Skip to main content
NCBI home page
Journal of Anatomy logoLink to Journal of Anatomy
. 2014 Jun 10;225(3):317–327. doi: 10.1111/joa.12208

Comparative functional anatomy of the epaxial musculature of dogs (Canis familiaris) bred for sprinting vs. fighting

Emma L Webster 1, Penny E Hudson 2, Sarah B Channon 3,✉
PMCID: PMC4166972  PMID: 24917310

Abstract

The axial musculoskeletal system of quadrupedal mammals is not currently well understood despite its functional importance in terms of facilitating postural stability and locomotion. Here we examined the detailed architecture of the muscles of the vertebral column of two breeds of dog, the Staffordshire bull terrier (SBT) and the racing greyhound, which have been selectively bred for physical combat and high speed sprint performance, respectively. Dissections of the epaxial musculature of nine racing greyhounds and six SBTs were carried out; muscle mass, length, and fascicle lengths were measured and used to calculate muscle physiological cross-sectional area (PCSA), and to estimate maximum muscle potential for force, work and power production. The longissimus dorsi muscle was found to have a high propensity for force production in both breeds of dog; however, when considered in combination with the iliocostalis lumborum muscle it showed enhanced potential for production of power and facilitating spinal extension during galloping gaits. This was particularly the case in the greyhound, where the m. longissimus dorsi and the m. iliocostalis lumborum were estimated to have the potential to augment hindlimb muscle power by ca. 12%. Breed differences were found within various other muscles of the axial musculoskeletal system, particularly in the cranial cervical muscles and also the deep muscles of the thorax which insert on the ribs. These may also highlight key functional adaptations between the two breeds of dog, which have been selectively bred for particular purposes. Additionally, in both breeds of dog, we illustrate specialisation of muscle function by spinal region, with differences in both mass and PCSA found between muscles at varying levels of the axial musculoskeletal system, and between muscle functional groups.

Keywords: axial, back, Canis, greyhound, locomotion, muscle, vertebral column

Introduction

The epaxial spinal muscles have been proposed as important during galloping in mammals, to facilitate coordinated limb movements, and to lengthen the stride during the extended aerial phase (Hildebrand, 1961). Studies of the electromyographic activity of these muscles suggest that their function is somewhat broader than that. Collectively, in addition to mobilising the trunk in extension, torsion and lateral bending, the epaxial muscles are also thought to play a role in controlling or resisting movements of the spine (dynamic stabilisation) as well as ensuring the structural integrity of the vertebral column (local/static stabilisation; English, 1980; Carlson et al. 1979; Ritter et al. 2001; Schilling & Carrier, 2009).

The three main epaxial muscle systems are, from most medial to lateral, the transversospinalis, longissimus and iliocostalis systems (Evans, 1993; Fig. 1). The transversospinalis system includes the m. spinalis, the m. semispinalis, and the deep multisegmental muscles such as the mm. multifidi, which are considered, based on their activity patterns and fibre metabolic profile, to be equipped for both dynamic spinal stabilisation, as well as producing spinal extension (Carlson et al. 1979; English, 1980; Ritter et al. 2001; Schilling, 2009; Schilling & Carrier, 2009). Similarly, the longissimus and iliocostalis systems (which are not easily distinguished from one another in the lumbar region of the dog) are also considered to have multiple roles. They have been implicated in increasing spinal stiffness (Carlson et al. 1979; English, 1980; Schilling & Carrier, 2009), counteracting sagittal flexion (Ritter et al. 2001) and in active generation of spinal extension during asymmetric gaits (English, 1980; Schilling & Carrier, 2010) such as the rotary gallop used at high speed by the greyhound. Studies suggest that vertebral level (Wakeling et al. 2007; Schilling & Carrier, 2009) as well as the gait utilised by the animal (English, 1980; Ritter et al. 2001; Schilling & Carrier, 2010) heavily influence the functional role of the various epaxial muscles.

Figure 1.

Figure 1

Open in a new tab

Diagrammatic representation of muscles of the vertebral column coloured by muscle group: red – longissimus; blue – transversospinalis; green – iliocostalis; orange – other.

As the function of the spine and spinal musculature is not consistent along its length, it follows that musculoskeletal anatomy may also vary accordingly. This morphological diversity is well illustrated by variation in the structure of vertebrae across mammals (e.g. Slijper, 1946; Kida et al. 1999) and also in non-mammalian vertebrates (e.g. Long et al. 1997). Studies of mammalian muscle features such as fibre metabolic profiles also indicate a cranial to caudal differentiation in fibre type distribution along the vertebral column (Moritz et al. 2007; Schilling, 2009) as well as within individual epaxial muscles (Gellman et al. 2002). We may therefore also expect regional variation in the quantitative morphology of muscles along the trunk in quadrupedal mammals. Such regionalisation of muscle architectural characteristics has been illustrated within a single epaxial muscle – the m. longissimus dorsi – in the horse (Ritruechai et al. 2008); however, whether regional specialisation in muscle architecture exists in other epaxial muscles and in other mammals is yet to be addressed.

There is a general paucity of quantitative comparative information regarding the anatomical structure of the muscles of the trunk of dogs and of quadrupeds in general. To date, only one study has considered muscle architecture in the canine back (Sharir et al. 2006), considering only the neck muscles of mixed-breed dogs. The aim of this study was therefore to provide detailed quantitative and comparative data regarding muscle functional anatomy throughout the axial musculoskeletal system of two breeds of dog bred for contrasting purposes: sprinting vs. fighting.

The racing greyhound is an elite sprinting athlete that has been under intense artificial selection for both high speed running and aerobic stamina. It has been shown that these dogs can often reach speeds of 17 m s−1 for short periods during sprinting (Usherwood & Wilson, 2005), and can accelerate at 10 m s−2 (Williams et al. 2009a). Laboratory experiments have shown sub-maximally accelerating greyhounds to produce total limb peak powers of ca. 300 W kg−1 of hindlimb extensor muscle mass (Williams et al. 2009b), and greyhounds during racing to have centre of mass (CoM) powers (over a stride) as high as 60 W kg−1 (body mass specific; Williams et al. 2009b). These figures are towards the top end of published values for potential maximum muscle power output in mammals (Weis-Fogh & Alexander, 1977) and in the company of animals which show particularly impressive values for maximum muscle power, such as ascending quail (Askew et al. 2001).

