“NO REPRODUCTION NO PRODUCTION” on this basis the reproduction management is an economic determinant in the success of any dairy enterprise. Among many components of reproduction management, estrus detection is the crucial one, as it contributes towards the ultimate pregnancy rate and survival of the embryo
(Layek et al., 2011a). Inadequate heat detection has been identified as a major limit to herd reproductive performance over many years. The effective reproduction management implies that within a determined period a pregnancy is achieved for each cow. High rates of detection of estrus with satisfying accuracy have an important impact on realization of gestation (
Boyd, 1984). It has been reported that 11.5% cattle and 20.75% buffaloes are inseminated at improper time
(Kumaresan et al., 2001), which clearly indicates that the cow were wrongly detected as in estrus and insemination of these cow incurs heavy loss in term of wasteful expenditure of quality male germplasm, production loss and increased risk of introducing genital infection in female. Each missed heat represents the loss of a complete estrus cycle of approximately 21 days that in a seasonally calving herd represents 21 days of lost potential production, so each missed heat has a significant financial loss. In this regard, attaining higher estrus detection efficiency and accuracy is an important key to improve individual animal along with overall herd fertility. The selection of proper estrus detection method for a particular dairy farm is dependent on several factors like scale of operation, availability of man power, type of animal
etc. and single aid can’t be used invariably. Various heat detection aids have been developed over the years for helping dairy producers to detect cows in heat
viz., chin-ball-marker, tail paint, pedometer, accelerometers
, Kumar patches and radio telemetric systems (Heat Watch). By implementing an automatic detection system such as pedometers, detection rates could improve, which would increase other reproductive measures such such as conception rates, pregnancy rates and calving intervals. It has been determined that pedometers can accurately detect estrus
(Firk et al., 2002; Roelofs et al., 2005; Madkar, 2013).
Estrus signs
The term estrus is derived from a Greek word ‘oistros’, it means ‘gatfly’whose buzzing during summer causes the cow to frenzied behavior (
Short, 1994). The restricted period of sexual receptivity, characterized by intense sexual desire, when the female will seek the male and even have periods of homosexual activity, in which a cow mimics the bull (
Thomas and Dobson, 1989). Estrous cycle can be divided into estrus, met estrus, diestrus and proestrus (Senger, 2005). The estrus period is characterized by sexual receptiveness and is distinguishable from the other stages due to specific behavioral signs (
Senger, 2005). A number of authors have described the common behaviors expressed by dairy animals when in estrus
(Lyimo et al., 2000; Senger, 2005;
Roelofs et al., 2005; Gunasekaran et al., 2007; Mangal, 2009;
Layek et al., 2011; Madkar et al., 2014; Madkar et al., 2015) the following are the most commonly reported behavioral characteristics:
- Standing to be mounted.
- Mounting other cows.
- Rubbed rump and tail-head.
- Chin resting.
- Restlessness.
- Increase in agonistic interactions (
e.g. head to head fights).
- Sniffing of the vagina of herd-mates.
- Flehmen reaction (wrinkling of the nose and curling of the lip).
- Frequent micturition.
- Tail raising.
- Bellowing.
- Mounted but not standing.
Senger (1994) proposed a standard model of an electronic system for detection of estrus, with the following characteristics:
1- The system must allow, through electronic, chemical, or visual means, continuous surveillance of quantifiable behavioral or physiological changes occurring during estrus.
2- The technology should provide automatic animal identification, capable of storing information related to the estrus event for future data retrieval. This identification should be permanent, allowing for monitoring throughout the animal‘s lifetime.
3- The system should be cost-effective with minimal human intervention.
4- Lastly, the monitoring device should measure a parameter that is highly correlated with the time of ovulation, ensuring high specificity.
There are currently two methods commercially available that possess the above-mentioned characteristics: heats mount detectors and automated activity monitoring systems. These will be described in more detail, with emphasis on automated heat detection systems.
Different methods for heat detection
Heat mount detectors
Heat mount detectors are pressure-sensors placed on the cow‘s rump and are stimulated each time the cow is mounted. The electrical stimulus is transmitted
via a radiotelemetry to a central transponder, which collects data on the frequency of mounts over time and the duration of each mount. Estrus is signaled when a minimum of three successive mounts lasting 2 seconds or longer occurs within a 4-hour period
(Rorie et al., 2002). Animals reared in different management systems may show the deferent intensities of estrus behavior
(Xu et al., 1998; Palmer et al., 2010). This affects the accuracy and efficiency of a system for detecting heat.
Xu et al., (1998) evaluated the HeatWatch® system, a commercial automated heat mount detector, in a pasture-based Friesian herd and demonstrated that the system identified 91.2% of the periods of estrus. These periods had a mean duration of 9.7 hours (h) with 13.6 mounts per period.
At-Taras and Spahr (2001), using non-synchronized cows fitted with HeatWatch® in a free stall barn found that the total mounting activity averaged 5.8 0.78 h with 6.7 mounts per heat. Recently,
Palmer et al., (2010) compared estrus characteristics between Holstein-Friesian cows housed in a free stall or on pasture. Again, estrus was measured with Heat Watch® and a higher incidence of sub-estrus events.
