The Veterinary Journal 198 (2013) e137–e142

Contents lists available at ScienceDirect

The Veterinary Journal

journal homepage: www.elsevier .com/ locate/ tv j l

The effect of two preparation procedures on an equine arena surfacein relation to motion of the hoof and metacarpophalangeal joint

1090-0233/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.tvjl.2013.09.048

⇑ Corresponding author. Tel.: +44 845 1962068.E-mail address: alison.northrop@anglia.ac.uk (A.J. Northrop).

Alison J. Northrop a,⇑, Laura-Anne Dagg b, Jaime H. Martin b, Charlotte V. Brigden b, Andrew G. Owen b,Emma L. Blundell b, Michael L. Peterson c, Sarah J. Hobbs d

a Anglia Ruskin University, East Road, Cambridge, Cambridgeshire CB1 1PT, England, UKb Myerscough College, St Michaels Road, Bilsborrow, Preston, Lancashire PR3 0RY, England, UKc University of Maine, Mechanical Engineering, 5711 Boardman Hall, Orono, ME, USAd University of Central Lancashire, Centre for Applied Sport and Exercise Sciences, Preston, Lancashire PR1 2HE, England, UK

a r t i c l e i n f o

Keywords:Equine

KinematicsSurface preparation

a b s t r a c t

A link between surface characteristics and injury has been identified in equine disciplines. Maintenanceprocedures are reported to affect surface characteristics and could influence horse movement. The studyinvestigated limb and hoof movement on a synthetic surface following two different preparations (har-rowing and rolling). Nine horses were recorded using infrared cameras and retro-reflective markers atwalk, trot and canter on two surface preparations in a cross-over design. Hoof rotation and displacement,metacarpophalangeal joint (MCPJ) extension and third metacarpal (McIII) inclination (roll, pitch and yaw)were analysed using a general linear model. Surface hardness and traction were also measured. No dif-ferences in hoof rotations or hoof displacements were found between preparations. However, followingharrowing, greater (P < 0.05) MCPJ extension at mid-stance and greater (P < 0.05) McIII adduction atimpact was found when gait was grouped. Hardness and traction were statistically similar for both prep-arations. Alteration to the surface cushion appears to be sufficient to produce subtle changes in stridecharacteristics.

� 2013 Elsevier Ltd. All rights reserved.

Introduction

Performance horses are expected to compete on arena surfacesprepared in a variety of ways using mechanical equipment, such asharrows and rollers, to alter the surface cushion. Harrowing a sur-face has been reported to decrease substrate hardness and shearresistance and should therefore increase surface deformation(Thomason and Peterson, 2008; Setterbo et al., 2011). Rolling asurface would be expected to produce the opposite effect.Currently, there is limited scientific evidence to recommend anideal preparation; however, optimum surfaces for any givendiscipline need to minimise concussion through energy absorption,whilst still returning energy to the limb in rebound to aidperformance (Barrey et al., 1991; Peterson et al., 2012). Greaterunderstanding of the effects of preparation on the biomechanicsof the horse is therefore needed.

The mechanical properties of a surface are strongly implicatedin the forces encountered at the impact phase of all gaits (Burn,2006; Gustås et al., 2006; Chateau et al., 2010). The hoof plays animportant part in damping the shock during impact; however,

the top layer of the surface (the cushion) can influence hoof decel-eration during initial impact when force is relatively low and load-ing rate is high (Peterson et al., 2012). Hoof deceleration producesbraking forces that withstand sliding during initial impact (Reiseret al., 2000) and, as the horse moves through the stance phase,there is likely to be a degree of hoof displacement before it stops.The body then travels forward over the limb, which acts as a sup-portive strut, causing an increase in ground reaction forces thatconsist of both vertical and horizontal components.

Peak force is known to vary with surface, since the load bearingcapacity of the substrate will influence the ground reaction forces(Chateau et al., 2010) and, consequently, affect the amount of met-acarpophalangeal joint (MCPJ) extension (Crevier-Denoix et al.,2010). The ability of the surface to resist horizontal movementand force is considered in terms of shear strength. The shearstrength and impact resistance of a surface directly affect theamount of hoof displacement and rotation (Gustås et al., 2006;Orlande et al., 2012). The terminal event of the stance phase startswith break-over and ends in toe-off, which involves the rotation ofthe heels over the toe, disengaging contact with the surface.

