Degradation of Cattle Waste. RODERICK I. ... energy is retained in the methane produced, with a relatively low yield of .... five, volume turnovers was allowed for stabilization after ... production rate for CH4 was then calculated from the.
APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Jan. 1990, p. 87-92 0099-2240/90/010087-06$02.00/0 Copyright © 1990, American Society for Microbiology
Vol. 56, No. 1
Efficiency of Bacterial Protein Synthesis during Anaerobic Degradation of Cattle Waste MACKIE* AND MARVIN P. BRYANT Department of Animal Sciences, University of Illinois, Urbana, Illinois 61801 RODERICK
I.
Received 10 July 1989/Accepted 23 October 1989
The rate of ['5N]ammonia ('5NH3) uptake or incorporation into bacterial cells was studied, using stirred, 3-liter benchtop digestors fed on a semicontinuous basis with cattle waste. The fermentations were carried out at 40 and 60°C and at four different loading rates (3, 6, 9, and 12 g of volatile solids per liter of reactor volume per day). The rate of NH3-N incorporation for the period 1 to 5 h after feeding at the four different loading rates was 0.49, 0.83, 1.05, and 1.08 mg/liter per h in the mesophilic digestor and 0.68, 1.07, 1.17, and 1.21 mg/liter per h in the thermophilic digestor. Values were lower 7 to 21 h after feeding in both digestors and were related to the rate of fermentation or CH4 production. In the mesophilic digestors, the rate of bacterial cell production ranged from 3.97 to 8.72 mg of dry cells per liter per h, 1 to 5 h after feeding at the different loading rates. Corresponding values for the thermophilic digestors ranged from 5.46 to 9.77 mg of dry cells per liter per h. Cell yield values ranged from 2.3 to 3.1 mg of dry cells per mol of CH4 produced in the mesophilic and thermophilic digestors at the two lower loading rates. The values were higher (2.8 to 3.4) in the mesophilic digestors at the two higher loading rates because of the accumulation of propionate and a consequent reduction in CH4 production. Low cell yields such as those measured in this study are characteristic of lowspecific-growth rates under energy-limited conditions.
acids, and NH3 can then be resynthesized into microbial protein (1, 6). In the rumen, ammonia is the major end product of the digestion of dietary protein and nonprotein nitrogen. A large portion (60 to 92%) of the dietary N is converted to ammonia-N in the rumen (16, 20, 21). Many ruminal bacteria, and the cellulolytic bacteria in particular, utilize ammonia-N in preference to amino acid-N; indeed, ammonia is an absolute growth requirement for some species (8, 9). About 60 to 80% of the microbial-N is derived from ammonia-N in the rumen (16, 20, 21). The major difference between the rumen and digestors appears to be the bacteria involved. In the rumen, proteins are degraded by carbohydrate-fermenting species, and the fermentation of peptides or amino acids does not provide sufficient energy for growth, whereas in anaerobic digestors, specialized bacteria such as the proteolytic clostridia are important in the energy-yielding process (18). The use of a "NH3-N tracer technique to measure microbial protein synthesis in anaerobic digestors is logical in view of the relatively high NH3-N pool sizes and its obvious importance as the major precursor for the synthesis of bacterial protein. In this paper, the term ammonia (NH3) is used to refer to the compound without defining its state of protonation. This paper presents the results of experiments in which the rate of uptake or incorporation of ('5NH4)2SO4 in stirred, benchtop fermentors fed on a semicontinuous basis with cattle waste was measured. The fermentations were carried out in mesophilic and thermophilic digestors at different loading rates.
