Effect of mastitis and related-germ on milk yield and composition during naturally-occurring udder infections in dairy cows

Transcription

1 Aim. Res. 51 (2002) INRA, EDP Scieces, 2002 DOI: /aimres: Origial article Effect of mastitis ad related-germ o milk yield ad compositio durig aturally-occurrig udder ifectios i dairy cows Jea-Baptiste COULON a *, Patrick GASQUI b, Jacques BARNOUIN b, Alai OLLIER a, Philippe PRADEL c, Domiique POMIÈS a a Uité de Recherches sur les Herbivores, INRA, Theix, Sait-Geès-Champaelle, Frace b Uité d Épidémiologie Aimale, INRA, Theix, Sait-Geès-Champaelle, Frace c Domaie de la Borie, INRA, Marceat, Frace (Received February 2002; accepted 27 September 2002) Abstract The effect of mastitis ad related-germs o milk chemical compositio (protei ad lactose cotets) ad milk somatic cell cout (SCC) was ivestigated i 501 milk quarter samples durig two cosecutive years i cows from three experimetal herds. Each ifected quarter was matched by a healthy oe i the same udder, as a cotrol. Milk protei ad mieral assays were performed i a subsample of 128 milks. Staphylococci were the most frequetly isolated germs (Staphylococcus aureus: 27%, coagulase-egative Staphylococci: 26%, Streptococci: 21%). Major milk pathoges (Staphylococcus aureus, Streptococcus uberis or Escherichia coli) associated with cliical sigs of mastitis were accompaied by higher SCC (+1.6 log ml 1, P < 0.01), lower lactose cocetratio ( 7.6 g kg 1, P < 0.01), higher protei cocetratio (+3.3 g kg 1, P < 0.01) ad higher soluble protei cocetratios (IgG ad BSA), hece a sharp decrease i the casei/protei ratio ( 10 percetage poits, P < 0.01). Chages were more marked whe Escherichia coli was preset. Coryebacterium bovis did ot alter milk chemical compositio whereas coagulase-egative Staphylococci slightly reduced lactose cocetratio ( 1.8 g kg 1 ) ad icreased SCC (+0.37 log ml 1 ). Calcium ad phosphorus milk cotets were hardly modified by the presece of microorgaisms. The decrease i milk yield durig cliical mastitis varied from 1.6 kg d 1 i the presece of Staphylococcus aureus to 15 kg d 1 i the presece of Escherichia coli. mastitis / milk compositio / microorgaism Résumé Effet du type de mammite et du germe sur la productio et la compositio du lait lors d ifectios mammaires aturelles chez la vache laitière. L effet de la ature des germes pathogèes présets das le lait sur la compositio chimique (taux protéique, taux de lactose) et la umératio cellulaire du lait a été étudié à partir d u échatillo de 501 laits de quartiers prélevés au *Correspodece ad reprits jbc@clermot.ira.fr

2 384 J.-B. Coulo et al. cours de 2 aées cosécutives sur les vaches de 3 troupeaux expérimetaux. Pour chaque quartier ifecté, u quartier sai de la même mamelle a servi de témoi. Sur 128 de ces prélèvemets, des aalyses de la compositio miérale et protéique des laits ot été réalisées. Les germes les plus fréquemmet observés ot été les staphylocoques (27 % de Staphylococcus aureus et 26% de staphylocoques à coagulase égative) et les streptocoques (21 %). Lorsqu elle a été associée à des siges de mammites cliiques, la présece d u germe majeur (Staphylococcus aureus, Streptococcus uberis ou Escherichia coli) s est accompagée d ue augmetatio de la umératio cellulaire (+1,6 log ml 1, P < 0,01), d ue dimiutio de la teeur e lactose ( 7,6 g kg 1, P < 0,01), d ue augmetatiodutauxprotéique(+3,3g kg 1, P < 0,01) et du taux de protéies solubles (IgG et BSA), de sorte que le rapport caséies/protéies a fortemet dimiué ( 10 poits de pourcetage, P < 0,01). Ces modificatios ot été les plus importates e présece d Escherichia coli.la présece decoryebacterium bovis a pas modifié la compositio chimique du lait, alors que celle de staphylocoques à coagulase égative a légèremet mais sigificativemet etraîé ue dimiutio de