Ruminal microbiota is associated with feed-efficiency phenotype of fattening bulls fed high-concentrate diets

Ruminal mic obio a is associa ed wi h eed-e ficiency pheno ype o a ening bulls ed high-concen a e die s S. Cos a-Rou a A , D. Villalba A,D , M. Blanco B,C , I. Casasús B,C , J. Balcells A and A. R. Se adj A A Depa amen de Ciència Animal, Uni e si a de Lleida, A inguda Alcalde Ro i a Rou e 191, 25198, Lleida, Spain. B Unidad de P oducción y Sanidad Animal, Cen o de In es igación y Tecnología Ag oalimen a ia de A agón (CITA), A enida Mon añana 930, 50059, Za agoza, Spain. C Ins i u o Ag oalimen a io de A agón –IA2 (CITA-Uni e sidad de Za agoza), Za agoza, Spain. D Co esponding au ho . Email: [email p o ec ed] Abs ac Con ex . Imp o ing eed e ficiency in li es ock p oduc ion is o g ea impo ance o educe eeding cos s. Aims. To examine he ela ionship be ween uminal mic obio a and a ia ion in eed e ficiency in bee ca le ed concen a e-based die s. Me hods. Residual eed in ake o 389 a ening bulls, supplied wi h co n-based concen a e and o age ad libi um, was used oes ima eanimals’ eede ficiency.Faecesand uminalfluidsamples, om48bullschosena andom,we ecollec ed o es ima e hei o age in ake and o de e mine hei appa en diges ibili y, uminal e men a ion and mic obio a. Those animalswi hex eme alueso eede ficiency(high-e ficiency(HE,n=12)andlow-e ficiency(LE,n=13))we esubjec ed o u he compa isons. Alphabiodi e si y was calcula ed on he basis o heno malised sequence da a. Be a di e si y was app oached h ough pe o ming a canonical co espondence analysis based on log- ans o med sequence da a. Gene a di e en ial abundance was es ed wi h an ANOVA-like di e en ial exp ession analysis and gene a in e ac ions we e de e mined applying he spa se co ela ions o composi ional da a echnique. Key esul s.No di e ences in d y ma e in ake we e ound be ween he wo ca ego ies o eed e ficiency (P=0.699); howe e , HE animals had highe appa en diges ibili y o d y ma e (P= 0.002), o ganic ma e (P= 0.003) and c ude p o ein (P= 0.043). The concen a ion o ola ile a y acids was una ec ed by eed e ficiency (P= 0.676) bu bu y a e p opo ion inc eased wi h ime in LE animals (P= 0.047). Ruminal mic obio a was di e en be ween HE and LE animals (P= 0.022); bo h abiodi e si y and gene a ne wo k connec ance inc eased wi h ime in LE bulls (P= 0.005 o Shannon index and P= 0.020 o Simpson index), which sugges s ha LE animals hos ed a mo e obus uminal mic obio a. Ce ain gene a usually ela ed o high ene gy loss h ough me hane p oduc ion we e ound o es ablish mo e connec ions wi h o he gene a in LE animals’ umen han in HE ones. Mic obio a unc ion capabili y sugges ed ha me hane me abolism was dec eased in HE finishing bulls. Conclusions.Rumenmic obio awasassocia edwi h eede ficiencypheno ypesin a eningbulls edconcen a e-based die s. Implica ions. The possible ade-o be ween eed e ficiency and obus ness o uminal mic obio a should be aken in o accoun o he op imisa ion o ca le p oduc ion, especially in sys ems wi h in insic cha ac e is ics ha may cons i u e a dis u bance o umen mic obial communi y. Keywo ds: appa en diges ibili y, bee ca le, eed e ficiency, umen mic obial communi y. Recei ed 5 June 2020, accep ed 27 No embe 2020, published online 28 Janua y 2021 In oduc ion Imp o ing eed e ficiency (FE) in bee ca le p oduc ion sys ems p o ides an oppo uni y o cu down on he cos o eeding li es ock. In ha sense, esidual eed in ake (RFI) can be used as an index o FE ha is independen o a ia ion in bodyweigh (BW) and a e age daily gain (ADG; A hu e al. 2001; Schenkel e al.2004; A hu and He d 2008), being he gold s anda d index o examine biological mechanisms associa ed wi h in e -animal di e ences in FE. Mo eo e , some s udies ha e demons a ed he possibili y o selec ion o low RFI as a s a egy o g eenhouse gas mi iga ion, as i has been co ela ed wi h lowe me hane emission and g ea e die diges ibili y (He d and A hu 2009). Limi a ions in conduc ing RFI ials ( eco ding BW and eed in ake o a CSIRO PUBLISHING Animal P oduc ion Science h ps://doi.o g/10.1071/AN20344 Jou nal compila ion CSIRO 2021 Open Access CC BY www.publish.csi o.au/jou nals/an SPECIAL ISSUE long ime) and sea ching o umen mic obial ma ke s o iden i y e ficien animals wi h low RFI ha e become a con empo a y challenge. Resea ch in ca le has ocused mos ly on he mic obial esponse o die a y changes and managemen p ac ices, whe eas ials o unde s anding he ela ionship be ween hos FE pheno ype and umen mic obio a a e sca ce and ye o be unde aken (Mye e al.2015). P e ious s udies ha e shown ha umen mic obes a e esponsible o ene gy supply h ough p oducing o ganic acids (Hun ing on 1990), and mos axa associa ed wi h a ia ion in FE ha e been ela ed o celluloly ic, e men a i e and me abolic ac i i ies (Mye e al.2015). The e o e, di e ences in he p oduc ion a e o o ganic acids lead o a ia ion in nu ien diges ibili y and e men a ion ha ul ima ely change animals’pheno ypic e ficiency (He d and A hu 2009). This expe imen aimed o unde s and he ela ionship be ween uminal mic obio a and a ia ion in FE o bee ca le ed concen a e-based die s. Ma e ials and me hods Animals, die s and housing Residual eed-in ake da a om wo eeding expe imen s comp ising 389 a ening bulls we e used o explo e ela ionships be ween uminal mic obio a and FE. This da ase included 317 animals aised a he esea ch acili ies o Coope a i a d’I a s d’U gell,SCCP (I a s d’U gell, Spain, 414105000N, 05805300E) and 72 animals om he CITA-La Ga cipolle a Resea ch S a ion (Jaca, Spain, 423703400N, 03001000W). All p ocedu