Functional morphology and flight kinematics of Artibeus jamaicensis (Chiroptera, Phyllostomidae)


Material Information

Functional morphology and flight kinematics of Artibeus jamaicensis (Chiroptera, Phyllostomidae)
Physical Description:
x, 160 leaves : ill. ; 28 cm.
Hermanson, John W
Publication Date:


Subjects / Keywords:
Artibeus jamaicensis   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1983.
Includes bibliographical references (leaves 153-159).
Statement of Responsibility:
by John W. Hermanson.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000365967
notis - ACA4795
oclc - 09967091
System ID:

Full Text







Copyright 1983


John W. Hermanson


Financial support for this study was provided in part by a

Grant-in-Aid of Research from the Society of Sigma Xi; from the

Department of Zoology, University of Florida; and from the Department

of Natural Sciences, Florida State Museum. I received support in the

form of a Museum Assistantship in the Department of Natural Sciences,

Florida State Museum, and a Graduate Teaching Assistantship in the

Department of Physiological Sciences, College of Veterinary Medicine.

Sr. Tomas Blohm kindly allowed me to study on his ranch in

Venezuela in 1981. Dr. John Robinson invited me to travel with him to

Venezuela, and then took care of the necessary procedures to obtain

permission for me to collect and transport live bats in that country.

I also acknowledge the Department of Biology, University of New Mexico,

for allowing me free access to the facilities of the Museum of

Southwestern Biology.

I thank my committee members for advice and attention at a time

when none of them needed the additional burden of a dissertation to

review. My committee included Drs. John Anderson, Donald Dewsbury,

David Webb, Ronald Wolff, and Charles Woods. Pamela Johnson typed and

then retyped the manuscript. I appreciate her patience in this task.

I offer thanks to the four men in biology who have most influenced

my path. Dr. Scott Altenbach, of the University of New Mexico, opened

his laboratory and his home to me during our joint venture. We will

have many enjoyable moments to reflect on during the years to come.

Dr. David Klingener, of the University of Massachusetts, taught a

fabulous course in comparative anatomy and opened my eyes to a career

in biology. Dr. Ted Goslow, of Northern Arizona University, provided

me with high goals in both my personal and academic life. Finally, my

advisor Dr. Charles Woods has provided friendship and a wealth of

experiences in natural history and anatomy on which to build a career.

I am fortunate to have studied with these four gentlemen.

I thank my peers for those good moments that make graduate

education unique. These moments include late nights out on the silent

Arizona desert, evenings of light snow or Great-horned owl hootings in

the Green Mountains, or simply industrious evenings in the labs at the

Museum. In particular, I will miss those evenings out on the back

porch with the folks in our household at Newnan's Lake. All of these

times are irrevocably tied together in my mind.

Last but not least I wish to acknowledge my family. My mother and

father endured my endeavor and provided encouragement and support. My

aunt, Catherine Leonard, has provided inspiration and the gift of

laughter. My "other aunt," Ruth Cushman, has been a friend

throughout. Finally, I remember my companion of many treks, Japhy.




LIST OF FIGURES . . ... vi

LIST OF TABLES . . ... .viii

ABSTRACT . . . ix




Results . . ... 15
Discussion . . .. 24

OSTEOLOGY . . .. 31


Trapezius Group . . ... .53
Costo-spino-scapular Group . ... 62
Latissimus Group . . .. 71
Deltoid Group . . 79
Supraspinous Group . . 84
Triceps Group . . 88
Pectoralis Group . . .. 91
Flexor Group of Arm . .. 103
Antebrachial Extensor Group . .. 105
Antebrachial Flexor Group . ... 123


CONCLUSIONS . ... .. 149




1. The wingbeat cycle of Artibeus jamaicensis. 7

2. Wing and body movements during slow forward flight in 17
Artibeus jamaicensis.

3. Wing and body movements during slow forward flight in 19
Artibeus jamaicensis.

4. Wing and body movements relative to still air during 21
slow forward flight in Artibeus jamaicensis.

5. High- and low-aspect ratio wingshapes. 28

6. Lateral view of the thoracic and axillary skeleton. 33

7. Cranial view of the right pectoral girdle. 35

8. Dorsal view of the scapula. 37

9. Medial view of the left humerus. 41

10. Articular surfaces of the humerus. 43

11. Dorsal view of the radius and ulna. 48

12. Dorsal and ventral views of the carpus. 51

13. Dorsal view of the shoulder and arm of Artibeus jamaicensis. 56

14. Activity patterns of shoulder and arm muscles in Artibeus 58
during slow flight.

15. Lateral view of serratus ventralis musculature. 66

16. Ventral view of the shoulder and arm of Artibeus 94

17. Electromyographic data for six regions in the pectoralis 99
muscle during slow flight.

18. Lateral view of the muscles of the elbow region. 108


19. Dorsal view of the muscles of the carpal region. 110

20. Medial view of the muscles of the elbow region. 125

21. Ventral view of the muscles of the carpal region. 127

22. Numbers of coactive muscles during the wingbeat cycle. 145


1. Aerodynamic parameters of the wing in two fruit bats, 22
Artibeus jamaicensis.

2. Electromyographic data for activity patterns of the 59
shoulder musculature in Artibeus jamaicensis during
slow flight.

3. Electromyographic data for activity patterns in six 100
regions of the pectoralis muscle of Artibeus
jamaicensis during slow flight.


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



John W. Hermanson

April 1983

Chairman: Dr. Charles A. Woods
Major Department: Zoology

The functional morphology of the pectoral girdle and arm of

Artibeus jamaicensis (Chiroptera, Phyllostomidae) is interpreted on

the basis of gross dissection and electromyographical analysis (EMG).

Electromyographic data obtained during flight for 15 muscles elucidate

several temporal patterns of activity associated with the wingbeat

cycle that are not similar to the patterns of flexor, extensor, and

bifunctional muscles observed in terrestrial mammals. Abductor

muscles exhibit intense activity associated with the early upstroke

phase of the wingbeat cycle and include clavotrapezius, acromiotrapezius,

latissimus dorsi, teres major, acromiodeltoideus, spinodeltoideus, and

triceps brachii (long and lateral heads). All abductors except for the

triceps brachii exhibit a secondary period of low-amplitude activity

associated with the early downstroke. Adductor muscles exhibit primary

activity immediately prior to and during the early downstroke phase.

The adductors include serratus ventralis thoracis, pectoralis, and

clavodeltoideus. Bifunctional muscles exhibit a single period of

activity through most of the wingbeat cycle, or two separate periods

of high-amplitude activity during a wingbeat cycle. The bifunctional

muscles include spinotrapezius, supraspinatus, infraspinatus, and

subscapularis. Dissection of all other muscles of the shoulder and

antebrachium form the basis of interpreting musculoskeletal movements

during flight in Artibeus. The major muscles of support in Artibeus

include serratus ventralis thoracis, pectoralis, and the trapezius

group. These muscles support the trunk between the wings during

flight or terrestrial locomotion. Propulsion during flight differs

from that observed during stepping in terrestrial mammals. During

the wingbeat, pectoralis provides the major component of thrust,

both by adducting and pronating the wing. Although latissimus dorsi

is a major propulsive muscle during stepping in terrestrial mammals,

its major function in Artibeus is to abduct the wing and reposition

the wing prior to the beginning of a downstroke.


Bat flight is energetically expensive (Thomas and Suthers, 1972;

Carpenter, 1975; Thomas, 1975, 1981) and requires extensive alterations

of the generalized mammalian morphology (Miller, 1907; Vaughan,

1970a). In order to understand how bats fly, one must both study their

anatomy as well as the subtle interactions between the nervous and

muscular systems of several bats must be studied. This study provides

information about the neuromuscular and osteological mechanisms

associated with flight in a frugivorous bat, Artibeus jamaicensis

(Microciroptera, Phyllostomidae). The objectives of the project are to

describe in detail the musculoskeletal system of Artibeus, to

illustrate the kinematics of the normal wingbeat cycle, and to obtain

electromyograms during flight for the major muscles of the shoulder

region and arm.

Three factors make Artibeus jamaicensis desirable for

morphological analysis. First, Artibeus is a ubiquitous neotropical

bat whose basic biology has been well documented. The available

information regarding the foraging and flight habits of this bat

permits deduction of the demands placed on the bat. Second, Artibeus

jamaicensis is easily maintained in captivity during the duration of an

experimental period (Rasweiler, 1977). Third, Artibeus is placed in

the family Phyllostomidae (Jones and Carter, 1976), a family that has

not been employed in previous EMG studies.

The diet of Artibeus includes fruits, flower products, and leaves

(Goodwin and Greenhall, 1961; Villa-R., 1967; Heithaus et al., 1975;

Gardner, 1977; Bonaccorso, 1979). Insectivory is insignificant in this

species even though Tuttle (1968) observed Artibeus jamaicensis in

Mexico feeding on blackflies within a cave. The bats usually transport

relatively heavy fruits from their source to adjacent feeding roosts

except when too large to carry efficiently (Morrison, 1978a;

Bonaccorso, 1979). Thus the wings of Artibeus need to be capable of

producing lift at low speeds without a great deal of maneuverability.

The flight style of Artibeus differs from that of Desmodus,

Antrozous, and Myotis, three bats previously studied with

electromyography. The vampire, Desmodus, is characterized by its

direct, swift flight (Altenbach, 1979). Antrozous and Myotis are both

insectivores but each employs slightly different foraging behaviors.

Antrozous is a long-eared bat and either gleans insects from foliage or

preys upon large, slow-moving insect species at low altitudes. Myotis

myotis tends to also be an opportunistic gleaner, searching for and

obtaining insects from the surface of foliage or the ground (Findley,

1972). Desmodus, Antrozous, and Myotis all forage for extended

periods of time during the night. Antrozous was reported to engage in

continuous foraging on the wing for one to four hours during early

evening and later for several hours before dawn (O'Shea and Vaughan,

1977). Myotis lucifugus similarly feeds for several hours during

early evening, and again for several shorter periods before dawn

(Anthony and Kunz, 1977). I expect behavioral studies will show Myotis

myotis has a similar time-energy budget. Turner (1975) reported that

Desmodus spends one to three hours foraging per night and found that

the bats spent a mean of one hour per night searching for their prey.

Artibeus, in contrast, tends to spend relatively little time in flight

during the night (Morrison, 1978a; 1978b). Artibeus also is not very

maneuverable and is easily captured in nets. There are two brief

flights between the day roost and the food source: one at dusk to the

food and one shortly before sunrise back to the day roost (Morrison,

1978a). Morrison found that day roosts were never very distant from a

source of ripe fruit. At night, Artibeus selects fruit at the source

tree and then carries the fruit to an adjacent feeding roost, often in

another tree or in a cave (Dalquest, 1953; Morrison, 1978a; Bonaccorso,

1979). After chewing the fruit and swallowing juice and edible parts,

the bat ejects the remaining fruit pulp and seeds before returning to

the source tree for another fruit. These feeding flights are repeated

many times during the night.

Artibeus jamaicensis transports fruit on average 270 m from the

parental trees to feeding roosts (Janzen et al., 1976).

