Conditions stimulating neutral detergent fiber degradation by dosing branched-chain volatile fatty acids. III: Relation with solid passage rate and pH on prokaryotic fatty acid profile and community in continuous culture

Cerrato-Sánchez et al., 2008

Roman-Garcia Y.
Denton B.L.
Mitchell K.E.
Lee C.
Socha M.T.
Firkins J.L.

Conditions stimulating neutral detergent fiber degradation by dosing branched-chain volatile fatty acids. I: Comparison with branched-chain amino acids and forage source in ruminal batch cultures.

,b) explained the role of branched-chain VFA (BCVFA) to stimulate fiber digestibility because they are required growth factors for cellulolytic bacteria; these cellulolytics also degrade hemicellulose. This benefit from adding exogenous BCVFA was expected to interact with pH and the solids passage rate (k_p), which can be assessed independently in continuous culture. These same factors were assumed to influence lipolysis and biohydrogenation (

Qiu et al., 2004

Qiu X.
Eastridge M.L.
Griswold K.E.
Firkins J.L.

Effects of substrate, passage rate, and pH in continuous culture on flows of conjugated linoleic acid and trans C18:1.

;

Fuentes et al., 2009

Fuentes M.C.
Calsamiglia S.
Cardozo P.W.
Vlaeminck B.

Effect of pH and level of concentrate in the diet on the production of biohydrogenation intermediates in a dual-flow continuous culture.

). That the magnitude of time with depressed pH was more inhibitory than the actual pH per se (

Cerrato-Sánchez M.
Calsamiglia S.
Ferret A.

Effect of the magnitude of the decrease of rumen pH on rumen fermentation in a dual-flow continuous culture system.

) suggests a shift in bacterial physiology to overcome this stress. Response to stress has been suggested to affect the bacterial fatty acid (FA) profile, including those branched-chain fatty acids (BCFA) that were elongated from BCVFA primers (

Vlaeminck B.
Fievez V.
Cabrita A.R.J.
Fonseca A.J.M.
Dewhurst R.J.

Factors affecting odd- and branched-chain fatty acids in milk: A review.

).

The BCFA, particularly anteiso-FA, have lower melting points than unbranched odd-chain and even-chain SFA (

Parsons J.B.
Rock C.O.

Bacterial lipids: Metabolism and membrane homeostasis.

). Anaerobic bacteria have limited ability to desaturate FA or incorporate UFA and therefore rely on a mix of straight-, odd-, and branched-chain FA to maintain membrane fluidity. According to

Giotis et al., 2007

Giotis E.S.
McDowell D.A.
Blair I.S.
Wilkinson B.J.

Role of branched-chain fatty acids in pH stress tolerance in Listeria monocytogenes..

, a more acidic environment makes the bacterial membrane more fluid, so theoretically less BCVFA would be elongated to BCFA with decreasing pH; on the other hand, higher pH makes the bacterial membrane more rigid, and bacteria respond by incorporating more BCFA into their membrane to maintain fluidity.

Vlaeminck B.
Fievez V.
Cabrita A.R.J.
Fonseca A.J.M.
Dewhurst R.J.

Factors affecting odd- and branched-chain fatty acids in milk: A review.

explained that the relative concentration of primers for FA synthesis or enzyme affinity for those primers could influence BCFA profile, but they concluded that bacterial species would have a much larger influence on FA composition than environmental factors. However, ruminal bacteria can incorporate exogenous FA into their membranes in response to availability of exogenous FA and substrate (

Moon and Anderson, 2001

Moon M.
Anderson K.L.

Eubacterium cellulosolvens alters its membrane protein, lipoprotein, and fatty acid composition in response to growth on cellulose.

), with palmitate decreasing fluidity and oleic acid increasing fluidity. Few other studies with ruminal bacteria have evaluated FA uptake, particularly when compared with changing BCVFA availability. Some exogenous FA are stored intracellularly (

Bas et al., 2003

Bas P.
Archimède H.
Rouzeau A.
Sauvant D.

Fatty acid composition of mixed-rumen bacteria: Effect of concentration and type of forage.

), and BCFA are still needed even when high amounts of unsaturated fat are fed (

Alves et al., 2013

Alves S.P.
Santos-Silva J.
Cabrita A.R.J.
Fonseca A.J.M.
Bessa R.J.B.

Detailed dimethylacetal and fatty acid composition of rumen content from lambs fed lucerne or concentrate supplemented with soybean oil.

). Thus, BCVFA elongation to BCFA is likely much more important to maintain membrane fluidity than incorporation of UFA such as oleate.

In ruminants, increasing forage in the diets increased BCFA in bacteria passing to the duodenum (

Bas et al., 2003

Bas P.
Archimède H.
Rouzeau A.
Sauvant D.

Fatty acid composition of mixed-rumen bacteria: Effect of concentration and type of forage.

;

Vlaeminck et al., 2006b

Vlaeminck B.
Fievez V.
Demeyer D.
Dewhurst R.J.

Effect of forage:concentrate ratio on fatty acid composition of rumen bacteria isolated from ruminal and duodenal digesta.

), but few studies have distinguished a change in FA composition in rumen bacterial populations resulting from changing forage:concentrate independent of a change in ruminal pH. For example,

Wang et al., 2018

Wang C.
Liu Q.
Guo G.
Huo W.J.
Pei C.X.
Zhang S.L.
Wang H.

Effects of concentrate-to-forage ratios and 2-methylbutyrate supplementation on ruminal fermentation, bacteria abundance and urinary excretion of purine derivatives in Chinese Simmental steers.

noted that supplementation of 2-methylbutyrate to steers increased NDF degradation and certain fibrolytic bacterial rRNA gene copies more in moderate-concentrate diets than in high-concentrate diets, but starch:fiber and pH effects were indistinguishable. Before 16S rRNA gene profiling techniques, bacterial systematics often relied on profiles of specific FA (

Miyagawa E.

Cellular fatty acid and fatty aldehyde composition of rumen bacteria.

;

Kopečný et al., 2003

Kopečný J.
Zorec M.
Mrazek J.
Kobayashi Y.
Marinsek-Logar R.

Butyrivibrio hungatei sp. nov. and Pseudobutyrivibrio xylanivorans sp. nov., butyrate-producing bacteria from the rumen.

), whereas bacterial FA composition also likely shifts under differing ruminal conditions. Because BCVFA are precursors for bacterial BCFA in membranes and branched-chain AA (BCAA) in protein, BCVFA uptake likely increases with increasing bacterial growth rate, which was expected to increase with higher k_p (

Roman-Garcia Y.
Mitchell K.E.
Denton B.L.
Lee C.
Socha M.
Wenner B.A.
Firkins J.L.

Conditions stimulating NDF degradation by dosing branched chain volatile fatty acids. II: Relation of branched-chain volatile fatty acids supplementation with solid passage rate and pH on NDF degradation and microbial function in continuous culture.

).

The objective of this study was to test for interactions in how decreasing pH and increasing k_p affect the FA profile and recovery of dosed ¹³C-BCVFA in bacterial FA and relate these factors to the prokaryotic community by using 16S rRNA gene sequencing. We hypothesized that increasing growth rate with high k_p should increase the percentage of BCVFA transfer into lipids of faster-growing bacteria than at low k_p. Also, we hypothesized that lower pH would decrease the elongation of BCVFA, especially 2-methylbutyrate, into anteiso-FA that increase membrane fluidity. Decreasing pH was projected to decrease the relative abundance of fibrolytic bacteria such as Fibrobacter and Ruminococcus spp., but BCVFA dosing was expected to help lessen this decrease, particularly when increasing k_p would increase growth pressure on bacteria.

MATERIALS AND METHODS

Continuous Culture Operation

Experimental Design and Treatments

In this study, 8 dual-flow continuous culture fermentors were used in an incomplete block design with 4 experimental periods. The 12-d period comprised 8 d of adjustment followed by 4 d of sampling. We tested a 2 × 2 × 2 factorial arrangement of 8 treatment combinations. Treatment factors were (1) control or BCVFA supplementation, (2) high pH (ranging from 6.3 to 6.8) or low pH (ranging from 5.7 to 6.2), and (3) low (2.5%/h) or high (5.0%/h) solids k_p. All conditions and the rationale for the BCVFA treatment including 2 mmol/d each of isovalerate, isobutyrate, and 2-methylbutyrate are described in the companion study (

Roman-Garcia Y.
Mitchell K.E.
Denton B.L.
Lee C.
Socha M.
Wenner B.A.
Firkins J.L.

Conditions stimulating NDF degradation by dosing branched chain volatile fatty acids. II: Relation of branched-chain volatile fatty acids supplementation with solid passage rate and pH on NDF degradation and microbial function in continuous culture.

). Briefly, adding H₃PO₄ or NaOH as necessary, the low pH buffer was maintained between 6.3 and 6.4, but the high pH buffer was maintained between 6.7 and 6.8 before infusion into the fermentors, resulting in minimum pH >5.7 and >6.3, respectively. Buffer input and fluid efflux were adjusted to manipulate solids k_p while maintaining fluid k_p at 10%/h.

Doses of ¹³C-Labeled BCVFA

Due to cost and availability, we chose [methyl-¹³C]-2-methylbutyrate above other potential labeled products. A pilot study dosed 0, 10, or 20 mg/d of ¹³C-labeled isoleucine (which is deaminated and decarboxylated to 2-methylbutyrate) followed by subsequent analysis of ¹³C enrichment in FA to achieve ≥0.04 atom percentage excess in bacterial FA using techniques described subsequently (data not shown). On a molar equivalency basis, 40 mg/d of [methyl-¹³C]-2-methylbutyrate (racemic mix) was deemed appropriate. The same dose was scaled based on carbons labeled (approximately 5 mg/d of ¹³C for each BCVFA): 10.3 mg/d of [2,3,4-¹³C, methyl-¹³C]-isovalerate and 11.8 mg/d of [2,3-¹³C, methyl-¹³C]-isobutyrate. All isotopes were 99% enriched and purchased from Cambridge Isotope Laboratories. Dosing of BCVFA started on d 2 (1 mmol of each of the 3 BCVFA at each of 2 daily feedings), allowing for 7 d of adaptation to the treatment. On d 5, ¹³C-labeled BCVFA were dosed for 4 d, which was enough for >2 turnovers of the slower solids k_p before sampling.

Bacterial Collection and Analysis for FA and Prokaryotic Community

On sampling days (d 8 to 11), whole effluent was collected at 12 h after feeding (0800 and 2000 h) on ice for each fermentor and mixed; a 30-mL sample was taken, frozen immediately, and stored at −80°C for subsequent DNA extraction. Following thawing, DNA was extracted and purified (

Wenner et al., 2020

Wenner B.A.
Wagner B.K.
St-Pierre N.R.
Yu Z.T.
Firkins J.L.

Inhibition of methanogenesis by nitrate, with or without defaunation, in continuous culture.

) before sequencing of the 16S rRNA gene. Computational procedures were as described by

Lee et al., 2021

Lee C.
Copelin J.E.
Park T.
Mitchell K.E.
Firkins J.L.
Socha M.T.
Luchini D.

Effects of diet fermentability and supplementation of 2-hydroxy-4-(methylthio)-butanoic acid and isoacids on milk fat depression: 2. Ruminal fermentation, fatty acid, and bacterial community structure.

. Briefly, the Illumina MiSeq 2 × 300 paired-end protocol was used, followed by filtering, annotation, and cluster analysis using amplicon sequencing variants (ASV) clustered at 99% similarity. The α and β diversity indices were derived as described by

Lee et al., 2021

Lee C.
Copelin J.E.
Park T.
Mitchell K.E.
Firkins J.L.
Socha M.T.
Luchini D.

Effects of diet fermentability and supplementation of 2-hydroxy-4-(methylthio)-butanoic acid and isoacids on milk fat depression: 2. Ruminal fermentation, fatty acid, and bacterial community structure.

except for Bray-Curtis dissimilarity (

Wenner et al., 2020

Wenner B.A.
Wagner B.K.
St-Pierre N.R.
Yu Z.T.
Firkins J.L.

Inhibition of methanogenesis by nitrate, with or without defaunation, in continuous culture.

).

Another 2 subsamples of 250 mL each were taken from effluents (above) for bacterial pellet collection; the pH was brought down to 2 (monitored using a pH meter) by adding 5 N HCl dropwise and then stored at 4°C until particle-associated bacteria were harvested in 2 separate samples per sampling day, as described previously (

Sukhija and Palmquist, 1988

Roman-Garcia Y.
Denton B.L.
Mitchell K.E.
Lee C.
Socha M.T.
Firkins J.L.

Conditions stimulating neutral detergent fiber degradation by dosing branched-chain volatile fatty acids. I: Comparison with branched-chain amino acids and forage source in ruminal batch cultures.

). One of the pellets was extracted and methylated according to

Sukhija P.S.
Palmquist D.L.

Rapid method for determination of total fatty acid content and composition of feedstuffs and feces.

, with the modification of

Jenkins, 2010

Jenkins T.C.