To achieve such impressive power production, greyhounds might possess superior physiological properties of limb muscles. Greyhounds have a higher proportion of fast muscle fibres than other breeds of dog (Rodriguez-Barbudo et al. 1984). However, even the fastest land mammal the cheetah does not have muscles with extraordinary physiological properties (92.5 W kg−1 muscle compared with 119.7 W kg−1 in the rabbit; West et al. 2013). It therefore follows that greyhounds may also possess a multitude of other strategies to enhance power and enable rapid accelerations. Elastic elements within the limbs may act to augment or amplify power production by their muscles (Peplowski & Marsh, 1997; Lichtwark & Wilson, 2005), with such a mechanism proposed to amplify power by up to 15 times (Aerts, 1998). A further and likely possibility is that the greyhound spinal musculature performs mechanical work on the CoM during locomotion, thus supplementing that produced by the limb musculature.

In contrast to greyhounds, Staffordshire bull terriers (SBTs) have historically been subject to selection for bloodsports and physical combat and as such, dogs of this type possess substantial specialisations for fighting. Clear adaptations have previously been noted in pit bull terrier limb anatomy; they exhibit relatively more muscle mass in their distal limbs compared with greyhounds, have stronger muscles in their forelimbs than their hindlimbs and a much smaller capacity for elastic storage in the extensor muscle-tendon systems of their ankle joints (Pasi & Carrier, 2003). The limb bones of pit bull terriers also appear well adapted for fighting, demonstrating high resistance of failure compared with the relatively stiff and brittle bones of greyhounds (Kemp et al. 2005). It follows that the spinal musculature of SBT dogs may also be adapted for the enhanced strength and dynamic stability that might be required for physical combat.

We hypothesised that whereas the SBT would exhibit adaptations for high muscle strength and postural stability, greyhound epaxial muscles would show a greater specialisation for work and power production and for facilitating a wide range of spinal motion. Additionally, we hypothesised that both breeds may exhibit regional specialisation of muscle function throughout the axial musculoskeletal system, with this being reflected by differences in muscle architecture between regions of the vertebral column and between different functional muscle groups.

Materials and methods

This study was approved by the University of Liverpool Veterinary Research Ethics Committee (Reference number VREC104). Nine mature greyhounds (mass 27.3 ± 1.7 kg; mean ± SD), euthanised due to traumatic racing injury to a limb, and six mature Staffordshire bull terriers (SBT; 21.4 ± 1.6 kg), euthanized for reasons not pertaining to the musculoskeletal system, were used for the study. Cadavers were frozen at −20 °C within 24 h of euthanasia and removed from the freezer and stored at 4 °C no more than 48 h prior to dissection.

For each dissection, the lefthand side of the cadaver was skinned, and the extrinsic muscles of the forelimb removed. Muscles of the vertebral column were systematically isolated and any external tendons associated with the muscle removed. Muscle mass was determined using an electronic balance accurate to 0.01 g (Mettler Toledo, Greiffensee, Switzerland), and muscle belly length was measured from origin to insertion using a flexible plastic tape measure accurate to 1 mm. Where a muscle had multiple insertions, this measurement was made to the insertion point furthest from the origin, ensuring that muscle belly length reflected the entire line of action of the muscle. A longitudinal incision was made in the muscle belly along the plane of the fascicles, and the lengths of a minimum of five, and up to 10, randomly selected fascicles were measured using a plastic measuring tape. In the case of large muscles, spanning multiple spinal segments, fascicle lengths were sampled along the full length of the muscle belly.

Muscle volume was calculated by dividing muscle mass by muscle density of 1.06 g cm−3 (Mendez & Keys, 1960). Physiological cross-sectional area (PCSA) was calculated for each muscle as muscle volume divided by mean fascicle length. The Architectural Index (AI), an index normalising fascicle length for muscle length, was calculated for each muscle to allow comparisons of fascicle lengths to be made between muscles of varying sizes. It was calculated as mean fascicle length divided by muscle belly length.

Resulting individual data were scaled in accordance with geometric scaling, with muscle mass scaled in proportion to body mass (g kg−1), PCSA scaled to body mass2/3 and muscle belly and fascicle lengths scaled to body mass1/3 (Alexander et al. 1981; Schmidt-Nielsen, 1984). The scaling method utilised in comparative anatomical studies is an important consideration (see Myatt et al. 2011 for a full discussion). To ensure that geometric scaling was the appropriate method for this study, an independent samples t-test (significance level of 0.05) was carried out to establish that body mass differed between the two breeds of dog (P < 0.0001). Body mass and muscle parameters (mass, PCSA and fascicle length) were log-transformed and linear regression carried out to establish the scaling relationships (exponents) for each parameter. These were not found to be different from geometric scaling relationships (P > 0.05 for all parameters) and therefore, given the well recognised sensitivity of allometric scaling methods to low sample sizes (Hui & Jackson, 2007), it was considered most appropriate to base subsequent analyses on the established geometric scaling relationships.

Scaled muscle architecture variables were compared between greyhound and SBT breeds via independent sample Student's t-tests, with a significance level set at 0.05. Muscles were grouped for further analysis based on two features: (i) position along the spine from cranial to caudal (based on location of cranial-most muscle origin: cervical, thoracic or lumbar) and (ii) muscle system [transversospinalis, iliocostalis, longissimus, and other muscles (predominantly of the neck)]. Data were evaluated for normality using a Shapiro–Wilk test and found to be normally distributed (P > 0.05). One-way analyses of variance (anova) followed by Tukey's multiple comparisons tests were conducted to determine differences between muscle variables at regions of the vertebral column, and between functional systems of muscles for each of the greyhound and SBT, respectively. All statistical procedures were undertaken in graphpad prism (Version 6.00 for Windows; GraphPad Software, La Jolla, CA, USA).

Results

Thirteen muscles were considered in this study (see Fig. 1 for a representation of their locations along the spine). Muscle architecture parameters for each muscle in both greyhounds and SBTs are provided in Table 1.

Table 1.

Quantitative muscle data (unscaled) from the epaxial musculature of the greyhound (GH) and Staffordshire bull terrier (SBT).