la teeur e lactose ( 1,8 g kg 1 ) et ue augmetatio de la umératio cellulaire (+0,37 log ml 1 ). Les teeurs e calcium et e phosphore ot été que peu modifiées par la présece de germes pathogèes. La chute de productio au momet de la mammite cliique a varié de 1,6 kg j 1 e présece de Staphylococcus aureus à 15,5 kg j 1 e présece d Escherichia coli. mammites / compositio chimique du lait / germes 1. INTRODUCTION The ecoomic impact of udder ifectios ca be highly detrimetal [3, 6], sice losses resultig from the treatmet-related ba o milk commercialisatio ad the very costs of treatmet add to direct productio losses. Studies ivestigatig the impact of mastitis o milk productio have produced widely variable results accordig to cow physiological coditios, seasoal factors or cow milk potetial [11, 16]. Also, mastitis is accompaied by sigificat modificatios of milk chemical compositio [15, 23] with both a reduced sythesis ad altered cell permeability [1]. Such modificatios affect protei ad mieral fractios, i particular, carry major cosequeces for milk techological properties [23] ad appear liked with the mastitis germ. If a umber of studies ivestigatig possible liks betwee germ ad milk compositio have bee coducted i experimetally iduced mastitis [9, 20, 29, 32, 40], few have bee carried out o aturally-occurrig mastitis [21]. The study, as a part of a wider programme o mastitis risk of recurrece [8], was aimed at assessig the relatioship betwee a set of milk chemical parameters (icludig protei ad mieral cocetratios), the mastitis germ ivolved ad the type of mastitis (cliical or subcliical) ad was based o observatios ad measuremets performed o three experimetal dairy herds. Milk specimes were sampled from each quarter of cows with mastitis to assess the idividual aimal effect withi the statistical aalysis of the results. 2. MATERIALS AND METHODS This study was coducted i three experimetal INRA herds (Theix, Marceat, Orcival) made up of Holstei (75%), Motbéliarde (15%) ad Taretaise (10%) cows (3000 to 7500 kg per lactatio). Durig two cosecutive years (September 1997 March 1999), the cliical mastitis cows were systematically subjected to sterile milk quarter samplig just before mastitis

3 Mastitis ad milk compositio 385 treatmet ad, whe possible, 30 days later. Out of 412 available lactatios, 155 cliical mastitis were visually diagosed through the udder ad milk, accordig to Gasqui ad Baroui s ratig table [7]. Mastitis was treated with atibiotics admiistered topically to the udder or by systemic admiistratio i severely affected cows. At dryig off, the cows systematically received a udder atibiotherapy. No ati-mastitis vaccie was admiistered. Most calvigs took place betwee September ad Jauary (85%). I the witer, the cows were kept either i tethered stables (Theix ad Marceat) or i cubicles (Orcival). The diets were composed of forage represetig 70 80% of the total dry matter (grass silage ad/or hay i Marceat ad Orcival, grass silage, hay, ad maize silage i Theix) ad were supplemeted with cocetrates based o cereal ad soyameal. The cows were tured to the pasture betwee May 1st ad November 1st. Milkigs were performed o-site (Theix) or i milkig parlours (Orcival ad Marceat) Milk samplig ad aalysis Two types of milk specimes were sampled from each quarter, sterile or ot. Sterile specimes were sampled accordig to the method described by Mialot [19]. Stadard procedures accordig to the Iteratioal Dairy Federatio guidelies [25] were applied for milk bacteriological aalysis. Microorgaism idetificatio used API kits. Milk bacteriology was performed at the INRA Aimal Epidemiology Research Uit (Sait-Geès-Champaelle, Frace). No-sterile specimes, collected after the sterile oes, were divided ito two samples. A preservative (bioopol) was added to oe of those samples that was used