During methanogenesis in anaerobic environments, degradable organic matter is converted to CH4 and CO2. During this anaerobic process, about 90% of the available energy is retained in the methane produced, with a relatively low yield of microbial cells (17). The relatively small release of energy must be divided and distributed among the different bacteria involved in the sequential process of anaerobic degradation (7, 19). Most of the information on anaerobic degradation comes from studies on ruminal fermentation, bacterial enrichment cultures, and fermentation of sewage sludge. Very little research has been performed on methane fermentation of animal wastes, especially with respect to nitrogen metabolism and the growth of bacteria in this type of ecosystem. The substrate or waste used as feedstock contains large numbers of bacteria, some of which are dead or nonviable and others which have been growing either in the body of the host animal or in collection tanks and are likely to grow in the digestor. However, it is difficult to assess how many of these viable bacteria actually grow in the digestor and what proportion is metabolically inactive and can be presumed dead. Several studies have analyzed input and output samples from digestors and demonstrated little change in crude protein (CP = [N x 6.25]) during digestion (13, 14). However, it is difficult to distinguish between the microbial proteins and other proteins in substrate and effluent, and it is therefore not possible to assess whether any degradation of nonmicrobial protein has occurred. The presence of proteolytic bacteria in fermentors (24, 28) suggests that the degradation of substrate protein is balanced by the utilization of the hydrolytic products for the synthesis of bacterial protein. There seems to be little doubt that protein degradation in digestors follows a similar pattern to that in the rumen, where proteins are hydrolyzed to peptides and amino acids, which are subsequently deaminated, with the production of NH3, C02, and volatile fatty acids (1, 6). Peptides, amino *
MATERIALS AND METHODS Digestors. Microferm benchtop fermentors (5 liter; New Brunswick Scientific Co., New Brunswick, N.J.) were used, each with a 3-liter working volume. One fermentor was maintained at 40°C for the mesophilic experiments, and the other was maintained at 60°C for the thermophilic experiments. Their design and operation have been described elsewhere (15).
Corresponding author. 87
88
MACKIE AND BRYANT
APPL. ENVIRON. MICROBIOL.
TABLE 1. Loading rates used in experiments with the mesophilic and thermophilic digestors Period no.
RT (days)
vol (mI/day)
VS in feed (%)
1 2 3 4
13 10 9 5
231 300 333 600
4 6 8 6
Effluenta
V
Loading rate (g of
per day)
per day)
3 6 9 12
a 3-liter working volume.
Substrate and loading rate. The collection and preparation of the cattle waste (feces and urine without bedding) used as substrate have been described previously (15). Four different loading rates were employed in the experiments and each digestor, starting from the lowest and progressing to the highest loading rate (Table 1). It should be noted that retention time (RT) decreased as loading rate increased. The digestors were fed once daily except at the 5-day RT, when they were fed every 12 h. A minimum of three, and usually five, volume turnovers was allowed for stabilization after each increase in loading rate. Infusion experiments with "5N. Sterile infusion solution in a 1-liter flask was infused at a constant measured rate into the fermentor vessel, using a peristaltic pump (model 600-1200; Harvard Apparatus Co., Dover, Mass.). The infusion solution was prepared by using 0.9 g of ('5NH4)2SO4 made up in 1 liter and sterilized by membrane filtration (0.2-,m pore size). This volume of infusion solution was sufficient for the duplicate experiments at each loading rate in the two digestors. The infusion was started 8 to 10 h before the daily batch feeding to allow "5N enrichment to reach equilibrium and then continued until the next feeding. Infusion of (15NH4)2SO4 had a negligible effect on NH3 pool size. The rate of NH3-N incorporation into bacterial cells was calculated as the infusion rate (milligrams of 15N per hour) divided by the specific activity of 15N (percent 15N excess) in the
