<%BANNER%>

Skeletal Biomechanics of the Florida Manatee (Trichechus manatus latirostris)


PAGE 1

SKELETAL BIOMECH ANICS OF THE FL ORIDA MANATEE ( Trichechus manatus latirostris ) By KARI BETH CLIFTON A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Kari Beth Clifton

PAGE 3

to all thosepeople and animal alike who have loved, supported, encouraged, instructed, amazed, inspired, and empowered me

PAGE 4

ACKNOWLEDGMENTS This project was an enormous undertakingone that I could not have completed alone. First I would like to acknowledge my supervisory committee chair (Roger Reep) and co-chair (Jack Mecholsky). I thank Roger for his unqualified generosity, encouragement, and guidance; and for feeding the spirit with food and wine. I benefited greatly from being allowed to wander off the path in pursuit of new ideas. I thank Jack for his patience in teaching a biologist to think like an engineer, and for the tailgates. My supervisory committee members (MaryBeth Horodyski, Donna Wheeler, and Tom Wronski) provided guidance, instruction, and support for which I am grateful. Many individuals contributed to the success of this project. While I was working at the Florida Department of Environmental Protection, Dr. Sentiel Rommel suggested I take a look at manatee bone. He and Dr. Tom Koob at the Shriners Hospital for Children introduced me to the field of bone mechanics. The Save the Manatee Club funded a pilot study that led to the inception of this project. I am indebted to Maggie Stoll for her invaluable technical expertise and administrative support. To my fellow graduate students in the fracture mechanics group, I am grateful for the weekly discussions of ceramics and other things. I learned a great deal about fractography and bone mechanics from Jiahau Yan. From the College of Engineering, Chuck Broward provided access to the air cannon, as well as his expert engineering skills. Billy Schultz performed the strain gauge testing. Bryan Conrad from the Department of Orthopaedics taught me to operate the iv

PAGE 5

biomechanical testing equipment, and provided invaluable technical support. Joan Yonchek lent her amazing expertise in hard and soft tissue histology. I thank Allyson Barrett and Ben Lee from the Department of Dental Biomaterials for their assistance with technical issues and data collection. This project made great strides forward with the help of veterinary students Julia Golden and Beth Carson; and undergraduate volunteers Johnattan, Rebecca, Jessica, Stephanie, Jaime, Travis, and Chris. Tissue samples were provided by the FWCs Marine Mammal Pathobiology Lab in St. Petersburg, and by Sea World of Florida. I extend my thanks and appreciation to Dr. Sentiel Rommel, Mr. Tom Pitchford, Mr. Ken Arrison and the staff of the MMPL, and to Dr. Beth Chittick of Sea World. I wish to thank the college of Veterinary Medicine, the Marine Animal Health program, and in particular Dr. Charles Courtney for his interest and enthusiasm for my project. This project was funded by a College of Veterinary Medicine Consolidated Faculty Research award and a Marine Conservation Trust Fund award. Additional funding was provided by grants from the Sigma Xi Scientific Research Society, a College of Veterinary Medicine Graduate Student Seed Grant, and a Merck-Merial Veterinary Scholars Research Grant. The Universitys Graduate Student Council and Graduate School provided travel awards so that I could attend conferences. Last, but not at all least, I would like to express my gratitude and appreciation for the love and support of my family throughout my many years of schooling. I am indebted to my nephews David, Adam, and Brain, for never, ever letting me get a lick of work done while at their house. There are, after all, more important things in life. v

PAGE 6

TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................iv LIST OF TABLES ...........................................................................................................viii LIST OF FIGURES ...........................................................................................................ix ABSTRACT .......................................................................................................................xi CHAPTER 1 INTRODUCTION........................................................................................................1 Skeletogenesis...............................................................................................................5 Bone Shape and Macrostructure...................................................................................7 Microscopic Structure of Bone.....................................................................................8 Mechanical Properties of Bone Related to Microscopic Structure.............................11 Variations in the Microscopic Structure of Vertebrate Bone.....................................14 Manatees in Perspective.............................................................................................24 2 MATERIAL PROPERTIES OF MANATEE RIB BONE.........................................29 Introduction.................................................................................................................29 Materials and Methods...............................................................................................32 Bone Sample Collection......................................................................................32 Specimen Preparation..........................................................................................32 Material Properties Testing.................................................................................33 Bone Density and Mineral Content Determination.............................................35 Statistical Analyses..............................................................................................36 Results.........................................................................................................................36 Within-Animal Comparisons...............................................................................37 Material Properties Tests.....................................................................................37 Bone Density and Mineral Content.....................................................................38 Discussion...................................................................................................................39 The Material Properties of Manatee Bone...........................................................39 The Effect of Mineral Content on the Material Properties of Bone....................45 Other Factors Influencing Bone Material Properties...........................................46 Summary..............................................................................................................50 vi

PAGE 7

3 STRUCTURAL PROPERTIES OF MANATEE RIBS TESTED IN IMPACT........59 Introduction.................................................................................................................59 Materials and Methods...............................................................................................60 Bone Sample Collection......................................................................................60 Whole Rib Impact Testing...................................................................................61 Strain Gauge Testing...........................................................................................62 Fractal Dimensional Increment...........................................................................63 Kinetic Energy Calculations................................................................................64 Statistical Analyses..............................................................................................65 Results.........................................................................................................................65 Whole Rib Impact Tests......................................................................................65 Strain Gauge Tests...............................................................................................66 Kinetic Energy Calculations................................................................................67 Discussion...................................................................................................................68 The Fractal Dimensional Increment....................................................................68 Fractal Dimensional Increment and Fracture Toughness....................................70 R-curve Behavior and Viscoelasticity in Bone...................................................71 Impact Energy Absorption of Soft Tissues.........................................................73 Summary..............................................................................................................74 4 CONCLUSIONS AND FUTURE RESEARCH........................................................85 Summary and Conclusions.........................................................................................85 Further Research.........................................................................................................87 Additional Avenues of Research................................................................................90 APPENDIX A DATA TABLES.........................................................................................................92 B DERIVATION OF WORK OF FRACTURE..........................................................101 LIST OF REFERENCES.................................................................................................103 BIOGRAPHICAL SKETCH...........................................................................................115 vii

PAGE 8

LIST OF TABLES Table page 2-1 Mean material properties for 26 manatees...............................................................51 3-1 Mechanical and fractal data for 20 manatee ribs fractured in impact......................76 3-2 Mean mechanical and fractal data by sex and age class...........................................76 3-3 Mechanical data for strain gauged ribs....................................................................77 3-4 Boat data for kinetic energy calculations.................................................................77 3-5 Kinetic energy calculations for vessel configurations and impact projectile...........77 A-1 Mortality information for animals used in this study...............................................92 A-2 Means for four mechanical properties by body region............................................93 A-3 Critical crack measurements for whole manatee ribs fractured in impact.............100 viii

PAGE 9

LIST OF FIGURES Figure page 2-1 Sample sites for biomechanical analyses.................................................................52 2-2 Flexural strength vs. total length..............................................................................52 2-3 Fracture toughness vs. total length...........................................................................53 2-4 Work of fracture vs. total length..............................................................................53 2-5 Elastic modulus vs. total length................................................................................54 2-6 Flexural strength vs. modulus..................................................................................54 2-7 Load-displacement curves for small sample specimens tested in 3 point flexure....55 2-8 Density vs. total body length....................................................................................56 2-9 Percent mineral content vs. total body length..........................................................56 2-10 Modulus vs. density..................................................................................................57 2-11 Modulus vs. percent mineral content.......................................................................57 2-12 Small sample crack size vs. total body length..........................................................58 3-1 Setup for impact tests of whole manatee ribs...........................................................78 3-2 Fracture surface of a manatee rib fractured by collision with a boat.......................79 3-3 Procedure for making epoxy replicas of manatee rib fracture surfaces...................80 3-4 Example of a fracture surface contour used for calculation of fractal dimensional increment (D*).........................................................................................................81 3-5 Fractal dimensional increment (D*) vs. total body length.......................................82 3-7 Kinetic energy as a function of boat speed..............................................................83 ix

PAGE 10

3-8 Fracture toughness (K C ) vs. fractal dimensional increment (D*)............................84 3-9 Whole rib crack size vs. total body length...............................................................84 x

PAGE 11

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy SKELETAL BIOMECHANICS OF THE FLORIDA MANATEE (Trichechus manatus latirostris) By Kari Beth Clifton December 2005 Chair: Roger L. Reep Cochair: John J. Mecholsky, Jr. Major Department: Veterinary Medicine Watercraft-related mortality of the Florida manatee (Trichechus manatus latirostris) accounts for 25% of all deaths from 1974-2004, comprising 85% of anthropogenic-related deaths. Of those, the proportion of deaths due to impact accounts for 67% of all watercraft fatalities. Reducing this preventable source of mortality is a high priority in state and federal manatee recovery efforts, which focus primarily on regulating boating activities. However, with more boats using the waterways, the potential threat posed by watercraft continues to increase. To establish safe boat speeds for manatee protection, an estimate of the forces required to fracture their bone is needed. This study is the first to quantify the biomechanical effects of boat strikes on manatees. Our goal was to estimate material and structural properties of manatee rib bone. To measure bone material properties, machined specimens were tested in 3-point flexure. Strength, toughness, modulus, and work of fracture were determined for three size classes, by sex. Mean flexural strengths ranged from 62 MPa, fracture toughness xi

PAGE 12

from 1.4.9 MPam 1/2 elastic moduli from 4 GPa, and work of fracture 3 MJm -3 indicating that manatee bone is less strong and tough than other mammalian bone, including human. There were no differences in mechanical variables between the sexes. Density and mineral content increased with body size, suggesting that material properties are likely correlated bone tissue quality, as well as to microstructure. For whole bone structural tests, ribs were impacted to estimate failure stress and kinetic energy needed to cause fracture. Fractographic analysis was used to calculate the stress at failure. Strengths ranged from 106 MPa. Strain gauges were used to validate these calculations. Fractal geometry was used to calculate the fractal dimensional increment (D*). Values for manatee bone (.08.22) were at the lower end of the range for ceramics, suggesting that a low amount of energy is needed to cause fracture. There was no relationship between D* and age. The ability to resist bone fracture does not appear to increase as animals grow. Impact energy calculations suggest that boats may inflict fatal injuries to manatees at speeds of 13 mph. xii