Morrison (1978a) observed that the distance between several Ficus

insipida and Ficus yaponensis, two species of figs important as food

trees, and the adjacent feeding roosts ranged from 25-400 m (mean 175

m). It is not clear why these bats carry fruits away from the parent

tree. Janzen et al. (1976) speculated that branch morphology and odor

in Andira inermis evolved to insure dispersal of the seeds away from

the parent tree. Seed mortality resulting from fungal infections and

weevil predation was most severe beneath the parent trees but declined

with distance away from the tree. They were unable to demonstrate a

clear cause and effect relationship, but did speculate that bats were

intolerant of the odor associated with many ripe Andira fruit. As an

alternative hypothesis, several authors suggested that Artibeus carried

fruit to a distinct feeding roost to decrease predation (Fenton and

Fleming, 1976; Morrison, 1978a; August, 1979). Artibeus avoid visually

orienting predators such as owls and opossums, by feeding only on dark

nights or during dark periods of the night (Morrison, 1978b). Specific

cases of predation by Didelphis marsupialis upon Artibeus were reported

in Venezuela (August, 1979). In any case, Artibeus make a number of

short flights during the night to and from a ripe fruit tree.

The neuromuscular control of flight can be correlated with similar

mechanisms in operation in the control of terrestrial locomotion.

Analysis of terrestrial involved resolution of the locomotor activity

of each limb into successive strides or step cycles, and quantification

of the interaction of these events between the limbs (gait analysis)

(Wetzel and Stuart, 1976). Philippson (1905) described four stages in

the step cycle of dogs, beginning with foot liftoff: F or flexion

phase is when most joint angles in the limbs decrease; El is the final

portion of the "swing" or non-support phase of the stride; E2 is the

first portion of the "stance" or support phase and includes some

passive joint flexion as a result of the limbs yielding under the

animal's weight; E3 is the final portion of the "stance" phase during

which all joints are actively being extended and maximum thrust is

being generated by the limb. Neurophysiologists studying vertebrate

locomotion have long focused on events surrounding the two reversal

points in the stride. For example, Sherrington (1910) made the

observation that a graded alternation of activities occurred in the

flexor and extensor muscles and that this alternation could be invoked

in a model of reflexive control of the step cycle (cf. Wetzel and

Stuart, 1976).

As an analogy, Hermanson and Altenbach (1981) expressed the

chiropteran wingbeat cycle as being a quantifiable and repeated unit of

normal bat flight. Individual wingbeats exhibit little variation among

conspecific individuals performing the same locomotory tasks under

similar conditions. The chiropteran downstroke, or adductor phase, is

similar in basic form and timing to the E2-E3 phases of the Philippson

step cycle and produces components of both lift and thrust (Norberg,

1976). The chiropteran upstroke, or abductor phase, parallels the F-El


cO 5

bo C -
60 0)
1 -i
HO (

cd 0

* o

o0 a

0 -r-4 40


r&. toi

1 0

4 0W

6C C
Sco Q
9 c

-H 14
40 0-

u a)


tn 4-> E

4-i4 Wd



* C

4-H 4-
0 Q0

3 CL

3 0
0 O

C o

) Cf

$3 0




phases of overground locomotion in its function as a recovery stroke.

The last portion of the upstroke often includes a thrust-producing

"flick phase" (Norberg, 1970, 1976). Except for the flick phase, it is

not practical to subdivide the upstroke or downstroke because of the

smooth arc followed by the wing during each phase (Figure 1).

Engberg and Lundberg (1969) demonstrated that cat limb muscles

exhibit characteristic activity profiles during overground locomotion:

extensor or flexor muscles produce a cumulative effect on the skeletal

system through overlap or coactivation, and muscles that span two

joints produce a complex function including both flexion and

extension. Engberg and Lundberg reported extensor activity prior to

foot touchdown during normal stepping. This important observation

indicated that muscle activity could be preprogrammed at the spinal or

supraspinal level, taking precedence over reflexive control of the step

cycle (Wetzel and Stuart, 1976). Subsequent studies demonstrated that

the EMG profiles for homologous mammals are conservative among mammals,

with subtle differences in timing related to differing functional

demands being placed upon the musculoskeletal system in various species

(English, 1978a, 1978b; Gambaryan, 1974; Rasmussen et al., 1978;

Tokuriki, 1973a, 1973b).

Hermanson and Altenbach (1981, 1983) assembled EMG data for muscle

activity during flight in an insectivorous bat, Antrozous pallidus

(Vespertilionidae). Of 17 muscles studied, 11 muscles demonstrated

patterns of either adductor or abductor activity. Only six muscles

exhibited a complex bifunctional pattern of activity. The adductor

muscles demonstrated an onset of activity on average 15-25 msec prior

to initiation of the downstroke. This pattern is similar to the El

extensor coactivation observed prior to foot touchdown in cats (Engberg

and Lundberg, 1969) and suggests that similar neural control mechanisms

are involved in regulating both the terrestrial step cycle and the

aerial wingbeat cycle. Despite gross anatomical modification during

the evolution of the chiropteral musculoskeletal system, the patterns

of limb movement and muscle activity remained relatively unchanged from

the basic mammalian condition.

This study presents the descriptive anatomy of the locomotor

system of A. jamaicensis, including the shoulder girdle, arm, and

forelimb portions of the wing. Descriptions are accompanied by

functional hypotheses for each muscle following a format employed in a

number of studies on the mammalian musculoskeletal system (Howell,

1926; Rinker, 1954; Klingener, 1964; Woods, 1972). Additionally, EMG

data are presented for 15 of the 17 muscles that were previously

analyzed in Antrozous pallidus (Hermanson and Altenbach, 1983). These

data are compared with simialr EMG observations upon terrestrial



Ten skeletons and ten alcoholic specimens of Artibeus jamaicensis

were used for descriptive study. Dissection specimens were preserved

in a solution of 10 percent solution of formalin for 10-14 days and

were later stored in 70 percent ethanol or 40 percent isopropanol.

Complete dissections performed on the shoulder, arm, and antebrachial

regions form the basis of descriptions and illustrations presented in

this study and for the design of electromyography experiments.

Terminology used in my study conforms to recommendations by the

International Commission on Veterinary Anatomical Nomenclature (Nomina

Anatomica Veterinaria, 1975). Although this procedure initially

appears to pose difficulty in making comparisons with previous studies

on chiropteran morphology, relatively few new terms are introduced.

Strict adherence to the terminology of the Nomina Anatomica Veterinaria

(NAV) will provide a consistent vocabulary for effective communication

about the subject and will facilitate communication with researchers

who have not mastered the specialized jargon of bat anatomy.

Several terms are used throughout this text to indicate direction

along the body axis or direction of movement at a joint. The terms

cranial and caudal have replaced anterior and posterior, respectively,

in all cases where the latter terms have been used in the chiropteran

literature. Cranial refers to the cranial end of the animal, or

indicates that a structure passes from a point on the body along a line

towards the transverse plane of the head. Caudal indicates the

opposite direction to cranial: to or towards a transverse plane

containing the caudal vertebrae. Medial and lateral are used to

indicate position or movement towards or away from the medial plane of

the body. Flexion and extension refer to a decrease or increase,

respectively, in the angle between two bones. Adduction is used to

denote movements of the wings, humerus, or clavicle towards the medial

plane on the ventral surface of the body. Abduction of the wings,

humerus, or clavicle is a movement away from the median plane of the

ventral surface of the body. Rotation of a bone about its long axis is

described as supination, or outward rotation of the top or lateral

surface of the bone, and pronation, an inward rotation of the top or

lateral surface of the bone. To facilitate comparison with terrestrial

mammals, the arm and antebrachium are described in the position of

maximal adduction. Thus these bones have cranial, caudal, lateral, and

medial aspects. The carpus and digits are described as if they are

held horizontally at the side of the body. These elements have

cranial, caudal, dorsal, and ventral palmarr) aspects. The apparent

inconsistency in terms and orientation between the anatomical position

of the arm and carpus is in fact the traditional and accepted format

for descriptive purposes.

Skeletons were measured with dial calipers in order to describe

shape and size of individual bones. This information is useful in

estimating position or attachments of soft tissues with respect to the


Muscle origins and insertions were described relative to named

structures or surfaces of individual bones: the extent of muscle

attachment to long bones is expressed as a percentage or proportion of

the total length of the bones. Fiber orientation was expressed with

respect to the long axis of a muscle, along a line fit by eye between

the lines of attachment.

Synchronized film and EMG records were obtained with a system

initially described in Altenbach (1972, 1979) and modified by Hermanson

and Altenbach (1981). A Hycam high-speed cine camera (Red Lakes

Laboratories, Inc.) photographed flight sequences at film speeds of

300-400 frames per second. Myopotentials were recorded from a bipolar

fine copper wire electrode implanted in the muscle belly. Implantation

was achieved by a simple microsurgical procedure. An incision was

first made in the dorsal integument superficial to the lumbar region

and a silicone plug assembly was sutured to the muscle and ligaments of

the paravertebral region. Electrodes were passed subcutaneously to a

second incision at the level of the specific muscle being studied. A

shielded cable surrounding the electrode assembly served as a tissue

ground. With the aid of a microscope, the electrode tips were

inspected to insure than less than

0.5 mm of the tip was bared. The tips were then placed in the muscle

belly with a 27 guage hypodermic needle, and both incisions were

sutured shut with 6-0 Ethicon nylon suture. Methoxyfluorane was used

as a short term anesthetic agent with excellent results. Data were

recorded shortly after the bat exhibited normal locomotor behavior and

within three hours of surgery. Extended delays between the time of

implantation and recording sessions resulted in less than satisfactory

results, apparently as a result of toxic reactions between the copper

electrode material and adjacent tissues. Inspection of one

implantation that had been in place for 24 hours showed significant

tissue necrosis in the muscle for a 2 mm diameter around the electrode

tip. Only one muscle was studied per experiment. Each implant was

visually inspected at the conclusion of each recording session to

verify that the electrode remained in place. Each animal was allowed

to recover for at least one week before being used in further


Normal flight was studied with single frame analysis of films of

normal horizontal flight sequences. Single film frame images were

projected with an L and W motion analyzer (Model 224A) and were traced

to ascertain wingtip motion and joint position in the wing. Vertical

displacement of the wingtips was used to determine the turnover points

of adduction and abduction. The beginning of a downstroke was treated

as the beginning of a wingbeat cycle and was assigned a numerical value

of 0.000 (Figure 1, frame B). A complete wing cycle included values

between 0.000 and 1.00, with 1.00 representing the final frame of an

upstroke (Figure 1, frame M). The transition between downstroke and

upstroke (Figure 1, frame H) and periods of muscle activity are

expressed as a mean percentage of a complete cycle.

Estimation of aerodynamic parameters was accomplished by measuring

the dimensions of the wing and body of two fresh specimens of Artibeus

jamaicensis. The mass of each bat was obtained to the nearest 0.5 g.