Technical note: Common analytical errors yielding inaccurate results during analysis of fatty acids in feed and digesta samples.

to better distinguish biohydrogenation intermediates. The FA methyl esters were analyzed as described previously (

Roman-Garcia Y.
Denton B.L.
Mitchell K.E.
Lee C.
Socha M.T.
Firkins J.L.

Conditions stimulating neutral detergent fiber degradation by dosing branched-chain volatile fatty acids. I: Comparison with branched-chain amino acids and forage source in ruminal batch cultures.

) using GC, with identification using the same bacterial standard mix and the same quantification and derivation of ¹³C enrichment using ion-ratio MS. The other pellet was used as reported in the companion paper (

Roman-Garcia Y.
Mitchell K.E.
Denton B.L.
Lee C.
Socha M.
Wenner B.A.
Firkins J.L.

Conditions stimulating NDF degradation by dosing branched chain volatile fatty acids. II: Relation of branched-chain volatile fatty acids supplementation with solid passage rate and pH on NDF degradation and microbial function in continuous culture.

). Background samples were taken before dosing to correct enrichments to atom percentage excess. The effluent flow of ¹³C (corrected for background) was divided by the daily total dose of ¹³C in the BCVFA treatments. The ¹³C recovered in individual and total FA was derived as outlined by

Roman-Garcia Y.
Denton B.L.
Mitchell K.E.
Lee C.
Socha M.T.
Firkins J.L.

Conditions stimulating neutral detergent fiber degradation by dosing branched-chain volatile fatty acids. I: Comparison with branched-chain amino acids and forage source in ruminal batch cultures.

, except that daily bacterial FA flow was used instead of FA recovered in bacterial pellets.

Statistical Analysis

Data were analyzed using PROC MIXED in SAS 9.4 (SAS Institute Inc.) according to this model:

Yijskl = µ + Bi + Hj + Ks + (B × H)ij + (H × K)js + (B × K)is + (B × H × K)ijs + Fl + pk + εijskl,

where Yijskl is the dependent variable; μ is the overall population mean; Bi is the fixed effect of BCVFA dose (i = control, BCVFA); Hj is the fixed effect of pH treatment (j = high, low); Ks is the fixed effect of k_p (s = high, low); (B × H)ij, (H × K)js, (B × K)is, and (B × H × K)ijs are the respective interactions; Fl is the fixed effect of fermentor (l = 1 to 8); pk is the random effect of kth period (k = 1 to 4); and εijskl is the random error. Differences were declared at P ≤ 0.05 and trends at P ≤ 0.10.

RESULTS

There were no interactions among treatments for total bacterial FA flow or FA percentages in bacterial OM or DM (Table 1). However, the total FA tended (P ≤ 0.09) to be 11 and 12% higher in bacterial OM and DM, respectively, for the main effect of dosing BCVFA. Increasing k_p decreased (P ≤ 0.02) bacterial FA flow and FA percentage of DM by 12 and 14%, respectively.

Table 1Fatty acid (FA) profile of bacteria in continuous cultures that were dosed without or with branched-chain VFA (BCVFA) and maintained at either high or low pH and either low or high solids passage rate (k_p)

¹

Treatments are High pH: range 6.8 to 6.3, Low pH: range 6.2 to 5.7, Low kp: 2.5%/h, High kp: 5.0%/h.

Item	Control				BCVFA				SEM	P-value ² P-values reported for the main effects of BCVFA dose, pH, kp, and their 2-way interactions. All 3-way interactions were P > 0.10. NS = P > 0.20.
	High pH		Low pH		High pH		Low pH
	Low k_p	High k_p	Low k_p	High k_p	Low k_p	High k_p	Low k_p	High k_p		BCVFA	pH	BCVFA × pH	k_p	BCVFA × k_p	pH × k_p
Total FA, g/d	0.793	0.630	0.805	0.633	0.722	0.679	0.850	0.768	0.102	NS	NS	NS	0.01	NS	NS
FA, % of OM	7.04	6.27	6.86	5.77	6.76	7.41	7.85	6.81	0.85	0.09	NS	NS	0.14	NS	NS
FA, % of DM	3.57	3.19	3.68	2.90	3.82	3.75	3.96	3.37	0.51	0.07	NS	NS	0.02	NS	NS
FA, g/100 g of total FA
11:0	0.0398	0.0275	0.0621	0.0653	0.0281	0.0205	0.0548	0.0674	0.0130	NS	<0.01	NS	NS	NS	0.15
10:0 2-OH	0.00382	0.00010	0.00010	0.00237	0.00022	0.00269	0.00057	0.00188	0.00220	NS	NS	NS	NS	NS	NS
12:0	0.730	0.725	0.773	0.919	0.622	0.584	0.818	0.822	0.142	0.17	<0.01	NS	NS	NS	NS
iso-13:0	0.454	0.462	0.391	0.536	0.404	0.368	0.374	0.431	0.106	0.07	NS	NS	0.19	NS	0.09
anteiso-13:0	0.0867	0.0921	0.0802	0.0994	0.0854	0.0784	0.0753	0.0851	0.0154	NS	NS	NS	NS	NS	NS
13:0	0.323	0.311	0.391	0.399	0.353	0.330	0.335	0.384	0.038	NS	0.01	0.08	NS	NS	0.18
12:0 2-OH	0.0569	0.0510	0.0423	0.0443	0.0612	0.0397	0.0402	0.0457	0.0144	NS	0.07	NS	NS	NS	0.08
12:0 3-OH	0.0496	0.0548	0.0800	0.0824	0.0454	0.0438	0.0837	0.0958	0.0210	NS	<0.01	NS	NS	NS	NS
iso-14:0	1.20	1.23	1.08	1.00	1.49	1.45	0.84	1.41	0.20	0.08	<0.01	NS	0.16	0.11	0.16
anteiso-14:0 ³ Anteiso-FA are expected to be odd chain, so these are likely fatty aldehydes and are not included in sums of anteiso-FA or BCFA.	0.0548	0.0570	0.0415	0.0382	0.0558	0.0614	0.0392	0.0534	0.0145	NS	0.01	NS	NS	NS	NS
14:0	4.42	4.43	4.05	4.74	4.15	4.21	4.27	4.13	0.45	0.20	NS	NS	NS	NS	NS
iso-15:0	3.49	4.30	2.34	3.36	3.21	3.57	2.65	3.11	0.68	NS	<0.01	NS	0.01	NS	NS
anteiso-15:0	5.15	5.24	5.61	6.41	5.08	4.83	5.36	6.45	0.78	NS	<0.01	NS	0.06	NS	0.03
15:0	4.20	3.84	5.72	4.99	5.41	4.69	4.35	5.92	0.53	0.16	0.01	0.02	NS	0.09	0.08
14:0 2-OH	0.806	0.984	1.093	0.760	0.971	0.947	0.732	0.691	0.256	NS	NS	0.16	NS	NS	0.18
14:0 3-OH	0.659	0.726	0.690	0.802	0.576	0.631	0.811	0.725	0.188	NS	0.05	NS	NS	NS	NS
iso-16:0	1.15	1.10	1.14	0.93	1.31	1.26	0.93	1.28	0.14	0.08	0.03	NS	NS	0.03	NS
anteiso-16:0 ³ Anteiso-FA are expected to be odd chain, so these are likely fatty aldehydes and are not included in sums of anteiso-FA or BCFA.	0.136	0.134	0.156	0.132	0.125	0.160	0.125	0.174	0.057	NS	NS	NS	NS	0.20	NS
16:1 cis-9	0.464	0.492	0.328	0.350	0.440	0.625	0.245	0.381	0.071	NS	<0.01	NS	<0.01	0.04	NS
16:0	28.0	28.2	30.3	31.1	29.6	28.4	28.8	30.5	3.2	NS	0.06	NS	NS	NS	NS
iso-17:0	0.454	0.411	0.463	0.533	0.413	0.404	0.497	0.516	0.071	NS	0.01	NS	NS	NS	NS
anteiso-17:0	0.865	0.747	0.847	0.853	0.758	0.786	0.756	0.909	0.105	NS	0.20	NS	NS	0.09	0.12
17:0	0.761	0.731	0.870	0.886	0.765	0.637	0.848	0.914	0.104	NS	<0.01	NS	NS	NS	0.11
16:0 2-OH	0.0216	0.0161	0.0950	0.2964	0.0923	0.0534	0.0263	0.0872	0.0903	NS	0.20	0.13	NS	NS	NS
18:2 cis-9,cis-12	1.69	2.21	1.64	2.21	2.55	2.08	1.69	1.85	0.66	NS	NS	NS	NS	NS	NS
18:1 cis-9–12 ⁴ Sum of all 18:1 FA with cis bonds at carbons 9 to 12, and sum of all 18:1 FA with trans bonds at carbons 10 to 12.	4.74	4.45	3.78	2.58	4.57	6.29	2.69	3.49	0.92	NS	<0.01	NS	NS	0.09	NS
18:1 trans-10–12 ⁴ Sum of all 18:1 FA with cis bonds at carbons 9 to 12, and sum of all 18:1 FA with trans bonds at carbons 10 to 12.	8.26	9.96	7.62	6.30	7.85	10.1	7.92	5.68	1.14	NS	<0.01	NS	NS	NS	0.02
18:0	30.6	28.3	29.8	26.0	28.6	29.1	31.7	27.9	3.1	NS	NS	NS	0.12	NS	NS
iso-19:0	0.0616	0.0670	0.0643	0.0426	0.1162	0.0639	0.0851	0.0673	0.0442	NS	NS	NS	NS	NS	NS
19:0 cis-9 ⁵ 19:0 cis-9 = cis-9,10-methylene octanoate.	0.00977	0.00000	0.02685	0.03600	0.00603	0.00000	0.00229	0.00000	0.01648	0.18	0.21	0.18	NS	NS	NS
19:0	0.0478	0.0694	0.0649	0.0371	0.0607	0.0392	0.0426	0.0412	0.0181	NS	NS	NS	NS	NS	NS
20:0	0.646	0.567	0.563	0.594	0.761	0.490	0.629	0.555	0.137	NS	NS	NS	0.06	0.19	0.15
Total unknown	5.46	5.47	6.36	6.62	5.39	5.56	5.80	6.53	1.10	NS	NS	NS	NS	NS	NS
Total OCFA ⁶ OCFA = sum of all odd straight-chain FA, ECFA = sum of all even straight-chain FA.	5.36	4.98	7.11	6.38	6.61	5.72	5.63	7.33	0.53	NS	<0.01	0.04	NS	0.13	0.06
Total ECFA ⁶ OCFA = sum of all odd straight-chain FA, ECFA = sum of all even straight-chain FA.	64.4	62.2	65.4	63.2	63.6	62.7	66.2	63.9	1.9	NS	0.13	NS	0.05	NS	NS
Total BCFA ⁷ Sum of all branched-chain FA that include iso-FA and anteiso-FA.	12.9	13.6	12.0	13.8	12.9	12.8	11.6	14.3	1.0	NS	NS	NS	0.02	NS	0.09
Total anteiso-FA	6.10	6.07	6.59	7.37	5.93	5.69	6.19	7.45	0.86	NS	<0.01	NS	0.07	NS	0.03
Total iso-FA ⁸ Total iso-FA include all isomers, but even iso-FA are restricted to iso-14:0 + iso-16:0, and odd iso-FA are restricted to iso-15:0 + iso-17:0.	6.82	7.56	5.48	6.41	6.94	7.12	5.38	6.81	0.87	NS	<0.01	NS	0.02	NS	NS
Total even iso-FA ⁸ Total iso-FA include all isomers, but even iso-FA are restricted to iso-14:0 + iso-16:0, and odd iso-FA are restricted to iso-15:0 + iso-17:0.	2.35	2.32	2.22	1.94	2.79	2.71	1.77	2.68	0.24	0.06	0.01	NS	NS	0.05	0.18
Total odd iso-FA ⁸ Total iso-FA include all isomers, but even iso-FA are restricted to iso-14:0 + iso-16:0, and odd iso-FA are restricted to iso-15:0 + iso-17:0.	3.95	4.71	2.80	3.90	3.62	3.97	3.15	3.63	0.73	NS	0.01	NS	0.01	NS	NS

1 Treatments are High pH: range 6.8 to 6.3, Low pH: range 6.2 to 5.7, Low k_p: 2.5%/h, High k_p: 5.0%/h.

2 P-values reported for the main effects of BCVFA dose, pH, k_p, and their 2-way interactions. All 3-way interactions were P > 0.10. NS = P > 0.20.

3 Anteiso-FA are expected to be odd chain, so these are likely fatty aldehydes and are not included in sums of anteiso-FA or BCFA.

4 Sum of all 18:1 FA with cis bonds at carbons 9 to 12, and sum of all 18:1 FA with trans bonds at carbons 10 to 12.

5 19:0 cis-9 = cis-9,10-methylene octanoate.

6 OCFA = sum of all odd straight-chain FA, ECFA = sum of all even straight-chain FA.

7 Sum of all branched-chain FA that include iso-FA and anteiso-FA.

8 Total iso-FA include all isomers, but even iso-FA are restricted to iso-14:0 + iso-16:0, and odd iso-FA are restricted to iso-15:0 + iso-17:0.

