Comparative aspects of volatile fatty acid production in the rumen and distal fermentation chamber in Svalbard reindeer

M i c r o b i a l fermentation end products were investigated in Svalbard reindeer at two different locations, on Nordenskioldland ( N L ) (»=7) and in a marginal area on Nordaustlandet ( N A ) (n= 11), at different seasons. The p H ranged from 6.51-6.70 in rumen contents and from 6.78-7.17 in the distal fermentation chamber (DFC=caecum and prox imal part of the colon) on N L compared to 6.10-6.71 in rumen contents and 6.50-7.35 in D F C contents on N A . The ruminal volatile fatty acid concentration ( [VFAJ) was 84.5 ± 9 . 5 mmol/1 compared to 63-9 ± 17.6 mmol/kg in the D F C on N L in winter. In autumn, ruminal and D F C [ V F A ] was h igh at 113.5 ± 13.0 mmol/1 and 90.4 ± 10.9 mmol/kg, respectively. O n N A ruminal [ V F A ] was 85.7 ± 12.4 mmol/1 and 59-6 ± 1 . 3 mmol/kg in the D F C in winter, compared to 107.3 ± 18.4 mmol/1 and 102.0 + 12.7 mmol/kg in rumen and D F C , respectively, in summer. Mean acetate/propionate (A/P) ratios in the rumen indicate fermentation i n favour of plant fibre digestion in winter (4.8) but not i n autumn (3.0) on N L . O n N A , the mean A/P ratio was 5.1 in winter, compared to 4.6 in summer. In all D F C investigated the A/P ratio was higher than 8.9The in i t ia l ruminal [ V F A ] d id not reflect the V F A production measured. O n N L , the production rate of V F A was low or not detectable i n rumen and D F C in winter, while in autumn the total production rate of V F A was 59-3 k J/kgW 0 7 5 / d , of which 6 .5% originated from the D F C . O n N A i n winter, a total of 121.3 k J/kgW 0 7 S/d was estimated of which 17% originated from the D F C , compared to a total of 380.4 k j / k g W " 7 '/d in summer where the D F C only contributed 2.7%. Plants (grasses and mosses) w i t h low quality i n winter do not seem to contribute significantly to the V F A production in rumen and D F C . V F A production in the D F C seems to be of significant importance in reindeer when pastures have low availability but h igh quality. The concenttation and the rate of V F A production i n the D F C contents were not related to the size of the chamber, but to the diet eaten. K e y w o r d s : Rangifer tarandusplatyrhyncbus, v o l a t i l e fatty acids , energy p r o d u c t i o n . Rangifer, 17 (2): 8 1 9 5


Introduction
Ruminants satisfy their energy requirements by utilising end products from fermentation of microorganisms living in the gastrointestinal tract (Barcroft Rangifer, 17 (2), 1997 et al, 1943;Annison & Armstrong, 1970).The rumen of domestic animals has been subject of many publications and shows that ruminal volatile fatty acid (VFA) concentration and pool size are cor-related with VFA production rate (Leng & Brett, 1966;Leng et al, 1968;Weston & Hogan, 1968;White & Staaland, 1983).Research on fermentation in the rumen and in the distal fermenting chamber (DFC = caecum and proximal part of the colon, before the colon enters the colon coil) in wild animals is limited.Lechner-Doll et al. (1991) found only small seasonal variation in the VFA concentration in African domestic ruminants which reflected the food availability and quality.The Svalbard reindeer {Rangifer tarandus platyrhynchus) live under the most austere nutritional conditions on the high-arctic archipelago of Svalbard (77-81°N).Their distribution range from high quality pastures on Nordenskioldland (NL) to that of Nordaustlandet (NA) with a marginally vegetation, characterised as an arctic desert (Staaland & Punsvik, 1980).
According to Hofmann (1985;1989), reindeer are classified as adaptable intermediate feeders based on their gastrointestinal (GI) anatomy and feeding habit which express their ability to use a mixed diet with low fibre content (Hofmann & Stewart, 1972).
The wet weight of the reticulo-rumen contents in adult female Svalbard reindeer was 77.6% of the total GI content (Staaland et al., 1979) in summer, 76.6% in autumn and 76.3% in winter (S0rmo et al., unpubl.).The DFC (distal fermentation chamber) of adult female Svalbard reindeer is large contributing 9.8% of the GI contents in autumn and 9.0% in winter (S0rmo et al., unpubl.).Orpin et al. (1985) and Mathiesen et al. (1987) found highly active microbial organisms in the rumen and caecum of Svalbard reindeer, with strong seasonal changes in the number and composition of bacteria.
Svalbard reindeer are faced with seasonal changes in photoperiod and have a corresponding pattern in food intake, which is low in winter compared to summer (Larsen et al., 1985).Using the predicted data from Nilssen et al. (1984) and from White & Staaland (1983), the ruminal production in summer on NL (575 kj/kg W 07 7d) contributed 80.5% more than fasting metabolic rate in Svalbard reindeer, but the importance of VFA produced in the DFC is not understood.The contribution of fermentation products from the rumen in winter is still unclear.We therefore wanted to examine the rumen and DFC contribution of VFA's, lactate and succinate to the daily energy supply in Svalbard reindeer in different locations and at different times of the year.The ruminal and DFC fermentation in Svalbard reindeer were therefore evaluated by investigating production rates of VFA using the zero-time in vitro techni-que (Carroll & Hungate, 1954;Hungate et al., 1961;Hungate, 1966;White & Staaland, 1983;Olsen & Mathiesen, 1996).

