Sequestration, migration and lysis of protozoa in the rumen

Transcription

1 ~ ~~~~~ Journal of General Microbiology (1990), 136, Printed in Great Britain 1869 Sequestration, migration and lysis of protozoa in the rumen P. ANKRAH, S. C. LOERCH" and B. A. DEHORITY Department of Animal Science, Ohio Agricultural Research and Development Center, The Ohio State University, Wooster, OH 44691, USA (Received 21 November 1989; revised 16 March 1990; accepted 16 May 1990) A container with nylon and stainless steel mesh screens (10 pm and 20 pm pores, respectively), which allowed the passage of bacteria and soluble substrate but restricted the passage of protozoa, was used to investigate the phenomena of sequestration, migration and lysis of protozoa in the rumen of steers. After feeding, the concentration of Isotrichidae in the rumen increased 8-7-fold within 40 min and then decreased 88% by 4 h; however, the concentration of Isotrichidae inside the containers remained almost constant. Fluctuations in concentrations of Isotrichidae were shown to be due to migration and sequestration within the rumen. Ophryoscolecidae did not exhibit the phenomena of sequestration and migration. When steers were fed once a day, about 50% of the decrease in the concentration of Ophryoscolecidae immediately after feeding could be attributed to the dilution effects of feed and water and/or passage out of the rumen. The remaining 50% of the decrease appeared to be from cell lysis, resulting in a wasteful recycling of protozoal protein. In an experiment to determine the effects of feed restriction on concentrations of Ophryoscolecidae, it was shown that decreased concentrations associated with substrage restriction are due to cell lysis rather than to passage out of the rumen. Introduction Ciliate protozoa appear to wash out of the rumen at a much slower rate than bacteria (Hungate et al., 1971; Weller & Pilgrim, 1974; Punia et al., 1987). Protozoa may avoid passing out of the rumen by sequestering, but quantitative experimental evidence to support this theory is limited (Bauchop & Clarke, 1976; Orpin & Letcher, 1978; Abe et al., 1981 ; Demeyer, 1981). Abe et al. (1981) observed a thick protozoal mass on the wall of the reticulum of steers slaughtered after an overnight fast and, based on this and counts of protozoa in the rumen, suggested that Isotrichidae ordinarily sequester on the wall of the reticulum with subsequent chemotactic migration into the rumen shortly after feeding. The diurnal variations observed in the numbers of Isotrichidae in the rumen support the sequestration hypothesis (Purser, 1961 ; Warner, 1962; Clarke, 1965; Dehority & Mattos, 1978; Abe et al., 1981; Dehority & Tirabasso, 1989). An incubation container with screens large enough to be permeable to rumen bacteria but small enough to restrict influx of protozoa should allow the study of protozoal sequestration in the rumen by permitting simultaneous monitoring of numbers of protozoa inside and outside the container when it is with rumen contents and placed in the rumen. Because the container would not provide any niche for sequestration of protozoa, differences in the numbers of protozoa between the container and the rumen should provide information on the sequestration and migration of the protozoa. The present study was undertaken to evaluate these hypotheses. Methods Selective container. The selective container consisted of a plastic (Enfield Plastics, UK) cylinder (7 x 6 x 7 cm; 6-5 cm i.d.) with screwon caps at both ends and a spout at the side of the cylinder (Figs 1 and 2). The similar to that developed by Fina et al. (1962). Three screens were used. The inner screen was a twill nylon mesh (Tetko Inc., USA) with lopm pores and the middle screen was a stainless steel wire mesh (Tetko) with 20 pm pores. The outer screen was a fibreglass screen with 2 mm pores. The middle and outer screens were used to protect the finer inner screen from rupture and to prevent clogging of the inner screen with particulate matter. The containers were assembled according to the sequence shown in Fig. 1. A fairly rigid piece of rubber tubing (9 mm id.) was connected to the spout of the container and the other end of the tubing was inserted through a hole in the rubber cap of the rumen cannula. The tubing was stoppered to prevent influx of fluid from the container into the tube. The length of the tubing was such that when the placed in the rumen, it rested on the bottom of the rumen cavity. In addition, the weighted by tying a steel weight around the spout to ensure that it remained in a consistent location at the bottom of the rumen. The purpose of the tubing was to hold the containers in place in the rumen and to aid in their retrieval. Containers were sampled after retrieval by unscrewing an end-cap (Fig. 1) SGM

