HAL Id: hal-00899820https://hal.archives-ouvertes.fr/hal-00899820

Submitted on 1 Jan 1996

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Effect of a high sulfur diet on rumen microbial activityand rumen thiamine status in sheep receiving a

semi-synthetic, thiamine-free dietLaurent Alvès de Oliveira, C Jean-Blain, V Dal Corso, V Bénard, A Durix, S

Komisarczuk-Bony

To cite this version:Laurent Alvès de Oliveira, C Jean-Blain, V Dal Corso, V Bénard, A Durix, et al.. Effect of a highsulfur diet on rumen microbial activity and rumen thiamine status in sheep receiving a semi-synthetic,thiamine-free diet. Reproduction Nutrition Development, EDP Sciences, 1996, 36 (1), pp.31-42. �hal-00899820�

Original article

Effect of a high sulfur diet on rumen microbialactivity and rumen thiamine status in sheepreceiving a semi-synthetic, thiamine-free diet

L Alves de Oliveira C Jean-Blain V Dal Corso,V Bénard, A Durix S Komisarczuk-Bony

Équipe associée Inra de physiopathologie du rumen, École nationale vétérinaire de Lyon,BP 83, 69280 Marcy-l’Étoile, France

(Received 3 April 1995; accepted 2 November 1995)

Summary-A semi-synthetic thiamine-free diet was used on weaned lambs to test the effect of a highsulfur level on the rumen, microbial activity and on the microbial production of thiamine. In vivo and invitro kinetic studies, as well as the determination of the thiamine concentrations and thiaminase activ-

ity in the rumen, were performed during the 16 week experiment. A high sulfur level (0.6%) in the diet,in comparison with a normal sulfur level (0.2%), did not modify the microbial activity of the rumen withthe exception of a slightly retarded decrease in the volatile fatty acid (VFA) rumen concentration. Therumen thiamine level and the thiaminase activity were not modified by the dietary sulfur level. In con-trast, the rate of sulfate reduction into sulfide in the rumen increased progressively with the 0.6% sul-fur diet. In conclusion, a high sulfur level (0.6%) in the diet of sheep did not modify the thiamine statusof the rumen. It strongly increased the production of sulfides but an adaptation period of several weekswas required by the rumen microflora to reduce sulfate at a maximal rate.

rumen / sulfur / thiamine / synthetic diet

Résumé ― Effet d’un régime excédentaire en soufre sur l’activité microbienne du rumen etsur le métabolisme ruminal de la thiamine chez le mouton recevant un régime semi-synthétiquedépourvu de thiamine. Un régime semi-synthétique dépourvu de thiamine a été utilisé chez l’agneausevré, pour étudier l’effet d’un apport excédentaire de soufre sur le métabolisme microbien et la pro-duction de thiamine dans le rumen. L’activité microbienne normale du rumen n’est pas modifiée par unrégime à 0,6 % de soufre comparé à un régime normal à 0,2 %, à l’exception d’une très légère et tar-dive diminution de la concentration ruminale en AGV. Des cinétiques effectuées in vivo et in vitroainsi que des déterminations de la teneur en thiamine et de l’activité thiaminasique dans le rumenont été effectuées pendant 16 semaines. La concentration en thiamine reste identique dans les deuxrégimes pendant toute la durée de l’expérience ; il n’y a pas non plus de modification de l’activité thia-

*

Correspondence and reprints

minasique. En conclusion un régime à 0, 6 % de soufre chez le mouton n’a pas d’effet sur la concen-tration ruminale de thiamine d’origine microbienne. Il provoque une très forte augmentation de la pro-duction de sulfures mais celle-ci n’est maximale qu’après une période d’adaptation de la micropopu-lation du rumen.

rumen / soufre / thiamine / régime synthétique

INTRODUCTION

An adequate concentration of sulfur (about0.2% of the dry matter (DM)) in the diet ofruminants is required to allow the rumenmicroflora to synthesize the sulfur aminoacids for microbial proteins and sulfur-con-taining vitamins such as thiamine, at an opti-mal rate. In contrast, an excess of dietarysulfur (above 0.4% DM) may induce anti-nutritional effects which could decrease theanimal’s performance and lead to clinicalsigns of intoxication and eventually death.This aspect has been reviewed by Kandylis(1984). The consequences of a high sulfurdiet depend not only on the dietary level ofsulfur but also on the type of sulfur com-pound, on its mode of administration andon the composition of the ration. A sulfurdietary concentration above 0.4% usuallydecreases both feed intake and animal per-formance. The effects of an excess of sulfur ron the rumen microbial digestion are poorlydocumented.

