adding water to change the DM from 90 to 50% increased DMI, ADG, and VFA production in dairy calves [68,69]. More recently, hay processing by reconstituting with water was evaluated in a series of studies published by Kargar et al. [58,59,70,71]. Hay was soaked in water for 24 h and mixed every 6 h to obtain a theoretical DM content of 20% [71]. This method has been used previously in the diets of mature cows to increase fecal consistency [72] and reduce digestion lag time in the rumen as a result of a higher fiber digestibility [73]. Kargar et al. [70,71] replaced dry alfalfa hay (10%) with a similar amount of reconstituted alfalfa hay, resulting in similar DMI and ADG, but higher NDF digestibility during the pre-weaned period. Furthermore, a greater improvement in health status (fecal score and general appearance score) was obtained with reconstituted hay, possibly due to decreased dustiness, similar to corn silage [70]. Therefore, corn silage, reconstituted hay, and beet pulp can be used interchangeably in dairy calf diets based on availability and the relative feed price [58,59].


*Animals* **2020**, *10*, 188


**Table 3.** *Cont.* geometricparticlelength,usingforage total amount fed during the whole trial period, while milk feeding rate equals the average daily milk allowance/birth body weight (%). 4 Outcomes = effect of forage inclusion,indicates a significant effect (*<sup>p</sup>* < 0.05), \* indicates a tendency (*<sup>p</sup>* < 0.1), N indicates a negative effect, *p* indicates a positive effect, and NS shows no significant effect. DMI and ADGevaluated by comparing calves fed with or without forage.

 \*\*

 were

#### *Animals* **2020** , *10*, 188

#### 2.1.3. Time and Method of Offering Forage

Time [24,75,76] and method [25,77,78] of offering forage are essential factors that can influence how dairy calves utilize forage. While Wu et al. [24] found no differences in DMI, ADG, and rumen development in calves fed alfalfa hay or oat hay either at day 3 or 15 of age, different outcomes were shown in two other studies, investigating the effect of age at which alfalfa hay [75] and oat hay [76] were introduced to calves. Both studies observed improved DMI and growth performance with forage provision, and the greatest growth performance and rumen development were obtained in calves offered hay from the 2nd week rather than the 4th or 6th week of age [75,76]. Based on these studies, suggestions may be made to include alfalfa or oat hay in diets of calves as early as week 2 or even right after birth in order to improve DMI and ADG in dairy calves [24,75].

Forage have been provided as a mixture with concentrate in previous studies, while recent studies investigated calf preference to different feeds by providing forage and concentrate separately. Castell et al. [21] observed a greater DMI and ADG in calves provided forage rather than without forage. In this study, calves consumed around 5% of forage when it was offered ad libitum and separately from the concentrate. However, several studies did not observe a positive effect on DMI when part of the concentrate was substituted for forage before weaning [43,47]. In research by Ülger et al. [47], who mixed forage at 20% with the concentrate and Suárez et al. [43] at either 30 or 60%, calves consumed at a predetermined fixed forage to concentrate ratio which was much greater compared to calves fed free choice. Therefore, the greater forage proportion in the mixture diet (i.e., containing 30 or 60% forage) [43] might mask the positive effect on DMI and growth performance compared with forage consumed voluntarily (around 5% forage) by calves [21]. A meta-analysis study has proven this inference; when forage was offered separately, starter feed intake and ADG increased compared to a mixed ration diet [34]. On the contrary, some studies claimed that DMI and growth performance were not different between two feeding methods (mixed vs. separate) [77,78], possibly due to the low proportion of forage (15 or 10%) in mixed ration, which was similar to that (11 or 10%) with separate forage provision in studies by Overvest [77] and EbnAli [78]. Moreover, the high level of milk feeding (around 13 or 26% of birth weight) might have decreased the solid feed consumption [77,78]. Although the forage feeding method may not always lead to better performance, it certainly affects the expression and development of dairy calf behavior [25,79], as discussed below.



total amount fed during the whole trial period, while milk feeding rate equals the average daily milk allowance/birth body weight (%). 4 Outcomes = effect of forage inclusion, \*\* indicates a significant effect (*<sup>p</sup>* < 0.05), N indicates a negative effect, *p* indicates positive effect, and NS shows no significant effect. DMI and ADG were evaluated by comparing calves fed with or without forage.

#### *Animals* **2020**, *10*, 188

#### *2.2. Concentrate and Milk Factors*

#### 2.2.1. The Physical Form of Concentrate Feed

There are different forms and types of calf starters. Porter et al. [37] reported that whether forage was included or not, calves on coarse mash (average particle size: 2014 μm) ate and gained more than those on pelleted diets (average particle size: 888 μm). Moreover, rumination was initiated earlier. Hence, up to 8 weeks of age, calves raised on a complete concentrate diet without forage did not experience a significant depression in growth performance, which might be due to the long particle size of coarse mash feed initiating rumination early and preventing bloat and parakeratosis in the rumen [65]. Two consecutive experiments were conducted by Terré et al. in 2005 to evaluate the influence of the physical form of concentrate feed (textured or pelleted) with or without forage inclusion on the performance of young calves. Calves receiving pelleted concentrate feed with straw exhibited a greater solid intake and higher rumen fluid pH compared with those receiving a pelleted concentrate feed without straw. Calves that received the texturized (containing whole corn) starter feed had equivalent rumen fluid pH to those fed a pelleted concentrate with straw. However, rumen fluid pH and performance were not improved when another texturized (containing rolled mixed grains) concentrate feed was offered [80]. These results show that the physical forms of concentrate feed may affect the calf performance and rumen environment differently. For example, calves fed a texturized concentrate feed containing whole corn had a greater rumen fluid pH than steam-flaked corn, dried-rolled corn, and roasted-rolled corn [81], likely because the calves spent a longer time chewing the whole corn feed, which increased saliva production, hence neutralizing the rumen pH and acids. In another study involving a mashed (with or without corn silage) and a textured concentrate (with or without corn silage), regardless of the physical form of concentrate feed, forage inclusion resulted in greater DMI, ADG, and final BW than non-included calves [67]. It was evident that forage provision had more effect on the growth performance than the physical form of the feed, whereby steam-flaked grains were the main component in the concentrate. In agreement, Mojahedi et al. [82] reported that including alfalfa hay could improve DMI and ADG of calves fed steam-flaked corn, as opposed to a cracked corn-based diet, probably because of higher amounts of gelatinized starch in the steam-flaked corn (44.1 vs. 12.5% of total starch, respectively). Possibly, forage inclusion enhanced starch fermentability of the steam-flaked corn through the provision of effective fiber. Collectively, a decrease in solid feed consumption in calves fed finely ground or pelleted starter on commercial farms compared with those fed textured concentrate [83] warrants forage provision to improve the solid feed intake, growth performance, and rumen environment to a greater extent [80,82].


**Objectives**

**Physical**

**Form of** **Concentrate**

Pellet, Mash

Pellet,

3

 11

 49

 7

 Ryegrass hay

> Texture

Pellet,

3

 20

 52

 8

 Ryegrass hay

> Texture

Mash,

4

 12

 56

 3

 Corn silage

 0, 15%

 4

 16

 28

 3

 Beet pulp

 0, 16%

Ad libitum (0, 6.8,

11.6%) Ad libitum (0, 4.3%)



GMPL: 0.5, 1.1, 3.0,

TMR

291 L

*p* **\*\***

*p* **\*\***

 Free choice

 Free choice

 -

 TMR

80 L (Around 7%) 274 L (Around 16%) 233 L (Around 13%)

**NS**

 **NS**

*(Trail 2)* [80] *Mirzaei* et al.,

**NS**

 **NS**

*(Trail 1)* [80]

*Terré* et al., 2015

*Terré* et al., 2015

**NS**

 **NS**

*Porter* et al., 2007 [37]


Texture and 4.0 mm (Around 13%) 2016 [67] 1Trt = Treatment. 2 GMPL refers to geometric mean particle length, which was determined using ASABE forage sieve methods (ANSI/ASAE S424.1) [45]. 3 Amount of milk fed is the total amount fed during the whole trial period, while milk feeding rate equals the average daily milk allowance/birth body weight (%). 4 Outcomes = effect of forage inclusion, \*\* indicates a significant effect (*<sup>p</sup>* < 0.05), N indicates a negative effect, *p* indicates positive effect, and NS shows no significant effect. DMI and ADG were evaluated by comparing calves fed with or without forage.

**Reference**


**Table6.**Asummaryofstudiesonforageinclusionindairycalvesbasedondifferentvolumesandmethodsofmilk

16

indicates a significant effect (*<sup>p</sup>* < 0.05), indicates a tendency (*<sup>p</sup>* < 0.1), N indicates a negative effect, *p* indicates positive effect, and NS shows no significant effect. DMI and ADG were evaluated by comparing calves fed with or without forage.

#### *Animals* **2020**, *10*, 188

#### 2.2.2. The Amount and Method of Milk Feeding

Most studies that suggested exclusive concentrate feeding were conducted with calves fed low amounts of milk [15,36]. For example, in the Hill et al. [15] study, only 120 L of milk was fed to calves before 28 days of life (weaning date), averaging around 4 L/d of milk (approximately 10% of birth body weight) which is insufficient for optimal growth. This low milk feeding rate might stimulate greater concentrate intake in calves to make up for the deficit in energy requirements. Indeed, a strong negative correlation between liquid and concentrate feed intake has been elucidated in a meta-analysis that shows calves fed high milk or milk replacer resulted in limited daily starter intakes [28].

As discussed earlier, compared to concentrates, forage are bulkier and are less digestible and have lower fermentation rates [5], which can lead to a low voluntary intake when low energy forage is offered separately or as a mixture with the concentrates [8,86]. Castells et al. [21] claimed that when calves were offered different forages (hay, straw, and silages, respectively) with concentrates ad libitum and separately, forage consumption was only 4–6% of the total solid feed intake. Interestingly, the proportion of hay consumed across studies seems to range from 3 to 45% of total solid feed intake [20,21,25,52,87]. The difference in the proportion of forage consumed across multiple studies may also depend upon milk feeding amounts. In two studies feeding different amounts of milk, Castells et al. [52] reported that calves consumed 3% of total solid feed as forage at a low level of milk feeding (214 L from d 0 to d 57, averaged 4 L/d, around 10% of birth body weight), while Xiao et al. [52] reported a greater ratio of forage to total solid feed intake, approximately 45% when a high amount of milk was offered (376 L from d 0 to d 56, averaged 6.8 L/d, around 17% of birth body weight) [25]. Milk contains a high content of fat and sugar, which provide the energy required by the calves, and greater milk amounts might alter concentrate requirements [25]. This speculation concurs with another study in which calves fed low amounts of milk consumed more concentrates, resulting in a lower ratio of forage to total solid feed intake in a low compared to a high milk feeding group (13.2% vs. 18.6%) [84].

Feeding patterns and methods could also affect forage intake in calves. When investigating the effect of either step down (fed at around 15% of birth body weight per day) or conventional (approximately 10% of birth body weight per day) feeding patterns in dairy calves, Khan et al. [88] found that the former had better performance. In a different study, Daneshvar et al. [23] reported that when similar amounts of milk were fed using different feeding patterns (step down vs. conventional), solid feed intake did not differ between treatments. Horvath et al. [85] showed that the feeding method (bucket vs. teat feeding) did not alter the forage and concentrate intake. Hence, milk allowance might have a greater impact on solid feed consumed by calves as opposed to the milk feeding pattern or method.

Limited studies directly investigating the relationship between milk allowance and forage consumption in pre-weaned calves are available, which calls for scientists to turn their attention to this area, especially with more farms leaning towards high milk volume feeding. Forage inclusion can promote total solid feed consumption and BW gain in calves, but factors such as the amount of forage, forage sources, forage feeding method, physical form of forage and concentrate, and milk allowance might confound these benefits. Calves should be slowly introduced to relatively low levels of forage while guarding against the use of low digestible forage (i.e., straw), which may depress total DMI and BW gain. Moreover, forage should be available free-choice and in separate containers from concentrate feed.

Table 7 shows a summary of selected studies that determined the effects of forage inclusion on performance, rumen fermentation and development, and expressive behavior in dairy calves. While the effects of feeding forage on performance, such as DMI and ADG, were controversial, relatively consistent results were obtained in other parameters, like rumen fermentation and expressive behavior.



Forages included dry hay, silage, straw, and by-products (e.g., cottonseed hulls) and reconstituted hay. Positive, negative, and no effect on a parameter was determined by adding forage in the diet compared with no forage inclusion in those studies. "Positive effect" represent an increase or improved effect (*<sup>p</sup>* < 0.05), "Negative effect" represent a decreased effect in the related parameter (*<sup>p</sup>* < 0.05), "/" means no studies were found to affect this parameter in the current review (*<sup>p</sup>* > 0.05).2 Parameters were measured in dairy calves within 3 months of age. Non-nutritive oral behavior/abnormal behavior included tongue rolling, licking buckets, pen or surface, sni ffing, vocalizing, and eating beddings; Satisfaction behavior included tail swishing, self-grooming, and rubbing.

1

#### **3. Rumen Environment**

#### *3.1. Rumen Fluid pH and Fermentation*

The rumen is the largest and most crucial compartment of the digestive system in adult ruminants, as it is vital for acquiring metabolic substrates through microbial fermentation. Although young calves have an undeveloped rumen, fermentation begins at a very early age [89] and may affect the development and health of the rumen. Prolonged low rumen fluid pH may cause subacute ruminal acidosis (SARA) in adult cows, which is well defined in beef feedlot cattle (pH < 5.8 for 3 h/d) [90] and dairy cows (as periods of moderately depressed pH, from about 5.0 to 5.5) [91]. However, in young calves, ruminal acidosis has not been clearly defined. Previous studies reported that rumen fluid pH in young calves is often well below 5.8 [89,92]. Some researchers believe that, as in mature cows, dairy calves can experience ruminal acidosis [34,89], probably due to the high amount of concentrate fed [92] in artificially rearing systems and the relatively low saliva [93] secreted at a young age.

Concentrate feed, high in rapidly fermentable carbohydrates, such as sugar and starch, provides energy for optimal growth, but the fermentation rate tends to generate lots of VFA and lactic acid, resulting in low rumen fluid pH [51]. Forage, high in fiber, may play a role in mitigating this challenge. Most of the studies (21 studies, accounting for 84% of summarized studies) explored in this review showed a positive effect of forage inclusion on rumen fluid pH in dairy calves, while very few reported no difference (four studies, accounting for 16% of summarized studies) or negative impact (Table 7). In agreement with a previous meta-analysis, our literature search showed that forage could improve the rumen fluid pH when supplemented to calves, though it might be dependent on the forage source [38,40]. Alfalfa hay is more likely to modulate rumen fluid pH during the milk-feeding period than other types of forages [33]. Maktabi et al. [39] reported that increasing fiber content by adding beet pulp (10% and 20%) in the concentrate diet failed to improve rumen fluid pH, while supplementing alfalfa hay (10%) resulted in a significant improvement in this parameter by providing more effective NDF. Terré et al. [53] also demonstrated that increasing NDF content (18.2 vs. 26.7%) by adding soybean hulls in the pelleted starter could not alter rumen fluid pH, but adding chopped oat hay, containing more effective fiber, could improve the ruminating behavior resulting in a higher pH. In agreement, Laarman et al. [94] reported a positive relationship between forage intake and rumen fluid pH, while SARA (rumen fluid pH below 5.8) could be exacerbated when calves are fed less than 0.08 kg/d [64], suggesting that even small amounts of forage consumption (timothy hay, 0.08 kg/d) can reduce rumen acidosis in calves.

