Automated electronic systems for the detection of oestrus and timing of AI in cattle

Animal Reproduction Science 60 61 2000 713 723 www.elsevier.comrlocateranireprosci Automated electronic systems for the detection of oestrus and timing of AI in cattle R.L. Nebel ), M.G. Dransfield, S.M. Jobst, J.H. Bame Department of Dairy Science, Virginia Polytechnic Institute and State UniÕersity, Blacksburg, VA 24061-0315, USA Abstract For the majority of dairy herds where artificial insemination AI is practiced, the limiting factor toward obtaining efficient reproductive performance is the failure to detect oestrus in a timely and accurate manner. Periodic visual observation has been the dominant method used to identify cows in oestrus. New approaches are being developed to provide automated systems of detection of oestrus using electronic technology. The goal of an oestrus detection program should be to identify oestrus positively and accurately in all cycling animals and consequently to identify animals not cycling. The ultimate goal should be to predict the time of ovulation, thus allowing for insemination that will maximize the opportunity for conception. Unfortunately, most studies designed to evaluate the optimal time of AI generally contained two technical deficiencies: inadequate numbers of cows for valid statistical comparisons and inaccurate knowledge of the onset of oestrus because of low frequency of visual observations andror efficiency of methods used for the detection of oestrus. Studies using pedometry and a pressure sensing radiotelemetric system will be reviewed as each have independently obtained an optimal time of AI of 5 to 17 h after either the increase in locomotive activity or following the first standing event associated with the onset of oestrus. q 2000 Elsevier Science B.V. All rights reserved. Keywords: Oestrus; Artificial insemination; Pedometry; Radiotelemetry; Bovine 1. Introduction Artificial insemination AI was one of the first and most influential biotechnologies of animal agriculture. The benefits of AI have ranged from reducing disease transmis- ) Corresponding author. Tel.: q1-540-231-4432; fax: q1-540-231-5014. E-mail address: rnebel@vt.edu R.L. Nebel. 0378-4320r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S0378-4320 00 00090-7

714 ( ) R.L. Nebel et al.ranimal Reproduction Science 60 61 2000 713 723 sion, thus increasing health and longevity to increasing yield traits afforded by genetic selection. Oestrus detection is often cited as the most costly component of an AI program. An estimated annual loss of ) US$300 million to the USA dairy industry occurs because of the failure to detect oestrus or the misdiagnosis of oestrus ŽSenger, 1994.. The efficient and accurate detection of oestrus and the timing of resulting AI remain major challenges to improving reproductive and economic efficiencies of many dairy farms ŽFoote, 1974; Gwazdauskas et al., 1986; Heersche and Nebel, 1994; Senger, 1994.. Methods of oestrus detection used in investigating time of ovulation have varied. Some investigators have used frequent visual observation ŽDeSilva et al., 1981; Larsson, 1987., frequent exposure to teaser animals or other cows ŽBrewster and Cole, 1941; Nalbandov and Casida, 1942; Mattoni et al., 1988., or a combination of frequent visual observation with oestrus detection aids Ž Rajamahendran et al., 1989.. Physical verification of ovulation in most studies has been by palpation of the ovaries per rectum at frequent intervals ŽBrewster and Cole, 1941; Nalbandov and Casida, 1942; Trimberger, 1948; Mattoni et al., 1988. or by frequent ultrasonography ŽLarsson, 1987; Rajamahendran et al., 1989.. Ovulation has been timed from the cessation of oestrus ŽBrewster and Cole, 1941; Nalbandov and Casida, 1942; Trimberger, 1948. and from onset of oestrus Ž Larsson, 1987; Mattoni et al., 1988; Rajamahendran et al., 1989.. In general, these studies did not contain either an adequate number of observations or continuous observation for behavioral oestrus to accurately determine the relationship of behavioral oestrus and occurrence of ovulation to determine the optimal time of AI. Traditionally, AI has followed the a.m. p.m. guideline established in 1948, which recommended that cows observed in oestrus in the a.m. should be submitted for AI in the p.m., and cows observed in oestrus during the p.m. should be submitted for AI the following a.m. Ž Trimberger, 1948.. Although ovulation time was reported from the end of oestrus, it was noted that initiation of oestrus was important for AI. Timing of AI from the onset of oestrus is important and quite evident when reviewing results of recent research using electronic methods that determine the onset of standing oestrus ŽMaatje et al., 1997; Dransfield et al., 1998.. There are many physiological, behavioral, and hormonal changes that take place during the oestrous cycle of the heifer or cow Ž Allrich, 1994., but many of these changes are highly variable among animals and are impractical to measure routinely. This manuscript will emphasize several manifestations of oestrus, which can be measured and have been correlated with ovulation and succeeding conception. 