Open table in a new tab

Dosing BCVFA tended to decrease (P = 0.07) iso-13:0 percentage of total FA and tended to increase (P = 0.08) iso-14:0 and iso-16:0 percentages of total FA (Table 1; main effect comparison). Decreasing pH decreased (P < 0.04) percentage of iso-14:0, iso-15:0, and iso-16:0 but increased (P = 0.01) anteiso-15:0 and iso-17:0. Increasing k_p increased (P = 0.01) iso-15:0 and tended (P = 0.06) to increase anteiso-15:0, although this effect was most pronounced for anteiso-15:0 with low pH (pH × k_p interaction, P = 0.03). There was an interaction (P = 0.03) between BCVFA and k_p such that increasing k_p decreased iso-16:0 without dosed BCVFA but increased iso-16:0 percentage when BCVFA were dosed. Adding BCVFA tended to increase anteiso-17:0 (P = 0.09) and increased total even-chain iso-FA (P = 0.05) when combined with high k_p (BCVFA × k_p interactions).

Decreasing pH decreased (P < 0.01) 16:1 cis-9 and the undistinguishable FA annotated as 18:1 cis-9 to cis-12 and also trans-10 to trans-12. These overlapping peaks could not be distinguished reliably because of our short column length. The main effect of low pH increased (P < 0.01) 15:0, 17:0, and total odd-chain SFA compared with high pH. The percentage of 16:0 also tended (P = 0.06) to increase with decreasing pH. High k_p increased (P = 0.02) total BCFA; high k_p increased (P = 0.02) iso-FA but only tended to increase (P = 0.07) anteiso-FA. Low pH decreased (P < 0.01) total iso-FA, odd iso-FA, and even iso-FA but increased anteiso-FA (P < 0.01).

No isotope was administered to control continuous cultures because they were not dosed with BCVFA. For the BCVFA treatments, there were no effects of pH or k_p on recovery of label in total bacterial C or FA flowing to the effluent (Table 2). Low pH decreased (P < 0.01) ¹³C recovery of anteiso-13:0 from dosed ¹³C-labeled BCVFA. Decreasing pH decreased (P ≤ 0.01) recovery of ¹³C in iso-14:0, iso-16:0, and total even-chain iso-FA, whereas recovery in iso-17:0 increased (P = 0.01) without any change in total odd iso-FA. Low pH also decreased recovery of ¹³C in FA eluting at the same time as our anteiso-14:0 standards; however, as will be discussed later, this identification is questionable. Increasing k_p tended to increase (P = 0.06) recovery of label in iso-14:0.

Table 2Recovery of total ¹³C or ¹³C in bacterial fatty acids (FA) after dosing ¹³C-enriched branched-chain VFA in continuous cultures maintained at either high or low pH and with either low or high solids passage rate (k_p)

¹

Treatments are High pH: range 6.8 to 6.3, Low pH: range 6.2 to 5.7, Low kp: 2.5%/h, High kp: 5.0%/h. Control treatment was not dosed with 13C.

Item	High pH		Low pH		SEM	P-value ² P-values reported for the main effects of pH, kp, and their 2-way interactions.
Item	Low k_p	High k_p	Low k_p	High k_p	SEM	pH	k_p	pH × k_p
Total ¹³C recovered in bacterial C flow, μg/mg of ¹³C dosed	99.9	88.8	106.4	96.7	1.0	NS	NS	NS
Total ¹³C recovered in bacterial FA flow, μg/mg of ¹³C dosed
Total FA	2.93	2.72	2.20	2.13	0.93	NS	NS	NS
11:0	0.00036	0.00117	0.00028	0.00044	0.00082	NS	NS	NS
12:0	0.00121	0.00126	0.00102	0.00140	0.00047	NS	NS	NS
iso-13:0	0.0293	0.0282	0.0190	0.0331	0.0138	NS	NS	NS
anteiso-13:0	0.00457	0.00651	0.00186	0.00331	0.00135	0.06	NS	NS
13:0	0.00084	0.00106	0.00154	0.00271	0.00099	NS	NS	NS
12:0 2-OH	0.00231	0.00201	0.00128	0.00202	0.00085	NS	NS	NS
12:0 3-OH	0.00045	0.00020	0.00010	0.00086	0.00058	NS	NS	NS
iso-14:0	0.1026	0.1881	0.0453	0.0743	0.0290	0.01	0.06	NS
anteiso-14:0 ³ Anteiso-FA are expected to be odd chain, so these are likely fatty aldehydes and are not included in sums of anteiso-FA or branched-chain FA.	0.00505	0.00499	0.00085	0.00346	0.00214	NS	NS	NS
14:0	0.119	0.054	0.106	0.111	0.045	NS	NS	NS
iso-15:0	0.218	0.153	0.171	0.251	0.065	NS	NS	0.20
anteiso-15:0	0.436	0.399	0.306	0.414	0.096	NS	NS	NS
15:0	0.106	0.090	0.019	0.158	0.073	NS	NS	NS
14:0 2-OH	0.0520	0.0540	0.0134	0.0353	0.0133	0.05	NS	NS
14:0 3-OH	0.0052	0.0103	0.0192	0.0097	0.0088	NS	NS	NS
iso-16:0	0.157	0.166	0.080	0.083	0.030	0.02	NS	NS
anteiso-16:0 ³ Anteiso-FA are expected to be odd chain, so these are likely fatty aldehydes and are not included in sums of anteiso-FA or branched-chain FA.	0.0065	0.0073	0.0032	0.0071	0.0047	NS	NS	NS
16:1 cis-9	0.0101	0.0055	0.0116	0.0107	0.0042	NS	NS	NS
16:0	1.048	0.774	0.851	0.410	0.530	NS	NS	NS
iso-17:0	0.0129	0.0108	0.0401	0.0427	0.0098	0.01	NS	NS
anteiso-17:0	0.0517	0.0587	0.0104	0.0466	0.0146	0.11	0.15	NS
17:0	0.0172	0.0077	0.0145	0.0170	0.0114	NS	NS	NS
16:0 2-OH	0.00096	0.00017	0.00043	0.00003	0.00035	NS	0.12	NS
18:2 cis-9, cis-12	0.0214	0.0086	0.0281	0.0144	0.0128	NS	NS	NS
18:1 cis-9–12 ⁴ Sum of all 18:1 FA with cis bonds at carbons 9 to 12, and sum of all 18:1 FA with trans bonds at carbons 10 to 12.	0.0197	0.0403	0.0080	0.0557	0.0314	NS	NS	NS
18:1 trans-10–12 ⁴ Sum of all 18:1 FA with cis bonds at carbons 9 to 12, and sum of all 18:1 FA with trans bonds at carbons 10 to 12.	0.185	0.212	0.093	0.052	0.125	NS	NS	NS
18:0	0.301	0.281	0.365	0.230	0.322	NS	NS	NS
19:0	0.00008	0.00028	0.00003	0.00010	0.00012	NS	NS	NS
20:0	0.00057	0.00234	0.00229	0.00271	0.00139	NS	NS	NS
Total unknown FA	0.175	0.161	0.162	0.145	0.041	NS	NS	NS
Total OCFA ⁵ OCFA = sum of all odd straight-chain FA; ECFA = sum of all even straight-chain FA.	0.131	0.109	0.031	0.173	0.075	NS	NS	NS
Total ECFA ⁵ OCFA = sum of all odd straight-chain FA; ECFA = sum of all even straight-chain FA.	1.63	1.35	1.25	0.56	0.85	NS	NS	NS
Total BCFA ⁶ Sum of all branched-chain FA that include iso-FA and anteiso-FA.	1.019	1.055	0.629	0.939	0.155	0.15	NS	NS
Total anteiso-FA	0.491	0.462	0.321	0.464	0.099	NS	NS	NS
Total iso-FA ⁷ Total iso-FA include all isomers, but even iso-FA are restricted to iso-14:0 + iso-16:0, and odd iso-FA are restricted to iso-15:0 + iso-17:0.	0.521	0.563	0.341	0.481	0.102	NS	NS	NS
Total even iso-FA ⁷ Total iso-FA include all isomers, but even iso-FA are restricted to iso-14:0 + iso-16:0, and odd iso-FA are restricted to iso-15:0 + iso-17:0.	0.258	0.353	0.123	0.161	0.045	<0.01	0.15	NS
Total odd iso-FA ⁷ Total iso-FA include all isomers, but even iso-FA are restricted to iso-14:0 + iso-16:0, and odd iso-FA are restricted to iso-15:0 + iso-17:0.	0.253	0.163	0.253	0.338	0.088	NS	NS	0.18

1 Treatments are High pH: range 6.8 to 6.3, Low pH: range 6.2 to 5.7, Low k_p: 2.5%/h, High k_p: 5.0%/h. Control treatment was not dosed with ¹³C.

2 P-values reported for the main effects of pH, k_p, and their 2-way interactions.

3 Anteiso-FA are expected to be odd chain, so these are likely fatty aldehydes and are not included in sums of anteiso-FA or branched-chain FA.

4 Sum of all 18:1 FA with cis bonds at carbons 9 to 12, and sum of all 18:1 FA with trans bonds at carbons 10 to 12.

5 OCFA = sum of all odd straight-chain FA; ECFA = sum of all even straight-chain FA.

6 Sum of all branched-chain FA that include iso-FA and anteiso-FA.

7 Total iso-FA include all isomers, but even iso-FA are restricted to iso-14:0 + iso-16:0, and odd iso-FA are restricted to iso-15:0 + iso-17:0.

Open table in a new tab

Diversity indices were largely unaffected except for a trend (P = 0.09) for increasing k_p to increase Simpson's index (Table 3). The main effect of k_p or its interaction with other factors was much less common than shifts in abundance of ASV assigned at the taxonomic rank of genus resulting from the pH treatment. Nonmetric multidimensional scaling plots of the Bray-Curtis distance matrix documented that pH (Figure 1A) and period (Figure 1B) had important effects (both P < 0.01) on prokaryotic community structure.

Table 3Bacterial diversity and relative abundance (% of total amplicon sequence variants, ASV) of genera within phylum Bacteroidetes in continuous cultures that were dosed without or with branched-chain VFA (BCVFA) and maintained at either high or low pH and either low or high solids passage rate (k_p)

¹

Treatments are High pH: range 6.8 to 6.3, Low pH: range 6.2 to 5.7, Low kp: 2.5%/h, High kp: 5.0%/h.

Item	Control				BCVFA				SEM	P-value ² P-values reported for the main effects of BCVFA dose, pH, kp, and their 2-way interactions. All 3-way interactions were P > 0.10. NS = P > 0.20.
	High pH		Low pH		High pH		Low pH
	Low k_p	High k_p	Low k_p	High k_p	Low k_p	High k_p	Low k_p	High k_p		BCVFA	pH	BCVFA × pH	k_p	BCVFA × k_p	pH × k_p
Diversity indices
ASV, n	1,448	1,192	1,267	1,247	1,263	1,301	1,534	1,210	150	NS	NS	NS	0.18	NS	NS
Chao1	1,535	1,244	1,340	1,319	1,366	1,385	1,668	1,277	174	NS	NS	NS	0.16	NS	NS
Evenness	0.770	0.806	0.801	0.793	0.775	0.781	0.801	0.793	0.022	NS	NS	NS	0.61	NS	NS
Faith's phylogenetic diversity	64.8	57.9	59.7	57.9	60.6	61.4	64.7	57.4	3.5	NS	NS	NS	0.11	NS	NS
Shannon's index	8.07	8.23	8.25	8.13	7.94	8.08	8.48	8.11	0.293	NS	NS	NS	0.77	NS	NS
Simpson's index	0.972	0.987	0.974	0.979	0.977	0.982	0.977	0.986	0.008	NS	NS	NS	0.09	NS	NS
Good's coverage	0.995	0.997	0.996	0.996	0.996	0.996	0.994	0.996	0.001	NS	NS	NS	0.18	NS	NS
Bacteroidetes	53.2	49.2	50.8	47.1	49.8	47.4	46.1	47.1	2.8	NS	NS	NS	NS	NS	NS
Prevotella	13.8	13.0	25.5	26.6	16.2	15.7	25.8	26.6	3.2	NS	<0.01	NS	NS	NS	NS
UN(Prevotellaceae) ³ UN = unclassified genus (lowest classified taxon within parentheses).	1.092	0.923	1.216	0.710	1.305	0.827	0.232	0.710	0.322	NS	0.10	0.15	NS	NS	NS
UN (Paraprevotellaceae)	1.86	1.52	1.12	0.98	1.92	2.02	1.32	0.98	0.27	NS	<0.01	NS	NS	NS	NS
YRC22 (Paraprevotellaceae)	1.45	2.16	1.37	1.32	1.29	1.71	1.41	1.32	0.29	NS	0.10	NS	NS	NS	0.08
CF231 (Paraprevotellaceae)	0.664	0.616	1.565	0.802	0.799	0.782	1.356	0.802	0.242	NS	0.01	NS	0.11	NS	0.17
UN(BS11 Gut Group)	16.06	9.40	2.33	3.42	10.35	9.24	0.95	3.42	1.91	0.18	<0.01	NS	NS	0.16	0.03
UN(S24–7)	4.45	7.13	6.37	3.54	5.10	6.51	4.87	3.54	1.15	NS	NS	NS	NS	NS	0.07
BF311(Bacteroidaceae)	0.665	0.535	0.334	0.202	0.641	0.552	0.183	0.202	0.114	NS	<0.01	NS	0.19	NS	NS
UN(RF16)	0.551	0.325	0.343	0.214	0.369	0.285	0.339	0.214	0.100	NS	0.06	NS	0.01	NS	NS
UN(Bacteroidales)	9.1	10.2	9.1	7.5	8.7	7.7	7.4	7.5	1.9	NS	NS	NS	NS	NS	NS
UN(Bacteroidales)	3.19	3.00	1.28	1.55	2.68	1.72	1.97	1.55	0.52	NS	0.01	0.08	NS	NS	NS

1 Treatments are High pH: range 6.8 to 6.3, Low pH: range 6.2 to 5.7, Low k_p: 2.5%/h, High k_p: 5.0%/h.

2 P-values reported for the main effects of BCVFA dose, pH, k_p, and their 2-way interactions. All 3-way interactions were P > 0.10. NS = P > 0.20.