2 1870 P. Ankrah, S. C. Loerch and B. A. Dehority Fig. I. Components of the container and their sequence of assembly. From left to right: screw-on cap, 2 mm pore fibreglass screen, glass wool, neoprene sealing ring, 20 pm pore stainless steel wire mesh, 10 pm pore nylon mesh, neoprene sealing ring, tapered sealing ring and container with an exposed end. Measurements of sequestration and migration. To determine the extent of sequestration in the rumen shortly after feeding, differences in the concentrations of protozoa between the rumen and containers were measured 4 h after feeding. Containers were (1 50 ml) with mixed rumen contents 40 min after feeding and were incubated until 4 h after feeding. ph and numbers of protozoa were determined at 0 h (before feeding), 40 min and 4 h after feeding (in the rumen) and 4 h after feeding (in the containers). Measurements of migration into the rumen digesta in response to feeding were made by filling the containers with rumen contents 6 h after feeding and comparing concentrations of protozoa in the rumen and containers 24 h 40 rnin after the previous days feeding. ph and numbers of protozoa were measured 6 h, 24 h and 24 h 40 rnin after feeding in the rumen and 24 h 40 rnin after feeding in the containers. Fig. 2. Assembled containers ready for incubation in the rumen. Evaluation of container design. The extent of influx of protozoa from the rumen into the containers was measured during a 24 h incubation. Rumen contents were obtained from a ruminally cannulated steer just before feeding (0 h), and the numbers of protozoa (Dehority, 1984) and the ph were measured. The steer was then fed 5.4 kg of alfalfa hay. Two containers, each containing 10 g ground (2 mm screen) alfalfa hay and 100 ml water, were assembled as previously described and placed in the rumen while the steer was eating. After 24 h incubation, the containers were removed from the rumen and a sample of mixed rumen contents was obtained from a location adjacent to the containers. ph and numbers of protozoa were determined on the rumen sample and the contents of the containers. Experimental design. In all experiments, a ruminally cannulated steer was fed 5.4 kg of chopped-first cutting alfalfa hay once per day at 0800 h. It took approximately 40 rnin for the steer to consume the ration. Two containers were incubated simultaneously in each experiment and each with rumen contents (150 ml) prior to being incubated in the rumen for a specified period of time. Each experiment was repeated twice. ph and numbers of protozoa were determined for all samples. The experimental design for these experiments was a randomized complete block with repetitions as blocks. Data from these experiments were analysed using ANOVA, with treatment means compared by the Multiple Comparison Method of Least Significant Difference protected by a significant (P < 0.10) F value (Steel & Torrie, 1960). Measurement of growth and division of protozoa. Changes in numbers of protozoa as a result of growth and division were determined when sequestration and migration associated with feeding were minimal (Dehority & Tirabasso, 1989). Containers were 6 h after feeding; concentrations of protozoa and ph were measured at this time and when containers were removed 24 h after feeding. Measurements of lysis of protozoa. Concentrations of protozoa in the rumen are at their lowest approximately 6-8 h after feeding (Purser & Moir, 1959; Warner, 1966; Potter & Dehority, 1973). The contribution of lysis to this decline was determined by filling the containers at 0 h (before feeding) and measuring differences in concentrations of protozoa between the rumen and containers