Myology of the pelvic limb of the brown‐throated three‐toed sloth (Bradypus variegatus)

Abstract

Tree sloths rely on their limb flexors for bodyweight support and joint stability during suspensory locomotion and posture. This study aims to describe the myology of three‐toed sloths and identify limb muscle traits that indicate modification for suspensorial habit. The pelvic limbs of the brown‐throated three‐toed sloth (Bradypus variegatus) were dissected, muscle belly mass was recorded, and the structural arrangements of the muscles were documented and compared with the available myological accounts for sloths. Overall, the limb musculature is simplified by containing muscles with generally long and parallel fascicles. A number of specific and informative muscle traits are additionally observed in the pelvic limb of B. variegatus: well‐developed hip flexors and hip extensors each displaying several fused bellies; massive knee flexors; two heads of the m. adductor longus and m. gracilis; robust digital flexors and flexor tendons; m. tibialis cranialis muscle complex originating from the tibia and fibula and containing a modified m. extensor digitorum I longus; appreciable muscle mass devoted to ankle flexion and hindfoot supination; only m. extensor digitorum brevis acts to extend the digits. Collectively, the findings for tree sloths emphasize muscle mass and organization for suspensory support namely by the hip flexors, knee flexors, and limb adductors, for which the latter two groups may stabilize suspensory postures by exerting appreciable medially‐directed force on the substrate. Specializations in the distal limb are also apparent for sustained purchase of the substrate by forceful digital flexion coupled with strong ankle flexion and supination of the hind feet, which is permitted by the reorganization of several digital extensors. Moreover, the reduction or loss of other digital flexor and ab‐adductor muscles marks a dramatic simplification of the intrinsic foot musculature in B. variegatus, the extent to which varies across extant species of two‐ and three‐toed tree sloths and likely is related to substrate preference/use.

The pelvic limb of sloths has numerous myological traits related to suspensory function. Most evident is the investment of knee flexor musculature and modifications to flexor and extensor muscles in their distal limb and feet.

1. INTRODUCTION

Tree sloths are members of the order Pilosa and superorder Xenarthra. Extant xenarthrans represent a basal clade of placental mammals (anteaters, armadillos, and sloths) that share a number of cranial features, namely a septomaxilla, rudimentary teeth, and postcranial xenarthrous articulations that reinforce the lumbar region of their vertebral column (Vaughn et al., 2013). Among the members of this clade, the sloths are arguably the most distinctive in their structure and function. Sloths are well‐known for their deliberately slow movements and use of suspensorial modes of arboreal locomotion, and yet, despite the similarities in habits displayed by these taxa, the lineages of two‐toed (Family: Choloepodidae) and three‐toed (Family: Bradypodidae) sloths are distantly related (Delsuc et al., 2001; Greenwood et al., 2001; Pujos et al., 2007), diverging nearly 29 MYA (Delsuc et al., 2019). Thus, their overall shared form, physiology, and mutual employment of nearly obligatory suspensory habits is strong evidence of evolutionary convergence (Gaudin, 2004; Gaudin & McDonald, 2008).

The brown‐throated three‐toed sloth (Bradypus variegatus: Schinz, 1825) is found throughout Central and South America (Gardner, 2008) and has a greater northern range than Hoffman's two‐toed sloth (Choloepus hoffmanni), for which it overlaps considerably in geographical distribution (Hayssen, 2010, 2011). Each species shows modifications to hindlimb bones and muscles, as well as a joint structure that allow for suspensory postures and below‐branch locomotion (Jouffroy et al., 1962; Vassal et al., 1962). A suite of morphological features common in the pelvic limb of tree sloths include a rigid pelvis (i.e., presence of a synsacrum: sacro‐ishiadic joints), elongate limb bones, unfused tibia/fibula, and three syndactylous digits on the hind feet, each with long, recurved claws (displayed on both sets of feet) (Grand, 1978; Mendel, 1981a, 1981b, 1985a, 1985b). These features for suspension were secondarily adapted in tree sloths as the basal stock of xenarthrans displayed a strong selection for fossorial habits (Nyakatura, 2012). However, a number of exceptional traits distinguish between three‐toed and two‐toed forms. For example, B. variegatus has disproportionate limb lengths (i.e., hindlimbs are shorter than the forelimbs: Wislocki, 1928), contains 8–9 cervical vertebrae for neck mobility, and retains hairy volar pads on both sets of feet (Mendel, 1985a), which is direct opposition to having limb pairs of similar length, fewer than seven cervical vertebrae, and bare (leathery) volar pads on the feet in Choloepus (Mendel, 1985b). Three‐toed sloths additionally have lower body mass than two‐toed sloths. Remarkably, the total (relative) skeletal muscle mass in Bradypus is reported to account for only ~24% of their body mass, while their thick skin accounts for nearly 16% and bone 13% (Grand, 1978). Lower body mass is associated with arboreal habits (Cartmill, 1985), especially in high‐canopy dwellers (Rupert et al., 2014), but it is not completely understood how reduced low muscle mass is related to specialization for suspensory locomotion in sloths, and this may further seem counterintuitive to the flexor muscle strength that is required from the limbs as a whole for suspension (Fujiwara et al., 2011; Olson et al., 2018). Moreover, comparatively less is known about the physiology and behavior in species of the genus Bradypus (e.g., Cliffe et al., 2015, 2018; Pauli et al., 2016) than Choloepus.

The fundamental presence or absence of muscles and their relative mass distributions are informative about limb function (Ercoli et al., 2013; Fisher et al., 2008; Marshall et al., 2021b; Olson et al., 2016), especially that of sloth forelimbs (Diniz et al., 2018; Miller, 1935; Nyakatura & Fischer, 2011; Olson et al., 2018). Myological descriptions are also available for the pelvic limb of sloths, but it is much less studied in relation to limb function (Nyakatura, 2012). The first complete accounts of post‐cranial myology in three‐toed sloths (B. tridactylus: Meckel, 1828; Cuvier, 1835; Cuvier & Laurillard, 1849) provided fundamental illustrations of muscle topography and descriptions of the musculature. Macalister (1869) described the limb muscles, including the entire pelvic limb of B. tridactylus, whereas Humphry (1870) reviewed those details with his own dissection of the same species, in addition to describing and partially illustrating pelvic limb myology in C. didactylus, all while providing interpretive details on muscle homology. Pelvic limb muscle masses were reported for B. tridactylus (Mackintosh, 1875a) and C. didactylus (Mackintosh, 1875b) for previously identified muscles, while work by Windle and Parsons (1899) summarized all previous myological descriptions for sloths and systematically compared them across taxa once classified in the now‐defunct Edentata. Therefore, descriptions and illustrations for these rare xenarthrans that have not been dissected in over a century need to be expanded and standardized with modern anatomical terminology, as all existing accounts contain conflicting or inaccurate nomenclature, making resolution of some muscle identities and homologies with other mammalian taxa difficult. The available records also often represent observations from a single specimen and are limited to two species for which the age, sex, and body mass are unknown. Last, the historical monographs on sloth limb myology do not provide muscle maps and do little to place the observations into a functional context, much less one that corresponds with a modern understanding of functional morphology in suspensory taxa.

For the first time, we thoroughly describe the myology in the pelvic limb of B. variegatus and determine mass distributions of the major functional muscle groups. Our subsequent companion study will evaluate the architectural properties of the musculature. The pelvic limbs of sloths have been shown to play a role in braking and stability during quadrupedal suspensory locomotion (Granatosky & Schmitt, 2017) and may often be the main limbs by which sloths support their body mass (Goffart, 1971; Granatosky et al., 2018). This function would be performed specifically by strong limb flexors that serve as anti‐gravity muscles in suspensory taxa (Fujiwara et al., 2011; Gorvet et al., 2020; Jouffroy & Stern, 1990; Nyakatura, 2012; Nyakatura et al., 2010). Therefore, the hip and knee flexors are expected to be well‐developed for stability, while the form of the hip extensors should reflect a role in propulsion during suspension and/or vertical climbing. We further hypothesize modifications to the leg and pes regions of the pelvic limb that reflect the evolution of the ability to strongly purchase grip on arboreal substrates with the hind feet. In sum, the observations made in this study will refine the framework for which to evaluate biomechanical structure/function in the tensile limb system of sloths by providing complete illustrations of muscle topography and much‐needed muscle maps detailing the arrangements of the pelvic limb musculature in these distinctive suspensorial mammals.

2. METHODS

Eight three‐toed sloths (6 adults/sub‐adults: 3.77 ± 0.5 kg; and 2 juveniles: 1.06 ± 0.1 kg) were dissected for this study. Sloths either died of natural causes or were euthanized for reasons unrelated to conditions with their pelvic limbs, and after which, the animals were frozen post‐mortem and stored until observation. Specimens were allowed to thaw for 24−36 h at ~4°C prior to dissection and measurement. One limb was used for myological observations, muscle measurement, and harvesting of muscle tissue blocks for myosin fiber type analysis (see Spainhower et al., 2021). Following the completion of dissection, the limb bones were cleaned and preserved at −20°C, whereas the cadavers were either refrozen or disposed of by burning. This work was conducted at The Sloth Sanctuary in Penshurt‐Limon, Costa Rica in the Spring 2015 and 2017. All procedures complied with protocols approved by the Costa Rica Ministry of the Environment, Energy, and Technology (MINAE: R‐033‐2015 to R. Cliffe; MINAE: R‐008‐2017 to M. T. Butcher) and adhered to the legal requirements of the United States. Morphological data for each individual studied are presented in Table 1.

TABLE 1.

Morphometric data for study specimens of Bradypus variegatus

Individual	Sex	Age	Limb	Body mass (kg)	Femur length (cm)	Tibia length (cm)	Metatarsal III length (cm)	Calcaneus length (cm)	Claw 3 length (cm)
Bv1	M	Sub‐Adult	R	3.40	9.53	8.11	1.61	3.87	5.22
Bv2	F	Adult	L	4.00	9.99	8.73	1.74	3.89	5.84
Bv3	M	Juvenile	R	1.01	5.98	5.06	0.96	2.21	4.18
Bv4	M	Adult	R	4.50	9.38	10.2	1.32	3.85	6.13
Bv5	M	Adult	L	3.90	8.30	7.90	1.17	3.41	4.92
Bv6	F	Juvenile	R	1.10	6.87	6.41	0.61	2.32	4.60
Bv7	F	Sub‐Adult	R	3.20	9.38	9.10	1.17	3.54	5.80
Bv8	M	Sub‐Adult	R	3.60	9.93	9.50	1.32	3.85	5.48

Values are means from n = 3 measurements.

Muscle names, origins, and insertions for B. variegatus (Bv) dissection a priori followed those of Macalister (1869), Humphry (1870), and Mackintosh (1875a, 1875b). Updated muscle nomenclature closely followed the Nomina Anatomica Veterinaria (International Committee on Veterinary Gross Anatomical Nomenclature 2017) with a few exceptions. The names for the m. tibialis cranialis fibular heads (caput fibulare et tibiale and caput fibulare) were coined to designate modification of a fused muscle complex; m. biceps femoris femoral (and largest) head and ischial head are used to designate the two parts of the muscle; the two subdivisions of m. flexor digitorum profundus are named after those of Fisher et al. (2008) and Ercoli et al. (2013). Synonyms of numerous muscles (e.g., m. plantaris and m. peroneus longus) are provided in Section 3 to facilitate comparisons with previous works on both sloth and mammalian myology. Muscle names and abbreviations are presented in Table 2.

TABLE 2.