The possible reasons for the increased rumen fluid pH when adding forage in the diet are multifactorial. On the one hand, forage is bulkier and has lower digestibility compared to the concentrate. The higher forage consumption leads to increased intake of effective fiber, which in turn stimulates the chewing (ruminating and eating) activity of calves [56,95], and subsequently improves the saliva production and rumen buffering [38,63]. On the other hand, the rapidly fermentable carbohydrates generate abundant VFA that may exceed and overwhelm the absorptive capacity of the undeveloped rumen [96]. Feeding forage in pre-weaned calves could reduce the concentration of VFA [19,23,25,35,40,52–54,67] and decrease rumen plaque formation [42,52], increasing the absorptive surface area of the rumen epithelium and hence reducing the accumulation of VFA and maintaining the appropriate rumen fluid pH. In addition, an increased passage rate in the rumen was observed in calves fed forage compared to those fed concentrate only, which lowered the feed retention time in the gastrointestinal tract (28.4 h for concentrate feed vs. 18.8 h for oat hay group), fermentation time and VFA concentration [52]. In the same study, calves fed forage tended to have a higher expression of monocarboxylate transporter-1 in the rumen wall [52], which plays a central role in transporting acetate, lactate, and protons from the rumen lumen to the bloodstream [97,98], hence alleviating VFA accumulation in the rumen as well.

A greater acetate [22,23,40,43,52,53,67], and a lower propionate [25,34,40,43,54,67], butyrate [23,25,27,39,40,53], and valerate [23,25,52,53,67] concentration/proportions have been reported

in calves fed with forage than those fed only concentrate. These dynamics in fermentation patterns are probably related to the changes in the rumen microbial ecosystem. For example, cellulolytic microbes, such as *Ruminococcus flavefaciens* and *Ruminococcus albus*, are more prevalent in animals fed high forage diets, which increase fiber degradation and elevates the proportion of acetate in the rumen [42,52]. Both propionate and butyrate stimulate and enhance rumen epithelial development [12,13,99], with butyrate serving as the preferred energy source as well as modulating the gene expression in the rumen epithelium [96]. A low proportion of these two VFA may limit the growth of rumen papillae [42,52]. Due to its relatively low proportions, the valerate has received little attention. It has been suggested that cellulolytic microbes utilize valerate in the rumen [100], which might explain its decrease in calves supplemented with forage in their diets.

Lactic acid decreases in the rumen when forage is included in the diet [51], which might end up positively altering rumen fluid pH. Terré et al. [53] found an interesting relationship between rumen VFA and rumen fluid pH. When rumen fluid pH was above 5.1, total VFA and rumen fluid pH were linearly correlated; however, when it fell below 5.1, the correlation disappeared. The implications are that lactic acid (a much stronger acid than VFA) may alter rumen fluid pH at a pH below 5.1 [53,101]. In adult cows, an acute ruminal acidosis was observed, with excessive consumption of concentrate feed leading to a sudden and uncompensated drop in rumen fluid pH (below about 5.0). Owens et al. [60] showed that lactic acid concentrations increased with a decline in rumen fluid pH. However, when rumen fluid pH (around 5.0–5.5) was moderately depressed, lactic acid accumulation was inconsistent [102] or transiently fluctuated [103]. Hence, although moderate depression in rumen fluid pH may cause SARA in dairy cows, it is not because of the lactic acid accumulation, but might be due to the accumulation of VFA alone [102].

Collectively, previous studies demonstrate that forage provision has a positive effect on rumen fluid pH and alters rumen fermentation in calves. However, the majority of these studies assessed rumen fluid pH and VFA at only a single time point. Further research is encouraged to test the dynamic changes in rumen fluid pH and VFA when different types, amounts, timing, and particle sizes of forage are supplemented in the diet, which may help us define SARA in young calves more accurately.

#### *3.2. Rumen Microbes*

A developed rumen is full of microorganisms that ferment and degrade multiple nutritional fractions (sugar, starch, fiber, protein, fat, and so on) and provides necessary metabolic substrates and nutrients to the dairy cattle. The microbial ecosystem differs between young calves and adult cows [104]. At birth, young calves possess no anaerobic microorganisms in the rumen [105], with recent evidence suggesting that colonization occurs immediately after birth [104,106]. Dominant microbes that are involved in normal rumen function of mature cows are present as early as one day of age [106]. This colonization of microorganisms and the presence of substrates trigger fermentation activity, which then provides indispensable nutrients for rumen development. It might take as long as a year for the rumen to mature and for calves to establish a stable rumen microbiota system with many factors involved [104].

Both liquid and solid feed appear to affect the microbial community in young calves [107]. In this review, our discussion is restricted to the effect of forage feeding on the microbiota of young calves. Castells et al. [52] reported that forage supplementation (alfalfa) numerically increased cellulolytic microbes (*Ruminococcus albus*) compared with calves fed only concentrate feeds. Similarly, Kim et al. [54] observed significantly higher copy numbers of cellulolytic bacteria (*Ruminococcus flavefaciens* and *Ruminococcus albus*) in calves supplemented with forages (orchard and timothy hay). Early studies also evaluated the effects of the physical form (finely ground, 1 mm theoretical particle size vs. unground, 0.64 cm theoretical particle size) of diet on rumen microbiota with two identical diets (25% alfalfa hay and 75% grain) that varied only in particle size. Calves offered the ground diet had a relatively lower rumen fluid pH and lower number of cellulolytic bacteria than calves fed the unground diet [89]. These results revealed that effective fiber might play a crucial role in changing the rumen environment, hence altering the microbial populations.

The next-generation sequencing (NGS) analysis has revealed that the major phyla in the rumen are *Firmicutes* (around 43%), *Bacteroidetes* (around 21%), *Actinobacteria* (around 18%), and *Proteobacteria* (around 4%) [108]. Relatively higher abundance of *Bacteroidetes* and lower abundance of *Actinobacteria* was observed in calves supplemented with forage compared to those fed only concentrate [54]. *Bacteroidetes*, the second most dominant phyla in calves' rumen, may stimulate the development of the digestive tract [109]. In adult cows fed concentrate feed, the relative abundance of *Bacteroidetes* dropped and cows were more susceptible to SARA [110]. The lower level of *Bacteroidetes* in calves fed only concentrate could be partly explained by greater feed intakes resulting in low rumen pH. Furthermore, Kim et al. [54] found that the most dominant genus in *Bacteroidetes* phylum was *Prevotella*, a highly active hemicellulolytic and starch-degraders [111] that mainly produce acetate. A relatively higher abundance of *Prevotella* may be related to the greater acetate proportion in calves offered forage [54]. Jami et al. [104] reported that *Prevotella* was the predominant genus in animals fed high-fiber diets rather than high-caloric diets. *Olsenella* is an important lactic acid-producing bacterium under the phylum *Actinobacteria* [112]. Forage inclusion decreases *Olsenella* (hay, 3.9% vs. concentrate, 13.2%) significantly, which contributes to the lower abundance of *Actinobacteria* (hay, 4.7% vs. concentrate, 13.9%) [54]. Thus, it can be speculated that forage inclusion in the diet might affect the growth of lactic acid-producing bacteria (such as *Olsenella*) by limiting the proportion of rapidly fermentable substrates (e.g., starch) replaced by fiber. In contrast, although numerical differences were observed in *Bacteroidetes* and *Actinobacteria* when evaluating the effect of forage supplementation, Lin et al. [108] reported that neither alpha nor beta diversity indices and microbiota were significantly different among the dietary groups [108], possibly because of the volume of milk (Lin, 252 L, 10% vs. Kim, 88 L, 4% of birth bodyweight) fed to calves, resulting in varying solid feed consumed. Alternatively, different forage sources (Lin, Oat hay vs. Kim, Timothy) and feeding levels (Forage/Total Solid: Lin, 6% vs. Kim, 20%) (Tables 2 and 4) might have led to insufficient forage consumption causing the changes in the composition of rumen microbiota in the Lin et al. study. These results indicate that the population of most predominant microbiota (e.g., *Bacteroidetes* and *Actinobacteria*) in the rumen is closely related to the amount and type of solid feed consumed. Further studies need to focus on these major groups to illustrate the relationship between rumen microbiota, fermentation and feed consumption in dairy calves.

It is important to note that the effect of various nutritional (sugars, starch, rumen degradable protein, NDF, or ADF) and physical (effective forage fiber or non-forage fiber) fractions on rumen microorganisms in young calves is not clearly defined in literature. Furthermore, most of the studies evaluating microbial and molecular changes in young calves have only studied a small number of microbes of interest. With the rapid development of the NGS technologies in the past 20 years, we recommend the exploration of the global changes in microbial abundance to better understand substrate fermentation and absorption and epithelium development in the rumen as forage are supplemented to calves in early life.

#### **4. Feeding Behavior**

#### *4.1. Ruminating and Eating Behavior*

Feed experiences and behavior development in early life might affect the behavioral expression of adult ruminants [79]. In the last decade, researchers have increased their attention on the development of calf behaviors, such as eating, standing, lying, and ruminating when forage is included in the diet. Forage inclusion in the diet undoubtedly increases chewing in calves even before weaning when only a small amount of solid feed is consumed. Increased chewing activity may be as a result of higher rumination [21,39,40,53,55,56,75–78] or the total time spent eating [39,40,57,76,77] when calves are fed forage.

In newborn ruminants, rumination is initially absent and emerge a few weeks after birth [113]. Providing forage to young calves can accelerate the development of rumination behavior [114,115]. van Ackeren et al. [116] observed that chewing time was lower in calves receiving a low NDF diet (26.2%) compared with those receiving a high NDF diet (31.3%). Porter et al. [37] claimed that calves began ruminating by week 4 of age when fed a more physically effective solid feed, while those fed a finely pelleted feed began ruminating from week 6. Rumination is crucial in ruminants helping maintain the rumen fluid pH by stimulating saliva production that neutralize VFA and lactic acid in the rumen and thus to maintain a healthy rumen environment [105].

Meal feeding patterns (meal size, frequency, and duration) can also impact the rumen environment. Generally, rumen fluid pH declines rapidly after feed ingestion, and the rate of decrease is associated with meal size and feeding frequency [117]. Large meal sizes and infrequent meals may result in a greater drop in rumen fluid pH post-ingestion. Horvath et al. [57] illustrated that the provision of forage not only increased the total eating time but also influenced the solid feed meal patterns. An improved meal frequency and duration were observed in their study, which leads to relatively slower post-prandial drops in rumen fluid pH, potentially decreasing the risk of SARA [63].

#### *4.2. Sorting Behavior*

Feed sorting is well demonstrated in adult cows, since they are highly sensitive to sweet taste [118]. Probably, the preference for sweetness reflects the inclination towards higher energy demands, hence the tendency to sort out for concentrates (sweet, high energy-density) in a total mixed ration [119]. The sorting out of the mixed ration can result in an unbalanced nutrient intake, whereby cows sort out for the rapidly fermentable cereals as opposed to forage, leading to a drop in rumen fluid pH, and hence inducing SARA [120]. Ingesting excessive fermentable concentrate feed can result in rumen acidosis [63]. In turn, the sorting behavior is altered further, leading the animals to choose the part of diets with longer particle size and slower fermentable rate [121]. These results suggest that ruminants develop feed preferences based on post-ingestive feedback [122] and they may be biased towards choosing certain nutrients as the situation demands.

Feed sorting is also seen in the early life of calves. When feeding concentrates and forage free-choice, variation in the proportion of forage to total solid intake was observed (ranging from 5 and 45%) [21,25,87]. The changes in dietary selection across multiple studies may depend on forage related factors and milk feeding allowance, as has been discussed above. It is interesting to note that feed preference and sorting can be established early and persist later on in life. Miller-Cushon et al. [123] reported that calves fed either concentrate or forage before weaning were likely to consume the feed that they were already familiar with, even when switched to a mixed diet after weaning. Similarly, our research group found that calves are likely to eat feed they were originally introduced to and familiar with even after switching to a free-choice diet, though this effect only lasted for a short period. However, after switching the diet at weaning, the provision of both concentrate and hay separately early in life led to a greater hay intake ratio (35.6%) than providing concentrate (17.7%) or hay (16.5%) solely before weaning. Furthermore, exposure to a diet of both concentrate and hay early in life could numerically improve the calves' ability to sort for long particles 6 months later [25]. Therefore, these results suggest that early exposure to feed experience can affect the feed preference immediately after switching diets and may have a long-lasting effect. The feeding method may also play an essential role in influencing the learning of feed sorting behavior. When we compared three different feeding methods (solely concentrate, separated concentrate and forage, mixed concentrate and forage for the first month; data unpublished) in 2 month old calves, the lowest sorting activity was observed in calves fed concentrate and forage separately. Hence, calves exposed early to a diet of concentrate or mixed ration are likely to sort for fine-grain particles, probably because these calves have already established their sorting behavior, which can last even after changing to a new mixed diet (data unpublished). Similarly, the provision of solid feed in pre-weaned calves as separate components reduced the extent of feed sorting after weaning compared to offering the diet as a mixed ration [87]. As already stated,

feed sorting is likely to influence rumen fluid pH and may lead to SARA. Separately feeding different solid feed components at the same time may avert sorting for fine particles when compared to feeding solely concentrate or mixed diets and might lead to a more stable rumen fluid pH and a healthier rumen in the calf. However, we cannot ignore the fact that the effect of forage inclusion on sorting behavior is dependent on a myriad of other factors (e.g., forage source, level, physical form, dry matter, and milk allowance).

#### *4.3. Other Behaviors*

Access to forage by dairy calves may also reduce the occurrence of other non-nutritive oral and abnormal behaviors [21,40,53,55–57,76,85], such as tongue rolling, licking of buckets, pen or surface, sniffing, vocalizing, and eating the bedding material. Horvath et al. [85] demonstrated that providing forage decreased the non-nutritive oral behaviors, and when combined with feeding milk by teat, the effects were more significant compared with bucket feeding. These results further buttress the fact that liquid and solid feeding can influence the development of pre-weaned calf behaviors. Furthermore, supplementation of good quality forage increased other behaviors that may indicate satisfaction (tail swishing, self-grooming, and rubbing) [56]. Worthy to note is that the decline in non-nutritive oral behavior may have also reduced the formation of hair and fiber balls in the rumen [4], which have been associated with poor health and growth of calves. Further research is encouraged to explore whether forage inclusion early in life would have a long-term effect on sorting and other behaviors.

#### **5. Conclusions**

Understanding factors that influence responses to forage inclusion in pre-weaned calves is of significant importance from a management point of view because the effect of offering forage on calf feed intake and growth rate has been inconsistent. In recent studies, a small amount of good quality forage such as alfalfa supplemented in the diet is likely to improve the DMI and ADG. However, these performances are dependent on the type of concentrate and the amount of milk offered. Although controversy remains on whether forage improves growth rate, it has been well documented that its inclusion early in life can help with the establishment of feeding behavior, leading to greater rumination and eating behavior as well as lowering the abnormal feed sorting behavior. All these positive effects can result in a higher rumen fluid pH and a more stable rumen environment, with a corresponding positive effect on rumen microbiota and fermentation. Further research is required to understand the long-term effects of offering forage to pre-weaned calves, since animal-related factors, such as feed selection and sorting, established early in life may persist later on in life.