2. Pedometry Almost 70 years elapsed between the first published recognition that female mammals display a predictable increase in physical activity when in oestrus and the first potentially useful field application of pedometry Ž Kiddy, 1977.. The first critical description of the relationship between physical activity and stage of the oestrous cycle utilized 13 Guernsey cows on pasture Ž Farris, 1954.. Cows equipped with mechanically activated pedometers were characterized as having a 218% higher physical activity

( ) R.L. Nebel et al.ranimal Reproduction Science 60 61 2000 713 723 715 during oestrus than during late diestrus and proestrus or during metestrus. More than 20 years elapsed after this report before research at USDA in Beltsville, MD revived interest in pedometry as a practical tool for oestrus detection of dairy cows ŽKiddy, 1977.. The average increase in activity at the time of oestrus was 393%, or approximately four times the activity of cows not in oestrus when housed in a free stall barn. When cows were housed in comfort stalls, cows in oestrus were about 2.76 times more active during oestrus, indicating that the type of housing influences the magnitude of change in physical activity. It was noted that the daily activity for each cow must be monitored and activity associated with oestrus compared to that obtained during the other stages of the oestrous cycle for pedometry to be most effective in identifying cows in oestrus. Individual cows differed significantly in the amount of activity expressed under the same conditions. Hurnik et al. Ž 1975. and Amyot and Hurnik Ž 1987. studied the activity of cows continuously monitored with time-lapsed video recording. Their investigations revealed that cows spend considerably more time walking when in oestrus and less time resting and eating than when they were not in oestrus. Continuous visual observation combined with pedometer measurements confirmed that continuous observation is more reliable than pedometry and periodic observations Ž Pennington et al., 1986.. A pedometer modification was made that internally compared the activity change during a specific time interval to the five previous a.m. or p.m. activities recorded. This modification was implemented to account for individual activity variation reported in previous studies. Sexual activity during periods of moderate weather was shifted markedly by management practices, but only slightly by diurnal tendencies. It was recommended that when establishing a schedule for periods of visual observation more concern should be given to disruptions of activities caused by management practices such as feeding and milking than the possible increase of the nocturnal occurrence of oestrus. Their recommendations for visual observations of oestrus focused on location with least crowding and best footing being primary to time of occurrence. Using a similar pedometry system simultaneously with time-lapsed video recording, Varner et al. Ž 1994. reported that walking activity increased in the 4 h prior to the onset of oestrus. Peak mounting activity occurred during times of low walking activity for the entire herd. This may have been influenced by the social dynamics of mounting behavior. Cows may be able to locate a sexually willing partner easier when there is less overall herd activity. Alternatively, the oestrual cow may be occupied with eating and milking activities, similar to her herdmates, which prevents them from engaging in mounting activities. Behavior trends of high mounting activity during times of low pedometer readings should be considered by system designers to increase the sensitivity of pedometry as a tool for the detection of oestrus. A commercial pedometry system designed for dairy cows where activity is recorded as steps per hour Ž Afimilk Pedometer, Afikim, Israel. was used to determine the effects of cow age, parity, milk yield, and days in milk on activity identified at oestrus ŽArney et al., 1994.. Mean dioestrous activity was not correlated with any other animal characteristics. Activity measurements were significantly greater in the p.m., 160 steps y1 y1 y1 h, than the a.m., 145 steps h. Arney et al. 1994 determined that steps h from 72 to 16 h before oestrus gradually increased linearly. From 16 h to oestrus, steps h y1