Material and methods
The study area The investigation was carried out in one lush and one marginal area of Svalbard.The rich area was located between 78°05'-78°17TSf and 15 W- Glaciers covers 80% of the island, and snow covers the ground for 9-10 months of the year.The vegetation is scarce and the area is characterised as an arctic desert (Staaland & Punsvik, 1980).(1967) and Goering & Van Soest (1970).Nitrogen contents were determined by the Kjeldahl method (Horwitz, 1980) and converted to crude protein Rangifer, 17 (2), 1997

Animals
(CP) by multiplication by 6.25.The amount of water-soluble carbohydrates (WSC) was determined as described in Olsen et al. (1994).Dry samples were ashed at 550 °C for 12h to convert all measurements to an organic matter (OM) basis.

In vitro fermentation of ruminal and DFC contents
When obtaining the rumen and the DFC contents, standard methods for rumen microbial studies were used (Orpin 1985;Olsen & Mathiesen, 1996).
The total rumen and DFC fluid volume was found by determining the dry matter content.To obtain the concentration and production rate of acetate, propionate and butyrate, which constitute the total volatile fatty acids (VFA), the whole rumen and the whole DFC was emptied into pre warmed thermos flasks (10 1 and 5 1) which were sealed.The zerotime in vitro technique was used to obtain the concentration and production rate of VFA in Svalbard reindeer (Carrol & Hungate, 1954;Hungate et al, 1961;Hungate, 1966;White &-Staaland, 1983;Olsen & Mathiesen, 1996).By incubating the The VFA's produced were assumed to be absorbed in the rumen or down the gastrointestinal tract.
Therefore the contribution of VFA's from the rumen and DFC to the total energy budget could be calculated.Energy available from the VFA pool was calculated using energy values of 874, 1534 and Rangifer, 17 (2), 1997 2190 kj/mole for acetic, propionic and butyric acids, respectively (Blaxter, 1962).

Measurements of pH
Ruminal and DFC contents were transferred to pre-  (1994).Each sample was homogenised and an internal standard (IS): (0.5 ml of 0.2136 g 2-ethyl butyric acid dissolved in 250 ml distilled water) and 0.6 ml 12M HC1 were added to 1ml of the homogenised sample, mixed for 0.5 min and extracted with 2 ml of diethyl ether for 1 min.The sample was centrifuged in aLabofuge, speed 5 for 5 min., the ether phase was removed and collected, and the sample was extracted again for 1 min using 1 ml of diethyl ether.Again the ether phase was removed after 5 min centrifuging.N-tert-butyldimethylsilyl-Nmethyltrifluoracetamide (MTBSTFA; lOul) was added to the combined ether extract in a test tube, which was sealed.The acids were derivated by heating for 20 min at 82 °C.Samples were kept at room temperature before injecting 0.5 ul into a gas chromatograph fitted with a CP SIL 8 CB column (Cromepack no 7452, 30 m, 0.2 mm ID) containing a film (0.25 um) of silica gel.The carrier gas was H2 at 45 cm/s.Injector' temperature was 250 °C and detector temperature was 270 °C.

