Lait (1995) 75, 513-521© Elsevier/INRA

513

Original article

Effect of forage type on milk yield,chemical composition and clotting properties of milk

JB Coulon 1, P Pradel 2, 1Verdier 3

1 Laboratoire adaptation des herbivores aux milieux, INRA, Theix,63122 Saint-Genès-Champanel/e;

2 Domaine de la Borie, INRA, 15190 Marcenat;3 Laboratoire de recherches fromagères, INRA, 36, rue de Salers, 15000 Aurillac, France

(Received 7 March 1995; accepted 28 June 1995)

Summary - Fifty-Iour lactating multiparous cows were used in a 3 x 3 latin square trial. During each3-week period cows were offered a diet composed either 01 perennial ryegrass silage (11 kg dry mat-ter/dl, or second-cutting cocksloot hay (13 kg dry matter/dl or maize silage (11 kg dry matter/dl. Eachdiet was supplemented with a mixture 01 barley, soya-bean meal and urea in proportions calculated tokeep the overall rations isocaloric and isonitrogenous. The cows led grass silage yielded, respec-tively, 0.4 (P> 0.05) and 1.9 kg/d (P < 0.01) more milk than those led maize silage or hay, but their milkwas signilicantly poorer in proteins and caseins. This result is Iinked to a distribution 01 ingested energyin lavour 01 milk yield rather than the deposition 01 body reserves with the grass silage-based diets. Thecasein/protein ratio, calcium concentration, pH and rheological characteristics 01 the milk were notsignilicantly different between the types 01 lorages.

dairy cow / coagulation property / milk composition 1 feeding

Résumé - Effet de la nature de la ration de base sur la composition chimique du lait et son apti-tude à la coagulation. Cinquante-quatre vaches laitières multipares en milieu de lactation ont étéutilisées dans un essai en carré latin 3 x 3. Pendant chaque période de 3 semaines, les animaux ontreçu une ration de fourrage isoénergétique, à base d'ensilage de ray-grass anglais (11 kg de matièresèche/j), de foin de seconde coupe de dactyle (13 kg de matière sèche/j) ou d'ensilage de maïs (11kg de matière sèche/j). Chaque ration a été complémentée avec la même quantité d'un mélanged'orge, de tourteau de soja et d'urée dont les proportions ont été calculées de manière à ce que lesapports azotés soient voisins et non limitants dans les 3 traitements. Les vaches recevant de l'ensilaged'herbe ont produit respectivement 0,4 (P > 0,05) et 1,9 kg/j (P < 0,01) de lait de plus que cel/es rece-vant de l'ensilage de maïs ou du foin, mais leur lait a été significativement plus pauvre en protéines eten caséines. Ce résultat est lié à un partage de l'énergie ingérée plus favorable à la production de lait

514 JB Coulon et al

qu'à la reconstitution des réserves corporelles avec la ration à base d'ensilage d'herbe. Le rapportcaséines/protéines, la teneur en calcium, le pH et les caractéristiques rhéologiques du lait n'ont pas étésignificativement différents selon le type de ration de base.

vache laitière 1 aptitude à la coagulation 1 composition du lait 1nature de l'alimentation

INTRODUCTION

The cheesemaking ability of milk is a com-plex notion encompassing different aspects,in particular chemical composition, clottlnqproperties and capacity to produce a ripenedcheese. The influence of production factorson this cheesemaking ability is currently asubject of much debate, particularly in thecase of cheeses with "Appellation d'Orig-ine Contrôlée" (AOC), for which the possi-bilities of modifying the raw material duringprocessing are Iimited, if not forbidden. Theinfluence of production factors (feeding inparticular) on the chemical composition ofmilk (essentially fat and protein concentra-tion) has been studied extensively (Journetand Chilliard, 1985; Rémond, 1985; Sutton,1989; DePeters and Cant, 1992; Murphyand O'Mara, 1993). Amongst these stud-ies, a certain number demonstrated that thetype of forages could lead to differences inanimal performances. However, in themajority of these studies, the effects couldbe attributed to modifications in nutrient sup-ply, particularly energy, linked to a change inforage type (Roffler et al, 1967; Holter et al,1976; Nicks et al, 1979; Sertilson andSursted, 1983; Sroderick, 1985; Coulon etal, 1985). Nevertheless, several trialsrevealed that for a same nutritive supply,grass silage-based rations lead, compari-tively to hay-based or maize silage-basedrations, to higher milk yields, due to a dis-tribution of ingested energy more favourableto milk secretion than to buildup of bodyreserves (Vérité and Journet, 1971;Sp6rndly, 1989). Under these conditions,due to a dilution effect, the fat and proteinconcentrations are subsequently lower with