PAGE 13

CHAPTER 1 INTRODUCTION The West Indian manatee (Trichechus manatus) is one of four extant species belonging to the order Sirenia. It occurs throughout the Caribbean, from the southeastern United States to northeastern Brazil. Manatees inhabit shallow, coastal estuarine or riverine waters, feeding primarily on seagrasses and other aquatic vegetation. Their reliance on shallow-water habitats has resulted in two distinct, geographically isolated subspecies: the Florida manatee (Trichechus manatus latirostris) is found primarily in Florida, whereas the Antillean manatee (Trichechus manatus manatus) is found in disjunct populations in eastern Central America, the Greater Antilles, and northeastern coastal South America (Marmontel, 1993). The Florida manatee is listed as endangered by the U.S. Department of the Interior, and is considered one of the most endangered marine mammals found in U.S. waters (U.S. Fish and Wildlife Service, 1996). Manatees are protected by state and federal legislation. Florida provided the first protection in 1893, when a law was passed prohibiting the killing of manatees (Reynolds and Odell, 1991). The Florida Manatee Sanctuary Act of 1978 established the entire state of Florida as a manatee refuge, and mandated the creation of boat speed regulatory zones (Reynolds and Odell, 1991). The Marine Mammal Protection Act of 1972 established federal guidelines for maintaining marine ecosystems, including marine mammals (Reynolds and Odell, 1991). The Endangered Species Act of 1973 mandates the development and implementation of recovery plans for wildlife identified as threatened or endangered with extinction. The 1

PAGE 14

2 main objective of the Florida manatee recovery plan, first released by the U.S. Fish and Wildlife Service in 1989, is to downlist, and to ultimately delist the species when the population has reached maximum sustainable levels, and threats to the habitat and mortality are controlled (Eberhardt and OShea, 1995). The plan also identifies the actions needed to meet those criteria. However, the rapid growth of the human population in Florida, and the continued decline in manatee habitat and increasing numbers of deaths makes it unlikely that the species will be downlisted. Sources of manatee mortality have been documented in Florida since 1974 (OShea et al., 1985; Ackerman et al., 1995; Wright et al., 1995). Watercraft-related mortality, caused by propeller wounds or by impact with a boats hull, accounts for 25% of all deaths from 1974 through 2004, and comprises 85% of anthropogenic-related deaths (Florida Fish and Wildlife Conservation Commission, 2005). Wright et al. (1995) examined necropsy records for 406 manatees killed by boats from 1979 through 1991. Deaths were broken down into three sub-categories: deaths from propeller wounds; from impact-related injuries; or from a combination of propeller and impact injuries. The proportion of impact-related mortality has increased over time, and now accounts for more than half (66%) of all watercraft-related deaths (Florida Fish and Wildlife Conservation Commission, 2005). Regardless of cause of death, most manatees bore scars of healed propeller wounds. Fatal propeller wounds were usually more severe than non-fatal wounds, and animals killed by propeller cuts rarely suffered skeletal damage. Conversely, approximately 66% of the carcasses attributed to impact injuries had fractured and/or luxated ribs. In addition, healed fractures were rarely observed in any

PAGE 15

3 carcasses. These data suggest that while many propeller wounds are sub-lethal, impact injuries sufficient to inflict skeletal damage are usually fatal. The number of watercraft-related deaths has increased at a rate of 7% annually from 1992 to 2004 (Florida Fish and Wildlife Conservation Commission, 2005). Reducing watercraft-related mortality is identified as a high priority in the manatee recovery plan. To date, recovery efforts have focused primarily on regulation of boating activities by establishing speed zones in areas where manatees and boats coexist. The Florida Fish and Wildlife Conservation Commission (FWC) is responsible for establishing boat speed regulatory zones in any body of water in Florida where manatee protection is warranted (Florida DNR, 1989). The Florida Manatee Sanctuary Act of 1978 and subsequent amendments give the FWC the authority to regulate boat speeds and to limit boating activities in areas essential to manatees. The 1989 Recommendations to Improve Boating Safety and Manatee Protection for Florida Waterways identified 13 key counties and gave them the option to develop site specific speed zones, adopt 300 foot buffer slow speed zones within all inshore/coastal waters, or adopt 30 mph in channels and 20 mph outside the channels. Predominantly, the counties chose to develop their own zones, and most were adopted and approved by the state of Florida by the early 1990s. To designate boat speed zones, the FWC primarily uses manatee aerial survey data, mortality records, satellite telemetry, and boat use data. Based on these datasets, speed zones are delimited by visual estimation of the areas used by manatees. Maximum speed limits are typically established by examining the number and pattern of manatees that utilize the area and the nature and extent of watercraft usage (e.g., a nearby boat

PAGE 16

4 channel indicates frequent, high speed use) if known. The proposed speed zones are then reviewed and modified in a series of public and legal hearings. The speed zones that are ultimately implemented by law are a compromise of the rules originally proposed by the FWC and the affected parties (e.g., commercial boating industries, recreational boaters). Thus, creation of boat speed zones is a highly subjective process, not based on any information pertaining to the biological effects of boat strikes on manatees. Although boat speed regulatory zones have been in existence since the inception of the Florida Manatee Sanctuary Act, the increasing human population along the Florida coastline has resulted in more boats utilizing the waterways. From 1996 through 2004, the number of watercraft on Florida waters has increased by 3% per year (U.S. Coast Guard, 2005). Florida is the fastest-growing recreational boating state in the U.S., ranking number 1 in the number of registered boats. In 2004, there were over 928,000 registered boats in the state (Florida Department of Motor Vehicles, unpublished data). Furthermore, new boat and engine designs allow boaters to utilize shallow areas once inaccessible (Wright et al., 1995). Thus, the potential threat to manatees posed by watercraft continues to increase, and regulatory efforts to date have not been successful in reducing the number of watercraft-related deaths. It is of considerable interest, to conservationists and boaters alike, to identify ways to reduce this major source of manatee mortality. In an effort to improve the efficacy of boat speed zones, the second revision of the U.S. Fish and Wildlife Services manatee recovery plan calls for research on mechanical characteristics of manatee bone (U.S. Fish and Wildlife Service, 2000). As discussed above, two-thirds of manatee deaths attributed to boats are a result of impact with some

PAGE 17

5 part of the vessel. Ribs broken by the impact damage the surrounding soft tissues severely enough to result in death. Information on the amount of force needed to break manatee ribs would be valuable in refining boat speed zones to reduce lethal impacts. Currently, there is no information on the amount of energy needed to break manatee ribs in impact. Manatee bone is unique in the animal kingdom. The ribs are solid cortical bone. They are pachyostotic, osteosclerotic, and appear to have a unique microstructure. It is these features that likely make manatee bone susceptible to fracture when struck by boats. The bony tetrapod endoskeleton first appeared some 400 million years ago (Carter and Beaupr, 2001). Throughout evolution, the basic form was constrained by genetics. However, nature has allowed a great deal of latitude in the details. Genetic mutations and also epigenetic and environmental factors have altered the mechanical environment of the skeleton (Carter and Beaupr, 2001). The result is immense variation in skeletal form and function, at both the gross structural level and the microscopic level. The remainder of this chapter presents a brief overview of the tremendous variety in bone tissues found among the vertebrates. It provides the framework for understanding where the manatee falls in the spectrum, and why we find it worthy of study. Skeletogenesis The bones of the skeleton are formed one of two ways. Intramembranous bone forms directly within the mesenchyme (Hall, 2005). There are a few exceptions, but generally the craniofacial bones and cartilages, as well as dentine and alveolar bone, are all formed intramembranously, and are derived from neural crest. In contrast, endochondral bone forms by ossification of a pre-existing cartilaginous model. The bones

PAGE 18

6 of the limbs, vertebral column, and the ribs are endochondral in origin. Endochondral bones are derived from mesoderm. Dermal bones are derived from the dermis of the skin, and are present in some members of every vertebrate class except birds (Kent and Miller, 1997). They are membrane bones, although they are derived from ectoderm or endoderm, unlike intramembranous bone which originates from mesoderm (Hall, 2005). Meckels cartilage and some bones of the chondrocranium are dermal bones. The pectoral girdle in most vertebrates is comprised of both dermal and endochondral bone (Romer and Parsons, 1977). Other dermal bones in modern vertebrates take the form of scales or plates. Dermal bones in fishes are comprised of various combinations of lamellar bone (cellular or acellular), dentin, and an enamel-like substance; all of which may be covered with epidermis (Kent and Miller, 1997). A few fishes possess bony plates. Some tetrapods have minute bony scales called osteoderms (Kent and Miller, 1997). Caecilians and some tropical toads have them. Crocodilians have osteoderms below cornified epidermis (Romer and Parsons, 1977). Some lizards have osteoderms that fuse with the skull shortly after birth, and lizards and crocodilians both possess abdominal ribs that are remnants of ventral bony scales (Romer and Parsons, 1977). The carapace and plastron of turtles are comprised of bony plates with immobile sutures (Kent and Miller, 1997). Armadillos and some whales are the only mammals that normally form dermal bone (Romer and Parsons, 1977; Kent and Miller, 1997). The armor of armadillos consists of polygonal bones with immobile sutures that are covered with epidermal scales. Heterotopic bones are not part of the endoskeleton proper. They typically form in fibrous connective tissues and are endochondral, intramembranous, or dermal in origin