716 ( ) R.L. Nebel et al.ranimal Reproduction Science 60 61 2000 713 723 increased more rapidly and linearly to peak oestrus, followed by an exponential decay with no refractory period post-oestrus. The optimal time of AI based on pedometer readings was predicted using mathematical models and rectal palpation of 171 cows Ž Maatje et al., 1997.. The pedometers Ž w Boumatic Heat-seeker-TX ; Dairy Equipment, Madison, WI. were set to produce an alarm signal Ž flashing light. when the mean activity of the last six 2-h periods was more than double the mean activity of the last six corresponding 2-h periods for the previous 2 days. Retrospectively, the time of the onset of oestrus and the number of hours from the beginning of increased activity to the time of AI was determined. The onset of oestrus was determined as the first 2-h period in which the mean pedometer reading for the current 12-h window was twice the running mean of a comparable period in the previous 2 days. Chance of pregnancy was highest between 6 and 17 h after increased pedometer activity, and the calculated optimum time of AI was 11.8"1.7 h. The aim of most published applications of pedometers until Maatje et al. Ž 1997. has been to improve rates of oestrus detection. A review by Lehrer et al. Ž 1992. stated that 70% to 80% of cows in oestrus are detected by pedometer measurements. Pedometry systems, which allow identification of the onset of oestrus in addition to efficient and accurate oestrus identification, will increase the usefulness of such technology in animal breeding. However, all current pedometry systems do not use real-time data transfer, thus requiring the activity information to be retrieved by an interrogation device. Therefore, system requirements dictate that retrieval of activity measurements can occur two or three times daily, usually at milking, reducing the effectiveness of determining the timing of insemination. 3. Pressure sensing radiotelemetric system Radio frequency data communications is the base technology employed by the w commercially available pressure sensing radiotelemetric HeatWatch system ŽDDx, Denver, CO.. A radiotelemetric device attached to each cow consists of a miniaturized radiowave transmitter, powered by a lithium 3-V battery and linked to a pressure sensor enclosed in a hard plastic case 5.3=8.1 cm and 1.8 cm in height. Each device is secured in a water-resistant pouch, attached to a 35 = 20 cm saddle-shaped nylon mesh patch that is glued with contact-type adhesive to the hair caudal to the sacral region. Activation of the pressure sensor by weight of a mounting herdmate for a minimum of 2 s produces a radiowave transmission Ž 0.4-km range.. Transmitted data consists of sensor identification, date Ž month, day, and year., time Ž h and min., and duration of sensor activation Ž. s. Transmitted signals are sent to a microcomputer via a fixed radio antenna. The remote signal receiver should be centrally located on each farm to maximize transmission area and situated to minimize transmission interference. Transmitted data from a remote receiver are chronologically stored in a buffer external to the microcomputer and transferred to a microcomputer at request of the software. The software generates both fixed management reports and individual cow files that can be viewed or printed.

( ) R.L. Nebel et al.ranimal Reproduction Science 60 61 2000 713 723 717 Duration of oestrus, determined by video recordings Ž Hurnik et al., 1975., varied with the number of dairy cows in oestrus simultaneously, increasing from 7.5 to 10.1 h, with one or three cows in oestrus, respectively. Duration of oestrus ranged from 2.6 to 26.2 h and averaged 14 h for 50 oestrus-synchronised beef heifers monitored with the pressure sensing radiotelemetric system Ž Stevenson et al., 1996.. Many heifers Ž 20.5%. had periods of oestrus activity - 10 h in duration. Monitoring two seasonal dairy herds on pasture with a similar radiotelemetric system, Xu et al. Ž 1998. reported the average duration of oestrus for 89 cows was 8.6 h. There were 11.2 recorded mounts during each oestrus for a total mounting duration of 29 s with an average duration of 2.5 s for each standing event. Using the radiotelemetric system to monitor mounting activity and ultrasonography to determine the time of ovulation, a significant, positive relationship between duration of oestrus and time of ovulation was reported Ž Walker et al., 1996.. A prolonged duration of mounting activity was associated with an extended interval from first mount to