Influences of starter NDF level on growth performance and rumen development in lambs fed isocaloric and isonitrogenous diets

Abstract

Neutral detergent fiber (NDF), as the main component of structural carbohydrate in forage, is a vital factor impacting the performance of young ruminants. The objective of this study was to evaluate the effect of starter NDF level on the rumen fermentation and morphometrics of lambs fed isocaloric and isonitrogenous diets. One hundred Hu sheep lambs (18 days of age and 6.1 ± 0.1 kg body weight [BW]) were randomly assigned to one of the four starter diets differing in their NDF levels on a dry matter (DM) basis: 1) diet with 14% NDF (NDF14), 2) diet with 18% NDF (NDF18), 3) diet with 22% NDF (NDF22), and 4) diet with 26% NDF (NDF26). Lambs were fed milk replacer with designed starters from 21 to 60 d of age and then transitioned to the same starter feed until 90 d of age. Six lambs from each treatment (close to the average BW of the corresponding treatment) were selected and slaughtered at 90 d of age for measuring rumen fermentation, anatomical development, and morphometric characteristics. Lambs receiving NDF22 and NDF26 starter diets had greater (P < 0.05) intake of starter and total DM, and average daily gain during the overall period, thereby heavier (P < 0.05) final weights at 90 d of age compared with those fed NDF14 starter diet. At 90 d of age, lambs fed NDF22 and NDF26 starters had higher rumen pH, followed by lower propionate, higher acetate concentrations (P < 0.05), and a higher acetate-to-propionate ratio compared with those fed NDF14 and NDF18 starters (P < 0.05). No significant differences of the full and empty weight of reticulorumen were observed among treatments; however, the reticulorumen weight expressed as a percentage of the complex stomach weight was lower (P < 0.05) in lambs fed NDF22 and NDF26 starters compared with those fed NDF14 starter. Furthermore, NDF14 lambs had the thickest keratin layer and epithelium compared with those in the other treatments but showed the thinnest muscle layer (P < 0.05). These results suggest that increasing starter NDF levels can ameliorate the rumen fermentation environment and alleviate hyperkeratosis and plaque formation in the rumen epithelium. In summary, including 22% and 26% NDF in the pelleted starter can improve the performance of lambs.

Introduction

Neonatal ruminants have an undeveloped and nonfunctional rumen and initially depend on milk to meet their nutrient demands for maintenance and growth (Khan et al., 2016). The development of the rumen is affected by many factors (Baldwin et al., 2004), the most important factor being the nature of solid feed (Yang et al., 2018; Zou et al., 2018). The provision of forage to young ruminants in early life can provide benefits to rumination and ruminal pH and can work to enhance the rumen development by improving muscular development and preventing papillae clumping (Yang et al., 2015; Nemati et al., 2016), thereby positively affect the growth performance (Imani et al., 2017). As the main constituent of structural carbohydrate in forage, neutral detergent fiber (NDF) is highly important in terms of growth performance in adult ruminants (Uwituze et al., 2010; Quinn et al., 2011). Studies have shown that growth performance and rumen development can be positively regulated by adjusting dietary NDF in the early life stages of calves and lambs, including adjustments in levels and sources of NDF (Castells et al., 2012; Nemati et al., 2016). Gallo et al. (2019) found that high-concentrate diets with 15% NDF and different levels of energy resulted in a better weight gain in finishing lambs. Although some studies have focused on the effects of dietary NDF levels on rumen development and growth performance of young ruminants, no trials have directly provided persuasive evidence to determine the optimal NDF levels in isocaloric and isonitrogenous diets and their impacts on rumen fermentation and tissue morphology, especially for early-weaned lambs. Therefore, the objective of this study was to evaluate the effects of different starter NDF levels on performance, rumen fermentation parameters, anatomical development, and morphometric characteristics of the rumen in lambs during the transition from a liquid to solid feed.

Materials and Methods

The experiment was conducted at the Hailun sheep farm in Jiangsu Province, China, from November 2016 to January 2017, according to the guidelines of the Animal Ethics Committee of the Chinese Academy of Agriculture Sciences (AEC-CAAS-20160920). Humane animal care and handling procedures were followed throughout the experiment.