PAGE 19

7 (Romer and Parsons, 1977; Kent and Miller, 1997). In mammals, sesamoids are heterotopic bones that form in tendons at points of friction, such as where they pass over joints. Sesamoid bones are also common in reptiles (Hall, 1984). Tendons in the legs of turkeys and pheasants often ossify (Currey, 1984; Kent and Miller, 1997). Kangaroos have ossified tendons on the mid-ventral surface of the tail (Murray, 1936). The baculum is a heterotopic bone found in insectivores, bats, rodents, carnivores, and in nonhuman primates (Romer and Parsons, 1977). Crocodilians have a bony plate in the eyelids, called the os palpebrae. The os cordis forms in the heart of deer and bovids. Rostral bones are formed in the snout of some mammals, such as the pig (Romer and Parsons, 1977; Kent and Miller, 1997). The os falciforme functions as a supernumerary digit in the digging hand of moles (Romer and Parsons, 1977). In birds, heterotopic bones are found in the gizzard of some doves and in the syrinx of others. Some lizards possess a cloacal bone in the ventral wall of the cloaca. One species of bat has a bone in its tongue, and bone is found in the diaphragm of camels (Kent and Miller, 1997). Some heterotopic elements are cartilaginous. Cartilage and bone have been reported in the hearts and aortae of snakes, turtles, crocodiles, alligators, birds, rats, and hamsters (Hall, 1984; Lopez et al., 2003). Hyaline cartilage is normally found in the aorta of rats, and fibrocartilage is found in the hearts of dogs and humans (Hall, 1984). Bone Shape and Macrostructure Bones are classified into four groups, based on their shape (Marieb, 1995). Long bones are roughly cylindrical in shape. They are generally longer than they are wide. The ends of the bones (epiphyses) are more expanded than the shaft (diaphysis). Bones of the limbs are long bones (such as the femur and humerus; as well as smaller bones like the metacarpals, metatarsals, and phalanges). Short bones such as carpals and tarsals are

PAGE 20

8 somewhat cuboidal in shape. Bones with complicated shapes (such as vertebrae and bones of the pelvic girdle) are classified as irregular. Flat bones, such as the frontal and parietal, are comprised of cancellous bone sandwiched between two layers of compact bone. At the gross structural level bone is classified into two types based on its density (Carter and Hayes, 1976; Hayes and Bouxsein, 1997). Compact, or cortical bone ranges in density from 1.6.0 g/cm 3 and has a porosity of 5%. Cortical bone forms a shell around the outside of bones, encasing the inner cancellous bone, and constitutes the bulk of the shaft of long bones (Martin and Burr, 1989). Cancellous, or trabecular bone is much more porous (30% porosity), with density from 0.07.0 g/cm 3 Cancellous bone is found in the interior of flat, short, and irregular bones and in the ends of long bones. It takes the form of needles or plates called trabeculae (Jee, 1988). The pores are generally filled with bone marrow. Microscopic Structure of Bone Bone, whether compact or cancellous, is comprised of two types of bone tissue: primary bone, of which there are two typeswoven and lamellar; and secondary bone (Jee, 1988). Woven bone is found in sites of initial bone formation, during early stages of development. It forms directly in mesenchyme (intramembranous bone), or by ossification of a cartilaginous model (endochondral bone). Woven bone forms rapidly, but as a result it is structurally inferior to secondary bone (Martin and Burr, 1989). Collagen fibers are randomly oriented within the matrix, and osteocytes are distributed irregularly. It is generally low in mineral content, although it has the potential to become more highly mineralized than lamellar bone (Martin and Burr, 1989). Woven bone can be formed de novo, with no previous bone or cartilage model (Martin and Burr, 1989). It

PAGE 21

9 forms the callus during fracture repair, and is formed in disease processes that involve the skeleton. It is also formed on the periosteal surfaces of bones that have been subjected to excessive strains (Martin et al., 1998). Primary lamellar bone, whether cortical or cancellous, must be laid down on a pre-existing surface of either cartilage or bone (Martin and Burr, 1989). Trabeculae are thickened by deposition of lamellae on woven bone. Compact bone is formed by continued deposition until the intertrabecular spaces are filled in (Jee, 1988). There are three major patterns (Jee, 1988) of lamellar arrangement in cortical bone: 1) Circumferential lamellae surround the internal and external surfaces of the bone (inner and outer circumferential lamellae, respectively), 2) Haversian systems, and 3) Interstitial lamellae, fragments of previous Haversian systems and circumferential lamellae that have been partially resorbed, fill in the gaps between secondary Haversian systems. Primary Haversian systems, or osteons, are formed by compaction of woven bone, or when blood vessels on the periosteal surface become incorporated into new bone as a result of appositional growth. Bone is deposited in concentric rings around the vessel, forming a vascular channel. Plexiform bone is another type of primary bone that contains both woven and primary lamellar bone (Currey, 1960; Enlow and Brown, 1956, 1958; Martin and Burr, 1989; Martin at el., 1998). It forms on either periosteal or endocortical surfaces. Subperiosteal or subendosteal bone forms buds perpendicular to the bone surface that grow for a short time, then join laterally, creating vascular spaces. The vascular spaces are then filled in with primary lamellar bone. The resulting tissue has the appearance of a

PAGE 22

10 brick wall. Plexiform bone is most commonly found on the periosteal surface of the diaphysis of long bones (Martin and Burr, 1989). Over time, bone must adjust to variable loading conditions and it must repair fatigue damage, a result of repetitive loading (Martin at al., 1998). Secondary lamellar bone is a result of bone remodeling, the replacement of existing bone with new bone. The secondary Haversian systems of remodeled bone are formed by tandem processes of bone removal and replacement (Jee, 1988). In cortical bone, groups of cells collectively referred to as a bone multicellular unit (BMU) first remove existing bone; then new bone is deposited in concentric lamellae, leaving a canal in the center. The lamellae are separated from the surrounding interstitial bone by the cement line, a layer of mineralized matrix devoid of collagen fibers (Jee, 1988; Martin and Burr, 1989). The resulting structure is a secondary Haversian system. So like primary lamellar bone, secondary lamellar bone is comprised of parallel sheets of lamellae. However, this bone is formed more slowly and is more organized than primary bone (Martin et al., 1998). Osteocytes are more regularly arranged. Within each lamella, the collagen fibers are all parallel, but the fibers in adjacent lamellae are oriented at 90 angles (other patterns of collagen fiber arrangement have also been documented). Primary and secondary Haversian systems can be distinguished from one another by their size and structure (Martin and Burr, 1989). Primary osteons have a smaller canal, fewer concentric lamellae, and no cement line. The trabeculae of cancellous bone are also lamellar in structure. In contrast to cortical bone, trabeculae are remodeled on the surface. The remodeled trabecular packets have a lamellar construction similar to secondary osteons (Jee, 1988).

PAGE 23

11 Mechanical Properties of Bone Related to Microscopic Structure Three mechanical properties are typically measured in bone: elastic modulus, strength, and toughness. The elastic modulus, or the Youngs modulus, is a measure of the stiffness of the material; it is a measure of the relationship between the external load applied, expressed in terms of stress; and the deformation response, in terms of strain. The strength of a material is the load capacity at fracture per unit area. For this calculation, knowledge of the type of loading and geometry of the structure is critical to proper comparison of strengths among different investigators. Toughness is a measure of the energy absorbed before fracture and released during fracture. With isotropic, monolithic materials, there is a direct relationship between the elastic modulus and toughness. Because of the complex, composite nature of bone, the relationships among elastic modulus, toughness, and strength may not be the same as with monolithic materials. Additionally, because bone is a hierarchical structure, these properties may vary as a function of scale of measurement. This section gives a brief overview of the mechanical properties of bone as related to its organization. The mechanical properties of bone are affected by its porosity and degree of mineralization (Martin and Burr, 1998). Porosity affects bone strength: the more porous the bone, the weaker it is. Hence cancellous bone is weaker than compact bone. Bone strength and stiffness increase with increased mineral content. Currey (1969) found that a small increase in mineralization results in a large change in modulus, a result of the organization of the mineral crystals. He showed that the optimal mineral content is about 66%. Above this, toughness decreases with increased mineralization. The organization of both cortical and cancellous bone influences its strength, stiffness, and toughness (Martin et al., 1998). Woven bone is weaker than lamellar bone,

PAGE 24

12 because of its randomly arranged collagen fibers, osteocytes, and mineral crystals (Martin and Burr, 1998). However, the random arrangement makes it behave as an isotropic material. This makes it an ideal tissue for structures that must be equally strong in all directions, such as rapidly growing bones that will soon be remodeled (Wainwright et al., 1982). The parallel collagen fibers of lamellar bone make it stronger than woven bone, but strength is affected by the orientation of the fibers in relation to the direction of the applied load (Martin and Burr, 1989). Generally, fibers parallel to the applied load are strongest in tension, and fibers perpendicular to the load are strongest in compression (Martin and Burr, 1989). Thus, anisotropic bone is stronger and stiffer in the direction of the predominant orientation of the collagen fibers. In most long bones, the fibers, lamellae, and osteons are parallel to the long axis. Hence they are stronger in that direction, but may be significantly weaker when loaded in other directions (Wainwright et al., 1982). In the transverse plane, plexiform bone is weaker when loaded radially (perpendicular to lamella) than tangentially (in the plane of the lamella). In the tangential plane plexiform bone fails near the interlamellar blood vessels. However, Haversian bone is equally strong in all orientations in the transverse plane, due to the concentric arrangement of the lamellae. Therefore, osteonal bone is transversely isotropic (Reilly and Burstein, 1974). Plexiform bone confers two mechanical advantages on structures (Martin and Burr, 1989). The rapid increase of the surface area for primary lamellar bone deposition greatly increases the strength of the bone in a short period of time. Second, adding this

PAGE 25

13 bone on the periosteal surface quickly increases the diameter of the structure; which increases its rigidity, thereby increasing its resistance to bending. Remodeled bone (i.e., that containing Haversian systems) is weaker than purely lamellar bone in compression, tension, shear, and bending (Martin and Burr, 1989). There are three possible reasons for this (Currey, 1984). First, Haversian bone is younger than the bone it replaced: therefore, it is less mineralized. Second, the creation of Haversian canals increases the porosity of the bone. Third, the cement lines surrounding Haversian systems do not contain collagen, so they are lower in strength than the lamellae. The fact that cement lines fracture easily imparts another important property to remodeled bonetoughness (the ability to resist failure). The tougher a material, the more energy required to break it (Gordon, 1978). The low-strength cement lines dissipate energy and stop cracks from propagating (Currey, 1984; Martin et al., 1998). The strength of cancellous bone is affected by the structure of the individual trabeculae, and by the arrangement of the trabeculae (Martin et al., 1998). Primary trabeculae may be stronger than remodeled ones, as the cement lines associated with trabecular packets of remodeled bone decrease its strength. Orientation of the trabeculae is also important: they are stronger when oriented parallel to applied loads (Martin et al., 1998). The trabeculae in the ends of long bones are oriented either along the lines of stress or in support of those that are (Thompson, 1942; Hildebrand, 1988; Martin et al., 1998). The porosity and trabecular structure of cancellous bone that make it weaker than cortical bone also make it more compliant, thus making it a good shock absorber (Martin and Burr, 1989).

