Effects of forage rape at various levels on growth performance, carcass traits, meat 1 quality, rumen fermentation and rumen microbiota of Hu lambs

11 Background: The objective of this study was to investigate the effects of forage rape 12 ( Brassica napus ) in total mixed ration (TMR)-based diet on growth performance, 13 carcass traits, meat quality, rumen fermentation and rumen microbiota of Hu lambs. 14 Methods: A total of 50 Hu lambs (20.86 ± 3.00 kg, three-month-old) were randomly 15 allocated into five dietary treatments: Ctrl, T1, T2, T3 and T4 with 0, 10%, 20%, 30% 16 and 40% of forage rape, respectively. Each treatment had 10 replicates of one lamb 17 each and the study lasted for 60 days. 18 Results: The results showed that T2, T3 and T4 increased the average daily gain (ADG) 19 and decreased the feed conversion ratio (FCR) significantly compared with Ctrl and T1 20 ( P < 0.05). Moreover, the final body weight and ADG were increased, and the FCR 21 were decreased linearly ( P < 0.05) along with increasing forage rape levels. The relative 22 weight of liver was increased in T3 and T4 compared with Ctrl ( P < 0.05). With 23 increasing forage rape levels, the relative content of intramuscular heptadecenoic acid 24 and α-linolenic acid, and the composition of various amino acids in the muscle of lambs 25 were increased linearly ( P < 0.05), while ruminal concentration of ammonia nitrogen 26 was decreased linearly ( P < 0.05). No difference on carcass traits, meat quality or 27 ruminal profile of short-chain fatty acids were observed among groups ( P > 0.05). In 28 addition, the inclusion of forage rape altered the rumen microbial community, and 29 increased the relative abundance of cellulolytic bacteria and short-chain fatty acid 30 producers, including genera Family_XIII_AD3011_group and Anaerovorax in T1, Succiniclasticum and Fibrobacter in T2, Ruminiclostridium_5 in T3, and members of 32 family Lachnospiraceae and genus Shuttleworthia in T4. 33 Conclusion: TMR included with forage rape could improve growth performance, meat 34 nutritional value and rumen microbial community of Hu lambs. 35

quite high [3][4][5]. Therefore, forage rape is gradually being used as a high-quality forage 48 for ruminants in Southern China in recent years [2]. 49 Actually, forage rape has been found to support the rapid growth of ruminants. When 50 fed as the sole diet, the average growth of young sheep on forage rape was 225 g/d, 51 which was much higher than on kale and swedes (120 g/d and 95 g/d, respectively) [6]. 52 As a supplementation, forage rape was reported to reduce the acetate-to-propionate ratio 53 and energy losses in the rumen, mainly methane emissions, resulting in improved feed 54 efficiency [7]. 55 However, to the best of our knowledge, no study has been conducted to explore the 56 utilization of forage rape in the form of total mixed ration (TMR) pellets, and little data 57 10 RDP Classifier for taxonomic classification (at 80% confidence threshold) at the 162 kingdom, phylum, class, order, family, and genus levels. 163

Statistical Analysis 164
Results were expressed as treatment means and pooled standard error of the mean 165 (SEM). Phenotypic data (growth performance, carcass traits, meat quality and 166 nutritional value, and ruminal fermentation parameters) and the alpha diversity indexes 167 of the rumen microbiota were analyzed using one-way ANOVA and Duncan's multiple 168 comparison test by SPSS 21.0 software (IBM Inc., NY, US). Polynomial contrasts were 169 performed to determine the linear and quadratic effects of increasing dietary forage rape 170 on the measured traits. P ≤ .05 was considered significant. For the relative abundances 171 of rumen bacteria, LEfSe analysis was utilized to determine the difference, and a 172 significant change was observed with an LDA (linear discriminant analysis) score > 2.0 173 calculated by LEfSe. 174

Growth performance and carcass traits 176
The dietary inclusion level of forage rape had no influence on the DMI of lambs (P = 177 0.855), but group T2, T3 and T4 increased the final BW and ADG significantly 178 compared with Ctrl (P < 0.05), resulting in remarkably decreased FCR (P < 0.05) as 179 shown in Table 2. In addition, the final BW (P = 0.005) and ADG (P < 0.001) increased 180 linearly with increasing forage rape levels, and FCR decreased both linearly (P < 0.001) 181 and quadratically (P < 0.001) with increasing forage rape levels. 182 11 There was no difference in the dressing percentage or LD area among the five dietary 183 treatment groups (P > 0.10). However, the liver index was increased significantly in 184 group T3 and T4 (P < 0.05), and the liver index, the slaughter weight and HCW of the 185 lambs exhibited a linear increase with increasing forage rape levels (P = 0.003, 0.012 186 and 0.036, respectively). A quadratic effect of forage rape levels was found on the index 187 of kidney (P = 0.029), and group T2 and T3 showed the highest kidney index. 188

