Prepulse Inhibition of the Acoustic Startle Reflex
Janneke E. van der Laan
University of Lincoln, Department of Biological Sciences, Riseholme Park, Lincoln UK
(
j.e.vdlaan@gmail.com
)
Abstract
The neurological basis of stereotypic behaviour in horses (like crib-biting and weaving) is a source of some debate. Dopamine is frequently cited in relation to its potential involvement in the
development and maintenance of stereotypic behaviour across species. It has been suggested that equine stereotypies are the result of increased dopaminergic activity (‘sensitization’) within the basal ganglia. This brain system is also involved in sensorimotor gating. A reliable diagnostic mechanism for measuring sensorimotor gating deficits is by quantifying prepulse inhibition (PPI) of the acoustic startle reflex. For this study, a field test was developed to examine whether horses that show stereotypic behaviour also exhibit dysfunctional PPI, which might reflect changes in sensorimotor gating due to enhanced dopamine transmission in the basal ganglia. Also, the aim was to distinguish between different types of stereotypies during analysis. Horses were tested for mild acoustic startle response and PPI. It was hypothesized that horses that show stereotypic behavior will show a reduced PPI of the acoustic startle response compared to control horses. 20 horses (control n = 8, cribbers n = 8, other stereotypies n = 4) were tested in their own stable for PPI. The stimuli consisted of loud pulses (80-85 dB, 1 KHz, the mild startling pulse (SP)) and the same pulses combined with a preceding non-startling pulse (60-65 dB, 1 KHz, prepulse plus startling pulse (PSP)). The horse’s behaviour was captured on camera in order to record the characteristics of a startle response.Data was analyzed using the non-parametric Wilcoxon signed-rank test to compare between control horses and the different stereotypies. Both stimuli (SP and PSP) significantly increased the occurrence of startle elements in the behaviour of the horses (P < 0.05), indicating that the pulses evoked a startle reaction. One of the startle characteristics (the instant eye blink) and the overall number of startle characteristics were significantly more likely in the SP compared to the PSP condition for control (p < 0.05) but not stereotypic horses. These results suggest that horses that show stereotypic behaviour might have reduced PPI of the acoustic startle reflex compared to non-stereotypic horses. This is consistent with the suggestion that stereotypic behaviour in horses is associated with dopamine dysfunction in the basal ganglia.
LP: The results of this study suggest that stereotypic horses show a neurological change associated
with a change in reactivity. Thus, these behaviours should not be considered simply character faults or management problems, but require a more encompassing biological evaluation, in order to develop treatments which seek to address the underlying mechanism, rather than simple the signs.
1. Introduction
Stereotypic behaviour in horses
It has been estimated that repetitive stereotypic behaviour in horses, like crib-biting, weaving and box-walking, has an average prevalence of 10% in stabled horses (Nicol, 1999) and increases up to 85% when relatively ‘mild type’ stereotypies (e.g. tongue playing) are included (Hausberger et al., 2009). Equine stereotypy is often associated with captivity, since free living or feral horses rarely exhibit this behaviour unless they are kept in captive housing conditions. Stereotypic behaviour seems functionless and is undesirable, both for horse owners who consider these behaviours as ‘vices’, as well for the horse. It has been suggested that there is a possibility of physiological damage to the horse like tooth wear, weight loss and excessive wear on the hooves and the musculoskeletal system as a result of these stereotypic behaviours (Cooper and Mason, 1998; Cooper and McGreevy, 2002). Often, the behaviour is managed through preventive means, for example through cribbing collars and anti-weaving bars, instead of addressing the potential cause. Enabling the horse to perform its stereotypic behaviour may result in distress and displacement behaviour, possibly followed by rebound behaviour (McGreevy and Nicol, 1998), suggesting compromised welfare (as reviewed by Cooper and McGreevy, 2002). As a cause, it seems that the environment of the horse (the horse management) is not suitable in fulfilling their individual ethological and evolutionary needs (e.g. the ability of foraging, social behaviour) and that the horse has very limited degree of control over the given situation leading to behavioural frustration (Winskill et al, 1995; Fraser and Broom, 1990). External (environmental) and internal (e.g. blood pressure, glucose level) factors determine how high the motivational state of an animal is, and a high motivation might result in frustration-related distress.This frustration might lead to the performance of stereotypic behaviour, for instance in order to ‘cope’ with the environment (Cooper and Nicol, 1993).
Development and persistence of stereotypic behaviour in horses
There are several types of stereotypic behaviour, usually divided into oral and locomotor
stereotypies. Oral stereotypies (for example crib-biting and wood-chewing) seem to be particularly associated with the diet of the horses, which often consists of concentrated food (Johnson et al., 1998). The oral stereotypy might be an expression of the lack of opportunity to forage, since feral horses spend 16 to 18 hours a day foraging (Picket, 2009). Also, concentrated feeding can cause digestive problems due to the lack of dietary fiber. It has been suggested that oral stereotypies may increase saliva production, therefore reducing the acidity of gastric tract, preventing stomach
ulceration (Nicol, 1999b). This is also supported by the study of Johnson et al. (1998), who found that stereotypies were reduced when given additional supplement (virginiamycin) which suppresses lactic acid production in the hindgut and increases hindgut pH. Similarly, Mills and MacLeod (2002) found a reduction in cribbing activity after feeding an antacid diet to neutralize stomach acidity in adult horses. However, the exact association between digestive problems and stereotypies has yet to be revealed. Locomotor stereotypies (e.g. weaving, box-walking) are often seen during phases of higher arousal, for example when other horses are taken out and around feeding times (Cooper et al., 2000). The development of this stereotypic behaviour might be due to restricted social behaviour or
frustration of other activity requiring locomotion. In a study by Cooper et al. (2000), the time spent weaving in weavers dropped to zero when the horses had opportunities for social interaction on all four sides of their enclosure. The development of both types of stereotypies is also cited in relation to certain types of weaning, especially oral stereotypies, with weaning by confinement in a stable or
barn associated with an increased risk for abnormal behaviour compared to paddock weaning (Waters et al., 2002).
Pathological considerations
There appears to be a role for endorphins (opioids) in the first stages of stereotypies, especially crib-biting (McGreevy, 2004). It has been suggested that stress stimulates the release of endorphins, triggering excessive dopaminergic activity within the striatum (Rendon et al., 2001). Also, altered dopamine physiology in the nigro-striatal dopamine system (which connects the substantia nigra with the striatum) may be the consequence of animals performing ritualized stereotypies over several years as a habit due to reinforcement, since this system appears to be an important structure that is involved in habit formation (Haber et al., 2000). Dopamine is considered to be the neurotransmitter that is most involved with the development and maintenance of stereotypic behaviour across species (as reviewed by McBride & Hemmings in 2009). High levels of dopaminergic activity within the basal ganglia might play an important role in many stereotypies (Dodman, 1998). The basal ganglia are critical for the development and maintenance of stereotypy (as reviewed by Robbins et al. in 1990). It is suggested that stereotypies are the result of increased neural transmission (‘sensitization’) of the basal ganglia, as stereotypy (crib-biting) horses have different dopamine D1 and D2 receptor densities in specific areas of the basal ganglia compared to non-stereotypic controls (McBride and Hemmings, 2005). Since the basal ganglia dopamine pathways are the primary centre for the control of motor responses, motivation, reward system (leading to addiction) and goal-directed behaviour, dopamine may activate basal ganglia motor systems to reinforce crib-biting via a high motivational state towards goal attainment and reward mechanism (reviewed by McBride and Hemmings in 2005 and 2009). In this way, expression of stereotypic behaviour can be self-rewarding for the horse.