Hindlimb muscles per limb region, their abbreviations, actions, and fiber architecture for Bradypus variegatus

Muscle	Abbreviation	Action	Fiber architecture
Hip region
Iliopsoas a	ILPS	Hip flexion	Parallel/unipennate
Psoas minor	PMN	Hip flexion	Parallel
Tensor fascia latae	TFL	Hip flexion	Parallel
Gluteus superficialis	GLS	Hip flexion (main portion of belly); limb abduction	Parallel
Gluteus medius	GLM	Hip extension; limb abduction	Parallel
Gluteus profundus b	GLP	Hip extension; limb abduction	Parallel
Piriformis	PFM	Hip extension; limb abduction	Parallel
Gemelli	GEM	Femoral stabilization, lateral rotation	Parallel
Quadratus femoris c	QF	Femoral stabilization, lateral rotation
Obturatorius externus	OE	Femoral stabilization, lateral rotation	Parallel/bipennate
Obturatorius internus	OI	Femoral stabilization, lateral rotation	Parallel
Thigh region
Sartorius	SRT	Hip flexion; knee flexion	Parallel/unipennate
Gracilis	GRC	Knee flexion; limb adduction; crural fascia tension	Parallel
Pectineus	PCT	Limb adduction	Parallel/bipennate
Adductor
Brevis	ADDB	Limb adduction	Parallel
Longus d	ADDL		Parallel
Magnus	ADDM		Parallel
Semitendinosus	ST	Hip extension; knee flexion	Parallel
Semimembranosus	SM	Hip extension; knee flexion; leg fascia tension	Parallel
Biceps femoris—ischial	BFI	Hip extension; knee flexion; leg fascia tension	Parallel
Biceps femoris—femoral	BFF	Knee flexion; leg fascia tension	Parallel
Quadriceps femoris
Rectus femoris	RF	Knee extension; hip flexion; femoral stabilization	Parallel
Vastus intermedius e	VI	Knee extension	Parallel
Vastus lateralis	VL	Knee extension	Unipennate
Vastus medialis	VM	Knee extension	Unipennate
Leg region
Gastrocnemius—medial	MG	Ankle extension; knee flexion	Parallel
Gastrocnemius—lateral	LG	Ankle extension; knee flexion	Parallel/unipennate
Soleus	SOL	Ankle extension	Parallel/unipennate
Popliteus	POP	Knee flexion; leg lateral rotation	Unipennate
Flexor digitorum superficialis	FDS	Digital flexion; knee flexion	Unipennate/bipennate
Flexor digitorum profundus—lateralis/medialis	FDP‐L‐M	Digital flexion	Unipennate
Tibialis caudalis	TCD	Supination (inversion) of hindfoot; ankle extension	Parallel
Fibularis longus	FL	Pronation (eversion) of hindfoot; ankle flexion	Parallel
Fibularis brevis	FB	Pronation of hindfoot	Parallel
Extensor digitorum lateralis f	EDLA	Pronation of hindfoot	Parallel/bipennate
Fibularis quartus	FQ	Ankle extension; pronation of hindfoot	Parallel
Extensor digitorum longus g	EDLO	Ankle flexion	Parallel/unipennate
Tibialis cranialis—tibial h	TCT	Supination of hindfoot; ankle flexion	Parallel/unipennate
Tibialis cranialis—fibular et tibial h	TCFT	Supination of hindfoot; ankle flexion	Parallel
Tibialis cranialis—fibular i	TCF	Supination of hindfoot; ankle flexion	Parallel
Foot region
Extensor digitorum brevis	EDB	Extension digits II–IV	Parallel
Flexor digitorum brevis j	FDB	Flexion digits II, IV	Parallel
Quadratus plantae	QP	Flexion digits II–IV	Parallel
Abductor digiti V k	ABD5	Abduction digit IV	Parallel
Interossei l	IOS	Abduction digits II & IV; metatarsal stabilization; extension of digits II–IV	Bipennate

^a In two specimens, the m. psoas major was separated from m. iliopsoas once the muscle had been excised.

^b Found to be separate in one female specimen with greater amounts of inter‐muscular fat storage.

^c Found to be separate from m. obturatorius internus and m. obturatorius externus in one female specimen.

^d Observed as both a single muscle head and two separate heads (cranial and caudal) depending on the individual.

^e Muscle head was difficult to separate from other heads of the m. quadriceps femoris.

^f This muscle appears to be a separate division of fibers from m. fibularis longus. They inserted onto the base of metatarsals IV and V and pronated the hindfoot. It is now correctly identified as m. extensor digitorum lateralis (previously named m. peroneus quinti) than m. fibularis tertius. Digit V is absent (remnant of metatarsal V) and the attachment on digit IV is potentially modified.

^g Muscle extended to only the base of metatarsal III.

^h Muscle bellies fuse and have a broad tibio‐fibular origin. Digit I is absent and the attachments are modified by insertion onto the remnant of metatarsal I.

ⁱ Previously named m. extensor digiti I longus (is a modifed part of m tibialis cranialis in B. variegatus).

^j Muscle slips were observed, but not clearly described, and absent in most individuals. Passed to laterally toward digit IV in three individuals (Bv2, 3, 8) and medially toward digit II in only Bv2 and may be remnants of m. flexor digitorum brevis.

^k Muscle slips were observed, but not clearly described, and present in individuals Bv1, 2, 4, and 7. Passed to the lateral aspect metatarsals IV–V and may be remnants of m. abductor digiti V.

^l Only dorsal interossei were clearly observed and muscle mass was taken for only three individuals (Bv6, 7, 8).

Briefly, the pelvic limbs of sloths were skinned and systematically dissected (proximal‐to‐distal). All muscles were identified and their attachments, actions (estimated from their anatomical position and manual manipulation during dissection), and fiber orientation were recorded. Muscle fiber architecture (e.g., pennate vs. parallel fibers) was verified ex vivo by micro‐dissection. Muscles with fascicle angles <15° (relative to the axis of force production) were generally classified as parallel regardless of belly shape, whereas muscles with fascicle angles >15° were classified as either unipennate, bipennate, or multipennate depending on fiber orientation and the number of intramuscular divisions or tendinous inscriptions observed (Marshall et al., 2021b). Dissections were documented by photographs taken with an EOS Rebel T5i D‐SLR camera (Cannon, USA) and these images were used to create limb muscle maps and illustrations of the limb muscle topography. Muscles were periodically moistened with a saline solution to prevent desiccation during dissection. If dissections and measurements could not be completed in one day, the limb was wrapped in gauze soaked in saline and stored overnight in a refrigerator.

Following the removal of each muscle, wet muscle belly mass (to the nearest 0.1 g) was recorded using a digital balance (Model: Scout‐Pro; Ohaus, USA). All muscle belly masses (see Table S1) were summed for each individual and calculated as a percentage of body mass (BM). Values for total limb muscle mass (TLMM) were log‐transformed and regressed against log BM via Model I regressions, thus yielding the slope (scaling exponent) of the relationship to assess if TLMM in sloth scales in proportion to body mass (isometric prediction = BM^1.0) (Payne et al., 2006). Muscles were then categorized into major functional groups: femoral rotators /stabilizers, hip flexors/extensors, limb abductors/adductors, knee flexors/extensors, ankle flexors/extensors, digital flexors/extensors, and hindfoot supinators/pronators. The relative mass of each functional group was quantified as a percentage of TLMM.

3. RESULTS

The pelvic limb contains 45 muscles (counting multiple heads or parts as separate muscles) that were studied. The initial observations of the pelvic limb in B. variegatus were as follows: the musculature was overall dark by being rich in red pigmentation; the size of most of the muscle bellies appeared to be proportional to the length of the limb segment rather than hypertrophied; many limb muscles were simple in their form by containing generally long and parallel fascicles. Indeed, relatively few muscle bellies displayed pennate fascicles, and among the pennate‐fibered muscles, unipennate fascicles were observed most frequently, but this form was inconsistent across muscles and individuals. Muscle bellies with bipennate fascicles were observed in some individuals and were isolated to the distal limb, namely the flexors of the digits, whereas no muscle bellies had multipennate fascicles except for those intrinsic to the foot composed of muscle slips. All muscles are organized by limb region and details of their myology are described below. Muscle actions and observed fiber architecture are also reported in Table 2.

3.1. Pelvic limb myology

3.1.1. M. iliopsoas (ILPS)

The m. iliopsoas (formerly m. psoadiliacus: Macalister, 1869) is a largely fused muscle that originates from the iliac fossa on the ventral surface of the ilium (m. iliacus) (Figures 1A and 2B), and the body and transverse processes of lumbar vertebrae L1–L4 as well as the ventrolateral margins of the sacrum (m. psoas major). A distinct m. psoas major (formerly m. psoas magnus: Humphry, 1870) was not observed in situ in our specimens as it was previously, and the origins of each belly appear non‐segmented in accordance with Macalister (1869) and Windle and Parsons (1899). The m. iliopsoas is thus considered one muscle along its entire length (in adults and juveniles) and it inserts onto the lesser trochanter of the femur and additionally, has an extensive attachment to the medial surface of the femur that exceeds the length of mid‐shaft (Figure 3A,B). In two individuals (adult males), the m. iliopsoas could be divided into two bellies once the massive muscle was excised. A separate mass for the m. psoas major is reported in Table S1.

FIGURE 1.

Muscle maps of the ventral (A) and dorsal (B) pelvis for Bradypus variegatus. Muscle nomenclature abbreviations are given in Table 2

FIGURE 2.

Muscle maps of the caudal (A) and lateral (B) pelvis for Bradypus variegatus. Muscle nomenclature abbreviations are given in Table 2. Colored in light gray and unlabeled is the acetabulum

FIGURE 3.

Muscle maps of the cranial (A) and caudal (B) femur as well as the cranial (C) and caudal (D) tibia‐fibula for Bradypus variegatus. Muscle nomenclature abbreviations are given in Table 2. Encircled on the medial epicondyle of the tibia in panel B is the alternative insertion for m. semitendinosus

3.1.2. M. psoas minor (PMN)

The m. psoas minor (formerly m. psoas parvus: Macalister, 1869) was not clearly observed in our dissections of B. variegatus. A small muscle slip resembling the m. psoas minor was separated from the m. psoas major (or m. iliopsoas) in only the two juvenile specimens and one adult individual. In two other adults, the two bellies were separated ex vivo. Its origin extended from the lateral aspect of the body of the terminal thoracic vertebrae (and ribs) to that of L1 and it was inserted on either the iliopubic ramus (Bv3, 6) or most clearly on the iliopectineal margin of the ilium (Bv7) (Figures 1A and 2B). The presence of a small m. psoas minor was noted by Macalister (1869) in B. tridactylus, and its presence was suggested, but not described by Humphry (1870).

3.1.3. M. tensor fasciae latae (TFL)

The m. tensor fasciae latae (formerly m. tensor vaginae femoris: Macalister, 1869; m. tensor fasciae femoris: Windle & Parsons, 1899) has mainly a fleshy origin from a small area on the lateral aspect of the tuber coxae (Figures 2B and 4), and in one individual (Bv7), some attachment to inguinal ligament was also observed. Its origin is immediately lateral to that of the m. sartorius (Figure 1B), and the two muscles are distinct in their attachments either on or about the cranial ilium. It inserts via the fascia lata onto the lateral femur, just distal to the greater trochanter and in common with fibers of the m. gluteus superficialis (Figures 3A and 4). Though small, its muscle belly is clearly separable from the gluteal muscles. Cuvier (1835) incorrectly identified this superficial muscle as the m. gluteus minimus.

FIGURE 4.

Lateral view of the hindlimb musculature for Bradypus variegatus. Muscles: (a) m. iliopsoas; (b) m. tensor fasciae latae; (c) m. gluteus superficialis; (d) mm. gemelli; (e) m. rectus femoris; (f) m. vastus lateralis; (g) m. semimembranosus; (h) m. biceps femoris‐ischial head; (i) m. biceps femoris‐femoral head; (j) m. flexor digitorum superficialis (or m. plantaris); (k) m. gastrocnemius‐lateral head; (l) m. soleus; (m) m. extensor digitorum longus; (n) m. fibularis longus; (o) m. fibularis quartus; (p) m. quadratus plantae; (q) m. extensor digitorum brevis; (r) m. interosseous lateralis. Colored in black is the sciatic nerve. Note: all superficial fibers of the m. biceps femoris‐ischial head that inserts onto crural fascia of the lateral leg is not illustrated to show the proximal bellies of the m. flexor digitorum superficialis and m. soleus

3.1.4. M. gluteus superficialis (GLS)

The m. gluteus superficialis (formerly m. gluteus maximus: Macalister, 1869, Mackintosh, 1875a; m. gluteus magnus: Humphry, 1870; m. ectogluteus: Windle & Parsons, 1899) is a large, fan‐shaped muscle that has origins from the dorsocaudal region of the iliac crest and the lateral margin of the sacrum (Figures 1B and 2B). In two male individuals, a primary and more extensive origin from the sacrum was observed. It inserts onto the femur distal to the greater trochanter (Figures 3A and 4) and also onto the thin band of fascia lata along the lateral femoral shaft. Humphry (1870) described the muscle to have its “posterior‐most” fibers that were in close contact with the m. semitendinosus and m. adductor. In our dissections, the m. gluteus superficialis was in close relationship to the m. biceps femoris on the caudolateral (not caudomedial) aspect of the hip joint.

3.1.5. M. gluteofemoralis (m. caudofemoralis, m. abductor cruris cranialis)

The m. gluteofemoralis (also referred to m. femorococcygeus) was not observed in our dissections of B. variegatus. Cuvier and Laurillard (1849) identified a small m. gluteofemoralis, while it was indicated to be fused with GLS in B. tridactylus (Windle & Parsons, 1899).

