Adhesion of Cellulolytic Ruminal Bacteria to Barley Straw

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Sept 199, p. 2698-273 99-224/9/92698-6$2./ Copyright 3 199, American Society for Microbiology Vol. 56, No. 9 Adhesion of Cellulolytic Ruminal Bacteria to Barley Straw SIVA BHAT, R. J. WALLACE,* AND E. R. RSKOV Rowett Research Institute, Bucksburn, Aberdeen AB2 9SB, United Kingdom Received 23 January 199/Accepted 21 June 199 Adhesion of the cellulolytic ruminal bacteria Ruminococcus flavefaciens and Fibrobacter succinogenes to barley straw was measured by incubating bacterial suspensions with hammer-milled straw for 3 min, filtering the mixtures through sintered glass filters, and measuring the optical densities of the filtrates. Maximum adhesion of both species occurred at ph 6. and during mid- to late-exponential phase. Adhesion was saturable at 33 and 23 mg (dry weight) g of straw-' for R.flavefaciens and F. succinogenes, respectively. Methyl cellulose and carboxymethyl cellulose inhibited adhesion by 24 to 33%. Competition between species was determined by measuring characteristic cell-associated enzyme activities in filtrates of mixtures incubated with straw; p-nitrophenyl-,-d-lactopyranoside hydrolysis was used as a marker for F. succinogenes, while either j-xylosidase or carboxymethyl celiulase was used for R. flavefaciens, depending on the other species present. R. flavefaciens had no influence on F. succinogenes adhesion, and F. succinogenes had only a minor (<2%) effect on R. flavefaciens adhesion. The noncellulolytic ruminal bacteria Bacteroides ruminicola and Seknomonas ruminantium had no influence on adhesion of either cellulolytic species, although these organisms also adhered to the straw. We concluded that R. flavefaciens and F. succinogenes have separate, specific adhesion sites on barley straw that are not obscured by competition with non-cellulolytic species. Digestion of plant cell walls in rumina occurs predominantly by the action of cellulolytic bacteria (1, 25). Close association of cellulolytic bacteria occurs while celluloserich plant cell wall materials are degraded (2, 3, 11-13, 17, 18), suggesting that bacterial adhesion is an important aspect of fiber breakdown. Adhesion of cellulolytic bacteria can be inhibited, and even reversed, by substrate analogs such as methyl cellulose (MC) and carboxymethyl cellulose (CMC) (11, 14, 15, 2). MC also inhibits cellulolytic enzyme activity (8, 19, 24), so by implication it seems likely that the cellulolytic enzymes are involved in adhesion (11, 14, 15), although this has by no means been confirmed (6, 24). Morris and Cole found that strains of Ruminococcus albus that did not adhere to cellulose generally showed little cellulolytic activity (16). This was not true of all strains, however, and adhesion to cellulose powder was not sufficient to enable digestion to occur (16). Adhesion of Fibrobacter succinogenes has been found to be specific and protein mediated (6). Gong and Forsberg isolated three classes of nonadherent mutants of F. succinogenes, each having different growth characteristics, which suggested that lesions had occurred in the regulation of synthesis of a cellulose-binding factor (6). The colonization of straw by cellulolytic bacteria has been utilized as a way to predict the quality of different batches and cultivars. Microbial carboxymethyl cellulase (CMCase) activity extracted from straw which had been incubated in nylon bags in vivo for 24 or 48 h was highly correlated with degradability (22). Therefore, it was surprising that, when attempts were made to find a rapid in vitro method of evaluating different straw samples by the initial adhesion of cellulolytic bacteria, no relationship was found between adhesion and degradability (2). In most quantitative studies of adhesion workers have used pure cellulose to define interactions between cellulolytic bacteria and their substrates (6, 1, 11, 14-18, 2, 21). Little is known about how these mechanisms relate to adhesion to the natural substrates which enter rumina. For * Corresponding author. 