Wing area was estimated by drawing the outline of the body and fully

outstretched wing of each bat on paper. The cut-out outline of the

airfoil was then weighed to the nearest 0.0005 g and then compared to

the mass of a .100 m2 square of the same paper to calculate actual

wing area. Wingspan was measured along a line between the two

wingtips. Aspect ratio is the ratio of length to width of the

airfoil. The equation for aspect ratio is b2/A (Norberg, 1981).

Wingloading is the ratio of body weight to wing area, W/A (Norberg,




The temporal relationship between the downstroke and upstroke is

critical in the interpretation of EMG data. In part, this relationship

reflects whether or not the activity of a given muscle is coincident

with or occurs before or after a particular event, such as the

beginning of a downstroke. During all filming sessions, flight was

observed to detect any grossly abnormal flight behaviors. All flight

sequences exhibiting turning behavior or altitude loss were eliminated

from consideration. The wingbeat cycle of Artibeus appeared to be

conservative during slow horizontal flight, based upon unaided visual

observation as well as initial study of the film records. However,

analysis of the film revealed slight differences in the flight of

restrained and unrestrained animals.

During unrestrained flight in two bats, the downstroke encompassed

55.1 percent of the wingbeat. In contrast, bats that carried an

electrode and plug assembly exhibited a mean downstroke of 49.44

percent. During unrestrained locomotion, the bats averaged 10.08

wingbeats per second. Records analyzed from four EMG experiments

(clavotrapezius, spinotrapezius, clavodeltoideus, and infraspinatus)

demonstrated a mean wingbeat frequency of 10.45 wingbeats per second.

S- 0 C
-0 Ta
4J O O

4-1 i I 1 w
4 t) 000 3
o 41 :3 *o
4 -1 H
0 -
Sil4- Q)

0 41 H 0
-cO ~~O

0 S m
; -1 0 0
m0 C. "
A e r-I bo 4 o
i 0 o c C -O

> "a U $-4 z0

0 0 : 0
o g o "0
,a > 0 0

40 4) 0 O :30
0(U il 6 CS

I 4-1 0 W u
S .-C 4 '- 0
*) 0 1-1 -

w0 Po 0 0
1 0 4J o en
p Z C C- r dl

W1 c 6a W 4
I- CB -) r. (U C H





" I


x 0



* '


400 cd
*1-4 >

4-1 O

c- 0 a)

4 -4
4-4 0

00 4-1
4.j 41
to-i o
94-4 4J
0) Q)

.-0 40

a) .u -H

C oo
o W

0M O
0dQ o
M a ) 4-1

C *4-1 Q 0
0) ) UD

00 w >

*) CO 0

rI 4
T= < o




.i o

/ e 0

x O


\ 0



/ --

/ /-* ^ \ ^~~~1

Figure 4.--Wing and body movements plotted relative to still
air during slow forward flight in Artibeus jamaicensis. The
bat flew from left to right at a velocity of 2.15 m/sec.
(A) The movements of the wingtip are plotted relative to the
body. (B) The movement of the wrist is plotted relative to
the body.


/ I
/ I
/ \

5 cm

Table 1. Aerodynamic parameters of the wing in two fruit bats,
Artibeus jamaicensis. All measurements are recorded in
meter-kilogram-second units.



wing span (b)

wing area (A)

aspect ratio (b2/A)

wing loading (W/A)

0.04 kg

0.40 N

0.39 m

0.02 m2


16.74 N/m2

0.03 kg

0.32 N

0.32 m

0.02 m2


15.47 N/m2

Analysis of one sequence of unrestrained flight is provided to

illustrate airspeed and movement of several parts of the body and wing

(Figures 2-4). Plots of head movement exhibited forward velocity

ranging from 2.19-2.85 m/sec. At the beginning of the downstroke, the

wingtip was positioned above and caudal to the center of the body. The

wing was adducted, but moved with an initial caudal sweep of the

wingtip relative to the center of the body. After approximately 15

msec, the adduction of the wingtip exhibited an increasing cranial

component of movement. The wrist moved in synchrony with the wingtip

but travelled through a smaller excursion. The wingtip continued

moving ventrally and cranially through 62.5 percent of the wingbeat

(Figure 2, Symbols 2-12). The wingtip and wrist synchronously showed

abduction at the beginning of the upstroke. There was a cranial

movement of both elements for approximately 5 msec. Subsequently,

abduction of the wingtip and the wrist included a caudal component of

movement for about 26 msec until the beginning of the flick phase.

During the flick phase (Figure 2, Symbols 18-21) the outer portion of

the chiropatagium was rapidly pronated. Cranial movement of the

wingtip occurred during the flick phase, which lasted approximately 5

msec, or until the beginning of the downstroke. During the entire

upstroke, movement of the wrist was in a dorsal and caudal direction.

Thus, during the wingbeat cycle the wrist followed an elongated

elliptical path. The path of the wingtip was best characterized as a

tight and complex figure eight when viewed laterally.

Several measures are used in the description and comparison of the

aerodynamic properties of Artibeus jamaicensis. Based on two specimens

obtained in Haiti, the following means were obtained for these

parameters: mean wing area, inclusive of the body, 0.02 m2;

wingspan, 0.38 m; aspect ratio, 6.67; wingloading, 16.11 N/m2.


The wing and flight characteristics of Artibeus jamaicensis are

generalized in comparison to many other bats. Analysis of the flight

kinematics and wingshape demonstrates that Artibeus is capable of slow

flight in the forest canopy but not the faster, more maneuverable

flight associated with the pursuit of insects among dense vegetation or

in open airspace.

The data obtained during EMG experiments on Artibeus represent a

departure from the mean downstroke times observed during the

unrestrained flight sequences. These EMG flight data are comparable to

downstroke phase relationships reported during EMG studies in Antrozous

where a range of 39.7 to 52.3 percent was reported for the downstroke

period (Hermanson and Altenbach, 1983). The downstroke of flying

animals lasts longer than the accompanying upstroke (Aymar, 1935;

Brown, 1963; Norberg, 1976; Brandon, 1979). Flying birds or bats can

be seen to oscillate in a vertical plane during flight, losing a slight

amount of altitude during each upstroke phase, and regaining altitude

during the subsequent downstroke (Eisentraut, 1936). The upstroke does

not provide sufficient lift to counteract the effect of gravity except

in certain species that employ a hovering mode of flight (Norberg,


The flight speeds observed in the present study are similar to

those observed in Plecotus auritus by Norberg (1976). In her study,

the bats were filmed while flying unrestrained within a room and flight

speeds ranged between 2 a 3 m/sec. Higher velocities were exhibited by

Noctilio albiventris (3.51-10.38 m/sec) and Tadarida brasiliensis

(4.33-9.40 m/sec) while flying in a wind tunnel (Brandon, 1979).

Leptonycteris sanborni, a facultative hovering bat, flew between 2.0

and 6.0 m/sec in a wind tunnel study. The higher speeds reported in

the wind tunnel studies were not duplicated in the present study.

Predictable flight, suitable for EMG analysis, was facilitated by

filming the bats shortly after takeoff and before they reached maximum

velocities. Also, it is not known if Artibeus is capable of flying up

to 4.47 m/sec as was generalized for many bats (Hayward and Davis,


The path of the wingtip during the wingbeat cycle changes with

increasing airspeed. During slow flight in Plecotus auritus, the

wingtip path was directed upwards and caudally relative to still air

during the upstroke (Norberg, 1976). At higher velocities (2-2.5

m/sec), the upstroke of Plecotus was directed almost vertically, and at

higher velocities (greater than 3.0 m/sec) exhibited a cranial and

dorsal movement relative to still air. The same tendency was present

in Artibeus flying between 2-2.85 m/sec. The wing was directed

caudally during the upstroke at low speeds (Figure 2) and more

vertically during the upstroke at higher velocities (Figure 3). If

wing movements are different at several airspeeds, these relationships

must be considered when evaluating EMG recordings. The difference in

wing movements may be correlated with differences in muscular activity.

Flying animals have been classified according to several

descriptors of wingshape including the aspect ratio (Saville, 1962).

Vaughan (1959) discussed the relationship between aspect ratio and

foraging habits. Struhsaker (1961) was the first author to describe

the aspect ratios of a diversity of chiropteran wings. Aspect ratios

were also provided for bats by Farney and Fleharty (1969), Findley et

al. (1972), Lawlor (1973), Smith and Starrett (1979), and Norberg

(1981). My calculation of an aspect ratio for Artibeus jamaicensis,

6.67, is slightly greater than but comparable to the value of 6.36

obtained by Norberg (1981). In general, bats with low aspect ratios

are common in heavily forested habitat, whereas high aspect ratio bats

are commonly associated with flight in open, uncluttered airspace

(Findley et al., 1972). For example, Vaughan (1959) compared the

high-altitude flight of Eumops perotis with that of swifts. These bats

fed high over the forest canopy at high airspeeds. Vaughan reported

that the aspect ratio of Eumops was 11.9, the highest value among the

three species that he studied. Intermediate aspect ratio wings are not

well correlated with any particular behavioral repertoire or foraging

environment (Norberg, 1981). Wing shape in Artibeus probably evolved

Figure 5.--High- and low-aspect ratio wingshapes. Tadarida
brasiliensis (Molossidae) exemplifies the long and narrow
outline of a high aspect ratio wing. In contrast, the wing
of Artibeus jamaicensis exhibits a lower aspect ratio,
characterized by a broad and stubby appearance.



in response to the needs of the bat in carrying large food items and

transporting infants. Artibeus females have been observed carrying

young weighing 10.0 g (personal observation). Low aspect ratio wings

promote increased lift at lower flight speeds (Findley et al., 1972),

permitting more careful scrutiny of potential food or of potential

feeding sites. Maneuverability does not appear to be characteristic of

the flight pattern of Artibeus in comparison with some more agile

bats. This was supported by repeated netting success while trying to

collect bats in mist nets (personal observation). In contrast, other

phyllostomids such as Micronycteris have low aspect ratio wings (Smith

and Starrett, 1979) and were frequently observed eluding my nets during

field work in Venezuela.

Wing loading is an estimate of the relationship between the weight

of a bat and the lift-producing potential of a wing. To fly slowly,

which would be adaptive to an Artibeus investigating a fig tree as a

potential food source, two strategies might be employed: the species

might exhibit low body weight or a greater airfoil surface area

(Findley et al., 1972). My data for wing loading in Artibeus were

slightly less than the 16.65 N/m2 reported by Norberg and greater

than the 15.94 N/m2 reported by Smith and Starrett (1979). Norberg

found that the wingloading of frugivorous microbats (this included most

phyllostomid fruit eaters) was intermediate between the insectivorous

Molossidae and Vespertilionidae. The molossids of equivalent weight

have high aspect ratios and must maintain high airspeeds to avoid a

stall (Vaughan, 1966; Norberg, 1981). In contrast, the vespertilionids

generally have a large wing area for a given body size (Norberg, 1981)

and fly at relatively lower airspeeds than similar-sized molossids

(Hayward and Davis, 1964; Brandon, 1979). Thus, the wing of Artibeus

represents a compromise between the two strategies suggested by Findley

et al. (1972). The bats have large body weights but have not

substantially increased wing surface area like vespertilionids have.