Muscle Abbv. Mass (g) Length (mm) Fascicle length (mm) PCSA (mm2)
GH SBT GH SBT GH SBT GH SBT
Obliquus capitis cranialis OCCr 9.4 (4.9) 7.8 (1.2)* 97 (2) 70 (3)* 89 (6) 54 (10) 102 (59) 129 (17)*
Obliquus capitis caudalis OCCa 32.2 (1.4) 34.8 (4.7)* 81 (24) 72 (5) 72 (31) 50 (2) 463 (181) 608 (42)*
Splenius Sp 124.8 (16.2) 87.6 (10.6) 342 (28) 249 (13)* 227 (64) 188 (8) 550 (156) 442 (64)
Semispinalis capitis SSC 86.9 (13.8) 103.8 (13.4)* 366 (30) 238 (12)* 50 (12) 84 (25)* 1715 (391) 1577 (330)
Longissimus capitis LCa 24.9 (5.3) 21.5 (3.8) 270 (36) 187 (21)* 174 (68) 55 (9)* 166 (95) 382 (110)*
Longissimus cervicis LCe 53.0 (9.8) 37.1 (7.5) 252 (33) 161 (18)* 110 (37) 87 (6) 504 (205) 397 (64)
Spinalis et semispinalis cervicis SetSC 33.5 (22.4) 14.4 (3.2) 222 (54) 140 (24)* 118 (66) 103 (27) 441 (514) 146 (69)
Serratus dorsalis SD 42.4 (11.3) 22.2 (2.3)* 222 (80) 188 (30) 80 (25) 61 (1) 501 (191) 351 (44)
Scalenus Sc 36.4 (4.4) 24.5 (2.9)* 273 (24) 169 (15)* 171 (28) 128 (12)* 206 (50) 190 (44)
Spinalis et semispinalis thoracis SetST 67.6 (11.3) 48.6 (7.0) 271 (16) 202 (38)* 86 (22) 92 (14)* 802 (293) 506 (105)
Multifidus thoracis MT 62.9 (32.3) 32.2 (6.1)* 304 (67) 200 (23)* 49 (15) 27 (17) 1346 (920) 1178 (446)
Iliocostalis thoracis ILT 35.2 (6.7) 25.3 (2.3) 332 (41) 273 (50)* 44 (10) 69 (20)* 778 (149) 358 (88)*
Longissimus dorsi LD 680.6 (218.3) 272.8 (39.3)* 652 (92) 422 (20)* 89 (28) 68 (15) 8139 (3700) 3923 (1080)*
Iliocostalis lumborum ILL 37.9 (17.4) 113.2 (7.2)* 335 (54) 209 (9)* 184 (84) 93 (3)* 209 (101) 1178 (104)*
Open in a new tab

Values indicate means. Values in brackets indicate standard deviation.

*

Significant difference after scaling for body mass (P < 0.05).

Breed differences

The longissimus thoracis et lumborum muscle (the thoracic and lumbar portions of which are fused at around the level of the 13th rib, and hereafter referred to as m. longissimus dorsi) was the largest muscle in the axial musculoskeletal system of both breeds (Fig. 2A; greyhound 681 ± 218 g and SBT 273 ± 39 g). The longissimus dorsi muscle was over five times the mass of the next largest muscle in the greyhound (m. splenius; 125 ± 16 g) and accounted for over 50% of the mass of the epaxial musculature. In the SBT, the m. longissimus dorsi contributed more than twice the mass of the next largest muscle, the m. iliocostalis lumborum (113 ± 7 g). There was a significant difference in the mass of the m. longissimus dorsi between the two breeds of dog, after scaling for body mass, with the greyhound found to have a significantly heavier longissimus dorsi muscle (P = 0.001). This muscle was found in both breeds of dog to be fused with the iliocostalis lumborum muscle caudally; the iliocostalis lumborum muscle was significantly larger in the SBT (P < 0.0001). However, combined with the mass of the m. longissimus dorsi, the total muscle mass was found to be larger in the greyhound (P = 0.018). Significant differences in mass of other muscles existed between breeds of dog (Fig. 2A); a number of muscles in the cervical region, namely the m. obliquus capitis cranialis, m. obliquus capitis caudalis, and m. semispinalis capitis were larger in the SBT (P = 0.004, < 0.0001 and < 0.0001, respectively), and a number of muscles in the thoracic region of the spine, the m. serratus dorsalis, m. scalenus and m. multifidus thoracis, were larger in the greyhound (P = 0.001, 0.009 and 0.005).

Figure 2.

Figure 2

Open in a new tab

(A) Scaled muscle mass. (B) Scaled PCSA (physiological cross-sectional area). (C) Scaled fascicle length. (D) Architectural Index for muscles of the vertebral column. Bars represent mean ± 95% confidence intervals for Staffordshire bull terriers (dark red; n = 6) and greyhounds (red; n = 9). Muscles are arranged from cranial to caudal along the vertebral column. Significant differences (P < 0.05) between dog breeds are indicated by asterisks.

The scaled PCSA of some muscles in the neck region was considerably larger in the SBT than in the greyhound (m. obliquus capitis cranialis, m. obliquus capitis caudalis; P = 0.003 and 0.004; Fig. 2B) as was the iliocostalis lumborum muscle (P < 0.0001). The PCSA of the m. iliocostalis thoracis, m. longissimus dorsi, and the combined PCSA of these two muscles were all found to be greater in the greyhound than in the SBT (P = 0.0002, < 0.0001 and < 0.0001). This difference was particularly large in the longissimus dorsi muscle alone and also when this muscle was paired with the m. iliocostalis lumborum (58 and 43% greater in the greyhound, respectively).

Despite the inherent variability of fascicle length within a muscle, the mean scaled fascicle length was found to differ significantly between breeds in a number of muscles (Fig. 2C). In the majority of cases (m. longissimus capitis, m. scalenus, m. spinalis et semispinalis thoracis, m. multifidus thoracis, m. iliocostalis thoracis, m. iliocostalis lumborum and the combined iliocostalis lumborum and longissimus dorsi muscle complex), fascicle lengths were found to be longer in the greyhound than in the SBT (P < 0.05 in all cases), with only the semispinalis capitis muscle exhibiting the opposite trend (P = 0.018). AI was found to be larger in the m. semispinalis capitis, m. spinalis et semispinalis thoracis and m. longissimus dorsi in SBTs (P = 0.0004, 0.02, and 0.049; Fig. 2D), and larger in the m. longissimus capitis and the combined longissimus dorsi – iliocostalis lumborum muscle complex in greyhounds (P = 0.005 and 0.03).

