A thesis submitted to the University of Manchester for the degree of Doctor of Philosophy in the Faculty of Life Sciences

Transcription

1 CONSTRAINTS ON THE ENERGETICS AND MECHANICS OF LOCOMOTION IN LEGHORN CHICKENS (GALLUS GALLUS DOMESTICUS) A thesis submitted to the University of Manchester for the degree of Doctor of Philosophy in the Faculty of Life Sciences 2015 Kayleigh A. Rose

2 Contents List of Abbreviations... 5 Abstract... 7 Declaration... 8 Copyright Statement... 9 Author information Acknowledgements Introduction The metabolic cost of terrestrial locomotion Measuring the metabolic cost of locomotion Factors influencing the metabolic cost of locomotion Gait Body size Scaling of limb posture Scaling of gait kinematics Scaling of locomotor energy metabolism Linking energetics and mechanics The rate of force generation hypothesis Intraspecific variation in body size Sex Domestic leghorn chickens (Gallus gallus domesticus) Overview and thesis aims Alternative format Intraspecific scaling of the minimum metabolic cost of transport in leghorn chickens (Gallus gallus domesticus): links with limb kinematics, morphometrics and posture Sex differences in gait utilization and energy metabolism during terrestrial locomotion in two varieties of chicken (Gallus gallus domesticus) selected for different body size Differential sex-specific gait dynamics in leghorn chickens (Gallus gallus domesticus) selectively bred for different body size Ontogeny of sex differences in the energetics and kinematics of terrestrial locomotion in leghorn chickens (Gallus gallus domesticus)

3 6. Variety, sex and ontogenetic differences in the pelvic limb muscle architectural properties of leghorn chickens (Gallus gallus domesticus) and their links with locomotor performance Sex differences in the metabolic cost of terrestrial locomotion are not accounted for by the rate of muscle force generation in two varieties of leghorn chicken (Gallus gallus domesticus) Discussion Summary of findings Intraspecific scaling of the minimum metabolic cost of transport The cost of muscle force production Sexual dimorphism and the metabolic cost of locomotion Sex constraints on maximum aerobic speed and gait Sex constraints on walking dynamics Understanding the scaling of limb posture and gait kinematics Conclusion References Appendix Word count: 60,300 3

4 List of Figures Figure 1.1 Figure 1.2 Figure 1.3 Figure 1.4 Figure 1.5 Figure 1.6 Examples of how the mass-specific metabolic cost of transport varies according to speed and gait in diferent species...21 Schematic of the effective mechanical advantage (EMA) of the limb..24 Duty factor versus Froude number in (a) birds and (b) the forelimbs of quadrupedal mammals.28 Interspecifc allometry of the minimum metabolic cost of transport.32 Interspecific scaling of the total mass-specific external mechanical energy (E mech ) of the centre of mass (CoM) and mass-specific minimum metabolic cost of transport (CoT min )..32 Metabolic rate, the rate of force generation and the cost coefficient in birds and humans 36 4

5 List of Abbreviations B B c CoL CoT min CoT net CoT tot DF E kh E kv EMA E mech E met E p E tot E trans ˆF Fr f stride h back h hip h hip : Σl segs J J JuvL JuvL L f l fem l stride L female bantam male bantam cost coefficient metabolic cost of locomotion minimum metabolic cost of transport net metabolic cost of transport total cost of transport duty factor horizontal kinetic energy vertical kinetic energy effective mechanical advantage mechanical cost of transport metabolic cost of locomotion potential energy total kinetic energy metabolic cost of transport normalised stride frequency Froude number stride frequency back height hip height posture index juvenile female juvenile male juvenile female standard breed juvenile male standard breed fascicle length femur length stride length female standard breed 5

6 L ˆL l tars l tib l stride M M M b Net-P met PCSA P met R r t stance ˆT stance t swing ˆT U swing male standard breed normalised stride length tarsometatarsus length tibiotarsus length stride length mature female mature male body mass net metabolic power effective physiological cross sectional area metabolic power moment arm of the ground reaction force extensor muscle moment arm stance duration normalised stance duration swing duration normalised swing duration speed!v O2 oxygen consumption!v CO2 W b w fem w tib w tars Σl segs carbon dioxide production body weight femur width tibiotarsus width tarsometatarsus width sum of the hind limb long bone length 6

7 Abstract Abstract of a thesis by Kayleigh Ann Rose submitted to the University of Manchester for the degree of PhD in the Faculty of Life Sciences and entitled Constraints on the energetics and mechanics of terrestrial locomotion in leghorn chickens (Gallus gallus domesticus) September 2015 Interspecific allometry of locomotor morphology, gait kinematics and energy metabolism is well described, but the underlying reasons for these patterns are not yet fully understood. Between disparate species, a plethora of confounding morphological and physiological variables precludes identifying the independent effects of body size, particular morphological characteristics or gait kinematics on metabolic costs. The main objective of my thesis is to elucidate the links between these integrated components of locomotion by exploiting a range in morphological and physiological similarities and differences in a single species to have resulted from selective breeding. My model species, the leghorn chicken (Gallus gallus domesticus), is selectively bred for egg laying productivity and varieties of different body size which are physiologically and geometrically similar: standard (large) and bantam (small). At sexual maturity, leghorns also exhibit male-biased sexual size dimorphism and males and females differentially bias tissue allocation to pelvic limb muscle and viscera/reproductive organs, respectively. Using respirometry measurements, I first demonstrate that between males of standard and bantam leghorn varieties, the cost of transport is independent of body mass. Using limb posture, muscle architecture and kinematic measurements, I provide the first empirical evidence in support of a potential mechanism behind a lack of scaling in the cost if transport at the intraspecific level. I also provide the first evidence of a greater metabolic cost of transport in a male compared to a female in any species. I demonstrate that the sexual dimorphism in energy metabolism cannot be accounted for by comparison at dynamically similar speeds, limb posture, muscle architectural properties or the rate of muscle force generation. Furthermore, by comparing with a younger cohort of standards, I demonstrate that sex differences in locomotor energy metabolism manifest before the onset of sexual maturity and must, therefore, result from alternative sub-organismal sexual dimorphisms. I suggest that these differences may relate to physiological specialisation for to economical egg-load carriage and performance in intermale aggressive combat. I also provide the first detailed comparison of the gait dynamics of the sexes in any species. I demonstrate that female leghorns walk with proportionally more foot-ground contact than their conspecific males. Furthermore, these kinematic differences are concomitant with the increase in anatomical mass at sexual maturity loading the swing and stance muscles of males and females, respectively. Perhaps the most striking finding in this thesis is that, while small cursosrial birds generally walk with crouched limbs, female standard leghorns walked with upright limbs. Just as larger species tend to have upright limbs to support a disproportionate amount of weight, compared to smaller more crouched species, these females may do the same due to artificial selection for increased egg production. Together, the locomotor measurements of the leghorns presented this thesis appear to support a recent hypothesis, that animals select postures and gaits to optimize muscle mechanical work and power demands and minimize active muscle volume. 7

8 Declaration No portion of the work referred to in this thesis has been submitted in support of an application for another degree or qualification of this or any other university or other institute of learning; Signed: Date: 8

9 Copyright Statement i. The author of this thesis (including any appendices and/or schedules to this thesis) owns certain copyright or related rights in it (the Copyright ) and s/he has given The University of Manchester certain rights to use such Copyright, including for administrative purposes. ii. iii. iv. Copies of this thesis, either in full or in extracts and whether in hard or electronic copy, may be made only in accordance with the Copyright, Designs and Patents Act 1988 (as amended) and regulations issued under it or, where appropriate, in accordance with licensing agreements which the University has from time to time. This page must form part of any such copies made. The ownership of certain Copyright, patents, designs, trade marks and other intellectual property (the Intellectual Property ) and any reproductions of copyright works in the thesis, for example graphs and tables ( Reproductions ), which may be described in this thesis, may not be owned by the author and may be owned by third parties. Such Intellectual Property and Reproductions cannot and must not be made available for use without the prior written permission of the owner(s) of the relevant Intellectual Property and/or Reproductions. Further information on the conditions under which disclosure, publication and commercialisation of this thesis, the Copyright and any Intellectual Property and/or Reproductions described in it may take place is available in the University IP Policy (see in any relevant Thesis restriction declarations deposited in the University Library, The University Library s regulations (see and in The University s policy on Presentation of Theses 9

10 Author information Kayleigh A. Rose Date of birth: 19 th September 1988 Nationality: British Address: Michael Smith Building, Oxford road, M13 9PT Education Ysgol Craig yr Wylfa School, Borth, Ceredigion, UK Ysgol Penglais School, Aberystwyth, Ceredigion, UK B.Sc. (Hons) in Zoology, Faculty of Life Sciences, University of Manchester, UK PhD, Faculty of Life Sciences, University of Manchester, UK Supervisor: Dr J. R. Codd. 10

11 Publications (1) Tickle, P. G., Lean, S. C., Rose, K. A. R., Wadugodapitiya, A. P. and Codd, J. R. (2013). The influence of load carrying on the energetics and kinematics of terrestrial locomotion in a diving bird. Biology Open, 2, (2) Rose, K. A., Tickle, P. G., Lees, J. J., Stokkan, K. A. and Codd, J. R. (2014). Neither season nor sex affects the cost of terrestrial locomotion in a circumpolar diving duck: the common eider (Somateria mollissima). Polar Biology, 37, (3) Rose, K. A., Nudds, R. L, and Codd, J. R. (2015). Intraspecific scaling of the minimum metabolic cost of transport in leghorn chickens (Gallus gallus domesticus): links with limb kinematics, morphometrics and posture. The Journal of Experimental Biology, 218, (4) Rose, K. A., Nudds, R. L, Butler, P. J. and Codd, J. R. (2015). Sex differences in gait utilization and energy metabolism during terrestrial locomotion in two varieties of chicken (Gallus gallus domesticus) selected for different body size. Biology Open, 4, (5) Rose, K.A., Codd, J. R., and Nudds, R. L. (in review). Differential sexspecific gait dynamics in leghorn chickens (Gallus gallus domesticus) selectively bred for different body size (6) Rose, K.A., Bates, K. T., Nudds, R. L. and Codd, J. R. (in review). Ontogeny of sex differences in the energetics and kinematics of terrestrial locomotion in leghorn chickens (Gallus gallus domesticus) (7) Rose, K.A., Nudds, R. L and Codd, J. R. (in review). Variety, sex and ontogenetic differences in the pelvic limb muscle architectural properties in leghorn chickens (Gallus gallus domesticus) and their links with locomotor performance (8) Rose, K.A., Codd, J. R. and Nudds, R. L. (in preparation). Sex differences in the metabolic cost of terrestrial locomotion are not accounted for by the rate of muscle force generation in two varieties of leghorn chicken (Gallus gallus domesticus) (9) Rose, K.A., McGhie, H. A., Codd, J. R., and Nudds, R. L. (in preparation). Pelvis width scales according to locomotor specialization in birds 11

12 Conference Presentations University of Manchester Manchester Organismal Biology Symposium 2014: Size and the metabolic cost of transport Faculty of Life Sciences PhD Student Symposium 2013: Size, sex and the metabolic cost of transport International Conference Society for Experimental Biology Annual Conference 2014 (Manchester): Scaling of the metabolic cost of transport: insights from selective breeding 12

13 Acknowledgements My greatest thanks go to my supervisor Jonathan Codd who inspired me to undertake this PhD. In particular, I would like to thank him for his immeasurable encouragement, support and friendship over the last five years, which have helped lead to where I am now and will surely influence what I do next. I would also like to thank my co-supervisor Robert Nudds for his valuable advice, guidance and interesting conversations (especially on statistics) throughout my studies. His excitement for science (and life in general) is infectious! Thanks also go to Karl Bates for his contribution to this thesis. Without the birds and resources he provided, two of my research chapters would not have been possible and I wouldn t know half as much about bird anatomy. I am also grateful to the journal editors and anonymous reviewers who have contributed towards the development of my manuscripts, from which I learnt a great deal. I gratefully acknowledge the funding provided by the National Environmental Research Council and the Manchester Museum, which made my research possible. To the whole Codd-Nudds lab medley including Pete, Lees, Nate, Mi, Jamie and Charlotte thank you for the help, the friendship and the laughs. Most of all, I will miss our regular lab meetings in Big Hands and the bolognese. I would also like to thank many friends - Vick, Karlina, Tom, Kelly, Ed, Jean, Paul, Nige and the Cummingses - for their support and the treasured memories made along the way. Last but not least, I thank my awesome parents and friends Wendy and Steve for supporting me in everything I have chosen to do! 13

14 Dedicated to my parents, Wendy and Stephen Rose 14

15 1. Introduction 15

16 1. Introduction 1.1 The metabolic cost of terrestrial locomotion Locomotion is integral to the evolutionary fitness of animals for myriad reasons; for example, via foraging/prey capture, evading predation or adverse environmental conditions, migrating and impressing/combatting for mates. The locomotor systems (i.e. musculoskeletal systems) and supporting organ systems of animals are therefore under multiple and varying demands. Evolution is expected to favour compromises between a species differing locomotor attributes, such as maximum performance (e.g. top speed, aerobic capacity), the economy of metabolic energy use, stamina and manoeuverability (Alexander 2003). The specific adaptations and constraints of a species locomotor system are the product of their life histories within the constraints of their environment, or multiple environments, within which they locomote (Nudds et al. 2010; Fish et al. 2001, 2000). Animals must apportion their energy between growth, maintenance, activity and reproduction; however, their energy budgets will always be limited (Weiner 1992). Since the metabolic requirements of locomotion are an integral component of an animal s energy budget, they have important consequences for fitness (Tolkamp et al. 2002; Goldstein 1988). Understanding the factors that influence the metabolic requirements of locomotion therefore provides insight into the behavior and evolution of animals. The metabolic cost of locomotion is inherently linked to morphology and physiology; therefore, investigations into the factors that underlie metabolic costs use integrative approaches by combining anatomical, biomechanical and metabolic measurements. Terrestrial locomotion is perhaps the most studied form of locomotion using this approach. Comparisons of the locomotion energetics and mechanics of disparate species from animals ranging in body size from invertebrates to elephants has identified common scaling patterns from which predictions can be generated for different species using morphological measurements (Taylor et al. 1982; Kram and Taylor 1990; Langman et al. 1995; Full et al. 1990b). Furthermore, it has drawn attention to unique morphological 16

17 and physiological adaptations and constraints in species whose measurements deviate from these predictions linked to dominant locomotor modes (Fish et al. 2000, 2001; Griffin and Kram 2000; Nudds et al. 2010), ecological niches (Bruinzeel et al. 1999), climates (Janis and Wilhelm 1993) or activity patterns (Nudds et al. 2011; Watson et al. 2011). Understanding of the factors that determine the metabolic cost of locomotion, however, is still limited. The majority of understanding on the structural basis for locomotor metabolic cost is based upon differences observed through interspecific comparative research. A single species, however, may show considerable variation in body form and physiology according size, sex or ontogenetic stage, for example. Few studies have exploited intraspecific variation in form and physiology as a means of understanding constraints on the energetics and mechanics of terrestrial locomotion (Lees et al. 2012; Abourachid 2000; Griffin et al. 2004; Langman et al. 2012; Paxton et al. 2013; Weyand et al. 2010). In this thesis, using domestic leghorn chickens (Gallus gallus domesticus) as a model species, I use an integrative and comparative approach to explore different factors, linked to anatomical structure and physiological specialization, that influence locomotion energetics, mechanics and maximum performance. Chapter 2 investigates the influence of intraspecific variation in body size on the metabolic cost of locomotion between leghorn varieties artificially selected for large and small body size. Chapter 3 investigates the influence of sexual dimorphism on maximal performance, metabolic costs and gait mechanics. In Chapter 4, I examine the gait dynamics of male and female leghorns of small and large varieties and investigate their links with sexual dimorphisms. Chapter 5 investigates the influence of the ontogeny of sexual dimorphism, including the onset of female gravidity and increased male muscles mass, on various aspects of locomotor performance. In Chapter 6 I quantify the hind limb skeletal muscle architectural properties of the sexes and varieties in these studies and link them to locomotor performance attributes. Finally, in Chapter 7 I investigate whether a kinematic mechanism accounts for intraspecific variations in the metabolic cost of locomotion in these birds. This thesis, altogether, presents novel insight into different factors influencing the locomotor attributes of a species. 17

18 1.2 Measuring the metabolic cost of locomotion Energy is the capacity to do work (the product of force and distance). The metabolic rate of an animal is the rate at which it converts chemical energy in the form of ATP into heat and external work in essential life processes. Metabolic rates can be directly quantified by measuring the heat loss of animals (Lighton 2008; Kleiber 1961). An indirect method of quantifying this heat loss (indirect calorimetry) is generally used, however, when quantifying the metabolic cost of terrestrial locomotion. Hess law of constant heat summation states that, irrespective of the number of steps in a chemical reaction, the enthalpy change of an overall reaction is the sum of the enthalpy changes of its individual steps. Therefore, for an animal respiring aerobically, if the organic substrate (protein, carbohydrate or lipid) being oxidized is known, as well as the quantity of respiratory gases to enter or leave the animal, metabolic rate can be calculated. The rate of oxygen consumption ( V! O2, ml min -1 ) can be converted into the rate of energy metabolism using known equivalents of heat produced per unit volume of oxygen produced for the oxidation of a given substrate (Lighton 2008; Brody 1945). The ratio of the rate of carbon dioxide production ( V! CO2, ml min -1 ) to V! O2 (respiratory exchange ratio, RER) allows the determination of the substrate being oxidized. RERs equal to 1, 0.7 and are indicative of pure carbohydrate, lipid and protein oxidation, respectively (Lighton 2008; Kleiber 1961). In the absence of one of V! O2 or V! CO2, however, an RER is often assumed in order to calculate metabolic rate (Watson et al. 2011). In order to understand the morphological and biomechanical factors that influence locomotor cost, instantaneous measurements of metabolic rates are required for a given movement pattern and speed. Generally, open-flow respirometry systems are used for taking accurate and momentary gas measurements as an animal locomotes, usually upon a treadmill, within a chamber, or wearing a loosely fitted mask from which gases are sampled (Lighton and Halsey 2011). The method requires an animal to be in a steady state of aerobic energy metabolism. Full details of the respirometry set ups used in my research are given in Chapters 2, 3 and 5. 18

19 1.3 Factors influencing the metabolic cost of locomotion Gait Gaits are defined by footfall patterns and the mechanical energy fluctuations of the body centre of mass (CoM). Striding locomotion is a complex interaction between muscles, tendons, limb segments and the posture of the limb, which vary considerably across species. Two simple mechanical paradigms, however, are used to describe a wide variety of gaits used by bipedal and quadrupedal vertebrates; both of which serve to minimize the positive mechanical work that must be performed by the muscles (Cavagna and Kaneko 1977). These models relate three principal forms of mechanical energy: kinetic energy, potential energy and elastic energy. Walking gaits, usually selected at slow speeds, are defined by the horizontal kinetic energy of forward motion (E kh ) and the sum of the vertical kinetic and gravitational potential energies (E kv + E p ) of the CoM fluctuating half a cycle out of phase, analogous to an inverted pendulum (Cavagna et al. 1977). In the inverted pendulum model, E kv + E p is at its greatest during mid stance, when the CoM is at its maximum height above a stiff limb, and the E kh is at its minimum. During the swing-phase of the limb E kv + E p is at a minimum and E kh is at its maximum (Cavagna et al. 1977). The pendular-like exchange of kinetic and potential energy acts to conserve mechanical energy across a stride reducing the amount of work that the muscle must perform to lift and accelerate the CoM, expected to reduce metabolic demands (approximately 70% of the mechanical energy is conserved in humans (Cavagna et al. 1977)). Another defining factor of a walking gait is that the duty factor (DF), the proportion of the stride that the foot is in contact with the ground, is > 0.5, i.e. there is always at least one foot in contact with the ground (Hayes and Alexander 1983). In running gaits, and the trots and gallops of quadrupedal mammals, the interchange between the mechanical energies of the body CoM is analogous to a bouncing ball and the elastic properties of muscle and tendon are exploited as an energy saving mechanism (Cavagna et al. 1977). Throughout the duration of the stride, E kh and E kv + E p cycle in phase with one another, being at their lowest 19

20 during the stance, or contact phase and at their highest during the swing phase of the limb. By contrast to the stiff limb during mid stance in walking gait, the limbs are more compliant. In the first half of the contact phase, total mechanical energy (the sum of the potential and kinetic energy) is converted to elastic energy via the stretching of tendons. In the second half of the contact phase, some of that mechanical energy is returned to the system via tendon recoil. Aerial running gaits are characterized by having a phase with no foot-ground contact, i.e. DF < 0.5. In some species, including elephants (Hutchinson et al. 2003), birds and primates, an intermediate gait termed Groucho running (Mcmahon et al. 1987) or grounded running (Hancock et al. 2007; Rubenson et al. 2004; Andrada et al. 2013; Gatesy and Biewener 1991; Nudds et al. 2011) is also used in which DF > 0.5, characteristic of walking, but the mechanical energies of the CoM cycle in-phase. (Cavagna et al. 1977). Transitions from one gait to another are generally associated with increases in speed (U). For example, a horse (Equus ferus caballus) will walk, trot and gallop at slow, intermediate and fast speeds, respectively (Hoyt and Taylor 1981). A number of differing explanations for why animals select different speeds and gaits have been proposed, relating to reducing musculoskeletal forces (Farley and Taylor 1991) and bone strain (Biewener and Taylor 1986). Another reasoning may be to minimizing energy expenditure (Hoyt and Taylor 1981; Wickler et al. 2003; Margaria et al. 1963). The potential for energy savings via gait changes are often investigated by comparing the metabolic energy required to move a unit body mass M b over a unit distance, the metabolic cost of transport (CoT, J kg -1 m - 1 ), across speeds and gaits (Nudds et al. 2011; Rubenson et al. 2004; Wickler et al. 2003). Horses, for example incur a minimum metabolic cost of transport (CoT min ) within the mid-range of speeds for each of their gaits (Fig 1.1A). With an increase in U past their walking optimum, their CoT increases curvilinearly until a U is met at which the CoT would be equivalent to that of trotting at the same U. When trained to walk at speeds greater than this threshold, their walking CoT exceeds that of trotting. Natural gait transition speeds in horses tend to occur at speeds where the CoT is equivalent between gaits (Hoyt and Taylor 1981). Furthermore, 20

21 Figure 1.1 (taken from Nudds et al (2011)): Examples of how the massspecific metabolic cost of transport varies according to speed and gait in diferent species. The individual graphs represent (A) horses; (B) camels, donkeys and humans; (C) kangaroos (pentapedal walking and bipedal hopping); and (D) a solid line for small mammals (walking only) and the platypus and a dotten line for barnacle geese. the natural speeds selected in any gait in horses tend to incur the minimum possible metabolic cost (Hoyt and Taylor 1981). Different species show different trends in the CoT versus U relationships, however. For example, humans (Homo sapiens), and large desert ungulates (Maloiy et al. 2009) also have optimum U within the mid range of their walking speeds at which they incur their CoT min ; however, upon transition to aerial running gaits, their CoT drops and becomes independent of U (Fig 1.1B). Kangaroos (Dawson and Taylor 1973) show a similar decline in CoT upon making the transition from pentapedal walking to bipedal hopping; however, walking bipedally, their CoT increases linearly (Fig 1.1C). Contrastingly, the relationship between CoT and U in many small mammals (Taylor et al. 1970), and avian species with specialisation to aquatic environments (Rose et al. 2014; Nudds et al. 2010), which only use walking gait, is an ever decreasing curve (Fig 21

22 1.1D). The mechanisms underlying the trends observed in different species are not fully understood. There is also evidence for a lack of influence of a gait transition upon the CoT (Nudds et al. 2011). For example, in the platypus (Ornithorhynchus anatinus) the CoT response to U does not decrease any further from the walking CoT min (Fig 1.1D) (Fish et al. 2001). In the only small cursorial avian species to have been examined, the Svalbard rock ptarmigan (Lagopus muta hyperborea), significant energy savings were associated with the transition from grounded running to aerial running, but not from walking to grounded running (Nudds et al. 2011). Nudds et al (2011) suggested that grounded running gaits may not be evolutionarily significant, but perhaps only used when locomoting upon a treadmill. A reduction in CoT at the walk-run transition speed was observed in the ostrich (Struthio camelus), however (Rubenson et al. 2004). Avoiding an aerial phase has been suggested to increase stability, which may be of particular importance on uneven terrain (Gatesy and Biewener 1991; Gatesy 1999). Furthermore, a more crouched rather than stiff limb may increase the control of head movement and enhance the stability of vision (Hancock et al. 2007), or reduce the mechanical associated with bouncing viscera (Daley and Usherwood 2010). Nevertheless, when comparing the links between speed and metabolic rates it is important to consider the gaits separately Body size Body size has important consequences for the morphology and physiology of animals (Schmidt-Nielsen 1975; Schmidt-Nielsen 1984). Large and small animals have different metabolic and structural demands. To understand the influence of body size upon form and function, allometry is generally used to understand basic principles. Because body size influences morphology, limb arrangement and, in turn, gait kinematics and mechanics, which are each linked to the metabolic cost of locomotion, it is useful to first consider the influence of body size on each, separately. 22

23 Scaling of limb posture The smallest and largest vertebrates differ markedly in body form linked to the constraints imposed by size on their shared biological materials. The material properties of muscle or bone (and hence, their abilities to withstand stress or strain) are inherent and are, therefore, the same across all vertebrate species irrespective of body size (Biewener 1982). During locomotion, muscles must generate propulsive and breaking forces and bones must resist compressive forces. These abilities in muscle and bone are dependent upon their cross-sectional area. If all animals were geometrically similar, however, anatomical masses would scale amongst them M 1 b, lengths M 1/3 b, and areas M 2/3 b. Therefore, the abilities of muscle and bone, with regards to generating and resisting forces, respectively, would increase at a slower rate with size, than would body weight (W b, N). Increasing locomotor stresses with increasing body size across geometrically similar animals would be expected to mean that the larger the animal, the closer they are operating towards their skeleton s maximum strength, at lower safety factors (the ratio of absolute strength to actual applied load). In order for a large animal to perform the same way as a smaller animal, modifications in either limb structure or arrangement would be required, otherwise the stresses imposed on the muscles and bones would be disproportionately large (Mcmahon 1973). The interspecific scaling of limb muscles and bones in mammals is almost geometric, however (Alexander et al. 1979, 1981). Peak stresses (force per unit cross-sectional area) upon the muscles and bones would therefore also be expected to scale geometrically ( M 1/3 b ). On the contrary, however, Biewener (1989) demonstrated that peak stresses on the skeleton during running and jumping were fairly consistent across a range of mammals from ground squirrels to horses, with safety factors between 2-4 regardless of body size. Modifications in limb arrangement, with increasing body size, are common to vertebrate species, however. With increasing body size, limb long bone segments become increasingly more aligned with one another, i.e. limb posture becomes more erect. The more erect a limb is, the more closely the joints and the bones are aligned with the ground reaction force vector (Fig 1.2). 23

24 Figure 1.2 (taken from Biewener (1989)): Schematic of the effective mechanical advantage (EMA) of the limb. F g is the ground reaction force vector, R is the mechanical advantage acting about the ankle, F m is the force exerted by the ankle with a mechanical advantage of r to counteract the moment exerted by F g. Moments exerted about the joints are therefore reduced, meaning the bones are subjected to less bending under forces. Furthermore, the muscles are not required to exert forces as large in order to prevent the limbs from collapse. A more erect limb increases a muscle s effective mechanical advantage (EMA, the ratio of the extensor muscle moment arm, r, to the moment arm of the ground reaction force, R) about a joint (Fig 1.2). The ground reaction force varies in direct proportion to body weight and together with the interspecific scaling of EMA ( M 0.26 b ), this indicates that muscle force should scale M 0.75 b and therefore the mass-specific muscle force (N kg -1 ) should scale M b. Biewener (1989), therefore posited that the increase in limb erectness and hence EMA of limb muscle across mammals (also observed in birds (Gatesy and Biewener 1991) with increasing size was to act to decrease mass-specific muscle force in larger animals. It was also posited that the forces that muscles must exert are likely a key factor in determining the amount of ATP consumed by the muscles (Biewener 1989). It is unclear, however, why within a given taxon of animals, which span a wide range in body size, there may be consistent limb posture. For example, 24

25 across felid species from <4 kg to nearly 200 kg in M b, limb posture is consistent (Day and Jayne 2007)). Furthermore, it is unclear why smaller animals have crouched limb posture when they could benefit from more upright limbs, which could be lighter and have higher safety factors. Suggestions for the benefits of a crouched posture have included improved stability and manoeuverability (Gatesy and Biewener 1991; Blum et al. 2011; Biewener 1989). A better understanding of the benefits of crouched and upright postures can be gained from considering how they impact upon gait kinematics (Usherwood 2013; Gatesy and Biewener 1991; Alexander and Jayes 1983). Scaling of gait kinematics The kinematic parameters of locomotion influence muscle work, power and force. With increasing speed (U) of locomotion, different species show broadly consistent changes in kinematic parameters (Abourachid 2000, 2001; Abourachid and Renous 2000; Biewener 1983; Alexander and Jayes 1983). Generally, both stride frequency (f stride ) and length (l stride ) show continuous linear responses to increasing U. In some species, however, the rate of increase in f stride reduces with the transition from pendulum to bouncing mechanics (White et al. 2008; Gatesy and Biewener 1991). Furthermore preferential mechanisms for increasing U according to greater incremental increases in f stride over l stride or vice versa are observed between some species (Abourachid 2001). Corresponding decreases in the durations of both the swinging of the limb (t swing ) and the stance phase of the limb (t stance, which is steeper) also occur (Gatesy and Biewener 1991) with increasing U. Furthermore, in some species the t swing can be relatively constant across speeds (observed more in track-way (Abourachid 2001; Paxton et al. 2013), than in treadmill, studies (Nudds et al. 2011). The main differences in the kinematic responses to speed between disparate species, however, are in slope and magnitude, which are largely influenced by differences in body size. A smaller animal, for example, with shorter legs, takes smaller and more frequent strides, with shorter swing and stance durations than a larger one with longer legs. In order to investigate whether interspecific differences in gait kinematics are due to body size alone, or are also influenced by body form, meaningful comparisons 25

26 can be conducted by normalizing for the body size and speed dependency of the kinematics parameters. Locomotor speed can be made dimensionless when represented as the ratio of inertial to gravitational forces acting upon the body CoM. This ratio reduces to the dimensionless Froude number (Fr = (mu 2 /L)/(mg) =U 2 /gl, where g is gravitational acceleration [9.81 m s -2 ] and L is a characteristic length [hip height or leg length, m]) (Alexander and Jayes 1983; Alexander 1976). Fr incorporates the E p and E k of the CoM associated with pendulum walking mechanics, but does not account for any elastic energy associated with bouncing gaits. h hip (also referred to as effective limb length) is the most suitable measure for L since it represents the maximum height of the limb as a strut and dictates the maximum E p of the body CoM. A prerequisite for dynamic similarity of locomotion is geometric similarity of body form (Alexander and Jayes 1983). Geometrically similar but different sized individuals moving in a dynamically similar fashion share equal phasing relationships of the limbs; their limbs would swing through the same angle; and they would share equal duty factors. Additionally, they would share similar values of size-normalized kinematics parameters, ground reaction forces and mechanical power outputs (Alexander and Jayes 1983; Hof 1996). Any remaining differences in gait dynamics following correction for differences in speed and body size are attributed to differences in body form (Alexander and Jayes 1983; Gatesy and Biewener 1991). Comparison of the gait dynamics of differing species has highlighted dynamic differences in locomotion linked to the interspecific scaling of limb posture. For example, across ground-dwelling avian species from the painted quail to the common ostrich ( kg), for a given Fr, larger species had lower DF, smaller limb excursions angles, greater relative stride frequencies and lower relative stride lengths than smaller ones (Gatesy and Biewener 1991). This corresponded to h hip being a greater proportion of total hind limb skeletal length in the larger compared to smaller species i.e. having more upright limbs (Gatesy and Biewener 1991). With a more crouched limb posture, the limb extends further during a stride, relative to h hip, allowing longer relative stride lengths and footground contact periods. Consequently, DF is relatively greater in smaller, 26

27 crouched postured animals, compared with larger, upright postured animals (Fig 1.3). Similar patterns relating gait dynamics and limb posture have been observed amongst different species of ratite (Abourachid and Renous 2000), across birds of varying locomotor specialization (Abourachid 2001), mallard ducks with differing postures resulting from artificial selection (Abourachid 2000), and cursorial versus non-mammals (Alexander and Jayes 1983). Based on the interspecific differences in DF for a given Fr (Fig 1.3), an alternative hypothesis to the muscle force hypothesis of Biewener (Biewener 1989) was generated by Usherwood for the interspecific scaling of limb posture (Usherwood 2013). The duration of the stance determines the rate at which muscle mechanical work must be done (i.e. the power requirements), and the posture of the limb determines the amount of work that must be done (Usherwood 2013). The active muscle demand is high for brief stance push-offs and the active muscle demand due to work is higher for long stance durations because of large horizontal impulses and fluctuations in E k (slowing down and speeding up again). Usherwood (Usherwood 2013) posited that if the costs of locomotion are related to the volume of active muscle, the specific postures and gaits of small and large animals may act to optimise muscle work and power demands in order to reduce the volume of active muscle required during locomotion. Due to the high power demands of being small (because shorter legs mean more frequent stances of shorter durations), a more crouched limb would allow for gait dynamics, which increase the proportion of a stride during which there is contact (decreasing the power demand). On the other hand, the large work demands of being large can be reduced via an upright limb, which reduces the externally applied moments and the mechanical loading on the supporting tissues as well as speeding up the stance so that less work is down slowing down and speeding back up again (Usherwood 2013). In support of this hypothesis, a recent study comparing the kinematics of self-selected gaits in toddlers and adults showed that aspects of gait scaled in a speed and size dependent manner consistent with minimizing muscle activation required for work and power (Hubel and Usherwood 2015). In chapter 4 of this thesis, I investigate the influence of intraspecific variation in body size, anatomical proportions and limb posture on the dynamics of walking gait in male 27

28 Figure 1.3 (taken from Usherwood (2013)): Duty factor versus Froude number in (a) birds and (b) the forelimbs of quadrupedal mammals. Larger animals >10 kg have lower duty factors than smaller ones <10 kg. The uncertainty of Froude number in the mouse is given as the leg lengths were assumed to be between 2-3 cm. References include [2]: (Gatesy and Biewener 1991); [8]: (Reilly 2000); [9]: (Rubenson et al. 2004); [10]: (Biewener 1983); [11]: (Robilliard et al. 2007); [12]: (Witte et al. 2006); and [13]: (Heglund et al. 1982b) and female layer chickens artificially selected for different body size. I provide empirical evidence to support the hypothesis of Usherwood (2013) associated with the work and power demands associated with female egg laying. Scaling of locomotor energy metabolism Metabolic rates during terrestrial locomotion are size, speed and gait dependent (Rubenson et al. 2007; Taylor et al. 1982). Comparison of the locomotion energetics of different animals therefore requires methodology for the standardisation of measurements for meaningful comparison between different species. The increase in mass-specific metabolic rate (P met, W kg -1 ) with forward speed (U m s -1 ) is linear for the majority of species to have been examined (Taylor et al. 1982; Taylor et al. 1970; Brackenbury and Avery 1980; Fedak et al. 1974; Roberts et al. 1998c; Nudds et al. 2010; White et al. 2008; Maloiy et al. 2009; Bamford and Maloiy 1980). This is despite the wide variety in M b, body form and number of legs among terrestrial organisms. The consistency in this pattern of increase in P met with U provides a constant slope value for each species, equal to 28

