Estrus in Dairy Cattle - Signs and Detection
Gordon King, Animal Science, University of Guelph

Infertility occurs in all herds but reproductive problems may become more obvious with artificial insemination since mating related events are recorded and summarized. Animal attendants often blame breeding failures on reproductive abnormalities such as acyclicity, silent estrus, follicular defects or insemination deficiencies. However, the inability to detect estrus so that females can be mated at the opportune time is, although often overlooked, the most frequent cause whenever artificial insemination is the principal mating procedure. Those people responsible for executing insemination procedures need a comprehensive understanding of reproductive function and sexual behavior to appreciate how important their estrous detection activities are in contributing toward the success of the entire program.

Estrous Cycles and Associated Sexual Behavior in Farm Mammals.

By the time of birth, female domesticated mammals possess a full complement of oocytes (eggs) contained with primordial or primary follicles located at the surface of their ovaries. No externally detectable evidence of ovarian activity appears through the initial growing period until the animal reaches sufficient physical size to undergo successful conception and gestation. At this time, synchronization of the appropriate gonadotropin controlling hormones from the anterior pituitary gland, plus the local regulatory mechanisms within the ovary, allow complete follicular maturation and ovulation. Thereafter, successive small groups of selected follicles commence final growth and development toward maturity so the associated estrous cycles commence. Female domesticated animals are anything but promiscuous, displaying only isolated and usually short periods of sexual activity (estrus or heat) coinciding with final follicular maturation. Estrogen, secreted by specialized endocrine cells located in tissues forming the walls of growing follicles, stimulates both the physical and behavioral signs of estrus. Thus, under natural conditions, females only become sexually motivated and receptive near the time when ovulation is imminent and conception is possible. Healthy, nonpregnant dairy cows and sexually mature heifers are polyestrus, showing regular periods of estrus throughout the year, unless pregnancy or pathology suppresses follicular development and ovulation, thereby suspending the cycle. Estrogen injections can produce estrous-like signs without follicular maturation and ovulation in both intact cycling, pregnant and even ovariectomized females. Unfortunately, veterinarians or herdspersons sometimes overlook this side effect accompanying the therapeutic use of estrogen.

Estrous Cycle

The bovine estrous cycle usually covers 17 to 23 days so normal, nonpregnant cows should show signs of estrous behavior (heat) every three weeks.

The most infallible sign of estrus is a female's willingness to stand immobilized and allow copulation to take place (a heterosexual interaction) or, in the case of cattle and occasionally in other species, to accept mounting by another female (a homosexual interaction). Many older references claim that the bovine estrous period lasts for 18 hours or more but somewhat more recent evidence indicates that it is considerably shorter for confined dairy cows (Hurnik, King & Robertson, 1975, App. Anim. Ethol. 2:55). The occurrence of short estrous periods in high producing dairy cows has been confirmed by a number of subsequent research investigations conducted in both temperate and tropical regions. Although considerable information on the relatively short duration of estrus now exists in scientific journals, it has not yet been widely incorporated into textbooks or extension publications.

A number of physical and behavioral signs may indicate that a cow or heifer is in or near estrus. Some of the things to watch for are:

Sexual Behavior

Associated Behavior

Other (Physical) Signs

The best indicator of estrus is when any cow or heifer repeatedly stands and accepts mounting by one of her herdmates. Unfortunately, they do not do this on demand. Those responsible for estrous detection must watch for any of the above indicators and, combining what they see with their own previous knowledge and experience, make decisions on whether to inseminate or not.

What Do We Know About Getting Cows Mated at the Correct Time?

Whenever cows and bulls graze together on pasture or range, they interact freely so that females in estrus rarely experience difficulty in locating sexual partners and mating near the time of ovulation. Confinement housing, however, places restrictions on normal reproductive behavior and patterns since animals cannot mix continuously with members of the opposite sex. Under these conditions the animal attendants must determine if and when service is given. Selection of the appropriate time for mating would not seem to present a problem since textbook descriptions of estrous behavior in cattle usually cite the classical references of Hammond (1927) or Trimberger (1948) who each give an average duration of 18 hours for temperate breeds.

In contrast to these older reports, a number of more recent investigations indicate that the period of sexual receptivity and related behavior demonstrated by confined Bos taurus dairy cows is considerably shorter than 18 hours (Table 1). Thus, a substantial number of females may pass through a follicular phase and ovulate without being detected, so matings are not performed at all (Williamson et al., 1972; King et al. 1976).