The acute toxicity of sulfur is mostlyrelated to a high sulfide production due tothe microbial reduction of sulfate in therumen. It consists of severe diarrhea, res-piratory distress and nervous symptomswith muscular tremors, a staggering gaitand, finally, coma. More recently, brainlesions identical to those observed in clas-sical thiamine-related cerebrocortical necro-

sis or polioencephalomalacia (CCN) havebeen described in ruminants receiving highsulfur diets. Numerous cases of CCN havebeen identified by veterinarians in France, inbeef cattle fattened with sugar beet pulpsenriched in aluminium sulfate. Several casesof CCN were observed in the USA in 6-18-

month-old cattle with high sulfate diets (Rais-beck, 1982). In Canada, field studiesreported outbreaks of CCN in cattle con-suming water with elevated sodium sulfatelevels (Harries, 1987; Gooneratne et al,1989a; Hamlen et al, 1993). Moreover, CCNhas been experimentally induced in sheepfed a 0.63% sulfur diet (Gooneratne et al,1989b; Rousseaux et ai, 1991 Experi-mentally, an intraruminal infusion of sodiumsulfide in the rumen can induce brain lesionssimilar to the classical lesions of CCN withinsome hours (McAllister et al, 1992). A pos-sible relationship between a high dietaryintake of sulfur and the animal’s thiaminestatus have been considered, as the patho-genesis of classical thiamine-dependentCCN, first described in 1961 (Terlecki andMarkson, 1961 is not completely under-stood. Some field studies intended to estab-

lish a relation between a high intake of sul-fur and a lowering of the blood thiaminecontent (Olkowski et al, 1991). Furtherexperimental investigations did not show arelation between high sulfur diets and dis-turbance of the thiamine status (Rousseauxet al, 1991; Olkowski, 1992). However, thediets used in these experiments either hada normal thiamine level or were enriched

with large amounts of this vitamin. In otherrespects, investigations on the effect of highsulfur diets on the microbial metabolism of

thiamine in the rumen are limited. In exper-iments in vitro a low nitrogen/sulfur ratio(Bick et al, 1978) or an addition of sulfate(Olkowski et al, 1993) increases the thi-amine-destroying activity of rumen juice ortotal rumen content cultures.

The aim of the present work was thereforeto investigate the effect of a high sulfur diet

on the rumen microbial activity and on itsthiamine status in weaned lambs. A semi-

synthetic thiamine-free diet was used in orderto avoid interference by dietary thiamine.

MATERIALS AND METHODS

Animals and experimental design

Seven female Texel lambs averaging 30 kg wereused in this experiment. Before the experimentthey were fed with a classical diet composed ofhay and concentrate. After a serological controlfor brucellosis and an antiparasitic treatment, theywere equipped with a permanent rumen canulae.They were adapted over a two-month period to asemi-synthetic diet, whose composition was iden-tical to the normal sulfur diet (NS) used during theexperiment, with the exception of the presenceof thiamine (2 mg/kg). They were then divided intotwo groups: a control group (three animals)received the semi-synthetic thiamine-free dietthroughout the experiment (16 weeks) with a nor-mal (0.2% sulfur content NS) (table I); and anexperimental group (four animals) received a semi-synthetic thiamine free diet with 0.4% sulfur (50%NS + 50% HS) for 4 weeks, and then the HS diet(0.6% sulfur) (table I) for the next 12 weeks.