grass silage-based rations (Rémond, 1978;Coulon and Garel, 1993). These modifica-tions can affect both the ability of milk tocoagulate, for which protein concentration isa determining factor (Remeuf et al, 1991;Macheboeuf et al, 1993; Martin and Coulon,1995), and cheese characteristics.

The aim of this study, conducted with 3different genetic types, was to describe andanalyse performances in production, chem-ical composition and clotting properties ofmilk in cows fed, at identicallevels of nutri-tive supply, a diet based on hay, grass silageor maize silage. The milk yielded by theseanimais was then used in the manufacturingof Saint-Nectaire cheeses. The results con-cerning the characteristics of these cheesesfigure in another publication (Verdier et al,1995).

MATERIALS AND METHODS

AnimaIs and feeding

Fifty-Iour multiparous dairy cows, belonging tothe breeds Holstein (n = 18), Montbéliarde (n =18) and Tarentaise (n = 18), were used. Theseanimais had calved between 11 November and 10January (5 December on average). In early lac-tation, ail cows were led a diet 01 grass silagesupplemented according to INRA recommenda-tions (1989). Three weeks prior to the beginning01 the experiment, ail animais were led, inrestricted quantities, a diet composed 01 second-cutting cocksloot hay (H), perennial ryegrasssilage (GS) and maize silage (MS), in equal pro-portions (ie around 4 kg dry matter [DMVd 01 eachlorage). The characteristics 01 the different leedsare specilied in table 1. The trial began on 12February. Three groups 01 6 cows per breed were

Forage type and milk composition

Table 1. Composition and nutritive value of forages.Composition chimique et valeur nutritive des fourrages.

515

Ryegrass si/age Maize si/ageCocksfoot 2nd eut hay

Dry matter (DM, %)Organic matter (g/kg DM)Crude protein (g/kg DM)Crude fiber (g/kg DM)Organic matter digestibilityUFL (/kg DM)PDIN (g/kg DM)PDIE (glkg DM)pHNH3-N (% total N)

21.79131632910.740.9298833.96.5

85.29111043030.670.776580

32.194992

2140.700.8655654.08.7

formed on the basis of production performancesof the 3 pre-experimental weeks and of thegenetic variant of the K casein. Three experi-mental periods, each of 3 weeks, were then con-ducted, du ring which each group was sucees-sively fed the 3 types of forage. To avoid digestivedisorders, 1 kg of hay was incorporated into themaize silage-based ration. Throughout eachperiod, the quantities of each forage providedwere calculated so that energy supplies wereidentical. They were, respectively, 11, 13 and 11kg DM/d for groups GS, H and MS. The variousdiets were supplemented with a mixture of barley,soya-bean meal and urea. The quantities of con-centrate given were determined for the wholeexperiment, so as to cover the maintenance andmilk production requirements monitored over thepreexperimental period, according to INRA rec-

., ommendations (1989). These quantities remained. unchanged for any given animal throughout the

... experiment, and from one group to the other. Therespective proportion of each ingredient in theconcentrate differed according to the type of for-age, so that the nitrogen supply levels were asclose as possible and non-restrictive. On aver-age, concentrate was composed of 90% barleyand 10% soya-bean meal in GS and H treat-ments, and of 75% barley and 25% soya-beanmeal in MS treatment. One hundred and fifty g/dof urea was added fn H and MS treatments, and100 g/d in GS trèatment: Changes from the dif-ferent diets were carried out over 3 d. On 25 April,ail animais were again fed the preexperimentalmixed diet, for a 2-week period. Throuqhout the

entire experiment, the cows were given 200 gld ofa minerai additive (6P-22Ca).