This combination leads to reduced maneuverability, different from the

butterfly-like flight observed in some vespertilionids. Also, I have

never observed Artibeus hovering. The bats tend to fly directly along

aerial pathways to and from a feeding tree, and fly past a tree several

times when investigating or searching for food.

In conclusion, the wing of Artibeus represents a generalized

condition relative to certain other bats. I suggest that this reflects

the heterogeneous nature of the airspace through which the bat must

travel every night. A high aspect ratio wing is not suited to travel

through the cluttered environment of most neotropical forests. Also,

at slow airspeeds, the larger lift production that accrues from a low

aspect ratio wing permits the bat to forage upon and to transport large

energy-rich foods. A secondary consequence of the large body mass and

low aspect ratio of Artibeus may relate to post-natal development and

maternal care of the young bats.




The clavicle in five specimens had a mean length of 16.10 mm and a

mean diameter at the midshaft of 1.15 mm. The bone articulates

proximally with the manubrium by a synovial joint. Distally, the

clavicle articulates with the scapula both by a synovial joint at the

base of the coracoid process and by a ligamentous connection with the

tip of the acromian process. Viewed cranially, the bone is straight,

with slight lateral curvature near its distal end. Viewed laterally,

the bone has a slight caudal curvature at the distal end. The distal

one-third of the cranial border of the bone is expanded for insertion

of part of the subclavius muscle. In lateral view the distal end of

the bone lies cranial relative to the proximal end.


The scapula is a thin, relatively flat bone positioned over the

dorsal surface of the cranial part of the ribcage. It articulates

proximally with the clavicle as described above. The scapula














.0 '
L -,

o w

3 E

z 0

*C "





Figure 7.--Cranial view of the right pectoral girdle.
UF 16265.

acromion process


glenoid fossa

coracoid process

craniomedial flange

) -.- caudal angle

.-- subscapular fossa

----------- clavicle

manubrium ---

5 MM







.. o o

0o E M

7o 7

% %
0 0 c E 0

O_ O
L 0 -

0 C c \

o 0 0 E

.. ,% .

o / I"

0 ,I

1 I
0 0- I
0 a

C 0
P O, Q Q

articulates with the humerus at the glenoid fossa, which forms the

shoulder joint. The dorsal surface of the scapula is divided into two

major subdivisions by the scapular spine. Cranially, the supraspinous

fossa forms a small flat shelf immediately cranial to the scapular

spine. Caudal to the scapular spine is the infraspinous fossa which

is, in cranial to caudal sequence, divided into three subdivisions:

craniomedial facet, intermediate facet, and caudolateral facet. The

ventral surface of the scapula is the subscapular fossa, composed of

two concave facets. The cranial subscapular facet corresponds with the

contours of the combined supraspinous fossa and craniomedial facet of

the dorsum. The caudal facet of the subscapular fossa corresponds with

the contours of the combined intermediate and caudolateral facets of

the dorsum. There are two borders on the scapula. Medially, the

vertebral border lies in a sagittal plane extending from the caudal

angle to the scapular spine. Cranial to the scapular spine, the

vertebral border curves laterally and becomes continuous with the

craniomedial flange of the scapula. On the lateral aspect of the

scapula, the axillary border is convex, coursing cranially from the

caudal angle to the glenoid fossa. The infraglenoid tubercle is a

small projection from the axillary border 1-5 mm caudal to the glenoid

fossa. The supraglenoid fossa is slightly concave and lies between the

glenoid fossa and the dorsal base of the coracoid process. This

surface abuts with the greater tubercle of the humerus when the wing is

fully abducted.

Several processes are evident on the scapula of Artibeus. The

cartilaginous expansion is triangular and projects from the caudal

angle and curves ventrally. The acromion process projects cranially

and dorsally from the scaular spine, and curves cranially. A 2 mm long

flared part of the acromion process lies dorsal to the clavoscapular

articulation. The coracoid process extends cranially and ventrally

from the supraglenoid region of the scapula and curves laterally to

terminate 2 mm ventral to the glenoid fossa.


The humerus averaged 31.90 mm in length (measured between the

articular surfaces) and 1.98 mm in diameter (along the craniocaudal

axis) at the midshaft. In cranial view the distal end of the shaft

curves slightly lateral. The shaft is unmarked. Both ends of the

humerus are modified to form articular surfaces and to accommodate

muscle attachments.

The shoulder joint is formed in part by the head of the humerus.

The head is a convex elliptical surface that articulates proximally

with the glenoid fossa. The greater tubercle lies lateral to the head

and projects 1 mm proximally. The greater tubercle provides attachment

for the suprascapular muscles and abuts with the supraglenoid fossa of

the scapular when the humerus is fully abducted. This lock occurs when

the humerus is abducted about 32 degrees dorsal to the horizontal plane

of the scapula. The lesser tuberosity also extends slightly proximal

to and lies on the medial aspect of the head. Several ridges extend

Figure 9.--Medial view of the left humerus. UF 16265.

greater tubercle

lesser tubercle

head ---------

lateral ridge ------'

medial epicondyle -----------

apex of pectoral crest

pectoral crest


5 MM

Figure 10.--Articular surfaces of the humerus. (A) Caudal
aspect of the proximal left humerus. (B) Cranial aspect of
the distal left humerus. UF 16265.

greater tubercle

lateral ridge



medial process

,- lesser tubercle

.-- ------- medial ridge

Sepicondylar ridge

lateral epicondyle





distally from the lesser and greater tuberosities. From the lesser

tuberosity, the medial ridge extends 6 mm along the proximal medial

border of the humerus and serves as the point of attachment of the

teres major and latissimus dorsi muscles. A second indistinct ridge

crosses the proximal surface of the humerus and courses along the base

of the ventral surface of the pectoral ridge. This ridge is the site

of insertion of the pectoralis abdominalis. The lateral ridge is a

small indistinct ridge on the lateral surface of the humerus courses 7

mm distal to the greater tuberosity. This forms the dorsolateral

border of the origin of triceps brachii, lateral head. The most

prominent ridge from the greater tuberosity is continuous distally with

the pectoral ridge, a large flat projection on the lateral aspect of

the humerus. The pectoral ridge extends 8 mm along the lateral edge of

the humerus. The widest portion of the pectoral ridge, 2-4 mm distal

to the head, is termed the apex. The pectoral ridge provides insertion

sites for the pectoralis muscles medially, and the deltoid muscles


At the elbow the radius articulates with the cranial surface of

the humerus. The ulna articulates along the distal and caudal surface

of the humerus. The capitulum and trochlea are fused cranially to

present a synovial surface for the humeroradial articulation. The

trochlea extends around the distal and caudal surface of the humerus to

provide an articular surface for the ulna. Medial to the trochlea and

composing one-third of the distal end of the bones is a medial

process. This region provides an attachment surface for several flexor

muscles of the antebrachium. On the lateral end of the capitulum is

the spool-shaped lateral epicondyle. Extending from the lateral

epicondyle proximally onto the shaft of the humerus is the short

epicondyle ridge. The lateral epicondyle and epicondylar ridge provide

an area of attachment for several antebrachial extensor muscles. The

elbow articulation is not a pure hinge joint. Slight rotation of the

radius relative to the capitulum is evident in fresh specimens of



The radius is slightly curved along its length. In five

specimens, the mean length is 53 mm. At the midshaft, it is thickest

in the dorsoventral axis (1.8 mm) to resist bending forces associated

with the wingbeat cycle. The proximal articular region is thickened to

provide a surface for articulation with the humerus and to provide a

strong site for attachment of the biceps brachii. The distal articular

region is wide and flat. This provides a structural mechanism for

separation of the tendons of insertion of the antebrachial muscles and

increases the mechanical advantage of several of these muscles as they

cross the carpus.

The proximal articular surfaces is oriented cranially and includes

a large concave fossa for articulation with the capitulum. Laterally a

smaller elliptical concavity articulates with the trochlea. A ridge on

the caudal aspect of the radius courses 5 mm distal from the elbow

joint. A ridge on the medial aspect of the proximal radius courses 3

mm distal to the elbow joint and provides the site for insertion of the

biceps brachii.

The styloid process and pseudostyloid process are small bumps on

the distal articular end of the radius but are not greatly developed in

bats. The pseudostyloid process is on the medial side of the distal

articular surface, analogous in position to the styloid process of the

ulna in other mammals.

The axis of the distal articular surface is laterally rotated

about 20 degrees from the orientation of the proximal articular

surface. The distal articular surface presents a concave surface for

articulation with the proximal carpal bones. When the wing is

laterally outstretched, the radiocarpal joint permits flexion and

extension in a mediolateral plane and no movement in the dorsoventral

plane. This restriction braces the joint against displacement caused

by the force of the airstream.


The ulna is reduced in bats to a length approximately one-half the

length of the adjacent radius. The proximal articular surface

comprises the cranial surface of the olecranon process and articulates

with the caudal extension of the trochlea of the humerus. An

interosseous membrane is found between the radius and ulna. The ulna

tapers towards its distal articulation with the radius and presents no


Figure 11.--Dorsal view of the radius and ulna.





10 mm


The diminutive size of and the tightly bound joint cavities

between the bones of the carpus make it difficult to adequately

describe the bones of this region. A perspective on the carpal anatomy

can be obtained by studying Figure 12. The carpus of Artibeus includes

eight bones as is typical of many mammals.

The proximal row of carpal bones includes the scaphoid, lunar, and

triquetrum. The largest carpal bone, the lunar, articulates proximally

with the radius and distally with the trapezoid, trapezium, and

capitate. On the lateral aspect of the distal articular surface of the

radius, the triquetrum (-cuneiform of Vaughan, 1959) is bound to the

radius and forms the caudoventral border of the canal traversed by the

antebrachial extensor muscles. The scaphoid is a small bone located

craniodorsally in the carpus along the cranial surface of the lunar.

The scaphoid, unlike the lunar and triquetrum, is not bound to the

distal surface of the radius.

The distal row of carpal elements includes the trapezium,

trapezoid, capitate, hamate, and pisiform. The trapezium articulates

with the proximal bases of metacarpals I and II, and with the

craniodistal surface of the lunar. The trapezoid is compressed

craniocaudally and articulates with the proximal base of metacarpal

II. The capitate (=magnum of Vaughan, 1959) is a large bone that has

an hourglass shape when viewed dorsally. It articulates with the

proximal bases of metacarpals II and III. Lateral to the capitate, the

Figure 12.--Dorsal and ventral views of the carpus.

styloid process






mc II

mc III



mc I

mc II


mc III

mc IV

mc V


Ic I

mc IV

', 'triquetrum


mc V

5 mm

hamate uniformom of Vaughan, 1959) articulates distally with the bases

of metacarpals IV and V, and proximally with the lunar and triquetrum.

The pisiform is a small rod-shaped bone on the palmar surface of the

carpus and is firmly bound to the ventral surface of the capitate and


The distal part of the wing is supported by digits II through V.

The metacarpals diverge within the wing membrane and are the largest

elements in each digit. Movement at the carpometacarpal joints of

digits II through V is restricted to flexion and extension. Flexion of

the carpometacarpal joints brings the digits parallel to and into

juxtaposition with the radius. The metacarpophalangeal and

interphalangeal joints permit only palmar flexion or dorsiflexion.