2698 example, straw contains abundant xylan and other polysaccharides, as well as cellulose (7). In this study we investigated some of the properties of adhesion of cellulolytic bacteria to barley straw, and we also investigated how different species, including noncellulolytic organisms, compete for binding sites. MATERIALS AND METHODS Bacterial strains. F. succinogenes S85 and Ruminococcus flavefaciens FD1 were gifts from M. P. Bryant, University of Illinois, Urbana. Bacteroides ruminicola M384 and Selenomonas ruminantium Z18 were isolated from sheep rumen fluid and are maintained at the Rowett Research Institute, Aberdeen, United Kingdom. Growth. Bacteria were grown in Hungate tubes under CO2 in ruminal fluid-containing medium 2 of Hobson (9) supplemented with 2 g of glucose per liter, 2 g of maltose per liter, and 2 g of cellobiose per liter. The inoculum used varied from 2 to 4% in order to produce mid- to late-exponentialphase cells after 14 to 18 h of incubation at 39 C. Adhesion measurements. All experiments were done with a single sample of barley straw (cultivar Corgi). The straw was ground in a laboratory hammer mill so that it passed through a 1-mm screen. Adhesion was allowed to occur by adding 7 ml of bacterial suspension to.25 g of straw in a Hungate tube, inverting the tube several times, and incubating the preparation at 39 C under CO2 for 3 min. This incubation time allows maximal adhesion of cellulolytic bacteria to cellulose (1, 16, 2, 21) and gave maximal adhesion in our experiments. The mixture was then filtered under a vacuum through a sintered glass filter, and the optical density at 65 nm of the filtrate was determined (2). Optical density values were converted to dry weight per milliliter by harvesting cells by centrifugation at 48, x g for 15 min from identical cultures and drying the resulting pellets to constant weight at 7 C. Levels of adhesion were calculated from the optical densities of filtrates of uninoculated medium (A), uninoculated medium to which straw had been added (B), a bacterial suspension (C), and a bacterial suspension to which straw

VOL. 56, 199 ADHESION OF RUMINAL BACTERIA TO BARLEY STRAW 2699 2 4 6 8 1 ph FIG. 1. Effect on ph of adhesion of rumen cellulolytic bacteria to barley straw in vitro. Each point is the mean of four determinations of two different cultures. Symbols:, R. flavefaciens FD1;, F. succinogenes S85. had been added (D), as follows: proportion of bacteria adhering = 1 - (D - B)I(C - A). The influence of ph on adhesion was determined by harvesting cells and suspending them in medium 2 to which we added.1 M Na2HPO3,.5 M citric acid, and HCl or NaOH to provide ph values ranging from 4 to 8. The influence of cell density was determined by suspending bacteria in different volumes of fresh medium. The inhibitory effects of MC (viscosity of 2% aqueous solution, 15 cp; Sigma Chemical Co., St. Louis, Mo.) and CMC (low viscosity; Sigma) were determined by dissolving these compounds in medium 2 and mixing the resulting 1.2.8 E '.4 (o 1.6 = 1.2 3: a.8.4.6p- I a) DV- C, cn cn >, (a) I1/1 -- % - _ ĊO, c,'_/_- "_. - (b) 1. r.8f.4 - o ) m C.2k i._ I / I I I /c -''.. _. s,..o,o ~ ~ ~ 4 8 1 2 16 2 24 2?8. 1.2.8 E.4 3: E 1.6- la 11.2 "c.8 ', ).4.8 ( Time (h) FIG. 2. Influence of stage of growth on the initial adhesion of R. flavefaciens (a) and F. succinogenes (b) to barley straw. Each point is the mean of two determinations of the same culture. Symbols:, growth;, adhesion. O.D., Optical density. - E - a) a) V cr- CZ, cc E m 1.H.5p- IL le I/ - -. -. - f I I I 1. 1.5 2. 2.5 Total biomass (mg dry wt mlv1) FIG. 3. Effect of bacterial cell density on adhesion to barley straw in vitro. Each point is the mean of four determinations of the adhesion of two different cultures to.25 g of straw suspended in 7 ml of medium. Symbols:, R. flavefaciens FD1;, F. succinogenes S85. preparations with bacterial suspensions to give final concentrations of.1% MC or CMC and bacterial optical densities of.5. Influence of growth phase on adhesion. Bacteria were grown at 39 C in 1 ml of medium 2 contained in serum bottles that were fitted with rubber septa and crimp caps (Phase Separations Ltd., Queensferry, United Kingdom). Samples were withdrawn anaerobically, and the optical densities were adjusted to.7 for measurement of adhesion. Competition experiments. The relative numbers of F. succinogenes and R. flavefaciens cells in mixtures with each other or with other bacteria in filtrates after incubation with straw were determined by measuring the activities of aryl glycosidases which were characteristic of different species. Overnight cultures in Hungate tubes were centrifuged at 1,2 x g for 3 min, and the resulting supernatants were replaced with fresh medium to give the required cell density. Increasing cell densities of potentially competitive bacteria were added to constant cell densities of F. succinogenes (1.9 Addition TABLE 1. Effects of MC and CMC on adhesion of R. flavefaciens and F. succinogenes to barley straw _ Adhesion (%Y) R. flavefaciens F. succinogenes None 55 1.2 42 1.6 MC (.1%) 37 2.2b 28 ± 3.8" CMC (.1%) 42 ± 2.3b 29 ± 1.8 a Average ± standard deviation of three determinations from three separate cultures. b Values differ from control values (P <.1) as determined by Student's t test.