Lambs, management, and treatments

One hundred healthy Hu sheep male lambs at 18 ± 2 d of age and average body weight (BW) of 6.1 ± 0.1 kg were purchased from the same farm. All lambs were weaned from natural sucking of the ewe on to milk replacer (MR) with pelleted starter during 18 to 20 d of age. According to BW and age, lambs were randomly assigned to one of four treatments differing in their starter NDF level: 1) starter with 14% NDF (NDF14), 2) starter with 18% NDF (NDF18), 3) starter with 22% NDF (NDF12), and 4) starter with 26% NDF (NDF26) on a dry matter (DM) basis. There were five pens per treatment and five lambs per pen (4 × 4 m). The pens were equipped with stainless steel nipple drinker, rectangular trough (3.0 × 0.3 × 0.2 m), and bamboo floor. The pens were disinfected (2% glutaraldehyde and 5% povidone-iodine) every half month to keep a dry and clean environment. MR was fed twice per day, and the pelleted starter was provided ad libitum. The starter NDF contents were adjusted by varying the alfalfa inclusion levels. Lambs were fed MR with starter from 21 to 60 d of age and then weaned to the same starter feed until 90 d of age. The experiment lasted 73 d with a 3-d transition period.

The MR, containing 25.6% crude protein (CP) and 12.6% fat, was purchased from the Beijing Precision Animal Nutrition Research Center (Beijing, China). MR was fed twice a day at 0800 and 1800 hours from 21 to 60 d of age, and the feeding level of the MR solids was adjusted in accordance with 1% BW of lambs. Before feeding via nipple bottles, the MR was dissolved in hot water to obtain a 40 °C solution (16.67% DM). The pelleted starters (Diameter, 4 mm; Length, 10 mm) were formed by the addition of a liquid media onto the powdered ingredients and then stored after the drying phase of a wet granulation and formulated to be isoenergetic and isonitrogenic. The ingredients and chemical composition of starters are shown in Table 1. The analyzed starter NDF values are close (<5% variance) to the test design. Daily starter allotments to each pen were adjusted to allow minimal residuals (<5% of the total offered) remaining in the trough just prior to the morning feeding. All lambs were freely allowed to access to water.

Table 1.

Ingredients and chemical composition of experimental starter feed (%, DM basis)

	Treatments
Items1	NDF14	NDF18	NDF22	NDF26
Ingredients
Corn	63.00	51.80	42.00	36.80
Soybean meal	23.50	17.50	8.00	0.00
Wheat bran	8.70	12.00	12.00	2.00
Fat power	0.00	1.40	2.20	3.00
DDGS	0.00	6.00	12.00	18.00
Extruded soybean	0.00	0.00	4.70	8.90
Alfalfa hay	0.00	6.80	15.00	27.50
Limestone	1.80	1.70	1.45	1.00
Dibasic calcium phosphate	1.40	1.20	1.05	1.20
Salt	0.60	0.60	0.60	0.60
Premix2	1.00	1.00	1.00	1.00
Nutrient composition
ME (MJ/kg)	11.33	11.33	11.33	11.33
DM	86.86	87.05	85.17	85.61
CP	19.63	19.46	19.94	18.83
EE	2.81	5.66	7.37	8.07
Ash	4.73	4.92	4.84	5.51
NDF	14.64	18.40	22.41	24.97
ADF	4.01	6.05	8.43	10.86
NFC	58.19	51.56	45.44	42.62
Ca	1.67	1.67	1.67	1.53
Total Phosphorus	0.74	0.77	0.76	0.75

¹Abbreviations: ADF, acid detergent fiber; Ca, Calcium; DDGS, distiller’s dried grains with solubles; EE, ether extract; ME, metabolic energy; NFC, non-fibrous carbohydrates, NFC = 100-NDF-CP-EE-Ash.

²Contained followings per kilogram of starter: vitamin A, 12,000 IU; vitamin D₃, 3,000 IU; vitamin E,30 IU; Cu, 12 mg; Fe, 64 mg; Mn, 56 mg; Zn, 60 mg; I, 1.2 mg; Se, 0.4 mg; cholinechloride, 0.4 mg.

Data collection and feed samples analysis

The offered amounts and residual of MR and starter were recorded daily. Lambs were weighed at 20, 60, and 90 d of age. The average daily intake of starter and total dry matter intake (DMI), average daily gain (ADG), and feed efficiency (kg BW gain / kg DMI) were calculated accordingly.