3.1.6. M. gluteus medius (GLM)

The m. gluteus medius is a thick muscle that originates mainly from the entire wing of the ilium and the lateral margin of the sacrum (Figures 1B and 2B). In only one individual (Bv7), the belly origin did not extend to the iliac crest. It inserts onto the greater trochanter of the femur (Figure 3A). The m. gluteus medius lies deep to GLS, and the caudal (or distal) and deep aspect of its muscle belly is commonly fused with either or both the m. gluteus profundus and m. piriformis (Figure 5). This condition was originally described by Meckel (1828) and later verified by both Macalister (1869) and Humphry (1870).

FIGURE 5.

Illustration of the deep hip muscles for Bradypus variegatus. Muscles: (a) m. gluteus medius; (b) m. piriformis; (c) m. gemellus superior; (d) m. gemellus inferior; (e) m. quadratus femoris; (f) m. obturatorius externus

3.1.7. M. gluteus profundus (GLP)

The m. gluteus profundus (formerly m. gluteus minimus: Macalister, 1869) originates from the caudal end of the dorsal ilium close to the sciatic foramen and cranial to the acetabulum (Figures 1B and 2B). It inserts either in common with or immediately medial to m. gluteus medius onto the greater trochanter of the femur (Figure 3A). In seven of the eight individuals, the m. gluteus profundus was extensively fused with the deep belly of GLM. Only in one adult female (Bv2) could the two muscles be separated due to the presence of a thin layer of fascia and deposition of inter‐muscular fat similar to previous descriptions of Mackintosh (1875a) and Windle and Parsons (1899).

3.1.8. M. piriformis (PFM)

The m. piriformis (formerly m. pyriformis: Mackintosh, 1875a) was observed as a separate belly from the m. gluteus medius in some of our specimens of B. variegatus. It originates deep from the ventral border of the sacrum and membrane of the sciatic foramen (Figures 1A and 2B) and inserts onto the caudal aspect of the greater trochanter in common with GLM (Figure 3A). The presence of a separable m. piriformis was noted by Macalister (1869) and Mackintosh (1875a) in B. tridactylus, but it was observed by Humphry (1870) as being inseparable from GLM and GLP. Our dissections are generally in accordance with those of Humphry (1870) that these three muscles are closely associated (or fused) and share an insertion on the greater trochanter of the femur (Figure 5).

3.1.9. Mm. gemelli (GEM)

The mm. gemelli (m. gemellus superioris and m. gemellus inferioris) originate from the lateral surface of the ischial tuberosity and dorsal margin of the obturator foramen (Figure 2B). In one individual (Bv8), its origin appeared to be modified (i.e., more cranial) to the ischial border of the sciatic foramen. These two muscles originate deep to, but may fuse superficially with the m. quadratus femoris, and insert proximal and laterally relative to the m. quadratus femoris onto the base of the greater trochanter and distomedial trochanteric fossa (Figures 3B and 5). Separate muscle bellies (or heads) and insertion tendons for the two mm. gemelli were clearly distinguished in only the female juvenile specimen. Two similarly sized bellies of the mm. gemelli were observed by Macalister (1869), whereas a single, fused muscle was described by Humphry (1870).

3.1.10. M. quadratus femoris (QF)

The m. quadratus femoris originates from the lateral surface of the ischium and dorsocaudal margin of the obturator foramen near to the ischial ramus (Figure 2B). It inserts onto the caudal aspect of the greater trochanter (or just distal to the greater trochanter) and distal‐most margins of the trochanteric fossa (Figure 3B) via both fleshy and tendinous fibers. The presence of a small m. quadratus femoris was noted by Macalister (1869), but it was later determined to be a large muscle (Mackintosh, 1875a) or fused (or not observed) with the mm. gemelli (Humphry, 1870); muscles in close proximity that share a similar origin from the ischium. Our observations partially agree with these previous descriptions. The m. quadratus femoris is appreciable in size and lies superficial to the m. obturatorius externus. These two muscle bellies run in opposing directions at nearly a 90° angle to each other along the caudal, proximal femur, thus acting to stabilize the head of the femur in the acetabulum by a crisscross arrangement (Figure 5).

3.1.11. M. obturatorius externus (OE)

The m. obturatorius externus is a large rotator muscle that originates broadly from the caudolateral aspect of the pubis, pubic ramus, and obturator membrane (Figure 2B). The muscle appears to have two parts to its belly, with an internal part exiting the obturator membrane and passing ventrally toward pubic symphysis joining an external part to form the sole tendon of insertion. It inserts via prominent tendinous fibers onto the caudal aspect (base) of the greater trochanter and the trochanteric fossa (Figures 3B and 5). The m. obturatorius externus runs medial‐to‐lateral in its attachment to the proximal, caudal femur. Mackintosh (1875a) previously described two parts of the m. obturatorius externus, although this was not noted by Macalister (1869) in B. tridactylus.

3.1.12. M. obturatorius internus (OI)

The m. obturatorius internus is a small rotator muscle that partially lies deep to the m. obturatorius externus with origins from the obturator membrane and superficially from the ventral surface of the lateral pubis, near the midpoint of the iliopubic ramus. The former portion of its origin was observed to span the cranial‐to‐caudal (and deep ventral) borders of the obturator membrane (Figure 2B). It inserts primarily onto the trochanteric fossa with some additional fibers attached to the joint capsule of the hip (Figure 3B). The m. obturatorius internus was observed as a separate muscle in half of our dissections, whereas and it otherwise appeared completely fused and inserted with OE. Additionally, in three of the eight specimens, the slight superficial portion of its belly was observed to take origin from a more cranial location on the iliopubic ramus. The lack of a medial (or internal) origin from the pelvis for the m. obturatorius internus was originally noted by Meckel (1828) and later ascribed to the condition observed in B. tridactylus by Cuvier (1835) and Macalister (1869). The muscle was also described as a second belly of OE (Windle & Parsons, 1899) with origins that match well with our dissections.

3.1.13. M. sartorius (SRT)

The m. sartorius (formerly m. iliotibialis: Windle & Parsons, 1899) originates from the cranial region of the tuber coxae via an aponeurosis continuous with the m. obliquus externus abdominis and fibers arising from the inguinal ligament medially (Figures 1B and 2B). The single belly of the muscle lies superficial to the m. quadriceps femoris and passes (lateral‐to‐medial) across the cranial thigh to the medial aspect of the knee joint in a parallel arrangement with the m. iliopsoas. It inserts by two slips of fleshy fibers onto the distal femur, where one slip shares fibers with the distal belly of the m. vastus medialis (Figure 6), and the second slip that attaches to the medial aspect of the distal femoral shaft immediately cranial and distal to the insertion of m. adductor longus (Figure 3A). The m. sartorius also has a thin tendon that dives deep along the medial knee and inserts onto the medial condyle of the tibia (Figure 3C). Mackintosh (1875b) described two parts of the m. sartorius (primus and secondus) that are not clearly observed in B. variegatus despite it having a dual insertion on the femur and tibia consistent with observations of both Humphry (1870) and Mackintosh (1875a). Slips of muscle inserting onto the distal femur were first noted by Cuvier (1835), while the dual origins for m. sartorius, namely the abdominal aponeurosis and/or inguinal ligament, were described by Meckel (1828).

FIGURE 6.

Medial view of the hindlimb musculature for Bradypus variegatus. Muscles: (a) m. sartorius; (b) m. pectineus‐adductor brevis; (c) m. adductor longus‐cranial head; (d) m. adductor longus‐caudal head; (e) m. adductor magnus; (f) m. vastus medialis; (g) m. semitendinosus; (h) m. semimembranosus; (i) m. gracilis‐external part; (j) m. gastrocnemius‐medial head; (k) m. flexor digitorum (profundus) medialis; (l) m. tibialis cranialis; (m) m. quadratus plantae; (n) m. extensor digitorum brevis; (o) m. interosseous medius. Hatched regions indicate the origins of m. gracilis (belly for the internal part not illustrated). Colored in gray proximally is the femoral artery and in black distally is the tibial nerve

3.1.14. M. gracilis (GRC)

The m. gracilis is a broad, superficial muscle that has an extensive origin from the fascia of m. obliquus externus abdominis (along the linea ilio‐pectinea), iliopubic ramus, pubis, and pubic symphysis (Figures 1A, 2B and 6), as well as the inguinal ligament in some specimens. In other individuals (Bv7, 8), it specifically took origin from the entire length of the iliopubic ramus and pubic tubercle. It is effectively divided into two parts: One part (internal‐femorotibial) of the muscle forms a slight tendon that inserts close to the medial condyle of the tibia (proximal one‐fifth of the medial tibia) and another fleshy portion that attaches to the proximal one‐half of the medial tibia (Figure 3C), while the second part (external‐crural: the majority of the muscle) continues distally forming a musculotendinous sheath that is superficial to all muscles of the caudal leg (Figure 6). This sheath becomes more tendinous just proximal to the ankle joint and terminates with a thin layer of the crural fascia at the calcaneus (Figure 7B). The m. gracilis also merges with the fascial insertion of the m. biceps femoris ischial head (at approximately the mid‐shaft of the tibia) to weakly insert onto the fibula and crural fascia of the lateral leg (Figure 3D). The m. gracilis was previously described most consistently as having two parts (Humphry, 1870; Windle & Parsons, 1899) instead of a single belly (Macalister, 1869; Mackintosh, 1875a), and this form of functional compartmentalization with distinct muscle parts arising from either fascial or bony origins with an extensive crural fascia insertion along the caudal leg (Cuvier & Laurillard, 1849; Meckel, 1828) is largely confirmed by our dissections.

FIGURE 7.

Muscle maps of the dorsal (A) and plantar (B) pes for Bradypus variegatus. The medial and lateral most metatarsals are remnants of digits I and V, respectively. Muscle nomenclature abbreviations are given in Table 2

3.1.15. M. pectineus (PCT)

The m. pectineus originates from the pectineal ridge along the cranial aspect of the iliopubic ramus (Figures 1A and 2B). A portion of its origin extended to just caudal to the acetabulum in one individual (Bv7). It inserts along the entire length of the femoral mid‐shaft immediately medial to the belly of the m. vastus medialis (Figure 3A). The m. pectineus was previously described as having a superficial portion and a deep portion (Cuvier & Laurillard, 1849; Macalister, 1869). Our dissections are generally in accordance with Macalister (1869), but we identify the superficial muscle as the m. pectineus and the deep muscle as the m. adductor brevis. We note in five specimens that these two muscles are nearly inseparable. Also, the femoral artery lies deep to the m. pectineus and is closely associated with the deep belly portion (or the m. adductor brevis when present as a separate muscle) along its length as this large blood vessel passes to the deep medial aspect of the knee joint (Figure 6).

3.1.16. M. adductor

The adductor muscle group (formerly m. adductor femoris: Windle & Parsons, 1899) is most commonly composed of four bellies. The m. adductor brevis (ADDB) originates from the cranial (or proximal) one‐fourth of the iliopubic ramus and inserts distal to the lesser trochanter onto the deep, caudomedial aspect of the femur (as much as the proximal two‐thirds) (Figures 1A, 2B and 3B). When clearly present, this muscle passes along the medial thigh just deep to PCT (Figure 6). Mackintosh (1875a) previously described the origin of m. adductor brevis from the ischial tuberosity in B. tridactylus that is inconsistent with our dissections.

The m. adductor longus (ADDL) originates broadly from the iliopubic ramus and is most commonly observed to have two heads: cranial and caudal (Figure 6). The smaller, deep cranial head originates from the cranial one‐third of the iliopubic ramus, whereas the larger, superficial caudal head originates from caudal (or distal) one‐third of the iliopubic ramus (Figures 1A and 2B). Humphry (1870) previously described a narrow portion (i.e., cranial head) and a broad portion (i.e., caudal head) of the m. adductor longus that is consistent with our dissections. In specimens lacking a division of the muscle bellies, the m. adductor longus formed a fleshy insertion onto the distal femur, immediately caudal to the insertion of ADDB. In the presence of a divided muscle belly, the caudal head inserted slightly more distal and nearer to the medial femoral epicondyle (Figures 3B and 6). In one adult male (Bv8), the caudal head shared an insertion with the m. semitendinosus, whereas the m. adductor longus was attached to the abdominal fascia at its origin in the female juvenile (Bv6). The latter individual also showed that the combined PCT and ADDB were additionally fused with ADDL.

