Effects of Shock Collars 2
E. Schalkea, , , J. Stichnotha, S. Otta and R. Jones-Baadeb
aDepartment of Animal Welfare and Behaviour, Veterinary School of Hannover,
Buenteweg 2, 30559 Hannover, Germany
bClemensstrasse 123, 80796 Muenchen, Germany
Available online 11 December 2006.
Abstract
The use of electric shock collars for training dogs is the subject of considerable
controversy. Supporters claim that they are a reliable means of eliminating self-rewarding
behaviour and that they can be used over greater distances and with less risk of stress and
injury than mechanical devices, such as choke chains. Opponents cite the risk of incorrect
or abusive use and temptation to use electric training collars without thought or time
given to alternative training methods, regardless of the fact that their use may be
associated with pain and fear. The aim of this study was to investigate whether any stress
is caused by the use of electric shock collars or not and in this way to contribute to their
evaluation with respect to animal welfare.
Fourteen laboratory-bred Beagles were used to ensure standardised breeding, raising and
training. Heart rate and saliva cortisol were used as stress parameters. The research
project lasted for 7 months, during which each dog was trained for 1.5 h per day. To
exclude circadian deviations of salivary cortisol values, each individual was allocated a
rigid timeslot. Training as well as the experiments themselves were conducted in a
secluded storage building to exclude the influence of external stressors.
Three experimental groups were used. Group A (Aversion) received the electric shock
when the dogs touched the prey—a rabbit dummy fixed to a motion device. Group H
(Here) received the electric shock when they did not obey a previously trained recall
command during hunting. Animals of group R (Random) received the electric shock
arbitrarily, i.e. the shock was administered unpredictably and out of context.
The main experiment lasted for 17 days. All animals were allowed to hunt unimpeded for
the first 5 days. For the next 5 days the dogs were stopped from hunting by a leash. Every
day, the stress parameters were determined. These values were compared with the values
that were obtained during the use of the electric training collars. The collars were used
over a period of 7 days as described previously. After 4 weeks the dogs were brought back
into the research area without receiving an electric pulse.
Group A did not show a significant rise in salivary cortisol levels, while group R and group
H did show a significant rise. When the animals were reintroduced to the research area
after 4 weeks, the results remained the same.
This led to the conclusion that animals, which were able to clearly associate the electric
stimulus with their action, i.e. touching the prey, and consequently were able to predict
and control the stressor, did not show considerable or persistent stress indicators.
Keywords: Dog; Training; Stress; Welfare; Electric training collars
Article Outline
1. Introduction
2. Materials and methods
2.1. Animals
2.2. Test area
2.3. Equipment and sampling
2.3.1. Electronic training collar
2.3.2. Heart rate measurement
2.3.3. Salivary cortisol measurement
2.4. Procedure
2.4.1. Adaptation phase
2.4.2. Main experiment
2.4.2.1. Base levels
2.4.2.2. Preliminary test
2.4.2.3. Main test
2.4.2.4. Post-test
2.5. Data processing
2.5.1. Cortisol level
2.5.2. Heart rate
2.5.2.1. Experiment with electric pulses
2.6. Statistical analysis
3. Results
3.1. Establishing reference values
3.2. Preliminary tests without application of electric pulses
3.2.1. Main tests with application of electric pulses
3.2.1.1. Comparison of salivary cortisol values
3.2.1.2. Dataset “r1–r3, n1”
3.2.1.3. Datasets “r1–r2, n1–n3”, “r1, n1–n3” and “r1–r2, n1”
3.2.1.4. Dataset “r1, n1”
3.2.2. Comparison of heart rate values
3.2.2.1. Datasets “r1–r2, n1”, “r1, n1” and “r1–r3, n1”
3.2.2.2. Dataset “r1–r2, n1–n3”
3.2.2.3. Dataset “r1, n1–n3”
3.3. Post-tests
3.3.1. Comparison of saliva cortisol values
3.3.2. Comparison of heart rate values
3.3.3. Comparison between, preliminary test, main test, and post-test
4. Discussion
5. Conclusion
Acknowledgements
References
1. Introduction
Dogs frequently exhibit undesirable behaviour, which their owners would like to
eliminate. Methods used to achieve this include a change of voice, training the dog to stop
the behaviour to a specific signal, frightening the dog with loud noises, and electronic
training (shock) collars. In Germany, these collars are sometimes applied in police dog
training as well as in training hunting dogs and dogs for competitive sport. Electronic
collars are also used in the training of pet dogs to stop unwanted hunting behaviour for
example. However, the use of these collars in dog training is a matter of considerable
controversy.
Supporters of the collars see them as an effective means of reliably eliminating self-
rewarding behaviour, such as unwanted hunting behaviour. According to Christiansen et
al. (2001), the use of such collars on dogs is an efficient way of preventing dogs from
chasing or attacking grazing sheep. In addition to this, supporters claim that the use of
these devices means less strain for dogs than the use of any other, mechanically operated
means of education such as, for example, choke chains. Klein (2003) states that some
authors stress the advantage of these instruments, claiming that their application does not
cause any physical injury, whereas mechanically operated equipment and inappropriate
training techniques can result in small injuries such as bruises or even severe damage such
as a ruptured spleen. Another advantage of devices like electronic collars according to
their supporters is the ability to use them over greater distances. Weick (1976) even goes
as far as to call their invention the fulfilment of a dream in dog training: contiguousness of
unwanted behaviour and punishment.
