The use of polyethylene glycol as a marker for measuring rumen water volume and the rate of flow of water from the rumen of grazing sheep

Transcription

1 New Zealand Journal of Agricultural Research ISSN: (Print) (Online) Journal homepage: The use of polyethylene glycol as a marker for measuring rumen water volume and the rate of flow of water from the rumen of grazing sheep M. J. Ulyatt To cite this article: M. J. Ulyatt (1964) The use of polyethylene glycol as a marker for measuring rumen water volume and the rate of flow of water from the rumen of grazing sheep, New Zealand Journal of Agricultural Research, 7:4, , DOI: / To link to this article: Published online: 06 Jan Submit your article to this journal Article views: 251 View related articles Citing articles: 12 View citing articles Full Terms & Conditions of access and use can be found at Download by: [ ] Date: 02 January 2018, At: 00:40

2 713 THE USE OF POLYETHYLENE GLYCOL AS A MARKER FOR MEASURING RUMEN WATER VOLUME AND THE RATE OF FLOW OF WATER FROM THE RUMEN OF GRAZING SHEEP By M. J. ULYATT Plant Chemistry Division, Department of Scientifi c and Industrial Research, Palmerston North (Received 4 August 1964) ABSTRACT 1. The use of polyethylene glycol as a marker for measuring rumen water volume and the flow of water out of the reticulo-omasal orifice has been examined in grazing sheep. 2. With grazing sheep the semi-logarithmic plot of marker concentration against time was curvilinear and so quadratic instead of linear regressions were fitted to the data. This procedure increased the statistical precision of estimating initial marker concentration. 3. The error involved in estimating rumen water volume and flow rate in grazing sheep was greater than that of Hyden (1961a), whose animals were fed chaffed hay and kept under constant environmental conditions. 4. Examples of rumen water volumes, total rumen DM contents, and flows of water out of the reticulo-omasal orifice have been presented for sheep grazing three different pasture types. INTRODUCTION In studies of ruminant nutrition it is frequently desired to measure both the quantity of digesta in the rumen and the rate of movement of digesta in the rumen and the rate of movement of digesta out of the rumen. Techniques involving slaughter of the experimental animals have been used (Makela 1956), but these cannot be employed in intensive studies with individual animals. Rumen water volume can be estimated from the dilution of a suitable water-soluble marker. If water volume is measured in this way and a representative dry-matter (DM) sample obtained, then the DM content of the rumen can be calculated. Estimation of rate of flow of digesta from the rumen is made difficult by the fa:t that various fractions of the digesta pa~s through the alimentary tract at rates which depend on characteristics such as solubility, particle size, and specific gravity (Balch 1950, 1960; Corbett, Greenhalgh, and Florence 1959; Campling and Freer 1960). Thus physical characteristics will determine the volume of distribution of any N.Z. ]. agric. Res. 7:

3 714 Rumen water measurements marker substance, the fraction of the digesta it is associated with, and its rate of flow through the digestive tract. The use of stained food particles as a reference is commonly made for ruminants under stall-feeding conditions (Balch 1950) but is unsuitable for grazing sheep because stained grass would have to be fed as a marker and the problem of recovering this from faeces is formidable. The criteria, assumptions, and mathematical relationships involved in the use of polyethylene glycol (mol. wt. 4,000) (PEG) as a watersoluble marker have been described by Hyden (1961a). In Hyden's method two important assumptions are made:- 1. The volume of water in the rumen remains constant during the experiment. 2. The rate of flow of water into and out of the rumen is continuous and constant. In this paper an attempt was made to test Hyden's technique when applied to determining rumen water volume and rate of flow of water out of the reticulo-omasal orifice in grazing sheep. Animals and feed EXPERIMENTAL METHODS The sheep used in the experiments were adult Romney Marsh wethers prepared with permanent rumen fistulas by the method of Jarrett (1948). Five experiments were carried out, each experiment comprising three consecutive measurements on each of six sheep. In Experiments 1 and 2 all animals were grazing a mixed rye grass-white clover pasture, while in Experiments 3, 4, and 5 three sheep were grazing short-rotation and three perennial ryegrass. The animals were removed from pasture two hours prior to dosing with PEG. This enabled the rumen contents to become more fluid than when the animals were fully fed, and facilitated mixing. A dose of approximately lag of PEG in 250 ml of water was administered via the rumen fistula by means of a 50 ml syringe. The syringe was fitted with a long piece of polythene tubing and the marker solution delivered into various regions of the rumen. Immediately after dosing the sheep were returned to pasture. Samples of rumen contents were collected with long forceps from a position immediately below the fistula 0, 1.5, 3, 6, 12, and 24 hours after dosing with marker. The rumen liquor was strained through muslin to remove plant material. A nalytical methods PEG in rumen liquor was determined by the turbidimetric method of Hyden (1956a), with some modifications by Smith (1959) and the author. The strained rumen liquor samples were centrifuged for 10 minutes at 1,000 g to remove particulate matter. The supernatant was clarified with barium hydroxide and zinc sulphate (Smith 1959), followed by centrifugation at 15,000 g for 10 minutes. From 0.5 to 5 ml of the resulting supernatant (depending on PEG concentration)