Samples and measurements

The quantity of milk produced was individuallyweighed alter each milking and the fat and proteinconcentrations were determined 5 d a week, ateach mil king (infrared method). Concentrate andforage intakes were measured individually everyday. The dry matter content of the forages wasmeasured 5 d a week. DM content of silage wascorrected to allow for DM losses during oven-dry-ing according to Dulphy and Demarquilly (1981).The chemical composition of the forages wasdetermined 3 times du ring the trial, as were thedigestibilities of dry and organic matter, by usinggroups of 4 wether sheep per forage type du ringmeasurement periods of 1 week alter a 2-weekadaptation period. These data enabled the cal-culation of the nutritive value of the forages(feed unit for lactation [UFL] and protein digestedin the small intestine when energy [PDIE] ornitroqen [PDIN] is not Iimiting in the rumen),according to INRA recommendations (Andrieuand Demarquilly, 1987) (table 1). The cows were

. weighed once a week, on the Tuesday, at thebeginning of the afternoon.

Once in each period, in an average sampleof the evening milking of the 6 cows belongingto each group (whose milk would be used in themanufacturing of Saint-Nectaire cheese [Verdier

S16 JB Coulon et al

et al, 1995)), milk pH (at 20°C), milk protein con-centration and urea concentration (colorimetricmethod using dimethyl amino benzaldehyde)were measured. Clotting properties (rennet clot-ting time, curd firming time, curd firmness) werealso measured, using a Formagraph apparatusaccording to the method proposed by McMahonand Brown (1982). Milk (10 ml) was added with200 J.!! of a 13 mg activated chymosin/l rennetsolution. Measurements were performed at themilk natural pH, then after pH standardization at6.60 by addition of a few drops of a 10% lacticacid solution. Ali measurements were duplicated.Total bacterial count (method NF V08051),somatic cell count (Somacount, Bentley), butyricspore count (Cerf and Bergère, 1968) and col-iform count (identified on desoxycholate lactoseagar [Leifson, 1935)) were measured in each ofthe average milk sampi es. As the levels of thesedifferent counts were always low, none of thesamples were discarded. During the last week ofeach period, individual milk samples were takenat the morning milking for the analysis of casein(Rowland's method [1938)) and calcium (flameemission method [International Organization forStandardization, 1987)) concentrations.

Energy and nitrogen balance were calculatedfrom the difference between food input and theanimais' requirements, according to INRA rec-ommendations (Andrieu and Demarquilly, 1987;Vermorel, 1989; Vérité and Peyraud, 1989).Energy supply was corrected for the decrease indiet energy value resulting from the increasedfeeding level and higher proportion of concen-trate in the diet (Vermorel, 1989). The energycontent of the milk produced was calculated fromits fat and protein concentrations according tothe formula proposed by Sjaunja (1989). Theenergy value of live-weight gain was estimatedas 4.5 UFL per kg (Chilliard et al, 1987).

Analysis of results

The analysis of results was carried out using themean data of the last week of each experimentalperiod. These data were processed by analysis ofvariance (SAS, 1987). Fixed effects included inthe model were the type of forage, breed, animal(nested within breed) and period, as weil asbreed x treatment interaction. In the absence ofsignificant breed x treatment interaction, onlyresults concerning the type of forage are pre-sented.

RESULTS

As planned, the energy and nitrogen sup-plies were never limiting (table Il). Through-out the trial, the animais showed positiveenergy and nitrogen balances. The energyand nitrogen supplies were, however, slightlyhigher in groups MS and H because of aslight overestimation of the POlE value ofgrass silage and a lower ingestion than fore-casted (10 vs 11 kg DM/d), essentially dueto a slight overestimation of DM content ofthe offered feed. The passage from one typeof diet to another was not accompanied bymarked refusais of feed.