Lopez et al., (2005) observed that high producing primiparous and multifarious Holstein cows (³39.5 kg/day), had a shorter duration of estrus (6.2 h) measured by Heat Watch® when compared to lower producing cows (10.9 h). Also, the same study showed a reduction in the number of standing events between high (6.3) and low producing cows (8.8). This suggests that cow productivity affects expression of estrus and may also affect heat mount detectors’ efficiency.
Development of activity monitoring systems
The increase in physical activity during estrus in dairy cows was first described by Farris (1954) in a study on 13 Guernsey cows. Later,
Kiddy (1977) evaluated the mean variation in physical activity of dairy cows in estrus using pedometers for human use. The devices were covered with a plastic case and attached to the cow‘s rear ankle using a strap. The study collected data on 87 estrous periods observed in 40 Holstein cows in a free stall barn, in which the average increase in movements during estrus was 393% compared to activity during non-estrus periods. These two studies were the basis for the development of activity systems for heat detection.
Heatime® (SCR Engineers, Natanya, Israel) is an activity monitoring system. It calculates the mean activity of rolling seven day periods, which is used as the reference period and is compared to the same time interval of the current day. In other words, the activity pattern for Day 8 is compared to the mean activity of Days 1 through 7, which is further divided into time intervals (
Bar, 2010). Afimilk®, another activity monitoring system, uses the mean value of the previous 10 days (
Galon, 2010). This reference period is included in the calculation for the heat-alarm detection of each commercial tag another point of consideration with regards to activity systems is the methodology used to measure activity. Some activity systems are based on a mercury-switch technology. Depending on the animals’ movement, the mercury tilt is switched on and off
(Firk et al., 2002) and a step counter component is linked to that. This system is probably obsolete, when compared to newer technologies using accelerometers. These are commonly used in electronically devices such as iPods, smart phones, GPS,
etc and are based on gravitational forces (
i.e. ―G-forces).
Heat detection intensity and accuracy with activity systems
Peralta et al., (2005) reported an efficiency of only 37% using ALPRO (DeLaval) in a large free stall Holstein herd during summer,
At-Taras and Spahr (2001), also evaluating cows in free stall conditions, found estrus detection efficiency to range from 79% to 87%. The first study calculated system efficiency based on the number of possible estrus periods that could have occurred during the study period, considering a 21-day cycle. Meanwhile, the second study used a combination of estrus signs to define if a true estrus had happened. The differences might be due to distinct definitions of a real estrus and
(Peralta et al., 2005) evaluating cows only during heat stress.
The sensitivity (
i.e. efficiency) of the activity system used in detecting heats by using the following formula
(Dohoo et al., 2009).
Silent heats in activity monitoring systems
Another variable that may influence the efficiency of estrus detection with activity monitors are silent ovulations not preceded by a sufficient increase in walking activity to signal a heat. Only one study was found exploring the possible association between silent ovulations and activity estrus
(Ranasinghe et al., 2010). This study used 161 Holstein-Friesian cows in a free stall equipped with Afimilk pedometers to analyse cow-level factors that may be associated with the occurrence of silent ovulations. Parity, season of ovulation (temperature range was -9.6° to 19.9°C) and postpartum diseases were not found to be risk factors for silent ovulations. Conversely, milk yield was significantly associated with an increase in silent ovulation for the second to fourth ovulations postpartum in high producing dairy cows. Considering that the voluntary waiting period usually lasts 50 days, the most relevant frequencies of silent ovulations would be the third and fourth estrus periods. In the study, the frequencies of silent ovulations for these estrus cycles were 21.3% and 10.5% respectively.
(Palmer et al., 2010) found an average incidence of 35% of apparent silent ovulations in Holstein-Friesian cows housed in a free stall using the Heat Watch system.
Walking activity related to ovulation
Estradiol is the hormone responsible for the manifestation of estrus behaviour (
Senger, 2005).
Shemesh et al., (1972) demonstrated that during estrus the estradiol concentration reached its peak 4 hours before the cow accepted mounting activity by a bull. This increase and subsequent steep decline of estradiol
(Shemesh et al., 1972; Dieleman et al., 1986) precedes the LH surge and consequently, ovulation. Only one study was found correlating the peak of estradiol to walking activity.
Lyimo et al., (1996) found that the peak of walking activity occurred 8 hours after the peak in estradiol concentrations. However, on closer inspection, this study may have failed to correctly analyze the pedometer data because only the raw count of steps was used in the analysis. It is well documented that cows’ activity follows a circadian rhythm (
Farris, 1954, 1994;
Lovendahl and Chagunda, 2010). Therefore, raw activity data should not have been used, because deviation in activity needs to be compared to the same time interval of the day it was measured. While it is possible that some time lag between the peaks of estradiol to activity estrus detection occurs, this should be less than 8 hours.