to determie protei ad lactose cotets (ifra-red method, Combifoss 5400, Foss Electric, Demark) ad somatic cell cout (fluoro-opto-electroic method, Fossomatic, Foss Electric, Demark). These aalyses were performed at the Laboratoire Iterprofessioel Laitier (Sait-Geès-Champaelle, Frace). Aother sample, which was immediately froze, was used to determie casei [30], calcium (atom absorptio spectrometry, [13]), phosphorus (spectrometry, [12]), immuoglobuli G (IgG) ad serum albumi (BSA) [18]. These aalyses were performed at the Laboratoire Iterprofessioel Laitier (Sait-Geès-Champaelle, Frace) ad at the INRA Herbivore Research Uit (Sait- Geès-Champaelle, Frace) Data processig Statistical aalysis was performed o the samples ( = 154) for which a ifected quarter ad a cotrol quarter were available, so that the aimal effect could be take ito accout. I 105 pairs of healthy quarters from the same udder, we had previously checked that the quarter positio (forward or rear) had o ifluece o milk compositio, whether the other quarters were ifected or ot by a pathoge (Tab. I). We also verified that milk compositio of the healthy quarters was ot affected by the ifectio of other quarter(s) of the same udder, regardless of the existece or ot of cliical mastitis i that quarter (Tab. II). Somatic cell cout, however, was slightly, but sigificatly icreased i a healthy quarter whe oe or several other quarters of the same udder were ifected by a pathoge (Tab. II). Out of the 154 specimes, 81 matched cliical mastitis ad 73 were cotrols collected 30 days after cliical sig occurrece. Oly 501 of the 616 milk specimes from correspodig quarters were of ay use, the remaiig 115 beig either cotamiated by a mixture of germs ( = 77) or very ufrequet germs ( = 25) or samples that could ot be aalysed ( = 13). Protei ad lactose assays ad somatic cell cout were performed o those 501 milk specimes. For casei, mieral, IgG ad BSA

4 386 J.-B. Coulo et al. Table I. Compositio of o-ifected quarter milk accordig to the positio of the quarter (forward or rear). Lactatio stage (d) Protei (g kg 1 ) Lactose (g kg 1 ) Cells (log ml 1 ) Quarter RSD 1 Sigificace 2 Forward Rear Residual stadard deviatio. 2 : P > Table II. Milk compositio of o-ifected quarters accordig to cliical ad ifectious status of the other quarters of the same udder. Ifectious status RSD 3 Sigificace 4 Negative 1 Positive 2 Positive 2 Cliical mastitis o o yes Lactatio stage (d) Protei (g kg 1 ) Lactose (g kg 1 ) Cells (log ml 1 ) a b 1 The 3 other quarters were ot ifected. 2 At least oe of the 3 other quarters was ifected. 3 Residual stadard deviatio. 4 : P > 0.05, *: P < 0.05, : P < b evaluatio, 128 quarter samples were selected, so as to make up a sufficiet umber of specimes for each of the mai germs: Escherichia coli ad Streptococcus uberis i cliical mastitis, coagulase-egative staphylococcus ad Coryebacterium bovis whe there was o cliical sig, ad Staphylococcus aureus i the presece or absece of cliical sigs. The data were aalysed usig aalysis of variace [31]. The factors itroduced i the model were the germ, raked accordig to the presece or absece of cliical sigs, ad the cow. Sice our statistical model adjusted for the cow effect, it took ito accout lactatio umber, lactatio stage ad other idividual ad herd effects. To assess the effect of the germ o the milk yield, the cases that occurred after the milk yield peak ad i which a sigle germ was ivolved ad for which the daily milk yield was available, were selected ( = 36). 3. RESULTS 3.1. Milk bacteriology Out of the 501 milk specimes aalysed, 270 were free of ay germs. The bacteriological results of the other 231 specimes are displayed i Table III. Staphylococci were the most preset, either Staphylococcus aureus as major pathoges or coagulaseegative staphylococci for mior oes.