Prepulse inhibition
The dopamine system also plays a role in sensorimotor gating, which refers to the regulation of transmission of sensory information to a motor output system (Peng et al., 1990). The acoustic startle reflex is a motor response to an intense acoustic stimulus. This startle response can be inhibited by presenting a non startling pre-stimulus milliseconds before the startle stimulus; a mechanism known as prepulse inhibition (PPI). A reliable diagnostic mechanism for measuring sensorimotor gating deficits is by quantifying PPI of the acoustic startle reflex (Ludewig et al., 2002). PPI of the acoustic startle reflex is significantly reduced in patients with neuropsychiatric disorders associated with dopaminergic, and possibly serotonergic and/or glutamatergic dysfunctions, such as schizophrenia (Parwani et al., 2000; Ludewig et al., 2002) and obsessive-compulsive disorder (Hoenig et al., 2005). The significance of dopamine to the PPI response is evidenced by the finding that thedopamine agonist amphetamine disrupts PPI in pigs, rats and humans(Lind et al., 2004; Zhang et al., 2000; Swerdlow et al., 2003). As sensorimotor gating is disrupted in patients with dopaminergic system dysfunctions, the question arises as to whether animals that exhibit dopamine dysfunction based stereotypic behaviour also show altered sensorimotor gating.
Prepulse inhibition in horses
Stereotypic behaviour is associated with reduced expression of natural behaviour and frequently with compromised welfare, therefore the welfare of horses could be improved by determining and eliminating the causes of stereotypic behaviour and by meeting the horses’ need to perform important natural behaviours. Since stereotypic behaviour correlates with an increased dopamine transmission, late stage stereotypies may represent some form of psychological disease, comparable
with obsessive-compulsive disorder, schizophrenia or autism in humans (Mills and Nankervis, 2000). Given that excessive dopaminergic activity appears to be associated with stereotypy and impairs PPI, a useful step in being able to differentiate stereotypy from other possible forms of repetitive
behaviour is to test sensorimotor gating in horses that show this behaviour. This study will examine sensorimotor gating in horses that show stereotypic behaviour versus neurotypical horses as a commonly accepted measure of dopamine activity in the dorsal ganglia with PPI of the acoustic startle response. PPI is usually studied in laboratory settings, but in order to have a practical field test, some adaptations to allow testing in the field is necessary. Therefore the aim of this study was: first to develop a version of the PPI test that could be applied in the field using a very mild startle reflex, so as not to significantly compromise the welfare of subjects and secondly examine the face validity of this test by examining the response in stereotypic and neurotypical horses. It was hypothesized that horses that exhibit stereotypic behaviour should show a reduced PPI of the acoustic startle response compared to control horses if the response is a true stereotypy and the test is valid.
2. Materials and methods
2.1 Subjects
In this study, 8 control and 12 stereotypic horses (see table 1) were used. The age of the animals within each group ranged from 6-20 years (M = 10.9) in the control group and 5-19 years (M = 12.8) in the stereotypy group. The horses were recruited from 5 different yards, including The University of Lincoln Equestrian Centre, two private yards, a horse sanctuary and a racing stable. Both control horses and stereotypic horses were obtained from all yards, with the exception of the racing stable and a private yard. Due to circumstances beyond the control of the experimenters, only stereotypic horses could be obtained from these yards. The onset of stereotypies could not be determined, since the behaviours were noticed by most owners from the moment they obtained the horses. Prior to the experiment, horse owners were informed about the procedures and approved participation by signing an informed consent.
Table 1.
Details of horses used in the study
Horse Stereotypies known Age (years) Breed Sex
Charlie Cribber 19 Thoroughbred Gelding
Ernie Cribber 18 Thoroughbred Gelding
Penny Cribber 13 Thoroughbred Mare
Percy Cribber 8 Thoroughbred Gelding
Tommy Cribber 8 Thoroughbred Gelding
Ellie Cribber 13 Shire X TB Mare
Monty Cribber 13 Cob X Gelding
Karl Cribber, weaver 17 Thoroughbred Gelding
Surface Weaver 9 Thoroughbred Gelding
Tom Wind-sucker 5 Thoroughbred Gelding
Secret Box-walker 12 Connemara X Mare
Jacob
pre-feeding & arousal locomotor
stereotypies 18 Thoroughbred Gelding
Sam - 17 Thoroughbred Gelding
Whizz - 8 Part bred Thoroughbred Mare
Archie - 7 Cleveland Bay X Gelding
Frankie - 10 Cob Mare
Simon - 6 Oldenburg X Hannovarian Gelding
Cassie - 6 Irish cob Mare
Fitz - 20 Warm blood Gelding
2.2 Experimental procedure
As a preliminary to this study, pilot tests were undertaken to optimize the experimental
procedures and set-up (see appendix). In the main study, the horses were tested in their stable, with loudspeakers fixed on the stable door at a specific distance from the horse’s head (~30 cm to one side of the horse’s head, with a 95 cm inter-speaker distance, see figure 1). The horses were all handled by a passive handler, who held the rope of the head-collar and made no other movements than
necessary to position the head of the horse centered above the stable door. The horse handler was told to be passive, and was not allowed to pet the animal or talk. The posture of the handler was important in order to prevent giving cues about the stimuli to the horses during the experiment, since horses are known to use several human cues and gestures during decision making (Krueger et al, 2011; Proops & McComb, 2010; Maros et al., 2008; Proops et al., 2010). However, the horse handler had no method for time referencing and thus could not predict pulses, so the chance that cues were given by the handler was minimised. In total, two horse handlers were used (alternating per horse) throughout this study. The observer was positioned at a distance of 1.5 meter of the horse (next to the stable) to regulate the stimulus presentation and did not move during the experiment except for gaze, which alternated between the horse and the pulse inducer.
The heart rate of the horses was recorded by a heart rate monitor (Polar® Equine Training Systems) fit around the body with a roller. Two cameras were positioned at a distance of ~2 - 3 m from the horse’s head to record the horse’s behaviour. The observer and handler started the cameras and the heart rate monitor simultaneously, before taking their predefined positions.
Prior to the experiment, the horses were exposed to 3 sounds with an increasing sound intensity (50, 60 and 70 dB, 1 kHz) with a stimulus interval of 5 seconds in order to determine their sensitivity to noise. In case of an unduly strong response, the horse would be excluded from the research in order to avoid compromising its welfare. However, no horses were excluded for this reason. Following the sound sensitivity test, the horses were exposed to the experimental stimuli.
In the case that horses were motivated to perform stereotypic behaviour during testing, this was allowed in order not to compromise their normal behaviour and increase their stress level. The stimuli were however not presented on those moments in order not to lose data. The stimuli were only presented when the horse’s head was in a resting position that reached the criteria (see 2.3 Stimuli). In total, two horses performed stereotypic behaviour during the experiment (one cribber and one weaver).
Figure 1. Experimental set-up.
2.3 Stimuli
The experimental stimuli consisted of startle stimuli (SP, startle pulse) and startle stimuli with a 60 ms preceding non-startling pulse, the prepulse (PP), this cluster of the PP and SP is referred to as PSP (prepulse plus startle pulse) in this study. The interval between the two types of stimuli (SP vs. PSP) was at least 2 minutes (see Figure 2) in order to allow the horse’s (possible) arousal to return to baseline level.The two minutes was defined from pilot studies based on the observation that no increase in heart rate (as a response to pulses or other environmental stimuli) took longer than 1 minute and 40 seconds to return to normal (previous) level. Two minutes was considered a minimal, rather than fixed interval, since a stimulus was only given when the horses head was in a resting position for longer than 3 seconds (criteria for pulse induction). The order in which the stimuli were presented was SP-PSP-SP-PSP-PSP-SP. This order was chosen to vary between the stimuli and to have a SP at the beginning and the end, to compare between the start of the experiment and the end to check for habituation or sensitization to the stimuli or the experiment, as determined by heart rate data. The experimental session started with a non-startling sound (65-70 dB) in order to aid the synchronization of video cameras, heart rate monitoring and stimulus presentation during data collection.