29 the metabolic energy required to move a unit of M b a unit distance, i.e. the minimum cost of transport (CoT min, J kg -1 m -1 ). This method for calculating CoT min is referred to as the slope method (Taylor et al. 1982; Taylor et al. 1970; Brackenbury and Avery 1980; Fedak et al. 1974; Roberts et al. 1998c; Nudds et al. 2010; White et al. 2008; Maloiy et al. 2009; Bamford and Maloiy 1980). CoT min is widely reported to scale with negative allometry across species (i.e. with increasing species M b, these slopes become less steep) (Taylor et al. 1982; Taylor et al. 1970). In a study by Taylor et al (Taylor et al. 1982), new and previously published data for CoT min were combined for over 90 different species of birds and mammals, ranging in M b from 7g 260 kg, and CoT min scaled amongst them M b (Fig 1.5). Larger species, therefore, carry a unit M b over a unit distance more economically than do smaller ones. A number of additional studies have provided data in support of this interspecific allometric trend, with crustaceans, myriapods, reptiles and amphibians also following suit (Full and Tu 1991). The upper and lower limits of M b for which there are empirical measurements of CoT min from these studies have been extended to <1g using invertebrates (Full and Tu 1991) and to 1542 kg using African elephants, Loxodonta africana (Langman et al. 1995). Asian elephants, Elephas maximus, exceed this upper limit in M b (up to 7500 kg in bulls) (Langman et al. 2012). However, due to a quasi-intraspecific scaling of the CoT min with M b among elephants, their CoT min fall above the 95% confidence limits of the line (Langman et al. 1995). The addition of their CoT min to the available data in the literature, have the impact of significantly reducing the scaling exponent of the interspecific allometric equation for CoT min ( M b ). Intraspecific scaling of the CoT min is discussed further in Alternative methods are also used for calculating and comparing the CoT min of different species. For example, in humans (Margaria et al. 1963), elephants (Langman et al. 1995; Langman et al. 2012), emus and ostriches (Watson et al. 2011) and horses (Hoyt and Taylor 1981), the incremental response of P met to increasing walking U is curvilinear, meaning a single slope value cannot be calculated. Furthermore, non-linearity of the relationship between P met and U also occurs due to gait transition in a number species (Nudds et al. 2011; Baudinette et al. 1992; Watson et al. 2011; Baudinette and Biewener 1998; 29

30 Rubenson et al. 2004; Margaria et al. 1963). For example, in humans (Margaria et al. 1963) and emu (Watson et al. 2011), the P met response to increasing U is curvilinear and upon transition from pendular to bouncing gait mechanics, the relationship becomes linear and with deflection in incremental increase (Rubenson et al. 2007; Margaria et al. 1963). In the absence of linearity of this relationship within and between gaits, a slope can be calculated for each individual U (P met /U) and re-plotted against U (as shown in Fig 1.1). The minimum measured cost of transport, incurred at an optimal speed/s can then be determined from the response of CoT to U, or, CoT min can be estimated from the equations which best describe the CoT versus U relationship (Langman et al. 2012; Zani and Kram 2008). Most often when using this method, resting metabolic rate is subtracted from P met, before division by U and this method is referred to as the subtraction method (Maloiy et al. 2009; Rubenson et al. 2007). A reevaluation of the interspecific allometry of CoT min according to gait and generated from data using the subtraction method highlighted gait-specific allometric trends (Rubenson et al. 2007). Rubenson et al (2007) demonstrated that walking and running gaits scaled M b and M b, respectively. Extrapolating from these shows that these allometric equations intersect at a M b of ~20 kg (Rubenson et al. 2007). For M b below 20 kg, the running CoT min is expected to be lower than walking CoT min, and for M b above 20 kg, the walking CoT min is expected to be lower than the running CoT min. These findings highlight the importance for considering the energetics of differing gaits separately (Rubenson et al. 2007). In another study to reevaluate the scaling of the CoT min using existing data in the literature, a statistical model best described the data as having a dichotomy between small animals (<1 kg) and large animals (>1 kg). The same dichotomy was found for the scaling of the mechanical cost of locomotion (E mech, CM ) to M b. As a result, the efficiency of locomotion (E mech, CM / E met ) did not scale with M b within each of the groups, but was 7% and 26% in the smaller (more crouched limbed) and large (more upright limbed) animals, respectively. Here, it was highlighted that the relationship between efficiency and M b is likely linked to kinematics and posture, rather than body size alone. 30

31 Ultimately, however, the links between CoT min, mechanics and morphology are not fully understood. New insights into the links between morphology, mechanics and CoT min are gained, however, when species are found to have CoT min that deviate from allometric predictions calculated using interspecific allometric equations and M b, termed secondary allometric signals (Fig 1.4). For example, for a given M b there can be considerable unexplained variation in CoT min (above and below the line) associated with variation in body form (Full et al. 1990a). Elevetad CoT min in penguins (more than double the expected value for their M b ) have been attributed to their relatively shorter limbs adapted for diving locomotion and a consequential increase in the numberof strides they must take to cover a given distance, or f stride required to move at a given speed (Griffin and Kram 2000). Furthermore, a steeper negative allometry in CoT min ( M b ) amongst the ratites has been linked to their ecology due to their high activity patterms, which are likely to contribute towards a large part of their daily energy budget (Watson et al. 2011). Ornate box turtles have a particularly low CoT min which has been linked to the coevolution of their shoulder morphology and their protective shell as well as very slow efficient muscle fibres (Zani and Kram 2008). Many of the large deviations from the line have been linked to ecological niche (Bruinzeel et al. 1999) and dominant locomotor mode (Dawson and Taylor 1973; Fish 2000; Fish et al. 2001; Griffin and Kram 2000; Pinshow et al. 1977; Nudds et al. 2010). In Chapter 3 of this thesis I demonstrate that gravid female leghorn chickens of bantam and large varieties incur a lower CoT min than predicted based on M b and interspecific allometry. In the chapters to follow, I investigate potential morphological and mechanical mechanisms for the more metabolically economical locomotion of the females than expected for their M b. 31

32 Figure 1.4 (taken from Kram (2012)): Interspecifc allometry of the minimum metabolic cost of transport. From ants to elephants, the mass-specific minimum metabolic cost of transport decreased in body mass in a systematic manner (primary allometric signal). For a given body mass there is variation in CoT min, with values greater or lower than expected (secondary allometric signal) Figure 1.5 (taken from Reilly et al (2007)): Interspecific scaling of the total mass-specific external mechanical energy (E mech ) of the centre of mass (CoM) and mass-specific minimum metabolic cost of transport (CoT min ). E mech and CoT min lines are dashed (=1.07M b ; (Full 1991, 1989)) and solid (=10.8M b ; (Taylor et al. 1982)), respectively. E mech is constrant across animals of different body size 32

33 1.3.3 Linking energetics and mechanics A long-standing research goal has been to find an explanation for why metabolic rates increase with locomotor speed and why larger animals use less energy to move a unit M b over a unit distance (Taylor et al. 1982; Roberts et al. 1998b; Kram and Taylor 1990; Heglund and Taylor 1988; Heglund et al. 1982c; Heglund et al. 1982a; Fedak et al. 1982; Pontzer 2007b, 2005). The total mass-specific mechanical cost of transport (E mech ), which incorporates both the internal work done to swing the limbs and the external work done to lift and accelerate the body CoM, is independent of M b across species from 30 g quail to a 100 kg ostrich (Fig 1.5). Therefore, whilst CoT min scales M b, the muscles of animals of all sizes perform work at the same rate. Therefore, neither the incremental increase in P met with U, nor the negative interspecific allometry of CoT min can be explained by assuming a constant efficiency between the energy consumed, and mechanical work performed by the muscles (Heglund et al. 1982c). Another avenue explored for explaining these trends was to consider the influence of f stride on P met. A greater f stride for a given speed requires more limb movements. Dividing P met by f stride gives the metabolic cost per stride (J kg -1 m -1 ). Heglund and Taylor (Heglund and Taylor 1988) found that for over 16 species of quadrupedal mammals ranging in M b from 30 g 200 kg, their cost per stride was the same at equivalent speeds (e.g. minimum, preferred and maximum sustainable speeds). However, the cost per stride increased with speed in each species. No link has yet been found between mechanical work performed by the muscles and these metabolic trends during locomotion. The findings of these studies to have attempted to explain locomotor metabolic costs using mechanics altogether suggest that the metabolic costs may, rather, be associated with generating muscle force and activating the muscles. 33

34 1.3.4 The rate of force generation hypothesis In addition to performing mechanical work, muscles must generate force to support body weight and accelerate the body forwards. In an early study by Taylor et al (1980) the metabolic rates of rats, dogs, humans and horses were measured as they moved carrying trunk loads to manipulate M b and investigate the cost of generating muscular force. It was found that carrying an extra gram of mass in these animals was the same as carrying a unit of original M b, as the proportional increase in metabolic rate, was equal to the proportional increase in M b (Taylor et al. 1980). It was also found that generating a unit of force was more metabolically expensive for smaller than for larger animals. It was suggested that the higher metabolic costs in smaller animals were due to their greater f stride which would require them to generate muscle force faster, recruiting more metabolically expensive muscle fibres (Taylor et al. 1980). Due to the directly proportional relationship between added mass and metabolic cost, it was also deduced that the swinging of the limb incurs a negligible cost. A less than proportional increase would have been expected if there were a considerable cost to swinging the limb (Taylor et al. 1980). Based upon the findings of Taylor et al (Taylor et al. 1980), the rate of force generation hypothesis of Kram and Taylor (1990) was developed. The forces that the muscles generate must be equal to the product of the total volume of active muscle and the rate of energy consumption of the muscles. The hypothesis links these two variables to the amount of force that is required to support body weight (W b, N) and the rate at which muscle force is applied (equal to the inverse of the t stance ), respectively (Fig 1.6). Here, the principle dictator of the metabolic cost of locomotion is considered to be the cost of generating muscular force, which is heavily influenced by the time available to do so. Across mammals (Kram and Taylor 1990) and birds (Roberts et al. 1998b) ranging in body size, absolute metabolic rate (E met, W) is proportional to W b and inversely proportional to t stance (Kram and Taylor 1990). A single proportionality constant, the cost coefficient (c = [E met W -1 b ]/[1/t stance ], J N -1 ) accounts for the majority of variation in metabolic rate between species of differing sizes and with speed. A 34

35 morphological basis for both the increase in metabolic rate with speed and the interspecific scaling of CoT min was suggested to be leg length, as a longer leg is generally linked to a longer t stance, which would allow the recruitment of slower more energetically economical muscle fibres (Kram and Taylor 1990). Shorter legged animals, by comparison, would take shorter steps and have shorter t stance, presumably recruiting faster, more metabolically costly muscle fibres (Kram and Taylor 1990; Roberts et al. 1998c). A number of assumptions must be met for this hypothesis to stand. Muscle forces were assumed to be exerted, primarily, to oppose gravity. Furthermore, it was assumed that independent of size and speed, a unit volume of active muscle exerts the same force on the ground and that muscles operate over similar force-velocity relationship ranges. The rate of force generation hypothesis has been well accepted as a simple explanation for the energetic cost of running for some time by many authors. A number of studies, however, have demonstrated that additional factors influence c including the EMA of the limb and the length of the active muscle volume (Sih and Stuhmiller 2003; Roberts et al. 1998b). For example, c is on average 1.7 times higher in avian bipeds compared to mammals for a given rate of force generation (1/t stance ) associated with their more crouched limbs (Roberts et al. 1998b). In humans, c is also lower than in avian bipeds (Fig 1.6 C), associated with their more upright limbs. Furthermore, c, although considered a constant, appears to increase with speed throughout the literature (White et al. 2008; Nudds et al. 2010; Roberts et al. 1998b). Potential explanations for this increase in c with U include decreases in muscle mechanical advantage with U or a change in the relative shortening velocities of muscles (higher muscle shortening velocities would require a greater volume of active muscle to provide the same force (Roberts et al. 1997). Furthermore, evidence from studies measuring blood flow to the limb muscles of guinea fowl, Numida meleagris, during locomotion suggest that the swinging of the limb may actually comprise as much as one quarter of the metabolic cost (Marsh 2004). Studies have shown that metabolic rates increase directly with an increased moment of inertia when loads are added to the limb (Martin 1985; Myers and Steudel 1985; Steudel 1990). In Chapter 7 of this 35

36 Figure 1.6 (taken from (Roberts et al. 1998b)): Metabolic rate, the rate of force generation and the cost coefficient in birds and humans. The individual graphs represent (A) weight-specific metabolic rate (E met W b -1, J N -1 ); (B) the rate of muscle force generation (1/t stance ) and (C) the cost coefficient of proportionality ([E met W b -1 ] / [1/t stance ]). thesis, I examine the applicability of the rate of force generation hypothesis for explaining intraspecific variation in E met associated with the size of a leghorn variety and with sexual dimorphism. I provide evidence to support that variation in c is linked to differences in EMA, and that the cost coefficient is too simplistic for accounting for interspecific differences in E met. 36

37 1.3.5 Intraspecific variation in body size Body size can differ considerably between phylogenetically close species or between individuals within a species; however, variation in body form is likely to be less at this level than is found between phylogenetically more distant species. For example, within the felid clade, from domestic cats (~4 kg) to tigers (~170 kg), the geometry of the appendicular skeletal element lengths was demonstrated to be relatively consistent, as well as limb posture, despite the 40-fold range in body size (Day and Jayne 2007). Furthermore, artificially selected horses of miniature, Arabian and draft breeds (Equus ferus caballus) span an 8-fold range in M b and a 2-fold range in leg length, yet scale geometrically in their trunk, fore limb and hind limb lengths (Griffin et al. 2004). In trying to understand the factors that dictate the metabolic cost of locomotion, comparison of intraspecific variability in CoT min is complementary to interspecific comparisons as it allows for the investigation of the influence of body size whilst controlling for large morphological and phylogenetic differences. Few studies, however, have examined the intraspecific influence of body size upon CoT min (Griffin et al. 2004; Langman et al. 2012; Maloiy et al. 2009; Weyand et al. 2010; Tullis and Andrus 2011; Walton et al. 1994). Amongst western toads (Bufo boreas halophilus), ranging in M b from g, CoT min was invariant with M b. Across the three breeds of horses ( kg) in the study of Griffin et al (2004) CoT min also scaled M b (Fig 1.9). Furthermore, a similar pattern was found in an investigation into the metabolic cost of locomotion in the one-humped camel (Camelus dromedaries) and the domesticated donkey (Equus asinus) (Maloiy et al. 2009). In camels, CoT min was within 1% of values previously reported in individuals on average 2.4-fold greater in M b (Yousef et al. 1989); while, in donkeys, CoT min was within 9% of values recorded by Yousef et al (1972) for individuals on average 1.5-fold greater in M b. Similarly, in study by Langman et al (2012), between sub-adult African (Loxodonta Africana) and adult Asian (Elephas maximus) elephants ( kg), CoT min scaled M b. Therefore a pattern exists across these quadrupedal species for a lack of intraspecific scaling in CoT min, indicating no independent 37

38 influence of M b or leg length on CoT min. According to the rate of force generation hypothesis, however, the variation in limb length would lead to differences in the time available to generate muscle force and presumably cost (Kram and Taylor 1990). Therefore an explanation is required to account for the scaling of CoT min both intra- and interspecifically. The dynamic similarity hypothesis of Alexander and Jayes (Alexander and Jayes 1983) postulates that geometrically and dynamically similar animals will incur similar optimal relative walking speeds at which they incur their CoT min. It is assumed that the lack of intraspecific scaling in CoT min is due to geometric and physiological similarity between the smallest and largest individuals (Langman et al. 2012). An alternative hypothesis by Griffin et al (2004) is that, between individuals of differing body size that are geometrically similar in muscle proportions, skeletal morphometrics and posture, muscle stress increases with size M 1/3 b, because muscle stress = force ( M 1 b ) / area ( M 2/3 b ), and there is no sizerelated change in EMA to prevent this increase. Therefore, any metabolic savings associated with the longer legs of larger individuals (Kram and Taylor 1990) would be counterbalanced due to requiring a greater volume of active muscle to support body weight (Griffin et al. 2004). However, since metabolic, kinematic and morphological (muscle, bone, posture) measurements have not been collected in tandem in any of the studies to date, there is not yet any empirical evidence for this. In contrast to findings within quadrupedal species, in humans, greater CoT min are consistently measured in smaller compared to larger walkers (Weyand et al. 2010). The results of the intraspeific investigations into the scaling of CoT min to date, altogether, indicate that data are lacking and attract further investigation into the relationships between between M b, leg length, posture and skeletal geometry. Furthermore, the majority of this research has been conducted on large mammalian species with upright limb posture. Chapter 2 of this thesis expands on the available data by investigating the intraspecific scaling of CoT min in a small, crouch postured avian species. 38

39 1.3.6 Sex While a single species can show considerable variation in body size, yet comparable physiology and form: at the same time, clear deviations from geometric and physiological similarity can also be present due to sex (Bonnet et al. 2001; Liao et al. 2012; Lourdais et al. 2006). The sexes are fundamentally different, since they commonly apportion their lifetime somatic and reproductive efforts differently (Scantlebury et al. 2006; Lourdais et al. 2006). Life efforts regarding seeking, competing for and securing mates (e.g. through courtship, territorial or aggressive behaviours), reproducing and parental care are usually sex-specific. The locomotor systems of the sexes are therefore often under differing demands. A number of the morphological variables have been linked to differences in the metabolic cost of locomotion, including M b, leg length, skeletal proportions and posture, which are known to be sexually dimorphic (Baumel 1953; Hoglund 1989; Smith et al. 2002; Cho et al. 2004). Differences may also manifest at even lower organismal levels. For example, females may allocate more tissue to trophic structures and digestive organs specialized for energy acquisition towards reproduction (Hammond et al. 2000; Camilleri and Shine 1990). Males, by comparison, may have greater heart size, muscle volume or differing muscle fibre-type characteristics compared to females to support their role in obtaining mates (Hammond et al. 2000). Despite extensive documentation of sexual dimorphism across vertebrate species, however, few studies have investigated sex differences in locomotor physiology or biomechanics (Lees et al. 2011; Rose et al. 2014; Brackenbury and Elsayed 1985). In Chapter 3 of this thesis, I investigate the influence of sex on the metabolic cost of locomotion, maximum aerobic speed and gait utilization in two varieties of leghorn chicken selected for small and large body size. I present the first finding of a lower metabolic cost of locomotion in a female when compared to a male. In the subsequent chapters of the thesis, I explore potential mechanisms for the sexual dimorphism in locomotion energetics further by examining gait dynamics (Chapter 4); making comparisons with birds at an earlier ontogenetic stage, preceding sexual maturity (Chapter 5); comparing male 39

40 and female skeletal muscle architectural properties (Chapter 6); and combining metabolic and kinematic data to examine the cost of muscle force production (Chapter 7). 1.4 Domestic leghorn chickens (Gallus gallus domesticus) The leghorn chicken (Gallus gallus domesticus) has a wide variety of body size form and physiological specialization. Leghorns are layer strains, artificially selected for egg production efficiency (to produce the maximum size and number of eggs for the minimum amount of food intake). Females consequently reach the onset of gravidity early in the ontogenetic process, which is then continuous throughout their lives rather than seasonal. The energy allocation of females is therefore biased towards reproduction. Male leghorns, by comparison, invest their energy in larger body size and greater somatic tissue allocation. Males are the more active sex, associated with mate competition. Males partake in aggressive, aerobically sustained combats in order to maintain social status, territory and access to females (Guhl et al. 1945). Males also perform a number of courtship activities that are not observed in females including flapping of the wings and tidbitting (Guhl et al. 1945). Linked to the differences in life efforts between the sexes, males generally outweigh females in anatomical components related to the locomotor system including muscle, bone, blood and heart mass. Females, contrastingly, have greater digestive components, skin and fat as well (Mitchell et al. 1931). The sexes, however, share similar initial post-hatch growth trajectories in M b until sexual maturation (4-5 months old) (Mitchell et al. 1931). Differentiation of secondary sexual characteristics is mediated by a rise in gonadal hormones and male muscle growth rates increase relative to females. Female growth also terminates before that of males, leading to strong male-biased sexual size dimorphism (Mitchell et al. 1931). Leghorns are also selected for standard breed (large) and bantam (small) varieties, providing an opportunity to investigate the influences of size and sex in these birds. 40

41 1.5 Overview and thesis aims Understanding the factors that can influence the metabolic requirements of locomotion provides insight into the behavior and evolution of animals and their locomotor systems. Given the diversity in body form across differing species to have resulted from long-term evolutionary processes, interspecific comparative approaches offer insight into morphological and physiological constraints on locomotor attributes. Although the evolutionary allometry of locomotor morphology, gait kinematics and metabolic costs is well reported, the underlying mechanisms are still a matter of debate. The plethora of confounding morphological and physiological variables (e.g. body size, geometry, number of legs, respiratory systems and muscle specialization) between disparate species makes it difficult to tease apart the independent effects of particular morphologies or gait kinematics on the metabolic cost of locomotion. A single species, however, can vary considerably in body size, but show relatively consistent morphological geometry and physiology (e.g. between artificially selected strains). Furthermore, a single species can also show variation in form and physiology influencing the locomotor system (e.g due to sexual dimorphism) whilst other variables remain consistent. Intraspecific comparative approaches therefore allow for the investigation into the independent effects of particular morphological characteristics. Furthermore, they provide a novel approach for further investigating existing hypotheses for the evolutionary allometry of locomotor morphology, gait kinematics and metabolic costs. Ultimately, explanations for both inter- and intraspecific scaling patterns are required. Little is understood, however, regarding the factors that influence locomotor attributes at the intraspecific level. 41

42 The principal objective of my PhD is to elucidate constraints on the terrestrial locomotor system within a single species using an integrative approach: quantifying and linking morphology, gait kinematics (using videography) and energy metabolism (using respirometry) during their locomotion. I use the leghorn chicken as my study species and make use of its similarities and differences in locomotor morphology and physiology to have resulted from selective breeding, linked to variety, sex and ontogenetic stage The research presented in this thesis is divided into six chapters, each suitable for stand-alone publication. The chapters address the following questions that have not yet been investigated: Chapter 2: How are locomotor morphology and kinematics linked to the intraspecific scaling of the minimum metabolic cost of transport? Chapter 3: How does sexual dimorphism influence locomotor energy metabolism and can any differences be explained by body size? Chapter 4: What are the selected gaits and postures of animals according to varied anatomical proportions influencing muscle force, work and power demands at the intraspecific level? Chapter 5: Do sexual dimorphisms in locomotor energetics and kinematics arise concurrently with secondary sex characteristics? Chapter 6: What are the links between pelvic limb muscle architectural properties and locomotor performance? Chapter 7: Can body weight and the rate of muscle force generation account for intraspecific variations in locomotor metabolic rates associated with body size and sexual dimorphism? 42

43 The specific aims of my thesis are: 1) To determine the intraspecific influence of body size on the minimum metabolic cost of transport (CoT min ) by comparing two varieties of leghorn artificially selected for large and small body size (standard breeds and bantams, respectively). I will test the hypothesis that CoT min will not scale hypoallometrically as is found between species, due to physiological, geometric and dynamic similarity. 2) To examine the effect of sex on energy metabolism during terrestrial locomotion in standard breed and bantam leghorns. I will test the hypothesis that males and females will differ in their patterns of energy metabolism with treadmill speeds, linked to their differential investments in somatic and reproductive tissue allocation. 3) To examine the influence of body size and sex on gait dynamics in standard breed and bantam leghorns. I will test the hypothesis that the sexes will employ differing gait dynamics associated with egg laying in females and greater limb muscle mass in males. 4) To investigate the ontogeny of sex-specific energetics and kinematics of terrestrial locomotion in standard breed leghorn chickens. I will test the hypothesis that the onset of sexual maturity (i.e. the onset of egg-laying and increased male muscle mass) will bring about further divergence in the energetics and kinematics of the sexes. 5) To quantify the hind limb skeletal muscle architectural properties of the birds used in these studies: mature and immature standard breeds of both sexes and bantams of both sexes. I will test the hypotheses that leghorn will exhibit variety, sex and age-specific muscle architecture. 43

44 6) To examine the relationships between metabolic and kinematic parameters for male and female bantam and standard breed leghorns. I will test the hypothesis that the rate of muscle force generation will account for intraspecific variation in the metabolic cost of locomotion. 44

45 1.6 Alternative format This PhD thesis is presented in the alternative format for examination at the University of Manchester. The six research chapters presented here have either been published as articles in peer-reviewed journals or are in the required format for submission for publication. Chapters 2-3 are presented as they are published in their journals and Chapters 4-7 are presented as final drafts in article format for submission. Page numbers and reference lists are exclusive to each results chapter. References provided in Chapters 1 and 8 (introduction and discussion) are listed at the end of the thesis. I am the first author of each of the chapters. The contributions of the other authors to each chapter are listed below. 2. Rose KA, Nudds RL, Codd JR (2015) Intraspecific scaling of the minimum metabolic cost of transport in leghorn chickens (Gallus gallus domesticus): links with limb kinematics, morphometrics and posture. Journal of Experimental Biology, 218, I conducted the experiments, analysed and interpreted the data and drafted the manuscript with editing assistance from Dr Jonathan Codd and Dr Robert Nudds. All authors contributed to the design of the study. The initial research idea was generated by Dr Jonathan Codd and DrRobert Nudds. 3. Rose KA, Nudds RL, Butler PJ, Codd JR (2015) Sex differences in gait utilization and energy metabolism during terrestrial locomotion in two varieties of leghorn chicken (Gallus gallus domesticus) selected for different body size. Biology Open, 4, I conducted the experiments, analysed and interpreted the data and drafted the manuscript with editing assistance from Dr Jonathan Codd, Dr Robert Nudds and Prof Patrick Butler. All authors contributed to the designed of the study. The study was conceived by Dr Jonathan Codd, Dr Robert Nudd and Prof Patrick Butler. 45

46 4. Rose KA, Codd JR, Nudds RL (in review) Differential sex-specific gait dynamics in leghorn chickens (Gallus gallus domesticus) selectively bred for different body size I conducted the experiments, analysed and interpreted the data and drafted the manuscript with editing assistance from Dr Jonathan Codd and Dr Robert Nudds. All authors contributed to the design of the study. 5. Rose KA, Nudds RL, Bates KB Codd JR (in review) Ontogeny of sex differences in the energetics and kinematics of terrestrial locomotion in leghorn chickens (Gallus gallus domesticus) I conducted the experiments analysed and interpreted the data and drafted the manuscript with editing assistance from Dr Jonathan Codd, Dr Robert Nudds and Dr Karl Bates. All authors contributed towards the design of the study. The initial research idea was my own. Dr Jonathan Codd provided the equipment and Dr Karl Bates provided the birds. 6. Rose KA, Nudds RL, Codd JR (in review) Variety, sex and ontogenetic differences in the hind limb skeletal muscle architectural properties of leghorn chickens (Gallus gallus domesticus) and their links with locomotor performance I conducted the experiments, analysed and interpreted the data and drafted the manuscript with editing assistance from Dr Jonathan Codd, Dr Robert Nudds. The initial research idea was my own. 46

47 7. Rose KA, Codd JR, Nudds RL (in preparation) Sex differences in the metabolic cost of locomotion are not accounted for by the rate of muscle force generation in two varieties of leghorn chicken (Gallus gallus domesticus) I conducted the analyses through combining data from Chapters 3 and 4. I interpreted the data and drafted the manuscript with editing assistance from Dr Jonathan Codd and Dr Robert Nudds. The research idea was my own. 47

48 2. Intraspecific scaling of the minimum metabolic cost of transport in leghorn chickens (Gallus gallus domesticus): links with limb kinematics, morphometrics and posture This chapter is a reprint of an article published in The Journal of Experimental Biology Rose, K. A., Nudds, R. L, and Codd, J. R. (2015). Intraspecific scaling of the minimum metabolic cost of transport in leghorn chickens (Gallus gallus domesticus): links with limb kinematics, morphometrics and posture. The Journal of Experimental Biology, 218,

49 2015. Published by The Company of Biologists Ltd The Journal of Experimental Biology (2015) 218, doi: /jeb RESEARCH ARTICLE Intraspecific scaling of the minimum metabolic cost of transport in leghorn chickens (Gallus gallus domesticus): links with limb kinematics, morphometrics and posture Kayleigh A. Rose, Robert L. Nudds and Jonathan R. Codd* ABSTRACT The minimum metabolic cost of transport (CoT min ;Jkg 1 m 1 ) scales negatively with increasing body mass ( M 1/3 b ) across species from a wide range of taxa associated with marked differences in body plan. At the intraspecific level, or between closely related species, however, CoT min does not always scale with M b. Similarity in physiology, dynamics of movement, skeletal geometry and posture between closely related individuals is thought to be responsible for this phenomenon, despite the fact that energetic, kinematic and morphometric data are rarely collected together. We examined the relationship between these integrated components of locomotion in leghorn chickens (Gallus gallus domesticus) selectively bred for large and bantam (miniature) varieties. Interspecific allometry predicts a CoT min 16% greater in bantams compared with the larger variety. However, despite 38% and 23% differences in M b and leg length, respectively, the two varieties shared an identical walking CoT min, independent of speed and equal to the allometric prediction derived from interspecific data for the larger variety. Furthermore, the two varieties moved with dynamic similarity and shared geometrically similar appendicular and axial skeletons. Hip height, however, did not scale geometrically and the smaller variety had more erect limbs, contrary to interspecific scaling trends. The lower than predicted CoT min in bantams for their M b was associated with both the more erect posture and a lower cost per stride (J kg 1 stride 1 ). Therefore, our findings are consistent with the notion that a more erect limb is associated with a lower CoT min and with the previous assumption that similarity in skeletal shape, inherently linked to walking dynamics, is associated with similarity in CoT min. KEY WORDS: Terrestrial locomotion, Size, Body mass, Geometric similarity, Energetics INTRODUCTION Body size has a significant influence on the morphology and metabolism of animals (Schmidt-Nielsen, 1975, 1984; Biewener, 1989). In animals that locomote terrestrially, the absolute amount of metabolic energy required to move a given distance increases with increasing body size, but not in direct proportion (slope <1) (Bruinzeel et al., 1999; Halsey and White, 2012). In relative terms, the mass-specific energy per unit distance (the cost of transport, Faculty of Life Sciences, University of Manchester, Manchester M13 9PT, UK. *Author for correspondence ( jonathan.codd@manchester.ac.uk) This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( which permits unrestricted use, distribution and reproduction in any medium provided that the original work is properly attributed. Received 18 July 2014; Accepted 26 January 2015 CoT; J kg 1 m 1 ) is lower in larger species than in smaller ones. Often, at optimal self-selected speeds within a gait, animals incur a minimum cost of transport (CoT min ) and it seems reasonable to expect natural selection to favour strategies that minimise the CoT min. For example, if the movement requirements of animals were similar, they would be expected to share optimum limb dynamics, and similar morphological proportions to allow it (Alexander and Jayes, 1983). The evolutionary allometry of CoT min with body mass (M b, kg) is widely reported. For example, across more than 90 species of mammals and birds (7 g to 260 kg), CoT min =10.7M 0.32 b (Taylor et al., 1982). Adding amphibians, reptiles and invertebrates (<1 g) to this data set yielded a similar result (CoT min =10.8M 0.32 b ; Full and Tu, 1991) and African elephants (Loxodonta africana, M b =1542 kg) fall within the 95% confidence intervals (CIs) of this equation (Langman et al., 1995). The scaling exponent, however, is known to differ between walking and running (Margaria et al., 1963; Minetti et al., 1999; Rubenson et al., 2004, 2007; Maloiy et al., 2009; Nudds et al., 2011; Watson et al., 2011), and also between small crouched- and large uprightpostured vertebrates (Reilly et al., 2007; Nudds et al., 2009). Furthermore, there is overlooked variation in CoT min at a given M b, associated with variation in body form (Full et al., 1990). The general trend of decreasing CoT min with M b, however, holds for over three orders of magnitude. Where outliers exist, their relatively more or less economical CoT min compared with other species of the same M b is attributed to adaptations associated with activity patterns (Watson et al., 2011), dominant locomotor mode (Dawson and Taylor, 1973; Fish et al., 2000, 2001; Griffin and Kram, 2000; Nudds et al., 2010), ecological niche (Bruinzeel et al., 1999), climate (Yousef et al., 1989; Maloiy et al., 2009) or having a protective shell (Baudinette et al., 2000; Zani and Kram, 2008). Ultimately, the reasons underlying the allometry of CoT min with M b and the factors that determine the CoT are not yet fully understood (Cavagna et al., 1977; Fedak et al., 1982; Heglund et al., 1982a,b; Heglund and Taylor, 1988; Kram and Taylor, 1990; Roberts et al., 1998; Pontzer, 2005, 2007a,b). Between disparate species, musculoskeletal morphology and shape vary with size (Schmidt-Nielsen, 1975, 1984; Biewener, 1989; Reilly et al., 2007), speed requirements (Garland, 1983), climate (Janis and Wilhelm, 1993), ecological niche (Bruinzeel et al., 1999) and locomotor mode (Griffin and Kram, 2000; Abourachid, 2001; Nudds et al., 2010). Within species or between closely related species, however, variation in shape is reduced, meaning insight can be gained into the factors that dictate the CoT and how it scales with M b independent of shape (Griffin et al., 2004; Day and Jayne, 2007; Langman et al., 2012). For example, miniature, Arabian and draft horses (Equus ferus caballus)showednodifferenceincot min when trotting, despite spanning 8- and 2-fold differences in M b and leg length, respectively (Griffin et al., 2004). Similarly, there was little The Journal of Experimental Biology 1028