5 Mastitis ad milk compositio 387 Table III. Bacteriological results observed o the 501 quarter milk samples. Type of germ Cliical mastitis No Yes No germ Major pathoges. Staphylococcus aureus. Streptococcus uberis. Escherichia coli. Other streptococci. Other germs 1 Mior pathoges. Coagulase-egative staphylococci. Coryebacterium bovis Maily Coryebacterium spp Milk yield evolutio accordig to the type of germ Whichever the germ ivolved, milk yield decreased oly from the date whe cliical sigs of mastitis were observed. The decrease (differece betwee the mea milk yield of days 3 ad 4 before mastitis detectio ad the mea milk yield of days 0 ad 1 after mastitis detectio) varied greatly accordig to the type of germ (Fig. 1): with Escherichia coli, milk yield decreased by a average of 15.5 kg per day over two days (P < 0.01) ad ever resumed its baselie level, eve after three weeks. With Streptoccocus uberis ad Staphylococcus aureus the decrease was low (2.9 kg d 1 ad 1.6 kg d 1, respectively) ad ot sigificat. With the other major germs, the decrease was itermediate (5.4 kg d 1, P < 0.01) ad did ot persist beyod 3 to 4 days. The sharp differeces i baselie productio levels were due both to the time of occurrece ad to the differet milk yield potetials accordig to the type of the germ: i our sample, Escherichia coli ad Staphylococcus aureus mastitis occurred i early lactatio (7 out of 8 cases ad 7 out of 10 durig the first 100 days, respectively) but the former ivolved cows with much higher milk yield potetials tha the latter (31.3 ad 24.6 kg per day o average durig the milk, kg.d D-7 D-5 D-3 D-1 days D1 D3 D5 D7 Figure 1. Milk yied patter aroud the day of detectio of cliical mastitis accordig to the germ: Escherichia coli,=8; Streptococus uberis, =5; Staphylococcus aureus, = 10; other major germs, = 13. 5th week of lactatio, respectively). Mastitis ascribable to Streptoccocus uberis ad other major germs affected high producig cows (29.2 ad 29.9 kg per day o average durig the 5th week of lactatio, respectively) but they occurred at a much later stage of lactatio (7 out of 13 cases ad 1 out of 5 durig the first 100 days, respectively).

6 388 J.-B. Coulo et al Effect of the germ o milk chemical compositio (Tabs. IV ad V) I 91% of the cases, the presece of a mior pathoge was ot accompaied by ay cliical sig of mastitis. Cliical sigs accompaied the presece of a major pathoge i 38% of the cases, with a wide variability accordig to the germ (from 22% with Staphylococcus aureus to 89% with Escherichia coli). No germ was detected at all i three percet of the cliical cases. The presece of a mior germ was accompaied by a icrease i somatic cell cout (+0.29 log ml 1, P < 0.01), although it did ot chage milk chemical compositio (Tab. IV). However, whe cosiderig the two types of mior germs foud i this study separately (coagulase-egative staphylococci ad Coryebacterium bovis), it was oted that o milk characteristic was degraded i the presece of Coryebacterium bovis, whereas lactose cocetratio was lower ( 1.8 g kg 1, P < 0.05) ad somatic cell cout was higher (+0.37 log ml 1, P < 0.01) i the presece of coagulaseegative staphylococci tha i healthy quarters. The presece of a major germ was accompaied by very sigificat chages i milk compositio whe it was associated with cliical sigs of mastitis. I particular, milk proteis icreased by 3.3 g kg 1 (P < 0.01), exclusively because of soluble proteis (IgG, BSA), so that the casei/protei ratio decreased by almost 10 percetage poits i the presece of cliical sigs. Also, lactose decreased by 7.6 g kg 1 ad somatic cell cout icreased by 1.6 log ml 1. These modificatios differed accordig to the germ ivolved (Tab. V): they were ideed more marked with Escherichia coli, whose presece i milk was liked to a decrease i lactose cocetratio ad casei/protei ratio, which reached 13.6 g kg 1 ad 19 percetage poits, respectively. Coversely, modificatios were less sigificat with Streptococcus uberis, IgG ad BSA cocetratios ot beig sigificatly differet with this germ from those observed i Table IV. Effect of the bacteriological ad cliical status o milk compositio (mea lactatio stage: 1 d). Bacteriological status RSD 1 Sigificace 2 Negative Mior pathoge Major pathoge Cliical mastitis No No No Yes Protei (g kg 1 ) Lactose (g kg 1 ) Cells (log ml 1 ) a 47.5a 4.70a a 46.6b 4.99b a 46.1b 5.58c b 39.9c 6.39d Casei (g kg 1 ) Casei/proteis (%) IgG (g kg 1 ) BSA (g kg 1 ) Calcium (g kg 1 ) Phosphorus (g kg 1 ) a 0.60a 0.17a a a 0.56a 0.17a a a 0.73a 0.22a ab b 2.09b 1.29b b 1 Residual stadard deviatio. 2 : P > 0.05, *: P < 0.05, : P < IgG: Immuoglobulis G; BSA: Bovie serum albumi. O the same row, results with differet letters are sigificatly differet (P < 0.05) *