**Author Contributions:** The paper was mainly conceived and designed by Z.C. Research articles were collected and analyzed by J.X. and G.M.A. The manuscript was mainly written by J.X. and G.M.A. and edited by J.L., Y.W., S.L., and Z.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Review* **Review of Strategies to Promote Rumen Development in Calves**

#### **Qiyu Diao 1,\*,**†**, Rong Zhang 1,2,**† **and Tong Fu <sup>3</sup>**


Received: 22 May 2019; Accepted: 23 July 2019; Published: 26 July 2019

**Simple Summary:** The rumen is an important digestive organ that plays a key role in the growth, production performance and health of ruminants. Promoting rumen development has always been a key target of calf nutrition. Current research reveals that an early feeding regime and nutrition have effects on rumen development and the establishment of rumen microbiota. The effects may persist for a long time, and consequently, impact the lifetime productive performance and health of adult ruminants. The most sensitive window for rumen manipulation may exist in the postnatal and weaning period. Thus, the early feeding regime and nutrition of calves deserve further research. The establishment of the rumen bacterial community is a mysterious and complex process. The development of microbial 16S rDNA gene sequencing and metagenome analysis enables us to learn more about the establishment of rumen microbes and their interactions in host gastrointestinal (GI) tract development.

**Abstract:** Digestive tract development in calves presents a uniquely organized system. Specifically, as the rumen develops and becomes colonized by microorganisms, a calf physiologically transitions from a pseudo-monogastric animal to a functioning ruminant. Importantly, the development of rumen in calves can directly affect the intake of feed, nutrient digestibility and overall growth. Even minor changes in the early feeding regime and nutrition can drastically influence rumen development, resulting in long-term effects on growth, health, and milk yields in adult cattle. Rumen development in newborn calves is one of the most important and interesting areas of calf nutrition. This paper presents a comprehensive review of recent studies of the gastrointestinal (GI) tract development in calves. Moreover, we also describe the effect of the environment in shaping the GI tract, including diet, feed additives and feeding management, as well as discuss the strategies to promote the physiological and microbiological development of rumen.

**Keywords:** calves; rumen; epithelium; microbiota; diet; feed additives; feeding management

#### **1. Introduction**

Rearing healthy calves is very important as it can have a significant impact on their growth and milk production performance in adult life. Adequate calf development is therefore crucially important for the entire dairy industry. Calves are challenged by a series of stress factors after they are born, including changes in their surroundings. Specifically, the living environment changes from the sterile uterus to natural outside conditions, in addition to changes in nutrition from that provided by the

mother to the digestion and absorption of feed by calves themselves. However, due to the poor immunity and the incomplete development of the digestive system in young calves, any interference from the external environment or changes to the nutrition can drastically affect the development of calves [1]. Some of the problems include diarrhea and slow weight gain, as well as respiratory tract disease, which can lead to high levels of morbidity and mortality, and pose significant challenges to breeding.

#### **2. Rumen Development**

Compared with monogastric animals, the forestomaches of ruminants have a specialized structure and function, which results in differences in digestion and physiology between ruminants and monogastric animals. Moreover, calves have additional unique systems that are present in their digestive tract during their development. At birth, the rumen is not completely developed, and significant changes in rumen development have to occur first before the calves can digest dry feed to guarantee their own growth needs. The specific changes include the development of the rumen organ and rumen epithelium, and the establishment of rumen microbiota. Understanding rumen development in newborn calves is one of the most important focus areas of calf nutrition.

#### *2.1. Rumen Organ Development*

The digestive system of young ruminants begins to develop during the embryonic period. For example, the stomach chambers are visible by day 56 in bovine embryos [2]. At birth, the weights of reticulorumen, omasum, and abomasum account for 38%, 13%, and 49% of the entire stomach weight, respectively [3]. By eight weeks of age, these proportions change to 61.23%, 13.40%, and 25.37% of the stomach weight, respectively [1]. Finally, at 12–16 weeks of age, they reach 67%, 18%, and 15% of the stomach weight, respectively [1,3] (Table 1). The esophageal groove, namely the rumoreticuler groove, is one of the unique features inside the gastrointestinal (GI) tract of calves. The majority of the liquid feed, such as colostrum, whole milk and milk replacer (MR), can bypass the rumen, reticulum and omasum, and flow directly into the abomasum as a result of the reflex closure of the esophageal groove. The abomasum of newborn calves is the only fully developed and functional stomach, and is also the most important digestive organ for calves at birth. The digestion of fat, carbohydrates, and protein is predominantly dependent on the digestive enzymes secreted by the abomasum and small intestine, which is similar to the digestive system in monogastric animals. Over time, with the increase in dry feed intake, the rumen begins to develop and starts to play more important digestive roles.


**Table 1.** The development of the forestomach.

<sup>1</sup> Each stomach compartment is expressed as a percentage of the total weight of the forestomach.

#### *2.2. Rumen Epithelium*

The ruminal epithelium performs many important functions and plays the key role in rumen development, including absorption, transportation, short-chain fatty acid metabolism, and protection. The proliferation and growth of the rumen squamous epithelium promotes the growth of papillae length and width, and increases the thickness of the interior rumen wall [4]. Work by Lesmeister and coworkers (2004) [5] considered the papillae length of the rumen as the most important factor

for the evaluation of rumen development, followed by the papillae width and rumen wall thickness. However, papillae per square centimeter is not used as an indicator of rumen development.

Newborn calves have a smooth epithelium with no prominent papillae. Calves fed solely with liquid feed have been shown to have limited rumen development characterized by decrease in rumen weight, papillary growth, degree of keratinization, pigmentation and musculature development [6,7]. Of note, increased intake of solid feed contributes to the rapid development of ruminal fermentation. As calves consume more starter feed, rumen digesta pH decreases, whereas volatile fatty acid (VFA) concentration gradually increases during the first two months. The molar proportion of acetate decreases during the first two months, and then starts to increase until nine months of age as forage intake increases [1,8]. The presence and absorption of VFAs in the rumen provides chemical stimuli required for the proliferation of rumen epithelium [6,9]. Importantly, intraruminal administration of acetate, propionate, and butyrate can stimulate the growth of rumen epithelium in young ruminants, with the effect of butyrate being the most prominent, followed by propionate [4,6]. Studies suggest that rumen papilla proliferation is associated with increased blood flow through the rumen wall [10,11] and a direct effect of butyrate and propionate on gene expression [12].

Despite many studies indicating that VFA can promote the development of rumen epithelium in vivo, the in vitro results suggest the opposite. For example, butyrate treatment decreases DNA synthesis of rumen epithelial cells in culture [13], while the proliferation of rumen epithelial cells is inhibited by rumen fluid in vitro [14]. The divergent in vivo and in vitro response may be linked with an indirect hormonal response to VFA metabolites. Several hormones, such as insulin, pentagastrin, and glucagon, have been implicated as possible VFA mediators that stimulate rumen epithelial proliferation [12,15]. A previous study by Baldwin (1999) reported that proliferation rates of rumen epithelial cells induced by insulin, epidermal growth factor, and insulin-like growth factor (IGF-1) were 75%, 97% and 96%, respectively [16]. Importantly, other studies also suggested that insulin, epidermal growth factor, and IGF-1 can overcome the inhibitory effect of butyrate [16,17].

#### *2.3. Ruminal Microbiota*

At birth, the GI tract of young ruminants is sterile. During the first hours of life, the forestomach becomes rapidly colonized with an abundant bacterial population. The neonates acquire bacteria from the dam, partners, feed, housing and environment. The early gut microbes of suckled lambs were mainly derived from the mother's teats (43%) and ambient air (28%), whereas those of bottle-fed lambs were dominated by bacteria from the mother's vagina (46%), ambient air (31%), and the sheep pen floor (12%) (Bi et al., 2019) [18]. Facultative anaerobes such as *Streptococcus* and *Enterococcus* are the early colonizers of rumen, which convert rumen to a fully anaerobic environment to promote the rapid establishment of strictly anaerobic bacteria [19]. By two days of age, the rumen microflora reaches 10<sup>9</sup> cells/mL with strictly anaerobic bacteria being predominantly found in the rumen of lambs [20]. The aerobic and facultatively anaerobic bacteria were 10- to 100-fold lower than the strictly anaerobic bacterial count observed during the first week, which continued to decrease afterwards [20].

Compared to older animals, the abundance of phylum Bacteroidetes was significantly lower in one-day-old calves and was mainly composed of the genus *Bacteroides*, whereas older animals were mainly colonized with *Prevotella* [19]. Work by Malmuthuge and coworkers (2014) [21] reported that the rumen contents of three-week-old calves contained a similar level of *Bacteroides* (15.8%) and *Prevotella* (15.1%), which may suggest that starter feed can propel rumen microbiome development to more mature status. The presence of cellulolytic and methanogenic bacteria was observed in lambs at three–four days of age, and the population of these bacteria reached a level similar to that observed in mature sheep within seven days of age [20]. Study by Jami and colleagues (2013) [19] reported that cellulolytic bacteria and other bacterial species important to rumen function can be detected as early as one day after birth. Thus, the establishment of these rumen bacteria occurs long before young ruminants have access to concentrated feed or forage. Dill-McFarland and coworkers (2017) [22] indicated that calves sampled a few days after weaning had a more diverse rumen community compared to calves sampled

during weaning. Several fungal operational taxonomic units (OTUs) observed in weaned calves are also present in adults. As fungi mainly colonize fibrous solids, this may suggest an introduction of forage allows previously low-abundant or transient fungi to persist and multiply.

The rumen bacterial population of two-week-old calves fed milk replacer (MR) was reported to contain 45 bacterial genera belonging to 15 phyla [23]. Similarly, 47 bacterial genera belonging to 13 phyla were observed in the three-week-old calves [21]. Interestingly, the rumen microbiota of the two-week-old calves has more heterogeneous microbiota and harbors more abundant yet transient bacterial species and genera compared to calves at 42 days of age [23]. Another study suggested that the diversity and intra-group similarity of rumen microbiota increases with age, suggesting a transition from a heterogeneous and less distinct community to a more homogeneous and diverse mature bacterial population [19]. This is further supported by a recent study, where gut communities showed higher alpha-diversity but lower beta-diversity with age [22]. Co-habitation facilitates individuals to acquire a shared microbiota [24]. The rumen microbiota was similar in weaned and adult goats that were co-housed pre-weaning [25]. This may also contribute to a convergence toward a similar microbiota in the adult animals.

The composition of the rumen bacterial community varied significantly among individual calves, suggesting a strong host-microbiota specificity in the rumen [19,23]. Similarly, the communities of archaea and fungi in rumen varied considerably among individuals [22,26]. This may suggest that the composition of the rumen microbial community is associated with the physiological condition of the host [19]. Moreover, work by Mayer and coworkers (2012) [27] found that fecal microbial composition was more similar between twin calves than between siblings, implying that host genetics partly define individual gut microbial composition.

Additionally, the bacterial composition was different among the gastrointestinal tract regions and between mucosa- and digesta-associated communities [21]. Colonization of calf rumen starts early in life with a distinct segregation of bacteria between digesta and epithelial surfaces. Similarly, the methanogen community also varies along the gastrointestinal tract [26]. This indicates that previous studies on fecal samples cannot adequately represent the complexity of the gut microbiome. Future studies should focus on both mucosa- and digesta-associated communities in rumen directly.

#### **3. Strategies to Promote Rumen Development**

Strategies to promote morphological structure and metabolic function of rumen in pre-ruminants are an ongoing issue which greatly attracts a lot of attention from the scientific community. Numerous studies and approaches attempt to modulate rumen fermentation and the microbial community in young ruminants to accelerate rumen development. These approaches include alteration of diet composition and physical forms, addition of new types of feed additives, and introduction of variables in the feeding management.

#### *3.1. Diet*

#### 3.1.1. Liquid Feed

Liquid feed may affect plasma concentration of hormones and growth factors, such as insulin and IGF-1, which play important roles in stimulating proliferation of rumen epithelial cells [16,28]. Colostrum contains many biologically active substances, mainly polypeptide growth factors and steroid hormones, including insulin, IGF-1, and transforming growth factor (TGF). Intake of colostrum has been associated with the development, digestion, and absorption ability of the GI tract in the newborn calves [16,28]. Moreover, a whole milk calf diet was shown to have a positive effect on milk yield during the first lactation of the adults compared to calves fed an MR diet. These results highlight the importance of biologically active milk-borne factors [29].

Soybean protein can be used as an alternative to milk protein in formulating MR [30]. Previous studies suggested that MR formulated with soy proteins can negatively affect the development of the small intestine [31,32]. The abomasal pH declines more slowly and pH is higher in calves fed MR containing soy flour compared to calves given whole milk [33–35]. Decreasing the pH of MR emulsion by addition of an acidifier reduces the pH of digesta pH in the rumen, reticulum, and omasum. Specifically, pH reduction of MR emulsion was found to be beneficial for the development of ruminal epithelium [36]. Work by Górka and coworkers (2011a) [37] reported a shorter papillae length of the cranial dorsal sac in calves fed MR compared to calves fed whole milk, and noted positive relationships between reticulorumen weight and small intestine weight, or with brush border enzyme activities. There is a close relationship between the development of the rumen and the small intestine. Importantly, different types of liquid feed affect the development of the small intestine, the intake of solid feed later in life, as well as the growth and metabolic status of calves, thereby indirectly affecting the development of forestomaches [37]. Enhancing the nutrition level of MR in calves induces changes in the expression of genes coding for proteins directly influencing rumen epithelial growth [38]. Moreover, liquid feed may flow into the rumen due to the closure of the esophageal groove. This can occur even in calves that are not clinically defined as rumen drinkers. Specifically, in veal calves that received large amounts of milk, the amount of leakage liquid was approximately 14–35%, which may induce ruminal and metabolic acidosis in a clinical case [39,40].

#### 3.1.2. Starter Feed

Feeding readily fermentable carbohydrates to calves increases VFA production in the rumen, which is necessary to stimulate the development of rumen epithelium [41,42]. Calves fed milk-only diet during the first three weeks present with a different microbial community in their GI tract and feces compared to calves given milk and solid feed [43]. Diets differing in carbohydrate composition lead to differences in rumen fermentation patterns and VFA profiles which may have a variable effect on rumen development [44,45]. For example, high concentrations of ruminal ammonia, acetate, propionate, and butyrate were detected in calves fed corn- and wheat-based diets compared to calves fed barleyand oat-based diets. Moreover, the forestomach weight and papillae growth were greater in calves fed corn- and wheat-based diets [46]. The mucosal thickness was greater in veal calves fed starchand pectin-based diets compared to calves on neutral detergent fiber (NDF)-based diets, however, a higher incidence of poorly developed mucosa was observed in calves fed starch-based diet than in animals fed pectin- and NDF-based diets [40]. It was reported that the stimulatory effects of VFAs are different, with butyrate being most stimulatory followed by propionate and then acetate [4,6]. Butyrate provides energy required for rumen wall thickening, formation of papillae and stimulating capillary development [47]. Butyrate can also increase the blood flow during nutrient absorption and metabolism and can directly affect gene expression in the ruminal epithelium [4].