27 BHAT ET AL. APPL. ENVIRON. MICROBIOL. TABLE 2. Cell density (mg [dry wt] per ml) Influence of varying the cell density of R. flavefaciens on the adhesion of F. succinogenes to barley straw when P-D-lactosidase activity was used as a marker 13-D-Lactosidase activity" Activity added Activity in filtrate F. succinogenes R. flavefaciens R. flavefaciens + (mg [dry wt] per g of straw)' F. succinogenes R. flavefaciens F. succinogenes R. flavefaciens adealn added alone F.scngns F. succinogenes 2.1 8. ±.27 5.1 ±.13 21.3 2.1 1.4 8.7 ±.9.2 ±.2.1 ±.2 6. ±.18 18.9 2.1 4.2 8.3 ±.18 1.9 ±.1 1.1 ±.15 6.4 ±.29 21.2 2.1 7. 8.7 ±.29 3.7 ±.12 2.4 ±.8 8.2 ±.25 19.6 2.1 8.4 9.6 ±.21 4.5 ±.18 3. ±.15 9.4 ±.19 19.6 a Values are the means ± standard deviations for four determinations from two different cultures and are expressed as nanomoles of p-nitrophenol produced b Calculated from the activities measured in filtrates. The activity found with R. flavefaciens alone was subtracted from the activity measured in the filtrate of the mixture. The difference was assumed to be due to F. succinogenes, and the quantity of F. succinogenes that adhered to the straw was calculated on this basis to 2.1 mg [dry weight] per ml) or R. flavefaciens (2.2 to 2.5 mg [dry weight] per ml). The levels of adhesion of the potentially competitive bacteria were determined separately by using the same cell densities. The mixtures were incubated and filtered as described above. Filtrates were centrifuged at 48, x g for 15 min, and the bacteria were suspended in 1 mm sodium phosphate buffer (ph 6.8) to measure enzyme activities. Enzyme assays. Aryl glycosidase activities in bacterial cultures and filtrates were determined by using p-nitrophenyl-,3-d-lactopyranoside (Sigma) as a substrate for F. succinogenes preparations and o-nitrophenyl-,-d-xylopyranoside (Sigma) as a substrate for R. flavefaciens preparations; the former activity is referred to below as,-lactosidase activity, although we did not establish that a single enzyme was responsible for p-nitrophenol release. The R. flavefaciens activity was 3-xylosidase (EC 3.2.1.37). A reaction mixture containing 1 ml of bacterial suspension and 1 ml of substrate (1.5 g/liter) was incubated for 3 min at 39 C. The reaction was terminated by adding 2 ml of 5% sodium carbonate, and the A43 of the centrifuged (48, x g, 15 min) reaction mixture was measured. Aryl glycosidase activities are expressed below as nanomoles of nitrophenol released per minute when o- or p-nitrophenol was used as the standard. CMCase extraction and activity measurements were done as described previously (22). CMCase activity is expressed below as nanomoles of D-glucose released per minute. TABLE 3. Cell density (mg [dry wtj per ml) RESULTS Adhesion of both R. flavefaciens and F. succinogenes to barley straw showed a fairly sharp peak at ph 6. (Fig. 1). However substantial adhesion was found at both ph 4. and 8. (equivalent to one-half of the maximum level or more). The stage of growth influenced adhesion. Peak adhesion occurred during mid- to late-exponential phase with both bacterial species (Fig. 2). Adhesion did not decline to much less than one-half of the maximum level at other phases of the growth cycle. Adhesion of both species was a saturable process (Fig. 3). Under the conditions used in this study, in which 7 ml of bacterial suspension was added to.25 g of straw, maximum adhesion was reached when the cell density of either species was about 1.6 mg (dry weight) per ml. However, because the affinity of R. flavefaciens was greater than that of F. succinogenes, this value corresponded to different population densities on the straw. The maximum level of adhesion for