7 Mastitis ad milk compositio 389 Table V. Effect of the type of germ ad the cliical status o milk compositio (mea lactatio stage: 123 d). Germ 1 No germ Cbovis Scoag- Saureus Saureus Suberis Ecoli Cliical mastitis No Yes Protei (g kg 1 ) Lactose (g kg 1 ) Cells (log ml 1 ) a 47.7a 4.69a a 48.2a 4.79ad a 45.9b 5.06d a 45.6b 5.84b 34.5b 39.7c 6.58c 36.3c 42.0c 5.64b d 34.1d 7.09c RSD 2 Sigificace Casei (g kg 1 ) Casei/proteis (%) IgG (g kg 1 ) BSA (g kg 1 ) Calcium (g kg 1 ) Phosphorus (g kg 1 ) a 82.8a 0.60a 0.17a 1.27a 0.95a ab 82.7ab 0.44a 0.15a 1.29abc 1.03b ab 82.4ab 0.64a 0.19a 1.29ac 0.96ac a 80.0bd 0.73a 0.22a 1.27abc 0.93ac ac 70.8c 2.45b 1.11b 1.17b 0.90cd b 76.7d 1.17a 0.48a 1.37c 0.96abc c 64.7c 3.33b 3.20c 1.13b 0.80d Cbovis: Coryebacterium bovis, Scoag-: Coagulase-egative staphylococci, Saureus: Staphylococcus aureus, Suberis: Streptococcus uberis, Ecoli: Escherichia coli. 2 Residual stadard deviatio. 3 : P > 0.05, *: P < 0.05, : P < IgG: Immuoglobulis G; BSA: Bovie serum albumi. O the same row, results with differet letters are sigificatly differet (P < 0.05). * healthy quarters ad lactose decreasig oly by 5.7 g kg 1. With Staphylococcus aureus, itermediate modificatios were oted. Whe this germ was ot associated with cliical sigs, modificatios i milk chemical compositio were oticeable ( 2.1 g kg 1 lactose, 2 percetage poits i the casei/protei ratio, log ml 1 i somatic cell cout), but ot as marked as whe Staphylococcus aureus was associated to cliical mastitis. Calcium ad phosphorus cocetratios were egatively affected oly i the presece of major pathoges (Escherichia coli ad Staphylococcus aureus) associated with cliical sigs. I cotrast, milk phosphorus cotet was higher tha i the healthy cotrol i the presece of Coryebacterium bovis. Streptococcus uberis was accompaied by high calcium cotets. Cosiderig the etire set of poits available, the reductio of the casei/protei ratio betwee healthy ad affected quarters of the same udder was closely ad liearly liked to that of lactose (R 2 = 0.69, P < 0.001). That ratio ad milk lactose cocetratio decreased i a curviliear maer (R 2 = 0.39 ad 0.46, respectively) with the icrease i the somatic cell cout. 4. DISCUSSION This study was based o data from aturally occurrig mastitis i experimetal farms, which provided a large umber of specimes ad esured homogeeous assessmet of mastitis cliical sigs. Also, whe workig with milk quarter samples which had healthy cotrols i the same udder, it was possible to optimally take the idividual aimal effect ito accout i data aalysis (icludig the aimal s geetic, physiological ad utritioal characteristics),

8 390 J.-B. Coulo et al. with that effect weightig very sigificatly i the iterpretatio of the chages i milk chemical compositio. To achieve this, it was hypothesised that the chemical compositio of healthy quarters was ot affected by the presece of ifected quarters i the same udder. This hypothesis may evertheless be questioed: through the observatio of the evolutio of milk characteristics with time, trials o mastitis iduced by ijectig edotoxis have show that such a idepedece is ot systematic, i particular regardig lactose [20, 34] whose precursors ca be reduced as a cosequece of ifectio. This observatio was ot cofirmed i this study because comparig the healthy quarter of a totally healthy udder with those which preseted at least oe ifected quarter, with or without cliical sigs, elicited o sigificat differece i either lactose or protei cotets, ad oly a slight icrease i somatic cell cout was observed. The germs idetified i this study durig cliical mastitis were those commoly observed i Frace [4, 32]. The proportio of