To date, most studies have compared either different surfacecompositions (Chateau et al., 2009; Setterbo et al., 2009) or moreextreme surface preparations than would normally be used in an

e138 A.J. Northrop et al. / The Veterinary Journal 198 (2013) e137–e142

arena (Gustås et al., 2006). The aims of this study were (1) to mea-sure and compare distal limb and hoof movement on two prepara-tions of a single synthetic arena surface; and (2) to investigate therelationship between distal limb and hoof movement and thecorresponding mechanical properties of each surface. It washypothesised that different surface preparations would alter themechanical properties of the surface cushion, which would influ-ence hoof rotation and hoof displacement, but not MCPJ extension,since the load bearing capacity of the surface was expected to besimilar. It was hypothesised that hoof rotation and hoof displace-ment would be greater on a harrowed surface than a rolled surfacedue to higher deformation and a softer cushion.

Materials and methods

Experimental design

Ethical approval for this project was obtained from University of Central Lanca-shire (approval number PG-11-02-SH). The study was performed outdoors on a spe-cially laid track that consisted of (by weight) 83% silica sand, 15% fibre and felt and2% wax that was laid to 120 mm depth.

Nine horses (4 geldings, 5 mares; mixed breed; mean ± standard deviation, SD,height 154.4 ± 9.2 cm, age 11.2 ± 2.1 years, weight 522.2 ± 84.1 kg) that weredeemed sound and fit were used. Horses were shod with standard iron shoes onall hooves. All horses underwent similar, regular workloads and were habituatedto the equipment prior to testing. Allocation of filming dates and surface prepara-tion order was randomised for each horse. One experienced rider was used through-out the study to reduce inconsistencies arising from riding styles.

For the dynamic trials, eight spherical and three hemispherical retro-reflectivemarkers (38 mm and 42 mm in diameter, respectively) were placed on the left fore-limb of each horse based on the marker set developed by Hobbs et al. (2010). Twoadditional spherical markers were positioned on the lateral and medial distal hooffor a standing image prior to the commencement of each trial (Fig. 1). An additionalmarker placed on the sternum of the horse allowed researchers to monitor velocity.Eight infrared cameras (QualisysOqus, Qualisys AB) were used to capture the leftforelimb of each horse through a calibrated volume of 7 m long, 3 m wide and1.5 m high at 245 Hz. Active filtering was used to extract the marker locations fromneighbouring frames to an accuracy (mean ± SD for all days) of 1.97 ± 0.56 mm for awand length of 749.4 mm.

Two standard preparations were used on the surface: (1) harrow with an inte-gral grading board (Arena Mate, GG Engineering); and (2) standard roller of 300 kgand 2 m diameter (Fig. 2). The surface was prepared using the first maintenance

Fig. 1. Full marker set used for data collection (1–11) including static markers(12 + 13). 1 and 2 (lateral and medial), proximal end of the 3rd metacarpus at thehead of the 2nd and 4th metacarpal bone; 3, offset medially from the middle of the3rd metacarpal; 4 and 5 (lateral and medial), distal end of the 3rd metacarpus overthe collateral ligaments of the MCPJ; 6, 1st phalanx over the common digitalextensor tendon; 7 and 8 (medial and lateral), distal end of the 1st phalanx; 9 and10 (medial and lateral), proximal end of the hoof wall at the coronary band at thelocation of the DIPJ; 11, proximal end of the dorsal hoof wall at coronary band; 12and 13 (medial and lateral), distal end of the dorsal hoof wall.

procedure for that horse and the horse was ridden at walk, trot and canter (left fore-limb was the leading limb) through the calibrated volume until three successful tri-als at each gait were recorded. No more than five trials for each horse werenecessary to obtain three successful captures. The second preparation was thenapplied immediately to the surface and the procedure was repeated. Surfacepreparation did not occur between gaits. Successful velocity matched trials weredeemed to be within ± 0.2 m/s of each other and where all of the markers werevisible through the volume.

