Secondary metabolites are found in many plants and confer a mechanism of chemical defence against microorganisms, insects and grazing ruminants. Initial interest in feeding secondary plant metabolites to manipulate digestive function in ruminants began with identifying forages that could serve as cost-effective supplements for ruminants on poor-quality pastures or crop residuesReference Leng, Bird, Klieve, Choo, Ball, Asefa, Brumby, Mudgal, Chaudhry, Haryono, Hendratno, Speedy and Pugliese1. More recently, interest in the antimicrobial properties of secondary plant metabolites has diversified as the livestock industry strives to find natural alternatives to the use of growth-promoting antibiotics to maintain animal health and environmental sustainabilityReference Wallace2.
Saponins are one type of secondary metabolites and occur as either triterprenoid or steroid aglycone structures linked to various carbohydrate moieties. The chemical structures of the saponins in Enterolobium cyclocarpum, a tropical multipurpose tree, are unknownReference Wina, Muetzel and Becker3, but other saponin-containing plants of the Leguminosae family generally have saponins of a triterprenoid-type structureReference Francis, Kerem, Makkar and Becker4. One of the mechanisms of action of saponins that may be of beneficial importance in ruminant nutrition is the antiprotozoal activity. The detrimental effects of saponins on ciliate protozoa are thought to be due to binding of the saponin with sterols present on the protozoal membrane surface causing cell ruptureReference Wina, Muetzel and Becker3. Removal of ruminal ciliate protozoa either completely, partially or selectively by specific genera offers a potential approach to increase N availability and efficiency of utilization, and to reduce N excretion in manure of ruminant livestock. Microbial protein is the major source of amino acids entering the small intestine and accounts for between 35 and 90 % of the protein flowReference Clark, Klusmeyer and Cameron5. The amount of microbial protein reaching the intestine, however, may represent only one-half of the total amount of microbial protein that is synthesized within the rumenReference Leng and Nolan6, Reference Koenig, Newbold, McIntosh and Rode7. The effect of protozoal predation on ruminal bacterial numbers and microbial protein flow is considerable. It has been shown that the removal of protozoa can result in increased bacterial numbers in the rumen, reduced intraruminal microbial protein recycling, and a 20–60 % increase in the flow of bacterial protein from the rumenReference Koenig, Newbold, McIntosh and Rode7, Reference Jouany8.
Ruminal ciliate protozoa are categorized as entodiniomorphs and holotrichs based on their physical properties and metabolismReference Williams, Coleman, Hobson and Stewart9. The entodinomorphs form the majority of the protozoal community in the rumen and feed principally by engulfment of particulate matter of which bacteria serve as the major protein sourceReference Williams, Coleman, Hobson and Stewart9. Holotrich protozoa (e.g. Isotricha spp. and Dasytricha spp.) utilize soluble protein sources but may also ingest and digest bacteriaReference Williams, Coleman, Hobson and Stewart9. Protozoa of the genus Entodinium are the most numerous comprising between 78 and 98 % of the protozoa in a mixed protozoal communityReference Jouany, Ivan, Papon and Lassalas10–Reference Hristov, Ivan, Rode and McAllister13. The Entodinia have been found to be the major contributor to bacterial protein turnover in the rumenReference Coleman and Sandford14, Reference Wallace and McPherson15.
Under experimental conditions the complete absence of ruminal ciliate protozoa improves the protein nutrition of the ruminantReference Koenig, Newbold, McIntosh and Rode7, Reference Jouany8. Under practical conditions, however, it is unlikely that complete protozoal removal can be achieved or maintained. Feeding the foliage of E. cyclocarpum causes a reduction in ruminal ciliate protozoal numbers in sheepReference Leng, Bird, Klieve, Choo, Ball, Asefa, Brumby, Mudgal, Chaudhry, Haryono, Hendratno, Speedy and Pugliese1, Reference Navas-Camacho, Laredo, Cuesta, Anzola and Leon16, although the antiprotozoal effect is of short duration, lasting only 12–14 d, and most effective in animals that have not been previously exposed to the antiprotozoal agentReference Ivan, Koenig, Teferedegne, Newbold, Entz, Rode and Ibrahim12. Due to the short duration of the antiprotozoal effect of E. cyclocarpum, it would have limited value as an antiprotozoal agent in the field, but it may be useful for short-term studies to determine the effect of a partial reduction of protozoal cell numbers on the protein nutrition of the ruminant. The objectives of the present study were to determine if a reduction in ruminal fauna due to E. cyclocarpum feeding can improve bacterial and microbial protein flow to the intestine and nutrient digestion in sheep faunated with total mixed fauna (TF) or Entodinium caudatum monofauna (EN). The effects were measured during the period when E. cyclocarpum was known to be effective at reducing protozoal numbers in the rumen of sheepReference Ivan, Koenig, Teferedegne, Newbold, Entz, Rode and Ibrahim12.
Materials and methods
Animals and housing
Fifteen 1.5-year-old Canadian Arcott fauna-free (FF) wethers fitted with ruminal and duodenal (placed proximal to the common bile and pancreatic ducts) cannulas were used. The ruminal cannulas measured 26 mm in internal diameter and were constructed of natural rubber (Macam Rubber Pty. Ltd, Baulkham Hills, NSW, Australia). The duodenal cannulas were a simple T-type with an 11 mm internal diameter, gutter-type flanges and constructed of polyvinylchloride.
At birth, the sheep were removed from the dam and reared as a group in a facility separated from other ruminants to maintain their natural FF status. The sheep were used in another studyReference Ivan, Koenig, Morgavi, Rode, Newbold and Entz17, prior to their use in the present experiment. For the prior studyReference Ivan, Koenig, Morgavi, Rode, Newbold and Entz17, the fifteen cannulated FF sheep were inoculated with 50 ml each of ruminal contents collected and combined from three donor sheep with a natural bacterial community. The bacterial community in the donor sheep was established by inoculation with ruminal contents that were collected from a naturally faunated cow and frozen to kill the protozoa but maintain the majority of the bacteria. Ten days after bacterial inoculation, the sheep were then randomly assigned to three groups of five sheep each and maintained as either FF or inoculated with EN or Polyplastron multivesiculatum genera to determine the effects of the monofaunas on the duodenal flow of bacterial and protozoal N, and their contribution to ruminal fermentation and digestion of feedReference Ivan, Koenig, Morgavi, Rode, Newbold and Entz17. At the termination of the prior study, the FF sheep and the sheep inoculated with EN were maintained with their respective fauna communities for use in the present experiment. The sheep that were monofaunated with Polyplastron were then inoculated with 50 ml of the ruminal fluid containing TF on the first day of the first period of the present experiment. The ruminal fluid used for the inoculation originated from three donor sheep faunated with mixed fauna of the Type A populationReference Eadie18.