Starter samples were taken weekly and dried at 60 °C for 48 h, and ground to pass a 1-mm screen (HK-08A ground mill). The chemical composition of starters, including DM, CP, ether extract, ash, calcium, total phosphorus, was analyzed according to the official methods of analysis (Association of Official Analytical Chemists, Washington, DC), whereas the NDF and acid detergent fiber were analyzed based on the method of Van Soest et al. (1991) and modified by Hintz et al. (1996) to include sodium sulfite during boiling.

Slaughter and rumen samples analysis

At 90 d of age, six Hu lambs from each treatment, weights close to the average weight of the treatment, were selected and slaughtered after fasting for 16 h. After slaughter, the weights of reticulorumen, omasum, and abomasum were measured with and without digesta, respectively. The capacity of rumen was measured by using the water-filling method (Abd-El-Khalek, 1986).

Immediately after slaughter, rumen was dissected, and its content was obtained for measuring pH using a portable pH meter (Orion 230 Aplus, Thermo Scientific, Beverly, MA, USA). Then the rumen contents were squeezed through four layers of cheesecloth to collect the rumen fluid (10 mL in tubes) and were stored at −20 °C until volatile fatty acids (VFA) analysis by gas chromatography (Li et al., 2009).

After removing the rumen digesta, one tissue sample approximately 2 cm² was taken from the ventral sac of the rumen for histological examination. The samples were then flushed with normal saline and fixed in 1.33 mol/L buffered formaldehyde until analysis. Histological measurements were carried out following the recommendations of Beiranvand et al. (2014a). Briefly, rumen tissue samples were dehydrated, embedded in paraffin, sectioned (5 μM), and stained with hematoxylin and eosin. Three pieces of the disconnected section of each sample were observed, and the papillae length, papillae width, keratin layer thickness, epithelium thickness, muscle layer thickness, and rumen wall thickness were measured using the Image-Pro express image analysis and processing system. Morphometric analyses were performed at a magnification of 4 × 10 times (Olympus BX-51; Olympus Corporation, Tokyo, Japan) using Image-Pro Plus 6.0 (Media Cybernetics, Silver Spring, MD, USA).

Statistical analysis

The growth performance data were analyzed based on a repeated measure mixed model of SPSS (SPSS, version 19.0; SPSS Inc., Chicago, IL, USA) with pens as the experimental unit as follows:

$${\displaystyle \begin{array}{l}{\displaystyle Y_{ijk}=\mu +M_i+T_j+M{T}_{ij}+c_k+{\epsilon }_{ijk}}\end{array}}$$

Where Y_ijk is the dependent variable, μ is the overall mean, M_i is the fixed effect of NDF level, T_j is the fixed effect of sampling time, MT_ij is the interaction effect, c_kis the random effect of the pen, and ε_ijk is the random residual error.

Other data were subjected to the one-way ANOVA procedure of SPSS with pen as the experimental unit. Duncan’s multiple range test was performed to detect the differences between treatments. Polynomial contrasts were conducted to determine the linear and quadratic effects of increasing NDF levels. Differences were declared significant when P < 0.05, and the tendency was declared with P-values between 0.05 and 0.1.

Results

Growth performance

No treatment × time interactions were observed for the intake of starter, NDF, total DM, ADG, and feed efficiency (P > 0.05), except the BW that had a treatment × time interaction (P < 0.05; Table 2). However, the total DMI, BW, and ADG of lambs were all significantly affected by diet (P < 0.05) and time (P < 0.001) factors. Lambs fed NDF22 and NDF26 starters have a greater (P < 0.05) starter intake and total DMI compared with those fed NDF14 starters during 21 to 60, 61 to 90, and 21 to 90 d of age. As a result of the greater starter intake and the increase of starter NDF content, NDF intake was higher (P < 0.05) in lambs fed NDF18, NDF22, and NDF26 starters than those fed NDF14 starters, with the highest NDF intake in the NDF26 treatment.

Table 2.