Table 1. Duration of Standing Estrus (hours) Demonstrated by Bos taurus Dairy Cows Recorded From Continuous or Frequent Observations and Including Substantial Numbers of Cycles.

Location Breed  Mean ± SE Remarks  Reference
USA Various 
11.9 ± 6.1
Winter & spring  Hall et al, 1959 
Costa Rica Jersey 
10.5 ± 3.2
AI & natural mating De Alba et al, 1961 
Canada Holstein 
7.5 ± 2.4 
10.1 ± 2.4
One in estrus 
Three in estrus 
Hurnik et al, 1975
Israel Fresian 
9.2 ± 0.4 
10.6 ±; 0.4
Folman et al, 1979
England Fresian 
10.2 ± 1.2
Heifers Esselmont et al, 1980 
Thailand Holstein 
5.6 ± 1.1 
5.1 ± 1.1
Cooler season 
Hotter season 
Pranee et al, 1996

Ovarian activity with follicular development, maturation and ovulation occurs within a few weeks after calving in adequately fed, healthy cows even under total confinement conditions. Unfortunately, the period when estrous signs appear is short and many mounting-mounted interactions occur at night when no one is present. Provided that ovarian activity commences early in the postpartum period, failure to detect the first or second ovulation doesn't depress performance since these should occur earlier than the dairyman normally wish to rebreed. However, the third ovulation generally occurs around 60 days and cows should be serviced by that time. In one study, conducted at the University of Guelph with ovarian activity monitored by sequential measurement of plasma progesterone concentrations and sexual behavior recorded with tme-lapse videotaping, animal attendants were detecting only 64% of the normal cycling cows at the third ovulation. Thus, one-third of the cows with normal estrous periods and ovulations were missed. This resulted in delayed rebreeding for a substantial number of animals and, even if conception occurred at the first insemination, prolonged intercalving intervals. Although the barn staff was certain that the cows were not functioning properly, the results indicated that failure to rebreed at the desired time usually resulted from inability of the herdsperson to detect estrus rather than because cows failed to cycle or had silent heats (King et al. 1976).

As illustrated in Table 2, however, cows were in estrus for short periods and had few mounting-mounted interactions unless other females were in or very near estrus at the same time

Table 2. Estrus Behavior in Totally Confined Holstein Dairy Cows Under Continuous Observation in a Temperate Environment.

One cow in or near estrus
Two cows in or near estrus
Three cows in or near estrus
Total mounting-mounted interactions 
Coming in (mounting), h 
2.5 ± 1.1
5.5 ± 1.4
9.8 ± 2.6
Standing when mounted, h 
7.5 ± 2.4
7.8 ± 1.4
10.1 ± 2.4
Going off estrus (mounting), h 
0.2 ± 0.1
4.5 ± 1.6
3.9 ± 1.8
Total mounting-mounted, h 

From Hurnik, King & Robertson, 1975

Most of the 72 cows observed in the University of Guelph dairy herd for collection of data presented in the above table had normal ovarian cycles established soon after parturition and ovulated regularly throughout the year. Acyclicity or ovulation without any detectable estrous signs were rare in this situation.  The postpartum ovarian activity and results should be similar in healthy, adequately fed dairy cows under almost all conditions.

Some years later a similar study was conducted to determine seasonal effects on postpartum ovarian activity, duration and intensity of sexual behavior for Holstein dairy cattle imported from Canada (a temperate climate) into northern Thailand (a tropical region). Animals, observed continuously during the cooler and hotter seasons for each of two years, were restricted to cement footing in an open-sided, pole-barn during the hotter season observation period in year one. The cows had access to both concrete and dirt footing during all other sessions in which they were under continuous observation. Sequential milk progesterone profiled provided an indication of when follicular phases occurred and recorded sexual behavior was compared with these to determine if estrous signs accompanied ovulations.

A number of females showed sexually associated behavior patterns such as sniffing and licking others during the hotter season observation periods. This pattern suggested that they were coming into estrus but the suspect cows never demonstrated any mounting-mounted behavior (Table 3). Although not conclusively proven, it appeared that herdmates who were not in or near estrus were sometimes reluctant to participate in homosexual interactions.