after 6 h incubation. The extent of lysis resulting from substrate limitation was determined by filling the containers 24 h after feeding and comparing concentrations of protozoa in the rumen and the containers 48 h after the steer had been fed. Terminology. Based on the classification scheme of Levine et al. (1980), the rumen protozoa encountered in this study are all in two families. Those protozoa in the genera Isotricha and Dasytricha are listed under the family Isotrichidae. All other genera (Entodinium, Diplodinium, Eudiplodinium, Polyplastron, Epidinium etc.) are listed under the family Ophryoscolecidae. Results and Discussion Evaluation of the container design In a preliminary study with inner screens having 10,15 or 20 pm pores, protozoa were observed in the containers after 8 h incubation in the rumen. Although most of the

3 Sequestration, migration and lysis of rumen protozoa rumen ciliate protozoal species are greater than 20 pm in width (Hungate, 1966), both large and small protozoa were found in the containers when the different screen sizes were used. Microscopic observation of rumen contents placed on a 15 pm screen indicated that protozoa were able to pass through the holes in the screens. Because the numbers and sizes of protozoa observed in the container with a 10 pm inner screen were greatly reduced compared to the other screen sizes (data not shown) this screen was used as the inner screen and the 20 pm and 2 mm screens were used as the middle and outer screens, respectively, in all subsequent studies. The screens used in these studies were quite effective in restricting movement of protozoa into the containers. Following a 24 h incubation, the total concentration of protozoa in non-inoculated containers was only 3% of that in the rumen (0.6 vs 20.6 x lo4 ml-l, respectively). Ophryoscolecidae constituted 67 % of the protozoa found in the containers, the remaining 33 % being Isotrichidae. The concentration of protozoa measured following the 24 h incubation is possibly a slight overestimate because it also includes growth and multiplication of protozoa within the containers. The concentration of Ophryoscolecidae in the containers was only 2.1% of that in the rumen (0.4 vs 19.4 x lo4 ml-l, respectively), while the concentration of Isotrichidae in the containers was 15.6% of that found in the rumen (0.2 vs 1.3 x lo4 ml-', respectively). Apparently, the screens were more successful in impeding the influx of Ophryoscolecidae than that of Isotrichidae. The rumen ph was 7.08 while the ph in the Protozoa in the rumen are usually larger than 20 pm (Hungate, 1966); however, passage of protozoa through small mesh fabric has been reported by Jouany & Senaud (1979), Lindberg et al. (1984) and Meyer & Mackie (1986). Although screen sizes less than 10pm might be capable of completely excluding protozoa, they may also greatly inhibit or retard exchange of rumen fluid and bacteria (Lindberg et al., 1984). Sequestration and migration of protozoa Sequestration was measured by comparing protozoal concentrations in the rumen and in the containers before and after feeding (Table 1). The concentration of total rumen protozoa was higher 40 min after feeding but fell back to prefeeding levels by 4 h. This increase at feeding time and subsequent sharp drop was primarily due to changes in the concentration of Isotrichidae. From before feeding to 40min after feeding, the concentration of Isotrichidae in the rumen increased by 8-7 fold, but then dropped to previous feeding levels by 4 h (Table 1). The concentration of Isotrichidae in the rumen decreased by 88 % in the period from 40 min to 4 h after feeding. In contrast, the concentration of Isotrichidae in the containers decreased by only 32% during this time period. If it is assumed that lysis and multiplication of Isotrichidae in the rumen were similar to that in the containers, then the difference 4 h after feeding must be attributed to a dilution effect of water, passage out of the rumen, sequestration or a combination of these factors, all of which were absent in the containers. If the sharp drop in the concentration of Isotrichidae in the rumen was due to dilution effects of water or passage out of the rumen, a similar response would be