The m. adductor magnus (ADDM) (formerly m. adductor primus: Mackintosh, 1875a) originates from the caudal one‐fourth of the iliopubic ramus and pubic tubercle, and it inserts onto the middle one‐third of the caudal femoral shaft, extending distally to the insertion of the m. adductor longus caudal head (Figures 1A, 2B and 3B). The relatively shorter belly of m. adductor magnus is situated deep to ADDL (Figure 6), and it is chiefly the distal arrangement of these two muscles that provide the passage (i.e., adductor hiatus) for the femoral artery to the popliteal fossa. Both Humphry (1870) and Mackintosh (1875a) described the m. adductor magnus to originate from the ischial tuberosity that is equally inconsistent with our dissections as that indicated for ADDB.

3.1.17. M. semitendinosus (ST)

The m. semitendinosus originates either via a tendon or fleshy fibers from the dorsal‐most portion of the ischiopubic ramus to the ischial tuberosity (Figure 2A,B). In one juvenile and one adult, the muscle originated in common with the m. biceps femorisischial head from the ischial tuberosity. Its distinct tendon passes deep to the m. semimembranosus to insert onto the caudomedial aspect of the medial condyle of the tibia (Figure 3D). However, in two specimens, it was observed to share an insertion onto the medial femoral epicondyle with the caudal head of ADDL (Figure 3B), or the distal‐most portion of ADDM in another individual (Bv8). The muscle belly also lies partially deep to the m. semimembranosus and runs nearly in parallel with the adductor muscles (Figure 6). No portion of the muscle originated from caudal vertebrae unlike the specimen of Humphry (1870), and it was not fused with the m. semimembranosus from a shared origin as described by Macalister (1869).

3.1.18. M. semimembranosus (SM)

The m. semimembranosus arises from an extended region of the ischial tuberosity and ischiopubic ramus, and it is both tendinous and fleshy at its origin (Figure 2A). A portion of its origin from the dorsal aspect of ischial tuberosity is shared with the m. biceps femoris ischial head (Figure 2A,B). Distally, the muscle belly has a small tendon that inserts onto the medial condyle of the tibia, a portion of which is just proximal and superficial to the insertion of ST (Figure 3D), and this arrangement was observed in all specimens. The m. semimembranosus also has fibers that make a fleshy insertion onto the caudomedial aspect of the femur in common with ADDM (Figure 3B), and other (distal) fibers that form a moderately expansive insertion onto the layer of the crural fascia just deep to the m. gracilis along the caudal leg (Figure 6). The latter insertion may extend from the proximal one‐third to the mid‐shaft of the tibia and makes a fleshy‐to‐membranous attachment onto the caudomedial surface of the tibia. Multiple insertions could correspond with a division of the muscle belly, where two, albeit adhered, bellies were observed in one adult specimen. Humphry (1870) and Windle and Parsons (1899) also described femoral and tibial insertions for m. semimembranosus but did not observe any investment with the crural fascia in B. tridactylus.

3.1.19. M. biceps femoris

The m. biceps femoris has two heads of origin: ischial (BFI) and femoral (BFF). The ischial head originates primarily from the dorsocaudal margin of the ischial tuberosity and partially from a fleshy attachment to the ischium (Figure 2A,B). It may additionally originate in part from the ischiopubic ramus via a thin layer of fascia, and this variation was observed in three individuals. This muscle can also be referred to as the m. flexor cruris lateralis (Windle & Parsons, 1899). The femoral head is well‐developed and originates from the distal end of the greater trochanter (distal to the insertion of GLS into the fascia lata) to the proximal one‐half of the caudolateral femoral shaft (Figure 3B). Although this origin was the most common condition, the femoral head was also observed to have a secondary, shared origin with the ischial head from the ischiopubic ramus via a prominent tendon in some specimens. Note that the sciatic nerve exits the sciatic foramen and passes caudolateral to the greater trochanter (Figure 4), and it continues distally along the medial border of the femoral head muscle belly.

The two heads share some fibers proximally but are not fused and collectively they insert onto a region of the leg spanning the lateral condyle of the femur (partially) to the lateral aspect of the proximal one‐fourth to one‐half of the fibula (mainly) (Figure 4). Additionally, BFI inserts onto the lateral expanse of crural fascia (superficial layer) along with m. gracilis (external part) and may lack a direct attachment to the fibula as observed in at least two adult males, whereas the attachment of BFF is more restricted to the cadual‐to‐caudolateral fibular shaft and crural fascia (deep layer) of the leg, which lies immediately superficial to the m. soleus (Figure 3D). In three individuals, the bony insertion of the femoral head notably extended to the middle two‐thirds and/or distal one‐half of the fibula, or further to the proximal calcaneal tendon and via its crural fascia investment, whereas the bony insertion of the ischial head onto the fibula was always more proximal at the knee, including the head of the fibula. An expansive insertion for the two bellies along the caudolateral leg was previously described in B. tridactylus (Humphry, 1870; Mackintosh, 1875a) and not limited to a common insertion on the head of the fibula as observed by Macalister (1869).

3.1.20. M. abductor cruris caudalis (m. tenuissimus)

The m. abductor cruris caudalis was not observed in our dissections of B. variegatus.

3.1.21. M. quadriceps femoris

The quadriceps femoris (formerly m. quadriceps extensor cruris: Humphry, 1870) is composed of four muscle heads: m. rectus femoris (RF), m. vastus intermedius (VI), m. vastus lateralis (VL), and m. vastus medialis (VM). The m. rectus femoris originates from the ilium immediate to the ridge of the acetabulum (Figure 2B). The m. vastus intermedius originates in part from the femoral neck, and the entire length of the cranial femoral shaft deep to the m. rectus femoris (Figure 3A). These two muscle heads were generally not distinct in our specimens and were clearly separable in only two adult males. The m. vastus lateralis originates from the femoral neck (and partially with fibers from the hip joint capsule), the craniolateral surface of the greater trochanter, and the craniolateral femoral shaft (Figures 3A and 4). The m. vastus medialis originates near the base of the femoral neck and the craniomedial femoral shaft (Figure 3A). This muscle head is observed to share some fibers with VI near mid‐shaft of the femur and SRT closer to the knee joint, although fiber sharing with SRT was not observed in the juvenile individuals. Moreover, bellies of the VI, VL, and VM were most easily excised as a single muscle complex in our dissections as they were in those of Mackintosh (1875a).

The bellies of all four muscle heads become more tendinous distally (proximal to the knee joint) and merge to form a quadriceps tendon that inserts onto an elongated, ovular patella (Figure 3A). The patella is present within the quadriceps tendon as a sesamoid bone that is situated into a deep patellar groove on the distal, cranial femur. A middle patellar ligament continues from the patella distal to the knee joint to attach onto an indistinctive tibial tuberosity (or the proximocranial tibia) and fascia surrounding the knee joint.

3.1.22. M. triceps surae

The m. gastrocnemius has medial (MG) and lateral (LG) heads of origin. The medial head (caput medialis) (formerly internus: Mackintosh, 1875a) originates from the caudal aspect of the medial condyle of the femur (Figure 3B). The proximal fibers of MG were observed to attach to a thin tendon of the m. tibialis caudalis in three of our dissections. The lateral head (caput lateralis) (formerly externus: Mackintosh, 1875a) originates from the caudolateral aspect of the distal femoral shaft just proximal to the lateral femoral condyle. Specifically, the origin of LG is medial and slightly distal relative to that of the large m. flexor digitorum superficialis. Neither head of m. gastrocnemius contained sesamoid bones at their origins as indicated by Humphry (1870). Each muscle belly (MG + LG) is strap‐like, and they merge proximal to the ankle joint to briefly form a single muscle belly with a common tendon (calcaneal tendon), which inserts onto the medial surface of the dorsal projection of the calcaneus (Figure 7A). The two muscles combined have a V‐shaped appearance along the superficial, caudal aspect of the leg (Figure 8A).

FIGURE 8.

Illustrations of superficial (A) and deep (B) caudal leg musculature as well as cranial leg muscles (C) for Bradypus variegatus. Muscles: (a) m. flexor digitorum superficialis (or m. plantaris); (b) m. gastrocnemius‐lateral head; (c) m. gastrocnemius‐medial head; (d) m. soleus; (e) m. popliteus; (f) m. flexor digitorum (profundus) medialis; (g) m. flexor digitorum (profundus) lateralis; (h) m. tibialis caudalis; (i) m. extensor digitorum longus; (j) m. tibialis cranialis‐tibial head; (k) m. tibialis cranialis‐fibular et tibial head; (l) m. tibialis cranialis‐fibular head; (m) m. extensor digitorum brevis. Note: m. quadratus plantae is not illustrated to show details of the deep flexor tendons in the plantar aspect of the pes

The m. soleus (SOL) originates from the area approximating the proximal one‐half to two‐thirds of the caudal fibula (Figure 3D). In one individual (Bv8), its origin spanned the entire length of the fibula. It inserts via a short tendon onto the dorsomedial surface of the calcaneus, immediately distal (and deep) to the insertion of the m. gastrocnemius (Figure 7A). Additionally, its muscle fibers may insert directly onto a portion of calcaneal tendon arising from the distal belly of LG. The m. soleus lies immediately deep to LG but superficial to the lateral part of mm. flexores digitorum profundi (i.e., m. flexor digitorum lateralis) and the medial border of its belly shares an investment of fascia specifically with the lateral head of the deep flexor muscle (Figure 8A). The m. soleus appeared to be absent in only the male juvenile (Bv3). Previous descriptions differ regarding an absolute proximal (Macalister, 1869) versus a mid‐fibular (Cuvier, 1835; Mackintosh, 1875a) or distal fibular (Windle & Parsons, 1899) origin for the m. soleus.

3.1.23. M. popliteus (POP)

The m. popliteus originates from the caudal aspect of the lateral femoral epicondyle (Figure 3B). Its broad tendon of origin contains a sizable sesamoid bone having facets that interact with both the femoral and tibial condyles along the caudal knee joint as indicated by Mackintosh (1875a). The fibers of the m. popliteus run lateral‐to‐medial to insert via a tendinous attachment onto the caudal aspect of the medial condyle of the tibia (Figures 3D and 8B), extending to near mid‐shaft of the tibia in two adult specimens. In one of those individuals, we noted a poplitofibular ligament spanning the proximal one‐half of the fibula. Also, the muscle belly is observed to have several tendinous inscriptions as seen in another two adults.

3.1.24. M. flexor digitorum superficialis (FDS)

The m. flexor digitorum superficialis (also referred to as m. plantaris) is a bulbous muscle that originates from an area on the caudal aspect of the distal femoral shaft and lateral epicondyle of the femur, immediately proximal and lateral to the origin of LG (Figures 3B, 4 and 8A). Its muscle belly joins together with the massive and broad bellies of the mm. flexores digitorum profundi to form the digital flexor muscle complex of the caudal leg. Distally, the belly of m. flexor digitorum superficialis has an independent tendon that emerges and inserts into the thick common tendon of both the mm. flexor digitorum medialis and flexor digitorum lateralis. Combined, this flexor tendon passes into the plantar side of the foot caudal and medial to the medial malleolus. The FDS does not directly provide a tendon to any digits in B. variegatus unlike previous descriptions of it inserting specifically into the tendon of m. flexor digitorum medialis (Mackintosh, 1875a) and serving digits III and IV in B. tridactylus.

3.1.25. Mm. flexores digitorum profundi (FDP)

The mm. flexores digitorum profundi comprise the m. flexor digitorum medialis (also referred to as m. flexor digitorum longus), m. flexor digitorum lateralis (also referred to as m. flexor digiti I longus), and m. tibialis caudalis. The bellies of the former two muscles immediately join and are highly fused into one large muscle belly with no clearly discernable fascial planes of separation in any of our specimens, thus both bellies were most easily excised and massed as a single digital flexor muscle complex (FDP: Table 2). Collectively, it broadly originates from the caudal‐to‐caudomedial tibial shaft (FDP‐M), interosseous membrane (FDP‐M and FDP‐L), and caudal fibular head and shaft (FDP‐L) (Figure 3D). The proximal portion of the FDP lies deep to the m. popliteus and FDS. Distally, the medial muscle gives rise to a thick tendon that is joined by a smaller tendon from the lateral muscle to form a common tendon (Figure 8B). The tendon from FDS joins the common tendon medially at the ankle joint along the plane of digit II before passing into the plantar side of the pes; this and the belly of FDS can be heavily invested with fibers from both deep flexors. This common flexor tendon then divides into three robust tendons, each of which passes through one of three tendon (volar) tunnels on the plantar pes to insert onto the distal phalanx bones of digits II–IV (Figures 7B and 8B). Prior to entering the tunnels, each of the three flexor tendons merges with one of the heavy tendinous sheaths of the m. quadratus plantae.