es we e pe o med unde P ojec Licence CEEA 01-07/16 and app o ed by he in-house E hics Commi ee o Animal Expe imen s a he Uni e si y o Lleida. Ca e and use o animals we e in acco dance wi h he Spanish Policy o Animal P o ec ion RD 53/2013, which mee s he Eu opean Union Di ec i e 2010/63 on he p o ec ion o animals used o expe imen al and o he scien ificpu poses. Animals aised a he esea ch acili ies o Coope a i a d’I a s d’U gell,SCCP we e dis ibu ed in he ollowing ou ba ches: ba ches Numbe 1 o Numbe 3 included 231 Hols ein bulls (63–83 animals pe ba ch) and ba ch Numbe 4 included 86 Mon belia d bulls. BW and eed in ake da a we e collec ed on a daily basis. Animals aised a CITA-La Ga cipolle a Resea ch S a ion we e dis ibu ed in he ollowing h ee ba ches: ba ches Numbe 5 and Numbe 6 included 28 and 32 Pa da de Mon aña bulls espec i ely, and ba ch Numbe 7 included 12 Pi enaica bulls. Fo hese animals, BW was measu ed weekly and eed in ake da a we e collec ed on a daily basis. Bodyweigh and eed in ake da a we e eco ded h oughou he en i e a ening phase; he fi s 150 days we e conside ed as he g owing phase (121 days old, s.d. 37 days; and 162 kg BW, s.d. 49 kg), ollowed by a finishing phase, which las ed un il animals eached slaugh e weigh (336 days old, s.d.: 31; and 501 kg BW, s.d. 56 kg). Animals we e ed concen a e and o age ad libi um,which we e p o ided sepa a ely in wo di e en bunke s, and hey had ee access o d inking wa e , ollowing he con en ional bee ca le eeding sys em in Spain. The concen a es used we e e y simila in composi ion and hei main ing edien s we e aw co n, co n glu en eed, aw ba ley, co n d ied dis ille s g ains wi h solubles and aw chickpea; whe eas o age used was mainly ba ley s aw (349 animals), oa s haylage (20 animals) and e ch haylage (20 animals). Feed chemical and nu i ional composi ion is shown in Table 1. Measu emen s and sampling In ake o concen a es was eco ded au oma ically a bo h esea ch acili ies, by using au oma ic eed s a ions ha we e equipped wi h a eedbunk (p o ided wi h a scale) and an indi idual eede . When a cal en e ed he eede , i was iden ified and i s concen a e in ake was ob ained by di e ence be ween ini ial and final eedbunk weigh . Feed s a ions a ailable a he esea ch acili ies o Coope a i a d’I a s d’U gell,SCCP we e addi ionally equipped wi h a scale unde he indi idual eede by which he animals we e au oma ically weighed a each isi , whe eas a CITA-La Ga cipolle a Resea ch S a ion, BW da a we e eco ded manually once a week. Faeces and uminal fluid samples om 48 bulls (selec ed a andom wi hin ba ches) we e collec ed wice, a mid-g owing pe iod (GRO, 159 days old and 225 kg BW) and mid-finishing pe iod (FIN, 266 days old and 434 kg BW), o o age in ake es ima ion and diges ibili y, uminal e men a ion and mic obial communi y cha ac e isa ion. Faecal exc e ion and o age in ake we e calcula ed on he basis o concen a e in ake and adap ing he wo indiges ible- ma ke sys em (Owens and Hanson 1992), by using ch omium oxide as an ex e nal ma ke and acid insoluble ash as in e nal ma ke . Then, appa en diges ibili y o d y ma e (DM), o ganic ma e (OM) and c ude p o ein (CP) we e es ima ed. De ailed in o ma ion abou ma ke adminis a ion, eed and aeces analy ical de e mina ions and appa en diges ibili y calcula ions ha e been desc ibed in Cos a-Rou a e al.(2020). Ruminal fluid was sampled in he mo ning by using an o al s omach ube connec ed o a acuum pump. Each sample was Table 1. Feed chemical and nu i ional composi ion Valuesa emeans, wi h minimumandmaximum gi en inpa en heses. ADF, acid de e gen fib e; CP, c ude p o ein; DM, d y ma e ; EE, e he ex ac ; NDF, neu al de e gen fib e; OM, o ganic ma e ; PDIN and PDIE, p o ein diges ible in he small in es ine allowed by p o ein and ene gy; UFV, o age uni o mea p oduc ion Pa ame e Concen a e Fo age Chemical composi ion (%DM) DM (% esh weigh ) 87.1 (85.6–87.9) 65.8 (48.9–85.7) OM 94.6 (94.0–94.8) 88.9 (84.3–92.6) CP 13.0 (11.2–14.0) 11.2 (7.2–16.4) EE 4.2 (2.5–7.2) 2.0 (1.4–2.7) NDF 16.6 (13.5–20.7) 57.1 (44.2–75.5) ADF 5.9 (4.7–7.9) 33.5 (28.2–43.8) Nu i ional composi ion UFV (UFV/kg DM) 1.02 (0.97–1.03) 0.55 (0.36–0.73) PDIN (g/kg DM) 91.8 (79.8–95.9) 65.4 (40.8–93.7) PDIE (g/kg DM) 87.4 (80.3–94.9) 56.0 (52.6–58.0) BAnimal P oduc ion Science S. Cos a-Rou a e al. ob ained h ough wo sequen ial collec ions. Fi s , uminal fluid (~200 mL) was collec ed and disca ded o a oid sample con amina ion wi h sali a ha could ge in o he ube du ing i s in oduc ion h ough he animal’s mou h and oesophagus. A e ha , uminal fluid (~200 mL) was e- ex ac ed, s ained h ough a cheeseclo h and i s pH was eco ded (Tes o 205, Tes o AG, Ge many). Then, uminal fluid was sampled o DNA ex ac ion, and de e mina ion o ammonia-ni ogen (N) and ola ile a y acid (VFA) concen a ions, and immedia ely ozen on d y ice. Sample p ese a ion condi ions and analy ical p ocedu es o ammonia-N and VFA de e mina ion a e de ailed in Cos a- Rou a e al.(2020). Ex ac ion and sequencing o DNA Ex ac ion o DNA was pe o med on eeze-d ied uminal fluid ( he ini ial amoun o he sample was 60 mg) h ough physical dis up ion (1 min) by using a bead bea e (Mini-bead Bea e 1, BioSpec P oduc s, USA) and subsequen DNA pu ifica ion was pe o med wi h he QIAamp DNA S ool Mini Ki (ID: 51504; QIAGEN N.V., Ge many), wi h he modifica ions o g ea e empe a u e (95C) and g ea e elu ion ime (3 min) o ensu e maximum DNA concen a ion in he final elu e. Amplifica ion o DNA was pe o med by using p ime s 341F and 805R, which a ge he V3 and V4 egions o he bac e ial and a chaeal 16S RNA. Sequencing was conduc ed on an Illumina MiSeq 2x300 pla o m by E a7 Bioin o ma ics (Spain). Assembly and fil a ion o sample eads, as well as ope a ional axonomic uni (OTU) p epa a ion ha e been de ailed in Cos a-Rou a e al.