Towards sensor-based calving detection in the rangelands: a systematic review of

credible behavioral and physiological indicators

Anita Z. Chang1, , David L. Swain, and Mark G. Trotter
Institute for Future Farming Systems, School of Health, Medical, and Applied Sciences, Central Queensland
University, Rockhampton North, QLD 4701, Australia

ABSTRACT: Calving is a critical point in both A systematic literature review was conducted

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

a cow and calf’s life, when both become more to identify changes in the dam’s behavioral and
susceptible to disease and risk of death. Ideally, physiological states in response to a calving event.
this period is carefully monitored. In extensive Articles (n = 104) consisting of 111 independent
grazing systems, however, it is often not eco- experiments were assessed following an inten-
nomically or physically possible for producers to sive search of electronic databases. Commonly
continuously monitor animals, and thus, calving reported indicators of parturition (n = 38) were
frequently goes undetected. The development identified, and temporal trend graphs were gener-
of sensor systems, particularly in these environ- ated for 13 of these changes. The results compare
ments, could provide significant benefits to the trends in behavioral and physiological changes
industry by increasing the quantity and quality across a variety of animal-related factors and
of individual animal monitoring. In the time sur- identifies several reliable indicators of parturition
rounding calving, cows undergo a series of be- for detection with sensors, namely calf grooming
havioral and physiological changes, which can behavior, changes in rumination duration, and
potentially be detected using sensing technolo- lying bouts. This synthesis of literature suggests
gies. Before developing a sensor-based approach, that variability exists between individuals and
it is worthwhile considering these behavioral and thus, combining several calving indicators may
physiological changes, such that the appropriate result in a more broadly applicable and accurate
technologies can be designed and developed. detection of parturition.
Key words: behavior, cattle, parturition, precision livestock management

© The Author(s) 2020. Published by Oxford University Press on behalf of the American Society
of Animal Science.
This is an Open Access article distributed under the terms of the Creative Commons Attribution
Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits
non-commercial re-use, distribution, and reproduction in any medium, provided the original work is
properly cited. For commercial re-use, please contact [email protected]
Transl. Anim. Sci. 2020.4:1-18
doi: 10.1093/tas/txaa155

INTRODUCTION processes for the production of milk. Similarly, in

beef cow–calf operations, the profitability of the
Reproductive efficiency is an important driver enterprise is reliant upon the successful produc-
of economic profitability in commercial cattle tion and growth of calves to weaning. Calving is a
production. In dairy production systems, preg- period of increased biological challenge and risk,
nancy and consequently calving are essential marked by a series of behavioral and physiological
changes. Complications during the calving period
Corresponding author: [email protected]
1 can result in cow and calf morbidity or death,
Received July 30, 2020. leading to reduced economic and animal welfare
Accepted August 13, 2020. outcomes. The most common complication is
1
2 Chang et al.

dystocia, which refers to an abnormal, prolonged, to undertake a meta-analysis (Grant and Booth,
and/or difficult calving, the period of time which 2009). In this review, however, the variation between
commences with the onset of uterine contractions data were too substantial and a true meta-analysis
(Meijering, 1984; Mee, 2008). Dystocia increases could not be conducted.
the likelihood of stillborn calves and decreases calf Electronic databases were searched in March
vigor during the neonatal period (Riley et al., 2004; 2018 for literature investigating changes in the be-
Barrier et al., 2012). Close monitoring of cattle havior and physiology of cows around a calving
around the time of calving can assist in the early event. The databases utilized for the search were
detection and adequate management of complica- BioOne, ProQuest, ScienceDirect, and Web of
tions. Traditionally, monitoring the behavioral and Science. Articles that met the set criteria under-
physiological changes associated with calving has went a further bibliographic search for relevant
been undertaken by experienced managers through articles. The terms used in the database search in-
careful observation in the field. The extensive nature cluded combinations of the following terms: “be-

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

of some production systems prevents close human havior*,” “birth*,” “calf,” “calving*,” “cow*,” and
observation of animals in the herds, and in some “parturition*.”
situations, the lack of available skilled labor further Articles had to meet the following criteria to be
limits the capacity to conduct these observations. included: 1) be written in English; 2) identified cattle
To overcome these limitations, the livestock as the primary subjects; 3) had to occur within 1 mo
industry has been exploring the potential for tech- prior to and/or following parturition; and 4) had to
nologies to remotely detect the behavioral and be focused on dam behavior and/or physiological
physiological changes associated with parturition. changes. Books and book chapters that were un-
There are several examples of sensor systems, which available electronically were excluded. Literature
have been developed or modified to detect partur- review articles were excluded, but any literature re-
ition, however, most remain relevant in a research views of a relevant topic underwent an extensive
context or, at best, in more intensive production bibliographic search, with appropriate references
systems (Trotter, 2010; Trotter et al., 2010). To op- subsequently included. Dissertations and theses
timize a parturition detection system for extensive were automatically filtered out on ProQuest; how-
grazing systems, an enhanced understanding of the ever, if a dissertation or thesis was identified dur-
anticipated behavioral and physiological factors ing a bibliographic search, the thesis was included.
likely to be expressed is needed. No other dissertations or theses were found using
A systematic literature review was conducted to any of the other database search engines. Similarly,
examine the results of scientific studies pertaining conference papers were also excluded unless these
to behavioral and physiological changes expressed were published or referenced in a subsequent
by a cow before, during, and after calving. The fol- peer-reviewed article. In experiments where human
lowing questions were explored: 1) How does cow intervention occurred, such as the administration
behavior and physiology change around a calving of pharmacological substances, the induction of
event? 2) How does methodology, breed, and pro- parturition, or caesarean sections, these articles
duction style vary in existing literature? 3) How were excluded. The exception to this was an art-
could the behavioral and physiological indicators icle where cow–calf separation occurred (Lidfors,
of calving best be detected using sensor technology? 1996). Because the intervention did not directly af-
fect the process of parturition, all prepartum and
calving changes were included, but all results were
MATERIALS AND METHODS
omitted following the dam–calf separation.
Search Strategy
Data Collection and Extraction
A systematic literature review methodology
based on that described in Williams et al. (2016) For each study that met the required search cri-
and Fogarty et al. (2018) was used for searching for teria, the author, year of publication, and title were
articles for inclusion in this review. This method- recorded. Articles with multiple experiments using
ology involved searching online databases to find different animals that were studied independently
articles that met a number of strict criteria (Grant of one another, were considered separate experi-
and Booth, 2009). Further evaluation was con- ments. In the event where each independent experi-
ducted to determine whether the results of the art- ment examined a different calving indicator and for
icles were quantitative and sufficiently comparable which there were inconsistent results, the experiment
Translate basic science to industry innovation
 Calving indicators for sensor development 3

in which the change was observed was recorded. For then categorized more broadly by methodology or
example, in Lidfors et al. (1994), which was made technology type (sensor technology and traditional
of three independent studies, isolation was only ob- methodology). Technologies were classified as sen-
served in experiment two and three, and would be sor, where the devices had the capacity to store or
labeled as so, whereas location changes were shown transmit data, although traditional methodolo-
in all three experiments and thus the experimental gies required manual data collection with further
number was not disclosed. The experimental loca- analysis. Rumination monitors were considered to
tion was recorded by country then by geographic be rumination detection systems that were micro-
region (Africa, Asia, Europe, North America, phone based. Accelerometers that were used for the
South America, and Oceania). If no experimental purposes of rumination detection were categorized
location was provided, the location was considered as being accelerometers. Rumination halters were
to be the address of the first author of the article. halters that utilized a pressure sensor to determine
The experimental animals were classified based on when a cow was ruminating.