Surface hardness and traction were measured using the same techniques asOrlande et al. (2012). Fifteen sections measuring 1.3 m � 1.8 m and considered toundergo the most traffic during data collection were tested immediately after eachhorse was filmed. Surface hardness was measured as peak deceleration (g) with aClegg impact hammer (drop height 0.45 m and weight 2.25 kg) and surface traction(Nm) was measured using a torque wrench with an adapted studded-horseshoebase plate (Blundell et al., 2010). Five locations were randomly selected for hard-ness and traction within each of the 15 sections of the track and a mean valuewas calculated. Moisture content was determined by testing surface samples underlaboratory conditions in a method similar to Peterson and McIlwraith (2008).

A limb model of the hoof, first phalanx (PI) and third metacarpal bone (McIII)segments were constructed from each standing trial and this was applied to thedynamic data of that horse using proprietary software (Qualisys Track Manager,Qualisys AB). Dynamic marker trajectories were smoothed using a low pass fourthorder recursive Butterworth filter with a cut-off frequency of 12 Hz. Impact of thehoof was identified as being mid-way between the vertical velocity minimum andvertical acceleration maximum of the hoof, in a method similar to that previouslyreported by Hobbs et al. (2010). Mid-stance was identified as the point that McIIIbecame vertical (Chateau et al., 2006a). Toe-off was established as being mid-waybetween the subsequent vertical velocity minimum and corresponding accelerationmaximum of the hoof. One stance phase was chosen and extracted from each trial.

McIII inclination, hoof rotation and displacement were calculated relative to theglobal coordinate system and MCPJ rotation was calculated as the rotation of PIabout McIII. Angles were extracted based on an XYZ Cardan sequence of rotationswhere X = flexion/extension or pitch, Y = ab/adduction or roll and Z = internal/exter-nal rotation or yaw. Velocities were derived from displacement data. MCPJ exten-sion and McIII inclination were extracted at impact, mid-stance and toe-off, andROM from impact to mid-stance and mid-stance to toe-off were calculated for hoofrotation and displacement. Peak MCPJ extension was also ascertained and McIIIinclination and the position of the sternum marker relative to the hoof at peak MCPJextension were extracted. Peak hoof, pitch velocity and mean velocity of the ster-num marker were determined from derivative data.

Data analysis

Mean ± standard error (SE) were established for each preparation. Data wereassessed for normality using a Kolmogorov-Smirnov test. A 3 (walk, trot, canter) � 2(harrowed vs. rolled) general linear model (GLM) was applied to test kinematic data(normally distributed). Hardness and traction measurements (normally distributed)were analysed for differences between preparations using a paired samples t test. Apartial correlation (controlling for gait) was used to investigate the relationshipsbetween variables calculated at peak MCPJ extension, which also included hardnessand mean velocity. For these data the difference between harrowed and rolled sur-faces (harrowed minus rolled) was used to reduce inter-horse variability. All datawere analysed (Minitab 16) with significance defined at P < 0.05.

Results

Kinematic variables

Measured differences between preparations (harrowed androlled) are summarised in Table 1. Mean velocity ± SE was 3.06 ±0.14 m/s for harrowed and 3.06 ± 0.14 m/s for rolled surfaces. Nosignificant differences (P > 0.05) were found between the twomaintenance preparations at hoof displacement, hoof rotation orhoof pitch velocity. For MCPJ extension, no significant difference(P > 0.05) was found between preparations at impact or toe-off.Significantly (P < 0.05) greater extension was found on theharrowed surface at peak extension and also extension at mid-stance, the latter occurring shortly after peak extension (Table 1).Mid-stance occurred at a similar percentage of stance phase(mean ± SD) for both surfaces across the group of horses at walk(39.4 ± 2.7% harrowed, 39.4 ± 2.7% rolled); trot (49.5 ± 4.5%harrowed, 50.2 ± 3.7% rolled) and canter (61.2 ± 5.7% harrowed,61.3 ± 3.4% rolled). A significant difference (P < 0.05) was foundbetween preparations at McIII inclination in roll at impact; a greateramount of limb adduction was evident on the harrowed surface

Fig. 2. Photograph of the test track demonstrating surface preparations. (a) Rolled. (b) Harrowed. The L frame present in (a) was used to establish orientation of the 3-dimensional axis in relation to the camera set-up.

Table 1Comparison of preparation effects on biomechanical characteristics (mean ± standard error).