For the duration of the experiment, the sheep were housed in an enclosed barn separated from other ruminants. The sheep were cared for according to the guidelines of the Canadian Council on Animal CareReference Olfert, Cross and McWilliams19 and the experimental protocol was approved by the Lethbridge Centre Animal Care Committee.
The experiment consisted of two periods of 28 d each with measurements and sample collection performed during the last 16 d (Fig. 1). The sheep were housed in individual pens in separate rooms of the barn for each protozoal group for the first 19 d of each period. The sheep were then housed in metabolism crates for four consecutive days: 1 d for adjustment to the crates and 3 d for total faecal collection. At the end of the 4 d, the sheep were returned to their individual pens for a minimum of 4 h in accordance with the guidelines of the Canadian Council on Animal Care, and then placed back into the metabolism crates to complete the remaining 5 d of the period. The metabolism crates were separated by a minimum distance of 2 m between the protozoal treatment groups.
Fig. 1 Time line for the two 28 d periods of the experiment.
Body weight was measured at the beginning and end of the first period, and again at the end of the second period at the same time on each day and averaged 68·2 (sd 6·0) kg over the experiment.
Dietary treatments
The dietary treatments were offered twice daily with one-third of the feed offered at 08.30 hours and two-thirds offered at 15.30 hours. For days 1–15, feed was offered in an amount to permit an ad libitum level of intake by ensuring that approximately 10 % of the feed offered was refused each day. From day 16 to day 28, the diet was offered at 85 % of the mean daily intake of the sheep with the lowest amount of intake as determined from days 1–15. Sheep had free access to water at all times.
The basal diet consisted of (DM basis) barley silage (600 g/kg) and a barley grain and soyabean meal pelleted concentrate (400 g/kg) with sufficient minerals and vitamins to meet recommendations20 (Table 1). The silage and concentrate were mixed daily prior to the morning feeding in a Data Ranger (American Calan, Northwood, NH, USA). In the first period, an additional 42·5 g soyabean meal (DM basis) was mixed into the 08.30 hour feed allotment for 12 d beginning on day 16 and continuing until the end of the period to make the diet (control) isonitrogenous to the diet fed in the second period. In the second period, 187 g (DM basis) of the basal diet was replaced with an equivalent amount of E. cyclocarpum leaves (187 g DM basis, 200 g as-fed basis) and was offered with the morning feed allotment for 12 d beginning on day 16 and continuing until the end of the period. Sun-dried branches of the E. cyclocarpum were imported from Costa Rica. The dried leaves were separated from the branches and twigs by hand. The saponin content of the E. cyclocarpum was 0·80 mg/g DM as determined by the haemolysis of sheep blood with digitonin as a standardReference Barbosa, Gross, Provan and Stermitz21. The sheep used in the present study had no prior exposure to E. cyclocarpum.
Table 1 Ingredient and chemical composition of the basal diet, and the chemical composition of the Enterolobium cyclocarpum and soyabean meal
ADF, acid detergent fibre; CP, crude protein; NDF, neutral detergent fibre; OM, organic matter.
* All ingredients, with the exception of the barley silage, were pelleted (6·4 mm in diameter).
† Contained per kg of trace mineralized salt: 923 g NaCl, 50 g/kg Dynamate (Pitmann Moore, Inc., Mendelein, IL, USA; 220 g/kg S, 180 g/kg K, 110 g/kg Mg, 10 g/kg Fe), 3·3 g Zn, 2·5 g Mn, 0·34 g Cu, 0·11 g I, 93 mg Co, 14·3 mg Se.
‡ Contained per g of vitamin mix: 3·44 mg retinyl acetate, 25 μg cholecalciferol, 100 mg α-tocopherol acetate.
§ Bear River Zeolite of Canada Corporation, Lethbridge, Alberta, Canada.
¶ ACF Flavouring Agent, Alltech, Inc., Guelph, Ontario, Canada.
The flow of digesta to the duodenum was determined using Yb (particle-associated marker) and Co (solute marker) as dual markersReference Faichney, McDonald and Warner22, Reference Siddons, Paradine, Beever and Cornell23. 15N and phosphatidylcholine were used as bacterial and protozoal markers for the calculation of the duodenal flow of bacterial and protozoal N, respectively. The flow and bacterial markers were incorporated into the concentrate prior to pelleting. About 1·3 kg YbCl3.6H2O (Rhône Poulenc Basic Chemicals Co., Shelton, CT, USA), 2·05 kg of the lithium salt of CoEDTAReference Udén, Colucci and Van Soest24 and 300 g [15NH4]2SO4 (minimum 10 atom percentage 15N; Isotec Inc., Miamisburg, OH, USA) were dissolved in 10 litres water and sprayed on 350 kg concentrate.
From days 14 to 28 of the period, the sheep received the marked concentrate in place of unmarked concentrate. On day 14, prior to offering the marked diet, each sheep received a prime dose containing half of the amount of marker consumed per day. The prime dose of the markers (1·4 g of the lithium salt of CoEDTA, 0·87 g YbCl3.6H2O, 0·25 g [15NH4]2SO4) was dissolved in 100 ml deionized water and administered to the rumen through the cannula.
Measurements and sample collection
Precautions were taken to avoid cross-contamination of the ruminal protozoal fauna between the treatments groups during feeding, sample collection and processing. Coveralls and gloves were changed between the handling and sample collection for each protozoal treatment group. The FF group was fed and sampled first, followed by the EN group and then the TF group of sheep. One room was set up for each protozoal treatment group with all the supplies and equipment necessary for sample processing.
Feed and orts samples
Samples of the barley silage and marked concentrate were collected on days 21, 24 and 27, dried in a 55°C oven for 48 h to determine DM content, ground through a 1 mm sieve and combined by period. Any feed refused when sheep were on a restricted intake from days 16 to 28 was weighed, subsampled, dried and ground as described for the feed samples and combined by sheep and period. Samples of the soyabean meal and E. cyclocarpum were also collected, dried and ground. Feed and orts were analysed for DM, organic matter (OM), N, neutral detergent fibre (NDF), acid detergent fibre (ADF), and the digesta flow and bacterial markers.