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

a combination of breed and experimental site—
dairy cattle, extensive beef cattle, and intensive beef Data Analysis
cattle. Dual purpose animals were classified based
on the descriptor in the article and if none was pro- The results of each independent experiment
vided, the experimental animals were considered to were analyzed to identify any behaviors or physio-
be beef cattle. For the purposes of the present lit- logical changes that were associated with partur-
erature review, extensive beef cattle were those that ition. Behaviors that were identified in three or
had access to pasture for the entirety of the trial, more articles were categorized as “commonly re-
although intensive beef cattle were any species that ported” indicators of parturition. Indicators that
did not have constant access to pasture or where were observed instantaneously as being present or
the experimental conditions were not specified. absent, were scored on a discrete scale, or had dis-
Number of experimental animals was recorded, as crete characteristics were excluded from temporal
well as birth type (dystocia, eutocia, both), cattle analysis. These were considered “categorical” indi-
breed, and parity (multiparous, primiparous, both). cators. For example, dam isolation from the herd
If the birth type was not specified, it was assumed was observed as being present or absent at the end
that a normal birthing process had occurred. For of the calving period, udder changes were observed
each independent experiment, the duration and in- on a scale corresponding to the state of the udder,
tensity of observations were recorded relative to the and location consisted of discrete, non-measura-
point of calving (the point at which the calf is fully ble characteristics, such as calving on the sand or
expelled). Experimental intensity was more broadly calving in a forest. A present or absent status was
categorized as being high intensity (50% or more of allocated based on the behavior or physiology dis-
the time monitored) or low intensity (less than 50% played by the majority of the animals. For example,
of the time monitored) For experiments that com- in Lidfors and Jensen (1988), 2 of 20 cows were ob-
menced based on a behavioral change, such as the served isolating themselves prior to calving, there-
first stage of labor (Selman et al., 1970; Wehrend fore, the study was recorded to have an “absent”
et al., 2006), behavioral signs of parturition (Owens status for isolation, as the majority of animals did
and Edey, 1985), the visibility of the amniotic sac not isolate themselves.
(Schuenemann et al., 2011), and the appearance of A simple vote counting method was used
the calf’s hooves (Lidfors et al., 1994), start and end to compare the number of experiments that re-
points were calculated based on the time intervals ported positive, negative, or no change trends for
identified in any of the articles that met the search each indicator with temporal data pre- and post-
criteria. Methodology and technology usage was parturition, where each observation was delin-
first recorded as methodology or technology type eated a “point” or “vote” for the corresponding
(accelerometer [leg-deployed], accelerometer [other, category (Williams et al. (2016). These indica-
e.g., collar or ear tag], blood sample, feed intake tors were considered “continuous” The trend of
monitor, global positioning system [GPS], heart change was determined based on the magnitude
monitor, liver biopsy, real time forage weighing, of change between the values at the start of the
visual observation, ruminal bolus, ruminal fluid observation to the end; therefore, if a spike was
sample, rumination monitor, spectrophotometer, detected midway through the observation period,
tail movement detector, temperature data logger, it was ignored. If there were multiple categories
thermometer, video, and water intake monitor), of animals with different trends, another “vote”
Translate basic science to industry innovation
4 Chang et al.

was allocated. Articles where the continuous vari- were plotted using a time series point plot. A simple
able was observed instantaneously, or its presence means of visualizing the overall trend in physio-
was simply noted, were omitted from the vote logical or behavioral characteristics was developed.
counting, but were still included in the corres- The values generated for each timeframe were aver-
ponding reference list. aged, and a rolling mean was created from the data,
Articles reporting trends over time were common incorporating the value for the time prior to and
in the literature, and so a method of broadly visu- following any one interval. This reduced the inci-
alizing the trends in these results was developed. dence of skewed and biased data, due to the lack of
Time series graphs were generated for behaviors data present for some timeframes. The visualization
and physiological changes where quantifiable data technique developed is not intended to represent the
was made available in five or more independent true or absolute likely changes expressed in the data
studies. Two temporal graphs were generated for but is designed to provide an overall indicator of
each of the described changes—a short-term graph likely variation over time. There are obvious limita-

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

depicting the 24 h prior to and following partur- tions to this technique, where large variation in the
ition, and a long-term graph spanning the 30 d range of data exists. This will be highlighted in the
prior to and following parturition. This approach results and discussed later.
ensured that there was sufficient data generated for
a baseline trend (control) to be established for com-
RESULTS
parison to parturition behaviors or physiology. The
latest timeframe any study used as a control period
was 1 d prior to parturition (Miedema et al., 2011a, Search Statistics
2011b). Data collected over a time period were The database search returned 594 articles for
plotted as occurring at the end of the observation analysis. Approximately 15.8% (n = 94) involved
period. For example, if lying behavior was observed other species that were not cattle, such as deer,
from −10 to −2 h, the data would be plotted for moose, and ungulates. There were 13.8% (n = 82)
the −2 h time point. In many studies, observations in which the topic was not within the designated
ceased at common time points, such as −2h, but time frame (1 mo prior to and/or following birth).
commenced over a variety of periods, for example, Another 24.1% (n = 143) did not focus on the dam’s
−10h and −5h. By allocating the data value to the behavior and/or physiological state, although a fur-
end point of the observation period, the number of ther 1.52% (n = 9) of articles were eliminated as
datapoints at commonly reported time points, such they involved some form of human intervention
as −2h, is maximized. As such, the trendline gener- to the normal calving process. The majority of art-
ated for the time series graphs experiences less bias icles (27.6%, n = 164) were eliminated based on the
as there are fewer single datapoints to skew the line. document type and a further 0.5% (n = 3) were ex-
A new series was generated for each experimental cluded as these were not written in English. A fur-
unit, for different parities, and/or for different birth ther 7.2% (n = 43) were not retrievable. There were
types. If figures were made available, but no data 9.4% (n = 56) of articles that were consistent with
were presented or discussed in the text, the data all of the selection criteria and were included in
were estimated from the figures. the initial count of eligible articles for review. The
On a subday scale, all sources that did not have bibliographies of these initial eligible articles were
a data point within the 24 h time period were ex- examined for additional eligible articles.
cluded from the graph to increase clarity. When The bibliographic search resulted in 135 unique
constructing the long-term graphs, measurements articles that were considering as being potentially
were only included if these were recorded at a min- relevant. Upon further investigation, only 35.6%
imum interval of 24 h. For example, if dry matter in- (n = 48) were considered eligible as per the criteria
take was measured from −4 to −1 d, this data point applied to select the first cohort of articles. There
would be excluded from the long-term graph, as the were 7.4% (n = 10) of articles that did not inves-
interval was for more than 24 h. In the event where tigate dam behavior around calving, 4.4% (n = 6)
two units were used for a physiological change, for that did not fall within the designated time period,
example, µg/mL and ng/mL, and a direct conver- and 3.0% (n = 4) were considered irrelevant be-
sion was not possible, a secondary axis was used to cause there was administration of pharmacological
plot the second measurement. substances or the induction of parturition. A large
To enable a basic visualization of the changes in proportion of articles (41.5%, n = 56) were irretriev-
physiological or behavioral characteristics, all data able, although one article (0.7%) was not written in
Translate basic science to industry innovation
 Calving indicators for sensor development 5

English, and a further 6.7% (n = 9) were excluded experiments, the most common being Holstein/
based on the type of article. Friesian varieties (n = 78), followed by crossbreed
A total of 104 articles in which there were re- animals (n = 13). The breed was not recorded for
ports on 111 independent experiments, were de- eight experiments (Supplementary Table S1).
termined to meet the selection criteria pertaining
to relevant dam behavior and physiology informa- Monitoring Techniques and Technologies
tion around the time of calving (Supplementary
Table S1). Traditional monitoring techniques, such as
observation and body fluid sampling, occurred in
Animals and Study Environment 64.2% (n = 122) of the studies, compared to sen-
sor-based monitoring, which occurred in 35.8%
The independent experiments spanned across 6 (n = 68) of the studies. Blood sampling was the most
continents and 25 countries: Africa (n = 1), Asia commonly used methodology (28.7%, n = 35), fol-