expected for Ophryoscolecidae. Since this did not occur, it was concluded that sequestration was the major cause of the decrease in the concentration of Isotrichidae shortly after feeding. The concentration of Ophryoscolecidae in rumen contents was fairly constant at all sampling times (Table 1), and did not differ from the concentration in the containers 4 h after feeding. This suggests that little or no sequestration occurred and that passage of Ophryoscolecidae out of the rumen was minimal during this period. This observation is in agreement with that of Clarke (1969, Warner (1966), Hungate et al. (197 1) and Weller & Pilgrim (1974). The latter authors estimated that protozoa actually pass out of the rumen at 20% or less of Table 1. Sequestration of protozoa in the rumen shortly after feeding a,b,cmeans in the same column with different superscripts differ at P < x No. of protozoa ml-i Rumen Oh 6.90" 11.96" " Rumen when 40 min 6-4P,b 18052~ *62b Rumen 4h 6-5 la,b 11.32" " Container 4h 60lb 15.38' " SE

4 1872 P. Ankrah, S. C. Loerch and B. A. Dehority Table 2. Migration of protozoa into rumen 'digesta' 40 rnin after feeding qb.cmeans in the same column with different superscripts differ at P < x No. of protozoa ml-l Rumen when 6h 6.32" 9-36" *76"2' Rumen 24 h 7~08~ 14-48b ' Rumen 24 h 40 min 6.35" 20.68' *36b Container 24 h 40 min 6~02~ 13.12b " SE the fluid rate based on direct counts of protozoa in the rumen fluid and rumen effluent obtained from continuously fed sheep. Their results indicated that protozoal nitrogen leaving the rumen amounted to 2% or less of the dietary nitrogen intake. Similar observations were reported by Punia et af. (1987). Although there was a trend for the ph in the containers at 4 h to be lower than that in the rumen, the ph in the containers remained above 6.0 and should not have resulted in lysis of protozoa (Dehority & Orpin, 1988). To determine if the increase in the concentration of Isotrichidae observed in the rumen 40 min after feeding was due to migration, an experiment was done to determine if this increase also took place in the containers. Containers were with rumen contents 6 h after feeding and incubated until 40 min after feeding on the second day. The total concentration of protozoa in the rumen increased in the period from 6 to 24 h and was greater still 40min after feeding on the second day (Table 2). The total concentration of protozoa in the containers 40 min after feeding on the second day was 37% lower than the concentration in the rumen at this time, and was similar to the concentration in the rumen 24 h after feeding. There were no differences between the concentrations of Ophryoscolecidae in the rumen 24 h after feeding and in the rumen and containers 40min after feeding on the second day. This again suggests that migration and sequestration of Ophryoscolecidae is minimal or non-existent. The increase in the total concentration of protozoa in the rumen 40 min after feeding on the second day was due to a 9.7-fold increase in the concentration of Isotrichidae. In contrast, concentrations of Isotrichidae did not increase in the containers over the entire time period. The increase in concentration of Isotrichidae in the rumen was thus attributed to migration of sequestered Isotrichidae into the rumen 'digesta'. The results of our migration and sequestration experiments are in agreement with reports that Isotrichi- dae begin to increase in numbers just before or at feeding time, reaching maximum concentrations at feeding or within 1 to 2 h after feeding (Purser, 1961 ; Warner, 1962; Clarke, 1965 ; Dehority & Mattos, 1978 ; Abe et al., 1981 ; Dehority & Tirabasso, 1989). Abe et al. (1981) observed a 4-fold increase in numbers of Isotrichidae within 1 h of commencement of feeding followed by a rapid decrease to the numbers observed before feeding. Murphy et af. (1985) observed a 10-fold increase in the concentrations of Isotrichidae 2 h after feeding, with a return to the numbers found before feeding 5 to 6 h later. Chemotactic response to soluble sugars in the diet has been suggested as a possible stimulus for the migration of Isotrichidae into the rumen contents at the time of feeding (Orpin & Letcher, 1978; Murphy et al., 1985). Abe