Regional variation

Both the greyhound and SBT showed regional variation in the location of muscle mass along the vertebral column (Figs 3 and 4). Both breeds had a higher proportion of muscle mass located in the lumbar region, compared with both the cervical and thoracic regions of the spine (P < 0.0001). Unlike the greyhound, however, the SBT had a larger amount of muscle mass located in the cervical region than in the thoracic region (P < 0.0001). This trend was mirrored by differences in PCSA between regions. Neither breed showed significant differences in either fascicle length or AI across regions of the vertebral column; however, there was a strong trend for the AI of the epaxial muscles to decrease caudally along the spine in the SBT in particular.

Figure 3.

Figure 3

Open in a new tab

Functional regionalisation of greyhound muscle architecture by region of the vertebral column (cervical – dark, thoracic – pale and lumbar – mid) and by muscle functional group (red – longissimus; blue – transversospinalis; green – iliocostalis; orange – other) for (A) scaled muscle mass, (B) scaled PCSA (physiological cross-sectional area), (C) scaled fascicle length and (D) Architectural Index. Bars represent mean ± SE.

Figure 4.

Figure 4

Open in a new tab

Functional regionalisation of Staffordshire bull terrier muscle architecture by region of the vertebral column (cervical – dark, thoracic – pale and lumbar – mid) and by muscle functional group (red – longissimus; blue – transversospinalis; green – iliocostalis; orange – other) for (A) scaled muscle mass, (B) scaled PCSA (physiological cross-sectional area), (C) scaled fascicle length and (D) Architectural Index. Bars represent mean ± SE.

The longissimus muscle system was substantially higher in mass and PCSA in both breeds of dog than the other respective muscle systems (P < 0.0001; Figs 3 and 4). Additionally, in both breeds, the transversospinalis system was larger in terms of mass and PCSA than the iliocostalis muscle system (P < 0.0001), and in the SBT also greater than the muscles categorised as ‘other’. No differences were seen between the respective muscle systems in terms of fascicle length or AI (P > 0.05) in either breed of dog.

Discussion

Muscle functional capacity

Measures of muscle architecture can be utilised to make predictions about a muscle's maximum functional capacity in terms of force production, performing work and generating power (e.g. Lieber & Friden, 2000; Payne et al. 2005; Smith et al. 2006; Williams et al. 2008a,b2008b; Hudson et al. 2011). Muscle architectural measurements alone cannot generate accurate figures for these parameters; such predictions require additional data, or assumptions to be made regarding muscle physiological properties such as contraction velocity. It is possible, however, using basic analytical and graphical procedures, to ascertain broad trends and relationships regarding maximal muscle functional capacity and specialisation within and between species and breeds. Figure 5A illustrates this concept by plotting scaled muscle PCSA against scaled fascicle length for both greyhound and SBT back muscles. Muscles located to the top left of this plot show high maximum capacity for force generation; those to the bottom right show specialisations for performing work across a wide range of motion, and those towards the centre/upper-right of the plot have a high maximum capacity for power production due to their large size and relatively long fascicles.

Figure 5.

Figure 5

Open in a new tab

(A) Scaled mean fascicle length against scaled physiological cross-sectional area (PCSA) for epaxial muscles of greyhounds (circular symbols) and Staffordshire bull terriers (triangular symbols). Bars represent SEM. Muscles towards the bottom right of the plot are suited to performing work over a wide range of motion; muscles to the top right are suited to work and power production; muscles to the top left of the plot have ability to generate high amounts of force. (B) Re-scaled plot, with the longissimus dorsi muscles removed to show, with improved clarity, the functional regionalisation of other back muscles. Point colours represent muscle group (red – longissimus; blue – transversospinalis; green – iliocostalis; orange – other). Muscle abbreviations are given in Table 1.

The longissimus dorsi muscle

The longissimus dorsi muscle shows an extremely high capacity for force generation within the axial musculoskeletal system of both breeds of dog (Fig. 5A), reflected by its relatively short fascicles and high PCSA. However, when normalised for size differences between breeds, the m. longissimus dorsi possesses a far higher propensity for muscle force production (significantly greater PCSA) in the greyhound than in the SBT (Fig. 2B; P < 0.05). The two canine breeds show similarity with one another in terms of the relative fascicle lengths of this muscle (P > 0.05). Early studies in cats (Carlson et al. 1979) indicated that the longissimus dorsi muscle is likely to be required to protect the spine from movements that may exceed its range of motion, rather than drive dynamic movements itself. More recent electromyographic studies of vertebral muscles in the dog (Ritter et al. 2001) and horse (Robert et al. 2001, 2002; Licka et al. 2004, 2009) also lead to the hypothesis that the m. longissimus dorsi appears to be geared towards stabilisation of the vertebral column against dynamic forces and limiting hyper-flexion of the spine. Further mass-loading experiments support this, suggesting that the m. longissimus dorsi stabilises the back/trunk during trotting locomotion, rather than actively initiating movement (Ritter et al. 2001). Our results support previous findings of a high propensity for force production in the longissimus dorsi muscle and suggest that this muscle in particular has an important role in support and stability.