50 RESEARCH ARTICLE The Journal of Experimental Biology (2015) 218, doi: /jeb List of symbols and abbreviations CoT min minimum cost of transport CoT net net cost of transport CoT tot total cost of transport f stride stride frequency h hip hip height l skel skeletal leg length l stride stride length M b body mass net-p met net metabolic power P met metabolic power RMR resting metabolic rate t stance stance duration t swing swing duration U speed V CO 2 rate of carbon dioxide production V O 2 rate of oxygen consumption difference in walking CoT min within camels (Camelus dromedaries, M b = kg) (Yousef et al., 1989; Maloiy et al., 2009) or donkeys (Equus asinus, M b = kg) (Yousef et al., 1972; Maloiy et al., 2009), or between adult Asian elephants (Elephas maximus) and sub-adult African elephants (M b = kg) (Langman et al., 1995, 2012). It is assumed that similarity in CoT min across individuals of differing body masses is due to their being geometrically, posturally and physiologically similar and locomoting with dynamically similar gaits (Griffin et al., 2004; Langman et al., 2012). Surprisingly, despite this explanation being widespread in the literature, there is no empirical evidence linking CoT min across a size range with similar limb kinematics and skeletal proportions for a walking gait (the only gait over which dynamic similarity can be investigated; Alexander and Jayes, 1983). In humans, the only bipedal species to have been examined across a size range (children adults), walking CoT min scaled in a similar manner to that found across species (i.e. M b 1/3 ) (Weyand et al., 2010), which is contrary to findings from within quadruped investigations where CoT min was similar across sizes. To fully understand these results, it is necessary to expand the available data for bipeds and to investigate the relationships between the CoT, M b, limb kinematics and skeletal proportions. Domestic leghorn chickens, Gallus gallus domesticus (Linnaeus 1758), are selectively bred for large and bantam (miniature) varieties, providing an opportunity to investigate how size influences CoT min independent of shape in an avian species. Rubenson et al. (2007) derived an interspecific scaling equation of walking CoT min against M b [CoT min =17.80(±2.98)M b 0.471(±0.032) ] using minimum measured values of the net cost of transport (CoT net ; the amount of energy required to move 1 kg over 1 m minus maintenance and postural costs) for a range of birds and mammals ( kg). The aim of this study was to investigate whether large (N=5; mean±s.e.m. M b =1.92±0.13 kg, range= kg) and bantam (N=9; M b =1.39±0.03 kg, range= kg) leghorns would show a 16% difference in CoT min as predicted by the Rubenson et al. (2007) equation, and to compare their CoT min with that of animals of a similar M b. Importantly, we simultaneously determined whether the two varieties of leghorn walked in a dynamically similar way and were geometrically and posturally similar to gain insight into the links between these integrated components of terrestrial locomotion. RESULTS Morphological measurements Mean linear dimensions measured from large and bantam leghorns are presented in Table 1. The skeletal measurements of the bantams Table 1. Hindlimb segment measurements and sternal keel lengths from the birds used in experiments and geometric predictions for appendicular measurements of the bantams Length/ width Bantam (mm) Large (mm) Bantam prediction (mm)* l keel 90.00± ±5.29 l fem 71.04± ± l tib ± ± l tars 75.45± ± l skel ± ± w fem 7.78± ± w tib 7.18± ± w tars 7.57± ± h hip ± ± l keel, keel length; l fem, femur length; l tib, tibiotarsus length; l tars, tarsometatarsus length; l skel, skeletal leg length; w fem, femur width; w tib, tibiotarsus width; w tars, tarsometatarsus width; h hip, hip height. *Predicted value ranges for the bantams were calculated as (large linear dimension±s.e.m.) 0.84 based on the percentage difference in keel length between the varieties. Bold values represent geometric predictions that were not significantly different from observed bantam measurements. N=6 for bantam sternum measurements. h hip measurements are given to the nearest millimetre because the measurements were made in metres to the nearest millimetre. were, on average, 83% of those of the larger variety. Predicted hindlimb dimensions (Table 1) for the bantams, based on the percentage difference in sternum length between the two varieties, all fell within the range predicted from the large variety data (mean± s.e.m.), indicating that the axial and appendicular skeletons of the two varieties were geometrically similar. Independent samples t-tests (equal variances assumed unless otherwise stated) showed that, represented as a proportion of total skeletal leg length (l skel =femur+tibiotarsus+tarsometatarsus lengths), the femur (0.28 in both varieties) was not significantly different (equal variances not assumed: Levene s test, F=13.71, P=0.003) between varieties (t=1.00, d.f.=4, P=0.374). Similarly, the tibiotarsus (t=0.07, d.f.=12, P=0.948) and tarsometatarsus lengths (t= 1.26, d.f.=12, P=0.233) were the same proportion of total leg length in the two varieties (0.42 and 0.30, respectively). Femur width, as a proportion of femur length was also similar (t=1.63, d.f.=12, P=0.128) between the two varieties (0.11 and 0.10 in bantam and large leghorns, respectively). Similarly, the tibiotarsus width:length ratio (0.07 in both varieties) did not differ (equal variances not assumed: Levene s test, F=5.25, P=0.041) between varieties (t=1.07, d.f.=5.70, P=0.326) and nor did the tarsometatarsus width/length ratio, which was 0.10 in both (t=0.00; d.f.=12, P=1.00). The two varieties therefore shared similar hindlimb skeletal proportions. The ratio of hip height to skeletal leg length, h hip :l skel, a measure of posture (Gatesy and Biewener, 1991), was on average 5% greater in the bantam compared with the large variety (0.79±0.02 and 0.74±0.01, respectively), but was not statistically different between varieties (t=1.96, d.f.=12, P=0.074). The predicted h hip for the bantams (Table 1), however, fell outside of the range predicted from the large variety s h hip data, being approximately 1 cm shorter than measured. Bantam h hip was 0.87 times that of the larger birds, which was a greater fraction than found for the skeletal element measurements. Therefore, the bantams adopted a more erect posture compared with the large variety. Walking kinematics Duty factor decreased linearly with speed (U,ms 1 ) and neither the slope nor the intercept of this relationship differed between varieties (Fig. 1A, Table 2). Stride frequency ( f stride, Hz) increased at the The Journal of Experimental Biology 1029

51 RESEARCH ARTICLE The Journal of Experimental Biology (2015) 218, doi: /jeb Duty factor lstride (m) A C fstride (Hz) Duration (s) U (m s 1 ) B D t stance t swing Fig. 1. Relationships between kinematics parameters and walking speed U. Filled circles and solid lines represent data for bantam leghorns and open circles and dashed lines represent data for large leghorns. The lines of best fit are (A) duty factor= 0.18U (bantam) and 0.18U+0.78 (large); (B) stride frequency, f stride =1.51U+0.83 (bantam) and 1.51U+0.46 (large); (C) stride length, l stride =0.36U+0.13 (bantam) and 0.36U+0.23 (large); and (D) swing time, t swing =0.16U 0.22 (bantam) and 0.21U 0.22 (large); and stance time, t stance =0.28U 0.64 (bantam) and 0.36U 0.64 (large). Data points are means±s.d. (s.e. are not large enough to be seen). same rate with U in the two varieties, but was 0.37 Hz greater in the bantam variety across all U (Fig. 1B, Table 2). Similarly, the incremental increase in stride length (l stride, m) with U was the same in the two size groups, whilst l stride was longer by 0.09 m across all U in the large variety (Fig. 1C, Table 2). The duration of the swing phase of the limb (t swing, s) decreased curvilinearly with U at the same rate in the two groups, but was 0.05 s longer in the large variety across all U (Fig. 1D, Table 2). Stance phase duration (t stance,s) also decreased curvilinearly with U and at the same rate in the two size groups. t stance was, however, 0.08 s longer in the large variety across all U (Fig. 1D, Table 2). Therefore, each parameter responded to increasing U the same way in the two varieties and differences in their absolute values (related to size) were fixed across all speeds. Metabolic power and CoT The positive relationship between mass-specific metabolic power (P met,wkg 1 ) and walking U (Fig. 2A) was similar (both the slopes and intercepts) for the two varieties (Table 2). Calculating CoT min as the slope of this relationship (slope method) therefore gives J kg 1 m 1 in each variety. During quiet standing, resting metabolic rate (RMR, W kg 1 ) did not differ (Fig. 2A, Table 2) between bantam and large leghorns (7.24±0.42 and 7.21± 0.48 W kg 1, respectively), indicating that they shared the same mass-specific energetic cost of general maintenance and maintaining their posture combined. Therefore, the relationship between net mass-specific metabolic power (net-p met,wkg 1 : the metabolic rate required for locomotion exceeding that required for standing quietly) and U (Fig. 2A) was also similar for the two size groups (Table 2). Total cost of transport (CoT tot,jkg 1 m 1 ) decreased curvilinearly with U, indicating that the highest walking speeds of the birds were most metabolically optimal. CoT net (J kg 1 m 1 ; net-p met /U), however, was not correlated with U and fell within a similar range for the two size groups (bantam: J kg 1 m 1 ; large: J kg 1 m 1 ) (Fig. 2B, Table 2). Calculating CoT min as the minimum measured CoT net (subtraction method), taken as the mean of all CoT net values across all speeds and both varieties, gives J kg 1 m 1. Predicted walking CoT min values for large and bantam leghorns based on Rubenson et al. (2007) were and J kg 1 m 1, respectively. Both varieties therefore shared a CoT min closer to that predicted for the larger variety, contrary to the 16% difference predicted. This corresponds to the bantams having a CoT min 14% lower than predicted for their M b, which fell within the 95% CIs of Rubenson et al. s (2007) equation. The net cost per stride (J kg 1 stride 1 ) was lower in bantams than in the larger variety by 1.17 J kg 1 stride 1 across all speeds (Fig. 2C, Table 2). DISCUSSION Across species, CoT min is reported to scale hypoallometrically with M b (Taylor et al., 1970, 1982; Fedak et al., 1974; Kram and Taylor, 1990; Full and Tu, 1991; Langman et al., 1995; Roberts et al., 1998). However, we found that bantam and large varieties of leghorn chickens have identical CoT min despite the smallest and largest individuals differing 1.7-fold in M b and 1.35-fold in leg length. An independence of CoT min from body size was previously reported within large quadrupedal species (>90 kg) spanning 1.5- to 8-fold ranges in M b and up to 2-fold ranges in leg length (Griffin et al., 2004; Maloiy et al., 2009; Langman et al., 2012). The present data represent the first evidence of a lack of correlation between M b and CoT min within an avian species. No effect of M b or leg length suggests that size itself does not influence the CoT but, rather, some other factor, perhaps correlated with body size, may be responsible. The simultaneous collection of kinematics and morphological data here allow us to investigate further previous hypotheses on what is driving the interspecific CoT min versus M b relationship. Larger species perform the same amount of mass-specific mechanical work as smaller species, whilst using less mass-specific metabolic energy during terrestrial locomotion (Fedak et al., 1982; Heglund et al., 1982a,b; Alexander, 2005). How this is possible is not fully understood. It is generally accepted that M b has no independent influence over CoT (Pontzer, 2005, 2007a,b). Leg length, however, is often discussed as the morphological factor explaining the allometry of CoT min (Kram and Taylor, 1990; Schmidt, 1984; Biewener, 2003; Alexander, 2003) as longer legs allow longer t stance for the muscles to apply force through recruiting slower, less metabolically expensive muscle fibres (metabolic rate is inversely proportional to t stance during which the muscles apply force) (Kram and Taylor, 1990). In addition, longer limbs allow lower f stride, requiring fewer muscle contractions. In the present study, however, the different sized birds shared the same mass-specific CoT min, despite the bantams having shorter limbs, shorter t stance and higher f stide compared with the larger variety. Using the maximum height of the limb as a strut (effective limb length, h hip ) as the indicator of size The Journal of Experimental Biology 1030

52 RESEARCH ARTICLE The Journal of Experimental Biology (2015) 218, doi: /jeb Table 2. Results of GLMs that tested for differences in metabolic and kinematic measurements between chicken varieties Parameter Covariate/ factor interaction GLM1 d.f. F P GLM2 d.f. F P 2 n p Observed power r 2 Duty factor U 1, < , < Variety 1, , Variety U 1, * * * * * f stride (Hz) U 1, < , < Variety 1, , < Variety U 1, * * * * * l stride (m) U 1, < , < Variety 1, , < Variety U 1, * * * * * log 10 t swing (s) log 10 U 1, < , < variety 1, , < Variety log 10 U 1, * * * * * log 10 t stance (s) log 10 U 1, < , < Variety 1, , < Variety log 10 U 1, * * * * * RMR (W kg 1 ) Variety 1, * * * * * 0.05 P met (W kg 1 ) U 1, < , < Variety 1, , Variety U 1, * * * * * Net-P met (W kg 1 ) U 1, < , < Variety 1, , Variety U 1, * * * * * log 10 CoT tot (J kg 1 m 1 ) log 10 U 1, < , < Variety 1, , Variety log 10 U 1, * * * * * CoT net (J kg 1 m 1 ) U 1, , Variety 1, , Variety U 1, * * * * * Net cost per stride U 1, < , (J kg 1 stride 1 ) Variety 1, , Variety U 1, * * * * * f stride, stride frequency; l stride, stride length; t swing, swing duration; t stance, stance duration; RMR, resting metabolic rate; P met, metabolic power; net-p met, net metabolic power; CoT tot, total cost of transport; CoT net, net cost of transport. Speed (U, ms 1 ) is a covariate, chicken variety is a fixed factor and variety U is the interaction term in the models. d.f. are represented as (d.f., error d.f.). The adjusted r 2 values are reported for second GLM analyses. Variables that did not have a significant effect on parameters were not included in second GLM analyses and are represented by an asterisk. has been shown to better predict CoT min across species (h hip, r 2 =0.98) than using the sum of the skeletal element lengths (l skel, r 2 =0.78) (Steudel and Beattie, 1995; Pontzer, 2007a). Over a small size scale of analysis, however, it has been demonstrated that between-individual differences in limb arrangement (e.g. limb excursion angle), the cost of swinging the limb and the coefficient of converting metabolic energy into muscle force k (which were not measured in this study) prevent a clear relationship between h hip and CoT min (Pontzer, 2005, 2007b). In agreement with Pontzer s (2005, 2007b) findings, despite the greater absolute h hip of the larger variety, compared with the bantams, they did not have a lower CoT min. It may be that variation in limb excursion angle (i.e. the difference in posture), rather than h hip, dominated variation in CoT min. Indeed, by using a model to predict the rate of force production associated with both supporting body weight and swinging the limb as a function of all of these parameters, Pontzer (2007a) found this was a better predictor of metabolic rate than contact time, limb length or M b at both interspecific and intraspecific levels. Equally, the shared CoT min of the two varieties may be due to their identical appendicular and axial skeletal geometry, consistent with previous assumptions in intraspecific analyses (Langman et al., 2012). Another potential explanatory factor is limb posture (linked to effective limb length). Across vertebrates, the limb bone lengths scale positively and almost geometrically with M b, but become increasingly more aligned with one another and less crouched (Biewener, 1989). A prominent step-change exists in the scaling of both CoT min and the mechanical cost of transport (E mech ; Jkg 1 m 1 ) across species associated with crouched postures in those <1 kg and upright postures in those >1 kg, making their efficiency of transport (CoT min /E mech ) approximately 7% and 26%, respectively (Reilly et al., 2007; Nudds et al., 2009). Unlike larger species with a more upright posture, small crouched-postured (noncursorial) species do not benefit from elastic energy savings or pendular mechanisms (Reilly et al., 2007). Furthermore, a more vertical limb decreases the muscular force required to support a unit of body weight and improves the mechanical advantage of the muscles (Biewener, 1989). The change in posture with increasing size means that muscle stress is nearly independent of M b across species (rather than M b 1/3 ). Griffin et al. (2004) suggested that between closely related individuals, consistent limb posture might account for consistent CoT min across a range of body sizes as muscle stress would in this case scale geometrically ( M b 1/3 ). The volume of active muscle would therefore increase with size and counter any metabolic savings associated with having longer legs (Griffin et al., 2004). However, in the present study the shared CoT min of the chicken groups did not correspond to a similar posture. When comparing the posture of the two size groups as h hip :l skel, the limbs The Journal of Experimental Biology 1031

53 RESEARCH ARTICLE The Journal of Experimental Biology (2015) 218, doi: /jeb Pmet (W kg 1 ) CoT (J kg 1 m 1 ) Net cost per stride (J kg 1 stride 1 ) B A C P met Net-P met CoT tot CoT net U (m s 1 ) Fig. 2. Relationships between mass-specific energetic parameters and walking speed. Data points and best-fit lines are as in Fig. 1. The lines of best fit are (A) metabolic power, P met =16.20U+6.93 (bantam) and 16.20U+5.86 (large); and net metabolic power, net-p met =16.00U 0.88 (bantam) and 16.00U 1.26 (large); (B) total cost of transport, CoT tot =22.39U 0.50 (bantam) and 19.95U 0.50 (large); and net cost of transport, CoT net =4.77U (bantam) and 4.77U (large); and (C) net cost per stride=7.10u+2.42 (bantam) and 21.21U+0.24 (large). Mass-specific resting (standing) metabolic rates are also included in A at 0 m s 1. Data points are means±s.e.m. were 5% more erect in the variety selected for smaller size. The shared CoT min in this case is perhaps better explained by the posture and lower cost per stride of the bantams. Across avian species, h hip represents a greater proportion of l skel with increasing M b (Gatesy and Biewener, 1991). One potential explanation for why we found the opposite to what would be expected, as well as the lower cost per stride in the bantams, may be that the two varieties differ in their derived muscle properties or architecture as a result of selective breeding. The kinematic data indicate that with U, the two varieties shared identical rates of change in all parameters, which would be expected to imply geometric, postural and dynamic similarity. Each kinematic parameter differed between the two varieties only by a fixed value across all speeds. The larger variety took longer strides by 9 cm, took less frequent strides by 0.37 Hz and had longer durations of both swing and stance phases of the limb by 0.05 and 0.08 s, respectively. At a given absolute U, duty factor is generally higher in larger species than in smaller ones (Gatesy and Biewener, 1991); however, the duty factors of the chickens were not significantly different between size groups. Similarly, a selection of felid species spanning a 46-fold range in M b were found to use similar duty factors at a similar walking speed (Day and Jayne, 2007). For what was previously an expectation (Griffin et al., 2004; Maloiy et al., 2009; Langman et al., 2012), the present data offer the first empirical evidence of a link between identical walking CoT min in individuals of differing size and similar limb dynamics and skeletal geometry. We can speculate that for a given skeletal shape, regardless of M b, walking CoT min may be consistent. Some additional studies in which shape was controlled for also support this idea. For example, adding back loads up to 50% of M b has a negligible effect on the CoT in quadrupedal rats, dogs and horses as well as bipedal humans, guinea fowl and other birds (Taylor et al., 1980; Ellerby and Marsh, 2006; Tickle et al., 2010, 2013). Furthermore, obese and thin humans of the same height (likely to be similar in skeletal proportions) show no difference in CoT min (Browning et al., 2006). In contrast to our findings, a comprehensive study of 48 humans spanning a 6-fold range in M b and 1.5-fold range in height concluded that CoT min was M b 1/3 (Weyand et al., 2010). This result, however, may be associated with ontogenetic differences in shape, because the human subjects ranged from 5 to 32 years of age and the data were intentionally separated into four size groups to reduce individual variability (Weyand et al., 2010). Indeed, dividing the CoT by body height accounted for the observed differences between the human size groups. Therefore, at any given speed, all subjects incurred the same CoT to cover the same horizontal distance relative to their own body height (Weyand et al., 2010). Small (2 g) ghost crabs (Ocypode quadrata), one of the few invertebrate species examined, were found to have a higher CoT than larger ones (47 g), despite their similar appearance in shape (Tullis and Andrus, 2011). In the absence of detailed kinematic and morphometric measurements, however, it is not possible to conclude much from this result. It is, of course, possible that the link we found here between energetics, kinematics and skeletal morphometrics may not be characteristic of species with more than two legs. Conclusions Leghorn chickens selectively bred for large and bantam varieties shared the same walking CoT min despite a 1.70-fold difference in M b and fold difference in total leg length between the smallest and largest individuals. These data represent the first evidence of CoT min being independent of M b within a small crouched-postured bipedal species. Our findings also provide the first evidence (for what was previously only assumed) of a link between this and similar walking dynamics and skeletal geometry. In contrast to interspecific trends, however, h hip did not scale geometrically between varieties and represented a greater proportion of total leg length in the bantam variety compared with the large variety. All birds shared a CoT min closer to that predicted for the larger variety and the CoT min of the bantams was approximately 14% lower than predicted from their M b. Our findings are therefore in agreement with the general consensus that for a given body size, CoT min decreases with limb erectness. The lower than predicted CoT min in the bantams was also associated with lower mass-specific energy requirements per stride, compared with the larger variety, which may be linked to differences in their posture and/or their derived muscle morphology/physiology. We emphasise the importance of intraspecific in addition to interspecific investigations as well as the combination of kinematics, morphometric and posture measurements towards gaining insight into the factors that dictate CoT. MATERIALS AND METHODS Study species Adult (>16 week) male bantam (N=9; mean±s.e.m. M b =1.39±0.03 kg) and large (N=5; M b =1.92±0.13 kg) leghorn chickens were purchased from a local The Journal of Experimental Biology 1032

54 RESEARCH ARTICLE The Journal of Experimental Biology (2015) 218, doi: /jeb breeder and housed in the University of Manchester s animal unit. All housing was maintained on a 13 h:11 h light:dark cycle, at C. Food and water were provided ad libitum, and the birds were not fasted prior to experiments. Birds were trained for 1 week to locomote on a motorised treadmill (T60 Tunturi, Finland) prior to data collection. All experiments were carried out in accordance with the Animals (Scientific Procedures) Act 1986, were approved by the University of Manchester Ethics Committee and performed under a UK Home Office Project Licence held by J.R.C. (40/3549). Respirometry An open flow respirometry system (all equipment Sable Systems International, Las Vegas, NV, USA) was used to measure the birds rates of oxygen consumption (V O2, ml min 1 ) and carbon dioxide production (V CO2, ml min 1 ). Perspex respirometry chambers were built (bantam: cm, large: cm) and mounted upon the treadmill. Air was pulled through the chambers using a FlowKit 500 at flow rates (FR) of 150 l min 1 (bantam) and 250 l min 1 (large). Excurrent airflow was sub-sampled (0.11 l min 1 ) for gas analysis. Water vapour pressure (WVP) was measured using an RH-300 water vapour analyser before the air was scrubbed of H 2 O with calcium chloride (2 6 mm granular, Merck, Darmstadt, Germany) and passed through a CO 2 analyser (CA-10A). The dry air was scrubbed of CO 2 using soda lime (2 5 mm granular, Sigma-Aldrich, Steinheim, Germany) and passed through a dual absolute and differential O 2 analyser (Oxilla II). Ambient air (scrubbed of H 2 O and CO 2 as before) was simultaneously passed through a second O 2 channel on the Oxilla II at 0.11 l min 1 by a pump (SS-3) to enable calculation of differential O 2 concentration (ΔO 2 ). CO 2 traces were baselined to calculate differential CO 2 concentration (ΔCO 2 ). Voltage outputs were recorded using a UI2 interface and analysed using ExpeData v software. The accuracy of the respirometry set up (±5%) across all speeds was determined using a N 2 dilution test (Fedak et al., 1981). Primary flow rates (FR) were adjusted to dry-corrected flow rates (FR c ), to account for the H 2 O scrubbed from air samples prior to gas measurements using: FR ðbp WVPÞ FR C ¼ ; ð1þ BP where BP is barometric pressure (measured with the Oxilla II) and WVP is water vapour pressure (Lighton, 2008). V O2 was calculated using (Lighton, 2008): V_ O2 ¼ FR CðDO 2 Þ 1 0:2095 and V O2 using (Lighton, 2008): _ V CO2 ¼ FR CðDCO 2 Þ 0:0004ð _ 1 0:0004 V O2 Þ ð2þ : ð3þ The birds were exercised over a range of randomised speeds (three per day) up to the maximum sustainable (bantam: m s 1, large: m s 1 ). Birds were given a rest of a minimum of 5 min to stand quietly between each period of exercise. RMRs were taken from the final rest period of each trial. Data were collected from stable gas readings lasting >1 min. Only data from speeds at which both varieties used a walking gait (0.28, 0.42, 0.56 and 0.69 m s 1 ) were included in analyses. Metabolic rate calculations Five values were calculated at each speed: (1) P met was converted from V O2, using respiratory exchange ratios (RERs: V O2 :V O2 ) and thermal equivalents taken from Brody (1945); (2) net-p met was calculated by subtracting RMR from locomotor P met (both from the same trial); (3) CoT tot was calculated as P met /U; (4) CoT net was calculated as net-p met /U; and (5) the cost per stride was calculated as net-p met /f stride. CoT min was calculated using two methods: first, as the slope of the linear relationship between P met and U (slope method) and, second, as the minimum measured CoT net (subtraction method). CoT min values calculated using the subtraction method were compared with predictions for walking birds and mammals of a similar M b using eqn 3 from Rubenson et al. (2007). Gait kinematics The birds were filmed (100 frames s 1 ) at all speeds in lateral view using a video camera (HDR-XR520VE, Sony, Japan). The left foot of each bird was tracked ( 10 strides) at each speed using Tracker software (v. 4.05, Open Source Physics) in order to quantify duty factor, f stride, l stride (U/f stride ), t stance and t swing. Fluctuations in the kinetic and potential energy of the centre of mass (CoM) across a stride were determined through frame-by-frame tracking of a marker positioned over the left hip joint of the birds (indicative of h hip ). To ensure that the birds were using a walking gait at all speeds analysed, the phase relationship between the horizontal kinetic energy (E kh ) and the sum of the potential and vertical kinetic energies (E p +E kv ) of the CoM (h hip ) was determined. An out-of-phase relationship, indicating a walking gait, was found for all speeds used in the analyses. Morphological measurements Keel length and the length and width (mid-shaft) of the right femur, tibiotarsus and tarsometatarsus was measured from the birds used in the respirometry experiments using digital vernier calipers (accuracy, ± 0.01 mm). Geometric similarity in linear dimensions between the two size groups was investigated by determining whether their axial and appendicular dimensions scaled 1:1. The mean appendicular dimensions of the bantams were predicted based on the ratio of their keel length to that of the large variety. Skeletal element lengths were also compared as a percentage of total leg length. The ratio of h hip to total skeletal leg length (l skel =femur+tibiotarsus+tarsometatarsus lengths) was calculated and used as a means of comparing posture between the two size groups, with a lower value indicating a more crouched posture. Back height (h back, m) was measured during the mid-stance as the distance from the hindtoe to the back at 90 deg to the direction of travel. Where birds (N=3) did not walk with easewith a hip marker, the ratio h hip :h back (bantam: 0.80±0.01, large: 0.77 ±0.00) was used to estimate h hip. Statistical analyses The slopes and the intercepts of the relationships between the dependent variables (metabolic or kinematics measures) and U were investigated for differences between chicken varieties using general linear models (GLMs). Models included variety as a fixed factor, U as a covariate and the interaction term variety U. If the interaction term was non-significant (indicating similar slopes between varieties), it was removed from the model and the updated model was re-run (assuming parallel lines) in order to test for differences in intercepts. Where the relationship between a dependent variable and U was curvilinear, the data were log 10 transformed. All best-fit lines were taken from coefficients tables produced by the GLMs. Between-variety differences in hindlimb skeletal element proportions (% total leg length) were investigated using independent samples t-tests. Hindlimb proportion data were tested for equality of variance using a Levene s test for equality of variance. Acknowledgements We would like to thank John Lees and Karlina Ozolina for their assistance with respirometry data collection. Competing interests The authors declare no competing or financial interests. Author contributions The study was conceived and designed by J.R.C. and R.L.N. K.A.R. executed the study. Data were interpreted and analysed by K.A.R. with assistance from R.L.N. and J.R.C. K.A.R., R.L.N. and J.R.C. drafted and revised the manuscript. Funding This research was supported through funding provided by the Biotechnology and Biological Sciences Research Council (BBSRC: G01138/1 and I /1 to J.R.C.). K.A.R. was supported by a Natural Environment Research Council (NERC) doctoral training account (DTA) PhD stipend and Collaborative Awards in Science and Engineering (CASE) partnership with The Manchester Museum. Deposited in PMC for immediate release. References Abourachid, A. (2001). Kinematic parameters of terrestrial locomotion in cursorial (ratites), swimming (ducks), and striding birds (quail and guinea fowl). Comp. Biochem. Physiol. A Mol. Integr. Physiol. 131, The Journal of Experimental Biology 1033

55 RESEARCH ARTICLE The Journal of Experimental Biology (2015) 218, doi: /jeb Alexander, R. M. (2003). Principles of Animal Locomotion. Princeton, NJ: Princeton University Press. Alexander, R. M. (2005). Models and the scaling of energy costs for locomotion. J. Exp. Biol. 208, Alexander, R. M. and Jayes, A. S. (1983). A dynamic similarity hypothesis for the gaits of quadrupedal mammals. J. Zool. 201, Baudinette, R. V., Miller, A. M. and Sarre, M. P. (2000). Aquatic and terrestrial locomotory energetics in a toad and a turtle: a search for generalisations among ectotherms. Physiol. Biochem. Zool. 73, Biewener, A. A. (1989). Scaling body support in mammals: limb posture and muscle mechanics. Science 245, Biewener, A. A. (2003). Animal Locomotion, pp Oxford: Oxford University Press. Brody, S. (1945). Bioenergetics and Growth, With Special Reference to The Efficiency Complex In Domestic Animals. New York: Reinhold. Browning, R. C., Baker, E. A., Herron, J. A. and Kram, R. (2006). Effects of obesity and sex on the energetic cost and preferred speed of walking. J. Appl. Physiol. 100, Bruinzeel, L. W., Piersma, T. and Kersten, M. (1999). Low costs of terrestrial locomotion in waders. Ardea 87, Cavagna, G. A., Heglund, N. C. and Taylor, C. R. (1977). Mechanical work in terrestrial locomotion - two casic mechanisms for minimizing energy-expenditure. Am. J. Physiol. 233, R243-R261. Dawson, T. J. and Taylor, C. R. (1973). Energetic cost of locomotion in kangaroos. Nature 246, Day, L. M. and Jayne, B. C. (2007). Interspecific scaling of the morphologyand posture of the limbs during the locomotion of cats (Felidae). J. Exp. Biol. 210, Ellerby, D. J. and Marsh, R. L. (2006). The energetic costs of trunk and distal-limb loading during walking and running in guinea fowl Numida meleagris: II. Muscle energy use as indicated by blood flow. J. Exp. Biol. 209, Fedak, M. A., Pinshow, B. and Schmidtn, K. (1974). Energy cost of bipedal running. Am. J. Physiol. 227, Fedak, M. A., Rome, L. and Seeherman, H. J. (1981). One-step N2-dilution technique for calibrating open-circuit VO2 measuring systems. J. Appl. Physiol. 51, Fedak, M. A., Heglund, N. C. and Taylor, C. R. (1982). Energetics and mechanics of terrestrial locomotion. II. Kinetic energy changes of the limbs and body as a function of speed and body size in birds and mammals. J. Exp. Biol. 97, Fish, F. E., Frappell, P. B., Baudinette, R. V. and MacFarlane, P. M. (2000). Energetics of terrestrial locomotion of the platypus: metabolic inefficiencies due to aquatic adaptation. Am. Zool. 40, Fish, F. E., Frappell, P. B., Baudinette, R. V. and MacFarlane, P. M. (2001). Energetics of terrestrial locomotion of the platypus Ornithorhynchus anatinus. J. Exp. Biol. 204, Full, R. J. and Tu, M. S. (1991). Mechanics of a rapid running insect: two-, four- and six-legged locomotion. J. Exp. Biol. 156, Full, R. J., Zuccarello, D. A. and Tullis, A. (1990). Effect of variation in form on the cost of terrestrial locomotion. J. Exp. Biol. 150, Garland, T. (1983). The relation between maximal running speed and body mass in terrestrial mammals. J. Zool. 199, Gatesy, S. M. and Biewener, A. A. (1991). Bipedal locomotion: effects of speed, size and limb posture in birds and humans. J. Zool. 224, Griffin, T. M. and Kram, R. (2000). Biomechanics: penguin waddling is not wasteful. Nature 408, 929. Griffin, T. M., Kram, R., Wickler, S. J. and Hoyt, D. F. (2004). Biomechanical and energetic determinants of the walk-trot transition in horses. J. Exp. Biol. 207, Halsey, L. G. and White, C. R. (2012). Comparative energetics of mammalian locomotion: humans are not different. J. Hum. Evol. 63, Heglund, N. C. and Taylor, C. R. (1988). Speed, stride frequency and energy-cost per stride - how do they change with body size and gait. J. Exp. Biol. 138, Heglund, N. C., Cavagna, G. A. and Taylor, C. R. (1982a). Energetics and mechanics of terrestrial locomotion.3. Energy changes of the center of mass as a function of speed and body size in birds and mammals. J. Exp. Biol. 97, Heglund, N. C., Fedak, M. A., Taylor, C. R. and Cavagna, G. A. (1982b). Energetics and mechanics of terrestrial locomotion.4. Total mechanical energy changes as a function of speed and body size in birds and mammals. J. Exp. Biol. 97, Janis, C. M. and Wilhelm, P. B. (1993). Were there mammalian pursuit predators in the Tertiary? Dances with wolf avatars. J. Mamm. Evol. 1, Kram, R. and Taylor, C. R. (1990). Energetics of running: a new perspective. Nature 346, Langman, V. A., Roberts, T. J., Black, J., Maloiy, G. M. O., Heglund, N. C., Webers, J. M., Kram, R. and Taylor, C. R. (1995). Moving cheaply - energetics of walking in the African elephant. J. Exp. Biol. 198, Langman, V. A., Rowe, M. F., Roberts, T. J., Langman, N. V. and Taylor, C. R. (2012). Minimum cost of transport in Asian elephants: do we really need a bigger elephant? J. Exp. Biol. 215, Lighton, J. R. B. (2008). Measuring Metabolic Rates: A Manual For Scientists. Oxford; New York: Oxford University Press. Maloiy, G. M. O., Rugangazi, B. M. and Rowe, M. F. (2009). Energy expenditure during level locomotion in large desert ungulates: the one-humped camel and the domestic donkey. J. Zool. 277, Margaria, R., Sassi, G., Aghemo, P. and Cerretelli, P. (1963). Energy cost of running. J. Appl. Physiol. 18, Minetti, A. E., Ardigo, L. P., Reinach, E. and Saibene, F. (1999). The relationship between mechanical work and energy expenditure of locomotion in horses. J. Exp. Biol. 202, Nudds, R. L., Codd, J. R. and Sellers, W. I. (2009). Evidence for a mass dependent step-change in the scaling of efficiency in terrestrial Locomotion. PLoS ONE 4, e6927. Nudds, R. L., Gardiner, J. D., Tickle, P. G. and Codd, J. R. (2010). Energetics and kinematics of walking in the barnacle goose (Branta leucopsis). Comp. Biochem. Physiol. A Mol. Integr. Physiol. 156, Nudds, R. L., Folkow, L. P., Lees, J. J., Tickle, P. G., Stokkan, K.-A. and Codd, J. R. (2011). Evidence for energy savings from aerial running in the Svalbard rock ptarmigan (Lagopus muta hyperborea). Proc. R. Soc. B Biol. Sci. 278, Pontzer, H. (2005). A new model predicting locomotor cost from limb length via force production. J. Exp. Biol. 208, Pontzer, H. (2007a). Effective limb length and the scaling of locomotor cost in terrestrial animals. J. Exp. Biol. 210, Pontzer, H. (2007b). Predicting the energy cost of terrestrial locomotion: a test of the LiMb model in humans and quadrupeds. J. Exp. Biol. 210, Reilly, S. M., McElroy, E. J. and Biknevicius, A. R. (2007). Posture, gait and the ecological relevance of locomotor costs and energy-saving mechanisms in tetrapods. Zoology 110, Roberts, T. J., Kram, R., Weyand, P. G. and Taylor, C. R. (1998). Energetics of bipedal running I. Metabolic cost of generating force. J. Exp. Biol. 201, Rubenson, J., Heliams, D. B., Lloyd, D. G. and Fournier, P. A. (2004). Gait selection in the ostrich: mechanical and metabolic characteristics of walking and running with and without an aerial phase. Proc. R. Soc. B Biol. Sci. 271, Rubenson, J., Heliams, D. B., Maloney, S. K., Withers, P. C., Lloyd, D. G. and Fournier, P. A. (2007). Reappraisal of the comparative cost of human locomotion using gait-specific allometric analyses. J. Exp. Biol. 210, Schmidt-Nielsen, K. (1975). Scaling in biology: the consequences of size. J. Exp. Zool. 194, Schmidt-Nielsen, K. (1984). Scaling, Why is Animal Size So Important? Cambridge; New York: Cambridge University Press. Steudel, K. and Beattie, J. (1995). Does limb length predict the relative energetic cost of locomotion in mammals? J. Zool. 235, Taylor, C. R., Schmidtn, K. and Raab, J. L. (1970). Scaling of energetic cost of running to body size in mammals. Am. J. Physiol. 219, Taylor, C. R., Heglund, N. C., Mcmahon, T. A. and Looney, T. R. (1980). Energetic cost of generating muscular force during running - a comparison of large and small animals. J. Exp. Biol. 86, Taylor, C. R., Heglund, N. C. and Maloiy, G. M. O. (1982). Energetics and mechanics of terrestrial Locomotion.1. Metabolic energy-consumption as a function of speed and body size in birds and mammals. J. Exp. Biol. 97, Tickle, P. G., Richardson, M. F. and Codd, J. R. (2010). Load carrying during locomotion in the barnacle goose (Branta leucopsis): the effect of load placement and size. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 156, Tickle, P. G., Lean, S. C., Rose, K. A. R., Wadugodapitiya, A. P. and Codd, J. R. (2013). The influence of load carrying on the energetics and kinematics of terrestrial locomotion in a diving bird. Biol. Open 2, Tullis, A. and Andrus, S. C. (2011). The cost of incline locomotion in ghost crabs (Ocypode quadrata) of different sizes. J. Comp. Physiol. B 181, Watson, R. R., Rubenson, J., Coder, L., Hoyt, D. F., Propert, M. W. G. and Marsh, R. L. (2011). Gait-specific energetics contributes to economical walking and running in emus and ostriches. Proc. R. Soc. B Biol. Sci. 278, Weyand, P. G., Smith, B. R., Puyau, M. R. and Butte, N. F. (2010). The massspecific energy cost of human walking is set by stature. J. Exp. Biol. 213, Yousef, M. K., Freeland, D. V. and Dill, D. B. (1972). Energetic cost of grade walking in man and burro, Equus-asinus - desert and mountain. J. Appl. Physiol. 33, Yousef, M. K., Webster, M. E. D. and Yousef, O. M. (1989). Energy costs of walking in camels, Camelus dromedarius. Physiol. Zool. 62, Zani, P. A. and Kram, R. (2008). Low metabolic cost of locomotion in ornate box turtles, Terrapene ornata. J. Exp. Biol. 211, The Journal of Experimental Biology 1034