Faeces and duodenal digesta collection
On day 20 the sheep were moved into metabolism crates and fitted with harnesses and faecal collection bags. The sheep were previously exposed to the crates and faecal collection bagsReference Ivan, Koenig, Morgavi, Rode, Newbold and Entz17, and were allowed 1 d to become accustomed to the crates and bags prior to sample collection in the present experiment. Total collection of faeces was performed for three consecutive 24 h periods from days 21 to 24 (the fifth to eighth day of feeding E. cyclocarpum) to determine total tract nutrient digestibility based on the flow of the Yb markerReference Siddons, Paradine, Beever and Cornell23. The faecal output was collected daily and weighed; 10 % was dried at 55°C in an oven to determine DM content, ground through a 1 mm sieve, and combined by sheep and period for analysis (DM, OM, N, NDF, ADF and Yb).
Duodenal samples were collected at various times from days 24 to 28 of each period (the eighth to the twelfth day of feeding E. cyclocarpum) to obtain digesta representative of the 24 h feeding cycle. Duodenal digesta was collected every 6 h, moving ahead 1 or 2 h each day to collect samples at 16 different times. Duodenal digesta (100 ml) was collected at 06.00, 12.00, 18.00 and 24.00 hours on day 24, at 07.00, 13.00 and 19.00 hours on day 25, at 01.00, 09.00, 15.00 and 21.00 hours on day 26, at 03.00, 10.00, 16.00 and 22.00 hours on day 27, and at 04.00 hours on day 28. Duodenal samples (100 ml for each time-point) were combined by sheep and period as collected and immediately frozen.
The combined duodenal samples were later thawed and a volume of 700–800 ml was separated into solid and liquid fractions by centrifugation at 1000 g, 4°C, for 15 min. The DM content of the fractions was determined by freeze-drying. The fractions were then ground through a 1 mm sieve. Marker and nutrient concentrations were determined in the whole, solid and liquid fractions of duodenal digesta to determine nutrient flow to the duodenum and pre-intestinal digestibility. An additional sample of the duodenal contents (15 ml) was centrifuged at 20 000 g and 4°C for 15 min and the supernatant was stored frozen until analysis of ammonia N.
Ruminal fermentation characteristics and bacterial and protozoal isolation
Ruminal contents for the determination of fermentation characteristics, and protozoal and bacterial isolation were collected at six different time-points (07.00, 10.00, 12.00, 16.00, 21.00 and 01.00 hours) from days 24 to 28. Ruminal contents (500 ml) were obtained through the cannula from multiple sites using a 60 cm3 syringe attached to tubing (20 mm internal diameter). The pH of the ruminal contents collected was measured immediately using a pH metre. Ruminal contents (60–70 ml) were then strained through one layer of cheesecloth. A volume of 5 ml of the strained ruminal fluid was acidified with 1 ml chilled 25 % (w/v) meta-phosphoric acid and the samples were stored frozen ( − 30°C) until analysis of volatile fatty acids (VFA). A volume of 10 ml of the ruminal fluid was added to 1 ml chilled 65 % (w/v) TCA to acidify the sample and precipitate the protein. The samples were kept in an ice-bath for 30 min, centrifuged at high speed (20 000 g, 4°C, for 20 min), and the supernatant was stored frozen until analysis of ammonia N. A volume of 5 ml of the strained ruminal fluid was mixed with 5 ml methylgreen formalin salineReference Ogimoto and Imai25 and the preserved samples were stored at room temperature in darkness until total cell numbers and species of protozoa were counted. In addition, during period 2, ruminal fluid samples were collected 2 h after the morning feed beginning 4 d before (days 12, 14 and 15 of the period) and during E. cyclocarpum feeding (days 16–28 of the period) for protozoal enumeration.
To determine microbial N and OM flow to the duodenum a representative sample of ruminal bacteria and protozoa were isolated and analysed for 15N enrichment, phosphatidyl choline (protozoa), N and OM. Approximately 200 ml of whole ruminal contents were mixed with an equal volume of 0·9 % saline and homogenized in a Waring blender (Waring Products Division, New Hartford, CT, USA) for two 30 s periods to dislodge particulate-associated bacteria. The homogenate was then strained through four layers of cheesecloth and the bacteria were harvested immediately by centrifugation. The filtrate was centrifuged at 800 g, 4°C, for 15 min to remove feed particles and protozoa and then the supernatant was centrifuged at 20 000 g, 4°C, for 45 min to obtain the bacterial pellet. The bacterial pellet was transferred to a container using a minimum amount of deionized water and freeze-dried.
Another 200 ml of the ruminal contents were strained through one layer of cheesecloth and the volume of the filtrate measured. An equal volume of 2 % formol saline was added to the feed particles, the particles were mixed, re-restrained and the two liquid fractions were combined for sedimentation of the protozoa. The liquid fraction was poured into a separatory funnel, allowed to stand for 45 min and then the sedimented protozoal layer was collected. The protozoal pellet was washed with 100 ml chilled (4°C) 0·9 % saline and centrifuged at 500 g, 4°C for 5 min. The supernatant was poured off and the protozoal pellet was washed two more times with a 100 ml volume of chilled saline followed by centrifugation. The washed protozoal pellet was then transferred to a container using a minimum amount of deionized water and freeze-dried. The dry samples were ground using a mortar and pestle, combined by sheep within period and stored frozen until analysis.
Chemical analysis
Analytical DM was determined by placing samples in an oven at 135°C for 2 h (Association of Official Analytical Chemists26 method 930·05) followed by hot weighing. The OM content was calculated as the difference between 1000 and the ash content (g/kg DM; Association of Official Analytical Chemists26 method 942·05). The NDF was determined as described by Van Soest et al. Reference Van Soest, Robertson and Lewis27 using heat-stable amylase and sodium sulphite, and ADF was determined according to the Association of Official Analytical Chemists26 (method 973·18) with modifications to each procedure for use in an ANKOM Fiber Analyzer (ANKOM Technology Corp., Fairport, NY, USA) and without correction for the ash content. Samples ground through a 1 mm sieve were further ground using a ball mill (Mixer Mill MM2000; Retsch, Haan, Germany) to a fine powder for measurement of N and 15N enrichment. Nitrogen was quantified by flash combustion with GC and thermal conductivity detection (Carlo Erba Instruments, Milan, Italy). Crude protein of the feed was calculated as N × 6·25. 15N enrichment (atom % excess) was measured in feed, duodenal fractions, bacteria and protozoa by flash combustion with isotope ratio MS (VG Isotech, Middlewich, UK). Phosphatidylcholine in the duodenal fractions and protozoa was determined according to Neill et al. Reference Neill, Dayrell, Kramer and Ivan28. The concentrations of the digesta flow markers (Yb and Co) were determined by inductively coupled plasma emission spectrometry (SpectroCiros CCD; Spectro Analytical Instruments, Kleve, KG, Germany).