mastitis without ay germ was low ad comparable to those commoly observed [27]. They could be due to either a o-ifectious iflammatory pheomea (traumatic i particular) or a detectio problem liked to the fact that we collected a sigle milk sample at the begiig of milkig [27]. O the cotrary to what has bee suggested i large-scale studies [2, 21], the chages i milk productio uder the effect of mastitis strogly deped o the type of germ. I particular, Escherichia coli mastitis iduces the greatest ad most durable milk yield reductios, whe compared with other types of mastitis. The magitude of this respose suggests that at least i the short term, the productio of all quarters i the udder was reduced by a systemic effect of Escherichia coli ifectio [33]. With the other germs, the milk loss of oe quarter was more or less compesated by a icrease i productio of the other quarters [32]. Milk yield evolutio patters that we observed were cosistet with those previously described o a weekly observatio scale [16]. They validated the hypothesis of a marked effect of the type of germ o milk productio ad appeared to cofirm that the productio level ad lactatio stage are promotig factors for certai types of ifectio [5]. These results also showed that precise estimates of the milk losses iduced by cliical mastitis require that measuremets be made daily or weekly to take ito accout the sometimes very fast resumptio (4 days) of pre-ifectio productio levels. The presece of major pathoges i milk quarters iduced very sigificat modificatios of milk chemical compositio, greater tha those described by Miller et al. [21] i aturally occurrig mastitis, but equal to those observed by Auldist et al. [1] i the same coditios, ad by several authors i experimetally iduced mastitis [9, 29, 33, 34, 40]. These modificatios were greater with Escherichia coli, icludig a g kg 1 decrease i lactose ad a 7 g kg 1 protei icrease. This icrease was ot the result of a cocetratio effect o the proteis sythesised by the udder because casei was sharply reduced, but rather to a icrease i blood proteis. Such chages were maily liked to two pheomea [28]: reduced secretory activity of the mammary cells ad icreased permeability of the mammary epithelium. The sharp icreases i serum albumi ad IgG cocetratios are proof of that pheomeo, which promotes the passage of blood proteis i the milk ad lactose escape to the blood [33]. Also, cliical mastitis is classically accompaied by a icrease i milk proteolytic activity, which cotributes to reducig the casei cotet [1,].Lastly,isevereEscherichia coli mastitis, these modificatios may also be liked to a reduced iput of the blood precursors (glucose, amio acids) ecessary for itramammary sythesis, uder the effect of the cows reduced appetite [33, 38].

9 Mastitis ad milk compositio 391 Out of the three major germs studied, Streptococcus uberis was the least cosequetial regardig milk characteristics, although its immediate effects o milk productio of the four quarters was greater tha with Staphylococcus aureus. With Streptococcus uberis, the somatic cell cout was similar to that observed with Staphylococcus aureus i the absece of cliical sigs. Also, the BSA ad IgG cocetratios of the milks where Streptococcus uberis was preset were o differet from those of healthy quarters. The weak effect of Streptococcus uberis o the compositio of quarter milk is hard to explai. Nevertheless, the respose time to ifectio may vary accordig to the pathogeic variability of the strais ad their resistace to atibiotics, as well as to the aimals ow resistace. That variability was especially revealed with Streptococcus uberis [22, 26, 37]. The icrease i somatic cell cout ad the chage i milk chemical compositio i the presece of Staphylococcus aureus were much more marked whe that germ was associated with cliical sigs. This result reflects the otio of severity, as already evideced with other germs i experimetal mastitis [9]. But eve i the absece of cliical sigs, the effects of the presece of Staphylococcus aureus o milk compositio were sigificat, icludig i particular a ear-to-three-percetage-poit reductio of the casei/total protei ratio. This result was cosistet with Urech et al. s