These movements adjust the camber of the wing during the wingbeat

cycle. Digit II has one phalanx. Digits III through V have three

phalanges. Digit I, the pollex, is unique because of the range of

motion possible at the carpometacarpal joint. Besides flexion and

extension, the pollical carpometacarpal joint permits the extensive

rotation necessary for use of the pollex in several functions: the

pollex is a prehensile organ used during landing maneuvers or while

manipulating food. The second phalanx of the pollex bears a claw. The

proximal and second phalanges of the pollex are both free of the

leading edge of the dactylopatagium minus and the propatagium.


Trapezius Group

Clavotrapezius and Acromiotrapezius

Form. The clavotrapezius originates by a fibrous attachment on

the dorsal surface of the seventh cervical vertebrae. Fibers of

acromiotrapezius arise from the dorsal surfaces of thoracic vertebrae

one through six. There is no apparent separation between adjacent

fibers of the clavotrapezius and acromiotrapezius. Insertion of the

clavotrapezius is by fibers on the distal 2.5 mm of the craniomedial

surface of the clavicle. The acromiotrapezius inserts separately from

the clavotrapezius: insertion is by fibers convergent upon the

anterior and dorsal surface of the acromion process and on a muscular

raphe caudal to the acromion process. Innervation of both muscles is

by cranial nerve XI (spinal accessory nerve).

Comparative aspects. The clavotrapezius and acromiotrapezius

muscles have been treated as two separate muscles in the Phyllostomidae

(Macalister, 1872; Walton, 1967), and also in the specialized vampire

bats (Altenbach, 1979). In contrast, the condition observed in many

other bats, including phyllostomatids, demonstrated fusion of these two

bellies (Strickler, 1978; Vaughan, 1959, 1970a). In the three species

he studied, Vaughan (1959) noted the fusion of the two bellies only in

Macrotus (Phyllostomidae). Macalister (1872; p. 138) most likely made

reference to the clavotrapezius of Artibeus sp. when he described "a

semi-detached upper slip passed from the two lowermost cervical spines

to the outer fifth of the clavicle." The two muscles are

distinguishable in Artibeus only by a fascial division coursing from

the C7-T1 articulation to the distal end of the clavicle. Despite

their separate insertion, I consider the two muscles to be fused and to

share a similar function. The clavotrapezius and acromiotrapezius form

a flat sheet of parallel fibers, thickest at the cranial end

(clavotrapezius portion).

Functional aspects. The clavotrapezius and acromiotrapezius

exhibit a similar pattern of EMG activity during the wingbeat cycle.

Clavotrapezius has a major period of EMG activity during the transition

between the late downstroke and early upstroke. A second period of

activity occurs during the middle to late upstroke phase, but the

myopotentials have less than one-half of the amplitude observed during

the earlier EMG burst. Acromiotrapezius also has two periods of EMG

activity, but both exhibit similar amplitude and frequency. The first

period of activity in acromiotrapezius precedes clavotrapezius activity

by approximately 10 m/sec. Both muscles are anatomically positioned to

effect the upstroke by drawing the scapula medially (Vaughan, 1959;

Strickler, 1978; Altenbach, 1979). The primary activity of each muscle

contributes to this upstroke function. The secondary burst of activity

possibly stabilizes the scapula in response to the adductor activity of


4 -4

0 01

1-1 ra-


4o c

> o

m cc


= CU

1 0
CO rb
o c

0) 414 M
00 m 0
H 0 ca

1 0 m 10
Cm C
r-1 (
tf rl -

a>uI co
F0 (X


-C c
o 00


0 0

- 0.
o co
.- 0.
a E
S Eo
0 c0


a o
co E
- 0
E !-


C O "-:

-0 "0 CO



0 0


oE r.
+- V)



z 10
(DO0 0 0 0
0) 0 'O 1 OC

- ar-H r-*
*i > U CU C
S> 0 Q)
0 0Ci 0 m r.

U) o 04-40Cu
a0 0 r 0
- 4u -i Cd
cE *o ,-q r) o
-H C.) (d r,- r -L
C- O I C
ca"0 *H *c

0 1 0 C

, co w > H w
o c 0 c 0 r- 4 J.
0 4-1 U 0
1 4 U 0c an)
3 u 1o Q) c au

o *c o ( 4 c1 a)
w -0 0 U 0 0
0 *0 ol c
MO *r CO rl 0
z m- 03 0
*4 D QI CO

t 0 > *H i
O 0 -H C*rI 0
C U 41 E 4-4 4-)
4) C: a) 4 0

*H -q VHW
i-l *O CO

i *rl -l O 0 f 0
I 2r cU rC0

* i0l 0 0
7 i-lc Tl 3 '-c
tIc !tl
1-1 0B 0i -i!< J





1o 0


- 4
S, | 0 C

C M 0 .I 7 c
a V- co C -

-2 c o ( n n o
> 0 0 C C 0 0 0 0 o 0
0 L O 0 > 0
0 Co C) Cc .- CL N 0
Vec o
_ E E E 0 0 2 0
vo o o C o Co C

< S0 V5

Table 2.--Electromyographic data for activity patterns of the shoulder
musculature in Artibeus jamaicensis during slow flight. N= number of
wingbeats analyzed per muscle. Mean duration of the downstroke is
expressed as a percentage of the wingbeat cycle duration. Mean times
of muscle activity onset and termination are recorded in relation to
the wingbeat cycle from 0.00 to 1.00, and one standard deviation is

Downstroke Onset Termination
Muscle N Duration Time (S.D.) Time (S.D.)



Latissimus dorsi

Teres major




Triceps brachii

12 0.515

9 0.477

13 0.428

3 0.412

10 0.536

10 0.532

11 0.481

3 0.594

0.332 (.099)
0.894 (.030)

0.236 (.033)
0.794 (.080)

0.329 (.028)
0.869 (.035)

0.259 (.022)
0.922 (.046)

0.355 (.056)
0.799 (.041)

0.315 (.053)
0.894 (.051)

0.309 (.030)
0.001 (.115)

0.564 (.017)

15 0.513 0.429 (.069)

0.631 (.045)
0.092 (.047)

0.572 (.104)
0.903 (.104)

0.644 (.025)
0.098 (.025)

0.519 (.016)
0.077 (.045)

0.628 (.040)
0.081 (.055)

0.556 (.062)
0.103 (.065)

0.608 (.030)
0.114 (.080)

0.741 (.053)

0.714 (.074)


Table 2.--continued

Downstroke Onset Termination
Muscle N Duration Time (S.D.) Time (S.D.)

Clavodeltoideus 19 0.544 0.853 (.059) 0.135 (.027)
0.481 (.038) 0.617 (.039)

Serratus ventralis 11 0.484 0.823 (.036) 0.091 (.058)
0.515 (.058) 0.599 (.109)

Pectoralis 7 0.526 0.781 (.027) 0.094 (.086)

Spinotrapezius 7 0.452 0.254 (.031) 0.672 (.032)
0.784 (.020) 0.123 (.047)

Supraspinatus 6 0.518 0.706 (.263) 0.142 (.084)

Infraspinatus 7 0.449 0.346 (.039) 0.607 (.046)
0.942 (.059) 0.120 (.039)

Subscapularis 10 0.450 0.420 (.060) 0.174 (.033)

the early downstroke, or causes active lengthening of the pectoralis

fibers during their early period of activity. Similar passive

lengthening, or "eccentric contractions," increased force production of

several locomotory muscles in other animals (Cavagna et al., 1977;

Heglund et al., 1979; Goslow et al., 1981). The biphasic activity

observed in Artibeus contrasts with the single EMG burst observed in

Antrozous (Hermanson and Altenbach, 1983). The heavier wingloading of

Artibeus may necessitate the extra muscular activity. In conclusion,

both clavotrapezius and acromiotrapezius show abductor activity, but at

slightly different times in the wingbeat cycle. The secondary activity

in each muscle stabilizes the scapula against the force of the



Form. The spinotrapezius originated by fibers on the dorsal

midline over lumbar vertebrae two through three. It inserts by an

aponeurosis on the vertebral border of the scapula, extending from the

caudal angle to the point 2 mm anterior to the scapular spine. The

spinotrapezius is a thin, parallel-fibered muscle. Based upon its

anatomy, the muscle tips the vertebral border of the scapula ventrally,

and retracts the scapula (Strickler, 1978). Innervation is by the

spinal accessory nerve.

Functional aspects. Spinotrapezius exhibits biphasic EMG

activity. Both periods of activity have approximately equal amplitude

and frequency characteristics. One period commences at 0.254 and

terminates at 0.672: This contributes to the overall abduction of the

wing along with the activity of clavotrapezius and acromiotrapezius. A

second period of activity commences at 0.784 and terminates at 0.123 of

the following wingbeat. Spinotrapezius is quiescent only during about

24 percent of the wingbeat, an observation similar to EMG data from

Desmodus (Altenbach, 1979), but in contrast to the shorter, single

activity period observed in Antrozous (Hermanson and Altenbach, 1983).

The extent of EMG activity throughout the wingbeat suggests a multiple

function for spinotrapezius as proposed by Vaughan (1959):

stabilization of the scapula and abduction of the wing and scapula are

likely functions.

Costo-spino-scapular Group

Serratus Ventralis Cervicis

The serratus ventralis cervicis was previously called the elevator

scapulae in the chiropteran anatomy literature.

Form. The serratus ventralis cervicis originates from fibrous

attachments along the dorsolateral aspect of the transverse processes

of cervical vertebrae three through six. The muscle has a fleshy

insertion on the ventral surface of the craniomedial flange of the

scapula, and along the vertebral border of the scapula craniad of the

scapular spine. The muscle is composed of two parallel-fibered slips:

the cranial slip originated from cervical vertebrae three and four, and

the caudal slip from five and six. Innervation is by a branch of the

dorsal scapular nerve, and by cervical nerve five.

Comparative aspects. In many mammals, the serratus ventralis is

composed of a continuous sheet of muscle, including fibers that

originate from the cervical vertebrae serratuss ventralis cervicis) and

from the lateral surface of the ribcage serratuss ventralis thoracis).

The terminology and comparative anatomy are discussed with the serratus

ventralis thoracis.

Strickler (1978) noted that although the cranial origin normally

extends to cervical vertebra five to seven the origin of serratus

ventralis cervicis varies in bats. Phyllostomoids were characterized

by a relatively large insertion, usually extending along the vertebral

border of the scapula caudal to the scapular spine. Walton (1967)

observed a similar insertion in Artibeus lituratus extending along the

"coracoid border" and vertebral border of the scapula to a point

immediately posterior to the spine.

Walton (1967) described the serratus ventralis cervicis of A.

literatus to be composed of three large slips. Macalister observed

only two slips in Artibeus sp. (Macalister, 1872). I concur with


Functional aspects. Strickler (1978) concluded that a correlation

existed between degree of humeroscapular "locking" observed in some

bats and the size of the serratus ventralis cervicis muscle. He

observed that this muscle was relatively enlarged while the rhomboideus

was relatively small in vespertilionids and phyllostomatids. In

contrast, the serratus ventralis cervicis was small and the

rhomboideus large in emballonurids and pteropodids. These two groups

possess poorly developed humeroscapular locking mechanisms (Vaughan,

1970a, 1970b; Strickler, 1978). Strickler suggested that the major

role of serratus ventralis cervicis is to stabilize the medial border

of the scapula against rotation about its longitudinal axis. No EMG

data are available to test this hypothesis.