56 3. Sex differences in gait utilization and energy metabolism during terrestrial locomotion in two varieties of chicken (Gallus gallus domesticus) selected for different body size This chapter is a reprint of an article published in Biology Open Rose, K. A., Nudds, R. L, Butler, P. J. and Codd, J. R. (2015). Sex differences in gait utilization and energy metabolism during terrestrial locomotion in two varieties of chicken (Gallus gallus domesticus) selected for different body size Biology Open, 4,

57 2015. Published by The Company of Biologists Ltd Biology Open (2015) 4, doi: /bio RESEARCH ARTICLE Sex differences in gait utilization and energy metabolism during terrestrial locomotion in two varieties of chicken (Gallus gallus domesticus) selected for different body size Kayleigh A. Rose 1, Robert L. Nudds 1, Patrick J. Butler 2 and Jonathan R. Codd 1, * ABSTRACT In leghorn chickens (Gallus gallus domesticus) of standard breed (large) and bantam (small) varieties, artificial selection has led to females being permanently gravid and sexual selection has led to male-biased size dimorphism. Using respirometry, videography and morphological measurements, sex and variety differences in metabolic cost of locomotion, gait utilisation and maximum sustainable speed (U max ) were investigated during treadmill locomotion. Males were capable of greater U max than females and used a grounded running gait at high speeds, which was only observed in a few bantam females and no standard breed females. Body mass accounted for variation in the incremental increase in metabolic power with speed between the varieties, but not the sexes. For the first time in an avian species, a greater mass-specific incremental cost of locomotion, and minimum measured cost of transport (CoT min ) were found in males than in females. Furthermore, in both varieties, the female CoT min was lower than predicted from interspecific allometry. Even when compared at equivalent speeds (using Froude number), CoT decreased more rapidly in females than in males. These trends were common to both varieties despite a more upright limb in females than in males in the standard breed, and a lack of dimorphism in posture in the bantam variety. Females may possess compensatory adaptations for metabolic efficiency during gravidity (e.g. in muscle specialization/posture/kinematics). Furthermore, the elevated power at faster speeds in males may be linked to their muscle properties being suited to inter-male aggressive combat. KEY WORDS: Birds, Metabolic rate, Sexual dimorphism, Gravidity, Posture, Mechanics INTRODUCTION Many avian species exhibit sexual dimorphism in morphology, physiology and behaviour, linked to differential specialization of the sexes towards mate competition, reproduction and parental care (Dunn et al., 2001). With the few exceptions of reverse sexual size dimorphism, where females are the larger sex (Reynolds, 1972; Hakkarainen et al., 1996; Pande and Dahanukar, 2012), males are often larger than females and these size differences are more pronounced in cursorial species (Hoglund, 1989). Furthermore, the 1 Faculty of Life Sciences, University of Manchester, Manchester M139PT, UK. 2 School of Biosciences, University of Birmingham, Birmingham B152TT, UK. *Author for correspondence ( jonathan.codd@manchester.ac.uk) This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( which permits unrestricted use, distribution and reproduction in any medium provided that the original work is properly attributed. Received 22 June 2015; Accepted 17 August 2015 relative proportions of the skeleton (Baumel, 1953), skeletal muscle and viscera may differ between the sexes (Hammond et al., 2000). Physiological performance traits (e.g. maximum aerobic capacity, maximum speed, endurance and metabolic costs) may also be expected to be sex-specific (Husak and Fox, 2008). Previous studies investigating physiological differences between the sexes in birds have focused on maximum performance and aerobic limits and/or scopes (Chappell et al., 1996, 2011; Hammond et al., 2000). Despite well documented influences of body size and shape on the mechanics and energetics of locomotion (Taylor et al., 1982; Alexander and Jayes, 1983), however, the influence of sexual dimorphism on locomotor performance in birds has been given little attention (Brackenbury and Elsayed, 1985; Lees et al., 2012; Rose et al., 2014). The metabolic cost of terrestrial locomotion has been investigated across a wide range of avian species. Most studies have focused on interspecific comparisons to understand scaling patterns with respect to body mass (M b ) and deviations from these patterns associated with body form and locomotor specialization. Usually in these studies, only one sex is considered (Nudds et al., 2010); the sex of the experimental animal is not specified (Taylor et al., 1971, 1982; Fedak et al., 1974; Pinshow et al., 1977; Roberts et al., 1998; White et al., 2008), or male and female data are pooled (Bamford and Maloiy, 1980; Bruinzeel et al., 1999; Ellerby et al., 2003; Rubenson et al., 2004; Ellerby and Marsh, 2006; Watson et al., 2011; Tickle et al., 2013). The potential for sex differences in locomotor performance has been investigated in very few avian species and different studies have produced varying results. For example, male Svalbard rock ptarmigan (Lagopus muta hyperborea) were shown to have lower mass-specific metabolic power (P met ;Wkg 1 ) requirements than females at any given treadmill speed, despite the sexes sharing similar M b (Lees et al., 2012). Furthermore, males achieved greater maximum sustainable speeds (U max ) by 50% and used aerial running gaits, whereas females did not (Lees et al., 2012). These results are consistent with the life history differences between the sexes, whereby male ptarmigan defend vast territories to secure mates and females, who are less active, provide parental care to chicks (Steen and Unander, 1985; Unander and Steen, 1985). In contrast, in the common eider (Somateria mollissima), a diving bird, no sex differences in gait choice, P met or U max were found despite males being 16 18% heavier than females (Rose et al., 2014). The similar locomotor performance of the sexes in eiders is consistent with the short amount of time that each sex spends using terrestrial locomotion, which is important for spring breeding and incubation, but not for securing mates (Portugal and Guillemette, 2011). Without knowledge on the morphological sexual dimorphisms of a species, however, it is difficult to understand any underlying mechanisms behind differences in locomotor performance. Biology Open 1306

58 RESEARCH ARTICLE Biology Open (2015) 4, doi: /bio List of abbreviations CoM centre of mass CoT min minimum cost of transport CoT net net cost of transport CoT tot total cost of transport E kh horizontal kinetic energy E kv vertical kinematic energy E p potential energy net-p met net metabolic power P met metabolic power RMR resting metabolic rate U speed U max maximum sustainable speed _V CO2 rate of carbon dioxide production rate of oxygen consumption _V O2 Domestic layer chickens (Gallus gallus domestics) are a useful species with which to investigate sex constraints on locomotor performance. Not only has artificial selection led to females being permanently gravid but male-biased sexual size dimorphism is common to both wild ancestral and derived chickens due to sexual selection (Remes and Szekely, 2010). The sex-specific behaviours (Guhl et al., 1945; Schutz et al., 2001), morphologies and physiologies (Mitchell et al., 1931; Whitehead, 2004; Remes and Szekely, 2010) of layer breeds are also well documented. For example, males compete with one another for social status, territory and access to females through sustained, aggressive, combats. Furthermore, males partake in courtship activities including feeding, crowing (Chappell et al., 1995; Horn et al., 1995; Wilson et al., 2008), wing dipping and flapping (Chappell et al., 1997). Females, in comparison, invest energy in reproduction (van Kampen, 1976a) and are the sole providers of parental care. To suit these specializations, males possess greater relative anatomical weights of the bones, skeletal muscles, heart and blood, whilst females outweigh males in digestive components, flesh and fat (Mitchell et al., 1931; Hammond et al., 2000). In a study by Brackenbury and Elsayed (1985), it was hypothesized that the sexes of layer chickens would differ in the metabolic cost of locomotion due to differences in the proportions of total metabolic energy devoted to reproduction (Brackenbury and Elsayed, 1985). Yet, no differences in mass-specific metabolic rates or the incremental cost of locomotion (also known as the minimum cost of transport, CoT min :Jkg 1 m 1 ) were found (Brackenbury and Elsayed, 1985). This lack of a difference is despite the fact that interspecific scaling of the CoT min with M b, would predict larger males to have a lower CoT min than smaller females. The male and female chickens in (Brackenbury and Elsayed, 1985), however, were from different strains meaning their results are difficult to interpret. Sex differences may not be consistent across chicken strains, which can differ markedly in size and other morphological measurements, depending on the reasons for which they were selectively bred (Paxton et al., 2010). In this study, we used videography and respirometry to compare male and female gait utilization, U max and metabolic rates over a range of treadmill speeds in standard breed (large, L and L ) and bantam (miniature, B and B ) varieties of leghorn chicken. We tested the hypothesis that sex would lead to greater differences in locomotor energy metabolism than variety, as the varieties are expected to be physiologically and geometrically similar (Rose et al., 2015). In addition, using morphological measurements taken from the birds, we compared the CoT of the birds at equivalent values of dimensionless speed defined by the Froude number (Fr=U 2 /gh hip, where U is walking speed, g is acceleration due to gravity and h hip is hip height) (Alexander and Jayes, 1983). Gravid females were expected to show a lower capacity for locomotion than males through a lower U max and fewer gaits utilized. RESULTS Sexual dimorphism As expected M b, h hip and Σ l seg (the sum of the hind limb skeletal element lengths) were greater in the standard than in the bantam variety (Table 1). M b was also 27% and 34% greater in males than in females in the small and large varieties, respectively (Table 1). Similarly, Σ l seg was 16% and 20% greater in males than in females in the small and large varieties, respectively (Table 1). Therefore, the sexual size dimorphism of these varieties did not scale geometrically, and was greater in the standard breed. An interaction between variety and sex for Σ l seg was found because of a greater difference in size between L and B (54.67 mm), than between L and B (38.63 mm). A significant interaction between variety and sex for h hip was also found because h hip was mm taller in B compared to B, whereas in the standard breed, h hip was mm taller L compared to L (the opposite pattern) (Table 1). Consequently, sexual dimorphism in limb posture index (h hip :Σ l seg ) was present in only the standard variety, whereby female limb posture was 23% more erect than that of the males (Table 1). Gaits With exception of L and B, which shared similar U max (Table 1), U max differed between groups (Χ 2 =17.41, d.f.=3, P<0.001) and was greater in males compared to females by 15% and 25% in bantam and standard breed leghorns, respectively. None of the birds in Table 1. Mean (±s.e.m) morphological measurements, maximum sustainable speeds and walk-grounded run transition speeds for the four chicken variety/sex combinations Group N a M b (kg) h hip (mm) Σ l seg (mm) Posture index b U max (m s 1 ) Transition (m s 1 ) Male bantam ± ± ± ± ±0.06 >0.69<0.97 Female bantam ± ±9.15 (N=5) ± ±0.03 (N=5) 0.75±0.08 >0.69<0.97 Male standard ± ± ± ± ±0.06 >0.97<1.25 Female standard ± ± ± ± ±0.04 No transition Two-way ANOVAs were performed to test for differences between varieties and sexes in M b (variety sex, F 1,24 =3.07, P=0.093; variety, F 1,25 =59.40, P<0.001; sex, F 1,25 =45.80, P<0.001), h hip (variety sex, F 1,22 =13.17, P=0.001; variety, F 1,22 =53.42, P<0.001; sex, F 1,22 =0.88, P=0.359), Σ l seg (variety sex, F 1,24 =5.11, P=0.033; variety, F 1,24 =170.76, P<0.001; sex, F 1,24 =137.02, P<0.001) and log posture index (variety sex, F 1,22 =20.13, P<0.001; variety, F 1,22 =6.39, P=0.019; sex, F 1,22 =8.87, P=0.007). Abbreviated measurements include body mass (M b ), hip height (h hip ), leg length (sum of hind limb skeletal element lengths, Σ l seg ) and maximum sustainable speed (U max ). a Sample size unless otherwise stated adjacent to the relevant mean value. b Posture indices were calculate as h hip :Σ l seg. Biology Open 1307

59 RESEARCH ARTICLE Biology Open (2015) 4, doi: /bio this study had duty factors below 0.5; therefore, they did not use aerial running gaits. In L, the maximum speed (U) at which the horizontal kinetic energy (E kh ) of the body centre of mass (CoM) was observed to fluctuate out-of-phase with the sum of the vertical kinetic and potential energy (E kv +E p ) of the CoM (walking gait mechanics, Fig. 1A) was 1.11 m s 1 (2 of 5 individuals). From m s 1 the E kh and E kv +E p of their CoM were in-phase (Fig. 1B), indicating that they used grounded running gaits. At the U max of the L, however, the E kh and E kv +E p of the CoM were out-of-phase indicating that they were still walking. In bantams of either sex, E kh and E kv +E p of the CoM were out-of-phase at speeds up to and including 0.83 m s 1, and in-phase from speeds of 0.83 m s 1 and greater, indicating that the sexes utilized walking and grounded running gait mechanics over similar speed ranges. However, only 3 of 7 females could sustain 0.83 m s 1, at which speed one individual was still walking. The same 3 B could sustain 0.97 m s 1 and were all grounded running at this speed. Therefore, most B and all L were either unwilling or incapable of performing a grounded running gait. Resting metabolic rates During quiet standing, RMR (P met, W) was positively correlated with M b (Table 2) and the slopes and intercepts of this relationship were similar between sexes and varieties (means were B : 10.70± 0.50, B : 8.54±0.41, L : 13.80±0.66 and L : 9.25±0.44). Likewise, mass-specific RMR (P met,wkg 1 ) was similar (Table 2) between sexes and varieties (means were B : 7.85±0.27, B : 7.13±0.57, L : 7.21±0.48 and L : 7.24±0.42). Walking metabolic power Absolute P met (W) was correlated with M b and U during walking (Fig. 2A-B) and increased curvilinearly (Fig. 3A-B) with U in all birds (Table 2). The incremental response to U was steeper in the bantams compared to the standards, but this difference was not significant when accounting for M b (Table 2). M b, however, did not explain the greater incremental response to U in males than in females (Table 2). A B normalized energy (J) normalized energy (J) percentage of stride Fig. 1. Examples of typical mechanical energy fluctuations of the CoM. (A) Walking gait (0.69 m s 1 in a L, 2.19 kg). (B) Grounded running gait (1.39 m s 1 in a L, 2.19 kg). Solid lines and the left y-axis represent horizontal kinetic energy (E kh ) of the CoM, and the dotted lines and the right y-axis represent vertical kinetic plus potential energies (E kv +E p ). normalized energy (J) normalized energy (J) Mass-specific P met (W kg 1 ) was positively correlated with U in all bird groups (Fig. 3C-D) and was best described by power curves. The exponents of these curves were common to both varieties with the incremental increase in mass-specific P met with U greater in males compared to females (Table 2). Mass-specific P met was lower across all U in the males of the larger variety than in males of the bantams, and likewise in females. Calculating mass-specific net-p met, by subtracting P met during quiet standing from P met, did not account for this sex difference (Table 2), but did reduce the net metabolic rates (intercepts) of the bantam variety relative to the large variety (Table 2). Again, net mass-specific P met increased with U, with higher exponents and intercepts in males than in females, and similar exponents, and intercepts for the males and females of each variety (Table 2). Therefore, the sexes shared similar metabolic rates at low speeds (Table 2); however, with increasing U, metabolic rates increased at a faster rate in males compared to females, indicating that to move at faster speeds is more costly to males than to females. As has been found previously in exercising domestic chickens (Brackenbury and Elsayed, 1985), respiratory exchange ratios (RERs) were close to 1 across all treadmill speeds (B,: 1.09 [ ], B : 1.10 [ ], L,: 1.09 [ ] and L : 1.14 [ ], means and [ranges]). RER increased positively with U, which may suggest a greater anaerobic contribution to metabolism with increasing U. No signs of fatigue (trouble maintaining balance, head or wing droopiness) or post exercise oxygen deficit on the gas traces were found however, to suggest a large amount anaerobic respiration by the muscles. Statistical analyses on mass-specific _V O2 with speed produced the same statistical outcomes as mass-specific P met (Table 2). Walking cost of transport The total metabolic cost of transport (CoT tot, J kg 1 m 1 ) decreased curvilinearly with U in both varieties and sexes (Fig. 3E & F). The rate of decrease in CoT tot was similar between varieties; however, the intercepts were lower in the larger variety compared to the bantams by 1 Jkg 1 m 1 (Table 2). The incremental decrease in CoT tot with U was greater in females than in males (Table 2). The change in mass-specific net metabolic cost of transport (CoT net,jkg 1 m 1 )withu(fig. 4) was almost independent of speed (small positive increase) in males, but decreased curvilinearly in females (Table 2). Consequently, the minimum measured CoT net in females occurred at their maximum walking speed and was and 8.67 J kg 1 m 1 in B and L, respectively (Fig. 4A). These values are lower than predictions (B =17.09 and L =15.40 J kg 1 m 1 ) based on interspecific allometry [CoT min =17.80M 0.47 b (Rubenson et al., 2007)] of the minimum measured CoT net for walking gaits (Fig. 4A). The CoT net of the females was lower than the CoT min predicted by interspecific allometry across the majority of their speed range, excluding the two slowest speeds (0.28 and 0.42 m s 1 ) (Fig. 4A). The CoT net values of the males were scattered either side of the CoT min prediction, uncorrelated with U and not significantly different between varieties (Fig. 4B). Froude corrections The sex differences in CoT tot at a given U may exist because the locomotion of the sexes is not dynamically similar. When calculated using weight (N) instead of M b, the CoT tot reduces to a dimensionless parameter (Fish et al., 2000). The dynamic similarity hypothesis poses that geometrically similar animals moving with equal ratios of gravitational and inertial forces acting Biology Open 1308

60 RESEARCH ARTICLE Biology Open (2015) 4, doi: /bio Table 2. Summary of the statistical models testing for variety and sex differences in resting or walking metabolic rate parameters Parameter a Non-significant interaction terms b Final ANOVA/ANCOVA/GLM Coefficients c logrmr (W) logm b variety sex (F 1,20 =0.53, P=0.473), variety sex (F 1,21 =0.77, P=0.391) logm b variety (F 1,22 =1.50, P=0.233), logm b sex (F 1,23 =0.20, P =0.660) logm b (F 1,24 =5.49, *P=0.028), variety (F 1,24 =0.04, P=0.841), sex (F 1,24 =2.50, P=0.127), r 2 =0.55 RMR (W kg 1 ) Variety sex (F 1,24 =0.00, P=0.955) variety (F 1,25 =2.65, P=0.116) sex (F 1,25 =0.52, P=0.477) r 2 =0.02 log _ V O2 (ml kg 1 min 1 ) logu variety sex (F 1,127 =1.44, P=0.233), variety sex (F 1,128 =094, P=0.333), logu variety (F 1,129 =0.89, P=0.350) logp met (W) logp met (W kg 1 ) lognet-p met (W kg 1 ) logcot tot (J kg 1 m 1 ) logcot net (J kg 1 m 1 ) logcot tot (J kg 1 m 1 ) logu variety sex (F 1,125 =2.59, P=0.110), logu variety (F 1,126 =0.16, P=0.693), variety sex (F 1,127 =2.88, P=0.092) logu variety sex (F 1,126 =2.29, P=0.133), logu variety (F 1,127 =0.13, P=0.721), variety sex (F 1,128 =0.43, P=0.514) logu:variety:sex (F 1,124 =0.34, P=0.563), variety sex (F 1,125 =0.20, P=0.654), logu variety (F 1,126 =0.448, P=0.505) logu:variety:sex (F 1,26 =2.21, P=0.140), logu variety (F 1,127 =0.11, P=0.736), variety sex (F 1,128 =0.38, P=0.537) logu variety sex (F 1,124 =0.32, P=0.573, variety sex (F 1,125 =0.19, P=0.667), logu variety (F 1,126 =0.454, P=0.502) logfr variety sex (F 1,126 =2.12, P=0.148), logfr variety (F 1,127 =0.00, P=0.951), variety sex (F 1,128 =1.53, P=0.218) logu (F 1,130 =118.75, *P<0.001), variety (F 1,130 =2.53, P=0.114), sex (F 1,130 =10.61, *P=0.001), logu sex (F 1,130 =14.31, *P<0.001), r 2 =0.51 logu (F 1,128 =118.83, *P<0.001), variety (F 1,128 =0.53, P=0.470), sex (F 1,128 =3.71, P=0.056), M b (F 1,128 =51.07, *P<0.001), logu sex (F 1,128 =11.05, *P=0.001), r 2 =0.76 logu (F 1,129 =118.75, *P<0.001), variety (F 1,129 =3.79, P=0.054), sex (F 1,129 =8.19, *P=0.005), logu sex (F 1,129 =13.51, *P<0.001), r 2 =0.51 logu (F 1,127 =87.28, *P<0.001), variety (F 1,127 =0.10, P=0.749 sex (F 1,127 =2.73, P=0.101), logu sex (F 1,127 =6.94, *P=0.009), r 2 =0.44 logu (F 1,129 =328.20, *P<0.001), variety (F 1,129 =3.65, P=0.058), sex (F 1,129 =7.88, *P=0.006), logu sex (F 1,129 =13.52, *P<0.001), r 2 =0.77 logu (F 1,127 =6.63, *P=0.011), variety (F 1,127 =0.09, P=0.762, sex (F 1,127 =2.67, P=0.105), logu sex (F 1,127 =6.96, *P=0.009), r 2 =0.10 logfr (F 1,129 =298.84, *P<0.001), variety (F 1,129 =30.04, *P<0.001), sex (F 1,129 =4.04, *P=0.046), logfr sex (F 1,129 =13.68, *P<0.001), r 2 = B =9.20M b 0.60 B =8.05M b 0.60 L =9.04M b 0.60 L =7.91M b B =7.85 B =7.57 L =7.23 L =6.96 B =58.97U 0.51 B =45.24U 0.45 L =56.11U 0.51 L =43.05U 0.45 B =29.00U 0.48 B =19.92U 0.32 L =41.25U 0.60 L =21.73U 0.26 B =21.39U 0.52 B =16.85U 0.26 L =20.17U 0.52 L =15.89U 0.26 B =13.35M b 1.04 B =9.08M b 0.62 L =13.03M b 1.04 L =8.87M b 0.62 B =21.28M b 0.49 B =16.79M b 0.73 L =20.09M b 0.49 L =15.84M b 0.73 B =13.27M b 0.03 B =9.04M b 0.39 L =12.98M b 0.03 L =8.84M b 0.39 B =18.89M b 0.24 B =13.59M b 0.37 L =16.05M b 0.24 L =11.54M b 0.37 a Parameter symbols are: resting metabolic power whilst standing (RMR), oxygen consumption rate ( V _ O2 ), metabolic power (P met ), net metabolic power (Net-P met ), total cost of transport (CoT tot ) and net cost of transport (CoT net ). b Non-significant interaction terms are presented in the order that they were removed from the models. c The coefficients were taken from the outputs of the final models and were back transformed to provide the best fit lines in Fig. 3A-F. *Statistically significant results. on their body CoM (i.e. at equal Fr) will incur a similar CoT (Alexander and Jayes, 1983). CoT min decreased curvilinearly with Fr at a faster rate in female than in male leghorns (Fig. 5A-B). The maximum Fr recorded, at which the females were still walking and incurred their CoT min was greater than that for males. At the Fr equivalent to the U max of the males, the CoT was already lower in females than in males. Female leghorns, therefore, carry a unit of their M b over a unit of distance with greater economy of energy use than males. Grounded running in males During grounded running gaits in the males, mass-specific P met (W kg 1 ), was 5.75 W kg 1 greater in the L, compared to B across all U (Table 3; Fig. 3C-D). Calculating net mass-specific P met (W kg 1 ) increased this difference between varieties to 9.18 W kg 1 (Table 3). Since P met during quiet standing was the same between varieties, the reduction in grounded running P met in the standard breed relative to the B upon calculating net-p met may indicate change in the postural cost of locomotion during a grounded running gait. CoT tot during grounded running was 7.76 J kg 1 m 1 greater in L, than in B. Similarly, CoT net was 6.27 J kg 1 m 1 greater in the standard variety. Neither P met, net mass-specific P met,cot tot nor CoT net changed with U in either variety (Table 3). When compared to interspecific allometric predictions of running using CoT min, =12.91M b (Rubenson et al., 2007), the measured B value is similar (B measured, predicted: 9.63 and J kg 1 m 1 ), but the measured L, value is greater (large measured predicted: and J kg 1 m 1 ). Therefore, during a grounded running gait, L, have a poorer economy of energy use than do B. Biology Open 1309

61 RESEARCH ARTICLE Biology Open (2015) 4, doi: /bio A m s m s m s m s -1 B m s m s m s m s m s -1 Fig. 2. Metabolic power versus body mass during walking gait. Black and white symbols represent males and females, respectively, in bantam leghorns (A) and standard breed leghorns (B). The size of the diamond represents the magnitude of the speed. P met (W) P met (W) m s m s M b (kg) M b (kg) DISCUSSION The principal aim of this study was to determine the influence of sex on locomotor performance in standard breed (large) and bantam (small) leghorn chickens. Differences in the incremental increase in walking P met with U between the varieties were negated by masscorrection, but mass-correction did not remove the observed sex differences. In both varieties, P met increased more rapidly with walking U in males than in females, indicating that to walk at faster speed was more costly in males, relative to females. This is the first evidence of a greater CoT min in a male bird when compared to a female. Our study is also the first to compare the CoT of the sexes over a similar range of Froude numbers in a species of bird. After negating the effects of body size and speed, the sex differences in CoT tot were shared by the two varieties, despite them exhibiting dissimilar sexual dimorphism in limb posture. While L were 23% more upright than L, no sex difference in posture was present in the bantam variety. In both varieties, females were lighter than males and had a lower CoT min, which contrasts to the expected negative allometry of CoT min with increasing M b (see solid line in Fig. 4A,B) across species (Taylor et al., 1982; Rubenson et al., 2007). It is widely accepted, however, that there is no independent effect of M b on CoT min (Pontzer, 2007). Furthermore, a growing body of evidence supports the hypothesis that the interspecific increase in limb erectness with M b is linked to the allometry of CoT min (Mcmahon et al., 1987; Griffin et al., 2004; Pontzer, 2007; Reilly et al., 2007; Nudds et al., 2009; Rose et al., 2015). At the intraspecific level, however, limb posture is not expected to change with M b (Griffin et al., 2004; Day and Jayne, 2007; Rose et al., 2015). Another reason why the measured sex differences in CoT min were unexpected is that females leghorns have lower ratios of skeletal muscle mass: visceral and reproductive mass, relative to males (Mitchell et al., 1931). Since the muscle force required to support body weight is considered the principal contributor to the metabolic cost of terrestrial locomotion (Taylor et al., 1980), above other costs such as swinging the limb (Marsh et al., 2004), and maintaining posture (Weyand et al., 2009), the females might be expected to incur a greater metabolic cost of locomotion per unit M b. Adding loads to the backs of mammals to manipulate M b, for example, leads to an increase in net locomotor metabolic rate, greater in proportion than the proportional increase in mass (McGowan et al., 2006). In the few avian species examined to date, however, an extra gram of back load was carried at a cost equal to (Tickle et al., 2010), or less than (Marsh et al., 2006; McGowan et al., 2006; Tickle et al., 2013) carrying a gram of original M b. If the hens carry each gram of reproductive load at a cost less than carrying each gram of the Table 3. Results of the ANCOVAs that tested for differences between varieties in the relationships between metabolic rate parameters and speed during grounded running Parameter Non-significant interaction terms (removed from final statistical model) Final ANCOVA Coefficients P met (W kg 1 ) logu:variety (F 1,26 =1.69, P=0.205) logu (F 1,27 =4.03, P=0.055), variety (F 1,27 =7.32, *P=0.012), r 2 =0.61 Net P met (W kg 1 ) logu:variety (F 1,26 =2.71, P=0.111) logu (F 1,27 =0.21, P=0.646), variety (F 1,27 =9.95, *P=0.004), r 2 =0.49 CoT tot (J kg 1 m 1 ) logu:variety (F 1,26 =1.70, P=0.203) logu (F 1,27 =1.47, P=0.236), variety (F 1,27 =7.34, *P=0.012), r 2 =0.20 CoT net (J kg 1 m 1 ) logu:variety (F 1,26 =0.70, P=0.409) logu (F 1,27 =0.00, P=0.932), variety (F 1,27 =6.82, *P=0.015), r 2 =0.32 All dependent variables and covariates were log transformed which improved the AIC of each model. The adjusted r 2 values from the final models are reported. *Statistically significant results B =17.78U 0.62 L =23.53U 0.62 B =10.50U 0.28 L =19.68U 0.28 B =17.75U 0.38 L =25.51U 0.38 B =9.63U 0.05 L = Biology Open

62 RESEARCH ARTICLE Biology Open (2015) 4, doi: /bio A P met (W) P met (W kg -1 ) C bantam leghorns * * walking bantam leghorns tran grounded running * * B P met (W) P met (W kg -1 ) D standard breed leghorns walking standard breed leghorns walking tran tran grounded running grounded running Fig. 3. Metabolic parameters as a function of treadmill speed. Black and white symbols represent males and females, respectively. Circles and squares represent bantam and standard breed leghorns, respectively. (A-B) Metabolic power (P met,w). (C-D) Mass-specific metabolic power. (E-F) Total mass-specific cost of transport (CoT tot, Jkg 1 m 1 ). Dashed vertical lines represent the greatest walking speed and the lowest grounded running speed. L used walking mechanics across their full speed range. Asterisks indicate where B sample size was N=3. Equations for the lines of best fit (males=solid lines and females=dashed lines) are given in Table 2. Data is represented as mean (±s.e.m). 5 0 walking tran grounded running 5 0 E CoT tot (J kg -1 m -1 ) bantam leghorns * * F CoT tot (J kg -1 m -1 ) standard breed leghorns walking tran grounded running 10 0 walking grounded running tran U (m s -1 ) U (m s -1 ) remaining M b, this could lead to the observed lower than expected CoT after dividing by total M b. Similar, to a previous finding in laying hens (van Kampen, 1976b), the CoT min of the females in this study was lower than that predicted using interspecific allometry. We expect, however, that more than just the exceptional load carrying ability of some birds compared to mammals is responsible for the low female CoT relative to M b and relative to male CoT. Sexual dimorphism in physiological performance is often associated with sex-specific adaptations that have resulted from the differential selective pressures on the sexes given their different life histories (Rogowitz and Chappell, 2000; Shillington and Peterson, 2002; Husak and Fox, 2008; Lees et al., 2012). Female chickens invest metabolic energy in gravidity (van Kampen, 1976a; Gloutney et al., 1996). Selection may be expected to act on the female s ability to carry eggs with metabolic economy of force generation. The evolution of compensatory traits that alleviate the potential costs of exaggerated sexually selected morphologies is usually considered from the male perspective (Husak and Swallow, 2011). Gravid female lizards have been shown to experience this type of selection (Shine et al., 1998). It is, however, unknown if this occurs in birds. One potential compensatory mechanism in females could be muscular adaptations that promote economical force generation (e.g. shorter fascicle lengths, or an increase in the proportion of slow oxidative muscle fibres). Furthermore, females may employ different gait kinematics (e.g. increased time of footground contact), which allow the recruitment of slower muscle fibres (Kram and Taylor, 1990). Male chickens, by contrast, invest more energy in terrestrial locomotion than females through maintaining territory, inter-male aggressive behaviour and intersexual courtship activity. Although the influence of these behaviours on daily energy budget is not known, it is interesting to consider why selection has not reduced the metabolic requirements of locomotion in leghorn males, relative to the less active females, as was found in another galliform species (Lees et al., 2012). Perhaps a stronger selection pressure on fighting ability promotes muscle architecture for fast, powerful, and sustained combats that are costly to use at intermediate to high walking speeds. Faster contracting, relatively longer muscle fascicles, and muscles with a greater capacity for force generation might be expected to have elevated power demands. There is precedence for this type of adaptation in birds as sex differences in flight muscle specialization have previously been identified in species where the males partake in fast volant courtship displays and females use high powered locomotion to a lesser degree than the males (Schultz et al., 2001). As expected, males achieved greater U max than females in common with many vertebrate species (Bhambhani and Singh, 1985; Brackenbury and Elsayed, 1985; Shine and Shetty, 2001; Finkler et al., 2003; Lees et al., 2012). Of course, the size difference between the sexes could explain this finding. However, a greater U mas in males compared to females is also common to species lacking sexual size dimorphism, but where the males have higher activity levels than females during the mating season (Lees et al., 2012). The greater U max in males is likely supported by their specializations for inter-male combat, including relatively larger skeletal muscles, hearts and lungs compared to females (Mitchell et al., 1931). At the same time, a reduction in U max and sprint speed in vertebrate females is often associated with the encumbrance of pregnancy or gravidity (Olsson et al., 2000; Shine, 2003; Knight, 2011). One benefit of a lower U is that it allows a longer stance phase during which sufficient force can be generated to support Biology Open 1311