Rumen development can also be affected by the dietary nutrient level. Interestingly, lambs fed a high protein diet had a higher concentration of ammonia nitrogen (NH3–N) but a lower proportion of total VFA and propionate [48]. Moreover, study by Shen and coworkers (2004) [49] identified that a high energy diet lead to rumen papillae proliferation, which was associated with IGF-1 receptors and increased plasma IGF-1 levels in baby goats. However, excessive consumption of rapidly fermentable starter feed may predispose calves to rumen acidosis. Specifically, it can reduce ruminal pH, decrease rumen motility, and result in keratinization of papillae, causing a decreased in VFA absorption [42,50,51].

#### 3.1.3. Forage

Forage is less energy-intensive than starter feed. The low digestibility of forage in the rumen increases gut fill and decreases voluntary intake of concentrated feed by calves, which results in insufficient levels of VFAs required to stimulate rumen growth [52]. However, forage consumption is associated with positive effects of fiber on rumination and salivation in the GI tract [53,54]. The inclusion of forage in the diet increases rumen pH in both pre-weaning and post-weaning calves [55,56]. Importantly, intake of forage was negatively correlated with the severity of subacute ruminal acidosis (SARA), suggesting that a small quantity of consumed forage (0.080 kg/day) can

alleviate rumen acidosis in calves [57]. The empty rumen weight was greater in calves supplemented with hay compared to calves fed a hay-free diet [54,56]. During weaning transition, feeding dietary forage in calves mitigates ruminal acidosis and induces changes in ruminal bacterial diversity and abundance [58]. Thus, two completely opposite opinions exist as to whether to feed forage to calves before weaning. To address this issue, several studies have been conducted to compare the effect of different initial time of forage provision on growth and rumen development in calves. Lin and colleagues (2017) [59] indicated that supplementation of oat hay to pre-weaned calves increased starter feed intake, ruminal pH, and reduced non-nutritive oral behaviors. Calves with hay supplementation initiated at two weeks of age showed the best productivity. Another study found that feeding forage to calves, either from 3 or 15 days of age, had no effect on growth rate, feed intake and rumen fermentation parameters compared to calves fed no forage, which also justified the supply of forage to young calves [60]. Inclusion of forage in the starter feed was positively linked with muscular development of the rumen [61,62] and morphological appearances of rumen epithelial cells, and caused decreased plaque formation [40,61]. Replacing 50% barley or corn with corn silage in the diet given to 10- or 90-day-old calves improved the thickness of the rumen wall, but had no significant effect on the papillae [63].

Different forage sources have different effects on stimulating chewing activity and saliva production [64]. Supplementation with NDF from alfalfa hay in the starter diet was shown to be more effective than beet pulp in increasing rumen pH and stimulating chewing activity [65]. A recent meta-analysis indicated that forage consumption can affect starter feed intake and performance in calves, which was modulated by forage level, sources, and physical forms of the starter [66].

#### 3.1.4. Physical Form

The physical form and particle size distribution of the diet exert significant influence on the anatomical and microbial development of the rumen [50,54]. For example, calves fed a ground diet had shorter papillae with a smaller surface area compared to calves fed the unground diet. Moreover, a decrease in cellulolytic bacteria and an increase in amylolytic bacteria were detected in calves fed the ground diet [50]. Consumption of finely ground diets can reduce ruminal pH [57] and lead to rumen parakeratosis [50,67]. Given these considerations, 75% of the particles in the starter feed should exceed 1190 μm in diameter [68]. Work by Lesmeister and Heinrichs (2004) [69] reported that calves fed texturized starter feed containing whole corn had higher ruminal pH compared to calves fed diet with dry-rolling corn, roasted-rolling corn, or steam-flaked corn. Increasing particle score of alfalfa hay from 1 mm to 3 mm can affect non-nutritive oral behaviours in calves fed finely ground starter feed [70]. However, research by Suarez-Mena and coworkers (2015, 2016) [8,71] suggested that increasing particle size of the starter diet by adding whole oat or straw of different lengths had no effect on rumen fermentation and calf development. Moreover, chopping of hay grass (~50% particles > 1.9 cm) decreased chewing time of calves [72], meanwhile, the richness and diversity of rectal microflora was reduced [73]. Provision of rations containing finely ground hay (2 mm) to calves may increase feed sorting and result in imbalanced intake of nutrients after weaning [74]. Increasing length of chopped hay from 2 mm to 3–4 cm reduced non-nutritive oral behaviors and improved nutrient digestibility [75]. The effect of the physical form and shape of the diet on calves is closely related to the inclusion rate, source, nutrient matrix and processing method of each ingredient. Importantly, the optimal calf diet specification designed specifically to promote rumen development has not been yet defined.

#### *3.2. Feed Additives*

#### 3.2.1. Probiotics

Probiotics are viable and beneficial microorganisms that help maintain GI microbial balance and promote rumen development. Feeding probiotics to calves around weaning age may facilitate the development of rumen bacterial communities and help calves with a transition from liquid feed to dry feed and forage [76,77]. Fermentation products of *Saccharomyces cerevisiae* have been shown to positively influence ruminal microbiota and improve ruminal morphology [78,79]. Specifically, effects of *Bacillus licheniformis*, *Saccharomyces cerevisiae* and their compounds can increase nitrogen utilization of the rumen microbial community and affect the fermentation pattern which was shown to be beneficial for growth of fattening lamb [80]. An oral dose of *Megasphaera elsdenii* NCIMB 41125 given to calves at 14 days of age increased ruminal butyrate, reticulorumen weight and papillae growth, suggesting an improvement in epithelial metabolism [81]. Supplementation of *Bacillus subtilis natto* in starter feed was shown to aid the development of rumen bacterial communities by increasing the growth of cellulolytic bacteria in calves after weaning [82].

However, feeding probiotics to calves has not always been shown to exert positive effects on the development of cellulolytic bacteria. For example, adding a mixture of *Lactobacillus plantarum* and *Bacillus subtilis* to MR and starter feed affected the denatured gradient gel electrophoresis (DGGE) fingerprint of the 16S ribosomal RNA genes, and reduced the number of *Ruminococcus albus* in calves [83]. In contrast, other studies reported that pH and enzymatic activities of rumen fluid were unaffected by three kinds of probiotic feeding in newborn calves [84]. Supplementation of *Candida tropicalis* in MR had no effect on the morphology of the forestomach and enzymatic activities of ruminal digesta [85]. Rumen and papillae measurements of Holstein bull calves were not affected by inclusion of *Aspergillus oryzae* fermentation extract in MR and starter feed [86]. Overall, the effects of probiotics on rumen development in calves are inconclusive, and frequently driven by differences in viable probiotic bacterial numbers, probiotics species, administration methods, and health status of animals.

#### 3.2.2. Effects of VFAs

VFAs are the primary products of rumen fermentation and contribute to rumen epithelium development in calves. Previous studies suggested that infusion of sodium propionate or sodium butyrate greatly promotes the development of the rumen papillae in calves [6,42,87]. Supplementation of MR with sodium butyrate was associated with increased reticulorumen weight and increased length and width of papillae [37,88,89]. Another study showed that calves receiving a blend of short and medium chain fatty acids as monoglycerides (0.2%) in MR had less degenerative tissue accumulation and a higher number of cytoplasmic protrusions on the exposed horn surfaces [90].

Branched-chain VFAs (BCVFA), such as isobutyrate, isovalerate and 2-methylbutyrate, are naturally derived from the catabolism of branched-chain amino acids. Adequate levels of BCVFA are essential for the growth of some cellulolytic bacteria and digestion of structural carbohydrates in the rumen [91–93]. Supplementation of isobutyrate and isovalerate in milk and concentrate feed can accelerate the growth of calves by improving ruminal fermentation, rumen enzyme activities and growth of cellulolytic bacteria [94,95]. Administration of VFAs have been proved to be effective in promoting rumen development in calves, however, the optimal inclusion rate of VFAs and BCVFA in feed deserved further researches.

#### 3.2.3. Plant Extracts

There are many studies focused on evaluating the potential of plant extracts as alternatives to feed antibiotics and growth promoters in ruminant nutrition. Plant extracts have been shown to favorably affect rumen microbiota [96] and modulate ruminal fermentation in ruminants [97–99]. However, studies evaluating how plant extracts affect rumen development in young ruminants are limited. A recent study revealed that adding Aloe barbadensis to milk was beneficial in increasing total VFA concentration and bacterial count in cross-bred calves [100]. Supplementation of mulberry leaf flavonoids in MR increased α-amylase activity in ruminal digesta and protease activity in abomasal digesta in calves [85]. Supplementation of caraway and garlic in concentrated feed can improve rumen fermentation parameters by increasing total VFAs, increasing rumen pH and decreasing rumen ammonia in growing buffalo calves [101]. Thyme and cinnamon essential oils were shown to decrease

the molar proportion of acetate and lower the ratio of acetate to propionate, as well as increase the level of propionate in Holstein calves consuming a high-concentrate diet. Finally, cinnamon essential oil was shown to increase rumen molar concentration of butyrate [102]. Plant extracts are among the most promising alternatives to antibiotics due to their extensive biological effects, and can be used in calf feed to prevent diarrhea. However, the efficacy of plant extracts is subject to a series of factors, including the composition of active components, addition levels, and physiological status of animals. The use of different types of plant extracts at various inclusion rates in the diet deserves further research. Moreover, effects of plant extracts on the colonization of microbial populations remains to be determined in calves.

#### *3.3. Feeding Management*

Weaning age can influence the development of rumen in pre-ruminants. For example, calves weaned at six weeks of age had longer and wider papillae compared to calves weaned at nine weeks of age [103]. In early-weaned calves, the ruminal pH, molar proportion of acetate and the ratio of acetate to propionate were lower, but the molar proportion of propionate and butyrate were greater [104]. The β-diversity of ruminal microbiota shifts rapidly in calves weaned at six weeks, while a more gradual shift is observed in calves weaned at eight weeks [105]. The colonization pattern substantially differs between newborn goats reared naturally with the dam and those reared artificially with MR. A higher bacterial diversity was observed in natural-fed goats [106]. Compared with suckling feeding, bottle feeding mode tended to increase the number of potential pathogens and delay the establishment of anaerobic microbes in the gut of lambs [18]. The total rumen bacterial population of lambs grazing at pasture with the nursing mother was lower compared to lambs weaned at 21 or 35 days of age, whereas methanogens and protozoa population were lower in early-weaned lambs compared to grazing lambs [107]. Kehoe and coworkers (2007) [108], however, reported that weaning age had no effect on rumen papillae length, width or rumen wall thickness. Different weaning methods (conventional weaning or concentrate-dependent weaning) result in similar rumen development [109]. The development of ruminal microbiome was not affected by the weaning strategy, and there was no effect of gradual or abrupt weaning [110]. The difference may mainly be associated with rumen development status. Due to the differences in feeding and management during the pre-weaning period, rumen development of calves may vary in different experiments. Calves with a well-developed rumen are able to utilize grains and forage efficiently. The effect of weaning age may only be detected in calves with undeveloped rumen. Additionally, pair-housed calves were shown to consume more solid feed at an earlier age compared to calves housed individually [111,112].

Intensive feeding of milk or MR may decrease starter feed intake, thereby delaying rumen development (Cowles, 2006) [113]. Hence, the amount of milk supplied to calves is normally restricted to promote starter feed intake and rumen development in conventional feeding practices [3,114]. However, calves fed limited amounts of milk had lower growth rates and abnormal behavior due to reduced nutrient intakes [55]. Schäff and colleagues (2018) [115] reported that compared to calves fed MR ad libitum, calves fed a restricted amount of MR had greater density of the rumen papillae in the atrium and ventral blind sac, but lower villus surface area and villus height/crypt depth ratio in the distal jejunum. Enhanced MR feeding increased the concentration of plasma IGF-1 and insulin [116,117], which may be beneficial for gastrointestinal growth in pre-weaning calves [16,28]. Furthermore, increasing nutrient intake from milk or MR resulted in enhanced milk yield in the first lactation [118]. Thus, intensive feeding practices have been widely adopted by producers; however, supporting feeding programs, such as a gradual weaning plan, need to be detailed to ensure optimum rumen development.

#### **4. Conclusions**

To summarize, it is beneficial for rumen development for calves to be fed high-quality liquid feed rich in biologically active substances. Minimization of the use of soy protein or appropriate acidification of MR may alleviate gastrointestinal epithelium lesions. Feeding readily fermentable carbohydrates to calves to increase VFA production can stimulate rumen development. A pellet or texturized starter feed is superior to a finely ground meal. Providing calves with high-graded forage, such as alfalfa hay, can reduce the occurrence of rumen acidosis and papillae keratinization. Moreover, additives can be used in calf feed due to their potential advantages in rumen manipulation, however, the types and the optimal inclusion rate deserve further study. More importantly, there is no fixed pattern of calf feed. The diet compositions and nutrient specifications should be matched with the feeding program and management to better promote rumen development.

The rumen is a unique part of the GI tract in ruminants. As the rumen develops and becomes colonized by microorganisms, a calf physiologically transitions from a pseudo-monogastric to a functioning ruminant. The development of rumen in calves can directly affect feed intake, nutrient digestibility and eventual growth of calves. Any changes in the early feeding regime and nutrition can influence rumen development, and thus, lead to long-lasting effects on subsequent growth, health, and milk production performance. Study by Moallem and colleagues (2010) [29] reported higher milk yields during the first lactation in heifers fed whole milk compared to heifers fed MR. Moreover, the same study suggested that MR did not impart any milk-borne effects in calves [29]. Increasing the nutrient intake from milk or MR prior to weaning resulted in an increased milk yield during the first lactation [118–120]. This phenomenon may be associated with epigenetic effects of early nutrition [118].

Additionally, an early feeding regime and nutrition can influence rumen development and rumen microbial composition, ultimately exerting an effect on the lifetime milk yield in cattle. Studies indicated that diets can modify the establishment of the bacterial community in lambs during weaning, which can persist for four to five months [121,122]. The postnatal period is frequently referred to as the most sensitive window for rumen manipulation [123–125], although studies evaluating ruminal imprinting are still limited. The majority of published studies focus on rumen organ development, rumen fermentation parameters, morphology, and changes in the population of cellulolytic bacteria. With the development of microbial 16S rDNA gene sequencing and metagenome analysis, additional studies will likely reveal the interactions between host GI tract development and establishment of rumen bacteria.

**Author Contributions:** Conceptualization, Q.D.; Writing—original draft, Q.D. and R.Z.; Writing—review & editing, Q.D. and T.F.

**Funding:** This research was supported by the Earmarked Fund for Beijing Dairy Industry Innovation Consortium of Agriculture Research System (BAIC06-2019) and the Collaborative Innovation Task of the Agricultural Science and Technology Innovation Program in Chinese Academy of Agricultural Sciences-Integration Innovation of Technology in Dairy Environment-Safe Rearing and Breeding.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **E**ff**ects of Pair Versus Individual Housing on Performance, Health, and Behavior of Dairy Calves**

## **Shuai Liu 1,**†**, Jiaying Ma 1,**†**, Jinghui Li 2, Gibson Maswayi Alugongo 1, Zhaohai Wu 3, Yajing Wang 1, Shengli Li <sup>1</sup> and Zhijun Cao 1,\***


Received: 30 November 2019; Accepted: 22 December 2019; Published: 25 December 2019

**Simple Summary:** In modern dairy farming systems, calves are often housed in individual pens or hutches, which results in less social interaction with their peers during the milk-feeding period. The aim of this study was to evaluate the effects of pair versus individual housing on performance, health, and behavior of dairy calves from the milk-feeding period to the first week after mixing. Results showed that pair versus individual housing had no effects on body weight, starter intake or average daily gain during the milk-feeding period, while pair housing increased the growth performance of calves during weaning and postweaning periods, and the beneficial effects of pair housing on growth faded after calves were mixed and moved to group housing. Paired calves showed higher diarrhea frequency only in week three. The behavior of calves was altered at different periods, including increased time spent in feeding, chewing and ruminating, and decreased self-grooming time, and a drop of non-nutritive manipulation for all calves after they were mixed and moved to group housing. We also found less social contact may lead to more non-nutritive manipulation.