Adversaries of electronic collars see a high risk of incorrect and abusive use since it is
tempting to apply these easy-to-use devices without spending much thought on alternative
training methods. They believe that, particularly in situations where they are applied to
achieve quick success in dog sport, their use is at the dog's expense and should be
prohibited. Schilder and van der Borg (2003) suggest banning these instruments from dog
sport completely. As an alternative they suggest that trainers and handlers should study
learning theory more thoroughly and reconsider the structure of their training.
Furthermore, their findings demonstrate that, apart from causing discomfort for the
animal being trained, the use of electronic collars is also associated with pain and fear,
when an electric shock is delivered. The parameter these two authors used for their
evaluation was the dogs’ body language. In 1998, Beerda et al. detected a correlation
between a very low body posture and salivary cortisol level. Salivary cortisol levels and
heart rate can be measured simultaneously to improve the interpretation of behavioural
data with regard to stress since together, they provide an indication of the activity of two
physiological systems that respond to acute stress in dogs, the sympathetic nervous
system and the hypothalamic pituitary adrenal (HPA) axis (Stichnoth, 2002).
The aim of the current study was to investigate the intensity of stress signs arising from
the use of electronic training collars. The level of stress was determined by measuring and
analysing the levels of salivary cortisol and heart rate.
In this study situations that often occur in dog training were reconstructed under
experimental conditions:
1. The elimination of hunting behaviour.
2. Punishment of a dog disobeying a verbal recall signal from a prey.
3. The application of electric shocks in a form that prevents the dog from associating the
stimulus either with its behaviour or a verbal signal, thus simulating inappropriate use by
the owner.
2. Materials and methods
2.1. Animals
In order to minimise variability among the dogs regarding their past experience, all dogs
in the study were chosen according to their hunting behaviour from Beagles from the
laboratory animal breeding stock of 25 animals from the “Tierfarm Kirchheimer Muehle”
of the University of Heidelberg. These dogs lived in groups of five or six, separated
according to gender. The kennels each measured 70 m2 and included a freely available
heated room. The dogs were fed commercial dry food twice daily, in the morning and in
the evening, and had water at their free disposal.
Five female and nine male Beagles, aged between 1.5 and 2 years took part in the research
project. Before starting any experiment, all dogs were examined by a veterinarian and
were found to be healthy. The veterinary examination included a general clinical
examination as well as a screen for salivary cortisol (see below). Prior to the start of the
study, the dogs only had contact with humans during the daily feeding and grooming
routines. They were not accustomed to being separated from their kennel mates.
2.2. Test area
All tests were carried out in a room on the second floor of a building located on the
“Tierfarm Kirchheimer Muehle”, accessible by some stairs. The building was situated at a
distance of about 120 m from the kennels. The room was 11.10 m long and 5.20 m wide
with a height of 2.50 m. It had one door, two radiators, two windows, two shelves and a
sink.
For the study, the dogs were removed one by one from their kennels and put into separate
transport kennels by the laboratory technician and researcher. The transport kennels
were brought to the test building by tractor and unloaded in front of it. To minimise stress
on the dogs concerned, dogs that lived together were taken to the test area together.
The dogs were taken individually from their transport kennel, put on a leash, and led
upstairs into the test area. Immediately after entering the room, the dog was equipped
with a belt for measuring the heart rate, and either an electronic training collar or a
dummy collar.
2.3. Equipment and sampling
2.3.1. Electronic training collar
The electronic training collar used in this research was a “Teletakt micro 3000”, produced
by Schecker GmbH & Co. The complete device consisted of a transmitter, a collar with a
receiver, and an identical dummy collar with a fake receiver. It had a maximum range of
500–600 m. The transmitter had a button to trigger the electric current, and could be run
at levels “0” (device switched off), and “1” (weakest impulse) up to “5” (strongest impulse).
The collar with the receiver had two electrodes on its inside, which had to be in close
contact with the dog's skin. According to the manufacturer of the device, the electric
impulse had a duration of less than 1 ms. Before using the collar on a dog the current,
voltage, and duration of the impulse were measured. Since these values are dependent on
the skin resistance, the devices were tested using resistances between 500 Ω and 2.2 kΩ,
to simulate skin resistance. When operating the device at level 5 the results were as
follows:
Resistance (Ω)
--------------------------------------------------------------------------------
Peak voltage (V)
--------------------------------------------------------------------------------
Peak current (A)
--------------------------------------------------------------------------------
500 700 1.25
2200 1760 0.82
In this investigation the device was run at level “5” in all of the experiments. This level was
chosen in order to investigate the dogs’ reactions under the highest electric pulse and as
such the worst condition possible.
2.3.2. Heart rate measurement
Measurement of heart rate was carried out with a Polar® Horse Trainer Transmitter and a
Polar® Vantage NVTM heart rate measuring instrument for horses. Preliminary studies
carried out at the Department for Animal Welfare and Behaviour had proved both devices
to be a reliable means for the heart rate measurement. The data was read into the Polar
Precision Performance™ programme (Polar® heart rate monitors, Baumann & Haldi SA)
via a serial interface. After activation, the receiver averaged the heart rate and recorded a
value in beats per minute every 5 s. Since the receiver included a stopwatch-function and
a second stopwatch was run simultaneously, the events occurring in the dog's
environment could be correlated with its heart rate. The dogs had to wear the transmitter
from the first day of the adaptation phase until the end of the post-test.