3 UN = unclassified genus (lowest classified taxon within parentheses).

Open table in a new tab

Figure thumbnail gr1 — Figure 1Bray-Curtis plots (nonmetric multidimensional scaling, NMDS axes) of a Bray-Curtis distance matrix for 16S rRNA gene relative abundance (rarefied data) from continuous cultures dosed with branched-chain VFA (P > 0.20) and maintained with high or low particulate passage rate (P > 0.20) and high or low pH (P < 0.01; panel A). The effect of incubation period (1–4) was P < 0.01 (panel B). There were no interactions (P > 0.20) among treatments.
View Large Image
Figure Viewer
Download Hi-res image
Download (PPT)

Phylum Bacteroidetes had the highest relative percentage of ASV and had no treatment differences (Table 3). All genera within this phylum above the abundance threshold were in order Bacteroidales, and all but 2 of these (S24–7 and an unclassified genus in family Paraprevotellaceae) were affected (P ≤ 0.10) by the main effect of pH treatment. Prevotella increased markedly, as did the CF231 genus in family Paraprevotellaceae, with decreasing pH. In contrast, the relative abundance of ASV assigned to a genus in family BS11 decreased with decreasing pH. Some pH × k_p interactions were detected (P ≤ 0.08), but the greatest shift was noted when the combination of low pH plus low k_p decreased the unclassified genus in BS11 while increasing the unclassified genus in family S24–7. The unclassified genus in RF16 was decreased (P = 0.01 for main effect) with increasing k_p.

The main effect of decreasing pH tended to decrease (P = 0.06) the relative ASV abundance assigned to phylum Firmicutes by an average of 14% (Table 4). Relative abundance of ASV assigned to Pseudobutyrivibrio, one unclassified genus in Lachnospiraceae, Oscillospora, both unclassified genera in Ruminococcaceae, and Mogibacterium decreased (P ≤ 0.09); all were in order Clostridiales. Conversely, relative abundance of ASV assigned to Shuttleworthia, Moryella, and Anaerostipes (all Clostridiales) increased (P ≤ 0.03) with decreasing pH.

Table 4Relative abundance (% of total amplicon sequence variants) of bacterial genera within the phylum Firmicutes in continuous cultures that were dosed without or with branched-chain VFA (BCVFA) and maintained at either high or low pH and either low or high solids passage rate (k_p)

¹

Treatments are High pH: range 6.8 to 6.3, Low pH: range 6.2 to 5.7, Low kp: 2.5%/h, High kp: 5.0%/h.

Item	Control				BCVFA				SEM	P-value ² P-values reported for the main effects of BCVFA dose, pH, kp, and their 2-way interactions. Only Shuttleworthia had a tendency for a 3-way interaction (P = 0.10). All other 3-way interactions were P > 0.10. NS = P > 0.20.
	High pH		Low pH		High pH		Low pH
	Low k_p	High k_p	Low k_p	High k_p	Low k_p	High k_p	Low k_p	High k_p		BCVFA	pH	BCVFA × pH	k_p	BCVFA × k_p	pH × k_p
Firmicutes	34.5	34.5	26.8	33.1	35.3	32.7	28.8	28.5	3.5	NS	0.06	NS	NS	NS	NS
Butyrivibrio	2.65	3.29	2.00	2.68	2.46	2.38	1.94	3.11	0.49	NS	NS	NS	0.06	NS	NS
Pseudobutyrivibrio	1.66	2.06	0.49	1.47	1.56	1.81	0.86	0.94	0.28	NS	<0.01	NS	0.03	0.19	NS
Coprococcus	1.236	1.469	0.683	1.207	0.691	0.824	0.931	0.893	0.293	0.16	NS	0.17	NS	NS	NS
Shuttleworthia	0.409	0.423	0.767	0.898	0.209	0.436	1.445	0.596	0.259	NS	0.01	NS	NS	NS	0.18
Moryella	0.106	0.077	0.144	0.146	0.110	0.108	0.213	0.198	0.035	0.15	0.01	NS	NS	NS	NS
Anaerostipes	0.058	0.052	0.302	0.159	0.089	0.118	0.122	0.159	0.063	NS	0.03	0.14	NS	NS	NS
UN (Lachnospiraceae) ³ UN = unclassified genus (lowest classified taxon within parentheses).	3.79	5.85	1.85	2.77	3.49	5.52	0.95	1.46	1.54	NS	0.02	NS	NS	NS	NS
UN (Lachnospiraceae)	1.30	1.74	1.37	1.42	1.93	1.41	1.18	1.15	0.34	NS	NS	NS	NS	NS	NS
Ruminococcus	2.25	2.85	2.75	2.41	2.28	2.80	3.12	2.60	0.45	NS	NS	NS	NS	NS	0.12
Oscillospira	0.611	0.428	0.207	0.188	0.313	0.465	0.136	0.159	0.089	0.20	<0.01	NS	NS	0.17	NS
UN(Ruminococcaceae)	2.60	2.19	2.28	2.12	2.72	2.30	1.64	1.79	0.43	NS	0.09	NS	NS	NS	NS
UN(Ruminococcaceae)	0.918	0.681	0.611	0.542	0.756	0.655	0.492	0.558	0.150	NS	0.04	NS	NS	NS	NS
Succiniclasticum	1.564	1.291	2.061	2.511	1.651	1.778	1.996	1.797	0.329	NS	0.04	0.17	NS	NS	NS
Dialister ⁴ Fermentor effect not included in the model to avoid negative values. See Statistical Analysis section in the Materials and Methods.	<0.001	<0.001	0.041	0.585	<0.001	<0.001	0.133	0.291	0.232	NS	0.10	NS	NS	NS	NS
Mitsuokella ⁴ Fermentor effect not included in the model to avoid negative values. See Statistical Analysis section in the Materials and Methods.	<0.001	<0.001	0.210	0.187	<0.001	<0.001	0.413	0.190	0.068	NS	<0.01	NS	0.18	NS	0.18
UN (Veillonellaceae)	0.080	0.056	0.388	0.340	0.076	0.071	0.485	0.277	0.057	NS	<0.01	NS	0.04	NS	0.10
UN (Veillonellaceae) ⁴ Fermentor effect not included in the model to avoid negative values. See Statistical Analysis section in the Materials and Methods.	0.013	0.000	0.142	0.287	0.037	0.028	0.214	0.247	0.075	NS	<0.01	NS	NS	NS	NS
Clostridium	0.626	0.811	0.839	0.680	0.774	0.748	1.116	0.778	0.126	NS	NS	NS	NS	NS	0.09
UN (Christensenellaceae) ⁴ Fermentor effect not included in the model to avoid negative values. See Statistical Analysis section in the Materials and Methods.	4.58	2.90	0.65	0.94	6.35	3.25	0.22	0.91	1.43	NS	NS	NS	<0.01	NS	NS
UN (Mogibacteriaceae)	0.630	0.456	0.565	0.517	0.527	0.470	0.358	0.564	0.099	NS	NS	NS	NS	0.12	0.09
Mogibacterium	0.141	0.093	0.119	0.056	0.120	0.080	0.028	0.041	0.038	NS	0.07	NS	0.16	NS	NS
UN (Clostridiales)	4.89	4.49	5.20	4.86	4.72	3.86	6.13	4.68	0.76	NS	0.16	NS	0.13	NS	NS
UN (Clostridiales)	2.97	2.80	2.59	3.59	2.96	3.01	2.65	2.90	0.37	NS	NS	NS	NS	NS	NS
Rummeliibacillus ⁴ Fermentor effect not included in the model to avoid negative values. See Statistical Analysis section in the Materials and Methods.	0.078	0.022	0.048	0.441	0.082	0.136	0.137	0.537	0.195	NS	0.12	NS	0.15	NS	0.14
Bulleidia	0.055	0.031	0.083	0.168	0.015	0.018	0.007	0.280	0.088	NS	0.10	NS	0.16	NS	0.13
Sharpea ⁴ Fermentor effect not included in the model to avoid negative values. See Statistical Analysis section in the Materials and Methods.	0.000	0.000	0.010	0.221	0.006	0.000	0.057	0.380	0.154	NS	0.12	NS	NS	NS	0.20

1 Treatments are High pH: range 6.8 to 6.3, Low pH: range 6.2 to 5.7, Low k_p: 2.5%/h, High k_p: 5.0%/h.

2 P-values reported for the main effects of BCVFA dose, pH, k_p, and their 2-way interactions. Only Shuttleworthia had a tendency for a 3-way interaction (P = 0.10). All other 3-way interactions were P > 0.10. NS = P > 0.20.

3 UN = unclassified genus (lowest classified taxon within parentheses).

4 Fermentor effect not included in the model to avoid negative values. See Statistical Analysis section in the Materials and Methods.

Open table in a new tab

Representatives of the Veillonellaceae family (including Succiniclasticum, Dialister, Mitsuokella, and 2 unclassified Veillonellaceae genera, all of which are in order Clostridiales) and Bulleidia (order Erysipelotrichales) increased with decreasing pH (P < 0.01 except Dialister and Bulleidia, which had trends of P = 0.10). Both Butyrivibrio and Pseudobutyrivibrio tended to increase (P ≤ 0.06) with increasing k_p, whereas an unclassified Veillonellaceae genus and an unclassified Christensenellaceae genus decreased (P ≤ 0.04). Trends (P ≤ 0.10) for interactions were noted when increasing k_p decreased relative abundance at low pH (an unclassified genus in Veillonellaceae and Clostridium) or high pH (unclassified genus in Mogibacteriaceae).

Fibrobacteres and its dominant genus Fibrobacter tended (P = 0.06) to increase, and Spirochaetes and its dominant genus Treponema increased (P = 0.01) with the main effect of dosing BCVFA (Table 5). The benefit from BCVFA for Fibrobacter was more prevalent at low pH (interaction, P = 0.08). The relative sequence abundance of phylum Proteobacteria increased (P = 0.01) up to 15% of total ASV with decreasing pH and high k_p. The unclassified genus in Comamonadaceae within class β-Proteobacteria decreased (P = 0.01) with decreasing pH. Desulfovibrio (class δ-Proteobacteria) increased (P = 0.01) with decreasing pH. In the λ-Proteobacteria class, genus Succinivibrio had an interaction (P = 0.03) such that increasing k_p increased its relative ASV abundance when pH was high but not when pH was low. For the other genera in this class, the unclassified Succinivibrionaceae increased (P < 0.01) with decreasing pH. However, Ruminobacter and Acinetobacter decreased (P ≤ 0.09) when pH decreased. Phylum TM7 was not affected by our treatments. Tenericutes and some of their represented genera decreased (P ≤ 0.01) in relative sequence abundance when pH decreased. Phylum Verrucomicrobia was not affected (P > 0.20) by treatment. Phylum Euryarchaeota (in domain Archaea) and its dominant genus Methanobrevibacter were also not affected (P > 0.20) by treatment.

Table 5Relative abundance (% of total amplicon sequence variants) of bacterial genera within other phyla (other than Bacteroidetes and Firmicutes) and the main archaeal phylum and genus in continuous cultures that were dosed without or with branched-chain VFA (BCVFA) and maintained at either high or low pH and either low or high solids passage rate (k_p)

¹

Treatments are High pH: range 6.8 to 6.3, Low pH: range 6.2 to 5.7, Low kp: 2.5%/h, High kp: 5.0%/h.

