The Journal of Experimental Biology 202, 23292338 (1999) 2329

Printed in Great Britain The Company of Biologists Limited 1999

JEB2261

THE RELATIONSHIP BETWEEN MECHANICAL WORK AND ENERGY

EXPENDITURE OF LOCOMOTION IN HORSES
A. E. MINETTI*, L. P. ARDIG, E. REINACH AND F. SAIBENE
Department of Physiology, Istituto Tecnologie Biomediche Avanzate, Consiglio Nazionale delle Ricerche, Via Fratelli
Cervi 93, 20090 Segrate, Milano, Italy
*Present address: Department of Exercise and Sport Science, Manchester Metropolitan University, Hassall Road, Alsager,
Cheshire ST7 2HL, UK (e-mail: a.e.minetti@mmu.ac.uk)

Accepted 1 June; published on WWW 9 August 1999

Summary
Three-dimensional motion capture and metabolic highest value to date for terrestrial locomotion. The
assessment were performed on four standardbred horses contribution of elastic structures in the horses limbs was
while walking, trotting and galloping on a motorized evaluated by calculating the elastic energy stored and
treadmill at different speeds. The mechanical work was released during a single bounce (WEL,BOUNCE), which was
partitioned into the internal work (WINT), due to the speed approximately 1.23 J kg1 for trotting and up to 6 J kg1 for
changes of body segments with respect to the body centre galloping. When taking into account the elastic energy
of mass, and the external work (WEXT), due to the position stored by the spine bending and released as WINT, as
and speed changes of the body centre of mass with respect suggested in the literature for galloping, WEL,BOUNCE was
to the environment. The estimated total mechanical work reduced by 0.88 J kg1. Indirect evidence indicates that
(WTOT=WINT+WEXT) increased with speed, while metabolic force, in addition to mechanical work, is also a determinant
work (C) remained rather constant. As a consequence, the of the metabolic energy expenditure in horse locomotion.
apparent efficiency (effAPP=WTOT/C) increased from
10 % (walking) to over 100 % (galloping), setting the Key words: biomechanics, energetics, locomotion, efficiency, horse.

Introduction
The horse (Equus caballus) is a very interesting species for galloping, a great deal of mechanical work is apparently
historical and athletic reasons. From ancient times to the generated, despite the rather low energy consumption required.
industrial revolution it was utilised by man as a work generator A major challenge in the physiology of locomotion today is the
in roles in public transportation, from the times of Alexander precise measurement of the mechanical work actually
the Great (fourth century BC) to the Pony Express. The highly generated by the muscles. These measurements are
specialised evolution, with lengthening of the distal limbs and complicated by two factors: (1) the ability to store and release
progressive reduction in the number of toes from five to one elastic strain energy in tendons and ligaments, and (2) the
(Hildebrand, 1994), and thousands of years of selective occurrence of isometric muscle contractions. The former
breeding for speed and endurance have also made the horse an affects the measured mechanical work, but not the metabolic
outstanding power generator. In fact, its maximum aerobic expenditure (as in a bouncing rubber ball). The latter,
power output is 3.5 times higher than the value predicted by conversely, increases the energy consumption and leaves the
general mammalian allometric equations (Taylor et al., 1981). mechanical work unaffected (as in a standing man with a
At speeds of approximately 10.5 and 16 m s1, for long-duration moderate degree of knee flexion). To complicate this scenario
and sprinting bouts, respectively, the horse is one of the fastest further, we need to consider that a muscle with in-series elastic
quadrupeds on Earth. Since the movement actuator, the muscle, structures can force oscillations (in periodic movements) that
is considered to be structurally similar across different species simultaneously involve both these factors, as has recently been
and sizes (Schmidt-Nielsen, 1984), there is a great interest in observed in running turkeys (Roberts et al., 1997).
the mechanisms underlying such an impressive performance. In this study, we concurrently assessed, at different gaits and
A few metabolic measurements have been made during speeds, the metabolic energy expenditure and the mechanical
locomotion on full-sized horses (e.g. Eaton et al., 1995), but work of horse locomotion using three-dimensional motion
no data on the corresponding mechanical work is available. As analysis. In addition, we attempted to separate the work
a consequence, the locomotory efficiency has never been produced by muscles from that derived from passive elastic
calculated. For bouncing gaits, such as running, trotting and structures and to evaluate the cost of force generation.
2330 A. E. MINETTI AND OTHERS
2
3
9

Links 1
12 Head 10
4
23 Neck 11
34 5 12
45 Humerus
56 Radius 6 13
67 Metacarpus
14
78 Fore-digit 7
39 Trunk 8
Fig. 1. Diagram showing the positions of 910
reflective markers on each side of the 1011 Femur
horses body. The list describes the 11 1112 Tibia
segments of each side that were used in 1213 Metatarsus
the calculation of mechanical work. 1314 Hind-digit