Cows fed grass silage (GS) yielded 004(P> O.OS) and 1.9 kg/d (P < 0.01) more milkthan those fed maize silage (MS) or hay(H), respectively. Their milk was significantlypoorer in proteins and caseins (P < 0.01)than those belonging to groups MS and H.The proportion of caseins in the proteinswas identical between ail 3 rations (82%).The animais fed maize silage presentedhigher fat concentration than those fed grasssilage or hay (+ 1A and +1.7 g/kg, respec-tively, P < 0.01). Overall, the yield of milksolids was similar for groups MS and GS,and about 10% lower for group H (table Il).The calcium concentration of the milk wassirnilar between ail 3 treatments. The ureaconcentration of the milk was high, andgreater (P < 0.01) in group GS in compari-son to MS and above ail H.

The pH of the milk was identical in ail 3treatments, as were the majority of its rhe-ological characteristics (table III). The c1ot-ting time of the milk at natural pH was, how-ever, slightly shorter with the hay-based diet(P < O.OS), and the firming time at stan-dardized pH slightly shorter with the maizesilage-based diet (P < O.OS).

The cows fed hay gained more weightthan those fed grass silage or maize silage(+284, +48 and -S7 g/d, respectively,P< O.OS).

Forage type and milk composition 517

DISCUSSION as already reported relative to hay (Journetand Dulphy, 1973; Burstedt and Lingvall,1977; Sp6rndly, 1989; Coulon and Garel,1993) or maize silage (Vérité and Journet,1971). In this trial, differences between milkproduction of the grass silage diet and the

This trial confirmed the beneficial effect ofgrass silage-based diets on milk productionto the detriment of milk solid concentration,

Table Il. Food intake, milk production and live-weight change data.Quantités ingérées, production laitière et variations de poids vif.

Ryegrass Cocksfoot Maize RSDIsi/age 2ndcuthay si/age

IntakeForage intake (kg DM/d) 10.0 12.4 11.42 0.6Concentrate intake (kg DM/d) 4.9 4.9 5.0 0.2Energy supply (UFUd) 14.P 14.6b 14.7b 0.6PDIN supply (g/d) 1635a 1534b 1623a 51POlE supply (g/d) 1386a 1536b 1443c 47

BalancesEnergy balance (UFUd) 0.8a 2.1b 1.3c 0.6PDIN balance (g/d) 326 316 315 60POlE balance (g/d) na 318b 135c 59Live-weight change (g/d)3 48ab 284a -57b 673

Milk productionMilk yield (kg/d) 18.9a 17.0b 18.5a 1.0Fat concentration (g/kg) 40.0a 39.7a 41.4b 1.8Protein concentration (g/kg) 31.3a 31.8b 32.2c 0.9Casein concentration (g/kg) 25.4a 26.2b 26.2b 1.0Casein:protein ratio 4 0.82 0.82 0.82 0.01Fat yield (g/d) 756a 676b 766a 43Protein yield (g/d) 588a 538b 593a 35Calcium concentration (g/kg) 1.30 1.31 1.33 0.08Urea concentration

(mg/100 ml)5 41.3a 30.2b 34.0c 3.5

1 Residual standard deviation. 2 Included 0.8 kg DM of hay (see Materials and Methods). 3 Calculated between the1st and the 3rd week of each period. 4 This ratio was computed by using the protein content of the rnilk sampleused for casein content determination (30.8, 31.9 and 31.9 g/kg in groups GS, H and MS, respectively). This proteincontent was slightly different fram that computed on the samples collected during the last week of each period,which figures in the table. 5 Measured on pooled sampi es. abc Values with different superscripts within rows differ signi-ficantly (P < 0.05).1 Écart type résiduel. 2 Y compris 0,8 kg MS de foin (voir Matériel et méthodes). 3 Calculé entre la [re et la 3e

semaine de chaque période. 4 Ce rapport a été calculé en utilisant le taux protéique de l'échantillon utilisé pour la déter-mination du taux de caséine (respectivement 30,8,31,9 et 31,9 glkg dans les lots GS, H et MS). Ce taux protéiqueest légèrement différent de celui qui figure dans le tableau et qui est la moyenne des valeurs relevées au cours dela dernière semaine de chaque période. 5 Mesurée sur des échantillons poolés. abc À l'intérieur de chaque ligne, lesvaleurs suivies d'une lettre différente sont significativement différentes (P < 0,05).