2.3.3. Salivary cortisol measurement
A small amount of citric acid was put into the dog's mouths to stimulate the secretion of
saliva. Afterwards, the saliva sample was taken from the dog's cheek pouches with a cotton
bud made by Salivette® Systems (Sarstedt AG & Co). Directly following this, the
Salivettes were centrifuged at a speed of 4000 rpm and cooled down to a temperature of
−20 °C. The measurement of the salivary cortisol values was carried out by the laboratory
of the University of Heidelberg. The laboratory staff carrying out the salivary cortisol
measurement were blinded as to which of the three groups of dogs a sample belonged. For
the measurement, a radio-immuno-assay with Tritium-marked cortisol and
autoantibodies was used. The Intraassay variance was 5%, the Interassay variance 10%,
and the lowest detection limit was at 0.1 ng/ml. The repeatability rate was >95%.
2.4. Procedure
The research project lasted from September 2000 to March 2001. The experiment was
carried out in two phases—adaptation: familiarisation with the environment, main
experiment: hunting behaviour
For the experiment, the dogs were divided into three groups:
Group A (“Aversion”) Consisted of three male and two female dogs that received an
electric pulse at precisely the moment they touched the prey, thus forming an association
between touching the prey and the electric stimulus
Group H (“Here”) Consisted of two male and two female dogs that received an electric
pulse when they did not obey a recall during hunting, having been previously trained to
recall during the adaptation phase
Group R (“Random”) Consisted of four male dogs and one female dog that received an
electric pulse arbitrarily, meaning that the stimulus was administered unpredictably and
out of context, either prior to orientation towards the prey, or while hunting, or after
having finished the hunting process when there was no prey in the room anymore. The
decision at which moment the electric pulse had to be administered was made by drawing
lots
2.4.1. Adaptation phase
All of the dogs initially experienced an adaptation phase that lasted 3 months. During this
period, the dogs were accustomed to the daily routine, the room in which the tests took
place, the persons carrying out the tests as well as the absence of their kennel mates.
Every one of the dogs was trained to hunt a rabbit dummy. Dogs belonging to group H
were additionally trained with the verbal signal “Here” as a recall signal. The adaptation
phase was concluded when every single dog was able to move around inside the testing
room without displaying signs of fear and stress, and showed clear forward movement on
the leash on its way to the room. Each dog was worked once a day for 1.5 h within a rigid
timeslot.
2.4.2. Main experiment
2.4.2.1. Base levels
In order to determine the baseline levels for salivary cortisol and heart rate, every dog
was led into the room on its own. It spent 50 min inside the room without anyone paying
attention to it. After these 50 min, continuous heart rate measurement was started and
saliva samples were taken every 5 min.
2.4.2.2. Preliminary test
The preliminary test was subdivided into two periods of 5 days each. During the first
period called “Simple Hunting”, each animal was allowed to hunt unimpeded. The hunting
sequence lasted for 1–2 min, and the dog was allowed to take hold of the prey and carry it
off. Ten minutes after the end of the hunting sequence, five saliva samples were taken at
intervals of 5 min. The heart rate was measured throughout the whole period. During the
second period called “Hunting Impeded”, the dogs were prevented from hunting by using
a leash. The prey was presented to the dogs, but hunting was made impossible by using a
1.50 m leash. The “Hunting Impeded” period was terminated after 2 min by removing the
prey. Saliva sampling and heart rate measurement were carried out in exactly the same
way and timing.
The preliminary test made sure that every dog was also examined in a situation in which it
had not been given an electric pulse, in this way, each dog could be used as its own control.
2.4.2.3. Main test
During the main test, electric pulses were administered to the dogs in accordance with the
group to which they belonged. Each dog was allowed a maximum of one electric pulse per
day. The main test was terminated for each individual dog as soon as it either did not show
any interest in the prey on three successive days, or obeyed the recall signal, or displayed
distinct signs of stress in the experimental environment, or after the third application of
an electric pulse. The heart rate was continually measured the whole time. The five saliva
samples were taken at intervals of 5 min; the first was taken 10 min after the application of
the electric pulse.
2.4.2.4. Post-test
For 4 weeks after the main test, the dogs did not have any contact with either the
experimental environment or the person conducting the experiments. At the end of the 4
weeks, they were again taken into the experimental environment, the five saliva samples
were taken, and the heart rate was measured.
The study was approved by the animal welfare officer of the Veterinary School of
Hanover and by the Governmental authorities according to the German Animal Welfare
Act.
2.5. Data processing
2.5.1. Cortisol level
The following key values were derived from the results of the cortisol measurement:
1. The cortisol values gained directly from the saliva samples were called “Absolute
Cortisol Values”.
2. In order to minimise influence arising from each dog's individual cortisol level, all
cortisol values were divided by the average of the cortisol values measured on the 5 days
of the period of “Simple Hunting”. The values gained in this way were called “Relative
Cortisol Values”. The values of the “Simple Hunting” periods were deliberately chosen as
reference values because they were lower than the baseline levels.