4 M. J. ULYATT 715 was pipetted into 10 ml spectrophotometer tubes and made up to 5 ml with distilled water. A turbidity with PEG was produced by transferring 5 ml of an aqueous solution of 30 per cent (w Iv) trichloroacetic acid and 5.9 per cent (w Iv) Ba C1 2. 2H 2 0 into each tube with a syringe pipette. The tubes were left for 20 minutes before reading the percentage transmittance in a Bausch and Lomb spectrophotometer at 525 mfl. If duplicates differed by more than 3 per cent transmittance the analysis was repeated. Under the conditions applying in the present work the time taken to reach maximum turbidity (i.e., minimum transmittance) was dependent on PEG concentration, but 20 minutes was the average time arrived at in this work. This differed from a time of five minutes by Hyden (1956a) and one hour by Smith (1959). Smith determined his time for maximum turbidity by adding the trichloroacetic acid solution to the PEG standard made up in distilled water. In the present work it was found that the addition of the clarifying reagents reduced this period to 20 minutes. 10.1athematical relationships These have been given in detail by Hyden (1961a). If the rumen water volume and the rate of flow of water into the rumen and out through the reticulo-omasal orifice are constant over the experimental period, and the added marker is mixed rapidly, then the relationship between marker concentration and time will be exponential. A straight line should result when the natural logarithm of marker concentration is plotted against time. Extrapolation of this line back to zero time (usually by linear regression) provides an estimate of marker concentration at the time of dosing. Rumen water volume (V) can thus be calculated from the relationship: p V C"-C' where p is the amount of marker administered and C' and C" are marker concentrations in the rumen before and after dosing. The rate of flow of water from the rumen can be calculated in two ways:- (a) as the percentage of rumen water volume leaving the rumen per hour (Vol. per cent), which is equivalent to 100 k, where k is the slope of the semi-logarithmic plot of the marker dilution curve; (b) as the volume of water leaving the rumen in unit time (u), expressed by the relationship: u - kv. Total rumen dry-matter content An approximation from the relationship: DM per cent Water per cent to total rumen DM content can be estimated Total rumen DM (g) Rumen water volume (ml).

5 716 Rumen water measurements Diurnal variation in rumen DM percentage was checked using three sheep grazing a ryegrass-clover pasture. There was a significant change in DM per cent during the day (P < 0.01), and the highest value was attained in the late afternoon. In the present work a sample of rumen contents was collected for marker analysis at 4.15 p.m. each day, so a sample for DM determination was obtained at the same time. 1. Rumen water volume RESULTS If the assumptions of constant rumen water volume and flow are valid, then the semi-logarithmic relationship between marker concentration and time will be linear. However, inspection of the plots of the natura! logarithm of concentration against time in the present work showed that in most cases the relationship was not linear but concave (Fig. 1). Estimates of initial marker concentration from both linear and 4 0 3'9 E 2 0 -Ol E~ L 0:::l 0' ::J c C>l E :::l 0:: l; 1 0 c 0,1 0 :;:; 0 L... C C>l u C 0 u ~ <.9 w D... (; Time After DOSing (hr-) Fig. 1.-Some examples of semi-logarithmic plots of PEG concentration in the rumen against time from the present experiments: open circle, triangle, and cross grazing sheep; solid circle = a sheep that had been starved for 16 hours prior to feeding and also during the sampling period. 24 quadratic fits of the dilution curves were calculated. The quadratic fit assumes that the curvilinear trend in flow rate between 1.5 and 24 hours started at time zero, which is in fact less restrictive than a straight line fit which assumes constant flow rate from zero time. Rumen water volumes and their standard errors estimated by both the:c methods from Experiment 1 are prc.cntcd in Table 1. The