518 JB Coulon et al

Table III. Clotting properties 01 milk.Aptitude à la coagulation du lait.

Ryegrass Cocksfoot Maizesi/age 2ndcut hay si/age

Protein concentrations (g/kg) 33.38 33.18 34.5bpH2 6.73 6.72 6.72

At natural pHRennet clotting time (min) 12.28 11.4b 12.18Curd lirming time (min) 5.9 5.2 5.0Curd lirmness (mm) 35.2 36.2 34.6

At standardized pH (6.60)Rennet clotting time (min) 10.1 9.9 10.0Curd lirming time (min) 4.48 4.48 4.1b

Curd lirmness (mm) 33.7 33.0 32.8

RSOI

0.80.02

0.80.75.5

0.60.46.1

1 Residual standard deviation. 2 01 the mean sample on which rheological characteristics were measured. ab Valueswith different superscripts within rows dilfer significantly (P < 0.05).1 Écart type résiduel. 2 De l'échantillon moyen sur lequel les caractéristiques rhéologiques ont été mesurées.ab À l'intérieur de chaque ligne, les valeurs suivies d'une lettre différente sont significativement différentes (P < 0,05).

other 2 diets were probably underestimated,and that of protein concentration overesti-mated, because of the slightly lower energysupply in the grass silage ration, relative tohay and maize silage rations. Indeed,energy supply is an important factor of vari-ation of milk protein concentration, evenwhen requirements are covered (Coulonand Rémond, 1991). This effect of the typeof diet on animal performances can beexplained 1) by a dilution effect of milk solidssecreted; this is the case between the maizesilage-based and the grass silage-baseddiets, for which the yields of secreted solidsare identical, although the concentrationsare significantly different; 2) by a differentdistribution of energy between milk and bodyreserves. Although it is always difficult tocalculate variations in liveweight over shortperiods, our results showed that animaisfed hay gained more weight than those fedgrass or maize silage. In the grass silagegroup, 58% of the energy intake was

excreted into milk, compared with 51 % inthe hay group and 56% in the maize silagegroup.

The reasons for the differences in thedilution of milk solids and in energy distri-bution according to type of forage have notyet been clearly established. They may belinked first to the balance in end productsof ruminai digestion (Saama et al, 1993):diets producing a high aceticlpropionic ratioin the rumen are generally beneficial to milkyield, and detrimental to weight gain andprotein concentration (Ekern and Reid,1963; Vérité and Journet, 1971; Thomas,1984a). The reverse is true when this ratiois low (Rémond and Journet, 1972). Dietswith a high acetic + butyric/propionic ratioare generally beneficial to fat concentration(Journet and Chilliard, 1985). Secondly, theobserved differences may originate from dif-ferences in nitrogen supply of the diets.A high level of nitrogen supply may favourmilk production (Vérité and Peyraud, 1989)