2.5.2. Heart rate
The data collected during the heart rate measurement period were plotted both for the
“Simple Hunting” and “Hunting Impeded” periods of the preliminary test as well as for the
main test. As each curve contained some fluctuation, the original curves were divided into
3 min segments. The average was calculated for each of these segments, and this produced
a new, smooth curve. This curve was then used to calculate the time necessary for the
heart rate to return to its normal value.
The following key values were derived from the results of the heart rate measurements:
1. The maximum value of the heart rate at the point the stressor (electric pulse) occurred
was determined. For each curve, there was one maximum value called “Max”.
2. In the period that began 15 min after the stressor had occurred and ended 30 min after
the stressor had occurred, the heart rate was measured. Of the values determined an
average value was formed. For each curve, there was one value called “Mw15”.
3. The ratio between “Max” and “Mw15” was formed in order to balance possible
individual heart rate levels. For each curve, there was one “Max/Mw15”.
4. The period of time between “Max” and “Mw15” was determined. For each curve, there
was one “Max–Mw15”.
During the measurement of baseline levels and during the post-test, possible stressors
(electric current or prey) were not presented. Here, the average of the entire curve was
calculated.
2.5.2.1. Experiment with electric pulses
The application of the electric current had a number of possible behavioural
consequences:
1. The dog stopped hunting after the first application and could not be stimulated to hunt
in the following trials.
2. The dog stopped hunting after the second application and could not be stimulated to
hunt in the following trials.
3. The dog stopped hunting after the third application and could not be stimulated to hunt
in the following trials.
In order to be able to compare the collected data, datasets were compiled according to the
number of days that electric pulses were administered and those without. This resulted in
the following five datasets:
“r1–r3, n1” Three days of application of the electric pulse, and the first day without the
stimulus. This set included data from two dogs of group R and from one dog of group A
“r1–r2, n1–n3” Day 1 and 2 with application of electric pulse, and day 1, 2 and 3 without
the stimulus. This set included data from three dogs of each group
“r1–r2, n1” Day 1 and 2 with application of electric pulse and day 1 without. This dataset
included data from four dogs of group A, three dogs of group H, and five dogs from group
R.
“r1, n1–n3” Day 1 with application of the electric pulse, and day 1, 2 and 3 without. This
dataset included data from four dogs from group A, four dogs from group H, and three
dogs from group R
“r1, n1” Day 1 with application of the electric pulse, and day 1 without. Data of all of the 14
dogs were compiled
2.6. Statistical analysis
Statistical analysis was carried out using the statistics programme SIGMASTAT® as well
as EXCEL 97. A significance level of p < 0.05 was used for all tests. When comparing
results of two dogs, a t-test for unpaired comparison was carried out. In cases where the
data were not normally distributed, the Mann–Whitney U-test was applied. When
comparing results of more than two dogs, a one-way ANOVA was conducted. In cases
where the data were not normally distributed, the Kruskal–Wallis H-test was performed.
When comparing a single dog's results collected on two different days, a paired t-test was
carried out. In cases where the data were not normally distributed, the Wilcoxon–Rank
test was used.
When comparing a single dog's results gathered from more than 2 days, a one-way or two-
way ANOVA for repeated measurements was conducted. In cases where the data were not
normally distributed, a one-way ANOVA for repeated measures (Friedman) was
performed.
In cases where significant differences were detected, the data were compared using Tukey-
test or Dunn's test.
3. Results
3.1. Establishing reference values
In order to determine whether the dogs’ salivary cortisol baseline levels could be used as
reference values for this research, they were compared to the cortisol values gathered
from both the periods “Simple Hunting” and “Hunting Impeded” of the preliminary test.
This showed that the salivary cortisol baseline levels were higher than the values gained
during the period of “Simple Hunting”. Therefore, the baseline levels were not suitable as
reference values. Instead, the average of the salivary cortisol values collected during the
period of “Simple Hunting” was used for every dog. As described in “Section 2 and
Methods”, this average was needed for calculating the “Relative Cortisol Values”.
Concerning the heart rate, the baseline levels of the heart rate were compared to the
averages of the “Mw15” values gained during the preliminary test. No significant
differences were detected (p = 0.57–1.0). Therefore, “Mw15” values were chosen as
reference values.
3.2. Preliminary tests without application of electric pulses
(i) Comparison of “Simple Hunting” and “Hunting Impeded” in general “Absolute Cortisol
Values” as well as “Relative Cortisol Values” gained from the period of “Hunting Impeded”
were found to be significantly higher than those during “Simple Hunting” (p ≤ 0.001 and p
≤ 0.001).
(ii) Comparison of “Simple Hunting” and “Hunting Impeded” concerning groups A, H and
R. When comparing the salivary cortisol results of groups A, H and R collected during the
preliminary tests, it was found that during “Simple Hunting”, the “Absolute Cortisol
Values” of group H were significantly less than those of both group A and R (p < 0.05).
During “Hunting Impeded”, the “Relative Cortisol Values” of group R were significantly
less than those of both group A and H (p < 0.05). No significant differences were
discovered in heart rates.
3.2.1. Main tests with application of electric pulses
3.2.1.1. Comparison of salivary cortisol values
For the three groups A, H, and R the “Absolute Cortisol Values” as well as “Relative
Cortisol Values” were analysed. The mean value of all dogs per group per day was taken.
Subsequently, those values for day 1, 2 and 3 with electric pulses and day 1 without were
compared and checked for differences.