The effect of measuring the longissimus dorsi and illiocostalis lumborum muscles (LD&IL) as a whole on the outcome measures of muscle function is highlighted in Fig. 5A. In both breeds of dog, combining both muscles into one measurement unit (reflecting the fusion of these two muscle bellies in situ) resulted in a right-shift on the graph for the muscle complex. This shift reflects the longer fascicles within the iliocostalis lumborum muscle portion, and hence the enhanced potential for power production when these two muscles are considered as a single functional unit. It is notable that the addition of the iliocostalis lumborum muscle results in slightly different alterations in muscle complex function for the two different breeds of dog considered here. The shift towards enhanced power-generating capability is more pronounced in the case of the greyhound, whereas in the SBT, the shift is smaller, and more in the direction of an increase in the force-generating capacity of the muscle complex. This may highlight a key functional difference in the more caudal component of this muscle complex in greyhounds compared with SBTs, with the greyhound caudal lumbar musculature more highly specialised toward facilitating rapid powerful extensions of the lumbar spine, as might be required for sprinting. This may reflect the relative greater specialisation for the use of the galloping gait and high speed locomotion in the greyhound compared with the SBT, historically bred for hunting and later fighting. When galloping, the greatest amplitude of sagittal bending occurs in the lumbar region, and the greatest increase in muscle activity of the longissimus lumborum muscle when moving up through the gaits from a walk to trot to gallop occurs at the level of L6 (Schilling & Carrier, 2010). The percentage of glycolytic muscle fibres also increase caudally along the epaxial musculature (Schilling, 2009), indicating a greater potential need caudally for increased muscle fibre contraction velocities and adaptation for muscle activity over a wider range of motion.

Sources of power for locomotion

We hypothesised that the greyhound axial musculoskeletal system would be specialised for power production to aid high speed locomotion and rapid accelerations. Many muscles of the greyhound limbs are suited to power production, in particular the large hip extensor muscles (Williams et al. 2008b); however, they are unable to match the exceedingly high power required during rapid accelerations (Williams et al. 2009a,b2009b). Our data suggest that the epaxial muscles of this breed, particularly the combined muscle unit of longissimus dorsi and iliocostalis lumborum, are likely to have the capacity to supplement the power produced by the muscles of the limbs. We can calculate the extent of this, since the instantaneous maximum muscle power can be estimated as one-tenth of the product of Fmax and maximum fibre contraction velocity, Vmax (Hill, 1938). We assume Vmax to be 3.8 Lo s−1 (Lo = muscle fibre resting length; Ameredes et al. 1992). Isometric Fmax can be estimated by multiplying 0.3 MPa – the maximum isometric stress of vertebrate skeletal muscle (Wells, 1965; Medler, 2002) – by PCSA.

The instantaneous maximum muscle power of the combined longissimus dorsi and iliocostalis lumborum muscles was calculated using the above method. When considered as a proportion of the total summed instantaneous maximum power of all the hindlimb muscles (564 W from Williams et al. 2008b) this suggests that the combined longissimus dorsi and iliocostalis muscle unit has the capacity to augment the power produced by the combined hindlimb musculature by 12%. A simpler estimation based on augmentation of simply muscle mass alone (summed total of hindlimb muscle mass is 5.42 kg taken from Williams et al. 2008b) provides a similar figure (11.7%). The total combined power-producing capability of the major epaxial muscle and hindlimb musculature therefore equals 106 W kg−1 muscle mass – a figure just below the maximum estimated operating power of greyhound muscles during maximal accelerations (120 W kg−1; Williams et al. 2009b). This latter value was calculated from the measured body mass specific power of 60 W kg−1 during maximal accelerations (Williams et al. 2009b), doubled to express in muscle mass specific terms, as 50% of greyhound body mass is muscle (Williams et al. 2008a,b2008b). Elastic elements within the limbs, or indeed the back, may play a role in amplification of muscle power in order to reach these high values, as has been illustrated in specific circumstances (e.g. Wilson et al. 2003). There is certainly compelling evidence to suggest that aponeuroses can act as efficient springs for storing and releasing energy (Azizi et al. 2009), and thus the thick thoracolumbar fascia that overlies m. longissimus dorsi may well play such a role within the back during locomotion (Alexander et al. 1985). Equally, we have not included the forelimbs in our calculations above, and these may contribute somewhat to powering accelerations.

Breed differences

Aside from differences between breeds in the m. longissimus dorsi and the m. iliocostalis lumborum, other muscles also showed significant architectural differences between SBTs and greyhounds (Fig. 2). A number of muscles in the cervical region, namely the m. obliquus capitis cranialis, m. obliquus capitis caudalis and m. semispinalis capitis, were larger in the SBT (P < 0.05), with considerably higher PCSAs. Of these muscles, the m. semispinalis capitis appears to be suited to generating reasonably high forces within the vertebral column (Fig. 5B). Head and neck motions are important for effective locomotor function, but are also imperative for a range of other behaviours: feeding, orientation (through visual, auditory and vestibular means) and defence, among others. The relative difference in size and morphology of these cranial neck muscles between breeds of dog may reflect selective breeding for rather different skull and neck conformations that facilitate sprinting vs. fighting in the greyhound and SBT, respectively.

A number of muscles of the thoracic region of the spine, the m. serratus dorsalis, mm. scaleni and m. multifidus thoracis, were larger in the greyhound (P < 0.05), with the mm. scaleni and m. multifidus thoracis also possessing relatively longer fascicles in the greyhound with respect to the SBT. These three muscles all have origins/insertions on the ribs and, as such, function not only to facilitate locomotion but also to varying extents in ventilation. Based upon gross morphological and experimental studies, these muscles can contribute to inspiration via expansion of the thoracic cavity by rotating the ribs cranially and laterally (e.g. Raper et al. 1966; De Troyer & Kelly, 1984; De Troyer et al. 1994). Locomotion and ventilation are intrinsically linked during galloping gaits (Bramble & Carrier, 1983), and it follows that an animal highly specialised for rapid sprinting may need to possess enhanced adaptations of musculature used to facilitate ventilation during galloping.

Studies such as this, which infer adaptation from direct comparisons between two species, are subject to well recognised limitations (Garland & Adolph, 1994), as this process involves the confounding of independent variables (the environmental factor, i.e. selective ‘regime’, and species membership). The specific choice of greyhound and SBTs as comparator breeds in this study, however, goes some way towards ameliorating these concerns, and has been discussed at length elsewhere (Pasi & Carrier, 2003; Kemp et al. 2005). The type of selection undergone by these two breeds has been very specific and extreme, and their environments (domestic, and influenced by a desire for the dog to be well housed, fed, and in good condition for sprinting or fighting) have been highly controlled. Thus, although it is plausible that differences between the two breeds seen in this study may exist due genetic drift, it is more likely that the adaptive differences indicated are caused directly by selection for fighting or sprinting.