63 RESEARCH ARTICLE Biology Open (2015) 4, doi: /bio A m s m s -1 CoT net (J kg -1 m -1 ) m s m s m s m s m s B 0.28 m s m s m s m s m s m s M b (kg) M b (kg) Fig. 4. Net cost of transport versus body mass across the range of walking speeds and mean minimum measured costs of transport. Grey and white diamonds represent bantam and standard breed leghorns, respectively. The size of the diamond represents the magnitude of the speed. Solid curves are Rubenson et al s (2007) interspecific allometric relationship between walking CoT min and M b. Mean CoT min is represented by a dotted line for B, a dashed line for L and a dotted and dashed line for all males. (A) Female CoT net decreased as a function of speed and the majority of their values were below the predicted CoT min. (B) Male CoT net was independent of speed and both varieties shared the same mean CoT min closer to the prediction for the standard breed variety. body weight. We suspect that the ability to generate sufficient force may limit female U max, relative to the males, given their lower muscle mass: visceral/reproductive mass ratio. Females of the two varieties were reluctant or unable to transition to grounded running gait mechanics. It is possible that they avoided higher U and grounded running gaits in order to reduce peak forces on their bones and avoid fracture as their bones may be weakened by the provision of medullary calcium towards eggshell formation (Bloom et al., 1941; Whitehead, 2004). This may be particularly pertinent in white leghorns, which are prone to osteoporosis during eggshell construction (Dacke et al., 1993). CONCLUSIONS The sexes of both standard breed and bantam varieties of leghorn chicken differed in all measured aspects of terrestrial locomotion. Males attained greater U max compared to females and used a grounded running gait at faster speeds, while gravid bantam females were reluctant to and standard breed females did not. These findings are consistent with the general consensus that gravidity and lower ratios of skeletal muscle:visceral mass in females, constrain locomotion. Our findings are likely the result of a combination of sex-specific adaptations and associated constraints that have resulted from differential selection pressures on the sexes. A CoT tot (J kg -1 m -1 ) 50 B Froude number Froude number Fig. 5. Total cost of transport versus Froude number. Bantam leghorn data is shown as circles. (A) and standard breed leghorns data as squares (B). Black and white symbols represent males and females, respectively. Data is represented as mean (±s.e.m) Biology Open

64 RESEARCH ARTICLE Biology Open (2015) 4, doi: /bio Furthermore, we suggest that gravid females may possess adaptations for greater metabolic economy of locomotion (e.g. in muscle specialization/posture/kinematics). MATERIALS AND METHODS Animals We acquired sexually mature (>16 weeks<1 year old) standard breed (5 male, 1.92±0.13 kg; 7 female, 1.43±0.02 kg, mean±s.e.m.) and bantam (9 male, 1.39±0.03 kg; 7 female, 1.09±0.04 kg, mean±s.e.m.) Leghorn chickens from local suppliers between March and May (breeding season) and housed them in the University of Manchester s Animal Unit. Hens were egg laying and males exhibited secondary sexual morphological characteristics, crowing and aggressive behaviour. Sexes and varieties were housed separately with ad libitum access to food (Specialist Poultry Breeder, Small Holder Range, Norfolk, UK: oils and fats: 6%; protein: 18%; fibre: 4.5%; Ash: 12.0%; calcium 4%) and water. Light-dark cycles were fixed at 13:11 h and temperatures at C. The birds were trained daily for one week to exercise on a treadmill (Tunturi T60, Turku, Finland), within a Perspex respirometry chamber. None of the birds was fasted prior to respirometry measurements. The male birds in this study were previously used in (Rose et al., 2015). A UK Home Office Project License held by Dr Codd (40/3549) covered all experimental procedures, which were undertaken with the ethical approval of the University of Manchester Ethics Committee. Respirometry Rates of O 2 consumption ( _V O2, ml min 1 ) and CO 2 production ( _V CO2, ml min 1 ) were measured from resting (standing) and exercising birds using a flow-through respirometry system (all equipment Sable Systems International, Las Vegas, NV, USA). Different sized chambers were built for large ( cm) and bantam leghorns ( cm) and a Flowkit 500 pulled ambient air through them at flow rates of 150 and 250 litres min 1 respectively. The Flowkit directed a sub-sample (0.11 litres min 1 ) from the main flow through the gas analysis system. Water vapour pressure (WVP) was measured by an RH300 before H 2 Owas scrubbed from the sample, using calcium chloride (2 6 mm granular, Merck, Darmstadt, Germany) and passed on to a CA-10A CO 2 analyser for CO 2 measurements. Dry air was scrubbed of CO 2 with a column of soda lime (2 5 mm granular, Sigma Aldrich, Steinheim, Germany) before passed on to an Oxzilla II O 2 analyser for O 2 and barometric pressure (BP) measurements. A pump (SS-3) sampled ambient air through a second channel at 0.11 litres min 1 and the sample was scrubbed of H 2 O and CO 2 (as previously described) before being passed through the Oxzilla. The accuracy of the set up (±5% across all treadmill speeds) was validated using an 2 injection test (Fedak et al., 1981). Differential O 2 concentration (ΔO 2, ambient O 2 box O 2 concentrations) was used in all calculations. CO 2 traces were base-lined in the absence of a bird, which allowed the calculation of differential CO 2 (ΔCO 2 ). Primary flow rates (F) were converted to corrected flow rates (F c ) to account for the H 2 O removed from the samples using Eqn 8.6 from Lighton (2008): FðBP WVPÞ F C ¼ ; ð1þ BP where WVP is water vapour pressure. _V O2 and _V CO2 were calculated using Eqns 10.1 and 10.8 from Lighton (2008), respectively: _V O2 ¼ F CðDO 2 Þ ð2þ ð1 0:2095Þ _V CO2 ¼ ðf CðDCO 2 ÞÞ ð0:0004ð _V O2 ÞÞ : ð3þ ð1 0:0004Þ RERs ( _V CO2 : _V O2 ) and their thermal equivalents (taken from Table 12.1 of Brody, 1945) were used to convert _V O2 into P met (W). To account for potential sex differences in body maintenance and postural metabolic requirements, net-p met (locomotor P met resting P met during quiet standing) was calculated using values taken from the same trial for each individual bird. Trials Experimental temperatures ranged from C (19.8±1.5 C, mean± s.e.m.). In a single trial, birds were exercised at a maximum of three randomly selected speeds and were given resting intervals of at least 5 min between each period of exercise to recover. The birds were walked at a minimum speed of 0.28 m s 1 and at increments of 0.14 m s 1 up to the maximum that they could sustain for steady _V O2 readings (>3 min). The final 1 min of the plateau was used for data analysis. All resting metabolic rates were taken from the final rest period of a trial and birds were given a day of rest between trials. Determining gait The gait mechanics of each bird was determined from video recordings (100 frames s 1 ; HDR-XR520VE, SONY, Japan) taken perpendicular to the direction of travel of the birds (from the left) in all trials. Using Tracker software v2.51 (Open Source Physics) a marked site over the left hip (the CoM) was tracked (min 3 strides) in every film frame to determine the mechanical energy fluctuations using temporal and spatial data. A calibration stick was positioned along the line of travel of a bird passing through digit 3 to avoid any error in measured dimensions that might have arisen due to a bird s displacement from it. The phasing of the CoM fluctuations in horizontal kinetic energy (E kh ) with the sum of its vertical kinetic and gravitational potential energies (E kv +E p ) was used to determine gait. An out of phase relationship is characteristic of walking gaits and an in-phase relationship of running gaits. Statistical analyses Statistical analyses were performed using the car package version (Fox and Weisberg, 2011) R GUI 1.42 Leopard build 64-bit (R Development Core Team, 2011). Morphological measurements were tested for the main effects of sex and variety as well as potential interaction effects using two-way ANOVAS. Resting P met and RERs were investigated for sex and variety differences using ANCOVA. M b was included in the models as a covariate to compensate for the effects of M b and variety and sex were included as fixed factors. The relationships between exercising metabolic rates and U were investigated for differences (in slopes and intercepts) between varieties and sexes (both factors) using linear models. Speed was included as the main covariate in each model. For non-mass-specific metabolic parameters, M b was included in the models as an additional covariate. For mass-specific metabolic rates, M b was not included in the models. All potential interaction terms were considered in the primary models before a step-wise backward deletion of non-significant interaction terms was conducted. For all parameters, the quality of our linear models according to the Akaike s information criterion was improved by log transforming the data. Shapiro Wilk tests were performed on the standardised residuals generated by each statistical model to ensure that the data conformed to a normal distribution. In the case of the U max comparison between groups, the residuals did not conform to a normal distribution even after transformation, so a Kruskal Wallis test with a Dunn post-hoc test was used. The adjusted r 2 values of the models are reported and unless otherwise stated the means are reported as ±s.e.m. The influence of speed on metabolic rate is gait dependent in some avian species (Rubenson et al., 2004, 2007; Nudds et al., 2011). Statistical analyses were, therefore, conducted on metabolic data from walking and grounded running gaits, separately. Sex comparisons were conducted for walking gaits only, since very little grounded running data were collected from the females. Competing interests The authors declare no competing or financial interests. Author contributions The study was conceived by J.R.C., R.L.N. and P.J.B. and all authors designed the study. K.A.R. collected, analysed and interpreted the data with assistance from R.L.N. and J.R.C. All authors contributed to preparation of the manuscript, approved and read the final submission. Biology Open 1313

65 RESEARCH ARTICLE Biology Open (2015) 4, doi: /bio Funding Funding was provided by the Biotechnology and Biological Sciences Research Council (BBSRC) [G01138/1 and /1 to J.R.C]. K.A.R was supported by a Natural Environment Research Council Doctoral Training Account (NERC DTA) stipend and Collaborative Awards in Science and Engineering (CASE) partnership with the Manchester Museum. References Alexander, R. M. and Jayes, A. S. (1983). A dynamic similarity hypothesis for the gaits of quadrupedal mammals. J. Zool. 201, Bamford, O. S. and Maloiy, G. M. O. (1980). Energy metabolism and heart rate during treadmill exercise in the Marabou stork. J. Appl. Physiol. 49, Baumel, J. J. (1953). Individual variation in the white-necked raven. Condor 55, Bhambhani, Y. and Singh, M. (1985). Metabolic and cinematographic analysis of walking and running in men and women. Med. Sci. Sport Exer. 17, Bloom, W., Bloom, M. A. and McLean, F. C. (1941). Calcification and ossification. Medullary bone changes in the reproductive cycle of female pigeons. Anat. Rec. 81, Brackenbury, J. H. and Elsayed, M. S. (1985). Comparison of running energetics in male and female domestic fowl. J. Exp. Biol. 117, Brody, S. (1945). Bioenergetics and Growth, with Special Reference to the Efficiency Complex in Domestic Animals. New York: Reinhold. Bruinzeel, L. W., Piersma, T. and Kersten, M. (1999). Low costs of terrestrial locomotion in waders. Ardea 87, Chappell, M. A., Zuk, M., Kwan, T. H. and Johnsen, T. S. (1995). Energy cost of an avian vocal display: crowing in Red Junglefowl. Anim. Behav. 49, Chappell, M. A., Zuk, M. and Johnsen, T. S. (1996). Repeatability of aerobic performance in Red Junglefowl: effects of ontogeny and nematode infection. Funct. Ecol. 10, Chappell, M. A., Zuk, M., Johnsen, T. S. and Kwan, T. H. (1997). Mate choice and aerobic capacity in red junglefowl. Behaviour 134, Chappell, M. A., Savard, J.-F., Siani, J., Coleman, S. W., Keagy, J. and Borgia, G. (2011). Aerobic capacity in wild satin bowerbirds: repeatability and effects of age, sex and condition. J. Exp. Biol. 214, Dacke, C. G., Arkle, S., Cook, D. J., Wormstone, I. M., Jones, S., Zaidi, M. and Bascal, Z. A. (1993). Medullary bone and avian calcium regulation. J. Exp. Biol. 184, Day, L. M. and Jayne, B. C. (2007). Interspecific scaling of the morphology and posture of the limbs during the locomotion of cats (Felidae). J. Exp. Biol. 210, Dunn, P. O., Whittingham, L. A. and Pitcher, T. E. (2001). Mating systems, sperm competition, and the evolution of sexual dimorphism in birds. Evolution 55, Ellerby, D. J. and Marsh, R. L. (2006). The energetic costs of trunk and distal-limb loading during walking and running in guinea fowl Numida meleagris: II. Muscle energy use as indicated by blood flow. J. Exp. Biol. 209, Ellerby, D. J., Cleary, M., Marsh, R. L. and Buchanan, C. I. (2003). Measurement of maximum oxygen consumption in Guinea fowl Numida meleagris indicates that birds and mammals display a similar diversity of aerobic scopes during running. Physiol. Biochem. Zool. 76, Fedak, M. A., Pinshow, B. and Schmidtn, K. (1974). Energy cost of bipedal running. Am. J. Physiol. 227, Fedak, M. A., Rome, L. and Seeherman, H. J. (1981). One-step N2-dilution technique for calibrating open-circuit VO2 measuring systems. J. Appl. Physiol. 51, Finkler, M. S., Sugalski, M. T. and Claussen, D. L. (2003). Sex-related differences in metabolic rate and locomotor performance in breeding spotted salamanders (Ambystoma maculatum). Copeia 2003, Fish, F. E., Frappell, P. B., Baudinette, R. V. and MacFarlane, P. M. (2000). Energetics of terrestrial locomotion of the platypus: metabolic inefficiencies due to aquatic adaptation. Am. Zool. 40, Fox, J. and Weisberg, S. (2011). An {R} Companion to Applied Regression, 2nd edn. Thousand Oaks, CA: Sage. Gloutney, M. L., West, N. and Clark, R. G. (1996). Metabolic costs of incubation and clutch size determination in the red junglefowl, Gallus gallus spadiceus. Comp. Biochem. Physiol. A Physiol. 114, Griffin, T. M., Kram, R., Wickler, S. J. and Hoyt, D. F. (2004). Biomechanical and energetic determinants of the walk-trot transition in horses. J. Exp. Biol. 207, Guhl, A. M., Collias, N. E. and Allee, W. C. (1945). Mating behavior and the social hierarchy in small flocks of white leghorns. Physiol. Zool. 18, Hakkarainen, H., Huhta, E., Lahti, K., Lundvall, P., Mappes, T., Tolonen, P. and Wiehn, J. (1996). A test of male mating and hunting success in the kestrel: the advantages of smallness? Behav. Ecol. Sociobiol. 39, Hammond, K. A., Chappell, M. A., Cardullo, R. A., Lin, R. S. and Johnsen, T. S. (2000). The mechanistic basis of aerobic performance variation in red junglefowl. J. Exp. Biol. 203, Hoglund, J. (1989). Size and plumage dimorphism in Lek-breeding birds: a comparative analysis. Am. Nat. 134, Horn, A. G., Leonard, M. L. and Weary, D. M. (1995). Oxygen consumption during crowing by roosters: talk is cheap. Anim. Behav. 50, Husak, J. F. and Fox, S. F. (2008). Sexual selection on locomotor performance. Evol. Ecol. Res. 10, Husak, J. F. and Swallow, J. G. (2011). Compensatory traits and the evolution of male ornaments. Behaviour 148, Knight, K. (2011). Pregnancy is a drag for bottlenose dolphins. J. Exp. Biol. 214, I. Kram, R. and Taylor, C. R. (1990). Energetics of running: a new perspective. Nature 346, Lees, J. J., Nudds, R. L., Folkow, L. P., Stokkan, K.-A. Codd, J. R. (2012). Understanding sex differences in the cost of terrestrial locomotion. Proc. R. Soc. B Biol. Sci. 279, Lighton, J. R. B. (2008). Measuring Metabolic Rates: A Manual for Scientists. Oxford; New York: Oxford University Press. Marsh, R. L., Ellerby, D. J., Carr, J. A., Henry, H. T. and Buchanan, C. I. (2004). Partitioning the energetics of walking and running: swinging the limbs is expensive. Science 303, Marsh, R. L., Ellerby, D. J., Henry, H. T. and Rubenson, J. (2006). The energetic costs of trunk and distal-limb loading during walking and running in guinea fowl Numida meleagris: I. Organismal metabolism and biomechanics. J. Exp. Biol. 209, McGowan, C. P., Duarte, H. A., Main, J. B. and Biewener, A. A. (2006). Effects of load carrying on metabolic cost and hindlimb muscle dynamics in guinea fowl (Numida meleagris). J. Appl. Physiol. 101, Mcmahon, T. A., Valiant, G. and Frederick, E. C. (1987). Groucho running. J. Appl. Physiol. 62, Mitchell, H. H., Card, L. E. and Hamilton, T. S. (1931). A technical study of the growth of White Leghorn chickens. Ill. Agric. Exp. Stn. Bull. 367, Nudds, R. L., Codd, J. R. and Sellers, W. I. (2009). Evidence for a mass dependent step-change in the scaling of efficiency in terrestrial Locomotion. PLoS ONE 4, e6927. Nudds, R. L., Gardiner, J. D., Tickle, P. G. and Codd, J. R. (2010). Energetics and kinematics of walking in the barnacle goose (Branta leucopsis). Comp. Biochem. Physiol. A Mol. Integr. Physiol. 156, Nudds, R. L., Folkow, L. P., Lees, J. J., Tickle, P. G., Stokkan, K.-A. and Codd, J. R. (2011). Evidence for energy savings from aerial running in the Svalbard rock ptarmigan (Lagopus muta hyperborea). Proc. R. Soc. B Biol. Sci. 278, Olsson, M., Shine, R. and Bak-Olsson, E. (2000). Locomotor impairment of gravid lizards: is the burden physical or physiological? J. Evol. Biol. 13, Pande, S. and Dahanukar, N. (2012). Reversed sexual dimorphism and differential prey delivery in barn owls (Tyto Alba). J. Raptor Res. 46, Paxton, H., Anthony, N. B., Corr, S. A. and Hutchinson, J. R. (2010). The effects of selective breeding on the architectural properties of the pelvic limb in broiler chickens: a comparative study across modern and ancestral populations. J. Anat. 217, Pinshow, B., Fedak, M. A. and Schmidt-Nielsen, K. (1977). Terrestrial locomotion in penguins: it costs more to waddle. Science 195, Pontzer, H. (2007). Effective limb length and the scaling of locomotor cost in terrestrial animals. J. Exp. Biol. 210, Portugal, S. J. and Guillemette, M. (2011). The use of body mass loss to estimate metabolic rate in birds. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 158, R Development Core Team (2011). R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing. Reilly, S. M., McElroy, E. J. and Biknevicius, A. R. (2007). Posture, gait and the ecological relevance of locomotor costs and energy-saving mechanisms in tetrapods. Zoology 110, Remes, V. and Szekely, T. (2010). Domestic chickens defy Rensch s rule: sexual size dimorphism in chicken breeds. J. Evol. Biol. 23, Reynolds, R. T. (1972). Sexual dimorphism in accipiter hawks: new hypothesis. Condor 74, Roberts, T. J., Kram, R., Weyand, P. G. and Taylor, C. R. (1998). Energetics of bipedal running. I. Metabolic cost of generating force. J. Exp. Biol. 201, Rogowitz, G. L. and Chappell, M. A. (2000). Energy metabolism of eucalyptusboring beetles at rest and during locomotion: gender makes a difference. J. Exp. Biol. 203, Rose, K. A., Tickle, P. G., Lees, J. J., Stokkan, K.-A. and Codd, J. R. (2014). Neither season nor sex affects the cost of terrestrial locomotion in a circumpolar diving duck: the common eider (Somateria mollissima). Polar Biol. 37, Rose, K. A., Nudds, R. L. and Codd, J. R. (2015). Intraspecific scaling of the minimum metabolic cost of transport in leghorn chickens (Gallus gallus domesticus): links with limb kinematics, morphometrics and posture. J. Exp. Biol. 218, Rubenson, J., Heliams, D. B., Lloyd, D. G. and Fournier, P. A. (2004). Gait selection in the ostrich: mechanical and metabolic characteristics of walking and running with and without an aerial phase. Proc. R. Soc. B Biol. Sci. 271, Biology Open 1314

66 RESEARCH ARTICLE Biology Open (2015) 4, doi: /bio Rubenson, J., Heliams, D. B., Maloney, S. K., Withers, P. C., Lloyd, D. G. and Fournier, P. A. (2007). Reappraisal of the comparative cost of human locomotion using gait-specific allometric analyses. J. Exp. Biol. 210, Schultz, J. D., Hertel, F., Bauch, M. and Schlinger, B. A. (2001). Adaptations for rapid and forceful contraction in wing muscles of the male golden-collared manakin: sex and species comparisons. J. Comp. Physiol. A Sens. Neural Behav. Physiol. 187, Schutz, K. E., Forkman, B. and Jensen, P. (2001). Domestication effects on foraging strategy, social behaviour and different fear responses: a comparison between the red junglefowl (Gallus gallus) and a modern layer strain. Appl. Anim. Behav. Sci. 74, Shillington, C. and Peterson, C. C. (2002). Energy metabolism of male and female tarantulas (Aphonopelma anax) during locomotion. J. Exp. Biol. 205, Shine, R. (2003). Effects of pregnancy on locomotor performance: an experimental study on lizards. Oecologia 136, Shine, R. and Shetty, S. (2001). Moving in two worlds: aquatic and terrestrial locomotion in sea snakes (Laticauda colubrina, Laticaudidae). J. Evol. Biol. 14, Shine, R., Keogh, S., Doughty, P. and Giragossyan, H. (1998). Costs of reproduction and the evolution of sexual dimorphism in a flying lizard Draco melanopogon (Agamidae). J. Zool. 246, Steen, J. B. and Unander, S. (1985). Breeding biology of the Svalbard Rock Ptarmigan Lagopus mutus hyperboreus. Ornis Scand. 16, Taylor, C. R., Dmiel, R., Fedak, M. and Schmidtn, K. (1971). Energetic cost of running and heat balance in a large bird, the rhea. Am. J. Physiol. 221, Taylor, C. R., Heglund, N. C., Mcmahon, T. A. and Looney, T. R. (1980). Energetic cost of generating muscular force during running - a comparison of large and small animals. J. Exp. Biol. 86, Taylor, C. R., Heglund, N. C. and Maloiy, G. M. O. (1982). Energetics and mechanics of terrestrial Locomotion.1. Metabolic energy-consumption as a function of speed and body size in birds and mammals. J. Exp. Biol. 97, Tickle, P. G., Richardson, M. F. and Codd, J. R. (2010). Load carrying during locomotion in the barnacle goose (Branta leucopsis): the effect of load placement and size. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 156, Tickle, P. G., Lean, S. C., Rose, K. A. R., Wadugodapitiya, A. P. and Codd, J. R. (2013). The influence of load carrying on the energetics and kinematics of terrestrial locomotion in a diving bird. Biol. Open 2, Unander, S. and Steen, J. B. (1985). Behaviour and social structure in Svalbard rock ptarmigan Lagopus mutus hyperboreus. Ornis Scand. 16, van Kampen, M. (1976a). Activity and energy expenditure in laying hens. 1. The energy cost of nesting activity and oviposition. J. Agric. Sci. 86, van Kampen, M. (1976b). Activity and energy-expenditure in laying hens. 2. The energy cost of exercise. J. Agric. Sci. 87, Watson, R. R., Rubenson, J., Coder, L., Hoyt, D. F., Propert, M. W. G. and Marsh, R. L. (2011). Gait-specific energetics contributes to economical walking and running in emus and ostriches. Proc. R. Soc. B Biol. Sci. 278, Weyand, P. G., Smith, B. R. and Sandell, R. F. (2009). Assessing the metabolic cost of walking: the influence of baseline subtractions. Conf. Proc. IEEE Eng. Med. Biol. Soc. 2009, White, C. R., Martin, G. R. and Butler, P. J. (2008). Pedestrian locomotion energetics and gait characteristics of a diving bird, the great cormorant, Phalacrocorax carbo. J. Comp. Physiol. B 178, Whitehead, C. C. (2004). Overview of bone biology in the egg-laying hen. Poult. Sci. 83, Wilson, D. R., Bayly, K. L., Nelson, X. J., Gillings, M. and Evans, C. S. (2008). Alarm calling best predicts mating and reproductive success in ornamented male fowl, Gallus gallus. Anim. Behav. 76, Biology Open 1315

67 4. Differential sex-specific gait dynamics in leghorn chickens (Gallus gallus domesticus) selectively bred for different body size This chapter is a draft of an article submitted to Journal of Experimental Biology Rose, K.A., Codd, J. R., and Nudds, R. L. (in review). Differential sex-specific gait dynamics in leghorn chickens (Gallus gallus domesticus) selectively bred for different body size 50

68 Differential sex-specific gait dynamics in leghorn chickens (Gallus gallus domesticus) selectively bred for different body size Kayleigh A. Rose, Jonathan R. Codd and Robert L. Nudds* Faculty of Life Sciences, University of Manchester, Manchester, M13 9PT, UK * Address for reprints and other correspondence: Dr. Robert Nudds Faculty of Life Sciences University of Manchester Manchester M13 9PT, UK robert.nudds@manchester.ac.uk Tel: +44(0)

69 ABSTRACT The differing limb dynamics and postures of small and large terrestrial animals may be mechanisms for minimising metabolic costs under scale-dependent muscle force, work and power demands; however, empirical evidence for this is lacking. Leghorn chickens (Gallus gallus domesticus) are highly dimorphic: males are the larger sex, with greater pelvic limb muscle volume, and females are permanently gravid. Furthermore, leghorns are selected for standard (large) and bantam (small) varieties and the former are sexually dimorphic in posture, with females having a more upright limb. Here, high-speed videography and morphological measurements were used to examine the walking dynamics of leghorn chickens of the two varieties and sexes. Hind limb skeletal elements were geometrically similar among the bird groups. The groups, however, did not move with dynamic similarity. In agreement with the interspecific scaling of relative duty factor (DF, proportion of a stride with ground contact) with body mass, bantams walked with greater DF than standards and females with greater DF than males. Greater DF in females than in males was achieved via variety-specific kinematic mechanisms, associated with the presence/absence of postural dimorphism. Females may require greater DF in order to reduce peak muscle forces and minimize power demands associated with lower muscle to reproductive tissue volume ratios and smaller body size. Furthermore, a more upright posture observed in the standard, but not bantam, females, may relate to minimizing the work demands of being larger and having proportionally larger reproductive volume. Lower DF in males relative to females may also be a work-minimizing strategy and/or due to greater limb inertia (due to greater pelvic limb muscle mass) prolonging the swing phase. KEY WORDS: Froude number; locomotion; posture; sexual dimorphism; walking 2

70 INTRODUCTION The walking kinematics of different sized animals can be compared equitably over speeds at which the ratios of inertial (centripetal) and gravitational forces acting upon the body centre of mass (CoM) are equal, using the Froude number: Fr = U 2 /gh hip (1), where U (m s -1 ) is horizontal speed, h hip (m) is hip height and g is gravitational acceleration (9.81 m s -2 ) (Alexander and Jayes 1983; Alexander 1976). Dynamic similarity of motion between different sized animals requires geometric similarity in body plan and equal values of dimensionless kinematic parameters (scaled appropriately to negate the effects of size) for a given value of Fr (Alexander and Jayes 1983; Alexander 1976; Hof 1996). The dynamic similarity hypothesis of Alexander and Jayes (1983) postulated that different quadrupedal mammals would locomote with dynamic similarity at equal Fr. Within non-cursorial (<1 kg) and cursorial (>10 kg) mammalian groups (Jenkins 1971) the hypothesis was supported; however, observed kinematic differences between the two groups were not accounted for (Alexander and Jayes 1983). Furthermore, between avian species of small and large body size, there is considerable deviation from dynamic similarity of locomotion (Gatesy and Biewener 1991; Abourachid 2001; Abourachid and Renous 2000). A general pattern, however, exists across these vertebrates, whereby smaller species move with greater relative duty factors (DF, proportion of a stride with ground contact) and relative stride lengths. These deviations from dynamic similarity of locomotion have been attributed to differences in limb posture and the relative lengths of the limb segments (Gatesy and Biewener 1991; Alexander and Jayes 1983; Abourachid 2001; Abourachid and Renous 2000). Crouched and upright limb postures are generally adopted by small and large vertebrate species, respectively, which are clear departures from geometric similarity in body form (Biewener 1989; Gatesy and Biewener 1991). An erect limb aligns body weight with the limb bones long axes reducing mechanical loading on the muscles associated with turning moments about the 3

71 joints (Biewener 1989). Furthermore, an erect limb requires shorter stance (pushoff) periods and reduces costly fore-aft speed fluctuations, further reducing muscle work requirements (Usherwood 2013). A more upright limb therefore suits the work (J kg -1 ) demands of being large given that body weight increases at a faster rate with size than the strength of the biological materials, which must support it (Biewener 1989). Although the same benefits of an upright limb would apply to smaller species, theoretically, their muscle power (J s -1 kg -1 ) requirements may be disproportionately high (Usherwood 2013). Therefore, a more crouched limb, requiring a longer push-off period, may act to minimise power requirements in smaller animals (Usherwood 2013). It is therefore suggested that animals of differing size optimise active muscle volume under differing muscle work and power demands. Indeed, in humans (Homo sapiens) toddlers are found to deviate from the work-minimising gaits of adults, via longer normalised stance periods for a give Fr (Hubel and Usherwood 2015). Morphological, physiological and behavioural adaptations often differ between the sexes (Lourdais et al. 2006; Salazar and Stoddard 2008; Scantlebury et al. 2006; Shillington and Peterson 2002; Owens and Hartley 1998) due to sexual selection and sex-specific reproductive roles (Shine 1989; Slatkin 1984). Sex-specific compensatory adaptations, which alleviate the costs of sexually selected traits, may also drive further divergence between the sexes (Shine et al. 1998; Husak and Swallow 2011; Carlson et al. 2014). In many vertebrate species, sex differences in body size (Lislevand et al. 2009; Remes and Szekely 2010) and the relative proportion of total body mass (M b ) allocated to different somatic and reproductive components (usually biased towards males and females, respectively) are common (Hammond et al. 2000; Lourdais et al. 2006; Shine et al. 1998). Furthermore, female reproductive specialisation may even require specific skeletal proportions (e.g. a wider pelvis (Baumel 1953; Smith et al. 2002; Cho et al. 2004)), or posture, during pregnancy (Franklin and Conner-Kerr 1998) or gravidity (Rose et al. 2015a). The demands of muscle force, work and power may, therefore, differ between the sexes. Any influence of sexual dimorphism upon locomotor performance (e.g. in the dynamics of movement or metabolic cost) is likely to have important consequences for evolutionary fitness (Tolkamp 4

72 et al. 2002). Most studies on vertebrate kinematics, however, have been conducted using either individuals of only one sex (Reilly 2000); without comparing sexes (Rubenson et al. 2004; Watson et al. 2011), or using individuals whose sexes were not reported (Griffin et al. 2004; Nudds et al. 2010; Abourachid 2000, 2001; Abourachid and Renous 2000; Gatesy and Biewener 1991). Previous studies have identified sex differences in walking kinematics in humans (Bhambhani and Singh 1985) and two species of bird (Rose et al. 2014; Lees et al. 2012), but whether size variations alone, or both size and additional unidentified sexual dimorphisms were behind the differences in kinematics was not determined. The leghorn chicken (Gallus gallus domesticus) is highly dimorphic, with males having larger body size and limb skeletal muscle volume than females (Mitchell et al. 1931). Female leghorns have greater visceral organ mass than males and remain permanently gravid (Mitchell et al. 1931). Furthermore, leghorns are artificially selected for standard (large) and bantam (small) varieties, and only the standards are dimorphic in limb posture, with females possessing a more upright limb during a walking gait (Rose et al. 2015a). Males of the two varieties are geometrically similar in their axial and appendicular skeletons, but the bantam males adopt a more upright posture than the standards during walking gait (Rose et al. 2015b). Leghorns are, therefore, ideal animals with which to investigate the effects of posture and differing proportions of locomotor muscle and visceral/reproductive tissue mass on the dynamics of terrestrial locomotion. Here, high-speed videography and morphological measurements were used to test the hypothesis that standard and bantam varieties of leghorn would show clear departures from dynamic similarity of motion associated with differences in posture and hind limb skeletal muscle and reproductive tissue proportions. 5

73 MATERIALS AND METHODS Animals and experimental protocol Male and female bantam brown leghorns (B and B ) and standard breed large white leghorns (L and L ) were obtained from local suppliers and housed in the University of Manchester s Animal Unit. All leghorns (> 16 weeks < 1 year) had reached sexual maturity and females were gravid. Sexes and varieties were housed separately with ad libitum access to food, water and nesting space. Birds were trained daily for a week to locomote for ~5 min within a Perspex chamber mounted upon a Tunturi T60 (Turku, Finland) treadmill. The kinematics of twenty-four of the twenty-six leghorns used for the simultaneously collected metabolic measurements described in (Rose et al. 2015a), are presented here (B : N = 9; 1.39 ± 0.03 kg; B : N = 5; 1.04 ± 0.03 kg, L : N = 5; 1.92 ± 0.13 kg, L : N = 5; 1.43 ± 0.06 kg, mean ± s.e.m). Each leghorn was exercised at a minimum speed of 0.28 m s -1 and at increasing increments of 0.14 (in a randomised order), up to the maximum they could sustain without showing signs of fatigue. The birds were rested between speed trials. All experiments were approved by the University of Manchester s ethics committee, carried out in accordance with the Animals (Scientific procedures) Act (1986) and performed under a UK Home Office Project Licence held by Dr Codd (40/3549). Kinematics and dynamics Since the concept of dynamic similarity based upon Fr is founded on invertedpendulum mechanics (Alexander and Jayes 1983), only kinematics data from walking gaits are included. Birds were filmed from a lateral view (HDR- XR520VE, Sony, Japan, 100 frames s -1 ). The left foot of each bird was tracked across ~10 continuous strides (constant speed and position) to obtain the times of toe-on and toe off, which were used to calculate duty factor DF, stride frequency (f stride, Hz), stride length (l stride = U/f stride, m), limb swing duration (t swing, s) and stance duration (t stance, s). U was normalised for size differences using Fr. Kinematic parameters were normalised based on the Hof (Hof 1996) record of non-dimensional forms of mechanical quantities as: relative stride length ( ˆL = 6

74 l stride /h hip ); relative stride frequency ( ˆF = f stride g hhip ); relative swing duration ( ˆT swing = t swing hhip g ); and relative stance duration ( ˆTstance = t stance hhip g ). Morphological measurements h hip were taken from Rose et al (2015a). Digital vernier callipers (± 0.01 mm) were used to measure hind limb long bone (femur, tibiotarsus and tarsometatarsus) lengths (l fem, l tib, l tars ) and widths (w fem, w tib, w tars ). The width of the pelvis (w pelv = the distance between the left and right acetabula) was also measured. To ensure the h hip, back height (h back ) and Σl segs were comparable, h back during mid stance at slow speeds 0.28 m s -1 was measured from the same videos that were used to measure h hip in Rose et al (2015a). Reproductive masses (developing eggs, ovaries and oviduct) were dissected from 5 standard females and 5 bantam females and weighed upon electronic scales (± 0.01 g). Statistical analyses All statistical analyses were conducted on R (v GUY 1.2 Snow Leopard build 558) (Team 2011). The Car package (Fox and Weisberg 2011) was used for all analyses of variance (ANOVA) in which variety and sex were included as fixed factors. Data were log transformed if it improved the quality of the models according to the Akaike s information criterion and in order to linearize curvilinear-trends for analyses. Shapiro-Wilk tests were performed on the standardised residuals generated by all statistical models to ensure the data conformed to a normal distribution. Where morphological data (Table 1) did not conform to a normal distribution even after transformation, a Kruskal Wallis test was conducted to compare the means of the four groups: B, B, L, and L. Dunn post-hoc tests were used to indicate which groups differed. The relationships between absolute and normalised kinematics variables with U and Fr were compared between the bird groups using linear models. U and Fr were included in the models as covariates and all potential interaction terms were considered before a stepwise backwards deletion of non-significant interaction terms was conducted to simplify the models. Outputs from the final models are 7