Table 3. Number of Ovulations and Detected Estrus Periods for Holstein Cows Under Continuous Observation in a Tropical Environment.

Cooler Season 
Hotter Season
Total cows 
Cycling cows
Total ovulations in observation period 
Ovulations with detectable estrous signs 
Quiet ovulations, % 

From Pranee, King, Subrod & Pongpiachan, 1996.

Constant observation of totally confined Holstein cows in a temperate climate showed that once cows established regular ovarian cycles after parturition, almost all ovulations were accompanied by some signs of estrus behavior (King et al. 1976). The general results obtained in Thailand during the cooler months were similar to findings in the Canadian study, but more ovulations without detectable estrous signs were recorded in the hotter months (Pranee et al, 1996).  As stated above, one subjective impression formed during the Thai study was that females in estrus may have been willing to engage in homosexual interactions during the warmer season but their nonestrus herdmates were less inclined.  This reluctance to participate in mounting activity would be even greater whenever animals were confined on slats, concrete or any other slippery surface.

Thatcher and Collier (1982) found an even more disturbing pattern occurring in Florida where half to two-thirds of potential estrous periods were missed in cooler months but this increased to over 75% in late spring and summer. Similarly, progesterone profiles compiled from Jersey and Holstein cows during the hotter months in Pakistan showed that all cows were cycling regularly (Imtiaz Hussain et al. 1992). However, with checking for estrus at 8 hours intervals, behavioral signs were observed in conjunction with only 36% of the ovulations. Although continuous observation can detect estrous signs for most ovulating cows under research station conditions, this surveillance intensity cannot be provided on commercial farms so the field situation is much closer to that described by Thatcher and Collier (1982) or Imtiaz Hussain et al. (1992). Farm operators must do whatever is practical to achieve reasonable detection efficiency, recognizing that some animals will always be missed (King, 1993). Whenever cows experience prolonged distress from any cause or animal attendants are distracted by other seasonal duties, the incidence of missed estrous periods will increase.

For those animals in Thailand that did engage in mounting-mounted interactions, the actual demonstration of sexual behavior was affected more by footing than by season. Cows that could move freely between the cement floored shed and dirt exercise yard showed mounting-mounted interactions for 11.5 ± 1.4 h in the cooler and 10.6 ± 1 h in the hotter seasons (Table 4). In contrast, animals confined to the shed only during the one hotter season showed sexual behavior for only 7.1 ± 1.4 h. Similarly, the actual time when cows stood passively and allowed herdmates to mount and the mean number of mounting-mounted interactions was significantly reduced during the observation period when females were restricted to the cement flooring (Table 4).

Table 4. Intensity and Duration of Estrus Demonstrated by Holstein Cows in a Tropical Environment.

Footing Cement & dirt  Cement & dirt Cement only 
Season Cooler  Hotter Hotter 
Mounting-mounted interactions, No.  24.1 ± 4.2a 26.5 ± 2.6a  11.4 ± 3.8b
Total duration of mounting-mounted, h   11.5 ± 1.4a 10.6 ± 1.0a  7.1 ± 1.4b
Total standing estrus, h 5.6 ± 1.1a  5.1 ± 1.1a 1.3 ± 1.1b 

Row means with same superscript not significantly different (P < 0.05).

These results certainly confirm the observations of Britt and coworkers (1986) who found that for dairy animals in North Carolina, season of the year had little influence on the duration or intensity of estrus. In contrast, the surface on which cows were confined (cement vs. dirt) had a profound effect on sexual behavior. The absence of a seasonal effect in North Carolina and Thailand does not support the observations from Florida which suggest that thermal stress alone produces a substantial reduction in the duration of estrus (Gangwar et al., 1965; Thatcher and Collier, 1982).