All animals received their respective experi-mental diets one week before the beginning ofthe experiment.

Kinetic studies in the rumen

To compare the effect of the dietary level of sulfur ron the rumen microbial activity, the postprandialkinetics of rumen pH, volatile fatty acids (VFA),ammonia (NH3), lactate, sulfate and sulfide con-centrations were measured once a week in eachanimal during weeks 2, 4, 6 and 8, to test theeventual adaptation of the rumen micropopula-tion to the semi-synthetic diet. To allow an acutecomparison of the kinetics between both groups,the morning meal was not distributed on the dayof the test. As a substitute, 300 g of NS or HSdiet were mixed with 1 300 mL artificial saliva pre-heated at 39 °C and were rapidly introduced intothe rumen through the canulae. Hourly samplingsof rumen content were undertaken for 7 h afterthe introduction of the artificial meal.

Microbial metabolism of thiaminein the rumen

In a preliminary assay, no consistent postpran-dial variation of the rumen thiamine concentra-tion was observed. Subsequently this parameterwas determined 5 h after the morning meal, for fiveconsecutive days for each animal at weeks 2, 4,6, 9 and 16 of the experiment. Rumen sampleswere collected after withdrawal and homoge-nization of the total rumen content. This was sub-

sequently reintroduced into the rumen.

In vitro incubations at week 8 of the experi-ment were carried out according to the methodof Jouany and Thivend (1986). An inoculum com-posed of 100 mL of undiluted rumen juice and100 mL of filtered rumen juice taken from sheepreceiving either the 0.2% S diet (NS) or the 0.6%diet (HS) was diluted with artificial saliva up to avolume of 400 mL. Soluble starch, cerelose, cel-lulose and ammonium chloride were added. Twoamounts of sulfate were added: normal (200 mg/L)and high (600 mg/L). The pH variations, VFA andfermented organic matter (FOM) production, gasproduction, ammonia and sulfate utilization weremeasured at t= 0, 2, 4 and 6 h. FOM was calcu-lated with the Demeyer and Van Nevel formula(1979). Each incubation was repeated six times.

Analytical procedures

The rumen juice and in vitro incubation fluidswere analysed for VFA by gas chromatography(Jouany, 1982), for NH3 according to the Weath-erburn method (1967), for sulfate and sulfide withthe methods of Sorb6 (1987) and Cline (1969),respectively, and for lactic acid according to theenzymatic procedure of Gutmann and Wahlefeld(1974).

The total thiamine content of the rumen juicewas determined according to the thiochromemethod of Strohecker and Henning (1963) afterthe hydrolysis of the phosphorylated compoundsand purification on a ion-exchange column Amber-lite CG 50. The potential and actual thiaminaseactivity was determined according to the methodof Thomas (1986) using pyridine as a co-substrateto measure the potential thiaminase I activity.

Statistical analysis

Statistical analysis was carried out using theStatview° program. Repeated measures analy-sis of variance was used to identify the effects ofsulfur level, duration of the experiment and inter-action between these factors in the in vivo experi-ments. The same statistical treatment was usedto identify the effect of the origin of inoculum andthe level of sulfate in in vitro incubations. The cor-relation between the potential thiaminase activityand the level of thiamine in the rumen was deter-mined.

RESULTS

In vivo experiment t

Growth, ingestion and pathologicalstatus of animals

The mean weights of animals at the end ofthe experiment were 28.55 ± 1.90 kg and28.24 ± 1.54 kg in the NS and the HSgroups, respectively. Total gain weight was0.55 kg in the NS group and negative(-0.64 kg) in the HS group. Both groups lostweight during the first 6 weeks of the exper-iment because of the poor efficiency of theprotein-free purified diet and its poor palata-bility. Normal growth began on week 11.

Average daily food intakes were respec-tively 736 ± 72 g and 701 ± 54 g in the NSand HS groups. The two groups were not

significantly different.