Ciliate protozoa in preserved ruminal fluid samples were counted using a Neubauer Improved Bright-Line counting cell (0·1 mm depth; Hausser Scientific, Horsham, PA, USA) and a light microscope. VFA in ruminal fluid were quantified using crotonic acid as an internal standard and GC (Model 5890; Hewlett Packard, Little Falls, DE, USA) with a capillary column (30 m × 0·25 mm internal diameter, 1 μm phase thickness, bonded PEG, Supelco Nukol; Sigma-Aldrich Canada, Oakville, ON, Canada) and flame ionization detection. Ruminal and duodenal fluid were analysed for ammonia N concentration by the salicyclate–nitroprusside–hypochlorite method using a flow injection analyserReference Sims, Ellsworth and Mulvaney29. Ruminal pH, protozoal counts (total and species), VFA and ammonia N were averaged for the samples collected at the six different time-points over days 24–28 for each sheep for statistical analysis.
Calculations and statistical analysis
Nutrient intakes were calculated for each sheep as the difference between the amount of diet DM offered and refused on days 16–28 and the chemical composition of the diet DM offered and refused for the corresponding period. Nutrient flow to the proximal duodenum was determined by mathematical reconstitution of the duodenal digesta based on the Yb and Co marker concentrations and the nutrient content of the solid and liquid fractionsReference Faichney, McDonald and Warner22.
The flow of protozoal N to the duodenum was calculated for each sheep by subtracting the mean duodenal flow of phosphatidylcholine measured in the FF group of sheep from the flow in each of the faunated sheep to correct for the flow of feed and endogenous phosphatidylcholine. The corrected duodenal flow of phosphatidylcholine that represented the flow of phosphatidylcholine from protozoa was then by multiplied by the ratio of N to phosphatidylcholine concentration in the protozoa isolated from each of the faunated sheep.
The flow of bacterial N was calculated by correcting the N flow to the duodenum and 15N enrichment in duodenal contents by subtracting the protozoal N flow and 15N enrichment in ruminal protozoa, respectively. Bacterial N flow to the small intestine was then computed by multiplying the corrected duodenal N flow by the ratio of the corrected 15N enrichment in duodenal contents and the 15N enrichment (atomic % excess) in ruminal bacteria.
Nutrient output in faeces was calculated by dividing the daily amount of Yb intake by Yb concentration in the faeces to determine DM output and then multiplying the DM output by the nutrient concentration of the faecal DM. Recovery of the Yb marker in faeces averaged 96 (sd 3) % and data were not corrected for recovery of the marker. Apparent digestibility of OM, ADF and NDF were calculated as the difference between the nutrient intakes and the amounts passing the duodenum or excreted in faeces.
The data for ruminal fermentation, microbial N flow and nutrient digestibility were analysed as a split-plot design using a mixed linear model (release 9.1; SAS Institute Inc., Cary, NC, USA) with sheep as a random effect, fauna as a fixed effect, E. cyclocarpum treatment and the E. cyclocarpum treatment × fauna interaction as fixed effects, and using sheep × fauna in the random statement. The control diet was fed in the first period and the diet supplemented with E. cyclocarpum was fed in the second period of the experiment, thus, diet is confounded with time. Least squares means were calculated for each treatment and when the main effect of fauna and the E. cyclocarpum × fauna interaction were significant the Bonferonni's test was used to evaluate differences among means. The data for the daily protozoal counts when E. cyclocarpum was fed were analysed as mixed linear model with sheep as a random effect, fauna as a fixed effect, time as a fixed effect and repeated measure, fauna × time interaction and using sheep × fauna as the subject. The variance and covariance error structure were estimated using the variance components and compound symmetry structure based on the lowest Akaike's information criterion. When the main effect of time was significant, Dunnett's test was used to compare the daily protozoal numbers with protozoal numbers prior to feeding E. cyclocarpum (the average of the counts measured on 3 d prior to feeding E. cyclocarpum). Differences among treatments were declared significant at P < 0·05 and trends were discussed at P < 0·10.
Results
Protozoal cell numbers and ruminal fermentation characteristics
There were no protozoa observed in ruminal fluid of any of the sheep of the FF group and only Entodinium were observed in sheep of the EN group. Protozoal cell numbers were reduced (P < 0·05; Fig. 2) in ruminal fluid 2 h after feeding the diet supplemented with the foliage of E. cyclocarpum from the third to the twelfth day of offering in sheep of the TF group, but protozoal numbers were only reduced from the third to the eighth day in sheep faunated with EN. The reduction in protozoal numbers relative to the numbers prior to feeding E. cyclocarpum ranged from 31 to 88 % in sheep with TF and from 38 to 82 % in sheep with EN. When protozoal numbers were expressed as the average of the six samples collected at various times throughout the 24 h feeding cycle from the ninth to the twelfth day of E. cyclocarpum supplementation, TF and EN cell numbers tended to be reduced (P < 0·09; Table 2) by approximately 25 % in sheep fed the diet supplemented with E. cyclocarpum compared to sheep fed the control diet.
Fig. 2 Protozoal cell numbers in ruminal fluid of sheep faunated with total mixed fauna (○) and Entodinium caudatum as a single-species monofauna (●) 2 h after feeding Enterolobium cyclocarpum. E. cyclocarpum was fed from days 16 to 28 of period 2 (for 12 d in total). Values are means with their standard errors depicted by vertical bars (n 5). Prefeeding cell numbers were determined as the average of cell numbers determined on experiment days 12, 14 and 15 for sheep faunated with each protozoal community. Mean values were significantly different from those of the prefeeding: *P < 0·05.
Table 2 Effects of feeding Enterolobium cyclocarpum on protozoal numbers and ruminal fermentation in sheep with no protozoa (fauna-free, FF), total mixed fauna (TF) or Entodinium caudatum as a single-species monofauna (EN)*
VFA, volatile fatty acids.
a,b Mean values within a row, within fauna, with unlike superscript letters were significantly different (P < 0·05).
* Protozoal cell numbers and ruminal fermentation characteristics were determined at six different time-points (07.00, 10.00, 12.00, 16.00, 21.00, 01.00 hours) from days 24 to 28 and averaged for each sheep.
† FF ≠ TF, FF = EN, TF = EN (P < 0·05).