This flexor muscle complex was previously identified as either the m. flexor digitorum longus (origin: tibia) and m. flexor hallucis longus (origin: fibula and interosseous membrane) (Macalister, 1869), or the combined m. flexor digitorum (Humphry, 1870). The muscle bellies and insertions are modified in B. variegatus, which have lost digits I and V (present are only the remnants of metatarsals I and V). For example, an origin from the craniolateral surface of the lateral condyle of the tibia and/or tibial tuberosity for the m. flexor digitorum lateralis was observed in one juvenile and one adult. Our observations of a common flexor tendon that crosses the ankle joint caudal to the medial malleolus and divides into three thick tendons are also in accordance with those of Macalister (1869). However, independent passage of tendons from each head of the digital flexor muscle complex to an assigned digit as previously described by Humphry (1870) and Macalister (1875a) was not noted in any specimen of B. variegatus.

The m. tibialis caudalis (TCD) (formerly m. tibialis posticus: Macalister, 1869) originates from the caudal surfaces of the shaft and the inner medial condyle of the tibia, with the condylar origin having a thin tendon that lies superficial to the m. popliteus (Figures 3D and 8B). In several individuals, an additional origin from the head of fibula just distal to tendon of POP is observed. The muscle belly was superficial to the origin of the m. flexor digitorum lateralis head of the FDP complex with fibers running from lateral‐to‐medial in those specimens. It inserts onto the plantar aspect of the base of the remnant of metatarsal I and/or entocuneiform via a long, thin tendon that passes in a deep groove (or tunnel) along the caudal aspect of the medial malleolus (Figure 7B). In one adult female (Bv2), the m. tibialis caudalis inserted onto the base of metatarsal II and the entocuneiform, while the male juvenile specimen (Bv3) showed a split muscle belly near its origin that merged distally into a single tendon of insertion onto metatarsal I. Some variation in the origin of this muscle was previously observed ranging from the proximal tibia and head of the fibula (Humphry, 1870) to the distal one‐half tibia shaft (Macalister, 1869), but it consistently had an insertion onto the entocuneiform of the plantar pes (Mackintosh, 1875a).

3.1.26. M. fibularis longus (FL)

The m. fibularis longus (also referred to as m. peroneus longus) has origins from the cranial surface of the lateral condyle of the femur (via a tendon in common with the m. extensor digitorum longus), head of the fibula, and proximal one‐half of the fibular shaft (Figures 3A,C and 4). Its attachments to the fibula are fleshy and associated both with two heads of the m. tibialis cranialis muscle complex and crural fascia of the lateral aspect of the leg. Also, it was fused proximally with the m. extensor digitorum longus in two adult male specimens, whereas its distal belly merges with (or gives rise to) what could be described as fibers of a m. fibularis tertius. This split belly form was observed in only the female juvenile and one adult female (Bv2). The m. fibularis longus has a thin tendon that passes at or slightly cranial to the lateral malleolus of the fibula to insert onto the remnant of metatarsal V, and/or base of metatarsal IV, as observed in two individuals (Figure 7A). Only Macalister (1869) previously described the m. fibularis longus as passing caudal to the lateral malleolus, while Cuvier and Laurillard (1849) did not illustrate the muscle, and Windle and Parsons (1899) indicated it was absent.

3.1.27. M. fibularis brevis (FB)

The m. fibularis brevis (also referred to as m. peroneus brevis) is a short muscle that has a fleshy origin from the distal one‐fourth to one‐half of the lateral fibular shaft and lateral malleolus of the fibula (Figure 3C). In one juvenile (Bv6), the muscle belly shared fibers with the m. fibularis longus at its origin, whereas in one adult individual (Bv8), no portion of its origin was from the fibular shaft. It inserts onto the lateral surface of the base of the remnant of metatarsal V (on the tuberosity) (Figure 7A) via an insertion tendon that passes cranial to the lateral malleolus and may be shared with that of FL. The myology of FL was mistaken for that of m. peroneus brevis by Cuvier (1835) and its presence was somewhat variable in other previously studied specimens (Windle & Parsons, 1899).

3.1.28. M. extensor digitorum lateralis (EDLA)

The m. extensor digitorum lateralis (formerly m. peroneus quinti: Macalister, 1869; m. peroneus quinti digiti: Windle & Parsons, 1899) is indistinct, arising as fibers from the caudolateral edge of the fibular shaft, with a few additional fibers taking origin from the fascia separating it from the m. fibularis longus (origin not shown). Its fibers run along the caudal aspect of the distal fibula and they insert onto metatarsal IV and the lateral tuberosity of the remnant of metatarsal V (Figure 7A). The muscle (or fibers representing the muscle) was clearly observed in only two individuals. A muscle belly representing the m. extensor digitorum lateralis was previously noted to insert onto metatarsal V in B. tridactylus (Macalister, 1869; Mackintosh, 1875a; Meckel, 1828).

Both Humphry (1870) and Windle and Parsons (1899) further described a muscle referred to as the m. peroneus tertius having a shared origin with fibers of the m. extensor digitorum longus and an insertion onto the dorsal surface of the base of metatarsal IV. These previous descriptions overlap with those for m. fibularis longus in B. variegatus, which was not clearly separable into two muscles with different tendons of insertion onto the bases of metatarsals IV and V, except in the female juvenile specimen. The m. peroneus tertius also was described in B. tridactylus by Mackintosh (1875a) and was indicated to be conjoined with an additional muscle slip (m. peroneo‐metatarsalis) spanning the lateral malleolus to the tuberosity of metatarsal V; however, no bellies with similar origins and insertions were identified in the specimen of Macalister (1869).

3.1.29. M. fibularis quartus (FQ)

The m. fibularis quartus (also referred to as m. peroneus quartus) is a small, triangular‐shaped muscle that is closely associated with the m. fibularis brevis (Figure 4). The two muscles share an origin from the distolateral fibula, and it inserts as a thin muscle slip onto the lateral aspect of the calcaneus (Figures 3C and 7A) and/or near the base of metatarsal V. The presence of the m. fibularis quartus as a differentiated belly was observed in one juvenile and three adult specimens, whereas it appeared to be either absent or indistinguishable from FB in the other four individuals. The muscle is also congruent with a previously named m. peroneo‐calcaneus, as it is associated with bellies of the fibularis (peroneal) muscle complex arising from the lateral fibula (Mackintosh, 1875a).

3.1.30. M. extensor digitorum longus (EDLO)

The m. extensor digitorum longus originates from the cranial aspect of the lateral tibial condyle and head of the fibula, as well as from the cranial surface of the lateral femoral condyle via short tendon (Figure 3A,C). In several individuals, there was an additional muscle slip that took origin from approximately the mid‐shaft of the fibula that shared fibers with the m. tibialis cranialis at its origin along the proximal one‐third of the craniolateral tibia. The m. extensor digitorum longus inserts via a short, thick tendon onto the dorsal surface of the proximal end of metatarsal III (Figures 7A and 8C). The tendon of insertion was bifurcated in one individual (Bv7) with an additional insertion on metatarsal II. No insertion onto the dorsal aponeurosis of the pes was observed. In general, the fibularis muscles and EDLO can have extensive overlapping fibers with their origins from the fibula. A sole femoral origin of EDLO was noted by Cuvier (1835) and variable insertions onto metatarsal II (Macalister, 1869), metatarsal III (Humphry, 1870), or metatarsal II and IV (Windle & Parsons, 1899) were previously described for B. tridactylus with no attachment to the digits.

3.1.31. M. tibialis cranialis

The m. tibialis cranialis (formerly m. tibialis anticus: Macalister, 1869) is an extensive muscle complex that is composed of one tibial head and two fibular heads. The tibial head (caput tibiale‐TCT) has origins from the medial condyle of the tibia, medial aspect of the tibial tuberosity (via a small tendon), and head of the fibula (Figure 3C), although an exclusive attachment to the head of the fibula was observed in the female juvenile (Bv6). Proximally, it shares fibers with EDLO where the origins of the two muscles overlap (Figure 8C). One fibular head (caput fibulare et tibiale‐TCFT) is rhomboidal in shape and has fibers that arise from distal one‐half to two‐thirds of the craniomedial fibula (Figure 3C). Its fibers span the interosseous membrane and merge with those of the tibial head near the mid‐belly of the latter muscle. In one individual (Bv7), the fibers from TCFT further attached to the craniomedial tibial shaft approximating one‐half the distal tibia to the medial malleolus. The combined muscle heads insert together onto the entire lateral margin of the remnant of metatarsal I via a single tendon (Figures 7A and 8C). The second fibular head (caput fibulare‐TCF) is dissimilar from the other in that it originates from the distal fibula, cranial aspect of the lateral malleolus, and distolateral tibia (Figure 3C). In only one individual (Bv8) was the tibial portion of its origin was not observed. The origin of this smaller fibular head is situated immediately medial (and deep) to that of the m. fibularis brevis and it appears to be the m. extensor digiti I longus now conjoined with the m. tibialis cranialis. Its fibers run lateral‐to‐medial across the ankle joint to insert onto the dorsomedial aspect of the base of metatarsal I and entocuneiform (Figures 7A and 8C). In one adult male (Bv8), this fibular head was subdivided into two parts with the proximal‐most portion being fused with TCFT.

The dual origin for the m. tibialis cranialis was initially noted by Meckel (1828) and this muscle was considered separate from the m. extensor digiti I longus, but the latter was described to insert onto either the remnant of metatarsal I (Cuvier & Laurillard, 1849) or the proximal phalanx of digit II (Macalister, 1869; Mackintosh, 1875a). Our dissections are generally in accordance with the origin and insertions for three heads of the m. tibialis cranialis as was previously described by Humphry (1870) and Windle and Parsons (1899).

3.1.32. M. extensor digiti I longus

The m. extensor digiti I longus (formerly m. extensor hallucis proprius: Macalister, 1869; m. extensor longus hallucis: Humphry, 1870; m. extensor hallucis: Mackintosh, 1875a) is not an independent belly with an extensor action in the pelvic limb of B. variegatus. It is merged with the two bellies of the m. tibialis cranialis and is now referred to as m. tibialis cranialis caput fibulare.

Several previous descriptions of B. tridactylus indicated a diminutive additional muscle arising from the distal‐most cranial aspect of the fibula (and/or interosseous membrane) and inserting onto the proximal phalanx of digit II (Macalister, 1869). This separate belly is absent in B. variegatus, or it is fused with the m. extensor digitorum brevis.

3.1.33. M. extensor digitorum brevis (EDB)

The m. extensor digitorum brevis is a broad muscle that originates from the dorsal surfaces of the tarsal bones (cuneiforms, cuboid, navicular) and proximal ends of the metatarsals (Figure 7A). An additional origin from the distolateral aspect of the calcaneus was observed in only one individual (Bv7). It inserts onto the distal end of the intermediate phalanx and/or base of the distal phalanx of digits II, III, and IV, joining with the attachments of the flexor tendons to form slight tendon expansions on each digit (Figure 7A). The muscle is closely associated with the dorsal mm. interossei that lie between the metatarsals (Figures 4 and 6) and it was difficult to clearly separate the two muscles (or groups) in some specimens. The m. extensor digitorum brevis is the only muscle of the pelvic limb that serves as a common extensor of the digits where the portions of its belly associated with digits II and IV are separated superficially by the insertion of the EDLO onto metatarsal III (Figure 8C). Macalister (1869) previously described an origin from the distal tibia and fibula along with an insertion onto only the proximal phalanx of digit II in B. tridactylus, which is identical to the description of the m. extensor digiti I longus by the same author.

3.1.34. M. flexor digitorum brevis

The m. flexor digitorum brevis (formerly m. flexor brevis digitorum: Humphry, 1870; m. flexor brevis: Mackintosh, 1875a) was not clearly observed in our dissections of B. variegatus. Muscle fibers arising from a tendon on the calcaneus and passing to lateral aspect of digit IV were observed in one individual (Bv8). Superficial slips of muscle from the calcaneus were also present in two other specimens (Bv2, 3) in the same lateral location, in addition to another slip passing to the medial aspect of digit II in only the adult female, but they appeared too damaged from the skinning of the hindfoot, and thus their exact attachments could not be determined. It is possible that these fibers represent remnants of cutaneous muscles previously described by Humphry (1870). However, a small m. flexor digitorum brevis was also noted in the specimens of Humphry (1870) and Windle and Parsons (1899). Specifically, it was described to originate from the calcaneus superficial to the origin of the m. quadratus plantae (Humphry, 1870), which is typically located deeper (i.e., second plantar layer) within the pes.