Predominantly nervous symptoms, con-sisting of head pressing, blindness, hyper-salivation, polypnea, uncoordination anddepression, were observed on week 7 dur-ing an approximately 10 h period for onelamb of the HS group after the introductionof the artificial meal into the rumen for kineticmeasures. The peak rumen concentrationof sulfide measured at this time (80 mg/L)was the highest observed in the animals inthe experiment. Gross pathology andhistopathology subsequently demonstratedsymmetrical necrotic lesions of the grey mat-ter on the anterior cortex.

Kinetic studies in the rumen

Effect of dietary sulfur level on microbialactivity of the rumen

Figure 1 indicates the mean kinetics of pH,VFA, NH3 and lactate determined on weeks

2, 4, 6 and 8, in the NS group, and on week4 and 8 in the HS group. Statistical analysisof the NS group showed no significant dif-

ference between weeks for all parameters,indicating that the rumen microbial activitywas, for this diet, stable throughout theexperiment. Consequently, the use of thisdiet was justified as a basis of comparisonfor the investigation of the effects of the highsulfur level. No significant difference wasobserved beween the different periods ofHS (0.4 vs 0.6% S) or between HS and NSdiets for pH and ammonia. A non-significantdifference was noted for lactic acid at week4 (0.4% S) in comparison with week 8 (0.6%S) and the control group NS. VFA kinetics inthe HS group were significantly different(P= = 0.01) from the NS group but only dur-ing week 8.

Effect of dietary sulfur level on sulfateuptake and sulfide formation

The kinetics of the sulfate and sulfide con-centrations throughout the experiment in theNS group and for weeks 4 and 8 in the HS

group are indicated in figure 2. As sulfatewas rapidly converted into sulfide, a peak ofsulfide appeared between the first and thethird hour of the kinetic. No significant mod-

ification of the sulfate and sulfide kineticswere observed between weeks 2 and 8 ofthe experiment in the NS group. In contrastthe slopes of the sulfate utilization curves inthe HS group increased between weeks 2and 8 and a corresponding increase in theproduction of sulfide was observed (P <

0.001 ). A noteworthy increase in sulfide pro-duction was observed between weeks 6 and8 in the HS group (P < 0.001 The peak ofsulfide was about three times higher at week8 than at week 6 despite the fact that thesame concentration of sulfur (0.6%) was pre-sent in the diet during those periods.

Thiamine status in the rumen

The thiamine concentrations in the rumencontent are indicated in table II. No signifi-cant evolution was observed during the 16 6weeks of the experiment for both groups.No significant differences were foundbetween the HS and NS groups. The actualthiaminase activity remained very low(< 0.01 mU/mL/min) in all the samples.Large variations in the potential thiaminase I

activity were recorded throughout the exper-iment (table III) but no significant differencewas found between the NS and HS groups.No negative correlation was found betweenthe potential thiaminase I activity and thethiamine concentration in the rumen (r =0.078).

In vitro incubations

Effect of inoculum origin and sulfate con-centration on in vitro microbial activity

The kinetics of the pH variations, VFA, FOM,gas production and NH3 utilization in theNS and HS fermentors with an initial sulfurconcentrations of 200 mg/L are indicated infigure 3. The fermentations were not modi-fied by the origin of the inoculum (NS or HSdonor sheep). In other respects, these kine-tics were not modified by a 600 mg/L sul-fur concentration.

Effect of inoculum origin and sulfate con-centration on in vitro sulfate utilization

The utilization of sulfate in the flasks con-

taining 200 mg/L sulfur was considerably

enhanced (x 3.5) in the HS donor flasks incomparison with the NS one (fig 3). With the600 mg/L sulfur concentration, the rate ofsulfate utilization was five times higher forthe HS donor flasks than the NS. The originof the inoculum, therefore, had a very strongeffect on the utilization of sulfate; inoculumissued from HS sheep adapted to the highsulfur diet utilized sulfate more rapidly thaninoculum issued from NS sheep receiving anormal amount of sulfur in their diet. Increas-

ing the concentration of sulfur in the flasksalso increased the rate of sulfate uptake.