PAGE 26

14 Variations in the Microscopic Structure of Vertebrate Bone Most vertebrates do not possess highly remodeled Haversian bone structure (Foote, 1916; Enlow and Brown, 1956, 1957, 1958; Currey, 1960). Foote (1916) made a histological comparison of the diaphysis of 440 femurs (294 nonhuman) encompassing all tetrapod vertebrate classes. He found that primary lamellar bone was the predominant type in the 39 species of amphibians examined. A small number had some plexiform bone, and a larger number had secondary Haversian systems. In general, the cortex was thin relative to the size of the bone. Among the 34 reptiles examined, primary lamellar bone predominated in the lizards, while Haversian bone was found in the turtles. Plexiform bone was found in the alligator and in some turtles. Most turtles follow the typical vertebrate pattern of long bone development. However, all turtles, whether terrestrial or aquatic, had medullary cavities densely filled with dense cancellous bone (Foote, 1916; Rhodin et al., 1981). In contrast, there was no cancellous bone in the lizards. Dinosaur bone was more similar to mammalian bone than to reptilian bone (Reid, 1984). Primary osteonal or plexiform bone was the most common types present. Secondary remodeling did occur, although the number and distribution of Haversian systems was highly variable. The degree of remodeling varied from none at all to extensive. Haversian distribution patterns ranged from randomly scattered to discreet regional concentrations. Areas with high densities of Haversian systems were associated with sites of muscular attachment, which is corroborated by the presence of Sharpeys fibers near these regions (Reid, 1984). Areas where tendons insert on bones are subject to heavy local stresses (Hildebrand, 1988). The diaphyseal medullary cavity was either empty or contained cancellous bone (Reid, 1984). Some of the trabeculae of primary

PAGE 27

15 cancellous bone were not remodeled, but rather lamellar bone was deposited onto the surfaces of primary endochondral trabeculae, and the cartilage core remained (Reid, 1984). Considerable variation in bone structure was noted among the 40 birds examined by Foote (1916). Although primary lamellar and Haversian bone were present in numerous species, he considered plexiform bone to be the most characteristic of the group. Cortical bone thickness varied considerably among species. He calculated that birds with empty bones (i.e., containing air sacs, no marrow) had thinner cortices relative to bone diameter than those with marrow, or those that had marrow cavities filled with cancellous bone, of similar size. There is relatively little variation in bone structure among birds compared to other vertebrate groups (Bellairs and Jenkin, 1960). The 188 mammals (nonhuman) examined by Foote (1916) were characterized by a great deal of structural diversity. Overall, secondary Haversian systems were present in a greater proportion of mammals than in any other class. However, some mammals were extensively remodeled, while others such as bats experienced little or none. Similarly, plexiform bone was found more often in mammals than in other classes. Among mammals, plexiform bone is found most commonly in the artiodactyls (Foote, 1916; Enlow and Brown, 1958; Martin and Burr, 1989). Manatees and their close relatives the elephants both have plexiform bone (Foote, 1916; Enlow and Brown, 1958; Marmontel, 1993). Deer antlers are plexiform, as are some long bones of large-breed canines, cats, and bears (Enlow and Brown, 1958; Currey, 1960). Plexiform bone has also been identified in the human mandible, and in the long bones of children who are under some degree of dietary stress (Martin and Burr, 1989). Foote (1916) found plexiform

PAGE 28

16 bone in many human fetal and juvenile femora. Currey (1960) reported a thin layer of plexiform bone on the periosteal surface of an adult human femur. Currey generalized that in many mammals larger than rats, some bones grow by deposition of plexiform bone at some stage of development. The extent to which primary bone is replaced by Haversian systems varies enormously from species to species, as well as with age, sex, and health, both within and among individuals (Foote, 1916; Currey, 1960). Haversian systems are almost completely absent from rodents and bats. Most bats possess primary lamellar bone except for the large Pteropus species, which do have some Haversian systems (Foote, 1916). Cats do not begin to remodel their bone until after reaching sexual maturity (Currey, 1960). Foote (1916) indicated the presence of cancellous bone in the femora of a number of large terrestrial and semi-aquatic mammals, such as the elephant, hippo, white rhino, tapir, otter, polar bear, bearded seal, and a number of artiodactyls. Wall (1983) also noted the presence of cancellous bone in the medullary cavities of the white rhino and the hippo. Bones in the bat wing have little or no medullary cavity and are poorly mineralized, making them less dense and stiff than most mammalian bone (Swartz and Watts, 1999). In contrast to terrestrial mammals, cetaceans have a reduced bone mass and density (de Buffrenil et al. 1990). The outer cortical layer of the humerus and radius were found to vary in thickness, but were relatively thin compared to terrestrial vertebrates, and the medullary cavities were filled with porous spongy bone (Felts and Spurrell, 1965, 1966). The forelimb is not a weight-bearing structure, nor is it used for propulsion, but rather it functions for maneuvering and attitude control (Felts and Spurrell, 1965, 1966;

PAGE 29

17 Domning and de Buffrenil, 1991). As such, it is exposed to significant tensile and compressive loads in the medio-lateral plane. The bones are correspondingly broad and flattened medio-laterally, modified from the tubular form typical of most animals. Foote (1916) observed that while all bones consist of the same structural units, they are combined in many different combinations and proportions. Because of this, he suggested that the lamella be considered the structural unit of bone. Other authors concur that bone tissues and their organizational patterns form a continuum rather than discreet categories (Currey, 1960; Moss, 1961; Martin et al., 1998). Moss (1961) stated that confusion arises when we try to classify skeletal tissue types in terms applicable to mammals alone. Many variations from what is considered the typical vertebrate bone structure exemplified by humans have been examined in detail. Some modifications characterize an entire class, while others are shared by species from more than one group. Some of these variations are described below. Three significant deviations from the classical vertebrate bone structure are found among the fishes. Elasmobranchs (sharks, skates, and rays) have a cartilaginous skeleton that is not replaced by endochondral bone (Budker, 1971; Compagno, 1999). In most elasmobranchs, however, the skeleton is strengthened by calcification of the cartilage. Calcified polygons of hydroxyapatite cover the surface of bones, and the vertebral centra are partially calcified in the interior (Compagno, 1999). There are three general patterns of vertebral calcification, and these patterns are so consistent that they have been used to classify sharks taxonomically (Goodrich, 1930; Budker, 1971). Additionally, the degree of calcification somewhat reflects age and habitat preference (Compagno, 1999). Skeletons of older animals are generally more calcified than younger ones. Inshore

PAGE 30

18 demersal sharks possess the most heavily calcified skeletons, while deep-water and some oceanic and inshore sharks are variably less calcified. The teleosts, or bony fishes, are unique in that their bone is largely acellular (Moss, 1961). Acellular bone is distinguished from cellular bone by the absence of osteocytes. It can be formed by two different mechanisms, through modified osteogenesis or chondroidal osteogenesis (Moss, 1961). During endochondral ossification, osteoblasts secrete matrix around themselves and become embedded. However, soon afterward the nuclei become pycnotic and die. The lacunae are subsequently filled in with bone. In appositional bone deposition the osteoblasts at the periosteal surface secrete matrix, but they continually withdraw from the surface. As a result, they never become embedded in the matrix, thereby producing consecutive layers of acellular bone (Moss, 1961). The process of acellular bone formation is not well understood, but it appears that there is a buildup of calcium salts inside the pycnotic cells. This intracytoplasmic mineralization leads to the death of the embedded cells (Moss, 1961).The mechanical properties of acellular bone are unknown (Wainwright et al., 1982; Currey, 1984). Chondroid bone, produced by a modified form of intramembranous bone formation, has characteristics of bone and cartilage (Fang and Hall, 1997; Kawakami et al., 2001; Witten and Hall, 2002). Chondroidal osteogenesis is a process of bone formation found primarily in fish (Moss, 1961). Cellular and acellular bone are formed from chondroid. The kype that appears in the Atlantic salmon during breeding season is composed of chondroid bone (Witten and Hall, 2003). This structure develops rapidly, and is structurally similar to bone formed under pathological conditions in humans such

PAGE 31

19 as periosteal osteosarcomas. Additionally, there is evidence that bone morphogenetic proteins are involved in the production of chondroid bone (Kawakami et al., 2001). Adventitious (secondary) cartilage forms on membrane bone in response to mechanical stimulation (Fang and Hall, 1997). This type of cartilage is found in skulls of birds and mammals, however, it does not form in amphibians and it is unclear whether or not it forms in reptiles. Additionally, amphibians do not form a cartilaginous callus in response to fracture, and whether or not reptiles do is unknown (Hall, 1984). It is also unclear whether these groups lack adventitial cartilage because their periostea have a reduced osteogenic potential compared to birds and mammals, or because the required mechanical stimulus for its formation is lacking (Hall, 1984). Reptiles commonly form metaplastic bone, in which chondrocytes are transformed into osteoblasts and cartilaginous matrix becomes bone (Hall, 1984). Additionally, metaplastic bone is also formed by chondrification and subsequent ossification of tendons and ligaments (Reid, 1984). It is possible that reptiles form metaplastic bone to compensate for the reduced boneand cartilageforming capabilities of the periosteum (Hall, 1984). Metaplastic bone similar to that in reptiles was also found in dinosaurs, typically in ossified tendons, but occasionally in the periosteal regions of limb bones (Reid, 1984). Osseous metaplasias have also been documented in a variety of soft tissues in humans, where the condition is always pathological (Dayoub et al., 2003; Siomin et al., 2003). A skeletal adaptation in birds is pneumatization, the inclusion of air sacs in the bones. Many bones of the skull and postcranial skeleton are hollow and contain air sacs lined with epithelium, which are continuous with the nasal sacs, tympanic cavities, or the