(2020). Es ima ion o RFI and clus e ing Weigh da a we e fi ed o a hi d-deg ee polynomials model in unc ion o age (Eqn 1) ha allows he es ima ion o he ADG o each animal a any age. Weigh i;age ¼P j¼3 j¼0bBATCH;jþAi;j  agejþei;age ð1Þ whe e b BATCH,j is he ba ch e ec (fixed); A i,j, is he j h andom coe ficien o he i h animal e ec ; age is he age o he animal (days) and e i,age he esidual e m. The ea e , ADG_de was ob ained as he fi s de i a i e o Eqn 1 o each mon h by using he mon hly a e age age o each animal (Eqn 2). The indi idual ADG de ia ion (ADG_de ) will accoun o hedi e enceo g ow ho heanimalcompa edwi h he a e age o he ba ch a each age. ADG de i;age ¼P j¼3 j¼1j·Ai;jage j-- 1ðÞ ð2Þ In o al, 86 eco ds (3%) ou o h ee s anda d de ia ions o he mean we e conside ed as ou lie s and excluded om he da ase . Residual eed in ake was modelled (Eqn 3) using he andom eg ession coe ficien app oach p oposed by Sa ie o e al. (2014). The model included ba ch, age (mon hs), ADGde , me abolic weigh (MW; mon hly mean BW 0.64 ) and was defined as ollows: FIij ¼B0;animal i þBa ch ·agejþðBa ch þB1;animal iÞ ·MWij þðBa ch þB2;animal iÞ·ADG de ij þeij ð3Þ whe e FI ij is DM in ake measu ed o Animal iin Mon h jand B k , animal i a e he andom coe ficien s o animal e ec modelled using an uns uc u ed ma ix o a iances be ween hem. The inclusion o ba ch e ec in Eqn 1and Eqn 3assu ed ha he FE calcula ed was no a ec ed by die di e ences. On he basis o he indi idual coe ficien s o ADG (B 1, animal i ) and MW (B 2, animal i ), animals we e seg ega ed in o ou ca ego ies o FE, as ollows: (1) Animals wi h posi i e coe ficien s o bo h ADG and MW belonged o ‘low-e ficiency in ADG and low-e ficiency in MW’ca ego y. (2) Animals wi h posi i e coe ficien o ADG bu nega i e coe ficien o MW belonged o ‘low-e ficiency in ADG and high-e ficiency in MW’ca ego y. (3) Animals wi h nega i e coe ficien o ADG bu posi i e coe ficien o MW belonged o ‘high-e ficiency in ADG and low-e ficiency in MW’ca ego y. (4) Animals wi h nega i e coe ficien s o bo h ADG and MW belonged o ‘high-e ficiency in ADG and high-e ficiency in MW’ca ego y. Fo he pu pose o he p esen s udy, he wo ex eme ca ego ies (1 and 4) we e conside ed as high-e ficiency (HE, posi i e RFI) and low-e ficiency (LE, nega i e RFI) animals, espec i ely. This clus e ing (HE s LE) was subjec ed o bioin o ma ic analyses o appa en diges ibili y, uminal e men a ion and mic obio a da a as explained below. Bioin o ma ics Sequence da a we e no malised and abiodi e si y indices we e calcula ed o measu e he a iabili y o OTUs wi hin a sample (R Co e Team 2020, Vegan package). To measu e di e ences in mic obio a composi ion among samples, bdi e si y was app oached h ough pe o ming a canonical co espondence analysis, based on log- ans o med OTU da a (ze os we e eplaced by adding 1 o each alue), and including FE (HE s LE), pe iod (GRO s FIN) and bo h ADG and MW coe ficien s as explana o y a iables (R Co e Team 2020, Vegan package). To ci cum en he composi ional bias p oblem (Tsilimig as and Fodo 2016; Gloo e al.2017;Calle2019), we applied he Ai chison’s cen ed log a io (cl ) ans o ma ion o ca y he da a o a Euclidean space, a e eplacing ze os by adding 1 o each alue. So as o es he significance o he ollowing e ec s: FE (HE s LE), pe iod (GRO s FIN) and bo h ADG and MW coe ficien s on mic obio a composi ion, a pe mu a ional mul i a ia e analysis o a iance (Adonis) was conduc ed on he basis o he cl Euclidean dis ance and calcula ing s a is ical significance a e 10000 andom pe mu a ions (R Co e Team 2020, Vegan package). So as o deciphe which gene a abundance we e esponsible o he di e ences among g oups, an ANOVA-like di e en ial Ruminal mic obio a and eed e ficiency in bulls Animal P oduc ion Science C exp ession (ALDEx) analysis was conduc ed o e hose gene a p esen a leas a 50% o he indi iduals (R Co e Team 2020, Aldex2 package; Fe nandes e al.2013). Finally, o desc ibe he in e ac ions wi hin umen mic obial communi y, we pe o med a ne wo k analysis h ough spa se co ela ions o composi ional da a (Spa CC) echnique (R Co e Team 2020, SpiecEasi package; F iedman and Alm 2012) o e hose gene a p esen a leas a 50% o he indi iduals. Mic obial ne wo ks we e g aphically ep esen ed (R Co e Team 2020; ig aph package) and hei complexi y was desc ibed in e ms o numbe o nodes (gene a), numbe o edges (significan posi i e o nega i e co ela ions), node deg ee (numbe o connec ions ha any node es ablishes wi h o he nodes) and be weenness (measu e o cen ali y in a g aph based on sho es pa hs). Mic obio a unc ional con en was assessed using a opic model app oach (R Co e Team 2020; heme agenomics package) ha consis s on (1) cap u ing g oups o co- occu ing axa e med ‘ opics’, (2) unco e ing wi hin- opic unc ional po en ial, and (3) linking hese opics and hei unc ional con en o specific sample ea u es (e.g. FE pheno ypes; Woloszynek e al.2019). S a is ical analyses The models o RFI we e sol ed using MIXED p ocedu e o SAS s a is ical so wa e (SAS 9.4, Ca y, NC, USA). In ake, appa en diges ibili y, uminal e men a ion pa ame e s and mic obial abiodi e si y da a we e analysed wi h a mixed model, including FE (HE s LE), pe iod (GRO s FIN) and hei in e ac ion as fixed e ec s and animal as a andom e ec , o accoun o epea ed measu emen s (R Co e Team 2020, lme4 package). Di e ences among