Statistical analyses
The non-parametric Wilcoxon rank sum test (Bhattacharyya & Johnson, 1977) was used to deter-

Chemical composition
Chemical composition of the rumen contents of Svalbard reindeer grazing on NL and on NA is calculated from S0rmo et al. (unpubl.) and is shown in Table 1.The cell wall contents from the rumens of NA reindeer were lower than in NL.Both the water-soluble carbohydrates (WSC) and crude protein (CP) were concomitantly higher in NA than in NL reindeer.Highest WSC and CP and lowest cell walls were noted for NA reindeer in summer.

PH
On NL, rate of pH decrease in DFC in winter was low (-0.12 ± 0.05 units/h), compared to -0.36 ± 0.02 units/h in autumn.On NA in winter, rate of pH decrease in DFC contents was 0.10 ± 0.03 and not different from that found in summer (-0.21 ± 0.13)(P>0.05),(Fig. 2).
Significant differences (P=0.05) in VFA pool size between winter and summer were observed on NA, but not on NL (Table 4 and 5 Rangi£er, 17 (2), 1997 total VFA, several other volatile fatty acids, in addition to lactate and succinate were found in the rumen and in the DFC of Svalbard reindeer (Table 6).In Tables 2 and 3  The major difference between VFA ratios in the rumen and DFC was the A/P ratio (Tables 2 and 3).
In the rumen of NL and NA reindeer, ratios were between 3.0 to 5.1, whereas those in the DFC were 10.7 to 14.8.In the rumen, lowest A/P ratios were associated with highest [VFA].

VFA production rate
In seven of ten data sets ruminal and DFC [VFA] increased with time to give regression coefficients of 0.047 to 0.480 mmol/h/ml (Fig. 3).The lowest value was driven by one of three animals with a significant trend, therefore no seasonal VFA production was determined for the DFC of NL reindeer in the winter (Table 4).For the rumen contents of NL reindeer in winter there was an overal decline in [VFA] with time (Fig. 3), again no in vitro VFA production rate could be estimated (Table 4).For all NA animals [VFA] increased with time (Table 5, Fig. 3) and mean VFA production rates were determined for both winter and summer (Table 5).Thus, on NL, total ruminal and DFC VFA rate of production was low or not detectable in winter but in autumn VFA production accounted for 59 ± 6 kJ/kgW 0 75 /d of metabolizable energy of which 6.5% originated from the DFC (Table 4).On NA in winter the total rate of VFA production accounted for 121.3 ( 69.5 kJ/kgW 07 7d metabolizable energy of which 17% originated from the DFC.In summer VFA metabolizable energy was 380.4 ± 182.9 kJ7kgW" 75 /d, a significant increase over winter (P=0.02,W=3, n} = 2, »2 = 9), but the DFC contributed only 2.7%.(Table 5).Initial ruminal VFA concentration was not correlated with ruminal VFA production rate (Tables 2 and 3).This was confirmed in the contents of the DFC.Rate of decrease of ruminal pH (Fig. 2) was related to increased VFA production rate (Table 4 and 5, Fig. 3), but was not significant for the DFC.