PAGE 32

20 bronchial system (Bellairs and Jenkin, 1960). The post-cranial skeleton is pneumatized by extensions of the respiratory air sacs. Typically, an extension of an air sac enters a bone through a pneumatic foramen. The sac extends into the marrow cavity, and as it grows around the degenerating trabeculae, it becomes broken up into a network of branching tubes. The trabeculae and marrow disappear from the bone shaft, but remain in the epiphyses in small quantities. The branches of the air sacs reunite to form a single sac that occupies the entire bone cavity (Bellairs and Jenkin, 1960). Pneumatization lightens the skeleton, and therefore is widely believed to be an adaptation for flight (Bellairs and Jenkin, 1960). The extent of pneumatization varies widely among species, but appears to be related to body size and lifestyle. Small birds, and aquatic birds such as penguins, are poorly pneumatized (Goodrich, 1930; Bellairs and Jenkin, 1960). In smaller species, the ratio of air sac volume to body surface area is small, so extensive pneumatization may not be economical. Diving in aquatic, flightless birds may be hindered by pneumatization. For larger, flighted birds the benefit of pneumatization may be significant, and this is especially beneficial because their skeleton is subjected to greater stress during flight (Bellairs and Jenkin, 1960). However, the cortical bone layer is about the same thickness in non-pneumatized bones of similar size. To compensate for this, pneumatized bones have a larger diameter than non-pneumatized bones of similar length, which increases its resistance to bending. Additionally, long bones of some large birds have extra internal support in the form of struts. Pneumatization has been documented in animals other than birds. The centra of some dinosaurs exhibit an unusual remodeled structure in which a few thin sheets of compact lamellae formed the external shell and a series of internal longitudinal partitions

PAGE 33

21 (Reid, 1984). The internal vertebral spaces were filled with air (Goodrich, 1930). This structure resulted in a substantial lightening of the centrum (Reid, 1984). Similarly, elephants have air-filled sinuses in the skull to lighten it (Feldhamer et al., 1999). In fishes, amphibians, reptiles, and birds, chondrocytes of the epiphyseal cartilages do not organize into vertical columns as seen in mammals (Bellairs and Jenkin, 1960; Moss, 1961; Rhodin et al., 1981; Hall, 1984). In reptiles and birds the epiphyses ossify differently from mammals (Bellairs and Jenkin, 1960; Hall, 1984). In small lizards, an area of epiphyseal cartilage never ossifies. In other lizards, there is no secondary center of ossification at all in the epiphysis, so ossification proceeds from the diaphysis. In these animals the articular cartilage functions as the growth plate. Most lizards do have secondary centers of ossification. However, ossification proceeds laterally, from the perichondreal surface inward (Hall, 1984). In contrast, the ossification patterns of large varanid lizards (e.g., Komodo dragon) are nearly identical to mammals. Unlike mammals, varanids retain the cartilage canals surrounding blood vessels of the epiphyses. Lizards that retain some cartilage in the epiphyses are capable of indeterminate growth (Hall, 1984). Like some other reptiles, turtles never develop secondary ossification centers in the epiphyses. As a result these animals are capable of continual endochondral growth throughout adulthood (Rhodin et al., 1981; Hall, 1984). Birds only have secondary centers of ossification at the proximal end of the tibia, which may be related to the invasion of the air sacs (Bellairs and Jenkin, 1960). Presence of secondary ossification centers might interfere with penetration of the sacs, and blood vessels associated with the ossification centers influence the arrangement of the chondrocytes.

PAGE 34

22 Pachyostosis is defined as an increase in cross-sectional area of bones by increasing the thickness of the periosteal cortices, and the absence of a free medullary cavity (Domning and de Buffrenil, 1991). Osteosclerosis refers to the replacement of cancellous bone with cortical bone (de Buffrenil et al., 1990; Domning and de Buffrenil, 1991). Increased bone density, mineralization, and reduced porosity are characteristic of this condition. Domning and de Buffrenil (1991) pointed out that the term pachyostosis is often, though erroneously, used to refer to both conditions simultaneously. The bones of some sirenians are, in fact, osteosclerotic without being pachyostotic. Fawcett (1942a, b) described the compact bone of manatee ribs as structurally less differentiated than that of other mammals; largely non-lamellar, periosteal bone laid down in an eccentric pattern, and the periosteum of the ribs continues to lay down new bone well after the animal has stopped growing. Large amounts of endochondral (primary) bone remain unabsorbed, and calcified cartilage matrix persists. The non-lamellar bone described by Fawcett (1942b) was later classified as plexiform bone by Enlow and Brown (1958). De Buffrenil and Schoevaert (1989) quantified pachyostosis in the dugong. They found that rather than being uniform, density and mineralization were greatest in the skull, anterior (cranial-most) vertebrae and ribs, and decreased caudad. A similar gradient existed in the forelimb, with greatest degree of pachyostosis in the scapula and humerus, and a lesser degree in the forearm and in the phalanges. In a review of sirenian phylogeny, Savage (1976) identified pachyostosis as the only truly autapomorph characteristic among the Sirenia. All sirenians, which first appeared in the middle Eocene, exhibited pachyostosis to some degree.

PAGE 35

23 Pachyostosis and osteosclerosis are not unique to sirenians. Pachyostosis is found in many species that are at least partially adapted to the aquatic environment (Wall, 1983). Nopsca (1923) identified the condition in dinosaurs. Pachyostotic bones have been described in pelicans and penguins (Foote, 1916; Meister, 1962). Walruses have a very dense rostrum, while the rest of the skull is comprised of spongy bone. This is hypothesized to aid the animals in diving for food (Kaiser 1965, 1967). De Buffrenil et al. (1990) described pachyostotic bone in two archaeocetes. Wall (1983) found that limb bone density was significantly greater in aquatic and semi-aquatic mammals than in terrestrial mammals. Stein (1989) compared limb bone densities among 15 species of marsupials and rodents. Fish and Stein (1991) compared the limb bone densities of several mustelids. Both studies reported that densities of limb bones of semi-aquatic species were significantly greater than those of terrestrial species. When Fish and Stein (1991) combined their data with that of Wall (1983), they found that some semi-aquatic mammals exhibited increased limb bone densities similar to fully aquatic mammals while others did not differ from terrestrial animals (Fish and Stein, 1991). Pachyostotic bone has also been found in a wide variety of fossil artiodactyls (Morales et al., 1993). Pachyostosis is believed to be an adaptation to promote reduced buoyancy in aquatic and semi-aquatic animals (Nopsca, 1923; Meister, 1962; Kaiser 1965, 1967; Felts and Spurrell, 1965, 1966; Rhodin et al., 1981; Wall, 1983; de Buffrenil and Schoevaert, 1989; Domning and de Buffrenil, 1991). Bone density in semi-aquatic animals appears to be related to the extent to which the animals are adapted to the aquatic environment (Wall, 1983; Stein, 1989; Fish and Stein, 1991). Wall (1983) postulated that selection for increased limb bone density occurred when total body submergence became a habitual

PAGE 36

24 part of the animals lifestyle. However, some fully aquatic mammals such as the cetaceans likely underwent a secondary reduction in bone density as they developed other methods of buoyancy control related to deep diving. The occurrence of pachyostotic bone in manatees has been likened to pathological conditions observed in humans. Fawcett (1942b) described it as resembling Albers-Schnberg disease, or marble bone disease, an autosomal dominant form of osteopetrosis. Increased bone density is a primary feature of the sclerosing diseases. However, many forms of osteopetrosis and osteosclerosis are accompanied by skeletal architectural deformities and a suite of soft tissue and physiological defects (Shapiro, 1993; Tolar et al., 2004) that are not present in manatees or other species exhibiting pachyostosis. Autosomal dominant osteopetrosis lacks the pathologies of other forms. Another candidate for the cause of pachyostosis is the recently discovered role of the LRP5 gene in skeletal mineralization. Some mutations of the gene have been shown to result in high bone mass, accompanied by an increase in bone volume and reduced marrow spaces, but without any other evident pathologies (Johnson et al., 2002; Koay and Brown, 2005). Manatees in Perspective Manatees share a number of developmental features with other marine animals that are not found in terrestrial vertebrates. Manatees (Fawcett, 1942a, b), cetaceans (Felts and Spurrell, 1965, 1966), penguins (Meister, 1962), and the marine turtle Dermochelys (Rhodin et al., 1981) all develop amedullary long bones that retain some endochondral bone that is not remodeled. In all these species except manatees the medullary cavity is filled with coarse-cancellous bone; manatees fill the cavity with compact bone. Additionally, these aquatic animals vascularize their epiphyses both perichondrally and transphyseally. Some pinnipeds (Versaggi, 1977) and extinct

PAGE 37

25 amphibians (Rhodin et al., 1981) also show similar patterns of development. Manatees and penguins both have an unusually thick periosteum that adheres tightly to the bone surface (Fawcett, 1942b; Meister, 1962). In manatees the periosteum of the ribs continues to lay down new bone well after the animal has stopped growing (Fawcett, 1942b). Felts and Spurrell (1965, 1966) identified some common structural and developmental characteristics of cetacean limb bones. They found that although cetaceans retain the typical mammalian pattern of fusion of ossification centers, the process is delayed. The zones of cartilage maturation are not as distinct or as organized as they are in terrestrial mammals. At no time during development are regular columns of chondrocytes observed near the growth plates, a feature shared with fishes, amphibians, reptiles, and birds. Secondary ossification of all cetacean limb elements begins after birth. Furthermore, there is an extended lag time between cartilage calcification and endochondral ossification. All of these features have been described in manatees (Fawcett, 1942b). Sirenians (e.g., manatees and dugongs) and cetaceans (e.g., whales) are the only fully aquatic marine mammals. As a result of their aquatic lifestyle, they are free of the mechanical constraints that influence limb bone architecture in animals that spend at least part of their time on land. In contrast to the cetaceans, manatees and dugongs are the only herbivorous, fully aquatic mammals (Domning and de Buffrenil, 1990). As herbivores, they are tied to the shallow waters of the euphotic zone, which has likely hindered them from taking advantage of the same hydrodynamic mechanisms of buoyancy control utilized by cetaceans (Domning and de Buffrenil, 1990). Domning and de Buffrenil (1991) offered reasonable evidence that pachyostosis and osteosclerosis in the manatee