**Abstract:** The aim of this study was to evaluate the effects of pair versus individual housing on performance, health, and behavior of dairy calves. Thirty female Holstein dairy calves were assigned to individual (n = 10) or pair housing (n = 10 pairs). The results showed that both treatments had a similar starter intake and average daily gain (ADG) during the preweaning period. During weaning and postweaning periods, paired calves had a higher starter intake, and the ADG of paired calves continued to increase but calves housed individually experienced a growth check. Paired calves showed higher diarrhea frequency only in week three. The results on behavior showed that feeding, chewing and ruminating time increased, and self-grooming time decreased with age during weaning and postweaning periods, and paired calves spent less time feeding, standing and self-grooming but more time lying during this time. After mixing, feeding, and chewing and ruminating time continued to rise, and self-grooming time continued to decline for both treatments. All calves spent less time standing and non-nutritive manipulation after mixing, and previously individually housed calves tended to increase non-nutritive manipulation. These results showed that pair housing improved growth during weaning and postweaning periods and that calves altered their behavior at different phases. Less social contact may lead to more non-nutritive manipulation.

**Keywords:** calf; pair housing; individual housing; behavior

#### **1. Introduction**

Under natural conditions, calves are nursed by the dam and tend to have social interactions with their peers or other animals [1,2]. In modern dairy farming systems, however, calves are often housed in individual pens or hutches. Hence, they are less likely to interact with their peers or other animals during the milk-feeding period.

Previous work has indicated that different housing systems (group versus individual housing) affect the performance and health of dairy calves. Some studies showed that compared with individual housing, group housing increased weight gains [3], starter intake [4] and hay intake of dairy calves [5]. Conversely, other studies showed no effects [6] or even negative effects on weight gain for group-housed calves [7]. Furthermore, respiratory diseases and diarrhea were reported to occur more frequently in group-housed veal calves [8]. On the contrary, Babu et al. [9] reported that rearing calves in a group resulted in a lower disease incidence. In other cases, health outcomes were similar between different housing systems [10]. The variability among studies may be related to differences in management (e.g., the number of animals per group, milk volume provided, duration of the feeding period, weaning method, and disease diagnosis). From a behavioral standpoint, weaning from a milk-based diet to a solid diet is one typical stressor faced by dairy calves, in which case, calves vocalize more (d 37 to 55) [11,12]. After weaning, calves are mixed with unfamiliar animals and moved to a novel environment, which may cause aggressive interactions (d 91 to 126) [13]. The stress resulted from weaning and mixing can negatively affect animal welfare [14]. Social housing during the milk-feeding period may have beneficial effects on behavior and cognition ability of calves even after they were weaned and mixed with unfamiliar animals in a group. Several studies have shown that social housing improved resilience to stress (d 51 to 53) [15], as well as increased competitive behavior (d 49 to 56) [16] and interactions (d 56 to 91) [17] after weaning. Furthermore, previous research has mainly clarified the effects of social housing on lying and feeding [17–19]. However, other behavioral responses, such as standing, chewing and ruminating, self-grooming, and non-nutritive manipulation have not been well characterized when calves were weaned and moved to group housing. The primary objective of this study was to compare growth, performance and health as well as evaluate the effects of paired versus individual housing on calves' behavior when they were weaned (d 42 to 56) and moved to group housing (d 63 to 70). We hypothesized that paired-housed calves would have better performance than individually housed calves.

When calves were weaned and moved to group housing, they experience changes in the way of feeding and management, especially when they are introduced to a different diet (from a milk-based diet to a solid diet or total mixed ration) and social environment. These changes may impact the behavior of calves. Overvest et al. [18] reported the day to day changes in lying and feeding during the weaning period (d 40 to 48), and Horvath and Miller-Cushon [14] described the day to day changes in standing time of calves mixed in a group (d 60 to 74 ± 5). However, how behavior would change from one period to another was still not clear. Therefore, the secondary objective focused on calf behavioral changes from the weaning period (d 42 to 56) to when they were mixed in a group (d 63 to 70). We hypothesized that calves would exhibit less socially affiliative behavior, such as self-grooming, which may be related to greater activity and exploratory behavior when calves were initially moved to group housing [14]. We also predicted that less non-nutritive manipulation would be observed after calves were moved to group housing, as non-nutritive manipulation often occurred among individually housed calves, especially during the milk-feeding period [20].

#### **2. Materials and Methods**

#### *2.1. Animals and Treatments*

This study was conducted at China Agricultural University's Dairy Education and Research Centre (Datong, Shanxi, China) in 2016, in accordance with protocols approved by the Ethical Committee of the College of Animal Science and Technology, China Agricultural University, Beijing, China (No. 2016DR07). Thirty female Holstein dairy calves were collected from the end of March to the mid-April and were assigned to individual (n = 10) or pair housing (n = 10 pairs) based on birthdate and body weight (mean ± SEM; 43.5 ± 0.59 kg). The age difference between calves in the same pair was within 48 h. Only calves with successful passive transfer of immunity (mean ± SEM; 6.24 ± 0.09 g/dL), determined by clinical refractometer 24 h after birth, were included in the study (serum total protein ≥5.5 g/dL).

#### *2.2. Housing and Management*

#### 2.2.1. Preweaning, Weaning, and Postweaning

All calves were born in a calving pen and separated from their dams within 1 h after birth and weighed. After that, calves were moved to a separate and clean straw-bedded nursery room adjacent to the calving facility. If the younger calf of a pair was born within 24 h of the older calf of the pair, then the two calves were moved directly to the experimental calf barn less than 200 m from the nursery room. If the younger calf of a pair was not born within 24 h of the birth of the older calf, then the older calf was kept in the nursery room until the younger calf of the pair was born, accepting an age difference within a pair of maximum 48 h. Calves were transferred to the calf pens by a cart; paired-housed calves were transferred together within 3 h to 5 h after the birth of the younger calf, whereas individually housed calves were transferred alone.

Individually reared calves were kept in individual pens (1.5 m × 2.0 m), while paired-housed calves were provided twice the area (3.0 m × 2.0 m). Calf pens were located under a 3-sided (solid, 1.1 m in height), roofed shelter with a metal gate at the front. Calves could hear calves and see calves in neighboring pens through the openings in the gate. Openings provided access to buckets (10 L for each one) placed 35 cm apart in the center of the pen for water and starter. Calves housed individually had two buckets (one for water, one for starter), while pair housing calves were provided twice the feeding facilities. All calves had free access to water and pelleted starter feed. All the feeding facilities were cleaned daily. The interior of each pen was bedded with sand and bedding was replaced weekly.

Colostrum was heated to 39 ◦C in a water bath. After that, colostrum was transferred to 4-L esophageal tubing bottles and fed to the calf through a tube within 2 h after the calf was born on d 1. From d 2 to 56, pasteurized waste milk (nonsaleable milk) was provided 3 times daily at 08:00, 15:00 and 20:00 and the volume of milk for each time was equal. During the preweaning period, calves were fed 6 L/d from d 2 to 7, 7.5 L/d from d 8 to 42. Weaning was carried out by reducing milk volume on d 43 and calves were fed 6 L/d from d 43 to 49 and then 3 L/d until d 56. At each milk feeding, the buckets for water were removed temporarily and milk buckets (5 L for each one) were placed in the same position of the pen. For each pair, two milk buckets were used at each milk feeding, whereas calves housed in individual pens had a single milk bucket. Milk buckets were cleaned after each feeding. After weaning, calves remained in their pens during the postweaning period (d 57 to 63 ± 1). No forage was offered before mixing.

#### 2.2.2. Mixing Period

On d 64 ± 1, individual calves were mixed with the paired-housed calves according to the age and moved to the calf barn. There were 5 groups and each group consisted of 6 calves: 2 previously housed in individual pens and 2 pairs previously housed in pairs. The age difference between calves in the same group was within 48 h. The back wall of the group pen (5.0 m × 4.0 m) was solid with two sides made from horizontal tubular metal bars (bar diameter: 5.0 cm; distance between bars: 12 cm) and a neck rail at the front. The length of the neck rail allowed all 6 calves to eat simultaneously (83 cm per calf). Total mixed ration (TMR) was delivered twice daily at 10:00 and16:00. Each group pen was equipped with one automatic water trough (length: 120 cm, width: 40 cm, height: 70 cm, depth: 20 cm) and calves had free access to water. Sand was used as bedding material and was replenished when the group was moved on d 70. Three Digital Thermometers (Deli Electronic Commerce Co., Ltd., Ningbo, China) were spaced evenly and mounted above (1.0 m) the sand bedding in the calf barn to record daily temperature and humidity (maximum and minimum).

The temperature and humidity fluctuated according to weather conditions (mean ± SD; 14.7 ± 9.3 ◦C and 23.2 ± 10.1% relative humidity).

#### *2.3. Sample Collection*

#### 2.3.1. Feed Sampling

Feed samples were collected weekly and immediately frozen at −20 ◦C until they were further analyzed. The nutritional composition (Table 1) of the dry matter, crude protein, neutral detergent fiber, acid detergent fiber, ether extract, crude ash, calcium and phosphorus were analyzed following the methods of AOAC International [21]. Throughout the study, starter intake was recorded daily based on the amount offered and refused by each calf from d 5 to 63 ± 1. TMR intake during the mixing period was not measured because of group housing.

**Table 1.** Nutrient compositions of milk, starter, and total mixed ration.


<sup>1</sup> The nutritive values are the means of the results of the analysis of samples collected each week. <sup>2</sup> Contained steam-flaked corn (33.5%), alfalfa hay (21.2%), oat hay (21.2%), soybean meal (19.7%), and premix compound (0.4%) on a DM basis.

#### 2.3.2. Body Measurements and Blood Sampling

Body weights (BW) were measured weekly (d 1, 7, 14, 21, 28, 35, 42, 49, 56, 63, 70). Body length (shoulders to pins), withers height, hip height, and heart girth were also recorded at the same time points. Blood samples were collected via jugular venipuncture using vacutainer serum collection tubes containing no anticoagulant 24 h after birth. The blood samples were then centrifuged at 3500× *g*, 4 ◦C for 15 min. Serum total protein (TP) was determined by an optical refractometer (Honneur Nutritional Technology Co. Ltd., Beijing, China).

#### *2.4. Health Check and Treatment*

The health check consisted of three parts: (1) fecal scoring, (2) clinical examination of the respiratory system, (3) rectal temperature. Fecal scores were recorded daily at 10:00 each day until d 63 based on a 1 to 4 system according to the guidelines outlined by Larson et al. [22]. Scores were, 1 = firm, well-formed (not hard); 2 = soft, pudding-like; 3 = runny, pancake batter; and 4 = liquid, splatters, pulpy orange juice. Fecal score data were collected by one independent trained observer. All fecal scores were recorded by observing fecal matter on the ground of the pen or the tail and hindquarters of the calf. Fecal scoring was not conducted during the mixing period, because the fecal scores of an individual could not be accurately identified due to group housing. A diarrheic day was defined when the fecal score was >2. Weekly diarrhea frequency was calculated with the following equation: Diarrhea frequency = [(number of diarrhea calves × days of diarrhea) / (total number of calves × days of trial)] × 100%. Any calf with a fecal score >2 was treated according to the protocols established by the farm veterinarian (e.g., by administering antibiotic drugs and electrolytic solutions). Respiratory health was checked before each morning feeding through visually inspecting nasal discharge and

listening to breathing difficulties with auscultation by the farm veterinarian and a member of the research team before morning milk feeding. If calves had signs of respiratory disease such as nasal discharge, cough and breathing difficulties, and a rectal temperature ≥39.5 ◦C, they were treated using an *Andrographis paniculata* injection (10 mL; Dazheng Tec-Phar. Co. Ltd., Changchun, China) for a maximum of 48 d; if respiratory disease or pyrexia was not alleviated, the calf received antibiotics treatment for a maximum of 48 d. Electrolytes were also administered intravenously to calves that had a severe respiratory disease until fully recovered. Throughout the study, one calf from individual housing was treated for 3 d during the mixing period because of nasal discharge and breathing difficulties, and no other calves had respiratory disease.

#### *2.5. Behavioral Observations*

A digital color camera (DS-7800, HIKVISION, Hangzhou, China) was placed above each selected pen (placed 2.5 m in front of the pens and 3.5 m from the pen floor), monitoring the behavior of the calves. During nighttime hours (from 17:30 to 07:30), the infrared monitoring function of the camera would turn on automatically. The recorded behaviors (Table 2) included feeding, chewing and ruminating, lying, standing, self-grooming, non-nutritive manipulation, and social contact.



<sup>1</sup> If one calf exhibit multiple behavior at one time point, then the multiple behavior were all recorded. Social contact was recorded for calves during the mixing period.

Six individually housed calves and 6 pairs of paired-housed calves were selected randomly for behavioral observations from d 43 to 70. Based on previous results [18,23,24], the sample sizes of behavior variables were estimated to obtain a power of 0.8 under a significance level of 0.05. During weaning and post-weaning periods, the behavioral data were recorded for 48 h on d 43, 50, and 57. For the mixing period, the behavior data were recorded for 48 h on the second day of mixing (d 65 ± 1) to avoid the effects of transition stress on calves. In order to clearly identify each selected calf from the groups during the mixing period, all selected calves were photographed from the front, back, left, right, and above. The observer could record behavior based on each calf's unique photos. For every 24 h duration (144 h in total for each calf), instantaneous scan-sampling with 5 min intervals was used to collect the lying, feeding, standing and chewing and ruminating data and continuous recording was used to collect the self-grooming, non-nutritive manipulation, and social contact data [25]. All behavioral data were recorded by one observer.

#### *2.6. Statistical Analysis*

#### 2.6.1. Starter Intake, Growth, and Health Data

Throughout the study, data were analyzed at the pen level (based on a single calf per pen for the individual treatment and the mean of the 2 calves per pen in the pair treatment). Starter intake data were averaged by the week, except for the first week data, which were averaged across the last three days (d 5 to 7). Continuous variables with repeated measurements, including starter intake, average daily gain (ADG), BW, and structural growth, were tested for normality using the UNIVARIATE procedure of SAS (version 9.2, SAS Institute Inc., Cary, NC, USA). These data were then analyzed from week 1 to week 10 (as a whole) using the MIXED procedure of SAS. The model included the fixed effects of time, treatment, and time × treatment interaction and the random effect of pen To account for the repeated measures within-subject, the covariance structures were chosen for each repeated variable on the basis of best fit which was determined from the Bayesian information criterion. The heterogeneous first-order autoregressive structure was selected for starter intake, BW, and structural growth data, and for ADG data, the first-order autoregressive structure was selected. Data for fecal scores were summarized by the week and analyzed using the Chi-squared test.