Discussion
In domesticated ruminants fed a high quality diet the in situ pH in the rumen is usually between 5.5 and 6.7 (Hungate, 1966).Ruminal pH is dependent on the diet, the rate of salivary secretion, rate of VFA production and absorption across the rumen wall (Church, 1983).The mean initial pH measured in the rumen of Svalbard reindeer did not vary significantly between seasons and locations, ranging from 6.48 to 6.68 (Table 1 and 2, Fig. 2).These values are similar to that measured by White & Staaland (1983) but high compared to that found by Orpin et al. (1985) where a mean (± SD) pH of 6.19 ± 0.16 in the rumen fluid was measured in summer.In winter Orpin et al. (1985) observed a mean pH of 6.75 ± 0.18 in the rumen fluid.The small differences in pH observed could be due to seasonal changes in capacity of salivary secretion.
The in vitro rate of decrease of pH in the rumen contents could indicate differences in rate of VFA production.On NA in summer, pH decreased 0.35 units/h compared to as much as 0.8 units/h on NL in summer (calculated from White & Staaland, 1983) and 0.48 units/h in autumn on NL (Fig. 2).
The difference in ruminal pH rate of decrease between seasons and locations are not simply explained by differences in plant quality (Table 1).As based on WSC and CP levels, the highest rate of change should be NA in summer, NA in winter and then NL in autumn (Fig. 1).Ruminal pH also reflects VFA production.In the DFC, pH was lowest (P = 0.05) in winter at NA indicating a more active microbial environment in this fermenting chamber compared to that found in the other seasons and areas (Tables 2 and 3).This is also reflected in the relatively high production rates of VFA from the DFC in this area in winter (Tables 4 and 5).On NL the DFC pH was not different in autumn and winter (Table 2).Mathiesen et al. (1987) found the pH in the caecum of Svalbard reindeer on NL to be low in summer (6.81 ± 0.12) and high in winter (7.14 ± 0.26) which contrasts our findings, which could be due to local variation in plant species and plant quality.Orpin et al. (1985) and Mathiesen et al. (1987) found that the ruminal and caecal bacterial composition is strongly affected by changes in diet quality and availability which in turn is reflected in the pattern and production rate of VFA and, hence, the pH.Regulation of pH in the DFC is, however, not well understood.PH in rumen and DFC influence on absorption of VFA.The mechanism for absorption of VFA across the caecal mucosa is by simple diffusion (Myers et al., 1967), but rate of absorption in the rumen and caecum of individual VFA was decreased 40-67% by increasing pH from 4.5-7.2 in the rumen (Dijkstra et al., 1993), and decreased by 44% when increasing the pH from 6.2 to 7.5 in the caecum (Myers et al, 1967).

Concentration and production rate of ruminal VFA
The concentration and composition of VFA in the Svalbard reindeer rumen (Table 2 and 3) is comparable to that found in the rumen fluid of domestic ruminants eating a poor quality forage (Hungate, 1966) and of domestic tuminants grazing in a thornbush savannah (Lechner-Doll et al., 1991).In ruminants the VFA concentration and production rate is determined by food quality and quantity (Hungate, 1966) No positive correlation between the chemical composition of rumen contents in the animals investigated and [VFA] could be observed (Table 1, 2 and 3), even if there were differences in plant cell wall constituents.The difference in ruminal chemical composition among locations is, however, reflected in ruminal VFA production (Table 1, 4 and 5 4, 5 and Fig. 3).
The difference probably reflects differences both in plant quality and availability (Weston & Hogan, 1968).These data support our understanding of the marginal summer and winter pasture where the availability seem to be low both summer and winter, but the quality seem to be high in NA.The low ruminal production tate on NL in winter seem to be related to a high proportion of ruminal plant cell wall and mosses (54%) (Table 1, S0rmo et al,   unpubl.).The concentration and pool size of VFA seem to be correlated with VFA production rate in domesticated ruminants (Weston & Hogan, 1968;Leng & Brett, 1966;Leng et al., 1968).This seems not to be the case in arctic ruminants like reindeer with strong seasonal changes in food intake and plant quality.Ruminal VFA concentration and pH were high in reindeer on NL in winter (Table 2), but VFA production rate was low (Table 4, Fig. 3).estimates of the rumen epithelium (Josefsen et al., 1996).We assume the absorptive surface of rumen epithelium in Svalbard reindeer changes from summer to winter in a similar pattern, and the change could influence on the rate of VFA absorption.