Yoshioka et al., (2010) demonstrated that, from the onset of increased activity (
i.e. above the threshold) to standing heat behavior, the interval was only 1 hour. Moreover, by using activity thresholds varying from 150% to 300% of the mean average activity from diestrus period, the interval from onset of increased activity to ovulation decreased by approximately 1 hour.
Activity monitoring systems on the neck versus on the leg
Kiddy (1977) described the first potential use of pedometers as an aid for heat detection, but indicated that the movement of the hind legs might not be well correlated with progressive movement. Interestingly, the patent of the first U.S. electronic estrus detection system for dairy cows was actually based on the invention of a motion detector and a digital counter applied in conjunction with an animal identification system attached to the cows‘ neck strap (
Rodrian, 1981). Therefore, the system would actually measure body movements, as opposed to steps.
Eradus, (1992) compared heat detection accuracy using a mercury switch based pedometer on the foreleg and on the neck and found more false positive alerts when the activity system was on the neck.
Sakaguchi et al., (2007) compared sensitivity and positive predictive value of a commercial Japanese pedometer attached to the neck or hind legs using 6 Holstein heifers in an open paddock. This study tested different activity thresholds and reference periods. Depending on these variables, there were times that the activity system on the leg performed better than the neck and vice versa, but there was a tendency of the leg system to perform better most times.
Løvendahl and Chagunda (2010), using a modified algorithm on ALPRO system (neck collar) and different thresholds, found detection rates between the ranges of 56.4% to 84.2%. However, the most interesting part of this work was the low error rate (false positives) associated with the algorithm, which ranged from 0.38% to 3.85%. This work has shown that by including smoothed deviations of the activity data in the heat detection function, this could account for most of the noise (
i.e. error) associated with activity being measured on the neck, resulting in high positive predictive values.
Holman et al., (2011) compared sensitivity and positive predictive values of two commercial systems (Afimilk® system; leg tag and Heatime collars; neck tag) in detecting estrus periods. The sensitivity did not differ between systems, but the Heatime collars had a significantly higher positive predictive value when compared to Afimilk® tags (93.5% and 73.5%, respectively). With the advances in technology used in activity systems and knowing that the tag attached to the leg is measuring motion and not necessarily only step, the correct terminology to define such systems would be activity monitoring system for heat detection. The term pedometer‘refers to the location of attachment of the tag, on the hind or forelegs.
Possible improvements in activity monitoring systems
For several years researchers have been trying to improve estrus detection and more recently, detection of ovulation using precision technologies, such as vaginal mucus resistance and accelerometer, temperature, daily milk yield, body and milk temperature, heart rate
(Firk et al., 2002). It is beyond the scope of this review to discuss these methods, but lying behaviour and rumination time have been recently implemented into newer activity systems. Heatime®, SCR has a new activity tag which also provides rumination data while Afikim® has introduced a measurement for lying behavior. However, such variables have yet to be used in conjunction with walking activity to improve heat detection rates, despite the availability of some data to support such applications
(Brehme et al., 2008; Brizuela, 2010;
Jónsson et al., 2011). By using a combination of measures in the formulation of estrus detection algorithms, heat detection rates greater than 90% would be likely achievable, while reducing the percentage of error rates. Additionally, the combination of more variables in the heat detection algorithm might help detect silent ovulations
(Firk et al., 2002).
Accelerometer
Accelerometers are devices capable of distinguishing between periods of no ambulation and periods of dynamic activity because of their ability to measure the force of gravity as well as accelerations due to movement (
Aminian 2004). This ability allows the monitoring of multiple species (Robert and White 2009). The positional location of the accelerometer is critical for defining various behaviors; accelerometers mounted to the lower limbs of the animal to describe behaviors such as lying, standing and walking (
Robert and White 2009;
Trenel et al., 2009). A sequence of studies was performed using one type of device in order to investigate the potential for three-dimensional (tri-axial) accelerometers to be used to describe both static and dynamic activities of free-moving humans and found sensitivity and specificity values of 99% and 94% respectively when using training and testing subsets of data
(Mathie et al., 2004). A similar study investigated the potential of a three dimensional accelerometer accompanied by embedded intelligence to perform onboard calculation of signals
(Narayanan et al., 2006).
Rumen temperature
It is well known that core body temperature of animal increases during in estrus. Increases in RT occurred at estrus compared with other periods
(Burns et al., 2002). The RT increase of 1.3°C at estrus in dairy cows. Increases in ruminal temperature (RuT) at estrus have been established and Measurement of RuT is achieved by administering a temperature bolus into the rumen of cows and information is sent to a computer
via telemetry
(Piccione et al., 2003). Ruminal temperature increased 0.4 to 1.0°C at estrus and RuT was not influenced by season
(Bailey et al., 2009; Cooper-Prado et al., 2011). An increase in RuT ³ 0.3°C and ³ 0.7°C during an 8 h evaluation period, compared with a mean pre-estrus RuT 12 to 84 h before the 8 h period, correctly identified 95% and 42%, respectively, of cows in estrus in December and 100% of estrous cows were detected in May by either a ³ 0.3°C or ³ 0.7°C RuT increase
(Bailey et al., 2009).