Harrowed(mean ± standard error)

Rolled(mean ± standard error)

Difference between preparation(P value)

Craniocaudal hoof displacement impact to mid-stance (cm) 1.62 ± 0.10 1.66 ± 0.10 0.824Mediolateral hoof displacement impact to mid-stance (cm) 0.67 ± 0.07 0.73 ± 0.09 0.963Vertical hoof displacement impact to mid-stance (cm) 2.44 ± 0.09 2.38 ± 0.09 0.184Mediolateral hoof displacement mid-stance to toe-off (cm) 1.59 ± 0.10 1.46 ± 0.1 0.439Vertical hoof displacement mid-stance to toe-off (cm) 5.16 ± 0.10 5.26 ± 0.01 0.812Hoof rotation impact to mid-stance: roll (�) 3.5 ± 0.25 3.1 ± 0.27 0.408Hoof rotation impact to mid-stance: pitch (�) 5.3 ± 0.56 5.5 ± 0.51 0.714Hoof rotation impact to mid-stance: yaw (�) 2.9 ± 0.25 2.7 ± 0.24 0.604Hoof rotation mid-stance to toe-off: roll (�) 10.5 ± 0.69 10.5 ± 0.64 0.717Hoof rotation mid-stance to toe-off: pitch (�) 85.1 ± 1.62 87.5 ± 1.35 0.122Hoof rotation mid-stance to toe-off: yaw (�) 11.7 ± 0.94 11.5 ± 1.03 0.895Fetlock angle impact (�) �16.35 ± 0.71 �15.69 ± 0.71 0.140Fetlock angle mid-stance (�) �47.20 ± 1.32 �46.04 ± 1.36 0.003*

Fetlock angle toe-off (�) 4.50 ± 0.94 5.24 ± 1.00 0.210McIII inclination impact (�) �4.60 ± 0.35 �3.68 ± 0.34 0.009*

McIII inclination mid-stance (�) �5.94 ± 0.25 �5.65 ± 0.27 0.240McIII inclination toe-off (�) �8.11 ± 0.49 �8.18 ± 0.43 0.98Peak fetlock angle (�) �50.16 ± 1.23 �49.12 ± 1.28 0.008*

Peak hoof pitch velocity (�/s) 1954 ± 53 1907 ± 48 0.240

* Significantly different (P < 0.05).

A.J. Northrop et al. / The Veterinary Journal 198 (2013) e137–e142 e139

compared to the rolled surface (Fig. 3). No other McIII inclinationvariables were significant.

Surface mechanical variables

No significant difference (P = 0.278) in hardness was foundbetween the two surface preparations (Table 2). Notably 7/9 trialsshow the rolled surface to be marginally harder than the harrowed,although only one of these produced significant results (Horse 9,P < 0.05). Torque wrench measurements were only taken after6/9 horses due to a mechanical fault with the torque wrenchmechanism. There was no significant difference (P > 0.05) overallbetween preparations, although the rolled surface had greatertraction during 4/6 trials (Table 2), two of which were significant(P < 0.05). There was no significant difference in moisture contentbetween the North (3.6 ± 1.1%) and South (3.0 ± 0.7%) of the track(P > 0.05) on any of the test dates.

Partial correlations

There was a significant positive relationship (P < 0.01) betweenMcIII roll inclination and mediolateral sternum position, suggest-ing that the greater the inclination (which is a larger negativevalue), the closer the sternum marker was to the hoof in the

mediolateral direction. No other relationships were significant(Table 3).

Discussion

The aim of this study was to investigate the movement of thedistal limb and hoof on an arena surface prepared using two stan-dard procedures. Overall, preparations did not significantly altersurface properties. However, small changes to the surface cushioninfluenced locomotion; an increase in MCPJ extension and McIIIroll inclination were found on the harrowed surface, which wasnot expected. The preparations did not alter hoof mechanics.

Hoof rotation is suggested to give a reliable indication of a sur-face’s ability to deform during the stance phase of the gait (Barreyet al., 1991; Chateau et al., 2006a). On a compliant substrate, suchas the synthetic one used for this study, the hoof deforms the sur-face as it rotates (Johnston and Back, 2006). Since no difference inhoof rotation was found between treatments, the effect of openingup the cushion on the harrowed preparation was not expected tohave altered surface deformation characteristics during brakingor propulsion. It was surmised that a single, maximal load of onelimb was sufficient to even out any changes due to preparation.Surface re-preparation between every trial for each horse, in addi-tion to a more discrete examination of hoof rotation during landing(when the forces are still low), may yield larger differences.