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

(n = 6), Europe (n = 47), North America (n = 45), lowed by visual observation (27.9%, n = 34). The
South America (n = 1), and Oceania (n = 6) most commonly reported technologies used were
(Supplementary Table S1). the feed intake monitor and the leg-based accel-
The majority of experiments were conducted erometer (23.9%, n = 16). The least commonly re-
using cattle of the Bos taurus genotype (98.2%, ported technologies used were the GPS, rumination
n = 109), with only two experiments utilizing Bos halter, and tail movement detector, which were only
indicus cattle (Supplementary Table S1). A combin- used in one experiment each (Figure 2).
ation of primiparous and multiparous cows were Sensor technologies were used at both a high
researched in 43.2% of the studies (n = 48). In an and low observational intensity in almost equal
additional 39 experiments, only multiparous cows measures (51.5%, n = 34 and 48.6%, n = 32 respect-
(35.1%) were studied and in 9 experiments only ively), although traditional sampling was predomin-
nulliparous cows were studied (8.1%). For 15 in- antly low intensity (less than 50% of the experiment
dividual experiments, there was no specification monitored; 70.3%, n = 85). Sensor technologies
of cow parity. In the majority of the studies, cows made up 48.6% of high intensity methodology
experienced a normal birthing process (76.6%, (more than 50% of the experiment monitored), al-
n = 85), but in 26 experiments, there were cows though traditional sampling comprised of 72.7%
included that also experienced dystocia (23.4%, of all low intensity trials.
n = 26). No studies exclusively reported the use of Of particular interest in this study was the use
cows with dystocia. of sensors in the articles reporting on extensive
Dairy cattle were the most common category grazing environments. Only one article was iden-
(81.1%, n = 91), with a comparatively small number tified that described the use of a sensor (GPS) in
of studies reporting on extensive beef cattle (9.9%, an extensive grazing environment (Flörcke and
n = 11) and intensively managed cows (8.1%, Grandin, 2014).
n = 10) (Figure 1).
A total of 23 different cattle breeds were Duration and Observation Intensity
used to conduct the research in 111 independent
Experiments started on average 28.2 d prior
to calving and concluded 20.8 d following calving,
for a total average duration of 48.5 d. The longest
study was conducted by Laporta et al. (2014), com-
mencing 165 d prior to parturition and finishing
190 d following parturition, for a total length of
355 d (Supplementary Figure S1). More than half
of the studies were considered long-term (longer
than 1 wk prior to and following calving; 61.3%,
n = 68).
More than half of the independent experiments
were considered low intensity (64.22%, n = 70).
Figure 1. Distribution of articles that were accessed in which there
were independent experiments conducted with extensive beef cattle,
Most of the long-term studies were done at a low
intensive beef cattle, and dairy cattle. The experimental animals were intensity (72.06%, n = 49), although the short-
dairy cattle in the vast majority of the retrieved articles. term studies were more commonly done at a high
Translate basic science to industry innovation
6 Chang et al.

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

Figure 2. Methodology and technology utilization in all studies. Sensor technologies utilized included devices capable of storing or transmitting
data to the observer, although traditional methodologies required manual data collection. Although a wider range of sensor technologies were used
in the various experiments, there was an overall preference for use of traditional methodologies.

intensity (53.49%, n = 23). One study did not spe- Most of the indicators were classified as being
cify the measurement intervals within their experi- “continuous” and could initially be summarized
ments and thus, could not be allocated an intensity as either increasing or decreasing (Supplementary
category. Table S4). The most commonly recorded indicator
of calving was proportion of time spent lying, which
Trend and Magnitude of Change in Calving was examined in 30 independent experiments. Of
Indicators these, 25 were conducted with dairy cattle, 4 with
beef cattle in extensive settings, and 1 study with
Indicators reported in three or more studies beef cattle in an intensive setting. The majority of
were more closely examined to identify patterns of studies reported time spent lying to decrease before
change relative to parturition. Some of the behav- parturition (12/16), but others reported an increase
ioral indicators were classified as being “categor- (3/16) or no change (1/16). Other commonly iden-
ical” and could only be summarized at the level of tified calving changes included dry matter intake
being present or absent. Of these, the most com- (n = 24), feeding (n = 21), lying bouts (n = 19), and
monly recorded changes were changes in udder rumination (n = 19) (Supplementary Table S4).
appearance (n = 15), followed by changes in vulva Other commonly identified calving indicators in-
appearance (n = 10), changes in location (n = 9), cluded changes in dry matter intake (n = 24), feed-
and the appearance of the amniotic sac (n = 9). ing (n = 21), lying bouts (n = 19), and rumination
Only two of these categorical behaviors were ob- (n = 19) (Supplementary Table S4). Of all of these,
served following parturition (arched back, placen- the most consistent indicators prior to parturition
tophagia; Supplementary Table S2). were a decrease in rumination (100% of studies)
Change in cow location was of particular and an increase in lying bouts (92.9% of studies).
interest, given [c] the intent of this article to ex- Calf licking behavior only commenced following
plore indicators in extensive grazing systems parturition and was reported to decrease over time
and was subsequently explored in more detail. in all eight studies.
Location at calving varied between studies, but Changes in behavior and physiological state
there was a preference for sheltered areas, such as were most commonly observed in the prepartum
buildings or covered yards, and open areas, such period (n = 24) as compared with the postpartum
as fields, observed. In two studies, there were also period (n = 8). For the majority of the reported
indications that cattle had a preference for calving indicators of calving, consistent results were re-
in areas where other cattle had previously calved ported between all the references. The exceptions
(Rørvang et al., 2017, 2018) (Supplementary to this were BHBA concentration postpartum,
Table S3). insulin concentration postpartum, lying duration
Translate basic science to industry innovation
 Calving indicators for sensor development 7

postpartum, rectal temperature postpartum, res- Behavioral indicators. Only two studies con-
piration rate both pre- and postpartum, and tained graphable dry matter intake data for the
stepping behavior postpartum (Supplementary short-term, subday graph. One study reported both
Table S4). pre- and post-calving, although the other study
Of particular interest from a sensor-based per- only reported on dry matter intake prior to partur-
spective is heart rate and body temperature. The ition. In both studies, a decrease in dry matter in-
number of studies reporting on heart rate were take was observed prior to calving (Figure 3g). One
low (n = 4), however, the majority reported an article reported an increase in dry matter intake fol-
increase in heart rate prior to parturition (75%; lowing calving (Figure 3g). Several studies reported
Supplementary Table S4). Likewise, the number changes in dry matter intake over the long-term.
of studies reporting on body temperature were low Most studies reported a slight decrease in intake
(n = 3), however, all studies found that body tem- prior to calving, with an increase following calving
perature decreased prior to and increased following (Figure 3h).

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

calving (Supplementary Table S4). Three studies reported on time spent drinking
in the short-term—one in the pre-calving period,
Temporal Changes one in the post-calving period, and one across the
pre- and post-calving period. The study reporting
Temporal graphs were generated for all eligible on drinking time in both the pre- and post-calving
indicators to enable a broad, visual understanding periods indicated a decrease prior to calving, before
of which changes might provide the most reliable increasing thereafter (Figure 3i). This post-calving
indication of a parturition event. trend was also observed in the study reporting on
Physiological indicators. For the long-term as- drinking time following parturition (Figure 3i).
sessment, most studies reported that BHBA in- Duration of drinking increased over time in the
creased during the pre-partum period, with a peak days before and after calving (Figure 3j).
occurring 5 d post-partum (Figure 3a). In some Feeding duration decreased in all but one
studies, a decrease in glucose concentration was study in the prepartum period before increasing
observed prior to calving, although in others, an in all studies after calving (Figure 3k). Over the
increase was observed, resulting in no discernible long-term, feeding duration remained relatively
overall trend across studies (Figure 3b). There was consisted in all but one study (Figure 3l).
a slight decrease in insulin concentration across In all studies, licking of the calf decreased at
articles as measured in ng/mL during the calving an exponential rate following calving, with the
period, although insulin concentration in studies most rapid decrease observed in the first 6 h after
measured using uIU/mg reported a positive quad- birth (Figure 3m). Cows began licking their calves
ratic curve, with the minimum concentration occur- on average 12 min following birth, with the longest
ring during parturition (Figure 3c). One study was latency to lick being 258 min following birth by a
excluded (Grum et al., 1996), because there was a dairy heifer as reported by Selman et al. (1970) fol-
magnitude inconsistency in concentrations com- lowing a long and difficult labor (Figure 3m). Only
pared with those in the other studies. The units of two studies reported on calf licking duration over
measure for NEFA concentrations were µEq/L and the long-term, however, the decreasing trend was
mmol/L. For both units of measurement, there was only observed in one of the studies, although the
a significant increase immediately prior to calving other study did not show a marked change over
with concentrations returning to basal in the first time (Figure 3n).
few days following calving in all studies (Figure 3d). The number of lying bouts increased leading
The short-term graph indicated that in the majority up to calving in all studies, with a peak occurring
of studies, there was an increase in progesterone in during parturition and a decrease in the number
the hours leading pre- and postpartum (Figure 3e). of lying bouts in the initial period after calving
In two studies, there was a significant decrease in (Figure 3o). For the long-term, lying bouts were
progesterone concentrations approximately 10 d quite consistent across most studies (Figure 3p).
before parturition (Figure 3f). Two other studies In one study, a peak was observed at the time of
showed similar rapid declines around 5 d prior calving (Figure 3p).
(Figure 3f). All studies reported that minimum There was no distinguishable pattern identi-
concentrations occurred at the time of calving and fied in duration of lying in the period prior to and
basal concentrations sustained during the early following parturition in the majority of studies
postpartum period (Figure 3f). (Figure 3q). In two studies, an increase was observed
Translate basic science to industry innovation
8 Chang et al.