NH3-N pool (2, 3). Serial samples (20 ml) of digestor contents were removed at different times before and after feeding by means of a 35-ml syringe with a 6-mm nozzle attached, by using butyl rubber tubing, to the harvest line of the fermentor. Care was taken to mix the fluid remaining in the harvest line with that in the rest of the digestor before removing a sample and adding it to a glass bottle containing 5.0 ml of 1 N H2SO4. A volume (5.0 ml) of 10% sodium tungstate was added to the sample bottle, and the contents were mixed gently and allowed to stand overnight. The samples were centrifuged (20,000 x g, 30 min, 4°C), and the supernatant was stored in rubber stoppered glass bottles at 4°C until required for analysis. Fermentor gas was sampled as described previously (15). Analytical procedures. Ammonia nitrogen concentration was determined, using the phenol-hypochlorite method (10). For "5N analyses, ammonia was liberated from acid-treated samples by the addition of excess alkali and steam distillation in a Markham apparatus. Ammonia was trapped in HCI, concentrated, and transferred to tubes. Cross-contamination was controlled by distilling 20 ml of ethanol between ammonia samples. The ammonia samples were oxidized by the Kjeldhal-Rittenberg procedure, using alkaline sodium hypobromite under vacuum on the mass spectrometer inlet system. 15N enrichment was analyzed by mass spectrometer (Nuclide 3-60 RMS; Nuclide Corp., State College, Pa.). The instrument was checked, using a series of standards (0.37 to
5.00% '5N) prepared by mixing known amounts of enriched (15NH4)2SO4 (99.75% 15N) with unenriched (NH4)2SO4 (0.37% 15N). The percent 15N in the samples was calculated from the ratio of 15N in pure N2 gas to that in the unknown sample analyzed simultaneously in the mass spectrometer. Gas composition was analyzed with an Aerograph A100 gas chromatograph (Wilkens Instrument & Research, Berkeley, Calif.), using a stainless steel column (1.52 m by 6.35 mm) packed with silica gel and connected to a thermal conductivity detector (bridge current, 200 mA) and recorder. Helium was used as the carrier gas (flow rate, 60 ml/min). Standard procedures (4) were used to determine total solids and volatile solids (VS). Chemicals. (15NH4)2SO4 (99.75 atom% '5N) was obtained from the Isotope Labelling Corporation, Whippany, N.J. All other chemicals were reagent grade and were obtained from commercial sources. RESULTS Total gas and methane production in digestors. Cumulative gas production was recorded at different times over the 24-h cycle, except in the case of twice daily feeding at the 5-day RT when it was recorded over a 12-h period. Both digestors were fed identical substrate at the four different loading rates. Total gas production over the 24-h period increased with each increase in loading rate and was higher in the thermophilic than in the mesophilic digestor. The gas production rates were calculated for each experiment by drawing a tangent to the total gas production curve at each of the sampling times to calculate the total gas production rate. The production rate for CH4 was then calculated from the percentage of CH4 in headspace gas of the fermentor vessel at each of the sampling times. The results are presented in Fig. 1. At each loading rate, the CH4 production rate increased rapidly, reaching a peak at the first sampling 1 h after feeding, and then gradually decreased to the prefeeding production rate. The rates were always higher in the thermophilic than in the mesophilic digestor. The rates also increased with each increase in loading rate, except in the case of the mesophilic digestor when the peak CH4 production rate was actually lower at a loading rate of 12 g of VS per liter of reactor volume per day (3.05 mM/h) than at a loading rate of 9 g of VS per liter of reactor volume per day (3.34 mM/h). Gas production rates did not decrease as fast after the peak in the thermophilic digestor compared with that of the mesophilic digestor. Ammonia concentration. The '5N tracer technique for determining the rate of '5NH3-N uptake from the NH3-N pool or the rate at which it is incorporated into microbial cell material is dependent on the existence of steady-state conditions with respect to the NH3-N pool size. If steady conditions exist in the digestor, which can be considered a closed compartment, the rate of ammonia production is equal to the rate of utilization. The results (Table 2) show that the concentration of ammonia was slightly higher in the period 1 to 5 h after feeding than in the 7- to 21-h period. The increase in pool size 1 to 5 h after feeding was 3.6 to 5.4% in the thermophilic and 4.7% in the mesophilic digestor at the loading rates of 3 and 6 g of VS per liter of reactor volume. At the two higher loading rates, pool size increased 7.6 to 9.3% 1 to 5 h after feeding in both digestors. This slight increase in pool size can be ascribed to the presence of both ammonia (0.52% of total solids) and nonammonia crude protein (10.0% of total solids) in the substrate. Ammonia concentration was higher in the mesophilic than in the thermophilic digestor at the same loading rate. This
BACTERIAL GROWTH DURING ANAEROBIC DIGESTION
VOL. 56, 1990 a
-r..