3.2.1.2. Dataset “r1–r3, n1”
Comparing the groups of dogs, the following was found: considering the 4 days “r1”, “r2”,
“r3”, and “n1” altogether, “Absolute Cortisol Values” and “Relative Cortisol Values” of
group R were significantly higher than those of group A (p < 0.05 and p < 0.05).
Considering only day “r1”, no significant difference between the values for group A and
those for group R was detected (difference of averages [MW-Diff] 0.39 ng/ml).
Considering the other 3 days, “r2”, “r3”, and “n1”, the “Absolute Cortisol Values” for
group R were significantly higher than those for group A (MW-Diff 2.57, 2.94, and 1.87
ng/ml).
Comparing the different days an electric stimulus was applied, in group A no significant
difference was detected. However, the highest value was measured on day “r1”. In group
R, the values recorded on day “r3” were significantly higher than those recorded on day
“r1” (MW-Diff 2.70 ng/ml) and those recorded on day “n1” (MW-Diff 1.19 ng/ml). Also,
the values recorded on day “r1” were significantly lower than those recorded on day “r2”
(MW-Diff 2.34 ng/ml) and those recorded on day “n1” (MW-Diff 1.50 ng/ml).
Altogether, the results described above show a continuous increase in “Absolute Cortisol
Values” for group R from day “r1” until day “r3”. On day “n1”, the values show a decrease,
however, the value is still higher compared to the day that the first electric pulse was
administered.
3.2.1.3. Datasets “r1–r2, n1–n3”, “r1, n1–n3” and “r1–r2, n1”
Concerning both “Absolute Cortisol Values” and “Relative Cortisol Values”, the results
collected from group “R” were significantly higher than those gained from both group A
and group H (p < 0.05). Furthermore, the “Relative Cortisol Values” gained from group H
were significantly higher than those gained from group A (p < 0.05).
3.2.1.4. Dataset “r1, n1”
Concerning both “Absolute Cortisol Values” and “Relative Cortisol Values”, the results
taken from group R were significantly higher than those taken from both groups A and H,
respectively.
3.2.2. Comparison of heart rate values
For the comparison of the heart rate measurements, the key figures “Max”, “Mw15”,
“Max/Mw15” and “Max–Mw15” were used.
3.2.2.1. Datasets “r1–r2, n1”, “r1, n1” and “r1–r3, n1”
For these datasets no significant differences in heart rate were found.
3.2.2.2. Dataset “r1–r2, n1–n3”
Concerning the key values “Mw15”, “Max/Mw15”, and “Max–Mw15”, no significant
differences were found between the groups of dogs or between the different days of the
experiment. However, concerning “Max”, significant differences existed.
3.2.2.3. Dataset “r1, n1–n3”
Significant differences were found between the values of “Mw15” in respect of the
application of the electric pulse (p = 0.02), as well as in respect of the values of
“Max/Mw15” when comparing group A with group H and R (p = 0.027).
3.3. Post-tests
3.3.1. Comparison of saliva cortisol values
In “Absolute Cortisol Values” and “Relative Cortisol Values”, significant differences (p <
0.001) between the groups of dogs were detected. The cortisol values gained from group
R were the highest ones, the values gained from group A the lowest ones recorded.
3.3.2. Comparison of heart rate values
Concerning the heart rates, no significant differences between the groups of dogs were
found.
3.3.3. Comparison between, preliminary test, main test, and post-test
For statistical reasons, only day “r1” and day “n1” were taken into account.
Within group A, significant differences were found for “Absolute Cortisol Values”. The
values collected on the fifth day of the period “Hunting Impeded” were significantly
higher than the values recorded during the post-test. No significant differences were
found for “Relative Cortisol Values”.
Within group H, both “Absolute Cortisol Values” as well as “Relative Cortisol Values”
gained during the post-test were significantly higher than the values recorded on the
second and the third day of the period of “Simple Hunting”. The salivary cortisol levels
measured during the preliminary test were not reached again during the post-tests.
Concerning group R, the “Absolute Cortisol Values” and the “Relative Cortisol Values”
measured during the post-test were higher than those gathered on all of the days of the
preliminary test as well as the first day of application of the electric pulse.
4. Discussion
Data recording in the preliminary test was carried out for the following reasons:
Simple Hunting To evaluate whether the parameters simply change due to the current
chase sequence
Hunting Impeded To investigate whether stress in a dog on the lead can be compared with
stress caused by the use of an electronic training collar
The complete experimental design of preliminary test, main test, and post-test was set up
in an unchanging, isolated room. Thus, exclusion of uncontrollable stressors from the
outside could be guaranteed and corruption of the rest results prevented.
The 14 dogs used in this experiment were exclusively laboratory animals obtained from
the same laboratory animal breeder. Before the experiment the dogs did not have any
intensive contact to humans. Therefore it might be assumed that the 14 Beagles were less
tolerant to stress than dogs living in a private household from a very young age on.
However, the advantage of laboratory animal husbandry is that standardisation of
breeding, upbringing, and husbandry is achieved. Therefore, varying behavioural
performances and a varying reactivity amongst the dogs cannot be readily attributed to
differences in their previous experience. The experimental groups can thus be more
reasonably compared with one another. Furthermore the animals’ environment can be
controlled more easily, which made the control of potential stressors on the day of the
experiment easier than it would have been in a private household and as a result,
uncontrolled influences on the experimental data are reduced to a minimum.