#### 2.6.2. Behavioral Data

Behavioral data obtained for individual calves from video were also averaged by pen (a single calf per pen for the individual treatment and the mean of the 2 calves per pen in the pair treatment) across the 48 h in each observation week (week 7 and week 8 during weaning, and week 9 during postweaning and 10 during mixing). For each 48 h behavioral observation period, the average duration of each kind of behavior per 24 h was calculated. Behavioral data were analyzed separately by two stages: (1) weaning and postweaning, and (2) mixing. The comparison of social contact between two treatments was only analyzed during the mixing period, as individually housed calves had no social interaction before mixing. For stage 1, the effect of housing on behavior was tested using the MIXED procedure of SAS. The model included the fixed effects of treatment, week, and week × treatment interaction, and the random effect of pen. To account for the repeated measures within-subject, the first-order autoregressive structure was chosen for each behavior on the basis of best fit, which was determined from the Bayesian information criterion. For stage 2, the effect of housing on behavior was tested using one-way ANOVA. Lying, standing, non-nutritive manipulation and social contact data were normally distributed. Behavioral data of feeding, and chewing and ruminating were analyzed after logarithm transformation, and self-grooming data were analyzed after square root transformation to meet the normality assumption. The transformed data were back-transformed to report.

All data were reported as least squares mean. Differences of *p* < 0.05 were considered significant and 0.05 ≤ *p* < 0.10 was considered a tendency.

#### **3. Results**

#### *3.1. Starter Intake and Growth*

As shown in Figure 1, starter intake showed an upward trend over time (*p* < 0.001) for both individually and pair housed calves with no difference in starter intake between treatments during the preweaning period (*p* > 0.05). During weaning and postweaning periods, starter intake tended to be higher for paired-housed calves during week seven (860.0 vs. 658.1 ± 80.1 g/d, *p* = 0.09), and than for individually housed calves during week eight (1461.4 vs. 1123.1 ± 97.0 g/d, *p* = 0.02) and week nine (2237.4 vs. 1899.5 ± 113.5 g/d, *p* = 0.04).

ADG increased over time (*p* < 0.001) for both treatments and no differences were found between treatments during the preweaning period (*p* > 0.05, Figure 2). During weaning and postweaning periods, the weight gain of paired-housed calves continued to increase, but individually housed calves experienced a growth check. The ADG for paired-housed calves tended to be higher during week seven (0.94 vs. 0.71 ± 0.07 kg/d, *p* = 0.08). Individually housed calves had higher ADG than calves housed in pairs during the mixing period (1.20 vs. 0.85 ± 0.09 kg/d, *p* = 0.01).

Throughout the study, the housing system (paired vs. individual) had no effects on BW (*p* = 0.50) and structural measurements (Table 3), including withers height (*p* = 0.55), heart girth (*p* = 0.38), abdominal girth (*p* = 0.14), and body length (*p* = 0.23).

**Figure 1.** Starter intake (LSM ± SEM) for calves housed individually (n = 10 calves) or in pairs (n = 10 pairs) before mixing. PH = calves housed in pairs; IH = calves housed individually; wk = week. TMR intake was not measured because of group housing during the mixing period (week 10). \* *p* < 0.05, † *p* < 0.10.

**Figure 2.** Average daily gain (LSM ± SEM) for calves housed individually (n = 10 calves) or in pairs (n = 10 pairs). PH = calves housed in pairs; IH = calves housed individually; wk = week. *p*-value: 0.90 (treatment), < 0.001 (week), 0.08 (treatment × week). \* *p* < 0.05, † *p* < 0.10.

**Table 3.** Least squares mean of structural measurements and BW for calves housed individually (n = 10 calves) or in pairs (n = 10 pairs) from week 1 to week 10.


<sup>1</sup> PH = calves housed in pairs; IH = calves housed individually. <sup>2</sup> Body length was measured from shoulders to pins.

#### *3.2. Health*

Throughout the study, one calf from individual housing during the mixing period suffered from respiratory disease, and no other calves had respiratory disease. Diarrhea frequency is shown in Figure 3. Pair housing increased diarrhea frequency in comparison with individual housing of calves during week three (18.0% vs. 6.0%, *p* = 0.03), yet no differences were found between treatments in other weeks (*p* > 0.05).

**Figure 3.** Effects of the housing system on diarrhea frequency before mixing for calves housed individually (n = 10 calves) or in pairs (n = 10 pairs). PH = calves housed in pairs; IH = calves housed individually; wk = week. Fecal scoring was not conducted during the mixing period (week 10), because the fecal scores of an individual could not be accurately identified due to group housing. \* *p* < 0.05.

#### *3.3. Behavior*

As shown in Figure 4, during weaning (week 7–8) and postweaning (week 9) periods, feeding time increased (*p* < 0.001) for both treatments. Overall, individually housed calves spent more time feeding (83.0 vs. 53.1 ± 1.15 min/d, *p* = 0.04) compared with paired-housed calves during this period. After mixing, feeding time decreased for individually-housed calves but increased for paired-housed calves, and the previous housing system had no effects on feeding time after mixing (*p* = 0.82). Ruminating time increased over weaning and postweaning periods (*p* < 0.001) for both treatments and individually-housed calves tended to have greater ruminating time than paired-housed calves during week seven (2.56 vs. 1.79 ± 0.26 h/d, *p* = 0.09). After mixing, ruminating time continued to increase with age for all calves with no differences found between treatments (*p* = 0.61).

Lying time increased during the weaning period and decreased during the postweaning period for all calves. Standing time increased during the postweaning period for all calves. Calves housed in pairs spent more time lying (17.3 vs. 16.4 ± 0.27 h/d, *p* = 0.03) and less time standing (6.33 vs. 7.11 ± 0.18 h/d, *p* = 0.01) compared with calves housed individually during weaning and postweaning periods. After mixing, lying time remained stable and standing time decreased for all calves, and the previous housing system had no effect on lying (*p* = 0.56) and standing time (*p* = 0.84). There was a decrease in self-grooming time for both treatments over weaning and postweaning periods (*p* = 0.01), and calves housed individually exhibited more self-grooming than calves housed in pairs (40.7 vs. 20.6 ± 4.10 min/d, *p* = 0.02). After mixing, self-grooming time continued to decrease for all calves, with no differences between treatments (*p* = 0.65). In addition, non-nutritive manipulation time did not change with calf age during weaning and postweaning periods (*p* = 0.62), and treatment had no effects

on non-nutritive manipulation time during this period (*p* = 0.10). After mixing, all calves decreased non-nutritive manipulation time, and the non-nutritive manipulation time tended to be longer for calves that were previously individually-housed (33.0 vs. 16.9 ± 5.62 min/d, *p* = 0.07). During the mixing period, the previous housing system had no effect on social contact (Figure 5; *p* = 0.53).

**Figure 4.** Effects of housing system on (**a**) feeding; (**b**) chewing and ruminating; (**c**) lying; (**d**) standing; (**e**) self-grooming; and (**f**) non-nutritive manipulation for calves housed individually (n = 6) or in pairs (n = 6). Stage 1 = from weaning to postweaning period, including weeks 7, 8, and 9. Stage 2 = mixing period, including week 10. PH = calves housed in pairs; IH = calves housed individually; wk = week. \* *p* < 0.05, † *p* < 0.10.

**Figure 5.** Effects of the housing system on social contact for calves housed individually (n = 6) or in pairs (n = 6) during the mixing period. PH = calves housed in pairs; IH = calves housed individually.

#### **4. Discussion**

#### *4.1. Starter Intake and Growth*

During the preweaning period, we did not observe any differences in starter intake, ADG or BW between treatments, yet during weaning and postweaning periods, pair housing improved growth performance. Our results were consistent with previous research that reported increased starter intake [6,26] and ADG [23] during the weaning period for paired-housed calves. Such improvements are likely due to social facilitation [12] and social learning [27], which allow calves housed in pairs to learn faster and eat more. Local enhancement is another factor affecting the feeding of calves, through which the behavior of one calf draws the attention of another in the same pair toward a particular food source [28,29]. In addition, paired calves might experience a lower level of stress during the weaning period because of social buffering [12]. The social buffering benefits of early pair housing have been discussed recently by Overvest et al. [18], who demonstrated that social housing might improve the ability to cope with the weaning stress via the positive effects on feed acceptance and behavioral flexibility. During the postweaning period, we observed greater starter intake in pair housing. Similar results were reported by Pempek et al. [20], who also attributed it to social facilitation. Besides, the competitive feeding environment among paired calves may also have resulted in more starter intake during the weaning and postweaning periods, as calves may increase the rate of feed intake in the competitive feeding environment [17]. Our results contribute to a body of evidence indicating that pair housing is particularly beneficial to solid-feed intake, growth, and supporting a smooth transition at weaning [3].

After mixing, calves were grouped together. Warnick et al. [30] and De Paula Vieira et al. [12] reported that calves previously housed in groups or pairs gained more than those previously housed in individual pens when they were mixed and placed together. Some studies attributed these results to the beneficial effects of social housing, such as reduced neophobia to new ration [31,32] and greater competitive success [16] when mixed with unfamiliar animals. On the contrary, we found that previously paired-housed calves had less ADG after mixing compared with calves housed individually, and the final BW was similar between treatments. Somewhat interestingly, Miller-Cushon and DeVries [4] reported that though paired calves had greater performance during the weaning period (d 39 to 49), previous housing (paired vs. individual) had no effect on DMI, ADG or final BW once previously individually housed calves were paired with unfamiliar calves after weaning (d 50 to 84). Similar results were reported by Overvest et al. [18], who also demonstrated that once calves previously housed individually were paired after weaning (d 49 to 56), they exhibited more feeding time and thus increased their solid feed DMI to a greater extent over time than paired-housed calves, and eventually resulted in similar DMI between treatments. These results suggested that previously individually housed calves could get the same performance (e.g., DMI, BW, and ADG) through modifying feeding behavior after they were exposed to social housing with unfamiliar calves. In this study, all calves experienced a sudden feed transition to TMR, in which case, the beneficial effects of social housing on food neophobia may be weakened by transition stress. Furthermore, we observed similar feeding time between treatments during the mixing period. Thus, we speculated that the higher ADG in calves previously housed individually may result from higher feeding rate, allowing them to consume sufficient TMR to meet or exceed their nutritional requirements and finally compensating for a previously lower starter intake. Further work to address this possibility is encouraged.

#### *4.2. Health*

In the current experiment, diarrhea frequency for calves housed in pairs was higher than that for calves housed individually in week three, yet no differences were found in other weeks. Some studies reported that housing calves in groups exhibited more health problems owing to higher levels of infectious agents and calf-calf transmission [33,34]. On the contrary, others reported a lower incidence of diarrhea for calves housed socially [9], and some found no differences in incidences of diarrhea and respiratory problems [35] between paired-housed calves and individually housed calves. The various results indicated that health problems were not consistently associated with social housing. The incidence of disease relies on many factors including calf immunity, environment management, disease diagnosis, and the ability of a calf to cope with stress [36]. These factors rather than the housing system may play a critical role in inducing health problems. Greater health problems in a group housing system may also stem from the difficulty of detecting disease in groups [23]. There is not enough evidence to support a diarrhea-increasing effect of pair housing, thus the higher incidence of diarrhea in pair housing in week three was probably due to low immunity to infection of calves aged from two to four weeks [37] and individual differences.

#### *4.3. Behavior*

Limited research has described how behavior would change at different periods from weaning to mixing, or the effects of paired or individual housing on behavior during these periods. Our results suggested that all calves experienced behavioral changes from weaning to mixing including increased feeding and ruminating time, and decreased self-grooming time.

The increase in feeding time and chewing and ruminating time over the weaning and postweaning periods aligned with the increase in feed intake. Besides, paired calves spent less time feeding but still had higher starter intake during weaning and postweaning periods, likely due to the competitive feeding environment as we discussed on starter intake and growth. Miller-Cushon et al. [17] found that calves housed in a competitive feeding environment had less time of feeding but an increased rate of feed intake compared with those housed in a noncompetitive feeding environment. Hence, paired-housed calves might increase their feeding rate rather than feeding time to consume more starter.

In the present study, lying time declined while standing time rose during the postweaning period, which could be attributed to the increase in feeding time with increasing age during this period. The previous study [18] suggested that calves may change their lying behavior to accustom themselves to feeding behavior. In addition, calves housed individually exhibited more lying time than paired calves during weaning and postweaning periods, which is contrary to previous studies [23,38]. Previous research [20] also mentioned no effects of individual vs. paired housing on lying. The variant space allowance for calves among studies may be responsible for the discrepancy in results, as space allowance was a vital factor for the expression of normal behavior [23,39]. Further research is encouraged to study the relationship between space allowance and lying. After mixing, standing time decreased for all calves. Previous studies reported that calves were more active and moved more followed by a reduction in activity after the first day of introduction to a group [14,29], and calves had diminished

behavioral reactions after the first 24-h period following regrouping [40]. Thus, we speculated that calves might not be as active as the first day of introduction to a group as the behavior data were recorded for 48 h on the second day of mixing to avoid the effects of transition stress on calves in this study.

Self-grooming is expressed by calves as caring for their own body, and this behavior may be a means of satisfying socialization [27]. More self-grooming activities were observed in individually-housed calves in the present study, which was consistent with previous research [27], as the socialization was absent in these calves. In addition, self-grooming can also be an expression of stress. Taking rodent as a research model, previous studies [41–43] has reported that the relationship between stress and self-grooming can be described as an inverted U-shaped: Self-grooming typically occurs spontaneously at low stress and becomes longer during moderate stress and can be inhibited by high-stress states. Thus, the higher self-grooming of individually housed calves may respond to the higher stress (moderate stress) they faced compared with paired calves during weaning and postweaning periods.

Non-nutritive manipulation commonly occurs within artificial rearing systems [44], which can be strengthened by social deprivation [45]. In the current study, a drop in non-nutritive manipulation time for all calves after mixing was observed, which was likely due to more social interactions among calves after mixing. Bokkers and Koene [46] also indicated that less social interaction was an important factor causing dairy calves to lick objects (a nonnutritive manipulation behavior). In this study, no effects were found on non-nutritive manipulation during the weaning or postweaning period, whereas individually-housed calves tended to spend more time on non-nutritive manipulation compared with paired calves during the mixing period, which was similar to the previous study [47]. Although the effect of the previous housing system on social contact was not significant during the mixing period, calves housed in pairs previously still exhibited more social contact numerically, which may result in less non-nutritive manipulation.

#### **5. Conclusions**

Paired versus individual housing had no effects on body weight, starter intake or ADG during the preweaning period, while pair housing increased the growth performance of calves during weaning and postweaning periods, and the beneficial effects of pair housing on growth was weakened after mixing. Paired calves showed higher diarrhea frequency only in week three. Calves altered their behaviors at different periods from weaning to mixing, including increased feeding time and chewing and ruminating time, and decreased self-grooming time, and a drop of non-nutritive manipulation for all calves after mixing. Furthermore, less social contact may result in more non-nutritive manipulation.

**Author Contributions:** Designed the experiments: Z.C., S.L. (Shuai Liu), and J.M. Conducted the experimental work: S.L. (Shuai Liu), J.M., J.L., G.M.A. and Z.W. Conducted the data analysis: S.L. (Shuai Liu) and J.M. Wrote the paper: S.L. (Shuai Liu) and J.M. Revised the paper: S.L. (Shuai Liu), J.M., J.L., G.M.A., Z.W., Y.W., S.L. (Shengli Li), Z.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by National Key Research and Development Program of China (2018YFD0501600).