Materials and methods displacements). This theorem states that the total kinetic
Four standardbred horses (mass 515.029.1 kg, mean S.D.; energy of a multi-link system can be divided into two parts:
range 498538 kg) walked (W), trotted (T) and galloped (G) (a) the kinetic energy of the segments arising from their change
on a motor-driven treadmill (Kagra 2000, Switzerland) at of speed with respect to the overall centre of mass, and (b) the
speeds ranging from 1 to 12 m s1. For each condition, kinetic energy of the overall centre of mass with respect to the
simultaneous mechanical and metabolic measurements were environment. The first term constitutes WINT, while the second
made. is included in the external work, WEXT, defined as the work
necessary to raise and accelerate the centre of mass of the body
Biomechanical measurements with respect to the environment. Other details of the
Each trial was recorded by means of an opto-electronic experimental design were reported by Minetti et al. (1993).
motion-analysis system (ELITE System, B.T.S., Italy) The three-dimensional displacement of the centre of mass of
consisting of four cameras, each sampling at a rate of 100 Hz. the body was calculated for each frame from the positions of
Fourteen reflective hemispherical markers (diameter 4 cm) the 22 body segments (the head, neck and trunk segment were
were placed on each side of the horse to identify the segments split into left and right parts). External work (WEXT) was
of interest: head, neck, trunk, humerus, radius, metacarpus, obtained by summing the increments of the total energy (ET)
fore-digit, femur, tibia, metatarsus and hind-digit (Fig. 1). The with respect to time: ET=EP+EKx+EKy (where EP is the
three-dimensional positions of segment extremities were potential energy of the centre of mass of the body and EKx and
calculated stereometrically by the system. Each data EKy are the horizontal and vertical components, respectively,
acquisition lasted for 5 s, and the stride frequency was of the kinetic energy of the bodys centre of mass). The total
calculated by analysing the periodicity of the vertical mechanical work (WTOT) was computed as the sum of WEXT
coordinates and by counting the number of frames for a stride. and WINT. Mechanical was expressed per kilogram of body
One representative stride was extracted, according to the mass and per unit distance (i.e. J kg1 m1).
regularity of the vertical coordinates of the markers with The energy recovery, an index of the ability of a system to
respect to time, from the 5 s acquisition period and was recover mechanical energy through the interchange between EP
successively fed into an algorithm for the determination of the and EK (equal to 100 % in an ideal pendulum), was obtained
mechanical work (see below). according to Cavagna et al. (1976).
Values for the segment mass (as a fraction of body mass), All the data-processing and statistics were performed on an
the position of the centre of mass and the radius of gyration Athena 486/66 DX2 computer and an Apple Macintosh Duo
were taken from Buchner et al. (1997). Positive internal and 2300 computer. Low-pass filtering of the spatial coordinates
external work (WINT and WEXT, respectively) were computed was performed using the procedure of DAmico and Ferrigno
using the method of Cavagna and Kaneko (1977), who used (1990), in which optimal filter cut-off frequencies are selected
Knigs theorem to account for the changes in the kinetic automatically. The range of cut-off frequencies was 5.09.0 Hz
energy of segments whose movements do not affect the using this procedure. A custom-built program, created using
position of the overall centre of mass (e.g. symmetrical limbs LabView2/MacOS (National Instruments, USA), was used to
 Mechanics and energetics of horse locomotion 2331
analyse the three-dimensional data and to calculate the 0.20
biomechanical variables for 209 strides. 4

Metabolic measurements

Metabolic cost (ml O2 kg-1 m-1)

0.15

Metabolic cost (J kg-1 m-1)