Experiments and data recordings were carried out over time periods of up to 1.5 h for
each dog. Distortion of the experimental results due to slight shifts in saliva sampling is
not to be expected. This is supported by the findings of Kemppainen and Sartin (1984) and
Thun et al. (1990) who found no circadian rhythm to ACTH-release in the dog.
Stress leads to activation of the hypothalamic pituitary adrenal axis (HPA-axis) resulting
in a release of corticosteroids (Henry and Stephans, 1977). In this study we used salivary
samples to determine the level of this steroid hormone so as to be able to repeat tests over
a relatively short period of time with the lowest amount of stress as possible. This method
simplifies experimental processes considerably and makes it possible to extend
investigations over a longer time period ([Greenwood and Shutt, 1992], [Zanella, 1992],
[Parrott et al., 1989], [Walker, 1989] and [Fell et al., 1985]).
The technique is valid because there is a high to very high correlation between overall
cortisol and salivary cortisol in dogs ([Vincent and Michell, 1992] and [Beerda, 1997]).
According to Vincent and Michell (1992), the cortisol concentration in canine saliva is
about 4–10% of the plasma cortisol. Beerda (1997) found saliva cortisol concentration to
have a value between 7.2% and 11% of plasma cortisol concentration.
The median of baseline cortisol level found by Vincent and Michell (1992) and Beerda
(1997) is higher than that recorded in this study. However, the median baseline values
found in this study was between 27.1% and 27.7% of their findings. The values measured
in this study before an electric pulse was administered correspond to only a fifth and a
third respectively of the values found in the literature. As the rate of cortisol recovery in
the test is as high as 95%, this cannot be attributed to the inter- and intra-assay variance of
5% and 10 %, respectively. One possibility is that there is a loss of cortisol due the
additional step in the extraction procedure, using citric acid. Stimulation of salivation with
citric acid does not alter the experimental results. Beerda (1997), for example, used citric
acid as a stimulant to increase salivation. So long as the current study does not attempt to
draw any conclusion from direct comparison of values with those stated in the literature
elsewhere, this is not a problem as the values of this study are consistent. Thun and
Schwartz-Porsche (1994) draw attention to the fact that released cortisol can mask a
circadian rhythm. Therefore it seems unlikely that testing the dogs within a time variant of
up to 1.5 h affected the results. Another possible explanation of the difference lies in the
proposition that the laboratory animals were more relaxed because they were adapted to
the experimental environment. However, the data for the electronic collar phase of the
study is not consistent with this idea.
The importance of collecting at least one cardiovascular parameter in addition to
neuroendocrine parameters has been indicated in the literature (Vincent et al., 1993). This
study used the suggested procedure of measurement of heart rate. It has been suggested
that heart rate frequency reaches higher values when tested animals are given the chance
to evade a stimulus ([Anderson and Brady, 1971], [Anderson and Brady, 1972], [Anderson
and Brady, 1973], [Anderson and Brady, 1976] and [Anderson and Tosheff, 1973]).
However, in previous studies the animals’ mobility has been restricted. This was not the
case in this study. Therefore, stress-induced and motion-induced changes of heart rate
frequency could overlap.
Kirschbaum and Hellhammer (1989), Laudat et al. (1988) and Riad-Fahmy et al. (1983)
state that there is a high individuality in baseline cortisol values as a result of
psychological stimulation as well as after the application of dexamethasone even under
standard experimental conditions. To exclude falsification of future measurement due to
the individual cortisol concentrations of the randomly assigned groups their cortisol
levels were tested in the preliminary test, where group R shows lower cortisol values than
the two other groups. In the Hunting Impeded period, groups H and A show significantly
higher relative cortisol values than group R. This suggests that the results obtained from
the experiment with an electronic collar and those from the post-test are not influenced by
the composition of the group.
The interpretation of the heart frequencies is more complicated since stress induced
reactions are likely to be overlapped by those triggered by physical activity. Yet it is
striking that on the days of stimulus application, an increase in the maxima of heart rate is
reached in group “R”. In group A and H this is the reverse. (Anderson and Brady, 1971),
(Anderson and Brady, 1972), (Anderson and Brady, 1973) and (Anderson and Brady,
1976) describe a decrease in heart frequency if animals are given the possibility to evade
the stressors.
Since the animals of group R unlike those of group A and H had no chance to associate
their behaviour or a warning signal like the “recall” with the punishing stimulus, this could
explain the increase in heart rate on the days of stimulus application.
The results of the cortisol determination show that there is least increase in the absolute
and relative values (between 22.45% and 31.31%) in group A. This increase is even lower
than that which occurred during the “Hunting Impeded” period. Group H follows with an
increase of up to 113.89% in terms of absolute cortisol levels and up to 160% concerning
the relative cortisol values. The highest increase is seen in group R with up to 336.36% in
the absolute cortisol values and up to 327.78% in the relative cortisol values. The values
of group R are significantly higher in each set of data than those of group H and A. In the
data set r1, n1, the relative cortisol values are significantly higher in group H than those of
group A. Beerda (1997) reports a cortisol increase due to unpredictable stressors, such as
sudden noise, of about 374%, while Palazzolo and Quadri (1987) reported an increase of
more than 250% if the animals were exposed to a temperature of −5 °C for an hour. Weiss
(1972) asserts that the predictability of the stressor is of crucial importance. Dess et al.