3.1.35. M. abductor digiti V

The m. abductor digiti V was not clearly observed in our dissections of B. variegatus. Muscle fibers arising from a proximal tendon along the plantar aspect of the pes and passing toward the lateral aspect of metatarsals IV–V were observed in four specimens (Bv1, 2, 4, 7). Neither Humphry (1870) nor Windle and Parsons (1899) described this muscle in B. tridactylus.

3.1.36. M. abductor digiti I

The m. abductor digiti I (formerly m. abductor interni digiti: Mackintosh, 1875b) was not observed in our dissections of B. variegatus. Windle and Parsons (1899) indicated this muscle was absent in B. tridactylus.

3.1.37. M. quadratus plantae (QP)

The m. quadratus plantae (formerly m. musculii accessorii: Macalister, 1869; m. accessories: Humphry, 1870; m. accessorius: Mackintosh, 1875a; m. accessories: Windle & Parsons, 1899) is a well‐developed muscle that shows an extensive origin along the medial, plantar, and lateral surfaces of the calcaneus (Figures 4, 6 and 7B). However, it originated from only the medial surface of the calcaneus in the female juvenile specimen. It is the main intrinsic muscle of the plantar aspect of the pes and is located superficial to the distal bellies and common flexor tendon of FDS and FDP (not shown). The muscle is most commonly compartmentalized into three divisions (i.e., muscle slips) distally, each with a heavy tendinous sheath that inserts into (or becomes continuous with) one of the three deep and robust digital flexor tendons. Specifically, the medial muscle slip of the m. quadratus plantae inserts on the flexor tendon to digit II, the plantar (central) slip to that of digit III, and the lateral slip to that of digit IV. Only one individual (Bv7) did not demonstrate this condition by having just two (medial/lateral) rather indistinct muscle slips. Each conjoined flexor tendon passes through with one of three separate tendon (volar) tunnels on the plantar side of the pes to strongly anchor the attachments of the flexor tendons to the distal phalanx of digits II, III, and IV. A well‐developed m. quadratus plantae was consistently described in all previous specimens of B. tridactylus (Humphry, 1870; Macalister, 1869; Mackintosh, 1875a; Windle & Parsons, 1899).

3.1.38. Mm. lumbricales

The mm. lumbricales were not observed in our dissections of B. variegatus. Humphry (1870) indicated that they were absent in B. tridactylus.

3.1.39. M. adductor digiti I

The m. adductor digiti I (formerly m. transversalis: Humphry, 1870; m. transversalis pedis: Mackintosh, 1875b) was not observed in our dissections of B. variegatus.

3.1.40. Mm. interossei (IOS)

The area of origin for mm. interossei (also referred to as mm. flexores brevis profundi, which includes m. flexor digiti I brevis) is from the dorsal aspect of the pes deep to the m. extensor digitorum brevis and situated between the metatarsal bones (Figure 7A). These muscles primarily arise in pairs on the medial/lateral sides of metatarsals II, III, and IV and extend distally to insert onto (or terminate into) the tendon expansions of EDB (Figure 7A). The bellies of the medial‐most (lateral aspect of metatarsal II) and lateral‐most (lateral aspect of metatarsal IV) interossei were clearly observed in the adult male and female specimens, and these appeared to arise more proximally from the tarsal bones. Thus, the distinctive tendons of the mm. interossei mainly spanned the metatarsophalangeal (MTP) and proximal interphalangeal (PIP) joints and acted at the distal interphalangeal joint (DIP) via the tendon expansions. Previous descriptions indicated that these two bellies extended between metatarsal II and the remnant of metatarsal I on the medial aspect of the pes, and between the metatarsal IV and the remnant of metatarsal V on the lateral aspect of the pes (Humphry, 1870). Our dissections confirm these descriptions (Figures 4 and 6). In addition, mm. interossei were previously noted to project to the plantar aspect of the pes and insert onto the medial aspects of digits II and IV to adduct these digits against digit III (Humphry, 1870; Mackintosh, 1875a). These attachments were not observed in any specimens of B. variegatus.

3.2. Functional mass distribution

Total limb muscle mass (TLMM) as a percentage of body mass is shown in Figure 9A. TLMM varies slightly from isometry (slope: 1.21) across the range of body mass sampled (Figure 9B). This body size scaling trend is consistent between proximal and distal limb muscle masses (data not shown). With respect to the main actions of the muscles in B. variegatus, the knee flexors are the most massive functional group, accounting for 30.8 ± 1.7% of TLMM. The robust femoral belly of m. biceps femoris and the m. semimembranosus, along with the large m. flexor digitorum superficialis, are major mass contributors to this group. The second most massive group is the digital flexors, which account for 18.1 ± 1.7% of TLMM (Figure 10). This functional group also contains the FDS, along with the massive and fused FDP‐L and FDP‐M bellies of the mm. flexores digitorum profundi and the large m. quadratus plantae. The limb adductors (e.g., m. adductor (all heads) and m. pectineus) account for 13.1 ± 1.5% of TLMM, and this mean value is moderately less than that for both the well‐developed hip flexor/extensor functional muscle groups which have equivalent mass (17%). The ankle flexors account for an appreciable amount of TLMM at 11.9 ± 1.3% as do the hindfoot supinators (largely the same bellies) at 7.32 ± 1.2%, which are just slightly more massive than both the knee and ankle extensors, but notably larger than the digital extensors (Figure 10). Except for the relatively larger mass of the hip flexors and smaller mass of the digital flexors in C. didactylus (N = 1), the functional group mass distributions are generally similar across the three species of sloths.

FIGURE 9.

Pelvic limb muscle mass as a percentage of body mass (A) and scaled with body mass (B) across seven individuals of Bradypus variegatus sampled. Total limb muscle mass (TLMM) was calculated as the summed mass of all muscles studied for each individual. The results of a Model I regression indicate that TLMM scaled with slight positive allometry largely due to the influence of one adult individual (Bv8) with hypertrophied muscles

FIGURE 10.

Distribution of functional group muscle mass to total limb muscle mass (TLMM) in three species of a tree sloth. Proximal‐to‐distal muscle group mass is expressed as a percentage, with bars representing means for each functional group. Error bars represent the SD (standard deviation) across N = 7 individuals of Bradypus variegatus (Bv). Muscles with synergistic functions are combined in one functional group. Bi‐articular muscles are included in more than one functional group (i.e., combined mass is greater than 100%). Data for Bradypus tridactylus (Bt: Mackintosh, 1875a) and Choloepus didactylus (Cd: Mackintosh, 1875b) are taken from the literature

4. DISCUSSION

Hindlimb muscles with long, parallel fascicles emphasize contractile excursion and the potential ability to move a limb segment through a large joint range of motion, the latter of which also depends on muscle moment arm (in‐lever) length which will be reported in our future sister study. Nonetheless, muscles capable of undergoing appreciable changes in length are useful for complex arboreal maneuvering as has been previously suggested for suspensory primates such as orangutans (Payne et al., 2006), in addition to being beneficial for braking function and agree with our predictions for pelvic limb function during suspensorial locomotion. For example, the long fascicles of the large m. sartorius may undergo strong lengthening contractions to slow the horizontal translation of the center of mass (CoM), and this postulated role is further consistent with hypotheses surrounding equal body weight distribution between fore‐ and hindlimb pairs during below‐branch walking in sloths (Granatosky et al., 2018; Spainhower et al., 2021). Other proximal flexor muscles of the pelvic limb would be expected to have a greater mechanical advantage by long muscle moment arms at the hip and knee joints (e.g., ILPS, SM, BFI/BFF: unpublished observations) and perform (more strictly) a role in body weight support, while long fascicles present in the digital flexors and supinators (invertors) of the hindfoot may reflect the functional demands of gripping/grasping (i.e., larger contractile excursion to accommodate foot position with different sized substrates). Bradypus is known to place a large amount of its foot area on the substrate during suspension and when vertical climbing (Gorvet et al., 2020; Mendel, 1985), in addition to purchasing or securing a strong grip by flexion of its digits/claws.

Few comparative data for muscle mass distribution are available for sloths. Yet, consistent with the two other species of sloth for which mass has been reported (Mackintosh, 1875a, 1875b), pelvic limb muscle mass in B. variegatus is primarily concentrated in the flexor musculature as expected for suspensory function (Fujiwara et al., 2011; Jouffroy & Stern, 1990; Nyakatura & Fischer, 2011; Olson et al., 2018). At the hip joint, however, orangutans (Payne et al., 2006) and howler monkeys (Grand, 1968) have proportionately larger hip extensors and smaller hip flexors than do sloths, although it is noted the latter species is not restricted to below‐branch locomotion. Moderately‐sized gluteal musculature (20%–28% of muscle mass at the hip joint) not only acts as hip extensors, but also abducts the limb at the hip joint. Large joint mobility in both abduction and axial rotation at the hip joint has been linked to a higher degree of arboreality in primates (Thorpe & Crompton, 2004) and tree sloths appear to be no exception to this functional relationship. Two‐toed sloths most frequently abduct their pelvic limbs when performing clambering and bridging positional behaviors (Granatosky et al., 2018), but do so to a lesser extent during suspensory walking (and posture) despite the thigh assuming a largely abducted position when the limb is retracted over the final one‐third of limb support. In general, the use of non‐parasagittal limb kinematics for suspension is reflective of the highly mobile hip, knee, and ankle joints in sloths (Mendel, 1985b; Nyakatura, 2012).

The hip flexors, digital flexors, and knee flexors are the most well‐developed functional groups in sloths, and combined, accounted for nearly 70% of the calculated TLMM in B. variegatus. A similar bias toward large distal flexor muscle development (e.g., knee, ankle, and digital flexors) is also observed in orangutans (Payne et al., 2006), whereas howler monkeys (Grand, 1968) have a nearly even flexor/extensor muscle mass distribution at their knee joints but equally massive digital flexors for a strong grip on arboreal substrates. These inclusive muscle groups are consistently distributed in three‐toed sloths, although the hip flexors are more developed relative to the digital flexors in C. didactylus (Figure 10). Specifically, the m. iliopsoas and m. gluteus superficialis in two‐toed sloths appear to be, on average, more massive than those in Bradypus, whereas the opposite pattern of mass distribution was found for FDS (not observed in Choloepus) and FDP. These findings may be partially explained by measurement from only a single specimen of C. didactylus (Mackintosh, 1875b) and/or the larger body mass of two‐toed sloths (Grand, 1978). Alternatively, they could be reflective of differences in ecological (Hayssen, 2010, 2011) and behavioral preferences (Granatosky et al., 2018) between genera. For instance, B. variegatus climbs more slowly and frequently, and it goes as high as the emergent level of the rainforest versus C. hoffmanni that engages in more frequent suspensory locomotion/posture (Urbani & Bosque, 2007) and has the capacity to move quicker with greater muscle power (Spainhower et al., 2021). Bradypus is also more selective with the types cecropia trees that it will inhabit (Chiarello, 2008) and prefers larger diameter substrates than does Choloepus (Mendel, 1985).

Differences in hip flexor vs. digital flexor functional group mass additionally might be related to variation in hindlimb length and proportions of the pes between genera (Mendel, 1981a, 1981b). Choloepus has longer distal limbs, a shorter calcaneus, and longer metatarsals compared with the pelvic limb of Bradypus, which has the inverse proportions and resembles more closely morphology for MA in digging taxa (Marshall et al., 2021a). The enhanced mechanical advantage of the pelvic limb musculature and robust digital flexors would be beneficial to support during prolonged vertical grasping/clinging to the substrate as three‐toed sloths rely on stealth as their main strategy for predatory avoidance (Carillo et al., 2009). Thus, the ability to use the hind feet for maintaining a strong grip on the substrate is likely a major selective pressure in suspensory/vertical climbing taxa (Payne et al., 2006). Nevertheless, despite any behavioral nuances, limb segment variation, and differences in mass of the digital flexors by the presence/absence of muscles, the flexor tendons of two‐ and three‐toed sloths are equally strong to support their bodyweight (Mossor et al., 2020) with little‐to‐no flexor muscle activation (Gorvet et al., 2020).