Effects of starter NDF levels on DMI and growth performance of lambs

	Treatments					P-value2
Items1	NDF14	NDF18	NDF22	NDF26	SEM	Trt	Time	Trt × Time
Total DMI, g/d
Days 21 to 90	417c	497bc	569ab	607a	21.31	<0.001	<0.001	0.064
Days 21 to 60	301b	367ab	426a	424a	12.07	0.003
Days 61 to 90	555c	671bc	759ab	856a	20.38	0.004
DMI of MR, g/d
Days 21 to 60	68	71	75	75
DMI of starter, g/d
Days 21 to 90	378c	457bc	526ab	564a	21.00	0.002	<0.001	0.489
Intake of NDF, g/d
Days 21 to 90	61d	92c	128b	152a	8.41	0.000	<0.001	0.614
BW, kg
Day 20	6.02	6.05	6.09	6.23	0.03	0.324	<0.001	0.008
Day 60	10.85b	12.26ab	13.04a	13.60a	0.27	0.014
Day 90	18.27c	21.15b	22.48ab	23.95a	0.43	0.041
ADG, kg/d
Days 21 to 90	176c	210b	235ab	247a	6.04	0.006	<0.001	0.246
Days 21 to 60	124b	151ab	172a	179a	21.70	0.048
Days 61 to 90	247c	290b	317ab	337a	15.77	0.019
Feed efficiency3
Days 21 to 90	0.42	0.42	0.41	0.41	0.01	0.682	0.700	0.409
Days 21 to 60	0.42	0.42	0.42	0.38	0.05	0.357
Days 61 to 90	0.37	0.45	0.42	0.42	0.04	0.225

¹Days mean days of the age of lambs.

²Trt, treatment effect; Time, time effect; Trt × Time, interaction effect of treatment and time.

³Feed efficiency was calculated using the following formula: Feed efficiency = weight gain (kg) / DMI (kg).

^a–dDifferences between treatments (P < 0.05).

At the initiation of the experiment (day 20), there was no difference (P > 0.05) of lambs’ BW between treatments. However, as a result of differences in DMI, lambs fed NDF22 and NDF26 starters were heavier (P < 0.05) compared with those fed NDF14 starter at 60 d of age. Thereafter, the difference still exists during the post-weaning stage. The lambs fed NDF18, NDF22, and NDF26 starters had a higher (P < 0.05) final BW than those fed NDF14 starter, and the highest BW was observed in NDF26 treatment.

Likewise, the ADG of lambs fed NDF22 and NDF26 starters was greater (P < 0.05) than those in NDF14 from 21 to 60 d of age. Additionally, lambs fed NDF18, NDF22 and NDF26 starters had greater (P < 0.05) ADG in comparison with lambs fed NDF14 starter during 61 to 90 and 21 to 90 d of age. However, the feed efficiency of lambs was not affected (P > 0.05) by increasing starter NDF levels during 21 to 60, 61 to 90, and 21 to 90 d of age.

Rumen fermentation parameters

With increasing NDF levels in the starter, rumen pH values, acetate-to-propionate ratio, and molar proportion of acetate increased linearly (P < 0.05; Table 3), while the total VFA (TVFA) concentration and molar proportion of propionate decreased linearly (P < 0.05). The rumen pH values were higher (P = 0.003) in lambs fed NDF22 and NDF26 starters than those fed NDF14 and NDF18 starters. Additionally, lambs fed NDF22 and NDF26 starters presented a greater (P = 0.012) molar proportion of acetate compared with those fed NDF14 starters. In contrast, the molar proportion of propionate was lower (P = 0.007) in lambs fed NDF22 and NDF26 starters than those fed NDF14 starters. The acetate-to-propionate ratio was greater (P < 0.05) for lambs fed NDF22 and NDF26 starters compared with those fed NDF14 and NDF18 starters.

Table 3.

Effects of starter NDF levels on pH and VFA concentrations in rumen fluid of lambs

	Treatments					P-value2
Items1	NDF14	NDF18	NDF22	NDF26	SEM	Trt	L	Q
pH	6.25b	6.10b	6.64a	6.62a	0.08	0.003	0.003	0.394
TVFA, mmol/L	43.69	45.54	25.39	33.65	2.49	0.109	0.013	0.118
Acetate, mol/mol	48.66c	48.22bc	55.19ab	57.99a	1.36	0.012	0.002	0.474
Propionate, mol/mol	46.54a	46.6a	37.87b	36.59b	1.47	0.007	0.001	0.762
Butyrate, mol/mol	4.80	5.09	6.94	5.42	0.37	0.155	0.239	0.206
A:P	1.09b	1.06b	1.48a	1.62a	0.08	0.015	0.003	0.529

¹A:P, the acetate-to-propionate ratio.