ovulation. However, this relationship existed over a relatively brief time interval Ž25 to 34 h.; therefore, differences in duration of oestrus would have limited importance in the timing of AI. The average oestrus duration was 9.6 h with a large standard deviation of 6.9 h. In a subsequent study involving 2661 inseminations, the duration of oestrus did not have a significant effect on conception rate Ž Dransfield et al., 1998.. Duration of oestrus, defined as the time interval from first to last standing event recorded by the radiotelemetric system, averaged 7.1" 5.4 h for 2055 oestrus periods. The duration of oestrus varies greatly not only among cows in the same herd but also among different studies. Differences in age, herd size, management conditions, frequency of observation, and definition of onset of oestrus may account for most of the variation in duration of oestrus among studies. Determination of the onset of oestrus is only possible with continuous monitoring for behavioral activity to accurately determine the first standing event when a herdmate is allowed to mount the animal in oestrus. Using time-lapsed video recording, Hurnik et al. Ž 1975. disclosed the highest frequency of onset of standing oestrus was observed during 1800 to 2400 h. After 12 years, when the study was repeated with more advanced video equipment, the hourly distribution of oestrus onset was tested against a normal distribution and no significant pattern of onset was revealed Ž Amyot and Hurnik, 1987.. Pooled 6-h intervals demonstrated the highest frequency of the onset of oestrus occurred between 1200 to 1800 and 0600 to 1200 h in the primiparous and multiparous groups, respectively. Conflicting data concerning diurnal and nocturnal oestrual activities are found in the literature. While the factors underlying these discrepancies are probably of a complex nature, a photoperiod effect directly or indirectly influencing oestrual expression may be a predominant one. In Amyot and Hurnik s Ž 1987. study, the sensitivity of the cameras permitted a nocturnal illumination of less than 3 lx, which is 90% less than the illumination used in the earlier study Ž Hurnik et al., 1975.. If oestrual activities are light-mediated, then suppression of oestrus expression should be expected during the nocturnal period. Using the radiotelemetric system to monitor mounting activity in pasture-fed cows, the onset of oestrus and distribution of total mounting activity occurred equally throughout the day when grouped into 6-h periods Ž Xu et al., 1998.. However,

718 ( ) R.L. Nebel et al.ranimal Reproduction Science 60 61 2000 713 723 individual hourly variation did occur with the greatest number of first mounts or oestrus onsets between 1200 and 1500 h and between 2100 and 2300 h. Total mounting activity did not parallel the hourly distribution of oestrus onsets and was more evenly distributed across the day despite a trend toward more mounting activity in the afternoon. Since 1994, the HeatWatch w system has been the exclusive method used for the identification of oestrus at the Virginia Tech University Dairy Center. The distribution in onset of oestrus is shown graphically for heifers Ž ns393. and cows Ž ns1075. ŽFig. 1 and Fig. 2, respectively.. There were no differences for onset of oestrus among hourly periods for either heifers or cows although it appeared that management practices influenced the onset of oestrus. Heifers were gathered between 0700 and 0800 h daily for concentrate feeding, which corresponded to the hour of peak initial activity Ž Fig. 1.. During the remainder of the day, heifers were on pasture with very limited human interaction and only a trend toward a diurnal pattern with more activity during the afternoon. A greater variation in the onset of oestrus was detected for cows. It appeared that the three periods of peak onset activity occurred when cows were either being moved in groups during milking Ž 2300 to 0000 and 1100 to 1300 h. or loafing on dirt lots during barn cleaning Ž 0800 to 0900 h.. When the precise onset of oestrus has been determined for a significant number of oestrus periods, all studies have revealed that initial standing behavior occurs evenly distributed throughout the day. The rise in estradiol-17b concentrations which occurs almost simultaneously with the onset of oestrus activity Ž Stevenson et al., 1998. is responsible for the initiation of behavioral oestrus. Estradiol-17b is also indirectly responsible for the release of LH by the modification in the amplitude and frequency of Fig. 1. Distribution in onset of oestrus at different hours of the day for 393 heifers monitored by the w radiotelemetric HeatWatch system DDx.