With the exception of the hip extensors, which had a relatively large distribution of mass that is likely associated with the force required for vertical propulsion during climbing and supplemental horizontal propulsion during suspensory walking, the distal muscles acting to both flex the ankle joint and supinate the hindfoot have a comparably appreciable muscle mass distribution. These findings provide added support for our predictions for suspensory function in the pelvic limbs of sloths, primates, and a diversity of climbing mammals that use hindfoot reversal while performing headfirst arboreal descents (Jenkins & McLearn, 1984), but notably in procyonids (Liu et al., 2016; Marsh et al., 2021) and felids (Morales et al., 2018) for which muscle mass distributions are available. Ankle flexors instead of ankle extensors must serve as anti‐gravity muscles in a tensile limb system and this is exemplified by the low ankle extensor to flexor muscle mass ratio of 0.42 for B. variegatus. Sloths, like ringtails (Liu et al., 2016) and margays (Morales et al., 2018), rely mainly on their distal supinators to enable stable suspension by their hind feet. Moreover, several muscles (namely the m. tibialis cranialis muscle complex, including m. extensor digiti I longus) perform both the actions of ankle flexion and hindfoot supination, matching the substrate gripping preferences of Bradypus. Comparative ratios of supinator to pronator muscle mass are 2.7 for B. variegatus, 2.5 for a ringtail and kinkajou (Marsh et al., 2021), and 7.5 for the margay (Morales et al., 2018) due to its massive m. tibialis cranialis but small pronators (evertors). In fact, the m. tibialis cranialis may account for as much as 4.1% of TLMM in carnivores (Gambaryan, 1974), a value that is still less than 6.1%, on average, in tree sloths. Critical to sloths is the ability to maintain their pes in the transverse plane, whereby a single functional unit of the tarsus, metatarsus, and phalanges is rotated to an orientation that is perpendicular to the substrate (Nyakatura et al., 2010), thus ensuring their specialized hook‐like feet have the most secure placement.

Among all muscles of the pelvic limb, however, the knee flexors are consistently the most massive in sloths and maybe most critical to suspensory support. It is possible that the large bi‐articular muscles acting to flex the knee joint mitigate tensile limb loading by sustained flexion. A similar function was evident by the strength (Olson et al., 2018) and muscle activation intensity (Gorvet et al., 2020) of the elbow flexors in the forelimb of B. variegatus. Sloths are incapable of using pendulum‐like exchanges of gravitational potential and kinetic energy during locomotion (Nyakatura & Andrada, 2013) meaning that movement occurs entirely by muscle work performed to translate the CoM, and because sloths deliberately move slowly to conserve metabolic energy, massive and slow‐contracting knee flexors (Spainhower et al., 2021) are best suited to support the bodyweight during the suspension. Additionally, sloths exert appreciable medially directed substrate reaction force (SRF) during below‐branch walking (Granatosky & Schmitt, 2017). As suggested for the forelimb elbow flexors (Gorvet et al., 2020), strong activations of the knee flexors could substantially contribute to the SRF patterns found during hindlimb contact. In addition, muscles acting to adduct the hip joint are comparably massive (~30% of muscle mass at the hip joint) in sloths, orangutans, and howler monkeys, as well as in the pelvic limb of other suspensory taxa (Liu et al., 2016), which suggests the ability of this functional group to facilitate grasping/clinging otherwise expected by universally strong digital flexors. It is further likely that the well‐developed limb adductors with four heads of the m. adductor in sloths also contributes to medial SRF and should act to maintain the horizontal position of CoM by more precise joint position control at the hip joint, all while preventing energy‐wasting by not oscillating the substrate (Alexander, 1991). This hypothesis needs to be verified with future analyses of SRF and muscle architectural properties in Bradypus.

4.1. New perspectives on muscle traits in sloths

Critical observation of pelvic limb myology in B. variegatus broadens our understanding of the family Bradypodidae, as well as potential structural modifications for suspensorial habits in both genera of tree sloths. Table 3 compares myological data obtained from B. variegatus with that of other species of sloth for which similar observations exist (e.g., Cuvier & Laurillard, 1849; Humphry, 1870; Macalister, 1869; Mackintosh, 1875a, 1875b; Meckel, 1828; Windle & Parsons, 1899). This information is useful for evaluating species both functionally and phylogenetically, and it is beneficial for analyses seeking to discover potentially novel postcranial characteristics that relate to suspensory habits in mammals.

TABLE 3.

Comparative hindlimb muscle traits in tree sloths

Trait	Bradypus variegatus*	Bradypus tridactylus a,b,c,d,e,g	Choloepus didactylus d,f,g
Iliacus + Psoas major	Fused	Fused	Fused
Psoas minor	Variable	Present	Absent
Gluteofemoralis	Absent	Absent	Unknown
Distinct belly of gluteus profundus	Variable	Absent	Variable
Piriformis	Variable (fused w/GLM)	Variable	Present
Quadratus femoris	Variable	Variable	Present
Obturatorius externus (# of bellies)	1 (or two parts)	1 or 2 (fused w/OI)	2 (fused w/OI)
Insertion of sartorius	Distal femur + proximal tibia	Distal femur + proximal tibia	Distal femur + proximal tibia
Gracilis (# of muscle parts)	2	Variable (1 or 2)	2
Pectineus	Variable (or fused w/ADDB)	Present	Variable
M. Adductor (# of longus muscle heads)	2 (well‐developed)	2 (well‐developed)	2 (well‐developed)
Semimembranosus (# of bellies)	1	1 or 2	1 or 2
Insertion of semitendinosus	Distal femur or proximal tibia	Tibia	Tibia
Biceps femoris (# of muscle heads)	2	2	2
Flexor digitorum superficialis (m. plantaris)	Present	Variable	Absent
Flexor digitorum brevis§	Unknown	Variable	Present
Quadratus plantae	Present	Present	Present
Tibialis cranialis† (# number of heads)	3	2 or 3	2 or 3
Distinct belly of extensor digiti I (or et digiti II or III)	Absent	Present	Present
Extensor digitorum lateralis‡	Variable	Variable	Variable
Fibularis brevis	Present	Present	Variable
Fibularis quartus	Variable	Absent	No data
Insertion of extensor digitorum longus¿	MT II–III (digit III most common)	MT II–IV (digits II, IV common)	Digits II–IV
Lumbricales	Absent	Absent	Variable

Data sources: ^aMeckel (1828); ^bCuvier and Laurillard (1849); ^cMacalister (1869); ^dHumphry (1870); ^e,fMackintosh (1875a, 1875b); ^gWindle and Parsons (1899); ^*present study; ^§previously identified as separate from m. quadratus plantae; ^†two fibular bellies of the muscle are present in B. variagatus; ^‡possibly a subset of fibers of the m. fibularis longus in B. variagatus; ^¿ muscle contains a subset of fibers previously identified as m. fibularis (peroneus) tertius.

In two‐ and three‐toed sloths observed thus far, there are mixed accounts on the presence of a separate m. gluteofemoralis (Cuvier & Laurillard, 1849) and m. gluteus profundus and m. piriformis (Mackintosh, 1875a, 1875b; Windle & Parsons, 1899), with the reported absence of each belly being an example of them becoming fused with the m. gluteus superficialis or m. gluteus medius, respectively. Specifically, it is possible that the absence of a true m. gluteofemoralis in sloths represents a modification whereby an appreciably large m. gluteus superficialis has a distal femoral insertion via fascia lata of the lateral thigh. Climbing carnivores (Fisher et al., 2008; Liu et al., 2016) and the slow loris (Nycticebus: McArdle, 1981) on the contrary, have a distinct m. gluteofemoralis to act as an additional hip abductor. Moreover, suspensorial apes, as well as pottos (e.g., Arctocebus: McArdle, 1981), have a cranioventral division of their m. gluteus profundus (i.e., m. scansorius) that is absent in tree sloths (Windle & Parsons, 1899). There are also variable observations of the m. quadratus femoris in sloths and its size when present relative to the immediately deep and distinct bipennate belly of the m. obturatorius externus (Table 3), which itself may subsume the small m. obturatorius internus and have two heads in C. didactylus (Mackintosh, 1875b) or present as a muscle with potentially two parts having the same insertion as it did in B. variegatus.

The general fusion of the m. piriformis and m. gluteus profundus with the m. gluteus medius to form a gluteal complex in Bradypus (Windle & Parsons, 1899) is indicative of simplification of the musculature about the hip. Sloths may rely minimally on their hip extensors to push against the substrate during below‐branch walking, and overall, the range of motion in extension at the hip joint is modest (Granatosky et al., 2018). Yet, the mass of the hip extensors is maintained nearly equivalent to that of the hip flexors in B. variegatus, which likely signifies their contribution to vertical support and active hip extension needed for propulsion during vertical climbing (Gorvet et al., 2020; Mendel, 1985). The aforementioned presence/absence and development of m. quadratus femoris may then be another example of simplifying form for numerous muscles that perform similar functions of stabilizing and laterally rotating the femur, which is also expected when sloths are climbing or suspensory maneuvering. We speculate that in individuals where the m. quadratus femoris is absent represents a condition where fusion has occurred between this belly and that of the m. obturatorius externus. It is additionally possible for both bellies of mm. gemelli and the m. quadratus femoris to unite to form another muscle complex (Humphry, 1870).

In contrast, an apparent specialization for use of the proximal pelvic limb for suspensory and climbing behaviors in sloths is reflected in the myology of the limb adductors. Complexity involving the number of heads of the m. adductor, including the large m. pectineus, as well as the divisions of the m. gracilis implies a greater degree of joint position control (Ercoli et al., 2015; Julik et al., 2012) that can be achieved in limb ab‐adduction. Similar observations, including two bellies of the m. adductor brevis and m. gracilis, as well as a notably distal insertion for m. pectineus exist in slow climbing Nycticebus (McArdle, 1981), which shares numerous evolutionary convergences with tree sloths. Thus, opposite to the muscle fusion observed in the hip extensors suggesting modification for gross movement rather than fine position control, the distinctive two heads of m. adductor longus in Bradypus and two parts of m. gracilis, with the latter acting on along the full length of the pelvic limb, emphasize the ability to stabilize the position of both the hip and knee joints in suspensory postures. Strong adduction of the pelvic limb would also be beneficial to the slow vertical climbing behavior observed in B. variegatus (Gorvet et al., 2020). Sustained vertical grasping/clinging via adduction of the thigh and flexion of the leg segment would supplement the contractile requirements for strong digital flexion and hindfoot supination to remain firmly in contact with the substrate. Moreover, the expansive distal insertion of the external part of m. gracilis (Table 3) into the crural fascia extending to the calcaneus (observed in both genera) may tense this connective tissue and facilitate the extension of the hindfoot at the ankle joint. This hypothesized action could assist in counter‐balancing ankle joint loading during suspension and/or vertical climbing. A comparably strong selection for suspensory locomotion and posture along with vertical climbing may have resulted in convergent evolution of the broadly similar observed myology of both the adductors and m. gracilis seen in Bradypus and Choloepus.

Joint position control for skillful arboreal maneuvering can likewise be observed in the form of the bi‐articular mm. semimembranosus, semitendinosus, and biceps femoris. Despite having dual insertions on the distal femur and proximal tibia in Bradypus (Humphry, 1870), the m. semimembranosus (SM) is most commonly a single muscle with no discernable divisions. However, Windle and Parsons (1899) described a second belly of SM (called m. presemimembranosus) as a part of the m. adductor arising from the ischial tuberosity and inserting onto the distal femur near the medial condyle, which approximates the location of the alternative femoral insertion of the sole m. semitendinosus (ST) that was observed variably in some individuals of B. variegatus. Howler monkeys have a m. semimembranosus accessorius (Grand, 1968), which appears to represent the arrangement of this second belly seen in two‐ and three‐toed sloths. Notably, the pelvic limbs of the red panda (Fisher et al., 2008), ringtail (Liu et al., 2016), and kinkajou (Marsh et al., 2021) have multiple bellies of SM and ST with variable degrees of fusion at or near both their origins and insertions. Whereas the SM can have up to three heads with at least one inserted near the medial femur condyle (Marsh et al., 2021), neither head of ST inserts on the distal end of the caudomedial femur in these scansorial‐to‐arboreal carnivores. It is, therefore, difficult to resolve the homology of an extra muscle previously observed in sloths. The presence of this muscle as documented may simply be examples of individuals where ST was inserted onto the distal femur rather than the tibia. Moreover, prior descriptions of the m. adductor (magus: Humphry, 1870; brevis: Mackintosh, 1875a) arising from the ischial tuberosity further confuses the identity and number of bellies of caudal thigh muscles possible in sloths. Future dissections of B. torquatus and C. hoffmanni should help resolve any remaining inconsistencies.