²L, linear effect; Q, quadratic effect; Trt, treatment effect.

^a–cDifferences between treatments (P < 0.05).

Stomach anatomical parameters

The BW before slaughter (BWS) and carcass weight of lambs increased (P < 0.05) linearly with increasing starter NDF levels. At 90 d of age, lambs fed NDF22 and NDF26 starters were heavier (P < 0.05) compared with those fed NDF14 starters.

Full weight and empty weight of stomach organs represent the weight of organs with and without digesta, respectively (Table 4). The full weights of the reticulorumen and omasum of lambs were similar among treatments (P > 0.05). But the full weight of the abomasum and empty weights of omasum and abomasum increased linearly with increasing NDF levels in the starter (P < 0.05). In contrast, the empty weight of reticulorumen decreased linearly with the NDF level (P = 0.024). The empty weight of the abomasum of lambs fed NDF26 starters was higher (P = 0.023) than those fed NDF14 and NDF18 starters.

Table 4.

Effect of starter NDF levels on anatomical development of stomach complex of lambs

	Treatments					P-value2
Items1	NDF14	NDF18	NDF22	NDF26	SEM	Trt	L	Q
BWS, kg	19.06b	21.15ab	22.12a	23.33a	0.55	0.030	0.004	0.648
Carcass weight, kg	8.47b	9.69ab	10.56a	10.88a	0.29	0.001	0.001	0.350
Full weight
Reticulorumen, kg	2.17	2.63	2.17	1.99	1.70	0.614	0.536	0.364
Abomasum, g	34.78	48.53	61.72	70.00	4.80	0.113	0.018	0.793
Omasum, g	187.30	194.77	189.05	196.05	16.61	0.998	0.906	0.995
Empty weight
WSC, g	749.45	722.75	619.98	632.77	28.11	0.277	0.080	0.723
WSC, g/kg of BWS	39.48a	34.30a	27.77b	26.82b	1.35	0.000	0.000	0.247
Reticulorumen, g	626.50	592.50	488.33	476.50	27.23	0.122	0.024	0.829
Reticulorumen, % of WSC	83.41a	81.78ab	78.50bc	74.28c	1.01	0.002	0.000	0.398
Reticulorumen, g/kg of BWS	32.91a	28.14a	21.83b	20.12b	1.33	0.000	0.000	0.384
Abomasum, g	30.23b	30.42b	34.78ab	41.25a	1.53	0.023	0.004	0.240
Abomasum, % of WSC	4.10c	4.25bc	5.59ab	6.87a	0.32	0.002	0.000	0.251
Abomasum, g/kg of BWS	1.62	1.45	1.56	1.77	0.06	0.304	0.304	0.119
Omasum, g	92.72	99.83	96.87	115.02	3.50	0.115	0.040	0.640
Omasum, % of WSC	12.49b	13.97b	15.91ab	18.86a	0.75	0.007	0.001	0.546
Omasum, g/kg of BWS	4.95	4.71	4.38	4.92	0.14	0.476	0.753	0.182
Rumen volume, L	6.06	6.75	6.35	5.82	0.34	0.821	0.744	0.413

¹Full weight or empty weight of stomach organs means the weight of organs with or without digesta.

² L, linear effect; Q, quadratic effect; Trt, treatment effect.

^a–cDifferences between treatments (P < 0.05).

When expressed as a percentage of the weight of the complex stomach (WSC), the reticulorumen weight decreased linearly (P < 0.05), whereas the omasum and abomasum weight increased linearly (P < 0.05) with increasing NDF levels in the starter. The reticulorumen weight of lambs expressed as a percentage of the WSC was lower (P = 0.002) when fed NDF22 and NDF26 starters compared with NDF14, whereas the omasum and abomasum weight of lambs as a percentage of WSC were higher (P < 0.05) fed NDF26 starters than those fed NDF14 and NDF18 starters. When expressed relative to BWS, the reticulorumen and stomach complex weight as a percentage of BWS decreased linearly (P < 0.001) with increasing NDF level in the starter. The reticulorumen and stomach complex weight as a percentage of BWS were reduced (P < 0.05) when lambs were fed NDF22 and NDF26 starters compared with NDF14 and NDF18.