leas squa e means we e assessed using Tukey mul iple-compa ison es (R Co e Team 2020, emmeans package). Indi idual samples ou o h ee s anda d de ia ions o he mean we e disca ded and no included in he s a is ical analysis. Resul s a e epo ed as leas squa e means and s anda d e o o mean. Significan e ec s we e decla ed a P<0.05 and endency o di e ence a Pbe ween 0.05 and 0.10. Resul s The ou defined FE ca ego ies based on andom eg ession coe ficien s included be ween 85 and 108 animals each; he wo ex eme ca ego ies co esponding o HE and LE animals had s a is ically di e en means o ADG and MW coe ficien s (Table S1, a ailable as Supplemen a y Ma e ial o his pape ). The se o 48 bulls ha we e sampled o appa en diges ibili y, uminal e men a ion and mic obio a cha ac e isa ion we e equally dis ibu ed wi hin he ou ca ego ies. The FE by pe iod in e ac ions we e no significan o any o he esponse a iables measu ed in he p esen s udy; hus, only he main e ec s means a e p esen ed and discussed. In ake, appa en diges ibili y and uminal e men a ion pa ame e s Da a on DM in ake (Table 2) indica ed ha bulls’concen a e and o age in akes we e simila be ween FE ca ego ies (HE s LE). Howe e , animals classified as HE had g ea e appa en Table 2. D y ma e (DM) in ake, nu ien appa en diges ibili y and uminal e men a ion pa ame e s Ob ained in in ensi ely ea ed bulls in he ollowing wo pe iods: g owing (GRO: 159 days old and 225 kg bodyweigh ) and finishing (FIN: 266 days old and 434 kg bodyweigh ). Residual eed in ake was modelled o classi y animals in o wo ca ego ies o eed e ficiency, namely, high e ficiency (HE, n = 12) and low e ficiency (LE, n = 13). S anda d e o o he mean (s.e.m.) and significance o eed e ficiency and pe iod e ec s a e shown. No eed e ficiency by pe iod in e ac ion was s a is ically significan (P>0.05) and hese a e no included in he able. A: P, ace a e- o-p opiona e a io; CP, c ude p o ein; N, ni ogen; OM, o ganic ma e ; VFA, ola ile a y acids. Mean alues wi hin a ow ollowed by di e en le e s di e significan ly (a P=0.05) Pa ame e Feed e ficiency Pe iod s.e.m. P- alue HE LE GRO FIN Feed e ficiency Pe iod In ake DM (kg/day) 7.52 7.64 5.88b 9.28a 0.235 0.699 <0.001 Concen a e DM 6.40 6.59 5.28b 7.72a 0.191 0.479 <0.001 Fo age DM 1.09 0.93 0.60b 1.42a 0.137 0.397 <0.001 Appa en diges ibili y coe ficien s (%) DM 75.16a 70.82b 72.88 73.09 0.955 0.002 0.875 OM 76.15a 71.94b 73.84 74.25 0.995 0.003 0.760 CP 71.53a 68.52b 69.64 70.41 1.020 0.043 0.579 Ruminal e men a ion pa ame e s pH 6.93 6.97 7.16a 6.75b 0.080 0.710 <0.001 Ammonia-N (mg/L) 12.36 19.27 12.03 19.59 3.618 0.181 0.095 VFA (mmol/L) 70.87 74.41 68.22 77.05 6.102 0.676 0.183 VFA (%) Ace a e 49.63 51.42 49.32 51.73 1.532 0.403 0.093 P opiona e 37.70 35.88 38.89a 34.69b 1.867 0.487 0.015 Bu y a e 8.18 8.38 7.70b 8.86a 0.425 0.737 0.025 B anched-chain VFA 1.93 1.99 1.66b 2.27a 0.128 0.724 <0.001 Ra io A:P 1.42 1.58 1.30b 1.69a 0.148 0.431 0.017 DAnimal P oduc ion Science S. Cos a-Rou a e al. diges ibili y coe ficien s o DM, OM and CP han did hei LE coun e pa s. Da a on uminal e men a ion pa ame e s (Table 2)showed no di e ences in uminal pH be ween HE and LE bulls. Ammonia-N concen a ion was low and a iable among animals; he e o e, no s a is ical di e ences be ween FE ca ego ies we e ound. Al hough he o al VFA concen a ion emained una ec ed by FE, nume ical di e ences we e ound in mola p opo ions o he main VFA; HE animals had a lowe p opo ion o ace a e and a highe p opo ion o p opiona e han did LE ones. Bu y a e p opo ion inc eased wi h ime in he case o LE animals (7.46% s 9.30% o GRO and FIN pe iods, espec i ely; P= 0.047) whe eas i emained equal o HE animals (7.95% s 8.42% o GRO and FIN, espec i ely; P= 0.908). In con as , b anched-chain VFA p opo ion (isobu y a e and iso ale a e) inc eased wi h ime only in he case o HE bulls (1.55% s 2.31% o GRO and FIN, espec i ely; P=0.006). Mic obial da ase ea u es Sequencing p ocedu e yielded an a e age o (mean s.e.m.) 19862 2215 sequences pe sample, esul ing in 973259 sequences in he whole s udy. In o al, 787 OTUs we e ob ained a he 98% sequence-simila i y cu -o le els, wi h 114 5 as he mean numbe o OTUs pe sample. Good’s co e age alue was 99.69 0.03%, sugges ing ha mo e han 99% o bac e ial and a chaeal phylo ypes we e iden ified. The unclassified a e o OTUs a genus le el was 0.75 0.09%. Sha ed OTUs by all indi iduals in each FE ca ego y and pe iod we e deemed o be co e bac e ial/a chaeal communi ies. Co e communi y ga he ed 69.90 2.94% o analysed sequences and was composed o fi e OTUs, namely, P e o ella uminicola,unclassified P e o ella (bo h ep esen ing mo e han 84% o sha ed sequences), unclassified Rosebu ia, Sha pea azabuensis and unclassified Me hanob e ibac e . Mic obial communi y biodi e si y Alpha biodi e si y (Table 3) was ound o be simila among bulls di e ing in hei FE; howe e , Shannon and Simpson index alues inc eased wi h ime only in LE animals (Shannon index 1.51 s 2.13, P= 0.005; Simpson index 0.53 s 0.70, P= 0.020, o GRO and FIN, espec i ely). Be a biodi e si y is g aphically ep esen ed in Fig. 1,as well as he e ec s o explana o y a iables included in he model, namely, FE, pe iod, and bo h MW and ADG coe ficien s in he RFI model. Samples a e clea ly clus e ed by pe iod and FE, wi h he e ec s o MW and ADG coe ficien s being less g aphically e iden . Adonis es esul s confi med he o eseen di e ences in uminal mic obio a composi ion when compa ing sampling pe iods (GRO s FIN, P<0.001), FE ca ego ies (HE s LE, P= 0.022) and MW coe ficien alues (P= 0.021), bu no