Fig. 3. Representative graphs of third metacarpal (McIII) inclination of roll (�), metacarpophalangeal joint (MCPJ) angle (�), mean speed (m/s) and hardness (g) for one horse atwalk, trot and canter on both preparations (harrowed and rolled).

Table 2Comparison of mean ± standard error arena hardness and traction for two preparations (harrowed and rolled) for each horse from the calibrated volume of the track (for eachhorse, n = 25 measurements).

Horse Hardness (g) Traction (Nm)

Harrowed mean ± SE Rolled mean ± SE P value Harrowed mean ± SE Rolled mean ± SE P value

1 63.3 ± 0.3 65.1 ± 0.9 0.4622 64.9 ± 0.5 65.0 ± 0.6 0.9563 67.0 ± 1.6 72.8 ± 0.5 0.1654 55.8 ± 0.6 58.2 ± 0.5 0.104 13.6 ± 0.1 13.3 ± 0.1 0.5085 66.5 ± 0.9 67.5 ± 1.1 0.474 12.6 ± 0.1 13.4 ± 0.1 0.0536 67.5 ± 0.4 66.3 ± 0.3 0.455 14.2 ± 0.1 13.8 ± 0.1 0.3877 73.7 ± 1.0 67.8 ± 0.6 0.129 13.6 ± 0.1 13.9 ± 0.1 0.3418 66.8 ± 1.2 71.4 ± 0.3 0.229 12.6 ± 0.1 14.4 ± 0.1 0.002*

9 69.1 ± 0.7 73.0 ± 0.6 0.002* 12.9 ± 0.1 13.7 ± 0.1 0.014*

Total 66.1 ± 0.3 67.4 ± 0.3 0.278 13.3 ± 0.04 13.7 ± 0.03 0.202

SE, standard error.* Significantly different (P < 0.05).

Table 3Partial correlation of the difference between harrowed and rolled measures (harrowed–rolled) for third metacarpal (McIII) inclination, roll at peak metacarpophalangeal joint(MCPJ) angle, peak MCPJ angle, hardness, mean speed, and sternum craniocaudal (CC) and mediolateral (ML) displacement relative to the hoof at peak MCPJ angle.

McIII inclination (�) Peak MCPJ angle (�) Hardness (g) Mean speed (m/s) Sternum CC (cm) Sternum ML (cm)

McIIII roll 1.00Peak MCPJ angle 0.03 1.00Hardness �0.08 0.13 1.00Speed �0.21 0.10 0.06 1.00Sternum CC displacement �0.08 0.07 �0.11 �0.01 1.00Sternum ML displacement 0.78* �0.14 �0.12 �0.09 �0.12 1.00

* Significantly different (P < 0.01).

e140 A.J. Northrop et al. / The Veterinary Journal 198 (2013) e137–e142

No differences for hoof displacement in craniocaudal (slip),mediolateral or vertical directions, from impact to mid-stance ormid-stance to toe-off were found between surface preparations.A small amount of craniocaudal displacement, especially at impact,is considered to be beneficial to the horse and aids in decreasingthe forces encountered during deceleration (Gustås et al., 2006;Orlande et al., 2012). Too much craniocaudal displacement may

be damaging to the supporting systems of the limb (Clanton et al.,1991); however it remains uncertain as to what level is detrimen-tal. The amount of displacement and rotation of the hoof can varydue to surface substrate characteristics. Burn (2006) suggestedthat, on sand surfaces, hoof penetration depth can affect displace-ment (craniocaudally) due to the forces exerted on the hoof wall bythe surrounding substrate. The current study used traction to

A.J. Northrop et al. / The Veterinary Journal 198 (2013) e137–e142 e141

indicate shear resistance and found traction to be statistically sim-ilar between preparations, which would explain the lack of differ-ence recorded for craniocaudal and mediolateral hoofdisplacement.