6 M. J. ULYATT 717 TABLE 1. Mean rumen water volumes and their standard errors (ml) calculated from linear and quadratic fits of marker dilution curves from sheep in Experiment 1. Each estimate of rumen water volume is the mean of three consecutive measurement days. Sheep No. Linear Type of regression Quadratic 1 4,970 ± 1,110 3,530 ± ,910 ± 890 3,770 ± ,070 ± 580 3,830 ± ,030 ± 410 4,420 ± ,070 ± 470 5,100 ± ,470 ± 950 2,280 ± 310 quadratic fit always gave a lower estimate of rumen water volume than did the linear regression because it gave a higher estimate of initial marker concentration. In most cases the quadratic fit gave a reduction in the standard error. Similar results were obtained for the other four experiments and it was apparent that use of quadratic regressions resulted in improved statistical precision in estimating initial marker concentration. For the above reasons the use of quadratic regressions was adopted. Hyden (1961a) gave no estimate of the error involved in his meaurements of rumen water volume, so standard errors of both linear and quadratic fits of data from his Experiment 1 (page 70) were calculated, and are presented along with their corresponding rumen water volumes in Table 2. No improvement in accuracy was gained by using a quadratic regression. Standard errors calculated from Hyden's (1961a) data were generally lower than in the present work. TABLE 2. Mean rumen water volumes and their stcndard errors (ml) calculated from linear and quadratic fits of marker dilution curves from sheep described by Hyden (l961a, Experiment 1, page 70). Each estimate of rumen water volume is the mean of five consecutive measurement days. Sheep No. Linear Type of regression \ Quadratic 3,710 ± 160 3,770 ± ,990 ± 210 5,070 ± 140 5,290 ± 120 5,330 ± 130 Mean estimates of rumen water volume, with their standard errors from sheep grazing three different types of pasture, are given in Table 3. Components of variance were obtained from analyses of variance of rumen water volumes. The components of variance for all five experiments in the present work were pooled, and, if the treatment component is excluded, 14 per cent of the variance in rumen water volume was associated with differences between sheep, 22 per cent was due to

7 718 Rumen water measurements TABLE 3. Estimates of rumen water volume, total rumen DM content, and the flow rate of water from the rumen, with their standard errors, from sheep grazing three different pasture types. The data for each sheep are the mean of three consecutive measurement days. Pasture type No. of Rumen water Total rumeni sheep volume (ml) DM (g) I Flow rate of water from the rumen Vol. % I ml/hour Perennial ryegrass plus white clover 12 3,160 ± ± ± ± 15 Perennial ryegrass 6 4,660 ± ± ± ± 31 Short-rotation ryegrass 6 3,270 ± ± ± ± 31 differences between measurement days, 51 per cent due to a sheep X day interaction, and 14 per cent was intrinsic to the method of measuring rumen water volume. A similar analysis was conducted on the rumen water volumes derived from the linear fitting of Hyden's (1961a) data. In this case 62 per cent of the variance was associated with sheep differences, 24 per cent with the sheep X day interaction, and 14 per cent was intrinsic to the method. A pooled estimate of the coefficient of variation per individual rumen water volume determination was 24 per cent in the present work and 19 per cent for that of Hyden (1961a). 2. Rate of flow of water from the rumen In some cases in the present work it was not possible to detect PEG in the 24-hour sample. Therefore, the rate of flow of water from the rumen was calculated from the slope of the PEG dilution curve in the 12 hours after dosing. This slope was derived from the relationship: Loge (C O-C12 ) k = 12 where Co and C 1 2, the PEG concentrations at times 0 and 12 hours after dosing, were calculated from the appropriate quadratic regression equation. This procedure assumed a straight line between 0 and 12 hours, which in most cases was not correct, so the flow rate calculated this way is only an approximation. This method probably over-estimates k over 24 hours because the rate of dilution of marker slowed between 12 and 24 hours (Fig. 1). Mean of rates of flow of water from the reticulo-omasal orifice with their standard errors from sheep grazing three different types of pasture are shown in Table Total rumen,om Estimates of total rumen dry-matter obtained in the present work are presented in Table 3.