Regional specialisation

Both the greyhound and SBT showed regional variation in the location of muscle mass along the vertebral column (Figs 3 and 4). Both breeds had a higher proportion of muscle mass located in the lumbar region than in either the cervical or thoracic regions of the spine. This trend was mirrored by differences in PCSA between regions. Our hypothesis, that location along the vertebral column influences muscle architecture parameters, appears to be supported by these data. Such regional specialisation of the epaxial muscle architecture has been seen in other species, notably the horse. Studies have indicated that in addition to variation between muscles along the length of the axial musculoskeletal system, an individual muscle also has the potential to perform varying functions along its length (Wakeling et al. 2007; Ritruechai et al. 2008). PCSA in the equine longissimus dorsi muscle increased caudally (Ritruechai et al. 2008), along with the apparent force-generating capacity of the muscle. It was not possible to segment the m. longissimus dorsi in this way (by vertebral level) in our study because the relatively longer fascicle lengths and different orientation in the canine specimens meant that fascicles spanned many spinal segments. Standardisation of any muscle segmentation would therefore have been problematic and was avoided.

Neither breed of dog showed a significant difference in fascicle length or AI between regions of the vertebral column; however, there was a strong trend for the AI of the epaxial muscles to decrease caudally along the spine in the SBT in particular. A trend for a caudal decrease in AI equates to the presence of shorter, more oblique muscle fibres in the caudal axial musculature. An oblique fibre orientation allows for both mobility and stability in all planes of the body simultaneously and thus better meets the complex needs for multiple diverse functions of the trunk in mammals (Schilling & Carrier, 2010). This appears to be particularly important towards the caudal end of the trunk, likely due to the requirement of this region to facilitate stabilization but also sagittal mobilisation during high-speed asymmetrical gaits. An oblique fibre orientation likely provides an advantage in the shortening velocity of the entire muscle (Brainerd & Azizi, 2005). This is because fibres in a pennate muscle rotate to greater angles of pennation as they shorten (Maganaris et al. 1998), meaning that the muscle's velocity along its line of action can exceed the velocity of the contracting fibres themselves. This suggests that despite their apparent adaptation for generation of force, the caudal epaxial muscles may still be capable of fast contraction and hence mobilisation. This is also supported by the presence of high proportions of glycolytic muscle fibres within caudal regions of the iliocostalis lumborum and longissimus dorsi muscles in some mammals (Schilling, 2009).

Limitations

Some muscle parameters in our data showed reasonably high standard deviations (Table 1), particularly for certain muscles, parameters (especially fascicle length, and thus PCSA, as it is derived from this) and in particular in the greyhound. Fascicle lengths are inherently variable throughout a muscle belly, and therefore this variation is somewhat expected in the average values presented. Other sources of this variation include differences in the mass and size of our specimens: for greyhounds, the largest dog was 7% heavier than the mean body mass, and the smallest, 11% lighter; for SBTs the largest dog was 8% heavier than the mean and the smallest, 7% lighter. We were restricted in our use of cadavers to those donated or sourced opportunistically, meaning that inevitably dogs used in the study will introduce some level of variation due to their wide range of backgrounds, age, training and exercise histories. We were unable to control for this variation in lifestyle features but our data remain representative of the wider domestic populations of these two breeds of dog.

Conclusion

Here we present quantitative anatomical data for the epaxial trunk muscles of dogs bred for sprinting (the racing greyhound) vs. fighting (the Staffordshire bull terrier) in a comparative context. We highlight that some greyhound epaxial muscles, the m. longissimus dorsi and m. iliocostalis lumborum in particular, appear to show specialisation for power production, when compared with the SBT. We estimate that these two epaxial muscles alone could augment hindlimb power production during locomotion by around 12% in the greyhound if used to facilitate lumbar extension during galloping gaits. Breed differences were also found within other muscles of the axial musculoskeletal system, particularly in the cranial cervical muscles and also the deep thoracic muscles which insert on the ribs. These may also highlight key functional adaptations between two breeds of dog which have been selectively bred for particular purposes. Additionally, we illustrate specialisation of muscle function by trunk region, with differences found in both mass and PCSA of muscles at varying levels of the trunk, and in different muscle functional groups.

Acknowledgments

The authors would like to thank Carol Hercock, Iain Young, Fay Penrose and Lee Moore for providing access to cadavers. Anthony Channon provided interesting and valuable comment on our data, and Rachel Davis, Stefanie Krauss and Daniel Rogers provided assistance with data collection.

Author contributions

E.W. was involved with data collection and analysis, and wrote the draft manuscript; P.H. was involved in data collection and editing of the manuscript; S.B.C. was involved with data collection, analysis, interpretation and editing of the manuscript.