Item	Control				BCVFA				SEM	P-value ² P-values reported for the main effects of BCVFA dose, pH, kp, and their 2-way interactions. Only Acinetobacter had a tendency for a 3-way interaction (P = 0.10). All other 3-way interactions were P > 0.10. NS = P > 0.20.
	High pH		Low pH		High pH		Low pH
	Low k_p	High k_p	Low k_p	High k_p	Low k_p	High k_p	Low k_p	High k_p		BCVFA	pH	BCVFA × pH	k_p	BCVFA × k_p	pH × k_p
Fibrobacteres	0.252	0.328	0.156	0.229	0.282	0.335	0.501	0.356	0.082	0.06	NS	0.08	NS	NS	NS
Fibrobacter	0.252	0.328	0.156	0.229	0.282	0.335	0.501	0.356	0.082	0.06	NS	0.08	NS	NS	NS
Spirochaetes	2.90	2.40	5.70	2.50	6.50	6.22	6.08	7.03	1.45	0.01	NS	NS	NS	NS	NS
Treponema	1.90	1.87	4.71	2.19	5.73	4.98	5.87	6.16	1.28	0.01	NS	NS	NS	NS	NS
Sphaerochaeta	0.353	0.117	0.161	0.112	0.231	0.357	0.166	0.172	0.100	NS	0.14	NS	NS	0.18	NS
UN (Spirochaetes) ³ UN = unclassified genus (lowest classified taxon within parentheses).	0.558	0.401	0.794	0.145	0.499	0.879	0.023	0.673	0.485	NS	NS	NS	NS	0.09	NS
Proteobacteria	5.2	10.0	12.8	12.1	5.3	10.0	15.1	11.0	3.1	NS	0.01	NS	NS	NS	0.07
UN (Comamonadaceae) ⁴ Fermentor effect not included in the model to avoid negative values. See Statistical Analysis section in the Materials and Methods.	0.687	1.289	0.035	0.082	0.684	1.350	0.060	0.112	0.339	NS	<0.01	NS	0.14	NS	NS
Desulfovibrio	0.163	0.118	0.427	0.538	0.076	0.057	0.354	0.711	0.180	NS	<0.01	NS	NS	NS	NS
Succinivibrio	0.33	3.84	1.47	1.59	1.41	4.80	1.06	1.47	1.11	NS	0.09	NS	0.01	NS	0.03
UN (Succinivibrionaceae)	2.23	3.01	9.94	9.37	1.75	1.51	13.11	7.33	3.13	NS	<0.01	NS	NS	NS	0.33
Ruminobacter	0.708	0.468	0.032	0.098	0.317	0.853	0.176	0.519	0.263	NS	0.06	NS	NS	0.19	NS
Acinetobacter	0.201	0.329	0.152	0.031	0.271	0.285	0.070	0.322	0.121	NS	0.09	NS	NS	NS	NS
Tenericutes	1.75	1.89	1.22	0.88	1.27	2.06	1.12	0.91	0.25	NS	<0.01	NS	NS	NS	0.03
Anaeroplasma	0.651	0.540	0.285	0.165	0.440	0.561	0.044	0.128	0.126	0.20	<0.01	NS	NS	NS	NS
UN (RF39)	0.582	0.600	0.623	0.637	0.520	0.615	1.030	0.649	0.168	NS	0.19	NS	NS	NS	NS
UN (ML615J-28)	0.319	0.604	0.275	0.047	0.210	0.770	0.018	0.093	0.177	NS	0.01	NS	0.17	NS	0.06
Actinobacteria ⁴ Fermentor effect not included in the model to avoid negative values. See Statistical Analysis section in the Materials and Methods.	0.06	0.03	0.71	1.75	0.04	0.03	0.38	1.97	0.79	NS	0.03	NS	0.20	NS	0.19
Bifidobacterium ⁴ Fermentor effect not included in the model to avoid negative values. See Statistical Analysis section in the Materials and Methods.	<0.01	<0.01	0.64	1.68	<0.01	<0.01	0.29	1.81	0.76	NS	0.03	NS	0.20	NS	0.20
TM7	0.481	0.501	0.489	0.643	0.524	0.582	0.797	0.667	0.244	NS	NS	NS	NS	NS	NS
UN (F16)	0.479	0.501	0.486	0.642	0.522	0.581	0.792	0.665	0.245	NS	NS	NS	NS	NS	NS
Verrucomicrobia	0.348	0.391	0.241	0.241	0.218	0.318	0.254	0.274	0.074	NS	NS	NS	NS	NS	NS
UN (RFP12)	0.307	0.309	0.217	0.207	0.187	0.267	0.210	0.243	0.067	NS	NS	NS	NS	NS	NS
Euryarchaeota	0.169	0.094	0.150	0.150	0.212	0.142	0.191	0.181	0.055	NS	NS	NS	NS	NS	NS
Methanobrevibacter	0.144	0.085	0.145	0.132	0.183	0.122	0.169	0.159	0.051	NS	NS	NS	NS	NS	NS

1 Treatments are High pH: range 6.8 to 6.3, Low pH: range 6.2 to 5.7, Low kp: 2.5%/h, High kp: 5.0%/h.

2 P-values reported for the main effects of BCVFA dose, pH, k_p, and their 2-way interactions. Only Acinetobacter had a tendency for a 3-way interaction (P = 0.10). All other 3-way interactions were P > 0.10. NS = P > 0.20.

3 UN = unclassified genus (lowest classified taxon within parentheses).

4 Fermentor effect not included in the model to avoid negative values. See Statistical Analysis section in the Materials and Methods.

Open table in a new tab

DISCUSSION

2-Methylbutyrate plus either isobutyrate or isovalerate were needed to maintain the highest fiber degradability during a 24-h incubation in batch cultures of mixed ruminal microbes, but isobutyrate more consistently stimulated NDF degradability than isovalerate, as supported by literature from characterized cellulolytic bacterial strains (

Roman-Garcia Y.
Denton B.L.
Mitchell K.E.
Lee C.
Socha M.T.
Firkins J.L.

Conditions stimulating neutral detergent fiber degradation by dosing branched-chain volatile fatty acids. I: Comparison with branched-chain amino acids and forage source in ruminal batch cultures.

). Dosing BCVFA tended to improve NDF digestibility by about 5.3 percentage units in the companion report (

Roman-Garcia Y.
Mitchell K.E.
Denton B.L.
Lee C.
Socha M.
Wenner B.A.
Firkins J.L.

Conditions stimulating NDF degradation by dosing branched chain volatile fatty acids. II: Relation of branched-chain volatile fatty acids supplementation with solid passage rate and pH on NDF degradation and microbial function in continuous culture.

). Therein, net production of isobutyrate and isovalerate decreased whereas net production of 2-methylbutyrate increased as a response to decreasing pH. We reasoned that the differences in net BCVFA accumulating inversely reflected their elongation into bacterial membranes. However, the BCFA profile could be affected both by our treatments and also by a shift in community structure toward more acid-tolerant amylolytics, as reported in other studies (

Vlaeminck B.
Fievez V.
Cabrita A.R.J.
Fonseca A.J.M.
Dewhurst R.J.

Factors affecting odd- and branched-chain fatty acids in milk: A review.

). We are among the first to evaluate the mode of action of BCVFA using both ¹³C-BCVFA dosing and the bacterial community.

BCVFA Effects on General Bacterial Fatty Acid Metabolism

The total FA flow in bacteria averaged about 1.0 g/d (Table 1), which is about 15% of the total bacterial CP flow (

Roman-Garcia Y.
Mitchell K.E.
Denton B.L.
Lee C.
Socha M.
Wenner B.A.
Firkins J.L.

Conditions stimulating NDF degradation by dosing branched chain volatile fatty acids. II: Relation of branched-chain volatile fatty acids supplementation with solid passage rate and pH on NDF degradation and microbial function in continuous culture.

). The decreased bacterial FA flow with increasing k_p could be a result of dilution of dietary FA that adsorb to bacterial cells or small particles that contaminate harvested bacteria. If increased concentration of BCVFA (by dosing) would increase substrate-level competition as primers in the fatty acid synthetase system, then we would expect an increased representation of BCFA in the total bacterial FA. However, bacterial FA flow was not increased statistically (numerical increase of 5.6%), even though dosing BCVFA tended to increase total bacterial FA by 11% of OM while having no effect on the BCFA percentage of total FA (Table 1). We noted similar increases in FA concentration when BCAA or BCVFA were dosed to batch cultures (

Robinson and Allison, 1969

Roman-Garcia Y.
Denton B.L.
Mitchell K.E.
Lee C.
Socha M.T.
Firkins J.L.

Conditions stimulating neutral detergent fiber degradation by dosing branched-chain volatile fatty acids. I: Comparison with branched-chain amino acids and forage source in ruminal batch cultures.

).

Robinson I.M.
Allison M.J.

Isoleucine biosynthesis from 2-methylbutyric acid by anaerobic bacteria from the rumen.

reported that increasing dose of 2-methylbutyrate increased its incorporation into isoleucine in Prevotella ruminicola with respect to transfer to lipid. They noted varying recoveries of ¹⁴C from 2-methylbutyrate in protein versus lipid fractions from several strains of ruminal bacteria and one methanogen. Because there was a very large difference between Ruminococcus albus and Ruminococcus flavefaciens (88 vs. 5% label recovery in lipid) despite similar taxonomy and niche, BCVFA incorporation into protein versus lipid fractions likely depends on needs for growth relative to precursor availability and any potential shifts in the FA:AA ratio in cells under those conditions.

Although little effect of imbalance was noted among individual BCVFA/BCAA pairs (

Allison et al., 1984

Allison M.J.
Baetz A.L.
Wiegel J.

Alternative pathways for biosynthesis of leucine and other amino acids in Bacteroides ruminicola and Bacteroides fragilis..

;

Atasoglu et al., 2004

Atasoglu C.
Guliye A.Y.
Wallace R.J.

Use of stable isotopes to measure de novo synthesis and turnover of amino acid-C and -N in mixed micro-organisms from the sheep rumen in vitro..

), imbalance can occur among BCAA (

Kajikawa et al., 2005

Kajikawa H.
Mitsumori M.
Tajima K.
Kurihara M.

Short communication: Amino acids antagonistic to the amino acids inhibitory for growth rate of mixed ruminal bacteria.

). All 3 of the BCAA are synthesized from pyruvate or a derivative thereof, such as threonine as a precursor for isoleucine (

Kaiser and Heinrichs, 2018

Kaiser J.C.
Heinrichs D.E.

Branching out: Alterations in bacterial physiology and virulence due to branched-chain amino acid deprivation.

). Relatively little is known regarding the regulatory control of FA synthesis in gut anaerobic bacteria, although those facultative anaerobes studied appear to control their lipid:protein ratio under differing growth conditions (

Parsons J.B.
Rock C.O.

Bacterial lipids: Metabolism and membrane homeostasis.

). The latter authors noted that starvation of leucine and isoleucine in cultures changed expression of genes involved in FA synthesis and membrane fluidity in gram-positive Bacillus subtilis and gram-negative Escherichia coli, respectively. Thus, increased total FA concentration resulting from dosing BCVFA in our study might result from stimulating FA synthesis in some unknown way, whereas we assume that imbalance is much less likely with provision of BCVFA compared with BCAA. Moreover, substituting BCVFA for BCAA increased recovery of dose in bacterial FA harvested from mixed ruminal microbes in vitro (

Roman-Garcia Y.
Denton B.L.
Mitchell K.E.
Lee C.
Socha M.T.
Firkins J.L.

Conditions stimulating neutral detergent fiber degradation by dosing branched-chain volatile fatty acids. I: Comparison with branched-chain amino acids and forage source in ruminal batch cultures.

).

Branched-Chain FA Profile and Recovery of ¹³C from BCVFA Doses

Recovery of label in AA was not assayed in the present experiment but must have been high because we recovered about 100 μg/mg of ¹³C-BCVFA dosed (about 10%), but <3% of that total recovered ¹³C was in total FA (Table 2).

Monticello and Costilow, 1982

Roman-Garcia Y.
Denton B.L.
Mitchell K.E.
Lee C.
Socha M.T.
Firkins J.L.

Conditions stimulating neutral detergent fiber degradation by dosing branched-chain volatile fatty acids. I: Comparison with branched-chain amino acids and forage source in ruminal batch cultures.

already justified why recovery should be higher in BCAA and why a numerically low recovery of ¹³C in FA is still biologically meaningful. Although the ¹³C enrichment was much higher in BCFA than in other FA in our study (data not shown), low enrichment in more abundant FA such as 16:0 (Table 2) suggests considerable labeling of acetyl-CoA derived from dosed ¹³C-BCVFA. Clostridium sporogenes interconverted radiolabeled valine with that of leucine, and leucine conversion to valine yielded acetate in their proposed pathway (

Monticello D.J.
Costilow R.N.

Interconversion of valine and leucine by Clostridium sporogenes..

). Degradation of BCVFA-CoA (originally from BCAA) to acetyl-CoA or propionyl-CoA is documented in several nonrumen Proteobacteria (

Kazakov et al., 2009

Kazakov A.E.
Rodionov D.A.
Alm E.
Arkin A.P.
Dubchak I.
Gelfand M.S.

Comparative genomics of regulation of fatty acid and branched-chain amino acid utilization in proteobacteria.

). Any metabolism of BCVFA to acetate or acetyl-CoA would allow distribution of ¹³C from our BCVFA into many compounds, especially bacterial FA (

Kristensen, 2001

Kristensen N.B.

Rumen microbial sequestration of [2-¹³C]acetate in cattle.

). Unfortunately, deconvolution analysis of isotopomers (

Atasoglu et al., 2004

Atasoglu C.
Guliye A.Y.
Wallace R.J.