8 ----,-~~~~- -~-~~ _..~,~- M. J. ULYATT The effects of rain In one experiment steady rain was experienced on the third day of PEG dosing and sampling. This had a marked effect on rumen water volume, DM content, and the flow of water out of the rumen (Table 4). TABLE 4. The effect of rain on rumen water volume, the rate of flow of water out of the rumen, and total rumen DM content in grazing sheep. Day 2 3 Rain) Sheep No. Item _~_~_ ~, - _I I I 4,870 3,770 4, (b) ml/hr 1, , Rumen water volume I (ml) Flow rate (a) Vol. % Total rumen DM (g) Rumen water volume (ml) Flow rate (a) Vol. % (b) ml/hr Total rumen DM (g) -- Rumen water volume (ml) Flow rate (a) Vol. % (b) ml/hr Total rumen DM (g) I~--I ,---- ~ ~_ iii 2 I I 4, , , , , ,020 4, ,680 3, , , '~-l , I 3,590 I 5, I I 770 2,490 I 1,610 4, ' 590 1, , Rain caused a reduction in rumen water volume and total Dl\1 content, while the percentage of rumen water leaving the rumen per hour was increased. In all cases but one the absolute flow rate of water from the rumen was decreased. 5. Relationship between the ;flows of water and organic matter In three of the present experiments, total faecal collections were made for 10 days prior to the marker dosing and sampling. There were no differences in pasture digestibility within each experiment, so faecal organic matter production was taken as a guide to intake. A correlation of (P < 0.05) was calculated between mean absolute flow of water from the rumen and faecal organic matter production for each sheep. 2, DISCUSSION The substance PEG appears to fulfil the criteria required of a water-soluble marker for use in the alimentary tract, and has been used as sueh by several workers (Corbett, Greenhalgh, and Florence 1959; Smith 1959; Gray, Jones, and Pilgrim 1960; Hyden 1961b; Oyaert and Bouckaert 1961; Sutton, McGilliard, and Jacobson 1962; Weller, Pilgrim, and Gray 1962). PEG is not toxic, not absorbed, not

9 720 Rumen water measurements metabolised by rumen micro-organisms, its distribution volume in rumen contents is approximately 95 per cent of total water, and it can be accurately estimated turbidimetrically if care is taken when clarifying the rumen liquor (Sperber, Hyden, and Ekman 1953 ; Hyden 1956a, 1956b, 1960, 1961a; Smith 1959). In the present work no trouble with interfering substances was encountered after clarifying by centrifuging at 15,000 g for 10 minutes the chemically clarified rumen liquor. Jacobson et at. (1963) found a small amount of PEG was adsorbed on or retained by the intestinal mucosa of rats. It is not known whether adsorption occurs in the rumen, but it might explain the small loss in faecal recovery of PEG noted by Hyden (1956b) and Corbett et at. (1958). Although PEG appears to fulfil the requirements of water-soluble marker, the main problems in its use are associated with theoretical aspects of the marker method of determining rumen water volume and flow rate. These problems are not peculiar to PEG; they apply to any water-soluble marker. The two assumptions that rumen water volume and the rate of flow of water through the rumen remain constant over the experimental period were not valid in the present work under grazing conditions. Rumen water volumes changed from day to day (22 per cent of the variation in rumen water volume being associated with day differences) and the semi-logarithmic plot of the marker dilution curve was curvilinear, indicating that flow rate changed during the day. Hyden (1961a) attempted to avoid these problems by providing "steady state" feeding conditions. His sheep were fed on chopped hay and were housed indoors under continuous artificial lighting and with free access to food and water. Under grazing conditions sheep are subject to the usual environmental fluctuations, and as can be seen in Table 4 factors such as rain can have marked effects on rumen water volume and flow rate. The sheep did not spread their eating evenly through the day in the present work but concentrated their grazing activity in the eight hours after dosing with PEG. Further, the nature of the feed probably influences the accuracy of measurement. The rate of flow of water from the rumens of pasture-fed sheep in the present work was three times as great as in Hyden's (196la) hay-fed animals. It is of interest that the figures for the rate of flow of water from the rumens of pasture-fed cows given by Hutton et at. (1964) of 21.6 and 24.7 percentage of the rumen water volume/hour are of the same order as in the present work. The uneven intake pattern and high rate of flow of water in pasture-fed sheep no doubt contributed to the curvilinear nature of the semilogarithmic plot to marker concentration against time and the greater error involved in measuring rumen water volume. It is difficult to obtain a representative sample of rumen liquor from pasture-fed sheep. Hyden (1961a) found that samples collected from an area immediately below the fistula gave a representative sample of marker concentration in the rumen. In the present work rumen contents were not fluid enough to allow sampling by suction tube, so long forceps were used to collect samples of rumen contents from below the fistula. A major criticism of the present work must be that although the statistical error in estimating rumen water volume was reduced by fitting