R. flavefaciens was 1.2 mg (dry weight) per ml, which was equivalent to 33 mg (dry weight) per g of straw, compared with values of.85 mg (dry weight) per ml and 23 mg (dry weight) per g of straw for F. succinogenes. Both MC and CMC inhibited the adhesion of R. flavefaciens and F. succinogenes to barley straw (Table 1). The degree of inhibition by MC was similar for both species Influence of varying the cell density of F. succinogenes on the adhesion of R. flavefaciens to barley straw when,b-d-xylosidase activity was used as a marker,-d-xylosidase activitya Activity added Activity in filtrate R. flavefaciens R. flavefaciens F. succinogenes R. flavefaciens F. succinogenes F. succinogenes F. succinogenes + (mg [dry wt] per g of straw)b added alone R. flavefaciens 2.5 8.2 +.16 3.9 ±.23 36.7 2.5.7 8.6 ±.15 3.9 ±.24 38.3 2.5 2.1 8.2 ±.16 1. ±.8.6 ±.4 4.1 ±.24 4.1 2.5 3.6 8.4 ±.19 2.2 ±.6 1.3 ±.13 5.2 ±.22 32.5 2.5 4.3 8.7 ±.9 2.5 +.18 1.7 +.15 5.4 ±.16 29.8 a Values are the means ± standard deviations for four determinations from two different cultures and are expressed as nanomoles of o-nitrophenol produced b Calculated from the activities measured in filtrates. The activity found with F. succinogenes alone was subtracted from the activity measured in the filtrate of the mixture. The difference was assumed to be due to R. flavefaciens, and the quantity of R. flavefaciens that adhered to the straw was calculated on this basis

VOL. 56, 199 ADHESION OF RUMINAL BACTERIA TO BARLEY STRAW 271 TABLE 4. Influence of varying the cell density of S. ruminantium on the adhesion of F. succinogenes to barley straw when,-d-lactosidase activity was used as a marker Cell density (mg [dry wt] per ml),-d-lactosidase activitya Activity added: Activity in filtrate: F. succinogenes F. succinogenes S. ruminantium F. succinogenes S. ruminantium Activity added: F. succinogenes + (mg [dry wt] per g of straw)b 1.9 9.2 ±.13 5.2 ±.22 23.1 1.9.6 8.7 ±.8 5.6 ±.17 19.7 1.9 1.9 8.8 ±.9 5. ±.36 23. 1.9 3.1 8.7 ±.11 5.2 ±.23 21.4 1.9 3.7 8.7 ±.9 5.2 ±.31 21.4 a Values are the means + standard deviations for four determinations from two different cultures and are expressed as nanomoles of p-nitrophenol produced b Calculated (about 33%). Inhibition by CMC was greater for F. succinogenes (31%) than for R. flavefaciens (24%). Competition between species for adhesion was determined by measuring aryl glycosidase activities which were characteristic of individual species. The activities measured were entirely cell associated in the original cultures. Cell suspensions were incubated with straw and filtered as described above. Enzyme activities were measured in the original bacterial suspension and in resuspended cells from the filtrate; this enabled us to calculate the composition of the mixture in the filtrate and hence the degree of attachment to straw. P-Lactosidase activity was much higher in F. succinogenes (4.1 nmol/mg [dry weight] per min) than in R. flavefaciens (.5 nmol/mg [dry weight] per min), S. ruminantium (. nmol/mg [dry weight] per min), and B. ruminicola (1. nmol/mg [dry weight] per min). Therefore, the cell density of F. succinogenes could be determined in the presence of S. ruminantium without correction. Corrections for the enzyme activity of the second organism were required to determine adhesion of F. succinogenes in the presence of the other two species. We assumed in these calculations that the percentage of adhesion of the lowactivity competing species remained constant as the cell density increased. Although this assumption was certainly not valid, particularly at higher cell densities, the errors that were introduced were likely to be small. Therefore, our assumption