Forage type and milk composition

without modifying fat and protein concen-tration (Rémond, 1985). Such a hypothesisis unlikely in this trial where POlE supplieswere slightly lower in group GS than in groupH. Thirdly, the metabolic utilization of thenutriments could be different, due to nitrogennutrition of cows (Vérité and Journet, 1975).A low proportion of essential amino acids(lysine and methionine) in total POlE canaffect protein concentration without modi-fying milk yield and fat concentration(Rulquin et al, 1993). However, in this trial,this hypothesis cannot be retained: methio-nine supply, computed accordinq to INRA-tion software (Coulon and Faverdin, 1995)was similiar in the 3 diets (1.84% POlE) andlysine supply was higher in the GS group(6.94% POlE) than in the H (6.89% POlE) orMS (6.91% POlE) groups. Finally, a spe-cific effect of the species cropped mighthave occurred (Oionne et al, 1981; Murphyand O'Mara, 1993) with or without alteringthe composition of the volatile fatty acid mix-ture in the rumen (Thomas, 1984b). Thebeneficial effect of the maize silage-baseddiet on milk fat concentration is th us well-established (Hoden et al, 1985). It is linkedboth to the relative richness in fatty acidsof this forage and to the ruminai fermenta-tions which it induces, favouring the syn-thesis of short or medium chain milk fattyacids. In the same way, an increase in milkyield to the detriment of body reserves andfat concentration was observed in thechangeover from a ryegrass-based diet to acocksfoot-based (Oemarquilly, 1963) orwhite c1over-based diet (Thompson et al,1985).

The type of basal diet supplied had nosignificant effect on either the pH of milk orits rheological characteristics, in agreementwith our former observations (Coulon andGarel, 1993; Coulon, 1995). This lack ofeffect between the GS and H groups ispartly due to the small differences in milkprotein concentration of the average sam-pies on which the rheological measurementswere carried out. Milk protein concentration

519

is, indeed, an essential factor of the variationin firming time and in curd firmness (Remeufet al, 1991; Martin and Coulon, 1995). Whenthe change in diet type is accompanied by alarge modification in milk protein concen-tration, its clotting ability is also modified(Grandison et al, 1985). Nevertheless, inour study, the maize silage-based diet,whose protein concentration was the high-est, did not show a much greater ability tocoagulate than the others.

Considering the forages selected, ourstudy does not allow the differentiation ofthe respective effects of storage method(silage or hay) and of type of flora (ryegrassor cocksfoot) on the characteristics of themilk. lt does reveal, however, that besidesthe well-known quantitative factors (nutri-tive supply level of the diet), qualitative fac-tors can significantly modify these charac-teristics. Comparisons conducted withidentical forages cropped in the form ofsilage or hay should enable the clarificationof the roles of these different factors.

ACKNOWLEDGMENTS

We thank E Albaret, M Javorek and their staff forthe samplings and measurements, and the PôleFromager AOC Massif-Central for supporting thestudy.

REFERENCES

Andrieu J, Demarquilly C (1987) Valeur nutritive deslourrages : tables et prévision. Bull Tech GRZV TheixINRA 70,61-74

Sertilsson J, Surstedt E (1983) Ellects 01 conservationmethod and stage 01 maturity upon the leeding value01 lorages to dairy cows. 1. Forage intake and effects01 concentrate/lorage ratio. Swed J Agric Res 13,189-200

Sroderick GA (1985) Allalla sil age or hay versus cornsilage as the sole lorage lor lactating dairy cows.J Dairy Sei 68, 3262-3271

Surstedt E, Lingvall P (1977) The effect 01 conservationsystem and stage 01 development 01 herbage on

520 JB Coulon et al

milk production. Proc 13th Int Grass/and Gongr,Liepzig, Germany, 610-616

Cerf 0, Bergère JL (1968) La numération des sporesde Clostridum et son application au lait et aux pro-duits laitiers. Numération des différents groupes deGlostridium. Lait 48, 501-519

Chilliard Y, Rémond B, Agabriel J, Robelin J, Vérité R(1987) Variations du contenu digestif et des réservescorporelles au cours du cycle gestation-lactation.Bull Tech GRZV Theix INRA 70,117-131

Coulon JB (1995) Effets respectifs du changement del'alimentation et de l'environnement lors de la mise àl'herbe sur la composition chimique et l'aptitude àla coagulation du lait. Ann Zootech 44, 153-160

Coulon JB, Rémond B (1991) Variations in milk outputand milk protein content in response to the level ofenergy supply to the dairy cow: a review. Livest ProdSei 29, 31-47