fidigs [36] o the effect of subcliical mastitis o milk protei compositio ad could be otherwise explaied by the chroicity ad relative icurability of most Staphylococcus aureus mastitis. The presece i milk of mior pathoges is ot frequetly associated with cliical sigs [39]. I this study, Coryebacterium bovis was ot accompaied by ay degradatio of milk productio or compositio, cosistetly with Le Va et al. s study o atural ifectios [17]. I cotrast, the lack of ifluece of Coryebacterium bovis o somatic cell cout is ot i accordace with the results from other studies ivolvig atural ifectios [10, 17, 24], which revealed a slight icrease i cell cout i the presece of Coryebacterium bovis. But i these studies, the presece of cliical sigs of mastitis was ot take ito cosideratio although it could have explaied the icrease i somatic cell cout. Coversely, eve i the absece of cliical sigs, the presece of coagulaseegative staphylococci i milk decreased lactose cocetratio ad icreased cell cout, o the cotrary to the results from Timms ad Schultz s study [35] where o effect o cell cout was foud. Isofar as these two parameters are the most sesitive to ifectio, these modificatios reflect a reactio, albeit slight, of the udder to the germ. I coclusio, this study showed that the modificatios of the milk chemical compositio are liked to the germ, but also that such chages are much more marked i the presece of cliical sigs tha i subcliical ifectios. Escherichia coli, which associates the strogest resposes, both productiowise ad compositiowise, appears to be the germ with the most deleterious ecoomic cosequeces o milk quality. ACKNOWLEDGMENTS We thak J.P. Chacorac for the realizatio of bacteriological aalysis, M. Barbet, E. Albaret, J.P. Pezat ad all the staff of the 3 experimetal farms for takig care of the cows ad the samplig ad measuremets, J.C. Boefoy for his participatio i chemical aalysis, D. Levieux for his helpful advice, ad the GALA associatio who supports this work. REFERENCES [1] Auldist M.J., Coats S., Rogers G.L., McDowell G.H., Chages i the compositio of milk from healthy ad mastitic dairy cows durig the lactatio cycle, Aust. J. Exp. Agric. 35 (1995)

10 392 J.-B. Coulo et al. [2] Bartlett P.C., Va Wilk J., Wilso D.J., Gree C.D., Miller G.Y., Majewski G.A., Heider L.E., Temporal patters of lost milk productio followig cliical mastitis i a large Michiga Holstei herd, J. Dairy Sci. 74 (1991) [3] Beck H.S., Wise W.S., Dodd F.H., Cost beefit aalysis of bovie mastitis i UK, J. Dairy Res. 59 (1992) [4] Berthelot X., Fabre J.M., Houffschmitt P., Lebreux B., Morva H., Estimatio de la fréquece des germes resposables de mammites chez la vache laitière e Frace, Rec. Rech. Rumiats 4 (1997) 283. [5] Chassage M., Baroui J., Chacorac J.P., Biological predictors for early cliical mastitis i Holstei cows uder field coditios i Frace, Prev. Vet. Med. 35 (1998) [6] Fouricho C., Évaluatio de l impact zootechique et écoomique des troubles de saté e élevage bovi laitier, Thèse, EA Rees, Frace, 2001, 252 p. [7] Gasqui P., Baroui J., Mise au poit et évaluatio e élevages expérimetaux d u système d eregistremet des mammites cliiques chez la vache laitière, Rec. Rech. Rumiats 5 (1998) 346. [8] Gasqui P., Pos O., Coulo J.B., A idividual modellig tool for costructive cliical mastitis durig the same lactatio i the dairy cows: a method based o a survival model, Vet. Res. 31 (2000) [9] Guidry A.J., Ost M., Mather H., Shailie W.E., Weilad B.T., Sequetial respose of milk leukocytes, albumi, immuoglobulis, moovalet ios, citrate, ad lactose i cows ifusios of Escherichia coli edotoxi ito the mammary glad, Am. J. Vet. Res. 44 (1984) [10] Hokae-Buzalski T., Griffi T.K., Dodd F.H., Observatios o Coryebacterium bovis ifectio of the bovie mammary glad. I. Natural ifectio, J. Dairy Res. 51 (1984) [11] Hortet P., Seegers H., Loss i milk yield ad related compositio chages resultig from cliical mastitis i dairy cows, Prev. Vet. Med. 37 (1988) [12] IDF, Determiatio of total phosphorus cotet: spectrometric method (mik), stadard 42B, It. Dairy Fed., Brussels, Belgium, [13] IOS, Iteratioal orgaizatio for stadardizatio, dried milk, Determiatio of sodium ad potassium cotets, Flame emissio