enabled us to determine the effects of competition with the principal, high-activity species. R. flavefaciens had a higher,-xylosidase activity (3.4 nmol/mg [dry weight] per min) than F. succinogenes (.6 nmol/mg [dry weight] per min) and S. ruminantium (.2 nmollmg [dry weight] per min). However, the R.flavefaciens activity was lower than the B. ruminicola activity (8.6 nmol/mg [dry weight] per min), and CMCase was used as a marker for R. flavefaciens (25 nmollmg [dry weight] per min) in the presence of B. ruminicola (.5 nmol/mg [dry weight] per min). Despite severalfold excesses of R. flavefaciens over F. succinogenes (Table 2) and smaller excesses of F. succinogenes over R. flavefaciens (Table 3), no inhibition of F. succinogenes adhesion was caused by R. flavefaciens, and only minor (<2%) inhibition of R. flavefaciens adhesion occurred with F. succinogenes. The noncellulolytic bacteria B. ruminicola and S. ruminantium had no influence on the adhesion of the cellulolytic species (Tables 4 to 7). The level of adhesion for B. ruminicola at its highest cell density was 26%, which was equivalent to an adhesion density of 24 mg (dry weight) per g of straw (Table 5). The same measurements were 33% and 29 mg (dry weight) per g of straw for S. ruminantium (Table 6). DISCUSSION The adhesion of F. succinogenes and R. flavefaciens to barley straw differed from adhesion to cellulose in several ways. Both of these cellulolytic species had maximum affinity for straw at ph 6., in contrast to the results of studies with cellulose in which ph values in the neutral range for the most part had little influence on adhesion (6, 15, 2, 21). Our results were much more similar to the ph profile TABLE 5. Influence of varying the cell density of B. ruminicola on the adhesion of F. succinogenes to barley straw when P-D-lactosidase activity was used as a marker P-D-Lactosidase activity' Cell density (mg [dry wtl per ml) Activity added Activity in filtrate F. succinogenes.b. rumiicola B. rminicola + (mg [dry wt] per g of straw)b added alone F. succinogenes F. succinogenes B. ruminicola F. succinogenes B. ruminicola B. ruminicola B. ruminicola + 1.9 7.3 ±.9 4.2 ±.2 22.6 1.9.6 7. ±.17.3 ±.1.2 ±.1 4. ±.18 24.3 1.9 1.7 6.7 +.12 1.5 ±.5 1.1 ±.1 5.1 ±.16 21.4 1.9 2.8 7.4 ±.5 2.8 ±.8 2.3 ±.11 6.4 +.16 23.7 1.9 3.3 7.1 ±.15 3.4 ±.6 2.5 ±.16 6.9 ±.12 2.2 a Values are the means ± standard deviations for four determinations from two different cultures and are expressed as nanomoles of p-nitrophenol produced b Calculated from the activities measured in filtrates. The activity found with B. ruminicola alone was subtracted from the activity measured in the filtrate of the mixture. The difference was assumed to be due to F. succinogenes, and the quantity of F. succinogenes that adhered to the straw was calculated on this basis

272 BHAT ET AL. APPL. ENVIRON. MICROBIOL. TABLE 6. Cell density (mg [dry wt] per ml) Influence of varying the cell density of S. ruminantium on the adhesion of R. flavefaciens to barley straw when P-D-Xylosidase activity was used as a marker P-D-Xylosidase activitya Activity added Activity in filtrate R. flavefaciens R. S. ruminantium S. ruminantium + (mg [dry wt] per g of straw)b flavefaciens S. ruminantium R. flavefaciens S. ruminantium added adealnr.fvfcis alone flavefaciens 2.2 8.9 ±.12 4.3 ±.15 31.8 2.2.5 9.1 ±.5.2 ±.2.1 ±.2 4.8 ±.14 29.8 2.2 1.6 8.8 ±.17.4 ±.2.2 ±.2 4.5 ±.31 31.5 2.2 2.6 8.5 ±.1.5 ±.2.3 ±.2 4.3 ±.23 32.6 2.2 3.1 9.2 ±.9.6 ±.2.4 ±.2 5. ±.49 3.8 a Values are the means ± standard deviations for four determinations from two different cultures and are expressed as nanomoles of o-nitrophenol produced b Calculated from the activities measured in