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

Figure 3. Long- and short-term temporal graphs generated for (a) β-hydroxybutyric acid (BHBA) (long-term), (b) glucose (long-term), (c) insulin
(long-term), (d) non-esterified fatty acids (NEFA) (long-term), (e) progesterone (short-term), (f) progesterone (long-term), (g) dry matter intake (short-
term), (h) dry matter intake (long-term), (i) drinking duration (short-term), (j) drinking duration (long-term), (k) feeding duration (short-term), (l)
feeding duration (long-term), (m) calf licking (short-term), (n) calf licking (long-term), (o) lying bouts (short-term), (p) lying bouts (long-term), (q)
lying duration (short-term), (r) lying duration (long-term), (s) rumination duration (short-term), (t) rumination duration (long-term), (u) standing
duration (short-term), and (v) standing duration (long-term). Graphs for short-term assessments had time ranges from the day prior to and following
parturition, although long-term graphs include a 30 d period from before and after calving. A secondary axis was generated where there was more
than one unit of measurement used. Data associated with the secondary axis are depicted with a cross marker, as opposed to the circle marker for data
corresponding to the primary axis. Thick grey lines depict the rolling average trendline, and darker grey thick lines correspond to the secondary axis.
Crosses used for graphing the short-term calf licking pattern refer to the time taken for the cow to initiate licking after birth. These data were excluded
from the trendline analysis. Numbers in the legend coincide with a reference, as listed in Supplementary Table S1.

around calving (Figure 3q). In the long-term, the lead up to parturition, before increasing thereafter
majority of studies reported on lying time prior (Figure 3s and t).
to calving (Figure 3r). In these, there was a slight Only three studies reported on standing time in
decrease in lying duration leading up to calving the short-term (Figure 3u). In the majority of stud-
(Figure 3r). ies reporting on time spent standing, no discern-
Studies reporting on rumination time in both ible change was observed on a subday or long-term
the short- and long-term reported a decrease in the level (Figure 3u and v).
Translate basic science to industry innovation
 Calving indicators for sensor development 9

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

Figure 3. Continued

DISCUSSION To focus the development of a sensor-based ap-

proach, this article has sought to identify the be-
The overarching objective of this article was
havioral and physiological changes associated with
to develop an understanding of the potential for a
sensor-based approach for detecting parturition in a parturition event and to synthesize the existing
extensive beef grazing systems. An essential compo- literature reporting on these changes. Analysis of
nent of this is the evaluation of changes in animal 104 articles identified 38 commonly reported indi-
physiology or behavior that might be measured cators that demonstrate variation around the time
using sensors. The initial analysis revealed that only of parturition, 23 of which were behavioral. Several
11 studies focused on monitoring parturition in ex- key indicators of parturition that will be discussed
tensively grazed cattle and only one study applied a have also been summarized in Supplementary Table
sensor (GPS). Comparing this to the 91 studies in S5, and have been assigned a low, medium, or high
dairy cattle suggests that there is a significant op- status depending on the likelihood of commercial
portunity for further research in this area. adoption for calving detection at this time. This
Translate basic science to industry innovation
10 Chang et al.

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

Figure 3. Continued

took into consideration the potential for practical more apparent. As such, other valuable calving in-
development in an extensive grazing context, the dicators may exist, however, given the lack of litera-
availability of a commercial device, the ease of de- ture available, they were not discussed in detail. The
ployment, and the consistency of reporting on the more consistent and apparent changes in behavior
indicator. or physiology are more likely to lend themselves to
The following discussion will focus on how being detected by a sensor system and consequently,
these indicators might be useful in the context of are the focal point of this discussion.
sensor-based systems aimed at detecting the par- Sensor technologies for monitoring livestock
turition event. In the context of investigating how are many and varied. Trotter et al. (2018) pro-
indicators might be detected in a sensor-based ap- vides a comprehensive list of the types of on-an-
proach, we have deliberately sought to focus on imal sensors being developed and applied in the
those behavioral or physiological characteristics, livestock sector and is worth reviewing to gain a
which are more consistently reported in the litera- broad understanding of the current state of the art
ture and those for which the change is clearer and in this field. This discussion, however, will refer to
Translate basic science to industry innovation
 Calving indicators for sensor development 11

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

Figure 3. Continued

specific sensor development and validation as they that these changes could be highly valuable in iden-
relate to each indicator reported. One key issue that tifying a calving event. A number of challenges
will be discussed is the practicality of sensor sys- exist, however, that impede the detection of these
tems in extensive grazing environments as commer- changes, particularly in extensive grazing systems.
cially viable solutions. Although some sensors lend The appearance of the amniotic sac of calf marks
themselves to deployment in these environments the initiation of the second stage of parturition and
at reasonable costs, for example, accelerometer ear is an obvious key component of the calving pro-
tag sensors, others, such as image analysis, cannot cess (Saint-Dizier and Chastant-Maillard, 2015).
be easily deployed in large areas and others remain A number of commercially available devices have
too expensive or impractical for whole herd deploy- been produced that utilize magnetic, mechanical, or
ment in a commercial context. New sensor systems, light sensors to identify the separation of the vulval
however, are constantly evolving and potential lips or are expelled prior to the calf, as described by
future developments enabling new deployment Saint-Dizier and Chastant-Maillard (2015). These
strategies and refinements in cost will also be ex- devices are generally expensive and deploying them
plored where relevant. across a whole herd would be prohibitive in a com-
mercial context. Some of these devices will also re-
Behavioral Indicators quire surgical or specialist deployment for accurate
detection and adequate retention and as such, this
Of the behavioral indicators identified, a small is unlikely to be easily adopted by industry (Saint-
proportion were non-continuous and were exclu- Dizier and Chastant-Maillard, 2015; Grodkowski
sively present or absent. Most of these features et al., 2018).
provided very clear and well-defined signals of Although udder and vulva changes have been
impending or occurring parturition. This is not sur- consistently observed across numerous studies,
prising, given that many represent the very obvious the ability to detect these indicators will be diffi-
physical conditions experienced during calving, cult to identify with a sensor. One potential can-
such as the emergence of calf feet. The high level of didate would be using automated image analysis
consistency across the varying references indicate (Ozkaya, 2015). Although only one system would
Translate basic science to industry innovation
12 Chang et al.

be required, achieving reliable collection of im- and subday scale suggest that it may be a key indi-
agery in extensive systems will be challenging. cator for inclusion in algorithms aimed at calving
Cattle must be presented in a consistent manner detection.
that allows for clear analysis, and visualization sys- Calf licking behavior is a commonly reported
tems would need to be deployed in areas that cattle event that is initiated soon after birth and de-
frequently visit. One potential location for deploy- clines exponentially in the hours following calving.
ment could be on a walk-over-weigh system enclos- Grooming behavior is an essential part of the ma-
ing a water trough or supplementation. ternal bonding process, and assists with drying
Cow isolation was commonly reported as being the calf of amniotic fluid and preventing heat
a potential indicator of impending parturition. loss (Hudson and Mullord, 1977; Edwards and
Two sensor systems exist that could be used to iden- Broom, 1982; von Keyserlingk and Weary, 2007).
tify isolation behavior—proximity loggers and GPS At this current time, no formal studies have been
devices. Although proximity loggers measure the conducted that investigate the use of accelerom-