in he case o ADG coe ficien alues (P= 0.276). S a is ical di e ences in abundance o gene a be ween FE ca ego ies (HE s LE) could no be de ec ed by ALDEx analysis, ega dless o he sampling pe iod (Fig. S1, a ailable as Supplemen a y Ma e ial o his pape ). Mic obial ne wo k Mic obial ne wo ks we e buil o es in e ac ions among bac e ial and a chaeal gene a (Fig. 2). Deg ee o in e ac ion Table 3. Ruminal mic obial abiodi e si y Ob ained in in ensi ely ea ed bulls in he ollowing wo pe iods: g owing (GRO: 159 days oldand 225 kg bodyweigh ) andfinishing(FIN:266daysoldand434kg bodyweigh ).Residual eedin ake wasmodelled oclassi yanimalsin o wo ca ego ies o eed e ficiency, namely, high e ficiency (HE, n = 12) and low e ficiency (LE, n = 13). S anda d e o o he mean (s.e.m.) and significance o eed e ficiency and pe iod e ec s a e shown. No eed e ficiency by pe iod in e ac ion was s a is ically significan (P>0.05) and hese a e no included in he able. Mean alues wi hin a ow ollowed by di e en le e s di e significan ly (a P= 0.05) Pa ame e Feed e ficiency Pe iod s.e.m. P- alue HE LE GRO FIN Feed e ficiency Pe iod Shannon index 1.76 1.82 1.53b 2.05a 0.149 0.760 <0.001 Simpson index 0.61 0.61 0.54b 0.68a 0.050 0.942 0.002 Richness 101.10 107.69 96.87b 111.92a 5.824 0.408 0.072 2 1 0 –1 –2 –2 0 2 CCA1 MW PERIOD HE_GRO LE_GRO HE_FIN LE_FIN ADG FEED EFFICIENCY CCA2 Fig. 1. G aphical ep esen a ion o canonical co espondence analysis (CCA) on bac e ial and a chaeal ope a ional axonomic uni s (OTUs) in uminal fluid, ob ained in in ensi ely ea ed bulls in he ollowing wo pe iods: g owing (GRO: 159 days old and 225 kg bodyweigh ) and finishing (FIN: 266 days old and 434 kg bodyweigh ). Residual eed in ake was modelled o classi y animals in o wo ca ego ies o eed e ficiency, namely, high e ficiency (HE) and low e ficiency (LE). The analysis included eed e ficiency, pe iod, and bo h a e age daily gain (ADG) and me abolic weigh (MW) coe ficien s as explana o y a iables. Ruminal mic obio a and eed e ficiency in bulls Animal P oduc ion Science E was s udied h ough he numbe o gene a (nodes) ha es ablished significan in e ac ions (edges) wi h o he gene a, as well as he numbe o in e ac ions es ablished pe node (node deg ee). Du ing he g owing pe iod, HE bulls had simila numbe o nodes aking pa in he mic obial ne wo k as did LE bulls (26 in HE s 24 in LE), bu highe numbe o edges(57inHE s28inLE)andahighe a e agenode deg ee (4.38 in HE s 2.33 in LE). Du ing he finishing pe iod, mic obial ne wo k a chi ec u e changed; HE bulls con inued o ha e mo e co ela ing nodes han did LE bulls (30 in HE s 21 in LE), bu LE animals d as ically inc eased hei numbe o edges (53 in HE s 59 in LE) and he node deg ee (3.53 in HE s 5.62 in LE). Mo eo e , we in es iga ed mic obial gene a ha ac as main in o ma ion ga eways in ne wo ks in e ms o be weenness cen ali y, i.e. he ex en o which one node lies on pa hs ha connec o he nodes. Ne wo ks o HE animals p esen ed highe be weenness cen ali y han did hose o LE animals in g owing (3.88 in HE s 1.04 in LE) bu no in finishing (2.37 in HE s 2.52 in LE) pe iod. Mic obial unc ional capabili y A e p edic ing unc ional con en o uminal mic obio a, wo pa hways we e ound o be di e en ially exp essed depending on he animal FE pheno ype. In he g owing pe iod, ABC anspo e pa hway (ATP-dependen anspo o molecules ac oss cell memb ane) was mo e ac i e in HE animals han in LE animals, and, in he finishing pe iod, me hane me abolism Eubac e ium Psych obac e T eponema T eponema Acine obac e Pseudo amibac e Pseudo amibac e Lac obacillus B e ibac e ium A h obac e Dialis e Dialis e Bi idobac e ium Bi idobac e ium Bac e oides Pseudobu y i ib io P e o ella Me hanob e ibac e Jeo ga icoccus Jeo ga icoccus Tu icibac e Tu icibac e Co ynebac e ium Co ynebac e ium Co ynebac e ium B achybac e ium S aphylococcus S aphylococcus Psych obac e S aphylococcus Pseudomonas Bu y i ib io Clos idium Eubac e ium Eubac e ium Mi suokella Mi suokella Mi suokella P e o ella Olsenella Lac obacillus Bi idobac e ium Bu y i ib io Bu y i ib io Dialis e Dialis e T eponema Acidaminococcus Pseudomonas Pseudobu y i ib io Pseudobu y i ib io Selenomonas Selenomonas O senella Rosebu ia Die zia P e o ella P e o ella Aga hobac e Aga hobac e Clos idium Clos idium Aga hobac e Pep oclos idium T eponema Rosebu ia Lachnoclos idium Allisonella Sha pea Sha pea Rosebu ia Ruminococcus Ruminococcus Ruminococcus Ruminococcus Bac e oides Acine obac e Fib obac e Selenomonas Me hanob e ibac e Me hanob e ibac e Pseudo amibac e Pseudo amibac e Megasphae a Megasphae a Megasphae a Mi suokella Allisonella Sha pea Acidaminococcus Olsenella Selenomonas Facklamia Facklamia Fib obac e Fib obac e Fib obac e (a)(b) (c)(d) Candida us_Phy oplasma Candida us_Phy oplasma Ae ococcus Weissella Fig. 2. Bac e ial and a chaeal gene a ne wo k in he umen o in ensi ely ea ed a ening bulls: (a–c) GRO: 159 days old and 225 kg bodyweigh ; (b–d) FIN: 266 days old and 434 kg bodyweigh . Residual eed in ake was modelled o classi y animals in o wo ca ego ies o eed e ficiency, namely, (a, b) high e ficiency (HE) and (c, d)lowe ficiency (LE). Ne wo ks we e gene a ed on he basis o hose gene a ha es ablished significan co ela ions ( >0.60 and P<0.05). G een and ed edges indica e posi i e and nega i e co ela ions, espec i ely. Node size is p opo ional o genus abundance in he uminal fluid. FAnimal P oduc ion Science S. Cos a-Rou a e al. was down egula ed in he umen o HE indi iduals when compa ed wi h LE animals. Discussion RFI and mechanisms unde lying he a iabili y o FE Va ia ions in RFI occu due o po en ial physiological mechanisms such as diges ion, e men a ion and me abolism (He d and A hu 2009). Ou findings showed ha HE animals appa en ly diges ed mo e eed, in e ms o DM, OM and CP, han did LE ones. These esul s a e in acco dance wi h p e ious s udies in which mo e e ficien animals showed a highe nu ien diges ibili y and a smalle nu ien loss h ough was e and me hane emission (Richa dson e al.1996; Nk umah e al.2006). Negesse e al.