filtrates. The activity found with S. ruminantium alone was subtracted from the activity measured in the filtrate of the mixture. The difference was assumed to be due to R. flavefaciens, and the quantity of R. flavefaciens that adhered to the straw was calculated on this basis found for adhesion of mixed ruminal bacteria to cellulose (1). Significant adhesion was found down to ph 4.. This capability probably enables cellulolytic bacteria, which cannot grow at ph values of 6. and below (23), to survive during transient decreases in ruminal ph values during the daily feeding cycle. Both of the cellulolytic species were inhibited from adhering by MC and CMC, although the degree of inhibition was less than has been observed previously with cellulose as a substrate (11, 14, 15, 2). It is not clear whether this was because the bacteria were binding to other, noncellulosic sites on the straw or because low-viscosity inhibitors were used (2). Ruminococci in particular are xylanolytic (4) and might be expected to adhere to hemicellulose as well as cellulose. Maximum adhesion was found during mid- to late-exponential phase, as was found previously with the attachment of F. succinogenes to cellulose (6, 14), and experiments were routinely done with cultures in this phase. Old cultures of cellulolytic bacteria tend to release cellulases into the extracellular medium (18), and this release may cause a loss of adhering ability. The methods which we used and which are used in most other studies owe much to the pioneering work of Minato and Suto (14). Optical density was measured after filtration of a straw-bacterium mixture through a sintered glass filter, and the quantity of bacteria that adhered was calculated by subtracting this value from the optical density of the filtered TABLE 7. Cell density (mg [dry wt] per ml) initial suspension, taking into account blanks determined with straw alone and uninoculated medium. Since the substrate was milled straw, entrapment of bacteria may have led to overestimation of adhesion, as pointed out by Rasmussen et al. (2). Alternative methods, such as the use of filter disks (2), are obviously not available for straw. Nevertheless, the finding that adhesion reached a plateau saturation density as cell density increased (Fig. 3) is inconsistent with large nonspecific effects of this type. Kinetic analysis of the type described by Gibbons et al. (5) might be useful in the future for determining the precise degree of specificity and, together with appropriate inhibitors, for resolving the chemical nature of these sites. Possible competition between bacteria for adhesion sites was investigated by using enzyme markers for each species. This method was developed as an alternative to radiolabeling. The enzymes were cell associated, and the activity was measured by using cells that were suspended from the filtrate after centrifugation. Thus, any interference from small amounts of cell-free enzymes was eliminated. A comparison of the data in Fig. 3, which were from an experiment in which adhesion was measured turbidimetrically, with the enzyme-derived data in Tables 2 to 7 shows that similar adhesion densities were obtained with the two methods. Competitive adhesion studies could obviously be done more conveniently with radioactively labeled bacteria, but no suitable labeled compound was found in a short survey of radiolabeled sugars and amino acids (Bhat, unpublished Influence of varying the cell density of B. ruminicola on the initial adhesion of R. flavefaciens to barley straw when CMCase activity was used as a marker CMCase activity' Activity added Activity in filtrate R. flavefaciens B. ruminicola B. ruminicola + (mg [dry wt] per