___
89
digestor for the period 1 to 5 h after feeding were 0.68, 1.07,
1.17, and 1.21 mg/liter per h. / (=- oFractional turnover rate or the fraction of the NH3-N pool turning over or being renewed per hour was calculated by the rate at which NH3-N was being removed from dividing 9zX _1 ~ / 1 o the pool by the pool size. In the mesophilic digestor, these values were 0.0007 to 0.0009/h for the 1- to 5-h period and 1 0.0005 to 0.0006/h for the 7- to 12-h period after feeding. In 0 the thermophilic digestor, the values were 0.0008 to 0.0012/h , _ , i 4 4 is 20 and 0.0006 to 0.0008/h. Turnover rate or the time taken for 0 18 N o_ renewal of the quantity of NH3-N equal to that present in the 5 B NH3-H pool was calculated as the reciprocal of the fractional turnover rate. These values ranged from 58 to 75 days in the >42 mesophilic digestor and from 44 to 62 days in the thermo.0 aI philic digestor. r4) 2 Rate of bacterial protein synthesis. The rate of bacterial protein synthesis from NH3-N (Table 3) was estimated from 0 1the rate of '5NH3-N incorporation multiplied by 6.25 and 0 .________________________ that 85% of the N is incorporated into amino acids , . .assuming O N 4 4 and therefore into bacterial protein. In the thermophilic 0 a 12 Is n o_ 54 digestor, the rates of bacterial protein synthesis were 3.59, p4 6 5.47, 6.18, and 6.43 mg/liter per h for the 1- to 5-h period C after feeding 3, 6, 9, and 12 g of VS per liter of reactor 4 volume per day. In the mesophilic digestor, the rates were 4.4) 9^-0_ 2.61, 4.41, 5.58, and 5.74 mg/liter per h. Values for the 7- to 21-h period were lower in both digestors and followed the ; *-~~same trend as the rate of NH3-N incorporation. Bacterial cell production from NH3-N was estimated, a value of 10.6% N (66% crude protein) in bacterial dry _____,___,____.___ .__ using o4-4 o '4'-'matter (25). The rate of bacterial cell production (Table 3) 0 4 12 20 S 16 2 was 3.97, 6.70, 8.48, and 8.72 mg/liter per h in the mesophilic digestor and 5.46, 8.31, 9.39, and 9.77 mg/liter per h in the ffi D thermophilic digestor for the 1- to 5-h period after feeding 3, D 6, 9, and 12 g of VS per liter of reactor volume per day. Values for the 7- to 21-h period were 2.52, 4.03, 5.73, and 35.87 mg/liter per h in the mesophilic digestor and 3.21, 4.89, 2 6.08, and 6.48 mg/liter per h in the thermophilic digestor. The increase in bacterial cell production decreased with each increase in loading rate. The largest increase occurred when - kl -= _ the loading rate was increased from 3 to 6 g of VS per liter of O tI. , , -4 16 O is8 18 X0 reactor volume per day. Efficiency of bacterial cell production. Cell yield was calHours culated, using the values for CH4 production rate (Fig. 1). Cell yields (Table 3) were low, ranging from 2.3 to 3.4 and FIG. 1. Methane production rates in the mesophilic (0) and 3.1 g of dry bacterial cells per mol of CH4 in the thermopohilic (0) digestors at loading rates of 12 (A), 9 (B), 6 (C), and from 2.0 to and mesophilic thermophilic digestors, respectively. In gen3 (D) g cof VS per liter of reactor volume per day. eral, the cell yields were higher 7 to 21 h after feeding than 1 to 5 h after feeding. The differences between the two time periods were smaller in the thermophilic digestor. For the differenice was not significant at the two lower loading rates, two lower loading rates (3 and 6 g of VS per liter of reactor but at tihe loading rates of 9 and 12 g of VS per liter of reactor volume per day), cell yields were higher in the thermophilic volume per day, the differences were much greater and than in the mesophilic digestor, while for the two higher rates the reverse was found. were, oon average, 150 to 200 mg/liter higher in the mesophilic than in the thermophilic digestor. Rate