The SP stimulus consisted of a pulse of 1 KHz with a sound intensity of 80-85 dB and a duration of 60 msec. The PSP was the combination of a prepulse (1 kHz, 40 msec, 65-70 dB) followed by (60 msec) the startle stimulus (SP). The characteristics of the stimuli were determined during the pilot trials (see appendix), and chosen to be non-startling (PP) and arousing (slightly startling, SP). A full startle was not necessary, since a slight startle (heightened arousal) was considered to be sufficient to be affected by the prepulse according to their theoretical bases. This is an innovation in the method used in this study, with the emphasis on welfare related refinement and field adaptation.
Figure 2. The experimental stimuli. The line represents a time slot of >14 minutes, with the big bars representing the startle stimulus (SP) and the smaller bars representing the prepulse (PP).
2.4 Behaviour measurements
In order to measure the startle response, the characteristics of a startle response in horses were defined prior to the experiment. Further, the heart rate of the horses was recorded, in order to note a degree of arousal. The behaviour of the horses was observed from the videos by the experimenter as well as by a second observer blind to the subjects and protocol.
The characteristics of a startle reflex in horses were defined as the following; eye blink, eye muscle movement, the Preyer reflex (quick, transient front-to-back retraction of the pinna) and slower ear movements, visible tension of the muscles or skin, head movement and movement of the muscles of the nose (for definitions of the characteristics, see table 2). The characteristics were scored as events either present (1) or not present (0). In order to measure the amplitude of the response, the intensity of the head movement (small, medium or large) and whether the eye blink was instant or delayed were also recorded. This was validated by assessing intra-observer reliability. For the complete data recording format sheet, see appendix 2. Also the overall number of startle characteristics that the horses showed after the stimuli was determined. The sum of startle characteristics was defined as the sum of the scores of the eye blink instant, eye blink delayed, eye movement, Preyer reflex, ear movement, muscle/skin, head movement and the nose movement (without taking in to account the intensity of the head movement) records.
Table 2.
The definitions of the characteristics of the startle reflex in horses used in this study
Charact. of startle Definition
Eye blink Eye blink, instant (onset within 500 ms) or delayed (onset within 1500 ms) after stimulus presentation. Eye movement If the muscles around the eye are moving and/or the eyeballs of the horse are moving
Preyer reflex Quick, transient front-to-back retraction of the pinna; quick movement of the ears.
Ear movement If Preyer reflex is observed, or other movements of the ear (slower, more directed movement of the ears). Muscle/skin Tension of the skin or muscles of the body (neck, legs, body), shake of the body or skin.
Head movement Movement of the head, fast or slow, short or long, in any direction. Intensity was S, M or L. Nose movement Tension of the muscles around the nose, enlargement of the nose holes.
As a baseline measure and control, 3 random samples in the first 2 minutes (before the first stimulus) and the last 2 minutes (after the last stimulus) were measured. These random moments
varied per horse. At these moments, the horse’s behaviour was recorded in exactly the same manner as after a stimulus. This resulted in 6 non-stimulus samples and 6 samples after a stimulus.
2.5 Statistical data analysis
The behaviour of the horses in the videos was observed using The Observer® XT (Noldus) and FrameShots 3 (FrameShots ™ Video Screen Capture, EOF Productions) and scored in Microsoft Excel (Microsoft Office). Data were transferred from Excel into SPSS (16.0, SPSS Inc.). Since data were not normally distributed (being event data frequencies), group sizes were relatively small (n = 12/n = 8/n = 4) and the magnitude of the difference between the values of the data was important, data were analyzed using a non-parametric Wilcoxon signed-rank test. This Wilcoxon analysis was used in order to compare the likelihood of startle characteristics after the SP, PSP and the random (no pulse) moments. This resulted in a comparison in reactions of the horses to the different pulses.
The heart rate of the horses was analyzed using a paired t-test in SPSS to compare the heart rate during the first two minutes and the last two minutes of the experiment.
2.6 Ethics
The experiments were approved by the relevant Departmental Ethics Committee (Biological Sciences) prior to the start of the study. The horses were not forced to participate in this experiment, and if a horse showed significant withdrawal from the sounds it would be eliminated from the study.
3. Results
Two control horses showed withdrawal from positioning their head over the stable door and were excluded prior to the study. None of the horse’s reaction to the sensitivity test stimuli appeared unduly strong, so we did not exclude a horse from the experiment on that ground and started the study with 20 horses. After data collection, one horse (Jacob, stereotypy group) was excluded since he failed to show a response to the pulses compared to the random moments. This could implicate that the horse was overly aroused (at that moment) to participate in the research or could indicate hearing problems. A response to the pulses is a necessary inclusion criterion for this study.
3.1 Effect of the tones Table 3.
The amount of startle characteristics shown after the pulses (SP and PSP) compared with the random moments (RB = Random Beginning, RE = Random End)
Startle Characteristic Result Significance
Instant Eye Blink SP > RE p < 0.05
Delayed Eye Blink -
-Eye Movement SP&PSP > RB&RE p < 0.01
Preyer Reflex SP&PSP > RB&RE p < 0.05
Ear Movement SP&PSP > RB&RE p < 0.001
Muscle/skin tension SP&PSP > RB&RE p < 0.05
Head Movement All SP&PSP > RE p < 0.05
Nose Movement -
Since 75% of the startle characteristics are significantly greater after tone than at random moments in all the horses as a group (n = 19, Table 3, for exact p-values see appendix 3), the tones (both SP and PSP) have an effect on the horse’s reaction. The overall number of startle characteristics was significantly greater after both SP and PSP tones than after random moments (p < 0.001).
3.2 Difference between the reaction to the SP and PSP Table 4.
The amount of startle characteristics shown after the SP compared with the PSP, for the different groups of horses. Only significant results are displayed, no other significant results (including for other startle characteristics) were found.
Horses
All Control Stereotypies Crib-biting Other stereotypies
n = 19 n = 8 n = 11 n = 8 n = 3
Instant Eye blink - SP > PSP - -
-p < 0.05
Overall (sum) Characteristics - SP > PSP - -
-p < 0.05
There is a significant difference between the reaction to the two types of pulse (SP & PSP) in the instant eye blink in the control group (n = 8, p < 0.05), with a greater instant eye blink rate after the SP than after the PSP (Table 4, for all and exact p-values see appendix 3). This implies that the prepulse decreases the chance of an instant eye blink compared with the startle stimulus alone in the control horses. Also, there is a significant difference between the reaction to the two pulses in the sum of signs in the control group (p < 0.05), with a greater number of signs after the SP compared to after the PSP. Neither effect is seen in the stereotypic horses (n = 11), and therefore the prepulse does not seem to have an effect on the probability of an instant eye blink rate (and sum of startle
characteristics) in stereotypic horses. When just the cribbers are considered (n = 8), there is no evidence of an effect of the PP in the PSP in this population. Also, when the horses that performed stereotypies during testing (n = 2) were excluded, there was no effect of the PP visible.
3.3 Habituation/sensitization to the pulses or experiment
A comparison of the first two minutes (without pulses) with the last two minutes (without pulses) found no difference between the separate startle characteristics shown at those moments. However, the sum of startle characteristics is higher at the random moments during the first two minutes of the test than at the random moments in the last two minutes of the experiment (p < 0.05).
Table 5.
The amount of startle characteristics shown after the first SP compared with the last SP and the first PSP compared with the last PSP, for the different groups of horses. Only significant results are displayed (blue), no other significant results (including for other startle characteristics) were found.
All (n = 19) Control (n = 8) Stereotypies (n = 11)
SP1 - SP3 PSP1 - PSP3 Direction SP1 - SP3 PSP1 - PSP3 Direction SP1 - SP3 PSP1 - PSP3 Direction Preyer Reflex - - - p < 0.05 PSP3 > PSP1 Head Movement Small p < 0.05 - SP1 > SP3 - - - p < 0.05 - SP1 > SP3
Differences in reaction between the first and the last pulses (both SP and PSP) were found (Table 5, for all and exact p-values see appendix 3). In the stereotypic horses, there is a difference in the head movements of the horse between the first SP and the last SP (p < 0.05). The horses show more (small) head movements in reaction to the first startle tone than to the last startle stimulus. Also, in the stereotypic group, the Preyer reflex was more present after the last PSP than after the first PSP (p < 0.05). These effects were not visible in the non-stereotypic horses.