g of straw)b added alone R. flavefaciens R. Rl flavefaciens fe B. ruminicola.me.r R. flavefaciens B. ruminicola adealnr.fvfcin 2.5 61.7 ± 1.6 35. ±.13 3.3 2.5.6 61.3 ± 1.14 33.6 ±.11 31.6 2.5 1.7 64.7 ± 1.55 36.3 ±.2 3.7 2.5 2.7 64.1 ±.79 37.4 ±.8 29.2 2.5 3.3 66.2 ±.65 1.7 ±.8 1.3 ±.6 38. ±.8 31.2 a Values are the means ± standard deviations for four determinations from two different cultures and are expressed as nanomoles of glucose released per minute per milliliter of culture. b Calculated from the activities measured in filtrates. The activity found with B. ruminicola alone was subtracted from the activity measured in the filtrate of the mixture. The difference was assumed to be due to R. flavefaciens, and the percentage of R. flavefaciens that adhered to the straw was calculated on this basis

VOL. 56, 199 ADHESION OF RUMINAL BACTERIA TO BARLEY STRAW 273 data). Clearly, this problem has now been overcome, at least for R. flavefaciens, by the use of ['4C]2-methylbutyrate as a result of the work of Rasmussen et al. (2). Enzyme markers indicated that there was little competition between R. flavefaciens and F. succinogenes for adhesion to straw and that although the noncellulolytic species B. ruminicola and S. ruminantium adhered to a similar extent, adhesion of the cellulolytic bacteria was not inhibited. Latham et al. (13) observed that R. flavefaciens and F. succinogenes appear to adhere at different sites of ryegrass cell walls, and the physical structures that mediate adhesion are quite different (3, 11, 12). Thus, our experiments provided quantitative confirmation that the bacteria adhere noncompetitively at different sites on straw and that noncellulolytic bacteria do not obstruct adhesion of cellulolytic bacteria. This should be confirmed with other species, however, particularly those such as Butyrivibriofibrisolvens, Megasphaera elsdenii, and Veillonella spp., which are known also to attach to cellulose (14). What seems certain is that attachment is largely a specific process and not simply a hydrophobic interaction between cellulolytic bacteria and the substrate (6, 15, 16). Recently, it was reported that adhesion of F. succinogenes was inhibited by the ATPase inhibitor N,N'-dicyclohexylcarbodiimide and by the ionophore lasalocid (21). It is unclear at present why adhesion of F. succinogenes should be energy linked. Adhesion of R. flavefaciens was unaffected by these inhibitors (21). However, the saturation capacity of the substrate seems to have little to do with its ultimate degradability (2). The quantity of bacteria that attach to straw initially is a small proportion of the final population that develops during digestion (22), and it is probably the development of the latter population that is critical in determining the rate and extent of degradation. LITERATURE CITED 1. Amos, H. E., and D. E. Akin. 1978. Rumen protozoal degradation of structurally intact forage tissues. Appl. Environ. Microbiol. 36:513-522. 2. Bhat, S., R. J. Wallace, and E. R. Orskov. 1988. Study of the relation between straw quality and its colonisation by rumen micro-organisms. J. Agric. Sci. 11:561-565. 3. Cheng, K.-J., C. S. Stewart, D. Dinsdale, and J. W. Costerton. 1983. Electron microscopy of bacteria involved in the digestion of plant cell walls. Anim. Feed Sci. Technol. 1:93-12. 4. Chesson, A., and C. W. Forsberg. 1988. Polysaccharide degradation by rumen microorganisms. p. 251-284. In P. N. Hobson (ed.), The rumen microbial ecosystem. Elsevier, London. 5. Gibbons, R. J., E. C. Moreno, and D. M. Spinell. 1976. Model delineating the effects of a salivary pellicle on the adsorption of Streptococcus mitior onto hydroxyapatite. Infect. Immun. 14: 119-1112. 