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

duration and frequency of interaction between two eters to identify calf grooming behavior. Despite
devices (Boyland et al., 2013), GPS systems provide this, it is theoretically possible to detect grooming
the animal’s finite position. Both of these sensors behavior using ear- or neck-based accelerometers.
are being pursued as potential commercial solu- Accelerometers detect variations in linear accel-
tions, as they can be integrated within collars and eration along the x-, y-, and z-axes to determine
ear tags (Trotter et al., 2018). distinct behaviors (Barwick, 2016). It is, therefore,
Calving location showed a great deal of vari- plausible to conclude that ear- or neck-based accel-
ation between the studies. The choice of calving lo- erometers could be used to the detect the distinctive
cation is dictated primarily by the production style, drawn out down-up head movement associated
where dairy and intensive beef enterprises often with licking a calf. One potential issue, however,
have more limited calving location options com- is the distinction between grazing and grooming.
pared to extensively grazed beef cattle. The ma- Similarly, grooming could also theoretically be de-
jority of the studies investigating location at calving tected using an audio sensor. Combining an audio
were conducted in a dairy environment and conse- and accelerometer device could also potentially
quently, the results reflect location preference when provide a method to distinguish between grazing
presented with limited options. More research into and grooming. If grooming could be detected using
calving location as an indicator in extensive envi­ sensors, it might be possible to explore other fea-
ronments needs to be undertaken to confirm the tures of parturition, such as maternal investment,
potential value of this indicator. which is known to impact later calf health, and po-
Numerous behavioral indicators expressed a tentially survival (Stěhulová et al., 2013).
pattern of change over time. In order to capture The studies examined showed that the rate of
these indicators, frequent monitoring is required. lying to standing transition peaked at the time of
Traditional methods of calving detection require calving (both at the multiday and subday scales).
experience to recognize the behavioral changes as- This behavior has been frequently measured in
sociated with parturition and is a labor and time the literature, often using leg-based accelerom-
intensive task. eters (Blackie et al., 2006; Chapinal et al., 2010;
Rumination is a key indicator of cow health Calderon and Cook, 2011; Jensen, 2012; Felton
and productivity, and as such, has been extensively et al., 2013; Sepúlveda-Varas et al., 2014; Borchers
explored by the dairy industry (Stangaferro et al., et al., 2015; Titler et al., 2015; Black and Krawczel,
2016; Beauchemin, 2018). Rumination monitoring 2016; Ouellet et al., 2016; Borchers et al., 2017;
has primarily been detected using an accelerom- Neave et al., 2017). It is more difficult, however, to
eter deployed in a collar or ear tag (Bikker et al., distinguish between a lying and standing position
2014; Pahl et al., 2014; Ouellet et al., 2016; Rutten with an ear- or neck-based accelerometer as the
et al., 2017) or an acoustic detection sensor, which two behaviors show similar accelerometer readings.
detects the distinctive sounds of rumination (Adin Leg-based accelerometers, however, show a distinct
et al., 2009; Kovács et al., 2017). Several commer- crossover of two axes and can easily be utilized to
cially available tools have been developed to detect determine a standing or lying position, and con-
rumination in the dairy industry (Reith et al., 2014; sequently, the number of lying bouts performed.
Eslamizad et al., 2018; Reynolds et al., 2019). The More recently, Busch et al. (2017), described the use
gradual decrease in rumination prior to parturition of a neck accelerometer in distinguishing between
and then its increase thereafter at both the multiday the lying and standing position via the detection
Translate basic science to industry innovation
 Calving indicators for sensor development 13

of the transitional movement, that is, a lying bout. physiological changes are often the key drivers of
This behavior is a strong candidate for integration the behavioral indicators previously explored.
into algorithms for detection of parturition. The studies reporting on NEFA and proges-
Dry matter intake, drinking time, and feeding terone concentration changes show a high degree
time decreased in the period prior to calving, be- of repeatability and indicate similar patterns of dis-
fore increasing again. These changes are frequently tinctive changes relative to parturition in different
reported in the literature as being measured, but experimental conditions, with varying breeds, and
exclusively using off-animal technologies, such nutritional regimes. Likewise, the reported results
as feed and water intake monitors (Huzzey et al., for estrogen show high levels of consistency be-
2007; Proudfoot et al., 2009; Chapinal et al., 2010; tween sources. At this current time, biosensors in
Jawor et al., 2012; Neave et al., 2017). Existing animals do not exist at a commercial level to allow
systems utilize an electronic identification tag at- for the detection of these hormones. Biosensors in
tached to the cow’s ear to provide an individual human medicine have been more extensively re-

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

with access to food or water. The system automat- searched compared with in livestock and could
ically measures the initial and final weight of the be adapted for use in animals. One successful and
contents of the bin to deduce how much water or commercially available biosensor in humans is the
food was consumed. To convert the amount of continuous glucose monitoring pump for patients
consumed food into dry matter intake, the forage with diabetes. These devices are commonly inserted
must be manually analyzed. These off-animal into the abdomen and arm and have the capacity to
technologies can only feasibly be utilized in inten- provide alerts when blood glucose falls below the
sive settings, where feed and water are available recommended level (Cappon et al., 2019). Future
exclusively from these troughs. In more extensive development into biosensing technologies capable
settings, it is difficult to control food and water of monitoring blood chemistry and providing alerts
consumption from other sources. Alternatively, in this manner would be a highly valuable asset in
drinking and feeding could potentially be meas- detecting impending calving.
ured using ear- or neck-based accelerometers. Temperature detection systems are a well-es-
A number of studies exist that identify grazing be- tablished technology with numerous commercially
havior from both ear- and neck-based accelerom- available options predominantly focused on the in-
eters (Pereira et al., 2018; Rahman et al., 2018). tensive livestock industries. An extensive list detail-
Another study by Greenwood et al. (2014) posits ing the company and sensor deployment location
the use of an on-animal sensor to quantify actual is available at Sellier et al. (2014). As indicated by
measures of feed intake, however, this concept Sellier et al. (2014), sensor deployment location is a
remains in development. Drinking behavior has key factor influencing the ease and invasiveness of
been measured using a neck-based accelerometer, deployment. Temperature sensors that require sur-
however, has not been compared with grazing be- gical deployment will require specialist training and
havior (Williams et al., 2017). While potentially incur additional time and medical costs. Rumen or
difficult to objectively measure, these indicators reticulum boluses, on the other hand, are less inva-
could be valuable components in a sensor-based sive and more easily deployed, but are influenced
system and warrant further investigation. by the ingestion of feed and water, resulting in tem-
perature fluctuations (Sellier et al., 2014). Similarly,
Physiological Indicators intravaginal deployment is often less invasive, but
require specialist deployment to ensure adequate re-
The vast majority of physiological indicators tention up until the time of calving (Menzies et al.,
were hormonal, which is unsurprising given the 2016). A potential noninvasive method of detecting
role that the endocrine system plays in parturition body temperature is via an infrared camera target-
(Shenavai et al., 2012). Although numerous tech- ing the eye (Scoley et al., 2019). These systems are
nologies exist that have the capacity to measure only available in an off-animal form, which may be
non-hormonal physiological traits, such as heart difficult to install and achieve reliable results in an
rate and temperature, on-animal biosensor devel- extensive environment.
opment for commercial deployment is non-existent Heart rate detection is common in the eques-
at this time. Despite this, it is worth discussing trian sporting industry, where animal stress re-
these indicators as future sensors may be devel- sponse, recovery, and training intensity are of high
oped to measure them and at the very least, these importance (von Lewinski et al., 2013). As such,

Translate basic science to industry innovation

14 Chang et al.

heart rate monitoring devices are more commonly data point at other times. As a result, the trend
designed for horses. Studies measuring heart rate in lines developed for each indicator became skewed
this literature review utilized these equine-designed to a greater or lesser extent. An attempt was made
devices, such as electrode belts (Kovács et al., to minimize this effect by using a rolling average,
2015). This technology, however, is commonly posi- but despite the transformation, bias still exists and
tioned around the chest of the animal and would needs to be considered when interpreting the re-
be unsuitable for deployment in systems where ani- sults. For example, in the temporal graph for feed-
mals are unaccustomed to being handled and for ing duration, there appears to be a peak 1 h prior to
long-term use. Additionally, animals must be pre- calving, however, this apparent peak could instead
pared by shaving excess hair and applying electrode be due to the lack of data at that time point. Only
gel for accurate generation (Lefcourt et al., 1999; one source was available for that time period, influ-
Kovács et al., 2015). Another potential method of encing the trendline upwards. The reliability of this
heart rate detection is an optical heart rate sensor. change is unknown, given the lack of available data.