3.4 Heart rate
Due to some loss of heart rate data during data collection, the heart rate data of only 14 horses (7 control and 7 stereotypic)could be analyzed. In these horses, a paired t-test showed that there was no significant difference in heart rate between the first two minutes and the last two minutes of the experiment (t = 1.237, df= 13, p > 0.05). Considering the reaction of the heart rate to the pulses, there was no difference in the heart rate (averaged over 3 to 4 seconds, depending on RR
measurement intervals) directly before and after the pulses (p > 0.05). 3.5 Second observer
Data was also observed by a blind second observer. Overall, the inter-observer reliability coefficient was 0.76. Pearson’s correlation between all data of the initial observer and second observer was r = 0.559, and was highly significant (p < 0.001). The inter-observer reliability coefficient of the data of the instant eye blink was much higher (0.91). The Pearson correlation for the instant eye blink data was r = 0.739 and was highly significant (p < 0.001). When the instant eye blink data of the second observer was analyzed, resulted in a trend for a higher instant eye blink rate after the SP than after the PSP for control horses (p = 0.08), but not stereotypy horses. This result is comparable with the results of the initial observer; however it is a weaker result.
4. Discussion and Conclusion
4.1 Discussion
In this study, investigating prepulse inhibition in control horses versus stereotypic horses, horses reacted both to the mild startle pulse (SP) and the startle plus prepulse (PSP). Furthermore, their heart rate did not increase after pulses, which might indicate that the startle evoked was in fact mild. A difference was found between the reaction to the SP and the PSP in terms of instant eye blink and overall characteristics in control but not stereotypic horses, which implies that there seems to be a form of PPI in control horses but not stereotypic horses. It seems that the horses were aroused slightly higher during the beginning of the test, showing more reaction at the random moments in the first two (non-pulse) minutes than during the last two (non-pulse) minutes. However, analyzing the heart rate of the horses revealed that there was no significant difference in the heart rate during the beginning of the test and the end. This suggests that the horses were not overly stressed during the experiment. Surprisingly, the habituation to the pulses found in terms of head movements was only found in the stereotypic horses, which suggests there is a difference in habituation to the startling pulses between stereotypic horses and control horses in terms of head movements. One
unanticipated finding was that the stereotypic horses showed the Preyer reflex more after the last PSP than after the first PSP, which could implicate sensitization to the PSP in terms of the Preyer reflex.
The dysfunction in sensorimotor gating in stereotypic horses may indicate wider neurological problems in handling sensory information. This could be an important issue for future research. With
respect to the neurological changes that seem to occur with stereotypy, one of the issues that emerge is the relation to the stereotypies. The neurological change can be the result of an addiction-like process, including up- and downregulation of dopamine receptors after being exposed to
increased dopamine transmission. However, the neurological difference might also be partly genetic, and appear immediately post-natal or develop later in life, resulting in a higher risk for stereotypies. The character of a breed is associated with stereotypy, for example Thoroughbreds show a higher rate of stereotypies than other breeds (Bachmann et al., 2003). The increased risk of stereotypies in Thoroughbreds could also be due to management regimes, since this breed is often managed for rapid growth and development in order to be trained for racing (Waters et al., 2002). The character of Thoroughbreds and other high stereotypy risk breeds is often described as quickly aroused and sensitive. In order to investigate the possibilities stated above, a longitudinally study could be conducted to measure PPI and thus sensorimotor gating in young horses that do not show
stereotypies (yet) and determine which horses develop stereotypies across their lifespan. This might give more insight in the onset of the neurological changes but also the relation with breed influences and management factors.
In order to optimize this experiment for future research studying PPI in horses, the study could be undertaken in a more controlled environment. This could increase the persistency of the data since the environment during these experiments was not totally controlled, consistency of
experimental setting was not credible throughout the study. Interruptions during the experiments were common, in terms of people talking in neighboring stables, other horses around the stables and even ducks wandering by. In addition, because the time provided to test a horse was just sufficient enough to perform the experiment, no time was left in order to habituate the horses to the
experimental settings. Finally, considering that both the handler and the observer were in the horses’ sight, minimizing the possible visual and auditory contact between the handler, observer and the horse could eliminate possible cues given to the horse. For example, in future research, headphones for the handler in order to prevent the handler from hearing the pulses could optimize the
experiment. However, previous mentioned inconsistencies were present throughout the whole study, both during the testing of control and stereotypic horses, and could thus be considered fairly
consistent between groups. Since the observer was not blind, the regulation of the pulses could have been biased.
Furthermore, the comparison between random moments and the pulse moments might be biased, since the pulse moments were chosen after moments that the horses had a still restful posture for 3 seconds. The random moments however were chosen on actual random moments determined by a random number generator (http://www.random.org/). This means that at the random moments, horses might also have been aroused, focused on extra-experimental cues or were just randomly moving. The random moments, therefore, represent a random state (any state) of the horse, instead of the resting state moments when the pulses were presented. However, this
implicates that the difference between random moments and pulse moments could be greater when the random moments were also chosen at rest moments, and thus the reaction to the pulses could be more distinct. For future research, the criteria for random moments should be the same as for pulse induction moments.
Data was also observed by a blind second observer, with an inter-observer reliability coefficient of 0.74. Better training of observers could increase inter-observer reliability in the future. The coefficient for the instant eye blink data was higher (0.91), therefore suggesting that the observation of the data important for the significant result of this study was consistent among observers.
4.2 Conclusion
The present study was designed to examine sensorimotor gating in horses that show stereotypic behaviour as a commonly accepted measure of dopamine activity in the dorsal ganglia. The aims were to develop a field test in which prepulse inhibition could be studied in horses, and to evaluate
this test in stereotypic horses. Horses reacted both to the mild startle pulse (SP) and the startle plus prepulse (PSP). A difference was found between the reaction to the SP and the PSP in terms of instant eye blink and overall characteristics in control but not stereotypic horses, which implies that there seems to be a form of PPI in control horses but not stereotypic horses. Returning to the hypothesis posed at the beginning of this study, the results of this study indicate that horses that exhibit
stereotypic behaviour show a reduced PPI of the acoustic startle reflex. This suggests that stereotypic horses, and especially crib-biters (as the sample size of others was too small to evaluate statistically) exhibit dysfunctional sensorimotor gating.
4.3 Implications
The findings of the current study are consistent with the theory that stereotypic behaviour in horses is associated with dopamine dysfunction in the basal ganglia. In general, therefore, it seems that stereotypic horses show a neurological change, and we should thus not only focus on character faults or management problems, but also search for a wider biological evaluation. Further research should be done to investigate the relation between the neurological changes and (the development of) equine stereotypies. Instead of preventing the signs, the underlying biological mechanism of stereotypy must be addressed in research focusing on treatments. Future studies on equine stereotypies are therefore recommended in order to eliminate physiological damage of the horses (like weight loss and colic) and thus improve the welfare of stereotypic horses.
Acknowledgements
The author would like to thank all colleagues and students that contributed to this study. First of all I would like to acknowledge professor Daniel Mills, for supervising this project with brilliant
suggestions and ideas and helping me to progress throughout this study. Second, I would like to thank the students (Jess Cook, Jenny Lee and Danielle Bynoe) that spend a great deal of their time helping with the data collection. Also, I would like to express my thanks to the staff of the Equine Unit at Riseholme for their assistance during the pilot studies and experiments. Further I would like to thank the staff of the Department of Biological Sciences of the University of Lincoln for their assistance, advice and providing discussions which helped to develop ideas. From this department, I would especially like to thank Jessica Hardiman for revising my English and her assistance throughout the study. Last but not least, this study would not have been possible without the cooperation of the horse owners.