Standard open-circuit respirometry was used to determine
3
the metabolic power requirements. The expired gas, flowing
through a mask firmly positioned around the nostrils (BRDL,
Birmingham, UK), was collected in Douglas bags, and
0.10
ventilation rate was assessed by using a chronometer and a dry 2
gas meter (Collins, USA). O2 and CO2 partial pressures were T G
measured using gas analysers (Binos 1, Leybold-Heraeus and
Oxynos 100, Rosemount, Germany). After converting values 0.05
W 1
to STPD (laboratory air temperature was always within the range
1416 C), the rate of O2 uptake was calculated. Values for the
standing horse were subtracted from the steady-state values,
reached after approximately 4 min of exercise, to obtain the net 0 0
metabolic work rate during each trial. To express the metabolic 0 2 4 6 8 10 12
work (C) in the same units as mechanical work (J kg1 m1),
-1
Speed (m s )
we divided the net metabolic work rate by the speed (m s1) Fig. 2. Relationship between the metabolic energy cost (per unit
and assumed that 1 ml of O2 is equivalent to 20.1 J (for a mean mass and distance) and speed during walking (W), running (R) and
respiratory quotient of approximately 0.95) to convert galloping (G). The thick curves refer to parabolic fits of the
metabolic into mechanical units. experimental data (the regression coefficients are given in Table 1),
The efficiency of locomotion, the ratio of mechanical to while the thin curves represent the parabolic fits of the data of Hoyt
metabolic work, was calculated by dividing WTOT by C. and Taylor (1981) for a 140 kg horse. Filled circles shows mean
galloping cost and bars refer to standard deviations (S.D.) of data
taken from the literature (Eaton et al., 1995; Potard et al., 1998;
Langsetmo et al., 1997; Butler et al., 1993; Wagner et al., 1989).
Results
Open circles represent data for walking, filled squares for trotting
The rate of energy consumption of standing horses was and open diamonds for galloping.
1.940.48 ml O2 kg1 min1 (mean S.D., N=4). Fig. 2 shows,
for each gait, the metabolic cost at different speeds. Because
of technical problems related to air leakage through the mask, and Taylor (1981). This is because the subtraction of basal
only a subset of all the measurements was considered. The (standing) metabolic rate from the metabolic measurements, as
main criteria for rejecting some measurements were the lack applied here, has a greater effect at low speeds than at higher
of an increase in ventilation with increasing speed (within each speeds (for a review of different techniques for calculating C,
gait) and a respiratory quotient exceeding 1.0. see Full, 1991). While exceeding (by approximately 30 %) the
Typical changes in mechanical energy for walking, trotting allometric equation obtained for mammalian and bird species
and galloping are shown in Fig. 3. By summing all the energy with masses ranging from 0.01 to 254 kg (Taylor et al., 1982),
increases in ET over the stride period and expressing the result the present values of C are comparable with previous
per unit mass and distance, we obtained WEXT. Fig. 4 shows measurements on horses (Farley and Taylor, 1991; Eaton et al.,
these values (upper curve) and WINT (lower curve). 1995; Potard et al., 1998; Langsetmo et al., 1997; Butler et al.,
Table 1 reports the regression coefficients for parabolic 1993; Wagner et al., 1989).
regressions for the main variables discussed in the paper.
Second-order regression was used for descriptive purposes, Mechanical work
since it can fit most linear and non-linear data without further The mechanical external work (Fig. 4, upper curves) during
increasing the complexity of the analysis. walking and trotting resemble the values for humans at
corresponding gaits (Cavagna et al., 1976). Fig. 5 shows the
three-dimensional range of movement of the centre of mass of
Discussion the horse during single strides at the different gaits and speeds.
The energetics of horses locomotion shows some It is surprising how the lateral displacement (dz) for a 500 kg
optimisation phenomena (Alexander, 1989). At each gait, a horse drops from 50 mm (slow walking) to only a few
speed exists at which C is minimal, as originally discovered in millimetres during fast galloping and how the vertical
a study on small horses (Hoyt and Taylor, 1981). In addition, displacement (dy) more than doubles when passing from
minima in C are comparable at the different gaits, showing an trotting to galloping.
overall constancy of the metabolic energy spent per unit As it is calculated (Cavagna et al., 1976), WEXT represents
distance travelled, despite the difference in speed, although the a good approximation of the mechanical work done by the
value of C during walking is lower than that reported by Hoyt muscles only for non-bouncing gaits such as walking. During
2332 A. E. MINETTI AND OTHERS

A Walk 1.35 m s-1

LH LH LF LH
LF RH RH
250 RF
RF

Mechanical energy (J)

200
ET
150

100
EKx
50
EP
0
0 0.4 0.8 1.2 1.6 2.0 2.4 2.8
Time (s)

B Trot 5.28 m s-1

LH LF LH LF
LF
RF RF
RH RH RH
Mechanical energy (J)

1200 ET

800
EKx
EP+EKy
400
EP
0
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
Time (s)

C Gallop 6.83 m s-1

Fig. 3. Typical examples of changes in mechanical
energy over time for each gait. The traces in each LH LH
LF RH LF RH
RH RF RF
graph were obtained by duplicating a single
experimental stride for illustrative purposes. Since the
Mechanical energy (J)