( ) R.L. Nebel et al.ranimal Reproduction Science 60 61 2000 713 723 719 Fig. 2. Distribution in onset of oestrus at different hours of the day for 1075 cows monitored by the w radiotelemetric HeatWatch system DDx. release of gonadotropin releasing hormone in the absence of progesterone, thus, estradiol-17b is ultimately responsible for ovulation. The rise in Estradiol-17b, which initiates this cascade of events, is probably independent of most environmental influences, thus the onset of oestrus is equally distributed during the day and should dictate, within management constraints, the timing of insemination. Biological events that affect the successful timing of AI are the length of the functional viable life of gametes Ž sperm and ova., transport time of viable sperm from the site of insemination to fertilization, and timing of ovulation in association with insemination as shown graphically in Fig. 3. The transport of viable spermatozoa to the oviducts requires a minimum of 6 h to obtain a population capable of fertilization, and Fig. 3. Average time relationships among reproductive events associated with fertilization in the bovine.

720 ( ) R.L. Nebel et al.ranimal Reproduction Science 60 61 2000 713 723 sperm numbers progressively increase over 8 to 18 h ŽThibault, 1973; Wilmut and Hunter, 1984; Hawk, 1987.. The functional viable life of bovine spermatozoa in the reproductive tract has been estimated at 24 to 30 h Ž Trimberger and Davis, 1943.. Ovulation following the onset of oestrus is approximately 27 h ŽChenault et al., 1975; Walker et al., 1996.. Although the maximum length of time the ovum may retain its capacity for fertilization is 20 to 24 h, the optimum period of retention of this capacity is remarkably transitory, estimated at 6 to 10 h Ž Brackett et al., 1980.. Thus, with the availability of a 24-h surveillance system to monitor behavioral events associated with oestrus, it seems appropriate to re-examine timing of AI in dairy cattle. Therefore, the primary objective of the study conducted by Dransfield et al. Ž 1998. was to evaluate timing of AI in dairy cows where the onset of oestrus was precisely determined using the radiotelemetric HeatWatch w system. Each farm selected a 3-h interval to inseminate cows identified in oestrus during the previous 24 h. Pregnancy status was determined by data for return to oestrus and palpation of the uterus 35 to 75 days following insemination. Logistic regression analysis for the probability of pregnancy was performed with a model including herd, interval from onset of oestrus to AI, standing events per oestrus, season, and days in milk at insemination. The time interval from onset of oestrus to insemination significantly influenced Ž P- 0.01. the percentage diagnosed pregnant 35 to 70 days following insemination. Odds of pregnancy resulting from AI increased approximately 34% for cows inseminated between 4 and 12 h after onset of oestrus when compared with a baseline interval of 0 to 4 h after onset. Intervals from onset of oestrus to AI)16 h were related negatively to the probability of conception. The bar graph shown in Fig. 4 represents the cows that were diagnosed pregnant relative to 4-h intervals from the first standing event to insemination. A curvilinear relationship between interval and pregnancy is unmistakable; conception rates were highest for cows inseminated from 5 to 16 h following the first standing event of oestrus. Inseminations performed between 4 and Fig. 4. Percentage pregnant by 4-h intervals relative to timing of AI from first standing event detected by the w radiotelemetric HeatWatch system Ž DDx. across 17 herds and 2661 inseminations. Number of inseminations for each 4-h interval is within parentheses.