Total protozoal cell numbers were lower (P < 0·05; Table 2) in sheep faunated with TF compared to sheep monofaunated with EN. In sheep faunated with TF, Entodinia spp. accounted for 94–96 % of the fauna, Polyplastron spp. accounted for 3–6 % and Isotricha spp. accounted for less than 1 %.
There was a significant interaction between E. cyclocarpum and fauna for ruminal pH (P < 0·05; Table 2), such that there was no effect of the diet on ruminal pH in the FF sheep, but ruminal pH was higher for sheep faunated with TF and EN when fed E. cyclocarpum compared to the control diet.
Total VFA concentration in ruminal fluid was higher (P < 0·05; Table 2) in sheep fed the diet supplemented with E. cyclocarpum compared to that of sheep fed the control diet. There was no effect (P>0·10) of feeding E. cyclocarpum on the proportion of acetic acid in ruminal fluid, and although there was a lower (P < 0·05) proportion of propionic acid compared to that in sheep fed the control diet, there was no effect (P>0·10) of the dietary treatments on the acetic to propionic acid ratio. There was also no effect (P>0·10) of diet on butyric acid, but the proportion of branched chain and longer acids was higher (P < 0·05) in ruminal fluid of sheep fed the diet supplemented with E. cyclocarpum.
Fauna had no effect (P>0·10; Table 2) on total VFA concentration, however, the VFA concentration in ruminal fluid of the FF sheep fed the control diet was low (37·5 mm) relative to the faunated sheep fed the same diet (62–66 mm). There was also no effect (P>0·10) of fauna on the proportions of the individual VFA, except for butyric acid. Sheep faunated with the TF had a higher proportion of butyric acid in ruminal fluid than FF sheep (P < 0·05).
The ruminal ammonia N concentration was lower (P < 0·05; Table 2) in the FF and the faunated sheep fed E. cyclocarpum compared to the control diet. The presence of TF increased (P < 0·05) ruminal ammonia N concentration above that of the FF sheep with the ruminal ammonia N concentration in the sheep monofaunated with EN being intermediate.
Organic matter and fibre digestibility
Intake of DM and OM did not differ (P>0·05; Table 3) between sheep fed the control diet and the diet supplemented with E. cyclocarpum, but because of the high fibre content of the E. cyclocarpum (589 g/kg NDF and 486 g/kg ADF), intake of NDF and ADF were about 7 and 24 % higher (P < 0·05), respectively, for sheep fed the diet supplemented with E. cyclocarpum.
Table 3 Effects of feeding Enterolobium cyclocarpum on intake and apparent digestibility of organic matter (OM) and fibre in sheep with no protozoa (fauna-free, FF), total mixed fauna (TF) or Entodinium caudatum as a single-species monofauna (EN)
ADF, acid detergent fibre; NDF, neutral detergent fibre.
a,b Mean values within a row, within fauna, with unlike superscript letters were significantly different (P < 0·10).
* Corrected for bacterial and protozoal OM.
† FF = TF, FF ≠ EN, TF = EN (P < 0·10).
The amount of OM apparently digested in the rumen (corrected for bacterial and protozoal OM) tended (P < 0·10) to be lower in sheep fed the diet supplemented with E. cyclocarpum, but the amount of NDF and ADF digested was not affected (P>0·10; Table 3). When expressed as the fractional coefficient of digestion, OM digested in the rumen was reduced (P < 0·05) from 0·69 in the sheep fed the control diet to 0·64 in sheep fed the diet supplemented with E. cyclocarpum. The reduction in the fractional coefficient of OM digestion tended (P < 0·10) to be greatest for the sheep with TF. There was also a tendency (P < 0·10) for a lower digestion coefficient for ruminal NDF digestion in sheep when fed the diet supplemented with E. cyclocarpum compared to the control diet (0·26 v. 0·33).
Fauna had no effect on the ruminal digestion of NDF (P>0·10; Table 3). For ruminal ADF digestion, the E. cyclocarpum × fauna interaction was significant (P = 0·05), but no differences (P>0·10) were found for the treatment comparisons of interest.
Apparent total tract digestion was reduced (P < 0·05; Table 3) for OM and fibre in sheep fed the diet supplemented with E. cyclocarpum. There was no effect (P>0·10) of faunation on total tract digestion.
Nitrogen intake, intestinal flow and digestion
Nitrogen intake was similar for sheep fed the control and E. cyclocarpum-supplemented diet and among the faunated treatment groups (P>0·10; Table 4) as was intended. There was no effect (P>0·10) of diet and fauna on the flow of ammonia N to the duodenum. The flow of non-ammonia N (NAN) to the duodenum was higher in sheep fed the diet supplemented with E. cyclocarpum (P < 0·05) compared to the control diet (28·7 v. 24·9 g N/d). There were also differences (P < 0·05) among the faunated groups of sheep in the flow of NAN, with the flow of NAN highest in the FF sheep, followed by sheep faunated with EN and then lowest in sheep faunated with TF.
Table 4 Effect of feeding Enterolobium cyclocarpum on nitrogen intake, intestinal flow and apparent digestibility in sheep with no protozoa (fauna-free, FF), total mixed fauna (TF) or Entodinium caudatum as a single-species monofauna (EN)
NAN, non-ammonia N; NMNAN, non-microbial non-ammonia N; OM, organic matter.
a,b Mean values within a row, within fauna, with unlike superscript letters were significantly different (P < 0·05).
c,d Mean values within a row, within diet, with unlike superscript letters were significantly different (P < 0·05).
* Corrected for bacterial and protozoal N flow.
† FF ≠ TF, FF ≠ EN, TF ≠ EN (P < 0·05).
‡ FF ≠ TF, FF = EN, TF ≠ EN (P < 0·05).
§ FF ≠ TF, FF ≠ EN, TF = EN (P < 0·05).
The higher flow of NAN in sheep fed the E. cyclocarpum treatment compared to the control diet was due in part to a tendency (P < 0·10; Table 4) for a higher flow of bacterial N (21·6 v. 20·3 g N/d, respectively). Protozoal N flow accounted for only 6 % of the microbial N flow in the faunated sheep and was not affected (P>0·10) by feeding E. cyclocarpum. In addition to being low (0·68 g N/d), protozoal N flow was also quite variable (sem 0·41 g N/d). Microbial N flow (the sum of the bacterial and protozoal N flow) was also not affected (P>0·10) by supplementing the diet with E. cyclocarpum. Bacterial and microbial efficiency, however, were higher (P < 0·05) for sheep fed the diet supplemented with E. cyclocarpum.