(positive) mechanical work is obtained by summing 2000

all the increases in the total energy (ET) curve over the ET
stride period (delimited by vertical bars), its main
components (the potential energy EP and the EKx
1000
horizontal component of the kinetic energy, EKx) have
been represented by removing an offset corresponding
EP+EKy
to their minimum value to accommodate all the curves EP
in the same graph. Horizontal bars correspond to the 0
contact time of the right-hind (RH), right-fore (RF), 0 0.2 0.4 0.6 0.8 1.0
left-hind (LH) and left-fore (LF) hooves. Time (s)

walking, a pendulum-like mechanism flattens the ET curve by WEXT increases with speed (Fig. 4), and the energy recovery
the sequential exchange of potential and kinetic energy of the parameter (related to the capacity to use pendulum-like energy-
centre of mass, resulting in a low WEXT. However, during saving mechanisms) (Cavagna et al., 1976) is higher during
bouncing gaits (such as trotting and galloping), WEXT galloping than during trotting (Fig. 6). This high level of
overestimates the mechanical work actually performed by the energy recovery, originally reported for rams and dogs in an
muscles because the increases in energy in the ET curve also earlier study (Cavagna et al., 1977), is not a completely
include the work performed by previously stretched elastic unexpected feature of the galloping gait. In the investigated
structures such as tendons and ligaments. It has been suggested speed range (612 m s1) of the gait (usually termed a slow
that, during galloping, both energy-saving mechanisms operate gallop or a canter), the footfall sequence is a ternary rhythm
simultaneously (Cavagna et al., 1977; Minetti, 1998a). In fact, with the middle beat formed by two diagonal hooves (left-hind,
 Mechanics and energetics of horse locomotion 2333
4 Fig. 4. Positive internal (WINT) and external (WEXT) mechanical
work as a function of speed during walking (W), trotting (T) and
W ext galloping (G). Filled symbols refer to trotting. The superimposed
curves represent the second-order regressions whose coefficients are
listed in Table 1. The different symbols represent the individual
3
Metabolic cost (J kg-1 m-1)

horses.

G
beat (LH, RF) behaves similarly to the trot (Fig. 3B). Because
2 of the dependence of EKx on speed and the independence of
EP from speed (EP=mgdy, where m is mass and g is the
T
acceleration due to gravity; dy is shown in Fig. 5A), as also
shown in other quadrupeds (Heglund et al., 1982), the energy
1 recovery during galloping tends to decrease at high speeds
W int (Fig. 6). A related similarity between walking and galloping is
W that the Lissajous contour of the centre of mass of the body
T G (i.e. a plot of vertical versus horizontal coordinates), when seen
0 from the left side of the animal, is a clockwise loop for both
0 2 4 6 8 10 12 14 gaits, while it is a counterclockwise loop for trotting (as for
Speed (m s-1) running humans, see Cavagna et al., 1964).
The mechanical internal work has never been assessed in
LH, and right-fore, RF, for instance) hitting the ground large (500 kg) horses. However, measurements in a 99 kg pony
simultaneously. This involves three downward peaks (braking) (Fedak et al., 1981) gave comparable results. The values for
in the horizontal kinetic energy curve of the bodys centre WINT in Fig. 4 show that the dependency on speed may be
of mass (Fig. 3C). Since the potential energy curve is different for each gait. This may be related to changes in limb
approximately sinusoidal, it is inevitable that EP and EKx are positioning, e.g. folding of the limbs during the swing phase
out of phase in some parts of the stride, thus increasing energy of bouncing gaits could reduce the moment arm and,
recovery (Minetti, 1998a). The energy exchange is particularly consequently, WINT. In fact, a recent WINT model (Minetti,
evident at the first (RH) and third (LF) beats, whilst the middle 1998b), when applied to the present data, is consistent with this

Table 1. Coefficients for parabolic regressions between the indicated variables and speed
a b c r2
Total work, WTOT (J kg1 m1) W 0.128 0.045 0.240 0.858
T 0.038 0.222 1.689 0.301
G 0.016 0.093 0.641 0.793

Metabolic work, C (ml O2 kg1 m1) W 0.081 0.212 0.212 0.595

T 0.014 0.104 0.291 0.304
G 0.011 0.146 0.595 0.614

Stride frequency, f (Hz) W 0.032 0.407 0.318 0.911

T 0.015 0.015 1.249 0.880
G 0.009 0.109 2.114 0.659

Vertical displacement, dy (mm) W 6.211 11.924 23.895 0.293

T 3.341 28.488 4.243 0.452
G 3.429 58.610 113.663 0.328

External work, WEXT (J kg1 m1) W 0.081 0.084 0.131 0.788

T 0.031 0.192 1.461 0.198
G 0.009 0.168 0.088 0.728

Internal work, WINT (J kg1 m1) W 0.047 0.038 0.109 0.923

T 0.007 0.030 0.228 0.720
G 0.007 0.076 0.552 0.838

Equations were of the form: y=ax2+bx+c, where x is speed (m s1). Sample size was 17 for walking, 24 for trotting and 27 for galloping for
all variables except C, for which values of N were 15, 16 and 10 respectively.
2334 A. E. MINETTI AND OTHERS
50
Vertical
A
150
40
dy (mm)

Energy recovery (%)

100
30

50
20

0
Longitudinal
100
B 10
dx (mm)

50 0
0 2 4 6 8 10 12 14
Speed (m s-1)
0
Lateral Fig. 6. Mechanical energy recovery representing the interchange
100
C between potential and kinetic energy of the centre of mass of the
body, calculated according to Cavagna and Kaneko (1977), plotted
dz (mm)

against speed. Symbols as for Fig. 4.