5. References
Bachmann, I., Audigé, L., & Stauffacher, M. (2003). Risk factors associated with behavioural disorders of crib-biting, weaving and box-walking in Swiss horses. Equine Veterinary Journal, 35, 158-163. Cooper, J.J. , & Mason, G.J. (1998). The identification of abnormal behaviour and behavioural
problems in stabled horses and their relationship to horse welfare: a comparative review. Equine Veterinary Journal Supplement, 27, 5-9.
Cooper, J. J., McDonald, L., & Mills, D. S. (2000). The effect of increasing visual horizons on stereotypic weaving: implications for the social housing of stabled horses. Applied Animal Behaviour Science,
Cooper, J. J., & McGreevy, P. (2002). Stereotypic behaviour in the stabled horse: causes, effects and prevention without compromising horse welfare. In: Waran, N. (ed.), The welfare of horses, Springer, 99-124.
Cooper, J. J., & Nicol, C. J. (1993). The “coping” hypothesis of stereotypies. Animal Behaviour, 45, 16-618.
Dodman, N. H. (1998). Veterinary models of obsessive-compulsive disorder. In: Jenicke MA, Baer L, Minichillo WA, eds. Obsessive-compulsive disorders: practical management. St Louis, MO: Mosby; 318-334.
Fraser, A. F., & Broom, D. M. (1990). Farm animal behaviour and welfare. Saunders, New York. Haber, S. N., Fudge, J. L., & Mcfarland, N. R. (2000). Striatonigrostriatal pathways in primates form an
ascending spiral from the shell to the dorsolateral striatum. Journal of Neuroscience, 20, 2369-82.
Hausberger, M., Gautier, E., Biquand, V., Lunel, C., & Jégo, P. (2009). Could work be a source of behavioural disorders? A study in horses. PloS One, 4, e7625.
Hoenig, K., Hochrein, A., Quednow, B. B., Maier, W., & Wagner, M. (2005). Impaired prepulse inhibition of acoustic startle in Obsessive-Compulsive Disorder. Biological Psychiatry, 57, 1153-1158.
Johnson, K. G., Tyrrell, J., Rowe, J. B., & Petherick, D. W. (1998). Behavioural changes in stabled horses given non-therapeutic levels of virginiamycin. Equine Veterinary Journal, 30, 139-142.
Krueger, K., Flauger, B., Farmer, K., & Maros, K. (2011). Horses (Equus caballus) use human local enhancement cues and adjust to human attention. Animal Cognition, 14, 187-201.
Lind, N. M., Arnfred, S. M., Hemmingsen, R. P., & Hansen, A. K. (2004). Prepulse inhibition of the acoustic startle reflex in pigs and its disruption by D-amphetamine. Behavioural Brain Research,
155, 217-222.
Ludewig, K., Geyer, M. A., Etzensberger, M., & Vollenweider, F.X. (2002). Stability of the acoustic startle reflex, prepulse inhibition, and habituation in schizophrenia. Schizophrenia Research, 55, 129–37.
Maros, K., Gácsi, M., & Miklósi, A. (2008). Comprehension of human pointing gestures in horses (Equus caballus). Animal Cognition, 11, 457-466.
McBride, S. D., & Hemmings, A. (2005). Altered mesoaccumbens and nigro-striatal dopamine physiology is associated with stereotypy development in a non-rodent species. Behavioural Brain Research, 159, 113-118.
McBride, S. D., & Hemmings, A. (2009). A perspective of equine stereotypy. Journal of Equine Veterinary Science, 29, 10-16.
McGreevy, P. (2004). Behavior and the brain; Neurophysiology and neurochemistry. Equine behaviour; a guide for veterinarians and equine scientist. Saunders, 65-69.
McGreevy, P., & Nicol, C.J. (1998). The effect of short term prevention on the subsequent rate of crib-biting in Thouroughbred horses. Equine veterinary journal, Supplement, 27, 30-34.
Mills, D. S., & MacLeod, C. A. (2002). The response of crib-biting and windsucking in horses to dietary supplementation with an antacid mixture. Ippologia, 13, 33-41.
Mills, D. S., & Nankervis, K. J. (2000). Equine behaviour: principles and practices. Blackwell Science Publications, Oxford.
Nicol, C.J. (1999). Stereotypies and their relation to management. In Harris, P.A., Gomarsall, G.M., Davidson, H.P.B and Green, R.E. (eds.), Proceedings of the British Equine Veterinary Association Specialist Days on Behaviour and Nutrition, 11-14.
Nicol, C.J. (1999b). Understanding equine stereotypies. Equine Veterinary Journal Supplement, 28, 20-25.
Parwani, A., Duncan, E. J., Bartlett, E., Madonick, S.H., Efferen, T.R., Rajan, R., et al. (2000). Impaired prepulse inhibition of acoustic startle in schizophrenia. Biological Psychiatry, 47, 662–9.
Peng, R.Y., Mansbach, R. S., Braff, D. L., & Geyer, M. A. (1990). A D2 dopamine receptor disrupts sensorimotor gating in rats. Implications for dopaminergic abnormalities in schizophrenia. Neuropsychopharmacology, 3, 211-218.
Picket, H. (2009). Horses: behaviour, cognition and welfare. Animal Sentience (www.animalsentience.com).
Proops, L., & McComb, K. (2010). Attributing attention: the use of human-given cues by domestic horses (Equus caballus). Animal Cognition, 13, 197-205.
Proops, L., Walton, M., & McComb, K. (2010). The use of human-given cues by domestic horses, Equus caballus, during an object choice task. Animal behaviour, 79, 1205-1209.
Swerdlow, N. R., Stephany, N., Wasserman, L. C., Talledo, J., Shoemaker, J., & Auerbach, P.P. (2003). Amphetamine effects on prepulse inhibition across-species: replication and parametric extension. Neuropsychopharmacology, 28, 640-650.
Rendon, R. A., Shuster, L., & Dodman, N.H. (2001). The effect of the NMDA receptor blocker,
dextromethorphan, on cribbing in horses. Pharmacology, Biochemistry and Behaviour, 68, 49-51. Robbins, T. W., Mittleman, G., Obrien, J., & Winn, P. (1990). The neuropsychological significance of
stereotypy induced by stimulant drugs. In: Cooper, S. J., Dourish, C. T., eds. Neurobiology of stereotyped behaviour. Oxford: Clarendon Press, 25-63.
Winskill, L., Waran, N.K., Channing, C., & Young, R.J. (1995). Stereotypies in the stabled horse: causes, treatment and prevention. Current Science, 69, 310-316.
Waters, A. J., Nicol, C. J., & French, N. P. (2002). Factors influencing the development of stereotypic and redirected behaviours in young horses: findings of a four year prospective epidemiological study. Equine veterinary journal, 34 (6), 572-579.
Zhang, J., Forkstam, C., Engel, J. A., & Svensson, L. (2000). Role of dopamine in prepulse inhibition of acoustic startle. Psychopharmacology, 149, 181-188.
Sensorimotor Gating in Equine Stereotypy
J.E. van der Laan
Appendices
Appendix 1: Pilot trials
Page 18
Appendix 2: Data sheet
Page 28
Appendix 3: Results section
Appendix 1 – Pilot trials
Dirty pilots
Loud sounds in the stables:
Telephone 72 - 80 dB
Stable doors <90 dB
People talking <80 dB
Dropping hayfork 85 dB
Cleaning with high pressure water 85 dB Materials:
Creative Zen x-fi (mp3 player) Loudspeakers
Stables
11-06-2010 check how horses react to different sounds in their own stables
Background noise (BN): ~35-50 dB (varies: planes, other horses, etc.)
Stimulus was given every 10-30 seconds, without loudspeakers (just the sound of the mp3 player). 4 different sound lengths were presented to the horses, in different orders and at different sound intensities. See table 1.