Hall et al. (1959) reported the average duration of estrus was 11.9 ± 6.1 h with a range between 2 and 46 hours for dairy cows maintained primarily on pasture in Louisiana. Although this study, conducted during the winter and spring seasons in a subtropical region, is sometimes cited as an indication that heat stress reduces the duration and intensity of estrus, the authors actually state: "Thus, it appears that the explanation for shorter estrous periods in this region is a complicated one, involving the possible interaction of nutrition, parasites, climate and endocrine gland performances in the cow." This extensive fertility study included many cows observed at reasonably frequent intervals so the findings clearly indicate that estrus duration for cows on pasture can be substantially shorter than the time periods previously reported by Hammond (1927) and Trimberger (1948). Similarly, continuous observation of Holstein cows, housed under total confinement in a temperate climate, indicated that standing estrus lasted only 7.5 ± 2.4 h and the total duration of mounting-mounted interactions rarely exceeded twelve hours unless two or more animals were in or near estrus at the same time (Table 2; Hurnik et al., 1975). The times for duration of sexual behavior reported by Hall et al. (1959) for dairy cows on pasture in a subtropical climate and by Hurnik et al. (1975) for lactating cows under total confinement in a temperate climate (Table 2) are similar to those found in Thailand whenever cows had access to exercise yards (Pranee et al., 1996).

The research findings described above and summarized in Table 1 indicate that many dairy animals in tropical and temperate regions show the classical signs of estrous behavior for less than 12 hours throughout the year. Collectively these reports certainly emphasize that the duration of estrus in confined dairy cows is almost always considerably shorter than the commonly quoted 18 hour period. It is likely that behavioral changes such as the shorter and less intense periods of sexual behavior demonstrated by dairy cows under various forms of stress, including total confinement, high production or adverse conditions of any type, are adaptive mechanisms to intensive management. Lindsay (1996), in his discussion of environmental effects on reproductive behavior, pointed out that one of the greatest challenges facing modern animal husbandry is to understand how the animal's surroundings influence behavioral patterns and to match genotypes to conditions that enhance productivity. Until such detailed knowledge and adapted genotypes are available, dairy farmers must do the best they can with conventional resources.

Problem Solving

Reproductive problems occur in herds of all sizes but practical solutions may differ between larger and smaller units. A logical beginning for any program attempting to improve reproductive efficiency should be an holistic assessment of the animals' environment to determine the physical and biological factors that may be depressing or enhancing current performance. This should be combined with a comprehensive analysis of breeding records to summarize past competence. In addition, the necessary clinical examinations and laboratory tests must be performed. With this information available, informed decisions can then be made regarding the major defficiencies and the best options for improvement. In most instances modifications will be needed so existing labor, genetics, feeding, housing, health and breeding practices must be examined in detail to identify major constraints and to devise practical solutions. Informed management decisions can only be made after a thorough evaluation of the environment, assessment of the resources available, the current husbandry system, the quantity and competence of existing or obtainable labor, and a consideration of the economic consequences that may be associated with any contemplated changes. Once these are defined, it is possible to decide on what additional inputs might be required and how the environment should be improved to enhance outputs.

The failure to detect estrus and have cows remated during the first six monthe of lactation is a common problem in dairy herds. Detection is difficult since the expression of estrous behavior varies between cows and also between the first postpartum and all subsequent cycles within cows. Factors such as group size and composition, age, health status, housing and especially footing, influence the duration and intensity of estrus. Thus, any recommendations on when to breed can only be elucidated in general terms. There are no magic bullets. The currently available technologies to assist with estrous detection, cycle control, artificial insemination and embryo manipulations are all tools to enable good managers to become more effective. They are not panaceas that will convert poor mangers into better ones. However, close attention to the following points could improve reproductive efficiency in many herds.

What Can We Do About Getting Animals Mated at The Right Time?

If most animals are to be successfully mated in time to maintain a 12 to 13.5 month calving interval, efficient estrous detection must be part of the regular management routine. The amount of time attendants can devote to this is limited so observation periods must coincide with periods when sexual activity is highest. Three, 20 minute observation periods, occurring just before the morning milking, sometime during late morning to early afternoon, and again late in the evening, would be the minimum time required. Time should not be wasted checking for estrus at feeding or when any equipment is operating since cows will be distracted and may not interact sexually. Both free and tied cows become more excitable and restless as they pass through estrus. Any animals that show increased activity or a change in normal behavior should be checked carefully for clear mucous discharge and an enlarged, reddened vulva plus other definitive signs of estrus when first noted and again at subsequent observation periods. Ideally, only animals that stand firmly and accept mounting by herdmates should be inseminated. However, on slatted floors or whenever the footing is poor, cows not in or near estrus may be reluctant to mount others. Thus, any female that persistently attempts to mount should be considered in estrus. If this practice is followed, it may result in the occasional animal being inseminated when she is just coming into estrus so cows still showing strong signs on the day following mating should be re-inseminated.