Concentration and production rate of DFC VFA.
Large amounts of the cell wall carbohydrates like cellulose and hemicellulose partially escape rumen digestion and thus become available for fermentation by DFC bacteria (Van Soest, 1994).The proportion of hemicellulose digested in the latge intestine is higher than that of cellulose (Ulyatt et al., 1975;Van Soest, 1994).This apparent resistance of hemicellulose for rumen fermentation could be an important factor limiting the rate of breakdown of cell wall carbohydrates in the rumen.Gray (1947) found that 17% of the digestible cellulose was digested in the caecum, 70% in the rumen and 13% in the colon of sheep.Chemical analyses of the plants in the rumen contents of the Svalbard reindeer suggest that the amount of plant cell wall is high on NL and low on NA ( ).The production rate of VFA in the DFC on NA in summer was, however, only half of that found in winter and contributed only 2,7% of the total energy produced from VFA in the GI system in summer compared to 17% in winter (Table 5).
This is probably due to the botanical composition of the plants eaten, since rumen contents consisted mostly of Saxifraga spp.(55%) and Draba spp.
(12%) in winter compared to grasses (53%) in summer (S0rmo et al., unpubl.).Caecal volume relative to ruminal reticulum volume is higher in concentrate selectors than in grazers (Hofmann & Stewart, 1972;Hofmann, 1973;1985;1989).In African ruminants there is no experimental evidence to support the view that caecal digestion is relatively more important in concentrate selectors than in grazers or intermediate feeders (Hoppe, 1984).We found, however, that in the Svalbard reindeer the DFC contributes substantially (17%) to the total production of VFA in winter in an area where food is scarce and that the importance of DFC seems less when food is abundant.In sheep, caecal VFA contain a relatively higher proportion of branched chain acids compared to the rumen, indicating greater conversion of protein to VFA (0rskov et al, 1970).In the Svalbard reindeer the concentration of branched, chained VFA in the DFC did not indicate a high fermentation of protein (Table 6).

Energetics
The energy contributions from VFA from the rumen and DFC of reindeer on NA in winter and summer were 8.0 and 67.3%, respectively, more than the fasting metabolic rate of reindeer (112.3KJ/kgW 07 7d) predicted by Nilssen et al. (1984), regardless of season.On NL in summer, VFA con-tributed as much as 80.5% more than the fasting metabolic rate (White & Staaland, 1983;Nilssen et al, 1984) but in autumn, VFA production contributed only 67.4% of fasting metabolic rate.
Annison & Armstrong (1970) estimated that in ruminants 50-70% of the basal metabolic rate (BMR) originates from VFA production from the digestive system.On NL in winter, however, the VFA production rates from the rumen and DFC was very low contributing 0-45.7% to the fasting metabolic rate calculated from Nilssen et al. (1984).This has also been observed for other ruminants like the Klipspringer (Oreotragus oreotragus) where the ruminal VFA as % of BMR contributed only 11% (Hoppe, 1984).Orpin et al. (1985)  Manuscript received 23 October, 1996accepted 21 April, 1991 Appendix: Tables 1-6 Table 1.Chemical composition of organic dry matter in the rumen of Svalbard reindeer (%, mean ± SD).Rangifer, 17 (2), 1997

A
total of 7 animals were investigated on NL (3 males in winter (April) 1994, 2 females in winter (April) 1995 and 2 females in autumn (October 1995)).On NA a total of 11 adult animals were investigated.Of these were two slaughtered in winter (April) 1994, one male and one female.In August 1995, 9 animals were investigated; 3 males and 6 females.All animals were shot while grazing and all samples were taken in the field immediately after killing.The whole animal and the ruminal and DFC contents were weighed and pH was recorded in the contents immediately after killing.The rumen contents also include contents from the reticulum, the DFC consisted of caecum and the proximal part of the colon, before it enters the colon coil.Chemical analysesSamplesof rumen contents were frozen (-20 °C) after death of the animal and were brought to the laboratory in Tromsø where it was dried at 115 °C until constant weight and the dry matter content was estimated.Analyses of the plant cell wall fraction (hemicellulose, cellulose and lignin) which was calculated from values of neutral detergent fibre (NDF), acidic detergent fibre (ADF) and acid detergent lignin (ADL), were carried out using the techniques of Van Soest (1963 a; b), Van Soest & Wine rumen and DFC contents anaerobically, a curve could be constructed from the change in VFA levels in the sample during the incubation period (70-120 min).The slope of the curve represents the rate of production in vivo, at the time the sample was collected from the animal.By extrapolating the regression lines to zero, we could determine the concentration of VFA in the fermenting chambers at the time of death.The contents were mixed thoroughly between each sampling.A sub sample of rumen contents (about 100 g) was sieved through two layers of muslin.The filtrate (10 ml) was fixed in 5 ml 0.5M HC1 and was frozen in counting tubes (-20 °C) until analysis.Individually marked and weighed counting tubes, each containing 10 ml 0.5M HC1 were added approximately 5 g of DFC contents at different time intervals in the field, and were sealed and frozen (-20 °C).After arrival at the laboratory at Department of Arctic Biology at Tromsø, the tubes were weighted and the added amount of DFC contents in each tube was calculated before analysis.All samples were taken in duplicate.From the one-hour estimate of VFA production, a total 24 h production of VFA was calculated.