DISCUSSION

While the toxicological effects of high sul-fur diets on animals are on the whole well

known, data related to the particular effect ofthese diets on the rumen microbial diges-tion are very limited. Kennedy et al (1971)used an in vitro technique with a washedcell suspension and found no deleteriouseffect on starch digestion with sulfur con-centrations as high as 11 mg/mL. Using thesame technique, Hubbert et al (1958)observed only a slight effect on cellulosedigestion with a sulfur concentration of

1 mg/mL and a marked inhibition with2 mg/mL. No modification in the digestibilityof the organic matter, crude protein, etherextract or crude fiber were noted by Johnsonet al (1968) in lambs fed a 0.5% sulfur diet.In steers receiving an all-concentrate fin-ishing diet, Rumsey (1978) found non-con-sistent modification of the ruminal pH,ammonia, VFA and lactic acid concentra-tion with dietary levels of sulfur up to 0.98%.

Our in vivo and in vitro experimental find-ings are in agreement with these data. Invivo, a 0.6% sulfur diet had roughly no effecton the rumen metabolism. Moreover, ourresults indicated clearly that a high dietarysulfur level resulted in a progressive adap-tation in the rumen microflora’s ability toreduce sulfate into sulfide at a faster rate.The resulting increase in sulfide concentra-tion might have been responsible for the

small decrease in the VFA concentrationsobserved in the HS group on week 8. How-

ever, it was not possible to determinewhether this was related to a direct inhibitingeffect of sulfide on the overall microbial activ-

ity or to an increase in the uptake of ener-getic substrates by the sulfate-reducing bac-teria.

In another respect, this progressive adap-tation towards reducing sulfate into sulfidecould explain the two to three month delaybefore the onset of nervous symptoms andlesions similar to classical CCN that wereobserved with high sulfur diets (Gooneratneet al, 1989a). Direct administration of sodiumsulfide is capable of inducing the samepathological events within some hours(McAllister et al, 1992). It might also explainwhy in our experiment, only the sheep thatexhibited the greatest efficiency in reduc-ing sulfate into sulfide and consequentlyhad the highest rumen concentration of sul-fide, presented the nervous symptoms andlesions similar to CCN.

With natural diets containing variousamounts of thiamine, the concentration ofthiamine in the rumen content cannot beused as a reliable indicator of microbial thi-amine synthesis. In fact, when the dietaryamount of thiamine is low, the ruminal syn-thesis is very active. In contrast, when thereis a high input of dietary thiamine, the netmicrobial synthesis in the rumen (microbialsynthesis minus microbial degradation) canbe negative (Breves et al, 1981; Miller et al,1986). These variations are not reflected inthe concentration of thiamine in the rumenas there is no differentiation between thi-amine from either origin. The use of a semi-synthetic thiamine-free diet eliminated thisproblem as all the thiamine found in therumen was of microbial origin. As thiamineis not absorbed through the rumen wall(H61ler et al, 1977; Steinberg et al, 1977),its concentration in the rumen can be con-sidered to give good indication of the amountsynthesized by the microbes. The thiamine

concentration has often been measured inthe rumen. Different authors using differentkinds of diets have found average thiamineconcentrations within a range of 0.10 to1.9 mg/g of rumen content. The rumen con-centrations found in our experimental ani-mals are in agreement with these data, butif compared with values observed by ElHindi (1977) on sheep fed a similar protein-free, thiamine-free diet, they are roughlythree times higher. This apparent discrep-ancy could be partly explained by method-ological differences. In fact, almost all thethiamine is localized in microorganisms.Most of the bacteria are attached to dietaryparticles. Therefore determination of the thi-amine concentration from rumen juice con-taining only a part of the coarse feed mate-rial must give lower results thandeterminations done on the whole rumencontent.