et al. (1983) suggested that the act of ingesting feed and the contractions of the reticulum during eating or the anticipation of feed may also be involved. Nakamura & Kurihara (1978) obtained evidence for migration and sequestration of protozoa, or their attachment to particulate matter, in vitro. Growth and division of protozoa Concentrations of protozoa in the rumen are at their lowest approximately 6-8 h after feeding (Purser & Moir, 1959; Warner, 1966; Potter & Dehority, 1973). Growth and division of protozoa was estimated by comparing concentrations of protozoa in the rumen, from 6 to 24 h after feeding, with the concentrations of protozoa in the containers. In this experiment the rumen ph increased from h after feeding to a high value of h after feeding (Table 3). Although the ph in the containers 24 h after feeding was lower than that in the rumen and the initial ph 6 h after feeding, it was apparently not low enough to cause death of protozoa in the containers. Both the total concentration of protozoa and the concentration of Ophryoscolecidae in the container and the rumen increased from 6 to 24 h after

5 ~~ ~ ~~~ Sequestration, migration and Iysis of rumen protozoa 1873 Table 3. Growth and division of protozoa in the rumen between 6 and 24 h afer feeding atb,cmeans in the same column with different superscripts differ at P < x No. of protozoa ml-l Rumen when 6h 6.49" 8.88" 8.24" Rumen 24 h 7.0lb ' Container 24 h 6.11' 12.16b 1 1*3gb SE Table 4. Lysis of protozoa in the rumen during thejrst 6 h after feeding a$bscmeans in the same column with different superscripts differ at P < x No. of protozoa ml-* Location feeding PH Total Op hryoscolecidae Isotrichidae ~~ Rumen when Oh 7.15" 20.88" 19.24" Rumen 6h ' Container 6h 6.63c 17.32c 15-98' SE *08 I feeding; however, no significant increase in the concentration of Isotrichidae was observed. The increase in the total concentration of protozoa and of Ophryoscolecidae was presumably due to cell division, because a similar increase occurred in both the rumen and the containers. The concentration of Isotrichidae in the rumen was low 6 h after feeding and remained low in both the rumen and the containers 24 h after feeding, although the concentration in the rumen 24 h after feeding, tended to increase slightly, probably as a result of some migration just prior to feeding (Dehority & Tirabasso, 1989). Lysis of protozoa The decline in the concentration of protozoa from 0 to 6 h after feeding may be due to lysis, the dilution effects of feed and water and/or passage out of the rumen. By comparing concentrations of protozoa at 0 h (before feeding) and 6 h after feeding in the rumen and in the containers, these causative factors can be differentiated. The concentration of Isotrichidae in the rumen and in the containers did not differ at 0 and 6 h; however, concentration of total protozoa and of Ophryoscolecidae in the rumen and the containers decreased from 0 to 6 h after feeding (Table 4). These decreases in total protozoal and Ophryoscolecidae concentrations were greater in the rumen than in the containers. Because the containers restrict protozoal passage, the decrease in concentration of total protozoa and Ophryoscolecidae in the containers was probably the result of cell lysis. Counts of protozoa in the containers at 6 h were intermediate between counts in the rumen at 0 and 6 h, suggesting that about 50% of the decrease in concentration in rumen fluid can be attributed to dilution and passage and 50% to lysis. The diurnal fluctuation in the concentrations of total protozoa and of Ophryoscolecidae in the rumen observed in this experiment, and in the protozoal growth and division experiment discussed previously, was similar to that reported by Warner (1966) and Potter & Dehority (1973). The latter authors attributed the fall in numbers of protozoa 1-6 h after feeding to dilution by feed, saliva, drinking water and passage of digesta from the rumen, but the increase in numbers of protozoa 6-24 h after feeding was attributed to protozoal cell division. No change in the concentration of Isotrichidae was observed 0 and 6 h after feeding, indicating that migration probably occurred after feeding and the Isotrichidae had sequestered by 6 h.