Serratus Ventralis Thoracis

Macalister (1872) observed and described the serratus magnus and

divided the muscle into superior and inferior portions. The same

muscle was described by Vaughan (1959) under the name serratus

anterior. He also divided the serratus anterior into two parts, the

anterior and posterior divisions. Subsequent authors followed his

precedent for bats (Walton, 1967; Norberg, 1970; Strickler, 1978;

Altenbach, 1979; Hermanson and Altenbach, 1981, 1983).

Serratus Ventralis Thoracis, Cranial Division

Form. The origin is a fibrous attachment on the dorsolateral

surface of the first rib, first intercoastal space, and cranial surface

of the second rib, medial to the origin of subclavius. The muscle has

a broad fleshy insertion along the ventral surface of the craniomedial

flange of the scapula, lateral to the insertion of serratus ventralis

cervicis. At the origin, the belly is small and rounded. Distally the

belly becomes wider and flatter, being thickest along the medial

border. The muscle is innervated by a branch of the dorsal scapular


Figure 15.--Lateral view of serratus ventralis musculature.
In the upper figure, subclavius is intact. In the lower
figure, subclavius is removed to reveal the position of the
deeper serratus ventralis cervicis.

serratus ventralis


coracoid process


serratus ventralis thoracis,

...---- cranial div.
\" .... caudal div.


serratus ventralis thoracis,
.... cranial div.
S caudal div.


Comparative aspects. The cranial division of serratus ventralis

thoracis is large in hovering bats and in slow-flying bats (Strickler,


Functional aspects. Vaughan (1959) proposed that the cranial

division functioned as a wing abductor by pulling ventrally on the

craniomedial edge of the scapula. The cranial division is large in

hovering bats, the glossophagines, and slow-flying bats, the

rhinolophids (Strickler, 1978). In both groups, the wingbeat amplitude

is large and may include extensive tipping of the lateral border of the

scapula ventrad during the downstroke. The cranial division is

situated to initiate abduction of the wing by arresting outward

rotation and by initiating inward rotation of the scapula. No EMG data

are available to test this hypothesis.

Serratus Ventralis Thoracis, Caudal Division

Form. The muscle originates by fibers along the dorsolateral

aspect of ribs two through 11, adjacent and dorsal to each

costochondral junction. A small, separate slip originates from the

caudolateral aspect of the first costal cartilage, caudal to the origin

of subclavius. The fibers of this small slip course caudally to insert

on the middle one-third of the axillary border of the scapula. The

fibers of the larger portion of the caudal division insert by fleshy

attachment along the caudal three-quarters to seven-eighths of the

axillary border and on the caudal angle of the scapula. The caudal

fibers inserting upon the caudal angle constitute the thickest portion

of the muscle. Fibers are arranged in parallel. The fibers

interdigitate at their origin with the scaleneus muscles cranially and

the external abdominal oblique muscles caudally. Innervation of the

caudal division is by the long thoracic nerve, coursing along the

lateral surface of the muscle.

Comparative aspects. Strickler (1978) found that the caudal

division is relatively large in most phyllostomids and small in the

Pteropidae and most of the Verpertilionidae. He felt that the function

and size of the muscle vary with respect to the development of the

humeroscapular locking mechanism.

Functional aspects. The shoulder locking mechanism described by

Vaughan (1959) provides a mechanical link between the action of the

caudal division upon the scapula and the resultant adduction of the

wing. Thus, contraction of the serratus ventralis thoracis was

hypothesized to power the downstroke. Electromyograms obtained from

Artibeus indicate that the posterior division has a major period of

activity that commenced before onset of the downstroke and ended during

the early downstroke. A second burst of activity is of shorter

duration and occurs during the early upstroke phase. Hermanson and

Altenbach (1981, 1983) reported a single period of activity for this

muscle in Antrozous during the transition between the upstroke and

downstroke. In both Antrozous and Artibeus, the pectoralis muscles are

active prior to the major burst of the caudal division of serratus

ventralis thoracis. The caudal division and pectoralis are coactive as

adductors throughout the early downstroke.

The second period of EMG activity occurs after the upstroke has

begun. This burst provides a force antagonistic to activity in

clavodeltoideus and latissimus dorsi. Clavodeltoideus activity tends

to protract the humerus, and thus advances the scapula craniad relative

to the thorax. Latissimus dorsi exhibits its most pronounced activity

as an abductor and pronator of the wing, causing retraction of the

scapula and arm.

English (1978a, 1978b) presented data for the serratus ventralis

cervicis and serratus ventralis thoracis of cats indicating biphasic

activity patterns. He concluded that serratus ventralis thoracis is

important in transmitting the weight of the body to the forelimbs

during the stance phase, an observation suggested by earlier anatomists

working only with cadaverous material (Davis, 1949; Gray, 1968). Of

greater interest was English's observation that coactivation of

cervical and thoracic portions of serratus ventralis during the swing

phase effected shoulder extension and repositioning of the limb for the

following step. Jenkins and Weijs (1979) found no significant

difference between EMG activity of different slips of serratus

ventralis thoracis in Didelphis. During walking, serratus ventralis of

Didelphis was active during the propulsive phase. These authors

commented upon a secondary period of activity during the swing phase in

Didelphis, but did not discuss its role. A secondary phase of activity

was not recorded in Antrozous (Hermanson and Altenbach, 1983).

Repositioning of the limb for the downstroke may be effected in

Antrozous during the upstroke by the force of the airstream on the wing

(Vaughan, 1959; Norberg, 1976) and by the action of the dorsal shoulder

musculature (Hermanson and Altenbach, 1983). The greater wingloading

in Artibeus may require an additional burst of muscle activity.


Form. The rhomboideus originates by fibers on the dorsal surface

of the transverse process of cervical vertebrae seven, lateral to the

cranial fibers of clavotrapezius, through thoracic vertebra six,

lateral to the proximal fibers of acromiotrapezius. The muscle fibers

course obliquely, caudodorsally through the interscapular region to

insert on the entire vertebral border of the scapula caudal to the

scapular spine. Innervation is by the dorsal scapular nerve.

Comparative aspects. The rhomboideus of the Phyllostomidae was

characterized by Strickler (1978) as being relatively small. In

contrast, both Macalister (1872) and Walton (1967) described the muscle

as robust in the Phyllostomidae. I concur with Strickler's

interpretation in part because of the muscle mass data presented in his

arguments, whereas the other authors relied upon subjective

interpretations. The romboideus is a thin, flat muscle in Artibeus.

Strickler (1978) hypothesized that either rhomboideus or elevator

scapulae are well developed in bats depending on locomotor

requirements. In the Molossidae, he observed that the rhomboideus was

relatively large and adapted to rotating the caudal angle of the

scapula medially relative to an axis perpendicular to the dorsal

surface of the scapula. He speculated that this action is important in

terrestrial movements and not in the scapular movements associated with

the phyllostomid shoulder during flight.

Functional aspects. Based upon the attachments and orientation, I

deduce that rhomboideus pulls the caudal angle of the scapula

ventromedially. There are no EMG data with which to elaborate upon

this hypothesis.

Latissimus-subscapular Group

Latissimus dorsi

Form. The muscle originates from the superficial fascia overlying

thoracic vertebra eleven through lumbar vertebra four. At the origin,

the belly is thicker over thoracic vertebrae and grades into a thin

aponeurosis caudally over the lumbar region. Latissimus dorsi inserts

by a thin, 1 mm wide tendon 3-4 mm distal to the shoulder joint on the

medial ridge of the humerus. The tendon lies ventral to and on the

proximal edge of the fibrous insertion of teres major. Fibers in the

latissimus dorsi are arranged in parallel at the origin, but converge

distally on the tendon of insertion. Innervation is received from the

thoracodorsal nerve.

Comparative aspects. Strickler (1978) did not observe much

variation in the latissimus dorsi attachments of bats, except in the

Rhinolophidae. In several phyllostomid species studied, the origin

of latissimus dorsi ranged between thoracic vertebra 10 and lumbar

vertebra 5, but no correlation was established between the attachments

and taxonomy or locomotor modes. Strickler's muscle mass data

indicated that the muscle was largest in slow fliers, and average-sized

in hovering and moderately fast frugivorous bats.

Functional aspects. The action of latissimus dorsi for Artibeus

and for bats in general is confusing because of conflicting evidence

from several sources. Latissimus dorsi was described by Vaughan (1959)

as a pronator and flexor of the humerus, and also as an important

retractor during terrestrial locomotion (1970b). Strickler (1978)

added that the muscle might also serve in abduction of the wing during

upstroke movements. For Desmodus, Altenbach (1978) obtained EMG data

suggestive of a major role in humerus pronation and adduction, but not

in abduction. In Antrozous latissimus dorsi functions primarily as a

pronator during the downstroke (Hermanson and Altenbach, 1983).

The EMG data for latissimus dorsi in Artibeus are difficult to

interpret in comparison with the data obtained from other studies. In

17 wingbeats, latissimus dorsi exhibits a biphasic activity pattern,

with highest intensity EMG myopotentials observed during the late

downstroke to early upstroke transition period (onset mean was 0.329,

average onset of downstroke movements was 0.428). A second period of

activity, of shorter duration and with lower amplitude mypotentials,

occurs during the late upstroke and early downstroke. These patterns

are consistent in all wingbeats. These data contrast with the EMG

pattern reported for Antrozous (Hermanson and Altenbach, 1983).

Although the greatest activity for latissimus dorsi in Artibeus was

observed during the late downstroke, the muscle was quiescent in

Antrozous during the same period.

English (1978a) observed variability in the EMG patterns of

latissimus dorsi in cats. Although the muscle was characterized by

English as monophasic and active predominantly during the propulsive

phase (EI-E2), he observed inconsistent bursts of activity during the

swing phase (F). These inconsistent bursts are similar in timing to

the late downstroke activity observed in Artibeus. Jenkins and Weijs

(1979) reported monophasic activity in the posterior part of latissimus

dorsi of walking Didelphis, however, they noted a consistent biphasic

pattern for recordings made with electrodes located in the anterior

fibers of the muscle. Placement of electrodes in intermediate

positions resulted in "hybrid" EMG patterns: this region of latissimus

dorsi was monophasic during some step cycles and biphasic during

others. Tokuriki (1973a, 1973b, 1974) observed two distinct phases of

activity in the latissimus dorsi of dogs at all gaits. One phase

occurred with other flexor muscles during the F phase, contributing to

lift the limb off the ground. A second and evidently more intense

period of activity occurred throughout the propulsive phase. In

terrestrial mammals, the activity of latissimus dorsi appears to be

biphasic, although the most intense activity occurs during a different

period of the locomotory cycle than observed in Artibeus.