Rumen morphometrics

When NDF levels in the starter were increased, a linear decrease (P < 0.05; Table 5) in the papillae width, epithelium thickness, and keratin layer thickness was observed, whereas a linear increase (P < 0.05) in muscle layer thickness of the rumen in the lambs was observed. Starter NDF levels did not affect papillae length (P > 0.05), but it influenced papillae width that was lower (P = 0.008) when lambs were fed NDF18, NDF22, and NDF26 starters compared with NDF14. Furthermore, lambs fed NDF18, NDF22, and NDF26 starters had a thinner keratin layer (P = 0.005) compared with those fed NDF14 starters. The muscle layer was thickest (P < 0.05) in lambs fed NDF14 starters compared with the other treatments. Epithelium thickness was also affected by starter NDF levels (P < 0.05), being thinnest for lambs fed NDF26 starters.

Table 5.

Effect of starter NDF levels on morphometrics of the rumen in lambs

	Treatments					P-value1
Items, μM	NDF14	NDF18	NDF22	NDF26	SEM	Trt	L	Q
Papillae length	1,628.16	1,577.31	1,623.34	1,614.19	45.85	0.984	0.992	0.830
Papillae width	353.35a	309.88b	297.49b	308.55b	6.52	0.008	0.008	0.033
Keratin layer thickness	64.23a	54.21b	57.11b	52.55b	1.29	0.005	0.004	0.273
Muscle layers thickness	458.25c	759.26a	586.04b	682.19ab	18.40	0.000	0.003	0.005
Epithelium thickness	126.23a	114.63a	117.16a	97.12b	2.35	0.000	0.000	0.335

¹L, linear effect; Q, quadratic effect; Trt, treatment effect.

^a–cDifferences between treatments (P < 0.05).

An example of representative images of the rumen wall is shown in Figure 1. Rumen images for lambs fed NDF14 and NDF18 starters showed the poor macroscopic evaluation of the rumen mucosa, containing focal or multifocal patches with coalescing and adhering papillae covered by a sticky mass of feed, hair, and cell debris. However, a better rumen mucosa is observed in lambs fed NDF22 and NDF26 starters without the phenomenon of feed plaque. Feed plaques were observed in the rumen of all the lambs (n = 6) fed NDF14 and NDF18 starters.

Figure 1.

Photographs of the rumen interior of lambs fed different NDF level starters. (A) NDF14; (B) NDF18; (C) NDF22; and (D) NDF26. NDF14, NDF18, NDF22, and NDF26 diets contained 14%, 18%, 22%, and 26% of NDF on a DM basis, respectively.

Discussion

Growth performance

As the main constituent of structural carbohydrates in forage, NDF is considered to be an important factor impacting the performance of young ruminants (Salinas-Chavira et al., 2013). For instance, Porter et al. (2007) found greater numerical differences in BW gain, starter intake, and rumen pH in pre-weaned calves fed high-NDF level (27%) in starters compared with those fed low-NDF level (20%) in starters. In the current study, with increasing starter NDF levels, the intake of starter and total DMI significantly increased during pre-weaning (21 to 60 d of age), post-weaning (61 to 90 d of age), and through the overall period (21 to 90 d of age). This might be associated with the ameliorating effects of starter NDF on the elevation of ruminal pH, which, at least in part, contributes to the improvement of the rumen fermentation environment. Likewise, the ADG and BW of lambs fed NDF22 and NDF26 starters were greater than those fed NDF14 starters during the pre- and post-weaning period in this study. Similar to our results, Nemati et al. (2016) reported that increasing alfalfa hay levels to 25% in starters (24.5% NDF) improved total DMI and ADG of post-weaned calves. Taken together, the NDF level of the starter can play a significant role in DMI and ADG of lambs. Exceptionally, Terré et al. (2013) observed that calves fed a low-NDF starter (18% vs. 27% NDF) tended to grow faster during the pre-weaning period. However, Terré et al. (2013) also mentioned that the provision of chopped hay is necessary to improve the performance. Considering the feed efficiency, the consensus among our results and previous studies suggest that differences in feed efficiency were not appreciable from the pre- and post-weaning period (Coverdale et al., 2004; Castells et al., 2012).