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

In these systems, light emitting diodes (LED) are The methodology used in this study allocated a
used to detect changes in blood flow to determine “vote” for each experiment that reported a positive,
heart rate (Schubert et al., 2018). The ear could negative, or null trend (Williams et al., 2016). For a
provide a viable location for these sensors. “vote” to be assigned to a trend, a majority of ani-
Although many physiological sensors could mals had to display the behavior or physiological
provide very good indicators of parturition these indicator. A minor proportion of experimental ani-
are generally more difficult to achieve in a sen- mals, however, may not have displayed the calving
sor-based approach targeted at commercial deploy- indicator in the same manner as the remainder of
ment in extensive grazing systems. the herd. For example, in Flörcke and Grandin
(2014), 95.2% of the cows isolated themselves more
Limitations and Future Directions than 50 m from the main herd at the time of calving,
although the remaining 4.8% did not participate in
The majority of studies examined in this lit- this behavior. Therefore, although the majority of
erature review were conducted with dairy cattle, animals displayed the behavior described, others
and almost all experiments utilized B. taurus cattle. did not. This individual variability is a key factor in
Dairy cattle and beef cattle, and B. taurus and producing accurate algorithms for sensor-based de-
B. indicus cattle, have been bred and used for dif- tection of parturition. In the majority of the stud-
ferent purposes and exhibit different physiology ies included in this article, however, this variability
and behaviors. The overrepresentation of dairy has not been described. Instead, particularly where
cattle could influence the perceived intensity of quantifiable data was made available, the average
behavioral changes in relation to calving com- value or behavior is provided. Future research with
pared with beef cattle, leading to over- or under- more in-depth analysis describing individual vari-
exaggerated calving parturition indicators. For ability would be of great benefit in the development
example, dairy cattle have been bred for high milk of sensing systems for precision livestock man-
production and therefore water consumption in agement. Consequently, the described changes in
these animals is far greater compared with beef detecting parturition must be interpreted with cau-
cattle (Australian and New Zealand Environment tion, however, using multiple indicators of calving
and Conservation Council and Agriculture and may increase the accuracy of parturition detection.
Resource Management Council of Australia and As discussed, individual variability exists when
New Zealand, 2000). Although acknowledging this performing calving behaviors. Some animals may
key limitation, there are clearly valuable insights to display more subtle behaviors as they approach
be gained from the current analysis in guiding fu- calving, although other animals may not display a
ture sensor-based developments. described behavior at all. The integration and de-
The majority of studies did not include data tection of multiple calving indicators, however,
that could be used for further analysis. For reli- could provide a viable solution to this issue. Two ap-
able time series graphs to be generated, and a true proaches can be used when integrating indicators:
meta-analyses to be conducted, more data was re- one multifunctional sensor might be used, such as
quired. The data that was available was unevenly an accelerometer; or multiple sensors, requiring
distributed across varying time periods, resulting coordination and/or data integration, can be uti­
in multiple data points at one period and only one lized. There are two possible ways of obtaining

Translate basic science to industry innovation

 Calving indicators for sensor development 15

data suitable for this integrated approach. Firstly, a these vary between individual animals. Thus, com-
single sensor, for example, an accelerometer, might bining several key indicators would result in a more
be used to derive multiple behavioral indicators. accurate detection of time of parturition. Future
Secondly, a number of different sensors might be research into automated biosensing technologies to
applied, such as accelerometer or GPS, from which detect hormone concentrations would further in-
the required data is derived. crease the accuracy of calving detection, given the
Potential behavioral indicator candidates for consistency of results between different breeds, nu-
calving detection using a single accelerometer in- tritional planes, and environmental settings. At the
clude lying bouts, rumination time, and calf groom- current point in time, however, on-animal sensor
ing behavior. Both lying bouts and rumination time technology could be used to automatically detect
have the capacity to be detected by ear- or neck- calving to enhance managerial decisions and maxi-
based accelerometer. Calf grooming behavior mize on-farm efficiency, productivity, and animal
could then be used to verify the conclusion of a welfare.

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

calving event. These behaviors could also be de-
tected in conjunction with isolation behavior, using SUPPLEMENTARY DATA
a GPS device.
When integrating indicators to monitor par- Supplementary data are available at
turition, the number of behavioral thresholds that Translational Animal Science online.
must be met before it is positively identified as a
ACKNOWLEDGMENT
calving event must be considered. When too few
behaviors are included, the threshold for calving The authors acknowledge J.E. Kinder and his
identification might become too low and oversen- editorial contributions and L.B. Ryall.
sitive. Many calving indicators might be mistaken
for diseased states, and as such, it is possible that FUNDING
a model with too few indicators will not detect
actual parturition events (Proudfoot et al., 2014; This work was supported by Central Queensland
Stangaferro et al., 2016). For example, an isolated University, Meat & Livestock Australia, and Telstra
cow with decreased rumination activity could be Corporation Limited. The funding sources had no
sick, instead of calving. Conversely, if too many involvement in the production of this research.
calving indicators are integrated, animals that have Conflict of interest statement. None declared.
more subtle or do not express the calving behaviors
may be overlooked. LITERATURE CITED
All of the described considerations need to Adin, G., R. Solomon, M. Nikbachat, A. Zenou, E. Yosef,
be taken into account in future research investi- A. Brosh, A. Shabtay, S. J. Mabjeesh, I. Halachmi, and
J. Miron. 2009. Effect of feeding cows in early lacta-
gating the development of sensors for parturition
tion with diets differing in roughage-neutral detergent
and dystocia detection on-farm. There is con- fiber content on intake behavior, rumination, and milk
siderable scope for future research in this area, production. J. Dairy Sci. 92:3364–3373. doi:10.3168/
particularly in the context of optimizing systems jds.2009-2078.
across the great diversity of grazing animal pro- Australian and New Zealand Environment and Conservation
duction systems. Council, and Agriculture and Resource Management
Council of Australia and New Zealand. 2000. Australian
and New Zealand guidelines for fresh and marine water
CONCLUSION quality No. 3.
Barrier, A. C., E. Ruelle, M. J. Haskell, and C. M. Dwyer. 2012.
As a cow approaches calving, she undergoes a Effect of a difficult calving on the vigour of the calf, the
number of behavioral and physiological changes. onset of maternal behaviour, and some behavioural indi-
cators of pain in the dam. Prev. Vet. Med. 103:248–256.
Measuring and detecting these changes could help doi:10.1016/j.prevetmed.2011.09.001
in informing an impending or ongoing calving Barwick, J. 2016. On-animal motion sensing using accelerom-
event and could improve production and welfare eters as a tool for monitoring sheep behavior and health
outcomes. The results of the systematic literature status [Ph.D]. University of New England.
review suggest that the automated monitoring and Beauchemin, K. A. 2018. Invited review: current perspectives
on eating and rumination activity in dairy cows. J. Dairy
detection of calving, as well as of dystocia inci-
Sci. 101:4762–4784. doi:10.3168/jds.2017-13706
dents, is possible. The results, however, also indicate Bikker, J. P., H. van Laar, P. Rump, J. Doorenbos, K. van Meurs,
that although there are many important behavioral G. M. Griffioen, and J. Dijkstra. 2014. Technical note:
changes that are associated with calving, many of evaluation of an ear-attached movement sensor to record
Translate basic science to industry innovation
16 Chang et al.

cow feeding behavior and activity. J. Dairy Sci. 97:2974– Greenwood, P. L., P. Valencia, L. Overs, D. R. Paull, and
2979. doi:10.3168/jds.2013-7560 I. W. Purvis. 2014. New ways of measuring intake, effi-
Black, R. A., and P. D. Krawczel. 2016. A case study of behavior ciency, and behavior of grazing livestock. Anim. Prod.
and performance of confined or pastured cows during the Sci. 54:1796–1804. doi:10.1071/AN14409
dry period. Animals 6(7):41. doi:10.3390/ani6070041 Grodkowski, G., T. Sakowski, K. Puppel, and T. Baars. 2018.
Blackie, N., J. R. Scaife, and E. C. L. Bleach. 2006. Lying Comparison of different applications of automatic herd
behavior and activity of early lactation Holstein dairy control systems on dairy farms—a review. J. Sci. Food
cattle measured using an activity monitor. Cattle Pract. Agric. 98:5181–5188. doi:10.1002/jsfa.9194
14(2):139–142. doi:10.3168/jds.2018-16045 Grum, D. E., J. K. Drackley, R. S. Younker, D. W. LaCount,
Borchers, M. R., Y. M. Chang, K. L. Proudfoot, and J. J. Veenhuizen. 1996. Nutrition during the dry
B. A. Wadsworth, A. E. Stone, and J. M. Bewley. 2017. period and hepatic lipid metabolism of periparturient
Machine-learning-based calving prediction from activity, dairy cows. J. Dairy Sci. 79:1850–1864. doi:10.3168/jds.
lying, and ruminating behaviors in dairy cattle. J. Dairy S0022-0302(96)76553-0
Sci. 100:5664–5674. doi:10.3168/jds.2016-11526 Hudson, S. J., and M. M. Mullord. 1977. Investigations of ma-
Borchers, M. R., Y. M. Chang, A. E. Stone, B. A. Wadsworth, ternal bonding in dairy cattle. Appl. Anim. Ethol. 3:271–
and J. M. Bewley. 2015. Predicting impending calving 276. doi:10.1016/0304-3762(77)90008-6