Extension of the MCPJ can be described as mostly occurring dueto the passive properties of the distal limb and caused by the ver-tical forces placed upon it (McGuigan and Wilson, 2003). It couldtherefore be expected that MCPJ extension would not alterbetween surface preparations. Instead, peak MCPJ extension wasgreater on the harrowed surface. The difference in magnitude ofMCPJ peak extension and extension at mid-stance was only 1�.Kicker et al. (2004) has tentatively suggested that differences inMCPJ extension as small as 1� could influence maximal force inthe suspensory structures; therefore, despite these small changes,further investigation is warranted. Initially it was thought thismay be related to limb loading, since there is a strong relationshipbetween the body’s centre of mass and the magnitude of loadingthat the limbs receive (Buchner et al., 2000). None of the variablestested at peak MCPJ extension and that may have influenced limbloading were correlated with peak extension. Passive loadingwould indicate that either a reaction from the ground or a changein limb loading, probably due to a change in dynamic posture, wasoccurring.

In humans, modifications in limb stiffness result from altera-tions in cushioning characteristics of running shoes or surfaces,which change peak vertical forces during running. This is usuallyrelated to varying the amount of knee flexion (Hardin et al.,2004) and, most commonly, larger peak forces are found in morecushioned shoes or surfaces (Ferris et al., 1998; Shorten, 2002).More recent findings in the horse suggest that surfaces composedof wax and sand vs. sand (Robin et al., 2009) or wax and sand vs.soil or turf (Setterbo et al., 2009) produce lower ground reactionforces in the wax-based surfaces, which are considered to bemore compliant. However, the current study investigated a surfacecomposed of the same materials and only the top cushion layeris altered. If horses actively alter forelimb stiffness due to proprio-ception, this must be happening at a more proximal level, whichwas not measured in this study. Further work in horses requiresa more extensive investigation of the variables that influenceforelimb loading during gait on different surface preparations.

McIII inclination in roll, pitch and yaw is used by researchers toidentify the placement of the limb in relation to the body (Chateauet al., 2006a; Hobbs et al., 2011). The present study found no differ-ences in McIII inclination for pitch or yaw at impact or toe-off, so itis probable that stride length did not change between preparations.It was expected these variables may alter, since previously horseshave been described as modifying their pattern of locomotion inrelation to substrate properties (Gustås et al., 2006; Chateauet al., 2010; Weishaupt et al., 2010). Chateau et al. (2010) foundthat horses alter stride length and stride frequency greatlybetween extremely different surfaces (asphalt and deep dry sand).Subtle changes to the surface cushion from two preparations ofthe same substrate do not appear to have the same effect.

At present, there is no information available regarding roll incli-nation of the limb in relation to surface preparation. However, sur-face inclination has been shown to influence limb inclination.Hobbs et al. (2011) reported significantly greater adduction ofMcIII on a flat surface than a banked surface on a turn. The workby Hobbs et al. (2011) demonstrated that a horse will alter place-ment of the hoof relative to the body and that this will affect limbadduction. In the current study, it appears that the horses werealtering foot placement in the frontal plane, which may be a pro-prioceptive alteration, but does not appear to be linked to MCPJextension. Greater adduction was found in the later part of thestance phase at all gaits and, in some horses, this was more notableat canter, but there was no overall surface effect. This may, in part,

be explained by the limb support patterns; during a single supportof the leading limb at canter, the sternum moves closer to the footin the mediolateral plane, which would bring the centre of massover the base of support and provide better dynamic balance. Addi-tionally, external rotation was observed in McIII after mid-stance,which corresponds with findings of Chateau et al. (2006b). Externalrotation will influence planar orientation and therefore measure-ment of McIII inclination in the mediolateral plane, which is a lim-itation of using this measurement.

The current study used one surface, but applied two prepara-tions that were hypothesised to influence the mechanical proper-ties of the cushion and therefore to affect stride characteristics.The Clegg hammer one drop technique has been used to identifydifferences in hardness characteristics in greyhound racetracks(Cook and Baker, 1998) and was deemed to be appropriate for thiscurrent study because of the interest in the surface cushion. Somedifferences in hardness were noted within days; however, data wasnormally distributed across the test dates. Overall, the rolled sur-face was marginally harder and had marginally more traction, asexpected. There were greater differences in the surface preparationon the last two days of testing, although no observed differences inhumidity, rainfall or air temperature were seen between test dates.It can be considered that the visually distinct surface preparationsused in this study do not alter surface characteristics measuredwith the Clegg hammer and traction device. Producing a deeperharrow and rolling with a heavier roller might have provided morecontrast in terms of preparations.