6 1874 P. Ankrah, S. C. Loerch and B. A. Dehority Table 5. Lysis of protozoa in the rumen in response to withholding feed from the host animal a.bmeans in the same column with different superscripts differ at P < x No. of protozoa ml-l Rumen when 24 h 7-16" 19.28" 17.04" 1-47" Rumen 48 h 7.88b 3~48~ 2.12b 2*24b Container 48 h 7.10" 4-12b 3.32b 1-36" SE An additional experiment was done to determine whether withholding feed from the host animal would cause protozoal lysis in the rumen. The concentrations of total protozoa and of Ophryoscolecidae in both the rumen and the containers decreased sharply when feed was withheld from the steer for an additional 24 h (Table 5). This indicated that Ophryoscolecidae lysed in the rumen when substrate was lacking. The concentration of Isotrichidae in the rumen was greater at 48 h than at 24 h or in the container at 48 h. This may have been due to migration of sequestered Isotrichidae into the rumen digesta (Dehority & Tirabasso, 1989). Total protozoal concentrations decreased by 82 % and 79% in the rumen and containers, respectively, between 24 and 48 h after feeding. The corresponding values for Ophryoscolecidae were 86% and 8 1 % respectively. Potter & Dehority (1973) also observed that numbers of protozoa decreased by 80% and fluid turnover rate was negligible when sheep were starved for 1 d. They concluded that the disappearance of protozoa after starvation must have been the result of cell lysis and not passage out of the rumen. Because passage out of the rumen is impeded for protozoa in the containers used in our study, the sharp decrease in the concentrations of protozoa in both the rumen and the containers at 48 h (Table 5) must have been due to lysis. From the results of these experiments, it was concluded that the observed diurnal fluctuations in concentrations of rumen Isotrichidae are primarily due to sequestration and migration, which confirms the results of Dehority & Tirabasso (1989). Ophryoscolecidae do not seem to exhibit the phenomena of sequestration and migration to any significant extent. For animals fed once a day, about 50% of the decrease in the concentration of Ophryoscolecidae immediately after feeding appear to result from dilution effects and from passage out the rumen. The remaining 50% of the decrease was attributed to cell lysis, which would be a wasteful recycling of protozoal protein in the rumen and would result in the loss of both protein and energy to the host animal. The results also suggest that decreases in the concentration of rumen Ophryoscolecidae during periods of substrate restriction are probably due to lysis, with passage out of the rumen being minimal. This research was supported by State and Federal funds appropriated to the Ohio Agricultural Research and Development Center, The Ohio State University, manuscript no References ABE, M., IRIKI, T., TOBE, N. & SHIBUI, H. (1981). Sequestration of holotrich protozoa in the reticule-rumen of cattle. Applied and Environmental Microbiology 41, ABE, M., SUZUKI, Y., OKANO, H. & IRIKI, I. (1983). Specific differences in fluctuation pattern of holotrich concentration in the rumen of cattle, goat and sheep. Japanese Journal of Zootechnical Science 54, BAUCHOP, T. & CLARKE, R. T. (1976). Attachment of the ciliate Epidinium crawley to plant fragments in sheep rumen. Applied and Environmental Microbiology 32, CLARKE, R. T. J. (1965). Diurnal variations in the number of ciliate protozoa in cattle. New Zealand Journal of Agricultural Research 8, 1-9. DEHORITY, B. A. (1984). Evaluation of subsampling and fixation procedures used for counting rumen protozoa. Applied and Environmental Microbiology 48, DEHORITY, B. A. & MATTOS, W. R. S. (1978). Diurnal changes