Meat quality and nutritional value 189
For lambs fed diets with various levels of forage rape, no difference was observed in 190 the pH value, water loss rate, cooked meat rate or proximate nutrition (moisture,CP,191 crude fat and crude ash) of the meat (P > 0.10, Table 3). 192 Compared with Ctrl and T1, the relative content of intramuscular heptadecenoic acid 193 (C17:1n-7) in the LD muscle was increased in T3 and T4 (P < 0.05, Table 4), and the 194 relative content of α-linolenic acid (C18:3n-3) was increased in T2, T3 and T4 (P < 195 0.05). In addition, the relative content of heptadecenoic acid and α-linolenic acid 196 showed a linear increase with increasing forage rape levels (P < 0.001), while a 197 quadratic effect of forage rape levels was found on the relative content of heptadecenoic 198 acid (P = 0.047). Overall intramuscular SFAs (saturated fatty acids), MUFAs 199 (monounsaturated fatty acids) and PUFAs (polyunsaturated fatty acids) contents were 200 not affected by feed (P > 0.10). In the current study, the contents of two long-chain 201 omega-3 fatty acids, eicosapentaenoic acid and docosahexaenoic acid, were too low to 202 be detected. 203 No difference in the composition of amino acid was found in the LD muscle among 204 the five dietary treatment groups (P > 0.05, Table 5). However, with increasing forage 205 rape levels, the contents of valine, threonine, leucine, isoleucine, methionine, arginine, 206 and umami amino acids (aspartic acid and glutamic acid) were increased linearly (P < 207 0.05). 208

Ruminal fermentation parameters 209
The ruminal fluid maintained a normal pH value (6.8 to 7.2, Table 6) for lambs fed 210 different diets, and the level of forage rape did not influence the ruminal pH value, the 211 concentration of total SCFA or the profile of SCFA (P > 0.10). However, the 212 concentration of ammonia nitrogen was found to be decreased linearly with increasing 213 forage rape levels (P = 0.019). 214

Rumen microbiota 215
In the present study, an average of 176777 raw reads were obtained from the rumen 216 microbiota, and an average of 88388 clean reads were remained. The rarefaction curves 217 of all the samples were nearly asymptotic (Supplementary Figure S1), indicating that 218 the depth of sequencing covered most of the microorganisms in the sample. In regard 219 to the alpha diversity, no difference was observed among the five groups except that 220 Chao1 and Ace indices were decreased in T1 compared with Ctrl (P < 0.05, Figure 1). 221 In addition, a quadratic effect of forage rape levels was found on Chao1 and Ace indices 222 genera Schwartzia, Lachnospiraceae_UCG_008, Lachnospiraceae_AC2044_group, 14 Family_XIII_AD3011_group and Succiniclasticum were the taxa that weighted most to 246 the differences among the communities, with an absolute LDA score more than 3. 247