Use of stable isotopes to measure de novo synthesis and turnover of amino acid-C and -N in mixed micro-organisms from the sheep rumen in vitro..

) was not possible with isotope-ratio MS, so these findings need further corroboration.

Even-chain iso-FA percentages in total FA were affected by BCFA, pH, and BCVFA × k_p, but results were mostly due to decreasing even-chain iso-FA when BCVFA was combined with low pH at low k_p (Table 1). In contrast to generally higher total concentrations, the recovery of ¹³C in even-chain iso-FA decreased markedly with decreasing pH (Table 2). The ¹³C recovery in anteiso-13:0 decreased in the low pH treatments, but it was a low percentage of total anteiso-FA recovery, which was not affected by pH. That total anteiso-FA was only decreased by low pH when k_p was low (Table 1) suggests that the supply of 2-methylbutyrate is important in rapidly growing bacteria, regardless of pH. The lower pH shifted relative abundance of many bacteria, notably increasing Prevotella, Veillonellaceae, and Succinivibrionaceae (Tables 3 to 5). These bacteria are not known to require BCVFA, and Prevotella ruminicola, Megasphaera elsdenii (family Veillonellaceae), and Ruminobacter amylophilus (family Succinivibrionaceae) are characterized proteolytics (

Stewart C.S.
Flint H.J.
Bryant M.P.

The rumen bacteria.

). A sequenced strain of Succinivibrio dextrinosolvens was suggested to grow on BCAA (

Hailemariam et al., 2020

Hailemariam S.
Zhao S.
Wang J.

Complete genome sequencing and transcriptome analysis of nitrogen metabolism of Succinivibrio dextrinosolvens strain Z6 isolated from dairy cow rumen.

). Therefore, more BCAA from RDP (which is obviously unlabeled) might have been decarboxylated and inserted into membranes, resulting in less mixing with extracellularly dosed ¹³C-BCVFA.

There are 2 potential explanations for why recovery of ¹³C was not affected for odd-chain iso-FA but decreased with decreasing pH for even-chain iso-FA. First, many Proteobacteria degrade isovaleryl-CoA with a different enzyme complex to acetyl-CoA than that used to degrade the CoA esters of isobutyrate and 2-methylbutyrate (

Kazakov et al., 2009

Kazakov A.E.
Rodionov D.A.
Alm E.
Arkin A.P.
Dubchak I.
Gelfand M.S.

Comparative genomics of regulation of fatty acid and branched-chain amino acid utilization in proteobacteria.

). Isobutyrate might have been catabolized more than isovalerate by the increasing members of Proteobacteria in our low pH treatments. Second, isovalerate could be metabolized differently in Prevotella, which also increased with decreasing pH. Dosing isovalerate profoundly increased expression of BCAA synthetic enzymes, perhaps as a stress response, when compared with isobutyrate (2-methylbutyrate was not tested) in Prevotella bryantii B14 (

Trautmann et al., 2020

Trautmann A.
Schleicher L.
Deusch S.
Gätgens J.
Steuber J.
Seifert J.

Short-chain fatty acids modulate metabolic pathways and membrane lipids in Prevotella bryantii B₁4.

). In that study, isobutyrate increased even-chain iso-FA at the expense of anteiso-FA and linear-chain FA, whereas isovalerate had a minimal effect on FA profile. Prevotella spp. contribute substantially (often predominantly) to proteolysis in the rumen (

Hartinger et al., 2018

Hartinger T.
Gresner N.
Südekum K.H.

Does intra-ruminal nitrogen recycling waste valuable resources? A review of major players and their manipulation.

). Prevotella spp. hydrolyze protein followed by transport of small peptides for intracellular deamination, yet they regulate ammonia assimilation enzymes as needed (

Kim et al., 2017

Kim J.N.
Méndez–García C.
Geier R.R.
Iakiviak M.
Chang J.
Cann I.
Mackie R.I.

Metabolic networks for nitrogen utilization in Prevotella ruminicola..

); they probably shift more to peptides as fuel as carbohydrate becomes depleted. Leucine (pairing with isovalerate) was metabolized to a greater extent than the other 2 BCAA in batch cultures of mixed ruminal bacteria (

Atasoglu et al., 2004

Atasoglu C.
Guliye A.Y.
Wallace R.J.

Use of stable isotopes to measure de novo synthesis and turnover of amino acid-C and -N in mixed micro-organisms from the sheep rumen in vitro..

). Thus, decreasing net production of isobutyrate and isovalerate relative to 2-methybutyrate (

Ifkovits and Ragheb, 1968

Roman-Garcia Y.
Mitchell K.E.
Denton B.L.
Lee C.
Socha M.
Wenner B.A.
Firkins J.L.

Conditions stimulating NDF degradation by dosing branched chain volatile fatty acids. II: Relation of branched-chain volatile fatty acids supplementation with solid passage rate and pH on NDF degradation and microbial function in continuous culture.

) is likely explained by increasing catabolism, thus explaining decreasing recovery of ¹³C (mostly or entirely from isobutyrate) in even-chain iso-FA (Table 2) even though the profile was not markedly reduced (Table 1).

Decreased pH tended to increase anteiso-17:0 and total anteiso-FA in bacteria harvested from continuous cultures (

Berthelot et al., 2019

Berthelot V.
Albarello H.
Broudiscou L.P.

Effect of extruded linseed supplementation, grain source and pH on dietary and microbial fatty acid outflows in continuous cultures of rumen microorganisms.

). In our study, low pH tended to increase the 16:0 percentage of bacterial FA, and pH interacted with both BCVFA and k_p for OCFA such that both OCFA and anteiso-FA tended to be increased the most by low pH when k_p was high. Bacillus subtilis inserts 15:0 and anteiso-FA primarily in the sn-1 and sn-2 positions, respectively, of its membrane's phospholipids (

Kaneda, 1991

Kaneda T.

Iso- and anteiso-fatty acids in bacteria: Biosynthesis, function, and taxonomic significance.

). Although little is known about such positional preference in rumen bacterial lipids, the anteiso-FA might be inserted in the sn-2 position to prevent excessive rigidity when SFA are increasingly inserted into the sn-1 position in rumen bacterial membranes.

Vlaeminck et al., 2006b

Vlaeminck B.
Fievez V.
Demeyer D.
Dewhurst R.J.

Effect of forage:concentrate ratio on fatty acid composition of rumen bacteria isolated from ruminal and duodenal digesta.

reported a higher proportion of anteiso-FA compared with iso-FA in bacteria from cows fed high-concentrate diets. Some amylolytic bacteria are highly enriched in anteiso-FA and contain low iso-FA (

Ifkovits R.W.
Ragheb H.S.

Cellular fatty acid composition and identification of rumen bacteria.

Miyagawa E.

Cellular fatty acid and fatty aldehyde composition of rumen bacteria.

;

Minato et al., 1988

Minato H.
Ishibashi S.
Hamaoka T.

Cellular fatty acid and sugar composition of representative strain of rumen bacteria.

). More work is needed to compare differences in BCFA profile in amylolytic and cellulolytic bacteria as mediated by shifts in their environmental pH.

The anteiso-14:0 and anteiso-16:0 are unlikely to be derived from 2-methylbutyrate as would be the odd-chain anteiso-FA (

Parsons J.B.
Rock C.O.

Bacterial lipids: Metabolism and membrane homeostasis.

). Further identification beyond our capacity was recommended before identifying even-chain anteiso-FA (

Kaneda, 1991

Kaneda T.

Iso- and anteiso-fatty acids in bacteria: Biosynthesis, function, and taxonomic significance.

); hence, these FA were not tallied in the total anteiso-FA. Although α-oxidation could have reduced the chain length by 1 carbon, which is well recognized in aerobic cultures and has been shown in a few pure cultures of ruminal bacteria, its significance was deemed to be minor in the rumen (

Vlaeminck B.
Fievez V.
Cabrita A.R.J.
Fonseca A.J.M.
Dewhurst R.J.

Factors affecting odd- and branched-chain fatty acids in milk: A review.

) and in batch culture of mixed rumen microbes (

Wu and Palmquist, 1991

Wu Z.
Palmquist D.L.

Synthesis and biohydrogenation of fatty acids by ruminal microorganisms in vitro.

). Ruminal bacteria are well known to make considerable amounts of fatty aldehydes (

Stewart C.S.
Flint H.J.
Bryant M.P.

The rumen bacteria.

), many of which can be available as dimethylacetals after standard methylation procedures (

Alves et al., 2013

Alves S.P.
Santos-Silva J.
Cabrita A.R.J.
Fonseca A.J.M.
Bessa R.J.B.

Detailed dimethylacetal and fatty acid composition of rumen content from lambs fed lucerne or concentrate supplemented with soybean oil.

). The majority of those had chain lengths of 14 to 16 carbons (

Ventto et al., 2017

Ventto L.
Leskinen H.
Kairenius P.
Stefański T.
Bayat A.R.
Vilkki J.
Shingfield K.J.

Diet-induced milk fat depression is associated with alterations in ruminal biohydrogenation pathways and formation of novel fatty acid intermediates in lactating cows.

). Discerning the identity of these peaks potentially eluting at the same time as anteiso-14:0 and anteiso-16:0 standards (which could be produced in aerobic bacteria) would require commercially available dimethylacetal standards.

The main effect of k_p for odd-chain iso-FA (elongated from isovalerate) and interactions (or trends) of BCVFA with k_p for even-chain iso-FA (from isobutyrate), anteiso-FA, and total BCFA with increasing k_p support our expectation that faster growth of mixed bacteria requires more BCVFA incorporation into BCFA. A higher k_p could also have shifted bacterial populations toward those with faster potential growth, possibly explaining the trend for a modest increase in Simpson's diversity index (Table 3). Even though total bacterial N production was not affected by treatment (

Roman-Garcia Y.
Mitchell K.E.
Denton B.L.
Lee C.
Socha M.
Wenner B.A.
Firkins J.L.

Conditions stimulating NDF degradation by dosing branched chain volatile fatty acids. II: Relation of branched-chain volatile fatty acids supplementation with solid passage rate and pH on NDF degradation and microbial function in continuous culture.

), faster growth post-feeding could be equalized during the 12-h feeding interval, so these effects could be accentuated with multiple feeding events in vivo. Although we dosed racemic 2-methylbutyrate, the S enantiomer is probably preferred for incorporation into anteiso-FA in Bacillus subtilis (

Kaneda, 1991

Kaneda T.

Iso- and anteiso-fatty acids in bacteria: Biosynthesis, function, and taxonomic significance.

) even though up to 5 to 10% could be derived from R enantiomers in ruminal bacteria (

Eibler et al., 2017

Eibler D.
Abdurahman H.
Ruoff T.
Kaffarnik S.
Steingass H.
Vetter W.

Unexpected formation of low amounts of (r)-configurated anteiso-fatty acids in rumen fluid experiments.

). Relative to isobutyrate and isovalerate, a benefit from exogenous 2-methylbutyrate is likely less reflected by a decreased ruminal concentration if the R enantiomers accumulate.

Our diet only had a moderate amount of dietary unsaturated fat. The decrease in cis-18:1 and trans-18:1 isomers with decreasing pH does suggest interruption of biohydrogenation at the lipolysis step, which would be supported by an approximate doubling of linoleic acid outflow in effluent from continuous culture (

Qiu et al., 2004

Qiu X.
Eastridge M.L.
Griswold K.E.
Firkins J.L.

Effects of substrate, passage rate, and pH in continuous culture on flows of conjugated linoleic acid and trans C18:1.

). Interactions of BCVFA with k_p (cis-18:1) and pH with k_p (trans-18:1) cannot be explained because our column could not consistently separate isomer peaks. Multiple cis-18:1 and trans-18:1 isomers result from biohydrogenation of linoleic and linolenic acid (

Dewanckele et al., 2020

Dewanckele L.
Toral P.G.
Vlaeminck B.
Fievez V.

Invited review: Role of rumen biohydrogenation intermediates and rumen microbes in diet-induced milk fat depression: An update.

).

Although valerate was assumed not needed and therefore not dosed in the BCVFA treatment (

Roman-Garcia Y.
Denton B.L.
Mitchell K.E.
Lee C.
Socha M.T.
Firkins J.L.

Conditions stimulating neutral detergent fiber degradation by dosing branched-chain volatile fatty acids. I: Comparison with branched-chain amino acids and forage source in ruminal batch cultures.

), OCFA increased with decreasing pH (Table 1). Valerate production also increased (

Roman-Garcia Y.
Mitchell K.E.
Denton B.L.
Lee C.
Socha M.
Wenner B.A.
Firkins J.L.

Conditions stimulating NDF degradation by dosing branched chain volatile fatty acids. II: Relation of branched-chain volatile fatty acids supplementation with solid passage rate and pH on NDF degradation and microbial function in continuous culture.

), thus increasing competition with even-chain primers for FA synthesis. Both butyrate and valerate increased considerably with decreasing pH of continuous cultures (

Berthelot et al., 2019

Berthelot V.
Albarello H.
Broudiscou L.P.

Effect of extruded linseed supplementation, grain source and pH on dietary and microbial fatty acid outflows in continuous cultures of rumen microorganisms.