References

  1. Aerts P. Vertical jumping in Galago senegalensis: the quest for an obligate mechanical power amplifier. Philos Trans R Soc Lond B Biol Sci. 1998;353:1607–1620. [Google Scholar]
  2. Alexander RMcN, Jayes AS, Maloiy GMO, et al. Allometry of the leg muscles of mammals. J Zool Lond. 1981;194:539–552. [Google Scholar]
  3. Alexander RMcN, Dimery NJ, Ker RF. Elastic structures in the back and their role in galloping in some mammals. J Zool. 1985;207:467–482. [Google Scholar]
  4. Ameredes BT, Brechue WF, Andrew GM, et al. Force-velocity shifts with repetitive isometric and isotonic contractions of canine gastrocnemius in situ. J Appl Physiol. 1992;73:2105–2111. doi: 10.1152/jappl.1992.73.5.2105. [DOI] [PubMed] [Google Scholar]
  5. Askew GN, Marsh RL, Ellington CP. The mechanical power output of the flight muscles of blue-breasted quail (Coturnix chinensis) during take-off. J Exp Biol. 2001;204:3601–3619. doi: 10.1242/jeb.204.21.3601. [DOI] [PubMed] [Google Scholar]
  6. Azizi E, Halenda GM, Roberts TJ. Mechanical properties of the gastrocnemius aponeurosis in wild turkeys. Integr Comp Biol. 2009;49:51–58. doi: 10.1093/icb/icp006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Brainerd EL, Azizi E. Muscle fiber angle, segment bulging and architectural gear ratio in segmented musculature. J Exp Biol. 2005;208:3249–3261. doi: 10.1242/jeb.01770. [DOI] [PubMed] [Google Scholar]
  8. Bramble DM, Carrier DR. Running and breathing in mammals. Science. 1983;219:251–256. doi: 10.1126/science.6849136. [DOI] [PubMed] [Google Scholar]
  9. Carlson H, Halbertsma J, Zomlefer M. Control of the trunk during walking in the cat. Acta Physiol Scand. 1979;105:251–253. doi: 10.1111/j.1748-1716.1979.tb06338.x. [DOI] [PubMed] [Google Scholar]
  10. De Troyer A, Kelly S. Action of neck accessory muscles on rib cage in dogs. J Appl Physiol. 1984;56:326–332. doi: 10.1152/jappl.1984.56.2.326. [DOI] [PubMed] [Google Scholar]
  11. De Troyer A, Cappello M, Brichant JF. Do canine scalene and sternomastoid muscles play a role in breathing? J Appl Physiol. 1994;76:242–252. doi: 10.1152/jappl.1994.76.1.242. [DOI] [PubMed] [Google Scholar]
  12. English AW. The functions of the lumbar spine during stepping in the cat. J Morphol. 1980;165:55–66. doi: 10.1002/jmor.1051650106. [DOI] [PubMed] [Google Scholar]
  13. Evans HE. Miller's Anatomy of the Dog. 3rd edn. Philadelphia: WB Saunders; 1993. [Google Scholar]
  14. Garland TJ, Adolph SC. Why not to do two-species comparative studies: limitations on inferring adaptation. Physiol Zool. 1994;67:797–828. [Google Scholar]
  15. Gellman KS, Bertram JEA, Hermanson JW. Morphology, histochemistry and function of epaxial cervical musculature in the horse (Equus caballus. J Morphol. 2002;251:182–194. doi: 10.1002/jmor.1082. [DOI] [PubMed] [Google Scholar]
  16. Hildebrand M. Further studies on locomotion of the cheetah. J Mammal. 1961;42:84–91. [Google Scholar]
  17. Hill A. The heat of shortening and the dynamic constants of muscle. Proc R Soc Lond B. 1938;126:136. [Google Scholar]
  18. Hudson PE, Corr SA, Payne-Davis RC, et al. Functional anatomy of the cheetah (Acinonyx jubatus) hindlimb. J Anat. 2011;218:363–374. doi: 10.1111/j.1469-7580.2010.01310.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hui D, Jackson RB. Uncertainty in allometric exponent estimation: a case study in scaling metabolic rate with body mass. J Theor Biol. 2007;249:168–177. doi: 10.1016/j.jtbi.2007.07.003. [DOI] [PubMed] [Google Scholar]
  20. Kemp TJ, Bachus KN, Nairn JA, et al. Functional trade-offs in the limb bones of dogs selected for running versus fighting. J Exp Biol. 2005;208:3475–3482. doi: 10.1242/jeb.01814. [DOI] [PubMed] [Google Scholar]
  21. Kida MY, Johnson DR, McAndrew TJ, et al. Adaptation in the vertebral column: a comparative study of patterns of metameric variation in seven species of small mammals. J Anat. 1999;194:207–214. doi: 10.1046/j.1469-7580.1999.19420207.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Lichtwark GA, Wilson AM. Effects of series elasticity and activation conditions on muscle power output and efficiency. J Exp Biol. 2005;208:2845. doi: 10.1242/jeb.01710. [DOI] [PubMed] [Google Scholar]
  23. Licka TF, Peham C, Frey A. Electromyographic activity of the longissimus dorsi muscles in horses during trotting on a treadmill. Am J Vet Res. 2004;65:155–158. doi: 10.2460/ajvr.2004.65.155. [DOI] [PubMed] [Google Scholar]
  24. Licka T, Frey A, Peham C. Electromyographic activity of the longissimus dorsi muscles in horses when walking on a treadmill. Vet J. 2009;180:71–76. doi: 10.1016/j.tvjl.2007.11.001. [DOI] [PubMed] [Google Scholar]
  25. Lieber RL, Friden J. Functional and clinical significance of skeletal muscle architecture. Muscle Nerve. 2000;23:1647–1666. doi: 10.1002/1097-4598(200011)23:11<1647::aid-mus1>3.0.co;2-m. [DOI] [PubMed] [Google Scholar]
  26. Long JH, Jr, Pabst DA, Shepherd WR, et al. Locomotor design of dolphin vertebral columns: bending mechanics and morphology of Delphinus delphis. J Exp Biol. 1997;200:65–81. doi: 10.1242/jeb.200.1.65. [DOI] [PubMed] [Google Scholar]
  27. Maganaris NC, Baltzopoulos V, Sargeant AJ. In vivo measurements of the triceps surae complex architecture in man: implications for muscle function. J Physiol. 1998;512:603–614. doi: 10.1111/j.1469-7793.1998.603be.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Medler S. Comparative trends in shortening velocity and force production in skeletal muscles. Am J Physiol Regul Integr Comp Physiol. 2002;283:R368–R378. doi: 10.1152/ajpregu.00689.2001. [DOI] [PubMed] [Google Scholar]
  29. Mendez J, Keys A. Density and composition of mammalian muscle. Metabolism. 1960;9:184–188. [Google Scholar]
  30. Moritz S, Fischer MS, Schilling N. Three-dimensional fibre-type distribution in the paravertebral muscles of the domestic ferret (Mustela putorius f. furo) with relation to functional demands during locomotion. Zoology (Jena) 2007;110:197–211. doi: 10.1016/j.zool.2007.01.004. [DOI] [PubMed] [Google Scholar]