The interaction between E. cyclocarpum and fauna for the flow of non-microbial NAN (NMNAN) was significant (P < 0·05; Table 4). Feeding the diet supplemented with E. cyclocarpum increased the flow of NMNAN in the FF and faunated groups of sheep. However, for sheep fed the control diet, the flow of NMNAN was lower for those faunated with TF and EN compared to the FF sheep, but when fed the diet supplemented with E. cyclocarpum, the flow of NMNAN in the sheep with TF and EN increased to the flow observed in the FF sheep. In the absence of protozoa, the flow of NMNAN was increased by 1·69 g N/d when the FF sheep were fed the diet supplemented with E. cyclocarpum compared to when fed the control diet. The flow of NMNAN was increased to a greater extent, however, when the sheep faunated with TF (4·09 g N/d) and EN (2·73 g N/d) were fed the diet supplemented with E. cyclocarpum compared to when fed the control diet.
Bacterial and microbial N flow were lowest (P < 0·05; Table 4) in the sheep faunated with TF, but did not differ (P>0·10) between the sheep faunated with EN and sheep that were FF. Bacterial and microbial efficiency were higher (P < 0·05) in the FF sheep than either of the faunated treatment groups.
The amount of N apparently digested in the rumen (corrected for bacterial and protozoal N) was lower (P < 0·05; Table 4) in sheep fed the diet supplemented with E. cyclocarpum compared to the diet without supplementation. The amount of N digested in the rumen was also lowest in the FF sheep (P < 0·05), however, there was a significant interaction between the dietary and fauna treatments for the coefficient of ruminal N digestion. As was found for the amount of dietary N digested in the rumen, the coefficient of ruminal N digestion was lower for all sheep fed the diet supplemented with E. cyclocarpum. When fed the control diet, the coefficient of ruminal N digestion was lower (P < 0·05) for the FF sheep compared to the sheep with TF and intermediate for sheep faunated with EN. There were, however, no differences among the faunated groups of sheep fed the E. cyclocarpum. Therefore, in the absence of protozoa, the coefficient of ruminal N digestion was reduced by 10 % when the FF sheep were fed the diet supplemented with E. cyclocarpum compared to when fed the control diet. The coefficient of ruminal N digestion was reduced to a greater extent, however, when sheep faunated with TF (18 %) and EN (12 %) were fed the diet supplemented with E. cyclocarpum compared to when fed the control diet. As was found for the ruminal digestion of N, total tract digestion was lower (P < 0·05) for sheep fed E. cyclocarpum, but there was no effect (P>0·10) of fauna on total tract N digestion.
Discussion
The antiprotozoal effect of E. cyclocarpum was first observed in in vitro studies as a reduction in protozoal numbers and bacteriolytic activity and this led to interest in its potential benefits for manipulating digestion and performance of ruminant livestockReference Leng, Bird, Klieve, Choo, Ball, Asefa, Brumby, Mudgal, Chaudhry, Haryono, Hendratno, Speedy and Pugliese1, Reference Teferedegne30. In in vivo studies with sheep and buffalo fed poor-quality forages supplemented with E. cyclocarpum, a reduction in ruminal fauna was observed when the foliage (10–34 % of diet DM) was fed for periods of short durationReference Leng, Bird, Klieve, Choo, Ball, Asefa, Brumby, Mudgal, Chaudhry, Haryono, Hendratno, Speedy and Pugliese1, Reference Navas-Camacho, Laredo, Cuesta, Anzola and Leon16. In the present experiment, the supplementation of E. cyclocarpum (16 % of diet DM) to a higher-quality barley-based diet was also effective in reducing total protozoal numbers. Protozoal numbers were reduced from 31 to 88 % 2 h after feeding E. cyclocarpum and by an average of 25 % over the 24 h feeding cycle when measured up to the twelfth day of supplementation. The present results are consistent with a previous experiment during which the same batch of material was fed to sheep fed a similar barley-based diet and with no prior exposure to E. cyclocarpum Reference Ivan, Koenig, Teferedegne, Newbold, Entz, Rode and Ibrahim12. In that study, protozoal numbers were reduced by 49 to 75 % 2 h after feeding the foliageReference Ivan, Koenig, Teferedegne, Newbold, Entz, Rode and Ibrahim12. However, the antiprotozoal effect dissipated and protozoal numbers returned to pre-supplementation or control levels when E. cyclocarpum was fed for more than 12–14 dReference Leng, Bird, Klieve, Choo, Ball, Asefa, Brumby, Mudgal, Chaudhry, Haryono, Hendratno, Speedy and Pugliese1, Reference Ivan, Koenig, Teferedegne, Newbold, Entz, Rode and Ibrahim12. The short-term antiprotozoal action of E. cyclocarpum and some other saponin-containing plants (e.g. Sesbania sesban) is attributed to the adaptation of the bacterial community to degrade the saponinReference Newbold, El Hassan, Wang, Ortega and Wallace31 and possibly by inactivation of the saponin by salivary proteinsReference Teferedegne30. Despite the short-term action of E. cyclocarpum, protozoal numbers were reduced in the present experiment by feeding E. cyclocarpum, which permitted the assessment of whether such reduced fauna could provide some of the same benefits as defaunation in improving the supply of bacterial N and efficiency of bacterial N synthesis in ruminants fed diets that are otherwise favourable for the growth of protozoa.
Saponins are recognized for their antiprotozoal action, but in addition, saponins have also been reported to inhibit ruminal bacteria and fungi. The inclusion of steroidal saponins from Yucca schidigera extract in the medium of pure cultures inhibited cellulolytic ruminal bacteria and fungi but had variable effects on non-cellulolytic bacteriaReference Wallace, Arthaud and Newbold32, Reference Wang, McAllister, Yanke and Cheeke33. It was suggested that the differential responses of bacterial species to steroidal saponins lies in the structure of the cell membrane, with gram-positive bacteria more susceptible than gram-negative bacteria to inhibitory effects, with the implication that the effects of saponins may be beneficial in much the same way as ionophoresReference Wallace, Arthaud and Newbold32. An inhibition of ruminal bacterial synthesis in vivo, however, could negate the beneficial effects of a reduced protozoal community on bacterial recycling and bacterial protein supply. In sheep dosed intraruminally with an alfalfa saponin extract, bacterial N flow and efficiency of bacterial N synthesis were decreasedReference Lu and Jorgensen34. In contrast, the feeding or administration of other saponin-containing plants or plant extracts have increased bacterial numbers and/or the supply of bacterial protein. Total and cellulolytic bacterial numbers in sheep fed S. sesban were increased when measured concomitantly with the reduction in protozoal numbersReference Newbold, El Hassan, Wang, Ortega and Wallace31. In the present experiment, bacterial N flow in both the FF and faunated sheep tended to be higher when fed the diet supplemented with E. cyclocarpum, indicating that bacterial biomass production was not negatively affected by the saponins.