Detecting estrus and proper timing of insemination in continuously tied cows is difficult. Placing animals who are between 50 and 100 days postpartum in a box stall or other loafing area between milkings provides opportunity to show sexual behavior. If these females are watched closely in the late evening, before the morning milking, and at some time in late morning or early afternoon, estrous detection efficiency and pregnancy rates should improve.

Estrus and ovulation are accompanied by numerous behavioral and physiological alterations. Some of these, like standing to be mounted, are either present or not present and can be detected readily through visual observation. Others, such as changing hormonal concentrations, increased mobility (restlessness), vaginal hyperemia or temperature fluctuations, require more sophisticated measurements. Advancement in electronic sensor technology, combined with increased use of computers on farms, creates greater opportunity for automated detection of estrus related phenomenon. Devices to record mounting activity, restlessness and milk temperature changes have already been marketed commercially. Unfortunately, these are not yet totally effective in detecting animals in estrus and produce a number of "false positive" responses. Further improvement can be expected as the sensor and telemetry technology improve and experience in design and operation increase. It should eventually be possible to provide practical electronic aids that, even if not 100% effective in predicting the optimum time for insemination, reliably focus attention on those animals who are in or near estrus. These are only aids, however, and cannot completely replace the eyes or the intelligence of a knowledgeable observer.

Ultimately, it may be possible to control precisely the time of ovulation so that acceptable pregnancy rates follow fixed-time insemination. Alternately, it might be feasible to control estrus and ovulation, leave the cow for six to eight days and then insert an IVF derived embryo. However, until reliable, fully automated detection arrives, or IVF transfers become practical on commercial farms, it is essential that routine observation for estrous behavior remains a part of the normal work schedule.  

Prerequisites for achieving satisfactory estrous detection in dairy herds:

GOALS: Establish a series of challenging but achievable goals for each of the important reproductive parameters such as heifer age at first mating, proportion of cows detected and mated by 75 or 80 d postpartum, pregnancy rate to first service, days open, services per pregnancy, adjusted calving-interval, and culling for infertility. The most common reason that dairy farmers fail to meet goals is because they do not have any.

IDENTIFICATION: Mark all heifers and cows by a method that allows immediate and positive recognition of each individual and her reproductive status. The use of different colored tapes placed around the base of the tail or on neck straps to indicate cows ready for mating or mated but not yet confirmed pregnant, work well in many herds

RECORDS: Maintain comprehensive records that include all estrous dates, mating dates and any gynecological observations. Various computer programs will summarize herd performance and highlight the current reproductive status for individual cows. These are only worthwhile if someone regularly enters current data and interprets the output. Remember, anything not measured accurately cannot be managed effectively.

RESPONSIBILITY: Assign all estrous detection duties and responsibilities to one experienced and motivated individual. This person must have ample time for observations, record keeping and all other related activities. Someone must be held accountable for performance and should receive a financial bonus for meeting or exceeding reproductive goals.

COMPETENCE: The owner-manager must insure that all individuals responsible for and involved with the detection of estrus have the necessary knowledge, training and experience to perform the task properly.

TIME: a) Devote a minimum of one hour per day entirely towards observing cows for estrous signs. Ideally, this should be divided into three, 20 minute periods occurring at times when the animals are not distracted by any other activities. More time will be required in larger herds or if animals are housed in several units.

b) An observation period late in the evening can often detect estrous females that did not show obvious signs earlier in the day or identify those demonstrating increased excitability associated with coming into estrus. To be successful, this late check must be done quietly without disturbing most of the herd. In tie stall barns look specifically for cows that are already standing or get up immediately on sensing the attendant's presence. This alone does not indicate estrus but serves to identify an animal that should be watched closely for the next 24 hours. When on pasture or in free stall housing, most cows will be lying and ruminating quietly during the evening. The observer should investigate closely any individuals that are up and clustered closely together since females in or near estrus form a sexually active group.