Conclusions

The present study indicates that small alterations in the surfacecushion can influence limb kinematics during stance. Only onesynthetic surface was assessed and therefore these findings arenot representative of all equine arena surfaces. Further work inves-tigating surface preparation is needed and could lead to a greaterunderstanding of subtle changes in stride characteristics.

Conflict of interest statement

None of the authors of this paper has a financial or personalrelationship with other people or organisations that could inappro-priately influence or bias the content of the paper.

Acknowledgements

The authors wish to thank Myerscough College, Bilsborrow,Preston, UK, for financial support for this project and N. Jennings,G. Crook and K. Owen for their assistance during the datacollection.

References

Barrey, E., Landjerit, B., Walter, R., 1991. Shock and vibration during hoof impact ondifferent surfaces. Equine Exercise Physiology 3, 97–106.

Blundell, E.L., Northrop, A.J., Owen, A.G., 2010. The use of Clegg hammer, tractionapparatus, Longchamp penetrometer and proctor penetrometer to test a sand/fibre/rubber/wax equestrian surface. In: Proceedings of the Science Technologyand Research into Sports Surfaces Conference, Loughborough, UK, 21–22 April2010.

Buchner, H.H., Obermuller, S., Scheidi, M., 2000. Body centre of mass movement inthe sound horse. The Veterinary Journal 160, 225–234.

Burn, J.F., 2006. Time domain characteristics of hoof-ground interaction at the onsetof stance phase. Equine Veterinary Journal 38, 657–663.

Chateau, H., Degueurce, C., Denoix, J.-M., 2006a. Effects of egg-bar shoes on the 3-dimensional kinematics of the distal forelimb in horses walking on a sand track.Equine Veterinary Journal Suppl. 36, 377–382.

Chateau, H., Degueurce, C., Denoix, J.-M., 2006b. Three dimensional kinematics ofthe distal forelimb in horses trotting on a treadmill and effects of the elevationof heel and toe. Equine Veterinary Journal 38, 164–169.

e142 A.J. Northrop et al. / The Veterinary Journal 198 (2013) e137–e142

Chateau, H., Robin, D., Falala, S., Pourcelot, P., Valette, J.-P., Ravary, B., Denoix, J.-M.,Crevier-Denoix, N., 2009. Effects of a synthetic all-weather track versus acrushed sandtrack on 3D acceleration of the front hoof in three horses trottingat high speed. Equine Veterinary Journal 41, 247–251.

Chateau, H., Holden, L., Robin, D., Falala, S., Pourcelot, P., Estroup, P., Denoix, J.-M.,Crevier-Denoix, N., 2010. Biomechanical analysis of hoof landing and strideparameters in harness trotter horses running on different tracks of a sand beach(from wet to dry) and on an asphalt road. Equine Veterinary Journal 42, 488–495.

Clanton, C., Kobluk, C., Robinson, R.A., 1991. Monitoring surface conditions of aThoroughbred racetrack. Journal of the Veterinary Medicine Association 198,613–620.

Cook, A., Baker, S.W., 1998. Survey of greyhound race tracks in the United Kingdom.II. Physical characteristics and track performance. Journal of Turfgrass Science74, 123–138.

Crevier-Denoix, N., Robin, D., Pourcelot, P., Falala, S., Holden, L., Estoup, P.,Desquilbet, L., Denoix, J.-M., Chateau, H., 2010. Ground reaction force andkinematic analysis of limb loading on two different beach sand tracks in harnesstrotters. Equine Veterinary Journal 42, 544–551.

Ferris, D.P., Louie, M., Farley, C.T., 1998. Running in the real world: Adjusting legstiffness for different surfaces. Proceedings of the Royal Society London B 265,989–994.

Gustås, P., Johnston, C., Drevemo, S., 2006. Ground reaction force of hoofdeceleration patterns on two different surfaces at the trot. Equine andComparative Exercise Physiology 3, 209–216.