The total saponin content of the E. cyclocarpum leaves used in the present study was determined semi-quantitatively by expressing the haemolytic activity of the saponin extract relative to the activity of digitoninReference Barbosa, Gross, Provan and Stermitz21. Haemolytic assays have been used for decades to detect and quantify saponins, however, the haemolytic activity can vary depending on the structure of the saponin and the assay technique usedReference Sen, Makkar and Becker35. Based on the haemolytic assay used in the present study, the active saponin content of the E. cyclocarpum leaves was 0·8 mg/g DM. This equated to 0·13 g saponins/kg DM for the E. cyclocarpum dietary treatment. The dietary concentration of saponins fed in the present study was low relative to other studies that have reported the saponin content of diets supplemented with E. cyclocarpum when determined using alternative methods. Hess et al. Reference Hess, Kreuzer, Díaz, Lascano, Carulla, Soliva and Machmüller36 reported the crude saponin content of the fruit of E. cyclocarpum to be 19 mg/g DM based on a method of extraction and gravimetric analysis. Navas-Camacho et al. Reference Wallace and McPherson15 reported feeding sheep 1·2 and 4 % of diet DM as saponins from E. cyclocarpum leaves (approximately 125 mg saponins/g E. cyclocarpum DM); the method of analysis, however, was not reported. Few studies have reported the saponin content of E. cyclocarpum and differences in methods of quantifying saponins make it difficult to draw conclusions regarding an effective dietary concentration of saponins from E. cyclocarpum.
The responses affecting protein nutrition that are often observed in defaunated ruminants include lower ruminal ammonia N concentration, increased flow of NAN and bacterial N to the intestine, and increased efficiency of microbial protein synthesis in the rumenReference Jouany8, Reference Eugene, Archiméde and Sauvant37. Improvement in the supply of NAN to the intestine of defaunated ruminants is due to a reduction in protozoal predation of bacteria, a reduction in intraruminal microbial N recycling, an increase in the size of the bacterial N pool and an increase in the net flow of bacterial N to the duodenumReference Koenig, Newbold, McIntosh and Rode7, Reference Demeyer and Van Nevel38–Reference Hsu, Fahey, Berger, Mackie and Merchen40. In addition, defaunation often decreases ruminal degradation of dietary protein and increases the flow of NMNAN to the duodenumReference Ivan, Neill, Forster, Alimon, Rode and Entz11, Reference Hsu, Fahey, Berger, Mackie and Merchen40, Reference Veira, Ivan and Jui41. The supplementation of the barley-based diet with E. cyclocarpum and the 25 % reduction in ruminal ciliate protozoal numbers increased the flow of NAN from the rumen to the duodenum by 16 % compared to the flow in sheep fed the control diet. The increased flow of NAN was due to an increase in the flow of both bacterial N and NMNAN. In addition, efficiency of bacterial protein synthesis in the rumen was improved by 13 %.
Feeding a diet supplemented with E. cyclocarpum improved NAN and bacterial N flow and the efficiency of bacterial protein synthesis in the rumen, but it also reduced ruminal and total tract digestion of OM and N. Supplementation of the diet with E. cyclocarpum increased the proportion of NDF and ADF in diet and could have affected cellulolytic bacterial numbers resulting in the reduction of OM and fibre digestion. The increase in the flow of NMNAN in the FF sheep fed the diet supplemented with E. cyclocarpum compared to when fed the control diet suggested that the protein of E. cyclocarpum was less ruminally digestible than that of the barley silage, barley grain and soyabean meal that it replaced. However, the increase in the flow of NMNAN was even greater in the TF and EN groups of sheep fed the diet supplemented with E. cyclocarpum compared to when fed the control diet and indicated that the reduction in ruminal N digestion and increase in flow of NMNAN was largely attributed to the reduction in protozoal numbers and the associated beneficial reduction in ruminal proteolysisReference Jouany8.
In addition to saponins, E. cyclocarpum may also contain condensed tannins which can range in concentration from 0 to 38 g/kg DMReference Ahn, Robertson, Elliott, Gutteridge and Ford42, Reference Monforte-Briceño, Sandoval-Castro, Ramirez-Avilés and Capetillo Leal43. Dietary concentrations of condensed tannins as low as 20–45 g/kg diet DM can form complexes with dietary proteins and reduce protein digestibilityReference Min, Barry, Attwood and McNabb44. The estimated dietary concentration of tannins fed in the present study may have ranged from 0 to 6·3 g/kg DM, and although the dietary concentration was considered low, the concentration of tannins present in E. cyclocarpum could have affected the digestion of the protein of E. cyclocarpum. When E. cyclocarpum with no condensed tannins was incubated in sacco in goats, fractional N disappearance at 24 h of incubation was extensive at 0·92Reference Ahn, Robertson, Elliott, Gutteridge and Ford42. However, when E. cyclocarpum containing 30 g/kg DM was incubated in sacco, N disappearance at 48 h was 0·52 in sheep fed E. cyclocarpum at 10 % of diet DM and was slightly lower at 0·40 in sheep fed E. cyclocarpum at 34 % of diet DMReference Navas-Camacho, Laredo, Cuesta, Anzola and Leon16. In the present study, the lower ruminal and total tract N digestion observed in sheep fed the diet supplemented with E. cyclocarpum may be due in part to the presence of tannins and the reduction in the digestion of the protein of E. cyclocarpum. Thus the effect of feeding a diet supplemented with E. cyclocarpum on diet digestibility would appear to be due to its nutritive and anti-nutritive composition in addition to its antiprotozoal properties. The design of this experiment permitted separation of the nutritive effects from the antiprotozoal effects, but it may be advantageous to avoid the overlap of these effects by feeding extracts of the potential antiprotozoal compounds.
The number of protozoa in ruminal fluid of the sheep monofaunated with EN was more than double the number in sheep faunated with the TF wherein Entodinium spp. account for 94–96 % of the total protozoal community. Higher numbers of protozoa in animals monofaunated with Entodinia spp. than in those faunated with TF have also been observed in other studiesReference Ivan, Neill, Forster, Alimon, Rode and Entz11. The present results may suggest that there is a considerable amount of predation of E. caudatum in a mixed protozoal community, and perhaps E. caudatum is less competitive in a mixed protozoal community where individual species of protozoa are influenced by other species that are presentReference Williams, Coleman, Hobson and Stewart9.