The arrangement of the m. biceps femoris in sloths demonstrates the opposite deviation from myology typical of carnivorans by having two bellies instead of one and more closely resembles the condition observed in monkeys and great apes, including humans (Payne et al., 2006). Although the ischial (or long) head has a consistent homology across mammals, our findings in B. variegatus support a previous assertion that the femoral (or short) head is homologous with m. abductor cruris caudalis (m. tenuissimus), but with a modified femoral origin in xenarthrans (Windle & Parsons, 1899). The importance of strong knee flexion may have been a selective pressure driving the rearrangement of this muscle in suspensory taxa such as sloths and primates. We may further speculate that having a little‐to‐no tail could be another factor related to modification in the presence/absence, arrangement, or size of several muscles of the caudal thigh that would otherwise arise from caudal vertebrae. Indeed, unlike the typically strap‐like m. abductor cruris caudalis (Liu et al., 2016), which does not coexist with a femoral head of m. biceps femoris, the latter is very well‐developed as distinctly evidenced in tree sloths and it is purportedly one of the strongest knee flexors. The femoral head of m. biceps femoris also has the potential to forcefully lateral rotate the leg due to its consistent distal insertion onto the fibula in sloths (Windle & Parsons, 1899). As much as 45° of rotation of the leg relative to the thigh may occur in sloth pelvic limbs when the knee is flexed at 90° (Mendel, 1985b). This position of the leg may facilitate alignment of the pes perpendicular to the substrate beyond enhancing medially directed SRF on the substrate important for stability in suspension.

Within the distal limb, FDS (or m. plantaris) was not observed in C. didactylus (Humphry, 1870; Mackintosh, 1875b), although it may have been previously mistaken for a portion of the lateral head of the m. gastrocnemius. Nevertheless, the absence of an additional muscle to both flex the knee and digits (primarily) could be a derived trait in two‐toed sloths and one that may be closely linked to their behavioral preferences. Accounts of hindlimb myology for C. hoffmanni are outstanding and needed to support this supposition. Despite the absence of FDS, independent control of the digits in the hind feet of sloths is not obviously necessary for their suspensory behaviors. This muscle gives rise to a tendon distally that merges with that of the common flexor tendon of FDP and this arrangement of the digital flexors as observed in B. variegatus provides strong, simultaneous flexion of the digits. Moreover, the presence of m. flexor digitorum brevis (FDB), which is a superficial intrinsic muscle of the pes in climbing carnivores (Fisher et al., 2008; Liu et al., 2016; Marsh et al., 2021), remains uncertain in sloths due to mixed accounts between two‐ and three‐toed forms, including our own evidence of muscle, slips with no clear insertion that could be remnants of m. flexor digitorum brevis (Tables 2 and 3). Descriptions of a small FDB originating as one to three superficial muscle slips from the calcaneus and inserted into the deep flexor tendons of digits II–IV (Humphry, 1870; Windle & Parsons, 1899) or digit III only (Mackintosh, 1875b) are confounded by a well‐developed m. quadratus plantae that was previously described as having similar origin locations and inserted into all three flexor tendons (Bradypus) or more strongly into that of digit IV in C. didactylus. A distinct belly for a second ancillary flexor of the digits is perhaps lost in B. variegatus due to enlargement of the m. quadratus plantae and this feature may represent another form of limb simplification involving sloth feet. However, because FDB or m. quadratus plantae could be described interchangeably for the previously named m. musculii accessorii (Macalister, 1869), and FDB was not previously observed in some specimens of B. tridactylus (Macalister, 1869; Mackintosh, 1875a), it is challenging to compare our observations with some historical accounts. That said, we consider FDB to be vestigial in B. variegatus and choose to maintain the identity of a well‐developed m. quadratus plantae as the main intrinsic flexor of the digits until future observations in additional Bradypus sp. may confirm otherwise.

With regard to the digital extensors, the m. extensor digiti I longus, which is potentially present as an independent muscle in species of sloth previously dissected, has been variably named and often incorrectly identified (Humphry, 1870; Macalister, 1869; Mackintosh, 1875a, 1875b; Windle & Parsons, 1899). Discrepancies in muscle nomenclature for sloths are indeed related to having only three remaining digits on their hind feet and the common use of human anatomical references. Furthermore, m. extensor digiti I longus has a modified condition in B. variegatus by: 1. inserting onto the rudimentary metatarsal I and no longer having an extensor action, and 2. merging with the two heads (one tibial and one fibular) of the m. tibialis cranialis to form a large muscle complex (Figure 8C). For these reasons, we renamed the proximal and distal fibular bellies observed in B. variegatus, the fibulare et tibiale head and fibulare head, respectively (Tables 2 and 3). The massive m. tibialis cranialis in tree sloths has historically been described as having three heads and, therefore, the muscle has a substantially modified origin and strong supinator function. This postcranial morphological trait is recognized as a specialization for suspensory habits in numerous climbing carnivores (Jenkins & McClearn, 1984) that are capable of hindfoot reversal related to having a broad tibia‐fibular origin for their well‐developed m. tibialis cranialis (Liu et al., 2016; Marsh et al., 2021; Morales et al., 2018). A robust mm. flexores digitorum profundi also facilitates hindfoot supination in those taxa, and most likely in sloths, and is an action for which in addition to these muscles, could be further supplemented by the m. gastrocnemius in Bradypus due to its medial insertion on the calcaneus.

The observation of multiple heads for m. tibialis cranialis that includes the homologous muscle representing the m. extensor digiti I longus, however, does not completely agree with all prior myological accounts for sloths, namely due to the absence of a separate (extrinsic) extensor of digit II in B. variegatus. In both B. tridactylus (Mackintosh, 1875a) and C. didactylus (Humphry, 1870) a second, small extensor muscle inserts onto the proximal phalanx of digit II or distal phalanges of digits II and III, respectively. Alternatively, in previous dissections where only two heads of m. tibialis cranialis were documented, the m. extensor digiti I longus was reported to either extend digit II (Macalister, 1869) or insert onto the base of the remnant metatarsal I (Mackintosh, 1875b). Although it is plausible that the loss of digit I resulted in a rearrangement of m. extensor digiti I longus (and thusly should be renamed m. extensor digit II or II et III), we are convinced that redundancies present in the existing descriptions for this muscle are largely attributed to subdivisions of m. tibialis cranialis (fibular heads), or more specifically the m. extensor digitorum brevis, which can be realized as muscle fusion or loss when its origin from the distal‐most fibula (Windle & Parsons, 1899) is not observed. This is most likely the condition in B. variegatus where the digital extensor muscle mass is restricted to an intrinsic origin from the dorsal pes.

Another muscle without an extensor action at the digits is the m. extensor digitorum lateralis and has a questionable presence in Bradypus (Tables 2 and 3), whereas it was not observed in C. didactylus (Mackintosh, 1875b; Windle & Parsons, 1899). The muscles of the lateral leg generally insert either at the base of or onto the tubercle of the remnant metatarsal V in sloth hind feet and there is debate over if the presence/absence of m. extensor digitorum lateralis involves the condition of m. fibularis longus (FL) and/or presence of a m. fibularis tertius. For example, m. fibularis tertius was previously described to insert onto metatarsals IV and V (Mackintosh, 1875a), and although it also was not clearly observed in B. variegatus as it was in C. didactylus (Mackintosh, 1875b), this dual insertion closely matches that of FL. To add further confusion, the small m. extensor digitorum lateralis as observed here could simply be a subdivision of fibers from the caudal belly of FL, whereas m. fibularis tertius (viewed as homologous with m. extensor digitorum lateralis: Macalister, 1869) would be expected to arise as a small belly from the distal, cranial fibula and possibly share fibers with m. extensor digitorum longus (EDLO) as it did in two specimens of Choloepus (Windle & Parsons, 1899). However, we interpret those previous observations as a division of EDLO by the insertion onto metatarsal IV rather than the presence of a separate muscle. Moreover, m. fibularis tertius is rarely identified in studies of pelvic limb musculature in primates (Channon et al., 2009; Demes et al., 1998; Marchi et al., 2018; Payne et al., 2006), which include evaluations of several suspensorial species. The available descriptions for the fibularis muscles of sloths arethusly overly complicated. To this end, rather than perpetuate erroneous complexity, we contend that functional compartmentalization of a large FL best accounts for the overall variation observed in the pelvic limb of B. variegatus, although FL, m. fibularis brevis, and m. extensor digitorum lateralis are the three muscles existing to pronate the hindfoot in sloths.

The insertion of EDLO onto the metacarpals rather than the phalanges of the digits is a final remarkable trait in Bradypus (Humphry, 1870; Macalister, 1869; Mackintosh, 1875a) and once more suggests that ankle flexion for anti‐gravity function is favored over digital extension. This modification, however, does vary by the number of tendons of insertion and the most common metatarsal of attachment (Table 3), but it equally represents a clear decoupling of ankle flexion and digital extension in three‐toed sloths that is not found in C. didactylus (Humphry, 1870; Mackintosh, 1875b). This disparity is remarkable given the overall high level of morphological convergence between the two genera (Nyakatura, 2012). Furthermore, the derived sole ankle flexor action of EDLO in Bradypus could be a consequence of a role of the massive m. tibialis cranialis muscle complex to both flex the foot at the ankle joint and supinate the hindfoot. Thus, the only muscle capable of digital extension is the intrinsic m. extensor digitorum brevis. This arrangement suggests that a limited range of force for extension of the digits as a functional unit is mainly required to counterbalance strong digital flexion. Yet, the m. extensor digitorum brevis does appear to be appreciable in size, but this may be due to difficulty in separating its belly from the well‐developed dorsal mm. interossei it overlies.

The mm. interossei are reported for all species of sloth for which data are available and they potentially allow for a limited degree of digital abduction, while the variable‐to‐absent condition of the mm. lumbricals (Table 3) and other intrinsic foot musculature in sloths further suggests the lack of fine control of the digits in their hind feet and instead, generalized use of their syndactylous digits as a functional unit. This feature strongly contrasts with the feet of numerous other clades of arboreal mammals that rely on both strong flexion and ab‐adduction of the digits for climbing (Marsh et al., 2021). Nonetheless, the hind feet of C. didactylus do show more complexity than those in Bradypus by the presence of mm. lumbricales arising mainly from the plantar insertion tendons of the m. tibialis cranialis (Humphry, 1870), which beyond a role in hindfoot supination, the latter muscle is notably modified for digital flexion function in two‐toed forms (Windle & Parsons, 1899). The elongate plantar pes of C. didactylus also contains m. adductor digit I as well as m. abductor digiti I (Mackintosh, 1875b) indicating that this species may require greater position control of its digits for use of smaller diameter substrates and more direct gripping with its extremely re‐curved claws. Our future dissections of C. hoffmanni aim to verify nuances in these functional traits. In contrast, the rudimentary muscle slips arising from the calcaneus or a proximal tendon along the plantar aspect of the pes and passing namely to the lateral aspect of metatarsals IV–V that were variably observed in the hind feet of B. variegatus most likely represent vestiges of intrinsic ab‐adductors of the lateral‐ and medial‐most digits.

4.2. Concluding remarks

Updated postcranial morphological data are severely lacking for numerous placental mammalian clades. The general lack of limb muscle characters from sloths (e.g., Bradypodidae) and other suspensory mammals too often limits our interpretations of their functional diversity. In this study, it was observed that the pelvic limb of brown‐throated three‐toed sloths demonstrates several modifications to the musculature that supported our hypotheses and predictions about their suspensory function. In addition to well‐developed flexor groups, most often the alterations in muscle structure are realized as simplifications of the limb system that involved muscle belly fusion, subdivision, or reduction. Perhaps the imprint of evolution for suspension is most apparent in the distal limbs and derived feet of tree sloths, where the loss of digits I and V has resulted in a reorganization and prioritization of flexor function for several digital “extensors” as well as a simplification of the intrinsic foot musculature overall. Our future studies will answer additional questions about muscle architecture and locomotor mechanics in sloth pelvic limbs.

AUTHOR CONTRIBUTIONS

M.T.B. developed the concepts and approach, collected and analyzed data, and drafted and revised the manuscript; D.M.M. performed data analysis, created illustrations, and edited and revised the manuscript; K.B.S. performed data collection and edited the manuscript; D.R.T. performed data collection and edited the manuscript; B.A.C. performed data collection and edited the manuscript; J.A.A.‐A. provided support and access to specimens, supervised data collection, and edited the manuscript; S.P.K. performed data collection and edited the manuscript; R.N.C. developed the concepts and experimental approach, collected and data, and edited the manuscript.

COMPETING INTERESTS

The authors declare no competing financial interests.

Supporting information

Table S1

Butcher, M.T. , Morgan, D.M. , Spainhower, K.B. , Thomas, D.R. , Chadwell, B.A. , Avey‐Arroyo, J.A. , et al (2022) Myology of the pelvic limb of the brown‐throated three‐toed sloth (Bradypus variegatus). Journal of Anatomy, 240, 1048–1074. Available from: 10.1111/joa.13626

DATA AVAILABILITY STATEMENT

All data generated for this study are included in this published article in adherence with disclosure policy of the journal. The authors are not sharing new data alongside the results of this report at this time.