75 reported. Best-fit lines were obtained from the coefficients tables of the final statistical models and were back transformed where data had been logged. RESULTS Morphological indices Each limb segment was a similar proportion of Σl segs in all of the leghorn groups excluding B, which had 1% more l tars, and concomitantly 1% less l fem, resulting in a small, but, nonetheless statistically significant difference (Table 1). The width of each limb segment was a similar proportion of its respective segment length in all groups (Table 1). w pelv, relative to Σl segs, was not significantly different between the sexes, but was ~1 % greater in the bantams, compared to the standards (Table 1). In the bantam variety, h hip : Σl segs did not differ between the sexes. Contrastingly, in the standard breed, the posture index of L indicated that their limb bones were completely aligned through their long axes (completely erect), and was ~27 % greater in L than in L (Table 1). The finding of a more erect posture in L when compared to the other three groups was further supported by indices incorporating h back during mid stance: h hip :h back did not differ between the bird groups, and Σl segs :h back was lower in L than in the other three leghorn groups (Table 1). Reproductive tissue mass was lower (F 1,8 = 33.44, P<0.001, R 2 = 0.78) in B (84.26 ± 7.53 g) than in L ( ± g) and similarly comprised a lower percentage of M b (F 1,8 = 8.74, P=0.018, R 2 = 0.46) in the B (8.40 ± 0.08 %) than in the L (1.49 ± 0.56 %). 8

76 Table 1. Mean (± s.e.m) morphometric indices and results of the statistical tests conducted to investigate whether the indices differed between varieties and sexes. The adjusted R 2 values of the final models are reported. Index B B L L Statistical results l fem :Σl segs Kruskal Wallis: X 2 = 19.6, df=3, P=<0.001 Dunn test: B v B : Z=3.71, P<0.001, B v L : Z=-0.61, P=0.272, B v L : Z=0.00, P=0.500, B v L : Z=-3.89, P<0.001, B v L : Z=-3.35, P<0.001, L v L : Z=0.55, P=0.292 l tib :Σl segs Kruskal Wallis: X 2 = 0.13, df=3, P=0.988 l tars :Σl segs Kruskal Wallis: X 2 = 12.47, P=0.006 Dunn test B v B : Z=-2.48, P=0.007, B v L : Z=0.90, P=0.18 B v L : Z=0.90, P=0.18, B v L : Z=3.05, P=0.001, B v L : Z=3.05, P=0.001, L v L : Z=0.00, P=0.500 w fem :l fem Kruskal Wallis: X 2 = 0.46, df=3, P=0.929 w tib :l tib Kruskal Wallis: X 2 = 7.20, df=3, P=0.066 w tars :l tars Kruskal Wallis: X 2 = 0.36, df=3, P=0.948 h hip :Σl segs 0.79 ± ± ± ± 0.03 a variety: F 1,19 =14.42, P=0.001, sex: F 1,19 =19.44, P<0.001, variety x sex: F 1,19 =42.30, P<0.001, R 2 =0.75 h hip :h back 0.78 ± ± ± ± 0.02 variety: F 1,20 =0.45, P=0.512, sex: F 1,20 =0.00, P=0.948, R 2 =0.00 Σl segs :h back 1.01 ± ± ± ± 0.02 variety: F 1,19 =20.58, P<0.001, sex: F 1,19 =22.16, P<0.001, variety x sex: F 1,19 =37.91, P<0.001, R 2 =0.78 w pelv :Σl segs 0.17 (N=6) variety: F 1,18 =11.368, P=0.003, sex: F 1,18 =0.347, P=0.079, R 2 =0.39 Standard errors are not presented where they were 0.00 to 2 decimal places The sample size is indicated in brackets if lower than the total sample size of the bird group a Mean posture index was calculated from a reduced sample of L taken from Rose et al (2015a), because there were no kinematic data for 2 individuals 9

77 Walking kinematics and dynamics Similar to previous findings in bipeds (Gatesy and Biewener 1991), absolute DF, t swing and t stance were negatively correlated, and l stride and f stride were positively correlated with U. Size correcting the kinematics parameters and U did not change the signs (positive or negative) of these correlations (Fig. 1). For all four groups of leghorns, the exponents describing the relationships between absolute or sizenormalised parameters and U or Fr were similar, unless otherwise stated below. Across all speeds, DF (Fig 1A) was greater in females than males (~ 2 %) and greater in the bantam than in the standard variety by ~ 2 % (Table 2). At comparable Fr, DF was, similarly, greater in females than in males in both varieties and this sex difference was greater in the bantams. In addition, DF was greater in the bantam than in the standard variety (Fig 1B; Table 2). The rate of decrease in DF with Fr was higher in the standard compared to the bantam variety and the divergence in DF between B and L was greater than that between B and L across the range of Fr. At any given U, l stride was greater in males than in females by a greater amount in the standard variety (70 mm) than in the bantam variety (20 mm) (Fig 1C; Table 2) and was associated with a greater difference between the males of the two varieties than between the females of the two varieties. At any given value of Fr, ˆL (Fig 1D), was greater in females than in males in the bantam variety, but, contrastingly, was greater in males than in females in the standard variety (Table 3). f stride (Fig 1E) was greater in females compared to males at any given U, but this sex difference was greater in the standard (0.21 Hz) than in the bantam (0.09 Hz) variety (Table 2). At comparable Fr, ˆF, was greater in males compared to females in the bantams, but, again, contrastingly, greater in females compared to males in the standard variety (Fig 1F; Table 2). At each U, t swing (Fig 1G) was greater in males than in females and this difference was lower in the bantam (0.02 s) than standard (0.04 s) variety (Table 2). In the bantams, ˆT swing (Fig 1H), was similar in the two sexes, whilst in the standard variety, it was significantly greater in males compared to females at a given Fr. 10

78 a B B L L b DF 0.7 DF c l stride (m) d ˆL e f stride (Hz) f ˆF g t swing (s) h swing ˆT i t stance (s) j stance ˆT U (m s -1 ) Fr Figure 1 Gait kinematics versus speed (left column) and size-corrected parameters and speeds (right column). (a-b) Duty factor. (c-d) Stride length. (e-f) Stride frequency. (g-h) Swing duration. (i-j) Stance duration. Each data point represents a trial for an individual bird. Best-fit equations (Table 2) were taken from the coefficients tables from the final linear models. 11

79 Table 2. Results of the linear models that tested for sex differences in absolute/normalised kinematics with speed/froude. Parameter Final model Lines of best fit DF U (F 1,111 =315.80, P<0.001),variety (F 1,111 =7.38, P=0.008), sex (F 1,111 =27.79, P<0.001), R 2 = 0.79 B : = -0.16U+0.79, B : = -0.16U+0.81, L : = -0.16U , L : = -0.16U+0.79 logdf logfr (F 1,109 =276.13, P<0.001), variety (F 1,109 =46.30, P<0.001), sex (F 1,109 =18.87, P<0.001), logfr x variety (F 1,109 =3.96, P=0.049), B : = 0.63Fr -0.05, B : = 0.66Fr L : = 0.59Fr -0.07, L : = 0.60Fr variety x sex (F 1,109 =6.69, P=0.011), R 2 = 0.77 logf stride logu (F 1,110 =883.84, P<0.001), variety (F 1,110 =238.46, P<0.001), sex (F 1,110 =32.60, P<0.001), variety x sex (F 1,110 =6.68, P=0.011), R 2 = 0.90 B : = 2.30U 0.52, B : = 2.39U 0.52 L : = 1.77U 0.52, L : = 1.98U 0.52 log ˆF logfr (F 1,110 =853.34, P<0.001), variety (F 1,110 =0.58, P=0.447), sex (F 1,110 =38.70, P<0.001), variety x sex (F 1,110 =119.10, P<0.001), R 2 = 0.90 B : = 0.39Fr 0.26, B : = 0.35Fr 0.26 L : = 0.33Fr 0.26, L : = 0.42Fr 0.26 logl stride logu (F 1,110 =662.71, P<0.001), variety (F 1,110 =222.07, P<0.001), sex (F 1,110 =40.77, P<0.001), variety x sex (F 1,110 =10.93, P=0.001), R 2 = 0.92 B : = 0.43U 0.46, B : = 0.41U 0.46 L : = 0.56U 0.46, L : = 0.49U 0.46 log ˆL logfr (F 1,110 =618.59, P<0.001), variety (F 1,110 =0.20, P=0.656), sex (F 1,110 =43.82, P<0.001), variety x sex (F 1,110 = P<0.001), R 2 = 0.88 B : = 2.50Fr 0.23, B : = 2.78Fr 0.23 L : = 2.95Fr 0.23, L : = 2.30Fr 0.23 logt swing logu (F 1,110 =94.88, P<0.001),variety (F 1,110 =225.41, P<0.001), sex (F 1,110 =71.69, P<0.001), variety x sex (F 1,110 =4.51, P=0.036), R 2 = 0.73 B : = 0.16U -0.22, B : = 0.14U L : = 0.22U -0.22, L : = 0.18U log ˆT swing logfr (F 1,110 =102.44, P<0.001), variety (F 1,110 =40.06, P<0.001), sex (F 1,110 =82.55, P<0.001), variety x sex (F 1,110 =61.91, P<0.001), R 2 = 0.69 B : = 1.04Fr -0.11, B : = 1.04Fr L : = 1.31Fr -0.11, L : = 1.00Fr logu (F 1,110 = , P<0.001), variety (F 1,110 =214.12, P=0.079), B := 0.28U -0.64, B : = 0.28U logt stance log ˆT stance sex (F 1,110 =22.35, P<0.001), variety x sex (F 1,110 =13.76, P<0.001), R 2 = 0.93 logfr (F 1,110 = , P<0.001), variety (F 1,110 =2.89, P=0.092), sex (F 1,110 =26.28, P<0.001), variety x sex (F 1,110 =173.83, P<0.001), R 2 = 0.93 L : = 0.36U -0.64, L : = 0.33U B : = 1.60Fr -0.32, B : = 1.85Fr L : = 1.85Fr -0.32, L : = 0.47Fr The adjusted R 2 values of the final models are reported. 12

80 t stance (Fig 1I) was similar in B and B, but was greater in L compared to L by 0.03s at all U (Table 2). Across all Fr, ˆT stance (Fig 1J; Table 2) was greater in B than in B, yet lower in L than in L. Therefore, none of the sex or variety differences in gait kinematics were accounted for correcting for body size differences using Fr. The two varieties differed in the mechanisms by which females had elevated DF relative to males. In the bantams, females had relatively longer stance durations than males (Fig 1J) and the sexes shared similar swing dynamics (Fig 1H). In the standard variety, however, females had relatively shorter swing and stance durations than males, but the sex difference in ˆT swing was much greater than that in ˆT stance (Fig 1H and J). DISCUSSION This study represents the first detailed comparison of the gait dynamics of the sexes in any species. Leghorn chickens, although all similar in their hind limb segment geometry, show considerable variation in limb posture and the relative contributions of anatomical components (skeletal muscle, visceral organs and reproductive tissues) to total body mass. In association with these morphological differences, and in agreement with our hypothesis, none of the leghorn groups walked with dynamic similarity. Incremental responses of absolute kinematics parameters to increasing U are generally greater in smaller species (Gatesy and Biewener 1991). All birds in the present study, however, showed similar incremental kinematic responses to U, despite the size differences associated with sex and variety. Most of the sex differences in absolute kinematic parameters paralleled inter-species differences, associated with body size (Gatesy and Biewener 1991). In females of the two varieties, f stride was greater, and l stride, t swing and t stance, smaller at any given U compared to that in their conspecific males, which had greater body size (except for t stance in the bantams, which was similar between the sexes). Similarity, f stride was greater, and l stride, t swing and t stance shorter at any given U in the bantams compared to the standards. The only absolute kinematics parameter that was not 13

81 comparable to inter-species patterns associated with body size was DF. Interspecific scaling patterns, would predict the heavier and more long-legged, animal to have a greater DF than the lighter, shorter-legged one, at the same U (Gatesy and Biewener 1991). In contrast, here, females walked with greater DF than males, and bantams walked with greater DF than standards. In agreement with inter-species differences in relative DF associated with body size (Alexander and Jayes 1983; Gatesy and Biewener 1991), at any given Fr, DF was still higher in the smaller bantam relative to the standard variety, and in females relative to males. Deviations from dynamic similarity of motion with increasing M b are usually associated with increasing limb erectness, i.e., an increasing h hip to Σl seg ratio (Biewener 1989, 1987; Gatesy and Biewener 1991). Smaller, more crouched, species can achieve greater l stride relative to their h hip because they can extend the crouched limb, which in turn, allows a greater DF (Gatesy and Biewener 1991). In contrast, a more erect limb is constrained in terms of the range of l stride and DF it can achieve, relative to a given h hip (Gatesy and Biewener 1991). That all of the birds in the present study shared similar pelvic limb skeletal geometry provides a control for the potential effects of differing limb segment proportions on walking dynamics and allows investigation of the influence of additional morphological characteristics on their walking dynamics, such as posture. Despite L having the most upright limbs, and the lowest relative stride lengths (Fig 1 E), however, they still produced a greater DF relative to h hip than did the L, whose limbs were more crouched. Furthermore, since sexual dimorphism in limb posture was exclusive to the standards, limb posture cannot explain the similar sex differences in DF between the two varieties. The sexual dimorphism in posture in the standard variety only was reflected in the opposite sex-specific dynamics of ˆL, ˆF, ˆT swing and ˆT stance at any given Fr between the two varieties (i.e the sex differences in gait dynamics were variety-specific, yet ultimately led to greater DF in females than males). The lower ˆL in L than in B and higher ˆF in L than in B are consistent with the general consensus that a more upright limb limits the length of a step relative to h hip (Gatesy and Biewener 1991). 14

82 By adopting a more upright limb posture, larger animals reduce the forces that the muscles must exert and that the bones must resist to counteract joint moments, which would otherwise scale geometrically ( M 1/3 b ) (Biewener 1989). Until recently, this has been considered the principal reason for the scaling of limb posture in vertebrates (Biewener 1989). Why smaller animals do not also have upright limbs so that they could have relatively more gracile and lighter bones, however, is not accounted for by this explanation. Explanations proposed to account for a more crouched limb include that it may improve manoeuvrability (Biewener 1989) stability (Gatesy and Biewener 1991) or minimize the cost of work associated with bouncing viscera (Daley and Usherwood 2010). Another potential explanation, however, which accounts for all postures, is that posture is not related to reducing muscle force with increasing size but instead, optimising work and power demands (which are scale-dependent) to minimise the volume of active muscle for a given size (Usherwood 2013). In this case, a crouched limb (allowing longer DF) for small animals may serve to reduce the power demands (which are high because at any given U shorter legs require shorter stances), whilst a more upright limb suits the work demands of being large (which are high because a disproportionate amount of body weight must be supported). The females in the present study may, therefore, benefit from greater DF, which would decrease the elevated power demands associated with having small limbs, yet greater body weight to support per unit of muscle mass because of gravidity (Mitchell et al. 1931). The L in the present study may have adopted kinematic and postural mechanisms for reducing both the elevated work demands due to gravity (via upright limbs) and the power demands of being small (via longer DF, achieved without a crouched posture). L may require a more upright limb than B because of their greater relative reproductive tissue mass. In B minimising power via a greater relative DF (exceeding that in L ) appears to be more important. Although manipulative trunk load carrying studies in birds are not associated with any changes in gait kinematics (Tickle et al. 2010; Tickle et al. 2013; McGowan et al. 2006), these experiments involve unnatural loads applied in backpacks and therefore may not represent a true gravid loading condition. Equally, males may benefit from lower DF relative to females via the 15

83 minimization of work demands associated with changes in fore-aft acceleration and deceleration, because of being larger. Alternatively/additionally, the greater DF in females relative to males may be in order to reduce peak muscle forces in supporting body weight, which may be particularly important when carrying the weight of developing eggs, yet with proportionally less muscle volume (Mitchell et al. 1931). Furthermore, layer chickens are well known to suffer from osteoporosis associated with the utilisation of calcium from limb bone medullary in order to form egg-shells (Whitehead 2004; Dacke et al. 1993). A greater proportion of ground contact throughout the stride to reduce peak forces exerted on the bones may reduce the risk of bone fracture. It is also possible that additional sexual dimorphisms, perhaps in muscle physiology or morphology, are linked to the sex differences in dynamics. For example, simply the greater relative pelvic limb muscle mass in males, relative to females (Mitchell et al. 1931), may increase limb inertia and prolong the swing phase of the limb, increasing its contribution to the stride period. In summary, this study represents the first detailed comparison of male and female gait dynamics in a bird. Clear departures from dynamic similarity of motion were evident between the sexes in standard and bantam varieties of leghorn chicken. Females walked with greater relative DF than males at any given Fr, but this sex difference was achieved through alternative kinematic mechanisms in each variety and linked to variety differences in sex-specific posture. L, who carry a greater relative reproductive mass than B, potentially represent the first documented example of an animal adopting mechanisms for minimising the demands of both work (via an upright limb, relative to the bantams) and power (via a longer DF than their heavier, more crouched male conspecifics). 16

84 Competing interests The authors have no competing interests Author s contributions R.L.N, J.R.C and K.A.R designed the study and contributed to the manuscript. K.A.R conducted the experiments and statistical analyses with advice from R.L.N and J.R.C. Funding This research was funded by the BBSRC (G01138/1 and /1 to J.R.C). K.A.R was supported by a NERC DTA stipend and CASE partnership with the Manchester Museum. 17

85 REFERENCES Abourachid A (2000) Bipedal locomotion in birds: the importance of functional parameters in terrestrial adaptation in Anatidae. Can J Zool 78: Abourachid A (2001) Kinematic parameters of terrestrial locomotion in cursorial (ratites), swimming (ducks), and striding birds (quail and guinea fowl). Comparative Biochemistry and Physiology a- Molecular and Integrative Physiology 131: Abourachid A, Renous S (2000) Bipedal locomotion in ratites (Paleognatiform): examples of cursorial birds. Ibis 142: Alexander RM (1976) Estimates of Speeds of Dinosaurs. Nature 261: Alexander RM, Jayes AS (1983) A dynamic similarity hypothesis for the gaits of quadrupedal mammals. J Zool 201: Baumel JJ (1953) Individual variation in the white- necked raven. Condor 55: doi:(doi: / ) Bhambhani Y, Singh M (1985) Metabolic and Cinematographic Analysis of Walking and Running in Men and Women. Med Sci Sport Exer 17: Biewener AA (1987) Scaling body support in sammals - limb posture and the relative mechanical advantage of limb muscles. Am Zool 27:A25- A25 Biewener AA (1989) Scaling body support in mammals - limb posture and muscle mechanics. Science 245:45-48 Carlson BE, McGinley S, Rowe MP (2014) Meek Males and Fighting Females: Sexually- Dimorphic Antipredator Behavior and Locomotor Performance Is Explained by Morphology in Bark Scorpions (Centruroides vittatus). Plos One 9 Cho SH, Park JM, Kwon OY (2004) Gender differences in three dimensional gait analysis data from 98 healthy Korean adults. Clin Biomech 19: Dacke CG, Arkle S, Cook DJ, Wormstone IM, Jones S, Zaidi M, Bascal ZA (1993) Medullary Bone and Avian Calcium Regulation. J Exp Biol 184:63-88 Daley MA, Usherwood JR (2010) Two explanations for the compliant running paradox: reduced work of bouncing viscera and increased stability in uneven terrain. Biol Lett 6: Fox J, Weisberg S (2011) An {R} companion to applied regression, second edition. Thousand Oaks CA: sage. 18

86 Franklin ME, Conner- Kerr T (1998) An analysis of posture and back pain in the first and third trimesters of pregnancy. J Orthop Sports Phys Ther 28: Gatesy SM, Biewener AA (1991) Bipedal locomotion - effects of speed, size and limb posture in birds and humans. J Zool 224: Griffin TM, Kram R, Wickler SJ, Hoyt DF (2004) Biomechanical and energetic determinants of the walk- trot transition in horses. J Exp Biol 207: Hammond KA, Chappell MA, Cardullo RA, Lin RS, Johnsen TS (2000) The mechanistic basis of aerobic performance variation in red junglefowl. J Exp Biol 203: Hof AL (1996) Scaling gait data to body size. Gait Posture 4: Hubel TY, Usherwood JR (2015) Children and adults minimise activated muscle volume by selecting gait parameters that balance gross mechanical power and work demands. J Exp Biol 218: Husak JF, Swallow JG (2011) Compensatory traits and the evolution of male ornaments. Behaviour 148:1-29 Jenkins FA (1971) Limb Posture and Locomotion in Virginia Opossum (Didelphis- Marsupialis) and in Other Non- Cursorial Mammals. J Zool 165:303- & Lees JJ, Nudds RL, Folkow LP, Stokkan KA, Codd JR (2012) Understanding sex differences in the cost of terrestrial locomotion. Proc R Soc B 279: Lislevand T, Figuerola J, Szekely T (2009) Evolution of sexual size dimorphism in grouse and allies (Aves: Phasianidae) in relation to mating competition, fecundity demands and resource division. J Evolution Biol 22: Lourdais O, Shine R, Bonnet X, Brischoux F (2006) Sex differences in body composition, performance and behaviour in the Colombian rainbow boa (Epicrates cenchria maurus, Boidae). J Zool 269: McGowan CP, Duarte HA, Main JB, Biewener AA (2006) Effects of load carrying on metabolic cost and hindlimb muscle dynamics in guinea fowl (Numida meleagris). J Appl Physiol 101: doi: /japplphysiol Mitchell HH, Card LE, Hamilton TS (1931) A technical study of the growth of White Leghorn chickens. Illinois agricultural experiment station bulletin 367: Nudds RL, Gardiner JD, Tickle PG, Codd JR (2010) Energetics and kinematics of walking in the barnacle goose (Branta leucopsis). Comp Biochem Physiol A Mol Integr Physiol 156: doi: /j.cbpa

87 Owens IPF, Hartley IR (1998) Sexual dimorphism in birds: why are there so many different forms of dimorphism? Proc R Soc B 265: Reilly SM (2000) Locomotion in the quail (Coturnix japonica): The kinematics of walking and increasing speed. J Morphol 243: Remes V, Szekely T (2010) Domestic chickens defy Rensch's rule: sexual size dimorphism in chicken breeds. J Evolution Biol 23: Rose KA, Nudds RL, Butler PJ, Codd JR (2015a) Sex differences in gait utilization and energy metabolism during terrestrial locomotion in two varieties of chicken (Gallus gallus domesticus) selected for different body size. Biol Open 4: doi: /bio Rose KA, Nudds RL, Codd JR (2015b) Intraspecific scaling of the minimum metabolic cost of transport in leghorn chickens (Gallus gallus domesticus): links with limb kinematics, morphometrics and posture. The Journal of experimental biology 218: doi: /jeb Rose KA, Tickle PG, Lees JJ, Stokkan KA, Codd JR (2014) Neither season nor sex affects the cost of terrestrial locomotion in a circumpolar diving duck: the common eider (Somateria mollissima). Polar Biol 37: Rubenson J, Heliams DB, Lloyd DG, Fournier PA (2004) Gait selection in the ostrich: mechanical and metabolic characteristics of walking and running with and without an aerial phase. Proc R Soc B 271: Salazar VL, Stoddard PK (2008) Sex differences in energetic costs explain sexual dimorphism in the circadian rhythm modulation of the electrocommunication signal of the gymnotiform fish Brachyhypopomus pinnicaudatus. J Exp Biol 211: Scantlebury M, Speakman JR, Bennett NC (2006) The energy costs of sexual dimorphism in mole- rats are morphological not behavioural. Proc R Soc B 273:57-63 Shillington C, Peterson CC (2002) Energy metabolism of male and female tarantulas (Aphonopelma anax) during locomotion. J Exp Biol 205: Shine R (1989) Ecological Causes for the Evolution of Sexual Dimorphism - a Review of the Evidence. Q Rev Biol 64: Shine R, Keogh S, Doughty P, Giragossyan H (1998) Costs of reproduction and the evolution of sexual dimorphism in a 'flying lizard' Draco melanopogon (Agamidae). J Zool 246: Slatkin M (1984) Ecological Causes of Sexual Dimorphism. Evolution 38:

88 Smith LK, Lelas JL, Kerrigan DC (2002) Gender differences in pelvic motions and center of mass displacement during walking: Stereotypes quantified. J Women Health Gen- B 11: Team RDC (2011) R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing, Tickle PG, Lean SC, Rose KA, Wadugodapitiya AP, Codd JR (2013) The influence of load carrying on the energetics and kinematics of terrestrial locomotion in a diving bird. Biol Open 2: doi: /bio Tickle PG, Richardson MF, Codd JR (2010) Load carrying during locomotion in the barnacle goose (Branta leucopsis): The effect of load placement and size. Comp Biochem Physiol Part A Mol Integr Physiol 156: doi: /j.cbpa Tolkamp BJ, Emmans GC, Yearsley J, Kyriazakis I (2002) Optimization of short- term animal behaviour and the currency of time. Anim Behav 64: doi: /anbe Usherwood JR (2013) Constraints on muscle performance provide a novel explanation for the scaling of posture in terrestrial animals. Biol Lett 9 Watson RR, Rubenson J, Coder L, Hoyt DF, Propert MW, Marsh RL (2011) Gait- specific energetics contributes to economical walking and running in emus and ostriches. Proc R Soc B 278: doi: /rspb Whitehead CC (2004) Overview of bone biology in the egg- laying hen. Poultry Sci 83:

89 5. Ontogeny of sex differences in the energetics and kinematics of terrestrial locomotion in leghorn chickens (Gallus gallus domesticus) This chapter is a draft of an article submitted to Scientific Reports Rose, K.A., Bates, K. T., Nudds, R. L Codd, J. R (in review). Ontogeny of sex differences in the energetics and kinematics of terrestrial locomotion in leghorn chickens (Gallus gallus domesticus) 51

90 Ontogeny of sex differences in the energetics and kinematics of terrestrial locomotion in leghorn chickens (Gallus gallus domesticus) Rose, K. A., Bates, K. T. Nudds, R. L. and Codd, J. R. Faculty of Life Sciences, University of Manchester, Manchester, M139PT *Address for reprints and other correspondence: Dr. Jonathan Codd Faculty of Life Sciences, University of Manchester, Manchester M13 9PT, UK Tel: +44(0)

91 ABSTRACT Sex differences in locomotor performance may precede the onset of sexual maturity and/or arise concomitantly with secondary sex characteristics. Here, we present the first study to quantify the terrestrial locomotor morphology, energetics and kinematics in a species, either side of sexual maturation. In domestic leghorn chickens (Gallus gallus domesticus) sexual maturation brings about permanent female gravidity and increased male hind limb muscle mass. We found that the sexes of a juvenile cohort of leghorns shared similar maximum sustainable speeds, while in a sexually mature cohort maximum sustainable speeds were greater by 67% (males) and 34% (females). Furthermore, compared to in juveniles, the absolute duration of leg swing was longer and shorter in sexually mature males and females, respectively. Consequently, the proportion of a stride in contact with the ground was higher in sexually mature females compared to males. Modulation of the proportion of ground contact with the development of secondary sex characteristics may act to minimize mechanical work in males; and minimise mechanical power and/or peak force in females. A greater incremental response of mass-specific metabolic power to speed in males compared to females was common to both age cohorts and, therefore, likely results from physiological sexual dimorphisms that precede sexual maturation. 2

92 INTRODUCTION Artificial selection in the domestic chicken (Gallus gallus domesticus) has led to derived morphology, physiology and behaviour distinct from that of its red jungle fowl ancestor (Fumihito et al. 1994). Artificial selection may result in unintended pathological consequences in farm animals (Rauw et al. 1998). For example, in the broiler chicken, selection for increased muscle growth rates has led to a compromise in the effectiveness of the reparatory apparatus (Tickle et al. 2014), abnormalities of the musculoskeletal system (Knowles et al. 2008), and the negative ontogenetic allometry of the heart and lungs associated with a number of pathologies (Tickle et al. 2014). The influence of this type of selection in broilers upon locomotor mechanics (Paxton et al. 2013; Reiter and Bessei 1997; Corr et al. 2003; Caplen et al. 2012) and morphology (Paxton et al. 2010; Talaty et al. 2010), across ontogeny (Tickle et al. 2014; Paxton et al. 2014), and also in comparison to less derived/wild-type strains (Paxton et al. 2010) is well studied. In layer chickens, however, selected for increased reproductive output (size and frequency of eggs laid), potential changes in locomotor physiology and mechanics associated with the developmental process have not yet been investigated. The energy budgets of animals are limited (Weiner 1992), and consequently, trade-offs in resource allocation exist at different life stages (Stearns 1989). In young animals, a bias in energy is allocated towards somatic tissue growth. An inherent trade-off exists, however, between growth rate and the maturation of tissues required for locomotion (Ricklefs 1979b, a; Ricklefs et al. 1994). Linked to this compromise, avian species exhibit distinct differences in ontogenetic strategy. Precocial birds (usually cursorial), for example, prioritise efficient locomotion over growth from hatch and will grow 3-4 times slower than altricial birds (often principally flyers) that do not begin locomotion until after the whole body growth period (Case 1978; Ricklefs 1979a, b). Other bird groups, with two or more principle modes of locomotion, such as the mallard (Anas platyrhynchos), which flies, walks and swims, exhibit differential ontogenetic strategies between the hind- and forelimbs, depending on which mode they use principally in early life (Dial and Carrier 2012). The precocial strategy is thought 3

93 to have arisen due to the strong evolutionary pressure posed by high predation rates on vulnerable juveniles (Wassersug and Sperry 1977; Arnold and Wassersug 1978), which are handicapped due to small body size, rapidly growing (softer) tissues and naiveté about their environment (Carrier 1996). Sexual maturation usually occurs later in the ontogenetic trajectory (DuPreez and Sales 1997; Parker and Garant 2005), at which point, energy no longer required for growth can be invested in reproduction (Stearns 1989). Although precocial, white leghorn (layer) chickens allocate their energy differently to wild precocial birds because food is no longer a limiting resource. The ability of the digestive and transport systems of the bird becomes the limiting factor in terms of energy intake and allocation to tissues. These environmental conditions combined with artificial selection have allowed for a shift in energy allocation in the sexes. In females, the emphasis is on reproduction, bringing forward the onset of egg-laying (Schutz et al. 2002), which is continuous throughout the lives of hens, rather than occurring only in breeding seasons. Male birds on the other hand, invest substantial energy in skeletal muscle tissue allocation, possessing greater muscle, bone, heart and blood masses compared to mature females, which outweigh males in digestive components, skin and fat as well as the reproductive system (Mitchell et al. 1931). These birds, therefore, exhibit distinct sexual dimorphisms in skeletal muscle and reproductive tissue allocation. The sexes, however, share similar initial post-hatch growth trajectories in body mass (M b ) (Mitchell et al. 1931). At the onset of sexual maturity (roughly 4-5 months old), the differentiation of secondary sexual characteristics is mediated by a rise in gonadal hormones and male skeletal muscle growth rates increase relative to females. Female growth also terminates before that of males, leading to strong male-biased sexual size dimorphism (Mitchell et al. 1931). Sexually mature leghorns exhibit sex differences in energy metabolism during locomotion, whereby the incremental increase in mass-specific metabolic power (P met, W kg -1 ) with speed (U) is steeper in males than in females (Rose et al. 2015a). It is therefore more metabolically costly for males to move at faster speeds compared to females. This difference in cost could not be explained by differences in body size by comparison over dynamically similar speeds, and it 4

94 was hypothesised to be the result of additional sexual dimorphisms in morphology and/or physiology (Rose et al. 2015a). Furthermore, males were able to sustain maximum speeds (U max ) approximately 25 % greater than females (Rose et al. 2015a). It is unclear whether these differences are associated with female or male specialisations or constraints, or a combination of these. It is also unclear whether these sex differences in locomotion are already manifest in the juvenile form or develop at the onset of sexual maturity. Furthermore, it is unclear whether the onset of continuous gravidity impedes upon the locomotor abilities of hens. The aim of this study was to investigate the effects of the ontogenetic differences in male and female morphology on the energetics and kinematics of locomotion in white leghorns. To achieve this, two age cohorts were selected for comparison: one prior to the onset of sexual maturity (juvenile [14-16 week-old] males and females: J and J ) and another that was sexually mature ( 20-weekold: M and M ). Sexual differentiation at maturity is a gradual process and may have already initiated in the younger cohort; however we were able to confirm that the mature cohort displayed male crowing and secondary sexual characteristics (large red combs and wattles) and female egg laying, whilst the juvenile cohort did not. We also quantified the accompanying sex-specific musculoskeletal and reproductive volumes and dimensions. The hypothesis tested was that none of the locomotor differences would be present in the juvenile cohort i.e., the sexual dimorphism in locomotor performance would develop concomitantly with the secondary sex characteristics in these birds. MATERIALS AND METHODS Animals Metabolic measurements from M and M ( 20-week-old) were taken from Rose et al (2015a). Juvenile (14-16 weeks-old) white leghorns (J : N=5; M b = 1.10 ± 0.10 kg, mean ± s.e.m) and J (N=7: M b = 1.05 ± 0.03 kg, mean ± s.e.m) were obtained from the same local suppliers and housed under the same conditions in the University of Manchester s animal unit with the same feeding regimes as the 5

95 birds in Rose et al (2015a). None of the J were crowing or exhibiting aggressive behaviour and J were not gravid when examined post mortem (whilst the opposite was true for the mature cohort (Rose et al. 2015a)), confirming that the birds had not reached sexual maturity. Experimental procedures were carried out under ethical approval from the University of Manchester Ethics Committee and in accordance with the Animals (Scientific Procedures) Act 1986, covered by a UK Home Office project licence (40/3549) held by Dr Codd. Respirometry Metabolic rates were measured from the birds at rest (standing) and during locomotion on a motorised treadmill (Tunturi T60, Turku, Finland). An open-flow respirometry set up (described in Rose et al (Rose et al. 2015a; Rose et al. 2015b)) was used to measure rates of O 2 consumption (!V CO2, ml min -1 ) and CO 2 production (, ml min -1 ). The chamber (97.5 x 53.5 x 48.0 cm) within which the birds exercised and the main flow rate (250 L min -1 ) directed through it were identical to those for the sexually mature leghorns in Rose et al (2015a). Juvenile respiratory exchange ratios (RERs:!V : VO2 CO2! ) were similar to those reported for the mature cohort in Rose et al (2015a). Thermal equivalents (Brody 1945) of the!v O2 RERs were used to convert!v O2 into metabolic power P met (W). Metabolic rates during quiet standing were subtracted from locomotor metabolic rates to determine the net metabolic cost of locomotion surplus to maintenance and postural costs. The cost of transport (CoT tot, J kg -1 m -1 ) was calculated as P met /U. The juveniles were exercised at speed intervals, up to the maximum that they could sustain: 0.28, 0.42, 0.56, 0.69 and 0.83 m s -1. Each bird completed two trials, comprised of 2-3 speeds in a random order, interrupted by 5 min resting periods during which the birds stood quietly and gas levels plateaued. Resting (standing) metabolic rate was taken from the final rest trace in each trial. 6