of NH3-N incorporation. The results are presented in DISCUSSION Table 2. The rate of NH3-N incorporation followed the same trends as the methane production or fermentation rate. The syntheses of macromolecules such as protein, DNA, Values were higher at 1 to 5 h than at 7 to 21 h after feeding or RNA have not been used in the investigation of in situ in both digestors and were higher in the thermophilic than in growth rates of bacterial populations in anaerobic digestors. the mesophilic digestor at each of the loading rates. Thus, A strong correlation exists between the rate of protein the rate of NH3-N incorporation into bacterial cells in the synthesis and cell growth, and the use of '5NH3-N is an ideal mesophilic digestor was 0.49, 0.83, 1.05, and 1.08 mg/liter method for analyzing the growth of complex microbial populations found in nature. The method is based on three per h for the period 1 to 5 h after feeding 3, 6, 9, and 12 g of VS per liter of reactor volume per day, respectively. The assumptions. (i) All, or most, of the bacteria in the digestor values were correspondingly lower for the period 7 to 21 h can utilize exogenous NH3-N. (ii) No increase in the NH3-N after feeding. Corresponding values in the thermophilic pool occurs with the addition of exogenous NH3-N. (iii) No 5
4-
A A
/ (-o
r
_,_.
APPL. ENVIRON. MICROBIOL.
MACKIE AND BRYANT
90
TABLE 2. Rate of incorporation into bacteria, pool size, fractional turnover rate, and turnover time of ammonia nitrogen in mesophilic and thermophilic digestors at different loading rates and at various times after feeding Digestor and loading reactor vol)
reactor vol) Mesophilic 3 13
6
10
9
9
12
5
Rate of NH3-N'
incorporation
NH3-N pool size (mg/liter)
1-5 7-21 1-5 7-21 1-5 7-21 1-2.5 5-12
0.49 0.31 0.83 0.50 1.05 0.71 1.08 0.73
580 560 1095 1036 1614 1476 1266 1170
1-5 7-21 1-5 7-21 1-5 7-21 1-2.5 5-12
0.68 0.40 1.07 0.61 1.17 0.76 1.21 0.80
Period after feeding (h)
RT (days)
~~~~~~~~~(mg/liter per h)
Fractional'
Turnover' time (days)
0.0008
62.3
turnover rate (per h)
0.0006 0.0008 0.0005 0.0007 0.0005
70.7
75.3
0.0009 0.0006
57.8
556 0.0012 531 0.0008 6 10 1067 0.0010 1019 0.0006 9 9 1428 0.0008 1327 0.0006 12 5 1053 0.0012 976 0.0008 a Calculated as infusion rate (milligrams of 15N per hour) divided by specific activity of '5N (per cent 15N excess) in NH3-N pool.
44.7
Thermophilic 3
13
55.6 61.9 43.6
b
Rate of the NH3-N incorporation divided by NH3-N pool size. c Calculated as the reciprocal of fractional turnover rate.
intracellular compartmentation of the NH3-N pool occurs. The ability of ruminal bacteria to utilize NH3-N has been well documented (8, 9, 16, 20, 21). It is likely that NH3-N is even more important as a starting point for bacterial protein synthesis in anaerobic digestors. However, there is no specific information on this aspect of N metabolism in bacteria isolated from anaerobic digestors. The infusion of (15NH4)2SO4 into the digestors had a negligible effect on NH3-N pool size because of the large pool size of NH3-N in the digestors, especially at the higher loading rates. The NH3-N pool size was slightly increased by the addition of
daily substrate containing some ammonia and nonammonia crude protein. The final assumption refers to the existence of metabolite pools within the bacterial cells that do not mix freely because of membrane or kinetic processes. Although compartmentalization of certain metabolite pools does exist, it is assumed that these pools are in rapid equilibrium with, and have specific activities identical to, the 15NH3-N precursor pool and, indeed, with the extracellular 15NH3-N pool. Thus, the rate at which '5NH3-N is removed from the NH3-N pool is the same as the rate at which it is incorporated into bacterial cell material.