and effect of ration on concentrations of the rumen ciliate Charon ventriculi. Applied and Environmental Microbiology 36, DEHORITY, B. A. & ORPIN, C. G. (1988). Development of, and natural fluctuations in, rumen microbial populations. In The Rumen Microbial Ecosystem, pp Edited by P. N. Hobson. London: Elsevier Applied Science. DEHORITY, B. A. & TIRABASSO, P. A. (1989). Factors affecting the migration and sequestration of rumen protozoa in the family Isotrichidae. Journal of General Microbiology 135, DEMEYER, D. I. (1981). Rumen microbes and digestion of plant cell walls. Agriculture and Environment 6, FINA, L. R., KEITH, C. L., BARTLEY, E. E., HARTMAN, P. A. & JACOBSON, N. L. (1962). Modified in vivo artificial rumen (VIVAR) techniques. Journal of Animal Science 21, 93&934. HUNGATE, R. E. (1966). The Rumen and its Microbes. New York: Academic Press. HUNGATE, R. E., REICHL, J. & PRINS, R. (1971). Parameters of rumen fermentation in a continuously fed sheep: evidence of a microbial rumination pool. Applied Microbiology 22, JOUANY, J. P. & SENAUD, J. (1979). Role of rumen protozoa in the digestion of food cellulosic materials. Annales de Recherches Veterinaires 10,

7 Sequestration, migration and lysis of rumen protozoa 1875 LEVINE, N. D., CORLISS, J. O., Cox, F. E. G., DEROUX, G., GRAIN, J., HONIGBERG, B. M., LEEDALE, G. F., LOEBLICH, A. R., 111, LOM, J., LYNN, D., MERINFELD, E. G., PAGE, F. C., POLJANSKY, G., SPRAGUE, V., VAVRA, J. & WALLACE, F. G. (1980). A newly revised classification of the protozoa. Journal of Protozoology 27, LINDBERG, J. E., KASPERSSON, A. & CISZUK, P. (1984). Studies on ph, number of protozoa and microbial ATP concentrations in rumenincubated nylon bags with different pore sizes. Journal of Agricultural Science 102, MEYER, J. H. F. & MACKIE, R. I. (1986). Microbiological evaluation of the intraruminal in sacculus digestion technique. Applied and Environmental Microbiology 51, MURPHY, M. R., DRONE, P. E., JR & WOODFORD, S. T. (1985). Factors stimulating migration of holotrich protozoa into the rumen. Applied and Environmental Microbiology 49, NAKAMURA, F. & KURIHARA, Y. (1978). Maintenance of a certain rumen protozoal population in a continuous in vitro fermentation system. Applied and Environmental Microbiology 35, ORPIN, C. G. & LETCHER, A. J. (1978). Some factors controlling the attachment of the rumen holotrich protozoa Isotricha intestinalis and I. prostoma to plant particles in vitro. Journal of General Microbiology 106, POITER, E. L. & DEHORITY, B. A. (1973). Effects of changes in feed level, starvation and level of feed after starvation upon the concentration of rumen protozoa in the ovine. Applied Microbiology 26, PUNIA, B. S., LEIBHOLZ, J. & FAICHNEY, G. J. (1987). The role of rumen protozoa in the utilization of paspalum (Paspulum dilututum) hay by cattle. British Journal of Nutrition 57, PURSER, D. B. (1961). A diurnal cycle for holotrich protozoa of the rumen. Nature, London 190, PURSER, D. B. & MOIR, R. J. (1959). Ruminal flora studies in the sheep. IX. The effect of ph on the ciliate population of the rumen in vivo. Australian Journal of Agricultural Research 10, STEEL, R. G. D. & TORRIE, J. H. (1960). Principles and Procedures of Statistics. New York : McGraw-Hill. WARNER, A. C. I. (1962). Some factors influencing the rumen microbial population. Journal of General Microbiology 28, WARNER, A. C. I. (1966). Diurnal changes in the concentration of micro-organisms in the rumen of sheep fed limited diets once daily. Journal of General Microbiology 45, WELLER, R. A. & PILGRIM, A. F. (1974). Passage of protozoa and volatile fatty acids from the rumen of the sheep and from a continuous in vitro fermentation system. British Journal of Nutrition 32,