Our results clearly indicated that themicrobial synthesis of thiamine remainedeffective with the semi-synthetic diet duringthe entire experimental period, either witha normal or a high sulfur level. The actualthiaminolytic activity in the rumen juiceremained very low with both levels of sul-fur despite that, theoretically, sulfite as anintermediate in the reduction of sulfate intosulfide can readily cleave thiamine. In fact,Bray (1969) indicated that sulfite is a verytransient intermediary which does not accu-mulate in the rumen. Considering the data ofBray as well as the negligible thiaminolyticactivity we found in the rumen juice it is

unlikely that a high sulfur diet increased thesulfite concentration in the rumen. On theother hand, high values of potential thiamin-ase-I activity recorded at weeks 10 and 16 inthe control group or at weeks 1, 2 and 3 inthe experimental group indicate that a highdestruction of thiamine would have occurredif an adequate co-substrate had been pre-sent. However, in the absence of a suffi-cient amount of co-substrate in the rumen

juice, no corresponding decrease in thiamine

was noted. In addition, the variations of thi-aminase I activity in our experiment werenot in relation to the sulfur level. These find-

ings can be compared with the data ofThomas et al (1990) who demonstrated highpotential thiaminase I activity in healthy aswell as in CCN sheep.

On the whole, our results are in agree-ment with those of Bick et al (1978) andOlkowski et al (1993). These authors foundthat high levels of sulfur did not impair thi-amine synthesis in vitro but slightlyincreased the level of thiamine destruction.

In our experiment, the net in vivo pro-duction of thiamine was not measured. It is

possible that the rate of thiamine synthesiswas high enough to maintain a normalrumen level even with a moderate increasein the thiamine destruction.

In conclusion, a large amount of sulfurin the diet, corresponding roughly to threetimes the normal level, had little negativeeffect on the rumen microbial activity andno effect on the level of thiamine in therumen. It is unlikely that such as amount ofsulfur could induce a thiamine deficiencyrelated to a lack of thiamine microbial pro-duction in the rumen. In addition, it must be

emphasized that the rumen ecosystemseemed much more resistant to a high sul-fide levels than the other tissues of the host

animal.

ACKNOWLEDGMENTS

The authors thank M Carcelen and C Huc fortheir skillful technical assistance. This work was

supported by a grant from the programme DGERFormation par la Recherche.

REFERENCES

Bick S, Breves G, Holler H (1978) Netto synthese vonmikrobiellem Protein und Thiamin im Panseninhalt

eiweissfrei ernahrter Schafe bei unterschiedlichen

Substrat-und Schwefelkonzentrationen in vitro. Zeitz

Tierphysiol 77erenanp Futtermittelk 41, 8-17 7

Bray AC (1969) Sulphur metabolism in sheep. li. The

absorption of inorganic sulphate and inorganic sul-phide from the sheep’s rumen. AustrJ Agric Res 20,739-748

Breves G, Brandt M, H61ler H, Rohr K (1981) Flow ofthiamin to the duodenum in dairy cows fed differentrations. J Agric Sci 96, 587-591

Cline JD (1969) Spectrophotometric determination ofhydrogen sulfide in natural waters. Limnol Oceanogr14, 454-458

Demeyer DI, Van Nevel CJ (1979) Effect of defauna-tion on the metabolism of rumen micro-organisms. BrJ Nutr 42, 515-524

El Hindi H (1977) In vitro Untersuchungen zur Umsatzvon Thiamin im Panseninhalt eiweissfrei ern6hrterSchafe. Thesis DM Vet Tierarztiiche Hochschule,Hannover, Germany

Gooneratne SR, Olkowski AA, Christensen DA (1989a)Sulfur-induced polioencephalomalacia in sheep: somebiochemical changes. Can J Vet Res 53, 462-467

Gooneratne SR, Olkowski AA, Klemmer RG, KesslerGA, Christensen DA (1989b) High sulfur-related thi-amine deficiency in cattle: a field study. Can Vet J 30,139-146

Gutmann I, Wahlefeld AW (1974) L-(+)-Lactate: deter-mination with lactate dehydrogenase and NAD. In:Methods of Enzymatic Analysis (HU Bergmeyer, ed),Academic Press Inc, New York, 2nd ed, vol 3, 1464-1468 .