Based on my EMG data and on comparative data for other mammals, I

conclude that latissimus dorsi in both Artibeus and Antrozous

stabilizes the shoulder and pronates the wing throughout the

downstroke. Primary EMG activity occurs during the transition from

downstroke to upstroke. Assuming a contraction time of 25 msec (Burke,

1978; Hermanson and Altenbach, 1981), the action of this muscle burst

pronates the wing during the first third of the upstroke. Analysis of

our film records indicates that the wing is continuously supinated

during the first two-thirds of the upstroke (Norberg, 1976; Altenbach,

1978), an action effected by the dorsal abductor muscles and by the

force of the airstream (Vaughan, 1959). The large burst of activity in

latissimus dorsi may therefore be necessary to counteract the

supinators and stabilize the shoulder during the early upstroke. The

temporal activity pattern of latissimus dorsi indicates that the muscle

is primarily a wing abductor.

Teres Major

Form. The teres major originates on the lateral and dorsal aspect

of the caudal angle of the scapula, and along the caudal 3 mm of the

axillary border of the scapula, by fibers. The origin does not extend

to the cartilaginous expansion at the caudal angle of the scapula. The

muscle inserts along the medial ridge of the humerus, 3-6 mm distal to

the shoulder joint. The insertion is primarily a fibrous attachment,

adjacent to but not conjoined with the tendon of insertion of the

latissimus dorsi. All myofibers are arranged in parallel. Innervation

is provided by several branches of the subscapular nerve.

Comparative aspects. Vaughan (1959) observed that the teres

major in Macrotus (Phyllostomidae) is larger relative to the latissimus

dorsi, than in Myotis (Vespertilionidae) and Eumops (Molossidae).

Strickler (1978) found that both the teres major and latissimus dorsi

exhibit positive allometry with respect to body mass. Triceps brachii

(long head) and biceps brachii (short head) were the only other muscles

in which Strickler observed positive allometry. He hypothesized that

whereas musculoskeletal specializations or lift provided by the

airstream permitted isometry to prevail in the primary adductors, and

abductors, respectively, the muscles associated with powering the flick

phase of the upstroke cannot be assisted by outside forces. Therefore,

a greater degree of positive allometry is exhibited in these four

muscles than in other muscles. Strickler did not include Artibeus in

his sample but did study other phyllostomids, including Carollia and

Phyllostomus, representative of locomotor styles broadly comparable to


Functional aspects. Electromyographic data exhibit two phases of

activity. High-amplitude activity occurs during the late downstroke to

upstroke transition. A second burst of one-half amplitude activity

occurs during the late upstroke to downstroke transition. The

association of muscle activity with turnover periods of the wingbeat

cycle indicates these muscles play an important role in stabilizing the

shoulder joint. Previous investigators used dissection studies and

determined the role of teres major is responsible for pronation of the

wing and a flexion of the shoulder joint (Vaughan, 1959; Norberg,

1970, 1972; Strickler, 1978; Altenbach, 1979). Hermanson and Altenbach

(1983) reported EMG data for four wingbeats in Antrozous but made no

specific conclusions beyond the classification of teres major as a

bifunctional muscle. In Artibeus, the biphasic activity pattern is

observed but is significantly different from that observed in

Antrozous. The primary activity of teres major in Antrozous, during

the final stages of the downstroke, probably pronates the wing

synergistically with latissimus dorsi. Together, the two muscles

counteract the onset of supination effected by the abductor muscles and

airstream, and provide a smooth transition of wing movements between

the downstroke and upstroke.


Form. The subscapularis fills the entire subscapular fossa. The

muscle originates along the periphery and on the elevated ridges of the

subscapular fossa, and from the ventral surface of the cartilaginous

expansion of the caudal angle of the scapula. The fibers exhibit a

multipinnate arrangement, with two main tendons converging to the

insertion on the lesser tuberosity of the humerus. The insertion is

composed of a broad tendinous attachment, and by fibers coursing

lateral of the tendon. Innervation is received from the two

subscapular nerves, derived from the dorsal division of the brachial


Comparative aspects. The attachments of the subscapularis are

conservative in all bats (Strickler, 1978). The large size of the

subscapularis relative to body mass was noted by Macalister (1872) as

being larger than observed in other mammals. Vaughan (1970a)

elaborated upon this point and provided data for the mass of

subscapularis relative to the total mass of the primary downstroke

muscles. This value ranged from 16.1 to 19.4 percent. Overall, the

relative weights of these primary downstroke muscles are consistent in

all bats. The data from Strickler (1978) are in agreement with Vaughan

except for in the Pteropodidae and for the phyllostomid genus

Desmodus. Strickler commented that pteropodids were unique among bats

in having small subscapularis muscles and small ventral muscle masses.

It was not established whether the small muscle masses represent a

primitive condition for chiropterans, or result from functional

constraints. In contrast, the large subscapularis muscle of Desmodus

appeared to be related to the terrestrial habits of the vampire bat,

for which there are increased demands for shoulder extension and wing

adduction (Strickler, 1978).

Functional aspects. Electromyography of the subscapularis in

free-flying Artibeus indicates that the muscle remains active through

approximately 75 percent of the wingbeat cycle. High-amplitude EMG

activity is observed through approximately 26 percent of the wingbeat

cycle and coincides with the second one-half of the upstroke and first

one-tenth of the subsequent downstroke. This activity precedes

myopotentials recorded from the other two downstroke muscles, serratus

ventralis thoracis and pectoralis. Subscapularis was discussed as a

component of the rotator cuff complex of primates and other mammals

(Tuttle and Basmajian, 1978) and is important in stabilizing the

shoulder joint. This stabilization includes the role of holding the

articular surface of the humerus within the glenoid fossa. Hermanson

and Altenbach (1981, 1983) concluded that subscapularis was primarily

responsible for fine control and stability of the shoulder joint in

Antrozous. They also observed that the expression of continuous versus

episodic activity in subscapularis was facultative and therefore

dependent upon the nature of the flight maneuver being attempted at any

given time. Kovtun and Moroz (1974) also used electromyography and

observed that subscapularis preceded pectoralis activity and ceased

activity 6-10 msec after mypotentials terminated in the "chest muscles"

of Myotis myotis. It was not clear whether their data were pooled from

several wingbeats or represented only one wingbeat.

The timing of subscapularis activity does not support Vaughan's

hypothesis (1959) that subscapularis is the third muscle among the

primary downstroke muscles to commence activity during a wingbeat.

However, the high-amplitude activity fit well with the temporal pattern

observed in downstroke muscles: myopotentials commence approximately

25 msec before wing adduction begins and terminates during the early

downstroke. Low-amplitude myopotentials during the early downstroke

and early upstroke indicate stabilization activity at the shoulder

joint. Low-amplitude activity at the transition between downstroke and

upstroke, and during the early upstroke stabilizes the shoulder against

forces produced by the abductor muscles during this period. Because of

these two periods of activity and the roles that the muscle carries

out, adduction and stabilization, the subscapularis is classified as a

bifunctional muscle and serves primarily to stabilize the shoulder


Deltoid Group

The deltoideus of bats has long been recognized as a tripartite

muscle composed of three parts: clavodeltoideus, acromiodeltoideus,

and spinodeltoideus (Macalister 1872). Because of functional

specializations associated with each of these parts, I will describe

and refer to each part specifically. All parts are innervated by the

axillary nerve.


Form. The clavodeltoideus originates by fibers along the

expanded, distal three-quarters of the clavicle. The muscle fibers

course dorsally, proximal to the apex of the pectoral ridge of the

humerus, and insert near the proximal insertion of acromiodeltoideus.

Fibers are arranged in parallel.

Comparative aspects. Vaughan (1959) pointed out that the

clavodeltoideus is separated from adjacent fibers of the pectoralis

only with great difficulty. Visualization of the axillary nerve is

necessary to correctly identify muscle bundles belonging to the

deltoideus group. Strickler (1978) illustrated some of the variation

in extent of origin that is possible within the order Chiroptera and

thus the need to undertake careful dissection of this region.

Functional aspects. The functional analysis of clavodeltoideus is

discussed along with the spinodeltoideus.


Form. The acromiodeltoideus has a fibrous origin along the entire

tip of the acromion process and proximally, along the base of the

acromion process as it intergrades with the scapular spine. Insertion

is by fibrous attachment along the lateral ridge of the humerus, from 1

mm distal to the greater tubercle along the proximal one-third of the

humerus. The myofiber architecture is parallel with some fibers

diverging distally along the insertion. A thin fascial plane bisects

the muscle into anterior and posterior portions and extends distally

from the middle of the acromion process.

Comparative aspects. The primitive condition for the origin of

the chiropteran acromiodeltoideus is restricted to the acromion process

(Strickler, 1978). In advanced families, the origin of the muscles has

extended medially along the scapular spine, a position that places

fibers farther from the shoulder joint and that increases the

effectiveness of acromiodeltoideus as an abductor and flexor of the

shoulder (Strickler, 1978). In Antrozous, fibers of acromiodeltoideus

and acromiotrapezius attach to each other along the dorsal scapular

ligament (Hermanson, 1978). The situation in Artibeus, in contrast, is

for both acromiodeltoideus and acromiotrapezius to insert upon a solid

mass of bone, adjacent to the acromion process.

Functional aspects. The action of the acromiodeltoideus is

discussed along with the spinodeltoideus.


Form. The spinodeltoideus originates by fibers from the dorsal

scapular ligament (sensu Hermanson and Altenbach, 1983), from the

vertebral border of the scapula caudal to the scapular spine, and over

the cartilaginous plate at the caudal angle of the scpaula. Insertion

is by an aponeurosis on the medial aspect of the dorsal ridge of the

humerus, deep to the insertion of the acromiodeltoideus and immediately

distal to the insertion of teres minor. The muscle is parallel-fibered

and forms a broad but thin sheet over the caudal scapular region.

Vaughan (1959) commented that spinodeltoideus was composed of two parts

in Macrotus.

Functional aspects. The deltoideus muscle of man is a single

muscle mass that was divided by anatomists into an anterior and

posterior portion. Each division can be correlated with differential

EMG activity while performing several actions (Inman et al., 1944).

Similarly, Hermanson and Altenbach (1983) presented functional

interpretations for all three heads in Antrozous and discussed these

data with respect to published data for several terrestrial mammals.

The acromiodeltoideus and spinodeltoideus EMG patterns for Antrozous

were in general agreement with the hypotheses of earlier descriptive

studies (Vaughan, 1959, Strickler, 1978): the muscles functioned as

abductors of the wing. The clavodeltoideus, however, was active almost

in synchrony with the primary downstroke muscles. This observation,

and dissection studies (Norberg, 1972; Strickler, 1978), led Hermanson

and Altenbach to conclude that the muscle was an important humeral

protractor and also a humeral adductor.

During the propulsive phase in Antrozous, a lack of muscle

activity in acromiodeltoideus and spinodeltoideus contrasted with

observations of EMG patterns in the deltoideus of terrestrial mammals.