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

using automatic activity and rumination measures in dairy Huzzey, J. M., D. M. Veira, D. M. Weary, and
cattle 7th European Conference on Precision Livestock M. A. von Keyserlingk. 2007. Prepartum behavior and
Farming; Milan, Italy; p. 406–414. dry matter intake identify dairy cows at risk for metritis. J.
Boyland, N. K., R. James, D. T. Mlynski, J. R. Madden, and Dairy Sci. 90:3220–3233. doi:10.3168/jds.2006-807
D. P. Croft. 2013. Spatial proximity loggers for recording Jawor, P. E., J. M. Huzzey, S. J. LeBlanc, and
animal social networks: consequences of inter-logger vari- M. A. von Keyserlingk. 2012. Associations of subclinical
ation in performance. Behav. Ecol. Sociobiol. 67(11):1877– hypocalcemia at calving with milk yield, and feeding, drinking,
1890. doi:110.1007/s00265-013-1622-6 and standing behaviors around parturition in Holstein cows.
Busch, P., H. Ewald, and F. Stüpmann. 2017. Determination of J. Dairy Sci. 95:1240–1248. doi:10.3168/jds.2011-4586
standing-time of dairy cows using 3D-accelerometer data Jensen, M. B. 2012. Behavior around the time of calving in
from collars. In: 2017 Eleventh International Conference dairy cows. Appl. Anim. Behav. Sci. 139(3–4):195–202.
on Sensing Technology, Sydney, Australia. von Keyserlingk, M. A., and D. M. Weary. 2007. Maternal be-
Calderon, D. F., and N. B. Cook. 2011. The effect of lameness havior in cattle. Horm. Behav. 52:106–113. doi:10.1016/j.
on the resting behavior and metabolic status of dairy cattle yhbeh.2007.03.015
during the transition period in a freestall-housed dairy herd. Kovács, L., F. L. Kézér, F. Ruff, and O. Szenci. 2017.
J. Dairy Sci. 94:2883–2894. doi:10.3168/jds.2010-3855 Rumination time and reticuloruminal temperature as pos-
Cappon, G., M. Vettoretti, G. Sparacino, and A. Facchinetti. sible predictors of dystocia in dairy cows. J. Dairy Sci.
2019. Continuous glucose monitoring sensors for diabetes 100:1568–1579. doi:10.3168/jds.2016-11884
management: a review of technologies and applications. Kovács, L., J. Tőzsér, F. L. Kézér, F. Ruff, M. Aubin-Wodala,
Diabetes Metab. J. 43:383–397. doi:10.4093/dmj.2019.0121 E. Albert, A. Choukeir, Z. Szelényi, and O. Szenci. 2015.
Chapinal, N., C. Goldhawk, A. M. de Passillé, Heart rate and heart rate variability in multiparous
M. A. G. von Keyserlingk, D. M. Weary, and J. Rushen. dairy cows with unassisted calvings in the periparturi-
2010. Overnight access to pasture does not reduce milk ent period. Physiol. Behav. 139:281–289. doi:10.1016/j.
production or feed intake in dairy cattle. Livest. Sci. physbeh.2014.11.039
129(1–3):104–110. doi:10.1016/j.livsci.2010.01.011 Laporta, J., A. L. Astessiano, C. López-Mazz, P. Soca,
Edwards, S. A., and D. M. Broom. 1982. Behavioral inter- A. C. Espasandin, and M. Carriquiry. 2014. Effects
actions of dairy cows with their newborn calves and the of herbage allowance of native grasslands in pure-
effects of parity. Anim. Behav. 30:525–535. doi:10.1016/ bred and crossbred beef cows: metabolic, endocrine
S0003-3472(82)80065-1 and hepatic gene expression profiles through the gesta-
Eslamizad, M., L. M. Tümmler, M. Derno, M. Hoch, and tion-lactation cycle. Animal 8:1119–1129. doi:10.1017/
B. Kuhla. 2018. Technical note: development of a pres- S1751731114000986
sure sensor-based system for measuring rumination time Lefcourt, A. M., B. Erez, M. A. Varner, R. Barfield, and
in pre-weaned dairy calves. J. Anim. Sci. 96:4483–4489. U. Tasch. 1999. A noninvasive radiotelemetry system
doi:10.1093/jas/sky337 to monitor heart rate for assessing stress responses of
Felton, C. A., M. G. Colazo, C. J. Bench, and D. J. Ambrose. bovines. J. Dairy Sci. 82:1179–1187. doi:10.3168/jds.
2013. Large variations exist in prepartum activity among S0022-0302(99)75341-5
dairy cows continuously housed in a tie-stall barn. Can. von Lewinski, M., S. Biau, R. Erber, N. Ille, J. Aurich,
J. Anim. Sci. 93(4):435–444. doi:10.4141/cjas2013-069 J.-M. Faure, E. Möstl, and C. Aurich. 2013. Cortisol
Flörcke, C., and T. Grandin. 2014. Separation behavior for release, heart rate, and heart rate variability in the
parturition of Red Angus beef cows. Open J. Anim. Sci. horse and its rider: different responses to training and
4(2):43–50. doi:10.4236/ojas.2014.42007 performance. Vet. J. 197(2):229–232. doi:10.1016/j.
Fogarty, E. S., D. L. Swain, G. Cronin, and M. Trotter. 2018. tvjl.2012.12.025
Autonomous on-animal sensors in sheep research: a sys- Lidfors, L. M. 1996. Behavioral effects of separating the dairy
tematic review. Comput. Electron. Agric. 150:245–256. calf immediately or 4 days post-partum. Appl. Anim.
doi:10.1016/j.compag.2018.04.017 Behav. Sci. 49:269–283. doi:10.1016/0168-1591(96)01053-2
Grant, M. J., and A. Booth. 2009. A typology of reviews: an analysis Lidfors, L., and P. Jensen. 1988. Behavior of free-ranging beef
of 14 review types and associated methodologies. Health Info. cows and calves. Appl. Anim. Behav. Sci. 20:237–247.
Libr. J. 26:91–108. doi:10.1111/j.1471-1842.2009.00848.x doi:10.1016/0168-1591(88)90049-4
Translate basic science to industry innovation
 Calving indicators for sensor development 17

Lidfors, L. M., D. Moran, J. Jung, P. Jensen, and H. Castren. Reith, S., H. Brandt, and S. Hoy. 2014. Simultaneous ana-
1994. Behavior at calving and choice of calving place in lysis of activity and rumination time, based on col-
cattle kept in different environments. Appl. Anim. Behav. lar-mounted sensor technology, of dairy cows over the
Sci. 42:11–28. doi:10.1016/0168-1591(94)90003-5 peri-estrus period. Livest. Sci. 170:219–227. doi:10.1016/j.
Mee, J. F. 2008. Prevalence and risk factors for dystocia in livsci.2014.10.013
dairy cattle: a review. Vet. J. 176:93–101. doi:10.1016/j. Reynolds, M. A., M. R. Borchers, J. A. Davidson,
tvjl.2007.12.032 C. M. Bradley, and J. M. Bewley. 2019. Technical note: an
Meijering, A. 1984. Dystocia and stillbirth in cattle—a review evaluation of technology-recorded rumination and feed-
of causes, relations, and implications. Livest. Prod. Sci. ing behaviors in dairy heifers. J. Dairy Sci. 102:6555–6558.
11(2):143–177. doi:10.1016/0301-6226(84)90057-5 doi:10.3168/jds.2018-15635
Menzies, D., K. P. Patison, N. J. Corbet, and D. L. Swain. Riley, D. G., C. C. Chase, Jr, T. A. Olson, S. W. Coleman, and
2016. Can an automated radiolocation device ascer- A. C. Hammond. 2004. Genetic and nongenetic influ-
tain date and location of calving in extensive beef sys- ences on vigor at birth and preweaning mortality of pure-
tems? RUN Regional Futures Conference, CQUniversity, bred and high percentage Brahman calves. J. Anim. Sci.
Rockhampton, Australia. 82:1581–1588. doi:10.2527/2004.8261581x
Miedema, H. M., M. S. Cockram, C. M. Dwyer, and Rørvang, M. V., M. S. Herskin, and M. B. Jensen. 2018. The