Hardin, E.C., Van den Bogert, A.J., Hamill, J., 2004. Kinematic adaptations duringrunning: Effects of footwear, surface, and duration. Medicine and Science inSports and Exercise 36, 838–844.

Hobbs, S.J., Orlande, O., Edmundson, C.J., Northrop, A.J., Martin, J.H., 2010.Development of a method to identify foot strike on an arena surface:Application to jump landing. Comparative Exercise Physiology 7, 19–25.

Hobbs, S.J., Licka, T., Polman, R., 2011. The difference in kinematics of horseswalking, trotting and cantering on a flat and banked 10 m circle. EquineVeterinary Journal 43, 686–694.

Johnston, C., Back, W., 2006. Hoof ground interaction: When biomechanical stimulichallenge the tissues of the distal limb. Equine Veterinary Journal 38, 634–641.

Kicker, C.J., Peham, C., Girtler, D., Licka, T., 2004. Influence of support boots onfetlock joint angle of the forelimb of the horse at walk and trot. EquineVeterinary Journal 36, 769–771.

McGuigan, P.M., Wilson, A.M., 2003. The effect of gait and digital flexor muscleactivation on limb compliance in the forelimb of the horse Equus caballus.Journal of Experimental Biology 206, 1325–1336.

Orlande, O., Hobbs, S.J., Martin, J.M., Owen, A.G., Northrop, A.J., 2012. Measuringhoof slip of the leading limb on jump landing over two different equine arenasurfaces. Comparative Exercise Physiology 8, 33–39.

Peterson, M.L., McIlwraith, C.W., 2008. Effect of track maintenance on mechanicalproperties of a dirt racetrack: A preliminary study. Equine Veterinary Journal40, 602–605.

Peterson, M.L., Roepstorff, L., Thomason, J.J., Mahaffey, C., McIlwraith, C.W., 2012.Racing Surfaces: Current Progress and Future Challenges Optimize Consistencyand Performance of Track Surfaces for Fewer Horse Injuries. White Paper.Racing Testing Surfaces Laboratory, Maine, USA. http://www.grayson-jockeyclub.org/resources/White_Paper_final.pdf (accessed 16 July 2013).

Reiser II, R.F., Peterson, M.L., Wayne, C., Woodward, B., 2000. Simulated effects ofracetrack material properties on the vertical loading of the equine forelimb.Sports Engineering 3, 1–11.

Robin, D., Chateau, H., Pacquet, L., Falala, S., Valette, J.-P., Pourcelot, P., Ravary, B.,Denoix, J.-M., Crevier-Denoix, N., 2009. Use of a 3D dynamometric horseshoe toassess the effects of an all-weather waxed track and a crushed sand track athigh speed trot: Preliminary study. Equine Veterinary Journal 41, 253–256.

Setterbo, J.J., Garcia, T.C., Campbell, I.P., Reese, J.L., Morgan, J.M., Kim, S.Y., Hubbard,M., Stover, S.M., 2009. Hoof accelerations and ground reaction forces ofThoroughbred racehorses measured on dirt, synthetic and turf track surfaces.American Journal of Veterinary Research 70, 1220–1229.

Setterbo, J.J., Yamaguchi, A., Hubbard, M., Upadhyaya, S.K., Stover, S.M., 2011. Effectsof equine racetrack surface type, depth, boundary area, and harrowing ondynamic surface properties measured using a track-testing device in alaboratory setting. Sports Engineering 14, 119–137.

Shorten, M.R., 2002. The myth of running shoe cushioning. In: Ujihashi, S., Haake,S.J. (Eds.), The Engineering of Sport 4: Proceedings of the 4th InternationalConference on the Engineering of Sport, Kyoto, Japan, 3–6 September 2002.Blackwell Science, Oxford, UK, pp. 1–6.

Thomason, J.J., Peterson, M.L., 2008. Biomechanical and mechanical investigationsof the hoof-track interface in racing horses. Veterinary Clinics of North America.Equine Practice 24, 53–77.

Weishaupt, M.A., Hogg, H.P., Auer, J.A., Wiestner, T., 2010. Velocity dependantchanges of time, force and spatial parameters in Warmblood horses walkingand trotting on a treadmill. Equine Veterinary Journal Suppl. 42, 530–537.