spectrophotometric method, ISO/DIS 8070, [] Kaartie L., Veijalaie K., Kuosa P.L., Pyörälä S., Sadholm M., Edotoxi-iduced mastitis ihibitio of casei sythesis ad activatio of the caseiolytic system, J. Vet. Med. 35 (1988) [15] Kitche B.J., Review of the progress of dairy sciece: bovie mastitis: milk compositioal chages ad diagostic tests, J. Dairy Res. 48 (1981) [16] Lescourret F., Coulo J.B., Modelig the impact of mastitis o milk productio by dairy cows, J. Dairy Sci. 77 (1994) [17] Le Va P.L., Eberhardt R.J., Kesler E.M., Effects of atural itramammary Coryebacterium bovis o milk yield ad secretios, J. Dairy Sci. 68 (1985) [18] Levieux D., Dosage des IgG du lait de vache par immuodiffusio radiale semi-automatisée, pour la détectio du colostrum, les laits de mammites ou de fi de gestatio, I Mise au poit, Lait 71 (1991) [19] Mialot J.P., Techique de prélèvemet de lait pour exame bactériologique, Rec. Méd. Vét. 159 (1983) [20] Michelutti I., Variatios des cellules somatiques et de la compositio biochimique du lait e relatio avec l iflammatio expérimetale de la glade mammaire, Thèse, INPL-EAIA, Frace, 1998, 150 p. [21] Miller R.H., Emauelso U., Brolud L., Persso E., Fuke H., Philipsso J., Relatioships of curret bacteriological status of the mammary glad to daily milk yield ad compositio, Acta Agric. Scad. 34 (1984) [22] Mile M.H., Barett D.C., Fitzpatrick J.L., Biggs A.M., Study of bacteria causig cliical mastitis ad pilot ivestigatio of the respose to treatmet of cases caused by Streptococcus uberis, Cattle Practice 1 (2001) [23] Muro G.L., Grieve P.A., Kitche B.J., Effects of mastitis o milk yield, milk compositio, processig properties ad yield ad quality of milk products, Aust. J. Dairy Techol. 39 (1984) [24] Ngatia T.A., Jese N.E., Berg B.B., Chages i the bovie udder quarters aturally ifected by Coryebacterium bovis, Brit. Vet. J. 177 (1991) [25] NMC, Laboratory ad Field Hadbook o Mastitis, Natioal Mastitis Coucil Ic., Arligto, VA 2201, USA, 1987, 34 p. [26] Puektes P., Masell P.D., Dyso R.S., Dick N.D., Browig G.F., Molecular epidemiology of Streptococcus uberis isolates from dairy cows with mastitis, J. Cli. Microbiol. 39 (2001) [27] Poutrel B., Gééralités sur les mammites de la vache laitière. Processus ifectieux, épidémiologie, diagostic, méthodes de cotrôle, Rec. Méd. Vét. 161 (1985) [28] Poutrel B., L amélioratio de la qualité du lait par la lutte cotre les mammites bovies, Med. Nut. 22 (1986) [29] Raiard P., Experimetal mastitis with Escherichia coli: kietics of bacteriostatic ad bactericidal activities, A. Rech. Vet. (1983) 1 11.

11 Mastitis ad milk compositio 393 [30] Rowlads S.J., The determiatio of itroge distributio i milk, J. Dairy Res. 9 (1938) [31] SAS Istitute Ic., SAS User s guide: Statistics, SAS Istitute Ic., Cary, NC, USA, [32] Seegers H., Meard J.L., Fouricho C., Mammites e élevage bovi laitier : importace actuelle, épidémiologie et plas de prévetio, Rec. Rech. Rumiats 4 (1997) [33] Shuster D.E., Harmo R.J., Jackso J.A., Hemke R.W., Edotoxi mastitis i cows milked four times daily, J. Dairy Sci. 74 (1991) [34] Shuster D.E., Harmo R.J., Jackso J.A., Hemke R.W., Suppressio of milk productio durig edotoxi-iduced mastitis, J. Dairy Sci. 74 (1991) [35] Timms L.L., Schultz L.H., Dyamics ad sigificace of coagulase-egative staphylococcal itramammary ifectios, J. Dairy Sci. 70 (1987) [36] Urech E., Puha Z., Schällibaum M., Chages i milk protei fractio as affected by subcliical mastitis, J. Dairy Sci. 82 (1999) [37] Vaarst M., Eevoldse C., Patters of cliical mastitis maifestatios i Daish orgaic dairy herds, J. Dairy Sci. 64 (1997) [38] Vadeputte-Va Messom G., Burveich C., Roets E., Massart-Leë A.M., Heyema R., Kremer W.D.J., Brad A., Classificatio of ewly calved cows ito moderate ad severe respoders to experimetally iduced Escherichia coli mastitis, J. Dairy Res. 60 (1993) [39] Waage S., Mork T., Roros A., Aaslad D., Hushamar A., Odegaard S.A., Bacteria associated with cliical mastitis i dairy heifers, J. Dairy Sci. 82 (1999) [40] Wheelock J.V., Rook J.A.F., Neave F.K., Dodd F.H., The effect of bacterial ifectios of the udder o the yield ad compositio of cow s milk, J. Dairy Res. 33 (1966) To access this joural olie: