1 THE MORPHOLOGY OF THE VENOUS SYSTEM IN THE HEAD AND NECK OF THE BOTTLENOSE DOLPHIN (TURSIOPS T RUNCATUS) AND FLORIDA MANATEE (TRICHECHUS MANATUS LATIROSTRIS) By ALEXANDER M. COSTIDIS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORID A IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012
2 2012 Alexander M. Costidis
3 To Sentiel Butch Rommel who contributed his intellect, time, and friendship and showed me the joys of anatomy. To the anatomists of the world who have experienced the wonder of these animals and have wo rked painstakingly to understand them.
4 ACKNOWLEDGMENTS I thank my loving family for their support and encouragement, and for giving me a solid foundation upon which stan d. Without them I would be lost. I am blessed for having Jennifers love, which got me thr ough many difficult times. Butch Rommels unwavering enthusias m about life is one of my most prized possessions and provided me with much of my drive. Brian Sta cys tireless pursuit of answers and grounded perspective regarding solutions has been suprem ely instructive and refreshing. Butchs, Brians and Micah Brodskys friendship has been priceless. Nicole Stacys warm thoughts and warmer smiles have been uplifting. Ann Pabst and Bill McLellan have been invaluable friends and colleagues whos e depth, breadth, and commitment are unparalleled. Roger Reeps gentle guidanc e and unwavering support made much of this possible and encouraged me through this process. I thank him for his calming presence, his rational intellect and fervent imagination. I thank my supervisory committee for investing their ti me and energy in helping me s ee this through. I thank Erick Roden for his jovial demean or and refreshing jest for life. I am grateful to the numerous stranding organizations around the country for their hard work against often seemingly insurmountabl e obstacles, and for their provision of such wonderful, sought after specimens. I am especially grateful to the biologists and staff of the Florida Fish and Wildlife Conservation Commission, the University of North Carolina Wilmington stranding network, the Vi rginia Aquarium stranding team, the Marine Mammal Conservancy, and Lowry Park Z oo for facilitating acquisition of many of my specimens. I am grateful to Dr. David Reese, Mary W ilson, Christine Fitzgerald and the rest of the radiology team at the University of Fl oridas College of Veterinary Medicine for their assistance with scanning of specimens at all hours of the day and
5 night. I am grateful to the University of Fl oridas College of Veterinary Medicine, Aquatic Animal Health Program and to Ruth Francis -Floyd for their support of my degree and their belief in me. Part of this research was funded by the University of Floridas College of Veterinary Medicine spring competition grant.
6 TABLE OF CONTENTS page ACKNOWLEDG MENTS .................................................................................................. 4 LIST OF TABLES .......................................................................................................... 10 LIST OF FIGURES ........................................................................................................ 11 LIST OF OBJECTS ....................................................................................................... 17 ABSTRACT ................................................................................................................... 18 CHA PTER 1 INTRODUCTION: COMPARATIVE ANAT OMY OF THE HEAD AND NECK VASCULATURE IN DOMESTIC AND MARINE MAMMALS .................................. 20 Significance of Vasculature ..................................................................................... 20 Notable Marine Mamm al Adaptat ions ..................................................................... 22 Terrestrial Mammalian Venous Morphology ........................................................... 25 Extrinsic Arterial Brain Supp ly in Domest ic Ma mmals ...................................... 27 Extrinsic Arterial Brain Supply in Marine Mamma ls .......................................... 30 Seali ons ..................................................................................................... 31 Seals .......................................................................................................... 32 Sireni ans .................................................................................................... 32 Cetaceans .................................................................................................. 33 Arterial Synopsis .............................................................................................. 36 Extrinsic Venous Brain Drainage in Domestic Mammals .................................. 38 Extrinsic Venous Brain Drai nage in Mari ne Mammals ...................................... 39 2 THE GROSS MORPHOLOGY OF THE VEN OUS SYSTEM IN THE HEAD AND NECK OF THE BOTT LENOSE DOLPHINS (TURSIOPS TRUNCATUS ) ............... 48 Chapter Fo reword ................................................................................................... 48 Materials and Methods ............................................................................................ 49 Results .................................................................................................................... 51 Neck ................................................................................................................. 51 Internal jugul ar veins .................................................................................. 54 External jugular veins ................................................................................. 56 Head ................................................................................................................. 57 Lingual and fa cial ve ins .............................................................................. 59 Branches of the facial vein inve sting the orbit and dorsal maxillary regions .................................................................................................... 63 CNS Ve ins ........................................................................................................ 64 Discuss ion .............................................................................................................. 67
7 Comparative V enous Anat omy ......................................................................... 67 Ontogenetic vs. Indi vidual Variability ................................................................ 68 Functional Implications of CNS Vasculature ..................................................... 69 3 DETAILS ON THE VASCULARIZATION OF THE AIR SINUSES AND FAT BODIES IN THE HEAD OF THE BOTTLENOSE DOLPHIN ( TURS IOPS TRUNCATUS ) ........................................................................................................ 91 Chapter Fo reword ................................................................................................... 91 Materials and Methods ............................................................................................ 96 Results .................................................................................................................... 99 Air Sinus and Fat Body Mor phology ................................................................. 99 Sinuses ...................................................................................................... 99 Air sinus extent .......................................................................................... 99 Bony walls ................................................................................................ 101 Bony re cesses ......................................................................................... 101 Eustachian tube ....................................................................................... 102 Blood-filled part (of the accessory si nus system) ..................................... 103 Fat bodies ................................................................................................ 103 Vascular A natomy .......................................................................................... 105 Facial vein ................................................................................................ 106 Internal jugular vein (#1, Figures 3-7, 3-9 to 3-10, and 3-13 to 3-14 ) ....... 107 External jugular vein (#2, Figures 3-9, 3-10, 3-14, 3-15) .......................... 108 Discuss ion ............................................................................................................ 109 Significance of the An at omy ........................................................................... 109 Accessory si nus system ........................................................................... 110 Fat bodies ................................................................................................ 114 Clarifications/In consistencies ......................................................................... 118 Synopsis ......................................................................................................... 124 4 THE GROSS MORPHOLOGY OF THE VEN OUS SYSTEM IN THE HEAD AND NECK OF THE FLORIDA MANATEE ( TRICHECHUS MANATUS LATIROST RIS) ..................................................................................................... 146 Chapter Fo reword ................................................................................................. 146 Materials and Methods .......................................................................................... 156 Results .................................................................................................................. 159 Arteries of the Head and Neck ....................................................................... 159 Veins of the Head and Neck ........................................................................... 163 Jugular Veins .................................................................................................. 165 Internal jugul ar vein .................................................................................. 165 External jugular vein ................................................................................ 169 Facial Veins .................................................................................................... 171 Lingual Veins and Hyoid Br anches ................................................................. 173 Maxillary Veins ............................................................................................... 176 Pterygoid, Deep Facial and Buccal Veins....................................................... 179 Facial Venous Structures ............................................................................... 183
8 Ventral massete ric ve in ............................................................................ 183 Lateral superficial branch 1 (tran sverse fa cial vein) ................................. 184 Lateral superficial branch 2 (ventral masseteric ve in) .............................. 185 Lateral superficial branch 3 (mas seteric v enous plexus) ......................... 186 Lateral deep branch (deep masseteric veins) .......................................... 187 Medial deep branch (pterygoid veins & deep facial vein) ......................... 187 Veins Of The Lips And Eyes ........................................................................... 188 Branch 1 (deep facial vein & anastomot ic branch of the ventral external ophthalmic vein) .................................................................................... 188 Branch 2 (fac ial ve in) ............................................................................... 189 Branch 3 ( angularis oris vein) .................................................................. 192 CNS Ve ins ...................................................................................................... 192 Brain veins ............................................................................................... 192 Epidural veins .......................................................................................... 194 Discuss ion ............................................................................................................ 195 Functional Implications of Facial Vasculature ................................................. 195 Functional Implications of Ophthalmic Va sculature ........................................ 200 Implications of Vasculature on Whole Body The rmoregulation ...................... 203 5 DETAILS ON SOME UNIQUE STRUCTU RES OF THE VENOUS SYST EM OF THE HEAD AND NECK OF THE FLORIDA MANATEE (TRICHECHUS MANATUS LATIROSTRIS) ................................................................................... 225 Chapter Fo reword ................................................................................................. 225 Materials and Methods .......................................................................................... 228 Results .................................................................................................................. 231 Mandibular Alveolar Vascular Bundle (MAB) .................................................. 231 Infraorbital Vascula r Bundle (IVB) .................................................................. 233 Ophthalmic Plex us ......................................................................................... 234 CNS Ve ins ...................................................................................................... 235 Brain veins ............................................................................................... 235 Epidural veins .......................................................................................... 237 Discuss ion ............................................................................................................ 239 Mandibular Alveolar Vascular Bundle (MAB) .................................................. 239 Infraorbital Vascula r Bundle (IVB) .................................................................. 240 Ophthalmic Plex us ......................................................................................... 244 Brain Ve ins ..................................................................................................... 248 Epidural Veins ................................................................................................ 249 Synopsis ......................................................................................................... 253 6 DISCUSSI ON ....................................................................................................... 276 Comparative Venous Morphology ......................................................................... 276 Species-specific Imp lications (D olphins) ........................................................ 278 Species-specific Implications (Manatees) ....................................................... 286 Thermoregulati on (brain) ......................................................................... 288 Thermoregulati on (oth er) ......................................................................... 294
9 Thermoregulati on (eyes) .......................................................................... 295 Vascular Growth and Pattern ing ........................................................................... 297 Primary Vascula r Plexus ................................................................................ 298 Molecular Te rminology ................................................................................... 300 Signaling Molecules ....................................................................................... 302 Vessel Growth Types ..................................................................................... 303 Patterning Dynamics (e.g. Clipping/Tri mming, Elongation, Linearity, etc .) ..... 304 Arteriovenous Differentiation of Primary Vascular Ple xuses ........................... 305 Physical and Physiological Guiding Pr inciples ............................................... 307 Organ Vascular Examples .................................................................................... 314 Observed Patterns in Marine Ma mmals ......................................................... 316 Function of Marine Ma mmal Vascula ture ....................................................... 325 APPENDIX LIST OF RE FERENCES ............................................................................................. 330 BIOGRAPHICAL SKETCH .......................................................................................... 353
10 LIST OF TABLES Table page 2-1 List of specimens us ed for this stu dy. ................................................................. 74 2-2 Structure label s and their names. ....................................................................... 74 3-1 List of specimens us ed for this stu dy. ............................................................... 126 3-2 List of soft tissue (blood vessels and nerves) structur e labels used in the figures and their co rre sponding na mes. ........................................................... 127 4-1 List of specimen numbers, total body length (TBL), stranding date, gender and description of research use for each specim en. ........................................ 208 4-2 List of figure labels and thei r corresponding stru cture n ames. ......................... 208 5-1 List of specimens us ed for this res earch. ......................................................... 254 5-2 List of structure labels and thei r names. ........................................................... 254 6-1 Structure label s and their names. ..................................................................... 328
11 LIST OF FIGURES Figure page 2-1 Ventral view of the venous branches (b lue) of the crani al vena cava (1) of a bottlenose dolphin (ECW-005). ........................................................................... 762-2 Simplified schematic repr esentation of Figure 21. ............................................. 772-3 Right lateral view of a 3D reconstruction of venous angiography of a bottlenose dol phin. ............................................................................................. 782-4 Ventral view of a three dimensional reconstruction of CT angiography of the veins of a bottlenose dolphin with the rostrum on shown on the left and caudal aspects to the right. ................................................................................. 792-5 Gross dissection of venous plexus ( venu plexi comitans arteria carotidis communis and externa ) (5) surrounding the common and external carotid arteri es. .............................................................................................................. 802-6 Right lateral view of internal and external caro tid arte ries. ................................. 812-7 Mid-sagittal section of dolphin head and neck with the head on the left and cranial thorax on the ri ght. .................................................................................. 812-8 Transverse slice of computed tom ographic angiography near apex of melon of a bottlenose dolphin, showing palatine (yellow arrows) and maxillary labial (red arrows) veins. .............................................................................................. 822-9 Left lateral view of maximum intensity projection from CT angiography of the head of a bottl enose dol phin............................................................................... 832-10 Dorsal view of three-dimensional reconstruction of the veins on the dorsal surface of the skull of a bottlenose dolphin. ........................................................ 842-11 Right lateral view of the head and ne ck of a bottlenose dolphin showing the caudal auricular vein (20) coursing behind the external auditory meatus and supplying the venous plexus (20) on the temporal and supraoccipital regions ............................................................................................................... 842-12 Left lateral view of three-dim ensional reconstruction of CT venous angiographic image of a bottlenose dol phin head, showing the ophthalmic plexus (15) emerging from the facial (4) vein and extending into the orbit. ........ 852-13 Close-up of a lateral view of the or bit showing the intr icate nature of the ophthalmic plexus (15), draining latera lly into the faci al vein (4)......................... 86
12 2-14 Dorsal view of maximum intensity projection of CT venous angiography of a bottlenose dolphin showing faint b one shadows and yell ow veins. .................... 872-15 Dorsal view of maximum intensity projection of CT venous angiography of a bottlenose dolphin showing faint b one shadows and yell ow veins. .................... 882-16 Excised epidural rete injected with red latex in the arterial component, and blue latex in the venous co mponent. .................................................................. 892-17 Excised epidural rete from a bottlenose dolphin sho wing ventral internal view and dorsal view. .................................................................................................. 903-1 Plate combining computed tomographic (CT) slices with volume renderings of the pterygoid and peribullar plexus (green), intramandibular fat body plexus (yellow), and anterior lobe (purple) to illustrate overa ll location of the structur es. ......................................................................................................... 1293-2 Schematic illustration of dorsal and ventral skeletal associations to the accessory sinus system in a bottlenose dolphin (left) and a pygmy sperm whale (ri ght). ..................................................................................................... 1303-3 Schematic illustrations of lateral and ventral views of the pterygoid sinus system in bottlenose dolphins (left) and pygmy sperm wh ales (right). .............. 1313-4 Medial view of the pterygoid venous pl exus that lines the lateral wall of the pterygoid sinus of a neonatal sperm whale ( Physeter macrocephalus ). ........... 1323-5 Cross-sectional comput ed tomographic image at the level of the eyes of a pygmy sperm whale with contrast enhanced veins. .......................................... 1333-6 Magnetic resonance imaging cross-sect ional view of the head of a bottlenose dolphin at the level of the eyes, showing association of the intra(IMFB) and extramandibular (EMF B) fat bodies. ................................................................. 1343-7 Three-dimensional angi ographic reconstruction of the left lateral aspect of the head of the bottlenose dolphin, showing associations of superficial veins and bony elem ents. .......................................................................................... 1353-8 Lateral view of right dentary (image has been flipped to simulate left lateral orientation for consistency between images) with a window cut out of the lateral wall in order to visualize the intramandibular fat body plexus. ............... 1363-9 Ventrolateral view of the right side of the bottlenose dolphin neck showing the jugular and facial veins, and the complex anastomoses between the structures (image has been flipped to si mulate left lateral orientation for consistency between images). .......................................................................... 137
13 3-10 Ventromedial view of the right half of a mid-sagittally sectioned bottlenose dolphin showing the jugul ar branching patterns................................................ 1383-11 Cross-sectional view of the intramandibular fat body (IMFB) of a bottlenose dolphin, progressing from caudal (left) to rost ral (right ). ................................... 1393-12 Medial view of gross dissection of a bottlenose dolphin with latex injected vessels ............................................................................................................ 1403-13 Lateral view of a volume renderi ng of computed tomographical angiography of a bottlenose dolphin head showing comp lexity of venous investment. ......... 1423-14 Mid-sagittal view of a volume rendering of computed tomographical angiography of a bottlenose dolphin head showing complexity of venous investment along medial aspect. ...................................................................... 1433-15 Medial view of gross dissection of the pterygoid and basicranial regions, identifying some of the key venous stru ctures outlined in the text (structure numbers correspond to those in previous images). .......................................... 1444-1 Ventral view of a CT angiographic reconstruction of the arterial system of the head and neck of a Florida manat ee showing major arteri es. .......................... 2104-2 Medial view of a CT angiographic rec onstruction of the major arteries of the right half of t he head and ne ck. ........................................................................ 2114-3 Ventral view of three-dimensiona l reconstruction of a Florida manatee showing complexity of jugul ar venous branc hes (blue). ................................... 2124-4 Ventral view of the head and neck of a Florida manatee showing gross morphology of branches (blue) of t he brachiocephalic veins (1) and their associations with the arterial structures (4, 14, 22, 23, 24). .............................. 2134-5 View of the right half of the head showi ng right internal jugular vein (3) as it receives the emissary of the jugular foramen (27) from the braincase. ............ 2144-6 Medial view of right maxillary vein (8) and its countless anastomosing branches encasing the exte rnal caroti d artery. ................................................. 2154-7 Ventral view of the head of a Florida manatee showing gross appearance of venous hyoid arch (26) draining into numerous laryngeal and pharyngeal branches and connecting to facial (7) and lingual (9) tributaries of the external jugul ar vein. ........................................................................................ 2164-8 Medial view of gross dissection of the neck region of a mid-sagittaly sectioned m anatee head. ................................................................................. 217
14 4-9 Right lateral view of gross disse ction of the neck region of a manatee showing the confluenc e of the ascendi ng pharyngeal vein (15) with the maxillary ve in (8). ............................................................................................. 2184-10 Left lateral view of the facial vein emerging laterad from the mandibular notch for the facial veins, and receiving deep facial and ventral masseteric vein caudally ........................................................................................................... 2194-11 Left lateral view of 3D reconstructi on of a CT venous angiogram of a Florida manatee head showing the pat h of the facial vein (7) and its numerous contributing branches draining the upper (37) and lower lips (36) and nasal passages ( 40, 41). ............................................................................................ 2204-12 Ventral view of 3D reconstruction of a CT venous angiogram of a Florida manatee head showing t he path of the facial veins (7). .................................... 2214-13 Dorsal view of 3D reconstruction of a CT venous angiogram of a Florida manatee head showing the origin s of the facial vein (7) as the angularis oculi (11). .................................................................................................................. 2224-14 Simplified schematic of a medial view of a manatee head summarizing the major venous structures descr ibed. .................................................................. 2235-1 Medial view of mandibular canal with medial wall removed to show arteriovenous nat ure of MAB. ........................................................................... 2565-2 Cross sectional view of the mandi bular vascular bundle showing arteries (red) and veins (blue) injected with latex. Note the general lack of arterial anastomoses and innumerabl e venous anast omoses. ..................................... 2565-3 Medial view of corrosion cast of veins of the left side of the manatee head showing veins of MAB investing soft tissues of mandible as mental veins. .... 2575-4 Medial view of the right half of a mid-sagittally sect ioned manatee head with the lateral wall of the bony naris re moved to visualize the IVB and its drainage into the maxillary vein (4). .................................................................. 2585-5 Cross sectional view of the infraorbital vascular bundle showing arteries (red) and veins (blue) inje cted with la tex. .................................................................. 2595-6 Photomicrograph of a Mas ons trichrome histologic preparation of a portion of the IVB in cro ss section. ................................................................................... 2605-7 Photomicrograph of a Mas ons trichrome histologic preparation of a portion of the IVB in cro ss section. ................................................................................... 2615-8 Right dorsolateral view of the maxillary lips showin g the distal branches of the IVB emerging from the infraorbital fo ramen and investing the lips in triads
15 (probe tip) composed of a central arte ry with two anastomosing satellite veins (5). .................................................................................................................... 2625-9 Left lateral view of ophthalmic venous plexus of the Florida manatee injected with liquid latex. ................................................................................................ 2635-10 Left lateral view of the left eye sho wing the investment of the ophthalmic venous plexus (8) in relation to the extrinsic eye muscl es. ............................... 2645-11 Dorsomedial close-up view of the proximal portion of the left ophthalmic vein (13) on its coarse to the calvarium. ................................................................... 2655-12 Medial view of the right half of a mid-sagittaly sectioned manatee showing bifurcation (3) of ophthalmic vein into the cavernous sinus (8) and durae (7) within the calvarium. ......................................................................................... 2665-13 Oblique dorsocaudal view of a la minectomized manatee head showing the intact dura on either side of the dors al sagittal sinus (17) and covering the brain. ................................................................................................................ 2675-14 View of an excised manatee brain with the dura covering the let cerebral hemisphere refl ected. ....................................................................................... 2685-15 Medial view of the right brain hemisphere of a mid-sagittaly sectioned manatee brain, showing extensive dural vasculature covering the dorsomedial surface of t he right brain hemisphere. .......................................... 2695-16 Medial view of the right brain hem isphere of a manatee showing a partially exposed choroi d plexus. ................................................................................... 2705-17 Cross-sectional view of the spinal cord and epidural rete at the level of the occipital condyles, showing extensive v entral and ventrolateral investment of arteries and veins. ............................................................................................ 2715-18 Ventral view of cervical epidural rete (14) showing enlarged ventral internal vertebral veins (23) along the periphery and a distinct midline separation of the left and right sides of the rete except in the most cr anial region (left). ...... 2715-19 Ventral view of the spinal cord with the epidural rete reflected along its natural ventra l division. ..................................................................................... 2725-20 Dorsal view of the cervical epidural rete showing arcuate veins (26) bridging the two sides of the rete and dorsal longitudinal anastomoses connecting arcuate ve ins. ................................................................................................... 2725-21 Left lateral view of an excised manat ee brain showing the ventral internal vertebral vein (23) bifurcating on the le ft. ......................................................... 273
16 5-22 Dorsal view of an excised manatee brain with intact durae showing the connections of the dorsal branc hes of the ventral internal vertebral veins (23) with the sigmoid sinuses (16). .......................................................................... 2745-23 Composite image showing location and morphology of the IVB, MAB, BVB, ICVB, and CVB. ................................................................................................ 2756-1 Simplified schematic representations of the superficial veins of the head of the cow (A) and horse (B) compared to the bottlenose dolphin (C) and Florida manat ee (D). ......................................................................................... 329
17 LIST OF OBJECTS Object page 3-1 Three-dimensional reconstruction of a bottlenose dolphin with a contrastenhanced venous system, showing spatial relations hips of the structures discuss ed. ........................................................................................................ 1454-1 Three-dimensional reconstructi on generated from computed tomographic angiography of a Florida manatee with a contrast enhanced venous system. 224
18 Abstract of Dissertation Pr esented to the Graduate School of the University of Fl orida in Partial Fulf illment of the Requirements for t he Degree of Doctor of Philosophy THE MORPHOLOGY OF THE VENOUS SYSTEM IN THE HEAD AND NECK OF THE BOTTLENOSE DOLPHIN (TURSIOPS T RUNCATUS) AND FLORIDA MANATEE (TRICHECHUS MANATUS LATIROSTRIS) By Alexander M. Costidis August 2012 Chair: Roger Reep Major: Veterinary Medical Sciences There is a marked paucity of informati on on the venous anatomy in the head and neck of the bottlenose dolphin (Tursiops truncatus) and Florida manatee (Trichechus manatus latirostris). Specimens from each species were injected with a mixture of liquid latex and barium sulfate suspension to facilitate computed tomographic angiography and gross dissection of venous structures. Angi ography was used in order to aid in visualization and dissection of vascular st ructures, and enable assessment of threedimensional relationships. Both specie s were found to have countless venous anastomoses throughout their head and neck. These anastomoses form a large collateral drainage system for blood leaving the tissues of the head and returning to the heart. Although many veins were identified as homologs of veins found in terrestrial mammals, numerous unique or elaborated vascular structures were found. Such structures included an ex pansive and intricate venous pl exus associated with the accessory sinus system and acoustic fat bo dies of the bottlenose dolphin and two remarkable arteriovenous vascular bundles associated with t he mandibular and infraorbital canals of the manat ee. Voluminous arteriovenous retia were also found
19 surrounding the spinal cord within the cerv ical portion of the neural canal of both species. All of the afor e mentioned structures hint ed at significant functional implications for such things as conservati on of body heat, regional heterothermy of the central nervous system, and volume compensat ion associated with air sinuses. The considerable venous investment of the head and neck of both species suggests that their venous system is not only an elaborate conduit for draining blood, but rather an intricate, dynamic system designed to fa cilitate and/or modulat e certain important physiological processes.
20 CHAPTER 1 INTRODUCTION: COMPARATIVE ANAT OMY OF THE HEAD AND NECK VAS CULATURE IN DOMESTIC AND MARINE MAMMALS Significance of Vasculature Understanding the functional morphol ogy of animals is often dependant on knowledge of their basic biology Knowledge of vascular an atomy is one of the most basic pieces of biological information and is integral to understanding how an animal functions. Vascular anatomy can often help explain physiological, patho logical, clinical and even behavioral manifestations. T he anatomy of blood vessels of humans, and laboratory and domestic mammals has been we ll known for decades, and is therefore often taken for granted. This knowledge has allowed scientists and clinicians to make considerable progress in understanding various physiolog ical processes and treat countless diseases (Batson, 1940; 1956; Beck & Plate, 2009; Carmeliet et al. 1996; Chaynes 2003; Clendenin & Conrad, 1979; de la Torre et al. 1959; 1962; Fujii et al. 1980; Groen & Ponssen, 1991; Moss, 1974; Mullan et al. 1996; Tatelman, 1960; Whisnant et al. 1956). The vascular morphology of mammals has vast implications on not only such obvious factors as oxygen and nutrient dist ribution and metabolite clearance, but also more subtle yet equally important factors such as thermoregulation, immunity and reproduction. Just as certain vascular structures (e.g. pampiniform plexus, nasal mucosal plexus, etc.) have impor tant thermoregulatory functi ons in terrestrial mammals, so have vascular structures (e.g. caudal va scular bundle, inguinal plexus, lumbocaudal plexus, etc.) in marine mammals been shown to have equally important thermoregulatory functions in whole body cooling and regional heterothermy of
21 thermally sensitive tissues such as the reproductive tract (Pabst et al. 1995; Rommel & Caplan, 2003; Rommel et al. 1992; 1993; 1994; 1995; 1998; 2001). Another crucial function of vasculature in diving marine mammals is its role in nitrogen absorption and eliminat ion from tissues (Fahlman et al. 2006; 2009). With the exception of human divers, nitrogen gas kinetics are rarely a concern for terrestrial mammals. Yet during diving marine mammals are regularly exposed to increased hydrostatic pressures that generate forces conducive to driving pulmonary nitrogen into circulation and subsequently into the ti ssues. Although some marine mammals have long been thought to be exempt from di ving nitrogen accumulation and its adverse effects, recent evidence suggests that this may not be entirely true (Fahlman et al. 2006; 2009; Hooker et al. 2009; 2011; Houser et al. 2001). Luckily, gas kinetics in human divers have been studied quite extensively and can provide a robust springboard from which to study marine mammal diving. Nonetheless, one of the key pieces of information in understanding diving gas kineti cs is knowledge of the vascular system, as it is one of the main factors determining the extent and distribution of gases in the body. Unfortunately, our knowledge of the vascu lar system of most marine mammals is limited at best. Fortunately, recent te chnological advances in computed tomography (CT) in conjunction with increased specimen recovery efforts and preservation ability have made it possible to examine some of the complicated vascular patterns in greater detail. As such, I undertook to describe t he gross morphology of the venous system in the head and neck of the bottlenose dolphin ( Tursiops truncatus) and Florida manatee ( Trichechus manatus latirostris ), by combining traditional gross dissection techniques with newer diagnostic imaging modalities.
22 Notable Marine Mammal Adaptations To start a discussion about comparativ e marine mammal vascular morphology, one has to familiarize themselves with some of the drastically different morphology expressed in some of the marine mammals. Of the seal s, manatees, and cetaceans, the cetaceans are arguably the most derived of the group. For instance, a dramatic evolutionary feature of cetacean skull morphology is manifested as the presence of signific ant telescopingdorsocaudal migration of the nasal, maxillary, and premaxillary bones and overlapping of the skull bonesto accommodate vertical positioning of the internal nares and dorsal location of the exte rnal nares. This telescoping necessitated migration of other skull features and altera tion of certain reference landmarks. For instance, the infraorbital foramina are no longer positioned below the orbit but are instead supraorbital and this morphology affect s the terminal branc hes of the maxillary arteries and veins. The laryngeal cartil ages have elongated epiglottal and corniculate cartilages that extend dorsad through the pharynx and seal on the roof of the pharynx. Fusion and thinning of cervical vertebrae limit degree of flexion of the neck. Mobility of the head has partially been restored by enl argement and rounding of the occipital condyles. Other integral changes in the biology of odontocete cetaceans are also reflected in their skull. The dentition--used no longer for chewing but rather for grasping prey--has acquired a homodont pattern. Replac ing the trabecular marrow cavities of the dentaries are completely hollow spaces invest ed with acoustically conducting lipid. The forehead or melon covering the dorsal aspect of the maxillary and prem axillary bones is composed of lipid with acoustic properties sim ilar to that investing the dentaries. The tympanic bullae are no longer firmly ankylosed to the temporal bones. The earbone complexes are loosely connected to the sku ll and are surrounded to a great extent by
23 air-filled peribullar sinuses. In odontocetes, a large accessory air-filled sinus system has developed on the ventral as pect of the skull in associ ation with much of the basicranial, pterygoid, and orbital region. Considering the less than trivial morphologi cal alterations, large size and cryptic lifestyles of marine mammals it should be no surprise that our understanding of them has progressed slowly. Little is known about the soft tissue morphology of marine mammal head and neck vasculature and even less is understood about it. This is especially true of cetaceans and sirenians. Furthermore, what is known about their vasculature varies considerably between groups This is no surprise considering the diversity in aortic branching patterns seen even in domestic mammals. More is known of cetacean arterial patterns, while seal venous patterns are much better understood than their arterial counterparts. The relative inconsistency of veins has apparently intimidated most people from attempting to describe them wit h much depth or breadth. While attempting to explain the paucity of information on venous drainage of the head of cetaceans, Fraser and Purves stated that anatomists such as Murie, Boenninghaus Carte & Macalister have, with justification, described the blood vessels only in very general terms, because the ramifications of the finer branches are exceedingly complex and form extensive retia mirabilia which are associated with the air space. (1960) Indeed when one examines the venous investm ent in the head and neck of cetaceans, what emerges is a convoluted arrangement of seemingly indirect pathways with numerous emissary veins, intercommunica ting sinuses, and plexiform elaborations. This design of multitudinous retial and bund le-type vascular aggregations can be found investing cervical, pharyngeal, epidural and in tracranial structures (Fraser & Purves,
24 1960; McFarland et al. 1979; Ommanney, 1932; Slijper, 1936; Walmsley, 1938) of all cetaceans. Most notable of the venous structures of cetaceans thus described are perhaps the epidural venous networks and the pterygoid associated fibrovenous plexuses (Fraser & Purves, 1960; Costidis & Rommel, 2012). Though quite variable in size and extent between species, the epi dural veins of cetaceans form a valve-less network of large, anastomosing veins (Barnett et al. 1958; Harrison & Tomlinson, 1956; Ommanney, 1932;) which comprises a collateral (to the jugular) connection of the brain to the cranial vena cava. Valve-less epidur al veins are not unique to cetaceans, but rather can be found in all mammals incl uding dogs and humans, and have been shown to form a large collateral path for venous re turn from the extremities (Batson, 1940; 1957; Breschet, 1819; Groen & Ponssen, 1991; Herlihy, 1947; Launay, 1896; Reinhard et al. 1962). As can be seen in Figures 312, 3-14, and 3-15, s ubstantial venous connections exist between the intracranial region and the epidur al, pterygoid and pharyngeal regions of the dolphi n. The fibrovenous plexus (Fraser & Purves, 1960) and its associated pterygoid vein s and pterygoid sinus vasculature can also be quite variable between different cetacean species, however their presence and connection to extensive venous investments of the intramandibular fat and walls of the accessory airfilled sinuses appears relatively consistent among at least odontocete species (Costidis & Rommel, 2012). Numerous additional venous peculiarities exist in marine mammals that at times confound the anatomist with a traditional kno wledge of vascular morphology. Such peculiarities include but are not limited to: (1) the lack of an azygos venous system in
25 cetaceans, (2) significantly enlarged hepatic veins converging into a massive hepatic sinus in some phocid seal species, (3) caval sphincters just cranial to the caval foramen of the diaphragm of seals and cetaceans, (4 ) elaborate venous components of vascular reproductive cooling systems in seals, manatees and cetaceans, (5) inguinal and cervical venous plexuses in seals, (6) extensive venous com ponents of the caudal vascular bundle of cetaceans and manatees, (7 ) elaborate pericardial venous plexuses in seals and cetaceans, (8) paired abdominal caudal venae cavae in seals, manatees, and dolphins, (9) arterial supply to the brain via convoluted thoracic and epidural retia rather than the through the carotid arterial system, and (10) counter-current vascular structures that extend fr om the aorta to the tip of the dorsal fin(Barnett et al. 1958; Harrison & Tomlinson, 1956; Hol et al. 1975; McFarland et al 1979; Fawcett, 1942; Pabst et al. 1995; Rommel et al. 1992; 1995; Ronald et al. 1977; Scholander & Schevill, 1955; St. Pierre, 1974). Terrestrial Mammalian Venous Morphology Due to the convoluted and inconst ant nature of the v enous system and the considerable diversity among species, it ma y prove helpful to first outline the venous pathways of better-studied species such as domestic mammals, before attempting to understand marine mammal venous patterns. Since no discussion of circulatory anatomy is complete without arterial consider ations, I will provide a cursory description of arterial circulation. Nonetheless, due to the focus of the research I will concentrate efforts on venous circulation. In the interest of searching for patterns based either on function or phylogeny--arguably two of t he most influential forces on organismal architectureI will describe the vasculatu re in artiodactyls, perissodactyls, and carnivores. I will try to make special note when individual species posses unique
26 vascular attributes, but only when those attr ibutes seem functi onally important or relevant. Of consequence is the fact that due to the varied human and veterinary medical backgrounds of researchers who conducted much of the historical work describing the vascular anatomy, there is considerable inconsistency in use of vascular terminology. Numerous examples exist of different nom enclature being used interchangeably for the same structure. Sometimes this was due to the nomenclature deriving from species in which a vessel had ancillary f unctions--the internal thoracic artery has frequently been called the internal mammary due to its impor tant role in blood su pply to the mammary glands of some species. Other times it has been based on differing nomenclature criteria such as the ascending pharyngeal ar tery of the external carotid artery being called the artery of the fora men lacerum due to its passage th rough that bony canal, or the costocervical trunk being called the posterior thoracic artery in cetaceans. Adding to the complexities of decipher ing the historical literatur e is the application of human spatial nomenclature to veterinary anatomy. Terminology such as superior, inferior, anterior, and posterior are frequent ly used in place of what might be considered more appropriate reference terminology such as cranial, caudal, ventral, and dorsal, respectively. Therefore, all terminology employed within this paper will be used under the assumption that all the animals in this discussion, whether terres trial or aquatic, are naturally found in ventral recumbency. Additionally, whenever possible, anatomical terminology will be based on that established by Schaller (2007) and Schummer et al. (1981) as it follows the established Nomina Anatomica Veterinaria
27 Extrinsic Arterial Brain S uppl y in Domestic Mammals Because the arterial and venous systems are often associated with each other either physically or functionally, an overview of the arterial supply to the brain seems prudent. Most mammalian arterial supply to t he brain follows one of three patterns. 1) The most common and seemingly most basal pa ttern is via bilateral internal carotid arteries. Primary blood supply to the brain via internal carotid arteries can be traced back from fish and amphibians to reptiles an d birds (Gillilan, 1967). This pattern of blood supply is seen in horses and primates (Gillilan, 1974; Gillilan & Marksbery, 1963). It is important to note that in some adult mammalian species, even when the proximal portions of the internal carotid arteries naturally degenerate at or near the time of birth, the distal most intracranial portions seem to always reconstitute, leading to the generic intracranial carotid pattern observed in mammals with a complete internal carotid (Gillilan, 1974). 2) The second pattern is seen in mammals such as cats and ruminants and occurs when the proximal or extracranial internal carotid artery does not directly supply the brain. Instead, the external ca rotid or one of its tributaries supplies the majority of the blood to the brai n. In such cases, a carotid rete mirabile (rete mirabile caroticum ) seems to always be present between the brain and the external carotid tributary (e.g. maxillary, a scending pharyngeal, etc.) supplying the brainas seen in the pig, sheep, and cow. 3) The third patte rn of brain supply might be considered intermediate, in which both the internal and ex ternal carotid arteries contribute to the circulation of the brain, albeit to a different degree. Such a pattern is seen in the dog and cat (Evans, 1993; Gillilan & Marksbery, 1963; Reinhard et al.1962,Schaller, 2007; Schummer et al. 1981). Aside from the obvious functioning as a conduit, any further
28 functional significance of the carotid rete is uncertain, but as Gillilan (1974) stated this hydraulic complex may alter the hemodyna mics of the blood reaching the brain. Owing to the fact that wit hout an increase in cross-sectional area, any diminution of caliber of pipes can result in increased resistance to flow, this seems like a reasonable argument. The consist ency of intracranial internal carotid arteries despite varying extracranial sources is suggestive of a highly conserved aspect of vascular morphology and may represent a strong structural and functi onal integration of those vessels with the anatomy of the in trinsic arterial blood supply to the brain. Altering such a structural and functional integration would likely require very drastic architectural changes in order to appropriately maintain the high degree of intrinsic vascularity required by the brain. It is important however to note that despite the fact that the intracranial arterial system appears fairly conserved among species, the extracranial system can be quite variable in its finer ramifi cations. In addition to the variation of the carotid arterial contribu tions to intracranial circulation, t he role of other arteries can also vary significantly. For instance the basilar arte ry is present in all domestic mammals as well as humans; however the size and origin can be quite different. Notably, Gillilan (1974) comments on the size and direction of the basilar artery in some animals, suggesting that the mere presence of the ar tery does not necessarily mean it supplies the brain. For instance, the basilar artery in the cow tapers as it progresses caudad (Gillilan, 1974). According to Gillilan (1974) blood must ther efore flow caudad from the internal carotid arteries and basilar supply of the brain is highly unlikely. Similarly, the basilar arteries of the pig and sheep taper significantly as they progress caudad and transition into an exceedingly small ventral (anterior) m edian spinal artery, with a
29 sudden reduction occurring in the sheep just at the junction of the caudal (inferior) cerebellar arteries (Gillilan, 1974). Intere stingly, Padget (1948) and Heuser (1927) noted that in the developing human and pig embr yo, the vertebral ar tery forms a major contribution to the basilar arte ry, therefore Gillilan s account is either in error, or is suggestive of ontogenetic changes in the vertebral contribution to the basilar. In opposition to Gillilan (1974), Schaller (2007) shows a very distinct and fairly substantial contribution of the bilateral vertebral arteries to the more caudal ex tension of the basilar arterythe ventral spinal arteryin the form of spinal rami, as we ll as directly to the basilar artery by way of a confluence with t he anastomotic branch of the occipital artery into an epidural rete mirabile It is not clear how much of this discrepancy between reports is due to individual variation or ex perimental error; however Schallers (2007) description seems more in accord with the developmental picture presented by Heuser (1927) and Padget (1948). Either way it is obvi ous that variation in arterial patterns can occur. Finally, it seems important to note that mere pres ence of vascular connections between two systems does not necessarily inform us of a specific direction of flow. As seen by various researchers cited by Gillilan (1974), the occipital and vertebral arteries of the sheep do not appear to contribute much to brain circulation until the external carotid (and therefore maxillary) arteries supplying the carotid retia mirabilia are occluded. Therefore vascular connections can at times be, not indications of normal blood flow patterns but rather examples of plumbing that is utilized either modestly as collateral flow, or during uniquely demanding physiological events. Research has shown that despite unilateral and even transient bilateral ligation of the common carotid
30 arteries in the dog, brain damage is avoided through collateralization of blood flow (Clendenin & Conrad, 1979; Moss, 1974; Tatelman, 1960; Whisnant et al. 1956). Perhaps even more striking is the apparent homol ateral isolation of blood supply to the brain of even the domestic mammals possessing a carotid rete despite the numerous emissary connections (Moore & Da lley, 2006). Indeed, just as in the horse which does not possess the numerous laterally anastomotic connections of the carotid rete Gillilan (1974) stated that only when the pressure relationships are di sturbed does admixture take place between the two sides. Extrinsic Arterial Brain Suppl y in Marine Mammals Considering the degree of variability in marine mammal forms, it should be no surprise that marine mammal arte rial supply to the brain is si milarly quite varied. What is striking, however, is that among the spec ies discussed here, there may be a relatively gradual transition from the more typical ma mmalian pattern to that most unique pattern exhibited by cetaceans. Curiously, little research has been published which describes the arterial pathways to the brain of phocid seals (Murie, 1873; Sobolewsky, 1986), but venous drainage is fairly well described. Conv ersely, arterial supp ly to the brain of cetaceans has been well described (Galliano et al. 1966; Lin et al. 1998; McFarland et al. 1979; Melnikov, 1997; Nagel et al. 1968; Nakajima, 1961; Ommanney, 1932; Slijper, 1936; Viamonte et al. 1968; Vogl & Fisher, 1982; Walm sley, 1938; Wilson, 1879) but our understanding of the venous drainage is seve rely lacking. Similarly, the arterial morphology in the head and neck of the manatee was descri bed to some extent by Stannius (1845) and Murie (1874), but the venous system of the region remains largely neglected. Despite the paucity of data, what little we know suggests a trend from the more typical mammalian carotid supply in most pinnipeds (sea lions and seals) to the
31 most unusual cetacean pattern involving comp lete substitution of carotid supply to the brain by epidural supply (Galliano et al. 1966; Lin et al. 1998; McFarland et al. 1979; Melnikov, 1997; Nagel et al. 1968; Nakajima, 1961; Ommanney, 1932; Slijper, 1936; Viamonte et al. 1968; Vogl & Fisher, 1982; Walmsley 1938; Wilson, 1879). Completing the transition between generic mammalian and unique cetacean patterns may be the sirenians (manatees and dugongs). In line wit h their fairly non-conformist biology, manatees (and perhaps sirenians in general ) appear to have an intermediate type of arterial pattern of blood supply to the brain, showing contributions from the internal carotid and epidural system (Stannius, 1845; Murie, 1874; Fawcett, 1942), though the degree of ontogenetic variat ion is not understood. Sealions Murie described the arterial pathways to the head of the sea lion, noting that although m inor contribution to the brain may occur via branches of the maxillary artery, the predominant arterial supply to the brain is via the inte rnal carotid artery (Murie, 1874) Dormer et al. (1977) also found the internal caro tid to form the main supply to the circle of Willis of the California sea lion ( Zalophus californianus ). Similarly, in the Australian sea lion ( Neophoca cinerea ) and New Zealand sea lion ( Phocarctos hookeri ), the internal carotid artery was found to travel through the carotid canal to the brain (King 1977). King (1977) however did not make m ention of the vertebrobasilar contributions nor did she investigate the termination of t he maxillary artery, therefore it is unknown what role they may play in supplying blood to the brain. Sobolewsky (186) described the aortic branching in two sea lion species however the distal extensions were not covered in detail. In the California sea li on, DuBoulay and Verity (1973) noted large contributions to brain supply by the in ternal carotid and vertebral arteries.
32 Seals Not much is known about phocid seal arte rial supply to the brain. There have been few public ations to date on the topic of seal vascular anatomy, with most descriptions covering the venous system (Bron et al. 1966; Folkow et al 1988; Hol et al 1975; Murdaugh et al. 1968; Nordgarden et al. 2000; Ponganis et al. 2006; Ronald et al. 1977; Sobolewsky, 1986). Sobolewsky (1986) co vered the arterial branching patterns in a number of seal species, however detail ed accounts of their c ephalic branches were not given. Based on previous latex injections and dissections, it appears as though blood supply to the brain may be predominan tly via the carotid system since no large arterial epidural retial system exists like t hat found in cetaceans and sirenians. Based on DuBoulay & Verity (1973), primary suppl y to the brain likely occurs through the internal carotid arteries but is signi ficantly supplemented by vertebrobasilar contributions. Sirenians Stannius (1845) described porti ons of the arterial system of the manatee. Three decades lat er, Murie (1874) similarly described portions of the arterial system of the head and neck. Neither description may be considered complete or exhaustive, however both of them provide a solid found ation from which to expand on, and given their logistical limitations were tremendous works. The general findings of Stannius (1845) and Murie (1874) were that the arterial supply to the head was fairly similar to that of domestic mammals, with a common carotid artery branching into internal and external carotid arteries, t he former traveling to the phar ynx and face while the latter enters the brain case to supply the circle of Willis. Although Murie (1874) described the presence of notable vertebral arteries, he failed to describe the presence of any form of
33 an epidural rete such as that described by Fawcett (1942), which given the size of the rete (Figures 5-17 to 5-19) is rather puzz ling. As discussed in Chapter 4, some significant differences were observed betw een Muries arterial description and my findings. Cetaceans As is common with most domestic mammals the external carotid arteries of cetaceans divide into arteries supplying the face and extracranial tissues. Specifically, the external carotid arteries branch into maxillary and mandibular arteries. The maxillary arteries supply various superfici al and deep facial structures, while the mandibular arteries supply the lingual tissu es and soft and hard tissues of each dentary before terminating as mental arteries. According to Fraser and Purves (1960) before passing forward within the rostrum, each maxi llary artery gives off numerous branches, most notably: (1) the mandibular artery to the ramus of the jaw, (2) th e pterygoid artery to the pterygoid and palatine mus cles which fo rms a rich plexus of vessels distributed to the submucosa of the ai r sacs, (3) the deep temporal artery, (4) the lachrymal arteries, (5) orbital arteries and finally after passing through the infraorbital foramen (6) branches to the muscles of the nasal cavi ty and blowhole. Walmsley (1938) also showed a superficial temporal artery branchi ng from the maxillary artery of the fin whale, however it is unclear if this occurs only in fin w hales or in mysticetes (baleen whales), or was simply not observed by Fraser and Purves (1960). As mentioned earlier, blood to the cetac ean brain does not follow the traditional route, but rather flows thr ough intricate arterial retial elaborations. Indeed, Wilson (1879) called these elaborations the most beautiful of anatomic al objects. After leaving the left ventricle of the heart, arterial blood destined for the brain flows around the aortic
34 arch and away from the brain. Although carotid tributaries to the facial and pharyngeal regions still exist, blood targeting the br ain mostly bypasses the carotid system and enters both costocervica l trunks (Slijper, 1936). It should be noted that the costocervical trunks in the case of cetaceans have historically been assigned vari ous names and appear to show considerable variability in their branching patterns among cetaceans (Wilson, 1879; Ommanney, 1932; Slijper, 1936). Wilson (1879) called the costoc ervical trunk of the narwhal ( Monodon monocheros ) a posterior thoracic artery and noted it s similarity to the subclavian artery of the human. He also called the entire ve ssel including its caudal extension past its arch a posterior thoracic artery, where by later convention it was termed a supreme intercostal artery, equivalent to the thorac ic vertebral artery discussed by Schummer et al. (1981). Ommanney (1932) and numerous other authors also called this vessel the posterior thoracic artery (Galliano et al. 1966; McFarland et al. 1979; Vogl & Fisher, 1982). As seen in Schummer et al. (1981), the supreme intercostal artery of some domestic mammals is most commonly a branch of the costocervical trunk that also gives off an ascending cervical branch. For clarity and owing to the fact that similar anatomy seems to be present in cetaceans with respect to this branching pattern, I will from here on refer to the pos terior thoracic arteries as costocervical trunks. Interestingly, Melnikov (1997) st ated that in the sperm whale ( Physeter macrocephalus ), the thoracic rete is supplied by the internal thoracic artery and makes no mention of a costocervical trunk. Giv en the elaborate and more ventrolaterally located extensions of the thoracic rete found in physeteriids and kogiids, this is not a surprise, however a complete loss of the co stocervical trunk seems unlikely. It seems
35 more likely that the costocervical trunks are reduced, with most of their functional role being taken over by branches of the internal thoracic arteries. Indeed, my preliminary observations in kogiids suggest that like the sperm whale, the thoracic retia of the pygmy and dwarf sperm whale are primarily s upplied by dorsal offshoots of the internal thoracic arteries rather than the costocervi cal trunks. Whatever the actual branching pattern and correct nomenclature is on an indivi dual or species acc ount, the rest of the pathway appears to be relatively consis tent among the cetaceans that have been studied. From here, blood flows caudally into the supreme intercostal arteries-tributaries of the costocervical trunks or in ternal thoracic arteriesand then medially into the thoracic retia (O mmanney, 1932; Slijper, 1936; Galliano et al. 1966; McFarland et al. 1979; Vogl & Fisher, 1982; Melnikov, 1997). The thoracic rete on either side is composed of a densely packed meshwork of sm all, highly convoluted, muscular arteries embedded in fatty connective tissue (Wils on, 1879; Nakajima, 1961; Walmsley, 1938; Vogl et al. 1981; McFarland et al. 1979). Loosely present throughout these retia are a few nerves and veins. A smaller contribution to the caudal portions of the thoracic retia is made by some of the dorsal segmental ar teries, though the degree of contribution can vary significantly between species. After flowing into the mid and c audal portions of the thoracic retia blood flows craniomediad to supply two pairs of connected retia The first pair of retia essentially cranial extensions of the thoracic retia is the cervical retia that invest the fascial planes between muscles in the cervical region and surround the cervical vertebrae. In some species the cervical retia extend up onto the occipital r egions of the skull, and these occipital retia also receive blood from occipital branc hes of the external carotid arteries
36 (Melnikov, 1997). The second pair of retia is the spinal or epidural retia that invest the regions surrounding the cervical and thoracic sp inal cord. With th e exception of some finer muscular branches, all blood flowing through the thoracic and cervical retia that is destined for the brain eventually fl ows medially into the epidural retia The epidural retia from either side remain separate all the way to the foramen magnum, at which point they supply bilateral spinal meningeal arteries These spinal meningeal arteries take a dorsolateral course around the cerebellar hemis pheres, only to join near the pituitary gland and optic chiasm (McFarland et al. 1979). As the spinal meningeal arteries join, they form a mass of convoluted arteries that surrounds the pituitary gland and becomes the point of origin for the bilateral intracranial internal carotid arteries that supply the cerebral hemispheres. Intere stingly, much like in cows and pigs, the extracranial internal carotid arteries are present and of c onsiderable caliber in cetacean fetuses, but regress prior to or soon after birth (Slij per, 1936; Fraser & Purves, 1960; Melnikov, 1997). Indeed, numerous investigators have doc umented internal carotid contributions to fetal cetacean blood supply to the brain-often via a transient stapedial arterywhich is replaced before birth with t he adult architecture (Galliano et al. 1966; Lin et al. 1998; McFarland et al. 1979; Melnikov, 1997; Nagel et al. 1968; Nakajima, 1961; Ommanney, 1932; Slijper, 1936; Viamonte et al. 1968; Vogl & Fisher, 1982; Walmsley, 1938; Wilson, 1879). Arterial Synopsis As one examines the various patterns of ar terial supply to the brain that can be found from more basal vertebr ate clades all the way up to the more derived mammalian taxa, it becomes evident that mammals em ploy numerous different strategies for supporting the often large and/ or comple x brain. Gillil an (1972) coins the term
37 mammalian shift, to denote the change in blood supply of the brainstem and cerebellum from the caudal rami of the internal carotid arteries as seen in submammalian vertebrates, to the vertebrobas ilar system as seen in mammals. Only two years later, Gillilan (1974) states that the basilar arte ry does not supply blood to the brain of cows, sheep, or pig but rather that the tapering basila r artery receives all of its blood from the caudal rami of the internal carotid arteries As discussed earlier, newer texts suggest that the basilar artery may not be as insign ificant as Gillilans accounts state, since it receives input s from the vertebral and occipital arteries. Certainly in the horse and dog there are consi derable contributions of t he vertebrobasilar system to brain blood supply. In light of these texts, Gillilans claim of a mammalian shift may to some degree be inaccurate. What exactly might be considered a more basal mammalian ar terial pattern, and what a more derived one is unclear. As one progresses from the more basal world of fishes and enters the progressively more derived amphibian, reptilian, bird, and mammal clades it appears that t he origin of the caudal (posterior) cerebral artery becomes an important landmark in development of blood supply to the brain. With the exception of fish, all other nonmammalian vertebrates such as the frog, turtle, lizard, and bird have caudal cerebral (o r posterior telencephal ic) arteries that branch off of the rostral rami of the internal carotid arteries (Gillilan, 1967). In contrast to these nonmammalian vertebrate clades, all mammals appear to have posterior cerebral arteries branching off of either the caudal rami of the internal carotid arteries as seen in the dog, cow, sheep, and pig, or the confluences of the caudal rami and t he basilar trunks as seen in the horse (Gillilan, 1972; 1974).
38 With substantial vertebrobasilar contributions to the brain, the horse has a pattern very similar to the more basal opossum and a rmadillo, rather than the artiodactyl clade, which like non-mammalian vertebrates depend on t he internal carotid arteries for most or all blood supply to the brain. Additionally the horse is similar to the presumably more basal extant mammals (edentates and marsupials ) in having bilateral basilar trunks that converge with the caudal rami of the internal carotid arteries, and from which arise the caudal cerebral arteries (Gillilan, 1972; 1974). Conversely, like non-mammalian vertebrates, artiodactyls do not possess bilatera l basilar trunks. The caudal cerebral arteries of artiodactyls therefor e arise directly from the caudal rami of the internal carotid arteries (Gillilan, 1972; 1974). Whether this is a more basal or derived condition is up for debate. Confusing the matter further, t he intracranial arterial pattern of artiodactyls seems more similar to that of the more deriv ed non-mammalian vertebrates rather than the primitive mammals, the loss of a proximal, extracranial internal carotid and subsequent formation of a carotid rete mirabile are quite likely a more derived characteristic than the consistent course of the equine, opossum, and a rmadillo internal caroti d arteries. So it appears as though the horse has st rong similarities to the more basal mammalian taxa as well as the non-mammalian vertebrates, whil e the artiodactyls share basal and more derived characteristics. It should be noted that although most artiodactyls appear to have a carotid rete two small, tropical ruminants, namely the greater and lesser mouse deer, to do not possess a carotid rete (Fukuta et al. 2007). Extrinsic Venous Brain Drainage in Domestic Mammals Most venous drainage described for the head in domestic mammals occurs either through the external jugular veins or a comb ination of external and internal jugular
39 veins. Nonetheless, even when both jugular veins are present the internal jugular is usually much smaller in caliber and dr ainage through the external jugular vein dominates. Drainage through the neural canal via the epidural veins does occur in a number of species but is thought to take on variable significanc e. In such cases, blood can flow out of the intervertebral forami na into the vertebral veins and down to the cranial vena cava. As with the arteries of these regions, anastomoses ( rami anastomoticus cum vena occipital, Schaller et al. 1981) often exist between the cranial most portions of the vertebral veins and the occipital veinstributaries of the internal jugular veins--therefore some epidural drainage may eventually occur via the internal jugular veins (Ghoshal et al. 1981). Extrinsic Venous Brain Drainage in Marine Mammals As previously mentioned, the venous dr ainage from the head of marine mammals is poorly understood. N onetheles s, it is glaringly obvious from the start that significant differences exist. I have found only one description of the venous system in the head of the sea lion (King, 1977). In her descripti on, King (1977) noted that numerous branches of the internal maxillary vein ( v. maxillaris Schaller 2007) drained the tissues of the brain through various ventral skull forami na. Although difficult to discern, the illustrations and text suggest that the largest emissary vein of the internal maxillary vein exits the braincase through the jugular foramen. This morphology seems problematic since in terrestrial carnivores, it is typically the internal jugular vein that connects to the emissary vein of the jugular foramen. Indeed, all anatomical lit erature I have found regarding venous drainage through the jugular foramen suggests that in the terrestrial species that might be considered most closel y related to sea lions (e.g. dog, cat, pig, bear), the internal jugular vein transits t he jugular foramen (Anderson, 1989; Evans,
40 1993; Ghoshal et al. 1981; Hegedus & Shackford, 1965; Reinhard et al. 1962; Schummer et al. 1981). It should be noted that in domestic ungulate species particularly those that either lack or have a significantly reduced internal jugular vein (e.g. ox, sheep, goat)the emissa ry of the jugular foramen connects to a tributary of the external jugular rather than internal jugular vein. Schaller (2007) noted that the emissary of the jugular foramen connects with various veins in ruminants and pigs but provides no further details. Given the af orementioned patterns observed in terrestrial mammals, it seems possible that the vein described by King (1977) as the internal maxillary is in fact the internal jugular vein, which branches off of the common jugular vein at the same level as the external maxillary (facial) vein If that were the case, the rostral continuation of the exte rnal jugular vein depicted by King (1977) would in fact be the internal maxillary vein. Such morpholog y would seem to better fit the terrestrial carnivore paradigm, however the numerous adaptations and modifications pinnipeds have undergone should allow for some flexibility in our expectations. Like sea lions, the literature on seal venous morphology is also quite limited, most of it having focused on postcranial venous anatomy (Barnett et al. 1958; Blix, 2011; Folkow et al 1988; Galantsev, 1991; Harri son & Tomlinson, 1956; Hol et al 1975; Nordgarden et al. 2000; Ponganis et al. 2006; Rommel et al. 1995; Ronald et al. 1977; St. Pierre, 1974; Watson & Dowd, 1972). Certai n structures such as the hepatic sinus and epidural venous sinus have received much of the attention (Barnett et al. 1958; Blix, 2011; Galantsev, 1991; Harrison & Tomlinson, 1956; Hol et al 1975; Nordgarden et al. 2000; Ponganis et al. 2006; Rommel et al. 1995; Ronald et al. 1977; Watson and Dowd, 1972) while few records seem to exist regarding the cephalic venous morphology
41 (Folkow et al. 1988). Within the neural canal, phoc id seals have a large venous sinus wrapping around the dorsal and dorsolateral aspec ts of the spinal cordthe extradural (epidural) intravertebral vein (Barnett et al. 1958; Harrison & Tomlinson, 1956). Some researchers have reported that in harbor seals ( Phoca vitulina ) the jugular venous system is poorly developed, emphasizing that the internal jugular veins are degenerate (Barnett et al. 1958; Harrison & Tomlinson, 1956, Ronald et al. 1977; St. Pierre, 1974). This appears to be similar to the condition of the horse and numerous small ruminants that have no extracranial internal jugular vein at all (Schummer et al. 1981). However, rather than the external jugular veins being responsible for venous drainage of the brain as in these terrestrial mammals, the epidural system appears to dominate in seals. Harrison and Tomlinson (1956) st ated that it is obvious that in this species [harbor seal] the main intracranial venous drainage is via the extradural intravertebral vein. Additionally, they mention that although the external jugular ve ins are quite small at their origins in the face and scalp, they enlarge subs tantially as they receive large veins from the cervical plexus. Indeed, on closer ex amination it becomes apparent that large vessels from the dorsal cervical plexus--lik ely homologs of the in tervertebral veins-depart from the extradural intravertebral vein (E IV) and join the external jugular veins. It should be noted that the extradural intraver tebral vein referred to by Harrison and Tomlinson (1956), is the equivalent of the epidural veins and/or retia discussed throughout this document. Therefore, although initially most of th e blood likely departs the brain through the foramen magnum via the EIV, a significant portion of it likely reaches the cranial vena cava through the jugul ar system. In fact, due to the mobile nature of the seal cervical region and t he afore-mentioned vascular connections,
42 Rommel et al (1995) suggested the possibility of t he cervical plexus acting as a noncardiac pump to assist return of low pre ssure post-capillary venous blood to the heart. Harrison and Tomlinson (1956) made a simila r suggestion about a potential non-cardiac function. Whether the connections between the EIV and jugular system have only a simple emissary function or mo re complex pumping ro les, it is evident that the cranial vena cava receives blood from the brain via the jugular and epidural vessels, but it all appears to be originally sourced from t he foramen magnum and epidural vessels. There has been some research and spec ulation regarding the flow of blood through the EIV. This may be important because craniad flow could mean that blood in the caudal-most regions of the EIVorig inating in the abdom en and hindquarters could have valve-less access to the crani al vena cava via t he EIV. Ronald et al. (1977) found bidirectional flow within the EIV, and noted a change in direction of flow from craniad to caudad upon initiation of simulated dives. Similarly, Nordgarden et al. (2000) using Doppler ultrasonography, found bidirectional flow, but predominantly in the cranial direction. It has been suggest ed that the method of restra int and dive simulation could significantly alter flow pa tterns and may account for the observed inconsistencies (Sentiel Rommel pers. comm.). For instance, if a large volume of intracranial drainage occurs through the cervical intervertebral veins and into the external jugular vein, physical restraint of the cervical region with tightened straps or belts could impair venous return through the traditional pathways. In such a case, blood would have to take alternate pathways to the heart, and the path of least resistance in such a case would likely be caudad.
43 Similarly, compression through restra int of the abdominal regions could conceivably modify circulatory patterns by way of a Valsalva-type of phenomenon (Batson, 1960). Additionally, Ronald et al. (1977) schematized the simulated dives, in which they show lifting and lowering of the seals head. It is known among biologists and clinicians working closely with manatees that lifting of the head of an apneustic individual can often stimulate breathi ng, presumably as an engrained autonomic response to the species common surfacing pos ture (pers. obs.). It is possible that similar stimuli exist in seals that could al ter circulatory pathways. Additionally, the physical effect that gravity has on blood flow is obvious, therefore, lifting and lowering the head might artificially effect the dire ctional changes observed in this valve-less venous system (Epstein et al. 1970; Mitchell et al. 2008; Ponganis et al. 2006; Schreiber et al. 2003; Ueda et al. 1964; Valdueza et al. 2000). The potentially complex interplay between all these behavioral and physical altera tions and restraint methods is therefore unknown and should at least insp ire caution when interpreting these results. A question that arose regarding the EIV is that if it provides a lar ge, valve-less path for venous blood from the abdomen and hind limbs, it may be subject to elevated pressures during axial locomotion or other exertion that i nduces a Valsalva-type of phenomenon. A Valsalva phenomenon that increases intrabdominal pressures would likely force blood out of the abdominal vena cava and into the EIV. Although it would likely depend on the amount of intrabdominal pressure and the volume of blood being shunted, it would seem counterproductive to have a large venous lake juxtaposed to the central nervous system (CNS). Increased blood volumes in the EIV could elevate pressures surrounding the spinal cord. As stated by the Monro-Kellie doctrine (Bahram, 2001),
44 such elevations would have to be met by a de crease in cerebrospinal fluid in order to avoid damaging the CNS. Unlike the patterns of arteri al supply to the brain, v enous drainage of the brain in cetaceans appears to share some traits in common with seals (Barnett et al. 1958; Harrison & Tomlinson, 1956; Ronald et al. 1977). Much like the EIV of seals, cetaceans posses a large venous rete extending from withi n the cranial cavity through the foramen magnum and into the epidural space of t he neural canal (Costidis and Rommel, 2012; also see Figures 3-13 and 3-14). Unlike seal s, the cervical portion of this venous mass is composed of more numerous, finer caliber, anastomosing retial veins, rather than a large dorsally located sinus, how ever they eventually coalesce into two large, ventrally located anastomosing veins (Barnett et al. 1958; Harrison & Tomlinson, 1956; Hoogland et al. 2012). Just like in the seal, the epidural ve ins connect to the cranial vena cava by way of large intercostals veins in the crani al thoracic region (Harrison & Tomlinson, 1956). In cetaceans the first few (typically three) intercostal veins on the right side are substantially larger than any other ones and fu se at about the level of the ventral border of the vertebral centra to form a single la rge descending intercostal vein that becomes the right costocervical vein and drains directly into the cranial vena cava (Figures 2-2 & 2-3). This drainage pathway however appears to be one of the few similarities between seal and cetacean venous drainage of the brai n. It appears as though seals have a jugular system that becomes relatively degenerate rostral to the cervical region (Harrison & Tomlinson, 1956). Conversely cetaceans have an intact jugular system that receives contributions from an elaborate venous system investing the pterygoid,
45 intramandibular and facial regions (Costi dis & Rommel, 2012). Though varied between species, direct connections and/or substantial anastomoses exist between the ophthalmic and pterygoid plexusesmuch like is seen in carnivores (Schummer et al. 1981)--and between pterygoid and bas icranial vasculature via emissary veins (Costidis & Rommel, 2012). The emissary veins connec t the extracranial venous system to the intracranial cavernous sinus in the pituit ary and tympanoperiotic regions (Figures 3-13, 3-15, and 3-16). At least some cetaceans posses a dural v enous sinus network similar to that found in terrestrial mammals (Costidis & Romme l, 2012). This network includes a dorsal sagittal sinus and transverse sinuses. The tr ansverse sinuses connect to the cavernous sinus at the base of the calvarium near the pitu itary. It is not cu rrently known whether manatees and seals possess similar sinuses however my results suggest that manatees have a dense plexiform investment of dural and meningeal veins that cover much of the surface of the brain, while still retaining many of the major dural sinuses (Figures 5-13 to 5-15). The similarity of the cetacean dural sinuses to generic mammalian pattern suggests that their origin from the anterior cardinal vein may also similar. Manatees once again appear to exhibit an in termediate pattern. However, this time it is between the pattern seen in ceta ceans and seals and that seen in terrestrial mammals. Similar to the more traditional terrestrial mammal pathways, large facial veins receive contributions from the orbital regions ( angularis occuli). The facial veins empty directly into external jugular veins. Like their cetacean counterparts previously discussed, manatees also possess a subst antial epidural venous component that
46 connects directly to the large dur al sinuses of the skull (Barnett et al. 1958; Harrison & Tomlinson, 1956; Fawcett, 1942). This epidural component drains through numerous costocervical and intervertebral veins, howev er the details remain as yet unverified. The format of the following chapters intentionally differs from chapter to chapter. The considerably divergent morphology of the veins in the bottlenose dolphin and Florida manatee necessitated that I present their information differently. To simplify the conceptual difficulties of di scussing the many branches of the venous system, I followed a perhaps unusual pattern of describing the lo cation along a vein in the same way I would for an artery. For example, rather than referring to the smallest peripheral veinsin reality the beginning of the venous returnas the proximal or parent veins, I named the more peripheral veins the distal ve ins, and the large vein s that they empty into the proximal veins. Therefore, any o ccurrence of the word pr oximal or distal in reference to a vein, automatically implies that proximal is more central or toward the core while distal is peripheral or away from where the larger draining vein is. Given the fact that the venous system of the head is valv e-less, bi-directional blood flow within the veins is likely and therefore use of this reference system for venous descriptions should not be considered functionally in appropriate. It is important to note that some of the terms used herein have been used quite variabl y in the historical literature and may therefore take on a different or related but inexact meaning. As such and in order to aid in clarity, I have defined a few key terms as they are used in this particular text. PLEXUS: A general term describing any vascu lar structure with a net-like or branching geometry. This term has no specificity as to st yle of branching, parent and terminal vessels, location or type of tissue investment.
47 RETE MIRABILE: An arterial or venous structur e composed of seemingly random or disorganized branching vessels which c an continue to subdivide into capillary beds or re-coalesce into a parent vessel. A rete is a type of plexus. VASCULAR BUNDLE: An arterial or venous structure composed of a parent vessel that subdivides suddenly within a small distance to produce numerous smaller vessels of seemingly organized, linear char acter. After prim ary branching, the secondary vessels undergo little or no dimi nution in caliber, giving the appearance of a paintbrush or broom-shaped stru cture. Often associated with a thermoregulatory function.
48 CHAPTER 2 THE GROSS MORPHOLOGY OF THE VENOU S SYSTEM IN THE HEAD AND NECK OF THE BOTTLENOSE DOLPHINS ( TURS IOPS TRUNCATUS ) Chapter Foreword The vascular anatomy of domesti c mammals has been well described for quite some time, however, due to limited access, protected status, and complex vascular patterning, cetacean vascular anatomy has received sporadic attention. Specifically, the arterial system of cetaceans has receiv ed considerable attention while the venous system has been largely neglected or ignored. Fraser and Purves stated that anatomists such as Murie, Boenninghaus Carte & Macalister have, with justification, described the blood vessels only in very general terms, because the ramifications of the finer branches are exceedingly complex and form extensive retia mirabilia which are associated with the air spaces. (1960) Indeed, the venous system in the head of cetaceans--and especially odontocete cetaceansforms an intricate system of interconnected retia that surround the complex accessory sinus system, often times filli ng every available space between bones, muscles and fascia (Costidis & Rommel, 2012). Nonetheless, despite this relatively unique complexity many researchers have studied parts of this system and have therefore formed a foundation to act as a springboard for this research. Researchers such as Boenninghaus (1904), Fras er & Purves (1960), McFarland et al (1979), Nakajima (1961), Ommanney (1932), Ridgway et al. (1974), Slijper (1936), Viamonte et al. (1968), Vogl & Fisher (1981; 1982), and Wa lmsley (1938), have all described various parts of the vascular morphology of the head of cetaceans. Most of these descriptions focused on the anatomy of the arterial system or were studies on distantly related species (e.g. fin whales), and therefore pr ovide only modest contributions to this particular discussion. Others such as Fras er and Purves (1960) provide a significant
49 foundation from which to begi n to understand the complex venous anatomy in the head of the bottlenose dolphin. Materials and Methods Specimens for this study were obt ained from deceased, stranded cetaceans recovered along the east coast of the United St ates and west coast of Florida (Table 1). Beach-cast carcasses were recovered by marine mammal stranding networks authorized by stranding agreement s from the National Marine Fisheries Service U.S. Marine Mammal Health and Stranding Program. All work was conducted under a parts authorization from National Marine Fisheries Service, pursuant to 50 CFR 216.22 and 216.37, and with prior approval from Univer sity of Florida s IACUC (Permit #: 200801345). Specimens used for describing the vascula r anatomy were injected with contrast medium into either the venous system or both the venous and arterial systems, following the procedural methodo logy outlined by Holliday et al (2006). Prior to injection of contrast media, all specim ens obtained a vascular flush using 0.9% phosphate buffered saline solution and so me received a subsequent 5% neutral buffered formalin vascular perfusion to enable a protracted dissecting period. Most specimens were targeted for examination of the venous system, and were therefore flushed through the arteries and out of the veins. Once the effluent ran clear, the flush was stopped and the specimens were refri gerated while being allowed to drain for several hours. Prior to injecting the vascular contrast material, 5mL balloon catheters were placed in the vessels to be injected and inflated unt il a good seal was formed. In specimens that were to be imaged via computed tomogr aphy (CT), the vascular system of interest
50 received a mixture of liquid latex and barium sulfate sus pension (Liquid Polibar Plus, Bracco Diagnostics Inc.), while vessels only destined for dissection received pure latex (Carolina Biological Inc.). All specim ens were refrigerated for 2 days following injections, to allow the latex cast to cure, and if unpreserved with formalin, where subsequently frozen at oC. Specimens that were imaged via CT were scanned at the thinnest slice thickness possible bas ed on the specimen length, and whenever possible the volumes were reconstructed to 0.5mm thickness to allow high resolution imaging of fine caliber vasculature. The resultant DICOM data was post-processed using Amira software (Visage Imaging Inc., San Diego, CA) on a Gateway desktop with memory and processor upgrades. Foll owing imaging and po st-processing, all specimens were thawed and dissected to va lidate and/or clarify imaged structures. It should be noted that given the elaborat e intracranial arterial and venous components reported on in the lit erature and the phylogenetic prox imity of the dolphin to the cow (compared to other domestic mammals), I chose to dissect two cow specimens. The head and neck portion was removed fr om two cow cadavers being used for veterinary teaching purposes at the Univer sity of Floridas College of Veterinary Medicine. I injected both specimens with a mi xture of liquid latex and barium sulfate suspension, according to the aforementioned protocol used for the dolphin specimens. Following CT angiographic imagi ng, the specimens were grossly dissected in order to outline certain structures of interest, namel y the plexus of the deep facial vein, the cavernous sinus, and the carotid rete
51 Results The morphology of the veins of the head and neck will be separated into sections, those branching off of the brachiocephalic tr unks and draining the regions of the neck, and those veins that are responsible fo r draining the tissues of the head. Neck The following section describes the veins in the region of the neck. Only two specimens used for thi s study included the major cranial branches of the vena cava, therefore, variability of those branches among individuals could not be assessed. The major cranial branches of the cr anial vena cava are the inter nal jugular, external jugular and facial veins (Figures 2-1 to 2-2). An addi tional large inflow tract to the cranial vena cava is formed by the confluence of the first three right intercostal veins. These veins merge shortly after their emergence from the in tervertebral foramina, at the level of the ventral margin of the cervical vertebral centra. After their fusion, a single large costocervical venous trunk travels in a slight ventrocaudal direction and connects directly to the right side of the cranial vena cava (Figure 2-3). Following the emergence of the cranial vena cava from the right side of the heart and just cranial to the fusion of the right co stocervical trunk, the vena cava bifurcates into two large trunks that fo rm the right and left brachiocephal ic trunks. As is common to domestic mammals, the brachiocephalic tr unks are formed by veins draining the pectoral limbs and head. Specifically, the ri ght brachiocephalic trunk is in line with the cranial vena cava, forming the visual continuation of the vena cava. The left brachiocephalic trunk forms an almost horizont al arch that emerges from the medial aspect of the vena cava and travels to the le ft side of the neck (Fi gures 2-1 & 2-2). Along its course to the left side, the left brachiocephalic trunk traverses over two
52 brachiocephalic arteries, giving off along it s course numerous small veins that emerge from its caudal border and travel caudad. These veins invest the tissues of the thymus and cranial mediastinum. On ei ther extreme, at least two slightly larger veins emerge from the left brachiocephalic trunk and travel caudad. After a distance of only about 1cm, these veins join and wrap around their co rresponding arteries, and travel along the medial surface of the sternum as the internal thoracic veins that supply branches to the sternum ( rami sternalis ), mediastinum ( Vv. mediastinales ), intercostal muscles ( Vv. intercostals ventralis ), diaphragm ( V. musculophrenica ) and ventral muscles of the thorax and abdomen ( V. epigastrica cranialis and V. epigastrica cranialis superficialis, Schummer et al. 1981; Schaller, 2007). After their proximal plexif orm arrangement, each internal thoracic vein forms a pai r of anastomosing ve ins that border the concomitant artery along much of its path caudally. From the cranial aspect of the middle region of the left brachiocephalic arch emerge numerous small veins that travel a short distance craniad into the substance of the adjacent thyroid gland. Although rather atypical, these veins likely constitute the caudal thyroid veins seen in domestic mammals ( Vv. thyroidea caudalis Schaller, 2007). The emergence of the facial and jugular branches appears to be variable and the exact morphology may be dependent on the ontogenetic stage. In the adult specimen (Figs 2-1 and 2-2), the right axillary, facial, in ternal and external jugu lar veins all split off of the right brachiocephalic trunk at the same location. Additionally, sizable anastomoses emerged from the beginning of the left brachiocephalic trunk and joined the right external jugular vein. On the left si de a slightly different pattern was observed. In the adult specimen, the left internal j ugular and facial veins emerged as a common
53 trunk at the terminal bifurcation of the left brachiocephalic (the other terminal branch was formed by the left axillary vein). The left external jugular vein emerged as a lone branch from the cranial aspect of the left branchiocephalic trunk, just medial to the combined trunk of the internal jugul ar and facial vein (Figure 2-2). In the young specimen, a different pattern was observed in which the bilateral asymmetry was more pronounced (Figure 2-4). On the left side, the internal jugular and facial veins emerged from the left brachiocephalic trunk as a common trunk and stayed as such until about the mid-cervical region. At this point, the facial vein took a lateral path, while the internal jugul ar vein continued on the same course. The left external jugular vein emerged directly from the cranial aspect of t he left brachiocephalic trunk and traveled in a straight path to the cervical region (Figure 2-4). On the right side, the branching morphology was slightly different. The right internal jugular and facial veins arose together from the same aspect of the ri ght brachiocephalic trunk. The facial vein immediately began traveling in a more lateral path, while the internal jugular traveled directly craniad in a path similar to its left counterpart. Like its le ft counterpart, the right external jugular vein emerged from t he cranial aspect of its corresponding brachiocephalic trunk and traveled straight craniad, along a path just ventrolateral to the internal jugular vein. Interestingly, alt hough a single slightly larger parent external jugular vein was found emerging from the brachiocephalic trunk, it was accompanied by numerous only slightly smaller veins that formed a plexus immediately upon emerging from the brachiocephalic trunk. This plexus which I called the venous satellite plexus of the common and external carotid artery ( vena plexi comitans arteria carotidis communis and externa ) surrounded the common carotid and s ubsequently the external carotid
54 artery and contributed to the formation of the external jugular vein (Figures 2-5 and 2-6). An additional plexustermed here the vena plexi commitans arteria carotidis interna was found surrounding each of the internal caroti d arteries (Figures 2-5 & 2-6). As is seen in domestic mammals, the facial veins tr avel lateral to the stylohyoid bones while the internal and external jugular had a medial position relative to the stylohoid bones. Internal jugular veins As mentioned earlier, the inte rnal jugular vein s emerge from the brachiocephalic trunks in common with the facial veins. In the adult specimen that included the cervical region, this pattern was bilaterally symmetr ical. In the younger specimen however, the left internal jugular and facial veins emerged as a common trunk, but did not divide off of the common trunk until the level of the midcervical vertebra. Despite the proximal branching pattern of the inte rnal jugular veins, they always send a sizable branch dorsorostrad toward the paraoccipital process of each exoccipital bone. This branch forms the emissary of the j ugular foramen that enters the calvarium and joins with the dural sinuses of the brain case (Figures 3-13 to 3-15). The intern al jugular veins also send a sizable rostral anastomotic branch. In both the adult and younger specimen, throughout their course across the cervical region, both internal jugular ve ins sent at least three sizable ventral anastomoses that contributed to the formation of the aforementi oned venous plexuses that surround the common carotid and external carotid arteries (Figures 2-5 and 2-6). The plexuses were described in brief by Ridgway et al. (1974) and Costidis and Rommel (2012). These peri-carotid plexusestermed here veni plexu comitans arteria carotidis communis and veni plexu comitans ar teria carotidis externa in order to follow standard veterinary anatomical nomenclature but also reflect their plexiform structure
55 are formed by the three large contributions from the internal jugular veins and innumerable contributions from the external jugular veins (Figures 2-5 and 2-6). In domestic mammals the vena comitans arteria carotidis externa a singular, nonplexiform vein, is typically a branch of the in ternal jugular vein. However, owing to the fact that in the dolphi n the external jugular veins are ti ghtly juxtaposed to the external carotid arteries and contribute countless anast omoses to the peri-carotid plexuses, it seems meaningless to assign either the internal or external jugular veins as the parent vein or primary targets of drainage for these pl exuses. Additionally, these distinctions of nomenclature are likely to be functionally insignificant. Although perhaps of lesser relevance to the discussion of head circulation, I should mention that in one of the specimens that I inject ed, which included the entire neck and much of the thorax, two interesti ng venous structure were discovered (Figure 2-6). Surrounding the cervical portion of the esophagus and extending into the submucosal tissues of the laryngopharynx is an extensive venous plexus. Although not as dense or voluminous as the plexuses a ssociated with the accessory air sinus system or CNS (Chapter 3), it is a considerable stru cture surrounding the ent irety of the cranial esophagus but located peripheral to the in trinsic esophageal musculature. As it progresses rostrad it presents a shallower po sition, eventually becoming submucosal in the pharyngeal region. Unfortunately the rostra l most portions of this plexus did not inject and could therefore not be followed. Nonetheless, delicate dissection revealed uninjected portions that appear to invest the root of the t ongue and potentially anastomose with the venous lingual vasculature. Interestingly, this pharyngeal plexus did not inject in any of t he other specimens, despite the good injection of the lingual
56 vasculature. This may suggest that either the two vascular systems are not well connected, or the connections involve vessel s with either intrinsic or extrinsic high resistance. In addition to the aforementioned per i-esophageal and pharyngeal plexuses, another prominent plexus was discovered withi n the cranial portion s of the trachea. These three plexuses had numerous rostra l connections between themselves at the level of the larynx. The endotracheal plexus --as considerable as that surrounding the esophagus--is located within the mu cosa or submucosa of the trachea, as it is visible through the luminal surface. I believe that this venous plexus represents the structure described histologically by Cozzi et al. (2005). The endotracheal plexus appears to be primarily drained by branches of the internal jugular veins that penetra te into the trachea just caudal to the larynx. Owing to the fact that the entire cadaver was not available and therefore not injected, it is not possible to determine if other venous contributions to the endotracheal plexus exist further caudad. External jugular veins Despite the aforementi oned differences in the formation of the trun k of the external jugular veins, the general morphology of the veins during their course across the cervical region is fairly consistent. After emerging from the brachiocephalic trunks, both external jugular veins travel on a straight and relatively hor izontal path toward the head, roughly paralleling the trachea. Along their course they give off numerous medial branches that invest the middle and crani al portions of the thyroid gland ( Vv. thyroidea medialis and cranialis ) (Figures 2-1 and 2-2). During th is course and before they reach the level of the laryngeal cartilages, both external jugular veins receive numerous large anastomoses from the internal jugular veins. These anastomotic branches fuse with the
57 external jugular veins at an oblique, caudodorsal angle (Figures 2-5 and 2-6). If one were to assume unidirectional flow of venous blood from the head to the cranial vena cava, the angle of these anastomoses would suggest that they facilitate drainage of the blood in the external jugular veins into the in ternal jugular veins. Interestingly, each external jugular vein is composed of a la rge vein that divides and rejoins along its length, forming a type of cradl e that embraces the external carotid artery. Throughout their course along the neck, each external j ugular vein is joined by multiple small anastomoses from the aforementioned vena plexu comitans arte ria carotidis communis and externa Head The following section describes the veins in the region of the head. At the level of the larynx, each exter nal jugular vein give s off numerous small br anches that travel ventromedial to the larynx the more rostral ones (v. laryngea cranialis) diving through the thyroid fissure between the thyroid and ar ytenoid cartilages, while the more caudal ones invest the muscles of the larynx ( v. laryngea caudalis ) (Schaller, 2007). No hyoid ( arcus hyoideus ) or laryngeal venous arches ( arcus laryngeus ) were observed in the adult specimens, however the young specimen had a venous structure just rostral to the basihyal bone of the hyoid apparatus which connected the left and right sides of the venous system and may have been equivalent to the hyoid venous arch. Additionally, this hyoid vasculature connected to the ex tensive pharyngeal and peri-laryngeal venous plexus described above. The extensive natur e of this plexus and poor injection quality precluded me from decipherin g its connections, however it could represent an elaboration of the laryngeal vasculature, including the laryngeal venous arches.
58 After giving off the laryngeal branches, t he external jugular veins curve around the stylohyoid bones of the hyoid ap paratus to become the maxill ary veins. The maxillary veins follow the pterygoid crest, giving off numerous pterygoid veins from their dorsal aspect. The pterygoid veins anastomose back forth with each other to form an intricate pterygoid plexus that lines the pterygoid sinu s and eventually all of it s lobes (Chapter 3). The maxillary veins then curve rostrodorsally sending numerous small branches to the palate. These branches form a palatine plexus that invests the soft palate and travels rostrad along the roof of the mouth. Just dorsal to the palatine branches, each maxillary vein sends a plexus of veins that projects past the hamular proc ess of the pterygoid bone, into the hamular lobe of the pterygoid sinus (Chapter 3). Dorsal to this structure, at least two large (~1mm diameter) veins emerge from the maxillary vein and travel through and invest the palatopharyngeal muscles ( mm. palatopharyngeus ). The maxillary veins then curve slightly la terad and continue to send off small veins that contribute to the extensive venous plexus of the a ccessory sinus system; specifically the ptery goid sinus and the lobes associated with it (preand post-orbital, and anterior). At this point, the maxillary ve ins lose their singular identities as large veins, breaking apart into countless small and large veins that form part of the dorsolateral component of the pterygoid plexus. This pl exus travels dorsad, merging on its lateral aspect with dorsal offshoot of t he intramandibular plex us and surrounding the maxillary artery on its way to the ventral infraorbital foramina (Figures 3-14 and 3-15). Near the ventral infraorbital foramina on the ventral aspect of t he maxillary bones, the plexus sends a large plexifo rm extension rostrad to become the plexus of the anterior lobe of the pterygoid sinus (Figures 3-13 to 3-15). The rest of the plexus undergoes a
59 diminution as it enters the ventral infraorbital foramina to become the infraorbital veins that exit on the dorsal aspec t of the maxillary bones. While traveling through the infraorbital canal, the infraorbital veins s end rostral veins within the maxillary bone (Figures 2-8, 2-9 & 2-12). These palatine ve ins travel within the maxillary bones to the distal tip of the rostrum, all the while pr oviding small veins t hat drain the teeth ( rami dentalis, Schummer et al. 1981). Once the infraorbital veins emerge from the dorsal infraorbital foramina, they break up into countless veins oriented in the rostral, dorsal, and caudal directions (Figure 2-9). The mo st ventral veins are oriented rostrad and travel along the surface of the maxillary and premax illary bones, investing the connective tissue on the rostrum (Figures 28 & 2-9). The dorsally oriented branches radiate in a fan-like fashion to invest the fatty tissue of the me lon. The caudally oriented branches form two distinct groups. The mo re rostral group invests the muscles and connective tissues associated with the blowhole as the lateral nasal veins. The more caudal group forms a sinus-like dilation that extends along the facial fossa of the ascending process of the maxillary bone, gi ving off branches here and there into the surrounding muscular and connective tissues as the dorsal nasal veins (Figure 2-10). Lingual and facial veins The origin of each facial vein is usually in common with the internal jugular vein. Immediately after its emergence from the brachiocephalic trunk, the facial vein takes a more lateral route than either jugular branch, emerging laterally between the brachiocephalic us and sternocephalicus (ste rnomastoideus) muscles. During its path along the cervical region, the facial vein gives off very few branches, those being muscular branches to surrounding muscles (e.g. m. sternocephalicus). Shortly after the facial vein separates from t he internal jugular vein and bef ore it exits laterad between
60 the brachiocephalicus and sternocephalicus muscles, it gives off its first branch consistent with the superficial cervical vein ( v. cervicalis superficialis ; Schaller 2007). The second branch of the facial vein curves ventrorostrad and travels directly to the submandibular region as the submental vein. The superficial cervical veins extend dorsad under cover of the brachiocephalicus muscle and give off a sizable dorsorostrally oriented branch that emerges from beneath the dorsal aspect of the brachiocephalicus muscle and travels superficial to the longissimus muscles that insert on the temporal and nuchal crests. This ve in and its many small twigs travel to the dorsal region of the supraoccipital and inte rparietal bones where they anastomose rostrad with the plexus of t he caudal auricular vein described below. In addition to the supraoccipital branch, the super ficial cervical vein also gives off near its origin a vertically directed vein that trav els to the dorsal surface of the longissimus muscles of the cervical region. Along its way it sends numerous rostrally oriented twigs to the large superficial cervical (prescapular) lymph nodes and a sizable dorsocaudally oriented vein that wraps around the craniodor sal border of the scapula ( v. prescapularis ; Schaller 2007). A few centimeters rostral to the emergence of the superficial cervical vein, the facial vein bifurcates to send a ventrorost ral submental vein, and a dorsorostral facial vein. Shortly after the subment al vein leaves, the facial ve in emerges onto the surface of the face by exiting between the brachiocephalicus and sternocephalicus muscles. As soon as it emerges from between those muscles, it sends its first large branch off of its dorsal aspect. This branch extends in a cu rved fashion, first dorsally and when it reaches the external auditory meatus dorsoc audally toward the temporal and occipital
61 regions. Along its course to the EAM, numerous small muscular branches leave the caudal aspect and ramify into the muscles that course under the parent vein (e.g. m. sternocephalicus ). At the level of the EAM, it give s off numerous rostral twigs to the connective tissues surrounding the EAM, and is therefore consistent in location and drainage field with the caudal auricular vein ( v. auricularis caudalis; Schaller 2007) or occipital vein. After sending the twigs to the EAM, the caudal auricular vein extends dorsorostrad, forming an extensive plexus t hat spreads out over the surface of the temporalis muscle and the insertion of the neck muscles (m. longissimus and multifidis ) along the nuchal and temporal crests and s upraoccipital bones (Figure 2-11). This extensive plexus forms countless rost ral anastomoses with veins surrounding and investing the blowhole and associated ti ssues, and caudal anastomoses with branches of the superficial cervical vein described ab ove. The plexus also heavily invests the dorsal palpebral tissues and connective tissues of the dorsal regions of the orbit. It should be noted that although I named the par ent vein the caudal auricular vein, its contribution to the external auditory meat us is relatively modest compared to the veins it sends to the temporal plexus. This should come as no surprise given the lack of external pinnae and the extensive dorsal ep axial musculature of the head associated with locomotion, and nasal musculature associated with vocalization. Perhaps more importantly, the caudal auricular vein bifurcates shortly after its origin at the facial vein, sending an even larger deep branch that ex tends caudad along the temporal crest to the occipital region where it anastomoses with a rostral branc h of the deep cervical vein ( v. cervicalis profunda; Schaller 2007). Given that the occi pital branch is larger than the caudal auricular branch and that in domestic ma mmals the two veins typically arise from
62 the maxillary vein in close proximity, it ma y be more appropriate to call the parent vein the occipital vein, and only its rostral tributary the caudal auricular vein. Interestingly, the caudal auricular vein in domestic species is a branch of the maxillary vein, however in those species th e external jugular ve ins and therefore the maxillary veins take a relatively s uperficial path through the neck. In Tursiops the external jugular and maxillary veins are de ep to muscles of substantial size (e.g. m. sternocephalicus sternohyoideus and brachiocephalicus ). This morphology is likely due to the unusual nature of cetacean locomotion, that being axial rather than appendicular, unlike that of domes tic mammalian species. Given this coverage of the maxillary vein by the muscles, a caudal auricular vein that arises fr om the maxillary vein would have to either pierce the muscle bel lies or transit between these large muscles, both of which situations would expose a small ve in like that to significant collapse forces during locomotion. Since dolphins spend much of their time swimming, such a strategy would seem counterproductive. The obvious solution would be to transition drainage of the caudal auricular vein, and any other smaller superficial vein for t hat matter, into the large, superficially located facial vein. Fo llowing these lines, I was not able to observe a distinct superficial cervical vein arising from either the external jugul ar or maxillary vein. As outlined above, the superficial cervical vein in Tursiops is a branch of the facial vein. After the facial vein gives off the caudal auricular/occi pital vein, it continues at an oblique angle across the caudal ramus of th e mandible, on a path toward the ventral aspect of the orbit. Just before it crosses lateral to the dentary, it sends a sizable mandibular alveolar vein just caudomedial to the mandibular condyle (Figures 2-9, 2-11 and 3-13 to 3-14). This vein fans out to fo rm a large venous plexus that invests the
63 intramandibular fat body within the hollow dentary. From this intramandibular venous plexus arises the continuation of the mandi bular alveolar vein that takes on a periarterial venous retial character as it travels to t he mental foramina, all the while surrounding the mandibular alve olar artery (Chapter 3). Branches of the facial vein investing the orbit and dorsal maxillary regions After giving off the mandibular alveolar vein the facial vein continues toward the orbit and follows the slender j ugal bone that forms the ventra l border of the orbit. Along this course it sends numerous small twigs that form an extramandi bular venous plexus that invests the extramandibular fat body (Chapt er 3). Numerous t wigs from the facial vein also forms a delicate plexus that surr ounds the adjacent facial nerve on its course across the ventral orbit to the antorbital notch. As the faci al vein reaches the ventral border of the orbit, it sends multiple large (2-6mm diameter) branches mediad. These medial branches travels ventral to the jugal bone and anastomoses with an extensive ophthalmic plexus (see description of external ophthalmic venous rete mirabile below) that surrounded the eyestalk and globe (Figures 2-9 2-12, & 2-13). At the same level as the ophthalmic anastomoses, the facial vein also sends rostral and ventrorostral twigs that form a fine plexus that invests t he tissues at the angle of the mouth. The dorsal portions of this plexus coalesce to form a mandibular labial vein that gently curves ventrorostrad to follow the lower lip and gum line (Figure 2-9 and 2-12). From there the facial vein continues on its oblique angle to the antorbital notch, at which point it wraps around the notch and ramifies into num erous veins. The largest of these veins immediately turns rostrad and tr avels within the upper lip to fo llow the lateral margin of the maxillary bone as the maxillary labial vein ( v. labialis superficialis and profunda Schummer et al. 1981) (Figures 2-9 and 2-12). The rest of the branches of the facial
64 vein fan out in the dorsorostral, dorsal, and dorsocaudal direction, forming a slight convexity as they disperse. This fan-shaped structure appears to follow the indistinct fascial layer between the blubber and melon fat. The lingual veins arise from the external jugular veins rather than from a common linguofacial trunk. Their terminations are typically found to be from the region where the external jugular veins turn into maxillary veins, around the level of the tympanoperiotic complexes. Sometimes joining the maxillar y veins as singular veins but usually as a few anastomosing veins, the lingual veins travel in a fairly direct path to the root of the tongue where they divide into the various li ngual, sublingual and deep lingual branches. Along their course through the tongue, they contribute to the formation of a peri-arterial venous rete that surrounds the lingual artery. CNS Veins The following sections describe the veins associated with the brain and cervical spinal cord. Given the complexity encountered in the rest of the veins of the head, it should be no surprise that the ve ins of the cervical spinal cord and brain are similarly quite complex. The neural canal is fill ed with a large mass of tortuous arteries and veins forming a retie that surrounds the spinal cord on all aspects (Figures 2-7, 2-14 & 2-15). This rete extends well into the calvarium, ev entually contributing heavily to the formation of the cavernous sinus. As the epidural venous rete enters the braincase through the foramen magnum, it sends an in tricate and rather voluminous plexus ventrad and ventrolaterad. This plexus foll ows the calvarial surface of the exoccipital and basioccipital bones to the clivus and extends laterad to the tentorium cerebelli on each side in association with the spinal m eningeal arteries, thereby forming a net-like structure that cradles the ventral and lateral aspects of the cerebellum (Figures 2-14 &
65 3-12). The plexus extends rostrad to fuse seamlessly with the caudal intercavernous sinus that lines the caudal aspect of the pitu itary gland (Figure 2-15 ). The dorsolateral portions of the cerebellar plex us re formed by larger, less plexiform veins (presumptive spinal meningeal veins) that enter the foramen magnum and travel along the inner surface of the ventral parie tal bone in close association with the spinal meningeal arteries. The more dorsolateral spinal m eningeal veins and the smaller veins of the denser ventral portion of the cerebellar plexus all anastomo se rostrolaterad with the large temporal sinus. Rostrally, all of these st ructures coalesce in the region of the oval foramina along the dorsal aspect of the basis phenoid bone. Where they coalesce, a thick venous plexus is formed as it receives rostral inputs from the cavernous sinuses, ventral input from the emissa ries of the oval foramina, and caudal inputs from the epidural rete (Figure 3-12 D-F). The plexus on t he basicranium then extends rostrad to become the cavernous sinus t hat surrounds the pituitary gl and (Figures 2-15 and 3-12). Superficially, the cavernous sinuses of the dolphin do not appear to be formed by two discrete bilateral sinuses joined by singular rostral and caudal intercavernous sinuses. Instead, the morpholog y of the cavernous and intercavernous sinuses is more reminiscent of that seen in the cow, in which an intricate venous plexus covers the ventral aspect of the brain and brain stem. Like the cow, each cavernous sinus is intimately associated with an arterial component. However, angiographic imaging allowed me to digitally dissect away the fi ne vessels of the cavernous sinus. Following digital removal of the finer ve ssels of the cavernous sinus, a distinct and fairly typical annular structure was seen surrounding the pitu itary gland (Figure 2-15). Despite some striking similarities between the CNS vasculatu re of the cow and do lphin, significant
66 differences do exist. Firstly, the cows ca vernous sinus pales in comparison to the sheer volume of the one pr esent in the dolphin. Secondly, the venous plexus coextensive nature with an arterial counterpart is also considerably more voluminous in the dolphin and extends from the orbital fissure s caudad past the cervical portions of the neural canal (Figure 2-14). This is unlike the cow in which the arterial component becomes much more diffuse at the level of the pons, after which it takes a character typical of domestic mammals. It is also unlike the cow because the intricate plexiform nature of the cavernous and intercavernous sinuses of the cow gets substantially simplified at the level of the pons at which point two rela tively discrete longitudinal basilar sinuses are formed. I found no such discrete nature to the basilar system of the dolphin, and could therefore ar gue that a traditional basilar system does not in fact exist in the dolphin. Instead, a widely dissemi nated, diffuse and coalescing venous plexus covers the entire floor of t he calvarium and acts as a drainage route for the various ventral cerebral, ponti ne and cerebellar veins. Rostrally, each cavernous sinus extended an intricate plexus at an oblique rostrolateral angle. These plexuses follow and surround the optic ner ves as they course through the orbital fissures toward the eyes, as the equivalents of the emissary veins of orbitorotunda (Ghoshal et al. 1981). Slijper (1936) and McFarland et al. (1979) described similar arterial components in the porpoise and bottlenose dolphin, respectively, and named them the internal ophthalmic retia mirabilia Given the coextensive nature of these venous retia, it seems r easonable to suggest they receive the same nomenclature (internal ophthalmic venous retia mirabilia ), distinguishing them from each other by the addition of a venous or arterial designation. Together, they may
67 be said to form the internal ophthalmic pl exus. These internal ophthalmic venous retia extend out of the orbital fi ssures to the orbit as the external ophthalmic venous retia Once at the orbit they contri bute largely to the formation of the ophthalmic plexus that surrounds each eye (see description above). Discussion Comparative Venous Anatomy Certain aspects of the venous morphology e ncountered in the head and neck of the bottlenose dolphin have elements very remi niscent of what is found in domestic mammals, while other structures appear to be either completely novel features or extensively elaborated struct ures of domestic mammals. With the exception of the anastomoses, the branching patterns of the pr oximal internal and ex ternal jugular veins hold no overwhelming surprises given t he general variability found in domestic mammals. Additionally, many of the more distal branches of the face and skull (e.g. facial vein, emissary of the jugular fora men, etc.) had very familiar morphologies. Despite these familiar traits, some subst antial differences do exist. Perhaps most notably, the dolphins head is invested with a surprising volume of venous plexuses associated with the accessory sinus syst em, acoustic fat bodies, and the central nervous system. The elaborate vasculature of the fat bodies and sinuses might be easily explained by the fact that both of those structur es appear to be enlarged and conceivably specialized adaptations specific to the demanding lifest yle of the dolphin. The vasculature of the CNS how ever may not be explained simp ly by the large brain of dolphins, since no vascular structures of that magnitude have ever been found associated with terrestrial ma mmals of lesser or greater encephalization. As discussed
68 below, I find the most plausible explanations for such magni ficent vascular investment to likely be related to either therma l or rheologic functions, or both. Ontogenetic vs. Individual Variability It was clear from the di ssection of the younger and adult specimens that although a certain degree of variability and possible ontogenetic modification likely occurs, the main patterns are remarkably consistent at l east from the juvenile or subadult stage and on. It is possible that because of the re latively advanced developmental state of the younger specimen, the venous morphology had already acquire d its adult character. Indeed, it has been well documented that certain parts of the arterial system undergo modifications as a cetacean transitions from the near-term fetal st age to the neonatal stage and then juvenile stage. Most notable perhaps is the regression of the distal extracranial internal carotid artery, and subs titution of blood supply to the brain via the epidural circulation. Although the adult morp hotype of blood supply to the brain is different in the cow, it shar es a common trait with respect to regression of the distal extracranial internal carotid artery. Despite the relative consistency I observed in the venous system, anastomoses between the major trunks abound, and these anastomoses may take on lesser or greater roles in draining blood. For instance, in the adult specimen that included the cervical porti ons of the veins (ECW -005), the root of the right external jugular vein was formed by two vein from two locations: 1) a large branch that emerged from the right brachiocephalic trunk in common with the internal jugular and linguofacial veins, and 2) at least two smaller branches that emerged from the left brachiocephalic trunk. The left ex ternal jugular vein was formed by two branches, a large and small br anch both of which emerged fr om the left brachiocephalic trunk proximal to the left axillary and common left internal jugular and linguofacial veins.
69 Therefore, one might envision that enlargement of the small branc hes of the right external jugular and reduction of the large branch would result in a more bilaterally symmetrical arrangement. Interestingly, the younger specimen showed a pattern more like that, in which the large contributions to the right external jugular vein were made by the branches from the proximal left brachiocephalic trunk. This difference may represent an ontogenetic shift in which the diameter of t he branches changes due to development of local pressure differences resulting in gradual atrophy or reduction of certain branches and enlargement of others. Fusion of the small and large branches and ac quisition of other small anastomoses may also result in the obser ved adult pattern. Alternatively, it may simply represent individual variation. Given the low sample numbers and known plasticity of veins it is impossible for me to know what the driving force behind these differences is. Functional Implications of CNS Vasculature The architecture of the venous structur es surrounding the brain and spinal cord provide tantalizing suggestions of functi onal implications beyond simple drainage of blood from the CNS. The sheer volume and complexity of the epidural veins and dural sinuses and their intimate juxtaposition to sim ilar arterial networks are features that are hard to ignore. Such intimate arteriov enous associations are commonly present in regions with specific thermoregulatory func tions. The simplest and perhaps closest terrestrial examples would be the pampiniform plexus of the scrotum of mammals and the carotid rete and associated cavernous sinus in the well-described cow and other smaller ruminants (Gillilan, 1974; Scholander, 1958). Both of these structures compose networks of tortuous, fine caliber vessels with a high surface area, quite similar to the
70 arteriovenous structures observed surrounding the CNS of Tursiops Indeed, I found it easy to envision the intracranial venous plexus of Tursiops as an extreme elaboration of the cavernous sinus found in the cow. Perhaps acting in accordance to similar embryologic or phylogenetic pre ssures, the epidural veins investing the cervical portion of the neural canal underwent similar el aboration to form the observed network. Whatever the complex mechanistic forces may be for such considerable development of the vasculature associated with the CNS, the benefits of such a system seem easier to speculate about. As menti oned above, the most obvious such benefit would appear to be the potential for thermal interactions. Such interactions might manifest as either thermal exchange between the arterial and v enous systems, therma l exchange between the vascular system and CNS, or both. Temperature exchange between the arterial and venous system s investing the brain is a well-documented occurrence. Indeed, numerous mammalian species have been shown to juxtapose cooler venous blood to warmer arterial blood in order to reduce the temperature of the arterial blood bound for the brain. Such a mechanism, termed Selective Brain Cooling (SBC) has b een shown to result in maintenance of normal brain temperature or at least reduc tion of elevated temperature in spite of considerable elevations in core body te mperature (Baker, 1972, 1979, 1982; Baker & Chapman, 1977; Baker & Hayward, 1968a; 1968b; Baker et al. 1974; Caputa et al. 1976; 1979; Fuller & Baker, 1983; Fuller et al. 2003; Maloney & Mitchell, 1997; Maloney et al. 2007; Mitchell et al. 2007). The benefits of such br ain cooling are considerable during periods of exposure to elevated ambi ent temperatures or under conditions of rigorous exertion. Given the higher metabolic rate of dolphins (Williams et al. 2001),
71 such an ability to cool the arterial blood bound for the brain would seem similarly beneficial. Some terrestrial mammals (e .g. rabbits) have been shown to use cooled venous blood to directly cool the brain, rather than the arterial blood destined for the brain (Caputa et al. 1976). Additionally, Hoogland et al. (2012) compared the morphology of the internal vertebral venou s plexus (IVVP) and its connections to veins draining warm muscles in various ma mmalian species including humans, and concluded that there was potent ial for the IVVP to have a sp inal cord warming function. Given the considerable volume and surface ar ea of the venous complexes at the floor of the braincase and surrounding the cerebellum, such a function may also be possible in Tursiops although the exact function (e.g. warmi ng vs. cooling) is likely debatable. Other less obvious yet potentia lly critical functions of CNS cooling might also be entertained. Three such functions could be hypoxic protection, facilitation of the mammalian dive response, and elicitation of unihemispheric sleep. It is well known that cetaceans spend considerable periods of ti me underwater, during which time they experience protracted apneustic periods. Substantial elevation of anaerobic metabolites have been documented in certai n cetacean species undergoing voluntary experimental dives (Kooyman et al. 1981). Given the regular and often extended periods of apnea, it seems reasonable to a ssume that tissues experience varying degrees of hypoxia. Although it is believed that blood flow to the brain is maintained during a dive, the degree to which this happens in Tursiops is unknown. Depending on the degree of perfusion and available oxygen within the circulating blood, the CNS may experience periods of oxygen de privation much the same way as other tissues do. A mechanism for reducing the temp erature of the CNS could provide protection from such
72 hypoxic incidents. Hypothermia has been shown to have numerous neuroprotective effects such as tissue hypometabolis m and reduction of hypoxic damage and reperfusion injury (Brooks & Duncan, 1940; Caputa et al. 1998; Marsala et al. 1993; Odden et al. 1999) Numerous researchers have shown that coo ling of different parts of the CNS can elicit profound physiological changes in re spiration, heart rate, and peripheral vascular tone (Gregor et al. 1976; Hales, 1983; Hammel et al. 1976; 1977; Jessen & Meyer, 1971; Kullmann et al. 1970; Simon, 1974; Wnnenberg, 19 73;). These physiological changes are all in one way or another characte ristic of the mammalian dive response (Andersen, 1966; Elsner et al. 1964; 1966; Kooyman & P onganis, 1998; Scholander, 1940). Different species appear to respond differently to heating and cooling of the CNS, however in terrestrial mammals the typical response to cooling of the CNS involves increased oxygen consumption, increase peripheral vasoconstriction and tachycardia, while warming of the CNS has the opposite effect. It should be noted that these responses are by no means simple and there appear to be numerous effectors that play an important role in how thes e afferent signals are integrated in the hypothalamus (Kullmann et al. 1970; Luzzati et al. 1999; Simon, 1971; 1974, 1981; Wnnenberg, 1973). Additional ly, we do not know how dolphins respond to these complex stimuli, though it is conceivabl e that they may have evolved different responses to similar stimuli. Res earch on harbor seals conducted by Caputa et al. (1977) suggests that differences do exist r egarding the threshold temperatures and how those thresholds are affect ed by ambient temperature. Therefore, the degree of
73 acclimatization to environmental temperatures also seems important in relation to how temperature signals are integrates. Another interesting consideration involv es the unusual form of sleep observed in dolphins. Numerous authors have described unihemispheric sleep in the bottlenose dolphin (Lyamin et al. 2008; Mukhametov, 1984; 1987; Ridgway, 2002; Ridgway et al. 2006; 2009;). Unihemispheric sleep has been described as a type of functional separation of the two brain hemispheres, allowing one to rest while the other maintains a certain degree of vigilance required for such things as locomotion and auditory surveying. Through functional imaging of a live sedated bottlenose dolphin, Ridgway et al. (2006) showed a reduction in the blood flow to the brain hemisphere believed to be resting. It is not currently known whet her that reduction of blood flow was caused endogenously by the brain or if it reflected some type of upstream limitation like vasoconstriction elicited through the sympat hetic system. It does however raise a question regarding the role of the epidural rete may play.
74 Table 2-1. List of specim ens used for this study. Specimen ID Species Common Name TBL (cm)Gender Date Stranded Research Use CMA1109 T. truncatus Bottlenose dolphin 191F 05/17/11 A, V Hubbs0909 T. truncatus Bottlenose dolphin 249F 02/27/09 A, V ECW005 T. truncatus Bottlenose dolphin 284M 12/16/10 A, V MMC-Tt0708 T. truncatus Bottlenose dolphin 256M 07/25/08V MMC-Tt0107 T. truncatus Bottlenose dolphin 271F 11/12/06 I Abbreviations for genus name are as follows: Tursiops ( T ), Feresa ( F), Physeter ( P ), Kogia ( K ), Mesoplodon ( M ). Abbreviations for research use are as follows: Angiography (A), MRI/CT Imaging (I), Vascular Dissection (V), Sinus Dissection (S). Table 2-2. Structure labels and their names. Structure Label Structure Name 1 Cranial vena cava 2 Left brachiocephalic vein 3 Axillary vein 4 Facial vein 4' Combined facial and internal jugular veins 5 External jugular vein 5' Vena plexu comitans arteria carotidis communis and externa 6 Internal thoracic vein 7 Aortic arch 8 Right brachiocephalic artery 9 Pulmonary trunk 10 Internal jugular vein 10' Anastomotic branches of the internal jugular vein 11 Maxillary vein 12 Mandibular alveolar vein 12' Intramandibular venous plexus 13 Submental vein 14 Vena plexu comitans arteria carotidis interna 15 Ophthalmic venous plexus 16 Melon veins 17 Maxillary labial veins
75 Table 2-2. Continued. Structure Label Structure Name 18 Mandibular labial veins 19 Dorsal nasal vein 20 Caudal auricular vein 20' Temporal and supraoccipital venous plexus 21 Epidural venous rete 22 Temporal dural sinus 23 Cavernous sinus 24 Internal ophthalmic rete 25 Venous plexus of the anterior lobe of the pterygoid sinus 26 Confluens sinuum 27 Palatine veins 28 Cerebellar plexus 29 Ventral internal vertebral veins 29 Midline anastomoses of vent ral internal vertebral veins 30 Intervertebral veins 31 Arcuate veins
76 Figure 2-1. Ventral view of the venous branches (blue) of the cranial vena cava (1) of a bottlenose dolphin (ECW-005). Visible ar e the left brachiocephalic vein (2), axillary veins (3), facial veins (4), external jugular veins (5) and internal thoracic vein (6). Major arteries vi sible are the aortic arch (7), right brachiocephalic trunk (8) and pulmonary trunk (9).
77 Figure 2-2. Simplified schematic representation of Figure 21. Numbers follow Table 22. Note the anastomoses between the in ternal (10) and external (5) jugular veins. The brown structure represents the th yroid gland.
78 Figure 2-3. Right lateral view of a 3D reconstruction of venous angiography of a bottlenose dolphin. Red arrow indicates the enlarged right costocervical trunk descending from the intercostal spaces to join the right brachiocephalic trunk and cranial vena cava.
79 Figure 2-4. Ventral view of a three di mensional reconstruction of CT angiography of the veins of a bottlenose dolphin with the rostrum on shown on the left and cauda l aspects to the right. Numbers fo llow Table 2-2. All visible gray structures represent veins draining in to the cranial vena cava (1). T he faint red structures represent the tympanoperiotic complexes of the ears.
80 A B Figure 2-5. Gross dissection of venous plexus ( venu plexi comitans arteria carotidis communis and externa ) (5) surrounding the common and external carotid arteries. Also visible are the lar ge anastomoses (10) between the internal (10) and external (5) jugular veins and the pericarotid plexus (5). A) Lateral view of the peri-carotid plexuses orient ed with the rostrum to the right. Note intimate association of external jugular vein and venous plexus. B) Medial view of the peri-carotid plexuses or iented with the rostrum to the left.
81 Figure 2-6. Right lateral view of internal and external carotid arteries. Note the venous plexus (14) surrounding the internal ca rotid artery located medial to the jugular branches and external carotid ar tery, and just ventral to vagus nerve. Figure 2-7. Mid-sagittal section of dolph in head and neck with the head on the left and cranial thorax on the right. Visible ar e the endotracheal plexus (red arrow) and plexus surrounding esophagus (yellow arrow).
82 A) B) Figure 2-8. Transverse slice of computed tomographic angiography near apex of melon of a bottlenose dolphin, showing palatine (yellow arrows) and maxillary labial (red arrows) veins. A) Vein s near the apex of the melon. B) Veins near distal extremity of rostrum.
83 Figure 2-9. Left lateral view of maximum intensity projection from CT angiography of the head of a bottlenose dolphin. Note the extensive venous investment of the melon (16). Also visible are th e facial (4), internal jugular (10), mandibular alveolar (12), submental (13), and maxillary (17) and mandibular (18) labial veins. Note the location of the intramandibular venous plexus (12) and ophthalmic plexus (15).
84 Figure 2-10. Dorsal view of three-dimensio nal reconstruction of the veins on the dorsal surface of the skull of a bottlenose dolphin Note the large lateral nasal veins (19). Figure 2-11. Right lateral view of the head and neck of a bottlenose dolphin showing the caudal auricular vein (20) coursing behind the external auditory meatus and supplying the venous plexus (20) on the te mporal and supraoccipital regions. The emergence of the facial (4) vein onto the superficial aspect can be seen just above the sternomastoid (Stm ) and sternohyoid (Sth) muscles.
85 Figure 2-12. Left lateral view of three-dimensional reconstruction of CT venous angiographic image of a bottlenose dol phin head, showing the ophthalmic plexus (15) emerging from the facial (4) vein and extending into the orbit.
86 Figure 2-13. Close-up of a late ral view of the orbit showi ng the intricate nature of the ophthalmic plexus (15), draining laterally into the facial ve in (4). Note that the eye stalk is reflected dorsad by the ti p of the forceps and the image has been flipped to maintain left lateral perspective (image is actually of the right orbit).
87 Figure 2-14. Dorsal view of maximum intensity projection of CT venous angiography of a bottlenose dolphin showing faint bone shadows and yellow veins. The epidural venous rete (21) can be seen entering the skull and connecting to the transverse dural sinuses (22) which in part drain the cavernous sinuses (23). Note that the ve ins of the basicranium and ventral head and neck have been cropped to simplify image.
88 Figure 2-15. Dorsal view of maximum intensity projection of CT venous angiography of a bottlenose dolphin showing faint bone shadows and yellow veins. The internal ophthalmic retia (24) can be seen extending caudomediad from the ophthalmic plexuses (15) and connecti ng to the cavernous sinuses (24).
89 Figure 2-16. Excised epidural rete injected with red latex in the arterial component, and blue latex in the venous component. A) Left lateral view of the excised cervical and cranial thoracic epidural rete (21) and intact dura of the left brain hemisphere. B) Ventral view of the ce rvical and cranial thoracic epidural rete (21) showing intimate juxtaposition of arteries and veins. Also visible are the distal cervical portions of the ventral internal vertebral veins (29) near the foramen magnum.
90 Figure 2-17. Excised epidural rete from a bottlenose dolphin showing ventral internal view and dorsal view. A) Ventral view of cervical epidural venous rete (21) with the thoracic arteri al component of the rete reflected laterally to show the large longitudinal ventral internal vert ebral veins (29). Together with their midline anastomoses (29), the longit udinal veins form the ventral internal vertebral plexus. B) Dorsal view of the cervical and cranial thoracic portion of the epidural rete (21) reflected laterally to show the spinal cord. Note that minimal midline anastomoses exist, su ch as the arcuate veins (31).
91 CHAPTER 3 DETAILS ON THE VASCULARIZATION OF THE AIR SINUSES A ND FAT BODIES IN THE HEAD OF THE BOTTLENOSE DOLPHIN ( TURSIOPS TRUNCAT US) Chapter Foreword Since the 1980s, numerous beaked whale mass strandings have been temporally and/or spatially associated wit h deployment of naval mid-frequency active sonar (Evans & England, 2001; Fernandez et al. 2004; 2005; Fratzis, 1998; Simmonds & LopezJurado, 1991). Research into potential caus al mechanisms underlying these events is logistically difficult, in part because of cons traints such as insufficient funding and the cryptic nature of beaked whales and because of legal and public constraints surrounding live-animal physiological experimentation (reviewed in Cox et al 2006). Most of the progress that has been made on this topic has been made either through tagging of live beaked whales (Tyack et al 2006) or from post-mortem morphological and pathological studies (Fernandez et al 2004; 2005; Hooker et al 2009). Hypotheses regarding the etio logy of the strandings r ange from physical acoustic trauma to gas and fat embolic trauma caused by behavioral alterations in dive profiles (Cox et al 2006; Hooker et al 2011: Rommel et al 2006). The tagging studies have generated invaluable insights into the diving behavior of certain beaked whale species, and these data have provided a platform for m odeling of gas kinetics such as nitrogen uptake and elimination (Fahlman et al 2006; 2009; Hooker et al 2009; Houser et al 2001; Tyack et al 2006; Zimmer & Tyack, 2007). Reprinted with permission from: Costidis, A.M. and Rommel, S.A. 2012. Vascularization of the air sinuses and fat bodies in the head of the bottlenose dolphin ( Tursiops truncatus): morphological implications on physiology. Front Physiol 3:243
92 Post-mortem examinations of beaked whales stranded in the Canary Islands have identified decompression sickness (DCS)-t ype sequelae, suggesting that gas bubble formation may be at the root of some of the observed strandings (Fernandez et al 2004; 2005). Additional findi ngs of apparent acute and chr onic embolization in other odontocete species suggest that gas bubble disease in cetaceans may be more common than initially thought (Jepson et al 2003; 2005). Much debate has focused on individual physiological, behavio ral, or anatomical traits that may predispose beaked whales to nitrogen saturation and subsequent DCS-like lesions (Cox et al 2006; Fahlman et al 2006; 2009; Hooker et al 2009; Rommel et al 2006). I suggest that a combination of physiological, anatomical, and behavioral characteristics contribute to this predisposition and that research shoul d focus on integrating all the aforementioned characteristics. Currently, our only insights into the functi onal implications of the anatomy are to be inferred from pathological findings. To begin this discussion, we chose to focus on the lesions that were observed in the beaked whal es that stranded in the Canary Islands in September of 2002 (Fernandez et al 2005). The most conspi cuous lesions discovered on post-mortem examination were intravascu lar emboli widely disseminated throughout the kidneys, lungs, liver, and central nervous system. Interestingly, these emboli were found to sometimes be composed of gas, other times of fat; however both were often present simultaneously within a single tissue. These findings suggest that either different body areas or types of tissue were compromised separately or a single region containing intimate gas and fat associations was damaged. In either situation, in addition to the involvement of gas-filled sp aces and fatty tissues, the other requirement
93 for such dissemination of the emboli is vascular introduction and transportation. Therefore, unless there wa s intravascular introduction of gas and fat throughout the body or generalized autochthonous embolus fo rmation, we suggest t hat the most likely location for vascular introduction of gas and fat emboli would be anatomical regions where gas, fat, and blood vessels are intimately associated (Figure 3-1; Movie 3-1). We therefore agree with Jepson et al (2005) who alluded to t he fatty tissues of the odontocete head as a reasonable source of fa t emboli due to the extensive vascular structures associated with bot h cranial air-filled sinuses and acoustic fat bodies and the frequent observation of hemorrhages within those tissues in DCS-like cases of strandings (Boenninghaus, 1904; Fras er and Purves, 1960; Fernandez et al 2004; 2005; McFarland et al 1979; Ridgway et al 1974; Slijper, 1936). There is a relative paucity of published information concerning beaked whale anatomy, especially the anatomy relating to diving physiology or issues of potential susceptibility to mid-frequency acti ve sonar (reviewed in Rommel et al 2006 ). Gas kinetic modeling has shown that knowledge of the vascular anatomy is integral to understanding the dynamics of nitrogen gas uptake and elimination (Fahlman et al 2006; 2009; Hooker et al 2009). Some of the most relevant beaked whale anatomy relating to diving physiology is therefore, likely to be the morphology and function of blood vessels, about which nothing has been pub lished. Although the ar terial system of cetaceans has received considerable attention, the venous system has remained largely undescribed (Fraser & Purves, 1960; McFarland et al 1979; Murie, 1873; Slijper, 1936; Ommanney, 1932; Viamonte et al 1968; Vogl & Fisher, 1973; Walmsley, 1938; Wilson, 1879). Given the lack of information on beaked whal e vascular anatomy, this
94 work will explore the vasculature associ ated with air spaces and fat bodies of the bottlenose dolphin head as it might relate to the formation of gas and fat emboli and the absorption and elimination of nitrogen gas. We hope that the information contained herein acts like a springboard from which our knowledge of the vascular anatomy of deep diving cetaceans can evolve. Before discussing the vascular anatomy of the region, a presentation of anatomical details of the accessory sinus system seems prudent. Cetaceans have lost the air-filled paranasal sinus system found in terrestrial mammals (Mead & Fordyce, 2009). However, cetaceans do have gas-f illed sinuses (the gas is derived from respiratory air but may vary in composition) that are simila r to those of the paranasal system. The cetacean accessory sinus system (Figures 3-2 and 3-3) is unique (Fraser & Purves, 1960; Mead & Fordyce, 2009); these un-pigmented mucosa-lined structures, which are located on the ventral aspect of t he skull, are typically associated with hearing and acoustic isolation of the ears (Houser et al 2004). The ventral sinus system is distinguished from the dorsal air sacs by appearance and function; the lining of the dorsal sacs is composed of pigmented epithelium (Reidenberg & Laitman, 2008) and these sacs are associated with sound production. Both sets of air-filled structures are confluent with the respiratory system. In their stunning monograph, Fraser and Purves (1960) outlined much of the anatomy of the pterygoid and peribullar sinuses and their surrounding auditory structures by piecing toget her information from corrosion casts and dissections of various delphinid species and compiling pub lished literature. Their anatomical descriptions included the bones, sinuses, an d vascular structures of the region and
95 have acted as the foundation of our review and much of our research on this topic as a whole. Fraser and Purves ( 1960) noted the extensive vascula r investment of the lining of the accessory sinus system present in del phinid species. The vascular nature of the pterygoid and peribullar tissues surr ounding and investing these sinuses has been highlighted in other cetaceans including mysticetes such as fin whales (Boenninghaus, 1904; Murie, 1873; Ommanney 1932; Walmsley, 1938). The accessory sinus system is therefore composed of a set of interconnected and physiologically dynamic structures (Figure 3-3) that have been described as both gasfilled and blood-filled (Fraser & Purves, 1960) This apparent contradiction can be explained because the spaces contain a series of gas-filled sinuses, the walls of which possess extensive vascular structures (Figure 3-4). Within this system the relative volumes of gas and blood can apparently be dynamically altered by hydraulic pressures of the respiratory system and the environment and by the amount of blood in the surrounding vascular structures (Fraser & Purves, 1960). Changes in volume of the gas-filled sinuses can be compensated for by corresponding changes in the adjacent soft tissues and with an incursion of blood if the sinus geometry is constrained by bone (Fraser & Purves, 1960; as in the hamular lobe of the Pty sinussee below). The sinuses occasionally contain a considerab le amount of viscous, stable foam whose source and composition remain elusive. The gas-filled part of the accessory sinus syst em is an extension of the middle ear cavity (Mead & Fordyce, 2009) that is connec ted to the upper respiratory system via the Eustachian tube (ET, auditory tube; Figure 32). The blood-filled portion (individual elements are often referred to as fibro venous plexuses) is an extensive network of
96 veins and venous blood sinuses (Fraser & Purv es, 1960) that we describe in detail for the first time. Fraser and Purves (1960) descri be a long history of anatomical interest in the accessory sinus system, starting with descrip tions of the peribullary sinus by Major in 1672 and Tyson in 1680. This system ha s been referred to by a variety of names such as: the pterygoid sinus system (Fra ser & Purves, 1960; Mead & Fordyce, 2009); pneumatic cavities (Boenninghaus, 1904); Eustachian system (Andersen, 1879); Eustachian sacs (Fraser & Purves, 1960); post palatine sinus (Flower, 1867); sinus cavities (Houser et al 2004); air-filled sinuses (Cranford et al ., 2008); air sac system (Reidenberg and Laitman, 2008). Recently, t he accessory sinus system and the nasal cavity were imaged in two live bottlenose dolphins ( Tursiops truncatus) by Houser et al (2004). The nomenclature that we use for the accessory sinus system originated with Beauregard (1894); it was refined by Fraser and Purves (1960) and clarified by Mead and Fordyce (2009). Although they are an invaluable resource, Fraser and Purves (1960) create considerable confusion with thei r interchangeable use of cavity, lobe, sac, and sinus (Mead & Fordyce, 2009). Materials and Methods Specimens for this study were obt ained from deceased, stranded cetaceans recovered along the east coast of the United St ates and west coast of Florida (Table 1). Beach-cast carcasses were recovered by marine mammal stranding networks authorized by stranding agreement s from the National Marine Fisheries Service U.S. Marine Mammal Health and Stranding Program. All work was conducted under a parts authorization from National Marine Fisheries Service, pursuant to 50 CFR 216.22 and 216.37, and with prior approval from Univer sity of Florida s IACUC (Permit #:
97 200801345) and University of North Carolina Wilmingtons IACUC (Permit #: A0809019). Specimens used for gross dissection of the accessory sinus system were either dissected fresh, or frozen, thawed and disse cted when logistics allowed. Schematic illustrations of the accessory sinus system and the simplified venous connections were made using EasyCAD (Evolution Comput ing, Phoenix Arizona 85020). Specimens used for describing the vascular anatomy were in jected with contrast medium into either the venous system or both the venous and arterial systems, following the procedural methodology outlined by Holliday et al (2006). Prior to injecti on of contrast media, all specimens obtained a vascular flush using 0. 9% phosphate buffered saline solution and some received a subsequent 5% neutral buffered formalin vascular perfusion to enable a protracted dissecting period. Most specimens were tar geted for examination of the venous system, and were therefore flushed thr ough the arteries and out of the veins. Once the effluent ran clear, the flush wa s stopped and the specimens were refrigerated while being allowed to drain for several hours. Prior to injecting the vascular contrast material, 5mL balloon catheters were placed in the vessels to be injected and inflated unt il a good seal was formed. In specimens that were to be imaged via computed tomogr aphy (CT), the vascular system of interest received a mixture of liquid latex and barium sulfate sus pension (Liquid Polibar Plus, Bracco Diagnostics Inc.), while vessels only destined for dissection received pure latex (Carolina Biological Inc.). All specim ens were refrigerated for 2 days following injections, to allow the latex cast to cure, and if unpreserved with formalin, where subsequently frozen at oC. Specimens that were imaged via CT were scanned at
98 the thinnest slice thickness possible bas ed on the specimen length, and whenever possible the volumes were reconstructed to 0.5mm thickness to allow high resolution imaging of fine caliber vasculature. The resultant DICOM data was post-processed using Amira software (Visage Imaging Inc., San Diego, CA) on a Gateway desktop with memory and processor upgrades. Foll owing imaging and po st-processing, all specimens were thawed and dissected to validate and/or clarify imaged structures. Although the focus of this st udy was to elucidate the venous morphology in regions of interest in the head of Tursiops the authors felt that a co mparative examination of deep-diving odontocete cetaceans would be valuable given the association of deep divers and sonar-related strandings. The authors therefore opport unistically obtained specimens from pygmy and dwarf sperm whales ( Kogia breviceps and sima ), sperm whales ( Physeter macrocephalus ), and Gervais beaked whales ( Mesoplodon europaeus ). Some of the sperm whale and pygmy sperm whale specimens were of sufficient quality and from young enough animals that could fit into the CT gantry. Prior to vascular dissection, those specimens we re imaged according to the aforementioned angiographic protocol and the data obtained was used to guide the dissections. As the focus of this study was on Tursiops only cursory mention is made with respect to findings from the other species.
99 Results1 Air Sinus and Fat Body Morphology Sinuses What is known about the accessory sinus system of th e cetacean head varies considerably by species; however, delphi nid and phocoenid species remain the best described. It is however clear that in a ll odontocete species there are extensive gasfilled sinuses on the ventral side of the skull. Interestingly, our preliminary research shows that the accessory air sinuses of deep diving odontocetes such as beaked whales, sperm whales, and pygmy and dwarf sperm whales are much larger, relatively, than those of even large delphinid specie s like pilot whales, and are invested with intricate and seemingly more voluminous venous arrangements. Much of the internal surface of these sinuses is lined with copious masses of convoluted, intercommunicating, and valve-less veins separat ed from the air spaces by walls so thin that they are translucent (Figures 3-4 and 3-5). Air sinus extent The access ory sinus system is mucosalined and juxtaposed (on at least one side) to bones of the basicranium. The sinus system exists as a bilaterally paired system of blind structures. These sinuses (Figure 3-3) may extend rostrally to cover the ventral aspect of the palate, parts of the orbit, and pharynx an d caudally to surround the tympanic bulla of the ear (T ym, Figure 3-2), the temporomandibular joint, and the joint between the skull and the hyoid apparatus. The accessory sinus system is dominated by the pterygoid sinus (Pty sinus). This sinus extends from just ro stral to the orbit back 1 In-text references to vascular structures that are labeled in figures will be followed by a number representing the label assigned in the figures. Impor tant tributaries of structures maintain the parent vessels number followed by apostr ophes. Non-vascular structures are abbreviated with three letters.
100 to the region of the tympanic bu lla (Tym). The Pty sinus has two more or less distinct lateral lobes2 that may extend dorsad around the ey estalk (preorbital and postorbital lobes). In Tursiops these lobes are expanded and may meet dorsal to the eye as a supraorbital lobe. The presence of an air-f illed sinus on the ventral aspect of the supraorbital process of the frontal bone dorsal to the eye necessitates a fenestration (optic infundibulum; Fraser & Purves, 1960) through the sinus to accommodate passage of the optic nerve, because the sinus extends over the regi on where the nerve exits the braincase. The Pty sinus in Tursiops has a hamular lobewithin t he hollowed hamulus of the pterygoid boneand an anterior l obe. These two lobes are formed by an indentation of the pterygoid sinus, caused by the lateral laminae of the pterygoid bone and palatine bones. This bony structure is lacking in the pygmy sperm whale ( Kogia breviceps ) and thus, distinction of those lobes is meani ngless. At its caudal end, the Pty sinus connects to three smaller sinuses, all of which are relatively close to the region occupied by the ear, the temporomandibular joint (TMJ), and the attachment of the hyoid apparatus via the tympanohyal cartilage: the most lateral of these three sinuses (the middle sinus) is associated with the TM J; a slightly more caudomedial sinus (posterior sinus) is associated with t he tympanohyal joint and is bordered by the paroccipital crest; and a caudomedial sinus ( peribullar sinus) that helps separate the tympanic bulla from the adjacent bones. 2 Lobes are partitions or outgrowths of a sinus.
101 Bony walls Despite the variable geometry of the re cesses and fossae of the si nus system, the bony associations of the sinuses follow some similar patterns among all odontocetes studied thus far. Delphinid species show marked similarity among the different species, while kogiids, physeteriids, and ziphiids all show similarity between themselves. In delphinids, the lateral wall of the pterygoid sinus is partly encased by the bony lateral laminae of the pterygoid and palatine bones. Conversely, in deep diving species like ziphiids, kogiids, and physeteriids, the latera l wall of the Pty sinus is composed entirely of soft tissue capable of collapsing onto itse lf. When manually manipulated, the Pty sinus is easily closed under the weight of the surrounding tissues, suggesting that maintaining it in an expanded state may requ ire a degree of pressurization. Although the significance of these featur es is unknown, it is difficult to ignore the common threads among the non-delphinid deep diving cetaceans. Shallow divers such as Tursiops have a deep indentation in the Pty sinus. In contrast, deep divers such as Kogia lack or have less distinct anterior and hamular lobes. Bony recesses The sinuses of the accessory sinus system fo llow the contours of slight depressions (fossae; Mead & Fordyce, 2009) in the skull bones on the ventral basicranium (Figure 3-2); t hese depressions typically have a smoother surface than the other regions of these skull bone s. The relatively large Pty sinus extends from the palatine and maxillary bones rostral to the orbit and caudally to the region of the tympanic bulla (Tym) of the middle ear. The Tym sits in a deep recess bordered medially by the basioccipital bone. The pharyngeal crest (Mead and Fordyce, 2009), a ridge of bone that delineates the bony lateral margins of the pharynx, is formed by the
102 ventrally projecting crests of the Pty and Boc bones. Note how the lobes of the pterygoid sinus in Tursiops extend much fart her into the orbit than those of Kogia; these lobes join distally to form the supraorbital lobe of the Pty sinus, on the ventral aspect of the supraorbital crest of the frontal bone. Eustachian tube The Eustac hian tube (ET) connects one hal f of the accessory sinus system and the ipsilateral middle ear to the respiratory sy stem (Figure 3-2). The ET is a distinct, soft tissue, mucosa-lined, hollo w tube located mostly within the lumen of the Pty sinus, attached to its ventral mucosa. The dorsal o pening of the ET is within the essentially vertical nasal cavity just above the palatopharyngeal muscle, which acts as a sphincter to isolate the nasal cavity from the rest of the upper respiratory system during a dive. The ET extends ventrally within the nasal cavity to the Eustachian notch (Mead & Fordyce, 2009; pterygoid notch Fraser & Purves, 1960; tubal notch Schulte, 1917), which is a distinct cleft in the pterygoid bone on the ventral aspect of the skull (Figure 32). Near the apex of the Eu stacean notch, the ET enters the accessory air sac system by passing through the wall of the Pty sac and a layer of connective tissue (which spans the Eustacean notch) to enter t he internal bony nares of the respiratory system. The ET enters (but does not open to) the accessory sac system at the lateral aspect of the Eustachian notch. Within the Pty sinus t he Eustachian tube extends caudally, roughly parallel to the pharyngeal crest, to open just rostral to the opening of the middle ear cavity. In Tursiops we have observed that the Eustac hian tube is surrounded by part of the adjacent pterygoid and perib ullar venous plexus, however the luminal surface of the ET is also trabecular but has never injected with casting latex, suggesting it is a non-
103 vascular structure. The caudal end of the ET is open to the lumen of the Pty sac rostral to the bony opening of the Tym proximally in Tursiops distally in Kogia Blood-filled part (of th e accessor y sinus system) All of the afore-mentioned l obes of the accessory sinus system are, in some form, associated with a venous plexus. The Pty sinus proper is associated with the most extensive venous plexus; however, all of t he lobes are surrounded by or associated with a sizeable venous plexuses. Interestingly, the portions of the accessory sinus system along the basicranium are associated with venous plexuses directly connected via sizable emissary veins to the intr acranial veins draining the brain. Fat bodies I have observed extensive vasc ularization in three of the largest acoustic fat bodies of dolphins, namely the melon, in tra-mandibular fat body (IMFB), and extramandibular fat body (EMFB) (Figures 3-1 and 3-6 to 3-8) in agreement with the observations of Fraser and Purves (1960), Maxia et al (2007), and Slijper (1936) (for more complete descriptions of the fat bodies see Cranford et al 1996; 2008; Harper et al 2008; Koopman et al 2003; McKenna et al 2011; Norris, 1969; Norris & Harvey, 1974). The melon fat is located on the dorsoro stral aspect of the skull, extending from the rostral border of the nasal passage to t he apex of the melon. Ventrally, the melon fat is separated from the ventral structur es of the head by the vomer, maxillary, premaxillary, pterygoid, and palatine bones and the mesorostral cartilage. Each IMFB is located along the medial margin of the ipsilateral dentary bone. The dentary is hollow medially, along the caudal two thirds of its length, and this cavity is filled with the IMFB which extends caudally to attach to the tympanoperiotic complex by way of dorsal and ventral branches of the fat body (Cranford et al 2008; Norris & Harvey, 1974; Ridgway,
104 1999). Medially, the lateral pterygoid muscles flank the IMFB, while its lateral border is defined by the dentary (Fraser & Purves, 1960). Each EMFB borders the lateral surface of each dentary. Along their ventral aspects, which extend ventral to the dentary, the EMFB and IMFB merge without any grossly visibl e distinction. On its lateral aspect the EMFB integrates seamlessly with the blubb er surrounding the lower jaw. Although the individual veins draining each of these fat bodies are different, all three drainage fields eventually converge into the external jugular (#2, Figures 3-7 to 310) and facial veins (#3, Figures 3-7 to 3-9) and via anastomoses ( #1, Figures 3-9 to 3-10) into the internal jugular veins. The melon fat is primarily drained by a multitude of veins that converge with the veins draining the rest of the tissues of the region (#9, Figu re 3-7 and 3-13 to 3-14) such as the nasal apparatus, maxillae, and maxillary (dorsal) lips. These veins converge, somewhat, as they travel through the dorsa l infraorbital foramina of the maxillary and premaxillary bones, emerging on the ventral aspect of the skull as part of the plexus of the accessory sinus system and determined herei n to be the dorsal continuation of the maxillary vein component of the pterygoi d venous plexus (#6, Figures 3-14 and 315E). From there, the plexus receives input from the nasa l (#6), palatine veins (#6), cavernous sinus (#16), ventral petrosal sinus (#27), intramandibular plexus (#10), pterygoid venous plexus (#11), before coalesci ng into a single maxillary vein (#6) that follows the pterygoid crest to become the exter nal jugular vein (Figures 3-7 and 3-12 to 3-13). Additional but less volu minous drainage of the melon also likely occurs through various terminal branches of the facial vein that wraps around the antorbital notch (#3, Figures 3-7 and 3-9).
105 The IMFB is invested with an intricate, voluminous plexus of anastomosing small caliber veins (#10, Figure 3-8) On cross section, the plexus can be seen throughout virtually the entire substance of the IMFB (# 10, Figure 3-11), except for a small roughly circular region near its ventral margin (r ed asterisks, Figure 311). Histological quantification of vessel dens ity has not yet been perfo rmed; however, numerous specimens with latex-injected vasculature ha ve consistently shown this well-defined region of reduced vascularizatio n. Throughout its length, this un-vascularized region of fat forms a circular t ube extending along much of the length of the IM FB. Interestingly, its position changes caudally as the tube approaches the ear, where it bends dorsad and attaches directly on the tympanoperiotic complex. Although the venous plexus appears undifferentiated in the wa y it invests the IMFB, the dorsal aspect of the plexus appears to have an increased density of longi tudinally oriented veins that extend rostrally. A portion of the increased dorsal venous density contributes to the formation of a peri-arterial venous rete (PAVR), a rosette of veins that surrounds and follows the mandibular alveolar artery rostrad and has therefore been te rmed the mandibular alveolar plexus (#10, Figures 3-11and 3-13) The venous investment of the EMFB (#20, Figure 3-9) appears more diffuse and the drainage field less singular than that of the IMFB. Unlike the IMFB, the venous in vestment of the EMFB does not appear as dense, however despite its more diffuse nature, the veins of the EMFB still maintain a plexiform arrangement of notable extent. Vascular Anatomy As is common to mo st domestic mammals (e .g. cow, horse, pig, cat), the external jugular vein of the bottlenose dolphin is the main draina ge route of the pterygoid vasculature although the internal jugular vein can be the pr imary drainage (e.g. dog)
106 (Evans, 1993; Ghoshal et al. 1981; Schummer et al. 1981; Schaller, 2007). What is different in the dolphin, however, is that the linguofacia l vein branches off of the brachiocephalic trunk instead of the extern al jugular vein, and in some cases as a common branch with the internal jugular vein. Additionally, t he facial vein gives off the mandibular vein that in domestic mammals or iginates from the ma xillary veinthe main continuation of the external j ugular veinor one if its tributaries. As we are just now beginning to elucidate the vascular anatom y of deep diving cetacean species, the following description of venous branching patte rns is based solely on that observed in Tursiops Additional clarification can be obtained in the supplementary movie (Movie 31) provided. Table 2 shows the structures labeled in figures 3-7 to 3-15 and their corresponding names. The plexuses asso ciated with the mandibular fat bodies and accessory sinus system are drained primarily by three parent veins, namely the facial, external jugular, and internal jugular veins. Facial vein At about the level of the paroccipit al crest and just lateral to the skull attachment of the tympanohyal element, the fa cial vein trifurcates into large lateral and ventromedial veins and a midline coalescing pl exiform mass of small veins3 (#3, 5 & 6, Figures 3-7 to 3-9 and 3-13). The lateral branch of the facial vein (#3, Figures 3-7 to 3-9 and 3-13) extends dorsolaterally around the dentary, pa ssing at an oblique angle under the mandibular condyle and crossing the condylo id crest on its path to the external ophthalmic plexus (#7) and the antorbital notch of the maxillary bone. In addition to its medial connection through the orbi t to the external ophthalmic plexus, the vein continues around the antorbital notch to the dorsal aspect of maxillary and 3 Traditionally venous branching descriptions follow a distal to proximal path representing the retrograde flow of venous blood; however for conceptual simplic ity we have chosen to describe branching patterns starting at proximal main branches and progressing out to more distal branches.
107 premaxillary bones to drain the tissues of the melon, nasal plugs and dorsal lips (#9, Figs. 7 & 13). This vein is consis tent in location and drainage field to the facial vein of terrestrial mammals. The middle branch of the facial vein (#5, Figures 3-7 to 3-9 and 3-13) can arise as either a single branch or a brush-like sp ray of small veins t hat extend from the rostral margins of the of the facial vein trifurcation. Whether single or plexiform, the structure eventually becomes a spray of numerous small veins that enter the dense connective tissues associated with t he ear and lower jaw attachment to the skull. Upon entering the connective tiss ue, the veins contributealong with the pterygoid plexus described below--to t he formation of the fibrovenous plexus described by Fraser and Purves. This fibrovenous plexus is composed of an intricate network of small (~1-5mm diameter) caliber anastomosing veins that invest much of the connective tissue and wrap around part of the tympanic bulla of the ear bone complex. The fibrovenous plexus4 extends rostrally, gradually losing its connective tissue component as it transit ions into the soft fa tty tissues of the intramandibular fat body (IMFB). As the ventrolateral portion of the fibrovenous plexus extends into the IM FB to become the mandibular plexus, the veins appear to distribute throughout most of the fatty substance (#10, Figures 3-8, 3-11, and 313 to 3-14). Along its dorsal margins the IMFB plexus coalesces to form a PAVR (#10, Figures 3-11 and 3-13) that surr ounds the mandibular artery and courses rostrad toward the mental foramina of the distal dentary and mandibular symphysis, occasionally giving off branches to the teeth. This PAVR is consistent with the mandibular (inferior) alveolar vein described in terrestri al mammals and is therefore responsible for drai ning blood from the teeth and rostral portions of the lips of the lower jaw. The third and most medial branch of the fa cial vein (#6, Figures 3-9 to 3-10 and 314 to 3-15) is similar in si ze to the lateral branch. Approximately 5cm after the ventromedial branch emerges fr om the trifurcation it is jo ined by the maxillary vein, the main terminus of the large external j ugular vein (#2, Figures 3-9 to 3-10 and 314 to 3-15). This medial branch of the faci al vein might, therefore, be considered an anastomotic branch to the maxillary vein. The two veins join to form a larger vein that travels roughly horizontally to t he level of the ceratohyal cartilages of the hyoid apparatus. At that level, the vein gives off smaller lingual and pharyngeal branches before curving dorsad along the pterygoid crest to supply the pterygoid veins. Internal jugular vein (#1, Figure s 3-7, 3-9 to 3-10, and 3-13 to 3-14) The internal jugular vein can arise either singularly or as a common trunk with the linguofacial vein, directly from the brachiocephalic vein. On ce the internal jugular vein 4 Fibrovenous plexus a complex system of venous channels and connective tissue forming a plexus that extends from the connective tissue of the ear to the pt erygoid region; it is mostly associated with the lining of the sinuses.
108 has traversed the short ne ck and reached the head, it cont ributes to the drainage of three regions. The first and most proximal (caudal) contribution is formed by numerous anastomoses (#1, Figures 3-9 to 3-10 and 3-14 to 3-15) between the internal and external jugular veins. These anastomose s arise from the ventral aspect of the internal jugular vein and travel obliquely ventrorostrad to fuse with the external jugular plexus surrounding the external caroti d artery (#19, Figures 3-9 to 3-10), in agreement with the findings of Ridgway et al. (1974). Like the mandibular alveolar vein, the plexus surrounding the external jugular vein forms a periarterial venous rete From this point rostrad, the external jugular vein (#2) becomes the maxillary vein (#6) and drains the majority of the pterygoid plexus as well as parts of the nasopharyngeal, palatine, and dorsa l nasal regions. Due to the anastomoses, the internal jugular vein might therefore be considered to facilitate this drainage. The second and third branches arise in co mmon from a bifurcation of the distal internal jugular vein. The proximal of the two branches curves sharply rostrad to become the ventral petrosal sinus (#27, Fi gure 3-15D) that lines the dorsal aspect of the peribullar plexus ( #12, Figures 3-14 to 3-15). The third branch extends vertically from t he bifurcation of the internal jugular vein, to enter the jugular foramen as the termi nus of the internal jugular vein (#25, Figure 3-15A-D). As it enters the caudal portion of the ca lvarium, it fuses with the temporal/sigmoid/transverse dural sinus, to form one of the main drainage paths for blood from the brain case. External jugular vein (#2, Fi gure s 3-9, 3-10, 3-14, 3-15) The external jugular vein arises directly from the brachiocephalic trunk as a plexiform structure composed of small and large veins that su rround the external carotid artery. The ventral portion of this plexus is highlighted by a distinct vein of considerably larger caliberthe external j ugular veinthan the rest of the plexus veins. As the external jugular vein traverses the ne ck and approaches the head, it receives the anastomotic branches (#1, Figures 3-14 to 3-15) of the internal jugul ar (#1) and facial veins (#3). It finally receives an anasto motic branch from the facial vein and then curves medially to follow the basioccipital an d pterygoid crests as the maxillary vein (#6). Along its course thr ough the pterygoid region, the ex ternal jugular vein forms
109 numerous branches and countless small anasto moses with other venous tributaries. The main branches invest the palatine, nasopharyngeal, pterygoid, and dorsal head tissues as follows: On its course to the rostroventral aspec t of the brain case, the maxillary vein sends dorsorostrally oriented veins of a plexiform nature (#6, Figures 3-12B-D) that invest the tissues on the roof of the oral cavity. These veins were considered to be the palatine veins which are shown to form a palatine plexus. Before it breaks up into the numerous ve ins that compose the pterygoid plexus (#11, Figures 3-12C-E and 3-14 to 3-15) t hat invests the lining of the pterygoid sinus and drain the substance of the pterygoid muscles, the maxillary vein sends rostrodorsad oriented branches (#6, Figure 3-12B-F) into the palatopharyngeal muscles where they anastomose with other palatopharyngeal veins supplied by the external jugular vein and a large pharyn geal plexus that surrounds the rostral esophagus and laryngeal cartilages (red asteri sks, Figure 3-12A). At this level the maxillary vein loses its singular identit y as it breaks up to become part of the pterygoid plexus associated with the acce ssory sinus system (#11, Figures 3-12CF and 3-14 to 3-15). Nonetheless, the latera l region of the plexus extends laterally under the external pterygoid musclesuch that is forms a cradle for the muscle and fuses with dorsal extensions of the intramandibular plexus. The combined plexuses extend dorsad toward the maxi llary bone and narrow considerably as they approach the ventral aspect of the maxillary and premaxillary bones. The plexus surrounds all but the medial aspec t of the maxillary artery as both structures travel to the infraorbital fora mina (#6, Figures 3-14 and 3-15E). This dorsal extension of the venous plexus maintains its course to the ventral infraorbital foramina, in juxtaposition wit h the corresponding maxi llary artery, and is therefore considered herein as the term inus of the maxillary vein. Just before the plexus reaches the infraorbital forami na, its rostral aspect sends a large mass of highly convoluted veins rostrad. T hese veins form a robust venous plexus (preorbital and anterior lobe plexus) that travels along the v entral lining of the anterior lobe of the accessory sinus syst em (#8, Figures 3-7 and 3-13 to 3-15). Discussion Significance of the Anatomy Despite the significant reorganizat ion of the skull bones to accommodate dorsally located nares and a large melon, many of the major veins in the heads of odontocetes find analogs in domestic mammals. Similarly, although we have no developmental data on these venous structures, the st riking similarity in distribution and location of many of
110 those structures suggests that they are al so homologs of the veins seen in domestic mammals. Nonetheless, a few significant elaborations on st ructures were observed, especially in regions associated with acousti c fat bodies and air sinuses. Specifically, given the novel evolutionary nature of the cetacean mandibular fat bodies, the venous plexuses of the intraand extramandibula r fat bodies find no notable homologue in domestic mammals. Although it is unknown whet her or not the venous plexuses of the accessory sinus system might find homologues in the pterygoid plexuses of various domestic mammals such as horses, cows, a nd dogs, the complexity and volume of the plexus system of cetaceans appears unparalleled. Accessory sinus system Since the accessory sinus system is not completely enc ased in robust, rigid bony compartments, it is presumably exposed to the effects of changing barometric pressures encountered during a dive. In order to avoid pathologies associated with physical injury of tissues exposed to dysbar ic changes it seems safe to assume that there exists a significant amount of physiological flexibility within those structures. The apparent susceptibility of the accessory sinus system to external pressures may have significant implications on diving gas kineti cs. The intimate anatomical association between large venous masses with thin linings and pockets of air exposed to changing pressures begs a question regarding absorption and clearance of gases during a dive (Reidenberg & Laitman, 2008). Wi th increasing hydrostatic pressures, gases in air are progressively driven into solution. This phenomenon is responsible for increased absorption of nitrogen in the lungs of a sc uba diver. Marine mammals had long been thought to be exempt from si gnificant accumulation of nitrogen during dives because they do not breathe compressed air at dept h, and because their pulmonary alveoli are
111 believed to progressively collapse with depth, thereby isolating pulmonary gases away from absorptive respiratory surfaces. Since the pressurized pulmonary air is thought to be segregated from the pulmonary blood circul ation, nitrogen absorption is believed to be limited. Yet these facts s eem contradictory to the findings of gas emboli in beaked whales, as diving-related gas emboli in hum ans are typically associated with pulmonary injury or autochthonous bubble growth due to improper management of nitrogen gases and/or air absorbed at depth (Brubakk et al 1999; Neuman, 1999). Indeed, recent studies have begun to show that gas embo li within stranded deep diving cetaceans are composed primarily of nitrogen gas, adding furt her support to the notion that under certain circumstances, nitrogen saturati on may be possible (Bernaldo de Quirs et al. 2011; 2012). Additionally, rec ent gas modeling in cetaceans has shown that the risk for nitrogen saturation may in fact be a concern (Fahlman et al 2006; 2009; Hooker et al 2009, 2011; Houser et al 2001). Fraser and Purves (1960) stated that the elaborate plexus investing the air sac system was apparently entirely subservient to the proper functioning of the latter. Given the intimate association of the a ccessory sinus system wit h the aforementioned venous plexuses, this seems like a reasonable conclusion. We suggest that the large venous investment of the air sinuses ma y provide an alternate mode of nitrogen absorption or eliminat ion during the course of a dive Such gas kinetics may be beneficial during normal diving, but may pose a threat of decompression sickness or venous gas embolism when diving conditions are ar tificially or inappropriately altered. During descent, the increasing pressure coul d drive nitrogen out of the sinus and into the surrounding venous plexus, as happens in the lungs. Admittedly, whether such
112 diffusion across the venous wall is possibl e will likely depend on t he thickness of the venous wall and the nitrogen partial pressure differential between the venous blood and sinus air. Although Fras er and Purves (1960) conduc ted some histological examinations of the lining of these sinuses, we suggest that future studies conduct histomorphometry in order to facilitate evaluation of t he likelihood of gas exchange across those surfaces. Since the venous pl exuses are not composed of capillary beds but rather larger veins, diffusion across thei r walls at normal pressures is unlikely, but would instead require considerab le pressure differentials to drive gas across. Modeling of such a function may provide insights in t he absence of physiological data. Although a reasonable assumption may be that only as mu ch nitrogen as is present in the sinus could ultimately be absorbed, the intr anarial connection between the pulmonary and accessory sinus system should not be ignored, since decreasing gas volume in the sinuses could be supplemented by pulmonary air in order to enable equalization of the middle ears. Interestingly, if indeed nitrogen gas can be ex changed at the lining of the accessory sinuses, a reduction in pressureas happens during ascentcould result in reversal of nitrogen flow from the blood to the sinuses, providing a non-pulmonary mechanism for elimination of nitrogen from the blood. During ascent, reducing pressure in the sinuses may be able to draw nitrogen out of the venous blood lining the sinuses. Nitrogen could then be clear ed through the Eustachian tube and out of the nasal passages. A recent analysis of the gas f ound in the pterygoid sinuses of stranded cetaceans showed that the sinuses have consistently high nitrogen gas levels (Bernaldo de Quirs, 2012). Additionally, elevated levels of CO2 were also found, suggesting that diffusion of CO2 across the sinus membranes shoul d also be examined. Given the
113 complexities of working with and interpreting results from postmortem specimens, more research on this topic is needed, however it may provide support for our suggestions. Another possible function of the accessory sinus vasculature that was postulated by Fraser and Purves (1960) may involve t he redistribution of blood into the venous plexuses to accommodate the reduction of air volume during descent. As suggested by Fraser and Purves (1960), this could allow th e lost volume of ai r to be replaced with blood to avoid dysbaric trauma to the sinus tissues as well as facilitating hydrostatic equilibrium of the ears whic h are surrounded by the sinus system. Such a pressurerelated redistribution of blood to the si nus plexuses may help explain the common presence of foam in the accessory sinus system, conceivably created by venous transudate that could help gen erate and stabilize the foam. Although such a bloodredistribution function may seem plausible in delphinids that have rigid bony lateral pterygoid laminae (Fig. 2), it may not explain the presence of an intric ate plexus in deep diving odontocetes that have a flexible late ral wall that can likely deform medially to accommodate the reduction in air volume c onsequent to compression. There is little doubt that deep diving marine mammals have evolved mechanisms for limiting nitrogen gas absorption and mitigating or managing bubble formation, whether in situ or intravascular. Nonetheless, given the extreme nature of the diving life styles of some of the deep diving cetaceans, it is possible that they live at the lim its of physiological tolerance, with small margins for error regarding gas management. Retention of these vascular plexuses despite the modified a ccessory sinus system anatomy may reflect their need for greater control over gas management.
114 Fat bodies The acoustic fat bodies of the lower jaw are i nteresting on many levels and have been a topic of considerable re search and debate. At the fo refront is their presumed function as analogs to the external pinnae of other mammals, receiving and channeling sound to the ears (Cranford et al 1996; 2008; Norris, 1969; Norris & Harvey, 1974). This suggested function naturally implicat es these structures in any discussion concerning auditory impairment resulting fr om intense ensonification. Perhaps less obvious is the possible role of the acoustic ja w fats as a source of fat emboli or as a nitrogen sink (Jepson et al. 2005). Koopman (2007) sugges ted that deep diving odontocetes such as beaked whales and sperm whales may fill their blubber with lipids that provide some type of physiological or mechanical advantage to diving (e.g. nitrogen sink). Mammalian fats have traditiona lly been considered poorly vascularized structures, and given the gener al paucity of cetacean vascular information, this paradigm has understandably be en propagated in the field of cetacean biology and discussions of vascular inve stment of fats (Houser et al 2001; Fahlman et al 2006; 2009; Hooker et al 2009). In contrast, our recent vascular research has shown that some cetacean fatty tissues are very well vascu larized. As noted by Fraser and Purves (1960), extraand intra-mandibular jaw fats are proving to be extensively vascularized by veins (see Figures 3-8 to 3-9 and 3-11) and often well vascularized by arteries. Jepson et al (2005) alluded to the possibility that damage to the tissue barrier between the acoustic lipid and the venous lumen could introduce fat into the circulation and lead to the formation of fat emboli. Such damage might result from physical trauma such as intense ensonification blunt or sharp force tr auma, or from traumat ic expansion of gas bubbles within nitrogen-saturated adipose tissue. Two factors make the acoustic jaw fat
115 interesting to us from this perspective. Fi rstly, fat is known to absorb nitrogen well and therefore can act as a sink for nitrogen that can expand once hydr ostatic pressure is reduced (Lango et al 1996). Secondly, the close anatom ical association of these fat bodies and their venous plexuses to the gas-filled sinuses and extensive pterygoid vascular networks may place them at increa sed risk of receiving either 1) elevated nitrogen levels absorbed and accumulated thr ough the sinus lining, or 2) gas emboli generated in the veins lining of the sinuses. Nitrogen absorbed thr ough the sinus lining could travel to the fat bodies via the robust connections described, being absorbed by the fat body and expanding within the adipose tissue, or expanding within the IMFB veins and disrupting the vascular barrier between the blood and lipid. Any of the aforementioned situations could conceivably re sult in physical damage of the fat with subsequent release of lipid into circulation. The high solubility of nitrogen in fat tiss ue is well-documented and results in high levels of nitrogen absorpti on in adipose tissue (Lango et al. 1996). Traditionally however it has been thought t hat nitrogen loading and unloading of fatty tissues is limited by the poor perfusion of the tissue (Fahlman et al 2006; 2009). Since the acoustic jaw fats are indeed well vascularized, it is conceivable that they may exchange nitrogen at higher rates than are typically asso ciated with fatty tissues Interestingly, there appears to be an exceptionally disproportionate number of veins relative to arteries relative to most ot her tissues. Though no functional studies exist from which to draw any conclusions, we suggest that an elevated venous density would allow for faster nitrogen elimination and may provide a mechanism for rapid nitrogen clearance in a vital tissue that when damaged can significantly reduce an individuals chances of
116 survival. We also cannot ignore the fact that the large volume and high surface area construction of the intramandibular fat body venous plexus may facilitate regional heterothermy of the intramandibula r fats that the veins inves t. Altering the temperature of the fat could affect the degree of solubility of nitrogen and therefore not only affect nitrogen loading and unloading but also influence the ease with which lipid can be mobilized and introduced into the vasculature, resulting in fat embolization. Another possibility is that by modulati ng the volume of venous blood in the mandibular acoustic fats, cetaceans may be able to alter the density of the fats either due to increased blood density or temperat ure change, thereby effecting a change in acoustic properties of the fat. This coul d allow cetaceans to either fine tune their hearing to specific frequencies or adjust the densit y of the fat to maintain their hearing in altered ambient temperatur e environments, as may happen during a transition from warm to cold water. Such functions may be supported by the finding that false killer whales are able to actively control their hearing based on the received echo from a target they are echolocating on (Nachtigall & Supi n, 2008). Finally, we cannot ignore the possible effect that regional mandibula r heterothermy may have on modulating other sensory components. A large por tion of the inferior alveol ar nerve of the mandibular branch of the trigeminal nerve passes through the intra-mandibular plexus on its course to the brain. If the intr amandibular plexus is capable of producing temperature changes within the mandible, those changes may be abl e to modulate the sensory input from the nerve by affecting the amplit ude, speed and/or duration of t he action potentials (Inman & Peruzzi, 1961; Ishiko & Loewenstein, 1961).
117 Numerous researchers have shown that temperatures of the blubber and extremities of cetaceans are often near ambient water temperatures, and since the mandible is a peripheral stru cture with high surface area it may frequently reach temperatures well-below normal mammali an core body temperature (Barbieri et al 2010; Meagher et al 2002; 2008; Noren et al 1999). Temperatures below 20oC have been shown to cause sharp decreases in the am plitude of the action potential of rapidly adapting Pacinian corpuscles, while t he action potential o ften disappears at temperatures below 15oC (Inman & Peruzzi, 1961; Ishiko & Loewenstein, 1961). Therefore, it may be possible for the mandibu lar plexuses to modulate the temperature of the surrounding tissues in order to ov ercome the adverse affects suboptimal temperatures may have on proper functioning of peripheral sensory nerves. Although this kind of temperature modulation of peripheral nerves might serve no apparent benefit in terrestrial mammals, th e role of the inferior alveolar nerve is not understood in cetaceans and this implication should perhaps not be completely discounted given the conductive heat loss that occurs in wate r and the highly derived adaptations of cetaceans. Interestingly, the melon fat appears to be drained by veins that are concentrated along the periphery of the fat body with large veins avoiding a direct course through the main substance of the melon fat. Conversely the IMFB is densely invested with veins. As we must assume that anatomical struct ures are not the result of evolutionary processes resulting in meaningless manifest ations, we are compelled to believe that this distinct difference may be reflective of a functional role. It is possible that this reflects different functional needs to modulat e physical characteristics of the fat bodies.
118 Clarifications/Inconsistencies Interestingly, Fraser and Purves (1960) noted the presence of distinct pterygoid and maxillary veins, with the maxillary vein tr aveling lateral to the paroccipital process and tympanohyal, and merging with the IMFB pl exus. They then noted that it merges with the deep temporal vein as it travel s caudad. They m ade no mention of the maxillary veins proximal connec tion to the external jugular vein as is common in domestic mammals, and noted that the pter ygoid vein eventually joined with the mandibular and internal jugular veins. They described the pterygoid vein as running along the pharyngeal crest for a consider able length, eventually terminating as countless ramifications of the fibrovenous pl exus associated with the accessory sinus system. We found this description to be conf using for a number of reasons. First, in most domestic mammals (except dogs) the pter ygoid vein is a branch of the maxillary vein, which like the linguofacial vein is one of the terminal branches of the external not internal jugular vein. Indeed, in our delphi nid specimens the intramandibular plexus and subsequent mandibular alveolar veins were s een consistently arising as a branch of the facial vein. Secondly, our findings in Tursiops suggest that the structure described by Fraser and Purves (1960) as t he pterygoid vein is in fact consistent in location and course with the maxillary veintraditiona lly considered the terminal branch of the external jugular veinand the pterygoid ven ous plexus is instead formed by numerous small pterygoid veins that branc h off of the maxillary vein and by the subdivision of the maxillary vein itself. Thirdly, since the inte rnal jugular vein enters the jugular foramen, the only large veins we have observed trave ling lateral to the tympanohyal cartilage and proximal stylohyal bone are t he continuation of the external jugular vein that wraps
119 around the tympanohyal cartilage, and the facial ve in (#3, Fig. 3-7) which is far removed from this location as it courses just deep to the blubber layer and subcutaneous fat. My findings are consistent with the patte rns observed by Ommanney (1932) in fin whales ( Balaenoptera physalus ) and by Walmsley (1938) w ho noted that the maxillary vein of the fetal fin whale is formed by ve ins accompanying the external carotid artery and tributaries from the pter ygoid venous plexus. Additi onally, the vein described by Fraser and Purves (1960) as the pterygoid vein formed in our specimens the terminal branch of the external jugular vein and passed dorsad through the infraorbital foramina, a pattern consistent with that of the maxilla ry vein of domestic mammals. Finally, we were not able to identify any substantial ve ins in the location identified by Fraser and Purves (1960) as the maxillary vein, unless t hey were referring to one of the veins of the external ophthalmic plexus (#7; rete vena ophthalmica externa of Slijper, 1936) or one of the many sizable veins of the pterygoid venous plexus. The aforementioned differences could be due to the fact that the specimens used by Fraser and Purves (1960) were decapi tated specimens that may have been missing the more proximal branches of the jugular veins. Nonetheless, the illustrations presented by Fraser and Purves (1960) sh ow the vascular and skull morphology from the occipital condyles rostrad, including the internal jugular vein, however the external jugular vein has been completely omitted, as has the origin of the maxillary vein. It may be that the proximal most trunk labeled as the internal jugular vein is in fact the common jugular, from which the internal and external jugular veins branch. The ventral branch (external jugular) would then give off the maxillary vein as it s primary branch. It is also possible that due to the use of decapita ted specimens, Fraser and Purves (1960)
120 considered the anastomotic branch of the in ternal jugular veinseen in all of our specimensas the main drainage path of the mandibular and maxillary veins, since they neither mention nor illustrate the external jugular vein. Another source of confusion for us was the reference made by McFarland et al (1979) to the rete vena ophthalmica externa first noted by Slijper (1936). McFarland et al (1979) showed a venous vascular cast of a large extracranial retial structure adjacent to the cranium and connecting to the spinal veins via the first three intervertebral foramina. McFarland et al state that this rete lies in the lower jaw and is probably the rete vena ophthalmica externa described by Slijper (1936). In none of our specimens did the intramandibular plexus connect directly to the epidural vein s. Although difficult to discern from the photographi c perspective provided, the positional references seem inappropriate and we therefore respectfully s uggest that the vascular cast may be of the plexus investing the parieto-frontal region that in our specimens had anastomoses with the superficial cervical and occipital veins an d ultimately the epidural circulation. Our findings suggest that the rete pictured by McFarland et al. (1979) may in fact be composed of numerous structura lly and regionally distinct retia that warrant separate nomenclature. Interestingly, despite the numerous fine caliber veins and plexuses that injected in our specimens, the only direct connections we observed between the spinal veins and the plexuses on ventral aspect of the skull were via the internal jugular emissary vein of the jugular foramen and the em issary vein of the fo ramen ovale. It is unclear at this time if this was an artifact of the injection medium, an individual variation, or a misidentification of structures; however it appears to be a significant difference.
121 Given the degree of plasticity inherent in venous connections, it is possible that these connections manifest differently betw een individuals of a given species. At the level of the medial aspect of the orbit, the late ral portion of the plexus associated with the Pty sinus fuses with dorsal offshoots of the intramandibular venous plexus and progresses dorsad to become the anterior lobe venous plexus. The largest portion of the anterior lobe venous plexus is located ventral and rostral to the eyestalk, and just ventral to the maxilla ry and premaxillary bones, in close association to the ventral lining of the preorbit al and anterior lobes of the accessory sinus system. The anterior lobe venous plexus acts as a crossr oads for numerous veins that converge on it. It has connections to intracranial bloo d via the ophthalmic plexuses that exit the orbital fissure, the melon via the facial vein and the infraorbital veins, and the ventral skull via the plexus of the accessory sinus system and maxillary vein. These collateral connections may be important for understandi ng compartmentalization of gases and observed distribution of gas and fat emboli. Upon initial examination of t he anatomy of the vasculature in the region of the Pty and peribullar sinuses, it becomes apparent t hat although there ar e homologies that may be drawn, the general pattern of venous investment seems relatively random. However, upon closer examination, it becomes evi dent that the locations and connections of the plexuses ar e relatively consistent bet ween individuals, with only the localized branching patterns within the plexus es showing much vari ability. Nonetheless, it appears that in addition to those struct ures with likely homologies to generic mammalian structures, almost every availa ble space between the jugular veins and distal head structures is f illed with plexuses of anastomosing veins, forming what can
122 only be considered a complex of countless collateral pathways. Although deciphering the exact anatomy of the venous structures themselves is important, the connections between those structures may be equally impor tant from a functional perspective, as they form collateral pathways and alter nate routes for trans port of emboli and compartmentalization of gases. The paired cavernous sinuses of domes tic mammals form a ring-like venous structure around the pituitar y gland (#16, Figures 3-12 and 3-15), just dorsal to the basisphenoid bone (Ghoshal et al 1981). Rostrally the sinuses connect to the ophthalmic veins via the orbital fissure, while caudally they run conf luent with the ventral petrosal and basilar sinuses. The ventral petrosal sinuses connect the caudolateral aspect of the cavernous sinuses to the vent ral margin of the sigmoid sinuses (Evans, 1993). In Tursiops these sinuseslocated at the floor of the brainconnect the cerebral venous circulation to the plexuses of the accessory sinus system. Fraser and Purves noted that the petrosal and cavernous sinuses of cetaceans are divested of the bony cranial protection found in terrestrial mammals, due to the displacement of the tympanoperiotic from participation in the wall of the cranium. (1960) I found this statement confusing since the cavernous sinus of most domestic mammals lies in a similar positionrelative to the basicranial bonesas what we observed in dolphins (Evans, 1993; Ghoshal et al 1981; Schaller, 2007), namely on the dorsal aspect of the preand basisphenoid bones. Fras er and Purves (1960) then stated that in cetaceans t he cavernous tissue body or spongy mass of Beauregard is the homologue of the caver nous sinus of terrestrial mammals. This too seems problematic since the spongy mass of B eauregard was illustrated by Boenninghaus (1903) as surrounding the tympanic bulla of the ear (if it is the same as the structure
123 labeled corpus cavernosum ), ventral to the ventral petro sal sinus. This is an unusual location for the cavernous sinus. We suggest that the cavernous sinus of t he bottlenose dolphin is in fact in the expected intracranial location just rostral to the clivus, surrounding the sela turcica and pituitary gland, and that the spongy mass of Beauregard is in fact a novel cetacean structure that represents a complex of interconnected veins between the ventral petrosal sinus, peribullar venous plexus and pterygoid venous plexus. Although our search was not exhaustive, we have f ound no homologous structure in domestic mammals. The cavernous sinuses of Tursi ops are directly connected to the middle meningeal veins, the ophthalmic plexuses, t he ventral petrosal sinuses, pterygoid plexuses and the emissary veins of the foramina ovalia (emissary vein of foramen lacerum medium of Boenninghaus). The tw o lateral components are connected across the midline via structures much like the in tercarvernous sinuses of domestic mammals. This venous loop is therefore consistent in location and drainage with the cavernous sinus of terrestrial mammals, and seems contradictory to the stat ements of Fraser and Purves (1960), as it is loca ted well within the calvarium. An additional source of confusion arising from the de scriptions of Fraser and Purves (1960) was found in their statement that under the hydrostatic pressures available, the corpus [cavernosum ] could be erected by way of the internal carotid. This appears to conflict with our findings as well as Boenni nghaus illustrat ion that the corpus cavernosum is venous in nature, connected by ventral tributaries to th e pterygoid and maxillary veins (#6, Figure 3-15A-D) and dorsal tributaries to the ventral petrosal sinus (#27, Figure 3-15D), rather than the carotid arterial system. Althoug h an arterial component was observed, it
124 appears modest and paled in comparison to the venous component. Additionally, the corpus cavernosum appeared grossly identical to the surrounding peribullar and pterygoid venous plexuses that connect directly to it. Therefore, it seems reasonable to assume that if the corpus cavernosum does indeed function as erectile tissue, this function can likely also be attributed to the venous plexuses surrounding it. Synopsis Intriguing features of the veins in the head of the dolphin are the sheer volume and complexity they display. Indeed, the veins show similarly complex patterns throughout the dolphin body, often times occupying ev ery available space within cavities and between tis sues. In the head and neck, venous plexuses are seen investing the intraand extramandibular fat bodies, the cranial sinuses, ophthalmic regions, nasal passages, tracheal mucosa, surrounding t he esophagus, within the epidural spaces, and inside the brain case. It is possible t hat the venous system has simply formed a network of collateral drainage pathways throughout the body, t hat can facilitate adequate drainage of blood from t he central nervous system dur ing periods of elevated pressure (e.g. Valsalva phenomenon), in or der to avoid damaging pressure-sensitive nervous tissue (Monro-Kellie doctrine). Al ternatively, it may be that despite the plexiform similarity, different portions of the venous system serve different purposes, whether collateral drainage, regional heterothermy, gas exchange, or some other function. For instance, the venous plexus found within the tracheal mucosa has been hypothesized to act as either a compensati ng mechanism for the reducing air volume during diving-related compression or as erec tile tissue that modifies the deformation properties of the tr acheal wall (Cozzi et al 2005). Similarly, Fraser and Purves (1960)
125 suggested that the role of the pterygoid plexuses was tied to the proper functioning of the pterygoid sinuses. The predominant piece of information missi ng from this anatomic al picture is the lack of functional physiological data on blood flow and gas characteristics in the aforementioned structures. We str ongly suggest that further studies of these structures may shed light on important physiological processe s previously discounted or neglected. We believe that a considerable amount of information can be garnered from further postmortem studies, however we feel that rela tively-non invasive, ethical, live cetacean experiments on functional morphology can be conducted and may be the only way to clarify some of the large data gaps that currently exist.
126 Table 3-1. List of specimens used for this study. Specimen ID Species Common Name TBL (cm)Gender Date Stranded Research Use CMA1109 T. truncatus Bottlenose dolphin 191F 05/17/11 A, V Hubbs0909 T. truncatus Bottlenose dolphin 249F 02/27/09 A, V ECW005 T. truncatus Bottlenose dolphin 284M 12/16/10 A, V MMC-Tt0708 T. truncatus Bottlenose dolphin 256M 07/25/08V MMC-Tt0107 T. truncatus Bottlenose dolphin 271F 11/12/06 I BRF164 T. truncatus Bottlenose dolphin 262F 07/23/07 V, S RJM003 T. truncatus Bottlenose dolphin 188F 06/28/08 V, S VMSM20031104 T. truncatus Bottlenose dolphin 204M 12/26/03 S VAQS20061067 T. truncatus Bottlenose dolphin 174M 08/06/06 V,S PBN003 T. truncatus Bottlenose dolphin 246F 02/14/08 V,S VAQS20051086 T. truncatus Bottlenose dolphin 195M 07/17/05 S MML0802 F. attenuata Pygmy killer whale 208M 06/30/05 A, V MMCPm 0908 P. macrocephalus Sperm whale 324F 09/29/08 A, V VAQS2008 1002 K. sima Dwarf sperm whale 160M 01/28/08 S FMMSN0906 K. breviceps Pygmy sperm whale 167M 07/10/09 A, V VAQS 20071006 K. breviceps Pygmy sperm whale 263F 02/27/07 V,S
127 Table 3-1. Continued. Specimen ID Species Common name TBL (cm)Gender Date stranded Research use CLP001 K. breviceps Pygmy sperm whale 225M 11/22/07 V,S KLC059 K. breviceps Pygmy sperm whale 224M 12/16/09 S KLC025 K. breviceps Pygmy sperm whale 213F 12/22/08 V,S MDB056 K. breviceps Pygmy sperm whale 264M 12/15/09 S MARS0903 M. europaeus Gervais' beaked whale 229M 08/02/09 A, V WAM593 M. densirostris Blaineville's beaked whale 423M 01/28/04 S MDB023 M. densirostris Blaineville's beaked whale 434F 09/15/08 S VAQS20091107 M. bidens Sowerby's beaked whale 397M 11/08/09 S Abbreviations for genus name are as follows: Tursiops ( T ), Feresa ( F), Physeter ( P ), Kogia ( K ), Mesoplodon ( M ). Abbreviations for research use are as follows: Angiography (A), MRI/CT Imaging (I), Vascular Dissection (V), Sinus Dissection (S). Table 3-2. List of soft tissue (blood vesse ls and nerves) structure labels used in the figures and their corresponding names. Structure Label Structure Name 1 Internal jugular vein 1' Anastomotic branch to ex ternal jugular vein 2 External jugular vein 3 Facial vein 4 Submental vein 5 Mandibular vein 6 Maxillary vein 6' Palatine plexus
128 Table 3-2. Continued. Structure Label Structure Name 6'' Nasopharyngeal veins 6''' Dorsal maxillary vein continuation 7 External ophthalmic plexus 8 Anterior lobe plexus 9 Melon veins 10 Intramandibular venous plexus 10' Mandibular peri-arterial venous rete /mandibular alveolar veins 11 Pterygoid plexus 12 Peribullar plexus 13 Temporal sinus 14 Dorsal sagittal sinus 15 Epidural veins 16 Cavernous sinus 17 External carotid artery 18 Internal carotid artery 18' Regressed terminus of in ternal carotid artery 19 Venous plexus surrounding ex ternal carotid artery 20 Extramandibular venous plexus 21 Vagus nerve 22 Hypoglossal nerve 23 Corpus cavernosum of Boenninghaus 24 Emissary vein of oval foramen 25 Emissary vein of jugular foramen 26 Bulbous venosus epibularis of Boenninghaus 27 Ventral petrosal sinus
129 Figure 3-1. Plate combining computed tomogr aphic (CT) slices with volume renderings of the pterygoid and peribullar plexus (g reen), intramandibu lar fat body plexus (yellow), and anterior lobe (purple) to illustrate overall location of the structures. Bright white areas in CT slices, and dark blue structures in volume renderings represent contrast enhanced venous structures. Panel (A) shows the plane of section for panels (B -D). Panel (B) shows a transverse view, panel (C) shows a coronal view, and panel (D) shows a sagittal view. Panel (D) shows a reference outline of the external surface of the dolphin overlaid on a three dimensional reconstruction of the venous system (blue) and structures of interest. Panel (E ) shows only the venous system so the more detailed structures can be observed. Panel (F) shows a medial view of
130 the veins of the right side of the h ead, as seen from a midsagittal plane of section. Note that the small inset panels within each larger panel show the orientation of the plane of section for each panel. Figure 3-2. Schematic illustration of dorsa l and ventral skeletal associations to the accessory sinus system in a bottlenose dolphin (left) and a pygmy sperm whale (right). Three letter abbreviations refer to bone names as follows: Als alisphenoid; Boc basioccipital; Ex o exoccipital; Frn frontal; Lac lacrimal; Mas mastoid (mastoid process of the periotic bone) ; Max maxilla; Osp orbitosphenoid; Pal palatine; Pty pterygoid; Sqa squamosal; Tym tympanic bulla, Vom vomer.
131 Figure 3-3. Schematic illustrations of late ral and ventral views of the pterygoid sinus system in bottlenose dolphins (left) and pygmy sperm whales (right). Note the more complex lobes observed in dolph ins, relative to the simple geometry of the system in pygmy sperm whales.
132 Figure 3-4. Medial view of t he pterygoid venous plexus that lines the lateral wall of the pterygoid sinus of a neonatal sperm whale ( Physeter macrocephalus ). Note the considerable volume and complexity, and the very thin wall separating the venous blood from the sinus air.
133 Figure 3-5. Cross-sectional computed tomographic image at t he level of the eyes of a pygmy sperm whale with contrast enhanced veins. The green regions represent the pterygoid venous plexus that surrounds much of the pterygoid sinus. Note that during postmortem ex amination the ptery goid venous plexus is usually empty and the green-colored space being occupied by it in this image is occupied by air in the expanded air sinus.
134 Figure 3-6. Magnetic resonance imaging cross-sectional view of the head of a bottlenose dolphin at the level of the ey es, showing association of the intra(IMFB) and extramandibular (EMFB) fat bodies. Also labeled are portions of the melon (Mln), pterygoid sinus (P ty) and connection between the preorbital (Pro) and postorbital (Pso) lobes of the pterygoid sinus system.
135 Figure 3-7. Three-dimensiona l angiographic reconstruction of the left lateral aspect of the head of the bottlenose dolphin, showing associations of superficial veins and bony elements. Veins were assigned names as follows: (1) internal jugular, (3) facial vein, (4 ) submental vein, (5) mandibular vein, (6) maxillary vein, (7) external ophthalmic plexus, (8) anterior lobe plexus, (9) melon veins. The red structure repres ents the tympanoperiotic complex. The external jugular is located medial to the faci al vein and can t herefore not be seen.
136 Figure 3-8. Lateral view of right dentary (image has been flipped to simulate left lateral orientation for consistency between images) with a window cut out of the lateral wall in order to visualize the intramandibular fat body plexus. Veins were assigned names as follows: (3) facial, (5) mandibular, (10) intramandibular fat body plexus. Note that red asterisks show the main connection between the intramandibular and extramandibular fat body veins.
137 Figure 3-9. Ventrolateral view of the ri ght side of the bottlenose dolphin neck showing the jugular and facial veins, and the complex anastomoses between the structures (image has been flipped to si mulate left lateral orientation for consistency between images). Structures were assigned names as follows: (1) internal jugular, (1) anastomotic branches to the external jugular and to plexus surrounding external carotid artery (2) external jugul ar, (3) facial, (5) mandibular, (6) maxillary, (17) external carotid arte ry, (18) internal carotid artery with surrounding venous plexus, (19) venous plexus surrounding the external carotid artery ( vena plexi commitans arteria carotidis externa ), (20) extramandibular fat body plexus, (21) vagus nerve, (22) hypoglossal nerve.
138 Figure 3-10. Ventromedial view of the right half of a mid-sagittally sectioned bottlenose dolphin showing the jugular branching patte rns. Veins were injected with blue latex and arteries with red. Structur es were assigned names as follows: (1) internal jugular, (1) anastomotic br anches to the external jugular and to plexus surrounding external carotid artery (2) external jugul ar, (6) maxillary, (11) pterygoid plexus, (12) peribullar plexus, (13) temporal sinus, (15) epidural veins, (16) cavernous sinus, (18) internal carotid artery with surrounding venous plexus, (18) regressed terminus of internal carotid artery, (19) external carotid artery with su rrounding venous plexus, (24) emissary vein of foramen ovale, (25) emis sary vein of jugular foramen.
139 Figure 3-11. Cross-sectional view of the intramandibular fat body (IMFB) of a bottlenose dolphin, progressing from caudal (l eft) to rostral (ri ght). Note the extensive investment of the fat with a venous plexus (10). Also note the periarterial venous rete (10-PAVR) surrounding the red mandibular alveolar artery in the dorsal region of the fat body, and the regions of un-injected fat highlighted with red asterisks near the ventral aspect. Due to the location and association with the concomitant ar tery, the PAVR has been termed the mandibular alveolar plexus.
140 Figure 3-12. Medial view of gross dissection of a bottlenose dolphin with latex injected vessels. Approximate scale bars have been added whenever a ruler was not present in the photograph. Panel (A) s hows the location of the pharyngeal plexus (red asterisks) in relation to the esophagus (Eso) and laryngeal cartilages (Lnx). The red structure represents the nasal passage surrounded by the palatopharyngeal muscle (Pph). Panel (B) shows the pharyngeal plexus (red asterisk), the maxillary vein (6) sending a palatine plexus (6) that invests the roof of t he oral cavity, and contributing to the nasopharyngeal veins (6) that drain the palatopharyngeal muscles (Pph). Also visible are the
141 cavernous sinus (16), pituitary gl and (Pit), and presphenoid (Psp) and basisphenoid (Bsp) bones that form part of the floor of the brain case. Panel (C) shows the contribution of the maxillary vein (6) to the formation of the lateral wall of the pterygoid plexus (11). Also visible is the pharyngeal plexus (red asterisk). Panel (D) shows the m edial aspect of the peribullar plexus (12) and dorsal pterygoid plexus (11) with the bone of the pterygoid crest removed. Also visible are the emissary vein (24) traveling through the oval foramen, the temporal si nus (13) and epidural venous plexus (15). The red structure represents the Eustachian tube which ext ends from the tympanic bulla to the pharynx along the pharyngeal crest. Note the visible blue latex within the hamular (Ham) lobe of the Pt y. This portion of the plexus can displace the entire volume of air within the hamular lobe. Panel (E) shows the dissection of the dorsal pt erygoid and peribullar plexuses. Panel (F) shows the opened pterygoid and peribullar sinus, ex posing part of the internal sinus lining and the connection between the late ral portion of the pterygoid venous plexus (11) that connects to the IMFB plexus. Note that the small inset panels within each larger panel show the orientation of the plane of section for each panel.
142 Figure 3-13. Lateral view of a volu me rendering of computed to mographical angiography of a bottlenose dolphin head showing complexity of venous invest ment. The bone has been removed in the rendering to allow viewing of the spatial relationships between the superficial and deep venous structures. Veins were assigned names as follows: (1) internal jugular, (3) fa cial, (3) anastomotic branch to the maxillary vein, (4 ) submental, (5) mandibular, (6) maxillary, (7) external ophthalmic plex us, (8) anterior lobe plexus (9) melon veins, (10) intramandibular fat body plexus, (10) m andibular alveolar plexus (13) te mporal sinus, (14) dorsal sagittal sinus, (15) epidural venous plex us, (16) cavernous sinus.
143 Figure 3-14. Mid-sagittal view of a volume rendering of computed tomographi cal angiography of a bottlenose dolphin head showing complexity of venous investment along medi al aspect. The bone has been removed from the rendering to allow viewing of all venous structures. Veins were assigned names as follows: (1) internal jugular, (1) internal jugular anastomosis with external jugular (2) external jugular, (3) facial, (4) submental, (5) mandibular, (6) maxillary, (6) dors al continuation of the maxillary vein (8) anterior lobe plexus, (9) melon veins, (10) intramandibular fat body pl exus, (11) pterygoid sinus plexus, ( 12) peribullar sinus plexus, (13) temporal sinus, (14) dorsal sagittal sinus, (15) epidural venous plexus, (16) cavernous sinus. Note the two red asterisks showing the two basicranial emissary connections between intracranial and ex tracranial veins. These emissaries form robust connections bet ween the plexuses of the pterygoi d sinus system and the dural sinus system.
144 Figure 3-15. Medial view of gross dissection of the pter ygoid and basicranial regions, identifying some of the key venous stru ctures outlined in the text (structure numbers correspond to those in previous images). Panel (A) shows the medial wall of the plexuses associ ated with the accessory sinus system and their emissary connections (24 & 25) to the intracranial dural sinuses. Panel (B) shows the pterygoid (11) and peribullar (12) pl exuses being opened, to see the lateral wall of the pterygoid sinus plexus and the corpus cavernosum of Boenninghaus (23). Panel (C) is sim ilar to Panel (B) but with the peribullar sinus opened further to expose the bulbous venosus epibularis of Boenninghaus (26) emerging from within the tympanic bulla (red asterisk) and merging with the dorsal part of the pterygoid and peribullar plexuses. Panel (D) is similar to Panel (C) but with t he dorsal aspect of the peribullar plexus opened further to expose the ventral petrosal sinus (27) and its caudal
145 connection to the internal jugular emissary vein of the jugular foramen (25). The tympanic bulla is marked with a red asterisk. Panel (E) shows an oblique rostral view of the left dor solateral extension of t he pterygoid plexus (11) around the external pterygoid muscle (Pte) as it forms the dorsal continuation of the maxillary vein (6) that passe s through the infraorbital foramina. Also shown is the extensive and complex natur e of the anterior lobe plexus (8) and the anterior lobe (red asterisks) it is associated with. Panel (F) shows a simplified schematic illustrat ion of the medial view based on a modification of Boenninghaus (1903) illustration of the venous connections in the pterygoid and basicranial regions. Shown are t he main patterns of venous connections of the plexuses associated with the in tramandibular fat body, accessory sinus system and the intracranial dural system Lighter blue vessels are more lateral (behind) darker structures. The brown structure represents the pituitary gland. Note that the corpus cavernosum of Boenninghaus (23) is not illustrated here due to its location latera l (behind) the peribullar plexus (12). Object 3-1. Three-dimensional reconstruc tion of a bottl enose dolphin with a contrastenhanced venous system, showing spatial relationships of the structures discussed. As the left lateral structur es are clipped away, the medial aspect of the pterygoid region comes into view Note that the colors that appear correspond to the same colors presented in figure 1: pterygoid and peribullar venous plexus (green), anterior lobe venous plexus (purple), and IMFB plexus (yellow).
146 CHAPTER 4 THE GROSS MORPHOLOGY OF THE VENOU S SYSTEM IN THE HEAD AND NECK OF THE FLORIDA MANATEE ( TRICHE CHUS MANATUS LATIROSTRIS) Chapter Foreword Manatees are herbivorous aquatic mammals that inhabit shallow c oastal and inshore waters of their tropica l and subtropical distribution. They are members of the order Sirenia, a monophyletic group that includes the extant dugongs and extinct Stellar sea cows (Marsh et al. 2012). Florida manatees ( Trichechus manatus latirostris ) are a subspecies of the Antillean manatee--one of three extant manatee species found worldwideand as an endangered marine mamma l are protected under the U.S. Endangered Species Act and the Marine Ma mmal Protection Act. The phylogeography of the Florida manatee suggests that current Florida manatee populations arose from the Protosirenidae family in the early middle to late Eocene epoch (Marsh et al. 2012). Although most molecular and morphological in vestigations place manatees in a clade related to elephants and hyraxes (Springer et al. 1999; Soshani, 1986), the similarities are not always apparent and phylogenetic in ferences regarding their anatomy and physiology have been challenging. What is clear is that manatees have been evolving their permanently aquatic lifestyle for over 60 million years (Marsh et al. 2012). Due to the Florida manatee s protected status, the Florida Fish and Wildlife Conservation Commission manages a statewi de stranding network responsible for rescuing all injured manatees and recoveri ng all dead manatees within the state. The causes of morbidity and mortality vary fr om anthropogenic causes such as watercraft interactions and entanglements, to natural causes such as cold stress and red tide intoxication (Lightsey et al. 2006). Manatee cold stress syn drome is a unique pathology among marine mammals and is believed to be a cascade of physiological dysfunction
147 brought about by continued exposure to water temperatures below 20oC (68oF). The syndrome is chronic in nature and is disti nguished from acute hy pothermia by prolonged survival, chronic supporative dermal ulcerations, gastrointestinal quiescence and constipation, emaciation, secondary in fections, and eventual death (Bossart et al. 2002). The skin lesions begin as focal bleachi ng or loss of pigment primarily along the rostrum, pectoral flippers and tail, followed by progressive ly more disseminated lesions along the flanks and ventrum. If exposure to cold continues, the bleached areas begin to form ulcers, which subsequently disrupt the skins barrier to infection (Bossart et al. 2002). It is not known why manatees are so suscept ible to cold temperatures when other marine mammals thrive in much cooler wate rs, however some have postulated that their lower than average metabolic heat production is to blame. Manatees show lower basal metabolic rates than terrestrial mammals of the same mass, while other marine mammals such as odontocetes, seals, and sea otters are believed to have two to three times the rate of terre strial mammals (Costa & Trillmich, 1988; Costa et al. 1989; Gallivan & Best 1980; Hampton et al. 1971; Irvine, 1983; Morrison et al. 1974; Ridgway & Patton 1971; Scholander & Irving 1941; Scholander et al. 1950;). An elevated metabolic rate is a useful adaptation to an aquatic lifestyle, since water conducts heat from the body about 25 times faster than air does. Therefore, the significantly lower metabolic rate of manatees may predispose them to cold stress, a fact that could explain their warm water distribution. What is interesting in the context of this research is that the dermal ulcerations that manifest in cold-stressed manatees may be the result of ischemic necrosis of the ski n. Peripheral vasoconstricti on is a natural physiological
148 response in all mammals exposed to cold te mperatures, and should therefore not come as a surprise (Bornmyr et al. 2001; Folkow et al. 1963; Hughes et al. 1984; Meyer & Webster, 1971; Webster & Johnson 1984;). Notably, the cold-stress ulcerations initially manifest along the rostrum, fluke, and pecto ral flippers--the extrem ities. If then the ulcerations are caused by vascular or hemodynamic changes and the lesions first appear along the distal extremities, it would seem reasonable to postulate that the vasculature in those extremities carries some inherent significance regarding the manifestation of cold stress lesions. Unfortunately, although the vasculature in the pectoral flippers and fluke has been described, li ttle is known about the vasculature in the manatee rostrum (Murie, 1874; Reep et al. 1998; 2001; Stannius, 1845). Another related thermal imp lication of the facial vasculature is with respect to conservation of heat during cold exposure. Using a thermal infrared camera, Dehnhardt et al. (1998) showed that harbor seals immers ed in water with mean temperatures of 1.2oC maintained significantly elevated ther mal signatures along the eyes and facial vibrissae despite the near ambient temperatures observed along the rest of their external surfaces. They interpreted thes e findings as proof that no vasoconstriction occurs in these sensory areas during cold acclimation, and justified it based on the thermal needs of the vibrissal mechanoreceptor s to function proper ly. Temperatures below 20oC have been shown to cause sharp decreas es in the amplitude of the action potential of rapidly adapting Pacinian corp uscles, while the action potential often disappears at tem peratures below 15oC (Inman & Peruzzi, 1961; Ishiko & Loewenstein, 1961; Necker, 1983). Dehnhardt et al. (1998) also noted that the mechanical properties of the tissues surrounding the vibrissae are sensitive to temperature and would
149 therefore be adversely affected by cold tem peratures. The findings of Dehnhart and his colleagues are of great interest in this discu ssion, because manatees have similar facial vibrissae that play an integral part in t he sensory exploration of their environment (Marshall et al. 1998a; 1998b; Reep et al.1998; 2001). Since manatees often find themselves in thermally challenging environments, it seems likely that they may face a similar conundrum as seals, that being the choice between lim iting their convective heat loss and maintaining adequate vibrissal sensor y function. Therefore, manatees may also need to maintain elevated vibrissal temperatures in order to maintain optimal mechanoreceptor function. Interestingly, 20oC water temperaturesat which t he action potential of Pacinian corpuscles is affected (Inman and Peruzzi, 1961)--are considered the threshold below which manatees begin to get cold stress syndrome. Cold stressed manatees may, therefore, face challenges in maintaining adequate function of the vibrissal mechanoreceptors, in light of their need to conserve body heat. Although Dehnhardt et al. (1998) showed that seals sacrifice body heat to maintain elevated vibrissal and ocular temperatures, the lower metabolic heat production of manatees may not allow them the luxury of dissipati ng heat to the environment. Therefore, at low ambient temperatures, manatees may be forced to al low vibrissae and their surrounding tissues to reach sub-optimal temperatures by lim iting either the am ount of blood or the temperature of the blood reaching the vibrissae. The vasocontrictive mechanism suggested by Dehnhardt et al results in reduced blood fl ow that can have multiple detrimental effects such as ischemic or hypoxic necrosis and reduced mechanoreceptor function due to decreased hydrostatic pressure within the blood sinuses of the vibrissae.
150 Instead of reducing the blood flow, an alternative thermal conservation strategy commonly employed by mammals is through va scular counter-current heat transfer. Such anatomical adaptations are elaborate in the extremities of marine mammals and even sloths, and are known to play an important role in facilitating heat conservation by drawing heat from the warm efferent arteries and returning it to the core via the afferent veins (Scholander, 1958; Schol ander & Krog, 1957; Scholan der & Schevill, 1955). Such vascular structures can be found in the chevron canal of sirenians and cetaceans, the dorsal fins of cetaceans, and the pecto ral flippers of manat ees (Elsner, 1966; Fawcett, 1942; Murie 1872; Pabst et al. 1995; Rommel & Caplan, 2003; Rommel et al. 1992; 1995; 1996; 2001; Scholander & Schevill, 1955; Stannius, 1845). It stands to reason that if manatees have t he ability to maintain warm vi brissal temperatures during mild environmental temperatures but must lim it convective heat loss through the face during cold temperatures, they may have vascular structures that facili tate such regional heterothermy. Although there has been considerable long-te rm effort placed on the collection and postmortem examination of Florida manatees to ascertain cause of death and collect life history data, some of the mo st basic biological information is still missing. Because the vascular morphology of domestic mammals has been well-defined for decades (Evans, 1993; Ghoshal et al. 1981; Hegedus & Schacke lford, 1965; Schummer et al. 1981), it often comes as a surprise to learn about the paucity of published information describing the vasculature of marine mammals. The paucit y of anatomical information is especially true for manatees, presumably due to their limit ed global distribution, large size, their elusive nature and protected status. To dat e, the vasculature in the head of manatees
151 has remained mostly ignored. Descriptions of the gross anatomy of blood vessels in the manatee have been limited and the majori ty of the available literature is difficult to either obtain or translate. This is especially tr ue for the veins in the head of the manatee. Stannius (1845) described aspects of the arterial system of the Florid a manatee, however the venous system remained almost untouched. In his wonderfully illustrated and scholarly monograph, Muries sole references to head veins stated that branches [of the veins of the face and head] were observed to return from the submaxillary region and outside of the jaw; these converge below the parotid gland and join the external j ugular opposite the paramastoid (1874) and the veins of the face and head were not followed in detail. Similarly, Fawcett briefly made mention of a retial-type of arteriovenous arrangement in the head of manatees when he stated the external carotid breaks up into a terminal spray of vessels which radiate, each in association with two veins, to supply various superficial structures of the side of the head. (1942) Echols (1984) described the branching of the cranial vena cava and veins of the pectoral flippers, but did not address head and neck veins. Finally, Rommel et al. (2001) and Rommel and Caplan (2003) descr ibed the vessels associated with the reproductive tract and the chevron canal respectively, but did not address the morphology of the venous syst em in the head and neck. Notably, even though literature on the vascular morphology of cetaceans and pinnipeds does exist (Fraser & Purves, 1960; Harrison & Tomlinson, 1956; Hol et al. 1975; King, 1977; McFarland et al 1979; Murie 1973; Ommanney; 1938; Ridgway et al. 1974; Slijper, 1936; Walmsley 1932), the striking morphologic al and phylogenetic differences between manatees and other members of the marine mammal group necessitate separate considerat ion of the anatomy. Indeed, what is known about the
152 extensive diversity of permanently aquat ic marine mammal groups should provide strong impetus to avoid generalizations and assumptions regarding the anatomy and physiology of these species. In thei r seminal work, Fraser and Purves (1960) suggested that our general lack of knowledge of venous morphology in cetaceans is due in large part to the extreme complexity t hat is found in their venous system. This theory may also ring true in the case of manatees. Before venturing into a discussion of marine mammal vasculature, it seems prudent to discuss the vascular patters obser ved in domestic mammals, as they often act as a reference point. The venous mor phology in the head of the various domestic mammals follows certain patterns. As woul d be expected, considerable variation exists between different species, however veins such as the maxillary a nd facial veins are found in all, and most of their major branches can be followed relatively easily in all species (Figure 6-1). As its name implies, one of the main tributaries of the external jugular vein-the facial vein-is responsible for draining most of the superficial structures of the face. This vein branches either sole ly from the external jugular vein or in combination with the lingual vein as the lingu ofacial vein, and extends around the lateral surface of the lower jaw to the rostrum. Along the rostrum the facial vein emits numerous branches to the lower and upper lips and tissues surrounding the nasal passages, before turning back in an arching fashion over the eye to anastomose with the eye vasculature and continue to the front al region. The aforementioned pattern is highly conserved and seen in all domestic mammals despite their various facial modifications and anatomical specializations.
153 The other main tributary and typically cons idered to be the terminal branch of the external jugular vein, is the maxillary vein Though there is considerable variability in the path and exact branching of the maxillary vein of different domestic mammal species, the general location and drainage field of the vein is fairly consistent, though undoubtedly less consistent than the facial vein. The maxillary vein can be considered the main vein responsible for draining the majority of deep struct ures of the head. Specifically, drainage fields include parts of the brain, internal nasal passages, pterygoid regions, and upper and lower jaws, as well as structures on the top of the head such as the external ear pinnae and horns. It should come as no surprise that the maxillary vein and its branches vary more between species than does the facial vein, since there is substantial diversity of mammalian morphotypes, with some bearing large ears and horns while others hav e small ears and no horns. Additionally, the degree to which a species masticates its food also appears to bear some influence on the degree of venous complexity in the pterygoid, bu ccal, and deep facial regions. As such, animals that chew their food (e.g. horses, ruminants, pigs) show enlargement of certain veins in this region (e.g. buccal and deep facial veins) compared to carnivores. Contraction of the chewing muscles su rrounding these enlarged buccal and deep facial veins and venous sinuses is believed to serve a non-cardiac pumping role for returning venous blood from the face to the heart (Dyce et al. 2002). The internal jugular veins are much more variable, playing important roles in the dog and pig, while being completely abs ent in small ruminants and most horses (Schummer et al. 1981). When present, the internal j ugular vein usually passes through the jugular foramen at the c audal region of the basicranium, to form a jugular emissary
154 vein that connects to the veins at the floor of the brain, specific ally the sigmoid dural sinus. These emissary veins usually act as pathways for drainage of some of the blood returning from the brain. The internal jugular vein also gives off the occipital vein in some species, while in other species the o ccipital vein branches off of the external jugular or maxillary veins (Schummer et al. 1981). Despite its parent vein, the occipital vein usually connects to the vertebrobasilar circulation via anastomotic branches of the vertebral veins as well as to the floor of the brain via emissary veins. Hegedus and Shackelford (1965) noted that as the structures of the face, nose, mouth, and ears developed to a greater degree, th e external jugular vein of the dog, sheep, rabbit, ox and horse took on a predominant role in the drainage of the brain. Conversely, in humans, monkeys, cats and pigs the inter nal jugular veins dom inate in the adult. Although I could find no dire ct statements from other aut hors corroborating the exact contributions mentioned by Hegedus and Schackelford (1965), Reinhard et al. (1962) are in agreement regarding the dog and human and Gillilan and Markesbery (1963) agree about the cat, human, and monkey. No specific mention was made by Schummer et al. (1981) about relative contributions of the jugular branches to drainage of the blood from the brain, however in their illustrations they represent the internal jugular veins of the dog and ox as being of relatively significant size, though clearly not as large as the external jugular veins. Given the substantial muscular and sensory investment of the face of t he manatee, it seems reasonable to expect similar dominance of the external jugular veins ov er the internal jugular veins. Since manatees share some similarities in habitat with cetaceans and pinnipeds, it is likely prudent to highlight what is known about the vasculature in those other very
155 derived marine mammal morphotypes. Of the pinnipeds and cetaceans, the pinnipeds are arguably less derived than cetaceans. Their external and internal gross morphology is much more reminiscent of terrestrial carnivores, especia lly canids and bears. What little literature exists regarding the venous morphology of their head and neck suggests that there are significant similariti es with the terrestrial phenotype (Barnett et al. 1958; Folkow et al. 1988; Harrison & Tomlinson, 1956; King, 1977; Murie, 1973; Ronald et al. 1977; Sobolewski, 1986). It shoul d be mentioned that the Order Pinnipedia includes seals and sealions, which although relatively si milar in external form, bear some striking anatomical differences. Seals have developed significantly derived venous structures such as enlarged hepatic and epidural venou s sinuses and elaborate, voluminous plexiform cervical and ingu inal plexuses (Barnett et al. 1958; Harrison & Tomlinson, 1956; Rommel et al. 1995; Ronald et al. 1977), whereas sealions appear to be a little more similar to their terrestrial ancestors Therefore, although the jugular venous branches and their major tributaries appear to be present in all pinnipeds and account for much of the same branching seen in dome stic mammals (King, 1977; Murie, 1973), seals have reduced jugular veins whose function has largely been replaced by the epidural and vertebrobasilar system. An exp ansive epidural venous sinus located dorsal and dorsolateral to the spinal cord is responsible for draining much of the blood from the brain (Barnett et al. 1958; Harrison & Tomlinson, 1956; Ronald et al. 1977). Detailed literature on cetacean venous morphology of the head is even more sparse than that found for pinnipeds (B oenninghaus, 1904; Costid is & Rommel, 2012; Fraser & Purves, 1960; Harrison & Tom linson, 1956; Hosokawa & Kamaya, 1965; Walmsley, 1938), however it suggests the presenc e of substantial modifications to parts
156 of the venous system. It should not be surprisi ng that the significant reorganization that accompanied telescoping of the skull to accommodate a dorsallyl ocated blowhole has resulted in reorganization of the corre sponding venous system. Additionally, specialized structures such as the acoustic fat bodies found in the melon and within the dentaries and complex accessory air sinus system also hint at suggestions of significant vascular modification (Costi dis & Rommel, 2012; Fraser & Purves, 1960). This does not even account for such extreme physiologic al adaptations as unihemispheric sleep and deep diving (Lyamin et al. 2008; Mukhametov, 1984; 1987; Ridgway, 2002). Nonetheless, the general venous branching patterns in the heads of cetaceans do indeed show homology to terrestrial mammals and can be traced with some certainty (Costidis & Rommel, 2012; Walmsley, 1938; McFarland et al. 1979; Ridgway et al. 1974). Materials and Methods The experimental procedure inv olved postmor tem latex injection and dissection of Florida manatees. All specimens were obtained postmortem, under a U.S. Fish and Wildlife Service permit (#MA067116-1), and all experimental procedures were conducted under University of Florida s IACUC (Permit #: 200801345). Florida manatees were used because the presence of a statewide stranding and rehabilitation network facilitates the avail ability of fresh specimens. Fr esh specimens were used in order to enable adequate flushing and inject ion of blood vessels of interest, since clotted blood in suboptimal specimens adver sely affects filling of vessels. Four manatee specimens were used for this study The experimental methodology used for the four specimens varied slightly due to im provement of the tec hnique over time as well as alternate experimental priorities and differing specimen types. All injections
157 were performed on specimens obtained from rescued animals that were euthanized due to untreatable medical conditions, as det ermined by the veterinary staff at the rehabilitation institutions. Following euthanasia, all specimens received a vascular flush using 0.9% phosphate buffered saline (PBS) so lution. The volume of PBS that was used varied based on specimen size, volume r equired to obtain clear effluent from the draining vessels, and degree of tissue edema during flush. Saline flush volumes ranged from 10L to 40L depending on the size of t he specimen and degree of blood clotting. Since the venous system was the primary target, saline flushes were always begun through the arterial system, in order to help force blood through the capillary beds and out of the veins. Once clear effluent drained from the veins, the flush was reversed in order to clear the arteries of any blood clot s. Following the PBS fl ush, specimens were allowed to drain for 2 to 4 hours, after which time one of two procedures was followed. Two of the specimens received vascular late x injections. The other two specimens received arterial perfusions of 18L of 4% neutral buffered formalin (NBF), in order to help preserve the tissues and prolong the ava ilable dissection time. Following the NBF perfusion and another 2 to 4 hour draining period, the sp ecimens were injected with latex mixtures. For angiographic imaging of the blood vesse ls in a computed tomography (CT) scanner, a mixture of liquid la tex (Carolina Biological, In c.) and 98% w/w barium sulfate suspension (Liquid Polibar Plus, Bracco Diag nostics) was injected into the vessels of interest, according to a modification of the protocol presented by Holliday et al. (2006). Latex injections varied slightly based on imaging goals (arter ial vs. venous) and specimen characteristics (head vs. head and crani al thorax). Fresh specimens received
158 venous injections of a 60:40 latex to barium sulfate suspension. Specimens that were perfused with NBF received 5% larger volume of barium sulfate suspension in the latex mixture. This increase in contrast agent wa s performed in order to increase the signal difference between the injected blood vessels and the soft tissues, because preserved tissues have greater radiopacity on CT, effe ctively reducing the differentiation between vessels and surrounding tissues. Three of the five specimens were imaged through a CT scanner, using a thin slice protocol. CT imaging was c onducted at the University of Florida College of Veterinary Medicines diagnostic radiology department and at the Baptist Mariners Hospital in Tavernier, FL. Axial slices were obtained from the specimens at 3mm thickness, with 1mm slice intervals. Whenever possible (spec imen size allowing) the volume data were reconstructed to 0.5mm slice thicknesses (alternatively 1.0mm). Since postmortem specimens were used, radiation exposure leve ls were not a concern. The resulting DICOM data was post-processed using Amir a (Visage Imaging, Inc.) software on a Gateway Precision T3500 with memory and processor enhancements. Post-processing was carried out in order to visualize t he data in 2D and 3D formats and gain an understanding of vessel locations and relationsh ips prior to gross dissection. Once three-dimensional images we re generated to guide the dissections, gross dissection was carried out on each specimen in order to validate and/or clarif y structures observed on CT. Findings on dissection were photo-documented. A fifth manatee specimen was flus hed with PBS and perfused with Mercox corrosion casting material (Ladd Research, Inc. ). Blue solution was injected in the veins and red solution in the arteries Following a 2hr curing submer sion in a cold water bath,
159 the specimen was gently lowered into a vat containing 15% KOH (Sigma-Aldrich, Inc). Only the rostrum was submerged initially, in order to corrode away the dense, heavy tissues of the snout. Following a week of rinses and resubmersions, the head was completely submerged in the vat with a fres h solution of 10% KOH. Numerous rinses and resubmersions were conducted over the per iod of three more weeks, until most of the tissue was corroded. The resulting spec imen was rinsed in a water bath for several days and then allowed to air dry. Results Before covering the venous system, I will give a brief d escription of the arterial system as it was investigated in one of my specimens. The reason for providing an arterial description is twofold. Firstly, it provides a reference fo r some of the venous structures. Secondly, interesting structur es were discovered in which arteriovenous juxtaposition is intimate and suggests that th ere may be an important physiologic role. Arteries of the Head and Neck Unlike some of the venous branc hing patterns described below, the aortic branching patterns showed more consistency with those seen in domestic mammals (Figures 4-1 & 4-2). The first major branch off of t he aortic arch is the right brachiocephalic artery which subsequently bifurcated into the right common carotid artery oriented rostrad and the right sublavian artery oriented rostrolaterad for a short distance before curving caudolaterad. From the right subclavian ar tery arise numerous branches including the right costocervical ar tery, the brachial artery supplying the brachial vascular bundle described by Mu rie (1874) and Fawcett (1942), and the large internal thoracic artery. The vertebral arte ry arises from the costocervical artery and supplies numerous segmental arteries to the epidural rete described below. A
160 significant size difference between the right and left vertebral arteries as described by Murie (1874) was not observed. The right common carotid travels a long distance in a relatively straight and horizontal path to the level of the occipital condyles (Figure 4-2), though this path is likely affected by head and neck posture. Along its course the right common carotid artery gives off innumerable small branches to the various arterial plexuses investing fascial planes between t he muscles of the proximal pectoral limbs and neck, as well as providing numerous th yroid branches to the thyroid gland, laryngeal branches, and glandular branches to the various lymph nodes of the neck (e.g. axillary lymph nodes, retropharyngeal lymph nodes, etc.). At the level of the gl enohumeral joint of the pectoral lim b, a large artery leaves the common carotid and invests the tissues medial and rostral to the scapula. Although this artery was not investigated extensively, it appeared to contain combined supply fields for the suband suprascapular arteries, potentia lly a result of the loss of a large singular axillary artery. A few centimeters rostral to the emergenc e of the scapular artery, the internal carotid artery emerges and travels at an oblique angle toward the tympanoperiotic earbone complex (Figure 4-2). Although Murie (1874) claimed that the internal and external carotid ar teries were similar in size, I found the internal carotid to be significantly smaller in diameter than the ex ternal carotid. So much so, that in my specimen the external carotid formed the cont inuation of the common carotid, while the internal carotid could be considered as a sm all tributary. It is possible that the developmental stage of Muries and my s pecimens differed enough to reflect such a difference, since Muries oldest of the two specimens was a female only a few months old, while my specimens were all obtained from adult animals. Such an ontogenetic
161 change may reflect a similar occurrence to that found in cetaceans and cows, in which the internal carotid artery is progressively substituted by alternat e blood supply to the brain. As the arterial component of the epidural rete of the manatee is quite voluminous, it appears as though it may be responsible for much of the adult blood supply to the brain (Chapter 5). Just ventral to the basisphenoid-basioccipital synchondrosis, the internal carotid artery bifurcates, sending a dorsocaudal and dorsorostral branch. The caudal branch was not followed in detail but appeared to connect to the cervical rete described by Murie (1874). T he rostral branch enters the brain case through the peribullar sinus and cranial hiatus and joins the arterial vasculature at the base of the brain. Following the emergence of the internal carotid artery, the common carotid artery continues rostrad as the external carotid artery (Figure 4-2). Just caudal to the angle of the mandible, the external caro tid artery gives off a sizable superficial temporal artery and became the maxillary artery. This is contrary to Mu ries (1874) description of a small temporal artery. The super ficial temporal artery travels dorsolaterally to invest the lateral aspect of the mandibular ramus as part of the masseteri c plexus described below. Once again the difference between Muries and my observations could be reflective of an ontogenetic difference in our specimens, since the superficial temporal artery invests the large masseteric plexus a nd may therefore gradually develop in size as the masseter muscle is increasingly recr uited with age. Shortly after the emergence of the superficial temporal arte ry, the maxillary artery divides into its vertical terminal continuation and a rostroventral branch (Figur e 4-2). The ventral branch gives off a rostromedial lingual artery to the tongue and travels to t he mandibular foramen as the
162 mandibular alveolar artery and branches into a vascular bundle as it enters the foramen, ultimately exiting through the mental foramina to supply the soft tissues of the lower lip and oral pad. The vertical branch eventually curves ro strad and enters the maxillary foramen as the infraorbital artery. Like the mandibular alveolar artery, the maxillary artery branches into a vascular bundle that enters the maxillary foramen and exits the infraorbital foramen to invest the tissues of the upper lip. Along its course to the infraorbital foramen, the infraor bital arterial bundle gives off a sizable ventrolateral spray of vessels that invests the muscles on the lateral aspect of the oral cavity (e.g. buccalis ). A dorsal spray of vessels also in vests the deep tissues of the temporal muscle ( A. temporalis profunda Schaller, 2007) and anastomoses laterally with tributaries of the superficial temporal arte ry and its masseteric plexus. Perhaps most notable is the absence of a detectable facial arte ry in any of my specimens (Figure 4-1). Although Murie (1874) did not specifically note the orig in of the facial artery, he implied that it emerged from the distal external carotid arte ry and wrapped around the facial notch of the mandible, in way similar to domestic mammals. Although only three of my five manatee specimens received arterial injections, they all appeared to have good arterial casts of much more distal arteries as well as of the facial veins, yet none exhibited even a hint of a facial artery. It is possible that the facial artery degenerates with age as chewing becomes more vigorous and anc illary arterial supply is established through the ventrolateral buccal branches of the infraorbital bundle and the various ramifications of the masseteric plexus. Alternatively, it ma y be that the facial artery contains a significant amount of smooth muscle in its wall and due to postmortem contracture failed to inject. Nonetheless, no such artery was ever appreciated grossly in
163 any of my specimens. Interestingly, DuBoul ay & Verity (1973) were not able to detect a facial artery in either the California sea lion ( Zalophus californianus ) or Baikal seal ( Pusa sibirica ) however, they did find a facial artery in the gray seal ( Halichoerus grypus ). Veins of the Head and Neck The following descriptions represent a summation of the branching patterns observed in all specimens, except where marked differences occurred. Unlik e the relatively simple venous branching patte rns observed in most domestic mammals, manatees have complex branching patterns t hat combine some of the traditional mammalian branches with more unique branchi ng arrangements. Mu ch of the venous branching in the head and neck of the manatee occurs in the form of tortuous, anastomosing structures. Although individual parent veins (e.g. external jugular and facial veins) are identifiable, many of the venous pathways involve vascular bundles and/or plexuses that anastomose freely, of ten making the anatomical identification and functional distinction of individual or specific drainage pathways challenging and perhaps meaningless. The small number of avail able specimens and lack of any literature on the venous morphology resulted in an incomplete descr iption of some of the venous anatomy. Although every attempt was made to detail a ll of the major veins of the head and neck, gaps still exist, especially with respect to the small tributaries of the nasopharyngeal regions, and salivary and lymphoid glands. T he veins of the nasal mucosa remained consistently poorly injected in all specim ens, suggesting a relatively high resistance system. Whether this resistance was due to i nherent structural features or postmortem vasoconstriction is unknown, however mo st specimens were flushed with phosphate
164 buffered saline solution containing potassium nitrate to relax vascular smooth muscle. Similarly due to specimen limitations, the mor phology of the cranial vena cava was not explored. Nonetheless, the morphology of the proximal brachiocephalic veins was obtained from two of the five specimens. Just cranial to the cranial vena cava, two large venous trunks emerge and travel rostrad, giving off branches with surprising bilateral symmetry (Figures 4-3 & 4-4). Ba sed on the destinations and paths of the various branches of these two trunks, they were named brachiocephalic trunks, since they eventually branch into ve ins that travel to the head, pectoral flippers, and thorax. Echols (1984) illustrations showed the pr esence of bilateral cranial venae cavae extending from the right atrium of the heart. Although I did not specifically investigate the cranial caval system, this morpholog y does not appear in any domestic mammal and based on my findings and those of Murie (1874) I suggest that the cranial vena cava be considered a single, roughly horiz ontal structure from which bilateral brachiocephalic trunks emer ge. From caudal to crani al, numerous branches emerge from the lateral aspects of each brachiocephalic trunk (Fi gure 4-3). The major lateral branches are the paired internal thoracic veins, the caudal brachial vein, brachial vascular bundle, and cranial brachial or cepha lic vein, respectively. It does not seem appropriate to name any portion of the trunks as the subclavian veins since the group of three venous structures to each pectoral limb branches separately off of the trunks that become the jugular veins. Within the thorax, numerous small veins from the brachiocephalic trunks invest the thymic and parietal pericardial tissues. Shortly after the brachiocephalic trunks emer ge from the thoracic inlet and send their branches to the
165 thorax and appendages, they give off numerous ramifica tions to the tracheobronchial lymph nodes and surrounding muscles (e.g. m. sternocephalicus, m. sternohyoideus). Jugular Veins The jugular veins branch early on from t he brachiocephalic trunks, in a manner resembling a cross between the pig and ox (Schummer et al. 1981). Approximately 3cm rostral to the apex of the aortic arch at the level of the third cervical vertebra and the emergence of the veins to the pecto ral flippers, each brachiocephalic trunk bifurcates, giving off a ventral and dorsal branch. The ventral branch is the rostral continuation of the brac hiocephalic trunk that invests t he structures of the face and was therefore identified as the ex ternal jugular vein, while th e dorsal branch was identified as the internal jugular vein based on its cour se to the jugular foramen (Figures 4-3 & 44). Since the internal jugul ar veins emerge at roughly the same level as the flipperassociated veins, it seems reasonable to c onsider them as direct branches of the brachiocephalic trunks, and therefore to thin k of the common jugular veins as absent. It should be noted that the internal and extern al jugular veins communicate freely with each other via numerous sizable anastomose s. Therefore, although they may be structurally distinct along much of their course, this distinction may be functionally meaningless. As they travel rostrad, all major jugular branches pass lateral to the stylohyoid bones before ramifyin g in the various tissues. Internal jugular vein The internal jugular veins are often of s ubstantial size. In some specimens the proximal portions of the in ternal jugular veins almost equaled the diameter of the external jugular veins --a fact supported by Muries findings. Each internal jugular vein travels cranially in a roughly horizontal m anner paralleling the common carotid artery
166 and external jugular vein on their dorsa l aspect, and in close proximity to the ventrolateral aspect of the cervical vertebrae. A medium-sized branch (~2mm diameter) emerges from the ventral aspect of the internal jugular and fu ses either with the external jugular vein or the peri-carotid venous plexus that it forms. The internal jugular vein then bifurcates into a large ventral branc h and smaller dorsal branch. The dorsal branch curves dorsorostrally and enters the jugular foramen. After entering the jugular foramen it travels through the peribullar sinus to join with the sigmoid dural sinus within the calvarium, a few centimet ers rostral to where the sigmoid dural sinuses receive large inputs from the neural canal (Figure 45). The ventral branch of each internal jugular vein curves ventrad for just a few centimeters to run adjac ent to the external carotid artery and then bifurcates again. At this bifurcation, two large branches of roughly equal size emerge, one heading dorsolate rad while the other curves ventrad. The ventral branch can fuse with the underlying maxillary vein but more typically joins a large ascending pharyngeal vein and the plexus surrounding t he external carotid artery (Figures 4-6, 4-8 & 4-9). The dorsal branc h extends obliquely rostrad and lateral to the digastricus muscle to fuse with the superficial tem poral vein that drains the masseteric plexus described below. Although in one of t he more complete specimens the ventral branch that anastomosed with the maxillary vein had the appearance of being a direct continuation of the internal jugular vein, before doing so it sent branches to the pharyngeal region in a manner similar to th e ascending pharyngeal vein observed in the other more complete specimen (see External Jugular vein descripti on). I therefore believe that the ventral anastomotic branch of the internal jugular with the external jugular is in fact the ascending pharyngeal vein, which connects caudally to the internal
167 jugular vein and rostrally to the superficial temporal vein. It appears as though the role of the ascending pharyngeal vein may be variable depending on the size of its anastomoses with the internal an d external jugular veins. At this level, this dorsal anastomosis bet ween the internal jugular and ascending pharyngeal veins forms a large alternate drai nage tract for the masseteric plexus. As this anastomosis approaches its fusion with t he masseteric plexus, it sends a few small veins dorsad to invest the region behind the external auditory meatus (EAM). Since these tributaries invest the tissues behind the EAM, it seems reasonable to consider them the equivalents of the caudal auricular veins of dome stic mammals. In domestic mammals, the caudal auricular veins are branches of the maxillary vein rather than the internal jugular or ascending pharyngeal vein s. This difference might be explained by the fact that the terminati on of the manatees caudal auricular veins was variable, sometimes being at the junction between the super ficial temporal veina branch of the maxillary vein--and the large rostral anasto mosis from the ascending pharyngeal vein, while in other specimens it was a direct branch of the superficial temporal vein. Therefore, determination of the parent vein of the caudal auricular veins in the manatee might depend on the degree of dev elopment of the terminal anastomosis of the internal jugular vein. Due to the countless anastomoses between these veins it seems unreasonable to definitively identif y a parent draining vein for the caudal auricular veins. The rostral anastomosis of the ascending pharyngeal vein with the superficial temporal vein can vary considerably in size--either due to injection artifact or natural variation. This manifests as an observed difference in formation of the superficial temporal vein. If the super ficial temporal vein should go lateral to the digastricus
168 muscle (like LPZ102904), then the vein labeled as superficial temporal in MSTm1001 is actually an anastomosis between the maxillary vein (medial to di gastricus) and the superficial temporal vein (lateral to digastr icus). In MSTm1001 (Fi gures 4-6 & 4-7), the large vein lateral to the digastricus muscl e appeared as a continuation of the internal jugular vein but was in fact the anastomo sis of the ascending pharyngeal with the superficial temporal vein, as seen in the other specimens. It should be noted that a rostral anastomosis of the superficial temporal vein with the ventral external ophthalmic vein ( r. anastomoticus cum plexus ophthalmico Schummer et al. 1981) that is found in carnivores was not observed in any of my specimens, either due to absence or artifact. Additionally, a frontal vein, which in the cat anastomoses with the caudal auricular vein and in carnivores also connects to the dorsal external ophthalmic vein, did not inject in any meaningful extent in any of my specimens and could therefore not be evaluated. However, due to the extensive filling of the ophthalmic vasculature and concomit ant lack of indication of any rostral anastomoses in all specimens, I suspect such a connection between the frontal and ophthalmic veins does not exist in the manateea condition ty pical of the pig, horse, and small ruminants. The medial branch of the inte rnal jugular vein that fo llows the external carotid artery was named the concomitan t or satellite vein of the external carotid artery ( vena comitans arteria carotidis externae, Schummer et al. 1981). Within the domestic mammals, this vein is only found in carnivor es and pigs. In the manatee this branch follows the external carotid ar tery to the level of its bi furcation into the mandibular alveolar and maxillary veins, forming along its course anastomoses with the maxillary
169 vein and the plexus that surrounds the exter nal carotid artery. At this point the concomitant vein of the external carotid artery contributes modestly to the facial vein, pterygoid plexus and maxillary vein. External jugular vein The external jugular veins begin at t he point where the br achiocephalic trunks branch into internal and external jugular ve ins just rostral to the emergence of the cephalic vein. After only about a centimeter or so, numerous (0.5-5mm diameter) veins emerge from the ventromedial as pect of the external jugular vein and extend mediad for 1 to 5cm (Figures 4-3 & 4-4). Some of these veins break up into small retia that invest surrounding muscles (e.g. m. sternohyoideus ), while the larger ones take more direct routes. Some of these la rger branches anastomose back and forth with each other as they wrap around the common carotid artery, forming a plexus that su rrounds the artery. One of the medial branches of the external jugular vein arches rostrally and extends on a direct route to the level of the angle of the mandible, ju st caudal to the facial vessel notch of the mandible ( incisura vasorum facialum ) (Schaller, 2007). At this point it bifurcates, sending a rostral branch that fu ses with the lingual and mandibular alveolar vessels and a medial branch that forms an ar ch across the midline to connect to the equivalent contralateral vein (Figure 4-4) (for more detail see hyoid arch description below). At approximately the level of the atlanto-occipital joint, the external jugular vein bifurcates into two large branches. T he ventral branch travels to the face and sometimes the tongue as either th e facial or linguofacial vein ( v. linguofacialis Schaller, 2007) described below (Figures 4-3, 4-11 & 4-12). The dorsal branch forms the maxillary vein as it is the terminal continuatio n of the external jugular vein that invests
170 the tissues of the pharynx, deep mandible, nasal passages, palate, deep nasal tissues and maxillary teeth (Figur e 4-6). Just before or at the level of the bifurcation of the external jugular vein into the maxillary and either facial or linguofacial veins, a vein between 3-6mm in diameter emerges from the dorsal aspect. The vein immediately divides into veins that invest the occipital region (v. occipitalis), the stylopharyngeus muscle ( r. stylopharyngeus ), and the pharyngeal region at the level of the laryngeal cartilages ( v. pharyngea ascendens ) (Figures 4-6 & 4-8). Although in bovines the ascending pharyngeal vein is considered a branc h of the occipital vein, I considered the ascending pharyngeal vein of the manatee as the parent vein because in all my specimens its rostral continuation to the pharynx was considerab ly larger than the continuation of the occipital vein. In additi on to sending numerous twigs to the wall of the pharynx, the ascending ph aryngeal vein anastomoses--via its main terminus--with the maxillary vein and associated peri-caroti d venous plexus. Tributaries of this anastomosis sometimes extend to the lingual region and anastomose with some of the veins forming the lingual plexus at the base of the tongue. As is common in domestic mammal specie s, branches of the linguofacial veins invest the lingual and facial tissues, however in the manatee the lingual and facial veins did not always emerge from the ex ternal jugular vein as a comm on trunk. In fact, in only one of the five specimens I examined was t here a common linguofacial trunk. In that specimen, each linguofacial ve in branched into two predomi nant veins that paralleled the maxillary vein and external carotid artery up to the level of the laryngeal cartilages. At that point the lar ger of the two branches curved slight ly mediad to follow the ventral margin of the insertion of the internal pt erygoid muscle along the pterygoid fossa of the
171 mandible. Shortly thereafter the vein curved around the mandibular notch for the facial vessels (incisura vasorum facialum Schaller 2007; Schummer et al. 1981) and emerged on the lateral aspect of the ventral margin of the body of the mandible ( corpus mandibulae ). In none of my specimens was th is branch found with an accompanying facial artery described by Murie (1874), how ever it was consistent in location and drainage field with the facial vein of domestic mammals ( v. facialis, Schaller, 2007; Schummer et al. 1981). After its emergence from t he linguofacial vein, the smaller of the two branches was typically located just m edial to the facial ve in and adjacent to the ventral aspect of the external carotid artery As this vein approached the base of the tongue it began to take on a more plexiform appearance, receiving numerous anastomoses from adjacent veins (e.g. mandi bular vascular bundle, hyoid arch, facial, maxillary). Eventually the lingual vein ( v. lingualis Schaller, 2007; Schummer et al. 1981) invested the muscles of the tongue (e.g. m. lingualis) as the lingual plexus. Distally the lingual plexus divided into tw o separate clusters of veins, one cluster investing the center of the tongue as the deep lingual vein (or deep lingual plexus), the other coursing ventrally and emerging vent rolaterally between the genioglossus muscle and dentary of the mandible as the sublingual vein (or sublingual plexus). At the level of the angle of the mandible, the lingual plexus received coun tless anastomoses from the mandibular alveolar veins and their resulting mandibular alveolar bundle. These anastomoses often times formed a greater contribution to lingual drainage than the lingual vein itself (see below for further detail). Facial Veins Throughout its course, the facial vein has numerous anastomoses with the external jugular vein as well as other smaller tributaries of the external jugular vein,
172 lingual vein and hyoid arch, a pattern s een occasionally even in humans (Gupta et al. 2003). As each facial vein courses around from the medial to the lateral aspect of the corresponding angle of the mandible, it gives off a sizable branch that extends in a roughly horizontal path toward the soft tissues of the lower lip (Figure 4-11 & 4-12). This vein subdivides extensively in the soft tissues of the lower lip, consistent with the mandibular labial vein ( vena labialis mandibularis Schummer et al. 1981). The distal extremity of the mandi bular labial vein sends medial twigs that anastomose with the mental veins, the terminal branches of the mandibular alveolar vascular bundle (Chapter 5). This pattern of connection is consistent with that seen in domestic mammals (Ghoshal et al 1981). The facial vein then continues dorsad for a short distance and near the angle of the mouth forms a dilated sinus-like structure that receives contributions from the pterygoid plexus and deep facial vein on the medial aspect of the dentary, and from the numerous veins of the medial and lateral aspect of the masseter muscle (e.g. ventral massete ric, transverse facial, deep masseteric, masseteric plexus) (Figures 4-4 and 4-5). Given the loca tion and contributing veins, I called this sinus-like dilatation of the facial vein the deep facial venous sinus (Figure 411), not to be confused with t he plexus of the deep facial vein described below or the sinus of the deep facial vein found in the horse (Schummer et al 1981). While the general patterns of the facial veins appeared similar between individual manatees, the exact positioning of the facial veins as they c ourses through the region of the salivary gland varies. When viewed from the ventral aspect, a roughly 5cm portion of the left facial vein of MSTm1001 was hi dden behind the ventrom edial portion of the mandibular salivary gland while the entire ri ght facial vein was exposed and separate
173 from the ventromedial surf ace of the salivary gland by approximately 2cm. In LPZ102900, there was signifi cant heterogeneity between the paths of the two facial veins. At the level of the glottal cartilages both facial veins traveled ventromediad away from the external jugular vein from which they sourced. As the left facial vein approached the large salivary gl and it traveled along the vent rolateral border of the gland before crossing the ventral part of the left dentary at an oblique angle. Conversely, although the right facial vein also traveled on a ventromedial path from its origin at the external jugular vein, it maintained a slightly deeper course that forced it to cross the dorsomedial surfac e of the salivary gland. T herefore, on a ventral approach with both salivary glands exposed, the entirety of the left facial ve in was visible while only the distal portions of the right facial vein were visible as it emerged around the dentary, the rest being hidden by the salivary gland. The very large salivary gland of manatees was therefore thought to exert considerable infl uence on the location of the various jugular branches. Lingual Veins and Hyoid Branches Although the venous invest ment of the tongue was fa irly consistent between specimens, its primary and sec ondary drainage routes were o ften less discer nible. In one specimen the lingual veins coalesced into a common linguofacial trunk, a branch of the external jugular vein that is shared with the facial vein, while in the other specimens multiple lingual veins emptied directly into the external jugular vein caudal to the termination of the facial vein or into the ma xillary vein rostral to the facial veins terminus. In the specimen with a common lin guofacial trunk, the lingual and facial veins took a separate path shortly after their branching, although they followed each other closely until the facial vein coursed around the mandible to the lateral aspect of the
174 dentary. Along their course from the tongue, the lingual veins formed numerous anastomoses with surrounding veins, including the mandibular alveolar, facial, thyroid, and maxillary veins. The most prominent structure emanating from the rostral part of the lingual veins was a venous bridge between the left and righ t lingual veins, which formed the hyoid arch (Figure 4-4). Like in the dog, the hyoid arch in the manatee is a fairly prominent structure located just rost ral to the laryngeal cartilages and dorsal to the median mandibular lymph node, at the level of the rostral margin of the basihyal bone (Schummer et al. 1981). In some specimens the majo r contribution to the hyoid archs formation was made from medial tributaries of the lingual vein s, and in these cases the arch may be considered as formed by an anastomosis between the lingual veins as seen in dogs, pigs, and ruminants (Schummer et al. 1981). In two of the specimens the lingual contributions to the hyoid arch were modest, comprised of only a few very small venous twigs. In those spec imens, the primary contributi on to the arch was made by fairly sizable veins draining directly into the external jugular veins and the maxillary veins and traveling juxtaposed to the external carotid artery and in parallel to the trachea. Although these could be elaborations of the caudal an d cranial thyroid veins, I could not with certitude find an equivalent vein in domestic mammals. This leaves me to believe that they are either novel anastomoses or significantly enlarged and elaborated veins that normally have a minimal role or more caudal drainage field in domestic species. In two of the five specimens, the lingual veins were less identifiable as they were multiple small branches of t he external jugular veins or maxillary veins and almost
175 immediately fused with the max illary plexus that surrounds th e external caro tid artery. Their contributions to the veins of the tongue were modest and heavily supplemented by anastomoses from the mandibular alveolar ve ins and vascular bundle, features similar in some ways to the domestic cat. It should be noted that at the leve l of the hyoid arch, medial anastomoses from the facial veins were more comm on than not. These sizable anastomoses forms large collateral drainage paths for the lingual blood, often times equaling or dominating the lingu al veins contribution. In addition to the lingual vein contri bution to the hyoid arch, two manatee specimens had an equal or lar ger contribution made by a pair of long, straight veins draining into the external jugular veins just caudal to the emergence of the facial veins and traveling along the ventral aspect of the sternohyoideus muscles. These long straight veins course along the ventrolatera l aspect of the trachea, consistent with the caudal thyroid veins described in domestic mammals. At the level of the caudal border of the mandibular ramus, a vein from each si de traveled from the li ngual vein medially at a transverse angle to invest the thyroid glan d lobes, consistent with the cranial thyroid veins. Unlike domestic mammals, the cranial thyroid veins appeared to originate from the lingual veins rather than the internal jugular veins. This modification may be related to the fact that the internal jugular veins course along a more dorsal path across the short neck. No midline anastomosis between the cranial thyroid veins was observed, and therefore unless the caliber of the veins is exceedingly small, I have considered the caudal laryngeal arch as being absent in the m anatee. It is possible that my inability to identify a large caudal thyroid c ontribution to the hyoid arch in two of the specimens was
176 due to the fact that they were decapitat ed specimens and theref ore were missing the cervical portions of the jugular branches in their entirety. A network of small veins enters the caudal border of the thyroid gland, along the region of the isthmus of the two lobes. These veins appeared to connect to numerous larger veins in the region of the aortic arch, including the brachiocephalic trunks, however their numerous ramifications could not be followed. Maxillary Veins The external jugular veins becom e the ma xillary veins followi ng the emergence of the large facial vein, which happens shortl y after the emergence of the axillary and cephalic veins (Figures 4-3, 4-6 & 4-8). Aside from the anastomoses received from the ascending pharyngeal and internal jugular veins, the first large branch of the maxillary vein is usually the superficial temporal vein ( v. temporalis superficialis, Schummer et al. 1981) that takes a dorsolateral path to em erge on the lateral aspect of the dentary (Figure 4-6, 4-8 & 4-9). After passing bet ween the digastricus and caudal border of the mandibular ramus and emerging laterally, it is always joined by a large (sometimes larger) anastomotic vein that travels along the lateral aspect of the digastricus and usually sources from anastomoses betw een the internal jugular and ascending pharyngeal veins. Together t he superficial temporal and anastomotic branches form the majority of the drainage path for the masseteric plexus (Figure 4-9 & 4-10). The morphology of the superficial temporal vein was found to be quite variable. In some specimens the superficial temporal vein was formed quite ventrally by a paired structure that embraced the homonymous artery on its course to the temporal region and masseteric plexus. In other specimens the proximal connection of the superficial
177 temporal vein was less obvious, receivi ng numerous anastomoses from the maxillary vein, ascending pharyngeal vein, and rostral anas tomosis from the internal jugular vein. After the emergence of the superficial te mporal veins, at the point where the maxillary vein curves dorsad on a vertical path, it sends a few large anastomosing veins from its ventral region that travels to the mandibular foramen as the mandibular alveolar veins ( vena alveolaris mandibularis Schummer et al. 1981). Approximately 2cm before they enter the mandibular foramen, the mand ibular alveolar veins subdivide into many small caliber veins forming the mandibular alveolar vascular bundle (Figure 4-6) that enters the mandibular canal and emerges from the mental foramina. Following the emergence of the mandibular alve olar veins, the maxillary vein obtains its vertical orientation and sends numerous small branches to the pterygoid region to form the pterygoid plexus. The maxillary vein t hen curves rostrad on a roughly horizontal path toward the maxillary foramen, where it becom es the infraorbital vascular bundle. In the region of the curvature from ve rtical to horizontal orientati on, two to three sizable vein emerge from the dorsal aspect and immedi ately invest the deep aspects of the temporalis muscle, consistent with the deep temporal vein of domestic mammals ( v. temporalis profunda Schummer et al. 1981). Shortly after the emergence of the deep temporal veins, two separate clusters of ve ins join from the rostral and caudal aspects of the external pterygoid muscle near its origin on the pterygoid process of the alisphenoid (Domning, 1978). As described below, these clusters fuse to form the pterygoid sling As in domestic mammals, the maxillary ve ins form the main terminal branches of the external jugular veinsthe other term inal branches being the lingual and facial
178 veins. What is unusual, however, is that in the manatee the maxillary veins terminate into the infraorbital venous bundles that enter the maxillary foramina, travel through the infraorbital canals with the infraorbital ar teries and nerves and exit the infraorbital foramina rostrally. Distinct infraorbital ve ins were not found in any of the manatee specimens. As soon as or just proximal to the maxillary veins entrance into the infraorbital canals they break up into countless small caliber parallel veins that form a vascular bundle like those described by Fawcett (1942), Murie (1872), and Rommel and Caplan (2003). All domestic mammals studied thus far lack an infraorbital vascular bundle, but instead have a single vein and artery that travel through the infraorbital canal. Additionally, the infraorbital veins of domestic mammals are tributaries of either the deep facial veins (e.g. dog, horse, cow) or the pterygoid pl exus and ophthalmic plexus (e.g. cats) (Schummer et al. 1981; Ghoshal et al. 1981). Another difference is that the maxillary veins and infraorbital venous bundl es of the manatee intimately surround their arterial counterparts. Conv ersely, in none of the domestic mammals studied does the beginning of the infraorbital vein follow the infraorbital artery, likely due to their different origins fr om the deep facial vein and maxi llary artery, respectively (Schummer et al. 1981). Interestingly, although t he infraorbital venous bundle appears to be a continuation of the maxillary vein, a sizable contribution is made to the bundle just caudal to its entrance into the maxillary foramen. This contri bution is from the deep facial vein--morphology more co nsistent with domestic mammals. Before the maxillary veins reach the maxillary foramina, they send a ventrorostrally-angled bundle-like mass of small veins that fan out on the lateral surface of the buccalis muscle. This mass of veins sends twigs into the muscle it covers and
179 also continues ventrad to anastomose variabl y with the various rostral branches of the facial vein (e.g. angularis oris, maxillary labial, etc. ). Once the maxillary veins break up into many linear, small caliber veins that enter the maxillary foramina as the infraorbital venous bundles, they travel through the infraor bital canal and exit at the infraorbital foramen, just ventroro stral to the orbit (Figures 4-9 to 4-11). Along their course through the infraorbital canal, the infraorbital bundl es send numerous ventral twigs to the maxillary teeth. Additionally, at a minimum of two locations the infraorbital bundle sends dorsomedial branches that suppl y a sizable nasal mucosal plex us. This plexus invests the mucosa of the main bony nares as well as the mucosa covering the modest scrolls of the ethmoturbinate bones. As they exit the infraorbital foramina, the veins on the periphery of the infraorbital v enous bundle begin to gently fan out into the tissues of the upper lip, in some instances anastomosing with branches of the maxill ary labial veins. Pterygoid, Deep Facial and Buccal Veins After a rather short but to rtuous course just medial to the caudal r amus of the mandible, the maxillary vein splits into pl exiform branches oriented ventromedially (mandibular alveolar & lingual) and branches oriented dorsally. The ventromedial branches become the mandibular alveolar veins that for t he mandibular alveolar venous bundle. The dorsal branches eventually se parate into two large clusters; the caudolateral one forming the superficial tempor al veins while the more rostral one is the continuation of the ma xillary vein. It is from t he ventromedial branch (mandibular alveolar) and dorsorostral branch (maxillary vein ) that the pterygoid veins emerge. No single pterygoid vein was found in any of the specimens. In stead, numerous small veins emerge from various parts of the maxi llary and mandibular alveolar veins, forming a delicate plexus that invests the pterygoid region and ra mifies in the internal and
180 external pterygoid muscles. All of these vein s invest the internal and external pterygoid muscles at various places. The majority of the veins investing t he internal pterygoid muscle emerge from the rostral aspect of the maxillary vein along its vertical section near the emergence of the superficial te mporal vein, and from the mass of anastomosing veins of the mandibular alveolar and lingual veins. Conversely, the veins investing the external pterygoid muscle em erge predominantly from the more distal vertical and proximal horizontal segments of the maxillary vein. Following the emergence of the various veins, the proximal horizontal portion of the maxillary vein gives off a rostroventrally oriented plexus of large, sinusoid veins. These sinusoid veins form the aforementioned pterygoid sling that wraps rostrally around the belly of the external pterygoid muscle near its insertion on the mandibular ramus, and heads dorsad again, sending br anches back into the maxillary vein and laterally to the temporalis muscle where they anastomose with tributaries of the deep temporal vein. Given the intimate associ ation of the venous sling with the external pterygoid muscle, I have called it the pterygoid sling to distinguish it from the pterygoid plexus composed of much finer widely distri buted veins. On its rostral aspect, just before it merges back with the maxillary vein, the pterygoid sling sends numerous small and/or large veins rostrad to fuse with a large plexus that covers the rostral aspect of the mandibular ramus and the rost ral and medial aspects of the temporalis muscle. This large plexus is a product of the deep facial vein that originates from the facial vein on the lateral aspect of the angle of the mouth, and was therefore na med the plexus of the deep facial vein. Such a plexus can al so be found in cows, though its location is not identical (Schummer et al. 1981). At the medial surface of the angle of the mouth, the
181 plexus of the deep facial vein coalesces to form a single (sometimes multiple) large vein that emerges on the lateral surf ace of the angle of the mouth as the deep facial vein (v. profunda faciei, Schaller 2007). The deep facial vein travels a short (0.5-2cm) distance rostrad where it fuses with the facial vein in the region of the deep facial sinus described above. An interesting finding was that the distribution pattern of the deep facial vein was unlike that seen in horses and cows. Although the origin of the deep fa cial vein from the facial vein is consistent with that seen in ung ulates, the rest of it s course deviated from that expected from domestic mammals. This difference is likely due to the shape of the mandible and its position relative to the skull. The cow and horse both have fairly elongated skulls and dentaries, resulting in an elongated buccal region. This morphology results in the formation of long deep facial and buccal veins, especially in the horse (Schummer et al. 1981). Conversely, the manatee has a relatively short, stout skull and mandible that may, at least partially, be the driving force behind the different morphology of those veins. Since the manatee lacks a facial crest and possesses a relatively short zygomatic proce ss, the extent of t he masseter muscle is comparatively shorter. This means that when the facial vein wraps around the facial notch to the lateral side of t he mandible, it does so at a fair ly sharp angle that results in a short deep facial vein. Interestingly, a singular buccal vein was not found in any of the manatee specimens examined. Unlike the horse, whic h has a long buccal vein originating from the maxillary vein and extending along the buccalis muscle to the rost rum, the manatee seems to satisfy the dr ainage requirements of the buccalis through other means.
182 Numerous small veins are found arising from the pterygoid plexus and plexus of the deep facial vein and extending rostrad to the caudal buccal region, but these veins are small in caliber and appear insign ificant when compared to the buccal vein of the horse. This morphology may be more like that of the cow, in which the buccal vein is a vein of modest size that originates from the deep facial vein. Gi ven the consistency of this morphology between specimens, I am led to belie ve it is not artifactual but rather the normal character of the buccal vein. I could think of only two explanations for such morphology. Firstly, because of the relati vely short skull, the buccal region of the manatee may not require as prominent a dr ainage field as that found in the horse. Secondly, drainage of the buccal field may be supplemented by alternate venous channels. Indeed, a sizable venous bundle is found to emerge from the ventrolateral aspect of the maxillary vein as it transiti ons into the infraorbital vascular bundle just before it enters t he maxillary foramen. This venous structure, composed of numerous small caliber veins that emerge as a cluster and eventually fan out, spreads out over much of the lateral surface of the buccalis muscle. Although many of these veins anastomose ventrally with veins traveling to the lips, they also send numerous twigs into the buccalis Such an arrangement likely provides a sizable collateral drainage path to the modest buccal veins and may have resulted in diminution of the buccal veins. Rostrally, the pterygoid plexus gives o ff numerous small and large branches that invest the tissues of the soft and hard palat e, including the caudal portions of the maxillary tooth capsule. In the region of and just lateral to the connections to the infraorbital bundle, the pterygoid sling sends dorsolateral branches that invest the temporal muscle and anastomose with the tri butaries of the deep temporal vein. The
183 pterygoid plexus is in fact continuous dorsa lly with the maxillary vein just before it becomes the infraorbital bundle and ventrall y with the mandibular alveolar bundle, separated from those two struct ures only by the presence of the internal and external pterygoid muscles. Facial Venous Structures Ventral masseteric vein A rather sizable vein (~2-3mm di ameter) emerges from the exte rnal jugular vein either in conjunction with or in close proximity to the emergence of the facial vein. This trunk travels rostrolaterad in a direct path to the lateral surface of the masseter muscle. Once at the masseter muscle it travels alon g the ventral half of the lateral aspect, sending numerous branches dorsad to anastomose with the masseteric plexus and transverse facial vein (Figure 4-9 & 4-10). Based on location and branching, this vein was considered to be the ventral masseteric vein ( v. masseterica ventralis Schaller 2007) seen in ruminants and horses (Schummer et al. 1981) in which it branches off of the maxillary vein. In the manatee, the mass eteric portion of the ventral masseteric vein is especially reminiscent of t hat seen in the horse, for a coupl e of reasons. Firstly, the dorsal offshoots that invest the lateral surface of the masseter muscle are very similar in location, orientation, and connection to those of the horse. Secondl y, as in the horse, the ventral masseteric vein fuses with the facial vein distally, just below the anastomosis with the buccal vein. At the angle of the mouth where the deep facial vein dives mediad, at least two large caliber veins connect either to the deep facial vein or the deep facial sinus. These veins travel in a tortuous but roughly horiz ontal path across the medial surface of the masseter muscle, in cover of th e large masseteric plexus (Figur e 4-10). I could not find
184 any equivalent veins in domestic mammals that resembled such a pattern and, therefore, given their location and variable presence I called them the deep masseteric veins. These deep masseteric veins often ti mes divide into smaller veins that merge with the masseteric plexus in the middle of the masseter muscle. Other times however the deep masseteric veins extend caudad th roughout the entirety of the masseteric plexus, to fuse with the transve rse facial vein and/or its parent superficial temporal vein. The exact location and confluence pattern of the facial vein and its tributaries as it wraps around the rostral aspect of the ma sseter muscle is quite variable between and within individual manatees. As the facial vein curves dorsad around the ventrorostral aspect of the masseter muscle, in some specimens it traveled dorsad in a rostral fashion similar to the cow, while in other specimens or on the cont ralateral side of the same specimen it angled dorsad or dorsocaudad in a similar manner to the horse. The dorsad and dorsocaudad angled morphology was usually associated with a smaller continuation of the facial vein toward the ro strum, the larger cont ribution being made by the confluent veins (e.g. deep facial, transvers e facial, ophthalmic anastomosis, etc.). In specimens with a vertical orientation, t he facial vein continued toward the rostrum unaffected in caliber, receiving along the way the same common tributaries. In addition to the facial and transverse facial veins, numerous other superficial and deep structures occupy the lateral and medial tissues of the face and cheeks (Figures 4-4 to 4-6). Lateral superficial branch 1 (transverse facial vein) The largest of the superficial veins on t he lateral aspect of the masseter muscl e most closely follows the facial nerve and is consistent with a transverse facial vein of the ox (vena transversa faciei Schummer et al. 1981). This vein drains into the superficial temporal vein ( vena temporalis superficialis, Schummer et al. 1981), a short large trunk
185 that branches off of the maxillary vein, ju st caudomedial to the ramus of the mandible (see External Jugular Vein description above). The transvers e facial vein can occur as either a singular vein or as a parent vein with multiple small venous branches that ramify into the masseteric plex us (Figures 4-8 & 4-9). I i dentified the bran ches of the transverse facial vein as dorsal masseteric veins such as those seen in the horse ( v. masseterica dorsalis Schummer et al. 1981). In those specim ens that had a modest or poorly injected transverse facial vein, the countless ramifications from the surrounding masseteric plexus that entered the transvers e facial vein often times obscured its identification as a distinct vein. At approximately its distal third, the transverse facial vein of one of the specimens gave off a small vein that emerged from its dorsal border and traveled obliquely toward the orbit, where it invested the lateral palpebral tissues. Since manatees have a sphincter-like palpebr al structure, the tr aditional distinction between medial and lateral canthus is problem atic. Nonetheless, the aforementioned small dorsal tributary of the transverse facial vein inve sted the caudal half of the dorsal and ventral palpebral sphincter, consistent with the parent vessel of the lateral dorsal and lateral ventral palpebral veins of the horse ( vena palpebralis superior lateralis and inferior lateralis, Schummer et al. 1981). Near its rostral fusion with the facial vein at the deep facial venous sinus, numerous other veins from the medial and lateral aspects of the mandibular ramus join the transverse facial vein. As will be explained later, these veins are consistent with the deep facial, v entral masseteric, and facial veins seen in domestic mammals. Lateral superficial branch 2 (ventral masseteric vein) The second largest single vein on the la teral surface of the masseter muscle i s located ventral to the transverse facial ve in and follows a similar longitudinal but
186 ventrally curved path across the masseter muscle to fuse directly with the facial vein just ventral to the deep facial venous sinus, or at the deep facial sinus itself. Before this vein merges with the facial vein or deep faci al sinus, it sends dorsad at least 3 short anastomotic branches to the transverse faci al vein in addition to numerous small connections to the masseteric pl exus (Figure 4-10). This ve in is consistent with the ventral masseteric vein of the ox and hor se, although its distribution on the masseter muscle is more similar to the horse (vena masseterica ventralis Schummer et al. 1981). Lateral superficial branch 3 (masseteric venous plexus) The third vein is in fact not a single ve in but rather an extensive venous plexus composed of countless anastomosing, small caliber veins (Figures 4-9 & 4-10). Near its drainage into the superficial temporal vein, the plexus appears to split into a few distinct clusters or plexus branches. From dorsal to ventral, th e first pl exus is angled almost vertically, traveling along the caudal bor der of the masseter muscle and mandibular ramus to terminate at the c audal margin of the stem of the squamosal bone. The second branch invests the dorsal and mid regi ons of the lateral masseteric surfaces, and extends medially into the depths of the masseter muscle. This medial extension forms a plexus of equal or greater size that invests the fasci a between the medial surface of the masseter muscle and the lateral surface of the mandibular ramus. Dorsal portions of the plexus travel to the te mporomandibular joint and dorsally to the temporalis muscle to anastomose with the deep temporal vein. The third plexus branch takes a more ventrorostral path eventually anastomosing with the ventral masseteric vein. All of the aforementioned branches of the plexus send numerous twigs into the body of the masseter muscle and therefore presumably play a role in draining the
187 muscle itself. Additionally, numerous rostral anastomoses occur with the facial vein and deep facial sinus. Lateral deep branch (deep masseteric veins) As mentioned above, the lateral masseteric plexus als o deeply invests the body of the masseter muscle itself, forming an equal or larger plexus on the medial aspect of the muscle, between the muscle and the ramus of the mandible. Alon g the rostral half of the masseter muscle, many of the ra mifications of the deep masseteric plexus coalesce into a few large veins (Figure 4-10) These veins--typically varying in number between 1 and 4--coalesce at or near the rostral border of the angle of the mandible, just caudal to the caudal-most teeth. Ju st ventral to the attachment of the zygomaticomandibularis muscle (D omning, 1978), the veins join the facial and/or deep facial vein. Medial deep branch (pterygoi d veins & deep facial vein) Medial to the mandibular ramus is a diffu se pterygoid plexus that invests the internal and external pterygoid muscles. The pterygoid ple xus does not appear to originate from a single pterygoid vein but rather receives numerous small pterygoid veins from the maxillary vein caudally and dorsally, and from the mandibular alveolar and lingual veins ventrally. The maxillary ve in contributions are made as a spray of small veins that emanate at various angles fr om the anastomosing cluster of large veins that form the vertical and horiz ontal portions of the maxillary vein before it becomes the infraorbital vascular bundle. The pterygoid pl exus sends clusters of veins rostrad along the medial margin of the ventral half of the internal pterygoid muscle where they appear to terminate, and dorsorostrally to invest t he external pterygoid and dorsal aspects of the internal pterygoid muscles.
188 Rostrad, near the angle of t he mouth the delicate pterygoi d plexus veins coalesce into a few small veins which extend rostrad to contribute to the fo rmation of either a single or a few slightly larger veins that join the plexus of the deep facial vein before it emerges laterally through the rostral angle of the mandible as the deep facial vein (Figure 4-10). Veins Of The Lips And Eyes After the deep facial s inus receives caudal c ontributions from the veins of the face (facial, transverse facial, deep facial, deep masseteric, ventra l masseteric, and masseteric plexus) at the rost ral margin of the lateral tendon of origin of the masseter muscle, it then trifurcates. Two large branches emerge dorsally and a third smaller branch travels dorsorostrally at an oblique an gle toward the maxillary labial tissues. The branches consist of the dorsomedially or iented deep facial vein, the dorsorostrally oriented continuation of the facial vein and the angularis oris vein (Figures 4-10 & 4-11). From caudal to rostral the branches were as follows: Branch 1 (deep facial vein & anastomot ic branch of the ventral external ophthalmic vein) The most caudal of the three branches forms the deep facial vein ( vena profunda faciei, Schaller 2007) that emerges in either a dorsomedial or horizontal medial fashion from the deep facial sinus of the facial vein, often times hid den lat erally from view by the jugal bone of the zygomatic arch and/or the rostral margin of the masseter muscle. At the angle of the mouth, the deep facial vein travels to t he medial side of the mandibular ramus to form the deep facial pl exus described below (Figure 4-10). After leaving the facial vein and before diving to the medial side of the mandi ble, the deep facial vein sends a large dorsal branch. This branch tr avels vertically for approximately 3-5cm and
189 then curves sharply rostrad to enter the orbit along the caudal aspect of the floor. Once in the orbit it fuses with t he ventral portion of a large opht halmic plexus. This dorsal branch of the deep facial vein was theref ore called the anastomotic branch of the ventral external ophthalmic vein ( ramus anastomoticus cum vena ophthalmica external ventrale, Schummer et al. 1981). Such an anastomosis between the deep facial and ventral external ophthalmic vein is common in carnivores. Additionally, in horses and pigs the deep facial vein forms the primary drainage route for the ophthalmic plexus or sinus (Schummer et al. 1981). This appears to be partial ly true in manatees, in which the anastomotic branch of the deep facial vein forms a drainage path of equal or greater caliber than the ophthalmic veins. Interest ingly, with the except ion of dogs, the deep facial veins of most domestic mammals extend dorsocaudal ly from the anastomosis with the facial vein to the anastomosis with t he ventral external opht halmic vein. In the manatee and at least some dogs, the deep facial vein extends dorsorostrally toward the orbit. This difference is likel y due to the architectural diffe rences of the skulls, in which the manatees orbit is positioned significantly more rostral than the lesser curvature of the angle of the jaw where the deep fa cial vein usually originates. Branch 2 (facial vein) The second most caudal of the three branc hes is the functional continuation of the facial vein. This vein emerges in a vertical fashion from the deep facial sinus but in line with the proximal facial vein. After conti nuing through the deep facial sinus it extends dorsad for only a short distance before turning to a more oblique dorsorostral path and then a roughly horizont al path as it follows the ventral aspect of the jugal bone and the zygomatic process of the maxi lla. Near the antorbital no tch, at the level of the infraorbital foramen, the faci al vein bifurcates. One tr ibutary continues rostrad on a
190 roughly horizontal path along the lateral aspect of the maxillary and premaxillary bones toward the maxillary lips to bec ome the maxillary labial vein ( Vv. labialis maxillaris Schummer et al. 1981) that breaks up into numerous tr ibutaries that invest the tissues of the upper lip (Figures 4-11 to 4-14). The other tributary and functional continuation of the facial vein wraps dorsocaudad around the cranial part of the bony orbit and toward the nasal and frontal bones. As they reach the lips, the maxillary l abial veins form many anastomoses between themselves as well as with the infraorbital vascular bundle described in the section on the maxillary vein. As these anastomotic branches extend rostrad they invest the superficial and deep tissues of the maxillary labi a to form a maxillary labial plexus. Due to the multiple anastomosing branches and appar ent lack of distinct or consistent branching, distinction betw een superficial and deep maxill ary labial veins seems meaningless. In place of the multiple super ficial and deep labial branches originating directly off of the facial vein as in the ox, it appears that in manatees venous drainage of the maxillary labia is accomplished primarily though a single parent maxillary labial vein, the lateral nasal vein, and the infraorbital vascular bundle on either side of the maxillae (Figures 4-11 to 4-14). As the facial vein wraps around the rostra l part of the orbit, it sends one sizable branch rostrally from the apex of its rostral curve, and one or more smaller dorsorostral and dorsomedial branches from its dorsal aspec t before it continues caudally along the medial aspect of the orbit. The rostral branch takes a slig htly more dorsal path than the maxillary labial vein, but soon af ter its emergence it parallels t he labial vein to the tip of the rostrum. This vein was identified as the lateral nasal vein ( vena nasalis lateralis
191 Schummer et al. 1981) based on its origin just dorsal and distal to the maxillary labial vein as well as its general location latera l to the nasal passage and extending toward the external nares. The finest branches of this vein did not inject likely due to the dense and abundant nature of the connecti ve tissue in the region. Nonetheless, despite the anatomical relation, the countless anasto moses with the terminal branches of the maxillary labial veins suggest that there ma y be little functional distinction between the two vessels, with venous flow ultimately depending on local pressures established during utilization of the lips and opening of the nasal passages. The dorsorostral and dorsomedial branches ju st proximal to t he beginning of the angularis oculi vein are consistent in location and investment with the dorsal nasal veins described in domestic mammals (Figures 4-11 to 4-14). In the region of the dorsal nasal branches, small ventrolaterally oriented tributaries emerge and travel toward the palpebrae along the rostromedial border of the orbit. These branches compose the medial ventral and medial dorsal palpebral veins. After giving off the nasal and palpebral branches, the facial vein conti nues dorsocaudad along the medial aspect of the dorsal orbit in a pattern unmistakably reminiscent of the angularis oculi vein ( vena angularis oculi Schummer et al. 1981). An anastomosis between the incompletely filled angularis oculi veins and the dorsal external ophthalmic veins, as is seen in domestic mammals, was not observed in the manatee, however artifactual under-representation remains a distinct possibility. No distinct frontal vein was observed as the terminus of the angularis oculi vein, however, a dorsocaudally oriented branch which ended abruptly was seen via CT angi ography of the region.
192 Branch 3 (angularis oris vein) The third and most rostral branch of the de ep facial sinus is the most slender and extends dorsorostrad at an ob lique angle toward the upper lips (Figure 4-11). This vessel follows the angle of the mouth and is consistent wit h the angularis oris vein ( vena angularis oris Schummer et al. 1981) of domestic mammals. A t its distal extremity, the vein anastomoses with the maxillary l abial veins and the resultant plexus. CNS Veins Brain veins The cerebral vasculat ure was only well inje cted in three of the five specimens. Despite the fact that jugular emissary connections exist between the dural veins and the internal jugular veins, the intracranial va sculature did not inject well in decapitated specimens. This fact alone suggests that the missing component in those specimens, namely the epidural veins, play an import ant role in drainage of blood from the calvarium. Indeed, in the s pecimens containing the cervic al portion of the vertebral column, an epidural venous plexus of considerable size was observed. The cervical epidural rete contains two large ventrolateral veins. These veins source in large part from the sigmoid dural sinuses inside the brain case and from dural veins at the base of the brain consistent in location and course with the basilar sinuses of domestic mammals ( sinus basilaris, Schaller 2007). The bilateral basilar sinuses extend rostromediad to become the cavernous sinuses similar to those seen in small ruminants ( sinus cavernosus, Schaller 2007). A rete-like cavernous sinus like that found in the cow was not present in any of the manatee specimens examined. Ventral petrosal sinuses were not observed connecti ng the basilar sinuses to the cavernous sinuses and it is therefore be lieved that they are not present as no extracranial segment
193 of the dural sinus system was found. This coul d be due to the fact t hat the intracranial bony ridges (e.g. tentorium cerebelli) are greatly reduced or absent in the manatee and much of the ventral braincase is occupied by a large cranial hiatus that is part of the voluminous peribullar sinus. This anatomy ma kes identification of a ventral petrosal sinus problematic. Numerous small anas tomoses were observed between the two cavernous sinuses, however only one large a nastomosis on the caudal border of the pituitary gland was observed, consistent with the caudal intercavernous sinus. A large rostral intercavernous sinus does not appear to exist. Instead, numerous minute veins were observed crossing between the two cave rnous sinuses where they diverge to become the ventral cerebral veins ( Vv. cerebri ventrales, Schaller 2007). The largest tributaries of these rostrally oriented ventra l cerebral veins extend through the orbital fissures in company with an artery, to event ually contribute to the formation of the ophthalmic plexus (for details on the ophthalmic plexus see C hapter 5). In addition to the ventral cerebral veins, the rostral portion of the ca vernous sinus sends numerous small ethmoidal veins that penetrate the rostral calvariu m and participate in draining the caudal-most portions of the ethmoturbinates. Although no discernible ossified falx cer ebri is present, a wedge shaped piece of connective tissue does exist along the longitudi nal fissure of the cerebrum, and contains a prominent dorsal sagittal sinus that extends many ventral twigs that form a plexus. The plexus penetrates through the dura and dives into the longitudinal cerebral fissure to invest the dorsal and medial faces of the cerebral hemispheres. Caudally the dorsal sagittal sinus forms a triangul ar plexiform confluence (confluens sinuum, Schaller 2007) from which bilateral transverse sinuses ( sinus transverses, Schaller 2007) emerge.
194 Laterally, each transverse sinus meets wit h a corresponding temporal sinus, while caudoventrad the transverse sinuses give off the sigmoid sinuses. Each sigmoid sinus extends ventrad through the peribullar air sinus and runs confl uent with the jugular emissary vein of the internal jugular vein described above, but also give off a large caudolateral vein that enters the neural canal and travels al ong the ventrolateral border of the cervical epidural rete as the internal vertebral venous plexus ( plexus vertebralis internus ventralis, Schaller 2007) and receives many of the segmental intervertebral veins that enter the ne ural canal (Chapter 5). Epidural veins The cervical epidural space is filled with a large tortuous mass of veins and arteries similar to the rete found in cetaceans. This epidural rete cradles the ventral and lateral aspects of the spinal cord throughout the entire cervical region, and follows the spinal cord into the braincase. The cervical epidural rete drains in part through lateral connections to the segmental in tervertebral veins that l eave via the intervertebral foramina. The venou s component of the rete is composed of a large mass of small, anastomosing veins as well as two sizable lo ngitudinal internal vertebral veins and a few smaller parallel ones. As the rete approaches the foramen magnum, two distinct components can be seen. The first componentthe ventral internal vertebral veins-ultimately gives off a dorsal and ventral branch on either side of the spinal cord. The ventral branch fuses with the basilar sinus while the dorsal branch fuses with the simgoid dural sinus. The second component of the rete is composed of the tortuous rete veins, which enter the braincase and spr ead out over the brain to invest the dura matter. More details on the anatomy of these structures are provided in Chapter 5.
195 Discussion Functional Implications of Facial Vasculature Despite the relatively derived gross mor phology of manatees, the general patterns of the veins of their head bear some similarity to those seen in dom estic mammals (Figure 6-1). Although the par ent veins of the head such as the internal and external jugular veins show notable differences in thei r courses, their distal branches are often recognizable and invest the tissues in a m anner familiar from domestic mammals. The veins of the face, the lower and upper jaw and the eye are all connected to the jugular system through at least two routes. For inst ance, the ophthalmic plexus is connected to the facial vein via the anastomotic branch of the deep facial vein, and to the maxillary vein and epidural plexus via the ophthalmic vein. This organization presumably allows for blood to be drained through either of thos e routes, allowing for balancing of altered hemodynamic pressures within the ophthalmic pl exus or structures connected to it. Such a phenomenon might be encount ered during utilization of the dexterous lips for manipulation of food. During such events, the deep facial veins may be exposed to elevated surrounding tissue pressures that collapse the lumen and impede blood flow. Blood draining from the tissues of the eye and eye lids into the ophthalmic plexus would not be able to drain into the deep facial sinus and would therefore follow the caudomedial path toward the brain case. Conver sely, it is possible that manipulation of the lips and surrounding tissues helps return blood through the facial veins. This noncardiac pump function is seen in horses, in which chewing causes blood to move through the facial veins from the lips to the ex ternal jugular veins. In domestic species like the horse, this function is aided by the presence of a valve within the facial vein, a structure that was not observed in any of the manatee specimens. Interestingly, there is
196 a marked absence of venous valves in the head of the manatee. In fact, venous valves are generally absent from the peripheral veins of manatees and cetaceans. It is possible that this absence is simply an adapt ive evolutionary response to an aquatic lifestyle in which venous return is not opposed by strong gravitational forces. Alternatively, the absence of valves may be a means for enabling collateral venous flow through any available structure during the elevated pressures associated with diving and axial locomotion. The pr esence of multiple collat eral drainage pathways is not unique to manatees. Countless ar terial collateral pathways to the brain are known to exist in dogs (Clendenin & Conrad, 1979; Moss, 1974; Tatelman, 1960; Whisnant et al. 1956;), and collateralization of venous dr ainage from the head of humans has been known for some time (Batson, 1940; 1956; Breschet, 1819; Epstein et al. 1970; Herlihy, 1947). The innumerable intricate anastomoses between adjacent venous branches are encountered at almost every le vel, a fact that raises one overarching issue. Although a certain degree of analogy and perhaps even homology can be found between the veins of the manatee and those of domestic mammals, those structures may not bear similar functional significance in manatees. With the tremendous collateralization of venous paths that exists in manatees, one cannot help but question whether any of the traditional venous paths are solely responsible for enough drainage to be considered the main functional path. For example, the mandibular alveolar vascular bundle is drained by the mandibular alveolar vein which in domestic mammals would be considered the main drainage vein for the blood inside the mandibular canal. However, as soon as the mandibular alveolar vein leaves the mandibular foramen it is joined by
197 many large anastomoses from the lingual ve in and the pterygoid plexus. Therefore, although the direct anatomic connection may be the mandibular alveolar vein, the functional path may in fact be any one or all of those tributaries. In spite of what specific nomenclature is given herein to the various veins and venous networks, it is clear that the manatee head is invested with a multitude of tortuous, anastomosing, ramifyin g retia and vascular bundles. Many of these structures are associated with similar arterial structures while others encase the major arteries of the head. In most terrestrial mammals such a condition might be considered pathological since the most ext ensive intraand extracranial retia that are seen in the cow, pale in comparison. However, in the world of marine mammals, one does not have to look far to find si milar examples. Ridgway et al. (1974) and Costidis and Rommel (2012) described a venous plexus surr ounding the external carotid artery of the bottlenose dolphin and communicating freely with the internal jugular vein. Although Ridgway et al. (1974) did not call this t he concomitant vein of the external carotid artery, its positioning and drainage into the internal jugular vein may warrant calling it an analog. Ridgway et al .s (1974) illustration and the im ages of dissected specimens provided by Costidis and Rommel (2012) of the structure are strikingly similar to some of the venous encasing of the external ca rotid artery that I have observed in the manatee, though admittedly not as intricate. As one studies marine mammals in greater detail, it becomes increasingly apparent that the permanently aquatic manatees, dolphins, and whales are all invested with extensive v enous networks in their head and neck (pers. obs.). Convoluted vascular patte rns often associated wit h congenital and/or
198 traumatic deformities in te rrestrial mammals (Hassler et al. 1989; Mullan et al. 1996; Rosenblum et al. 1987; Baev et al. 1999) abound in healthy marine mammals. There is little doubt that these elaborate vascular structures represent adaptations or specializations for their aquatic lifestyle s, as they likely represent considerable energetic expenses that requi re building and maintenance th roughout the animals lives. Despite the fact that dolphins and manat ees may be exposed to some similar, though albeit milder, environmental pressures (e.g. hydrostatic pressures), the location, extent, and morphology of the complex retia and vascular bundles is quite different, but always present in some form. I find this to be s uggestive of an adaptive sp ecialization, rather than an evolutionary remnant or accidental resurfacing of atavistic traits. As regards the general anastomosing nature of the jugular veins seen in both species, the most parsimonious explanation seems to be that the multitudinous connections provide a high degree of vascular redundancy through which ve nous return can be collateralized. This may find support in the fact that thes e collateral branches show a fair amount of variability, in the form of bilateral asymmetr y as well as individual variability. This consistent presence but inconsistent expres sion would be in line with the mechanisms of formation of collateral venous pathwa ys, as has been shown in domestic species such as the dog (Clendenin & Conrad, 1979; Moss, 1974; Tatelman, 1960; Whisnant et al. 1956). It has been shown that even domestic mammals possess numerous intrinsic collateral venous pathways that remain small under normal conditions, but enlarge and become well established under pathologic conditions that recruit them on a consistent basis. It then seems reasonable to sugges t that the jugular anastomoses seen in dolphins and manatees may in fact begin as small anastomotic connections that
199 eventually become large, wellestablished veins due to their continual recruitment as collateral pathways. Such conditions might be encountered with the progressive development of diving ability, as an individual transitions from in utero life to an active diving lifestyle. In other words, as an indi vidual begins to swim and dive on its own, its tissues will be exposed to varying pressures, both hydrostatic and locomotory. This notion might easily be investigated by ex amining numerous specimens of different ontogenetic stages. Such an explanation of consistent recrui tment as the driving force behind the generation of the venous struct ures seems to fall short when trying to envision the mechanisms underlying the form ation of the more predict able and consistent venous structures found in the heads of dolphins and manatees. Spec ifically, structures such as the plexus of the accessory sinus syst em of the dolphin (Chapter 3) and the mandibular alveolar and infraorbital vascu lar bundles of the manatee may not be explained by simple exposure to differential hydrostatic pressures. These structures represent relatively elaborate, organized structures that s how considerable consistency in location and structure and appear inexorably tied to other structures such as nerves, fat bodies and air sinuses. In the manatee, the intimate associati on of arterial and venous vascular bundles and retia seems much too organized to be aimle ss or wasteful. Without physiologic experimentation the following postulates wil l remain just that. Nonetheless, the infraorbital and mandibular vascular bundles of the manatee have an unmistakable resemblance to the caudal vascular bundles found in the chevron canal of manatees (Figure 4-12) and cetaceans and the limbs of sloths (Elsner, 1966; Rommel & Caplan,
200 2003; Scholander & Krog, 1957; Scholander & Schevill, 1955); a structure convincingly shown to exert significant thermal influenc e on the regional blood and reproductive temperatures. A similar function is postula ted here for the infraor bital and mandibular alveolar vascular bundles. This thermor egulatory capacity could manifest as: 1) a means for conserving body heat by reducing c onvective heat loss from the face, 2) a means for modulating sensory i nput from the highly dexterous and very sensitive lips of the manatee or 3) a means for affect ing the temperatur e of the brain. Functional Implications of Ophthalmic Vasculature Fritches et al. (2005) showed that retinal heati ng in swordfish significantly improves the temporal resolu tion and therefore the visual det ection of rapid motion. Conversely retinal cooling must have the oppos ite effect. Additionally, low temperature has been shown to reduce the amplitude and ve locity of compound action potentials (Luzzati et al. 1999). Although these data establish a basis for temperature-dependant ocular function, they offer no conclusi ve evidence. Ninom iya and Yoshida (2007) described a complex arterial ophthalmic rete in the cetacean eye and postulated that it served important thermoregulatory functions. Folkow et al. (1988) described an elaborate arteriovenous ophthalmic rete in harp and gray seals, and suggested that it served as a counter-current heat exchanger that lim ited the amount of heat lost through the eyes. Conversely, Dehnhart et al. (1998) found the thermal signature of the seal eyes to remain constant at cold and wa rm temperatures, suggesting that temperature was maintained and that ophthalmic conserva tion of heat via counter-current heat exchange was not occurring. It seems possible that the rete may serve the purpose of maintaining an optimal ophthalmic temper ature despite fluctuating ambient temperatures rather than lim iting the convective heat loss to the environment.
201 Manatees are neither predatory nor typica lly fast swimming and may therefore not require great visual acuity or temporal resolu tion. In fact, manat ees often occupy very turbid and/or tannic waters with limited visi bility, and are theref ore most dependant on their tactile senses. Manatees may therefore be sacrificing their visual acuity in order to gain the thermal benefits of reduced convec tive heat loss from the eyes, and such a function may be accomplished through the ophthalmic plexus. There is also ample evidence that selective brain cooling occurs in various terrestrial mammals studied to date. Indeed, it has been shown that blood cooled at the periphery is used as a thermal sink that absorbs heat from the arterial system. This allows animals to reduce the damaging effects of brain hyperthermia during strenuous exercise. In addition to cooling of the ar terial blood supplying the brain of dogs and ungulates, Caputa et al. (1976) found evidence that selective brain cooling in rabbits was dependent on direct venous blood pools such as the pterygoid plexus rather than cooled arterial supply. Specifically, they found that pterygoid venous blood was crucial in determining the temperat ure of the basal brain of the rabbit, and that brain temperature grew proportionally to the dist ance from the pterygoid plexus. Caputa et al (1976) postulated that the ophthalmic plexus had a role in this type of cooling, presumably due to its anatomic connections to the nasal mucosa and brain. This proposition seems especially tantalizing in the manatee due to the extensive and voluminous distribution of the ophthalmic plexus and its direct connections to both the infraorbital vascular bundle and the cerebral v enous circulation. In light of all the existing evidence regarding selective brai n cooling in mammals, one factor seems puzzling. The venous portion of the infraor bital vascular bundle returns blood from the
202 very muscular and likely thermogenic lips. Additionally, the venous bundle is intimately juxtaposed to an arterial counterpart that is presumably warm and therefore would provide ample potential for counter-current heat exchange. Both of these factors suggest that the veins of the infraorbital venous bundle may in fact be significantly warmer than the rest of the veins of the fa ce. Since the infraorbital venous bundle is connected to the veins of the brain via emissary veins as well as anastomoses with the ophthalmic plexus, it seems possible that warmed venous blood may reach the brain. Such a concept is at odds with the generic mammalian paradigm of brain cooling and would therefore be suspect if it werent for some peculiarities of manatees. Manatees are thought to produce 1/3rd to a 1/6th of the metabolic heat of a terrestrial mammal of similar size (Irvine, 1983). The lower me tabolic heat production and considerable body heat lost through the water ma y put manatees at a lower thermal budget than other marine mammals, as evidenced by the large cold-stress related manatee mortalities during cold winter months (Bossart et al. 2002). The observed vascular patterns may at least in part, be a thermal coping mechanism and would constitute the first example of mammalian selective brain heating, rather than cooling. A dditionally, live manatees that have been documented with thermal cameras dur ing medical exams on cold days, have shown large heat bursts associated with the lips (Joe Gaspard, pers. comm.., 2012). The pronounced thermal signature of the lips suggests increased availability of heat to be returned to the brain via the infr aorbital bundle and ophthalmic plexus. Interestingly, chronic cold stress ulcera tions appear first in the upper lips of the manatee. The ulcerations are thought by so me to be caused by ischemic necrosis of the tissue due to thermoregulatory peripheral vasoconstr iction and redistribution of
203 blood to the body core (Bossart et al. 2002). It seems possibl e that the elaborate maxillary labial venous plexus and it s associated drainage pathway through the infraorbital vascular bundle may have the capability to draw heat away from the outgoing arteries, thereby facilitating hy pothermic damage to lips during periods of extreme or protracted cold expos ure. This could help explain the early manifestation of cold stress lesions in the lips. The degree of involvement of the facial vibrissae is unknown, however the vibrissae contain blood sinuses, and preliminary results suggest that the maxillary labial plexus connects directly to the sinus follicles. It is therefore tempting to suggest that sensor y modulation could also occu r either through modulation of temperature or pressure within the sinus follicles. Implications of Vasculature on Whole Bod y Thermoregulation As discussed earlier, a possible me chanism for maintaining adequate temperatures in the vibrissae during moder ate ambient temperatur es but limiting heat loss during cold ambient temperatures might employ a vascular counter-current heat exchange structure. Indeed, my findings of the presence of an arteriovenous mandibular alveolar and infraorbital vascu lar bundles suggest that manatees may be able to control the amount of heat lost at the lips by capturing the outgoing arterial heat and returning it to the core. There is pres ently no information regarding the microscopic anatomy or vasoconstrictive ability of the vessels of the mandibular alveolar or infraorbital vascular bundles, therefore no in ferences can be made regarding the degree of vasoconstriction during cold exposure. Nonetheless, empirical observations using thermal cameras on manatees suggest that during acute cold exposure manatees maintain elevated temperatures in the vibrissa l fields (Gaspard, pers. comm. 2012). It is not known what effects chronic exposure to suboptimal temperatures has on blood
204 supply in the face of t he manatee, however in human extremities acute exposure typically leads to vasoconstriction, followed by vasodilatation (Bornmyr et al. 2001; Brown et al. 2003; Folkow et al. 1963). The vasodilatation is typically produced after chronic or extreme cold exposure, and is believed to be due to a combination of sympathetic autonomic input and suppression of local vasoconstrictive signals (Folkow et al. 1963). Although in human extr emities the hunting phenomenon of alternating periods of vasoconstriction and vasodilatation is co mmon, it was not observed by Folkow et al. (1963) in cat extremities. Instead, that cat paws show ed an initial vasoconstriction followed by sustained vasodilatation. Addi tionally, by cooling and heating the arterial inflow to the paw prior to lo cal external cooling, Folkow et al. (1963) demonstrated that local cooling of extremities led to consi derably less pronounced blood flow decreases in cats that had been cooled than in warmed individual s. If chronic expo sure to cold leads to sustained vasodilatation of facial veins in manatees, it is perhaps even more important for them to have a mechanism to limit the amount of heat lost at the lips. A counter-current heat exchange structure could conceivably satisfy that requirement by maintaining blood flow but conserving heat.. A nother factor that should not be ignored is the fact that unlike the relatively thin and delicate lips of seals, manatees have large, very muscular lips. Those lips are used frequent ly to investigate objects and manipulate food into the oral cavity, and may produce s ubstantial quantities of heat. It therefore may be possible that the use of the muscular lips produces heat which can then not only help maintain elevated vibrissal temperatures when blood flow or blood temperature are reduced but also heat the blood returning to the head.
205 The possibility of warmed venous blood return from the facial periphery brings about another question. If the low metabolic ra te of manatees combined with the high conductivity and specific heat of water place manatees at a thermal deficit during cold weather, the generic mammalian paradigm of selective brain cooling may not apply to manatees. The venous structur es in the head of the manatees provide hints at possible mechanisms that would facilitate selective brai n warming rather than cooling. Selective brain cooling has been described in panti ng terrestrial mammals and mammals with relatively high respiratory rates, since the cooling effect on the bl ood is generated by evaporative cooling. Since manatees are neither panting mammals nor do they possess elevated respiratory rates or elaborate nasal conchae with high surface area, evaporative cooling in the nasal passages is highly unlikely. Instead, countless emissary veins connect the intracranial vein s to the veins draining the lips and face. Specifically, the largest emissary connecti ons are found between the intracranial veins and the ophthalmic and infraorbital plexuses. As discussed, the infraorbital plexus potential for returning warmed venous blood means that warmed blood could travel through the emissary veins and into the in tracranial veins t hat bathe the brain. Finally, it seems as though the venous morphology in the head and neck of the manatee shares a number of common charac teristics with many of the domestic mammal and marine mammal species (Figure 6-1) The various facial veins resemble those seen in the horse, while the plexifo rm arrangement of the veins of the rostrum resembles that of the cow. The internal jugul ar veins show similarities to those of the pig, while the masseteric retial structures are most like cows and cetaceans. Nonetheless, the infraorbit al and mandibular alveolar vascular bundles are not
206 encountered in any of the terrestrial or marine mammals examined to date, and constitute a unique trait in manatees. Although no arteriovenous anastomoses were observed in any of the specimens, the viscosity of the latex vascular casting mate rial used in this research would likely not allow passage through small ana stomoses. This was int entional since arteriovenous cross over would have obfuscat ed the results, however it means that the presence or absence of arteriovenous anastomoses cannot be assessed. Interestingly, Elsner (1966) noted the presence of numerous si zable arteriovenous anastomoses in the brachial vascular bundle of the dugong. Gi ven the relatively proximate phylogenetic relationship between dugongs and manatees and t he similar vascular structures, such anastomoses may also exist in the manatee. Whether those anastomoses are present only in the brachial vascular bundle or al so in the facial bundles is not known. The exact morphology of the venous branches investing the lateral aspect of the face was quite variable both within and bet ween individuals. Although the same branches were always present, the location and size of the branches varied considerably. This created a challenge in deciphering the patterns and subsequently describing a general pattern for manatees, however, the mere presence of this inconsistency did not seem problematic. The venous patterns observed in the region suggest that their formation is quite labile. The many collateral pathways likely provide numerous drainage routes whose differential recr uitment or inactivity results in them being pronounced or atrophied. It is known from experimental work in dogs that preexisting collateral pathways of drai nage are employed when major drainage pathways are ligated. When this ligation is chronic, collateral pathways enlarge and
207 become well-establish ed routes of drainage (Clendeni n & Conrad, 1979; Moss, 1974; Tatelman, 1960; Whisnant et al. 1956). Given the variability observed in just five manatee specimens, it seems r easonable to suggest that similar forces may guide the development of the veins in the facial/masse teric region of the manatee. A combination of ontogenetic and life history features (e.g. diving frequency, muscular use, etc.) may help drive the enlargement of different veins while others regress. As mentioned earlier, the external jugular veins were expected to dominate the head and neck due to the considerable muscular investment of the face (Hegedus & Shackelford, 1965). Yet what I found was that the internal jugular veins could obtain substantial caliber, often appr oaching that of the external jugular veins. This arrangement might seem problem atic were it not for the fact that numerous large anastomoses were found between the internal and external jugular system. These anastomoses likely help explain the large size of the proximal internal jugular vein as it appears that it forms a sizable collateral drai nage path for blood returning from the facial and external jugular system. Th e large size of t he proximal internal jugular veins is therefore likely more reflective of its colla teral facial drainage role than its role in draining blood from the brain which appears to happen through emissary veins of relatively modest caliber. It should be noted that although the viscous nature of the latex/barium mixture did not allow for injection of finer vessels and capillaries in the lip, 3D reconstruction of the maxillary labial veins suggests the presence of an extensive plexus. It is tempting to suggest that this plexus is in some signifi cant form associated with the blood sinuses surrounding the hair follicles and vibrissae in the lip.
208 Table 4-1. List of specimen numbers, to tal body length (TBL), stranding date, gender and description of research use for each specimen. Specimen ID TBL (cm) Gender Research Use LPZ102654 275 F A, G LPZ102900 251 M A, G MSTm1001 302 M A, G LPZ102904 315 F G LPZ102962 307 F C LPZ102961 209 M Histology A=Angiography; G=Gross dissection; C=Corrosion cast Table 4-2. List of figure labels and their corresponding structure names. Structure label Structure name 1 Brachiocephalic v. 2 External jugular v. 3 Internal jugular v. 3' Anastomosis to ascending pharyngeal v. 4 Brachial vascular bundle 5 Caudal cubital v. 6 Ventral masseteric v. 7 Facial v. 7' Deep facial sinus 8 Maxillary v. 8' Maxillary a. 9 Lingual v. 9' Lingual a. 10 Internal thoracic v. and a. 11 Angularis oculi v. 12 Pterygoid sling. 13 Deep facial plexus 14 Masseteric plexus 15 Ascending pharyngeal v. 15' Anastomosis to maxilary v. 16 Angularis oris v. 17 Mandibular alveolar v. and vascular bundle
209 18 IVB 19 Mental v. 19' Mental a. 20 Epidural rete 21 Aortic arch 22 Right brachiocephalic a. Table 4-2. Continued Structure label Structure name 23 Right subclavian 24 Common carotid a. 25 Pharyngeal plexus 26 Hyoid venous arch 27 Emissary of jugular foramen 28 Vagus n. 29 Superficial temporal v. 29' Superficial temporal a. 30 IVB 31 Infraorbital a. 32 Internal carotid a. 33 Mandibular alveolar a. 34 Pterygoid v. and a. (or plexus) 35 Palatine v. 36 Mandibular labial v. 37 Maxillary labial v. 37' Midline anastomosis 38 Deep facial v. 38' Anastomosis to ophthalmic plexus 39 Transverse facial v. 40 Lateral nasal v. 41 Dorsal nasal v. 42 Ophthalmic plexus 43 Deep masseteric v. 44 Deep temporal v. 45 Vertebral a. 46 External Carotid a. 47 Meningeal v.
210 Figure 4-1. Ventral view of a CT angiographi c reconstruction of the arterial system of the head and neck of a Florida m anatee showing major arteries.
211 Figure 4-2. Medial view of a CT angiogr aphic reconstruction of the major arteries of the right half of the head and neck.
212 Figure 4-3. Ventral view of three-dimensional reconstruction of a Florida manatee showing complexity of jugular venous br anches (blue). Aortic arch and main rostral branches are shown in red. Numbers follow Table 4-2.
213 Figure 4-4. Ventral view of the head and neck of a Florida manatee showing gross morphology of branches (blue) of t he brachiocephalic veins (1) and their associations with the arterial structures (4, 14, 22, 23, 24). Note that this specimen is the same spec imen imaged in Figure 4-1.
214 Figure 4-5. View of the right half of the head showing right inte rnal jugular vein (3) as it receives the emissary of the jugular foramen (27) fr om the braincase. Also visible is the epidural rete (20) entering the brain case to connect to the many meningeal veins (47) exiting the dura to perforate the surface of the brain. Caudal is to the right and rostral to the left.
215 Figure 4-6. Medial view of right maxi llary vein (8) and its countless anastomosing branches encasing the external carotid artery. The pharyngeal plexus can be seen being form ed by convergent anastomoses from the lingual veins (9), mandibular alveolar bundle (17), pt erygoid veins (34), ascending pharynge al vein (15), facial vein (7), maxillary vein (8) and superficial temporal veins (29) Rostral is to the left. Caudal is to the right.
216 Figure 4-7. Ventral view of the head of a Florida manatee showin g gross appearance of venous hyoid arch (26) draining into numerous laryngeal and pharyngeal branches and connecting to facial (7) and lin gual (9) tributaries of the external jugular vein.
217 Figure 4-8. Medial view of gross dissect ion of the neck region of a mid-sagittaly sectioned manatee head. Visible is t he ascending pharyngeal vein (15) draining into the maxillary vein (8). Note the large caudal anastomosis (3) between the internal jugular vein and the ascending pharyngeal vein. Also note the large rostral anastomosis ( 15) between the ascending pharyngeal vein and the maxillary vein (8) medial to the digastricus muscle. Rostral is to the left. Caudal is to the right.
218 Figure 4-9. Right lateral view of gross dissection of the neck region of a manatee showing the confluence of the ascendi ng pharyngeal vein (15) with the maxillary vein (8). Note the lar ge caudal anastomosis (3) between the internal jugular vein and the ascending pharyngeal vein. Also note the large rostral anastomosis (15) between t he ascending pharyngeal vein and the maxillary vein medial to the digastricus muscle. Rostral is to the right. Caudal is to the left.
219 Figure 4-10. Left lateral view of the facial vein emerging later ad from the mandibular notch for the facial veins, and receiv ing deep facial and ventral masseteric vein caudally. The numerous labial br anches (36, 37) can be seen emerging from the rostral (lef t) aspect of the facial vein on its course around the eye as the angularis oculi (11). The extensive masseteric plexus (14) and associated transverse facial (39), deep masseteri c (43), ventral masseteric (6) and deep facial (38) appear prominently. The deep fa cial sinus (7) is also visible. Rostral is to the left. Caudal is to the right.
220 Figure 4-11. Left lateral view of 3D reconstruction of a CT venous angiogram of a Florid a manatee head showing the path of the facial vein (7) and its num erous contributing branches draining the upper (37) and lower lips (36) and nasal passages (40, 41).
221 Figure 4-12. Ventral view of 3D reconstruction of a CT venous angiogram of a Florida manatee head showing the path of the facial veins (7). Also visible is the association of the maxillary labial veins (37) with the infraorbital vascular bundle (18).
222 Figure 4-13. Dorsal view of 3D reconstruction of a CT venous angiogram of a Florida manatee head showing the origins of the facial vein (7) as the angularis oculi (11). The nasal (40, 41) and ma xillary labial veins (37) can be seen draining into the facial veins.
223 Figure 4-14. Simplified schem atic of a medial view of a manatee head summarizing the major venous structures described.
224 Object 4-1. Three-dimensional recons truction generated from computed tomographic angiography of a Florida manatee with a contrast enhanced venous system. The blue structures re present the v enous system; red stru ctures represent the arterial system. Note the complexity of branching patterns in the neck region. Also note the considerable venous investment of the lips.
225 CHAPTER 5 DETAILS ON SOME UNIQUE STRUCTURES OF THE VENOUS SYST EM OF THE HEAD AND NECK OF THE FLORIDA MANATEE (TRICHECHUS MANATUS LATIROSTRIS) Chapter Foreword As discuss ed in the previous chapter, the Florida manatee possesses an intricate system of interconnected veins throughout its head and neck. Despite this seemingly haphazard arrangement, certain regions contain remarkably organized vascular structures that imply specif ic functional purposes. The organized structures can be found in the mandibular canal, the infraorbita l canal, the bony orbit, and surrounding the cranial and cervical central nervous system (CNS). These venous complexes represent two different and discrete types of vascula r structures, namely vascular bundles and retia Unfortunately, there has been considerable historic inconsistency in the use of the term rete So much so, that Wilson was compelled to note that it seems unfortunate that the term rete mirabile should have been applied so indiscriminately to all vascular plexiform arrangements, for, both anatomically and teleologically, when t he various plexuses are compared with each other, a good deal of confusion and great want of harmony are at once observable. (1879) Galen (129-201 AD) first coined the term rete mirabile (wonderful network) to represent the intricate, tort uous nature of the net -like arrangement of arteries found at the base of the ox brain (Fawcett, 1942; Viale, 2006). Alt hough the original rete mirabile of Galen was a term used for the rete mirabile caroticum at the base of the brainan exclusively arterial structure it later acquired a more general meaning not specific to arteries but rather to a cert ain geometric arrangement of bl ood vessels, be they arterial or venous in nature. The term was late r refined by numerous authors who described different types of retia based on their origins, terminations, and branching patterns. For
226 instance, Wilson (18 79) suggested that retia be divided into bilateral types having two margins each supplied by a different parent vessel, and axial types found in the course of any single vessel. He furt her subdivided axial types into mediate and terminal retia based on whether they re-coalesce or not, and complete and incomplete depending on whether the entirety of the parent trunk subdivides into the rete Under Wilsons nomenclature, the original rete mirabile caroticum described by Galen would likely have been called a complete mediate axial type since it coalesces distally into intracranial internal carotid arteries. Although Hunter (1787) was the first to describe vascu lar plexuses in whales, Richard Owen (1868) was the first to use the term retia mirabilia to describe the thoracic retia of cetaceans. Unfortunately, Owen (1868) later stated that he was unable to detect vascular plexuses in the manatee and Hu xley (1866) similarly claimed they were absent in dugongs, a close relative of the m anatee. This led to a general belief that retia indeed vascular plexuses as a whole, we re absent in sirenians. Nonetheless, investigators like Murie (1874) and Fawce tt (1942) later discovered that vascular plexuses were in fact present in manatees however, most of th em attained a geometry quite different from the retia known in cetaceans. Murie (1874) and Fawcett (1942) both described an arteriovenous plexus in the pecto ral flipper of the manatee that resembled the shape of a paintbrush com posed of approximately 600 arteri es intimately associated with approximately 1200 veins, All of these veins drain into the subclavian vein. Rather than the tortuous, anastomosing vessels described in retia of cetaceans and terrestrial mammals, manatee extremities have numerous vascular bundles of similarly sized, roughly linear and parallel arteri es and veins that issue almo st all at once from their
227 parent vessel. Fawcett decided that these structures were most similar to those described only a few years earlier in the extremities of sloths (Wislocki & Straus, 1932), and suggested that they be si milarly named as vascular bundles. The distinguishing feature of these va scular bundles from retia is that the parent vessel undergoes a relatively complete and sudden subdivision in to innumerable small tributaries that run parallel to each other for substantial distanc es, rarely encountering any further major reduction in size. Fawcett (1942) discove red these vascular bundles to be most prominent in the pectoral flippers of the ma natee, but also noted finding smaller ones in the intercostal spaces, the chevron canal, and as terminal branches of the external carotid artery on the side of the face. The terminal branches of the external carotid artery that Fawcett (1942) referred to were likely the same ones described by Stannius (1845) as the infraorbital arteries. Unfortunately, neither author examined the venous system of that region and ther efore both missed the intimate arteriovenous associations of that structure. It is in teresting that Fawcett placed rela tively modest attention on the caudal vascular bundle in the chevron canal of the manatee, as later research showed it to be a structure of considerable size and functional significance (Rommel & Caplan, 2003). In addition to the fact that manatees po ssess vascular bundles, they also appear to have certain vascular structures more like the retia of cetaceans, a fact which makes Owens (1868) and Huxleys (186 6) assertions of a lack of vascular plexuses in manatees more troublesome. In fact, Fawc ett (1942) stated the beliefs of two such eminent comparative anatomists as Owen and Huxley have been difficult to eradicate.
228 Barnett et al. (1958) found that the cranial portions of the cervical spinal cord of the dugong were surrounded by retial tissue, and as I have described in Chapter 4, I found a similar structure in the neur al canal of the Florida manatee. Materials and Methods The experimental procedure inv olved postmor tem latex injection and dissection of 4 Florida manatees, and corrosion casting of 1 Florida manatee. All specimens were obtained postmortem, under a U.S. Fish and Wildlife Service permit (#MA067116-1), and all experimental procedures were cond ucted under University of Floridas IACUC (Permit #: 200801345). Flor ida manatees were used because the presence of a statewide stranding and rehabilitation network facilitates the availability of fresh specimens. Fresh specimens were used in order to enable adequate flushing and injection of blood vessels of interest, si nce clotted blood in suboptimal specimens adversely affects filling of vessels. Four manatee specimens were used for imaging and dissection. The experimental methodology used for the four specimens varied slightly due to improvement of the technique over ti me as well as alternate experimental priorities and differing specimen types. All injections were performed on specimens obtained from rescued animals that were euthanized due to untreatable medical conditions, as determined by the veterinary st aff at the rehabilitati on institutions. Following euthanasia, all specimens received a vascular flush using 0.9% phosphate buffered saline (PBS) solution. The volume of PBS that was used varied based on specimen size, volume required to obtain cl ear effluent from the draining vessels, and degree of tissue edema during flush. Sali ne flush volumes ranged from 10L to 40L depending on the size of the specimen and degree of blood clotting. Since the venous system was the primary target, saline flushes were always begun through the arterial
229 system, in order to help force blood through t he capillary beds and out of the veins. Once clear effluent drained from the veins, t he flush was reversed in order to clear the arteries of any blood clots. Following the P BS flush, specimens were allowed to drain for 2 to 4 hours, a fter which time one of two procedur es was followed. Two of the specimens received vascular latex injecti ons. The other two specimens received arterial perfusions of 18L of 4% neutral buffered formalin (NBF), in order to help preserve the tissues and prolong the availa ble dissection time. Following the NBF perfusion and another 2 to 4 hour draining period, the sp ecimens were injected with latex mixtures. For angiographic imaging of the blood vesse ls in a computed tomography (CT) scanner, a mixture of liquid la tex (Carolina Biological, In c.) and 98% w/w barium sulfate suspension (Liquid Polibar Plus, Bracco Diag nostics) was injected into the vessels of interest, according to a modification of the protocol presented by Holliday et al. (2006). Latex injections varied slightly based on imaging goals (arter ial vs. venous) and specimen characteristics (head vs. head and crani al thorax). Fresh specimens received venous injections of a 60:40 latex to barium sulfate suspension. Specimens that were perfused with NBF received 5% larger volume of barium sulfate suspension in the latex mixture. This increase in contrast agent wa s performed in order to increase the signal difference between the injected blood vessels and the soft tissues, because preserved tissues have greater radiopacity on CT, effe ctively reducing the differentiation between vessels and surrounding tissues. Three of the five specimens were imaged through a CT scanner, using a thin slice protocol. CT imaging was c onducted at the University of Florida College of Veterinary
230 Medicines diagnostic radiology department and at the Baptist Mariners Hospital in Tavernier, FL. Axial slices were obtained from the specimens at 3mm thickness, with 1mm slice intervals. Whenever possible (spec imen size allowing) the volume data were reconstructed to 0.5mm slice thicknesses (alternatively 1.0mm). Since postmortem specimens were used, radiation exposure leve ls were not a concern. The resulting DICOM data was post-processed using Amira (Visage Imaging, Inc.) software on a Gateway Precision T3500 with memory and processor enhancements. Post-processing was carried out in order to visualize t he data in 2D and 3D formats and gain an understanding of vessel locations and relationsh ips prior to gross dissection. Once three-dimensional images we re generated to guide the dissections, gross dissection was carried out on each specimen in order to validate and/or clarif y structures observed on CT. Findings on dissection were photo-documented. A fifth manatee specimen was flus hed with PBS and perfused with Mercox corrosion casting material (Ladd Research, Inc. ). Blue solution was injected in the veins and red solution in the arteries Following a 2hr curing submer sion in a cold water bath, the specimen was gently lowered into a vat c ontaining 15% KOH (Sigma-Aldrich, Inc.). Only the rostrum was submerged initially, in order to corrode away the dense, heavy tissues of the snout. Following a week of rinses and resubmersions, the head was completely submerged in the vat with a fres h solution of 10% KOH. Numerous rinses and resubmersions were conducted over the per iod of three more weeks, until most of the tissue was corroded. The resulting spec imen was rinsed in a water bath for several days and then allowed to air dry.
231 Results Unlike most of the veins of domestic mamm als, many of the veins in the head of the manatee form a tortuous network of intercommunicating veins and vascular bundles. The veins associated with sensory structures (e.g. eyes and lips) of the manatee abound with elaborate venous structures, some of which are intimately associated with similar arteri al counterparts, while others appear to stand alone. The structures of greatest intere st here are the mandibular alveolar and infraorbital vascular bundles, the ophthalmic plexus, and the epidural retia It should be noted that the estimates of luminal diameter may be exaggerated since they are based off of latex injected specimens rather than histol ogic measurements and depending on the degree of pressurization during latex injecti on, veins may be over expanded or under represented. Mandibular Alveolar Vascular Bundle (MAB) The mandibular canal of the Florida manatee is occupi ed by two structures: the mandibular alveolar branch of the trigeminal nerve and the mandibular alveolar vascular bundle. The vascular bundle is an arteriovenous structure composed of numerous fine caliber (<1mm), occasionally anastomosing arteries embedded in a network of similarly sized anast omosing veins (Figure 5-1). The frequency of anastomoses between veinsseemingly far greater than that of the arteries--forms the likeness of a venous lake surrounding the arteries (Figure 5-2), with a striking similarity to the caudal vascular bundle (CVB) of the manatee (Ro mmel & Caplan, 2003). In the center of this vascular bundle can be found the mandibular alveolar ner ve (Figure 5-2). The MAB extends the entire length of the m andibular canal, exiting distally via the mental foramina and proximally via the m andibular foramen.
232 Distally, the veins and arteri es of the MAB exit the m ental foramina and invest the muscles (e.g. mm. mentalis, orbicularis oris and pars oris of sphincter coli superficialis ; Domning, 1978) and skin of the lower lip, and the oral pad on the dorsal surface of the entrance of the oral cavity (Figure 5-3). Proximally, the MAB exits the mandibular foramen at which point the arterial and v enous contributions separate somewhat. The arterial component coalesces into a single mandibular alveolar artery that gently curves dorsocaudad to fuse with the maxillary artery (Figures 4-1 & 4-2). The more complicated venous component of the MAB exits the mandibular foramen and coalesces into numerous large veins but never into a single large mandibular alveolar vein. These large veins travel caudad for a short distance, all the while receiving numerous anastomoses from the lingual, facial, and maxillary veins (Fi gures 5-1 & 5-3). Approximately 1.5-2cm caudal to the mandibular foramen, the drainage fields diverge, forming relatively discrete paths through the lingual venous branches and maxillary veins. The connections to the maxillary veins are via at least two sizable, continuously anastomosing veins that surround the mandibular alveolar artery and may therefore be named the mandibular alveolar veins. The largest and most direct proximal connection of the venous component of the MAB is to the maxillary vein via twosometimes more --mandibular alveolar veins. It should be noted that at the level of the mandibular foramen where the venous component of the MAB exits t he foramen, some of the num erous anastomoses to the lingual veins are formed by a small plexus that invests the tongue. Therefore a similar but apparently smaller arteriovenous plexus ex ists within the root of the tongue and grossly has the appearance of a structure conf luent with the proximal potion of the MAB
233 (Figure 5-3). Indeed, in some specimens t he lingual venous plexus could be said to originate from the mandibular alveolar veins and their concomitant vascular bundle, rather than from di screte lingual veins. Infraorbital Vascular Bundle (IVB) The infraorbital cana l of the manatee is filled by two structures, namely the infraorbital branch of the trigeminal nerve, and the infraorbital vascular bundle. Like the MAB, the IVB is an arteriovenous structure composed of many anastomosing fine caliber arteries embedded in a venous lake -type structure formed by numerous anastomosing fine caliber veins (Figure 5-4). Despite their similarity, the IVB attains a cross sectional area 3 to 5 times larger t han the MAB. The arte rial component of the IVB is composed of over 100 frequently anasto mosing arteries rangi ng in diameter from 0.2 to 1.0mm. The venous component invest s and envelopes the arterial component so that on the periphery only the fi nest caliber arteries are seen, while the center contains more large caliber arteries (Figure 5-5). The IVB is in timately associated with the infraorbital branch of the trigeminal nerve, proximally being bordered by branches of the nerve on its ventrolateral side (Figure 5-5), while distally the nerve branches out and perforates part of the substance of the IVB as it ex its the infraorbital foramen to invest the maxillary labial tissues. Histologically (Figures 5-6 & 5-7), the IVB presents nearly identical to the MAB and to the CVB descr ibed by Rommel & Caplan (2003) (Figure 56). Distally, the IVB exits the infraorbital fo ramen and begins to branch into clusters containing a central artery bounded by two anastomosing satellite veins. The paired veins anastomose not only between themselves but also proximally with other pairs of veins from juxtaposed arteriovenous triads. These arteriovenous triads spread out,
234 investing the maxillary labial muscles ( mm. maxillonasolabialis lateralis nasi and levator nasolabialis ; Domning, 1978) and skin (Figure 5-8). On their path to the distal most parts of the lips, the veins of some of the triads anastomose with the superficial and deep maxillary labial veins that drain into the facial vein. Along its course through the infraorbital ca nal, small veins leave the IVB and enter minute ventral alveolar foramina in the max illary and jugal bones that form the ventral portion of the infraorbital canal. Although t he veins were too small to follow to their terminus, their source, location and investment is consistent with that seen in the maxillary alveolar veins or dental rami of dome stic mammals (Schummer et al. 1981). Dorsad, the IVB sends a cluster of veins into the body of the temporalis muscle as the deep temporal veins ( v. temporalis profunda Schummer et al. 1981). Proximally, the IVB exits the maxillary foramen and quickly c oalesces from the countless veins of the bundle into two or three main veins that anastomose freely with each other, partially encasing the concomitant maxillary artery. These veins form the distal end of the maxillary vein(s) which proximally curves ve ntrad to the pterygoid region, is joined by the pterygoid plexus and mandibular al veolar veins and eventually becomes the external jugular vein. Ophthalmic Plexus The ophthalmic plexus of the manatee is composed of numerous small and large veins that anastomose back and forth with each other (Figure 5-9). The largest caliber veins, in places as large as 1cm across, ar e located just inside the thick periorbita and surround the extrinsic muscles of the eye. As these lar ge veins anastomose with each other, they form sinus-like dilatations. Media lly these veins of the plexus send many small twigs, some of which ramify into the extrinsic eye muscles while others connect to
235 much smaller veins forming a delicate plexus medial to the extrinsic eye muscles. The majority of the plexus however is located lateral to the extrinsic eye muscles and just inside the periorbita. Ventrolaterad, the ophthalmic plexus receives a large connection from the anastomotic branch of the deep facial vein (Figures 5-9 and 5-10). The details of this anastomotic branch are discussed in C hapter 4. This branc h is considerable in size, often attaining diameter s of 0.5cm. At least two but usually three arteriovenous clusters connect the dorsal aspect of the IVB to the ventromedial aspect of the ophthalmic plexus (Figure 5-9) Caudally, the ophthalmic plexus coalesces into smaller veins that form a cone-like stru cture that follows t he tapering boundaries of the eyestalk. The plexus ultimately coalesces into a small cluster of veins surrounding a central ophthalmic vein (Figure 5-10 & 5-11), all of which course on a linear path to the orbital fissure, in common with the optic nerve. T he central ophthalmic vein of this venous cluster was not observed attaining a diamet er greater than approx imately 1mm, while the veins surrounding it are much smaller (Figur e 5-11). As soon as this cluster of veins entered the orbital fissure it bifurcates, sending a caudal branch in line with the parent ophthalmic vein, and a dorsolateral branch (Fi gure 5-12). The ventral branch merges with the cavernous sinus, while the dorsola teral branch invests the dura on the lateral aspect of the ipsilateral cerebral hemisphere, eventually joining with the various other dural tributaries that are pr esent. Along its path to the cavernous sinus, the ophthalmic vein receives a few minute venous twigs from the maxillary artery. CNS Veins Brain veins Like the spinal cord, much of the brai n is surrounded by a sizable venous plexus. Although the entire dura appears well invested with veins (Figure 5-13), certain regions
236 have considerably increased densities. T hese regions appear to coincide with areas where the dura dives medially along major sulci and in regions where the fibrous falx cerebri and tentoria cerebellii exist. It is likely that these fibrous wedges--often times being extensions of the falx cerebri and tentor ia cerebelli--that invest various partitions of the brain act as structural pathways for the veins that emerge from them and penetrate the brain. Nonethele ss, throughout the surface of the brain the dural plexus described in Chapter 4 sends multiple perfo rating branches that penetrate the dura and invest the surface of the brai n (Figure 5-14). In addition to those, the ventral aspect of the brain has an extensive investment of veins. Along the median plane of t he brain, a large straight sinus leaves the rostral border of the beginning of t he dorsal sagittal sinus. This happens at the caudal border of the longitudinal fissure of the cerebrum, in the creas e formed between the cerebral hemispheres and the vermis of the cerebellum. The straight sinus travels rostrad only a distance of about 1cm before a si ngle dorsal vein of the corpus callosum issues from it and extends along the dorsal margin of the corpus callosum (Figure 5-15). The straight sinus then curves slightly ventrad and subsequent ly bifurcates laterally into two sizable veins each of which travels directly into its ip silateral cerebral ventricle. As they enter the first and second ventricle these veins become the choroid plexuses that travel along the medial margin of each ventricle to its ro stral-most extent at which point they curve laterad to form two horseshoe shaped loops. These loops are composed of a large main choroid vein that forms the horseshoe, with numerous small plexiform veins that connect the various parts of the loop (Figur e 5-16). After curving around the lateral margin of the ventricle each choroid plexus courses caudad within its respective
237 ventricle, wrapping caudoventrad around the interthalamic adhesion toward the mesencephalic aqueduct. As each choroi d plexus wraps around the interthalamic adhesion, it undergoes a substant ial diminution in caliber. Epidural veins As mentioned in Chapt er 4, the calvar ium and neural canal are invested with a considerable mass of tortuous veins and arteries The veins and arteries of the cervical neural canal form an epidural plexus (rete mirabile epidurale, Schaller, 2007) composed of a dense network of vessels that cradle th e spinal cord on all but the dorsal-most aspect where the plexus is minimal (Figur e 5-13 & 5-17). Both arterial and venous components of the rete receive regular contributions at every intervertebral foramen, however the more caudal venous contributions appear to be larger in caliber (Figure 518). With both arterial and venous component s injected, near the foramen magnum the epidural plexus obtains a cross sectional th ickness of 0.5cm on the ventral aspect, while lateral walls were in places thicker than 1cm (Figure 5-17). At approximately the level of the first and second cervical vertebra, the pl exus becomes considerably thinner in cross section and a prominent bilatera l separation forms on the ventral aspect so that the left and right sides do not communicate ventrally (Figures 5-18 and 5-19). Therefore the left and right sides of the epidural rete do not appear to communicate until the rete approaches the foramen magnum at which point the two side anastomose intricately in a manner that makes them utte rly indistinguishable (Figur e 5-19). When the left and right sides of the segregated epidural rete are pulled apart, numerous small radicular veins can be observed traveling through the dur a in common with the v entral rootlets of the cervical nerves (Figure 5-19). It shoul d be noted that although the plexus becomes thinner in the more caudal cervical regi ons, it still attains a considerably more
238 voluminous nature than anything I have found described thus far in terrestrial mammals. Additionally, since the specimens used di d not have the thorac ic segment of the vertebral column, it is not possible to asse ss how much of the caudal diminution of the epidural plexus was due to experimental artifa ct such as clamping with subsequent poor filling. On the dorsal aspect the epidural veins ta ke on a very different morphology. Although numerous small longitudinal anastomoses exist, the median portion contain a few larger longitudinal veins that are present throughout much of the cervical section of the neural canal but appear of larger caliber in the caudal regions (F igure 5-20). These dorsal median veins received regular lateral contributions to their drainage from each side of the epidural rete at the level of each intervert ebral foramen, thereby forming venous arches. These venous arches are consistent in location and morphology with the arcuate branches of the internal vertebr al plexus of domestic mammals (Reinhard et al. 1962). If indeed homologous to the domes tic mammal arcuate veins, this morphology may in fact not be related to the epidural rete itself but rather the location at which both the rete and dorsal median veins re ceive anastomoses from the intervertebral veins that enter the neural canal. As t he dorsal median veins approach the foramen magnum they diverge laterally from each other and join the dural venous plexus that surrounds the dorsocaudal and v entrocaudal aspects of the cerebellum. This plexus sends many anastomoses to the large dural sinuses (e.g. confluens sinuum, temporal sinus, etc. ) and extensive dural plexus that surrounds the brain.
239 Discussion Mandibular Alveolar Vascular Bundle (MAB) I have been able to find only one mention of the arterial component of the mandibular alveolar vascular bundle, and no venous references. Murie (1874) stated that the inferior dental (presumpti ve mandibular alveolar ), lingual, and internal maxillary ( maxillary ) arteries originate from the pterygoid portion of the external carotid artery as retial bundles. He then stated that the vascular network (presumably only the mandibular alveolar portion) extends into the large vacuity (presumptive mandibular foramen) of the mandible resembling a mesh work of fibrous tissue which like in cetaceans has interstices partially occupi ed by fatty tissue and ner ves. These findings seem problematic for a couple of reasons. Firstly, in the two specimens of mine that received an arterial latex injection in addition to the venous injection, I was able to find discrete, singular branches of the mandibular alveolar, lingual, and maxillary arteries originating from the exte rnal carotid artery (Figures 5-1 to 5-2). Indeed shortly after their origin the lingual and mandibula r alveolar arteries ramified into finer plexuses, but in neither specimen did they arise from the ex ternal carotid arteries as a plexus. Secondly, Muries comparison of the fatty and nervous tissues associated with the mandibular alveolar plexus within the m andibular foramen to those seen within the dentary of cetaceans seems inap propriate. Although there is a small amount of fat, the largest cross sectional area of that space is without a doubt occupied by the centrally located mandibular branch of the trigeminal nerve, and the venous counterpart to the arterial component. This is entirely unlike the condition seen in odontocetes which have a massive mandibular foramen and canal occupied primarily by lipid of a special functional nature and by an expansive intramandibular venous plexus (Chapter 3).
240 Infraorbital Vascular Bundle (IVB) The only two references to the infraorbita l vas cular bundle that I have been able to find were those made once again by Murie (1874) as well as by Fawcett (1942). As Murie barely covered the venous system, his descr iption of the IVB wa s only arterial. Yet once again I found his description rather troublesome. Murie ( 1874) stated that the numerous capillaries of the internal maxillar y division pass on to the pterygo-maxillary fissure and send inwards superior dental arte rioles; whilst the main mass, lying the lateral groove of the maxilla, is cont inued on through the orbi t and emerges at the infraorbital foramen, spreading amongst the fles hy and other structures of the face and snout. Firstly, his assignation of the arteri es of the infraorbital vascular bundle as capillaries seems entirely unwa rranted, as they are much too large to be capillaries. Secondly, he stated that the superior dental ( maxillary alveolar ) branches are arterioles that branch off of the maxillary capillaries. This too seems inappropriate since arterioles are usually upstream of capillaries and indeed the maxillary alveolar arteries I observed were considerably smaller in caliber than the arte ries of the IVB. Fina lly, the path of the IVB described seems inconsistent. Though per haps a linguistic artifact, Murie (1874) states that the IVB travels thr ough the orbit on its way to the ro strum. In fact, I found the IVB to travel exclusively ventromedial to the eye and the bony orbit. This discrepancy might reflect the fact that t he infraorbital canal is in fact incomplete laterally and therefore does not have a bony separation from the more late ral orbit until it reaches the most rostral aspect of the orbit where t he jugal bone forms a complete infraorbital foramen. Given the fact that he made special mention of the nerves and connective tissue of the mandibular alveolar vascular bundle, it is surprising that he did not make mention of
241 the large cluster of infraorbital nerves and veins that accompany the arterial component of the IVB. Despit e these various inconsistencies and omissions, the general character of his description of the IVB correlates to what I observed in its arterial and venous components. Interestingly, Fawcett (1942) referred to a terminal spray of vessels radiating from the external ca rotid artery to supply the super ficial structures of the side of the head. As no further m ention was made it is impossible to determine if he was referring to the masseteric plex us (Chapter 4) located on the actual side of the head, or the terminal branches of the infraorbital vascular bundle which emerge onto the lips and supply much of the tissue of the lips and rostra l oral cavity. Since he called it a terminal spray of the external carotid artery, I am inclined to believe he was referring to the termination of the infraorbital vascular bundl e, however, like Murie (1874) he made no mention of the even more volu minous venous portion of the bundle. Instead he merely noted that when the bundle radiat es to supply the face, each artery is accompanied by two veins. As one examines the distribution of t he arteriovenous vascular bundles in the manatee, an overarching priority of heat conservation emerges. This comes as no surprise given the lower metabolic rate and aquatic lifestyle of the manatee. Scholander and Krog (1957) showed that the limbs of the sloth are invested proximally with arteriovenous vascular bundles and th rough direct heat measurements showed that these structures act as counter-cur rent heat exchangers. Even the juxtaposed deep arteries and veins in the human extrem ities have been shown to be capapble of counter-current heat exchange, and those structures provide much less surface area than the vascular bundles found in manatees (Bazett et al. 1948a; 1948b; Horvath et al.
242 1950). Although not through direct therma l measurements, t he thermoregulatory capacity of the caudal vascular bundle of the dolphin was shown by Elsner et al. (1974). Through anatomical and radiographic studies they showed that the vasculature of the tail of the dolphin is designed in such a way as to facilitate thermoregulation. This function is accomplished by either circumventing capillary beds in the tail via recruitment of arteriovenous anastomoses or by ut ilization of capillary beds and subsequent superficial venous return. T he first condition acts as a heat conserving mechanism that returns blood via the caudal vascular bundle while the second condition allows dolphins to dump heat to the environment. Sinc e manatees possess a caudal vascular bundle much like that seen in cetaceans (Rommel & Caplan, 2003), there is no reason to expect it to have a different function. The apparent similarities of the MAB and IVB argue also for a similar function. The dire ct thermal measurements in the limbs of sloths (Scholander and Krog, 1957)structures nearly identical to the vascular bundles found in manatees--established the thermoregul atory function of vascular bundles about which Scholander said No matter what other function they may have, they are by physical necessity heat exchangers and economizers of body heat (1958). Although an exhaustive comparison of t he vessel thicknesses, diameters and numbers of these structur es has not been conducted, the gross and microscopic similarities between the brachial and c audal vascular bundles to the MAB and IVB cannot be disregarded. When compared to the caudal vascular bundle, the MAB and IVB both exhibit nearly identical morpholog y, being composed of small arteries and veins packed tightly together, thereby providing a high degree of surface area for counter-current heat exchange. Given the lo w tolerance that manatees have for water
243 temperatures below 20oC, having heat-conserving structures in all the extremities that have high surface area to volume ratios and therefore a greater potential for convective heat losswould be a beneficial strategy for thermoregulation during exposure to sub optimal temperatures. Although the character of t he mandibular alveolar and infraorbital vascular bundles is quite similar, two structural differ ences stand out. Firstly, although the venous component of the MAB appears to have a similar degree of abundant venous anastomoses, the degree of arterial anastomoses is quite different. While the arteries of the MAB show very few anastomoses, th ose of the IVB have abundant anastomoses. The significance of this difference is unknown. It could presumably reflect a difference in either the developmental process or ph ysiological role. The second interesting difference between the two vascular bundles relates to the positioning of the accompanying nerves. Although two small nerv es are present along the periphery of the MAB, the main mandibular alveolar nerve is located in the very center of the MAB. Conversely, along much of the length of the IVB, the infraorbital nerves travel along the periphery of the IVB. In the more distal regions wher e the IVB exits the infraorbital foramen, the nerves take a more central path before branching out and innervating the soft tissues of the upper lips. Once again, the significance of this difference is not understood. If sensory modulat ion by regional heterothermy of the nerves is indeed possible, this positioning may reflect the thermal capacity of each vascular bundle. Since the MAB is considerably smaller in cross section and therefore contains fewer arteries and veins, the mandibular alveolar nerve may need to be centrally located in order to take advantage of the more modest thermal cap abilities of the MAB.
244 Conversely, the large size of the IVB may tr anslate into a greater thermal capability, so the associated nerves may benefit sufficiently from mere juxtaposition to the IVB and may not require a central location. It should however be noted that an extensive examination of the nerve loca tion relative to the vascular bundle was not conducted. Although it appeared that the ma ndibular alveolar nerve trav els through the center of the entire MAB, the IVB was not examined along numerous serial sections. Nonetheless, the distal portions of the in fraorbital nerves do exhibit a more central distribution, which may simply afford them with sufficient thermoregulatory capacity. Finally, given the likely difference in effi ciency of the two vascu lar bundles, it may be that the infraorbital nerve would be adver sely affected by being centrally located throughout the entire length of th e IVB. Such positioning within a very efficient thermal exchanger could translate into suboptima lly elevated temper atures. Prolonged recruitment of the IVB due to regional thermal needs coul d result in prolonged heating of the infraorbital nerve, which may be detrimental. By embedding the nerve in only a portion of the IVB, the tem perature of the nerve may be regulated more easily without having to alter the function of the IVB and ca use unwanted thermal effects to the tissues being affected by the IVB. Since the MAB is much less expansive and located in tissues likely to be more susceptible to co lder ambient temperatures, the mandibular alveolar nerve may need a central location throughout the entire ext ent of the MAB in order to maintain appropriate temperatures. Ophthalmic Plexus The above description of the ophthalmic ve n ous plexus constitutes the first mention of such a structure t hat I have been able to find. As such, there is no literature to which to compare it. Fortunately, the cons istent injection of t hat structure in my
245 specimens enabled a quality observati on of it and also implied t hat there are large or at least low resistance venous connections to it from the jugular system. Indeed, in all specimens I consistently found a large anas tomotic branch from the deep facial vein that appeared to be the primary filling route for the casting material. This finding and the consistently modest injection of the ophthalmic veins l eading to the calvarium are strongly suggestive that the main drainage r oute is likely via the deep facial vein. Interestingly, since this connection appears to be completely lacking of any venous valves, any significant oral and/or facial m anipulation may have the result of generating pressures which force blood in the facial and deep facial veins back into the ophthalmic plexus. I can only speculate as to the r epercussions of such a free flow mechanism, however it seems they could be to the benefit or detriment of t he animal. Clearly, elevated pressures could build in the ophthal mic plexus that could affect the proper functioning of the eye either through pressure effects on the optic nerve or physical damage to the globe itself. However, that does not appear to present much of a problem for the manatee, therefore I must consider this va lve-less venous feature to be beneficial or at the very least inconsequential. It seems likely that since there are no valves, pressurization of the ophthalmic plexus due to orofacial m anipulation would be only transient. As soon as relaxation wa s achieved, pressure would be released and blood would drain down the valve-less anastomotic branch. Should pressures in the ophthalmic plexus elevate to potentially harmf ul levels, the smaller but ever present connections to the veins of the infraorbital vascular bu ndle and dura should provide enough relief.
246 But what of potentially benef icial features of these connections? The large anastomosis with the deep facial vein on the si de of the face is connected to the veins of the face, which based on their size mu st handle considerable volumes of blood draining from the face and rost rum. Therefore blood from the face has the potential of being rerouted through the anastomotic branch of the deep facial vein into the ophthalmic plexus. This might allow modi fication of temperature or pressure surrounding the eye. There is significant ev idence that temperature can affect sensory tissues by affecting conduction speeds of action potentials, firing rates and amplitudes, and other features. Regional heterothe rmy of the eye has been suggested by numerous researchers w ho described ophthalmic retia in various avian, mammalian, and even tuna species (Aslan et al. 2005; Ninomiya, 2002; Ninomiya & Masui, 1999; Ninomiya & Yoshida, 2007) and could provid e a means for sensor y modulation of the eye, as is seen in certain tuna and some av ian species. Another possible effect of pressurizing the eye relates to intraoccula r pressures. It has been shown that in humans, increased intraoccular pressure can exert significant influence on the trigeminocardiac reflex (TCR) by way of t he oculo-cardiac reflex (Schaller, 2004). These influences are brought on by mechani cal stimulation of the ocular and/or periocular structures leading to stimulation of the ophthalmic branc h of the trigeminal nerve. According to Schaller (2004), the down stream effects of the trigeminocardiac reflex can lead to cardiac dysrhythmia up to asystole, arterial hypotension, apnea and gastric hypermotility. It undoubtedly seems lik e a stretch, however it may be possible for the manatee to actively or passively affect the trigeminocardiac reflex by modulating the ophthalmic plexus pressure and consequently the intraoccular pressure. It should
247 not be overlooked however, that elevated intraoccular pressures can have other downstream effects. Although the mechanisms underlying the effect are not well understood, Armaly and Araki (1975) showed t hat choroidal circulat ion within the brain drops during transient elevat ion of intraoccular pressure s in the cat and Rhesus monkey. Finally, a beneficial feature of this plexus may be that it is capable of trapping much of the heat emanated by the eye in order to return it to the core. The vascular bundles discussed above strongly suggest t hat the peripheral vasculature of the manatee is designed in such a way that it c an most efficiently conserve heat in all extremities. The ophthalmic plexus could form a functionally simila r structure that still allows adequate heating of t he eye. Unlike the vascular bundles found in all the extremities of the manatee, the ophthalmic plexus appears ent irely composed of veins. This might make sense when we consider opti mal temperature functioning for the eye. It seems reasonable to suggest that the extr emities that contain vascular bundles, namely the fluke, pectoral flippers, and rostrum, are invested mostly with skin, fat, and muscle. Conversely the eyes are sens itive sensory organs that likely require maintenance of optimal temperature for pr oper functioning. Since the sole thermal functional capability of a vascular bundle is to draw heat away from the outgoing arteries and return it via the incoming ve ins (Scholander, 1958; Scholander & Krog, 1957), a vascular bundle can only function by cooling the outgoing arte rial blood. Its only way of reducing the degree of arterial cool ing is by reducing the flow of either the outgoing warm arterial blood or the incoming cooled venous blood. Therefore, if a vascular bundle were to be the main structur e supplying blood to the eye, conservation
248 of heat at the eye would require either cooling of the arteri al blood supplying the eye or recruitment of a different venous return that would not draw heat away from the arteries. Use of the venous component of the vascular bundle could subsequently affect the eyes function by limiting the amount of heat that reaches it. Therefore, what the ophthalmic venous plexus might represent is an alternate heat conserving mechanism that attempts to balance the need for proper ocular heating and oxygen tensions with the need to conserve heat. Since the ophthalmi c plexus surrounds the eye, it might allow the eye to be supplied with warm, oxy genated arterial blood, while still absorbing much of the heat lost from the eye through convection. Brain Veins Like the veins in the dolphin, the m anatee braincase is invested with copious amounts of vasculariz ation, much of which pr oves to be venous in nature. I have been unable to find any such extreme investment described in the literature on terrestrial mammals and am therefore inclined to believe it may be unique to certain marine mammals. Indeed the picture that emerges is one of relatively complete bathing of the brain in a dura whose large part is compos ed of a venous plexus. It is tempting to suggest that this could offer some advantage for regional heterothe rmy of the brain, however I am in no position to evaluate such a potential given the lack of any functional or physiological data. Clearly such a t hermal function would be dependent on a number of factors, such as the tem perature differential between the dural venous plexus and the brain tissue, the rate of ex change or flow of the blood within the plexus, and the volume of blood within the plexus, to name a few. What I find interesting however is that much of the epidural venous rete that is connected to the dural plexus is also connected to the numerous intercostal vascular
249 bundles along the elongated pleural cavity (Fawcett, 1942). As described above, vascular bundles seem to have the inher ent capacity for temperature exchange between their arterial and venous counterparts Such temperatur e exchange could translate into modification of the blood within the venous portion of the vascular bundles, and subsequent flow of that blood into the epidural rete In addition to the potential thermal exchange bet ween the arterial and venous components of the intercostal vascular bundles, the fact that the bundles line the parietal pleurae of the lung cavities might conceivably mean that those vascular bundles are exposed to thermal influenced from the pleural cavities either through evaporative cooling or direct thermal convection and conduc tance from respired air. Although the manatee brain itse lf is quite lysencephalic, t he primary cerebral veins could still be identified with re lative certainty and, superfici ally at least, do not appear to warrant any further mention. One structur e however stood out beyond all other cerebral veins, namely the choroid plexus which was superbly enlarged. Though perhaps not entirely surprising given the sizable cerebral ventricles possessed by the manatee, it certainly begs the question of function. I s uppose the sheer size of the plexus could be a simple reflection of the size of t he ventricles and the consequent vascular requirements for adequate production of cerebr ospinal fluid. Though such a simple explanation may suffice, t he magnitude and complexity of this structure may warrant further study. Epidural Veins Interestingly, the cervi cal epidural vein s of domestic mammals show considerably different morphology than that found in t he manatee. Domestic mammals possess two main ventrolaterally located epidural veins ventral internal vertebral plexuses--that
250 occasionally connect with each other throug h ventral basivertebral veins and dorsal arcuate branches. Conversely, the manatee contains an intricate and voluminous venous rete that surrounds much of the spinal cord and is associated with a similar arterial component. Surprising ly, despite the extreme elaboration of the epidural veins of the manatee, there is a bilateral isolation of the cerv ical portions of the two intricate retia Dorsally, domestic mammals show little venous presence, having only one notable venous structure at the level of each intervertebral foramen. Forming in part from contributions from the intervertebral veins, these arcuate veins arch segmentally over the spinal cord to connect to the contra lateral ventral internal vertebral plexus and intervertebral vein (Reinhard et al. 1962). The spinal veins and epidural plexus re ceived very cursory mention by Murie (1874). In the Florida manatee, Fawcett ( 1942) found two spinal veins within the neural canal, ventral to the spinal cord Similarly, in the dugong Barnett et al. (1958) also found a pair of two spinal veins on the vent ral aspect of the spinal cord until they reached the cervical portion at which point they passed dorsal to the cord and entered the skull. They also noted t hat on cross section of the cerv ical vertebral column they found an even larger vein on the dorsal aspect of the cord, while more anteriorly all three veins were replaced by retial tissue that surrounded the cord. I cannot be certain of why Fawcett (1942) only m entioned the paired v entral spinal veins and omitted the dorsal vein or the epidural rete in the cranial most cervical region. I can only suggest that it could have been related to the qualit y of his specimens, their ontogenetic stage, or the cranial extent to which he may have limited his examination of the spinal cord. Since Fawcett (1942) stated that one of the fetal specimens was excellently preserved,
251 and his adult specimen was opportunistically ex amined following a physiological study, I can only assume that decomposit ional state was not a factor in his omission. It also seems unlikely that it was the result of ontogenetic variation since Fawcett (1942) examined an adult and two fetal manatee specimens and Barnett et al. (1958) observed the dorsal spinal vein in a dugong fetus co nsiderably smaller than the manatee fetuses examined by Fawcett (1942). I t herefore am led to believe th at Fawcett either chose not to mention or simply did not examine the cranial most re gions of the spinal cord. Even more confusing for me was t he fact that although the epidural rete that I found in the cervical portion of the Flori da manatee specimen did regress caudally, it appeared to extend through the entire cervical portion of the neural canal. Due to specimen limitations I was not able to determine at what point the rete coalesces in its entirety into the two ventral spinal veins, however, throughout the entire cervical portion of my specimen there were two large distinct ventral spinal veins visible (Figure 5-15). This is unlike what was observed by Barnett et al. (1958) in the dugong, in which the two ventral spinal veins curved dorsad when they reached the cervical portion of the spinal cord. Although these ve ins exhibited a similar dorsa l path, they did not obtain that path until they reached the atlanto-occipital joint. Addi tionally, when they did curve dorsad, only a dorsal branch of their bifurcat ion did so as it fused with the temporal dural sinus; the ventral one fusing with the epidur al plexus on the ventral aspect of the spinal cord (Figure 5-18). After curvi ng dorsad, the dorsal branch fused with the temporal dural sinus to form the confluens sinuum (Figure 5-19). Shortly before the temporal sinus fused with the confluens sinuum it was joined ventrally by the emissary vein of the jugular foramen. Therefore, t he two main drainage routes for blood in the
252 dural sinuses are in very close proximity. This might prove beneficial during altered vascular pressure profiles. If venous return through the emissary of the jugular foramen is impeded by collapse of the internal jugular vein, the blood can immediately be routed through the proximally connected branches of the ventral spinal vein and epidural plexus, thereby avoiding any transient pressu re elevations that could damage the CNS. It should also be noted that in the manat ee, I found that the two large ventral internal vertebral veins do not exist in isol ation during their course through the cervical neural canal. They are instead paralleled by numerous other sma ller yet still sizable veins which anastomose back and forth with eac h other and with the ventral internal vertebral veins. Theref ore the epidural venous rete is not composed of only small, coiled veins but rather a mix of large and sm all longitudinal veins that anastomose with each other, as well numerous much smaller re tial veins that bend back and forth in all directions and anastomose freely with adjacent veins. Interestingly, there was no mention by any of the aforem entioned authors of any bilatera l separation of the epidural rete As discussed previously, the ventra l aspect of the epidural arteriovenous rete remain segregated at the midline, e ffectively forming left and right epidural retia This separation was only oblit erated near the foramen magnum where the two retia fuse to become indistinguishable. Although it is tempting to consider these two retia as isolated from each other during the cervical segment, it should be recalled that their dorsal aspects are connected by segmental, yet fa irly sizable, dorsal arcuate veins. Nonetheless, it is known from domestic and laboratory species that the simple presence of vascular connections does not m ean that those connections are employed under normal circumstances. For instance, alt hough the arterial circle of Willis at the
253 base of the brain is a circular structure capabl e of allowing blood on the right side of the circle to flow into the left side of the brai n, under normal circumstances that is not the case. Instead, blood supply to each hemispher e is accomplished through the ipsilateral portion of the circle of Willis, and this holds true even in ungulates with a carotid rete (Baldwin & Bell, 1963a, b, c; Gillilan, 1958; 1974). Only when occlusion or impairment of flow occurs in one parent artery of the ci rcle, does blood flow from one portion of the circle of Willis into the contralateral hemis phere (Baldwin & Bell, 1963b,c). The ventral midline segregation of the cerv ical portion of the epidural rete may have a similar rheological separation under normal circumstances, which is only altered when pressure profiles are altered. As with the cetaceans, I am co mpelled to suggest that the elaborate epidural and intracranial venous retia of the manatee serve a higher purpose than simple drainage of blood from the CNS as that drainage could conceivably occur through simple linear pathways like those seen in true seals. I believe their high surface area and intimate association with the CNS provides support for t he notion of regional thermal influences on the CNS. Synopsis It is at times hard to fathom that such elabor ate structures have remained largely ignored or un-described unt il now. I find their voluminous nature to be suggestive of notable energetic cost to build and maintain, and am therefore inclined to believe they serve important physiological roles. Wi thout physiological experiments one can only speculate as to their function, however, the degree of organization I encountered in these structures provides ample ammunition for arguing vehemently against any notion of a pointless presence or haphazard construc tion. Given the present technology available to us and the high quality of m anatee specimens that can currently be
254 obtained, I venture to suggest that the mi ssing details of some of the vascular morphology (e.g. thoracic spinal circulation) could be discovered. Additionally, modern technology such as infrared thermography and magnetic resonance imaging could go a long way toward elucidating some of the functional or physiol ogical implications of the structures described herein. The intimate association of some of these elaborate vascular structures with areas that first manifest cold st ress lesions suggests that a connection between the potential thermoregulatory function of the vascular structures and the pathophysiology of cold stress syndrom e may exist. I therefore believe that further research into the transient and chronic adaptive thermal responses of the vasculature in the region of the face may shed light on the mechanisms underlying some of the observed cold stress lesions. Table 5-1. List of specim ens used for this research. Specimen ID TBL (cm) Gender Research Use LPZ102654 275 F A, G LPZ102900 251 M A, G MSTm1001 302 M A, G LPZ102904 315 F G LPZ102962 307 F C LPZ102961 209 M Histology A=Angiography; G=Gross dissection; C=Corrosion casting Table 5-2. List of structure labels and their names. Structure label Structure name 1 Mental v. & a. 2 Lingual v. and a. 3 Facial v. 4 Maxillary v. a. 5 Arteriovenous triads of IVB 6 Angularis oculi v. 7 Maxillary labial v. 8 Ophthalmic plexus 8' Anastomoses to IVB
255 Table 5-2. Continued. Structure label Structure name 9 Anastomosis from deep facial v. 10 Facial v. continuation 11 Mandibular labial v. 12 Angularis oris v. 13 Ophthalmic v. 13' Plexus around ophthalmic v. 14 Epidural rete 15 Transverse dural sinus 16 Sigmoid sinus 17 Dorsal sagittal sinus 18 Confluens sinuum 19 Meningeal v. 20 Straight sinus 21 V. of corpus callosum 22 Choroid v. 22' Caudal terminus of choroid plexus 23 Ventral spinal v. 24 Intervertebral v. 25 Radicular & medullary v. 26 Arcuate v. 27 Temporal sinus 28 Emissary of jugular foramen
256 Figure 5-1. Medial view of mandibular canal with medial wall removed to show arteriovenous nature of MAB. The inset in the top le ft shows a close-up of the arteries and veins of the MAB. Figure 5-2. Cross sectional view of the mandibular va scular bundle showing arteries (red) and veins (blue) injected with latex. Note the general lack of arterial anastomoses and innumerable venous anastomoses. Also note the central presence of the mandibular alveolar nerve.
257 Figure 5-3. Medial view of corrosion cast of veins of the left side of the manatee head showing veins of MAB investing soft tiss ues of mandible as mental veins. Note that the predominance of red vessels is an artifact of the non-viscous casting material crossing capillary beds.
258 Figure 5-4. Medial view of the right half of a mid-sagittally se ctioned manatee head with the lateral wall of the bony naris remov ed to visualize the IVB and its drainage into the maxillary vein (4). The rostrum is to the left. The inset in the top left corner shows a close-up of the arteries and veins of the IVB.
259 Figure 5-5. Cross sectional view of the infraorbital vascular bundle showing arteries (red) and veins (blue) injected with latex. Note how the venous anastomoses form a venous lake-like stru cture that surrounds the arteries and that the largest arteries appear toward the center of the bundle. Unlike the MAB, numerous arterial anastomoses are visible. Also note the infraorbital nerve seen as a brown structure on the top right.
260 Figure 5-6. Photomicrograph of a Masons tr ichrome histologic pre paration of a portion of the IVB in cross section. Note that the relatively large arteries (A) are surrounded by similarly sized veins (V). The inset on the bottom left shows a histologic cross-sectional view of the CVB for comparison. Note the presumptive arterial anastomosis (Aa) on the top left of the CVB. The degree of arterial (A) contraction is visibly gr eater in the CVB. Both specimens were recovered from the same animal and were treated similarly; therefore the reason for the contracture may be func tionally significant. Nonetheless, the pattern similarity is evident. The sca le bars represent 1 mm for the IVB, and 500 m for the CVB.
261 Figure 5-7. Photomicrograph of a Masons tr ichrome histologic pre paration of a portion of the IVB in cross section. Note how the nerve clus ter (N) in the center is surrounded by a roughly annular structure co mposed of small arteries (A) and veins (V). A venous anastomosis (Va) can be seen on the right. The vessels surrounding the nerve cluster appear much smaller than those in Figure 5-6 that do not directly surround nerves. The large separation between the vessels and the top portion of the nerves is a histologic processing artifact.
262 Figure 5-8. Right dorsolateral view of the maxillary lips showing the distal branches of the IVB emerging from the infraorbital fo ramen and investing the lips in triads (probe tip) composed of a central artery with two anastomosing satellite veins (5). Rostral is to the ri ght. Caudal is to the left.
263 Figure 5-9. Left lateral view of ophthal mic venous plexus of the Florida manatee injected with liquid latex. Note the large sinusoid veins that compose parts of the plexus (8). Also visible are t he numerous small ventral anastomoses with the IVB (8), and the single large lateral anastomosis with the deep facial vein (9).
264 Figure 5-10. Left lateral view of the left ey e showing the investment of the ophthalmic venous plexus (8) in relation to the ex trinsic eye muscles. The plexus has been separated midlaterad and reflected dorsally and ventrally to expose the muscles. The large anastomotic branch of the deep facial vein (9) can be seen fusing with the ventral external opht halmic vein at the ventral margin of the ophthalmic plexus.
265 Figure 5-11. Dorsomedial close-up view of the proximal portion of the left ophthalmic vein (13) on its coarse to the calvarium. Note the presence of a central vein (13) surrounded by a plexus (13) co mposed of fine caliber veins. The infraorbital vascular bundle is visible just ventral. Caudal is to the left. Rostral is to the right.
266 Figure 5-12. Medial view of the right half of a mid-sagi ttaly sectioned manatee showing bifurcation (3) of ophthalmic vein into the cavernous sinus (8) and durae (7) within the calvarium. Note that paper labels in picture do not represent structure labels from Table 5-2.
267 Figure 5-13. Oblique dorsocaudal view of a laminectomized manatee head showing the intact dura on either side of the dors al sagittal sinus (17) and covering the brain. Note the extensive venous invest ment of the dura. An excised cross section of the spinal cord with its extensive epidural rete (14) at the level of the foramen magnum can be seen on the left.
268 Figure 5-14. View of an excised manatee br ain with the dura covering the let cerebral hemisphere reflected. Note the ext ensive venous meningeal branches (19) perforating the cerebrum. This patte rn (in some places even denser) was observed throughout the entire surface of the brain.
269 Figure 5-15. Medial view of the right brain hemisphere of a mid-sagittaly sectioned manatee brain, showing extensive dural vasculature covering the dorsomedial surface of the right brai n hemisphere. Also visi ble are the choroid veins forming the internal cerebral veins (22) that along with the median vein of the corpus callosum (21) drain in to the straight sinus (20).
270 Figure 5-16. Medial view of the right brain hemisphere of a manatee showing a partially exposed choroid plexus. For comparis on, the corrosion cast of a choroid plexus has been placed at the bot tom, showing the horseshoe-shaped internal cerebral vein (22) becom ing the choroid vein and tapering caudolaterad (22).
271 Figure 5-17. Cross-sectional view of the spinal cord and epidural rete at the level of the occipital condyles, showing extensive v entral and ventrolateral investment of arteries and veins. Figure 5-18. Ventral view of cervical epidural rete (14) showing enlarged ventral internal vertebral veins (23) along the periphery and a distinct midline separation of the left and right sides of the rete except in the most cranial region (left). Note the large segmental connections to the intervertebral veins (24).
272 Figure 5-19. Ventral view of t he spinal cord with the epidural rete reflected along its natural ventral division. Note the sm ooth surfaces and lack of communicating veins or arteries between the two sides. Numerous radicular and medullary branches (25) can be seen entering subdur ally with the ventral nerve roots. Figure 5-20. Dorsal view of the cervical epidural rete showing arcuate veins (26) bridging the two sides of the rete and dorsal longitudinal anastomoses connecting arcuate veins. On th e left side, note that as the rete enters the foramen magnum, the dorsal rete becomes much more diffuse.
273 Figure 5-21. Left lateral view of an excised manatee brain showing the ventral internal vertebral vein (23) bifurcating on the left. Note that the dorsal branch fuses with the sigmoid sinus (16). The seve red vessel directed ventrad is the emissary of the jugular fora men (28) that drains into the internal jugular vein. Structures 23 and 28 repres ent the main drainage routes for blood in the dural sinuses.
274 Figure 5-22. Dorsal view of an excised m anatee brain with inta ct durae showing the connections of the dorsal branc hes of the ventral internal vertebral veins (23) with the sigmoid sinuses (16). The sigmoid sinus then connects to the temporal (27) and transverse (15) si nuses which together with the dorsal sagittal dural sinus form the confluens sinuum (18).
275 Figure 5-23. Composite image showing location and morphology of the IVB, MAB, BVB, ICVB, and CVB. Note the whole body distribution and marked presence of peripheral vascu lar bundles. Image of CVB and ICVM were used with permission by the authors: Rommel & Caplan (2003).
276 CHAPTER 6 DISCUSSION In the preceding chapters I have presented t he observations I made from a fai rly limited number of Florida manatee and bottlenos e dolphin specimens. It comes as no surprise that due to the difficulties of postmortem vascular in jections and the substantial vascular complexity I encountered, I was not able to identify all of t he finer ramifications of the veins in certain regions. Nonethele ss, numerous rather uni que structures were identified and/or expo unded upon, some of which prov ide tantalizing functional implications. Perhaps most notable of those were the infrao rbital and mandibular vascular bundles of the manatee and the v enous plexuses of the accessory sinus system of dolphins, however the extensive venous investment of the calvarium and neural canal in both species also raises interesting questions. Comparative Venous Morphology In humans, the facial vein typically drains into the internal jugular vein. This arrangement is unlike the condition seen in domestic mammal species, in whic h the facial vein is usually a branch of the linguof acial vein, a major tri butary of the external jugular vein. Despite this differing anatomy found typically in humans, numerous variations exist in which the facial vein drains either directly into t he external jugular vein or into the subclavian vein (Gupta et al. 2003). In those cases where the facial vein drains into the subclavian vein, anastomoses are still seen between the facial and external jugular veins. It should be noted t hat even in humans whose facial veins drain normally into the internal jugular veins, a sizable anastomosis usually exists between the external jugular vein and either the internal jugular or facial vein (Gupta et al. 2003). This seems especially relevant for two reasons. Firstly, it provides a tangible way of
277 conceptualizing the ontogenetic and phylogenetic mechanisms underlying this atypical mammalian arrangement. Secondly, it helps us better explain the numerous connections witnessed between the internal and external jugular and facial veins in manatees and dolphins, since the variability found in humans suggests a considerable inherent plasticity in the formation of t hese connections. Nonet heless, it appears as though manatees and dolphins have elaborated on th is character. Indeed, the general pattern that is observed is that thei r venous system forms anastomoses throughout much of the head and neck. Humans are not unique in their possessi on of anastomoses between the internal and external jugular veins. All domest ic mammals examined show a degree of anastomosis between all of the veins of the head (Ghoshal et al. 1981; Schaller, 2007; Schummer et al. 1981). These anastomoses occu r both between ipsilateral and contralateral veins so that the final pattern is of a venous system with multiple drainage paths. Some of these drainage paths have been elicited through experimental differential ligation of vessels. Such experiments have shown that anastomoses are not only a character of the venous system, but ar e also present in the arterial system (Clendenin and Conrad. 1979; Moss, 1974; Whisnant, 1956). What is interesting, however, is that many of these anastomoses appear to be relatively modest in domestic mammals. During experimental ligation that results in altered vascular pressure profiles, these small preexis ting anastomoses are progre ssively recruited as the obstruction to normal blood flow persists. Chronic recruitment of these anastomoses results in permanent establishment of some of those pathways ( Clendenin and Conrad. 1979; Moss, 1974; Whisnant, 1956). These feat ures all suggest that the vascular
278 system of the domestic mammals is designed wit h a type of hierarchical order in mind. Well-established primary paths ar e consistently responsible for providing specific supply and drainage fields for target tissues, while the various smaller vessels and anastomoses exist as ancillary collateral paths that facilitate supply and drainage during rare moments that the main drainage paths are obstructed or pressure profiles are altered. Interestingly, the intricate retia found in domestic mammals and marine mammals can occasionally be found, in one form or another, in humans. Henkes et al. (2007) reported on two cases of humans with agenesis of one internal carotid artery showing formation of a collateral carotid rete mirabile connecting the ipsilateral external carotid artery to the intracranial in ternal carotid artery. Concom itant hypoplasia of the carotid canal was also reported. Similarly, Kim et al. (2006) showed presence of bilateral carotid rete compensation in response to occlusion of the extracranial internal carotid arteries and the basilar artery, and coarctati on of the aorta is known to result in establishment of collateral thor acic arterial pathways (Kirks et al. 1986). Although the mechanisms of formation are not always unders tood, it is known that human carotid rete mirabile does not exist in any ontogenetic st age and is consequently not considered a developmental remnant. Therefore, t hese examples show that a carotid rete mirabile can form in a human as a response to a pathol ogic absence of normal internal carotid flow or as an atavistic phenomenon (Konno et al. 2001). Species-specific Impli cations (Dolphins) The anastomoses seen between the internal and external jugular veins of the dolphin wer e consistently present in the specimens I examined. Although the specimen number was small, such consistency suggests t hat they are not artifactual or pathologic,
279 but rather a normal occurrenc e. Notably, countless sizable anastomoses were found on every major vein in the cervical, gullar, fa cial and temporoccipital regions suggesting that venous anastomoses are not only common, but also even ubiquitous. Therefore, it is perhaps reasonable to suggest that unlike terrestrial mammals, which possess relatively modest numbers of large anastomose s that are likely employed sporadically, dolphins have proliferated their anastomoses in order to compen sate for regularly encountered venous obstructions. Such obstructions might be encountered during locomotion and diving. A well-known example of muscular cont raction causing obstruction of venous return is the Valsalva phenomenon, in which elevated intrabdominal pressures result in collapse of the abdominal vena ca va and impaired venous return In humans, activities that can elicit the Valsalva phenomenon are coughing, lifting of heavy objects, and other activities that require strong and/or sustained contraction of the abdominal musculature. Interestingly, dolphins are axial rather than appendicular locomotors and therefore regularly employ their trunk musculature (e.g. m. longissimus, iliocostalis, and rectus abdominus ) in order to propel themselves through the water. Since trunk muscles often do not work in isolation, it is likely t hat the muscles of t he neck and throat (e.g. m. sternohyoideus and sternocephalicus ) also play an important ro le in locomotion. Given the morphology of the venous system of the head and neck of t he dolphin and its associations with the muscles of the region, such regular contraction may conceivably result in regular obstruction of venous retu rn. The ubiquitous presence of collateral anastomoses may therefore provide a mechanism for alleviating the large but transient
280 obstructions. Such a function may help ex plain the presence of such large and ubiquitous anastomoses compared to thos e seen in domestic mammals and humans. Another situation commonly encountered by dolphins is exposure to elevated hydrostatic pressures of diving. Much of their life is spent underwater, and they are capable of attaining depths greater than 300m (Harrison & Kooyman, 1971; Ridgway & Harrison, 1986; Ridgway & Howard, 1979). Such pressures undoubtedly exert considerable influences on venous return, espec ially in places such as the ventral skull where air-fiiled sinuses exist. As with the lo comotion-induced collater alization that may occur, the numerous venous channels may also provide relief from the effects of divingrelated compression, essentially limiti ng or avoiding venous hypertension and subsequent tissue edema. Given the sizable c onnections of the veins of the head to the intracranial veins, such a mechanism might also be beneficial fo r avoiding hypertensive damage to the brain. In addition to the many venous anastomoses, other intricate venous structures of great interest were observed. As was di scussed in Chapter 3, the accessory air-filled sinus system of the bottlenose dolphin is line d and/or associated with an intricate venous plexus that surrounds and/or abuts almost all of the air spaces. So co-extensive is the venous investment of those air spaces that they seem inexorably linked. Any structural and/or functional modification in one structure would likely be followed by a concomitant change in the ot her. Indeed, Fraser and Purv es (1960) noted that the general impression of the vascular system in t he region of the base of the skull is of an elaborate plexus of vessels investing the whole of the air sac system, and apparently entirely subservient to the proper functioning of the latte r. The argument could be
281 made that there exists a rat her reciprocal relationship bet ween the two structures, each responding to changes in the other. For in stance, an increase in air volume or air pressure within the sinus system could result in a reduction of blood within the sinus plexuses, while a reduction in air volume would conceivably have the opposite effect, causing redistribution of blood into the plexus. Conversely, an increase in the blood volume within the sinus plexuses could result in pressurization of the sinus system. We currently have no functional concept of the rheological intricacies of the plexuses in the dolphin head, so speculation is unavoidable. Nonetheless, based on what we know from other mammals, some hypotheses might be garnered. Clearly the bl ood in the plexuses of the accessory sinus system is venous and since no arteriovenous anastomoses were found likely to be postcapillary, low pressure blood. Without the vis a tergo force from the heart and with an apparent lack of arteriovenous sh unts in the region, it woul d seem as though the venous pressure within the plexus is a product of a few key forces, namely the central venous pressure, the pressure afforded by the mu scular complex associated with the sinus system (e.g. m. pterygoideus internus and externus ) and the pressure of the gas within the sinus system. The central venous pressure is predominantly a function of the venous blood volume and the degree of venous compliance (e .g. sympathetic vasoconstrictive tone). Venous blood volume is in turn influenced by numerous variables such as cardiac output, pulmonary perfusion rate, contraction of skeletal muscles, and hydrostatic forces such as gravity. Given the lack of physi ological data for dolphins and the frequent physiological alterations they undergo wit h diving, it may be unwise to make
282 assumptions about some of af orementioned variables. What we do know, however, is that cetaceans spend much of their time diving, and during diving many of the aforementioned variables show a certain degr ee of alteration (Scholander, 1940; Elsner, 1966). Although some anatomical features such as the elastic aortic bulb are thought to provide some ancillary pressure drive during diastole, they are likely not sufficient to maintain the pressure at the levels gener ated during the normal cardiac cycle at the surface. Additionally, many dives in volve extended apneustic periods during which intrapleural pressures are presumably more negat ive. This would tend to create a more negative pressure in the right side of the he art and the thoracic ve ins, driving venous blood into the heart and reducing central v enous pressure. Conversely, the general lack of a gravitational force while in the wa ter likely increases central venous pressure by allowing more peripheral venous blood to exert its force rather than pooling in the extremities, similar to what happens to pilots when the g-forces are alleviated. All of these and the countless other factors result in a very complex picture regarding venous pressure in the dolphin and therefore offer mixed insights into what the pressures might be in the venous plexus of the accessory si nus system. Ignoring, by necessity, these complexities, a few factors c an reasonably be accounted for. It has long been known that venous return from the periphery is aided by movement of limbs and contraction of skeletal muscles (Broderick et al. 2010; Eisele et al. 2001; Hasegawa et al. 2011; McNally et al. 1997; Roberts et al. 1971). Venous blood in the limbs of mammals is pushed against gravity toward the co re by intermittent contraction of the limb muscles surrounding the peripheral veins. This action is
283 facilitated by the presence of venous valv es in the extremities that limit the gravitationally driven backflow of blood to the distal extremities. Such a system ensures unidirectional flow of blood from the distal extremities to the proximal extremities and then subsequently to the central venous syst em (Boisseau, 1997). Interestingly, cetaceans are notoriously lacking of such valves, and I have found no such valves in the head of the bottlenose dolphin, suggesti ng that the venous plexuses of the accessory sinus system composed of a valveless bi-directional venous system. This may seem like a trivial consideration given the fact that an aquat ic lifestyle affords alleviation from gravitys influence on venous return, however it comes with what seems like a fairly substantia l inherent implication. The lack of venous valves within the pl exuses of the accessory sinus system means that blood within that system could re spond to driving forces by traveling in either direction, retrograde or anterograde. Therefore, any contraction of muscles adjacent to the venous plexuses can generate a driving force for the blood within the plexuses, and that blood may be driven in either direction--proximally toward the jugular veins and heart or distally toward the face and brain. The same could be said of the effects of movement or def ormation of non-muscular struct ures (e.g. sinuses, tendons, connective tissues, mandible) adjacent to the venous plexuses. Given the lack of gravitational drive and venous valves, the aforementioned effects could be profound. Compounding these effects may be the fact t hat contraction of a muscle or deformation of a non-muscular structure may also result in obstruction of venous outflow in certain areas, resulting in the ability to pressurize the venous syst em upstream of the obstruction (Brescher, 1958; Oshima et al. 2007).
284 Such an obstruction would be similar to providing artificial compression to an extremity, thereby resulting in poolin g of blood and tissue edema distal to the obstruction. Contraction of the ventral neck muscles (e.g. m. sternohyoideus and sternomastoideus ) of the dolphin may result in compression and potential collapse of the jugular veins. Combined with relaxation of the pterygoid muscles, blood would likely pool inside the plexuses of the accessory sinus system (Oshima et al. 2007). Given the relatively modest collateral drainage pathway s (e.g. infraorbital veins and emissary veins of foramen ovale) available to blood in the plexuses, dolphins may be able to pool and subsequently pressurize the blood within the plexus by subsequent contraction of the pterygoid muscles. Interestingly, the pterygoid venous plexus has been implicated in alleviation of symptoms of patulous Eustachian tube (ET) in humans, a condition caused by incomplete closure of the ET (Oshima et al. 2007). Indeed, Oshima et al. showed that neck compressioneliciting venous pooling in the headcaused expansion of the pterygoid muscles and pterygoid venous plex us as viewed through magnetic resonance imaging (MRI). Upon endoscopic examination, the authors found that the anterior wall of the ET lumen adjacent to the pharyngeal orif ice protruded, and this change occurred within a few seconds of neck compression. These findings lead Oshima et al. (2007) to suggest that closure of the ET may be at least partially modulated by the degree of filling of the pterygoid venous plexus. Given the voluminous nature of the pter ygoid venous plexus in dolphins and the presumed need of diving mammals have for a high degree of control over Eustachian tube patency, these findings ar e hard to ignore. The lar ge nature of the pterygoid
285 venous plexus combined with the large pterygoid muscles may afford the dolphin the ability to control the amount of air that passes through the ET during a dive when barometric pressures may be driving air fr om the relatively incompressible upper respiratory tract found inside the bony nares to the compressible accessory sinus system to. Since the air in the upper resp iratory tract is needed fo r echolocation, the motivation to control the volume of air in thos e spaces is likely quite high. Interestingly, the findings of a fairly voluminous venous plexus within the substance of the palatopharyngeal muscle may support such a function, since engo rgement of this muscle could aid in sealing of the distal ET as it travels across the lateral aspect of the palatopharyngeal and salpingo-pharyngeus muscles to exit just above the palatopharyngeal sphincter (Fraser & Purves 1960). As mentioned earlier, Fraser and Purves believed that the vascular system of the accessory sinus system was tied to the proper functioning of the sinuses. As the ET is directly connected to the accessory sinus system, and occupies adjacent spaces, it may be affected by t he degree of filling of the venous plexuses associated with the region. In the interest of parsimony, it is nece ssary to mention that there are undoubtedly alternative explanations to t he functions of some of these structures. For instance, the venous investment of the pal atopharyngeal muscle might simp ly represent an increased perfusion and therefore drainage of a muscle that is employed so consistently and therefore has elevated metabolic byproducts that require clearance. Similarly, the venous plexuses of the accessory sinus system could merely represent a passive blood storing structure rather than a dynamically responsive pressure regulating mechanism.
286 There are currently no physiological data av ailable to begin to assess any of these potential roles, and until such data exists, t hese roles will remain purely speculative. Species-specific Imp lications (Manatees) As was found in the dolphin, manatees also possess many sizable venous anastomoses throughout the venous system of the head and neck. The maxillary vein presents as a single vein juxtaposed by countless anas tomosing veins that form a large plexus that surrounds the exte rnal carotid artery. The p haryngeal and pterygoid regions are dominated by intricate plexuses of anasto mosing veins whose distal tributaries also connect back and forth at numerous locations. Similar to the dolphin, it seems as though these multitudinous connections may serve the purpose of forming collateral drainage paths. Although manatees do not possess the elaborate accessory sinus system of dolphins, they do have considerable facial dexterity. Since they use their dexterous lips to manipulate food and investi gate objects, they have considerable facial mobility facilitated by extensive muscular inve stment. Use of these muscles and their associated connective tissues may result in transient obstruction of venous return, and the numerous collateral pathways may facilitat e alternate venous flow during times of altered regional venous pressure profiles. As mentioned in the introduction, Heged us and Shackleford (1965) stated that mammals which developed the structures of the face, nose, mouth, and ears to a greater degree showed greater dominance of external jugula r drainage of the brain. Interestingly, this was not found to be the case for either the dolphin or manatee. Temporarily ignoring the potent ial for hemodynamic effects on blood flow from the brain and focusing solely on the morphology of t he venous connections paints a different picture for the dolphin and manatee. In the dolphin, the caliber of the emissary veins of
287 the foramina ovaliatributaries of the external jugular veinsappears roughly equal to the emissary veins of the jugular foramina of the internal jugular veins. While dolphins may have reduced certain facial structures like the external pinnae, the melon represents a substantial investm ent of metabolically active ti ssue. I theref ore believe it appropriate to argue that the fa cial structures are as subs tantialif not moreas those referred to by Hegedus and Shackleford (1965). Similarly, the manatee brain had roughly equal--though less direct--caliber drainage paths to the external jugular ve ins via the anastomotic branch of the deep facial vein to the ophthalmic plexus, as it did to the in ternal jugular veins via the emissaries of the jugular fo ramina. Although manatees have al so lost their external ear pinnae, their lips represent a significantly voluminous and ac tive facial region. It therefore appears as though the notion that external jugular vein dominance is concomitant with facial development may not hold true for dolphins and manatees. This discrepancy might be explained in two complim entary ways. Firstly, the numerous anastomoses between the internal and external jugular system may have resulted in a functional equilibration of dr ainage duties of the two venous systems. If each of the venous systems represents a collateral drai nage path for the ot her system and they both regularly take on variable drainage ro les based on transient regional pressure profiles, they may ultimately share equal or similar roles in brain drainage. Secondly, the voluminous epidural venous plexus found in the neural canal provides a large valveless drainage route for blood returning from the brain and may therefore reduce the need for jugular drainage of the brain. This reduced need for jugular drainage may manifest as a similar contribution from eac h jugular system. It should be noted that
288 although the jugular contributions to brai n drainage may not fit the domestic mammalian pattern perfectly, the facial venous drai nage of dolphins and manatees was remarkably consistent with that of domestic mammals, despite some of the facial modifications (Figure 6-1). Thermoregulation (brain) The thermal biology of marine mammals is an exceedingly important concept when disc ussing their ability to cope with an aquatic lifestyle. In addition to their need to maintain an overall thermal balance that allows them to cope with increased convective heat loss to the water, they also face ot her challenges that ma y be met in part by specialized thermoregulatory adaptations. Ex amples of such other challenges include protracted apneustic periods and thermal sensit ivities of the reproductive tract. Numerous studies have illustrated the pres ence of vascular structures designed for maintenance of appropriate temperatures, whet her regional (e.g. reproductive tract) or whole body temperatures (Pabst et al 1995; Rommel & Caplan, 2003; Rommel et al. 1992; 1993; 1995, 2001; Scholan der, 1958; Elsner; 1966). What has not been clarified however is whether there is a connection between thermoregulation and diving physiology. Evidence regarding body temperatures and diving in marine mammals is sparse (Odden et al 1999). Furthermore, what data exists is reflective of whole body temperature or at least core temperatures and therefore does not provide any insights regarding the potent ial for regional heterothermy. This is important with respect to the current discussion because there is ample evidence showing the regional het erothermy of certain parts of the central nervous system can elicit profound physiological cascades throughout the whole body. Specifically, regions of the spinal cord and brain have been shown to respond to thermal stimuli by altering
289 key physiological mechanism such as periphera l vascular tone, respiration, metabolism, etc. (Carlisle & Ingram, 1973; Hammel et al. 1976; Jessen and Mayer, 1971; Kullmann et al. 1970; Miller & South, 1979; Simon, 1974; 1981; stnes & Bech, 1992; Wnnenberg, 1973; 1983). Wnnenberg (1973) show ed that electrical activity in the hypothalamus of guinea pigs was significantly increased by temperature elevations of only a couple degrees. stnes and Bech (1992) noted that cooling of the thoracic and cervical spinal cord of pigeons resulted in increases in metabolic heat production and body temperature, suggesting the presence of thermosensitive elements within the spinal cord. Interestingly, Hammel et al. (1977) found harbor seals to show different responses to hypothalamic temperature alterations compared to dogs and Adelie penguins (Hammel et al. 1976). Although in all thr ee species ambient and core temperatures exert an effect on the response to hypothalamic cooling, the effect is different. With lower ambient and/or core body temperatur es, the threshold temperature at which hypothalamic cooling elicits increased oxygen consumption is elevated in the dog and penguin. Conversely, in the harbor seal the threshold appears unaffected by ambient and core temperatur es. Instead, the rate at which oxygen consumption increases with dropping hypothalamic temperature is elevated when ambient or core temperatures are reduced. Jessen and Mayer (1971) found that the spinal cord and hypothalamus are roughly equivalent core s ensors of temperatur e in the dog, with cooling of either structure causing elevation of heat pr oduction through shivering and reduced respiratory evaporative heat loss. Holom et al (2008) showed that in rats, acti vation of temperature-sensitive meningeal afferents can regulate the blood flow within the dura matter through a
290 sympathetic reflect. This is especially inte resting to me given the extensive dural and meningeal vascularization observed in manatees and dolphins. If like in the rat (Holom et al 2008) warming of the dura elicits increa ses in dural blood flow, this might represent the situation that occurs upon term ination of a dive during which clearance of metabolic byproducts is necessary. Elevat ion of temperature within the epidural plexuses through either routing of warm bloo d or cessation of routing of cooled blood, may help elevate the temperature and in turn e licit a concomitant elevation in blood flow and cerebral perfusion. Additionally, Simon (1974) cited numerous transection experiments in which interruption of the spinal anterolateral tracts of the spinal cord abolished or reduced thermoregulatory effector responses to spinal thermal stimulation providing that the ascending sp inal neurons in the ventrolate ral tracts were responsible for conduction of a significant portion of t he thermal sensory stimulus. I found this interesting given that a majority of the epidural retia of the dolphin and manatee are in fact located along the ventral and ventrolateral aspects of the spinal cord, in close juxtaposition to the ventrolatera l tracts. Interestingly, the cold sensitive units within the ventrolateral tract of terrestrial mammals appear to be fewer and less sensitive to cold stimulation than the warm sensitive units (Gregor et al. 1976; Simon and Ir iki, 1971). It is not currently known if this condition is similar in marine mammals. Nonetheless, it appears as though some diversity in response to thermal stimulation of the CNS may exist. Such diversity may be related to spec ific adaptations for coping with a species ecological niche. An equally important component of regional heterothe rmy of the CNS is the potential neuroprotective benefits of cooling of neural ti ssues. The extended dive times
291 experienced by some mari ne mammals bring into question the degree of ischemic and/or hypoxic exposure of the CNS. Marsala et al (1993) showed substantial spinal cord protection from ischemic damage affor ded by panmyelic epidural cooling in dogs. The role that selective brain cooling may play in marine mammals is currently unknown. There exists considerable cont roversy regarding the existence and role of selective brain cooling in certain mamma ls and even birds (Maloney & Mitchell, 1997; Maloney et al. 2007; Mitchell et al. 2009; Fuller et al. 2003), however it is generally accepted that certain mammals use vascula r structures in order to modify the temperature of the brain and mitigate the damaging events of brain hyperthermia. Mizunuma et al (2009) that neurons in the human br ain are much more sensitive to transient elevated temperatures than they are to acute hyperbaric conditions. Though we have no functional data to support or re ject the existence of selective brain temperature control in dolphins and manatees, some of the vascular structures and their associations provide tantalizing hints of su ch potential. As ment ioned above, regional heterothermy of the brain may serve not only a thermal protective mechanism but may also modulate global physiologic mechanisms such as blood perfusion, heart rate and respiration. For instance, the connections between the infraorbital vascular bundles (IVB) and cavernous sinuses of manatees suggest, from a structural standpoint, the potential for regional heterothermy. The structure of the infraorbital foramina appears nearly identical to that of the caudal vascular bundl e located in the chevron canal, a structure widely accepted to serve a heat conserving f unction. By juxtaposing a large number of fine caliber warm arteries to an equal or gr eater number of small, cool veins, the
292 outgoing arterial heat can be trapped and returned by the incoming venous counterpart. Such a system would allow warmth to capt ured and returned the core by whichever path may be available. Though a single large path to the brain does not appear to exist from the IVB, the venous com ponent of the IVB connects to the cavernous sinus at mid length via at least two dor sal anastomoses wit h the ophthalmic venous plexus, and proximally via plexiform connec tions from the maxillary ve in that follow the cranial nerves that exit the ventral skull. Both of these routes provide significant potential routes that could drain into t he cavernous sinus at the base of the brain. An additional connection may be via the highly convoluted epidur al venous plexus that is intimately juxtaposed to an arterial counterpart. This high surface area arteriovenous structure may also be capable of considerable heat ex change and therefore play a role in selective brain control. Additionally, the venous epidural plexus of the manatee surrounds the cervical spinal cord and extends into the calvarium in association with the brain stem, both regions of the CNS know n to be influenced by heating and cooling. It should be clear at this point that t he aforementioned struct ures found in the Florida manatee would facilitate brain warming, a concept that is completely counter to the generic mammalian paradigm of brain cooling. At the risk of proposing a complete paradigm shift, I postulate t hat the manatee might be considered a viable candidate for such a unique strategy due to its rather unique problems with maintaining a thermal balance. Even in the relatively warm subtro pical waters of Florida, manatees find themselves at a therma l deficit during winter months. So much so that in some years hundreds of manatee deaths are a ttributed to acute and chronic cold exposure. This might be explained by a couple of important factors, namely the conductive/convective
293 heat loss caused by water and the fact that manatees are believed to have a third to a sixth of the metabolic rate of terrestrial mammals of simila r size. Marine mammals such as phocid seals and cetaceans compensate for conductive heat loss through a number of mechanisms, those most notable being incr eases of insulation and metabolic rates 23 times those of terrestrial mammals. It should therefore come as no surprise t hat manatees might require physiologic and anatomic thermoregulatory mechanisms for limiting heat loss and supporting proper neural function. The IVB may function in exactly such a manner, by limiting heat loss to the environment while simultaneo usly rerouting part of that warmth to the brain in order to maintain proper neural function. Support for such a function may be found in the fact that the winter temperatures encountered by Florida manateesco incidentally the same temperatures that cause morbidity and mortalityare in a range shown to affect the amplitude and velocity of action potentials in peripheral nerves. By absorbing arterial heat into its veins, the IVB would allow the intimately associated nerves to be maintained at a higher temperature, while th e distal portions of the muzzle are allowed to drop in temperature in order to limit conductive/convective heat loss. Regardless of whether or not this speculation is appropriate, one thing seems fairly certain. Every extremity in the m anatee is invested with two venous returns; one warmed and one cooled. This is not different from the extremit ies of terrestrial mammals, however the elaboration of the warm ed venous return is astonishing in both cross-sectional surface area and field of drainage. Indeed, the presence of elaborate counter-current heat exchange structures in every appendage with a high surface area to volume ratio and therefore high conductive/convective heat loss cannot be ignored.
294 The vascular bundles of the maxillary and mandi bular portions of the snout, the pectoral flippers, and the tail provide strong evidenc e in support of an overarching need to conserve heat, or at least the ability to do so efficiently. It is challenging to think of the ev olutionary drive that led to such a thermoregulatory mechanism, when other marine mammals appear to have adopted different mechanisms. It is possible that due to the fairly protracted periods of fasting during cold weather, dependence on insula ting subcutaneous fat deposits and high metabolic rates would be a poor strategy for cold tolerance. A hypometabolic state that utilizes fewer energy reserves combined with anatomical adaptations designed to reduce the amount of metabolic heat lost to the environ ment may represent a novel strategy for adapting to a marine environmen t. Florida manatees show considerable site fidelity in the regions they inhabit, and given their dependence on coastal vegetation and fresh water resources, they may not have the adaptive flexibility to radiate into more suitable environments. Thermoregulation (other) Vascular bundles are not unique to manat ees. Indeed the caudal v ascular bundle can also be found in the chevron canal of cetaceans, and has been widely accepted to function in the same way, namely heat conservation. Like the broad paddle of the manatee, cetacean tail flukes are highly vascularized region s with high-surface area to volume ratios, and are thus capable of radi ating considerable heat to the environment. This proves beneficial during periods of exertion or exposure to higher ambient temperatures. Conversely, when exposed to colder temperatures, cetaceans are capable of limiting the amount of heat lost at the tail by returning blood through the caudal vascular bundle. Interestingly, des pite the nearly identical architecture and
295 vascular origin of the caudal vascular bundle (CVB) of manatees and dolphins and the extensive investment of the cetacean body with vascular plexuses, vascular bundles cannot be found in any other part of the cetacean va scular system that has been described thus far. It is not known if this represents a manifestat ion of phylogenetic or functional significance, however the apparen t thermal challenges faced by manatees may go a long way towards explaining the pro liferation of vascular bundles compared to cetaceans. Another example of a stru cture likely capable of counter-current heat exchange can be in the tongue of the dolphin. The ma in lingual artery predominantly supplying the major tongue muscles ( m. genioglosus and m. hyoglossus ) is surrounded by a venous plexus composing a periarterial venous rete (PAVR). Additionally, the mandibular alveolar artery is similarly surrounded by a PAVR. Just like the caudal vascular bundle, PAVRs present in the dorsa l fin of dolphins are widely accepted as heat conserving structures (Elsner et al. 1974; Scholander & Schevill, 1955; Scholander, 1958). The lack of physiological temperature measurements in the lingual and intramandibular regions once again forces me to speculate as to their function, however given their structural similarity to the PAVRs of the dorsal fin, it seems reasonable to draw similar conclusions about their function. Thermoregulation (eyes) Another structure that calls for attention is the vo luminous and intricate ophthalmic plexus found in Florida manatees. Despite t he relatively modest size of their eyes and poor visual acuity, the eyes are surrounded by and invested with a considerab le venous plexus with connections to the veins of the face and the cavernous sinus of the calvarium. It is not clear fr om the present study what t he degree of contri bution of each
296 of those connections play in either drainage or filling of the ophthalmic plexus, however three important facts are known. Firstly, t he ophthalmic plexus is connected to potential sources of warmed venous blood (e.g. IVB). Secondly, the plexus also has substantial connections to the veins of t he face (e.g. facial, deep facial masseteric plexus, etc.). Thirdly, the plexus connects directly to the cavernous sinus and to veins forming an extensive plexus of fine caliber veins within the dura of the brain case. These connections have the potential fo r significant implications. The connection to a potential source of warmed venous blood within the IVB means that the temperature of the blood within the opht halmic plexusand therefore the eyes--may be elevated. Warming of the eyes could serve as sensory modulation allowing manatees to maintain their vision during exposure to colder temperatures when the rest of their face is undergoing cooling. The connecti on to the large, valve-less veins of the face might serve a hemodynamic f unction, providing some form of structural support or enhancement to the ey e (Samuelson, pers. comm. 201 2). Since the veins of the face are associated with and/or em bedded in large muscles used frequently, it seems possible that the muscles can act as a non-cardiac pump that increases venous pressure and forces blood toward the ophthal mic plexus. Such higher venous blood pressure in the ophthalmic plexus could provi de structural rigidity to the tissues within the orbit, much like a hydrostatic skeleton. This theory seems problematic because the ophthalmic plexus connects dire ctly to the cavernous sinus via a sizable vein. Unless an undiscovered venous valve or vasoconstric tive property exists in this vein, the pressure in the ophthalmic plexus would be likely be alleviated by passage of blood from the plexus to the cavernous sinus. More importantly, such elevated venous
297 pressures in this region might have deleteri ous effects on the brain tissues by way of the Monro-Kellie doctrine. The direct connection of the ophthalmic plexus to the cavernous sinus is normal even in domes tic mammals, however the plexus appears much larger than what is found in domestic mammals. If the facial veins of manatees respond to thermal stimuli in a manner similar to domestic mammals, they could constrict, ther eby facilitating labial and nasal venous blood flow through the IVB, essentially actuating the counter-current heat exchanger and subsequent enabling warmed venous return through the IVB. The connections between the IVB and ophthalmic plexus coul d then enable passage of the warmed blood from the IVB to the ophthalmic plexus. It is im portant to note that ophthalmic retia have been described in a number of mammalian and avian species. In all cases, these structures were described as arteriovenous counter-current heat exchangers that by definition contain intimately ju xtaposed arteries and veins. W hat is interesting in relation to this fact is that the ophthalmic plexus of the manatee appears to be almost entirely venous in nature, and therefore would not have the ability for counter-current heat exchange. Many of the aforementioned physiologic al questions might be answered with relatively simple, non-invasive methods. Use of thermal imagi ng cameras, Doppler ultrasonography, and thermal probes in conjunc tion with differential regional and whole body warming and cooling may provide answers regarding the thermal capabilities of some of the complex vascular structure. Vascular Growth and Patterning A topic of particular importance when exam ining comparative vascular morphology is that of driving dev elopmental forces and regulatory mechanisms. Perhaps equally
298 significant to observing the interand intraspe cific variations in blood vessel patterning are the mechanisms manifesting those patterns. In this day of molecular biology, no discussion of the morphological patterning of blood vessels can be considered in isolation of what is known about the involvem ent of molecular signaling. Indeed, when considering some of the vascular elaborati ons observed in certain marine mammals, a rudimentary understanding of mole cular signaling is at the very least beneficial. Nonetheless, before considering the mechanisms driving formation of the often complex and tortuous vascular patterns of marine mammals, it should prove fruitful to explore the mechanisms behind the more subtle or perhaps commonplace vascular patterns observed in laboratory animals. As such, I shall start by defining some key molecular terminology as it relates to vascular signaling, followed by a cursory explanation of some of the better established signaling pat hways. I will subsequently discuss other non-molecular regulatory and/or guiding me chanism behind formation of certain vascular patterns and will conclude the sequence by discussing some relevant connections to marine mammal vascular networks and possible mechanisms driving the observed vascular patterns. Primary Vascular Plexus The origin of any vascular system appears to commence with the establishment of an undifferentiated network of blood vessels (Heuser, 1923; Padget, 1948; Sabin, 1926). In the yolk sac, this network is fo rmed from the migrati on and specialization of progenitor cells of splancnic mesodermal origin After receiving the appropriate signals, those cells migrate to thei r appropriate location and form blood is lands, which are clusters of hemangiocytes. The blood island s are bordered by angiocytes destined to become vascular endothelial cells, while the central portions of the islands are occupied
299 by cells that become blood cells (Lars en, 2001). After once again receiving the appropriate instructions, the angiocytes c onnect to adjacent ones and ultimately form tubules with lumens. The tubules eventually connect to each other to form a primary vascular plexus. Unlike the yolk sac vasculogenic process, intraembryonic splanchnopleuric mesodermal cells do not fo rm blood islands (Larsen, 2001). Instead, angioblastsinduced by substances of endoderma l origindifferentiate into flattened endothelial cells that join to form angio cysts. The angiocysts subsequently connect to each other to form elongated tubular vessels ca lled angioblastic cords. The cords in turn coalesce to form angioblastic plex uses (Larsen, 2001). Shaping of the adult vascular networks is the result of numer ous processes including the continued generation and recruitment of new angiocysts. It is not until later that these primary angioblastic vascular plexuses are given instructions to begin the angiogenic process, during which arteriovenous differentiation, capillary bed formation, and large scale vessel landscaping occur. According to Evans (1909), the idea of a primary vascular plexus acting as the precursor to all adult vessels was first suggested by Aeby in 1868. It wa s not until years later that this hypothesis was truly embraced. Evans (1909) stated that nothing in the anatomy of the adult vascular system approaches a netlike conditi on until we reach the capillary bed, where it forms a characteristic feature and he asked if the primitive net postulated by Aeby is not in fact a capillary net? Evans first st atement about the lack of adult nets outside of capillary regions seems in error when c onsidering marine mammal vascular patterns (e.g. epidural rete ) as well as some seen in domestic species (e.g. carotid rete ). Evans second observation may with our current kno wledge seem pedestrian, however, for the
300 knowledge of the time was quite insightful. It is now known that capillary beds are formed from retention and devel opment of the primary vascula r nets. Through the work of these and other researchers (Schulte & Tilney, 1909; Mall, 1904, Sabin, 1926) our morphological understanding of the origin of capillary beds and large trunk vessels progressed by leaps and bounds. The next and presumably predictable step was to figure out the molecular signaling that guided these substantial modifications to the physical features of the vascu lature which inescapably have dramatic effects on the physiology of the devel oping and adult organism. Molecular Terminology The most basic functional units in this discus sion can be divided into two main groups, those being the extracellular si gnaling molecules, and the cell-surface receptors. The signaling mo lecules are usually proteins expressed or released by certain tissue types and which carry a specific message. In this context, the message is typically one of attracting or repelling vascu lar growth. This concept may prove more important and complex at a second glance. For instance, a blood vessel may grow in a certain direction either because a signal is a ttracting it that way, or because everywhere else there is a signal repelling it. Similarly, a vessel may not grow in a particular direction either because there is an inhibitory signal in that direction or because of a lack of an attractive signal. Whichever the case, this highlights a potentially important philosophical (and likely physiological) notion, that being that vessels may grow into a certain pattern in one of two ways: (1) by be ing initially instructed to proliferate wildly and subsequently being trimmed according to tissue demands, or (2) by minimizing the proliferation to only the regions of proliferative signaling. Certainly reckless proliferation seems wasteful and potentially harmful and aberrant arteriovenous connections are
301 often associated with certain malignancies and dangerous vascular malformations (Feindel & Perot, 1965; Rosenblum et al. 1987), however in a developing fetus in which little tissue differentiation and compartmentaliz ation has occurred, perhaps this reflects a less specialized or sophisticated va scular investment based on a less organized structure. Whichever method is used, there appears to be care ful orchestration of either the proliferative or deconstructive processe s of blood vessel formation, and the degree to which each strategy is used may be less dependent on species and more dependent on developmental stage and tissue type. T hese two differing strategies will be discussed further in subsequent sections. For the afore-mentioned signa ling molecules to work, ce ll-surface receptors must exist. Those receptors must be able to recognize and bind to the signaling molecules and convey the inhibitory or activating message to the target cell. In many cases, the signaling molecule is the same but t he presence, abundance and type of receptor molecules determine the direction of a cascade. This naturally adds another source of control over the system by way of contro lling the receptor numbers and types, as well as their binding affinities and specificities. For instance, response of a cell to an abundant signal molecule may be only as good as the number of receptors expressed on that cell. The likelihood of the receptor encountering its ligand is proportional to the numbers of each that is present (assuming no interference by different molecules--antagonistic or otherwise). Assuming an interaction between ligand an d receptor does occur, the signal strength will depend on numerous factors such as the number of activated receptors, the strength of each receptor s subsequent signal, and the strength of the ligand signal
302 itself. For instance, certain homodimeric ligands have been shown in some cases to stimulate greater responses t han heterodimers formed by two different ligands of similar function (DiSalvo et al. 1995; Cao et al. 1996). It therefore becomes evident that various levels of control are possible and indeed exist. These controls allow a developing or repairing body to fine tune various aspects of angiogenesis and vasculogenesis. Mutant models and targeted microinjection techniques in laboratory animals allow researchers to study these c ontrols by way of observing the effects of ablation or over-expression of cellular signals on aberrant growth of blood vessels. Signaling Molecules There are undoubtedly endless nu mbers of signaling m olecu les and it is not my intent to try to cover many of them. However, in order to begin to understand how some of these mechanism work, we must discuss some of the better-s tudied molecules. Perhaps the best studied one is vascular endothel ial growth factor (VEGF), which is a secreted glycoprotein first isolated in the 1980s (Achen & Stacker, 1998). Different VEGF isoforms are known to exist in humans, however the 165 amino acid long isoform is believed to be the most ubiquitous. Ea ch isoform appears to have a different degree of bioavailability, with some forms circulating as a solubl e protein while others remain associated with the cell membrane or ev en the extracellular matrix (Houck et al. 1992; Park et al 1993). Numerous experiments have si nce been conducted that show VEGF to have a crucial role in angiogenesis as well as tumor development (Carmeliet et al. 1996; Ferrara et al. 1991). Certain isoforms of VEGF are also known to induce vascular permeability and to promote survival of newly formed blood vessels (Alon et al 1995; Benjamin & Keshet, 1997). Since the initia l discovery of VEGF, additional molecules have been identified and grouped together based on similarity of their primary structure,
303 to comprise the VEGF family. Such mole cules include VEGF-B, VEGF-C, VEGF-D, and placenta growth factor (Achen & Stacker, 1998). Interestingly, VEGF-type molecules have also been isolated from the orf virus of the Poxvir us family, whose pathologic lesions in part consist of vascula r proliferation and swelling (Groves et al. 1991; Lyttle et al. 1994; Meyer et al. 1999). Both sequelae may be easily explained by endothelial cell proliferation and increas ed vascular permeability. It has recently been shown that VEGF is a very potent signaling molecule with effects on numerous physiological processes, and despite their similarity different isoforms appear to serve very different roles. It has been shown that VEGF-B is important in vasculogenesis, while VEGF-D is an integral proliferat ive stimulant during angiogenesis (Shalaby et al. 1995; Carmeliet et al. 1996; Ferrara, 2001). Vessel Growth Types Mature blood vessel growth can be separated into two types based on the origin of the cells and the typical timing relati ve to embryogenesis and maturation. Vasculogenesis refers to the differentiation of endothelial cell prec ursors into endothelial cells to form new blood vessels, as seen in early embryogenes is. The differentiation is induced by paracrine factors released by endodermal tissues (Larsen, 2001). Vasculogenesis is usually associated wit h the formation of vascular netsthe undifferentiated fetal vascular beds. In adults, circulating precursor cells can also be incorporated into newly growing vessels at sites of angiogenesis (Asahara et al. 1997; 1999). Angiogenesis refers to the formation of ma ture blood vessels and capillary beds from the preexisting primary vascular plexus es, and is considered the hallmark type of vascular growth in adults and later stages of embryogenesis (Achen & Stacker, 1998). Furthermore, Patan et al. (1996) noted two distinct types of angiogenesis; sprouting of
304 vessels from preexisting vessels and splitting of vessels to generate greater numbers of vessels--intussusception. It is unclear what exactly drives these two different types of angiogenesis, however it is likely the resu lt of complex crosstalk between different paracrine signaling molecules refe rred to in subsequent text. Patterning Dynamics (e.g. Clipping/Trimming, Elongation, Linearity etc.) So we know what causes creation and proliferation of vascular structures but what determines the shape of vascular structures? Clearly, signals of proliferation and/or elongation such as those provided by VEGF can guide vascular pattern development. Uncontrolled expression of VEGF has been shown to generate masses of convoluted vascular nets at various st ages of angiogenesis (Flamme et al. 1995). However, as discussed earlier, deletion can be just as infl uential as creation when it comes to pattern generation. Therefore, in opposition to or in coordination with VEGF-type signals, deletion or trimming of vessels can hav e a profound influence on vascular network shape by trimming pre-existing networks into a different patte rn. Additionally, blood flow itself has been shown to have drastic influence s on patterning of the arterial tree as well as on the regulation of certain molecular signal molecules such as EphrinB2 and EphB4 known to be involved in arteriovenous di fferentiation and shunt formation (le Noble et al. 2004). What about linearity? Semaphorins have been implicated in guiding the formation of linear, organized vascular patterns by guiding cell migration. The semaphorin family of proteins is bi-functional capable of promoting and inhi biting growth. Sema3A has been shown to bind to endothelial cells and inhibit their migration. It also is capable of inhibiting capillary sprouting and has been linked to retraction of lamellipodia. Conversely, Sema4D induces chemotaxis and tubulogenesis in endothelial cells and
305 enhances blood vessel formation in in vivo rat models. Plexin-D1, the receptor for Sema3E has been identified as essential fo r normal vascular patterning and endothelial positioning (Roth et al. 2009). Plexin-D1 deficient mi ce show excessive branching and loss of normal segmental blood vessel patterns seen in the intersomitic regions (Eichmann et al. 2005). Arteriovenous Differentiation of Primary Vascular Plexuses There is now evidence that arteries and veins differentiate from pre-existing primary vascular plexuses. Long before ar teriovenous identity has been established, the destiny of an angioblast and all its pr ogeny seems to have already been decided. Arteriovenous identity appears to be dete rmined around the time of VEGF-induced proliferation and differentiation of mesodermal cells into angioblasts (Zhong et al. 2001; Lawson et al. 2001, 2002). This has been shown to occur, at least in part, because of differential expression of cell surface molecules. A molecular class known as ephrins, specifically Ephrin-B2, is found only on the surface of angi oblasts destined to be arterial in identity, while the receptor (EphB4) for this ligand is present only on the surface of angioblasts destined to be venous in nature. The expression of these surface molecules is determined, at least in part, by the degree of activation of the Notch surface protein which in turn activates a tr anscription factor known as Gridlock (Lawson et al. 2001, 2002). Interestingly, it appears as though the basal condition in which there is no Notch activation, results in only venous formation. This may make sense evolutionarily since inappropria te or aberrant arterial form ation may be more detrimental than venous formation, and therefore likely requires greater control. Making this picture more complicated is the fact that VEGF appears to have a significant role in stimulating Notch activa tion. Since VEGF is required to initiate
306 vasculogenesis but also increases Notch acti vation which leads to arterial designation of angioblasts, it is unclear how VEGFs pres ence is controlled in a manner that does not preclude assignment of angioblasts to a venous fate. Some researchers (Weinstein & Lawson, 2003) have suggested that such events are choreographe d through temporal organization of vasculogenesis, stating that these events could work if arteriovenous differentiation is sequential ra ther than simultaneous. This wo uld require that arteries differentiate first and use the ephrin/Eph interaction to subsequently form their venous counterparts. This mechanism is supported by t he fact that the vitell ine arteries of the chick appear first within the cap illary network. Many tributaries of larger arteries such as the ventral and dorsal intercostal arteries and the internal thor acic arteries (aka internal mammary) after leaving the parent vessel are immediately bordered by a vein on either side. Additionally, many of the arteries of the extremities are bordered by two veins. Finally, this mechanism seems to al so fit well with the fact that arteries and nerves seem to follow very similar pattern s through the tissues, while veins do not. Since VEGF has been shown to also stimul ate nerve growth and development, it seems reasonable to assume that VEGF is responsible for laying down the initial arterial and nervous architecture, and the arteries comple te the vascular networking by inducing concomitant venous development. The implications of this ephrin and VEGF type of signaling mechanism cannot be overstated when one considers some of the vascular patterns observed in marine mammals. Certainly the seemingly disorgan ized architecture of the predominantly arterial thoracic retia of cetaceans seems rather pedestrian with respect to arteriovenous differentiation signals, but what of structures such as the brachial
307 vascular bundles of edentates and sirenians and caudal vascular bundles of sirenians and cetaceans? The caudal vascular bundle of a manatee is composed of approximately 1100 arteries juxtaposed to approximately 2200 anastomosing veins, allowing for what seems like an unmatched counter current heat exchanger (Rommel & Caplan, 2003). When one sees this or other similar structures in cross section, the close proximity and dense packing of arteries and veins become apparent. If indeed the molecular signals for arteriovenous differentiation are of paracrine origin, how does such a complex yet ex tremely organized, linear structure get generated from such densely packed vasculature? This also begs the question of whether the pattern forms before the differentiation, or the differentiation occurs before or during elaboration of the structure? Does the entire pattern form and then receive alternative differentiation signals (based on extracellular signaling molecules and surface receptors of immature vessels), or do the arteries and ve ins grow sequentially, or simultaneously? All aforementioned mechanism s seem possible, for if the arteries form first in a linear fashion, their paracrine inductive capabilities would likely also lead to linear venous formation. Alternatively, the differential trimming of a generalized vascular plexus could lead to similar patterns. The complication arises from the fact that these networks would have to coalesce into separate parent trunk vessels (e.g. caudal abdominal aorta and vena cava). Physical and Physiologi cal Guiding Principles Despite the clearly influential properties of genetically determined express ion levels of molecular signaling molecules and t heir respective receptors, considering them in isolation would at best be na ve. The physiologically relevant signals of chemical and physical forces are undeniable and should ther efore not be overlooked. Soft and hard
308 tissue remodeling is dynamic and constantly responding to chemical and physical cues. Bone remodels based on forces exerted upo n it and stomach lining thickens in response to elevated acidity. These and m any other dynamic tissue responses are the result of molecular and other messages co nveyed to the tissues and exerted by the tissues. Therefore, although there is cert ainly a genetically predetermined pattern of vascularity in each tissue, forces such as oxygen diffusion, oxygen availability, and tissue-specific metabolic demands can be very influential in determining the degree and perhaps pattern of vascularity. In many cases the molecular signals and pathways used for conveying such signals may be no different than those used during vasculoge nesis and angiogenesis. For instance, in addition to genetically predetermined levels of VEGF expression, up regulation of VEGF has been shown to occur in tissues exposed to hypoxic insult (Shweiki et al. 1992). Making use of already es tablished and successful mechanisms naturally makes sense from a parsimonious standpoint; nonetheless, it is neither clear how the genetically predetermined signals are differentiated from or managed in concert with the environmental cues, nor how the levels of influence of each are controlled. This question becomes increasingly relev ant when considering that both signaling pathways may use the same mechanisms and mole cules. Perhaps there is a means of hierarchical prioritization of signals, or an overarching genetic ally determined limiting mechanism that allows only so much signalin g despite whether the source is genetic or physiological. A lack of such control may be what distinguishes norma l from aberrant or cancerous growth?
309 Gdde and Kurz (2001) showed significant effects of shear and frictional stresses when they were applied to vascular growth modeling algorithms, not surprisingly suggestive of important effects of physical attributes that det ermine the efficiency of fluid transport through a tube. Aftera ll, the same rules of econom y of structure and attention to structural integrity should apply to vascular networks much like they apply to plumbing. The principle of structural optima that provide the best compromise between structural integrity, cost of building and maintenance, and co st of transport is not a new concept for the generation of any bi ological tissue. Indeed, Djonov et al. (2002) note that the theory of bifurcating vascular systems predicts vessel diameters that are related to optimality criteria like minimization of pumping energy or of building material. Along similar lines, Schulte and Tilney noted that the result of the substitution of large trunks for plexuses is the reduction of the impediment offered to the venous return by surface friction, consequently either a reduction of card iac work or, the work performed by the heart remaining the same, a more rapid circulation and potentially a higher rate of metabolism. (1909) Despite the guiding principle of econom y, there are likely to exist certain exceptions, where a specif ic adapted function (e.g. thermoregulation, diffusion, concentration) of a vascular structure ma y be of overarching importance that supersedes the rule of economy. For exam ple, a blood vessel that is exposed to elevated pressures (e.g. ascending aorta) ma y compromise on economy of construction and maintenance in order to establish the necessa ry structural integrit y. In this case, the wall thickening is necessitated by the hi gh-pressure cardiac outflow as well as the unavoidable mathematical relationship of disproportionately increasing wall thickness with increasing volume. Conversely, a vesse l exposed to low, post-capillary pressures (e.g. vena cava) may compromise on structural integrity for the sake of economy by way
310 of reduction of surface friction and low resistanc e to flowa larger diameter, thin-walled vessel. It is important at this stage to note that these examples are likely to be dreadful oversimplifications and that numerous other forces add to the grand scheme of influences guiding such physical properties as vessel thickness. For instance the elastin and collagen composition of the tunica media of a blood vessel changes as one progresses further distad from a parent artery, so that parent vessels such as the aorta have more elastin and smaller tributaries have more muscle. This composition directly affects the degree of elastic recoil of the vessel as well as its vasoconstrictive ability, and therefore has important dow nstream effects on the transmission of a pressure pulse and other forces such as shear stress (K-J Li, 2004). Modeling of shear stress showed a 25% reduction of shear in a compliant vesse l model over a rigid vessel model (K-J Li, 2004). At low pressures, elastin dominates the behavior while at high pressures where elastins anisotropic behavior becomes eviden t, fibrin becomes more important (K-J Li, 2004). Nonetheless, Li (1985) found that gr eater effects on pressure and flow through vascular junctions were caused by geometry than by elastic factors. Even more so than individual vessel char acteristics like composition, vascular network patterns are also likely, at least in part, guided by certain physical forces such as resistance to flow. The exponential incr ease of resistance with relatively small decreases in cross sectional area of bl ood vessels necessitates specific patterned responses. A common means of mitigating the harmful manifestations of this physical property is to increase the total downstream cross sectional area, however this has implications such as increasing resist ance with decreasing vessel diameter. For instance if a large vessel must be branched into smaller ones, a successful way of
311 doing that without detrimentally increasing blood velocity and shear stress is to branch enough times so that the total cross sectional surface area of all the branches is equal to or greater than that of t he parent vessel. Naturally t he shape of the bifurcation will also affect characteristics such as laminar and turbulent flow, shear stress, resistance, and velocity. To avoid tasking the heart with excessive backpressure, this may be what is occurring in the arterial component of the caudal vascular bundle of marine mammals, as the caudal aorta abruptly branches into the numerous smaller arteries of the caudal vascular bundle and the caudal artery. It should be noted that too significant a velocity/pressure reduction would be detriment al to delivery of the blood to capillary beds which can generate considerable back pressure, therefore a balance between heart loading and under pressurization is likel y necessary. Also important may be the flow generated by muscular contraction and undulatory locomotory movements of the region, however this component is currently unknown. Conversely, a reverse employ ment of the afor ementioned physical property of fluid dynamics might be used to increase velocity and aid in transport of low-pressure postcapillary blood returning to the heart, as may happen in the venous counterpart of the caudal vascular bundle. A reduction in total cr oss sectional area may result in a velocity increase which aids the return of venous blo od to the heart. A di fferent strategy for mitigating the effects of surface area r eduction during vessel branching may be to generate vessels with thicker, more elasti c walls that can expand and propagate the pressure pulse during diastole (Shadwick an d Gosline, 1994; K-J Li, 2004). The elastic expansion of the walls serves numerous roles: to (1) absorb and dissipate some of the energy of the increasing velocity thereby reducing harmful downstream effects of
312 pressure and velocity, and (2) avoid rupture of the vessel wall as it receives the high velocity stream, and (3) to reduce shear stress which can lead to atherosclerosis (K-J Li, 2004). Finally, creation of collateral flows can also hel p with management of some of the physical challenges encounter ed with reduction of total cross sectional surface area and its effects on blood flow. A dramatic such example can be seen in hum ans with coarctation of the aorta. In such cases it is common to see other vessels such as the supreme intercostal arteries or the internal mammary arteries establish collateral flow by connecting the proximal high-pressure portion of the aorta with the distal, low-pressure section, thereby maintaining adequate flow to the abdomen and caudae despite the aortic stenosis (Kirks et al. 1986). Kirks et al. report that mediastinal collateral flow is a common finding in children with aortic coarctation. This suggests that during development of the vasculature, the constriction of the aorta wit h subsequent up-stream pressure increase and down-stream oxygen deficits leads to ph ysical and/or physiologic signals that maintain and enlarge pre-existing collateral flows. Considering the embryonic changes in arterial supply to the cetacean brain, it may be possible that elaboration of the thoracospinal retia is in part due to diminution of the carotid supply. In addition to the obvious perfusion deficit of t he brain, such a diminution of the internal carotid system may also generate increased pressures in the aortic system which in turn necessitates collateral development and maintenance. Support for this idea may be present in the ol der literature. Sinclair (1967) noted a temporally progressive change in arterial bl ood supply to the brain during dissections of spotted dolphin ( Stenella sp. ) embryos. He noted that in a 4.2mm embryo primary
313 supply to the brain was initially conducted through the stapedial arteryan embryonic offshoot of the internal carotid artery. In the 12 and 22mm embryo this pattern was replaced by a vertebralbasilar system. McFarland et al. (1979) noted that their findings as well as Slijpers (1936) are in opposition to Sinclairs (1967), and that the blood supply to the brain of the dolphin was via the supreme intercostal arteries into the thoracospinal retia and through the foramen magnum. Though we know this pattern to be true in post-natal cetaceans, McFarland et al. (1979) did not examine fetal specimens. Hence, there is a distinct possibilit y that Sinclairs (1967) findings are in fact correct and represent the developmental stages en route to the collateralization of the arterial supply to the brain before the t horacospinal system takes over. Similar ontogenetic changes were documented in the s perm whale by Melnikov (1997) and the harbor porpoise by Slijper (1936). Sinclairs findings show a change from the generic mammalian phenotype to a progressively great er dependence of vertebral blood supply. This should not seem surprising in light of the fact that a gro up of closely related mammals, the artiodactyls, replaces embryonic internal carotid arterial supply to the brain with external carotid and occipital cont ributions (Schaller, 2007). Humans have supplemented internal carotid supply with enla rged vertebral arteries, and Gillilan (1972) stated the major developmental advance in the blood supply to the brains of primitive mammals is the addition of the vertebral arteries to supply the brainstem. Interestingly, Sinclair noted that the vertebral arteries in his dissected specimens were uniquely situated dorsal to the transvers e processes, rather than coursing through transverse foramina. This may be suggestive of a vertebral arterial system on its way to being absorbed into the extensive thoracic and epidural retia since no distinct vertebral
314 arteries can be found in the adult cetacean. If in fact this represents a gradual transition through different stages of blood supply to the brain, it raises an interesting question, that being, whether the thoracospinal retia are new, plesiomorphic adaptations or elaborated atavisms? Interestingly, it appears as though all non-mammalian vertebrates, including fish, have a basilar arterial connection to the brain, however negligible its contribution may be (Gillilan, 1967). This basilar system comes in many different forms, some with bilateral separations and loops and others with a single median vessel, but all these examples st ress the fact that numerous patterns of development of this vasculature ar e possible and may be rooted deep within the ancestral genes of cetaceans and sirenians. If that is indeed the case, it may be more likely that Sinclairs (1967) finding of a vertebral-basilar system early on in cetacean embryogenesis is even further support ed, and that this syst em gradually gets incorporated into the thoracospinal system. Organ Vascular Examples Although the information presented thus far suggests that all intraembryonic vasculature develops the same way, many organs appear to guide much of their own intrinsic vascularization by releasing their own set of angiogenic factors. Instead, and particularly within dev eloping organs, mu ch of the vasculature may be generated through intrinsic formation of angioblastic cords and plexuses and subsequent sprouting of the definitive vasculature rather than a primitive undifferentiated network. This perhaps makes sense from the standpoint that many of the organs do not develop until much of the circulatory system of the tr unk has been consolidated and organized. By this time, there may therefore be much better organization to the flow of blood through the system, and the development of disorganized rudimentary plexiform vascular
315 patterns within organs is poorly suited for t he task. At this point, generating a more targeted and organized pattern of vascularity is likely more appropriate and is best guided by the organ itself. Wislocki (1939) noted that the vessels which form in the brain of the opossum do not arise by reduction or simplification of a primary anastomosing capillary net. He stated that from the time of their first appearanc e the vessels supplying the brain consist of simple, non-anastomotic loops. These loops form through sprouting from parent vessels. (1939) Though among mammals this loop-type brain vascularization seems to be a peculiar trait of marsupials, it also is a trait of mo re basal vertebrates such as lamprays, hagfish, lungfish, tuataras, some amphibians and many of the lizards (G illilan, 1967). Perhaps more important to this discussi on however, is that it highli ghts the idea that not all vasculature arises from consolidation of primary vascular nets, especially within organs. Kidneys offer an example of intrinsic guidance of vascular development. The developing nephrons secrete VEGF which guides blood vessels into the developing reniculi, to form the vasculature necessary for filtration (glomerular apparatus) and reabsorption (arteria re cta). This again is an example of guided formation of individual vessels rather than coalescing of primary plexuses. The t heory of sequential arteriovenous differentiation proposed by Weinstein and Lawson (2003) once again makes sense when looking at the pattern of renal vasculature. It seems reasonable to suggest that the interlobular arteries with thei r glomeruli and arteria recta form first and then induce the formation of the adjacent interlobular veins and associated venous components of the vasa recta. Since there are no venous co mponents to the glomeruli, arteries would have to form first and veins may subsequently be formed only in regions
316 where arterial expression of ephrin occurs. The lack of veins in the glomerular region may be suggestive of a lack of ephrin expression in the glomerular arteri es. It is unclear however, whether the renal vessels form solely through guided migration of blood vessels into the developing reniculi, or if blood vessels form within the reniculi and then merge with the infiltrating vessels. Interesting associations have been found in relation to angiogenesis of adipose tissues. As mentioned previously, it is we ll known that VEGF is a crucial factor for stimulating angiogenesis, however the molecular interactions are significantly affected by cold exposure. Xue et al. (2009) showed that cold exposure of mice led to enhanced angiogenesis in brown and white adipose tissue. Additionally, hypoxia is known to upregulate VEGF expression (Huang et al., 2007) and downregulate angiogenesisinhibitors like thrombospondin (Laderoute et al. 2000). These findings are especially interesting in relation to this discussion, because not only are dolphins and manatees frequently exposed to cold temperatures, bu t they also have considerable adipose investment in their bodies, and due to their breathholding life styles are often exposed to bouts of diving-related hypoxia. It is unk nown what role any of these characteristics might play in influencing angiogenesis in specific body regions. Nonetheless, these types of interactions may prove crucial in understanding blood ve ssel development in areas like the intramandibular fat body of the dolphin, where adipose tissue abounds and those tissues are often exposed to cold ambient temperatures and hypoxic conditions. Observed Patterns in Marine Mammals Of particular interest in this discu ssion are the mechanis ms that guide the formation of some of the elaborate vascula r patterns observed in marine mammals.
317 Although there are numerous vari ations on the theme, there are three distinct patterns of vascular networks that have been found in ma rine mammals. The first is a relatively typical form of a rete mirabile much like the type seen in terrestrial mammals. This pattern is composed of either a parent artery or parent vein that branches into progressively smaller vessels, resembling an arboreal or fractal pattern. The branching pattern can either be symmetrical or stagger ed, ladder-like. A second type of retial pattern found in some marine mammals is found in the dorsal thoracic wall, cervical, and epidural regions (e.g. thoracic, cervical and epidural retia ). This type of pattern consists of highly convoluted vessels with a variable but usually more diffuse degree of anastomotic connections. Finally, the third type of vascular pattern observed in marine mammals is that which has been termed a vascular bundle. T he vascular bundle consists of a parent vessel which branches abruptly into numerous smaller vessels. There is little to no reduction in caliber of the branches as they progress distad, and t he apparent pattern is reminiscent of a broom or paint brush rather than a net (Murie, 1874; Fawcett, 1942; Elsner 1966). Until recently, this type of pattern has predominantly been associated with the brachial vascular bundle (e.g. lemurs, sloths, sirenians, cetaceans) and the caudal vascular bundle (e.g. sirenians, ce taceans, kangaroos), however intercostal vascular bundles also exist in sirenians (Murie, 1874; Wislocki & Straus, 1932; Fawcett, 1942. Rommel & Caplan, 2003). Additionally, the infraorbital and mandibular vascular bundles described herein also have the unmista kable form of a vascular bundle. These cephalic vascular bundles are present in t he heads of Florida manatees, and so far as I can tell from the size of t he infraorbital canals and foramina of their skullsapparently
318 representative of the vascular mass they accommodate--may also be present in other manatee species, in dugongs and the extinct St ellar sea cows. Although this is not a discussion about phylogeny or evolution, the functional implications of the vascular patterns observed are inexorably linked to the biogeography and phylogeny of the animals that possess them. It may therefore prove fruitful to examine the available osteological fossil sirenian evidence as it relates to vascular structures such as the infraorbital vascular bundles. Since the infraorbital vascular bundles are seemingly heat-conserving structures like those found in the chevron canal and pectoral flippers, and the size of the infraorbital canals is tied to the size of the vascular bundles, there may be clues in the fossil evidence relating to the diversification and/or radiation of certain sirenian ancestors. This may be espec ially relevant when one considers that the Eoceine epoch--when sirenians are thought to have first diverged from their ancestors-had a much warmer climate (Marsh et al. 2012). Perhaps the onset of that radiation is reflected in the fossil record, with si renian ancestors not showing an enlarged infraorbital foramen, while later predecessors inhabiting cooler climates have a welldeveloped one. It may, theref ore, be instructive to conduct an allometric dimensional comparison of infraorbital foramina and certain skull features representative of size (e.g. condylobasilar length). Comparing such allometric relationships not only between extinct and extant sirenians but also other mammals that do not possess an infraorbital vascular bundle may provide insights into th e evolution of sirenians. An interesting species to include in such a comparison may be the elephant which also possesses large infraorbital nerves that innervate the dexterous trunk.
319 Although Stannius (1845) does not appear to have recognized the infraorbital vascular bundle as such, he was the firs t to describe its bundle-like arterial arrangement. It is unclear to what extent anastomoses exist in the vascular bundles. Though we know the venous component is hi ghly anastomotic, the arterial component is up for discussion. In cross section of the infraorbital vascular bundle it was evident in my specimens that a certai n degree of arterial anastomo sis was present, yet modest compared to the venous anastomoses. Fawc ett (1942) stated that occasionally arterial anastomoses exist, but he observed no arteriovenous anastomoses. I similarly observed no arteriovenous anastomoses, yet as discussed earlier, Elsner (1966) claimed to have found arteriovenous anastomose s in the brachial vascular bundle of the dugong. Considering what we know about vessel growth and pattern development, it is possible that the three described vascular patte rns in marine mammals are the result of modification of a single type of primary vascu lar plexus. If this is indeed the case, the mechanisms of angiogenesis are likely to be sim ilar in all cases, up to the formation of the primary plexus. By c hanging the type of trimming of the primary vascular plexus, any number of configurations can be produc ed. As mentioned previously, it seems prudent for evolution to favor the most economical method for generating these structures. Therefore, using modifications of a singl e primary mechanism to form alternate patterns seems like a good idea. For instance, it may be possible to form the thoracic rete and the vascular bundle by trimming of the primary vascular plexus. To complete the vascular bundle and generate the straight vessels it contains, the structure would then have to be stretched out, as what may happen as the em bryo grows. A
320 gross morphological study of the vascular bundles in embryonic and fetal specimens may help shed light on this. Finding more conv oluted embryonic structures in place of the linear adult vascular bundles might s upport such a mechanism of development. Interestingly, the venous component of t he adult infraorbital vascular bundle of the manatee had a non-traditional branching pattern as it coalesced into its parent maxillary veins. Instead of the organized, broomlike branching pattern of the brachial, intercostal, and caudal vascular bundles, the proximal veins of the infraorbital and mandibular bundles coalesced into a rete -like plexus before becoming the maxillary veins. I found this to be unique to the mandi bular and infraorbital bundles and slightly at odds with the defining criteria of a vascular bundle put forth by Fawcett (1942). The broom-like bundles in the pectoral, intercostal, and caudal regions have a relatively planar or dorsoventral flatteni ng, while the cephalic vascular bundles are both relatively round. It is possible that this discrepancy in the proximal branching pattern is simply reflective of the cross-sectional shape of the bundle, since a broom-like branching pattern may not accommodate a round crosssectional structure and vice versa. It is unclear what role the extracellula r matrix plays in the case of vascular bundles, but unlike retia vascular bundles tend to be embedded in connective tissue (Fawcett, 1942). Since it has been shown that extracellular matrix can have significant influence on vessel development, it is possibl e that this association also aids in differentiation of the primary vascular pl exus into a vascular bundle rather than a rete I should however note that Fawcetts statem ent about connective tissue associated with vascular bundles may at least in part be in er ror, since much of the tissue he referred to was likely the venous counterpar t of the bundle whose presence he was not aware of.
321 The work conducted by Evans (1909) may shed some light on part of this picture. In his diagram of the developm ent of the anterior (ventral) spi nal artery of the pig he shows a vascular network on either of the two ventrolateral aspects of the spinal cord, with essentially no connections between the left and right sides. The subsequent images shows numerous fine capillary sprouts connec ting the two capillary beds across the midline. It is therefore not a challenge to envision simila r development of the arterial epidural retia being arrested in the first stage that precedes the formation of the midline connections, thereby retain ing the median segregation th roughout the length of the epidural retia as they enter the calvarium and suppl y the brain as the spinal meningeal arteries (Ommaney, 1938; Slijper, 1936; Fawcett, 1942; Murie, 1974; McFarland et al., 1979). As my findings suggest, the venous syst em in certain parts of the epidural retia of dolphins and manatees also maintain a certain degree of bilateral segregation. It seems reasonable at this stage to suggest, from a purely morphological standpoint, that the form ation of the spinal retia of cetaceans and sirenians may in fact represent an arrested early stage of development of the ventral spinal artery. This is further supported by the fact that in manatees at leas t, there appears to be no basilar artery formed by the confluenc e of the vertebral and ventral spinal arteries, but instead the spinal retia are joined by separate vertebral arte ries. Similarly, we know from numerous terrestrial species, that the subclavian arteries form at least in part from reduction and simplification of the dorsal intersegmental arteries--predominantly the 7th cervical dorsal intersegmental arteries, though McFarland et al (1979) state that the 6th intersegmental artery develops into the dist al part of the subclavian artery. These intersegmental arteries are bilaterally-paire d dorsal sprouts of the aorta which form
322 numerous anastomoses resulting in a rather tortuous plexus. The more rostrallylocated cervical dorsal intersegmental arteries anastomose between each other and subsequently lose their ventral aortic connecti ons to form the vertebral arteries, while the more caudal thoracic ones form the adult dorsal intercostal arteries (Wollard, 1936; Noden & deLahunta, 1985). It is conceivable then--both in location and origin--that these anastomotic dorsal intersegmental arteries may eventually form the thoracic retia observed in cetaceans and the intercostal vascular bundles observed in sirenians. Indeed the main arterial supply of the thoracic rete in the dolphin is the supreme in tercostal (via the costocervical trunk) artery, which is thought to be an elaborat ion of the first inte rcostal artery. So perhaps the thoracic rete represents elaborate anasto moses between the dorsal intersegmental arteries, and thei r connections to the epidural rete are remnants of the anastomotic connections between the vertebral and epidural networks. McFarland et al. (1979) state that Slijper (1936) proposed two mechanisms by which the cetacean retia developed, the first being thro ugh maintenance of embryonic plexuses and the second being through vascular sprouting and elaboration of preexisting plexuses. They further state t hat Slijper believed the internal carotid rete and part of the thoracic rete of cetaceans were formed by sprouting, while many of the venous retia of cetaceans were formed by persistenc e of embryonic plexuses. It is hard to know exactly which parts Slijper meant and many venous patterns do seem less organized, however, it seems more likely t hat a combination of both processes is responsible for generating the observed patterns. For instance, it seems wasteful and contrary to general evolutionary trends of conservation of energy to build a vascular
323 network only to tear it down and rebuild it again. It would instead seem prudent to differentiate a preexisting network into either arterial or venous, and then use selective trimming and sprouting to modify it. In fact, embryonic development seems to be dominated by gradual growth and change, which in principle would be better suit ed to similar development of vasculature. A cetacean that is building a wholly arterial plexus like the thoracospinal rete which will replace carotid blood supply to the brain wo uld therefore seem better served to follow this pattern of developmen t, rather than to degenerate and/ or coalesce the entire embryonic vascular plexus into a parent vessel (e.g. internal carotid artery) and then resprout another plexus to compensate for degener ation of the internal carotid. Clearly these processes will depend on the timing of various events. If the growth and development of tissues is not carefully tim ed with the various vascular changes, there is likely to be incongruence between vascular changes and tissue demands, resulting in insufficient vascular supply and drainage fields Indeed it is seen that vascular changes tend to follow tissue changes, and as mentioned earlier tend to be very responsive to tissue demands. It therefore follows that va scular development of structures such as the thoracospinal rete is likely the product of numerous simultaneous processes, including differentiation, trimming, and sprouting. Having said that, there is ample evidenc e that during development, blood supply to the brain of mammals is subject to considerab le phylogenetic plasticity. By this I mean that although there are evolutionarily predetermined vascular patterns that will eventually develop as the mature or adult fo rm, phylogenetic recapitulation is strongly present and tissue changes are necessary to eliminate or reverse the recapitulated
324 traits. In other word s, this recapitulation is subject to the temporal, spatial, and genetic forces that drive differentiation into the adult form. In order to form a complete and elaborate thoracospinal rete it may be necessary to limit al ternate arterial supply to the brain. If recapitulated struct ures such as the internal ca rotid and stapedial arteries are not eliminated, it is possible t he development of the thoracic rete may be hindered. Therefore although the adult fo rm may be genetically steered toward a thoracospinal rete it may not be possible to form it complete ly without the relative obliteration of the carotid supply. Indeed, McFarland et al. state that the brain is supplied with blood solely by the internal carotid arterial system early in ontogeny, but as the brain grow s in size its increasing blood supply needs are met by development of an additi onal route, the vertebral arterial system, in most mammals. Furthermo re, the pathways can be altered as evidenced by the regression of the stapedial and primitive trigeminal arteries while new blood supply pathways are also introduced. (1979) This embryonic stapedial artery is present even in cetac eans, despite the fact that they even regress its parent internal caroti d artery. The regre ssion of the stapedial artery may be due to ossifica tion of the tympanic cavity and therefore be limited by physiological/anatomic changes, while regression of the internal carotid may be more related to phylogenetic signals? We also know from some artiodactyl species that internal carotid arterial supply to the brain has been replaced by external carotid supply, and in the phylogenetically related cetaceans carotid supply to the brain has been replaced completely with thoracospinal retial supply. It is therefore my assumption that structures like the thoracospinal retia of cetaceans do not form due any single process but rather due to a complex interplay of messages including local and global tissue demands, genetically predetermined pa tterns, and the energetic budget.
325 It should be noted that not all looping va scular beds form from reduction or simplification of pre-existing primary vascular beds. Through injection of India ink in the heart of developing opossums, Wislocki (1939) showed evidence that the capillary loops investing the medulla and spinal cord devel op from paired arteriovenous outgrowths of an already reduced vascular bed. Wislocki elegantly showed regions of capillary sprouting consistent with what is now co mmonly identified as sprouting angiogenesis. He showed that arteries and their concomitan t veins sprouted concurrently in a dorsad manner from the ventral trunk arteries and vein s of the pial plexus, gradually investing deeper into the nervous tissue. From the primary capillary branches form lateral branches which develop in a similar fashion, investing nervous tissue between capillary loops, which is probably determined by the metabolic needs of the surrounding brain tissue (Wislocki, 1939). The various spr outsdorsal and lateral--finally merge to complete the arteriovenous capillary circuit. Though this provides strong evidence for spontaneous formation of arteriovenous loops, it is important to note t hat this manner of growth is occurring in small capillary beds, no t in arteriovenous structures comprised of large caliber vessels, and may therefore be limited to capillary beds. Function of Marine Mammal Vasculature It has been suggested by researchers that the thoracic, cervical, and spinal rete of cetaceans serves a pressure damping role for arterial blood destined for the brain (Nagel et al. 1968). The significantly shortened ne ck of cetaceans is thought to pose a risk of central nervous system (CNS) damage from the hearts pressure pulse. The argument then is that by forcing arterial blood from the heart to ta ke a path away from the brain and through a mass of convoluted, elastic arteries befor e entering the brain case, the pressure pulse can be absorbed by the vessel walls (e.g. windkessel effect)
326 (Shadwick & Gosline, 1994). This seems like a reasonable argument for it appears as though only the marine mammals with s hortened neckscetaceans and sirenians have these vascular retia while the long necked pinnipeds (seals and sea lions) do not. Nonetheless, there are some potential pr oblems with this theory. Though much of the brains arterial supply in manatees and adult cetaceans is indeed through these retia there are notable contributions from the internal and/or ex ternal carotid arteries in manatees and fetal cetaceans. This may mean that the pressure pulse is not as pronounced as has been previously suggested, despite the short neck. Another confounding variable is the issue of blood redi stribution during a dive It may be safe to assume that since manatees do not dive deep, redistribution is modest; however, cetaceans are known to dive to depths quite likely to lead to blood redistribution. Additionally, it has been shown that some of the sequellae of the dive response are peripheral vasoconstriction (Irving et al. 1938, 1941; Elsner et al. 1966) and maintenance of blood supply to the heart and CNS It is therefore possible that the combination of peripheral vasoconstricti on and blood redistribution could cause blood pressure to the CNS to elevate during a dive, however Elsner et al. (1966), Irving et al. (1941) and others have shown that main arterial blood pressure does not increase during a dive. This was thought to be t he result of sustained bradycardia and decreased cardiac output. It is not clear how a fetal cetacean in the womb of a diving cetacean will be affected by the building hydrostatic pressures. Hui (1975) suggested t hat the thoracic retia act as a place for displaced abdominal blood to redistribute during a dive in order to compensate for thoracic pressurization and limit the degree of pulmonary co llapse that accompanies thor acic collapse. Though this
327 may indeed be a viable suggestion, this too has its problems. Firstly, blood in the cervical and spinal retia would not benefit the pleural cavi ties during pressurization, so why have elaborate retia in the cervical and spinal regions, unless the volume of displaced blood warrants it? Secondly since the thoracic retia are directly connected to the cervical and spinal retia pressurization of the thoracic retia during thoracic collapse should lead to elevated pressures in the cervical and spinal retia and therefore pressurize the blood going to the CNS. This might be considered a detrimental effect however perhaps this helps with pumping of the blood to the brain during the bradycardic event. Finally, mana tees have cervical and spinal retia without having thoracic retia. Considering their lackluster diving performance, one might think it unlikely that much blood redistribution occurs. However, their massive pleural cavities may be more prone to redistribution due to the large air volume subject to compression. It is unknown what role the relatively rigid, non-compliant thorax pl ays in this. Their relatively distant phylogenetic proximity to ce taceans also begs the question of why they too have these vascular structures? Is it co nvergent evolution or phylogenetic proximity that produced that similarity and is it functional relative to short ne ck or diving? So are the thoracic, cervical, and spinal retia derived or basal characteristics?
328 Table 6-1. Structure labels and their names. Structure Label Structure Name 1 External jugular vein 2 Facial vein 3 Maxillary vein 4 Mandibular labial vein 5 Maxillary labial vein 6 Lateral nasal vein 7 Dorsal nasal vein 8 Angularis oculi vein 9 Mandibular alveolar vein 10 Deep facial vein 11 Superficial temporal vein 12 Transverse facial vein 13 Ventral masseteric vein 14 Caudal auricular vein
329 Figure 6-1. Simplified schemat ic representations of the s uperficial veins of the head of the cow (A) and horse (B) compared to the bottlenose dolphin (C) and Florida manat ee (D). Note that desp ite the significant skull modifications of the dolphin, numerous veins are still recognizable as corresponding to those seen in domestic mammals. Horse and cow images were adapted from Schummer et al. (1981).
330 LIST OF REFERENCES Achen MG, Stacker SA (1998) The vascular endothelial grow th factor family; proteins which guide the development of the vasculature. Int J Exp Pathol 79 255-265. Alon T, Hemo I, Itin A, et al. (1995) Vascular endothelial growth factor acts as a survival factor for newly fo rmed retinal vessels and has im plications for retinopathy of prematurity. Nat Med 1, 1024-1028. Andersen JE (ed) (1978) Grant's Atlas of Anatomy, 7th edition. Waverl y Pr. Baltimore, USA. Armaly M, Araki M (1975) Effect of ocular pressure on choroidal circulation in the cat and Rhesus monkey. Invest Ophth 14, 584-591. Asahara T, Murohara T, Sullivan A, et al. (1997) Isolation of putative progenitor endothelial cells for angiogenesis. Science 275, 965-967. Asahara T, Takahashi T, Masuda H, et al. (1999) VEGF contributes to postnatal neovascularization by mobilizing bone marro w-derives endothelial progenitor cells. EMBO J 18, 3964-3972. Aslan K, Krtl zcan S, et al. (2005) Morphometric c haracteristics of the ophthalmic rete in the goose living at high altitudes and cold climate conditions. Anat Histol Embryol 34, 294-297. Baev NI, Wong JH, Awad IA (1999) Vasculogenesis and angiogenesis in relation to arteriovenous malformations of the brain, In: Arteriovenous malformations in functional areas of the brain (ed. Yamada S), pp.241248. New York: Futura Publishing Co. Bahram M (2001) The Monro-Kellie hypothesis: applications in CSF volume depletion. Neurology 56, 1746-1748. Baker MA (1972) Influence of the carotid rete on brain temperature in cats exposed to hot environments. J Physiol 220, 711-728. Baker MA (1979) A brain-cooling system in mammals. Sci Am 240, 130-139. Baker MA (1982) Brain cooling in endot herms in heat and exercise. Ann Rev Physiol 44, 85-96. Baker MA, Chapman LW (1977) Rapid brain cooling in exercising dogs. Science 195, 781-783. Baker MA, Chapman LW, Nathanson M (1974) Control of brain temperature in dogs: effects of tracheostomy. Resp Physiol 22, 325-333.
331 Baker MA, Hayward JN (1968a) Intracranial heat exchange and regulation of brain temperature in sheep. Life Sci 7, 349-357. Baker MA, Hayward JN (1968b) The influence of the nasal mucosa and the carotid rete upon hypothalamic temperature in sheep. J Physiol 198, 561-579. Baldwin BA, Bell FR (1963a) Blood flow in the caroti d and vertebral arteries of the sheep and calf. J Physiol 167, 448-462. Baldwin BA, Bell FR (1963b) The effect of the blood pr essure in the sheep and calf of clamping some of the arteries contri buting to the cephalic circulation. J Physiol 167, 463-479. Baldwin BA, Bell FR (1963c) The anatomy of the cerebr al circulation of the sheep and ox. The dynamic distribution of the bl ood supplied by the carotid and vertebral arteries to the cranial region. J Anat 97, 203-215. Barbieri M, McLellan WA, Wells RS, et al. (2010) Using infrared thermography to assess seasonal trends in dorsal fin su rface temperatures of free-swimming bottlenose dolphins ( Tursiops truncatus) in Sarasota Bay, Florida. Mar Mammal Sci 26, 53-66. Barnett CH, Harrison RJ, Tomlinson JD (1958) Variations in the venous systems of mammals. Biol Rev Camb Philos 33, 442-487. Batson O (1940) The function of the vertebral ve ins and their role in the spread of metastases, Ann Surg 112, 138-149. Batson O (1956) The vertebral vein system. Am J Roentgenol 78, 195-212. Batson O (1960) The Valsalva maneuver and the vertebral vein system. Angiology 11, 443-447. Bazett H, Love L, Netwon M, Ei senberg L, Day R, Forster R (1948) Temperature changes in blood flowing in arteries and veins in man. J App Physiol 1, 3-19. Bazett H, Mendelson ES, Love L, Libet B (1948) Precooling of blood in the arteries, effective heat capacity and evaporative cooling as factors modifying cooling of the extremities. J App Physiol 1, 169-182. Beauregar d H (1894) Recherches sur l'appareil auditif chez les Mammiferes. Journal de l'anatomie et de la physiologie normale s et pathologiques de l'homme et des animaux 367-413. Benjamin LE, Keschet E (1997) Conditional switching of vascular endothelial growth factor (VEGF) expression in tumors: induction of endothelial cell shedding and regression of hemangioblastoma-lik e vessels by VEGF withdrawal. Proc Natl Acad Sci 94, 8761-8766.
332 Bernaldo de Quirs Y, Gonzlez -Diaz O, Saavedra P, et al. (2011) Methodology for in situ gas sampling, transport and laborator y analysis of gases from stranded cetaceans. Sci Rep 1, 1-10. Bernaldo de Quirs Y, Gonzl ez-Diaz O, Arbelo M, Sierra E, Sacchini S, Fernndez A (2012) Decompression vs. decompositi on: distribution, amount, and gas composition of bubbles in stranded marine mammals. Front Physiol 3, 1-19. Blix AS (2011) The venous system of seals, with new ideas on the significance of the extradural intravertebral vein (Commentary). J Exp Biol 214, 3507-3510. Blix AS, Walle L, Messelt EB, Folkow LP (2010) Selective brain cooling and its vascular basis in diving seals. J Exp Biol 213, 2610-2616. Boenninghaus FWG (1904) Das ohr des zahnwales, z ugleich ein bertieg zur theorie du schdleiting. Zool Jahrb 19, 189-360. Boisseau MR (1997) Valvules veineuses des members inferieurs: problemes hemodynamicques, blologiques et relations physiopathologiques. J Malad Vascul 22, 122-127. Bornmyr S, Castenfors J, Evander E, et al. (2001) Effect of local cold provocation on systolic blood pressure and skin blood flow in the finger. Clin Physiol 21, 570-575. Bossart GD, Meisner RA, Rommel SA, et al. (2002) Pathological features of the Florida manatee cold stress syndrome. Aquat Mammal 29, 9-17. Brecher GA (1956) Venous Return New York: Grune and Stratton. Breschet MG (1819) Essai sur les ve ine dur rechis. Paris: Mquignon-Marvis Breschet MG (1836) Histoire anatomique et physiologique dun organe de nature vasculaire decouvert dans les cetaces; su ive de quelques considerations sue la repiration de ces animaux et des amphibies Paris: Brechet Jeune. Broderick BJ, Corley GJ, Quondamatteo F, Breen PP, Serrador J, Laighin G (2010) Venous emptying from the foot: influences of weight bearing, toe curls, electrical stimulation, pa ssive compression, and posture. J Appl Physiol 109, 1045-1052. Bron KM, Murdaugh HV, Millen JE, Lenthall R, Raskin P, Robin ED (1966) Arterial constrictor response in a diving mammal. S cience 152, 540-543 Brooks B, Duncan GW (1940) The effects of temperat ure on the survival of anemic tissue. Ann Surgery 112, 130-137. Brown CM, Sanya EO, Hilz MJ (2003) Effect of cold face stimulation on cerebral blood flow in humans. Brain Res Bull 61, 81-86.
333 Brubakk AO, Vik A, Flook V (1999) Gas bubbles and the lungs. In: The Lung at Depth. (eds. Lundgren CEG, Miller JN), pp. 237-294. New York: Marcel Dekker Cao Y, Chen H, Zhou L, et al (1996) Heterodimers of plac enta growth factor-vascular endothelial growth factor: endothelial activity, tumor cell expression, and high affinity binding to Flk-1/KDR. J Biol Chem 271, 3154-3162. Cao X-J, Oertel D (2005) Temperature affects voltage-sensitive conductances differentially in octopus cells of the mammalian cochlear nucleus. J Neurophysiol 94, 821-832. Carey FG, Gibson QH (1983) Heat and oxygen exchange in the rete mirabile of the bluefin tuna, Thunnus thynnus Comp Biochem Physiol 74A, 333-342. Carey, F.G., Teal, J.M., Kanwisher, J.W., et al. (1971) Warm-bodied fish. Am Zool 11, 135-143. Carlisle HJ, Ingram DL (1973) The effects of heating and cooling the spinal cord and hypothalamus on thermoregulat ory behaviour in the pig. J Physiol 231, 353-364. Carmeliet P, Moons L, Dewerchin M, et al. (1996) Insights in vessel development and vascular disorders using targeted inactiva tion and transfer of vascular endothelial growth factor, the tissue factor receptor, and the plasminogen system. Ann NY Acad Sci 811, 191-206. Caputa M, Cabanac M (1979) Bradycardia during face cooling in man may be produced by selective brain cooling. J Appl Physiol 46, 905-907. Caputa M, Folkow L, Blix AS (1998) Rapid brain cooling in diving ducks. Am J Physiol 275, R363-R371. Caputa M, Kadziela W, Narebski J (1976) Significance of cr anial circulation for the brain homeothermia in rabbits: II the role of the cranial venous lakes in the defence against hyperthermia. Acta Neurobio. Exp 36, 625-638. Chaynes P (2003) Microsurgical anatom y of the great cerebral vein of Galen and its tributaries. J Neurosurg 99, 1028-1038. Clendenin MA, Conrad MC (1979) Collateral vessel development, following unilateral chronic carotid occlusion in the dog. Am J Vet Res 40, 84-88. Costa DP, Croxall JP, Duck C (1989) Foraging energetics of Antarctic fur seals, Arctocephalus gazella in relation to changes in prey availability. Ecology 70, 596606. Costa DP, Trillmich F (1988) Mass changes and metabolis m during the perinatal fast: a comparison between Antarctic ( Arctocephalus gazella ) and Glapagos fur seals ( A. galapagoensis ). Physiol Zool 61, 160-169.
334 Costidis A, Rommel SA (2012) Vascularization of air sinuses and fat bodies in the head of the bottlenose dolphin (Tursiops trunc atus): morphological implications on physiology. Front Physiol 3, 243. doi: 10.3389/fphys.2012.00243 Cox TM, Ragen TJ, Read AJ, et al. (2006) Understanding the impacts of anthropogenic sound on beaked whales. J Cetacean Res Manage 7, 177-187. Cozzi B, Bagnoli P, Acocella F, Costantino ML (2005) Structure and biomechanical properties of the trac hea of the striped dolphin Stenella coeruleoalba : evidence for evolutionary adaptations to diving. Anat Rec 284, 500-510. Cranford TW, Amundin M, Norris KS (1996) Functional morphology and homology in the Odontocete nasal complex: im plications for sound generation. J Morphol 228, 223-285. Cranford TW, Krysl P, Hildebrand JA (2008) Acoustic pathways revealed: simulated sound transmission and reception in Cuviers beaked whale ( Ziphius cavirostris) Bioinsp. Biomim 3, 1-10. Cranford TW, McKenna MF, Soldevilla MS, et al. (2008) Anatomic geometry of sound transmission and reception in Cuviers beaked whale ( Ziphius cavirostris ). Anat Rec 291, 353-378. Dehnhardt G, Mauck B, Hyvrinen H (1998) Ambient temperat ure does not affect the tactile sensitivity of mystacial vibrissae in harbour seals. J Exp Biol 201, 30233029. de la Torre E, Mitchell OC, Netsky MG (1962) Anatomic and angi ographic study of the vertebral-basilar arterial system in the dog. Am J Anat 110, 187-197. de la Torre E, Netsky MG, Meschan I (1959) Intracranial and extracranial circulations in the dog: anatomic and andgiographic studies. Am J Anat 105, 343-382. DiSalvo J, Bayne ML, Conn G, et al. (1995) Purification and characterization of a naturally-occuring vascular endothelial gr owth-factor placenta growt-factor heterdimer. J Biol Chem 270, 7717-7723. Djonov V, Baum O, Burri H (2003) Vascular remodeling by intussusceptive angiogenesis. Cell Tissue Res 314, 107-117. Djonov VG, Kurz H, Burri PH (2002) Optimality in the developing vascular system: branching remodeling by means of intu ssusception as an efficient adaptation mechanism. Dev Dynam 224, 391-402. Domning DP (1978) The myology of the A mazonian manatee, Trichechus inunguis (Natterer) (Mammalia: Sirenia). Acta Amazon 8(Supp), 1-81.
335 Dormer KJ, Denn MJ, Stone HL (1977) Cerebral blood flow in the sea lion ( Zalophus californianus ) during voluntary diving. Comp Biochem Physiol 58A, 11-18. Du Boulay GH, Verity PM (1973) The Cranial Arteries of Mammals. London: Heinemann Medical. Dyce KM, Sack WO, Wensing CJG (2002) Textbook of Veterinary Anatomy,3rd Ed. Philadelphia: Saunders. Echols MP (1984) The venous drainage of the thor acic flipper of the West Indian manatee. Senior Thesis, University of Florida, Gainesville, Florida. 20pp. Eichmann A, Le Noble F, Autiero M, Carmeliet P (2005) Guidance of vascular and neural network formation. Curr Opin Neurobiol 15, 108-115. Eisele R, Weichert E, Eren A, Kinzl L (2001) The effect of partial and full weightbearing on venous retun in the lower limb. J Bone Joint Surg 83B, 1037-1040. Elsner R (1966) Cardiovascular adjustments to diving. In: The Biology of Marine Mammals, (ed. Andersen HT), p.117-145. New York: Academic Press. Elsner R, Hanafee WN, Hammond DD (1971) Angiography of the inferior vena cava of the harbor seal during simulated diving. Am J Physiol 220, 1155-1157. Elsner R, Pirie J, Kenny DD, Schemmer S (1974) Functional circulatory anatomy of cetacean appendages. In: Functional Anatomy of Marine Mammals Vol 2 (ed Harrison RJ), p. 143-160. New York, Academic Press. Elsner R, Wright J, Hornbein TF (1964) Blood flow distri bution in diving mammals. J Fed Proc 23, 111. Epstein HM, Linde HW, Crampt on AR, Ciric IS, Eckenhoff JE (1970) The vertebral venous plexus as a major cer ebral venous outflow tract. Anesthesiology 32, 332337. Espinosa A, Reid RJ, Jaber JR, Mart in V, Cunningham AA, Fernndez A (2003) Gas-bubble lesions in stranded animals: wa s sonar responsible for a spate of whale deaths after an Atlant ic military exercise? Nature 425, 575-576. Evans HM (1909) On the development of the aort ae, cardinal and umbilical veins, and the other blood vessels of vertebr ate embryos from capillaries. Anat Rec 3, 498518. Evans HE (1993) Chapter 12: Veins, In Millers Anatomy of the Dog, 3rd edn. (ed. Evans HE), pp. 682-716. Philadel phia: W.B. Saunders).
336 Evans DL, England G (2001) Joint interim report on the Bahamas marine mammal stranding event of 15-16 March 2000 (December 2001) US Department of Commerce. 59pp. http://www.nmfs.noaa.gov/pr/pdfs/ health/stranding _bahamas2000.pdf Fawcett DW (1942) A comparative study of bl ood-vascular bundles in the Florida manatee ( Trhichechus latirostris) and in certain cetaceans and edentates. J Morphol 71, 105-133. Fahlman A, Olszowka A, Bolstrom B, et al. (2006) Deep diving mammals: Dive behavior and circulatory adjustments contribute to bends avoidance. Resp Physiol Neurobiol 153, 66-77. Fahlman A, Hooker SK, Olszokwa A, et al. (2009) Estimating the effect of lung collapse and pulmonary shunt on gas exch ange during breath-hold diving: the Scholander and Kooyman legacy. Resp Physiol Neurobi 165, 28-39. Feindel W, Perot P (1965) Red cerebral veins; a report on arteriovenous shunts in tumors and cerebral scars. J Neurosurg 22, 315-325. Fernandez A, Arbelo M, Deaville R, et al. (2004) Beaked whales, sonar and decompression sickness. Nature 428, U1-2. Fernndez A, Edwards, JF Rodriguez F, et al. (2005) Gas and fat embolic syndrome involving a mass strand ing of beaked whales (family Ziphiidae ) exposed to anthropogenic sonar signals. Vet Pathol 42, 446-457. Ferrara N, Houck KA, Jakeman LB, et al. (1991) The vascular endothelial growthfactor family of polyptides. J Cell Biochem 47, 211-218. Ferrara N (2001) Role of vascular endothelial growth factor in regulation of physiological angiogenesis. Am J Physiol-Cell Ph 280, C1358-C1366. Flamme I, Frolich T, Risau W (1997) Molecular mechanism s of vasculogenesis and embryonic angiogenesis. J Cell Physiol 173, 206-210. Flamme I, von Reutern M, Drexler HCA, et al (1995) Overexpression of vascular endothelial growth factor in the avian embryo induces hypervascularization and increased vascular permeability without alterati ons of embryonic pattern formation. Dev Biol 171, 399-414. Flower WH (1867) On the Osteology of the Chachalot or Sperm-Whale (Physeter macrocephalus). Trans. Zool. Soc. Lon 6, 309-372. Folkow B, Fox RH, Krog J, Odelram H, Thorn O (1963) Studies on the reactions of the cutaneous vessels to cold exposure. Acta Physiol Scand 58, 342-354.
337 Folkow LP, Blix AS, Eide TJ (1988) Anatomical and functi onal aspects of the nasal mucosal and opthalmic retia of phocid seals. J Zool 216, 417-436. Frantzis A (1998) Does acoustic testing strand whales? Nature 392, 29. Fraser FC, Purves PE (1960) Hearing in cetaceans: evolution of the accessory air sacs and the structure and function of the outer and middle ear in recent cetaceans. Bull Br Mus Nat Hist Zool 7, 1-140 + 53 pls. Fritches KA, Brill RW, Warrant EJ (2005) Warm eyes provide superior vision in swordfishes. Curr Biol 15, 55-58. Fukuta K, Kudo H, Sasaki M, et al. (2007) Absence of carotid rete mirabile in small tropical ruminants: implications for t he evolution of the arterial system in artiodactyls. J Anat 210, 112-116. Fuller CA, Baker MA (1983) Selective regulation of br ain and body temperatures in the squirrel monkey. Am J Physiol 245, R293-R297. Fuller A, Kamerman PR, Maloney SK, et al. (2003) Variability in brain and arterial blood temperatures in free-ranging os triches in their natural habitat. J Exp Biol 206, 1171-1181. Fujii K, Lenkey C, Rhoton AL (1980) Microsurgical anatomy of the choroidal arteries: Lateral and third ventricles. J Neurosurg 52, 165-188. Galliano RE, Morgane PJ, McFarland WL, et al. (1966) The anatomy of the cervicothoracic arterial system in the bottlenose dolphin ( Tursiops truncatus) with a surgical approach suit able for guided angiography. Anat Rec 155, 325-338. Gallivan GJ, Best RC (1980) Metabolism and respirat ion of the Amazonian manatee ( Trichechus inunguis ). Physiol Zool 53, 245-253. Ghoshal NG, Koch T, Pospesko P (1981) The Venous Drainage of the Domestic Animals Philadelphia: Saunders. Gillilan LA (1957) General principles of the arterial blood vesse l patterns to the brain. T Am Neurol Assoc 82, 65-68. Gillilan LA (1959) Significant superficial anastomoses in the arteri al blood supply to the human brain. J Comp Neurol 112, 55-74. Gillilan LA (1967) A comparative study of the ex trinsic and intrinsic arterial blood supply to brains of submammalian vertebrates. J Comp Neurol 130, 175-196. Gillilan LA (1972) Blood supply to primitive mammalian brains. J Comp Neurol 145, 209-222.
338 Gillilan LA (1974) Blood supply to brains of ungulates wit h and without a rete mirabile caroticum. J Comp Neurol 153, 275-290. Gillilan LA, Markesbery WR (1963) Arteriovenous shunts in blood supply to brains of some common laboratory animals-with special attention to rete mirabile conjucatum in cat. J Comp Neurol 121, 305-311. Gdde R, Kurz H (2001) Structural and biophysical simulation of angiogenesis and vascular remodeling. Dev Dynam 220, 387-401. Gregor M, Jnig W, Riedel W (1976) Response pattern of cutaneous postganglionic neurons to the hindlimb on spinal co rd heating and cooling in the cat. Pflg Arch 38, 135-140. Groen RJM, Ponssen H (1991) Vascular anatomy of the spinal epidural space: considerations on the etiology of t he spontaneous spinal epidural hematoma. Clin Anat 4, 413-420. Groves RW, Wilson-Jones E, MacDonald DM (1991) Human orf and milkers nodule: a clinicopathologic study. J Am Acad Dermatol 25, 706-711 Gupta V, Tuli A, Choudhry R, Agarwal S, Mangal A (2003) Facial vein draining into external jugular vein in humans: its variations, phylogen etic retention and clinical relevance. Surg Radiol Anat 25, 36-41. Hales JRS (1983) Thermoregulatory requirement s for circulatory adjustments to promote heat loss in animals. J Theor Biol 8, 219-224. Hammel HT, Elsner RW, Heller HC, Maggert JA, Bainton CR (1977) Thermoregulatory responses to altering hypothalamic temperature in the harbor seal. Am J Physiol 232, R18-R26. Hammel HT, Maggert J, Kaul R (1976) Effects of altering sp inal cord temperature on temperature regulation in the Adelie penguin, Pygoscelis adeliae Pflg Arch 362, 1-6. Hampton IFG, Whittow GC, Szekerczes J, Rutherford S (1971) Heat transfer and body temperature in the Atlantic bottlenose dolphin, Tursiops truncatus In t J Biometeorol 15, 247-253. Harper CJ, McLellan WA, Rommel SA Gay DM, Dillaman RM, Pabst DA (2008) Morphology of the melon and its tendinous connections to the facial muscles in bottlenose dolphins ( Tursiops truncatus). J Morphol 269, 820-839. Harrison RJ, Kooyman GL (1971) Diving in marine mammals. In: Oxford Biology Readers No. 6 (ed. Head JJ). Carolina Biological Supply Co.: Burlington. p. 1 16.
339 Harrison RJ, Tomlinson JD (1956) Observations on the venous system in certain Pinnipedia and Cetacea. P Zool Soc Lond 126, 205-233. Hasegawa Y, Ono T, Sakagami J, et al. (2011) Influence of vo luntary control of masticatory side and rhythm on cerebral hemodynamics. Clin Oral Invest 15, 113118. Hassler W, Thron A, Grote EH (1989) Hemodynamics of spinal dural arteriovenous fistulas; an intraoperative study. J Neurosurg 70, 360-370. Hayward JN, Baker MA (1969) A comparative st udy of the role of the cerebral arterial blood in the regulation of brain temperature in five mammals. Brain Res 16, 417440. Hegedus SA, Shackelford RT (1965) A comparative anatom ical study of the craniocervical venous systems in mammals, with special reference to the dog: relationship of anatomy to measurements of cerebral blood flow. Am J Anat 116, 375-386. Henkes H, Reinartz J, Fischer S, et al. (2007) Rete mirabile bei segmentaler Agenesie der A. carotis interna. Nervenarzt 78, 948-953. Herlihy WF (1947) Revisions of the venous system: the role of the vertebral veins. Med J Australia 1, 661-672. Heuser CH (1923) The branchial vessels and their derivatives in the pig. Contrib Embryol 15, 121-131. Hilario JD, Rodino-Kalpac LR, Wang C, Bettie CE (2009) Semaphorin 5A is a bifunctional axon guidance cue fo r axial motoneurons in vivo. Dev Biol 326, 190200. Hokkanen JEI (1990) Temperature regulat ion of marine mammals. J Theor Biol 145, 465-485. Hol R, Blix AS, Myhre HO (1975) Selective redistributi on of the blood volume in the diving seal ( Pagophylus groenlandicus ). Rap Proces 169, 423-432. Holliday CM, Ridgely RC, Balanoff AM, Witmer LM (2006) Cephalic vascular anatomy in flamingos ( Phoenicopterus ruber ) based on novel vascular injection and computed tomographic imag ing analyses. Anat Rec 288, 1031-1041. Holom VH, Messlinger K, Fischer MJM (2008) Temperature-dependent neuronal regulation of arterial blood flow in ra t cranial dura mater. J Neurosci Res 86, 158164.
340 Hoogland PVJM, Vorster W, Groen RJM, Kotz SH (2012) Possible thermoregulatory functions of the internal vertebral venous plexus in man and various other mammals: evidence from comparative anatomical studies. Clin Anat 25, 452-460. Hooker SK, Baird RW, Fahlman A (2009) Could beaked whales get the bends? Effect of diving behaviour and physiology on modeled gas exchange for three species: Ziphius cavirostris Mesoplodon densirostris and Hyperoodon ampullatus. Resp Physiol Neurobi 167, 235-246. Hooker, S.K., Fahlman, A. Moore, M.J., et al. (2011) Deadly diving? Physiological and behavioral management of decompression stress in diving mammals. P R Soc B Published online, doi:10.1098/rspb.2011.2088 Hosokawa H, Kamiya T (1965) Sections of the dolphins head ( Stenella caeruleoalba ). Sci Rep Whales Res Inst 19, 105-111 Houck KA, Leung DW, Rowland AM, Winer J, Ferrara N (1992) Dual regulation of vascular endothelial growth factor bi oavailability by genetic and proteolytic mechanisms. J Biol Chem 267, 26031-26037 Houser DS, Howard R, Ridgway S (2001) Can diving-induced tissue nitrogen supersaturation increase the chance of acoustically driven bubble growth in marine mammals? J Theor Biol 213, 183-95. Houser DS, Finneran J, Carder D, et al. (2004) Structural and functional imaging of bottlenose dolphin ( Tursiops truncatus ) cranial anatomy. J Exp Biol 207, 36573665. Hughes JH, Henry RE, Daly MJ (1984) Influence of ethanol and ambient temperature on skin blood flow. Ann Emergency Med 13, 597-600. Hui CA (1975) Thoracic collapse as affected by the retia thoracica in the dolphin. Resp Physiol 25, 63-70. Hunter J (1787) Observations on the st ructure and oeconomy of Whales. Phil Trans R Soc Londong 77, 371-450 Inman DR, Peruzzi P (1961) The effects of temperature on the re sponses of Pacinian corpuscles. J Physiol 155, 280-301. Irvine AB (1983) Manatee metabolism and its infl uence on distributi on in Florida. Biol Conserv 25, 315-334. Irving L (1938) Changes in blood flow through the brain and muscle during arrest of breathing. Am J Physiol 122, 207-214. Irving L, Scholander PF, Grinnell SW (1941) The regulation of arterial blood pressure in the seal during diving. Am J Physiol 135, 557-566.
341 Ishiko N, Loewenstein WR (1961) Effects of temperatur e on the generator and action potentials of a sense organ. J Gen Physiol 45, 105-124. Jessen C, Mayer ET (1971) Spinal cord and hypothalamus as core sensors of temperature in the conscious dog: I equivalence of responses. Pflg Arch 324, 189-204. Jepson PD, Arbelo M, Deaville R, et al. (2005) Acute and chronic gas bubble lesions in cetaceans stranded in the United Kingdom. Vet Pathol 42, 291-305. Kim MS, Lee SJ, Lee CH et al. (2006) Bilateral segmental absence of the internal carotid artery with rete compensation associated with absence of basilar artery: case report. Surg Neurol 65, 615-620. Kirks DR, Currarino G, Chen JTT (1986) Mediastinal collateral arteries important vessels in coarctat ion of the aorta. Am J Roentgenol 146, 757-762. K-J Li J (2004) Dynamics of the vascular system. In: Series on Bioengineering & Biomedical Engi neering-Vol.1. World Scientific. Konno K, Kurita H (2001) Carotid rete mirabile associated with aortic arch anomaly. Arch Neurol 58, 1694-1695. Koopman HN, Iverson SJ, Read AJ (2003) High concentrations of isovaleric acid in the fats of odontocetes: variation and pa tterns of accumulation in blubber vs. stability in the melon. J Comp Physiol B 173, 247-261. Koopman HN (2007) Ontogenetic, phylogenetic and ec ological factors influencing the biochemical structure of the blubber of odontocetes. Mar Biol 151, 277-291. Kooyman GL, Castellini MA, Davis RW (1981) Physiology of diving in marine mammals. Ann Rev Physiol 43, 343-356. Kooyman GL, Ponganis PJ (1998) The physiological basis of diving to depth: birds and mammals. Ann Rev Physiol 60, 19-133. Kullmann R, Schnung W, Simon E (1970) Antagonistic changes of blood flow and sympathetic activity in different vascular beds following central thermal stimulation: I. blood flow in skin, muscle and intestine during spinal cord heating and cooling in anesthetized dogs. Pflg Arch 319, 146-161. Kvadsh eim PH, Miller PJO, Tyack PL et al. (2012) Estimated tissue and blood N2 levels and risk of decompression sick ness in deep-, interm ediate-, and shallowdiving toothed whales during exposure to naval sonar. Front Physiol 3, 1-14. Laderoute KR, Alarcon RM, Brody MD et al. (2000) Opposing effe cts of hypoxia on expression of the angiogenic inhibitor th rombospondin 1 and the angiogenic inducer vascular endothelial growth factor. Clin Cancer Res 6, 2941-2950.
342 Lango T, Morland T, Brubakk AO (1996) Diffusion coefficients and solubility coefficients for gases in bi ological fluids: a review. Undersea Hyperb Med 23, 247272. Larsen WJ (2001) Chapter 8: Development of the Vasculature. In: Human Embryology (3rd Edition), pp.123-146. London: Churchill Livingstone. Launay P (1896) Veines jugulaires et arteres ca rotides chez lhomme et les animaux superieurs. Paris: Libraire s de lacademie de medicine. Lawson ND, Scheer N, Pham V, et al. (2001) Notch signaling is required for arterialvenous differentation during embryonic vascular development. Development 128, 3675-3683. Lawson ND, Weinstein BM (2002) In vivo imaging of embryonic vascular development using transgenic zebrafish. Dev Biol 248, 307-318. Le Noble F, Moyon D, Pardanaud L, et al. (2004) Flow regulates arterial-venous differentiation in the chick embryo yolk sac. Development 131, 361-375. Lightsey JD, Rommel SA, Costidis AM, Pitchford TD (2006) Methods used during gross necropsy to determine watercraft-related mortality in the Florida manatee (Trichechus manatus latirostris) J Zoo Wildlife Med 37, 262-275. Lin F, Lin R, Chou L (1998) Structure of the thoracic retia mirabilia of three cetacean species ( Mammalia: Cetacea ). Acta Zool Taiwanica 9, 111-118. Linthicum DS, Carey FG (1972) Regulation of brain and eye temperatures by the bluefin tuna. Comp Biochem Physiol 43A, 425-433. Luzzati V, Mateu L, Marquez G, Borgo M (1999) Structural and electrophysiological effects of local anesthetics and of lo w temperature on my elinated nerves: implication of the lipid chains in nerve excitability. J Mol Biol 286, 1389-1402. Lyamin OI, Manger PR, Ridgway SH, Mukhametov LM, Siegel JM (2008) Cetacean sleep: an unusual from of mammalian sleep. Neurosci Biobehav Rev 32, 14511484. Lyttle DJ, Fraser KM, Fleming SB, Mercer AA, Robinson AJ (1994) Homologs of vascular endothelial growth-factor ar e encoded by the poxvirus orf virus. J Virol 68, 84-92. Mall FP (1904) On the development of the blood-vessels of the brain in the human embryo. Am J Anat 4, 1-18. Maloney SK, Mitchell G (1997) Selective brai n cooling: role of angularis oculi vein in nasal thermoreception. Am J Physiol-Reg I 273, R1108-R1116.
343 Maloney SK, Mitchell D, Blanche D (2007) The contribution of carotid rete variability to brain temperature variability in sheep in a thermoneutral environment. Am J Physiol-Reg I 292, R1298-R1305. Marsala M, Vanicky I, Galik J, Radonak J, Kundrat I, Marsala J (1993) Panmyelic epidural cooling protects against ischemic spinal cord damage. J Surg Res 55, 2131. Marsh H, OShea TJ, Reynolds JE (2012) Affinities, origins and diversity of the Sirenia through time. In: Ecology and Conservation of the Sirenia, pp. 35-77. Cambridge University Press, Cambridge, UK. Marshall CD, Clark LA, Reep RL (1998a) The muscular hydrostat of the Florida manatee ( Trichechus manatus latirostris ) and its role in the use of perioral bristles. Mar Mammal Sci 14, 290-303. Marshall CD, Huth GD, Edmonds VM, Halin DL, Reep RL (1998b) Prehensile use of perioroal bristles during feeding and asso ciated behaviors of t he Florida manatee ( Trichechus manatus latirostris ). Mar Mammal Sci 14, 274-289. Maxia C, Scano P, Maggiani F, et al. (2007) A morphological and 13C NMR study of the extramandibular fat bodies of the striped dolphin ( Stenella coeruleoalba ). Anat Rec 290, 913-919. McFarland WL, Jacobs MS, Morgane PJ (1979) Blood supply to the brain of the dolphin, Tursiops truncatus with comparative observati on on special aspects of the cerebrovascular supply of other vertebrates. Neurosci Behav Rev 3, 1-93. McKenna MF, Cranford TW, Berta A, Pyenson ND (2011) Morphology of the odontocete melon and its implications for acoustic function. Mar Mammal Sci doi: 10.1111/j.1748-7692.2011.00526.x McNally MA, Cook EA, Mollan RAB (1997) The effect of active movement of the foot on venous blood flow after total hip replacement. J Bone Joint Surg 79A, 11981201. Mead JG (1975) Anatomy of the ex ternal nasal passages and facial complex in the Delphinidae (Mammalia: Cetacea). Smithson Contrib Zool 207, 1-72. Mead JG, Fordyce RE (2009) The Therian skull: a lexicon with emphasis on the Odontocetes. Smithson Contrib Zool 627, 1-249. Meagher EM, McLellan WA Westgate AJ Wells RS, Friers on D Jr., Pabst DA (2002) The relationship between heat flow and vasculature in the dorsal fin of wild bottlenose dolphins Tursiops truncatus. J Exp Biol 205, 3475.
344 Meagher EM, McLellan WA, Westgate AJ, Wells RS, Blum JE, Pabst DA (2008) Seasonal patterns of heat loss in wild bottlenose dolphins ( Tursiops truncatus ). J Comp Physiol B 178, 529. Melnikov VV (1997) The arterial system of the sperm whale ( Physeter macrocephalus). J Morphol 234, 37-50. Meyer M, Clauss M, Leppl e-Wienhues A, et al. (1999) A novel vascular endothelial growth factor encoded by Orf virus, VEG F-E, mediates angio genesis via signaling through VEGFR-2 (KDR) but not VEGFR-1 (Flt-1) recept or tyrosine kinases. EMBO J 18, 363-374. Meyer AA, Webster AJF (1971) Cold-induced vas odilatation in the sheep. Can J Physiol Pharm 49, 901-908. Midtgrd U (1983) Scaling of the brain and t he eye cooling system in birds: a morphometric analysis of the rete ophthalmicum J Exp Zool 225, 197-207. Miller VM, South FE (1979) Spinal cord, hypothalamic, and air temperature: interaction with arousal states in the marmot. Am J Physiol 236, R107-R116. Misunuma M, Takahashi N, Usam i A, Matsuki N, Ikegaya Y (2009) Hightemperature, but not high-pressure conditions alter neuronal activity. J Pharmacol Sci 110, 117-121. Mitchell G, Bobbitt JP, Devries S (2008) Cerebral perfusion pressure in giraffe: modelling the effects of head-raising and lowering. J Theor Biol 252, 98-108. Mitchell J, Mitchell G, Lust A (2009) Thermoregulat ory anatomy of pronghorn. Eur J Wildlife Res 55, 23-31. Moore KL, Dalley AF (2006) Chapter 7: Head. In: Clinically Oriented Anatomy (5th Edition) p886-1043. Philadelphia: Lippi ncott Williams & Wilkins. Morrison PM, Rosenmann M, Estes JA (1974) Metabolism and regulation in the sea otter. Physiol Zool 47, 218-299. Moss G (1974) The adequacy of the cerebral collateral circulat ion: tolerance of awake experimental animals to acute bilate ral common carotid artery occl usion. J Sug Res 16, 337-338. Mullan S, Mojtahedi S, Johnson D, MacDonald L (1996) Embryological basis for some aspects of cerebral vascular fistulas and malformations. J Neurosurg 85, 18. Mukhametov LM (1984) Sleep in marine mammals. Exp Brain Res 8, 227-238.
345 Mukhametov LM (1987) Unihemispheric slow-wav e sleep in the Amazonian dolphin, Inia geoffrensis Neurosci Lett 79, 128-132. Murdaugh HV, Cross CE, Millen JE et al. (1968) Dissociation of bradycardia and arterial constriction during diving in the seal Phoca vitulina Science 162, 364-365. Murie J (1872) On the form and structure of the manatee ( Manatus americanus). Trans Zool Soc Lon 8, 127-150. Murie J (1873) On the organization of the Caaing whale Globicephalus melas Trans Zool Soc Lon 8, 235-302. Murie J (1874) Researches upo n the anatomy of the Pinnipedia Part III Descriptive anatomy of the sea-lion ( Otaria jubata ). Trans Zool Soc Lon 8, 501-582. Nachtigall PE, Supin AY (2008) A false killer whale adjusts its hearing when it echolocates (review). J Exp Biol 211, 1714-1718. Nagel EL, Morgane PJ, McFarland WL, Galliano RE (1968) Rete mirabile of dolphin: its pressure-damping effect on cerebral circulation. Science 161, 898-900. Nakajima M (1961) In regard to the rete mirabile of the Cetacea: with emphasis especially on Grampidelphis griseus and Tursiops truncatus J Med Acad Toho 8, 1-23. Necker R (1983) Effects of temperature on affer ent synaptic transmission in the dorsal horn of the spinal cord of pigeons. J Therm Biol 8, 15-18. Neuman TS (1999) Pulmonary fitness for diving. In: The Lung at Depth. (eds. Lundgren CEG, Miller JN, pp. 73-90. New York: Marcel Dekker Ninomiya H (2002) Vasculature of the ophthalmic rete in night herons ( Nycticorax nycticorax ): scanning electron microscopy of corrosion casts. Vet Ophthalmol 5, 155-160. Ninomiya H, Masui M (1999) Vasculature of the orbital rete in the Japanese deer ( Cervus nippon ). Vet Ophthalmol 2, 107-112. Ninomiya H, Yoshida E (2007) Functional anatomy of t he ocular circulatory system: vascular corrosion casts of the cetacean eye. Vet Ophthalmol 10, 231-238. Noden DM, de LaHunta A (1985) The embryology of domestic animals. Developmental mechanisms and malformations. Balt imore: Williams & Wilkins. Nordgarden U, Folkow LP Walloe L, Blix AS (2000) On the direction and velocity of blood flow in the extradural intrav ertebral vein of harp seals ( Phoca groenlandica ) during simulated diving. Acta Physiol Scand 168, 271-276.
346 Noren DP, Williams TM, Berry P, Butler E (1999) Thermoregulation during swimming and diving in bottlenose dolphins, Tursiops truncatus. J Comp Physiol B 169, 93 99. Norris KS (1969) The echolocation of marine mammals. In: The Biology of Marine Mammals. (ed. Andersen HT), pp. 391-423. New York: Academic Press. Norris KS, Harvey GW (1974) Sound transmission in the porpoise head. J Acoust Soc Am 56, 659. dden Folkow LP, Caputa M, Hotvedt R, Blix AS (1999) Brain cooling in diving seals. Acta Physiol Scand 166, 77-78. Ommanney FD (1932) The vascular network ( retia mirabilia ) of the fin whale ( Balaenoptera physalus ). Discovery Rep 5, 327-62. Oshima T, Ogura M, Kikuchi T, et al. (2007) Involvement of the pterygoid venous plexus in patulous Eustachian tube symptoms. Acta Oto-Laryngol 127, 693-699. stnes JE, Bech C (1992) Thermosensitivity of different parts of the spinal-cord of the pigeon ( Columba livia ). J Exp Biol 162, 185-196. Owen R (1868) On the Anatomy of Vertebr ates. Vol. 3 Londong: Longmans, Green, 1868, pp. 513-576. Padget DH (1948) The development of the cranial arteries in the human embryo. Contrib Embryol 32, 205-261. Pabst DA, Rommel SA, McLellan WA, et al. (1995) Thermoregulation of the intraabdominal testes of the bottlenose dolphins ( Tursiops truncatus) during exercise. J Exp Biol 198, 221-226. Panneton WM, Gan Q, Le J et al. (2012) Activation of brainstem neurons by underwater diving in the rat. Front Physiol 3, 1-13. Park JE, Keller GA, Ferrara N (1993) The vascular endothelia l growth factor (VEGF) isoforms: differential deposition into t he subepithelial extracellular matrix and bioactivity of extracel lular matrix-bound VEGF. Mol Biol Cell 4, 1317-1326. Patan S (2004) Vasculogenesis and angiogenesis. Cancer Treat 117 3-32. Patan S, Haenni B, Burri PH (1996) Implementation of intussusceptive microvascular growth in the chicken chorioallant oic membrane (cam): 1. Pillar formation by folding of the capillary wall. Microvasc Res 51, 80-98.
347 Ponganis PJ, Knower-Stockard T, Levenson DH, Berg L., Baranov EA (2006) Intravascular pressure profiles in elephant seals: hypotheses on the caval sphincter, extradural vein and venous return to the heart. Comp Biochem Phys A 145, 123-130. Reep RL, Marshall CD, Stoll ML, Whitaker DM (1998) Distribution and innervations of facial bristles and hairs in the Florida manatee ( Trichechus manatus latirostris ). Mar Mammal Sci 14, 257-273. Reep RL, Stoll ML, Marshall CD, Homer BL, Samuelson DA (2001) Microanatomy of facial vibrissae in the Florida manatee: t he basis for specialized sensory function and oripulation. Brain Behav Evol 58, 1-14. Reidenberg JS, Laitman JT (2008) Sisters of the si nuses: cetacean air sacs. Anat Rec 291, 1389-1396. Reinhard KR, Miller ME, Evans HE (1962) The craniovertebral veins and sinuses of the dog. Am J Anat 111, 67-76. Ridgway SH (1999) An illustration of Norris aqcoustic window. Mar Mammal Sci 15, 926-930. Ridgway SH (2002) Asymmetry and symmetry in brain waves from dolphin left and right hemispheres: some observation afte r anesthesia, during quiescent hanging behavior, and during visual obstruction. Brain Behav Evol 60, 265-274. Ridgway SH, Harrison RJ (1986) Diving dolphins. In: Research on Dolphins (eds. Bryden MM & Harrison RJ). Oxford: Oxfo rd University Press. p 33-58. Ridgway S, Houser D, Finneran J, Card er D, Keogh M, van Bonn W, Smith C, Scadeng M, Dubowitz D, Mattrey R, Hoh C (2006) Functional imaging of dolphin brain metabolism and blood flow. J Exp Biol 209, 2902-2910. Ridgway SH, Howard R (1979) Dolphin lung collapse and intramuscular circulation during free diving: evidence from nitrogen washout. Science 206, 1182-1183. Ridgway S, Keogh M, Carder D, Finneran J, Kamolnic k T, Todd M, Goldblatt A (2009) Dolphins maintain cognitive per formance during 72 to 120 hours of continuous auditory vigilance. J Exp Biol 212, 1519-1527. Ridgway SH, McCormick JG, Wever EG (1974) A surgical approach to the dolphins ear. J Exp Zool 188, 265-276. Ridgway SH, Patton GS (1971) Dolphin thyroid: Some anatomical and physiological findings. Z vergl Physiologie 71, 129-141. Risau W (1995) Differentation of endothelium. FASEB J 9, 926-933.
348 Risau W, Flamme I (1995) Vasculogenesis. Annu Rev Cell Dev Bi 11, 73-91. Roberts VC, Sabri S, Pietroni MC, Gurewich V, Cotton LT (1971) Passive flexion and femoral vein flow: a study us ing a motorized foot mover. Br Med J 3, 78-81. Rommel SA, Costidis AM, Fernandez A, et al. (2006) Elements of beaked whale anatomy and diving physiology and some hypothetical causes of sonar-related stranding. J Cetacean Res Manage 7, 189-209. Rommel SA, Caplan H (2003) Vascular adaptations for heat conservation in the tail of Florida manatees ( Trichechus manatus latirostris ). J Anat 202, 343-353. Rommel SA, Early GA, Matassa KA, Pabst DA, McLellan WA (1995) Venous structures associated with thermoregulation of phocid seal reproductive organs. Anat Rec 243, 390-402. Rommel SA, Pabst DA, McLellan WA (1993) Functional morphology of the vascular plexuses associated with the cetacean uterus. Anat Rec 237, 538-546. Rommel SA, Pabst DA, McLellan WA (1998) Reproductive thermoregulation in marine mammals. Am Sci 86, 440-448. Rommel SA, Pabst DA, McLellan WA (2001) Functional morphology of venous structures associated with the male and female reproductive systems in Florida manatees ( Trichechus manatus latirostris ). Anat Rec 264, 339-347. Rommel SA, Pabst DA, McLellan WA et al (1992) Anatomical evidence for a countercurrent heat exchanger associated with dolphin testes. Anat Rec 232, 150156. Rommel SA, Pabst DA, McLellan WA, et al. (1994) Temperature regulation of the testes of the bottlenosed dolphin (Tursiops truncatus) J Comp Physiol B 164, 130134. Ronald K, McCarter R, Selley LJ (1977) Venous circulation in the harp seal ( Pagophilus groenlandicus ). In: Functional Anatomy of Marine Mammals Vol. 3. (ed. Harrison RJ), pp. 235-270. NewYork: Academic Press. Rosenblum B, Oldfield EH, Doppman JL, Di Chiro G (1987) Spinal arteriovenous malformations: a comparison of dural ar teriovenous fistulas and intradural AVMs in 81 patients. J Neurosurg 67, 795-802. Roth L, Koncina E, Satkauskas S, Cremel G, Aunis D Bagnard D (2009) The many faces of semaphorins: from development to pathology. Cell Mol Life Sci 66, 649666. Sabin FR (1926) Origin and development of the primitive vessels of the chick and of the pig. Contrib Embryol 18, 61-131.
349 Scano P, Maxia C, Maggiani F, Crnjar R, Lai A, Sirigu P (2005) A histological and NMR study of the melon of the striped dolphin ( Stenella coeruleoalba ). Chem Phys Lipids 134, 21. Shalaby F, Rossant J, Yamaguchi TP Ge rtsensstein M, Wu X-F, Breitman ML, Schuh AC (1995) Failure of blood-island and vasculogenesis in Flk-1 deficient mice. Nature 376, 62-66. Schaller B (2004) Trigeminocardiac reflex -A clinincal phenomenon or a new physiological entity? J Neurol 251, 658-665. Schaller O ed. (2007) Illustrated Veterinary A natomical Nomenclature (2nd Edition). Stuttgart: Ferdinand Enke. Scholander PF (1940) Experimental investigations on the respiratory function in diving mammals and birds. Hvalrd Skr 22, 1-131. Scholander PF (1958) Counter current exc hange; a principle in biology. Hvalrd Skr 44, 1-24. Scholander PF, Hock R, Walters V, Johnson F, Irving L (1950) Heat regulation in some Arctic and tropical mammals and birds. Biol Bull 99, 237-258. Scholander PF, Irving L (1941) Experimental investigations on the respiration and diving of the Florida manatee. J Cell Compar Physl 17, 169-191. Scholander PF, Irving L, Grinnell SW (1942) On the temperat ure and metabolism of the seal during diving. J Cell Compar Physl 19, 67-78. Scholander PF, Krog J (1957) Countercurrent heat ex change and vascular bundles in Sloths. J Appl Physiol 10, 405-411. Scholander PF, Schevill WE (1955) Counter-current va scular heat exchange in the fins of whales. J Applied Physiol 8, 279-282. Schreiber SJ, Lrtzing F, Gtze R, Do epp F, Klingebiel R, Valdueza JM. (2003) Extrajugular pathways of human cerebr al venous blood drainage assessed by duplex ultrasound. J Appl Physiol 94, 1802-1805. Schulte HVW (1917) The skull of Kogia breviceps Blainville. B Am Mus Nat Hist 37, 361-404. Schulte HVW, Tilney F (1909) A note on the organization of the venous return with especial reference to the iliac veins. Anat Rec 3, 555-577.
350 Schummer A, Wilkens H, Vollmerhaus B, Habermehl K (1981) Veins (venae). In: The Circulatory System, the Skin, and the Cutaneous Organs of the Domestic Mammals (Translated by Siller WG, Wight PAL). In: The Viscera of the Domestic Animals (Vol. 3). (eds Nickel R, Schummer A, Seiferle E), pp. 184-268. New York: Verlag Paul Parey. Shadwick RE, Gosline JM (1994) Arterial mechanics in the fin whale suggest a unique hemodynamic design. Am J Physiol 267, R805-R818. Schackelford RT, Hegedus SA (1966) Factors affectin g cerebral blood flow experimental review: sympathectomy, hypothermia, CO2 inhalation and pavarine. Ann Surg 163, 771-777. Shalaby F, Rossant J, Yamaguchi TP et al. (1995) Failure of blood-island and vasculogenesis in Flk-1-deficient mice. Nature 376, 62-66. Shoshani J (1986) Mammalian phylogeny: comparison of morphological and molecular results. Mol Biol Evol 3, 222-242. Shweiki D, Itin A, Soffer D, Keshet E (1992) Vascular endothelial growth factor induced by hypoxia may mediat e hypoxia-initiated angiogenesis. Nature 359, 843845. Simon E (1974) Temperature regulatio n; the spinal cord as a site of extrahypothalamic thermoregulatory functions. Rev Physiol Biochem Pharmacol 71, 1-76. Simon E (1981) Effects of CNS temperat ure on generation and transmission of temperature signals in homeotherms. Pfl Arch 392, 79-88. Simon E, Iriki M (1971) Sensory transmission of sp inal heat and cold sensitivity in ascending spinal neurons. Pflg Arch 328, 103-120. Simmons MP, Lopez-Jurado LF (1991) Whale and the military. Nature 351, 448. Sinclair JG (1967) Cerebral vascular system and ocular nerves of dolphin ( Stenella ) embryos. Tex Rep Biol Med 25, 559-571. Slijper EJ (1936) Die Cetacean: Vergleichend-Anatomisch und Systematisch Amsterdam: Asher and Company. [1972 reprint; in German]. Smodlaka H, Henry RW, Reed RB (2009) Macroscopic anatomy of the great vessels and structures associated with the heart of the ringed seal ( Pusa hispida ). Anat Histol Embryol 38, 161-168. Sommer LS, McFarland WL, Gallia no RE, Nagel EL, Morgane PJ (1968) Hemodynamic and coronary angiographic studies in the bottlenose dolphin ( Tursiops truncatus). Am J Physiol 215, 1498-1505.
351 Sobolewky EI (1986) Morphological investigations of the arterial system in pinnipeds. Zoologichesky Zhurnal 65, 1400-1410. Springer MS, Amrine HM, Burk A, Stanhope MJ (1999) Additional support for Afrotheria and Paenungulata, the performa nce of mitochondrial versus nuclear genes, and the impact of data partitions with heterogeneous base composition. Syst Biol 48, 65-75. St. Pierre H (1974) The topographical splanchnol ogy and the superficial vascular system of the harp seal Pagophylus groenlandicus (Erxleben 1777). In: Functional Anatomy of Marine Mammals, Vol. 2. (ed. Harrison RJ ), pp. 161-195. New York: Academic Press. Stannius FH (1845) Beitrge zur Kenntniss der Amerikanischen Manatis. pp. 37. Rostock: Adlers Eben. Tatelman M (1960) Pathways of cerebral collateral circulation. Radiology 75, 349-361. Tyack PL, Johnson M, Aguilar Soto N, Sturlese A, Madsen PT (2006) Extreme diving of beaked whales. J Exp Biol 209, 4238-4253. Ueda H, Koide T, Ito I, et al. (1964) Postural changes in blood distribution and its relation to the change in cardiac output. Jap Heart J 5, 115-126. Valdueza JM, von Mnste r T, Hoffmann O, et al. (2000) Postural dependency of the cerebral venous outflow. Lancet 355, 2000-2001. Viale G (2006) The rete mirabile of the cranial base: a millenary legend. Neurosurgery 58, 1198-1207. Viamonte M, Morgane PJ, Galliano RE, Angel EL, McFarland WL (1968) Angiography in the living dolphin and observations of bl ood supply to the brain. Am J Physiol 214, 1225-1249. Vogl AW, Fisher HD (1981) The internal carotid arte ry does not directly supply the brain in the Monodontidae (Order Cetacea). J Morphol 170, 207-214. Vogl AW, Fisher HD (1982) Arterial retia related to supply of the central nervous system in two small t oothed whalesnarwhal ( Monodon monoceros ) and beluga ( Delphinapterus leucas ). J Morphol 174, 41-56. Vogl AW, Todd ME, Fisher HD (1981) An ultrastructural and fluorescence histochemical investigation of the innervation of retial arteries in Monodon monceros J Morphol 168, 109-120. Walmsley R (1938) Some observations on the vascular system of a female fetal finback. Contrib Embryol 164, 109-78.
352 Watson AG, Dowd AD (1972) The azygos veins of the Weddell seal, Leptonychotes weddelli (Lesson). Canad J Zool 50, 1583-1587. Webster MED, Johnson KG (1984) Distribution of tem perature in the skin of sheep exposed to moderate environments. Nature 4915, 208-209. Weinstein BM, Lawson ND (2002) Arteries, vein s, notch, and vegf. In The Cardiovascular System, 67, 155-162. New York: Cold Spring Harbor Laboratory Press. Weyrauch HB, de Garis CF 1937. Normal and interrupted vascular patterns in the intestinal mesentery of the rat. An ex perimental study on collateral circulation. Am J Anat 61, 343-372. Whisnant JP, Millikan CH, Wakim KG, Sayre GP (1956) Collateral circulation to the brain of the dog following bilateral ligati on of the carotid and vertebral arteries. Am J Physiol 186, 275-277. Williams TM, Haun J, Davis RW, Fuiman LA, Kohin S (2001) A killer apetite: metabolic consequences of carnivory in marine mammals. Comp Biochem Physiol A 129, 785-796. Wilson HS (1879) The rete mirabile of the narwhal. J Anat Physiol 14, 377-398. Wislocki GB (1939) The unusual mode of developm ent of the blood vessels of the opossums brain. Anat Rec 74, 409-427. Wislocki GB, Straus WL (1932) On the blood vascular bundles in the limbs of certain edentates and lemurs. Bull Mus Comp Zool 74, 1-15. Wollard HH (1936) The development of the principal arterial stems in the fore-limb of the Pig. Contrib Embryol 14, 141-154. Wnnenberg W (1973) Alteration of EEG activity of hypothalamus by thermal stimulation of spinal cord. Pflg Arch 344, 75-82. Wnnenberg W (1983) Thermosensitivity of the preoptic region and the spinal cord in the golden hamster. J Therm Biol 8, 381-384. Zapol WM, Liggins GC, Schneider RC, et al. (1979) Regional blood flow during simulated diving in the conscious Weddel seal. J Appl Physiol 47, 968-973. Zhong TP, Childs S, Leu JP, Fishman MC (2001) Gridlock signaling pathway fashions the first embryonic artery. Nature 414, 216-220. Zimmer WMX, Tyack PL (2007) Repetitive shallow dive s pose decompression risk in deep-diving beaked whales. Mar Mammal Sci 23, 888-925.
353 BIOGRAPHICAL SKETCH Alex grew up in the northern suburbs of Athens, Greece. From a young age he had a passion for the ocean, whic h he pursued early on through diving. After high school, he came to the U.S. to attend Ecke rd College, a small college with a strong marine biology program. Without a com pass, he haphazardly navigated the labyrinth that was the undergraduate pr ogram, until he happened upon Sent iel Butch Rommel. Butch is a wonderful human being with an in fectious curiosity for life and an unparalleled enthusiasm and dedication to teaching. He quickly took Alex under his wing and showed him the joys of anatomy. It did not take long to get hooked. After graduating from college with a bac helors degree in marine science, Alex got a job as research staff at the Florida Fish and Wildlife Conservation Commissions Marine Mammal Pathobiology Lab (MMPL). The MMPL gave him a rare opportunity to get regular exposure to marine mammal specimens and build ease with anatomy and handling of post-mortem tissues that few peopl e have an opportunity to develop. It also gave him a unique set of skills revolving around forensic wildlife examinations and interpretations of watercraft wounds in marine mammals. Under the mentorship of Butch, Alex developed those skills and gradually took on greater responsibility teachi ng necropsy and anatomy classes for stranding biologists and veterinary students. He eventually had the pleasure of teaching countless such workshops around the country as well as other wonderful places such as Cuba, Puerto Rico and Belize. As he was exposed to more and more cetacean species, he became interested in large whales and began regularly attending lar ge whale stranding events. A few years later the National Marine Fisher ies Service named him an Atlantic Large Whale Necropsy Team Leader, a person res ponsible for leading the cause of death
354 examination of a deceased lar ge whale. During the later years at the MMPL, Alexs collaborations with the Univer sity of Floridas College of Veterinary Medicine Aquatic Animal Health Program result ed in many good relationships. Owing to his trepidation from his lackluster earlier academic years, he resisted th e notion of pursuing a PhD for quite some time. He finally succumbed to the temptation and entered the Aquatic Animal Health Program in 2007, while st ill being employed at the MMPL. Alex eventually moved to the University of Florida to pursue his degree more rigorously. He had the fortune of getting a most excellent advisor (Dr. R oger Reep) who gently guided him through the academic gauntlet. Throughout the last few year s at the University of Florida, Alex had the wonderful luxury of pursuing numerous research interests while still being actively involved in the national marine mammal stranding network. He had the benefit of making great friendships and meeting excellent new colleagues. Alex treasured and benefited tremendous ly from this experience, and regularly reminds himself how lucky he is to be able to pursue answers to questions that evolve from his innate curiosity and awe of these most intr iguing animals. Al ex earned his PhD in Veterinary Medical Sciences at the University of Florida in the summer of 2012. During his studies at the University of Florida, Alex obtained a federal grant to pursue a postodoctoral position at the University of North Carolina Wilmington. He will be spending his next two years pursuing simi lar vascular research in deep-diving cetaceans such as beaked whales and pygmy sperm whales.