Discussion 248
Owing to the fast growth rate and high nutritional value of forage rape, it has been 249 proposed as a promising feed crop for ruminants, which partially alleviates the lack of 250 high-quality forage in Southern China [2,3]. To our knowledge, this was the first study 251 to systemically evaluate the application effect of forage rape at various levels on sheep 252 in the form of TMR pellets. 253 Under the conditions of this study, Hu lambs fed diets with 20-40% of forage rape 254 had greater final BW, ADG and feed conversion efficiency than lambs fed diets without 255 or with 10% of forage rape. Since DMI was not affected, the result implied greater feed 256 efficiency of forage rape than peanut vine, which promoted the growth of lambs. Apart 257 from high rumen degradation rate of the main nutrients [3, 10], sheep and cattle fed 258 forage rape were found to emit less enteric methane, resulting in less energy loss [7]. 259 Glucosinolates are main anti-nutritional factors in forage rape and rapeseed, which 260 restrict their utilization as feed resources. Compared with rapeseed, the content of 261 glucosinolates is lower in vegetative tissues, like root, stem and leaf [11]. Besides, 262 ruminants have relatively stronger tolerance capability to glucosinolates [12]. In our 263 study, the glucosinolates concentration of diet included with 40% of forage rape was 264 4.07 μmol/g, within the tolerance threshold of lambs. Therefore, up to 40% of forage 265 rape in TMR had no adverse effect on the growth performance, carcass traits or the 266 15 index of thyroid of Hu lambs in the present study. However, the index of liver was 267 elevated when 30-40% of forage rape was included in TMR, which might be mild 268 hepatic edema caused by glucosinolates metabolites, like nitriles [12]. 269 Nowadays there is increasing consumer demanding for high-quality meat products 270 [14], the content of α-linolenic acid in the grilled loins was increased when forage rape 279 was used as finishing feed for lambs compared with ryegrass. In the current study, the 280 relative content of α-linolenic and heptadecenoic acid were increased when lambs were 281 fed 20-40% of forage rape. Despite that, the predominant SFAs were palmitic (C16:0) 282 and stearic (C18:0) acids, and the predominant MUFA was oleic acid (C18:1n-9c) for 283 lambs fed different diets, consistent with previous results [14,15]. And the predominant 284 fatty acids and overall SFAs, MUFAs and PUFAs were not influenced by diets. In the 285 present study, the amino acid composition was not affected by the dietary treatments. 286 However, five kinds of essential amino acid (EAA) and three kinds of non-essential 287 amino acid (NEAA) showed a linear increase with increasing forage rape levels. The 288 percentage of amino acids producing the tastes of umami was also increased linearly. 289 The EAA and NEAA requirements of an adult man are 0.18 g/kg per day and 0.48 g/kg 290 per day, respectively, which equals EAA/NEAA = 37.5%. In this study, the mean ratios 291 of EAA/NEAA of the LD muscle were 74-76%, which were much higher than those 292 recommended by FAO/WHO/UNU [16], therefore lamb appears to be an excellent 293 source of protein. linearly with increasing forage rape levels, implying a possible linear increase in 300 nitrogen utilization, which is in accordance with the linear increase in the growth 301 performance of lambs. According to the present study, increasing forage rape levels 302 from 0 to 40% did not affect the profile of SCFA, suggesting that the ruminal 303 fermentation pattern of the diets was unaffected by the inclusion of forage rape. 304 Although a few investigations have evaluated the nutritional properties of forage rape 305 used in ruminants [3,4,6,7], less information is available on the rumen microbiota. In 306 the current study, the divergence of the rumen microbial communities of Hu lambs fed 307 different diets confirmed that rumen microbiota could response to changes in the diets. 308 When lambs were fed diets with 10% of forage rape, the community richness of rumen 309 microbiota decreased since Chao1 and Ace indices were reduced, but the community 310 diversity remained unchanged since similar Shannon and Simpson indices were 311 observed compared with group Ctrl. 312 Out of the eighteen significantly abundant genera with LDA score larger than two, 313 Quinella is a propionate-producing bacterium [19]. Some members of the genus 314 Eubacterium_uniforme are cellulolytic bacteria capable of producing butyrate. When 335 30% of forage rape was included in the diet of lambs, the genus Ruminiclostridium_5 336 was found to be abundant in the rumen. Ruminiclostridium can effectively utilize 337 cellulose, cell wall polysaccharides and raw lignocellulose feedstocks to improve the 338 digestibility of feed nutrition [28]. Similar with Succiniclasticum, the genus Schwartzia 339 is another succinate-specific bacteria and propionate producer [29]. In addition, 40% of 340 forage rape promoted the abundance of several butyrate-producing bacteria, including 341 Lachnospiraceae UCG 008, Lachnospiraceae AC2044 group, Blautia (also from the 342 family Lachnospiraceae) and Shuttleworthia. Since the rumen fluid was collected after 343 24 hours of fasting, we did not observe a higher butyrate or propionate concentration in 344 the current study. However, it seemed that feeding of animals with forage rape 345 facilitated the growth of SCFA producers in the rumen and enhance the ruminal 346 fibrolytic function, thus elevated the ADG and feed conversion efficiency of lambs. 347

Conclusion 348
In conclusion, increasing dietary inclusion of forage rape from 0 to 40% improved the 349 growth performance and increased the content of α-linolenic acid and a variety of amino 350 19 acids in the muscle of Hu lambs linearly, while no detrimental impact on carcass traits, 351 meat quality or ruminal fermentation parameters was observed. Though limitations of 352 our study included that the maximum inclusion level of forage rape was 40%, the linear 353 effects indicate that better results could be expected with higher levels of forage rape. 354 Furthermore, this study also provides the first evidence that the inclusion of forage rape 355 altered the rumen microbial community, and increased the relative abundance of System. The funders had no role in study design, analysis or writing of this article. 385

Availability of data and materials 386
All data generated or analyzed during this study are available from the corresponding 387 author by request. The datasets supporting the conclusions of this article are included 388 in the article. 389

Ethics approval and consent to participate 390
The use of animals, including welfare, husbandry, experimental procedures, and the 391 collection of samples used for this study, were conducted according to the principles of 392 21 the Animal Care and Use Committee of the Hubei Academy of Agricultural Sciences 393 (Hubei, China), which approved the study protocol. 394

Consent for publication 395
Not applicable. 396