Rumen fermentation parameters

In agreement with previous investigations (Beiranvand et al., 2014a; Nemati et al., 2016), a greater ruminal pH in lambs fed NDF22 and NDF26 starters was observed compared with lambs fed NDF14 and NDF18 starters in this study. This could be attributed to fibrous materials (NDF) in the forage that do not ferment rapidly and may be beneficial in raising and stabilizing the ruminal pH (Khan et al., 2016). The increased rumen pH may also contribute to the reduction in the keratin layer thickness observed in this study (Pazoki et al., 2017). These results suggest that ruminal pH can be altered by starter NDF concentrations and that appropriate NDF levels play a significant role in maintaining pH stability. It is worth noting that different fat content in diets may affect VFA production by influencing bacterial attachment (Zhao et al., 2016). Consequently, TVFA concentrations were linearly decreased with increasing NDF levels in this study. As expected, high-NDF starters (NDF22 and NDF26) yielded more acetate, whereas high propionate production was generated by low-NDF starters (NDF14 and NDF18) in this study. Generally, types of carbohydrates (starch and NDF) are considered as vital factors affecting fermentation patterns (Žitnan et al., 1998). Greater rumen acetate-to-propionate ratios detected in NDF22 and NDF26 diets were mainly attributed to an increase in rumen acetate molar proportions, whereas the decrease in acetate-to-propionate ratios with low NDF level diets (NDF14 and NDF18) was mainly due to an increase in rumen propionate molar proportions.

Anatomical development

A linear decrease in reticulorumen weight in this study was observed with increasing starter NDF levels. It can be directly explained by the results of the variation of papillae width and epithelium thickness among treatments in this study. Furthermore, it may also be related to the decreases of non-fiber carbohydrates contained in the starter which is considered to promote the development of the rumen epithelium by producing more VFAs, especially butyrate and propionate (Baldwin and McLeod, 2000). The full weight of the reticulorumen was not affected by the starter NDF levels, indicating that the greater BW of lambs fed high-NDF level starter was not likely due to changes in gut fill but rather to tissue growth (Imani et al., 2017; Pazoki et al., 2017). Both the current study and previous studies (Khan et al., 2011; Mirzaei et al., 2015) noted a greater abomasal weight in the forage-supplemented diets (high NDF concentration) compared with non-forage-supplemented calves. It may be attributed to ruminal fractional passage rate because forage can increase ruminal fractional passage rate and result in more undigested feed reaching the abomasum, which might induce greater development of the abomasum tissue (Castells et al., 2013).

Morphometric characteristics

Focal patches with coalescing and adhering papillae covered by a sticky mass of feed were observed by rumen images from lambs fed low-NDF starters (NDF14 and NDF18) in this study. Similar to our results, Beiranvand et al. (2014b) reported that papillae width and epithelium thickness decreased in calves fed the diet with 14.8% NDF compared with those fed diet containing 11.6% NDF, suggesting that high NDF levels may have a negative effect on rumen epithelial tissue development. Previous reports (Norouzian and Valizadeh, 2014; Mirzaei et al., 2015) considered that physical factors from the forage might result in the decreasing keratinization of the inner surface of the rumen because the length of fiber increases the ability for physical removal of the dying epithelial cells. However, our results confirmed that starter NDF levels might be more beneficial to depress parakeratosis. The difference in muscle layer thickness among treatments could be attributed to different starter intake, because higher starter and NDF intake can lead to greater motility and contractions of rumen wall, thereby creating a larger and stronger rumen muscle in lambs.

Conclusions

Increasing starter NDF levels improved DMI, ADG, and final BW, but no effects were observed for feed efficiency in lambs. Furthermore, considering the 22% and 26% NDF levels in the pelleted starter, ameliorated rumen fermentation and morphometrics by improving ruminal pH and acetate concentration, the plaque formation alleviation, and reduction in the thickness of the keratin layer were observed. Therefore, the better rumen development resulted in enhanced performance in lambs. Our results also suggest that starter NDF levels appear to be a primary factor affecting performance and rumen development of lambs.

Glossary

Abbreviations

ADG

average daily gain

BW

body weight dry matter intake

BWS

BW before slaughter

CP

crude protein

DM

dry matter

DMI

dry matter intake

MR

milk replacer

NDF

neutral detergent fiber

NDF14

diet with 14% NDF

NDF18

diet with 14% NDF

NDF22

diet with 22% NDF

NDF26

diet with 26% NDF

TVFA

total VFA

VFA

volatile fatty acids

WSC

weight of the complex stomach

Conflict of interest statement

The authors declare no real or perceived conflicts of interest.