**Acknowledgments:** We are grateful to the staff of China Agricultural University's Dairy Education and Research Centre (Datong, Shanxi, P.R. China) for their assistance with the trial.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

#### *Communication*

## **E**ff**ects of Dietary Rumen-Protected Betaine Supplementation on Performance of Postpartum Dairy Cows and Immunity of Newborn Calves**

**Beibei Wang 1, Chong Wang 2, Ruowei Guan 1, Kai Shi 1, Zihai Wei 1, Jianxin Liu <sup>1</sup> and Hongyun Liu 1,\***


Received: 8 March 2019; Accepted: 12 April 2019; Published: 15 April 2019

**Simple Summary:** Betaine plays an important role in growth, lactation, protein synthesis, and fat metabolism in animals, but there are few studies on transition dairy cows and newborn calves. The aim of the current study was to evaluate the effects of rumen-protected betaine supplementation from four weeks before expected calving to six weeks postpartum regarding the lactation performance and blood metabolites of dairy cows and immunity of newborn calves. The results suggested that betaine supplementation tended to increase fat mobilization of postpartum dairy cows. Furthermore, compared to the control calves, the betaine calves had greater plasma total protein and globulin concentrations, which indicates that the immunity of the betaine calves might have improved.

**Abstract:** The objective of this study was to evaluate the effects of rumen-protected betaine supplementation on performance of postpartum dairy cows and immunity of newborn calves. Twenty-four multiparous Holstein dairy cows were randomly divided into the control (CON, *n* = 12) and rumen-protected betaine (BET, *n* = 12) groups after blocking by parity and milk yield during the previous lactation cycle. The cows were fed a basal total mixed ration diet without BET (CON) or with BET at 20 g/d per cow (BET) from four weeks before expected calving to six weeks postpartum. The results showed that betaine supplementation had no effect on dry matter intake and milk yield of the cows. The BET cows tended to increase feed efficiency (energy-corrected milk/dry matter intake) and body weight loss postpartum compared to the CON cows. The plasma β-hydroxybutyrate concentrations of the BET cows were greater at d seven after calving than those of the CON cows. Moreover, compared to the CON calves, the BET calves had greater plasma total protein and globulin concentrations. The plasma glucose concentrations of the BET calves tended to decrease relative to CON cows. In conclusion, rumen-protected betaine supplementation from four weeks before expected calving tended to increase fat mobilization of postpartum dairy cows, and might improve the immunity of newborn calves.

**Keywords:** betaine; dairy cows; newborn calves; fat mobilization; immunity

#### **1. Introduction**

During the transition period, dairy cows are in a state of great metabolic stress because of the increased demand for nutrients to maintain fetal growth and milk synthesis. Transition dairy cows tend to have negative energy and amino acid balance after calving, which leads to an increase in fat and protein mobilization in tissues [1,2]. A negative methyl donor balance also likely occurs in transition cows because milk is high in methylated compounds [3]. Moreover, the last two months of gestation, where 60% of the body weight gain before birth occurs [4], is critical for bovine fetal development.

Betaine functions as a methyl donor and an organic osmolyte [2,5], which plays an important role in growth, lactation, protein synthesis, and fat metabolism in animals [6]. Betaine supplementation in the diets of steers increased body weight gain and fat deposition [7]. In lactating dairy cows, feeding betaine improved the milk yield and milk protein [8,9]. Supplementing betaine reduced plasma concentrations of non-esterified fatty acids (NEFA) and β-hydroxybutyrate (BHB) of lactating dairy cows [10], but elevated the concentrations of NEFA and BHB of transition dairy cows to change lipid metabolism [11]. Furthermore, betaine is vital for fetal development [12], and is related to the offspring's weight and immunity [13]. However, due to the fast rumen degradation (approximately 45%/h) of betaine in vivo [14], unprotected betaine cannot be absorbed efficiently. Our previous study showed that dietary rumen-protected betaine supplementation in lactating dairy cows improved lactation performance and fat metabolism [15]. Whether it improves the performance of postpartum dairy cows and the immunity of newborn calves remains unexplored. Therefore, the objectives of the current study were to evaluate the effects of rumen-protected betaine supplementation from four weeks before expected calving to six weeks postpartum on the lactation performance and blood metabolites of dairy cows and immunity of newborn calves.

#### **2. Materials and Methods**

#### *2.1. Animals and Treatments*

All the experimental protocols used in this study were approved by the Animal Care Committee of Zhejiang University (Hangzhou, China) (Approval Number: ZJU20160379). Twenty-four multiparous prepartum Holstein dairy cows were selected and divided randomly into the control (CON, *n* = 12) and rumen-protected betaine (BET, *n* = 12) groups after blocking by parity (2.27, SD = 1.4) and milk yield during the previous lactation cycle (24.9 kg/d, SD = 6.0). The cows were fed a basal total mixed ration (TMR) diet (Table 1) without BET (CON) or with BET at 20 g/d per cow (BET), according to Zhang et al. (2014) [9] from four weeks before expected calving to six weeks postpartum. The basal diets were formulated based on the NRC (2001) [16]. The cows were fed three times daily at approximately 06:30, 13:30, and 19:30 h, and BET (BET with 30% purity, Hangzhou King Technology Feed Co., Ltd, Hangzhou, China) was supplemented twice daily in the morning and evening by top-dressing the TMR during feeding. All cows were housed in tie-stall barns and given access to fresh water ad libitum. After parturition, the cows were milked three times daily at approximately 07:00, 14:00, and 20:00 h. Fourteen calves randomly selected (CON calves: *n* = 7, BET calves: *n* = 7) were studied from birth to 24 h. All calves were weighed with a digital scale immediately after birth and were fed fresh colostrum from their dams within 2 h of birth. The calves were individually housed in hutches, and water was offered ad libitum.


**Table 1.** Ingredient and chemical composition of the diets fed during the prepartum and postpartum periods.


**Table 1.** *Cont.*

<sup>1</sup> DM = dry matter.<sup>2</sup> Formulated to contain (per kilogram of premix) 220 to 400 KIU of vitamin A, 50 to 100 KIU of vitamin D3, ≥2250 IU of vitamin E, ≥40 mg of D-Biotin, ≥380 mg of niacinamide, ≥40 mg of Beta-carotene, 0.2 to 0.7 g of Cu, 1.0 to 3.8 g of Zn, 0.8 to 3.0 g of Mn, 12.5 to 100 mg of I, 8.0 to 25 mg of Se, 2.5 to 50 of mg Co, 10.0% to 30.0% of Ca, 10.0% to 30.0% of NaCl, and ≥1.5% of total phosphorus.

#### *2.2. Sample Preparation*

The amounts of feed offered and refused were recorded according to Gu et al. (2018) [17] to determine dry matter intake (DMI). The TMR samples were collected weekly for dry matter (DM, 105 ◦C for 5 h), crude ash, ether extract (EE), crude protein (CP), and acid detergent fiber (ADF), according to AOAC procedures (method 942.05, 920.39, 988.05, and 973.18, respectively), and neutral detergent fiber (NDF) with sodium sulfite and amylase was analyzed [18]. Body weight (BW) was measured on d 0 and 42 after calving according to Wang et al. (2017) [19].

Milk yield was recorded for two consecutive days each week and milk samples from three consecutive milking were taken each week in the amounts proportional to the yield (4:3:3, composite from each daily milking). The samples were stored at 4 ◦C with bronopol tablets (D & F Control System Inc., San Ramon CA, USA) for later determination of protein, fat, lactose, total solids, and milk urea nitrogen (MUN) using a Combi Foss FT+ instrument (Foss Electric, Hillerød, Denmark). The 3.5% FCM (Fat-corrected milk) and ECM (Energy-corrected milk) were calculated by the formula [20]: 3.5% FCM = (milk yield, kg/d × 0.4324) + (milk fat, kg/d × 16.216), ECM = (milk yield, kg/d × 0.327) + (milk fat, kg/d × 12.95) + (milk protein, kg/d × 7.20).

Blood samples from the cows were collected from the coccygeal vein in sodium-heparinized tubes at approximately 4 h after the morning feeding on −21, −10, 0, 7, 14, 28, and 42 d relative to calving. Blood samples from the calves were collected via the jugular vein using a sodium-heparinized tube shortly after birth before colostrum feeding and at approximately 24 h after birth. The samples were centrifuged for 10 min at 3000 g at 4 ◦C to harvest plasma, which was stored at -20 ◦C until analysis. The plasma samples were analyzed using an Auto Analyzer 7020 instrument (Hitachi High-Technologies Corp., Tokyo, Japan) with colorimetric commercial kits (Ningbo Medical System Biotechnology Co., Ltd., Ningbo, China) for total protein (TP), albumin (ALB), alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), total bilirubin (TBIL), triglyceride (TG), cholesterol (CHOL), glucose (GLU), superoxide dismutase (SOD), NEFA, and BHB. The concentrations of globulin (GLOB) were calculated by the formula [21]: GLOB (g/L) = TP (g/L) − ALB (g/L).

#### *2.3. Data Analysis*

A randomized block design with repeated measures was used. The DMI, lactation performance, feed efficiency, and blood metabolites of the cows were analyzed with PROC MIXED of SAS 9.2 (SAS Institute Inc., Cary, NC, USA). Treatment, week, treatment × week, and block were included as the fixed effects in the model, and cow within treatment was used as a random effect. The blood metabolites of the calves were analyzed using the same procedure in SAS 9.2, except sampling hour instead of week was used as the repeated measure. The BW change of the cows, colostrum composition, and calves birth weight were analyzed using PROC MIXED of SAS 9.2 without the repeated statement. All associated interactions were removed from the model. The results are presented as least squares means. Statistical significance was determined at *p* ≤ 0.05 and tendencies at 0.05 < *p* ≤ 0.10.

#### **3. Results and Discussion**

Betaine supplementation had no effect on DMI, milk yield, and composition (*p* > 0.1, Table 2). Monteiro et al. (2017) [11] found that cows supplemented with betaine-containing molasses from 60 d before expected calving had higher milk yield, whereas no differences were observed in milk yield of cows supplemented with betaine-containing molasses from 24 d before expected calving, which is consistent with our results. The addition of betaine during the transition period increased the milk yield in a time-dependent manner, which might be related to the functions of betaine as an organic osmolyte to maintain the cell function by stabilizing cellular proteins and promoting proper protein folding [5,22]. The dry period is critical for the renewal and growth of mammary cells [23]. Hence, betaine addition during the far-off period has a positive effect on prepartum mammary growth, which increases the subsequent milk yield. The Lys: Met ratio in the postpartum diets was estimated to be 3.13:1 in our study, which had met the ideal Lys: Met ratio of 3.0:1 for an optimal milk protein content and yield [16,24]. Methyl donors (choline) additional supplementation had no detectable effect on cow performance when the Lys: Met ratio in diets had reached 3.0:1 [25]. This might also be a reason for BET additional supplementation, which has no effect on milk yield and composition.

Compared to CON cows, BET cows tended to increase fat-corrected milk/dry matter intake (FCM/DMI, *p* = 0.09), energy-corrected milk/dry matter intake (ECM/DMI, *p* = 0.08), and BW loss postpartum (*p* = 0.10) (Table 2). The plasma BHB concentrations of the BET cows were greater at d seven after calving than those of the CON cows (treatment × time: *p* = 0.07, Table 3). The greater number of animals in the study might have increased the statistical significance. The BET cows tended to have greater feed efficiency and BW loss postpartum in our study, coupled with greater concentrations of BHB at d seven after calving, which indicates that the BET cows might have an enhanced fat mobilization in early lactation due to higher milk yield numerically (milk yield was approximately 2.53 kg/d higher in BET cows than in CON cows) [11,26].




**Table 2.** *Cont.*

<sup>1</sup> 3.5% FCM (Fat-corrected milk) = (milk yield, kg/d × 0.4324) + (milk fat, kg/d × 16.216) [20]. <sup>2</sup> ECM = (milk yield, kg/d × 0.327) + (milk fat, kg/d × 12.95) + (milk protein, kg/d × 7.20) [20].

**Table 3.** Effects of supplementing cows without rumen-protected betaine (CON) or with rumen-protected betaine (BET) on blood metabolites from four weeks before expected calving to six weeks postpartum.


<sup>1</sup> TP = total protein. ALB = albumin. GLOB = globulin. A/G = albumin/globulin. ALT = alanine aminotransferase. AST = aspartate aminotransferase. ALP = alkaline phosphatase. TBIL = total bilirubin. TG = triglyceride. CHOL = cholesterol. GLU = glucose. NEFA = non-esterified fatty acids. BHB = β-hydroxybutyrate.

The plasma TP and GLOB concentrations of the BET calves were greater than those of the CON calves (*p* = 0.04, *p* = 0.05, respectively, Table 4), although no differences in calves birth weight were found between treatments (37.80 ± 1.68 kg vs. 36.03 ± 1.68 kg). The plasma TP and GLOB concentrations of calves increased significantly with maternal betaine supplementation, which indicates that it might improve the immunity of newborn calves because of maternal methyl donors supplementation [13,27]. Maternal dietary supplementation with methyl donors could program the health of offspring through the epigenetic regulation of the DNA molecule and cell signaling [27,28], which might improve the capacity for GLOB absorption of the intestine to improve the immunity of newborn calves. Furthermore, the lactose content of the BET colostrum tended to increase compared to the CON colostrum in our study (3.50% vs. 3.07%, *p* = 0.07), which might also have contributed to the results. Lactose plays a key role in the energy supply, absorption of minerals, and gastrointestinal functions of calves [29,30].

The plasma GLU concentrations of the BET calves tended to decrease compared with those of the CON calves (*p* = 0.09, Table 4). The plasma SOD concentrations of the BET calves were greater at 2 h after birth than those of the CON cows (treatment × time: *p* = 0.01, Table 4). A positive correlation between neonatal glucose and cortisol concentrations proved that the lower concentrations of glucose in the BET calves were likely to have less stress around calving [31]. In turn, the degree of stress might influence newborn calves' energetic mobilization [31,32]. The greater plasma SOD concentrations 2 h after birth in our study suggested that the BET calves were in a state of less stress [33]. The plasma glucose concentrations in newborn calves might be related to the uteroplacental transport of glucose via mTOR signaling [34] and hepatic gluconeogenic gene expression via epigenetic mechanisms [35], which deserve further study.


**Table 4.** The blood metabolites during the 24 h after birth of calves born to dams supplemented without rumen-protected betaine (CON) or with rumen-protected betaine (BET) during the peripartal period.

<sup>1</sup> TP = total protein. ALB = albumin. GLOB = globulin. A/G = albumin/globulin. ALT = alanine aminotransferase. AST = aspartate aminotransferase. ALP = alkaline phosphatase. TBIL = total bilirubin. GLU = glucose. SOD = superoxide dismutase.

#### **4. Conclusions**

Dietary rumen-protected betaine supplementation from four weeks before expected calving had no detectable effect on dry matter intake and milk yield, but tended to increase fat mobilization of postpartum dairy cows. Furthermore, the BET calves had greater plasma total protein and globulin concentrations, which indicates that the immunity of the BET calves might improve.

**Author Contributions:** Conceived and designed the experiments: B.W., C.W., J.L. and H.L. Conducted the experimental work: B.W., R.G. and K.S. Conducted the data analysis: B.W. and Z.W. Wrote the paper: B.W. Revised the paper: B.W., C.W. and H.L.