(2017) also obse ed imp o ed appa en diges ibili y coe ficien s o DM, OM and CP in HE hei e s; hese animals exc e ed a smalle p opo ion o N h ough aeces and hei N biological alue (diges ible N a io) was highe han ha o less e ficien hei e s, sugges ing ha CP diges ion and me abolism may be enhanced in HE animals. In compa ison, de Assis Lage e al.(2019)didno find di e ences in diges ibili y coe ficien s o such nu ien s bu hey epo ed a endency o HE hei e s o be e diges e he ex ac ac ion. Vola ile a y acids a e p oduc s o umen mic obial e men a ion o ca bohyd a es, cons i u ing he main ene gy sou ce o uminan s (Be gman 1990). Al hough di e ences be ween HE and LE bulls did no each significance o any uminal e men a ion pa ame e , nume ical alues indica ed ha LE animals had a e men a ion pa e n o ien ed owa ds he p oduc ion o highe mola p opo ion o ace ic and less p opionic acids han did HE animals, wi h he consequen e ec on ace a e- o-p opiona e a io. These obse ed di e ences in umen e men a ion pa e n may be playing a ole in he bulls’FE pheno ype, since me abolic hyd ogen p oduced in he fi s s ep o ace ic acid pa hway is la e aken up by me hanogens, inc easing ene gy loss h ough gas emissions (Unge eld 2020). Du ing he g owing pe iod, mola p opo ions o he majo VFA we e simila o hose obse ed by Yus e e al.(2020)in bee hei e s ed a simila ad libi um concen a e plus s aw die . Howe e , du ing he finishing pe iod, he highe amoun o o al VFA, concomi an wi h significan ly lowe uminal pH, we e oo ed in he inc eased DM in ake and, consequen ly, in he highe ex en o e men a ion p ocess. Howe e , p opionic acid showed a di e en end and i was highe in younge animals. He nandez-U dane a e al.(1976) epo ed ha o age- o-concen a e a io a ec s he mola p opo ions o VFA, which o high-concen a e die s changes owa ds dec eased ace ic and inc eased p opionic acid; he e o e, in ou expe imen , he lowe o age- o-concen a e a io du ing he g owing (11%) han in he finishing pe iod (18%) can explain he obse ed dec eased p opo ion o p opionic acid wi h ime. RFI and uminal mic obial communi y In he p esen s udy, Illumina sequencing echnology was used o analyse bac e ial and a chaeal composi ion, biodi e si y, connec ance and unc ional capabili y wi hin he umen o in ensi ely ea ed bulls di e ing in hei FE. A nega i e co ela ion be ween uminal mic obial a biodi e si y and FE has been p e iously desc ibed in milking cows (Shaba e al.2016), sugges ing ha e ficien mic obio as a e less complex bu mo e specialised in p o iding highe concen a ions o ele an ou pu me aboli es ha can be used o mee hos ’s ene gy equi emen s. In a simila manne , ou esul s showed ha mic obial adi e si y alues significan ly inc eased wi h ime only in LE bulls. Mic obial di e si y is also posi i ely co ela ed wi h communi y s abili y and obus ness, as bo h di e en ial esponse o a iable condi ions and unc ional edundancy o species a e enhanced (McCann 2000; Moya and Fe e 2016). Thus, i seems easonable o hypo hesise ha a biodi e si y o he mic obio a has posi i e and nega i e coexis ing e ec s on ecosys em obus ness and eed u ilisa ion e ficiency, espec i ely. The ac ha LE bulls conside ably inc eased hei gene a ne wo k connec ance wi h ime, while HE bulls kep i cons an o e en diminished i , suppo s he hypo hesis ha LE animals’ uminal mic obio a could be mo e obus and ha e an enhanced abili y o cope wi h possible dis u bances (Dunne e al.2002). E en hough bdi e si y ep esen a ion showed clea clus e ing o bulls’mic obial communi y, no s a is ical di e ences in he abundance o main gene a could be ound be ween FE ca ego ies. Conside ing ha some s udies ha e had success in epo ing a ela ionship be ween ce ain mic obial axa and animal’sFE(McCanne al.2014;Mye e al.2015; Pe ea e al.2017;Delgadoe al.2019), we conside ha he ollowing ac o s could hinde de ec ion o such ela ionship: (1) he e can be subs an ial animal- o-animal a ia ion in he umen mic obial communi y, hus equi ing a g ea e numbe o animals o be able o obse e a significan associa ion be ween mic obial axa and FE (B ulc e al.2009; Weime e al.2010), and (2) he lack o di e ences obse ed be ween FE ca ego ies a he main- gene a le el may indica e ha he impo an a ia ion in mic obial communi ies lies a a fine esolu ion (e.g. a species le el o low-abundance gene a). Ki elmann e al.(2014) desc ibed he exis ence o h ee uminal mic obial communi ies linked o di e en me hane yields in sheep; umino ype H was cha ac e ised by he highes me hane emissions and ha bou ed he highe abundance o species belonging o Ruminococcus, o he Ruminococcaceae, Lachnospi aceae, Ca abac e iaceae, Cop ococcus, o he Clos idiales, P e o ella, o he Bac e oidales and Alphap o eobac e ia. In a ecen s udy in sheep, Ghanba i Maman e al.