The proportions of milk stored in the gland cistern (cisternal) and within secretory alveolar tissue (alveolar) were determined for a group of ten cows in mid lactation, yielding ∼ 20 kg milk/d. As a proportion of total milk, cisternal milk, measured at 8 h after milking, ranged from 0·09 to 0·53. Milking frequency was then reduced from twice daily to once daily on two occasions, first for 1 week and then, 2 weeks later, for 2 weeks. Milk yield was reduced during once daily milking in all cows, by an average of 22·8% during the first week (P < 0·001). The amount of decrease varied among cows in a way that was unrelated to pretreatment yield, but the immediate decrease was significantly and inversely correlated with cisternal milk proportion (r = 0·81, P < 0·01). When milking reverted to twice daily, yield recovered to a value not significantly different from pretreatment, and the degree of recovery was inversely related to cisternal milk proportion (r = 0·74, P < 0·01). With the exception of one cow, yield decreases correlated well between the two periods of once daily milking. When treatment weeks were disregarded, lactation persistency (the rate of decline in yield with time) was unaffected by once daily milking. The results show that the ability of individual cows to tolerate once daily milking is related to their cisternal storage characteristics; yield is reduced less in cows that store a greater proportion of their total milk production within the cistern. Once daily milking for up to 2 weeks had no lasting effect on milk yield or lactation persistency, suggesting that there had been no detrimental effect on udder development.

The effect of milking interval on secretion rates of milk and milk constituents was studied in four milking camels during late lactation. An interval of 4 h produced the highest milk secretion rate compared with 8, 12 and 16 h intervals, and rates tended to decline with increasing milking intervals. Secretion rates of organic (lactose, SNF, fat and protein) and inorganic milk constituents (sodium, potassium, calcium and magnesium) generally followed similar patterns of decreasing secretion with increasing milking intervals. Many of the values for different intervals were significantly different.

Forty-four goats were milked and milk flow recorded without or with 1 min manual prestimulation in early, mid and late lactation. Ultrasound measurements of cross sections of the whole mammary gland were performed in a water bath. In additional experiments with 15 goats, milk flow was recorded and frequent blood samples were taken for the determination of oxytocin and prolactin concentrations. Milk yield increased from the first to the third lactation and decreased markedly during the course of lactation. Average and peak milk flow rates were closely related to the actual milk yield. The ultrasound cisternal area was 27·4±1·5% of the entire udder half cross section. Milking characteristics were scarcely different without or with prestimulation, although oxytocin was released within 30 s after the start of prestimulation, whereas oxytocin concentrations without prestimulation increased only after the start of milking. Concentrations of prolactin were higher during July and August than in April, and increased similarly with or without prestimulation during milking. In contrast to dairy cows, prestimulation and an opportune release of oxytocin during milking does not significantly influence the course of milk flow in goats, and this is probably because large amounts of cisternal milk allow milk ejection to be induced only after the start of milking without causing bimodal or otherwise reduced milk flow.

An artificial teat containing a force transducer was constructed to study the forces exerted by the liner on the teat during milking. The objective was to develop a method of comparing the action of different liners under a variety of different milking conditions. The rigid structure of the artificial teat did not provide an accurate representation of a real teat but deformable models used in the past have not proved practical. Initial tests of the teat showed that it gave repeatable results during calibration. When placed in a liner, however, repeatable results could not be obtained if the teat–liner interface was disturbed. The forces that were measured under simulated milking conditions varied with liner wall position and with time. There was a peak and trough prior to a stable asymptotie value being attained and a noticeable time lag behind pulsation chamber pressure. This technique can be used to compare liners when a reliable method of locating the artificial teat within the liner has been developed.

The aim of this work was to test the diffusion-controlled hypothesis of milk coagulation kinetics by reducing the diffusion coefficient of casein micelles. This has been achieved by increasing the solvent viscosity of milk through sucrose addition. Milk was reconstituted from skim milk powder and sucrose added at 100–300 g kg–1. Hydrolysis and coagulation were followed by chromatographic determination of caseinomacropeptide content and optical, thermal and viscoelastic measurements. Soluble and ionic calcium were determined by atomic absorption spectrophotometry and ionometry and micelle size was measured by dynamic light scattering. Addition of sucrose resulted in a substantial retardation of both enzymic and aggregation steps, a re-equilibration of calcium because of water reduction, and a micelle size increase. The enzymic rate constant was inversely proportional to the viscosity, according to a diffusion-controlled model, and the lag or characteristic times for the aggregation reaction were inversely proportional to the viscosity. These results are consistent with the involvement of diffusion-controlled steps in the sequence of reactions.

Naturally occurring monomeric κ-casein and αs2-casein in bovine milk were purified by ion-exchange chromatography in order to localize potential intrachain disulphide bridges. Enzymic cleavage of the proteins followed by mass spectrometry and amino acid sequence analysis of cystine-containing peptides revealed the presence of an intrachain disulphide bond in both proteins.

Two dimensional NMR spectroscopy was used to study the structure of a peptide composed of the N-terminal 42 amino acid residues of β-casein. The peptide was obtained by enzymic cleavage using endoproteinase Asp-N. Complete sequence-specific 1H NMR assignment was performed for the peptide at three Ca2+ concentrations (0, 22 and 37 mM). The NMR results show that the peptide was highly flexible and adopted multiple conformations. No stable secondary structures were present; however, the peptide had some regions with non-random structure. The region between residues Leu16 and Asn27 adopted conformations with an increased contribution of α-helical structure, a so-called nascent helix. Two regions, Glu11-SerP15 and Lys29-Phe33 showed an increased population of conformations with extended structures. Addition of Ca2+ induced chemical shift changes for the backbone amide protons, especially around the phosphoserine region and around the suggested α-helical structure, indicating that the addition of Ca2+ stabilized the structure already present in the apo form of the peptide.