Coulon JB, Garel JP (1993) A note on the effect of for-age type on the yield, chemical composition andclotting properties of milk. Anim Prad 57, 495-499

Coulon JB, Faverdin P (1995) Vaches lattières. ln: INRA-tion, logiciel de ca/cul et d'analyse de rations pour lesruminants, version 2.6 (J Agabriel, P Champciaux, CEspinasse, eds) CNERTA, Dijon, France

Coulon JB, Garel JP, Hoden A, Journet M, Liénard G(1985) Production laitière en zone de montagne:effets pluriannuels du type de ration hivernale et duniveau de complémentation. Bull Tech GRZV TheixINRA 61,31-48

Demarquilly C (1963) Influence de la nature du pâturagesur la production laitière et la composition du lait.Ann Zootech 12, 69-104

DePeters EJ, Cant JP (1992) Nutritionnal factors influ-encing the nitrogen composition of bovine milk: areview. J Dairy Sci 75, 2043-2070

Dionne JL, Roy G, Pelletier G, Genest J, Tremblay A,Fernet C (1981) Paturage versus ensilage pour laproduction laitière. Gan J Plant Sei 61,915-927

Dulphy JP, Demarquilly C (1981) Problèmes particuliersaux ensilages. In : Prévision de la valeur nutritivedes aliments des ruminants (C Demarquilly, ed)INRA Publications, Versailles, France, 81-104

Ekern A, Reid JT (1963) Efficiency of energy utilizationby young caille ingested diets of hay, silage, andhay supplemented with lactic acid. J Dairy Sci 46,522-529

Grandison AS, Manning DJ, Thomson DJ, AndersonM (1985) Chemical composition, rennet coagula-tion properties and f1avour of milks from cows graz-ing ryegrass or white clover. J Dairy Res 52, 33-39

Hoden A, Coulon JB, Dulphy JP (1985) Influence del'alimentation sur la composition du lait. 3. Effets desrégimes alimentaires sur les taux butyreux et pro-téiques. Bull Tech GRZV Theix INRA 62, 69-79

Holter JB, Urban WE, Davis HA (1976) Haycrop silageversus hay in a mixed ration for lactating cows.J Dairy Sci 59,1087-1099

Institut National de la Recherche Agronomique (INRA)(1989) Ruminant nutrition. Recommendedallowances and feed tables (R Jarrige, ed) INRAand John Libbey Eurotext, London, Paris

International Organization for Standardization (lOS)(1987) Dried milk. Determination of sodium andpotassium contents. Flame emission spectrophoto-metric method. ISO/DIS 8070

Journet M, Dulphy JP (1973) Utilisation des ensilagesd'herbe par les vaches laitières. Fourrages 56, 71-87

Journet M, Chilliard Y (1985) Influence de l'alimentationsur la composition du lait de vache. 1. Tauxbutyreux: facteurs généraux. Bull Tech GRZV TheixINRA 60, 13-23

Leifson E (1935) New culture media based on sodiumdesoxycholate for the isolation of intestinal patho-genes and for the enumeration of colon bacilli in milkand water. J Patho/ Bacterio/40, 581

Macheboeuf D, Coulon JB, D'Hour P (1993) Effect ofbreed, protein genetic variants and feeding on cows'milk coagulation properties. J Dairy Res 60, 43-54

McMahon DJ, Brown RJ (1982) Evaluation of Forma-graph for comparing rennet solutions. J Dairy Sei65,1639-1642

Martin B, Coulon JB (1995) Facteurs de production du laitet caractéristiques des fromages. 1. Influence desfacteurs de production sur l'aptitude à la coagula-tion des laits de troupeaux. Lait 75, 61-80

Murphy JJ, O'Mara F (1993) Nutritional manipulation ofmilk protein concentration and its impact on the dairyindustry. Livest Prod Sci35, 117-134