6. Gong, J., and C. W. Forsberg. 1989. Factors affecting adhesion of Fibrobacter succinogenes subsp. succinogenes S85 and adherence-defective mutants to cellulose. Appl. Environ. Microbiol. 55:339-344. 7. Gordon, A. H., J. A. Lomax, and A. Chesson. 1983. Glycosidic linkages of legume, grass and cereal straw cell walls before and after extensive degradation by rumen microorganisms. J. Sci. Food Agric. 34:1341-135. 8. Groleau, D., and C. W. Forsberg. 1983. Partial characterization of the extracellular carboxymethylcellulase activity produced by the rumen bacterium Bacteroides succinogenes. Can. J. Microbiol. 29:54-517. 9. Hobson, P. N. 1969. Rumen bacteria. Methods Microbiol. 3B: 133-149. 1. Kopecny, J., J. F. Jurcuk, and S. Bartos. 1983. The effect of ph and 1,4-dithiothreitol on the adhesion of rumen bacteria. Folia Microbiol. 28:13-133. 11. Kudo, H., K.-J. Cheng, and J. W. Costerton. 1987. Electron microscopic study of the methylcellulose-mediated detachment of cellulolytic rumen bacteria from cellulose fibers. Can. J. Microbiol. 33:267-272. 12. Latham, M. J., B. E. Brooker, G. L. Pettipher, and P. J. Harris. 1978. Ruminococcus flavefaciens cell coat and adhesion to cotton cellulose and to cell walls in leaves of perennial ryegrass (Lolium perenne). Appl. Environ. Microbiol. 35:156-165. 13. Latham, M. J., B. E. Brooker, G. L. Pettipher, and P. J. Harris. 1978. Adhesion of Bacteroides succinogenes in pure culture and in the presence of Ruminococcus flavefaciens to cell walls in leaves of perennial ryegrass (Lolium perenne). Appl. Environ. Microbiol. 35:1166-1173. 14. Minato, H., and T. Suto. 1978. Technique for fractionation of bacteria in rumen microbial ecosystem. II. Attachment of bacteria isolated from bovine rumen to cellulose powder in vitro and elution of bacteria attached therefrom. J. Gen. Appl. Microbiol. 24:1-16. 15. Morris, E. J. 1988. Characteristics of the adhesion of Ruminococcus albus to cellulose. FEMS Microbiol. Lett. 51:113-118. 16. Morris, E. J., and. J. Cole. 1987. Relationship between cellulolytic activity and adhesion to cellulose in Ruminococcus albus. J. Gen. Microbiol. 133:123-132. 17. Patterson, H., R. Irvin, J. W. Costerton, and K.-J. Cheng. 1975. Ultrastructure and adhesion properties of Ruminococcus albus. J. Bacteriol. 122:278-287. 18. Pettipher, G. L., and M. J. Latham. 1979. Characteristics of enzymes produced by Ruminococcus albus which degrade plant cell walls. J. Gen. Microbiol. 11:21-27. 19. Rasmussen, M. A., R. B. Hespell, B. A. White, and R. J. Bothast. 1988. The inhibitory effects of methylcellulose on cellulose degradation by Ruminococcus flavefaciens. Appl. Environ. Microbiol. 54:89-897. 2. Rasmussen, M. A., B. A. White, and R. B. Hespell. 1988. Improved assay for quantitating adherence of ruminal bacteria to cellulose. Appl. Environ. Microbiol. 55:289-291. 21. Roger, V., G. Fonty, and P. Gouet. 1989. Adhesion of rumen cellulolytic bacteria to cellulose: a different mechanism for Bacteroides succinogenes S85 and for Rutninococcus flavefaciens 7? Asian-Aust. J. Anim. Sci. 2:462-464. 22. Silva, A. T., R. J. Wallace, and E. R. rskov. 1987. Use of particle-bound microbial enzyme activity to predict the rate and extent of fibre degradation in the rumen. Br. J. Nutr. 57:47-415. 23. Stewart, C. S. 1977. Factors affecting the cellulolytic activity of rumen contents. Appl. Environ. Microbiol. 33:497-52. 24. White, B. A., M. A. Rasmussen, and R. M. Gardner. 1988. Methylcellulose inhibition of exo-p-1,4-glucanase A from Ruminococcus flavefaciens FD-1. Appl. Environ. Microbiol. 54: 1634-1636. 25. Windham, W. R., and D. E. Akin. 1984. Rumen fungi and forage fiber degradation. Appl. Environ. Microbiol. 48:473-476.