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

A. I. Macrae. 2011a. Behavioral predictors of the start motivation-based calving facility: social and cognitive
of normal and dystocic calving in dairy cows and heifers. factors influence isolation seeking behaviour of Holstein
Appl. Anim. Behav. Sci. 132(1–2):14–19. doi:10.1016/j. dairy cows at calving. Plos One 13:e0191128. doi:10.1371/
applanim.2011.03.003 journal.pone.0191128
Miedema, H. M., M. S. Cockram, C. M. Dwyer, and Rørvang, M. V., B. L. Nielsen, M. S. Herskin, and M. B. Jensen.
A. I. Macrae. 2011b. Changes in the behavior of dairy 2017. Short communication: calving site selection of
cows during the 24h before normal calving compared with multiparous, group-housed dairy cows is influenced by
behavior during late pregnancy. Appl. Anim. Behav. Sci. site of a previous calving. J. Dairy Sci. 100:1467–1471.
131(1–2):8–14. doi:10.1016/J.APPLANIM.2011.01.012 doi:10.3168/jds.2016-11681
Neave, H. W., J. Lomb, M. A. G. von Keyserlingk, A. Behnam- Rutten, C. J., C. Kamphuis, H. Hogeveen, K. Huijps,
Shabahang, and D. M. Weary. 2017. Parity differences M. Nielen, and W. Steeneveld. 2017. Sensor data on
in the behavior of transition dairy cows. J. Dairy Sci. cow activity, rumination, and ear temperature im-
100:548–561. doi:10.3168/jds.2016-10987 prove prediction of the start of calving in dairy cows.
Ouellet, V., E. Vasseur, W. Heuwieser, O. Burfeind, X. Maldague, Comput. Electron. Agric. 132:108–118. doi:10.1016/j.
and É. Charbonneau. 2016. Evaluation of calving indica- compag.2016.11.009
tors measured by automated monitoring devices to pre- Saint-Dizier, M., and S. Chastant-Maillard. 2015. Methods
dict the onset of calving in Holstein dairy cows. J. Dairy and on-farm devices to predict calving time in cattle. Vet.
Sci. 99:1539–1548. doi:10.3168/jds.2015-10057 J. 205:349–356. doi:10.1016/j.tvjl.2015.05.006
Owens, J. L., and T. N. Edey. 1985. Parturient behavior and Schubert, M. M., A. Clark, and A. B. De La Rosa. 2018. The
calf survival in a herd selected for twinning. Appl. Anim. Polar® OH1 optical heart rate sensor is valid during mod-
Behav. Sci. 13:321–333. doi:10.1016/0168-1591(85)90012-7 erate-vigorous exercise. Sports Med. Int. Open 2:E67–
Ozkaya, S. 2015. Prediction possibility of milk yield from E70. doi:10.1055/a-0631-0920
udder measurements using digital image analysis on Schuenemann, G. M., I. Nieto, S. Bas, K. N. Galvão, and
Holstein cows. Indian J. Anim. Res. 49(3):388–391. J. Workman. 2011. Assessment of calving progress and
doi:10.5958/0976-0555.2015.00050.3 reference times for obstetric intervention during dys-
Pahl, C., E. Hartung, A. Grothmann, K. Mahlkow-Nerge, and tocia in Holstein dairy cows. J. Dairy Sci. 94:5494–5501.
A. Haeussermann. 2014. Rumination activity of dairy doi:10.3168/jds.2011-4436
cows in the 24 hours before and after calving. J. Dairy Sci. Scoley, G. E., A. W. Gordon, and S. J. Morrison. 2019. Use of
97:6935–6941. doi:10.3168/jds.2014-8194 thermal imaging in dairy calves: exploring the repeatability
Pereira, G. M., B. J. Heins, and M. I. Endres. 2018. Technical and accuracy of measures taken from different anatom-
note: validation of an ear-tag accelerometer sensor to ical regions. Transl. Anim. Sci. 3:564–576. doi:10.1093/tas/
determine rumination, eating, and activity behaviors txy126
of grazing dairy cattle. J. Dairy Sci. 101:2492–2495. Sellier, N., E. Guettier, and C. Staub. 2014. A review of meth-
doi:10.3168/jds.2016-12534 ods to measure animal body temperature in precision
Proudfoot, K. L., J. M. Huzzey, and M. A. von Keyserlingk. farming. Am. J. Agr. Sci. Technol. 2(2):74–99. doi:10.7726/
2009. The effect of dystocia on the dry matter intake and ajast.2014.1008
behavior of Holstein cows. J. Dairy Sci. 92:4937–4944. Selman, I. E., A. D. McEwan, and E. W. Fisher. 1970. Studies
doi:10.3168/jds.2009-2135 on natural suckling in cattle during the first eight hours
Proudfoot, K. L., M. B. Jensen, D. M. Weary, and postpartum I. behavioral studies (dams). Anim. Behav.
M. A. von Keyserlingk. 2014. Dairy cows seek isola- 18:276–283. doi:10.4141/cjas90-047
tion at calving and when ill. J. Dairy Sci. 97:2731–2739. Sepúlveda-Varas, P., D. M. Weary, and M. A. von Keyserlingk.
doi:10.3168/jds.2013-7274 2014. Lying behavior and postpartum health status
Rahman, A., D. V. Smith, B. Little, A. B. Ingham, in grazing dairy cows. J. Dairy Sci. 97:6334–6343.
P. L. Greenwood, and G. J. Bishop-Hurley. 2018. Cattle doi:10.3168/jds.2014-8357
behavior classification from collar, halter, and ear tag Shenavai, S., S. Preissing, B. Hoffmann, M. Dilly, C. Pfarrer,
sensors. Inf. Process. Agric. 5(1):124–133. doi:10.1016/j. G. R. Özalp, C. Caliskan, K. Seyrek-Intas, and
inpa.2017.10.001 G. Schuler. 2012. Investigations into the mechanisms

Translate basic science to industry innovation

18 Chang et al.

controlling parturition in cattle. Reproduction 144:279– Trotter, M., A. Cosby, J. Manning, M. Thomson, T. Trotter,
292. doi:10.1530/REP-11-0471 P. Graz, E. S. Fogarty, A. Lobb, and A. Smart. 2018.
Stangaferro, M. L., R. Wijma, L. S. Caixeta, M. A. Al-Abri, Demonstrating the value of animal location and behavior
and J. O. Giordano. 2016. Use of rumination and activity data in the red meat value chain, Meat & Livestock Australia.
monitoring for the identification of dairy cows with health Trotter, M. G., D. W. Lamb, G. N. Hinch, and C. N. Guppy.
disorders: Part I. Metabolic and digestive disorders. J. 2010. Global navigation satellite system livestock tracking:
Dairy Sci. 99:7395–7410. doi:10.3168/jds.2016-10907 system development and data interpretation. Anim. Prod.
Stěhulová, I., M. Špinka, R. Šárová, L. Máchová, R. Kněz, Sci. 50:616–623. doi:10.1071/AN09203
and P. Firla. 2013. Maternal behavior in beef cows is in- Wehrend, A., E. Hofmann, K. Failing, and H. Bostedt. 2006.
dividually consistent and sensitive to cow body condition, Behavior during the first stage of labour in cattle: in-
calf sex, and weight. Appl. Anim. Behav. Sci. 144(3–4):89– fluence of parity and dystocia. Appl. Anim. Behav. Sci.
97. doi:10.1016/j.applanim.2013.01.003 100(3–4):164–170. doi:10.1016/j.applanim.2005.11.008
Titler, M., M. G. Maquivar, S. Bas, P. J. Rajala-Schultz, Williams, L. R., G. J. Bishop-Hurley, A. E. Anderson, and
E. Gordon, K. McCullough, P. Federico, and D. L. Swain. 2017. Application of accelerometers to re-
G. M. Schuenemann. 2015. Prediction of parturition in cord drinking behavior of beef cattle. Anim. Prod. Sci.
Holstein dairy cattle using electronic data loggers. J. Dairy 59(1):122–132. doi:10.1071/AN17052

Downloaded from https://academic.oup.com/tas/article/4/3/txaa155/5896568 by guest on 18 June 2025

Sci. 98:5304–5312. doi:10.3168/jds.2014-9223 Williams, L. R., E. L. Jackson, G. J. Bishop-Hurley, and
Trotter, M. G. 2010. Precision agriculture for pasture, range- D. L. Swain. 2016. Drinking frequency effects on the
land, and livestock systems. In: Food Security from performance of cattle: a systematic review. J. Anim.
Sustainable Agriculture 15th Australian Agronomy Physiol. Anim. Nutr. (Berl). 101:1076–1092. doi:10.1111/
Conference, Lincoln, New Zealand. jpn.12640

Translate basic science to industry innovation