TABLE 3. Rate and efficiency of bacterial protein synthesis in mesophilic and thermophilic digestors at different loading rates and at various times after feeding Digestor and loading rate (g of VS/liter of reactor vol)
RT (days)
Period aft feeding (h)
Rate of NH3-N incorporation (mg/liter per h)
Rate of bacterial" protein synthesis (mg/liter per h)
Rate of bacterial' cell production (mg/liter per h)
Cell yield' (g of cell/mol of CH4)
1-5 7-21 1-5 7-21 1-5 7-21 1-2.5 5-12
0.49 0.31 0.83 0.50 1.05 0.71 1.08 0.73
2.61 1.66 4.41 2.65 5.58 3.77 5.74 3.86
3.97 2.52 6.70 4.03 8.48 5.73 8.72 5.87
2.31 3.00 2.42 2.41 2.75 3.39 3.13 3.23
1-5 7-21 1-5 7-21 1-5 7-21 1-2.5 5-12
0.68 0.40 1.07 0.61 1.17 0.76 1.21 0.80
3.59 2.11 5.47 3.22 6.18 4.00 6.43 4.26
5.46 3.21 8.31 4.89 9.39 6.08 9.77 6.48
2.80 3.09 2.58 2.35 1.98 2.17 2.01 1.95
Mesophilic 3
13
6
10
9
9
12
5
Thermophilic 3
13
6
10
9
9
12
5
a Rate of NH3-N incorporation x 6.25 x 0.85. b (Rate of bacterial protein synthesis x 100)/66. '
Rate of bacterial cell production/rate of CH4 production.
VOL. 56,
199
BACTERIAL GROWTH DURING ANAEROBIC DIGESTION
The NH3-N in the digestors can essentially be considered a closed compartment which will be influenced by three factors: conversion of substrate nonammonia crude protein and nonprotein nitrogen to NH3, (ii) NH3 assimilation by bacteria, and (iii) bacterial lysis and degradation of cell constituents. The main disadvantage of the '5NH3-N technique is that it ignores direct incorporation of amino acidand peptide-nitrogen into the bacterial protein and would, therefore, underestimate bacterial protein synthesis by this amount. In the rumen, 20 to 40% of microbial protein is derived from sources other than ammonia (16, 20, 21). This is partly due to the requirement of the ruminal protozoa for preformed protein and nucleic acids (1). Furthermore, amino acids are rapidly fermented to short-chain or branched-chain fatty acids, ammonia, and CO2 (13, 14, 18), and only very low concentrations of amino acids are found in the rumen (6, 12, 16, 21). Thus, in anaerobic digestors which have no protozoa and high levels of NH3-N, it is unlikely that sources other than ammonia or N-containing compounds not equilibrating with the NH3-N pool account for more than a small percentage of the bacterial protein synthesized. There is no information on bacterial lysis and degradation, particularly of the viable bacteria growing in digestors. Any lysis of bacteria which had incorporated '5NH3-N during the infusion period would increase the difference between the total flux and irreversible loss rates. However, in the present experiments, the infusion period was short relative to the turnover rate of NH3-N from the pool and to the growth and death rates of viable bacteria in the digestors. Thus, the recycling rate was probably negligible, and the irreversible loss rate would have been similar to the total flux rate. The rumen is the most studied and best understood natural ecosystem in which anaerobic degradation occurs and provides values for comparison with the digestors. The rate of NH3-N uptake or incorporation in the mesophilic digestor averaged ca. 1 mg/liter per h for the two higher loading rates. A rate of 58 mg/kg of whole ruminal digesta per h was obtained for in vitro incubations with rumen contents obtained from sheep fed alfalfa hay (2, 3). Corresponding values for fractional turnover rate and turnover time were 0.11 to 0.31/h and 3.2 to 9.1/h, respectively, for the ruminal studies (2, 3). The values in the mesophilic digestor were characterized by large pool sizes (1,170 to 1,614 mg/liter) and low fractional turnover rates (0.0005 to 0.0009/h), resulting in long turnover times. The turnover calculated in Table 2 would be turnover due to bacterial uptake alone, since the rates were measured between two successive substrate additions. The dilution or turnover of the NH3-N pool by the removal of daily effluent