10 M. J. ULYATT 721 the marker dilution curves by quadratic regression, it is not known how accurately the PEG marker technique measures absolute rumen volume. It was not possible to compare rumen water volumes estimated by the marker technique with direct measurements obtained by removal of rumen contents through a rumen fistula: the jarrett-type rumen cannula (Jarrett 1948) used in the present work was too small for the latter purpose. Thus there is no certainty that the PEG marker technique as used in the present work measured absolute rumen water volume. However, the technique appears adequate for comparing the rumen water volumes of sheep on different feeds. It is difficult to relate the rate of flow of water from the rumen to the rate of flow of digesta through the digestive tract. Many of the substances passing through the reticulo-omasal orifice Me water-soluble, but the flow of solid particles is dependent on physical characteristics such as particle size and specific gravity (Balch 1950, 1960). Corbett et al. (1958) and Corbett, Greenhalgh, and Florence (1959) compared the rates of passage of chromium -oxide and PEG through the digestive tracts of cows and found that the passage of PEG was faster than chromium oxide, which was probably associated with the solid fraction of the digesta. Weller, Pilgrim, and Gray (1962) found that PEG passed out of the rumen more rapidly than lignin. Therefore it seems that the rate of flow of water through the digestive tract is faster than that of total digesta, Hyden (1961a) thought that increased food consumption should influence the rate of flow of water from the rumen but he could provide no clear indication of such a relationship. In the present work a correlation coefficient of was calculated between faecal organic matter production and the rate of flow of water from the rumen. It is evident that while PEG might be adequate as a marker for measuring the rate of absorption of water-soluble compounds and the rate of flow of water through compartments of the digestive tract, it cannot be used to measure the rate of flow of total digesta. ACKNOWLEDGMENTS The author wishes to acknowledge the help and guidance of Dr A. T. Johns and Professor A. L. Rae. Dr C. S. W. Reid provided much constructive criticism. Mr A. C. Glenday gave considerable help with the statistical analyses. The technical assistance of Mr P. Vlieg is gratefully acknowledged. REFERENCES BALCH, C. C. 1950: Brit. J. Nutr. 4: : In "Digestive physiology and nutrition of the ruminant." pp Ed. Lewis, D. Butterworths, London. CAMPLING, R. C.; FREER, M. 1960: Nature, Lond. 188: 670. CORBETT, J. L.; GREENHALGH, J. F. D.; FLORENCE, E. 1959: Brit. ]. Nutr. 13: 337. CORBETT, J. L.; GREENHALGH, J. F. D.; GWYNN, P. E.; WALKER, D. 1958: Ibid. 12: 266. GRAY, F. V.; JONES, G. B.; PILGRIM, A. F. 1960: Aust. ]. agric. Res. 11: 383.

11 722 Rumen water measurements HUTTON, J. B.; HUGHES, J. W.; NEWTH, R. P.; WATANABE, K. 1964: Proc. N.Z. Soc. Anim. Prod. 24: 29. HYDEN, S. 1956a: Lantbrllogsk, Ann. 22: b: Ibid. 22: : In "Digestive physiology and nutrition of the rurmnant." pp Ed. Lewis, D. Butterworths, London a: LantbrHiigsk, Ann. 27: b: Ibid. 27: 273. JACOBSON, E. D.; BONDY, D. C.; BROITMAN, S. A.; FORDTRAN, J. S. 1963: Gastroenterology 44: 761. JARRETT, I. G. 1948: I, Coun. sci. industr. Res. Aust. 21: 311. MAKELA, A. 1956: Acta agraria fennica 85: 1. OYAERT, W.; BOUCKAERT, J. H. 1961: Res. vet. Sci. 2: 41. SMITH, R. H. 1959: J. agric. Sci. 52: 72. SPERBER, I.; HYDEN, S.; EKMAN, J. 1953: LantbrHogsk, Ann. 20: 337. SUTTON, J. D.; MCGILLIARD, A. D.; JACOBSON, N. L. 1962: ]. Dairy Sci. 45: WELLER, R. A.; PILGRIM, A. F.; GRAY, F. V. 1962: Brit. J. Nutr. 16: 83.