(1983) postulate that the ability to control the stressor is even more important than its
predictability. The development of plasma cortisol values can therefore be explained
thus. The dogs of group A were able to associate the prey with the application of the
stimulus. Thus the animals were able to predict and to control the stimulus. This explains
the lowest increase of only up to 31.3% in relative cortisol values. In group H the audio-
signal ”Here“ had been trained but not in conjunction with prey being present. As a result
of this lack of specific training, application of the stimulus could be predicted by the dogs,
but they were not able to control their initial reaction, i.e. chasing the prey. This might be
the reason for the distinct increase of cortisol values up to 160% in the relative cortisol
values. However, the increase was not as high as that of 327.78% in group R. The animals
of group R were neither able to control nor predict the stimulus due to its random
application. The increase is similar to the values found in literature concerning an
unpredictable stressor. This corresponds with the result of the post-test and Polsky's
(1994) statement that a lack of timing and/or electric shock that lasts too long causes a
fear of the environment/of people in dogs.
The results are consistent with Feddersen-Petersen and Teutsch (1999) as well as
Grauvogl (1991) who found that poor timing i.e. the impossibility to associate electric
shock and stimulus leads to insecurity and extreme states of anxiety.
5. Conclusion
The results of this study suggest that poor timing in the application of high level electric
pulses, such as those used in this study, means there is a high risk that dogs will show
severe and persistent stress symptoms.
We recommend that the use of these devices should be restricted with proof of theoretical
and practical qualification required and then the use of these devices should only be
allowed in strictly specified situations.
Acknowledgment
We wish to thank the “Hans and Helga Maus Foundation” for supporting this research
financially.
References
Anderson and Brady, 1971 D.E. Anderson and J.V. Brady, Preavoidance blood pressure
elevations accompanied by heart rate decreases in the dog, Science 172 (1971), pp. 595–
597. View Record in Scopus | Cited By in Scopus (1)
Anderson and Brady, 1972 D.E. Anderson and J.V. Brady, Differential preparatory
cardiovascular responses to aversive and appetitive behavioural conditioning, Cond.
Reflex. 7 (1972), pp. 82–96. View Record in Scopus | Cited By in Scopus (5)
Anderson and Brady, 1973 D.E. Anderson and J.V. Brady, Prolonged pre-avoidance
effects upon blood pressure and heart rate in the dog, Psychosom. Med. 35 (1973), pp. 4–
12. View Record in Scopus | Cited By in Scopus (1)
Anderson and Brady, 1976 D.E. Anderson and J.V. Brady, Cardiovascular responses to
avoidance conditioning in the dog: effects of beta adrenergic blockade, Psychosom. Med.
38 (1976), pp. 181–189. View Record in Scopus | Cited By in Scopus (5)
Anderson and Tosheff, 1973 D.E. Anderson and J.G. Tosheff, Cardiac output and total
peripheral resistance changes during preavoidance periods in the dog, J. Appl. Physiol. 34
(1973), pp. 650–654. View Record in Scopus | Cited By in Scopus (4)
Beerda, 1997 Beerda, B., 1997. Stress and well-being in dogs. Dissertation, Utrecht,
Netherlands.
Beerda et al., 1998 B. Beerda, M.B.H. Schilder, J.A.R.A.M. van Hooff, H.W. de Vries and J.
A. Mol, Behavioural, saliva cortisol and heart rate responses to different types of stimuli
in dogs, Appl. Anim. Behav. Sci. 58 (1998), pp. 365–381. SummaryPlus | Full Text + Links
| PDF (217 K) | View Record in Scopus | Cited By in Scopus (43)
Christiansen et al., 2001 F.O. Christiansen, M. Bakken and B.O. Braastad, Social
facilitation of predatory, sheep-chasing behaviour in Norwegian Elkhounds, grey, Appl.
Anim. Behav. Sci. 72 (2001), pp. 105–114. SummaryPlus | Full Text + Links | PDF (169 K)
| View Record in Scopus | Cited By in Scopus (3)
Dess et al., 1983 N.K. Dess, D. Linwick, J. Patterson, J.B. Overmier and S. Levine,
Immediate and proactive effects of controllability and predictability on plasma cortisol
responses to shock in dogs, Behav. Neurosci. 97 (1983), pp. 1005–1016. Abstract | Full
Text via CrossRef | View Record in Scopus | Cited By in Scopus (35)
Feddersen-Petersen and Teutsch, 1999 Feddersen-Petersen, D., Teutsch, G.M., 1999.
Grundlagen einer tierschutzgerechten Ausbildung von Hunden. Verband für das Deutsche
Hundewesen (VDH), 1st ed., Dortmund, Germany.
Fell et al., 1985 L.R. Fell, D.A. Shutt and C.J. Bentley, Development of a salivary cortisol
method for detecting changes in plasma “free” cortisol arising from acute stress in sheep,
Aust. Vet. J. 62 (1985), pp. 403–406. View Record in Scopus | Cited By in Scopus (24)
Grauvogl, 1991 A. Grauvogl, Elektroschocks bei Haustieren, Dtsch. Tieraerztebl. 2 (1991),
p. 93.