Horse Distance (m) Sound (msec) Stand (mp3) dB Reaction horse Possible habituation
Shaun 1 120 20 Ear twitch 1 120 23 " "
1 120 25 63 50% Eye blink
Harry 1 120 25 65 50% Eye blink Yes, after 5 tones
1 500 25 50% Preyer reflex Otto 1 120 25 66 % Eye blink
1 500 25 Slight muscle tension nose/eye/ear Bill 1 40 25 20% Ear twitch
1 60 25 2nd & 3rd tone little response, > 3rd tone: no reaction Yes 1 120 25 30% Eye blink, 30% Preyer
1 500 25 15% Eye blink, 50% Preyer Chancer 1 40 25 Muscle tension around eye and ear (at rest) 1 60 25 " "
1 120 25 63 20% Eye blink, Preyer (at the first stimulus) Yes 1 500 25 65 30% Eye blink
Chicago 1,5 40 25 25% Eye blink, 35% eye muscle (at rest) 1,5 60 25 62 " "
1,5 120 25 25% Eye muscle
1,5 500 25 20% Ear twitch Yes
Megan 1 40 25 60% Eye blink (1st)/muscle (rest), ear twitch first tone Yes 1 60 25 40% Eye muscle
1 120 25 50% Ear twitch, eye muscle
1 500 25 20% Eye blink, Preyer (only 2nd time) Yes Honey 1 120 25 Eye blink
(eating)
Bias during experiment:
Telephone in the stables is louder than the stimuli given (80 dB – 72 dB, dependent of the distance between the stable of the horse and the phone).
Obvious presence of observer.
Habituation due to quick next stimulus (short stimulus interval).
In the stables, horses hear the sounds played to the other horses already before they were played to them.
Conclusions and discussion:
1. The horses experience sounds with the same intensity as the stimuli also in their home environment.
2. The gap between two stimuli should be bigger in order to avoid habituation to the stimuli (larger stimulus interval then 10-30 seconds).
3. The horses do not respond aversive to the sounds (could be because the sounds are too weak, <65 dB)
4. Next thing to do: first test different intensities of the sound instead of the length of the sounds, in a more regulated and isolated environment
Treadmill hall
21 & 25-06-2010 Play sound at different intensities to the horses in the treadmill hall.
Sound: 120 msec, background noise: ~ 45 dB. Stimulus was given every 10-50 seconds.
Loudspeakers were positioned in front/above the horses, the horses were monitored by cameras. The horses were handled by the caretakers (Emily, Liam).
See table 2. For analysis, the prevalence of the characteristics of a startle response was counted. When a characteristic was slight of delayed, it counted as 0,5 characteristic, while a skin/muscle contraction was counted as 1,5 characteristic, since this is a characteristic strongly associated with a startle reflex. The average number of characteristics per stimulus intensity is seen in figure 1. Bias during experiment:
Very noisy environment (dogs and people and cars make the background noise go up to ~ 40-45 (sometimes even 50) dB). The horses are extremely focused on the sounds outside and not at the sounds that are played, because they can hear everything outside, but cannot see it. The horses were relatively stressed, a lot of them have not (frequently) been in the treadmill
hall before.
The same stimulus was played several times for the horse, so the reaction of the horses observed is averaged over several stimuli.
Conclusions and discussion:
1. The treadmill hall is not a good place to do the experiment, it is too noisy and the horses are not relaxed because it is not familiar. They can hear but not see what happens outside of the treadmill hall. Move experiment to indoor arena.
2. The number of characteristics of a startle response increase with an increasing sound intensity.
3. The cameras were not good enough positioned, some data could not be rechecked on the video’s because (part of) the horses were out of sight in the video (see ‘?’ in table).
Date Horse Stand (mp3) dB Reaction horse
21-06-'10 Shaun 6 50
-(very stressed) 8 55 Preyer reflex ~
12 60 Preyer reflex ~ 14 65 -17 70 Preyer reflex 18 75 Preyer reflex 21 80 Movement head 22 82 ?
Beau 6 50 Ear twitch
(Calm) 8 55 Ear twitch
12 60 Ear twitch
14 65 Ear twitch
17 70 Ear twitch and Eye Muscle
18 75 Eye muscle/ Eye blink
21 80 Eye muscle and Preyer reflex 24 85 Skin reflex, ear twitch
Ellie 6 50
-(Middle stressed) 8 55 Ear twitch
12 60 Ear twitch
14 65 Eye blink and ear twitch 17 70 Eye blink and ear twitch
18 75 Eye blink
21 80 Eye blink
24 85 Eye muscle and Preyer reflex
25-06-'10 Megan 6 50 Ears twitch
(Calm) 8 55
14 65 Eye blink (?)
17 70 Ears twitch
18 75 Ears twitch (eye blink?), slight head movement 21 80 Ears twitch, eye blink
24 85 Eye blink, Preyer reflex, head movement
Chicago 6 50 Ears twitch, (eye blink?)
(very stressed) 8 55 Blink?
12 60 Eye blink, ear twitch, head movement
14 65
-17 70
-18 75 Ear twitch
21 80 Ear twitch
24 85 Ear twitch
Chancer 6 50 Eye blink?
(Calm) 8 55 Ear twitch, eye blink
12 60 Ear twitch (almost preyer)
14 65 Ear twitch, head movement, eye blink
17 70 Eye blink
18 75 Eye blink, little ear twitch and head movement
21 80 Ears twitch, eye muscle
24 85 Preyer reflex, eye muscle, head movement
Bill 6 50
-(Calm) 8 55 Ear twitch
12 60 ~ Ear twitch
14 65 Ear twitch
17 70 Ear twitch and head movement
18 75 Ear twitch and head movement
21 80 Ear twitch and head movement (fierce) 22 82 Ear twitch and head movement (fierce)
Frankie 6 50
-(Calm) 8 55 Very little ear twitch
12 60
-14 65 - , delayed eye blink 17 70 ~ little ear twitch
18 75 ~ eye blink, ear twitch, head movement
21 80 Ear twitch
24 85 Ear twitch
25 89 Skin shake, little ear twitch Table 2: reaction of horses to different sound intensities in the Treadmill Hall
Indoor arena
29-06-2010 + 01-07-2010 try out sound intensities in indoor arena
Sound: 120 msec, presented at different sound intensities, BN ~ 40-45 (50 with dogs).
See table 3. For analysis, the prevalence of the characteristics of a startle response was counted in the same way as in the treadmill hall, with the exception that the Preyer reflex was added as a characteristic strongly associated with a startle reflex and thus counted as 1,5 characteristic. The underlying reason is that at this point, the Preyer reflex was defined and recognizable in later video analysis, as it was not in the treadmill hall.
Bias during experiment:
In the indoor arena, there are still a lot of sounds. The door makes sounds up to 70 dB, there are also dogs (<50 dB), cars (<65 dB), people talking (<65 dB), rain (<49 dB) and airplanes flying over (<70 dB). The horses however seem less bothered and a lot less stressed than in the treadmill hall (they produce less droppings and stand in resting positions a lot more). Conclusions and discussion:
1. The camera’s were positioned too close, even from a distance the horses features and startle characteristics are still visible and recordable, so the camera’s need to be positioned
differently and further away.
2. Overall comment: number of droppings is lower in indoor arena compared to treadmill hall 3. There seems to be a trend for an increase of the number of characteristics when the intensity
Figure 2: mean characteristics of startle response of the horses showed per sound intensity in the indoor arena, including standard error bars.