). Valerate can be formed from degradation of some AA (

Gorosito et al., 1985

Gorosito A.R.
Russell J.B.
Van Soest P.J.

Effect of carbon-4 and carbon-5 volatile fatty acids on digestion of plant cell wall in vitro.

) but it is extensively formed from condensation of propionate with acetyl-CoA to consume [2H], much like valeryl-CoA is further elongated to OCFA (

Ungerfeld, 2015b

Ungerfeld E.M.

Shifts in metabolic hydrogen sinks in the methanogenesis-inhibited ruminal fermentation: A meta-analysis.

). The BCFA have higher melting points than the OCFA (

Enser, 1984

Enser M.

The chemistry, biochemistry and nutritional importance of animal fats.

); when in lower pH environments that increase membrane fluidity, we would expect more 15:0 and less iso-FA to be incorporated into membranes (

Fievez et al., 2012

Fievez V.
Colman E.
Castro-Montoya J.M.
Stefanov I.
Vlaeminck B.

Milk odd- and branched-chain fatty acids as biomarkers of rumen function—An update.

). Lower pH suppressed methanogenesis, which led to alternative [2H] sinks (

Ungerfeld, 2015a

Ungerfeld E.M.

Limits to dihydrogen incorporation into electron sinks alternative to methanogenesis in ruminal fermentation.

), as supported by shifts in bacterial populations (see later discussion).

Relative Sequence Abundance of Prokaryotes

Diversity Indices

Continuous cultures had lower diversity but were still represented by major taxa from the original inoculum (

Salfer et al., 2018

Salfer I.J.
Staley C.
Johnson H.E.
Sadowsky M.J.
Stern M.D.

Comparisons of bacterial and archaeal communities in the rumen and a dual-flow continuous culture fermentation system using amplicon sequencing.

). All 8 of the abundant core genera (

Xue et al., 2018

Xue M.
Sun H.
Wu X.
Guan L.L.
Liu J.

Assessment of rumen microbiota from a large dairy cattle cohort reveals the pan and core bacteriomes contributing to varied phenotypes.

) were also abundant in our continuous cultures; in addition, heritable taxa (

Wallace et al., 2019

Wallace R.J.
Sasson G.
Garnsworthy P.C.
Tapio I.
Gregson E.
Bani P.
Huhtanen P.
Bayat A.R.
Strozzi F.
Biscarini F.
Snelling T.J.
Saunders N.
Potterton S.L.
Craigon J.
Minuti A.
Trevisi E.
Callegari M.L.
Cappelli F.P.
Cabezas-Garcia E.H.
Vilkki J.
Pinares-Patino C.
Fliegerová K.O.
Mrázek J.
Sechovcová H.
Kopečný J.
Bonin A.
Boyer F.
Taberlet P.
Kokou F.
Halperin E.
Williams J.L.
Shingfield K.J.
Mizrahi I.

A heritable subset of the core rumen microbiome dictates dairy cow productivity and emissions.

) were nearly completely represented, suggesting strong representation of ruminal conditions. The ASV clustering approach was chosen to improve accuracy and reduce technical errors associated with sequencing and computational analyses using operational taxonomic unit clustering approaches (

Caruso et al., 2019

Caruso V.
Song X.
Asquith M.
Karstens L.

Performance of microbiome sequence inference methods in environments with varying biomass.

). The average ASV that were observed (n = 1,308) or derived by the Chao 1 nonparametric statistic (n = 1,392) indicate a robust prokaryotic community. Although we noted no change in our study resulting from dosing BCVFA, when BCVFA were fed to lactating dairy cattle, the number of ASV and most indices of α-diversity increased (

Lee et al., 2021

Lee C.
Copelin J.E.
Park T.
Mitchell K.E.
Firkins J.L.
Socha M.T.
Luchini D.

Effects of diet fermentability and supplementation of 2-hydroxy-4-(methylthio)-butanoic acid and isoacids on milk fat depression: 2. Ruminal fermentation, fatty acid, and bacterial community structure.

). Our β-diversity results (Figure 1) support the major shifts in relative ASV abundance associated with pH, and the period effect reflects the important role of differing inocula sources. Our inocula was prepared from 4 lactating dairy cows (

Hackmann and Firkins, 2015

Roman-Garcia Y.
Mitchell K.E.
Denton B.L.
Lee C.
Socha M.
Wenner B.A.
Firkins J.L.

Conditions stimulating NDF degradation by dosing branched chain volatile fatty acids. II: Relation of branched-chain volatile fatty acids supplementation with solid passage rate and pH on NDF degradation and microbial function in continuous culture.

).

Phylum Bacteroidetes

Prevotella and BS11 were highly represented in the Bacteroidetes, which represented about half of all ASV in our study. The Prevotellaceae are well known for the versatility of substrates used by its members (

Deusch et al., 2017

Deusch S.
Camarinha-Silva A.
Conrad J.
Beifuss U.
Rodehutscord M.
Seifert J.

A structural and functional elucidation of the rumen microbiome influenced by various diets and microenvironments.

) and for metabolic efficiency, such as making ATP via electron transport phosphorylation (

Deusch et al., 2019

Deusch S.
Bok E.
Schleicher L.
Seifert J.
Steuber J.

Occurrence and function of the Na⁺-translocating NADH:quinone oxidoreductase in Prevotella spp.

). Prevotella was strongly positively associated with subacute rumen acidosis, as was family S24–7, whereas family BS11 was inversely associated (

McCann et al., 2016

McCann J.C.
Luan S.
Cardoso F.C.
Derakhshani H.
Khafipour E.
Loor J.J.

Induction of subacute ruminal acidosis affects the ruminal microbiome and epithelium.

). Even subclinical acidosis would promote periods of low feed intake and therefore slow ruminal k_p, explaining some of the interactions, such that S24–7 might have outcompeted BS11 with low pH and low k_p combinations. The uncultivated BS11 likely is specific for a hemicellulose rather than starch (

Solden et al., 2017

Solden L.M.
Hoyt D.W.
Collins W.B.
Plank J.E.
Daly R.A.
Hildebrand E.
Beavers T.J.
Wolfe R.
Nicora C.D.
Purvine S.O.
Carstensen M.
Lipton M.S.
Spalinger D.E.
Firkins J.L.
Wolfe B.A.
Wrighton K.C.

New roles in hemicellulosic sugar fermentation for the uncultivated bacteroidetes family BS11.

), whereas S24–7 is likely more amylolytic and opportunistic (

Ormerod et al., 2016

Ormerod K.L.
Wood D.L.A.
Lachner N.
Gellatly S.L.
Daly J.N.
Parsons J.D.
Dal'Molin C.G.O.
Palfreyman R.W.
Nielsen L.K.
Cooper M.A.
Morrison M.
Hansbro P.M.
Hugenholtz P.

Genomic characterization of the uncultured Bacteroidales family S24–7 inhabiting the guts of homeothermic animals.

). Amylolytic bacteria should have more resilience to decreased ruminal pH. Little is known about Bacteroidales family RF16, but increasing dietary hay inclusion increased its relative abundance (

Klevenhusen et al., 2017

Klevenhusen F.
Petri R.M.
Kleefisch M.T.
Khiaosa-Ard R.
Metzler-Zebeli B.U.
Zebeli Q.

Changes in fibre-adherent and fluid-associated microbial communities and fermentation profiles in the rumen of cattle fed diets differing in hay quality and concentrate amount.

Google Scholar

).

Phylum Firmicutes

Many shifts were observed within the members of the Firmicutes resulting from decreasing pH (Table 4). Both Butyrivibrio and Pseudobutyrivibrio increased in relative sequence abundance with increasing k_p, perhaps resulting from their unique ways to make ATP (

Hackmann T.J.
Firkins J.L.

Electron transport phosphorylation in rumen butyrivibrios: Unprecedented ATP yield for glucose fermentation to butyrate.

). However, Pseudobutyrivibrio decreased with decreasing pH. The butyrivibrios have been grouped with respect to their ability to produce lactate (

Palevich et al., 2019

Palevich N.
Kelly W.J.
Leahy S.C.
Denman S.
Altermann E.
Rakonjac J.
Attwood G.T.

Comparative genomics of rumen Butyrivibrio spp. Uncovers a continuum of polysaccharide-degrading capabilities.

). Although characterized Pseudobutyrivibrio can produce lactate, their fibrolytic niche seems to be more restrictive than that of Butyrivibrio. The butyrivibrios are recognized for biohydrogenation through the trans-11 18:1 pathway, and the Butyrivibrio genus includes stearate producers that are more inhibited by linoleic acid when lactate accumulates (

Paillard et al., 2007

Paillard D.
McKain N.
Chaudhary L.C.
Walker N.D.
Pizette F.
Koppova I.
McEwan N.R.
Kopecný J.
Vercoe P.E.
Louis P.
Wallace R.J.

Relation between phylogenetic position, lipid metabolism and butyrate production by different Butyrivibrio-like bacteria from the rumen.

).

Decreasing pH increased genera of the Veillonellaceae (including Succiniclasticum, Dialister, and Mitsuokella) and Erysipelotrichaceae (Bulleidia), many of which are lactate and succinate producers or consumers (

Deusch et al., 2017

Deusch S.
Camarinha-Silva A.
Conrad J.
Beifuss U.
Rodehutscord M.
Seifert J.

A structural and functional elucidation of the rumen microbiome influenced by various diets and microenvironments.

). Mitsuokella is closely related to Selenomonas, and Succiniclasticum is noted for its main niche to convert succinate to propionate (

Stewart C.S.
Flint H.J.
Bryant M.P.

The rumen bacteria.

). Inhibition of methanogenesis increased expression of genes in propionate production via succinate particularly attributed to increases in Prevotella and in Selenomonas or other Veillonellaceae in goats (

Denman et al., 2015

Denman S.E.
Martinez Fernandez G.
Shinkai T.
Mitsumori M.
McSweeney C.S.

Metagenomic analysis of the rumen microbial community following inhibition of methane formation by a halogenated methane analog.

). Increased Prevotella and Veillonellaceae help explain the simultaneous 28% decreases in both methane production and acetate:propionate ratio when pH was decreased (

Roman-Garcia Y.
Mitchell K.E.
Denton B.L.
Lee C.
Socha M.
Wenner B.A.
Firkins J.L.

Conditions stimulating NDF degradation by dosing branched chain volatile fatty acids. II: Relation of branched-chain volatile fatty acids supplementation with solid passage rate and pH on NDF degradation and microbial function in continuous culture.

). Many of the Veillonellaceae were also correlated with a shift toward the trans-10 18:1 pathway (

Dewanckele et al., 2018

Dewanckele L.
Vlaeminck B.
Hernandez-Sanabria E.
Ruiz-González A.
Debruyne S.
Jeyanathan J.
Fievez V.

Rumen biohydrogenation and microbial community changes upon early life supplementation of 22:6n-3 enriched microalgae to goats.

).

The family Ruminococcaceae is widely recognized for its cellulolytic representatives, R. albus and R. flavefaciens; however, other members of this family have diverse substrates (

La Reau and Suen, 2018

La Reau A.J.
Suen G.

The ruminococci: Key symbionts of the gut ecosystem.

). Although the Ruminococcaceae are prevalent in the particulate fraction and typically associated with cellulose and hemicellulose degradation (

Denman et al., 2018

Denman S.E.
Morgavi D.P.
McSweeney C.S.

Review: The application of omics to rumen microbiota function.

;

Moraïs and Mizrahi, 2019

Moraïs S.
Mizrahi I.

Islands in the stream: From individual to communal fiber degradation in the rumen ecosystem.

), the ruminal Ruminococcus bromii seems to be specifically amylolytic and prevalent in ruminants fed high-grain diets (

Mukhopadhya et al., 2018

Mukhopadhya I.
Moraïs S.
Laverde-Gomez J.
Sheridan P.O.
Walker A.W.
Kelly W.
Klieve A.V.
Ouwerkerk D.
Duncan S.H.
Louis P.
Koropatkin N.
Cockburn D.
Kibler R.
Cooper P.J.
Sandoval C.
Crost E.
Juge N.
Bayer E.A.
Flint H.J.

Sporulation capability and amylosome conservation among diverse human colonic and rumen isolates of the keystone starch-degrader Ruminococcus bromii..

). In support, relative abundance of ASV assigned to R. bromii was increased (P < 0.01) from <0.01% to 0.57% with high versus low pH treatments (data not shown). Therefore, increased amylolytic Ruminococcus species at the expense of cellulolytic species probably explains the lack of response in the total genus, whereas Oscillospira (family Ruminococcaceae) and uncharacterized Ruminococcaceae were apparently inhibited by decreasing pH. Oscillospira can have unique morphologies and potentially diverse niches in the rumen, but it was particularly abundant with ruminants fed fresh forage diets (

Mackie et al., 2003

Mackie R.I.
Aminov R.I.
Hu W.
Klieve A.V.
Ouwerkerk D.
Sundset M.A.
Kamagata Y.

Ecology of uncultivated Oscillospira species in the rumen of cattle, sheep, and reindeer as assessed by microscopy and molecular approaches.