Greenwood and Shutt, 1992 P.L. Greenwood and D.A. Shutt, Salivary and plasma cortisol
as an index of stress in goats, Aust. Vet. J. 69 (1992), pp. 161–163. View Record in Scopus
| Cited By in Scopus (23)
Henry and Stephans, 1977 J.P. Henry and P.M. Stephans, Stress, health and social
environment: a sociobiologic approach to medicine, Springer Verlag, New York (1977).
Kemppainen and Sartin, 1984 R.J. Kemppainen and J.L. Sartin, Evidence for episodic but
not circardian activity in plasma concentrations of adrenocorticotrophin, cortisol and
thyroxine in dogs, J. Endocrinol. 103 (1984), pp. 219–226. View Record in Scopus | Cited
By in Scopus (42)
Kirschbaum and Hellhammer, 1989 C. Kirschbaum and D. Hellhammer, Response
variability of salivary cortisol under psychological stimulation, J. Clin. Chem. Clin.
Biochem. 27 (1989), p. 237. View Record in Scopus | Cited By in Scopus (17)
Klein, 2003 D. Klein, Telereizgeraete, Gutachten, Muenster, Germany (2003).
Laudat et al., 1988 M.H. Laudat, S. Cerdas, C. Fournier, D. Guiban, B. Guilhaume and J.P.
Luton, Salivary cortisol measurement: a practical approach to assess pituitary-adrenal
function, J. Clin. Endocrinol. Metab. 66 (1988), pp. 343–348. View Record in Scopus |
Cited By in Scopus (108)
Palazzolo and Quadri, 1987 D.L. Palazzolo and S.K. Quadri, Plasma thyroxine and cortisol
under basal conditions and during cold stress in the aging dog, Proc. Soc. Exp. Biol. Med.
185 (1987), pp. 305–311. View Record in Scopus | Cited By in Scopus (12)
Parrott et al., 1989 R.F. Parrott, B.H. Misson and B.A. Baldwin, Salivary cortisol in pigs
following adrenocorticotrophic hormone stimulation: comparison with plasma levels, Br.
Vet. J. 145 (1989), pp. 362–366. Abstract | View Record in Scopus | Cited By in Scopus
(34)
Polsky, 1994 R.H. Polsky, Electronic shock collars: are they worth the risks?, J. Am.
Anim. Hosp. Assoc. 30 (1994), pp. 463–468.
Riad-Fahmy et al., 1983 D. Riad-Fahmy, G.F. Read and R.F. Walker, Salivary steroid
assays for assessing variation in endocrine activity, J. Steroid Biochem. 19 (1983), pp.
265–272. View Record in Scopus | Cited By in Scopus (38)
Schilder and van der Borg, 2003 M.B.H. Schilder and J.A.R.A.M. van der Borg, Training
dogs with help of the shock collar: short and long term behavioural effects, Appl. Anim.
Behav. Sci. 85 (2003), pp. 319–334.
Stichnoth, 2002 Stichnoth, J., 2002. Stresserscheinungen beim praxisaehnlichen Einsatz
von elektrischen Erziehungshalsbändern beim Hund. Dissertation, Hannover, Germany.
Thun and Schwartz-Porsche, 1994 R. Thun and D. Schwartz-Porsche, Nebennierenrinde.
In: F. Doecke, Editor, Veterinaermedizinische Endokrinologie (3th ed.), Gustav Fischer
Verlag, Jena, Stuttgart, Germany (1994), pp. 315–329.
Thun et al., 1990 R. Thun, E. Eggenberger and K. Zerobin, Twenty-four hour profiles of
plasma cortisol and testosterone in the male dog: absence of circardian rhythmicity,
seasonal influence and hormonal interrelationships, Reprod. Dom. Anim. 25 (1990), pp.
68–77. Full Text via CrossRef
Vincent and Michell, 1992 I.C. Vincent and A.R. Michell, Comparison of cortisol
concentrations in saliva and plasma of dogs, Res. Vet. Sci. 53 (1992), pp. 342–345. View
Record in Scopus | Cited By in Scopus (38)
Vincent et al., 1993 I.C. Vincent, A.R. Michell and R.A. Leahy, Non-invasive measurement
of arterial blood pressure in dogs: a potential indicator for the identification of stress, Res.
Vet. Sci. 54 (1993), pp. 195–201. View Record in Scopus | Cited By in Scopus (26)
Walker, 1989 R.F. Walker, Salivary corticosteroids: clinical and research applications, J.
Clin. Chem. Clin. Biochem. 27 (1989), pp. 234–235. View Record in Scopus | Cited By in
Scopus (6)
Weick, 1976 Weick, F., 1976. Das Teletaktgeraet als Dressurhilfe. In: Der
Jagdgebrauchshund, vols. 3–5, special edition, p. 1.
Weiss, 1972 J.M. Weiss, Psychological factors in stress and disease, Scientific Am. 226
(1972), pp. 104–113. View Record in Scopus | Cited By in Scopus (52)
Zanella, 1992 Zanella, A.J., 1992. Sow welfare indicators and their inter-relationships. Ph.
D. Thesis, Cambridge, UK.
This paper is part of the special issue entitled “Veterinary Behavioural Medicine” guest
edited by Daniel Mills and Gary Landsberg.
Corresponding author. Tel.: +49 511 953 8494; fax: +49 511 953 8056.
Applied Animal Behaviour Science
Volume 105, Issue 4, July 2007, Pages 369-380
Veterinary Behavioural Medicine
| Clinical signs caused by the use of electric training collars on dogs in everyday life situations |