Date Horse Stand (mp3) dB Reaction horse
29-06-'10 Megan 6 54
-(stressed) 8 55 Ear twitch
-> foal outside 12 57
-14 60
-17 65 Eye blink
18 68 Ear twitch
21 74 Eye blink, preyer reflex, head movement 24 80 - (too aroused)
Ellie 6 54
-(Calm) 8 55 Eye blink, Preyer reflex, head movement
12 57 Ear twitch
14 60 Ear twitch, eye blink (delayed), head movement
17 65 Ear twitch
18 68
-21 74 Ear twitch
24 80 Eye blink, Preyer reflex, head movement
01-07-'10 Chancer 6 54
-(mild stressed) 8 55
-12 57 Ear twitch (very little) 14 60 Ear twitch (one ear) 17 65 Eye blink, ear twitch 18 68 Eye blink, ear twitch
21 74 Eye blink (delayed), ear twitch (little) 24 80 Eye muscle, ear twitch
Jeremy 6 54 Eye blink (delayed), ear twitch, head movement
8 55 Eye blink, ear twitch (delayed), head movement
14 60
-17 65 Muscle/skin, eye blink (delayed), head movement
18 68 Ear twitch
21 74 Eye blink, head movement
24 80 Eye blink, ear twitch, head movement
Prince 6 54
-(stallion) 8 55 Ear twitch
12 57 Ear twitch
14 60 Eye blink (delayed)
17 65
-18 68 Eye blink, ear twitch, head movement (slight) 21 74 Head movement (slight), eye muscle 24 80 Eye blink (delayed)
Fratna 6 54 Ears twitch
8 55 Eye blink (delayed), ear twitch 12 57 Preyer reflex
14 60 -
17 65 Eye blink (delayed), ear twitch
18 68 Eye muscle
21 74 Eye blink, Preyer reflex, head movement 24 80 Eye muscle, Preyer reflex
Table 3: reaction of horses to different sound intensities in the indoor arena
06 & 08-07-2010 try out sound lengths in indoor arena(08-07 with heart rate)
BN: ~ 40-45 dB. The audio file that was presented to the horses consisted of series of 4 different sound lengths (40, 60, 120 and 500 msec). These series were played at different sound intensities.
Date Name dB
Length
(msec) Reaction Startle Remarks
06-07-'10 Monty 68 40 Et 60 Eb (del.) 120 Eb (del.) 500 Pr 74 40 Muscle, hm, et Yes 60 Et (one ear) 120 Hm
500 Et, em Plane flying over
80 40 Skin, hm, et, eb (OOS) Yes
76 60 Et
Fratna 68 40 - Mowing loan
60 Hm, eb (del.)
120 Et Petting
500 ? (missing data, OOS)
74 40 Et
60 Skin, hm Yes Mowing loan
120 Hm Mowing loan, walking
500 Et Mowing loan, due to head shaking
80 40 Pr Mowing loan
60 Et Mowing loan
120 Et Mowing loan
500 Et Mowing loan
Murphy 68 40 Et, eb Dogs 55 dB
60 Hm Doors 60-65 dB
120 Et, hm Might be due to pulling on rope
500 Et Might be due to already moving head
74 40 Et
60
-120 Et, hm (slight) Distracted
80 40 Nose muscle
60 Et Aroused (snorking)
120 Et, eb (del.)
78 500 Et, eb (del.)
74 40 Eb Petting
08-07-'10 Bill 68 40 Et, eb, hm Yes
60 Et, eb, hm Yes
120 Et, eb, hm Yes
500 Et, eb, hm Yes
74 40 Et, eb (del.), hm Yes
60 Et, hm Yes 120 Et Distracted! 500 Eb (del.), hm, pr Yes 60 40 Et, hm 57 60 Et 56 120 Et (slight) 57 500 Et
80 40 Et, eb (del.), hm, pr Yes
60 Et, hm Yes Airplane flew over
120 Et, eb (del.), hm, pr Yes
500 Et, eb, hm, pr Yes
Chancer 68 40 Pr
60 Et (slight), eb 120 Et (slight), eb
500 Eb Interruption (people opening door)
74 40 Et
60 Slight et 120 Et, em, hm 500 Et
80 40 Pr, noseholes Yes Bigger silence break
60 Et
120 Et Vliegtuig vlieg over
500 Et Vliegtuig vlieg over
57 40 Et
60
-60 120
500
-65 40 Et (little)
Murphy 68 40 Et, eb (del.) Immediately after tone: petting
60 Et Petting
120 Eb (too early), et
500 Et Petting (heavily)
74 40
-60 Et, hm, skin/muscle Yes
120 - Petting 500 pr Dogs 80 40 Em, pr Dogs 60 Hm slight, et 120 Em, pr 500 Pr, hm, eb Yes Distracted 57 40 -60 60 Et (little) 65 120 -500 - Distracted
Table 4: reaction of horses to different sound intensities and sound lengths in the indoor arena. Et = ear twitch, eb = eye blink, hm = head movement, pr = preyer reflex, del. = delayed, OOS = out of sight.
Bias during experiment:
The handler of the horses was petting them regularly.
There was a lot of sound outside of the arena, especially the loan mower, but also birds, dogs and planes, the pilot jets produced sounds up to 70 dB.
Halfway the trial with Chancer, a group of people (a guided tour) walked around the indoor arena and disrupted the experiment by opening the door.
1. If possible, conduct the experiments in the weekend or at other times when the yard is a lot more silent (when there are no of less dogs/planes/people at the yard).
2. The number of characteristics of the startle response increases with an increasing sound intensity (as previously seen in the former pilots), with the maximum at 80 dB (the maximal dB that was presented) (see figure 3 and 4).
3. The number of characteristics of the startle was highest in response to the 40 msec and 500 msec sound at 80 dB (see figure 5).
4. The higher mean number of startle response characteristics at 40 msec at 80 dB compared to the 60 and 120 msec stimulus might be due to the fact that the 40 msec pulse was presented as the first stimulus of the highest sound intensity (80 dB), and the higher response to the 500 msec might be due to the long duration (and possible increased intensity) of the pulse. Considering previous research and the data until this point in the pilots, the pulse will be a 60 msec sound at 80-85 dB, and the prepulse will be a 40 msec sound at a sound intensity of 60-65 dB.
Figure 3: the mean number of startle response characteristics in response to a specific sound intensity (dB).
Figure 4: the mean number of startle response characteristics in response to three specific higher sound intensities (68, 74 and 80 dB), including standard error bars.
Figure 5: the mean number of startle response characteristics in response to four different time lengths (40, 60, 120 and 500 msec) at three specific higher sound intensities (68, 74 and 80 dB), including standard error bars.
21 & 22-07-2010 Prepulse and pulse stimuli with heart rate measurement in indoor arena
Audiofile: prepulse + pulse and pulse alone. The prepulse (PP) was a 40 msec sound at a sound intensity of 60-65 dB and was combined with a startle pulse, the startle pulse (SP) consisted of a 60 msec sound at an intensity of 80-85 dB. The pulses were played with an interval of 3 minutes, with the first pulse 1 minute after onset of the experiment. The order in which the pulses were presented was: SP-PSP-SP-PSP-PSP-SP.
Date Horse PreP/SP Reaction Startle Remarks
21-07-'10 Fratna SP - Distracted PSP - Distracted (hm) SP Em, pr PSP Pr (slight) PSP - Boxes falling SP Et (1 ear) Megan SP Muscle, et
PSP Et (very slight) Distracted
SP Et, nose muscle
PSP Skin/muscle, hm, et, eb Yes
PSP Et (slight), eb (del.)
SP Hm, et, em
PSP
-SP Eb (del.), et (slight), hm (del.)
PSP
-PSP Et Distracted by flies
SP Et, em, hm (del.) Moving
22-07-'10 Name PreP/SP Reaction Startle Remarks
Bill SP Hm, et (1 ear), eb
PSP Skin/muscle, hm, pr Yes Airplane
SP Hm Flies bothering horse
PSP Skin/muscle Yes
PSP Skin/muscle, et Yes Start rain (~45-49 dB)
SP Hm
Owen SP Et, em
PSP Skin, et, eb (del.), hm (del.) Yes
SP Et (slight)
PSP Et, em (slight) A bit distracted
PSP - A lot distracted
SP Hm, et
02-08-'10 Name PreP/SP Reaction Startle Remarks
Fratna SP Eb Pulled back
PSP Em, et (slight) People talking
SP ? Out of sight People talking
PSP ? Distracted People talking
PSP Et People talking
SP Muscle/skin, hm, et Yes
Chancer SP Skin, et, hm Moving
PSP Et
SP Hm Was already shaking head
PSP Hm (slight), pr, eb (del.)
