Abstract: The variation in conformation of 356 Swedish Warmblood horses is described, using a quantitative method of measuring horses. Thirty-three of the horses were elite dressage horses, 28 were elite showjumpers, 100 were riding school horses and 195 were unselected four-year-olds. Most horses had a long body form. The average height at the withers was 163.4 cm. Sixty per cent of the horses had a bench knee conformation, 50 per cent had a toe-in conformation of the forelimbs and 80 per cent had outwardly rotated hind limbs. The majority of these deviations were mild or moderate. Conformation was influenced by sex and age. Mares were smaller and had longer bodies and shorter limbs. The elite dressage horses and showjumpers had larger hock angles and more sloping scapulas than other horses. The showjumpers also had smaller fetlock angles in the front limbs. It is suggested that the larger hock angles among the elite horses may be because hocks with small angles are more prone to injury, and because small hock angles may negatively influence the ability to attain the degree of collection necessary for good performance in advanced classes.

Abstract: The purpose of this study was to demonstrate the relationships between jumping technique and dorsoventral acceleration measured at the sternum. Eight saddle horses of various jumping abilities competed on a selective experimental show jumping course including 14 obstacles. An accelerometric belt fastened onto the thorax continuously measured the dorsoventral acceleration during the course. At each jump, 11 locomotor parameters (acceleration peaks, durations and stride frequency) were obtained from the dorsoventral acceleration-time curves. The type of obstacle significantly influenced the hindlimb acceleration peak at take-off and the landing acceleration peak (P<0.01). The poor jumpers exhibited a higher mean forelimb acceleration peak at take-off, a higher forelimb/hindlimb ratio between peaks of acceleration (F/H), and a lower approach stride frequency than good jumpers. Knocking over an obstacle was significantly associated with a low hindlimb acceleration peak at take-off and a high F/H ratio (P<0.01). In order to observe the continuous changes in the frequency domain of the dorsoventral acceleration during the approach and take-off phase, a Morlet's wavelet analysis was computed for each horse jumping over a series of 3 vertical obstacles. Different patterns of time-frequency images obtained by wavelet analysis were found when the horse either knocked over a vertical obstacle or cleared it. In the latter case, the image pattern showed an instantaneous increase in stride frequency at the end of the approach phase, and a marked energy content in the middle frequency range at take-off.

Abstract: The objectives were to compare the temporal characteristics of canter pirouette strides with collected canter strides in elite dressage horses, and to determine whether the stride kinematics of the canter pirouettes fulfilled the requirements specified in the Federation Equestre Internationale Rules for Dressage Events. Eleven horses were videotaped (60 fields/s) during the individual medal competition at the 1992 Olympic Games. Temporal variables were extracted from the videotapes using standard methods. Two strides were analysed on each of the left and right leads and these were pooled to give mean values for the collected canter and the pirouettes. The pirouettes were completed in 4-9 strides, (mean of 6.4). In the collected canter strides, mean duration of the suspension was 0.013 s. There was no suspension in any of the pirouette strides, instead the stance phases of the leading forelimb and trailing hindlimb overlapped by a mean of 0.163 s. In 9 horses the trailing forelimb contacted the ground before the diagonal leading hindlimb in the collected canter, whereas in the pirouettes the leading hindlimb always made contact before the trailing forelimb (mean dissociation 0.164 s), giving the strides a distinct 4 beat rhythm. Due to increases in advanced placement between the diagonal limb pair and between the 2 forelimbs, the stride duration was longer in the pirouette (0.879 s) than the collected canter (0.629 s). It is concluded that the canter pirouette strides did not maintain the rhythm and timing of the the collected canter strides in any of the 11 horses.

Abstract: The object of this study was to determine the limb support sequence during the transitions from walk to trot and from trot to walk in dressage horses under saddle and to test the null hypothesis that the limb support sequence during the transitions is not related to the level of training. Sixteen dressage horses training at novice to FEI Grand Prix level were videotaped performing an average of 9 transitions each from walk to trot and from trot to walk. The 30-Hz videotapes were viewed in slow motion, and based on the limb support sequence the transitions were categorized into two types. In type 1 transitions there were no intermediate steps between the walk and trot sequences. Type 2 transitions were characterized by intermediate steps, including a single support phase. The Kendall rank-order correlation coefficient showed that a higher level of training was positively associated with an increased percentage of type 1 transitions for both walk-to-trot transitions (p < or = 0.05) and trot-to-walk transitions (p < or = 0.01). No significant preference for initiating or completing the trot on the left or right diagonal was found using the binomial test for individual horses and the Wilcoxon signed-ranks test for the group.

Abstract: The ability to jump over an obstacle depends upon the generation of work across the joints of the propelling limb(s). The total work generated by one hind limb of a horse and the contribution to the total work by four joints of the hind limb were determined for a jump. It was hypothesized that the hip and ankle joints would have extensor moments performing positive work, while the knee would have a flexor moment and perform negative work during the jump. Ground reaction forces and sagittal plane kinematics were simultaneously recorded during each jumping trial. Joint moment, power and work were determined for the metatarsophalangeal (MP), tarsal (ankle), tibiofemoral (knee) and coxofemoral (hip) joints. The hip, knee and ankle all flexed and then extended and the MP extended and then flexed during ground contact. Consistent with our hypothesis, large extensor moments were observed at the hip and ankle joints and large flexor moments at the knee and MP joints throughout ground contact of the hind limb. Peak moments tended to occur earlier in stance in the proximal joints but peak power generation of the hind limb joints occurred at similar times except for the MP joint, with the hip and ankle peaking first followed by the MP joint. During the first portion of ground contact (approximately 40%), the net result of the joint powers was the absorption of power. During the remainder of the contact period, the hind limb generated power. This pattern of power absorption followed by power generation paralleled the power profiles of the hip, ankle and MP joints. The total work performed by one hind limb was 0.71 J kg(-1). Surprisingly, the knee produced 85% of the work (0.60 J kg(-1)) done by the hind limb, and the positive work performed by the knee occurred during the first 40% of the take-off. There is little net work generated by the other three joints over the entire take-off. Velocity of the tuber coxae (a landmark on the pelvis of the animal) was negative (downward) during the first 40% of stance, which perhaps reflects the negative work performed to decrease the potential energy during the first 40% of contact. During the final 60% of contact, the hip, ankle and MP joints generate positive work, which is reflected in the increase of the animal's potential energy.