( ) R.L. Nebel et al.ranimal Reproduction Science 60 61 2000 713 723 721 12 h following onset of oestrus achieved a conception rate of approximately 50% vs. 30% for inseminations performed after 16 h from onset Ž Fig. 4.. From previous reports Ž Trimberger, 1948; Trimberger and Davis, 1943; Nebel et al., 1994; Maatje et al., 1997., near-optimal conception rates would be expected for cows submitted for insemination 12 to 18 h after detection of oestrus. Mathematical modeling to predict the optimal time for AI using activity pedometers and visual signs of oestrus, estimated 11.8 h from onset Ž Maatje et al., 1997., which coincides with the approximate midpoint of the 5 to 16 h optimum using the HeatWatch w system. 4. Conclusions The aim of most published applications of pedometers until Maatje et al. Ž 1997. has been to improve rates of oestrus detection. Pedometry systems, which allow identification of the onset of oestrus in addition to efficient and accurate oestrus identification, will increase the usefulness of such technology. Pedometry systems must use real-time data transfer to allow the identification of oestrus onset and accurate timing of insemination. Current pedometer systems only allow retrieval of activity data two or three times daily, usually at milking, reducing the effectiveness of determining the optimal timing of insemination. The onset of oestrus periods have been shown to be equally distributed during the day, and 24.1% of all oestrus periods were classified by Dransfield et al. Ž 1998. as Ž y1 having low intensity -1.5 standing events h. and short duration Ž -7 h.. These two characteristics strongly contribute to the low efficiency of oestrus detection. Guidelines for the timing of insemination set forth by Trimberger Ž 1948. suggest that the optimal interval for pregnancy results is for AI to occur approximately 12 h after initial observation of standing oestrus. Results reported here from systems that more accurately determine the precise onset of oestrus would suggest insemination should be performed earlier following the detection of oestrus. Using the a.m. p.m. guideline would lower the probability of resulting pregnancy, as many cows observed most likely have been in oestrus for several hours prior to the detection of oestrus. Previous studies have reported that, when onset of oestrus is not known, once daily insemination for cows observed in standing oestrus can be used as effectively as the a.m. p.m. guideline and results in no difference in conception rates. Results from our work with the HeatWatch w system would suggest that, if onset of oestrus is unknown, insemination should be performed within 6 h of initial observation of oestrus. References Allrich, R.D., 1994. Endocrine and neural control of oestrus in dairy cows. J. Dairy Sci. 7, 2738 2744. Amyot, E., Hurnik, J.F., 1987. Diurnal patterns of oestrous behavior of dairy cows housed in a free stall. Can. J. Anim. Sci. 67, 605 614. Arney, D.R., Kitwood, S.E., Phillips, C.J.C., 1994. The increase in activity during oestrus in dairy cows. Appl. Anim. Behav. Sci. 40, 211 218.

722 ( ) R.L. Nebel et al.ranimal Reproduction Science 60 61 2000 713 723 Brackett, B.G., Oh, Y.K., Evans, J.F., Donawick, W.J., 1980. Fertilization and early development of cow ova. Biol. Reprod. 23, 189 205. Brewster, J.E., Cole, C.L., 1941. The time of ovulation in cattle. J. Dairy Sci. 24, 111 115. Chenault, J.R., Thatcher, W.W., Kalra, P.S., Abrams, R.M., Wilcox, C.J., 1975. Transitory changes in plasma progestins, estradiol, and luteinizing hormone approaching ovulation in the bovine. J. Dairy Sci. 58, 709 717. DeSilva, A.W.M.V., Anderson, G.W., Gwazdauskas, F.C., McGilliard, M.L., Lineweaver, J.A., 1981. Interrelationships with oestrous behavior and conception in dairy cattle. J. Dairy Sci. 64, 2409 2418. Dransfield, M.B.G., Nebel, R.L., Pearson, R.E., Warnick, L.D., 1998. Timing of insemination for dairy cows identified in oestrus by a radiotelemetric oestrus detection system. J. Dairy Sci. 81, 1874 1882. Farris, E.J., 1954. Activity of dairy cows during oestrus. J. Am. Vet. Med. Assoc. 