Nicks B, Bienfait JM, Lambert J, Nogarede P, Forceille MJ(1979) Comparaison de la valeur alimentaire du foindéshydraté et de l'ensilage d'herbe dans des rationspour vaches laitières. Ann Méd Vét 123, 351-359

Remeuf F, Cossin V, Dervin C, Lenoir J, Tomassone R(1991) Relations entre les caractéristiques physico-chimiques des laits et leur aptitude fromagère. Lait71,397-421

Rémond B (1978) Influence de l'alimentation et de lasaison sur la composition du lait. In : La vache laitière.INRA ed, Paris, France, 231-244

Rémond B (1985) Influence de l'alimentation sur la com-position du lait de vache. 2. Taux protéique: facteursgénéraux. Bull Tech GRZV Theix /NRA 62, 53-67

Rémond B, Journet M (1972) Alimentation des vacheslaitières avec des rations à forte proportion d'ali-ments concentrés. II. Comportement alimentaire etdigestion dans le rumen. Ann Zootech 21, 191-205

Rofller RE, Niedermeier RP, Baumgardt BR (1967) Eval-uation of alfalfa-brome forage stored as wilted silage,low-moisture silage, and hay. J Dairy Sei 50, 1805-1813

Forage type and milk composition

Rowland SJ (1938) The determination of the nitrogendistribution in milk. J Dairy Res 9, 42-46

Rulquin H, Pisulewski PM, Vérité R, Guinard J (1993)Milk production and composition as a function ofpostruminal lysine and methionine supply: a nutrl-ent-response approach. Livest Prod Sci37, 69-90

Saama PM, Mao IL, Holter JB (1993) Sources of varia-tion in partitioning of intake energy for lactating Hol-stein cows. J Dairy Sci 76, 1334-1341

SAS Institute (1987) SAS/STA T user's guide. SAS Insti-tute Inc, Cary, NC, USA

Sjaunja La (1989) Methods for calculation of the energycontent of milk. International commission for record-ing the productivity of milk animal (ICRPMA). Brus-sels, Belgium

Spërndly E (1989) Effect of diet on rnllk composition andyield of dairy cows with special emphasis on milkprotein content. Swed J Agric Res 19, 99-106

Sulton JO (1989) Altering milk composition by feeding.J Dairy Sci72, 2801-2814

Thomas C (1984a) Milk compositional quality and therole of forages. In: Milk compositional quality and itsimportance in future markets (ME Castle, RG Gunn,ed). Occ Publ Br Soc Anim Prod, No 9

Thomas PC (1984b) Feeding and milk production. In:Milk compositional quality and its importance in future

521

markets (ME Castle, RG Gunn, ed) Occ Publ Br SocAnim Prod, No 9

Thompson DJ, Beever DE, Haines MJ, Cammell SB,Evans RT, Dhanoa MS, Austin AR (1985) Yield andcomposition of milk from Frlesian cows grazing eitherperennial ryegrass or white clover in early lactation.J Dairy Res 52,17-31

Verdier l, Coulon JB, Pradel P, Berdagué JL (1995)Effect of forage type and cow breed on the charac-teristics of matured Saint-Nectaire cheeses. Lait 6,523-533

Vérité R, Journet M (1971) Utilisation comparée de l'ensi-lage de maïs et de l'ensilage d'herbe pour la pro-duction laitière. Ann Zootech 20, 153-167

Vérité R, Journet M (1975) Alimentation des vacheslaitières avec de l'ensilage de maïs: influence de ianature de l'ensilage, de la suralimentation énergé-tique et de la nature de la complémentation azotée.1. Production laitière. Ann Zootech 24, 95-107

Vérité R, Peyraud JL (1989) Protein: the POl system.ln: Ruminant nutrition. Recommended allowancesand feed tables (R Jarrige, 00) INRA and John LibbeyEurotext, London, Paris, 33-47

Vermorel M (1989) Energy: the feed unit systems. In:Ruminant nutrition. Recommended allowances andfeed tables (R Jarrige, ed) INRA and John LibbeyEurotext, London, Paris, 23-30