50

Efficiency
0 The capacity to maintain a high mechanical energy of the
0 5 10 15 body by periodic storage and release of elastic energy (during
Speed (m s-1) trotting and galloping), the additional ability to exchange
Fig. 5. Vertical (dy) (A), horizontal (dx) (B) and lateral (dz) (C) potential and kinetic energy at faster gaits (galloping) and the
displacements of the centre of mass of the horses body during a very high aerobic power output (Taylor et al., 1981) are the keys
single stride at different gaits and speeds. Symbols as for Fig. 4. to the athletic performance of horses. The structural design of
the limbs and the actuator architecture, based on long tendons
coupled with short, thick muscles (Dimery et al., 1986), reflects
hypothesis during trotting (WINT is lower than during walking the development of fast gaits at reduced metabolic cost. The
by 19 %) and, to an even greater extent (37 % lower) during efficiency of locomotion can be calculated as:
galloping. Such a reduction in WINT, mostly deriving from the
WEXT + WINT
change in limb geometry during the swing phase, could also effAPP = . (1)
be the consequence of the evolution of limb structures with the C
centre of mass located very proximally with respect to the The efficiency values shown in Fig. 7A range from
pivoting point (as noted by DArcy Thompson, 1961). approximately 10 % (for walking) to more than 100 %
Furthermore, the model (Minetti, 1998b) shows that WINT (galloping). [Note that these values, obtained by disregarding
depends strongly on speed, stride frequency and duty factor an amount of negative work equal to the positive work in level
(i.e. the fraction of the stride for which one foot is in contact periodic movements, actually approximate the efficiency of
with the ground). Since the relationship between the last two positive work. It can be shown, by assuming that the efficiency
variables and speed is different for walking, trotting and of generating negative work is five times greater than that of
galloping, the same applies to WINT. While WINT appears to be generating positive work (Abbott et al., 1952), that the true
a minor contributor to the total mechanical work at high speed positive work efficiency values are underestimated by 17 %.]
(see Fig. 4), the removal of the elastic component from the While Taylor (1994) suggested that this could be the case by
measured external work is expected to make it worthwhile to extrapolating from mechanical data for different smaller
reduce WINT. The contributions of single limbs to the estimated animals, the present values for galloping are, to the authors
WINT may change from one gait to another. In fact, in knowledge, the highest ever measured for animal locomotion.
symmetrical gaits (walking and trotting), all limbs contribute In contrast to walking, the efficiencies of trotting and galloping
equally to the overall WINT, while during galloping the first- far exceed the value usually obtained from the thermodynamics
down hind limb (also called the leading rear) exhibits a greater of muscular contraction (2535 %; Woledge et al., 1985). This
WINT (2329 %) than the other limbs, giving it a predominant is the reason why such efficiency is sometimes termed
role in propulsion. apparent efficiency (effAPP; Full, 1991). Among the terms
 Mechanics and energetics of horse locomotion 2335
120 Woledge et al., 1985), the elastic contribution to the total
A mechanical work, as measured, can be calculated. In addition,
100 by considering that trotting and cantering (slow galloping)
G differ in the number of bounces per stride (nBOUNCE is 2 for
Efficiency, effAPP (%)

trotting and 1 for galloping), the elastic work per bounce

80 (WEL,BOUNCE) can be obtained:
T v
60 WEL,BOUNCE =WEL , (3)
f nBOUNCE

40 where v is speed and f is the stride frquency. Alexander and his

co-workers (Alexander, 1988b; Ker et al., 1987) estimated that
the amount of elastic energy stored during a running step was
20 W
35 J for the Achilles tendon and 17 J for the arch of the foot
(a total of 52 J for a mean body mass of 70 kg, or
0
0 2 4 6 8 10
0.75 J kg1 bounce1). The values for horses, computed using
equation 3 and the regression equations given in Table 1, have
8 been plotted in relation to speed in Fig. 7B as continuous lines.
B WEL,BOUNCE is negligible during walking, as expected, but is
Elastic energy (J kg-1 bounce-1)