96 Kinematics Kinematics parameters were obtained using the exact protocol used in Rose et al (Rose et al. 2015b). All trials were filmed with a high-speed (100 frames s -1 ) video camera (HDR- XR520VE, Sony, Japan) from the side. The tip of digit 3 on the foot closest to the camera was tracked over ~10 strides using Tracker software (v. 4.05, Open Source Physics). Temporal data were used to calculate f stride, l stride (U/f stride ), t swing, t stance and DF. Morphological measurements Juvenile carcasses were scanned using computed tomography (CT) at the Small Animal Teaching Hospital at the University of Liverpool. Three-dimensional reconstruction of full skeletons was carried out by image segmentation and meshing in 3D Slicer ( MeshLab ( was subsequently used to measure lengths of the femur, tibiotarsus and tarsometatarsus from the 3D skeletal models. Five frozen carcasses from each of the bird groups, excluding J (N=4), were defrosted for 24 hours prior to dissection. Thirteen major pelvic limb muscles (Table 1) with known contributions to the swing and stance phases of the limb (Marsh et al. 2004; Jacobson and Hollyday 1982) were chosen for comparison. The muscles were identified based on a description by Paxton et al (Paxton et al. 2010) and dissected from the right pelvic limb and weighed using electronic scales (± 0.01 g). Table 1. Major pelvic limb muscles of the chicken pelvic limb and their abbreviations. Muscle Abbreviation Part of the limb M. iliotibialis cranialis IC Proximal M. iliotibialis lateralis (pre and post acetabularis) IL Proximal M. iliofibularis ILFB Proximal M. flexor cruris lateralis pars pelvica FCLP Proximal M. flexor cruris medialis FCM Proximal M. iliotrochantericus caudalis ITC Proximal M. femerotibialis medialis FMT Proximal M. pubioischiofemoralis pars lateralis PIFL Proximal M. pubioischiofemoralis pars medialis PIFM Proximal M. gastrocnemius pars lateralis GL Distal M. gastrocnemius pars medialis GM Distal M. fibularis lateralis FL Distal M. tibialis cranialis caput tibiale and femorale TCT/F Distal 7

97 Statistical analyses Only data from the range of speeds utilised by all birds (i.e. up to the U max of the juvenile birds, 0.83 m s -1 ) were included in statistical analyses. All data from the mature birds are presented in the graphs, however, to show their capacities for U max and associated kinematics. The mature birds were known be using walking gait mechanics at speeds up to and including 0.83 m s -1 (Rose et al. 2015a). In the juveniles, a lack of an inflection in the relationships between all kinematic/metabolic parameters with U suggests that a gait transition did not occur. All statistical analyses were performed using the car package version (Fox and Weisberg 2011) on R GUI 1.42 Leopard build 64-bit (Team 2011). Shapiro-wilk tests were performed on the standardised residuals of the models to ensure that the data approximated a normal distribution. Where the data did not conform to a normal distribution, data were log transformed. Data were also log transformed if it improved the Akaike s information criterion of the models. Age-cohort and sex were included as fixed factors in all models. Twoway analyses of variance (ANOVAs) were used to test for differences in morphological measurements. Linear models were conducted to test for differences in the relationships between energetic or kinematic variables and U. U was included as a covariate in the models. All potential interaction terms were considered in all primary models before a step-wise backward deletion of nonsignificant interaction terms was conducted. The final model outputs are reported. Best-fit lines were obtained using the effect sizes from the coefficients tables output by the statistical models and were back-transformed where data had been log-transformed. 8

98 RESULTS Body mass M b (Table 2) was significantly greater in the sexually mature, compared to the juvenile cohort (Table 3). A significant age x sex interaction in M b was identified due to similar masses in the sexes of the juvenile cohort (1.05-fold greater in J than in J ), but a 1.34-fold greater M b in M than in M (Table 3). This was associated with a greater difference in M b between J and M (0.82 kg) than between J and M (0.41 kg). Limb bone lengths All absolute hind limb skeletal bone lengths (Table 2) were significantly longer in males compared to females (Table 3). A significant age x sex interaction was present in the sum of the three hind limb bone lengths (Σl segs ) due to a greater sex difference in the mature compared to the juvenile cohort. This was linked to a lack of difference Σl segs between J and M ; but significantly longer Σl segs in M compared to J (Table 3). Reproductive mass Reproductive mass in M was on average ± (s.d) g, which comprised ± 1.26 (s.d) % of M b. 9

99 Table 2. Mean (± s.e.m) morphological measurements cohort sex N M b (kg) l fem (mm) l tib (mm) l tars (mm) Σl segs (mm) Juvenile Female ± ± ± ± ± 5.08 Juvenile Male ± ± ± ± ± 8.64 Mature Female ± ± ± ± ± 2.44 Mature Male ± ± ± ± ± 6.06 l fem, femur length; l tib, tibiotarsus length; l tars, tarsometatarsus length; Σl segs sum of the hind limb bone lengths Results of the two-way ANOVAS conducted to test for age and sex affects are in Table 2 10

100 Table 3. Results of the two-way ANOVAs performed to determine whether age and sex affect the morphological measurements Measurement Final model M b age (F 1,16 =46.53, P<0.001), sex (F 1,16 =24.48, P<0.001), age x sex (F 1,16 =33.83, P<0.001) R 2 =0.84 l fem age (F 1,21 =3.46, P=0.077), sex (F 1,21 =26.78, P<0.001) R 2 =0.55 l tib age (F 1,21 =1.35, P=0.259), sex (F 1,21 =29.69, P<0.001) R 2 =0.56 l tars age (F 1,21 =2.00, P=0.172), sex (F 1,21 =53.54, P<0.001) R 2 =0.70 Σl segs age (F 1,20 =3.36, P=0.082), sex (F 1,20 =48.85, P<0.001), age x sex (F 1,20 =4.45, P=0.048) R 2 =0.70 log(m IL ) age (F 1,15 =22.56, P<0.001), sex (F 1,15 =22.45, P<0.001), age x sex (F 1,15 =22.87, P<0.001), R 2 =0.79 M IC age (F 1,15 =37.64, P<0.001), sex (F 1,15 =26.57, P<0.001), age x sex (F 1,15 =30.63, P<0.001), R 2 =0.84 M ILFB age (F 1,15 =17.23, P<0.001), sex (F 1,15 =28.54, P<0.001), age x sex (F 1,15 =18.35, P<0.001), R 2 =0.78 M FCLP age (F 1,15 =28.88, P<0.001), sex (F 1,15 =12.85, P=0.003), age x sex (F 1,15 =15.14, P=0.001), R 2 =0.74 M ITC age (F 1,15 =31.84, P<0.001), sex (F 1,15 =42.15, P<0.001), age x sex (F 1,15 =25.08, P<0.001), R 2 =0.85 M PIFL age (F 1,15 =19.19, P<0.001), sex (F 1,15 =26.21, P<0.001), age x sex (F 1,15 =17.15, P<0.001), R 2 =0.78 M PIFM age (F 1,15 =24.33, P<0.001), sex (F 1,15 =41.55, P<0.001), age x sex (F 1,15 =23.40, P<0.001), R 2 =0.83 M GL age (F 1,15 =46.40, P<0.001), sex (F 1,15 =43.80, P<0.001), age x sex (F 1,15 =47.14, P<0.001), R 2 =0.89 log(m GM ) age (F 1,15 =20.49, P<0.001), sex (F 1,15 =24.62, P<0.001), age x sex (F 1,15 =14.84, P<0.001), R 2 =0.77 M FL age (F 1,15 =21.71 P<0.001), sex (F 1,15 =25.88, P<0.001), age x sex (F 1,15 =31.50, P<0.001), R 2 =0.81 M TCT/F age (F 1,15 =7.60, P=0.015), sex (F 1,15 =16.84, P<0.001), age x sex (F 1,15 =10.73, P=0.005), R 2 =0.65 M FCM age (F 1,14 =5.96, P=0.027), sex (F 1,14 =3.40, P=0.085), age x sex (F 1,14 =7.06, P=0.018), R 2 =0.44 M FMT age (F 1,14 =32.23, P<0.001), sex (F 1,14 =41.18, P<0.001), age x sex (F 1,14 =33.75, P<0.001), R 2 =0.86 M IL : M b age (F 1,15 =1.98, P=0.180), sex (F 1,15 =8.79, P=0.010), age x sex (F 1,15 =20.70, P<0.001), R 2 =0.62 M IC : M b age (F 1,15 =9.77, P=0.007), sex (F 1,15 =8.04, P=0.013), age x sex (F 1,15 =25.31, P<0.001), R 2 =0.70 M ILFB : M b age (F 1,15 =1.08, P=0.315), sex (F 1,15 =18.52, P<0.001), age x sex (F 1,15 =14.46, P<0.001), R 2 =0.64 M FCLP : M b age (F 1,15 =14.39, P=0.002), sex (F 1,15 =3.67, P=0.074), age x sex (F 1,15 =18.48, P<0.001), R 2 =0.66 M ITC : M b age (F 1,15 =0.03, P=0.871), sex (F 1,15 =5.61, P=0.032), age x sex (F 1,15 =5.21, P=0.038), R 2 =0.31 M PIFL : M b age (F 1,15 =0.13, P=0.721), sex (F 1,15 =6.81, P=0.020), age x sex (F 1,15 =10.16, P=0.006), R 2 =0.44 M PIFM : M b age (F 1,15 =1.58, P=0.228), sex (F 1,15 =19.57, P<0.001), age x sex (F 1,15 =10.87, P=0.005), R 2 =0.62 M GL : M b age (F 1,15 =5.33, P=0.036), sex (F 1,15 =5.84, P=0.029), age x sex (F 1,15 =25.44, P<0.001), R 2 =0.66 log(m GM : M b ) age (F 1,15 =0.77, P=0.395), sex (F 1,15 =12.40, P=0.003), age x sex (F 1,15 =12.71, P=0.003), R 2 =0.56 M FL : M b age (F 1,15 =2.26, P=0.154), sex (F 1,15 =15.48, P=0.001), age x sex (F 1,15 =57.44, P<0.001), R 2 =0.80 M TCT/F : M b age (F 1,15 =0.00, P=0.970), sex (F 1,15 =8.76, P=0.010), age x sex (F 1,15 =6.51, P=0.022), R 2 =0.41 M FCM : M b age (F 1,14 =7.96, P=0.014), sex (F 1,14 =6.62 P=0.022), age x sex (F 1,14 =6.88, P=0.020), R 2 =0.54 M FMT : M b age (F 1,14 =0.79, P=0.388), sex (F 1,14 =10.57 P=0.006), age x sex (F 1,14 =21.88, P<0.001), R 2 =0.64 Multiple comparison adjustments were not conducted given the high significance levels (P<0.001) identified below our selected significance level threshold (P 0.05) The adjusted R 2 values are reported from the final models 11

101 Muscle measurements In each of thirteen measured pelvic limb muscles (see Table 1 for abbreviations) a significant age x sex interaction was present in absolute mass (Fig. 1a, Table 2), which did not differ significantly between the sexes of the juvenile cohort, but was greater in M than in M (Table 2). Each muscle was also of similar absolute mass in J and M, with the exception of the FCLP, which was greater in M than in J. Furthermore, each absolute muscle was greater in M than in J. A significant age x sex interaction was also present in the relative mass (%M b ) of each pelvic limb muscles (Fig. 1b, Table 2). In the mature cohort, the relative mass of each muscle was greater in M than in M. In the juvenile cohort, however, the majority of muscles were similar in relative mass between the sexes, with the exceptions of the IC, FL, GL, FCLP and FMT, in which it was greater in J than in J (the opposite sexual dimorphism to the mature cohort). Since the females of the two age cohorts did not differ significantly in the absolute masses of their individual muscles, the lower %M b of the muscles in M relative to J was due to the greater the M b of the M relative to J (attributed to increased mass in the body outside of the pelvic limb) and linked to gravidity. Therefore, in the sexually mature cohort, most of the relative muscle masses differed from those of the juvenile cohort and the direction in which they differed was opposite for each sex: M muscles had greater relative muscle masses compared to J, linked to an increase in absolute muscle mass that comes with sexual maturation; and M had lower relative muscle masses compared to J, linked to an increase in reproductive mass that comes with sexual maturation 12

102 a Muscle mass (g) J J M M FCM PIFL PIFM IC TCT/F ITC FL ILFB GL FCLP FMT GM IL b Series1 J * % body mass Series2 J LARGE M FEMALES LARGE M MALES * * * FCM PIFL PIFM IC TCT/F ITC FL ILFB GL FCLP FMT GM IL Pelvic limb muscle Figure 1 Pelvic limb muscle measurements in juvenile and sexually mature leghorns. (a) Absolute muscle mass. (b) Muscle percentage of total body mass. Muscle abbreviations are defined in Table 1. A significant age x sex interaction was identified in all measurements (Table 2). Asterisks denote where the sex differences are the opposite between the two age cohorts. Bars represent means (± s.e.m.). 13

103 Maximum sustainable speed Juvenile leghorns of both sexes reached a U max of 0.83 m s -1. In comparison, the U max of M (1.39 m s -1 ) exceeded that of M (1.11 m s -1 ) by 25 %. Mature leghorns of each sex achieved greater U max than the juveniles: the U max of M exceeded that of J, by 67 % and the U max of M exceeded that of J by 34 %. Standing metabolic rates Standing-P met was similar in the males and females within each cohort but was greater in the juvenile compared to sexually mature cohort by ~ 2 W kg -1 (Table 4). Metabolic rates during locomotion The incremental increase in gross mass-specific P met (Fig. 2a) with U was greater in males compared to females in both cohorts (Table 4) and also greater in the mature compared to the juvenile cohort (Table 4). Following the subtracting of standing-p met from gross locomotor P met to calculate net-p met (Fig. 2b), the same statistical differences were true. Therefore, faster speeds were more metabolically expensive for males than for females, and more expensive for mature, than for juvenile birds. CoT tot (Fig. 2c) decreased curvilinearly as a function of U in all four groups. The rate of decrease in CoT tot with U, however, was greater in the juvenile than in the mature cohort (Table 4). The rate of decrease was also greater in females than in males. CoT net (Fig. 2d) also decreased curvilinearly with U in all but M in which it was invariant with speed. 14

104 Table 4 Results of the final linear model outputs performed to investigate age and sex related differences in energetics and kinematics Parameter Final model Coefficients Standing (W kg -1 ) log P met (W kg -1 ) log Net-P met (W kg -1 ) log CoT tot (J kg -1 m -1 ) log CoT net (J kg -1 m -1 ) DF log t stance (s) log t swing (s) log f stride (Hz) log l stride (m) age (F 1,23 = 13.86, P = 0.001), sex (F 1,23 = 0.16, P = 0.691) R 2 = 0.33 logu (F 1,108 =53.25, P<0.001), age (F 1,108 =40.55, P<0.001), sex (F 1,108 =1.39, P=0.241), logu x age (F 1,108 =6.25, P=0.014), logu x sex F 1,108 = 8.53, P=0.004), R 2 = 0.48 logu (F 1,108 =50.93, P<0.001), age (F 1,108 =0.57, P=0.451), sex (F 1,108 =1.71, P=0.194), logu x age (F 1,108 =4.91, P=0.029), logu x sex (F 1,108 =7.49, P=0.007) R 2 = 0.35 logu (F 1,108 =462.05, P<0.001), age (F 1,108 =40.71, P<0.001), sex (F 1,108 =1.41, P=0.237), logu x age (F 1,108 =6.32, P=0.013), logu x sex (F 1,108 =8.60, P<0.001), R 2 = 0.82 logu (F 1,108 =25.39, P<0.001), age (F 1,108 =0.57, P=0.454), sex (F 1,108 =1.72, P=0.193), logu x age (F 1,108 =4.93, P=0.028), logu x sex (F 1,108 =7.50, P=0.007) R 2 = 0.24 U (F 1,100 =138.38, P<0.001), age (F 1,100 =0.02, P=0.898), sex (F 1,100 =1.08, P=0.302), age x sex (F 1,100 =5.02, P=0.027) R 2 = 0.58 logu (F 1,101 = , P<0.001), age (F 1,101 =5.87, P=0.017), sex (F 1,101 =62.84, P<0.001) R 2 = 0.93 logu (F 1,100 = 98.18, P<0.001), age (F 1,100 =2.37, P=0.127), sex (F 1,100 =43.75, P<0.001), age x sex (F 1,100 =8.05, P<0.006) R 2 = 0.59 logu (F 1,101 = , P<0.001), age (F 1,101 =5.23, P=0.024), sex (F 1,101 =103.25, P<0.001) R 2 = 0.93 logu (F 1,101 = , P<0.001), age (F 1,101 =6.02, P=0.016), sex (F 1,101 =98.97, P<0.001) R 2 = 0.93 J F : =9.44 J M : =9.19 M F : =7.17 M M : =6.92 J F : =16.62U 0.07 J M : =19.70U 0.28 M F : =15.75U 0.25 M M : =18.67U 0.46 J F : =7.08U 0.20 J M : =10.46U 0.66 M F : =8.63U 0.57 M M : =12.74U 1.03 J F : =16.51U J M : =19.58U M F : =18.56U M M : =15.65U J F : =6.93U J M : =10.60U M F : =8.82U M M : =13.49U J F : = -0.15U+0.78 J M : = -0.15U+0.78 M F : = 0.15U+0.79 M M : = 0.15U+0.77 J F : =0.34U J M : =0.37U M F : =0.32U M M : =0.36U J F : =0.19U J M : =0.21U M F : =0.18U M M : =0.22U J F : =1.95U 0.53 J M : =1.73U 0.53 M F : =2.01U 0.53 M M : =1.78U 0.53 J F : =0.51U 0.44 J M : =0.58U 0.44 M F : =0.49U 0.44 M M : =0.56U 0.44 The adjusted R 2 values are reported from the final model 15

105 a P met (W kg -1 ) Net-P met (W kg -1 ) b J J M M CoT tot (J kg -1 m -1 ) c d CoT net (J kg -1 m -1 ) U (m s -1 ) Figure 2 Mass-specific energetics parameters versus treadmill speed (U). (a) Metabolic power (P met ). (b) Net- metabolic power (Net- P met ). (c) Total cost of transport. (d) Net cost of transport. Data points are means (± s.e.m). Best-fit lines were taken from the coefficients tables of the linear models conducted to test for age and sex effects (Table 4). 16

106 Kinematics Each kinematic parameter responded (increased or decreased) with increases in U at a similar rate in all of the four bird groups, unless otherwise stated below. Duty factor (DF, the relative contribution of the stance phase to the stride period) decreased linearly with U. A significant age x sex interaction was present in DF (Fig. 3a) due to DF being greater in males than in females (<1%) in the juvenile cohort but greater in females than in males in the mature cohort (by 2%) (Table 4). Stance duration (t stance ) decreased curvilinearly with U (Fig. 3b, Table 4). Across all speeds, regardless of the age of the birds, t stance was 0.03 s greater in males than in females (Table 4). Much smaller, but significant differences were also observed between the age cohorts, with t stance being slightly greater in the mature compared to the juvenile group. Swing duration (t swing ) also decreased curvilinearly with U and at the same rate in all birds groups; however, t swing (the intercept) was greater in males compared to females at any given U by 0.02 s in the juveniles, and by 0.04 s in the mature birds (Fig. 3c, Table 4). A significant age x sex interaction was identified in t swing as the sex difference was greater in the mature compared to the juvenile cohort. The intercept was lower in M relative to J, and higher in M relative to J. Therefore, the sexes of the mature cohort deviate in t swing from their corresponding sexes in the juvenile cohort in different ways. Stride frequency (f stride ) increased with U and the trend was best described by a power function (Fig. 3d, Table 4). Sex significantly influenced f stride (Table 4), which was 0.22 Hz and 0.23 Hz faster in females than in males across all U in the juvenile and mature cohorts respectively. Again, smaller but significant differences in f stride were associated with age, which were greater in the mature compared to the juvenile cohort (by 0.06 in females and 0.05 in males). Stride length (l stride ) increased with U and the trend was also best described by a power function (Fig. 3e, Table 4). Strides were 0.07 m longer in males compared to females, and 0.02 m longer in the juvenile compared to the mature cohort across all U. 17

107 a DF J J M M b t stance (s) c t swing (s) d 2.1 f stride (Hz) l stride (m) e U (m s -1 ) Figure 3 Kinematics parameters versus treadmill speed (U). (a) Duty factor (DF). (b) Stance duration (t stance ). (c) Swing duration (t swing ). (d) Stride frequency (f stride ). (e) Stride length (l stride ). Data points are means (± s.e.m). Best-fit lines were taken from the coefficients tables of the linear models conducted to test for age and sex effects (Table 4). 18

108 DISCUSSION Here we report the first comparison of the locomotor energetics, kinematics and morphology of a species just prior to and just after the onset of sexual maturity. In comparison to a cohort of juvenile white leghorn chickens, whose sexes were similar in body form, a sexually mature cohort showed strong male-biased sexual size dimorphism, greater limb length and relative muscle mass in males and greater reproductive mass in females. Despite the large ontogenetic differences in hind limb skeletal muscle and reproductive masses identified in males and females, respectively, no age x sex interactions in locomotor energetics were identified. The lower incremental metabolic cost of locomotion in females relative to males, common to both juvenile and sexually mature leghorns must, therefore be due to sexual dimorphisms in physiology than precede the onset of sexual maturity. An age x sex interaction was identified in only U max and two kinematics parameters (t swing and DF). The sex differences in maximum performance can be linked to the measured sex differences in morphology. Maximum running speed scales with positive allometry against M b (Garland 1983); therefore the greater U max in mature, relative to juvenile, birds and in mature males relative to females, is expected simply because of greater body size. However, the lack of sex difference in U max in the juveniles (which were also dimorphic in leg length, but to a lesser degree) suggests a more important role of muscle in determining U max than leg length. Muscle physiological cross sectional area (largely influenced by muscle mass) is directly proportional to the maximum force and power that a muscle can produce (Lieber and Friden 2000). The greater volume of the hind limb skeletal muscles in M (on average 2.25-fold), therefore, likely contributed to their greater U max (which would require greater peak muscle forces) relative to the other three chicken groups. U max is not only dependent upon the maximum f stride and l stride that the birds can achieve, but also the ability of the birds to sustain the speed aerobically (~5min of locomotion is required for respirometry measurements). A lower U max in juveniles is consistent with previous findings from ontogenetic comparisons, whereby stamina is usually lower in juvenile forms (Carrier 1996). The greater 19

109 U max in M may be indicative of a greater aerobic capacity. In a study on red jungle fowl, maximum rate of oxygen consumption (!V O2 max ) in mature males exceeded that of mature females, but no sex difference was identified in chicks (Chappell et al. 1996). Therefore, it seems likely that with the onset of sexual maturity there are physiological changes in males, which increase their capacity for aerobic respiration in the muscles. There are a number of levels at which this physiological difference could manifest. For example, in the red jungle fowl,!v O2 max correlated with cecum, heart, pectoralis and hind limb skeletal muscle masses as well as pectoralis citrate synthase activity, indicative of system wide (peripheral and central organ) specialisation (symmorphosis)(hammond et al. 2000). Although there is evidence amongst vertebrate species for reductions in locomotor performance with pregnancy/gravidity (Brischoux et al. 2011; Knight 2011; Munns et al. 2015; Olsson et al. 2000; Shine 2003; Lee et al. 1996), U max was actually greater in M than in J. This is despite the two age cohorts of female sharing similar absolute limb lengths and muscle mass, whilst only M were gravid and were also 1.36-fold heavier than J. Therefore, the same quantity of muscle must generate greater forces to support the additional body weight of M. The relative muscle mass of some of the pelvic limb muscles were a lower in M than in J, meaning greater forces were actually generated from relatively less muscle in the M. Age differences in U max in female leghorns are, therefore, likely explained by differences in aerobic capacity that are not linked to muscle quantity. Interestingly, in the red jungle fowl,!v O2 max did not correlate with the mass or enzyme capacities of skeletal muscles (peripheral organs) but was correlated with haematocrit and the mass of the large intestine (a central organ) (Hammond et al. 2000). The sex differences in most of the kinematic parameters were common to both cohorts. This is likely due to the fact that the juvenile cohort also exhibited some sexual size dimorphism in hind limb skeletal length. t stance and l stride were greater, and f stride lower, in larger males relative to smaller females, as would be expected when comparing a larger animal with a smaller one (based on 20

110 interspecific comparison) (Gatesy and Biewener 1991). The sex difference in DF in the mature cohort (greater in females than in males) differed from what would be expected based on interspecific differences associated with size but can be attributed to mechano-physiological constraints imposed by the measured secondary sex characteristics. The distribution and placement of additional mass on the body have important consequences for locomotion. For example, in manipulative studies in which masses were added to the distal limbs of birds, a corresponding increase in t swing has been reported (e.g. a 5% M b load to the distal limb caused a ~16% increase in t swing in the barnacle goose (Tickle et al. 2010)). The greater muscle mass on the hind limbs (proximal and distal) and increase in leg length of M relative to that of J, may increase limb inertia, which might be expected to increase t swing. Gait kinematics which minimize muscle mechanical work requirements have shorter stances and hence, duty factors (Usherwood 2013; Srinivasan 2011; Srinivasan and Ruina 2006). It is also possible that males decrease their duty factors at sexual maturity in order to minimize muscle mechanical work demands, which would be expected to increase with the rise in muscle and bone mass. In opposition to the ontogenetic differences in t swing found in males, t swing was faster in M than in J. If all kinematics parameters were the same between M and J, peak external forces and muscle mechanical work and power would each be expected to be greater in M due to their greater reproductive and body masses, yet similar muscle masses compared to J. One potential reason for a faster t swing in gravid mature females might be to increase the relative contribution of t stance to the stride period (DF was greater in M than all other bird groups), which would allow more time for generating sufficient muscle force to support the increase in M b. A greater DF for a given U would decrease peak external forces, which may be important in hens, due to a reduction in bone strength associated with the utilisation of medullary bone calcium in egg-shell formation (Dacke et al. 1993; Whitehead 2004). Furthermore, a greater DF decreases muscle mechanical power requirements (Usherwood 2013). The ontogenetic differences in t swing in female leghorns may, therefore, also represent a power minimizing mechanism. Loads added to the backs of birds, which increase the amount of body 21

111 weight that the stance muscles must support, however, have not been associated with any changes in kinematics parameters (Tickle et al. 2010; Tickle et al. 2013). The location of the added load to the females during gravidity, however, has not yet been mimicked in any load carrying studies in birds. Pregnant humans (Branco et al. 2013) and wallabies carrying young in the pouch (Baudinette and Biewener 1998) are, like the leghorns here, known to increase DF with pregnancy. CONCLUSION Contrary to our hypothesis that sex differences in locomotor energy metabolism would be associated with sexual maturation in white leghorns; lower incremental metabolic costs of locomotion in females relative to males, were also found in juveniles. Sexual maturation in white leghorns is associated with large increases in hind limb skeletal muscle mass in males and reproductive tissue mass in females. Differences in the location of the additional tissues on the body following sexual maturity differentially impact upon the duration of the swing phase of the limb of the sexes. We suggest that the birds modulate the swing, and hence duty factor, in order to minimize muscle mechanical work (males) and mechanical power and/or peak force (females). A role of secondary sex characteristics in influencing maximum performance in males was indicated by maximum sustainable speeds 67 % greater in the mature compared to the juvenile cohort. Furthermore, no evidence was found in females for a constraint of gravidity on maximum sustainable speed. Unlike broiler chickens, which experience locomotor difficulties as they develop the muscle mass for which they were artificially selected, leghorns show a greater capacity for sustained locomotion with the onset of egg-laying. 22

112 Competing financial interests The authors have no competing interests Author s contributions KAR conceived the study. KAR, JRC, RLN and KB designed the study. JRC provided the equipment for the experiments. KB provided the birds, CT-scanned and generated 3D skeletal models, and processed kinematic data. KAR executed the experiments, performed the data and statistical analyses, interpreted the results and drafted the manuscript with advice from JRC, RLN and KB. All authors contributed to the editing of the final manuscript. 23

113 REFERENCES Arnold SJ, Wassersug RJ (1978) Differential Predation on Metamorphic Anurans by Garter Snakes (Thamnophis) - Social- Behavior as a Possible Defense. Ecology 59: Baudinette RV, Biewener AA (1998) Young wallabies get a free ride. Nature 395: Branco M, Santos- Rocha R, Aguiar L, Vieira F, Veloso A (2013) Kinematic analysis of gait in the second and third trimesters of pregnancy. J Pregnancy 2013: doi: /2013/ Brischoux F, Bonnet X, Shine R (2011) Conflicts between feeding and reproduction in amphibious snakes (sea kraits, Laticauda spp.). Austral Ecol 36:46-52 Brody S (1945) Bioenergetics and growth, with special reference to the efficiency complex in domestic animals. Reinhold, New York Caplen G, Hothersall B, Murrell JC, Nicol CJ, Waterman- Pearson AE, Weeks CA, Colborne GR (2012) Kinematic Analysis Quantifies Gait Abnormalities Associated with Lameness in Broiler Chickens and Identifies Evolutionary Gait Differences. Plos One 7 Carrier DR (1996) Ontogenetic limits on locomotor performance. Physiol Zool 69: Case TJ (1978) On the evolution and adaptive significance of postnatal growth rates in the terrestrial vertebrates. The Quarterly review of biology 53: Chappell MA, Zuk M, Johnsen TS (1996) Repeatability of aerobic performance in Red Junglefowl: Effects of ontogeny and nematode infection. Funct Ecol 10: Corr SA, McCorquodale CC, McGovern RE, Gentle MJ, Bennett D (2003) Evaluation of ground reaction forces produced by chickens walking on a force plate. Am J Vet Res 64:76-82 Dacke CG, Arkle S, Cook DJ, Wormstone IM, Jones S, Zaidi M, Bascal ZA (1993) Medullary Bone and Avian Calcium Regulation. J Exp Biol 184:63-88 Dial TR, Carrier DR (2012) Precocial hindlimbs and altricial forelimbs: partitioning ontogenetic strategies in mallards (Anas platyrhynchos). J Exp Biol 215: DuPreez JJ, Sales J (1997) Growth rate of different sexes of the European quail (Coturnix coturnix). Br Poult Sci 38: Fox J, Weisberg S (2011) An {R} companion to applied regression, second edition. Thousand Oaks CA: sage. 24

114 Fumihito A, Miyake T, Sumi SI, Takada M, Ohno S, Kondo N (1994) One Subspecies of the Red Junglefowl (Gallus- Gallus Gallus) Suffices as the Matriarchic Ancestor of All Domestic Breeds. Proc Natl Acad Sci U S A 91: doi:doi /Pnas Garland T (1983) The relation between maximal running speed and body- mass in terrestrial mammals. J Zool 199: Gatesy SM, Biewener AA (1991) Bipedal locomotion - effects of speed, size and limb posture in birds and humans. J Zool 224: Hammond KA, Chappell MA, Cardullo RA, Lin RS, Johnsen TS (2000) The mechanistic basis of aerobic performance variation in red junglefowl. J Exp Biol 203: Jacobson RD, Hollyday M (1982) A behavioral and electromyographic study of walking in the chick. J Neurophysiol 48: Knight K (2011) Pregnancy Is a Drag for Bottlenose Dolphins. J Exp Biol 214:I- I Knowles TG, Kestin SC, Haslam SM, Brown SN, Green LE, Butterworth A, Pope SJ, Pfeiffer D, Nicol CJ (2008) Leg disorders in broiler chickens: prevalence, risk factors and prevention. Plos One 3:e1545. doi: /journal.pone Lee SJ, Witter MS, Cuthill IC, Goldsmith AR (1996) Reduction in escape performance as a cost of reproduction in gravid starlings, Sturnus vulgaris. Proc R Soc B 263: Lieber RL, Friden J (2000) Functional and clinical significance of skeletal muscle architecture. Muscle Nerve 23: Marsh RL, Ellerby DJ, Carr JA, Henry HT, Buchanan CI (2004) Partitioning the energetics of walking and running: Swinging the limbs is expensive. Science 303:80-83 Mitchell HH, Card LE, Hamilton TS (1931) A technical study of the growth of White Leghorn chickens. Illinois agricultural experiment station bulletin 367: Munns SL, Edwards A, Nicol S, Frappell PB (2015) Pregnancy limits lung function during exercise and depresses metabolic rate in the skink Tiliqua nigrolutea. J Exp Biol 218: Olsson M, Shine R, Bak- Olsson E (2000) Locomotor impairment of gravid lizards: is the burden physical or physiological? J Evolution Biol 13: Parker TH, Garant D (2005) Quantitative genetics of ontogeny of sexual dimorphism in red junglefowl (Gallus gallus). Heredity 95: Paxton H, Anthony NB, Corr SA, Hutchinson JR (2010) The effects of selective breeding on the architectural properties of the pelvic limb in 25

115 broiler chickens: a comparative study across modern and ancestral populations. J Anat 217: Paxton H, Daley MA, Corr SA, Hutchinson JR (2013) The gait dynamics of the modern broiler chicken: a cautionary tale of selective breeding. The Journal of experimental biology 216: doi: /jeb Paxton H, Tickle PG, Rankin JW, Codd JR, Hutchinson JR (2014) Anatomical and biomechanical traits of broiler chickens across ontogeny. Part II. Body segment inertial properties and muscle architecture of the pelvic limb. Peerj 2 Rauw WM, Kanis E, Noordhuizen- Stassen EN, Grommers FJ (1998) Undesirable side effects of selection for high production efficiency in farm animals: a review. Livest Prod Sci 56:15-33 Reiter K, Bessei W (1997) Gait analysis in laying hens and broilers with and without leg disorders. Equine Vet J Suppl: Ricklefs RE (1979a) Adaptation, Constraint, and Compromise in Avian Postnatal- Development. Biol Rev Camb Philos Soc 54: Ricklefs RE (1979b) Patterns of Growth in Birds.5. Comparative- Study of Development in the Starling, Common Tern, and Japanese Quail. Auk 96:10-30 Ricklefs RE, Shea RE, Choi IH (1994) Inverse Relationship between Functional Maturity and Exponential- Growth Rate of Avian Skeletal- Muscle - a Constraint on Evolutionary Response. Evolution 48: Rose KA, Nudds RL, Butler PJ, Codd JR (2015a) Sex differences in gait utilization and energy metabolism during terrestrial locomotion in two varieties of chicken (Gallus gallus domesticus) selected for different body size. Biol Open 4: doi: /bio Rose KA, Nudds RL, Codd JR (2015b) Intraspecific scaling of the minimum metabolic cost of transport in leghorn chickens (Gallus gallus domesticus): links with limb kinematics, morphometrics and posture. The Journal of experimental biology 218: doi: /jeb Schutz K, Kerje S, Carlborg O, Jacobsson L, Andersson L, Jensen P (2002) QTL analysis of a red junglefowl x white leghorn intercross reveals trade- off in resource allocation between behavior and production traits. Behav Genet 32: Shine R (2003) Effects of pregnancy on locomotor performance: an experimental study on lizards. Oecologia 136: Srinivasan M (2011) Fifteen observations on the structure of energy- minimizing gaits in many simple biped models. J R Soc Interface 8:

116 Srinivasan M, Ruina A (2006) Computer optimization of a minimal biped model discovers walking and running. Nature 439:72-75 Stearns SC (1989) Trade- Offs in Life- History Evolution. Funct Ecol 3: Talaty PN, Katanbaf MN, Hester PY (2010) Bone mineralization in male commercial broilers and its relationship to gait score. Poultry Sci 89: doi: /ps Team RDC (2011) R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing, Tickle PG, Lean SC, Rose KA, Wadugodapitiya AP, Codd JR (2013) The influence of load carrying on the energetics and kinematics of terrestrial locomotion in a diving bird. Biol Open 2: doi: /bio Tickle PG, Paxton H, Rankin JW, Hutchinson JR, Codd JR (2014) Anatomical and biomechanical traits of broiler chickens across ontogeny. Part I. Anatomy of the musculoskeletal respiratory apparatus and changes in organ size. Peerj 2 Tickle PG, Richardson MF, Codd JR (2010) Load carrying during locomotion in the barnacle goose (Branta leucopsis): The effect of load placement and size. Comp Biochem Physiol Part A Mol Integr Physiol 156: doi: /j.cbpa Usherwood JR (2013) Constraints on muscle performance provide a novel explanation for the scaling of posture in terrestrial animals. Biol Lett 9 Wassersug RJ, Sperry DG (1977) Relationship of Locomotion to Differential Predation on Pseudacris- Triseriata (Anura Hylidae). Ecology 58: Weiner J (1992) Physiological Limits to Sustainable Energy Budgets in Birds and Mammals - Ecological Implications. Trends Ecol Evol 7: Whitehead CC (2004) Overview of bone biology in the egg- laying hen. Poultry Sci 83:

117 6. Variety, sex and ontogenetic differences in the pelvic limb muscle architectural properties of leghorn chickens (Gallus gallus domesticus) and their links with locomotor performance This chapter is a draft of an article submitted to Journal of Anatomy Rose, K. A., Nudds, R. L, and Codd, J. R. (in review). Variety, sex and ontogenetic differences in the pelvic limb muscle architectural properties of leghorn chickens (Gallus gallus domesticus) and their links with locomotor performance 52

118 Variety, sex and ontogenetic differences in the pelvic limb muscle architectural properties of leghorn chickens (Gallus gallus domesticus) and their links with locomotor performance Kayleigh A. Rose, Robert L. Nudds and Jonathan R. Codd* Faculty of Life Sciences, University of Manchester, Manchester, M139PT *Address for reprints and other correspondence: Dr. Jonathan Codd Faculty of Life Sciences, University of Manchester, Manchester M13 9PT, UK Tel: +44(0)

119 ABSTRACT Leghorn (layer) chickens (Gallus gallus domesticus) differ in locomotor morphology and performance due to artificial selection for standard (large) and bantam (small) varieties, sexual dimorphisms and ontogenetic stage. Here, the hind limb skeletal muscle architectural properties of mature and juvenile standard breeds and mature bantams are compared and linked to measures of locomotor performance. Mature males possessed greater relative muscle physiological crosssectional areas than their conspecific females, indicative of greater force generating capacity, and in line with their greater maximum sustainable speeds compared to females. Furthermore, some of the relative-fascicle lengths of the pennate muscles were greater in mature males than in mature females, which may permit greater muscle contractibility. Immature standard leghorns, however, did not share the same dimorphisms as their mature forms. The differences in architectural properties between immature and mature standard males indicate that with the onset of male sexual maturity, concomitant with increasing muscle mass in males, the relative fascicle lengths of pennate muscles and the relative physiological cross sectional areas of the parallel-fibred muscles also increase. The age-related differences in standard breed male muscle architecture are linked to the presence and absence of sex differences in maximum aerobic speeds. Males of bantam and standard varieties shared similar muscle proportions (% body mass), but exhibited intrinsic muscle differences with a tendency for greater forcegenerating capabilities in bantams and greater contractile (work producing) capabilities in standards. The metabolic costs associated with the longer fascicle lengths, together with more crouched limbs in standard than in bantam males may explain the lack of allometry in the minimum metabolic cost of transport between these birds of different size. KEY WORDS: artificial selection, sexual dimorphism, locomotion, muscle architecture, chicken, ontogeny, size 2

120 INTRODUCTION The domestic chicken (Gallus gallus domesticus) has a broad range in body size and anatomical proportions associated with sexual dimorphism, ontogenetic stage and breed (Tickle et al. 2014; Paxton et al. 2014; Paxton et al. 2010; Mitchell et al. 1931). Both sex and interbreed variation in aspects of locomotor performance (mechanics, kinematics, metabolic costs and maximum aerobic speeds) have also been identified in the chicken (Corr et al. 1998; Corr et al. 2003; Paxton et al. 2013; Rose et al. 2015a; Rose et al. 2015b). Skeletal muscle is integral to all aspects of locomotor performance. Quantification and comparison of muscle architecture among chicken breeds may reveal links between form and function following artificial selection. Most research in this area has focused on the limb muscle architectural properties of broiler chickens artificially selected for increased growth rates. The skeletal muscle architectural properties of broiler breeders and ancestral jungle fowl including at differing ontogenetic stages, have been comprehensively described (Paxton et al. 2010; Paxton et al. 2014). Intraand inter-breed variations in the muscle architecture of chickens selected for their egg laying productivity, however, have yet to be investigated. The architecture of skeletal muscle refers to the arrangement and geometric properties of its fascicles. Muscle mass (M m, proportional to fascicle number), fascicle length (L f ) and fascicle angle relative to the axis of force generation (pennation angle, θ) are the principle determinants of the overall function of muscle (Burkholder et al. 1994; Lieber and Friden 2000; Eng et al. 2008). L f influences the distance over which muscle can contract to do work (i.e. the working range, work (J) = change in length (m) x force (N)) and therefore, the contractile ability (excursion and velocity) of the muscle. A muscle s effective physiological cross-sectional area (PCSA, calculated using muscle volume, L f and θ) is representative of its total area of muscle fibres, which is directly proportional to the maximum force that a muscle can produce and is therefore indicative of force-generating capacity (Gans and Bock 1965; Powell et al. 1984). For a given mass of muscle, L f and PCSA are inversely proportional (i.e. a trade-off exists between the two), meaning that muscles are generally specialised towards either 3

121 greater working ranges or force production, according to their location on the limb and principal function (Lieber and Friden 2000). A general pattern identified in vertebrate species is for the specialisation of many proximal limb muscles for doing work (with long parallel fibred muscles), whilst distal muscles (which are pennate, with tendons) are specialised for generating large forces and storing and recovering mechanical energy as an energy saving mechanism. There is evidence, however, that distal pennate muscles also perform considerable amounts of positive mechanical work (Ellerby et al 2005). Powerful muscles able to produce large forces over a large working range (by doing large amounts of work) generally have greater cross-sections of fibres, are associated with greater metabolic costs and consequently are less numerous. Leghorn (layer) chickens (Gallus gallus domesticus) have a broad range in body size and anatomical proportions associated with sexual dimorphism (Rose et al. 2015a), ontogenetic stage (Mitchell et al. 1931) and variety (Rose et al. 2015b). Female leghorn chickens (layer strains) are artificially selected for egg production efficiency (number and quantity) and reach early sexual maturation and are permanently gravid, towards which they bias resource allocation (Schutz et al. 2002). Leghorns also exhibit male-biased sexual size dimorphism in body mass (M b ) and males have greater muscle masses as relative proportions of M b than females (Mitchell et al. 1931) in common with many other vertebrate species (Lislevand et al. 2009; Lourdais et al. 2006; Polak and Frynta 2009). Sex differences in skeletal muscle mass are common to species in which the males use locomotion more or to greater intensity than females (e.g. for obtaining mates) (Zhang et al. 2014; Welch and Altshuler 2009). In leghorns, males also fight involving fast and powerful jumping to compete for females (Guhl et al. 1945). Given the disparity in somatic tissue, reproductive tissue and utilisation of locomotion between the sexes in leghorns, it seems reasonable to expect that they will exhibit sexual dimorphism in the specialisation of their muscle architecture (e.g. powerful muscles for fighting in males and economical muscles for load bearing in females). Sex-specific muscle specialisation has been investigated at the microscopic level (e.g. differences in fibre-type proportions) in some species 4

122 (Welch and Altshuler 2009; Henry and Burke 1998). No studies to date, however, have determined whether sex influences muscle architectural properties. Early in the post-hatch growth trajectories of leghorns (prior to sexual maturity), the sexes are more similar in form (Mitchell et al. 1931). At the onset of sexual maturity, however, a rise in gonadal hormones triggers increased M m in males. At a similar time, the onset of female gravidity also occurs under hormonal control. Comparison of immature and mature leghorns may provide insight into whether the onset of sexual maturity leads to sexual dimorphisms in muscle architecture. A number of intraspecific differences in whole-animal performance (locomotor mechanics, kinematics, energetics and maximum speeds) have been identified in leghorn chickens in association with these morphological differences; however, mechanisms underlying them are unclear. Mature male leghorns of standard breed (L, large) and bantam (B, small) varieties, for example, are geometrically similar in their axial and appendicular skeletons and share similar mass-specific minimum metabolic costs of transport (CoT min ) and limb kinematic patterns (Rose et al. 2015b), contrary to the negative allomery of CoT min across species (Taylor et al. 1982; Rubenson et al. 2007; Roberts et al. 1998). L, however, have a more crouched limb posture and incur a greater metabolic cost per stride (Rose et al. 2015b). In both bantam and standard breed leghorns, greater incremental mass-specific metabolic costs of locomotion are found in males relative to their females (L and B ) (Rose et al. 2015a). The same sex difference in locomotor energy metabolism was identified in immature standard breeds (JuvL and JuvL ) (K. A. Rose, unpublished). Furthermore, L and B are able to sustain greater maximum aerobic speeds than their conspecific females (Rose et al. 2015a), which may be explained simply by the rise in male muscle mass at sexual maturity (Mitchell et al. 1931). It is possible, however, that all of these four previously observed differences in locomotor parameters may also be linked to muscle architectural differences. The aim of this study was to investigate the skeletal muscle architectural properties of some of the major pelvic limb muscles in mature male (L ) and female (L,) and juvenile (JuvL and JuvL ) standard breed and mature bantams 5

123 (B and B ) and to link any differences with previously measured parameters of locomotor performance from the same individual birds. The following hypotheses were tested: (1) that males would possess greater relative muscle PCSA and L f than females (indicative of superior but metabolically costly muscle function relative to females, with superior economy of force generation); (2) that immature birds would possess similar sex-specific architecture to mature birds, linked to their similar metabolic patterns (therefore, differing only in muscle volume between ontogenetic stages); and (3) that B and L would have relatively shorter and longer L f, linked to the lack of intraspecific scaling in CoT min between them despite their different body sizes. MATERIALS AND METHODS Anatomy All birds (B : N=5, B : N=5, L : N=5, L : N=5, JuvL : N=4 and JuvL : N=5) in the present study were previously used in locomotor experiments (Rose et al. 2015a; Rose et al. 2015b), following which they were euthanized and frozen (-20 C). Frozen carcasses of the birds were thawed for 24 h and reweighed prior to dissection. Thirteen major pelvic limb muscles (Fig. 1, Table 1) were identified based on (Paxton et al. 2010). The muscles were removed individually from the right pelvic limb and separated from fascia and tendon before being weighed (±0.001g) upon electronic scales (Denver Instrumental, Germany). Tissues were kept hydrated with 0.75% saline solution. Fascicle lengths (L f, mm) were measured using a measuring rule (± 1 mm), and pennation angles (θ) with a protractor (± 1 ) from the mature birds. Photographs (0.63x magnification) were taken of the muscles of juvenile birds using a microscope (LEICA MZ9.5). ImageJ (Schneider et al. 2012) was used to measure angles and lengths from the photographs. A minimum of 5 L f and θ were recorded from each muscle and an average was calculated. Female reproductive systems (developing eggs, ovarian tissue and oviduct) were also dissected from B and L and weighed. 6

124 IL IC ITC FMT PIFM PIFL ILFB GL GM FCLP FL FCM TCT/F Figure 1 Anatomical diagram of the major muscles of the chicken pelvic limb. Muscle abbreviations are defined in Table 1 7

125 Table 1 List of the major muscles of the chicken pelvic limb Muscle Abbreviation Location Functional group Active during Architecture a M. iliotibialis cranialis IC Proximal Hip flexor, knee extensor Swing Yes M. iliotibialis lateralis (pre and post acetabularis) IL Proximal Hip abductor, hip flexor, knee extensor Swing and stance Yes M. iliofibularis ILFB Proximal Hip extensor, hip abductor, knee flexor Swing and stance Yes M. flexor cruris lateralis pars pelvica FCLP Proximal Hip extensor, hip abductor, knee flexor Stance Yes M. flexor cruris medialis FCM Proximal Hip extensor, hip abductor, knee flexor Stance Yes M. iliotrochantericus caudalis ITC Proximal Hip flexor, medial rotator Swing and stance Yes M. femerotibialis medialis FMT Proximal Knee extensor Swing and stance X M. pubioischiofemoralis pars lateralis PIFL Proximal Hip extensor Stance Yes M. pubioischiofemoralis pars medialis PIFM Proximal Hip extensor Stance Yes M. gastrocnemius pars lateralis GL Distal Knee flexor, ankle extensor Stance Yes M. gastrocnemius pars medialis GM Distal Knee flexor, ankle extensor Stance Yes M. fibularis lateralis FL Distal Ankle extensor, digit flexor Stance Yes M. tibialis cranialis caput tibiale and femorale TCT/F Distal Knee extensor, ankle flexor Swing X Numbers correspond to figure 1 a No architecture data is included for FMT and TCT/F as measurements were not Functional group information was taken from (Paxton et al. 2010) Swing phase during which the muscles are active is taken from (Marsh et al. 2004; Jacobson and Hollyday 1982) 8

126 Muscle Architecture For each of the pelvic limb muscles, PCSA was calculated using the same equation as in previous studies (Sacks and Roy 1982; Powell et al. 1984; Paxton et al. 2010): PCSA = M m cos θ/ ρl f (1) where M m = muscle mass and ρ= muscle density. The standard density value for mammalian and avian muscle was used (1.06 g cm -3 ) (Mendez and Keys 1960; Paxton et al. 2010). Data were normalised for comparisons between groups by 1/3 negating size dependent effects. Since lengths are M b and areas M 2/3 b, L f was normalised using (Paxton et al. 2010): L f / M b 1/3 (2), and PCSA using (Paxton et al. 2010): PSCA/M b 2/3 (3) Statistical analyses All statistical analyses were performed using the car package version (Fox and Weisberg 2011) on R GUI 1.42 Leopard build 64-bit (Team 2011). Shapiro-wilk tests were performed on the standardised residuals of all models to ensure a normal distribution of the data. Non-normally distributed data was corrected using either a log(x), -1/(x) or (x) 2 transformation depending on the skew. All morphological measurements were compared using two-way analyses of variance (ANOVAs). Separate ANOVAs were conducted to determine variety and sex differences among standard and bantam leghorns, and to reveal any age and sex differences among standard leghorns. 9

127 RESULTS Anatomy The ratio of mature male to mature female M b was 1.26 and 1.43 in the bantam and standard varieties of leghorn, respectively (Table 2). The sum of thirteen hind limb muscle masses (Σ M m(limb) ) was greater in males than in females by and 2.25-fold in bantam and standard leghorns, respectively. Σ M m(limb) comprised 2% more of M b in B than in B, and 2.54% more of M b in L than in L (Table 2). By comparison, in the juvenile cohort of the standard variety, male M b exceeded that of females by only 10% (Table 2) and absolute Σ M m(limb) did not differ between the sexes (Table 2), but comprised ~0.5% more of M b in JuvL than in JuvL. Σ M m(limb) as a percentage M b was similar between the males of the two varieties (Table 2), but was 0.68% greater in B than in L. Reproductive mass (M rep ) was 93% greater in L than in B, whilst overall M b was 28% greater and Σ M m(limb) only 12% greater than that of B. Therefore, L have greater relative reproductive load to carry with relatively less muscle than B whilst males of the two varieties share similar relative Σ M m(limb). Muscle architecture The mean body-size normalised fascicle length (L f ) and effective physiological cross-sectional areas (PCSA) values of standard and bantam leghorns are summarised in Figure 2A, and those of juvenile and sexually mature standard leghorns in Figure 2B. The mean absolute muscle data for each of the bird groups are provided in supplementary Table 1. In agreement with findings in red jungle fowl and broiler chickens (Paxton et al. 2010), the pennate distal limb muscles (with greatest force generating capability), including GL, GM and FL, had shorter L f and greater PCSA than the parallel fibred proximal limb muscles (possessing greater capacity for fast contraction) (Fig 2A-B). Furthermore, the ITC, a proximal pennate muscle, shared L f and PCSA more similar to those muscles from the distal limb (Fig 2A-B). 10

128 Table 2 Mean (± s.e.m) anatomical masses of six groups of leghorn differing in variety, sex and ontogenetic stage Measurement B B L L JuvL JuvL M b (kg) 1.39 ± ± ± ± ± ± 0.03 Σ limb M m (g) a ± ± ± ± ± ± 1.22 Σ limb M m (%M b ) 7.39 ± ± ± ± ± ± 0.22 b M rep n/a ± 7.53 c n/a ± n/a n/a M rep (%M b ) n/a 8.40 ± 0.88 n/a ± 0.56 n/a n/a Symbols are for body mass (M b ), the sum of thirteen hind limb muscle masses (Σ limb M m ) and female reproductive (developing and calcified eggs and oviduct) mass (M rep ) a pelvic limb muscle is the sum of the masses of the thirteen muscles presented in Fig 1 and Table 1 for a single leg b reproductive mass was assumed negligible in mass in the context of constraints on terrestrial locomotion c B M rep (N=5) is calculated from measurements substituted for 4 individuals, by another 4 (same age and cohort and also used in locomotion experiments) as only eggs, but not oviducts were weighed in 4/5 individuals from which muscle data were taken. 11

129 a PCSA / M b 2/ GL GM FL GM ITC GM ITC GM GL GL IL ITC ITC FL FL FL GL IL IL ILFB ILFB IL FCLP PIFM PIFM ILFB FCLP IC ILFB FCLP PIFM PIFL FCM PIFL IC PIFM FCLP FCM IC IC PIFL PIFL FCM FCM Series1 Series2 Series3 Series L L B B b 4.0 Series1 JuvL 3.5 FL Series2 JuvL PCSA / M b 2/ FL ITC GL ITC FL GM GM GL GM GL ITC FL ITC GL GM IL IL MATURE L MALES IL L MATURE FEMALE S ILFB IL PIFM FCLP ILFB ILFB PIFM IC FCLP FCLP PIFM PIFM PIFL PIFL ILFB IC IC FCM FCM FCLP PIFL PIFL FCM FCM IC Fascicle length / M b 1/3 Figure 2 Normalised PCSA against normalised fascicle length for the pelvic limb muscles of leghorn chickens. (a) Male and female bantam and standard leghorns. (b) Male and female juvenile and sexually mature standard leghorns. Muscles that cluster in the top left corners of the graphs are the pennate muscles of the distal limb and those in the bottom right are the parallel fibred muscles of the proximal limb. Data points represent means for each bird group. 12

130 Variety and sex effects In the two varieties of leghorn, each absolute M m was significantly greater in males than in females (Table 3) by on average 2.27-fold (range: fold) in the standards and 1.80-fold in the bantams (range: fold). The majority of muscles were greater in M m in the standard compared to the bantam variety, with the exceptions of FL and IL in which the variety differences were not statistically significantly different (Table 3). Furthermore, variety x sex interactions in M m were identified in the PIFM, TCT/F, ITC, ILFB and FMT due to it being greater in mature males than mature females to a greater degree in the standard, compared to the bantam variety (Table 3). As expected, the %M b of each muscle (Fig 3; Table 3) was greater in mature males than in mature females in the two varieties (by % in the standards, % in the bantams). Correction for M b this way eliminated the variety differences in all muscles excluding the IC, FL, FMT (in which values were greater in the bantams compared to the standards) and PIFM (which was greater in the standards than in the bantams). The only muscle to exhibit a variety x sex interaction after size-correction was the PIFM as the sex difference in %M b was greater in the larger of the two varieties (Fig 3; Table 3). Therefore, differences in M b accounted for the majority of variety differences but none of the sex differences in absolute M m. In the two varieties absolute L f was greater in males than in females in each muscle excluding some of the proximal parallel fibred muscles (IL, ILFB, FCM, PIFM) in which it was similar between sexes (Table 3). Significant variety x sex interactions in L f were identified in the GM and FL in which the sex difference was markedly greater in the standard compared to the bantam variety. L f was greater in the standard compared to the bantam variety in each muscle (excluding the ITC, for which the difference was not statistically significant) (Table3). 13

131 Table 3 Results of the two-way ANOVAs testing for variety and sex differences in muscle architectural properties Measurement Final model M b variety (F 1,16 =52.92, P<0.001), sex (F 1,16 =48.60, P<0.001), variety x sex (F 1,16 =6.00, P=0.026), R 2 =0.85 log Σ limb M m variety (F 1,17 =59.71, P<0.001), sex (F 1,17 =54.86, P<0.001), R 2 =0.86 log Σ limb M m (%M b ) variety (F 1,17 =9.10, P=0.008), sex (F 1,17 =161.53, P<0.001), R 2 =0.90 M rep (one-way ANOVA) variety (F 1,8 =33.44, P<0.001), R 2 = 0.78 M rep (%M b ) (one-way ANOVA) variety (F 1,8 =8.74, P=0.018), R 2 = 0.46 M IC variety (F 1,17 =6.12, P=0.024), sex (F 1,17 =72.95, P<0.001), R 2 =0.80 M ILFB variety (F 1,16 =9.43, P=0.007), sex (F 1,16 =66.67, P<0.001), variety x sex (F 1,16 =5.41, P=0.033), R 2 =0.81 M FCLP variety (F 1,17 =5.88, P=0.027), sex (F 1,17 =22.37, P<0.001), R 2 =0.58 M PIFL variety (F 1,17 =20.45, P<0.001), sex (F 1,17 =78.92, P<0.001), R 2 =0.84 M PIFM variety (F 1,16 =41.60, P<0.001), sex (F 1,16 =75.51, P<0.001), variety x sex (F 1,16 =18.67, P<0.001), R 2 =0.87 M GL variety (F 1,17 =11.16, P=0.004), sex (F 1,17 =139.70, P<0.001), R 2 =0.89-1/(M GM ) variety (F 1,17 =16.93, P<0.001), sex (F 1,17 =118.79, P<0.001), R 2 =0.88 M FL variety (F 1,17 =2.34, P=0.144), sex (F 1,17 =66.28, P<0.001), R 2 =0.89 M IL variety (F 1,17 =4.38, P=0.051), sex (F 1,17 =50.38, P<0.001), R 2 =0.74 M TCTF variety (F 1,16 =12.88, P=0.002), sex (F 1,16 =26.78, P<0.001), variety x sex (F 1,16 =7.17, P=0.017), R 2 =0.70 M ITC variety (F 1,16 =36.00, P<0.001), sex (F 1,16 =146.82, P<0.001), variety x sex (F 1,16 =9.02, P=0.008), R 2 =0.91 M FCM variety (F 1,16 =9.94, P=0.006), sex (F 1,16 =38.37, P<0.001), R 2 =0.73 M FMT variety (F 1,16 =3.52, P=0.079), sex (F 1,16 =88.01, P<0.001), variety x sex (F 1,16 =8.03, P=0.012), R 2 =0.84 IC %M b variety (F 1,17 =5.03, P=0.038), sex (F 1,17 =69.46, P<0.001), R 2 =0.79 ILFB %M b variety (F 1,17 =0.65, P=0.431), sex (F 1,17 =88.92, P<0.001), R 2 =0.82 FCLP %M b variety (F 1,17 =0.31, P=0.583), sex (F 1,17 =13.81, P=0.002), R 2 =0.39 PIFL %M b variety (F 1,17 =0.13, P=0.727), sex (F 1,17 =39.24, P<0.001), R 2 =0.66 PIFM %M b variety (F 1,16 =18.56, P<0.001), sex (F 1,16 =74.24, P<0.001), variety x sex (F 1,16 =7.89, P=0.013), R 2 =0.84 GL %M b variety (F 1,17 =3.30, P=0.087), sex (F 1,17 =65.92, P<0.001), R 2 =0.78 log GM %M b variety (F 1,17 =0.09, P=0.767), sex (F 1,17 =41.04, P<0.001), R 2 =0.67 FL %M b variety (F 1,17 =18.43, P<0.001), sex (F 1,17 =115.72, P<0.001), R 2 =0.87 IL %M b variety (F 1,17 =4.08, P=0.059), sex (F 1,17 =72.73, P<0.001), R 2 =0.80 TCT/F %M b variety (F 1,17 =12.86, P=0.109), sex (F 1,17 =18.31, P<0.001), R 2 =0.50 ITC %M b variety (F 1,17 =0.11, P=0.749), sex (F 1,17 =40.49, P<0.001), R 2 =0.67 FCM %M b variety (F 1,16 =0.23, P=0.637), sex (F 1,16 =30.26, P<0.001), R 2 =0.62 FMT %M b variety (F 1,17 =13.06, P=0.002), sex (F 1,17 =39.44, P<0.001), R 2 =

132 Table 3 continued Results of the two-way ANOVAs testing for variety and sex differences in muscle architectural properties Measurement Final model L f IC variety (F 1,17 =5.11, P=0.037), sex (F 1,17 =7.82, P=0.012), R 2 =0.37 L f ILFB variety (F 1,17 =23.29, P<0.001), sex (F 1,17 =0.95, P=0.344), R 2 =0.54 L f FCLP variety (F 1,17 =55.16, P<0.001), sex (F 1,17 =6.18, P=0.024), R 2 =0.76 L f PIFL variety (F 1,17 =8.87, P=0.008), sex (F 1,17 =11.54, P=0.003), R 2 =0.49 L f PIFM variety (F 1,17 =27.62, P<0.001), sex (F 1,17 =0.55, P=0.469), R 2 =0.58 L f GL variety (F 1,17 =18.02, P<0.001), sex (F 1,17 =26.62, P<0.001), R 2 =0.69 L f GM variety (F 1,16 =9.72, P=0.007), sex (F 1,16 =19.23, P<0.001), variety x sex (F 1,16 =8.42, P=0.010), R 2 =0.64 log(l f FL ) variety (F 1,16 =0.81, P=0.381), sex (F 1,16 =12.63, P=0.002), variety x sex (F 1,16 =8.97, P=0.009), R 2 =0.51 L f IL variety (F 1,17 =9.24, P=0.007), sex (F 1,17 =3.77, P=0.069), R 2 =0.37 L f ITC variety (F 1,16 =2.58, P=0.128), sex (F 1,16 =19.86, P<0.001), R 2 =0.54 L f FCM variety (F 1,16 =16.09, P=0.001), sex (F 1,16 =0.18, P=0.680), R 2 =0.44 1/3 L f /M b IC variety (F 1,17 =0.01, P=0.937), sex (F 1,17 =0.65, P=0.431), R 2 = /3 L f /M b ILFB variety (F 1,17 =5.45, P=0.032), sex (F 1,17 =1.26, P=0.277), R 2 = /3 L f /M b FCLP variety (F 1,17 =23.15, P<0.001), sex (F 1,17 =0.00, P=0.954), R 2 = /3 L f /M b PIFL variety (F 1,17 =2.44, P=0.137), sex (F 1,17 =4.07, P=0.060), R 2 = /3 L f /M b PIFM variety (F 1,17 =16.30, P<0.001), sex (F 1,17 =0.16, P=0.691), R 2 = /3 L f /M b GL variety (F 1,17 =6.09, P=0.025), sex (F 1,17 =12.36, P=0.003), R 2 = /3 L f /M b GM variety (F 1,16 =3.62 P=0.075), sex (F 1,16 =10.58, P=0.005), variety x sex (F 1,16 =5.76, P=0.029), R 2 =0.47 1/3 L f /M b FL variety (F 1,16 =0.18 P=0.674), sex (F 1,16 =5.72, P=0.029), variety x sex (F 1,16 =6.08, P=0.025), R 2 =0.32 1/3 L f /M b IL variety (F 1,17 =2.39, P=0.141), sex (F 1,17 =0.24, P=0.628), R 2 = /3 L f /M b ITC variety (F 1,16 =0.44, P=0.518), sex (F 1,16 =10.80, P=0.005), R 2 =0.35 1/3 L f /M b FCM variety (F 1,16 =2.03, P=0.173), sex (F 1,16 =5.66, P=0.030), R 2 =

133 Table 3 continued Results of the two-way ANOVAs testing for variety and sex differences in muscle architectural properties Measurement Final model PSCA IC variety (F 1,17 =1.41, P=0.251), sex (F 1,17 =67.55, P<0.001), R 2 =0.78 PSCA ILFB variety (F 1,17 =0.76, P=0.396), sex (F 1,17 =58.45, P<0.001), R 2 =0.75 PSCA FCLP variety (F 1,17 =0.06, P=0.815), sex (F 1,17 =24.64, P<0.001), R 2 =0.54-1/PSCA PIFL variety (F 1,17 =1.77, P=0.201), sex (F 1,17 =15.91, P<0.001), R 2 =0.45 PSCA PIFM variety (F 1,17 =0.84, P=0.373), sex (F 1,17 =30.21, P<0.001), R 2 =0.60 PSCA GL variety (F 1,16 =0.15, P=0.701), sex (F 1,16 =25.90, P<0.001), R 2 =0.58 PSCA GM variety (F 1,17 =0.44, P=0.518), sex (F 1,17 =14.09, P=0.002), R 2 =0.40 PSCA FL variety (F 1,17 =0.09, P=0.762), sex (F 1,17 =14.37, P=0.001), R 2 =0.39 PSCA IL variety (F 1,17 =0.04, P=0.844), sex (F 1,17 =34.43, P<0.001), R 2 =0.63 PSCA ITC variety (F 1,16 =0.44, P=0.518), sex (F 1,16 =10.80, P=0.005), R 2 =0.35 PSCA FCM variety (F 1,16 =2.42, P=0.139), sex (F 1,16 =37.21, P<0.001), R 2 =0.68 2/3 logpsca/ M b IC variety (F 1,17 =4.35, P=0.053), sex (F 1,17 =46.77, P<0.001), R 2 =0.72 2/3 PSCA/ M b ILFB variety (F 1,17 =5.43, P=0.032), sex (F 1,17 =81.25, P<0.001), R 2 =0.82 2/3 PSCA/ M b FCLP variety (F 1,17 =12.12, P=0.003), sex (F 1,17 =14.75, P=0.001), R 2 =0.57 2/3 log PSCA/ M b PIFL breed (F 1,17 =4.51, P=0.511), sex (F 1,17 =4.27, P=0.054), R 2 =0.13 2/3 PSCA/ M b PIFM breed (F 1,17 =1.28, P=0.274), sex (F 1,17 =19.06, P<0.001), R 2 =0.49 2/3 PSCA/ M b GL breed (F 1,16 =14.46, P=0.002), sex (F 1,16 =13.79, P=0.002), R 2 =0.61 2/3 PSCA/ M b GM breed (F 1,16 =7.08 P=0.017), sex (F 1,16 =3.09, P=0.098), breed x sex (F 1,16 =6.03, P=0.026), R 2 =0.41 2/3 PSCA/ M b FL breed (F 1,17 =4.62, P=0.046), sex (F 1,17 =6.72, P=0.019), R 2 =0.33 2/3 PSCA/ M b IL breed (F 1,17 =6.33, P=0.022), sex (F 1,17 =24.47, P<0.001), R 2 =0.60 2/3 PSCA/ M b ITC breed (F 1,16 =3.10, P=0.099), sex (F 1,16 =1.50, P=0.239), R 2 =0.11 2/3 PSCA/ M b FCM breed (F 1,16 =0.23, P=0.641), sex (F 1,16 =41.06, P<0.001), R 2 =

134 B Series2 B Series1 S % body mass L Series4 L Series3 V S S S V S S S S V S S 0.4 * 0.2 S S 0.0 FCM PIFL PIFM IC TCT/F ITC FL ILFB GL FCLP FMT GM IL Pelvic limb muscle Figure 3 Variety and sex differences in the percentage body mass of the pelvic limb muscles. Bars represent means (± s.e.m). Significant influences of variety, sex and variety x sex interactions are denoted by V, S and *, respectively. Results of the two way ANOVAs conducted to test for variety and sex affects are in Table 3. Normalising each L f for differences in M b (Fig 4A; Table 3) eliminated the sex differences in all of the parallel fibred muscles of the proximal limb excluding the FCM, in which it was greater in females (the only muscle for which this trend was true). Sex differences in L f in the pennate muscles, however, were not accounted for by normalisation in the ITC nor GL in either variety (Fig 4A; Table 3). Furthermore, significant variety x sex interactions in normalised L f were identified in the GM and FL, since no sex differences were present in the bantams, but it was greater in L than in L. The normalised L f of the standard variety was greater than that of the bantams in the ILFB, FCLP, PIFM and GL (Fig 4A; Table 3). 17

135 a Fascicle length / M b 1/ V V S V S V S Series2 Series1 Series4 Series3 * * IC IL ILFB FCLP FCM PIFL PIFM ITC GL GM FL B B L L b PCSA / M b 2/ V S B Series2 B Series1 L Series4 Series3 L V S * V S S V S V S S S 0.0 IC IL ILFB FCLP FCM PIFL PIFM ITC GL GM FL Pelvic limb muscle Figure 4 Variety and sex differences in pelvic limb muscle architectural measurements. (a) Normalised fascicle length. (b) Normalised PCSA. Bars represent means (± s.e.m.). Significant influences of variety, sex and variety x sex interactions are denoted by V, S and *, respectively. Results of the two way ANOVAs conducted to test for variety and sex affects are in Table 3. 18

136 No variety differences were present in absolute PCSA of the muscles (Table 3). However, the PCSA of each muscle was significantly greater in males compared to females (Table 3). Normalised PCSA was greater in the bantams in all muscles excluding the IC, FCM, PIFL and PIFM (Fig 4B; Table 3), and was also significantly greater in males than in females in every muscle (excluding PIFL and ITC in which it was similar amongst the groups). Furthermore, a significant variety x sex interaction in normalised PCSA was present in the GM, which was greater in B than in all other groups (sex difference in bantams only) (Fig 4B; Table 3). Therefore, common to both varieties, a greater relative PCSA and hence force generating capacity was found in males, relative to females across a mixture of both parallel fibred and pennate muscles; however, there was a tendency in the pennate muscles for the dimorphism to be larger in the bantams, whilst either lacking, or reduced, in the standard breeds. Relative L f were similar between the sexes in the parallel fibred muscles indicative of a similar contractibility. There was a tendency, however, again in the pennate muscles, for some relatively longer L f in males than in females, but to a greater degree in the standards than the bantams. There was also a tendency for relatively longer L f in the parallel fibred muscles of standard breeds compared to bantams, common to both sexes. Overall, therefore, it seems that B and L specialise differentially towards force generation and contractibility, respectively, and the same applies to some, but fewer, of the muscles between females of the two varieties. 19

137 Age and sex effects A significant age x sex interaction was found in each M m (Table 4), due to a lack of sex difference in the juveniles compared to the large difference reported between L and L. These sexual dimorphisms were associated with greater M m in L than in all other groups, but no age related differences in M m in the females in any muscles except for the FCLP, which was heavier in L than in JuvL. A significant age x sex interaction was also present in the %M b of each muscle (Fig 5; Table 4). In the juvenile cohort, half of muscles (FCM, PIFL, PIFM, TCT/F, ITC, ILFB, GM) were of equal %M b within the two sexes, and the other half (IC, IL, FCLP, FMT, FL and GL) were a greater %M b in JuvL than in JuvL, compared to the greater %M b of all muscles in L than in L. % body mass Series1 Series2 JuvL JuvL LARGE L FEMALES LARGE L MALES * * * * * * * * * * * * * 0.0 FCM PIFL PIFM IC TCT/F ITC FL ILFB GL FCLP FMT GM IL Pelvic limb muscle Figure 5 Age and sex differences in the percentage body mass of the pelvic limb muscles.. Bars represent means (± s.e.m.). Significant age x sex interactions are denoted by an asterisk. Results of the two way ANOVAs conducted to test for age and sex affects are in Table 4. 20