PAGE 38

26 are adaptations to increase skeletal ballast for buoyancy control. Increased bone diameter, density, mineralization, and compactness all serve to increase the mass of the skeleton to offset the positive buoyancy of the lungs. Although increased mineralization will increase the mass of the bones, Currey (1969) demonstrated that as mineral content rises beyond a certain point, bone becomes more brittle in static and impact loading. Increasing bone diameter may offset the reduction in strength somewhat, by increasing flexural stiffness. In manatee ribs, pachyostosis is accomplished by the addition of plexiform bone, largely to the lateral surface. Mechanically, this may be advantageous compared to purely lamellar bone. When a load is applied to plexiform bone, the woven bone will experience less stress than the lamellar bone due to its lower modulus (Wainwright et al., 1982). Additionally, remodeling rates appear to be low in manatee ribs, as indicated by the low number of Haversian systems. This may be because manatees, being aquatic animals, do not experience loads that influence remodeling to the same degree as terrestrial animals. The crack-stopping ability of osteons increases bone toughness (Martin et al., 1998), so the low number of osteons in manatee bone may contribute to its low toughness compared to other bone. Most developmental and structural traits of manatee bone are shared with other animals, a result of phylogenetic relatedness and common evolutionary problems associated with the return to the marine environment. Other features such as long bones constructed of solid compact bone likely result from the manatees unique ecological niche. Although the material properties of bone types have been studied, manatees provide a unique combination of structure and function. There are two articles that

PAGE 39

27 qualitatively describe the histology of manatee bone (Fawcett, 1942b; Domning and Myrick, 1980). Marmontel (1993) qualitatively assessed remodeling rates in a number of bones. Domning and de Buffrenil, 1991 measured density, mineral content, and porosity for manatee rib bone for a range of animal sizes. Yan (2002) and Yan et al. (2005) calculated fracture toughness for rib bone. There is no information on the mechanical properties of whole bone, or of manatee bone tested in impact. This study is the first to comprehensively measure mechanical properties of manatee bone. The overall goal of this project was to quantify the biomechanical effects of boat strikes on manatees. This was accomplished by integrating basic research data on bone biomechanics with data from impact tests, which have a more practical application. We conducted two kinds of tests to measure both material properties and structural properties of manatee bone. Material properties are those of the tissue itself, independent of whole bone size or shape. Structural properties are those properties conferred by the size and shape of whole bones, in addition to the properties of the material. For example, a block-shaped vertebra will behave differently from a long, slender rib when tested in compression, even if they are constructed out of the same material. In Chapter 2, the objective was to conduct 3 point flexure tests using standard test protocols to measure material properties of manatee rib compact bone. Two hypotheses were tested: 1) manatee bone is less strong and tough compared to other species when loaded statically, and 2) the biomechanical properties of manatee bone vary between sexes and among age classes. We measured strength, Youngs modulus, fracture toughness, and work of fracture for both sexes, in three age classes. Chapter 3 addresses the hypothesis that forces generated by most watercraft under normal operation are sufficient to inflict fatal

PAGE 40

28 skeletal injuries to manatees. Impact tests were used to estimate the energy needed to fracture whole ribs and the stress at failure. The work presented here provides the first detailed information about material and structural properties of manatee bone. Data generated by the studies herein will contribute to our understanding of manatee-boat interactions, and will be instrumental in shifting the focus to a more objective approach for establishing regulatory boat speed zones adequate to reduce watercraft-related mortality.

PAGE 41

CHAPTER 2 MATERIAL PROPERTIES OF MANATEE RIB BONE Introduction The Florida manatee (Trichechus manatus latirostris) is listed as endangered by the U.S. Department of the Interior, and is considered one of the most endangered marine mammals found in U.S. waters (U.S. Fish and Wildlife Service, 1996). Sources of manatee mortality have been documented in Florida since 1974 (OShea et al., 1985; Ackerman et al., 1995; Wright et al., 1995). Watercraft-related mortality, caused by propeller wounds or by impact with a boats hull, accounted for 25% of all deaths from 1974 through 2004, and comprised 85% of human-related deaths. The proportion of deaths due to impact accounts for more than half (55%) of all watercraft-related deaths (Wright et al., 1995). Animals killed by propeller wounds rarely suffer skeletal damage. Conversely, approximately 66% of the carcasses attributed to hull impact injuries have fractured and/or luxated ribs (Wright et al., 1995). Furthermore, healed fractures are observed infrequently. This suggests that while many propeller wounds are sub-lethal, hull impact injuries sufficient to inflict skeletal damage are usually fatal. The number of documented watercraft-related deaths increased at a rate of 7% annually from 1992 through 2004 (Florida Fish and Wildlife Conservation Commission, 2003). Attempts at reducing watercraft-related mortality have focused primarily on regulation of boating activities, such as creating speed zones in areas where manatees and boats coexist. However, the increasing human population along the Florida coastline has resulted in more boats utilizing the waterways. From 1996 through 2004, the number of 29

PAGE 42

30 watercraft on Florida waters increased by 3% per year (U.S. Coast Guard, 2005). In 2004, with over 928,000 registered boats in the state; Florida moved ahead of California and Michigan to be ranked the number one recreational boating state in the U.S. (Florida Department of Highway Safety and Motor Vehicles, unpublished data; U.S. Coast Guard, 2005). Additionally, new boat and engine designs allow boaters to utilize shallow areas once inaccessible. Thus, the potential threat to manatees posed by watercraft continues to increase. Creation of boat speed zones is a highly subjective process, not based on any information pertaining to the biological effects of boat strikes on manatees. A major goal of the U.S. Fish and Wildlife Services manatee recovery plan is to refine the boat speed zones rules to reduce boat strike deaths (U.S. Fish and Wildlife Service, 2000). Toward this goal, the plan calls for research on mechanical characteristics of manatee bone. The data from this study will be useful in shifting the focus to a more objective approach for establishing regulatory boat speed zones adequate to reduce watercraft-related mortality. Manatee bone is unusual compared to that of other marine mammals. Compared to terrestrial mammals, the general trend in marine mammals has been a reduction of bone mass and density (de Buffrenil et al., 1990). In contrast to most marine mammals, the manatee skeleton exhibits pachyostosis, characterized by thickening of bone tissue, replacement of cancellous bone with compact bone, and absence of a free medullary cavity (Fawcett, 1942b; de Buffrenil et al., 1990; Domning and de Buffrenil, 1991). The mechanical properties of bone have been well studied for humans and some domestic animals (Wainwright et al., 1982; Currey, 2002); however, mechanical studies on the bones of marine mammals have been few. Mechanical property data are available for fin whale (Balaenoptera physalus), dense-beaked whale (Mesoplodon densirostris), narwhal

PAGE 43

31 (Monodon monocceros), polar bear (Ursus maritimus), walrus (Odobenus rosmarus), and dugong (Dugong dugon) (Currey, 1979a; Currey and Brear, 1990; Evans et al., 1990; Brear et al., 1993; Currey et al., 1994; Zioupos et al., 1997; Currey, 1999). Some information on the composition of manatee bone exists (Fawcett, 1942a, b; Domning and Myrick, 1980; Domning and de Buffrenil, 1991; Marmontel, 1993). Fawcett (1942b) and Domning and Myrick (1980) provide qualitative reports on the microstructure of rib bone. Domining and de Buffrenil (1991) measured density, mineral content, and porosity (measured as its inverse, compactness). They found that mineral content and density increased with body size through the subadults, and leveled off in adults. Additionally, de Buffrenil and Schoevaert (1989) examined bone density in the dugong, a close relative of the manatee. They found a bone density gradient that increased from craniad to caudad, and from dorsad to ventrad. In developing an aging technique for manatees, Marmontel (1993) examined histological sections of various bones. She reported differences in the number of osteons in males and females of similar age, and inferred that there may be differences in remodeling rate between the sexes. We recently reported on the use of fractography and Chevron notch beam tests to measure fracture toughness (Yan, 2002; Yan et al., 2005). The present study is the first comprehensive measure of the mechanical properties of manatee bone. The first step in quantifying the biomechanical effects of boat strikes on manatees was to measure some material properties of manatee rib compact bone. Based on the limited information on manatee bone available in the literature, we hypothesized that manatee bone is less strong and tough compared to other species when loaded statically, and that the material properties of manatee bone vary between sexes and among age

PAGE 44

32 classes. The objectives were to calculate flexural strength, toughness, Youngs modulus, and work of fracture in bending for rib bone, and to measure density and mineral content for manatees of both sexes in three age classes. Materials and Methods Bone Sample Collection Bone tissues were obtained from the Florida Fish and Wildlife Conservation Commissions Marine Mammal Pathobiology Laboratory. Collection and use of tissues for research was conducted under U.S. Fish and Wildlife Service permit #MA067116-0 issued to the University of Florida. Animals of both sexes in three age classes (calf, subadult, adult) were tested. Because it is difficult to age manatees, there is little age data available. By convention, total body length (cm) is used as a proxy for age (OShea et al, 1985). Calves were defined as 175 cm total body length, subadults were 176 cm total body length, and adults > 275 cm total body length (OShea et al., 1985). Three ribs were obtained from fresh carcasses at the time of necropsy. One rib was taken from the cranial, middle, and caudal thoracic regions (Figure 2-1). Ribs were taken from the same positions whenever possible, and stored frozen in plastic bags until machining. Freezing and thawing of bone does not significantly alter the mechanical properties of bone (Sedlin and Hirsch, 1966; Currey, 1988b; Pelker et al., 1984). They were brought to room temperature for machining and testing. Specimen Preparation A band saw was used to rough cut 60 mm segments from the proximal and middle 1/3 sections of the ribs under a constant spray of water, to keep the bone from heating and to keep it wet. Specimens with dimensions of 50 x 3 x 3 mm were machined with a diamond-edged band on a wet saw (Exakt Technologies, Hamburg, Germany), with the

PAGE 45

33 long axis of the test pieces parallel to the long axis of the rib. Dimensions were verified with digital calipers (Mitutoyo, Tokyo, Japan) to the nearest 0.01 mm. Machined specimens were stored in physiological buffered saline until testing, and they were tested wet. Material Properties Testing Three point bending tests were performed with a Mini-Bionix materials testing machine following ASTM standard D790M-92 for flexural testing of plastics (MTS Systems, Eden Prairie, MN; ASTM, 2001). For each specimen, a 10-cycle, 10-Newton preload was applied, then loaded to failure using a standard 3-point bending fixture at a displacement rate of 1mm/min. The load was applied to the surface that corresponded to the lateral surface of the rib (i.e., tested in the transverse orientation). Load and displacement were recorded at 5 Hz, and output to a personal computer for analysis. Mechanical variables measured were flexural strength, Youngs modulus (i.e., elastic modulus) in bending, fracture toughness, and work of fracture. The flexural strength of a material is the load required to break a piece of the material in bending (Wainwright et al., 1982). Flexural strength in 3-point bending was calculated as 2BD2FL3 (2-1) where F is peak load (Newtons) at failure, l is the distance (meters) between the supports of the bending fixture, B is the specimen width (m), and D is the specimen depth (m). A correction factor was applied to the strength calculation if the specimen fractured somewhere other than opposite the point at which the load was applied (Riley et al., 1999).The correction factor was calculated as

PAGE 46

34 L)x)2/L((2corr (2-2) where corr is the corrected flexural strength, L is the original length of the test specimen, and x is the length of the shorter of the two halves after testing, and is the flexural strength from Eq. 2-1. Youngs modulus is an indication of how easily the material strains elastically under a given stress; it is a measure of the stiffness of the material (Jackson, 1992). Youngs modulus in bending was calculated as I48L)x/F(E3 (2-3) where F/x is the slope of the load-displacement curve, l is the distance between the supports of the bending fixture (m), and I is the moment of inertia, calculated as 12BDI3 (2-4) Yan et al. (2005) showed that, unlike typical bone, manatee bone behaves more as a ceramic. Therefore, fractographic analysis and fracture mechanics can be used to calculate fracture toughness. Fracture toughness is the ability to resist fracture, measured as the critical stress intensity factor (K C ). This is an estimate of the amount of energy required to propagate a macrocrack that leads to fracture (Mecholsky, 1996); it is a measure of the behavior in the pre-yield region. The critical stress intensity factor was calculated as K C = Y(c) 1/2 (2-5) where Y is a geometric factor based on crack shape and loading conditions, is flexural strength, and c is the crack size (m 2 ). Crack size is calculated as c = (ab) 1/2 (2-6)

PAGE 47

35 where a and b are crack depth and half-width (m), respectively, as measured from the fracture surface with a light microscope. A correction factor was applied if the failure did not originate from the surface of the specimen (Mecholsky, 1993). Scanning electron micrographs were taken of some fracture surfaces for detailed examination and validation of light microscope measurements. The work of fracture is another measure of energy. Like fracture toughness, work of fracture is that energy required to propagate a crack through a specimen (Aksel and Warren, 2003). It is calculated from the area under the stress-strain curve; it is a measure of the behavior in both the preand post-yield regions. For materials that have little or no post-yield deformation, work of fracture and fracture essentially measure the same property, i.e., toughness. However, for materials that have post-yield deformation, or exhibit some inelasticity in the pre-yield region, the two variables may be different. For those materials, work of fracture gives a more complete picture of the behavior. Work of fracture was calculated [cf. Appendix B] for specimens tested in 3-point flexure as 2BD2W3WOF (2-7) where W is the area under the load displacement curve (N m). Bone Density and Mineral Content Determination Extra bone from the middle sampling region of each rib was used to measure Youngs modulus with ultrasound to compare to calculated values, and to measure density and mineral content. Since calculation of Youngs modulus from the load-displacement curve is prone to error, ultrasound was used to as a check for the calculated modulus (Vincent, 1990). Specimens were defatted with ammonia, dried at 95 for 48 h, and weighed to the nearest 0.001 g. Density was determined using an Accupyc 1330

PAGE 48

36 helium pycnometer (Micromeritics Instrument Corp., Norcross, GA). Ultrasound measurements were made with a Nuson Ultrasan (Nuson, Boalsburg, PA). Samples were then incinerated at 800C for 10 h, and weighed to the nearest 0.001 g. Mineral content was expressed as a percent of ash weight over dry weight (Kaufman and Siffert, 2001). Statistical Analyses Statistical analyses were conducted both within and among animals. For each individual, specimens within each sample site were compared (Figure 2-1B) to look for differences within sample sites. Next, specimens for each sample site were pooled to compare among sample sites within individuals. To examine ontogenetic changes in material properties, specimens were pooled across sites to calculate means by animal. For each variable, animal means were regressed against total body length to examine ontogenetic patterns. Analysis of covariance (ANCOVA) was performed, with length as the covariate, to look for differences between the sexes (Zar, 1996). For all tests, results were considered significant if p<0.05. Analyses were performed with Statistical Analysis Systems using Proc GLM (SAS Institute, Inc., 1989). Results Six males and six females were tested for the subadult and adult age classes (Table 2-1). Tissues were obtained from only two calves. Owing to their small size, it is difficult to recover calf carcasses in fresh condition. The ribs of the smallest calf (total body length 102 cm, a newborn calf) were too small to obtain more than one or two specimens each. Additionally, some ribs of the smaller subadults were too small to obtain samples from both the proximal and middle regions. For these animals, only the proximal region was taken.

PAGE 49

37 Within-Animal Comparisons A total of 3,220 specimens from 26 animals were tested in 3-point bending. Fracture toughness was determined for 1,650 of those. At each sample site, no significant differences were found among samples, indicating that the bone at each sample site was homogeneous. Specimens for each sample site were pooled Analyses comparing sampling locations within animals did not reveal any consistent patterns, either cranio-caudal or dorso-ventral, for any variable. Therefore, specimens were pooled for each animal for analyses by sex and total body length (a proxy for age). Mean values for each variable, by animal, are reported in Table 2-1. Material Properties Tests Figures 2-2 to 2-4 and 2-6 show the relationship of each mechanical variable to total body length. For each mechanical variable, ANCOVA indicated no difference between the sexes. Therefore, sexes were pooled for regression analyses of variables against total body length. For all variables, linear regression did not adequately describe the relationships between variables and total body length. Rather, the relationships were curvilinear; a second order polynomial regression was the best fit, and highly significant, in all cases. Flexural strength increased with body length up to 267 cm, then reached a plateau (R 2 = 0.79, p<0.0001, Figure 2-2). Fracture toughness exhibited a similar pattern (R 2 = 0.73, p<0.0001, Figure 2-3). Work of fracture increased with body size through the subadult size class, just as flexural strength and toughness (R 2 = 0.50, p = 0.0004, Figure 2-4). Unlike the other variables, work of fracture showed a strong decline with increasing body length for animals 265 cm total length or greater. Although there was a large amount of scatter among the subadults, the value for the smallest calf MEC0033 was much less than the

PAGE 50

38 others. The regression was run minus this animal to determine if it was responsible for the parabolic shape; it was not. Fracture energy, as measured by work of fracture, was greatest in animals of intermediate body length, about 200cm total length (i.e, the large subadults). Youngs modulus showed a different pattern, increasing with body length over the entire range of body sizes (R 2 = 0.86, p < 0.0001, Figure 2-5). Ultrasound was used to measure modulus of 13 bone samples from 11 animals. On average, the ultrasound modulus was 1 GPa greater than that calculated from the load-displacement curves. The good agreement between the ultrasound results and our calculated values indicates that our modulus values are reasonably accurate. Currey (2004) has shown that, in general, Youngs modulus and flexural strength are highly correlated (R 2 = 0.86). Our data reflected the same positive correlation, although the relationship was not as close (R 2 = 0.67, Figure 2-6). Figure 2-7 shows some example load-displacement curves for the small sample specimens tested here. Representative curves for each size class are given. To facilitate comparison among age classes, the values for ultimate load, Youngs modulus, and work of fracture are given for each curve. Bone Density and Mineral Content Extra bone was available to measure density and mineral content for 20 animals (Table 2-1). Mean density by animal was 2.17 g/cm 3 ( 0.04 SD, range 2.10.22 g/cm 3 ), and mean mineral content was 69% ( 2% SD, range 64%). Density showed a weak positive correlation with total length, however, the correlation was not significant (R 2 = 0.19, p = 0.0517, Figure 2-8). Percent mineral content also correlated weakly to total body length, although the relationship was significant (R 2 = 0.27, p = 0.0198, Figure 2-9) In contrast, density showed a strong, positive correlation with elastic modulus with the

PAGE 51

39 removal of one outlier (R 2 = 0.64, p<0.0001, Figure 2-10). Percent mineral content also showed a strong correlation with modulus (R 2 = 0.61, p<0.0001, Figure 2-11). Our results show that while modulus was highly correlated to both bone tissue density and mineral content, only mineral content increased with total body length. Discussion Reducing watercraft-related manatee mortality is identified as a Level 1 high priority in the Florida Manatee Recovery Plan of the U.S. Fish and Wildlife Service (U.S. Fish and Wildlife Service, 2000). The recovery plan calls for research on mechanical characteristics of bone to obtain a better understanding of the effects of watercraft-related impacts To meet this objective, the goal of this study was to measure some material properties of manatee rib compact bone. Based on the limited information available in the literature, we hypothesized that manatee bone is less strong and tough compared to other species when loaded statically, and that the material properties of manatee bone vary between sexes and with the size of the animal. The Material Properties of Manatee Bone The main finding of this study is that manatee bone is on average less strong and tough than other mammals. Mean flexural strength by animal ranged from 62 MPa. In comparison, typical flexural strengths for human and bovine bone specimens tested in 3-point bending are 209 MPa and 224 MPa, respectively (Martin et al., 1998). Flexural strengths have been reported for horse radius (154-249 MPa), black bear tibia (211 MPa), and ox femur (227 MPa) (Schryver, 1978; Wainwright et al., 1982; Reilly and Currey, 1999; Batson et al., 2000; Harvey and Donahue, 2004). We would expect normally manatee bone to have a bending strength comparable to that of other animals with similar organization. However, we predicted that out values would be lower than