remarkably high during trotting and galloping. Values for

6
trotting can be compared with human running, the
corresponding gait for bipeds. The 64 % higher WEL,BOUNCE
4 during trotting (1.23 J kg1 bounce1), however, turns out to be
more similar (18 %) to the values for human running when the
elastic work per bounce and per limb is considered. The highest
2 value of WEL,BOUNCE is for galloping, 6 J kg1 bounce1, and the
curve suggests a further increase with speed.
It has been proposed that additional storage and release of
0 elastic energy during galloping (but not during trotting) could
occur in the musculo-ligamentous apparatus along the spine,
0 2 4 6 8 10 which extends (storage) during the foot-contact time and recoils
Speed (m s-1) (release) during the flight period (Alexander, 1988a). It has been
postulated that this elastic energy saving affects only WINT. The
Fig. 7. (A) relationship between apparent efficiency (effAPP) and
energy stored in the spine during the contact phase would be
speed obtained from the present measurements and equation 1; the
shaded area represents the efficiency range of concentric muscle available for regrouping the limbs under the body during the
contraction as reported in the literature (Woledge et al., 1985). flight period. Alexander (1988a) estimated that the amount of
(B) The elastic energy stored and released during a single bounce energy released was 44 J stride1 in a 50 kg deer (Dama dama)
(trotting, two bounces per stride; galloping, one bounce per stride) or 0.88 J kg1 bounce1. In the lower graph of Fig. 7B, the
obtained using equation 3; the dashed line represents the elastic WEL,BOUNCE attributable to limb elastic structures for a
energy for galloping when the effect of spine bending on internal galloping horse (dashed line) has been calculated according to
work (WINT) is taken into account (see Discussion for details). this energy-saving strategy. To provide evidence for the
G, galloping; T, trotting; W, walking. bending of the spine at different speeds and gaits, we positioned
seven markers on the skin corresponding to the spinous
involved in equation 1, WEXT is the most likely to be inaccurate processes of vertebrae and analysed their three-dimensional
(i.e. overestimated with respect to the positive work actually motion in one horse walking at 1.41 m s1, trotting at 3.49 m s1
done by muscles, as already mentioned). We will attempt, in and galloping at three speeds (6.36, 7.80 and 9.17 m s1). Fig. 8
the following, to calculate the elastic energy (stored and) shows the time course of the vertical coordinates at all these
released during each bounce for the different gaits and speeds. speeds. It is apparent that the spine translates vertically without
The following equation: changing its shape during trotting, whereas the out-of-phase
waveforms suggest a bendingrecoil sequence within each
(WEXT WEL) + WINT
effMUS = , (2) stride during galloping. Moreover, the relationships among
C waveforms seem not to be affected by galloping speed. It can
where WEL is the elastic work (J kg1 m1) released by the limb be therefore be hypothesised that the energy saving per stride
tendons, is similar to equation 1 and deals with the muscle by this mechanism depends on stride frequency alone.
efficiency (effMUS) of transforming metabolic work into fresh The values of WEL,BOUNCE obtained in this study are
(rather than previously stored) mechanical work. Equation 2 expected to be quite realistic because the calculation has been
tells us that, by assuming effMUS to be constant (=0.30; derived from experimental measurements. However, the
2336 A. E. MINETTI AND OTHERS
2.0
Walk 1.41 m s-1 Trot 3.49 m s-1
1.9
1 2 3 4 5 6 7 8
1.8

1.7
Vertical position (m)

1.6

1.5
2.0
Gallop 7.80 m s-1 Gallop 9.17 m s-1

1.9

1.8

1.7

1.6

Gallop 6.36 m s1
1.5
0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0
Time (s)

Fig. 8. Vertical positions of seven markers located along the spine with respect to time. The lowest two speeds refer to walking and trotting,
while the other three pertain to galloping. The phase shift between the markers during galloping indicates bending of the spine.