PSP Hm (slight), et, eb
SP Pr Distracted
Table 5: reaction of horses to PSP and SP in the indoor arena. SP=startle stimulus, PSP=Prepulse + startle stimulus. Et = ear twitch, eb = eye blink, hm = head movement, pr = preyer reflex, em = eye muscle, del. = delayed.
Bias during experiment:
Again, there were several possible distracting sounds around the indoor arena. At 22-07 it was heavy raining, which resulted in a BN of ~57 dB.
At 02-08, during the trial of Fratna, Daniel Mills was present in order to check how the experiments were conducted. This leaded to talking and switching of a strategy during the trial, and might have influenced the results.
Conclusions and discussion:
1. The average number of characteristics of a startle response is higher after the SP (pulse alone) stimulus then in response to the PSP (prepulse + pulse) stimulus, however this difference is not significant.
2. Considering the level of stress during the trials in the indoor arena (which is low, but still there is a moderate level of stress) and the fact that probably horses from other yards will be tested as well, the official experiment will be conducted in the horses stable.
Figure 6: mean number of characteristics of a startle reflex as response to a pulse alone (SP) and a prepulse + pulse stimuli (PSP).
Horse Stimulus Mean
Fratna SP 1,00 PSP 0,33 Megan SP 2,50 PSP 2,00 Prince SP 2,33 PSP 0,33 Bill SP 1,50 PSP 2,50 Owen SP 1,50 PSP 1,67 Fratna SP 1,25 PSP 2,25 Chancer SP 2,00 PSP 2,00
Table 6: the mean number of startle reflex characteristics per horse that they showed in response to a a pulse alone (SP) and a prepulse + pulse stimuli (PSP).
Appendix 2 – data sheet
The data was collected by observing the video recordings and scoring the behaviour of the horses in the data sheet below. Each line represents a time point at which for 1500 msec the behaviour of the horse was observed for startle characteristics.
Eyes Ears Body
Eye blink
Eye muscle
Preyer
reflex Ear move Muscle/skin
Head movement Nose move Time Instant Y/N Delayed
Y/N Y/N Y/N Y/N
Directio
n Y/N Y/N S/M/L Y/N
SP 2:00:00 1 0 1 0 1 Back 0 1 S 1
PSP 4:00:00 0 0 1 0 1 Side 1 1 S 1
SP 6:00:00 0 0 1 1 1 Back 0 0 - 0
PSP 8:00:00 0 0 1 1 1 Back 1 1 M 1
SP 12:51:5 0 0 0 1 0 1 Back 0 0 - 1 RB1 0:11:00 1 0 1 0 1 Back 0 1 L 1 RB2 0:21:00 0 0 0 0 1 Front 0 0 - 1 RB3 0:26:00 0 1 0 0 1 Front 0 1 S 1 RE1 13:05:5 0 0 0 0 0 0 - 0 0 - 1 RE2 14:23:5 0 0 0 0 0 0 - 0 0 - 1 RE3 14:50:0 0 0 0 0 0 0 - 0 0 - 0 S = startle stimulus
P = prepulse + startle stimulus
Yes = Y = 1 No = N = 0 ? = data not observable
Charact. of startle Definition
Eye blink Eye blink, instant (onset withing 500 ms) or delayed (onset withing 1500 ms) after stimulus presentation. Eye movement If the muscles around the eye are moving and/or the eyeballs of the horse are moving
Preyer reflex Quick, transient front-to-back retraction of the pinna; quick movement of the ears.
Ear movement If Preyer reflex is observed, or other movements of the ear (slower, more directed movement of the ears). Muscle/skin Tension of the skin or muscles of the body (neck, legs, body), shake of the body or skin.
Head movement Movement of the head, fast or slow, short or long, in any direction. Intensity was S, M or L. Nose movement Tension of the muscles around the nose, enlargement of the nose holes.
Appendix 3 – results section
3.1 Effect of the tones
Appendix 3 - Table 1. Results from the Wilcoxon analysis, comparing the reaction of all horses per startle characteristic at pulse moments (SP and PSP) with the random moments (RB = random beginning and RE = random end). The significant results are displayed in blue.
SP-RB SP-RE PSP-RB PSP-RE Eye Blink Instant 0.145 0.046 0.830 0.187
Eye Blink Delayed 0.564 0.886 0.741 0.541
Eye Movement 0.001 0.001 0.006 0.001
Preyer Reflex 0.025 0.034 0.034 0.020
Ear Movement Back 0.000 0.000 0.000 0.000
Ear Movement Side 0.366 0.813 0.025 0.405
Muscle/Skin 0.016 0.017 0.018 0.014
Head Movement All 0.064 0.002 0.081 0.012
Head Movement Small 0.057 0.007 0.234 0.082
Head Movement Medium 0.397 0.819 0.490 0.361
Head Movement Large 0.655 0.317 0.317 0.157
Nose Movement 0.538 0.499 1.000 0.745
Overall (sum) Characteristics 0.000 0.000 0.000 0.000
3.2 Difference between the reaction to the SP and PSP
Appendix 3 - Table 2. Results from the Wilcoxon analysis, comparing the reaction of the horses per startle characteristic after the SP compared to the PSP per group. The significant results are displayed in blue.
SP - PSP All Control Stereotypies Crib-biting Other stereotypies
n = 19 n = 8 n = 11 n = 8 n = 3
Eye Blink Instant 0.284 0.034 0.792 0.589 0.317
Eye Blink Delayed 0.490 1.000 0.453 0.414 0.785
Eye Movement 0.465 0.334 0.705 1.000 0.157
Preyer Reflex 0.739 0.564 0.414 0.655 0.317
Ear Movement 0.157 1.000 0.157 0.157 1.000
Ear Movement Back 0.102 0.317 0.180 0.317 0.317
Ear Movement Side 0.414 1.000 0.414 0.180 0.317
Muscle/Skin 0.442 0.285 0.066 0.066
-Head Movement All 0.958 1.000 0.860 0.480 1.000
Head Movement Small 0.277 0.453 0.434 0.785 0.197
Head Movement Medium 0.368 0.577 0.442 0.461 0.450
Head Movement Large 0.317 0.317 1.000 1.000 1.000
Nose Movement 0.157 0.083 0.655 1.000 0.317
Overall Characteristics 0.668 0.027 0.339 0.340 0.465
3.3 Habituation/sensitization to the pulses or experiment
Appendix 3 - Table 3. Results from the Wilcoxon analysis, comparing the reaction of the horses per startle characteristic after the first SP compared to the last SP and the first PSP compared to the last PSP. Results are displayed for all horses, control horses and stereotypic horses. The significant results are displayed in blue.
All Contro l Stereotypies S1 - S3 P1 - P3 Directio n S1 - S3 P1 - P3 Direction S1 - S3 P1 - P3 Direction
Eye Blink Instant 0.705 0.083 - 1.000 1.000 - 0.655 0.083
-Eye Blink Delayed 0.317 0.655 - 1.000 1.000 - 0.180 0.564
-Eye Movement 1.000 0.655 - 1.000 0.157 - 1.000 0.564
-Preyer Reflex 0.564 0.180 - 0.317 0.317 - 1.000 0.046 P3 > P1
Ear Movement 0.317 1.000 - 1.000 0.317 - 0.317 0.317
-Ear Movement Side 0.083 0.670 - 0.317 0.257 - 0.157 0.888
-Muscle/Skin 1.000 0.317 - 0.317 1.000 - 0.317 0.317
-Head Movement All 0.035 0.564 S1 > S3 0.180 0.317 - 0.102 1.000
-Head Movement Small 0.035 0.564 S1 > S3 0.414 0.317 - 0.025 1.000 S1 > S3
Head Movement Medium 1.000 0.157 - 0.317 0.157 - 0.317 1.000
-Head Movement Large 1.000 0.317 - 0.317 1.000 - 0.317 0.317
-Nose Movement 0.655 1.000 - 0.564 1.000 - 1.000 1.000