  31. Myatt JP, Crompton RH, Thorpe SKS. Hindlimb muscle architecture in non-human great apes and a comparison of methods for analysing inter-species variation. J Anat. 2011;219:150–166. doi: 10.1111/j.1469-7580.2011.01383.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Pasi BM, Carrier DR. Functional trade-offs in the limb muscles of dogs selected for running vs. fighting. J Evol Biol. 2003;16:324–332. doi: 10.1046/j.1420-9101.2003.00512.x. [DOI] [PubMed] [Google Scholar]
  33. Payne RC, Hutchinson JR, Robilliard JJ, et al. Functional specialisation of pelvic limb anatomy in horses (Equus caballus. J Anat. 2005;206:557–574. doi: 10.1111/j.1469-7580.2005.00420.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Peplowski MM, Marsh RL. Work and power output in the hindlimb muscles of Cuban tree frogs Osteopilus septentrionalis during jumping. J Exp Biol. 1997;200:2861–2870. doi: 10.1242/jeb.200.22.2861. [DOI] [PubMed] [Google Scholar]
  35. Raper AJ, Thompson WT, Shapiro W, et al. Scalene and sternomastoid muscle function. J Appl Physiol. 1966;21:497–502. doi: 10.1152/jappl.1966.21.2.497. [DOI] [PubMed] [Google Scholar]
  36. Ritruechai P, Weller R, Wakeling JM. Regionalisation of the muscle fascicle architecture in the equine longissimus dorsi muscle. Eq Vet J. 2008;40:246–251. doi: 10.2746/042516408X273675. [DOI] [PubMed] [Google Scholar]
  37. Ritter DA, Nassar PN, Fife M, et al. Epaxial muscle function in trotting dogs. J Exp Biol. 2001;204:3053–3064. doi: 10.1242/jeb.204.17.3053. [DOI] [PubMed] [Google Scholar]
  38. Robert C, Audige F, Valette JP, et al. Effects of treadmill speed on the mechanics of the back in the trotting saddlehorse. Equine Vet J Suppl. 2001;33:154–159. doi: 10.1111/j.2042-3306.2001.tb05380.x. [DOI] [PubMed] [Google Scholar]
  39. Robert C, Valette JP, Pourcelot P, et al. Effects of trotting speed on muscle activity and kinematics in saddlehorses. Equine Vet J Suppl. 2002;34:295–301. doi: 10.1111/j.2042-3306.2002.tb05436.x. [DOI] [PubMed] [Google Scholar]
  40. Rodriguez-Barbudo V, Varmonde R, Aguiera E, et al. Estudio histoquimico y morfometrico del musculo tibial craneal en perros de diferentes aptitudes (galago, fox terrier y pastor alemain) Anat Histol Embryol. 1984;13:300–312. doi: 10.1111/j.1439-0264.1984.tb00260.x. [DOI] [PubMed] [Google Scholar]
  41. Schilling N. Metabolic profile of the perivertebral muscles of small therian mammals: implications for the evolution of the mammalian trunk musculature. Zoology (Jena) 2009;112:279–304. doi: 10.1016/j.zool.2008.09.007. [DOI] [PubMed] [Google Scholar]
  42. Schilling N, Carrier DR. Function of the epaxial muscles during trotting. J Exp Biol. 2009;212:1053–1063. doi: 10.1242/jeb.020248. [DOI] [PubMed] [Google Scholar]
  43. Schilling N, Carrier DR. Function of the epaxial muscles in walking, trotting and galloping dogs: implications for the evolution of epaxial muscle function in tetrapods. J Exp Biol. 2010;213:1490–1502. doi: 10.1242/jeb.039487. [DOI] [PubMed] [Google Scholar]
  44. Schmidt-Nielsen K. Scaling: Why Is Size so Important? 1st edn. Cambridge, UK: Cambridge University Press; 1984. [Google Scholar]
  45. Sharir A, Milgram J, Shahar R. Structural and functional anatomy of the neck musculature of the dog (Canis familiaris. J Anat. 2006;208:331–351. doi: 10.1111/j.1469-7580.2006.00533.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Slijper EJ. Comparative biological-anatomical investigations on the vertebral column and spinal musculature of mammals. Proc Kon Ned Akad Wetenschap. 1946;45:1–128. [Google Scholar]
  47. Smith NC, Wilson AM, Jespers KJ, et al. Muscle architecture and functional anatomy of the pelvic limb of the ostrich (Struthio camelus. J Anat. 2006;209:765–779. doi: 10.1111/j.1469-7580.2006.00658.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Usherwood JR, Wilson AM. Biomechanics: no force limit on greyhound sprint speed. Nature. 2005;438:753–754. doi: 10.1038/438753a. [DOI] [PubMed] [Google Scholar]
  49. Wakeling JM, Ritruechai P, Dalton S, et al. Segmental variation in the activity and function of the equine longissimus dorsi muscle during walk and trot. Equine Comp Exerc Phys. 2007;4:95–103. [Google Scholar]
  50. Weis-Fogh T, Alexander RM. The sustained power output from striated muscle. In: Pedley TJ, editor. Scale Effects in Animal Locomotion. London: Academic Press; 1977. pp. 511–525. [Google Scholar]
  51. Wells J. Comparison of mechanical properties between slow and fast mammalian muscles. J Physiol. 1965;178:252–269. doi: 10.1113/jphysiol.1965.sp007626. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. West TG, Toepfer CN, Woledge RC, et al. Power output of skinned skeletal muscle fibres from the cheetah (Acinonyx jubatus. J Exp Biol. 2013;216:2974–2982. doi: 10.1242/jeb.083667. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Williams SB, Wilson AM, Daynes J, et al. Functional anatomy and muscle moment arms of the thoracic limb of an elite sprinting athlete: the racing greyhound (Canis familiaris. J Anat. 2008a;213:373–382. doi: 10.1111/j.1469-7580.2008.00962.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Williams SB, Wilson AM, Rhodes L, et al. Functional anatomy and muscle moment arms of the pelvic limb of an elite sprinting athlete: the racing greyhound (Canis familiaris) J Anat. 2008b;213:361–372. doi: 10.1111/j.1469-7580.2008.00961.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Williams SB, Usherwood JR, Jespers KJ, et al. Exploring the mechanical basis for acceleration: pelvic limb locomotor function during accelerations in racing greyhounds (Canis familiaris. J Exp Biol. 2009a;212:550–565. doi: 10.1242/jeb.018093. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Williams SB, Tan H, Usherwood JR, et al. Pitch then power: limitations to acceleration in quadrupeds. Biol Lett. 2009b;5:610–613. doi: 10.1098/rsbl.2009.0360. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Wilson AM, Watson JC, Lichtwark GA. Biomechanics: a catapult action for rapid limb protraction. Nature. 2003;421:35–36. doi: 10.1038/421035a. [DOI] [PubMed] [Google Scholar]

Articles from Journal of Anatomy are provided here courtesy of Anatomical Society of Great Britain and Ireland

RESOURCES