assumption that effMUS is constant at the different speeds and Taylor, 1990) and in the gait choice (Farley and Taylor, 1991)
gaits is questionable, although previous studies on human have been recently pointed out. Experimental findings on
walking and running at different speeds and with different running turkeys (Roberts et al., 1997) suggest that extensor
gradients (Minetti et al., 1993, 1994; Minetti, 1995) succeeded muscles shorten very little during the stance phase of bouncing
in matching mechanics to energetics using such a gaits, whilst the elastic structures (tendons) stretch and recoil.
simplification. It should be noted that WEL as obtained from In this view, the muscles should contract almost isometrically
equation 2 represents the lowest estimate of the elastic and do little work during stance. If this were true, a remarkable
contribution to locomotion, since it would be higher if the proportion of the metabolic cost might be set by the cost
muscles were to operate at a lower than optimal efficiency. associated with the generation of tension. According to this
There is an unusual relationship between the apparent hypothesis, Kram and Taylor (1990) introduced the reciprocal
efficiency (effAP, Fig. 7A) and the vertical excursion of the of the stride length, defined as the distance travelled by the
centre of mass during the stride (dy; Fig. 5A). The effAP centre of mass during the support phase (their equation 4), as
maxima for trotting and galloping correspond to the speeds at a predictor of the cost of transport (corresponding to C in the
which dy is also maximal. This suggests that, at each gait, a present paper). We calculated and plotted this predictor for
resonant frequency (and thus a resonant speed) exists at trotting and galloping (Fig. 9). When comparing this graph
which the greatest elastic contribution (dy; i.e. the compression with a plot of C versus speed (Fig. 2), it emerges that the
of the spring) is obtained via the minimum forced-oscillation shapes of the curves are different, with (stride length)1
effort. decreasing monotonically with speed (Fig. 9) and C being
approximately constant for all gaits (Fig. 2). Conversely, the
The force hypothesis net mechanical work, calculated by removing from the
When comparing mechanical work and metabolic cost, measured mechanical work the estimated elastic energy
another issue should be considered. Muscles utilise metabolic released, increases with speed for each gait. It is possible,
fuel both for generating work and for maintaining force (with therefore, that a combination of the cost of generating force
no shortening, thus with no work). The roles of the force and the cost of producing mechanical work is necessary to
exerted in energy expenditure during running (Kram and explain the energy expenditure of locomotion in horses.
 Mechanics and energetics of horse locomotion 2337
1.2 physiological cost of negative work. J. Physiol., Lond. 117,
380390.
Alexander, R. McN. (1988a). Why mammals gallop. Am. Zool. 28,
Reciprocal stride length (m-1)
1.0
237245.
Alexander, R. McN. (1988b). Elastic Mechanisms in Animal
0.8 Movement. Cambridge: Cambridge University Press.
Alexander, R. McN. (1989). Optimization and gaits in the
0.6 locomotion of vertebrates. Physiol. Rev. 69, 11991227.
Buchner, H. H. F., Savelberg, H. H. C. M., Schamhardt, H. C. and
Barneveld, A. (1997). Inertial properties of Dutch Warmblood
0.4
horses. J. Biomech. 30, 653658.
Butler, P. J., Woakes, A. J., Smale, K., Roberts, C. A., Hillidge,
0.2 C. J., Snow, D. H. and Marlin, D. J. (1993). Respiratory and
cardiovascular adjustments during exercise of increasing intensity
0 and during recovery in Thoroughbred racehorses. J. Exp. Biol. 179,
0 4 8 12 159180.
Speed (m s-1) Cavagna, G. A., Heglund, N. C. and Taylor, C. R. (1977).
Mechanical work in terrestrial locomotion: two basic mechanisms for
Fig. 9. A graph showing the reciprocal of the stride length, a minimizing energy expenditure. Am. J. Physiol. 233, R243R261.
predictor of the cost of locomotion (see Kram and Taylor, 1990), Cavagna, G. A. and Kaneko, M. (1977). Mechanical work and
plotted against speed for trotting and galloping. Filled symbols efficiency in level walking and running. J. Physiol., Lond. 268,
represent trotting, open symbols represent galloping; the different 467481.
symbol shapes indicate the individual horses. Cavagna, G. A., Saibene, F. P. and Margaria, R. (1964).
Mechanical work in running. J. Appl. Physiol. 19, 249256.
Cavagna, G. A., Thys, H. and Zamboni, A. (1976). The sources of
In conclusion, horses probably represent the biggest external work in level walking and running. J. Physiol., Lond. 262,
quadrupeds for which metabolic and mechanical 639657.
measurements of locomotion can be made in a controlled DAmico, M. and Ferrigno, G. (1990). Technique for the evaluation
environment (constant-speed treadmill). The relationship of derivatives from noisy biomechanical displacement data using a
model based bandwidth-selection procedure. Med. Biol. Eng.
between these variables, which are associated with athletic
Comp. 28, 407415.
performance, sheds light on the role of the different gaits
DArcy Thompson, W. (1961). On Growth and Form. Cambridge:
(walking and galloping are the most common gaits in Cambridge University Press.
quadrupeds, while trotting is optional in some cases; Dimery, N. J., Alexander, R.McN. and Ker, R. F. (1986). Elastic
Pennycuick, 1975). extension of the leg tendons in the locomotion of horses (Equus
An apparent efficiency of locomotion exceeding 100 % at caballus). J. Zool., Lond., 205, 415425.
high speeds, the highest value reported to date, implies that a Eaton, M. D., Evans, D. L., Hodgson, D. R. and Reuben, J. R.
huge amount of elastic energy is continuously stored and (1995). Effect of treadmill incline and speed on metabolic rate during
released to maintain the body in motion. Additional elastic exercise in Thoroughbred horses. J. Appl. Physiol. 79, 951957.
structures, such as the ligaments in the spine, a highly Farley, C. T. and Taylor, C. R. (1991). A mechanical trigger for the
specialised limb design and the concurrent use of pendulum- trotgallop transition in horses. Science 253, 306308.
Fedak, M. A., Heglund, N. C. and Taylor, C. R. (1981). Energetics
like mechanics at high speeds, are responsible of the athletic
and mechanics of terrestrial locomotion. II. Kinetic energy changes
performance of horses. The combined energy-saving strategy
of the limbs and body as a function of speed and body size in birds
adopted during the gallop can potentially explain the and mammals. J. Exp. Biol. 97, 2340.
evolutionary advantage of introducing such a gait and deserves Full, R. J. (1991). The concepts of efficiency and economy in land
further investigation. locomotion. In Efficiency and Economy in Animal Physiology (ed.
E. W. Blake), pp. 97131, Cambridge: Cambridge University Press.
The authors wish to thank Colonel M. Reitano and his staff Heglund, N. C., Cavagna, G. A. and Taylor, C. R. (1982).
at the Scuola di Cavalleria, Italian Army, Montelibretti, Energetics and mechanics of terrestrial locomotion. III. Energy
Roma, for kindly sharing the exercise laboratory and changes of the centre of mass as a function of speed and body size
providing the horses. B.T.S. (Milano, Italy) is acknowledged in birds and mammals. J. Exp. Biol. 97, 4156.
Hildebrand, M. (1994). Analysis of Vertebrate Structure, 4th edn.
for helping in the calibration of the motion-analysis apparatus.
New York: Wiley.
We are also indebted to Professor Gabriele Cortili (Universit
Hoyt, D. F. and Taylor, C. R. (1981). Gait and the energetics of
Studi, Milano) who contributed improvements to the locomotion in horses. Nature 292, 239240.
metabolic apparatus and provided logistic support. Ker, R. F., Bennet, M. B., Bibby, S. R., Kester, R. C. and
Alexander, R. McN. (1987). The spring in the arch of the human
foot. Nature 325, 147149.
References Kram, R. and Taylor, C. R. (1990). Energetics of running: a new
Abbott, B. C., Bigland, B. and Ritchie, J. M. (1952). The perspective. Nature 346, 265266.
2338 A. E. MINETTI AND OTHERS
Langsetmo, I., Weigle, G. E., Fedde, M.. R., Erickson, H. H., Roberts, T. J., Marsh, E. L., Weyand, P. G. and Taylor, C. R.
Bartsow, T. J. and Poole, D. C. (1997). VO kinetics in the horse (1997). Muscular force in running turkeys: the economy of
during moderate and heavy exercise. J. Appl. Physiol. 83, minimizing work. Science 275, 11131115.
12351241. Schmidt-Nielsen, K. (1984). Scaling: Why is Animal Size so
Minetti, A. E. (1995). Optimum gradient of mountain paths. J. Appl. Important? Cambridge: Cambridge University Press.
Physiol. 79, 16981703. Taylor, C. R. (1994). Relating mechanics and energetics during
Minetti, A. E. (1998a). The biomechanics of skipping gaits: a third exercise. Adv. Vet. Sci. Comp. Med. 38A, 181215.
locomotion paradigm? Proc. R. Soc. Lond. B 265, 12271235. Taylor, C. R., Heglund, N. C. and Maloiy, G. M. O. (1982).
Minetti, A. E. (1998b). A model equation for the prediction of Energetics and mechanics of terrestrial locomotion. I. Metabolic
mechanical internal work of terrestrial locomotion. J. Biomech. 31, energy consumption as a function of speed and body size in birds
463468. and mammals. J. Exp. Biol. 97, 121.
Minetti, A. E., Ardig, L. P. and Saibene, F. (1993). Mechanical Taylor, C. R., Maloiy, G. M. O., Weibel, E. R., Langman, V. A.,
determinants of gradient walking energetics in man. J. Physiol., Kamau, J. M. Z., Seeherman, H. J. and Heglund, N. C. (1981).
Lond. 472, 725736. Design of the mammalian respiratory system. III. Scaling
Minetti, A. E., Ardig, L. P. and Saibene, F. (1994). Mechanical maximum aerobic capacity to body mass: wild and domestic
determinants of the minimum energy cost of gradient running. J. mammals. Respir. Physiol. 44, 2537.
Exp. Biol. 195, 211225. Wagner, P. D., Gillespie, J. R., Landgren, G. L., Fedde, M. R.,
Pennycuick, C. J. (1975). On the running of the gnu (Connochaetes Jones, B. W., Debowes, R. M., Pieschl, R. L. and Erikson, H.
taurinus) and other animals. J. Exp. Biol. 63, 775799. H. (1989). Mechanism of exercise-induced hypoxemia in horses. J.
Potard, U. S., Leith, D. E. and Fedde, M. R. (1998). Force, speed Appl. Physiol. 66, 12271233.
and oxygen consumption in Thoroughbred and draft horses. J. Woledge, R. C., Curtin, N. A. and Homsher, E. (1985). Energetic
Appl. Physiol. 84, 20522059. Aspects of Muscle Contraction. London: Academic Press.