(2020) also iden ified ce ain genes om Lachnospi aceae, Ruminococcus,Bu y i ib io and Selenomonas axa ha can ha e significan e ec s on me hane p oduc ion pa hway. In acco dance wi h hese s udies, ou co- abundance analysis showed ha ce ain gene a p e iously ela ed o high me hane emission (e.g. Me hanob e ibac e , Rosebu ia,Aga hobac e ,Bu y i ib io,Pseudobu y i ib io, Ruminococcus,Selenomonas) ei he we e mo e cen al o e ol ed o be mo e cen al in LE animal ne wo ks du ing he ansi ion om g owing o finishing pe iods (Tables S2, Table S3, a ailable as Supplemen a y Ma e ial o his pape ), which could a leas pa ially cause hei lowe FE. Ruminal mic obio a and eed e ficiency in bulls Animal P oduc ion Science G Recen s udies ha e highligh ed a possible ela ionship be ween mic obial me abolic unc ions and he animal’sFE, bu he na u e o such ela ionship is s ill unclea . Li e al. (2016) obse ed ha HE ca le had a mo e ac i e me abolism o nucleo ides, as well as o a ious ene gy-gene a ing molecules (e.g. p opanoa e, glyoxyla e and dica boxyla e, s a ch and suc ose), hypo hesising ha such inc eased me abolic ac i i y could enhance eed diges ion and p o ide he hos wi h mo e nu ien s. Li e al.(2016) and Elolimy e al. (2020) also epo ed ha umen mic obio a o he mos e ficien ca le was mo e ac i e in cell p oli e a ion and su i abili y, inducing cellula g ow h and inc easing ole ance o i al in ec ion; likewise, ou esul s showed enhanced cell memb ane anspo unc ions in HE g owing animals. Finally, he obse ed dec ease o me hane me abolism ac i i y in HE finishing bulls (Shaba e al. 2016) suppo s he p e ious idea ha high and low me hane emi e s can ha e a simila abundance o uminal me hanogens bu di e en ial exp ession and ansc ip ion o me hanogenesis pa hway genes (Shi e al.2014). Conclusions The explo a ion o he ela ionship be ween umen mic obial communi y and hos FE showed inc eased nu ien diges ibili y in HE animals. Alpha biodi e si y and gene a ne wo k connec ance inc eased wi h ime in LE bulls, highligh ing a possible ade-o be ween FE and uminal mic obio a obus ness. Mo eo e , ce ain gene a ha ha e p e iously been ela ed o high me hane emission we e mo e cen al in LE animals’gene a ne wo ks. Ou esul s ha e p o ided e idence ha he umen mic obio a could be one o he biological ac o s associa ed wi h a ia ion in ca le FE. Conflic s o in e es The au ho s decla e no conflic s o in e es . Acknowledgemen s Special hanks a e ex ended o J. R. Be olín Pa dos o hei labo a o y assis ance. This s udy is a pa o GenTORE p ojec and ecei ed unding om he Eu opean Union H2020 p og am unde g an ag eemen no727213 and Ins i u o Nacional de In es igación y Tecnología Ag a ia y Alimen a ia (RTA-14-038-C02). S. Cos a- Rou a is ecipien o a esea ch aining g an om Spanish Minis y o Science and Inno a ion (FPU 2016/03761). A. R. Se adj con ac is financed by GenTORE p ojec . Re e ences A hu JPF, He d RM (2008) Residual eed in ake in bee ca le. Re is a B asilei a de Zoo ecnia 37, 269–279. doi:10.1590/S1516-3598 2008001300031 A hu PF, A che JA, Johns on DJ, He d RM, Richa dson EC, Pa nell PF (2001) Gene ic and pheno ypic a iance and co a iance componen s o eed in ake, eed e ficiency, and o he pos weaning ai s in Angus ca le. Jou nal o Animal Science 79,2805–2811. doi:10.2527/ 2001.79112805x Be gman EN (1990) Ene gy con ibu ions o ola ile a y acids om he gas oin es inal ac in a ious species. Physiological Re iews 70, 567–590. doi:10.1152/phys e .1990.70.2.567 B ulc JM, An onopoulos DA, Be g Mille ME, Wilson MK, Yanna ell AC, DinsdaleEA,Edwa dsRE,F ankED,Eme sonJB,WacklinP,Cou inho PM, Hen issa B, Nelson KE, Whi e BA (2009) Gene-cen ic me agenomics o he fibe -adhe en bo ine umen mic obiome e eals o age specific glycoside hyd olases. P oceedings o he Na ional Academy o Sciences o he Uni ed S a es o Ame ica 106, 1948–1953. doi:10.1073/pnas.0806191105 Calle ML (2019) S a is ical analysis o me agenomics da a. Genomics & In o ma ics 17, e6. doi:10.5808/GI.2019.17.1.e6 Cos a-Rou a S, Balcells J, de la Fuen e G, Mo a-Gil J, Llanes N, Villalba D (2020) Nu ien u iliza ion e ficiency, uminal e men a ion and mic obial communi y in Hols ein bulls ed concen a e-based die s wi h di e en o age sou ce. Animal Feed Science and Technology 269, 114662. doi:10.1016/j.ani eedsci.2020.114662 de Assis Lage CF, Ges ei a Coelho S, Diniz Ne o H do C, Rocha Malacco VM, Pacheco Rod igues JP, Sac amen o JP, Sama ini Machado F, Ribei o Pe ei a LG, Ribei o Tomich T, Magalhães Campos M (2019) Rela ionship be ween eed e ficiency indexes and pe o mance, body measu emen s, diges ibili y, ene gy pa i ioning, and ni ogen pa i ioning in p e-weaning dai y hei e s. PLoS One 14, e0223368. doi:10.1371/jou nal.pone.0223368 Delgado B, Bach A, Guasch I, González C, Elcoso G, P yce JE, Gonzalez- Recio O (2019) Whole umen me agenome sequencing allows classi ying and p edic ing eed e ficiency and in ake le els in ca le. Scien ific Repo s 9, 11. doi:10.1038/s41598-018-36673-w Dunne JA, Williams RJ, Ma inez ND (2002) Ne wo k s uc u e and biodi e si y loss in ood webs: obus ness inc eases wi h connec ance. Ecology Le e s 5, 558–567. doi:10.1046/j.1461-0248.2002.00354.x Elolimy A, Alha hi A, Zeineldin M, Pa ys C, Loo JJ (2020) Residual eed in ake di e gence du ing he p eweaning pe iod is associa ed wi h unique hindgu mic obiome and me abolome p ofiles in neona al Hols ein hei e cal es. Jou nal o Animal Science and Bio echnology 11, 13. doi:10.1186/s40104-019-0406-x Fe nandesAD,MacklaimJM,LinnTG,ReidG, Gloo GB (2013)ANOVA- like di e en ial exp ession (ALDEx) analysis o mixed popula ion RNA-seq. PLoS One 8, e67019. doi:10.1371/jou nal.pone.0067019 F iedman J, Alm EJ (2012) In e ing co ela ion ne wo ks om genomic su ey da a. PLoS Compu a ional Biology 8, e1002687. doi:10.1371/ jou nal.pcbi.1002687 Ghanba i Maman L, Palizban F, Fallah A anaki F, Elmi Ghiasi N, A iaeenejad S, Gha a i MR, Hosseini Salekdeh G, Ka ousi