The role of interfacial protein in determining the heat stability of recombined milk was investigated by removing serum protein prior to homogenization and reincorporating it after homogenization. In addition, the surface protein composition of recombined fat globules was probed by analyses of protein load and by quantification of the individual surface protein components using FPLC. In the absence of serum protein, substantially more casein was bound to the fat surface during homogenization. Despite this, the detrimental effect of homogenization on heat stability did not occur when serum protein had been removed from the system. Reincorporation of serum protein after homogenization caused the heat coagulation time–pH profile to revert to a form very similar to that observed without removing serum protein from the system. Thus, adsorption of serum protein did not affect heat stability. It is more likely that heat-induced interactions of serum protein with surface-adsorbed casein promoted heat coagulation. Fat surface area rather than casein load affected these interfacial protein-protein interactions during heating.

Trans fatty acids (TFA) formed during biohydrogenation by ruminant animals were quantified in cows' milk fat by means of Fourier transform i.r. (FTIR) spectroscopy. When samples were analysed as the derived methyl esters, a spectral subtraction technique resulted in TFA values that were not biased by the unspecific absorption of intact triacylglycerols or influenced by the shape of the trans band. Austrian milk fat samples contained 20–50 g TFA/kg. Results obtained by analysis of TFA as triacylglycerols compared favourably with those obtained by using methyl esters. The findings of the FTIR method were further substantiated by applying argentation chromatography in combination with capillary gas chromatography to the same set of samples. Besides trans monoenoic acids, non-methylene interrupted cis, trans octadecadienoic acids contributed significantly to the total trans content. Two novel TFA, trans-11, cis-15 octadecadienoic and cis-9, trans-13 octadecadienoic acids, were identified in milk fat by means of their retention behaviour on argentation chromatography and gas chromatography as well as by an oxidative degradation procedure.

A total of 529 bacterial strains have been isolated from milk and cream sampled at different sites in a dairy production plant under conditions selective for aerobic sporeforming bacteria. Identification with classical methods based on morphological, physiological and biochemical criteria showed Bacillus licheniformis to be the most frequently occurring Bacillus sp. The investigation also revealed 62 unidentified strains. Classical identification methods were time consuming (3–7d), lacked specificity and —because of their dependence on phenotypic gene expression—sometimes produced ambiguous results. Consequently, a colony hybridization method developed for the identification of B. licheniformis strains and using non-radioactive labelled 23S rRNA targeted oligonucleotide probes was also used. Identification of B. licheniformis with borth classical and hybridization methods revealed diverging identification results for 70 strains.

A laboratory scale bioreactor was used for continuous acidification and inoculation of milk with a proteinase-negative, lactose-fermenting strain, Lactococcus lactis subsp. lactis C2S. Calcium alginate-entrapped cells were immobilized on a spiral stainless steel mesh incorporated into a column bioreactor and used to acidify and inoculate reconstituted skim milk. Characteristics of the immobilized cell bioreactor (ICB) were compared with those of a free cell bioreactor (FCB) during challenge with a virulent phage. Steady state biomass and lactate productivities were respectively 25-fold and 12-fold larger with the ICB than with the FCB. The ICB and the FCB were inoculated with the prolate phage c2 at multiplicities of infection of 0·25 and 0·02 respectively. Within 90 min of the infection, the FCB viable cell concentration dropped by five orders of magnitude and never recovered, while the plaque forming units/ml increased dramatically. In the ICB, released cells decreased immediately after infection, but subsequently increased, while the plaque forming units/ml steadily declined, indicating that phage were being washed out of the bioreactor. Productivity of FCB decreased to zero, whereas productivity of the ICB only decreased ∼ 60% and subsequently recovered to its initial steady state value.

The antimutagenic effects of uninoculated milk and milks cultured with Bifidobacterium or Lactobacillus strains towards the mutagenicity induced by two direct mutagens, 4-nitroquinoline N-oxide and 2-nitrofluorene, and three dietary indirect mutagens, aflatoxin B1, benzo(a)pyrene and quercetin, were investigated using the in vitro Salmonella typhimurium test. Each cultured milk sample and control milk had a significant antimutagenic effect, to an extent varying with the mutagen used. Uninoculated milk had a greater inhibitory effect than cultured milks towards dietary indirect mutagens.

The effect of giving yogurt supplements to Balb/c mice on the various gut-associated lymphoid cells was studied. Animals were fed for 2, 5, 7 and 10 consecutive days. The different lymphoid cell types were identified and counted by haematoxylin–eosin staining of histological slices. The numbers of cells secreting IgA, IgG and IgM and the numbers of T lymphocytes were determined by direct immunofluorescence. The degree of activation of the intestinal macrophages in the small intestine was assessed by measuring the β-glucuronidase (EC 3.2.1.31) released into the intestinal fluid, and also by a histochemical method. Throughout the feeding period, there were no histological alterations in the gut, but there was marked cell infiltration, mainly of plasma cells and lymphocytes. The number of macrophages on the small intestine increased significantly after feeding for 2 d, while the β-glucuronidase activity was only slightly higher that of the controls. After a 7 d feeding period, the number of IgA secreting cells increased, while the values for cells secreting IgM and IgG and for T lymphocytes remained similar to those of the controls. The effect of giving yogurt on lymphoid cells associated with the large intestine was mainly on the numbers of IgA secreting B cells and T lymphocytes, with a marked increase during the whole feeding period in the latter type of cell. Since giving yogurt mainly enhanced the IgA secreting B cells in both small and large intestines, this increase would strengthen the host's defence mechanisms in the intestinal mucosa. Although the number of macrophages was increased, there was no enhancement in their activity, which might have harmed the host by producing an inflammatory response.