125, 117 121. Foote, R.H., 1974. Oestrus detection and oestrus detection aids. J. Dairy Sci. 58, 248 256. Gwazdauskas, F.C., Whittier, W.D., Vinson, W.E., Pearson, R.E., 1986. Evaluation of reproductive efficiency of dairy cattle with emphasis on timing of breeding. J. Dairy Sci. 69, 290 297. Hawk, H.W., 1987. Transport and fate of spermatozoa after insemination of cattle. J. Dairy Sci. 70, 1487 1503. Heersche, G., Nebel, R.L., 1994. Measuring efficiency and accuracy of detection of oestrus. J. Dairy Sci. 77, 2754 2761. Hurnik, J.F., King, G.J., Robertson, H.A., 1975. Oestrous and related behaviour in postpartum Holstein cows. Appl. Anim. Ethol. 2, 55 68. Kiddy, C.A., 1977. Variation in physical activity as an indication of oestrus in dairy cows. J. Dairy Sci. 60, 235 243. Larsson, B., 1987. Determination of ovulation by ultrasound examination and its relation to the LH peak in heifers. J. Vet. Med. 34, 749 756. Lehrer, A.R., Lewis, G.S., Aizinbud, E., 1992. Oestrus detection in cattle: recent developments. Anim. Reprod. Sci. 28, 355 361. Maatje, K., Loeffler, S.H., Engel, B., 1997. Optimal time of insemination in cows that show visual signs of oestrus by estimating onset of oestrus with pedometers. J. Dairy Sci. 80, 1098 1105. Mattoni, M., Mukasa-Mugerwa, E., Cecchini, G., Sovani, S., 1988. The reproductive performance of East African Ž Bos indicus. Zebu cattle in Ethiopia: 1. Oestrous cycle length, duration, behavior and ovulation time. Theriogenology 30, 961 971. Nalbandov, A., Casida, L.E., 1942. Ovulation and its relation to oestrus in cows. J. Anim. Sci. 1, 189 195. Nebel, R.L., Walker, W.L., McGilliard, M.L., Allen, C.H., Heckman, G.S., 1994. Timing of insemination of dairy cows: fixed time once daily versus morning and afternoon. J. Dairy Sci. 77, 3185 3191. Pennington, J.A., Albright, J.L., Callahan, C.J., 1986. Relationships of sexual activities in oestrous cows to different frequencies of observation and pedometer measurements. J. Dairy Sci. 69, 2925 2934. Rajamahendran, R., Robinson, J., Desbottes, S., Walton, J.S., 1989. Temporal relationships among oestrus, body temperature, milk yield, progesterone and luteinizing hormone levels, and ovulation in dairy cows. Theriogenology 31, 1173 1182. Senger, P.L., 1994. The oestrus detection problem: new concepts, technologies, and possibilities. J. Dairy Sci. 77, 2745 2753. Stevenson, J.S., Lamb, G.C., Kobayashi, Y., Hoffman, D.P., 1998. Luteolysis during two stages of the oestrous cycle: subsequent endocrine profiles associated with radiotelemetrically detected oestrus in heifers. J. Dairy Sci. 81, 2897 2903. Stevenson, J.S., Smith, M.W., Jaeger, J.R., Corah, L.R., LeFever, D.G., 1996. Detection of oestrus by visual observation and radiotelemetry in peripubertal, estrus-synchronized beef heifers. J. Anim. Sci. 74, 729 735. Thibault, C., 1973. Sperm transport and storage in vertebrates. J. Reprod. Fertil., Suppl. 18, 39 53. Trimberger, G.W., 1948. Breeding efficiency in dairy cattle from artificial insemination at various intervals before and after ovulation. Nebr., Agric. Exp. Stn., Res. Bull. 153, 1 26. Trimberger, G.W., Davis, H.P., 1943. Conception rate in dairy cattle from artificial insemination at various stages of oestrus. Nebr., Agric. Exp. Stn., Res. Bull. 129, 1 14. Varner, M., Maatje, K., Nielen, M., Rossing, W., 1994. Changes in dairy cow pedometer readings with

( ) R.L. Nebel et al.ranimal Reproduction Science 60 61 2000 713 723 723 different number of cows in oestrus. In: Proc. 3rd Int. Dairy Housing Conf. Am. Soc. Agric. Eng. Orlando, FL. ASAE, St. Joseph, MI, pp. 434 442. Walker, W.L., Nebel, R.L., McGillard, M.L., 1996. Time of ovulation relative to mounting activity in dairy cattle. J. Dairy Sci. 79, 1555 1561. Wilmut, I., Hunter, R.H.F., 1984. Sperm transport into the oviducts of heifers mated early in oestrus. Reprod. Nutr. Dev. 24, 461 468. Xu, Z.Z., McKnight, D.J., Vishwanath, R., Pitt, C.J., Burton, L.J., 1998. Oestrus detection using radiotelemetry or visual observation and tail painting for dairy cows on pasture. J. Dairy Sci. 81, 2890 2896.