In the present experiment, the flow of NAN to the intestine of FF sheep was 25 % higher than in sheep faunated with TF and 12 % higher than in the sheep monofaunated with EN. The flow of bacterial N to the small intestine of the FF sheep was 17 % higher than in sheep faunated with TF, but was only 4·3 % higher (numerically) than in sheep faunated with EN. The similarity of the bacterial N flow between the FF sheep and sheep monofaunated with EN was unexpected and in contrast to the results of previous studiesReference Ivan, Neill, Forster, Alimon, Rode and Entz11, Reference Ivan, Koenig, Morgavi, Rode, Newbold and Entz17. In the previous study using the same sheep under similar experimental conditions, bacterial N flow was 22 % higher in the FF sheep compared to the sheep monofaunated with ENReference Ivan, Koenig, Morgavi, Rode, Newbold and Entz17. The sheep in each of these studies were fed a similar barley-based diet, except that the diet in the present experiment was higher in crude protein due to the additional soyabean meal or E. cyclocarpum. In a study with sheep fed haycrop and maize silage-based diets, bacterial N flow was increased by approximately 32 and 85 %, respectively, in sheep monofaunated with either Entodinium spp. or TF when compared to FF sheepReference Ivan, Neill, Forster, Alimon, Rode and Entz11. Differences in the size distribution of particles of insoluble protein and starch among the diets fed in these studies may contribute to the differential effects of the Entodinium monofauna on bacterial N flow. The rate of engulfment of bacteria by protozoa can be reduced by the uptake of particulate materials, especially starch grainsReference Coleman, McDonald and Warner45. At higher concentrations of starch, the engulfment of starch grains reduced the engulfment of bacterial cells by E. caudatum Reference Coleman, McDonald and Warner45. Engulfment by E. caudatum is not selective and E. caudatum will engulf all particles of the appropriate size, whereas other protozoal species may be more selective towards bacteriaReference Coleman, McDonald and Warner45. Thus, in the present experiment, despite a larger protozoal community in the sheep monofaunated with EN, the negative effect of protozoa on the NAN and bacterial N supply was greater for the sheep faunated with TF than those monofaunated with EN. In the study of Ivan et al. Reference Ivan, Koenig, Morgavi, Rode, Newbold and Entz17, the duodenal flow of NAN was reduced to the same extent in sheep monofaunated with Entodinia spp. or Polyplastron spp. compared to FF sheep. In in vitro studies, large protozoa isolated from the Type A population engulfed mixed and pure bacterial suspensions at a faster rate than Entodinium spp.Reference Coleman and Sandford14, although in vivo in a mixed Type A population, the numbers of large protozoa would be considerably lower than Entodinium spp. The above combined results suggest that Entodinia spp. may not have the greatest negative impact on bacterial N synthesis in a mixed protozoal community.
Protozoal N flow did not differ between the two faunated groups of sheep and accounted for less than 6 % of the microbial N flow. Protozoal N may account for 35–65 % of the microbial N pool in the rumen but because protozoa are largely retained within the rumen and undergo extensive intraruminal recyclingReference Harrison, Beever and Osbourn46, protozoal N flow to the intestine generally ranges from only 4 to 40 % of the total microbial N flowReference Harrison, Beever and Osbourn46–Reference Robinson, Fadel and Ivan49. As was found for bacterial N flow, the flow of microbial N (bacterial N plus protozoal N) was 14 % higher in the FF sheep than in sheep faunated with the TF, but it did not differ from sheep faunated with EN.
Defaunation can decrease OM and fibre digestibility in the rumen and total tractReference Eugene, Archiméde and Sauvant37. A decrease in ruminal OM digestion caused by defaunation can negate the positive effect of an increased intestinal flow of bacterial N on the improvement of bacterial or microbial efficiency. In the present experiment, there were no differences between the FF sheep and sheep faunated with TF for apparent ruminal and total tract OM digestibility, although there was a tendency for a lower amount of OM digestion in the FF sheep compared to the sheep monofaunated with EN. The coefficients of OM digestion, however, were similar for the FF sheep and the sheep monofaunated with EN. This is in agreement with results of a similar study, wherein Ivan et al. Reference Ivan, Koenig, Morgavi, Rode, Newbold and Entz17 found no differences in apparent ruminal and total tract OM and fibre digestion between FF sheep and sheep monofaunated with EN. Total VFA concentration was also not affected by the type of fauna. Acetate and butyrate are the major fermentation end products of protozoaReference Jouany, Demeyer and Grain50, but in the present experiment, there were no differences in the proportions of individual VFA, except for butyrate concentration which was higher in the sheep faunated with TF than in the FF sheep. Thus, with no reduction in OM fermentation, the higher intestinal flow of bacterial N resulted in greater bacterial and microbial efficiency in the FF sheep than in sheep faunated with TF. Many studies have reported that defaunation improved microbial N flow and efficiencyReference Jouany8, Reference Eugene, Archiméde and Sauvant37, however, not all have quantified the contribution of protozoal N to the total microbial N flow to the intestine. Phosphatidylcholine was used as a protozoal marker, but background levels of phosphatidylcholine can be quite high and variable, and the method is unable to quantify low protozoal flowsReference Ivan, Koenig, Morgavi, Rode, Newbold and Entz17. To assess accurately the implications of defaunation on microbial N supply and efficiency, accurate and sensitive methods will need to be developed to determine protozoal N flowReference Sylvester, Karnati, Yu, Newbold and Firkins51.
The supplementation of a barley-based diet with E. cyclocarpum reduced ruminal ciliate protozoal numbers by 25 %, increased the intestinal supply of NAN and bacterial N, and improved the efficiency of bacterial protein synthesis. The antiprotozoal effect of dietary E. cyclocarpum, however, is of short-term duration, and therefore E. cyclocarpum has limited usefulness as an antiprotozoal agent in ruminant diets. However, the antiprotozoal effects achieved in the present study suggest that a 25 % reduction in protozoal numbers is sufficient to achieve the beneficial effects of reduced fauna on N utilization in ruminant production.
Acknowledgements
The authors thank Dr J. Kastelic for professional help with the surgical preparation of the sheep and L. Neill, R. Moore, K. Andrews, R. Cadieu-Browne, R. Wuerfel and D. Vedres for technical assistance. C. J. Newbold and B. Teferedegne gratefully acknowledge financial support from the Scottish Office Agriculture, Environment and Fisheries Department and ODA Holdback Award R6397.