Deltoideus activity occurred during the stance phase in stepping cats

at several speeds of locomotion (English, 1978a). Biphasic activity

occurred in the deltoideus muscle in Didelphis: the largest amplitude

activity occurred during the swing phase, and a less intense period of

activity during the stance phase (Jenkins and Weijs, 1979). When the

acromioclavicular and spinal portions of the Didelphis deltoideus were

monitored simultaneously, the spinal portion exhibited intense activity

during the first one-third of the swing phase, while the

acromioclavicular portion was active during the final two-thirds of the

swing phase (Jenkins and Weijs, 1979). Recordings from the deltoideus

of Canis indicated continuous myopotentials throughout the step cycle

of walking, trotting, and galloping dogs (Tokuriki, 1973a, 1973b,


Data for the deltoideus of Artibeus provide contrasts and

similarities with the Antrozous observations. For example, the primary

(highest amplitude) EMG activity in the clavodeltoideus was evident

during the downstroke portion of the wingbeat. Myopotentials commenced

almost in synchrony with the pectoralis, contributing to wing adduction

but also to the anterior movement of the wing. This wing protraction

is an important aspect of the "Rudderflug" or rowing flight described

by Eisentraut (1936) and Norberg (1976). Low-amplitude activity of

brief duration occurs during the transition between downstroke and

upstroke and serves to stabilize the scapula. In contrast,

highest-amplitude activity in both the acromiodeltoideus and

spinodeltoideus occurred during the late downstroke and early

upstroke. These data therefore suggest an important role for

acromiodeltoideus and spinodeltoideus in wing abduction and agree with

observations in Antrozous (Hermanson and Altenbach, 1983). Both

acromiodeltoideus and spinodeltoideus were biphasically active. The

posterior part of acromiodeltoideus showed a low-amplitude phase of

activity roughly coincident with the primary burst of clavodeltoideus,

although of shorter duration and with less amplitude than in the latter

muscle. Spinodeltoideus exhibited low-amplitude activity during the

first one-third of the downstroke. These secondary EMG bursts in the

acromiodeltoideus and spinodeltoideus may stabilize the shoulder

against the powerful actions of the pectoralis.

Suprascapular Group


Form. Supraspinatus originates by fibers along the cranial and

medial perimeter of the supraspinous fossa, deep to the belly of

acromiotrapezius. The muscle fibers are arranged in parallel. This

muscle has a fibrous insertion on the tip of the greater tubercle of

the humerus. Innervaton is by the suprascapular nerve.

Comparative aspects. The supraspinatus exhibits little variation

in its attachments for all bat species (Vaughan, 1959, 1970a;

Strickler, 1978). The regression of supraspinatus mass against body

mass suggests that the size of this muscle is average to relatively

large size in slow-flying bats, and smaller in faster-flying species

(Strickler, 1978).

Functional aspects. The mammalian supraspinatus is considered one

of the rotator cuff muscles of the shoulder: its primary role is to

stabilize the glenohumeral articulation in concert with the

infraspinatus, subscapularis, and teres minor (Hollinshead, 1974;

Tuttle and Basmajian, 1978). For bats, Vaughan (1959) described the

general action of the supraspinatus to be shoulder extension and

humerus supination, a description accepted by subsequent authors (cf.

Strickler, 1978). Vaughan (1959) suspected that supraspinatus was most

effective as a shoulder extensor in the phyllostomid Macrotus, as

compared to Eumops and Myotis. The shoulder joint of Macrotus allows

more freedom of movement than is observed in Eumops and Myotis. Both

of the latter species had bony locking mechanisms that restricted

movement of the humerus in both the dorsoventral and craniocaudal

directions. Thus, the humerus of Macrotus could be extended cranial

to a transverse plane through the shoulder, while the humerus of Eumops

could only be extended forward to this plane.

Electromyographic data for the supraspinatus is available for only

six wingbeats, primarily because the small size and deep position of

the muscle make the surgical implantation a traumatic procedure for the

animal. The muscle exhibited a single, extended period of activity

commencing during the early upstroke, and continuing through the early

downstroke. During two of the wingbeats, supraspinatus exhibited two

periods of normal-amplitude EMG activity with a brief period of

quiescence interposed.

Biphasic EMG activity was observed in supraspinatus of Antrozous

(Hermanson and Altenbach, 1983). Biphasic activity was also observed

during the step cycle of Didelphis (Jenkins and Weijs, 1978), but not

during stepping in cats (English, 1978a) or dogs (Tokuriki, 1973a,

1973b, 1974). In several primate species, Tuttle and Basmajian (1978)

reported prominent activity of the supraspinatus during humeral

elevation and cautioned that the function of this muscle may be one of

acceleration of the humerus rather than as a simple stabilizer (i.e.,

rotator cuff). The variability observed in all of these studies may be

attributed to small sample sizes, at least for the two bat species

studied, and to an inconsistent pattern of locomotor movements studied

in the primate examples. In the latter study, movements of the animal

were not restricted to forward progression, but instead included

forward, lateral, and vertical movements. Support for a role in wing

supination is provided by EMG data indicating activity during the early

upstroke when the wing is supinated and abducted.


Form. The infraspinatus originates from the periphery of the

infraspinous fossa of the scapula: cranially from the spine of the

scapula; medially from the dorsum of the vertebral border of the

scapula; and laterally from the raised ridge of the infraspinous

fossa. All fibers insert upon a central tendon that courses through

the middle of the muscle. This tendon forms the sole insertion of the

infraspinatus upon the lateral ridge of the humerus. Fibers at the

medial end (nearest the vertebral border) are the longest in the muscle

and insert upon the central tendon at an acute angle. Fibers that

insert more distally on the central tendon are shorter and tend to

intersect the tendon at a larger angle than the medial fibers.

Innervation is by the suprascapular nerve, after that nerve provides a

branch to the supraspinatus and courses laterally under the tip of the

acromion process.

Comparative aspects. Strickler (1978) noted little variation in

the attachments of the infraspinatus in all bat species that he

studied. In contrast to the regression of muscle mass on body mass

obtained for supraspinatus, Strickler observed that the muscle was

larger in fast-flying species, but smaller in fast-flying frugivores.

No functional explanation was provided. Within families, there was no

consistent ratio observed between the mass of infraspinatus and


Functional aspects. Two phases of activity were evident in the

infraspinatus of Artibeus. The first period and the one exhibiting the

highest amplitude myopotentials commenced immediately prior to the

onset of downstroke movements, and continued during the first third of

the downstroke. A second, less intense period of activity was observed

in six wingbeats during the late downstroke and early upstroke. These

data were similar to observations made on the infraspinatus of

Antrozous (Hermanson and Altenbach, 1983). Although two distinct

phases were also observed in walking Didelphis (Jenkins and Weijs,

1979), only one period of activity was noted in cats (English, 1978)

and dogs (Tokuriki, 1973a, 1973b, 1974) at all gaits.

On the basis of anatomical position the infraspinatus is a

shoulder flexor, and a humerus abductor and supinator (Hermanson and

Altenbach, 1983; Strickler, 1978; Altenbach, 1979). The biphasic

activity pattern suggests an underlying stabilizing role at the

shoulder; the secondary EMG activity observed during the early

downstroke counteracts the action of the pectoralis and protects the

glenohumeral joint in association with other intrinsic muscles of the

shoulder. Because the wing of a bat is not capable of being

"feathered" during the upstroke like that of a bird, the role of the

abductors is accentuated in bats (Vaughan, 1959). Thus, the

infraspinatus exhibits strong activity during the non-propulsive

upstroke, and acts in concert with the more superficial spinodeltoideus

to abduct the wing.

Triceps Group

The triceps brachii is composed of three heads in most mammals.

Each head is innervated by branches of the radial nerve. Because of

the different origins of these heads, each is treated separately.

Triceps Brachii, Medial Head

Form. The medial head has a fleshy origin along the caudal

surface of the humerus, approximately 10-13 mm distal to the greater

tubercle. Muscle fibers pass medial to the radial nerve. Fibers

insert distally in a unipinnate fashion on the main triceps brachii

tendon. This is the smallest of the three heads of triceps brachii.

Comparative aspects. In most phyllostomids, the medial head

originates on the middle one-third of the ventral surface of the

humerus (Strickler, 1978). The attachments are conservative in most


Functional aspects. The medial head of triceps brachii spans one

joint, the elbow. Vaughan (1959) suggested that the medial head

stabilizes the sesamoid bones found at the insertion of the triceps

brachii tendons in many bats. No EMG data are available for any bats

to corroborate or refute the hypothesis.

Triceps Brachii, Lateral Head

Form. The lateral head originates by fibers on the proximal

one-quarter of the humerus, primarily on the caudal surface adjacent to

the greater tubercle, surgical neck, and lesser tubercle. All fibers

pass lateral to the radial nerve. Central fibers insert directly upon

a central tendon of insertion. Peripheral fibers insert on either side

of this tendon at an angle of less than 10 degrees.

Functional aspects. The lateral head of triceps brachii crosses

one joint, the elbow joint, and is capable to effecting extension of

that joint. EMG data is discussed along with the long head of triceps


Triceps Brachii, Long Head

Form. The long head arises from the axillary border of the

scapula, adjacent to the infraglenoid tubercle. The muscle is

pinnate. Fibers course distally to insert upon a central tendon in

common with the other two heads of triceps brachii. The main

insertional tendon contains a small sesamoid caudal to the elbow

joint. A synovial cavity is present deep to the tendon, permitting

movement of the tendon relative to the joint in this area. The

insertional tendon of the triceps brachii attaches to the caudal

surface of the olecranon region of the ulna.

Comparative aspects. The anatomy of the triceps brachii does not

depart significantly from the general pattern observed in other bats

(cf. Strickler, 1978): the long head spans the shoulder and elbow

joints, the lateral head spans only the elbow joint.

Functional aspects. In terrestrial mammals, the triceps brachii

function as anti-gravity muscles principally by maintaining elbow

extension during standing or locomotion (Armstrong et al., 1982;

English, 1978b; Jenkins and Weijs, 1979). EMG data for long and

lateral head of triceps brachii in Antrozous did not clearly support an

elbow extensor hypothesis (Hermanson and Altenbach, 1983). In A.

jamaicensis, muscle activity in the long head commenced prior to

initiation of the upstroke, and continued throughout the first one-half

of the upstroke. The lateral head also commenced activity prior to the

downstroke, but relative to the long head, EMG activity began later and

was of shorter duration. These temporal patterns are similar to

observations of triceps brachii activity in Antrozous (Hermanson and

Altenbach, 1983). In both bat species, the long and lateral heads of

triceps brachii exhibited activity patterns assoicated with wing

abduction. It is not clear how either portion of triceps brachii

effects elbow extension, a movement associated with the downstroke

phase of the wingbeat cycle (Eisentraut, 1936; Norberg, 1976; Hermanson

and Altenbach, 1983). Elbow extension automatically contributes to

spreading the distal wing during the downstroke in many bat species

(Vaughan, 1966; Vaughan and Bateman, 1970). In Artibeus, the amount of

time between EMG onset and downstroke initiation was approximately 55

msec. Armstrong et al. (1977) and Hermanson and Foehring (1982)

demonstrated the existence of fast type histochemical profiles for

triceps brachii in Myotis and Tadarida, respectively. In several

mammalian muscles, fast type profiles are correlated with contraction