FACILITIES: a) Turn tied animals out at regular intervals for exercise into an area providing good footing and freedom from distractions. Provision of an exercise pen in which females at the appropriate stage for rebreeding can spend part or all of the daytime or nighttime hours is a useful addition in many tie-stall herds with estrous detection difficulties. Exercise periods provide opportunity for cows to display behavioral signs but this only contributes to detection efficiency if someone is there to see and record mounting activity.

b) Cows in free stall or loafing units should also be observed regularly and at times when they are not disturbed by other activities. In larger herds, grouping of cows according to lactation stage and management requirements is an excellent practice since this places all animals between two and five-month postpartum together. Thus, observations for estrous detection can be focused on those in the appropriate period when rebreeding should take place.

DETECTION AIDS: a) Use visual reminders that provide obvious indication which animals are due for postpartum checking, mating, pregnancy diagnosis, special treatment or drying off. Items such as computer programs, breeding wheels, 21 day calendars, neck or ankle straps and individual colored tapes indicating reproductive status are all valuable management aids when used properly, but serve no useful purpose unless kept current and consulted daily.

b) Other estrous detection aids such as electronic sensors, chin ball markers, tail paint, heat mount indicators, teaser bulls, and even estrous cycle control procedures are available to help identification of proper time for mating. When used properly, estrous detection aids are tools that can assist an already competent herdsperson to do a better job. They are not crutches that will compensate for poor husbandry.

EVALUATION: The entire reproductive management program and performance should be evaluated at frequent intervals to insure that management goals are being met or to initiate corrective procedures whenever these are indicated.

Britt, J.H., R.G. Scott, J.D. Armstrong and M.D. Whitacre. 1986. Determinants of estrous behavior in lactating cows. J. Dairy Sci. 69:2195-2202.

De Alba, J., E. Villa Corta and G. Ulloa. 1961. Influence of natural service on length of estrus in the cow. Anim. Prod. 3:32-330.

Esslemont, R.J., R.G. Glencross, M.J. Bryant and G.S. Pope. 1980. A quantitative study of pre-ovulatory behaviour in cattle (British Friesian heifers). Applied Anim. Ethol. 6:1-17.

Folman, Y., A. Berman, Z. Herz, M. Kiam, M. Rosenberg, M. Mamen and S. Gordin. 1979. Milk yield and fertility of high-yielding dairy cows in a sub-tropical climate during summer and winter. J. Dairy Res. 46:411-425.

Gangwar, P.D., C. Branton and D.L. Evans. 1965. Reproductive and physiological responses of Holstein heifers to controlled and natural climatic conditions. J. Dairy Sci. 48: 222-227.

Hall, J.G., C. Branton and E.J. Stone. 1959. Estrus, estrous cycles, ovulation time, time of service and fertility of dairy cows in Louisiana. J. Dairy Sci. 42:1086-1094.

Hammond, J. 1927. The physiology of reproduction in the cow. Cambridge U. Press. London.

Hurnik, J.F., G.J. King and H.A. Robertson. 1975. Estrus and related behaviour in postpartum Holstein cows. App. Anim. Ethol. 2:55­68.

Imtiaz Hussain, S.M., J.W. Fuquay and M. Younas. 1992. Estrous cyclicity in nonlactating and lactating Holsteins and Jerseys during a Pakastani summer. J Dairy Sci. 75:2968-2975.

King, G.J. 1993. Reproductive efficiency. pp 239-252. In: Animal Production in Canada. Martin, Hudson & Young, editors. U of Alberta.

King, G.J., J.F. Hurnik and H.A. Robertson. 1976. Ovarian activity and estrus in dairy cows during early lactation. J. Anim. Sci. 42:688-692.

Lindsay, D.R. 1996. Environment and reproductive behaviour. Anim. Reprod. Sci. 42:1-12.

Roditan, P., G.J. King, S. Subrod & P. Pongpiachan. 1996. Oestrus behaviour of Holstein cows during cooler and hotter tropical seasons. Anim. Reprod. Sci. 45:47-58.

Thatcher, W.W. and R.J. Collier. 1982. Effect of heat on animal productivity. In: CRC Handbook of Agricultural Productivity Vol. II. M. Rechcigi, editor. CRC Press. Boca Raton. Pp 77-105.

Trimberger, G.W. 1948. Breeding efficiency in dairy cattle from artificial insemination at various intervals before and after ovulation. Nebraska Res. Bull. 153:3-26.

Williamson, N.B., R.S. Morris, D.C. Blood, C.M. Cannon and P.J. Wright. 1972. A study of oestrous behaviour and oestrus detection methods in a large commercial dairy herd. Vet. Rec. 91:58-62.