**Funding:** Grants from the National Key Research and Development Program of China (2018YFD0501600), the China Agriculture Research System (No. CARS-37), and the Zhejiang Provincial Natural Science Foundation (LY18C170002) supported this research.

**Acknowledgments:** The authors thank Hangjiang Dairy Farm (Hangzhou, China) for providing and caring for the cows.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## **The E**ff**ect of Intravenous Infusions of Glutamine on Duodenal Cell Autophagy and Apoptosis in Early-Weaned Calves**

**Xusheng Dong 1,**†**, Ruina Zhai 2,**†**, Zhaolin Liu 1, Xueyan Lin 1, Zhonghua Wang <sup>1</sup> and Zhiyong Hu 1,\***


Received: 14 March 2019; Accepted: 24 June 2019; Published: 1 July 2019

**Simple Summary:** The objective of this study was to determine the effects of intravenous infusions of L-glutamine (Gln) on the autophagy and apoptosis of duodenum cells in weaned calves. The results showed that the autophagy level of duodenal cells was increased with an increasing Gln infusion dose (0 to 20 g/d) and dropped when Gln was further increased to 40 g/d. We also found that the level of apoptosis was decreased with an increasing Gln infusion dose from 0 to 20 g/d, and then rose as the dose increased to 40 g/d. This knowledge will provide a reference for weaned calf health management.

**Abstract:** The objectives of this study were to determine the effects of intravenous infusions of L-glutamine (Gln) on the autophagy and apoptosis of duodenum cells in early-weaned calves. Holstein male calves were weaned at day 35 (20 male calves, birth weight 43 ± 1.8 kg; 35 ± 3 d of age) and randomly allocated to four treatments (5 calves/treatment). The treatments were: (1) infusion of NaCl, representing the control group (C); (2) infusion of 10 g/d of Gln solution (L); (3) infusion of 20 g/d of Gln solution (M); and (4) infusion of 40 g/d of Gln solution (H). The solutions were infused for 2 h daily for 3 consecutive days after weaning. All calves were killed on the third day post-weaning. The results showed that the autophagy level of the duodenal cells was increased as the Gln infusions increased from 0 to 20 g/d and dropped with a further increase in dose (40 g/d). We also found that the level of apoptosis was decreased with Gln infusion from 0 to 20 g/d and rose as the dose increased to 40 g/d. This knowledge provides a reference for weaned calf health management.

**Keywords:** calf; glutamine; autophagy; apoptosis

#### **1. Introduction**

Glutamine (Gln) is the most abundant amino acid in vivo and is a major respiratory fuel and metabolic precursor for many cell types [1]. Glutamine, which once was regarded as a nonessential amino acid, has recently been termed conditionally essential during injury or oxidative stress [2,3]. A previous study suggested that Gln could protect the small intestine from various harmful injuries in rats [4]. Kallweit et al. [5] showed that Gln protects intestinal cells from both heat and oxidant stress. Recently, researchers attempted to evaluate the impact of Gln on autophagy and apoptosis [6–8]. Sakiyama et al. [7] suggested that Gln could protect intestinal epithelial cells by enhancing autophagy.

Autophagy is a specific protein degradation process that functions in the bulk degradation of cellular components and has been recognized as an important mechanism for cell survival under conditions of stress [7,9]. When cells lack nutrients, autophagy is activated to supply amino acids in order to maintain cell survival [10]. In vivo apoptosis and autophagy are two forms of physiological and conserved programmed cell death [11]. Apoptosis is characterized by a series of morphological changes, including plasma membrane blebbing, nuclear condensation, and fragmentation, which lead to the formation of apoptotic bodies [12]. When cells are under stress, autophagy and apoptosis are activated [12]. In general, autophagy is activated first and maintains cell homeostasis [13]. When stress is prolonged or exceeds a threshold, apoptosis is activated [12,14]. The protein microtubule-associated protein 1 light chain 3-II(LC3-II), which is a useful marker of autophagic membranes, is essential for the expansion of the early autophagosome in the context of cellular house-keeping and autophagic cell death [15,16]. Caspase-3 is an executioner caspase, which is activated by apoptosis [12]. Furthermore, PI3K/Akt/mTOR signaling pathways, which inhibit autophagy, have been found to be essential for the regulation of autophagy [17]. A previous study found that caffeine could induce autophagy by abolishing AKT phosphorylation [17]. The kinase mammalian target of rapamycin (mTOR) is a downstream target of the PI3K/AKT pathway [17]. Deactivation of mTOR signaling induces autophagy [18]. Amino acids, which are provided by autophagy, can restore mTOR complex 1 (mTORC1) activity during amino acid starvation [8]. The restoration of mTORC1 in turn inhibits autophagy, which completes the feedback loop [8]. The feedback loop could protect the cells by mitigating damage from stress and starvation.

Weaning is the transition from the ingestion of milk to solid feed for calves with dramatic gastrointestinal transformations [19]. Weaning is a particularly vulnerable period for mammals, with an increased risk of malnutrition, intestinal infections, and poor growth [20,21]. The function and morphology of the small intestine are severely disturbed after weaning, such as villous shortening in pigs [22,23]. Our previous study found that an exogenous supply of Gln increased the autophagy level of liver cells and increased growth rates, villus height, and crypt depth of the duodenum in early-weaned calves [20,24]. Whether Gln induces autophagy and apoptosis of the duodenum cells in early-weaned calves remains unknown. The purpose of this study was to evaluate the effect of Gln on the autophagy and apoptosis of duodenum cells in early-weaned calves. We hypothesized that intravenous infusions of Gln would increase the level of autophagy and reduce the level of apoptosis. This would provide a reference for weaned calf health management.

#### **2. Materials and Methods**

Animal care and use were approved and conducted under established standards of the Ethics Committee on animals of Shandong Agricultural University (SDAUA-2018-012). The study was conducted during November of 2018 at the Shandong high-speed modern dairy farm in Ji Nan, Shandong, China. The animals were individually housed in a pen with free access to water and fresh calf starter. The ingredient and nutrient composition of the calf starter is given in Table 1.

All calves received 4 L of colostrum in the 2-hour period after birth and were then fed 6 L of whole milk 3 times daily until weaning. Fresh calf starter was offered ad libitum beginning at 3 d of age. Water was offered daily ad libitum. Holstein calves were weaned at day 35 (20 male calves, birth weight 43 ± 1.8 kg; 35 ± 3 d of age) and randomly allocated to four treatments (5 calves/treatment). Starting from day 35, the calves were given the following treatments for 3 consecutive days. The treatments were: (1) infusion of 1.5 L of 0.85% NaCl, representing the control group (C); (2) infusion of 10 g/d of Gln mixed with 1.5 L of 0.85% NaCl solution (L); (3) infusion of 20 g/d of Gln mixed with 1.5 L of 0.85% NaCl solution (M); and (4) infusion of 40 g/d of Gln mixed with 1.5 L of 0.85% NaCl solution (H). The dose of intravenous infusion Gln referred to that in a previous study [20]. At the beginning of the experiment, all calves had milk removed from their daily diet. The solutions were infused for 2 h daily for 3 consecutive days after weaning. Starter intake for each calf was measured daily during the infusion period.


**Table 1.** Ingredient and nutrient composition of the experimental starter of calves.

DM: dry matter; CP: crude protein; NDF: neutral detergent fiber; ADF: acid detergent fiber; ME: metabolizable energy. <sup>1</sup> Premix contained (mg/kg): vitamin A, 4035; vitamin D, 1740; vitamin E, 39; Fe, 18; Zn, 37; Cu, 10.6; Mn, 15.3; Co, 0.12; I, 0.47; and Se, 0.35.

All calves were euthanized following captive bolt gun stunning on the 3 d post-weaning day for measuring the autophagy and apoptosis of duodenum cells. After opening the body cavity, the samples of duodenum (entire wall from 6 cm distal to the pylorus) were immediately frozen in liquid nitrogen and stored at −80 ◦C until western blotting was performed.

Briefly, the tissue sample blocks (entire duodenum from 6 cm distal to the pylorus) were washed with phosphate buffer saline (PBS, Solarbio, P1020-500 mL, Beijing, China), cut into small pieces, homogenized in PBS at 4 ◦C using a Servicebio KZ-II homogenizer, kept on ice for 0.5 h, oscillated to ensure complete tissue cracking every 5 min, and then centrifuged (3000× g, 10 min, 4 ◦C). Protein concentration was determined in the supernatant (BCA Protein Assay Kit, G2026, Servicebio, Wuhan, China). The sample was then diluted with an equal volume of Laemmli sample buffer (Bio-Rad, 1610737, Shanghai, China) and boiled for 5 min. Sodium dodecyl sulfate-PAGE, electro-transfer of proteins, and immunoblotting were performed as previously described [25,26]. Antibodies used for immunoblotting were anti-LC3 (Sigma-Aldrich, L8918, Shanghai, China), anti-Caspase-3 (Sigma-Aldrich, C8487, Shanghai, China), anti-mTOR (Sigma-Aldrich, SAB2701843, Shanghai, China), anti-phospho-mTOR (Sigma-Aldrich, SAB4301526, Shanghai, China), anti-β-actin (Sigma-Aldrich, A2066, Shanghai, China), and appropriate secondary antibodies (Servicebio, GB23303, Wuhan, China). The chemiluminescence of bands of interest were detected with a digital G: Box imager (Syngene, Frederick, MD, USA). The band density was quantified with ImageJ software (National Institutes of Health, Bethesda, MD, USA).

The data were analyzed as a completely randomized design using one-way ANOVA of SAS 8.2 (SAS Institute Inc., Cary, NC). The individual calf was considered as the experimental unit. The analysis used the following model: y*ij* = μ + α*i* + ε*ij* (y = western blot data, μ = mean, i = dose of infusions, and ε = residuals). The means were compared using Duncan's multiple range test. Significance was declared at *p* < 0.05.

#### **3. Results**

The starter intake of group C, L, M, and H were 1.12 kg/d, 1.15 kg/d, 1.22 kg/d, and 1.19 kg/d, respectively. Starter intake was not different between treatments. The results reported in this research showed that the autophagy level of the duodenal cells was increased with an increasing Gln infusion dose (0 to 20 g/d) and dropped when Gln was further increased to 40 g/d (Figure 1). We also found that the level of apoptosis was decreased with an increasing Gln infusion dose from 0 to 20 g/d, and then rose with an increasing dose of Gln to 40 g/d (Figure 1). In group M, the level of autophagy reached the highest level; in contrast, the level of apoptosis reached a lowest point. The expression of mTOR was significantly decreased after Gln infusion (*p* < 0.05). The expression of p-mTOR in group M was lower than that in other groups (*p* < 0.05).

**Figure 1.** Effects of glutamine (Gln) infusions on the microtubule-associated protein 1 light chain 3-II (LC3-II), mTOR, p-mTOR, and caspase3 expression of duodenum in weaned calves. Treatment was as follows: (1) C: infusion of 1.5 L of 0.85% NaCl; (2) L: infusion of 10 g/d of Gln mixed with 1.5 L of 0.85% NaCl; (3) M: infusion of 20 g/d of Gln mixed with 1.5 L of 0.85% NaCl; (4) H: infusion of 40 g/d of Gln mixed with 1.5 L of 0.85% NaCl. Insets depict representative blots. Values represent means ± SD. Response from statistical result, *p* < 0.05. β-Actin was used to normalize the expression of target proteins. The letters below the bar graph indicate different treatments. Different letters above the bar indicate differences between different groups (*p* < 0.05).

#### **4. Discussion**

In this study, we demonstrated that low dose infusion of Gln could induce autophagy and retard apoptosis. We further increased the infusion of Gln to 40 g/d and found that the effect of Gln on calves was reduced. We concluded that Gln-induced autophagy is mainly dependent on the inhibition of mTOR phosphorylation. Gln is involved in stress protection by way of the stimulation of autophagy in intestinal cells [7,27]. A previous study found that Gln infusion increased growth rates, villus height, and crypt depth in the duodenum of early-weaned calves [24]. In this study, when the concentration of

Gln was increased from 0 to 20 g/d, the autophagy levels increased as the Gln infusion dose increased. The results suggested that Gln can promote autophagy in the duodenum. This finding is consistent with that of Sakiyama et al. [7], who confirmed that Gln is essential for maintaining autophagy and mounting an autophagic response under stress in intestinal cells. It has been suggested that Gln can induce autophagy in intestinal epithelial cells through restraining mTOR and p38 MAP kinase pathways [7]. The expression of mTOR and p-mTOR in our study is also consistent with this study. The expression of p-mTOR was significantly decreased after Gln infusion. To further investigate the effects of high dose Gln, we further increased the infusion of Gln to 40 g/d. Interestingly, we found that the autophagy level was decreased as the Gln infusion dose increased from 20 to 40 g/d. A possible explanation for this might be that the activity of Gln synthetase in the body gradually decreases when the blood concentration of Gln was excessive. In *Escherichia coli*, Gln synthetase activity is subject to inhibition by different end products of Gln metabolism [28]. A previous report suggested that over expression of Gln synthetase inhibited mTOR activity and activated autophagy [29]. Thus, the level of autophagy induced by Gln synthetase was decreased as the activity of Gln synthetase decreased.

Normally, autophagy restrains the activity of apoptosis, and apoptosis-associated caspase activation shuts off the autophagic process [12]. In our study, we also found that the level of apoptosis fell to a low point with the intravenous infusion dose of 20 g/d. The tendency of apoptosis was opposite to that of autophagy, which suggested that autophagy may inhibit the activity of apoptosis. Kallweit et al. [5] found that Gln protects intestinal cells from both heat and oxidant injury, which are key mechanisms in the prevention of apoptosis. Han et al. [30] demonstrated that the combination of Gln has the ability to maintain the integrity of the intestinal mucosal barrier by inhibiting the apoptosis of intestinal epithelial cells. These results are similar to our study. Our result suggests that the effect of Gln on apoptosis is contrary to that on autophagy. AKT is a kinase with dual autophagy–apoptosis regulatory potential, which can phosphorylate Beclin 1 and B-cell lymphoma-2 antagonists of cell death (BAD) to inhibit autophagic and apoptotic functions, respectively [12]. The activation of AKT could inhibit autophagy by inducing mTOR [12]. Thus, considerably more work needs to be done to determine the effect of Gln on Akt/mTOR signaling pathways in early-weaned calf.

In commercial dairy farms, dairy calves are weaned early to reduce milk costs. However, the gastrointestinal tract of the calf is not ready for early weaning [19]. In our research, intravenous infusions of low dose Gln could increases autophagy, which probably relieved weaning stress. A further study could assess the effects of diet supplement Gln on the early-weaned calf. This knowledge will provide a reference for Gln supplementation for weaned calf health management.

#### **5. Conclusions**

In conclusion, Gln could induce autophagy and decrease the level of apoptosis in the duodenum of early-weaned calves. The intravenous infusion moderate dose (20 g/d) of Gln is most effective. This knowledge will provide a reference for weaned calf health management.

**Author Contributions:** Z.H. and Z.W. conceived and designed the experiments; X.D. and R.Z. performed the experiments; X.D. and X.L. analyzed the data; X.D. and Z.L. wrote the paper.

**Funding:** This research was funded by the National Natural Science Foundation of China, grant number 31772624, and the National Key Research and Development Program of China, grant number 2018YFD0501600.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*