volume and addition of fresh substrate would, in fact, reduce the turnover time to the retention time, i.e., 13-, 10-, 9-, and 5-day RTs. The rate of bacterial protein synthesis (Table 3) was calculated assuming that 85% of the NH3 incorporated was utilized to synthesize protein and the remaining 15% was utilized to synthesize nucleic acids. The value of 10.6% N (66% crude protein) was obtained from experiments on anaerobic digestion of sewage sludge (25) at RTs similar to those in the present study. This is similar to the commonly used value of 10 to 10.5% N (62.5 to 65.6% crude protein) for ruminal studies (2, 3, 23). The composition of bacterial cells is influenced not only by medium composition but also by growth rate. At low growth rates, bacterial cells have been shown to contain less protein and more nucleic acid (12). However, cell composition has not been studied at growth rates characteristic of anaerobic digestors and sediments. Rates of bacterial cell production ranged from 5.73 to 8.72 (i)
0-
91
0.5 1
C)
0.44
-0 '-
1
ao
0.3 -
0.2 t 0
80
1 60
240
320
1/,u (hours) FIG. 2. Double reciprocal plot of estimated values Of YCH4 for the thermophilic digestor, where values against growth rate for 1 to 5 h (0) and 7 to 21 h (0) have been used to fit a linear regression line to the data points.
(,u)
and from 6.08 to 9.77 mg of dry cells per liter per h in the mesophilic and thermophilic digestors at the highest loading rates. A rate of 534 mg of dry cells per kg of whole ruminal digesta per h was obtained for the rumen (2, 3). The pattern and rate of NH3-N incorporation and bacterial protein synthesis was higher at 1 to5 h than at 7 to 21 h after feeding and was dependent on the rate of fermentation. In anaerobic degradation, the central theme in bacterial metabolism is the fermentation of substrate components in order to obtain energy for maintenance and growth. Baldwin and Denham (5) have demonstrated with modeling techniques that bacteria grow at a set rate in proportion to fermentation rate. Cell yield values ranged from 2.3 to 3.0 and 2.4 to 3.1 g of dry cells per mol of CH4 produced in the mesophilic and thermophilic digestors, respectively, at the two lower loading rates. The values then decreased at the higher loading rates in the thermophilic digestor and increased in the mesophilic digestor. This can be explained by the increasing amounts of volatile fatty acids (mainly propionate) accumulating in the mesophilic digestor, with a resultant decrease in CH4 production and thus a higher yield value. These values are very similar to those reported for a wide range of methanogens growing on acetate (1.4 to 3.8 g [dry weight] per mol of CH4) and H2-C02 (0.6 to 8.7 g/mol) (11, 29). Since the yield value (YCH4, grams of dry cells produced per mole of CH4 produced) is related to the specific growth rate(,u), the maintenance energy requirements of the bacteria (me, moles of CH4 per gram of dry cells synthesized per hour) and the theoretical maximum growth yield at a given maintenance requirement (YCH ) can be expressed by the following equation: = 1/Y4 + me/L.. A plot of1/YCH4 against will have an intercept of1/YCj'a and slope of me(22, 26). Although the present data setis limited, the results from the thermophilic digestor were used to calculate values forYa and me (Fig. 2). Similar estimates from the mesophilic digestor were not calculated because of the change in fermentation pattern at the two higher dilution rates and hence insufficient data points. These coefficients derived from semicontinuous chemostat experiments can be used to predict bacterial growth from the methane yield in digestors fed similar substrates and are of great importance in evaluating and predicting performance. At low specific growth rates, the relative amount of energy for maintenance is higher than that at higher growth rates (26, 27). At the low specific growth rates in the present experiments