Hamlen H, Clark E, Janzen E (1993) Polioencephalo-malacia in cattle consuming water with elevatedsodium sulfate levels: a herd investigation. Can VetJ 34, 153-158

Harries WN (1987) Polioencephalomalacia in feedlotcattle drinking water high in sodium sulfate. Can VetJ28, 717 7

H61ler H, Fecke M, Schaller K (1977) Permeability tothiamin of the sheep rumen wall in vitro. JAnim Sci44, 158-163

Hubbert F, Cheng E, Burroughs W (1958) Mineralrequirements of rumen microorganisms for cellulosedigestion in vitro. J Anim Sci 17, 559-568

Johnson WH, Meiske JC, Goodrich RD (1968) Influenceof high levels of two forms of sulfate on lambs. J AnimSci 27, 1166

Jouany JP (1982) Volatile fatty acids and alcohol deter-mination in digestive contents, silage juices, bacte-rial cultures and anaerobic fermenter contents. Sci

Aliment2, 131-144

Jouany JP, Thivend P (1986) In vitro effects of avoparcinon protein degradability and rumen fermentation.Anim Feed Sci Techn 15, 215-229

Kandylis K (1984) Toxicology of sulfur in ruminant:review. J Dairy Sci 67, 2179-2187

Kennedy LG, Mitchell GE, Little CO (1971) Influence ofsulfur on in vitro starch digestion by rumen microor-ganisms. JAnim Sci32, 359-363

McAllister MM, Gould DH, Hamar DW (1992) Sulphide-induced polioencephalomalacia in lambs. J CompPatho1106, 267-278

Miller BL, Meiske JC, Goodrich RD (1986) Effects ofgrain source and concentrate level on B-vitamin pro-duction and absorption in steers. JAnim Sci62, 473-483

Olkowski AA (1992) Role of thiamin status in sulphur-induced polioencephalomalacia in sheep. Res VetSci 52, 78-85

Olkowski AA, Rousseaux CG, Christensen DA (1991)Association of sulfate-water and blood thiamine con-centration in beef cattle: field studies. Can JAnim

Sci 71, 825-832

Olkowski AA, Laarveld B, Patience JF, Francis SI, Chris-tensen DA (1993) The effect of sulphate on thiamine-destroying activity in rumen content cultures in vitro.Int J Vit Nutr Res 63, 38-44

Raisbeck MF (1982) Is polioencephalomalacia associ-ated with high sulfates diet? J Am Vet Med Assoc180, 1303-1305

Rousseaux CG, Olkowski AA, Chauvet A, GooneratneSR, Christensen DA (1991) Ovine polioencephalo-

malacia associated with dietary sulfur intake. J VetMed A 38, 229-239

Rumsey TS (1978) Effects of dietary sulfur addition andsynovex-S ear implants on feedlot steers fed an all-concentrate finishing diet. JAnim Sci46, 463-477

Sorb6 B (1987) Sulfate: turbidimetric and nephelometricmethods. Methods Enzymol 143, 3-6

Steinberg W, Kaufmann W, Hagemeister H (1977) Mes-sungen zur Bestimmung der Resorption von Thi-amin in der Vorm5gen und Darm von Milchkühen.J Tierphysiol Tierernahr Futtermittelk 39, 282-288

Strohecker R, Henning HM (1963) Vitaminbestim-mungen. Merck AG Darmstadt Ed Chemie Gmbh,Weinheim, Germany

Terlecki S, Markson LM (1961) Cerebrocortical necrosisin cattle and sheep. Vet Rec 73, 23-27

Thomas KW (1986) Oral treatment of polioencephalo-malacia and subclinical thiamine deficiency with thi-amine propyl disulphide and thiamine hydrochloride.J Vet Pharmacol Ther 9, 402-411 1

Thomas KW, Kelly AP, Beers PT, Brennan RG (1990)Thiamine deficiency in sheep exported live by sea.Austr Vet J 67, 215-218 8

Weatherburn MW (1967) Phenol hypochlorite reactionfor determination of ammonia. Anal Chem 89, 971-974