), which should have higher ruminal pH. Characterized Ruminococcus spp. have clear requirements for BCVFA (

Stewart C.S.
Flint H.J.
Bryant M.P.

The rumen bacteria.

), but we detected no effects of supplementation in our study, perhaps because of a shift among uncharacterized ruminococci that might not have the same requirements or because Fibrobacter outcompeted the cellulolytic ruminococci for the cellulolysis niche.

Phyla Fibrobacteres and Spirochaetes

The relative abundance of Fibrobacter increased when BCVFA were dosed and helped prevent a decreased abundance when pH decreased (Table 5). Its most consistent response to individual BCVFA in the literature was to isobutyrate, although 2-methylbutyrate also might stimulate growth (

Stanton and Canale-Parola, 1980

Roman-Garcia Y.
Denton B.L.
Mitchell K.E.
Lee C.
Socha M.T.
Firkins J.L.

Conditions stimulating neutral detergent fiber degradation by dosing branched-chain volatile fatty acids. I: Comparison with branched-chain amino acids and forage source in ruminal batch cultures.

). Fibrobacter spp. are now known to be more diverse than once thought but are still primarily cellulolytic specialists (

Neumann and Suen, 2018

Neumann A.P.
Suen G.

The phylogenomic diversity of herbivore-associated Fibrobacter spp. Is correlated to lignocellulose-degrading potential.

). Fibrobacter spp. also have a major role in depolymerizing hemicellulose and pectin to support other members in the consortium even though they ferment only end products from cellulolysis. Both phylum Fibrobacteres and genus Treponema are considered core taxa in dairy cattle (

Xue et al., 2018

Xue M.
Sun H.
Wu X.
Guan L.L.
Liu J.

Assessment of rumen microbiota from a large dairy cattle cohort reveals the pan and core bacteriomes contributing to varied phenotypes.

). Treponema spp. can have a direct role in degradation of pectin and possibly hemicellulose but also indirect roles attributed to cross-feeding with cellulolytics (

Moraïs and Mizrahi, 2019

Moraïs S.
Mizrahi I.

Islands in the stream: From individual to communal fiber degradation in the rumen ecosystem.

). As those authors explained, Fibrobacter has a unique and complicated system to degrade fiber, including extracellular protein complexes and membrane-bound vesicles that pit plant cell walls. Ruminal treponemes probably are chemoattracted to cellulose and interact closely with or help push the nonmotile Fibrobacter into close contact with the plant cell wall, cross-feeding on the resultant oligomers or sugars (

Stanton T.B.
Canale-Parola E.

Treponema bryantii sp. nov., a rumen spirochete that interacts with cellulolytic bacteria.

,

Harwood and Canale-Parola, 1984

Harwood C.S.
Canale-Parola E.

Ecology of spirochetes.

). Treponema bryantii requires isobutyrate and 2-methylbutyrate (

Stanton and Canale-Parola, 1980

Stanton T.B.
Canale-Parola E.

Treponema bryantii sp. nov., a rumen spirochete that interacts with cellulolytic bacteria.

), whereas Treponema saccharophilum requires isobutyrate (

Paster and Canale-Parola, 1985

Paster B.J.
Canale-Parola E.

Treponema saccharophilum sp. nov., a large pectinolytic spirochete from the bovine rumen.

) for growth. Both of these treponemes produce formate, although T. bryantii also produces succinate. All ruminal strains of Fibrobacter can produce formate, but all strains mainly produce succinate, which is subsequently fermented primarily to propionate and therefore associated with decreased methanogenesis (

Neumann and Suen, 2018

Neumann A.P.
Suen G.

The phylogenomic diversity of herbivore-associated Fibrobacter spp. Is correlated to lignocellulose-degrading potential.

), supporting our previous discussion. Similarly, positive associations of succinate-producing Fibrobacter and Treponema with succinate-consuming Succiniclasticum and negative relationships with H₂-producing Ruminococcus (

Xie et al., 2018

Xie X.
Yang C.
Guan L.L.
Wang J.
Xue M.
Liu J.X.

Persistence of cellulolytic bacteria Fibrobacter and Treponema after short-term corn stover-based dietary intervention reveals the potential to improve rumen fibrolytic function.

) support greater considerations of networks to suppress methanogenesis without disrupting fiber degradation.

The expectation for Treponema to enhance adherence and fiber degradation by other community members (

Kudo et al., 1987

Kudo H.
Cheng K.J.
Costerton J.W.

Interactions between Treponema bryantii and cellulolytic bacteria in the in vitro degradation of straw cellulose.

), including Fibrobacter succinogenes, likely involves BCVFA availability. Two strains of F. succinogenes had a relatively high concentration of anteiso odd-chain FA or aldehydes some of which increased in their membranes when a BCVFA solution was provided (

Saluzzi et al., 1993

Saluzzi L.
Smith A.
Stewart C.S.

Analysis of bacterial phospholipid markers and plant monosaccharides during forage degradation by Ruminococcus flavefaciens and Fibrobacter succinogenes in co-culture.

). Roles of anteiso-15:0 in the glycolipid component of the membrane were noted for F. succinogenes and the oral spirochete Treponema denticola (

Vinogradov et al., 2001

Vinogradov E.
Egbosimba E.E.
Perry M.B.
Lam J.S.
Forsberg C.W.

Structural analysis of the carbohydrate components of the outer membrane of the lipopolysaccharide-lacking cellulolytic ruminal bacterium Fibrobacter succinogenes S85.

). Those 2 bacteria reportedly lack the lipopolysaccharide components that are typical for other gram-negative bacteria. Moreover, the FA composition in the membrane of the cellulolytic vesicles of Fibrobacter (

Moraïs and Mizrahi, 2019

Moraïs S.
Mizrahi I.

Islands in the stream: From individual to communal fiber degradation in the rumen ecosystem.

) has not been addressed. Although pure cultures of Fibrobacter had adherence inhibited by low pH, induced acidosis did not decrease its relative abundance (

McCann et al., 2016

McCann J.C.
Luan S.
Cardoso F.C.
Derakhshani H.
Khafipour E.
Loor J.J.

Induction of subacute ruminal acidosis affects the ruminal microbiome and epithelium.

). The dietary RDP in that study was not provided; however, based on the dietary composition, RDP was probably more than adequate to provide BCVFA. Based on our BCVFA × pH interaction, more work is needed to assess whether adequate BCVFA aid Fibrobacter's resilience to periodic low pH compared with other cellulolytics.

Other Phyla

The relative increase in Proteobacteria (gram-negative) cumulative ASV and many of ASV assigned as genera in that phylum was at the expense of the relative decrease in most Firmicutes (most are gram-positive with a few exceptions such as the Negativicutes). These results correspond with the relative increase in 3-hydroxy 14:0 (Table 1), which is an important component in the cell wall of gram-negative bacteria derived endogenously (

Parsons J.B.
Rock C.O.

Bacterial lipids: Metabolism and membrane homeostasis.

) and noted in many rumen bacteria (

Sohlenkamp and Geiger, 2016

Miyagawa E.

Cellular fatty acid and fatty aldehyde composition of rumen bacteria.

). The source of the 2-hydroxy 14:0 is less clear. It has been noted in family Burkholderiaceae in class β-Proteobacteria (

Sohlenkamp C.
Geiger O.

Bacterial membrane lipids: Diversity in structures and pathways.

); in our survey, Comamonadaceae (order Burkholderiales) almost disappeared when pH declined. The prevalence of 3-hydoxy 14:0 but not 2-hydroxy 14:0 was noted in isolates presumptively identified as Succinivibrio (

Miyagawa E.

Cellular fatty acid and fatty aldehyde composition of rumen bacteria.

), but occurrence of 2-hydroxy 14:0 has received little attention for ruminal bacteria to our knowledge. Succinivibrionaceae are recognized for their starch degradation and succinate production (

Denman et al., 2018

Denman S.E.
Morgavi D.P.
McSweeney C.S.

Review: The application of omics to rumen microbiota function.

), so their ability to thrive in low pH is expected. Succinivibrionaceae abundance is inversely proportional to methanogenesis (

Wallace et al., 2015

Wallace R.J.
Rooke J.A.
McKain N.
Duthie C.-A.
Hyslop J.J.
Ross D.W.
Waterhouse A.
Watson M.
Roehe R.

The rumen microbial metagenome associated with high methane production in cattle.

). This increase further supports the expanded role of succinate as an intermediate to explain the decreased methanogenesis, as described previously. Increased abundance of Desulfovibrio was also noted in cattle that emitted high amounts of methane (

Wallace et al., 2015

Wallace R.J.
Rooke J.A.
McKain N.
Duthie C.-A.
Hyslop J.J.
Ross D.W.
Waterhouse A.
Watson M.
Roehe R.

The rumen microbial metagenome associated with high methane production in cattle.

). Because aqueous H₂ is captured in sulfate reduction, Desulfovibrio compete with hydrogenotrophic methanogens, and their increase with decreasing pH might have lessened potential differences in aqueous H₂ concentration (

Martínez-Álvaro et al., 2020

Roman-Garcia Y.
Mitchell K.E.
Denton B.L.
Lee C.
Socha M.
Wenner B.A.
Firkins J.L.

Conditions stimulating NDF degradation by dosing branched chain volatile fatty acids. II: Relation of branched-chain volatile fatty acids supplementation with solid passage rate and pH on NDF degradation and microbial function in continuous culture.

).

Methanobrevibacter is typically the most abundant archaeal genus in the rumen (

Martínez-Álvaro M.
Auffret M.D.
Stewart R.D.
Dewhurst R.J.
Duthie C.-A.
Rooke J.A.
Wallace R.J.
Shih B.
Freeman T.C.
Watson M.
Roehe R.

Identification of complex rumen microbiome interaction within diverse functional niches as mechanisms affecting the variation of methane emissions in bovine.

) and was the dominant genus in our survey. Methanomicrobium mobile (hydrogenotrophic, not meeting our abundance threshold) requires all 3 BCVFA but especially 2-methylbutyrate in the highest concentration, whereas Methanobrevibacter ruminantium requires only 2-methylbutyrate (

Tanner and Wolfe, 1988

Tanner R.S.
Wolfe R.S.

Nutritional requirements of Methanomicrobium mobile..

). Methanobrevibacter ruminantium strain M1's requirement was predicted to result from a gene lacking for isoleucine synthesis (

Leahy et al., 2010

Leahy S.C.
Kelly W.J.
Altermann E.
Ronimus R.S.
Yeoman C.J.
Pacheco D.M.
Li D.
Kong Z.
McTavish S.
Sang C.
Lambie S.C.
Janssen P.H.
Dey D.
Attwood G.T.

The genome sequence of the rumen methanogen Methanobrevibacter ruminantium reveals new possibilities for controlling ruminant methane emissions.

). Although we did not detect any effect of BCVFA supplementation on Methanobrevibacter relative abundance, dosing BCVFA did increase methanogenesis, which was attributed in large part to improved NDF degradability (

Roman-Garcia Y.
Mitchell K.E.
Denton B.L.
Lee C.
Socha M.
Wenner B.A.
Firkins J.L.

Conditions stimulating NDF degradation by dosing branched chain volatile fatty acids. II: Relation of branched-chain volatile fatty acids supplementation with solid passage rate and pH on NDF degradation and microbial function in continuous culture.

). The suppressed methanogenesis at low pH might be a direct result of uncoupling of ATP synthesis resulting from changing the proton gradient needed for H⁺-transport-coupled ATP synthesis (

Russell, 1998

Russell J.B.

The importance of pH in the regulation of ruminal acetate to propionate ratio and methane production in vitro.

). Uncoupling of methanogenesis from ATP usage for growth has been proposed. Although it needs empirical testing (

Wallace et al., 2017

Wallace R.J.
Snelling T.J.
McCartney C.A.
Tapio I.
Strozzi F.

Application of meta-omics techniques to understand greenhouse gas emissions originating from ruminal metabolism.

), energy spilling was projected for ruminal methanogens based on nonrumen archaea (

Russell, 2007

Russell J.B.

The energy spilling reactions of bacteria and other organisms.

). In addition to uncoupling of methanogenesis from growth, a shifting of individual species within Methanobrevibacter could also explain the varying efficiency of methanogenesis per its relative abundance (

Danielsson et al., 2017

Danielsson R.
Dicksved J.
Sun L.
Gonda H.
Müller B.
Schnürer A.
Bertilsson J.

Methane production in dairy cows correlates with rumen methanogenic and bacterial community structure.

).

Some inconsistencies in our data remain unresolved by direct comparisons of treatment effects within individual genera. For example, Ruminobacter specifically uses starch as substrate, so its decreased relative ASV abundance with decreasing pH was unexpected but might be a result of displacement by other bacteria that have more versatile substrate accessibility (

Klevenhusen et al., 2017

Klevenhusen F.
Petri R.M.
Kleefisch M.T.
Khiaosa-Ard R.
Metzler-Zebeli B.U.
Zebeli Q.

Changes in fibre-adherent and fluid-associated microbial communities and fermentation profiles in the rumen of cattle fed diets differing in hay quality and concentrate amount.

Google Scholar

), perhaps because of Ruminobacter's cellular location of amylase (