K (2020) Co-abundance analysis e eals hidden playe s associa ed wi h high me hane yield pheno ype in sheep umen mic obiome. Scien ific Repo s 10, 4995. doi:10.1038/s41598-020-61942-y Gloo GB, Macklaim JM, Pawlowsky-Glahn V, Egozcue JJ (2017) Mic obiome da ase s a e composi ional: and his is no op ional. F on ie s in Mic obiology 8, 2224. doi:10.3389/ micb.2017.02224 He d RM, A hu PF (2009) Physiological basis o esidual eed in ake. Jou nal o Animal Science 87, E64–E71. doi:10.2527/jas.2008-1345 He nandez-U dane a A, Coppock CE, McDowell RE, Gianola D, Smi h NE (1976) Changes in o age-concen a e a io o comple e eeds o dai y cows. Jou nal o Dai y Science 59, 695–707. doi:10.3168/jds. S0022-0302(76)84260-9 Hun ing on G (1990) Ene gy me abolism in he diges i e ac and li e o ca le: influence o physiological s a e and nu i ion. Rep oduc ion, Nu i ion, De elopmen 30,35–47. doi:10.1051/ nd:19900103 Ki elmann S, Pina es-Pa iño CS, Seedo H, Ki k MR, Ganesh S, McEwan JC, Janssen PH (2014) Two di e en bac e ial communi y ypes a e linked wi h he low-me hane emission ai in sheep. PLoS One 9, e103171. doi:10.1371/jou nal.pone.0103171 LiF,ZhouM,OminskiK,GuanLL(2016)Does he umenmic obiomeplay a ole in eed e ficiency o bee ca le? Jou nal o Animal Science 94, 44–48. doi:10.2527/jas.2016-0524 HAnimal P oduc ion Science S. Cos a-Rou a e al. McCann KS (2000) The di e si y–s abili y deba e. Na u e 405, 228–233. doi:10.1038/35012234 McCann JC, Wiley LM, Fo bes TD, Rouque e FM, Tedeschi LO (2014) Rela ionship be ween he umen mic obiome and esidual eed in ake- e ficiency o B ahman bulls s ocked on be mudag ass pas u es. PLoS One 9, e91864. doi:10.1371/jou nal.pone.0091864 Moya A, Fe e M (2016) Func ional edundancy-induced s abili y o gu mic obio a subjec ed o dis u bance. T ends in Mic obiology 24, 402–413. doi:10.1016/j. im.2016.02.002 Mye PR, Smi h TPL, Wells JE, Kuehn LA, F ee ly HC (2015) Rumen mic obiome om s ee s di e ing in eed e ficiency. PLoS One 10, e0129174. doi:10.1371/jou nal.pone.0129174 Negesse T, Da C, Kundu SS (2017) Residual eed in ake, diges ibili y o nu ien s and e ficiency o wa e u iliza ions in Mu ah bu alo hei e s. Jou nalo Dai y,Ve e ina y&AnimalResea ch5,74–80.doi:10.15406/ jd a .2017.05.00138 Nk umah JD, Okine EK, Ma hison GW, Schmid K, Li C, Basa ab JA, P ice MA, Wang Z, Moo e SS (2006) Rela ionships o eedlo eed e ficiency, pe o mance, and eeding beha io wi h me abolic a e, me hane p oduc ion, and ene gy pa i ioning in bee ca le. Jou nal o Animal Science 84, 145–153. doi:10.2527/2006.841145x Owens FN, Hanson CF (1992) Ex e nal and in e nal ma ke s o app aising si e and ex en o diges ion in uminan s. Jou nal o Dai y Science 75,2605–2617. doi:10.3168/jds.S0022-0302(92) 78023-0 Pe ea K, Pe z K, Oli o SK, Williams A, Lachman M, Ishaq SL, Thomson J, Yeoman CJ (2017) Feed e ficiency pheno ypes in lambs in ol e changes in uminal, colonic, and small-in es ine-loca ed mic obio a. Jou nal o Animal Science 95,2585–2592. doi:10.2527/jas2016.1222 R Co e Team (2020) ‘R: a language and en i onmen o s a is ical compu ing.’(R Founda ion o S a is ical Compu ing: Vienna, Aus ia) Richa dson EC, He d RM, A hu PF, W igh J, Xu G, Dibley K, Oddy VH (1996) Possible physiological indica o s o ne eed con e sion e ficiency in bee ca le. P oceedings o he Aus alian Socie y o Animal P oduc ion 21, 103–106. Sa ie o D, Be y DP, F iggens NC (2014) Towa ds an imp o ed es ima ion o he biological componen s o esidual eed in ake in g owing ca le. Jou nal o Animal Science 92,4 67–476. doi:10.2527/ jas.2013-6894 Schenkel FS, Mille SP, Wil on JW (2004) Gene ic pa ame e s and b eed di e ences o eed e ficiency, g ow h, and body composi ion ai s o young bee bulls. Canadian Jou nal o Animal Science 84,177–185. doi:10.4141/A03-085 Shaba SKB, Sasson G, Do on-Faigenboim A, Du man T, Yaacoby S, Be g Mille ME, Whi e BA, Sh e ze N, Miz ahi I (2016) Specific mic obiome-dependen mechanisms unde lie he ene gy ha es e ficiency o uminan s. The ISME Jou nal 10,2958–2972. doi:10.1038/ismej.2016.62 Shi W, Moon CD, Leahy SC, Kang D, F oula J, Ki elmann S, Fan C, Deu sch S, Gagic D, Seedo H, Kelly WJ, A ua R, Sang C, Soni P, Li D, Pina es-Pa iño CS, McEwan JC, Janssen PH, Chen F, Visel A, Wang Z, A wood GT, Rubin EM (2014) Me hane yield pheno ypes linked o di e en ial gene exp ession in he sheep umen mic obiome. Genome Resea ch 24,1517–1525. doi:10.1101/g .168245.113 Tsilimig as MCB, Fodo AA (2016) Composi ional da a analysis o he mic obiome: undamen als, ools, and challenges. Annals o Epidemiology 26,330–335. doi:10.1016/j.annepidem.2016.03.002 Unge eld EM (2020) Me abolic hyd ogen flows in umen e men a ion: p inciples and possibili ies o in e en ions. F on ie s in Mic obiology 11, 589. doi:10.3389/ micb.2020.00589 Weime PJ, S e enson DM, Man o ani HC, Man SLC (2010) Hos specifici y o he uminal bac e ial communi y in he dai y cow ollowing nea - o al exchange o uminal con en s. Jou nal o Dai y Science 93,5902–5912. doi:10.3168/jds.2010-3500 Woloszynek S, Mell JC, Zhao Z, Simpson G, O’Conno MP, Rosen GL (2019) Explo ing hema ic s uc u e and p edic ed unc ionali y o 16S RNA amplicon da a. PLoS One 14, e0219235. doi:10.1371/jou nal. pone.0219235 Yus e S, Amanzouga ene Z, de Vega A, Fonde ila M, Blanco M, Casasús I (2020) E ec o p eweaning die on pe o mance, blood me aboli es and umen e men a ion a ound weaning in cal es o wo bee b eeds. Animal P oduc ion Science 60,1018–1027. doi:10.1071/AN19152 Handling edi o : Di Maybe y Ruminal mic obio a and eed e ficiency in bulls Animal P oduc ion Science I www.publish.csi o.au/jou nals/an