Fig. 2 .
Fig. 2. Regression lines plot for the pH in the rumen (O) and in the distal fermentation chamber (•) of Svalbard reindeer at Nordenskioldland (NL) and on Nordaustlandet (NA) different times of the year as a function of time.(W=winter.A=autumn.S=summer).

Fig. 3 -
Fig. 3-Regression lines plot for the concentrations of the main volatile fatty acids in the rumen (O) and in the distal fermentation chamber (•) of Svalbard reindeer at Nordenskioldland (NL) and on Nordaustlandet (NA) different times of the year as a function of time.(W=winter.A=autumn.S = summer).
Fig. 1).White & Staaland (1983) estimated a ruminal VFA rate of production corresponding to 575 kJ/kgW 07 7d on NL in summer.On NA in summer and on NL in autumn, ruminal VFA production represented 58 and 10%, respectively, of that found on NL in summer, which indicates a high forage Rangifer, 17 (2), 1997 VFA absorption across the rumen wall of Svalbard reindeer is low in winter when pH is high and could explain the high VFA levels measured.Lechner-Doll et al. (1991), observed that dilution rate, rumen fluid volume and absorption rate influenced VFA concentration far more than the production rate in seasonal Aftican domestic ruminants.Weston & Hogan (1968) also indicate that variations in diet and physiological state of the animal could influence on the blood flow supply to the rumen epithelium and could result in changes in absorption rates, pool size and rate of production of VFA relative to the VFA concentration.In Norwegian reindeer the absorptive surface of the rumen decreased between September and April with 48% determined by surface enlargement factor calculated that body fat in Svalbard reindeer can only supply 10¬ 30% of the daily energy expenditure during the winter and that the rest probably originated from the plants eaten.The mosses in the rumen contents of the NL reindeer did not seem to contribute significantly to energy production in winter.In conclusion, the seasonal and geographical differences in diet of the Svalbard reindeer are reflected in the fermentative activity in the rumen and DFC micro-organisms.In areas with abundant vegetation with high contents of plant cell walls but low WSC and protein levels, the production rate of VFA is low both in the rumen and DFC, even though the size of the DFC is large.In areas where cell wall material is low and WSC and CP relatively high, but plant availability is low, the production rate of VFA is high both in summer and winter and the DFC contributes as much as 17% of the total VFA produced in winter. in the analysis of fibrous feeds.IV.Determination of plant cell-wall constituents.-journal of the Association of Official Agric.Chemists.50: 50-55.Weston, R. H. & Hogan, J. P. 1968.The digestion of pasture plants by sheep.I. Ruminal production of volatile fatty acids by sheep offered diets of ryegrass and forage oats.-Austr.J. Agric.Res.19: 419-432.White, R. G. & Staaland, H. 1983.Ruminal volatile fatty acid producrion as an indicator of forage qualiry in Svalbard reindeer.-Acta Tool.Fennic.175: 61-63-0rskov, E. R., Fraser, C, Mason, V. C. & Mann, S. O. 1970.Influence of starch digestion in the large intestine of sheep on caecal fermentation, caecal microflora and faecal nittogen excretion.-Br.J. Nutr.24: 671-682.
CN CN ^H I/O ITS rCN m XT ON csTi d P CN P ON XT ITS ON XT ON i/o I/O r--I/O O o i/o ON r-ON 00 ON NO NO xr NO iwo NO GN NO ' -1 00 m wo od NO ni GN NO GN NO d 00 oo GN P 1-1 wo xr WO xr NO 00 xr >-l WO o oo