PAGE 52

40 values for other species, based on the high mineral content reported by others. We did find higher than average mineral content, is likely the reason for the low strengths (see below). The Youngs modulus, or the modulus of elasticity, is the measure of the stiffness of a material. Mean modulus ranged from 4 GPa, which was similar to that for equine metacarpals (14 GPa), horse radius (14 GPa), human femur (15 GPa), cow femur (20 GPa), and black bear tibia (16-30 GPa) (Martin et al., 1998; Zioupos and Currey, 1998; Currey, 2001; Les et al., 2002; Harvey and Donahue, 2004). Fracture toughness is the ability to resist fracture, one measure of which is the critical stress intensity factor (K C ). Toughness is an important property because a tough bone is more resistant to fracture, and this is most important when bone is impacted (Wainwright et al., 1982; Martin et al., 1998). Because bone is viscoelastic, it will behave differently when loaded at different rates (Martin et al., 1998). Generally, the more rapidly bone is loaded, the greater the modulus, and the less energy it is able to absorb. The bone mineral content and density previously described in manatees led us to predict that manatee bone may be on average less tough than other bone (Domning and de Buffrenil, 1991). As predicted, data reported here indicate that manatee bone (1.4.9 MPam 1/2 ) is generally less tough than other bone such as human tibia (2.2.7 MPam 1/2 ), bovine femur and tibia (2.2.5 MPam 1/2 ), and equine metacarpus (4.4.5 MPam 1/2 ) (Martin et al., 1998; Malik et al., 2003). The ontogenetic pattern in fracture toughness mirrored the pattern seen in flexural strength. Since fracture toughness is proportional to both the flexural strength and the flaw size (Equation 2-5), our results suggest that the flaw size is similar over the range of body sizes tested here. A plot of

PAGE 53

41 mean crack size by animal vs. total body length indicated that there was no difference in crack size in relation to body size (R 2 = 0.0009, p = 0.8886, Figure 2-12). For both strength and fracture toughness, the regressions may indicate a slight decline for the largest animals. It appeared as though the largest animal, MSW0245, may be responsible for this. This animal had been aged at 25, and many of the test specimens had a qualitatively different look from those of other adults. It may be that this possible decline in strength and toughness in the largest animals is a result of age-related decreases in material properties similar to those well-documented in humans. However, without more ages available for the other larger animals, and without more animals of advanced age, this cannot be determined conclusively. Another measure of toughness is work of fracture. The work of fracture for manatee bone varied from 3.0.0 MJ/m 3 which is considerably lower than the 28 MJ/m 3 for human femur, but well within the range of values reported for other mammals (Currey and Pond, 1989). Harvey and Donahue (2004) recorded values of 3.2-9.4 MJ/m 3 for black bear tibia. Currey (2002) reported works of fracture between 0.3.6 MJ/m 3 for a wide variety of terrestrial and marine animals. Typically, young bone has a low work of fracture that first increases with age, then declines. Manatee bone followed the same pattern. Mean manatee rib bone density of 2.17 g/cm 3 was greater than that reported for humans, which is typically 1.7.9 g/cm 3 but equal to that of cow femur (2.1 g/cm 3 ) and baboon femur (1.90.17 g/cm 3 ) (Yeni et al., 1998; Phelps et al., 2000; Currey, 2002). De Buffrenil and Schoevaert (1989) reported densities of 1.79.10 for ribs from three dugongs (2 subadults and 1 adult), a close relative of manatees who also exhibit

PAGE 54

42 pachyostosis of the skeleton. In similar fashion, mineral content for the manatee bone averaged 69%, greater than what is commonly reported for other species. With one exception, all values were between 68-71%. Even greater values of up to 75% mineral content (range 67%) were reported for dugong ribs (de Buffrenil and Schoevaert, 1989). According to Martin et al. (1998), mineral content is typically 65 3% for cortical bone. Currey and Butler (1975) reported mineral contents of 60% for human femora aged 2 years, and 64% for black bear tibias from animals 2 years (Harvey and Donahue, 2004). It is important to clearly define the terms density and mineral content. Density can be measured two ways, as the weight per unit volume of only the constituent material, or of the volume inclusive of the spaces, or voids (Martin et al., 1998). The latter is termed apparent density. Thus, bone that is highly porous will have low apparent density. If, however, just the bone tissue is considered, the true tissue density of the porous sample would be just as high as a volume completely filled with the same material. The pycnometer used for this study forces helium gas, under pressure, into the sample chamber to determine volume. In concept, this fills the sample voids if the permeability is adequate, thereby measuring true density. In similar fashion, mineral content can be construed as either the amount of mineral per unit sample volume (i.e., volumetric mineralization), or per volume of bone matrix, exclusive of voids (i.e., specific mineralization). Ashing bone, as done here, is a measure of specific mineralization. A more complete picture of manatee bone density and mineralization patterns may have been drawn if we had also calculated apparent density and volumetric mineralization.

PAGE 55

43 We can summarize our findings about manatees bone as follows: flexural strength was less than for other species; both stiffness and fracture toughness were roughly equal to or less than other species; and work of fracture was similar to values reported for a number of species. Material properties of manatee cortical bone increased with total body length up to about 267 cm, then leveled off, or declined as for work of fracture, with increasing body size. We expected this pattern to be related to an increase in density and/or mineral content with increasing body length. This appears to be what happened, as density and mineral content increased with body size. However, the relationship between density and body size was not significant. This is likely due to two things: that the variation in density was slight, and that comparatively our values were at the high end of the spectrum of values reported in the literature. Mineral content showed greater variation, and increased significantly with body size. Importantly, mineral content was higher than that seen in most mammalian bone. The high mineral contents, coupled with densities at the high end of published values may explain the patterns of the mechanical variables to total body length in animals greater than 265 cm. To better understand the material behavior of manatee rib bone, additional factors need to be evaluated, such as its plexiform structure, collagen content/quality, and its reportedly low remodeling rate (Marmontel, 1993). We can look to Domning and de Buffrenil (1991) to gain insight into our findings. The authors determined density, mineral content, and compactness of manatee rib for 12 animals. Compactness is the percent of the total cross sectional area that is occupied by bone; it is the inverse of porosity. Total body lengths were missing for the three largest animals. Of the nine animals of known body length, only one was an adult (324 cm TL);

PAGE 56

44 the others were calves and subadults (141 cm TL). It is important to note that the authors calculated apparent density, whereas we calculated the material density. They reported densities of 1.59.07 g/cm 3 (mean 1.85 g/cm 3 ), lower than our densities. However, mineral content for those animals was 67%, essentially identical to our data. There was a strong correlation between mineral content and total length, and apparent density and total body length. However, that was related to the fact that those animals were primarily calves and subadults (i.e, less than 275 cm TL); we saw a similar pattern in our data over the same size range. The authors noted that there seemed to be no relation of mineral content and density to body size greater than 275 cm. Compactness varied from 84%, which translates to 6% porosity. Porosity declined with increasing body length (R 2 = 0.53). The authors concluded that mineralization was not dependent on size, and that the increase in apparent density with size was due to the decrease in porosity. Considering both datasets together, we can draw some conclusions about the relationship between mineral content and material properties. Manatee bone tissue density, consistently high, varies little over a wide range of body sizes. The increase in apparent density and mineral content with body size, along with a decrease in porosity, indicates an increase in bone tissue volume. Furthermore, there appears to be little change in the quality of the material. As the animals grow, the interstitial spaces are filled in with more bone. Some additional mineral is added to the existing bone. However, because the existing bone is already so highly mineralized, the resulting bone has reduced strength and toughness. Thus, the ontogenetic patterns in mechanical variables appears to be related to the amount of material present, but the overall low values for strength and toughness are related to the quality of the material.

PAGE 57

45 The Effect of Mineral Content on the Material Properties of Bone Bending strength of compact bone is primarily determined by the Youngs modulus (Currey, 1999). In turn, the primary determinant of modulus is mineral content and to a lesser extent, porosity and bone architecture (Schaffler and Burr, 1988; Currey, 1998, 2003, 2004). The effect of mineral content on bone material properties has been well studied, both in terms of changes with age (Currey and Butler, 1975; Currey, 1979b) and interspecific variation (Currey, 1969, 1988, 1999, 2002, 2004a, b). Dramatic changes in material properties result from a relatively small change in mineral content. Generally, for mineral contents of 63%, static strength, toughness, and modulus increase as mineral is deposited in the collagen. However, above 68%, strength may actually decrease as mineral content rises, due to the fact that the hydroxyapatite crystals coalesce and are therefore unable to stop cracks from propagating (Currey, 1969). Additionally, the impact strength of highly mineralized bone is less than its static strength, and impact strength drops more precipitously than static strength as mineralization increases (Currey, 1969). A decrease in toughness is generally associated with an increase in mineral content (Currey, 1969; Malik et al., 2003). Specifically, toughness, as measured by work of fracture, increases with mineral content up to 66%, and then declines precipitously as mineral content continues to rise (Currey, 1969). Our mineral content ranged from 6471% (mean 69%). This may be why we saw the decrease in strength and toughness for the largest manatees. Degree of bone mineralization is known to increase with age in humans, and there is a corresponding increase in strength and stiffness, and a decrease in absorption of impact energy (Currey, 1979b, 1984, 2002; Currey et al., 1996). Although this pattern is well documented in humans, this is not the case for all animals. Harvey and Donahue

PAGE 58

46 (2004, 2005) found that although mineral content, hence modulus, increased in black bear tibia from animals aged 2 years, there was no decline in strength, toughness, or porosity with age. Furthermore, they found no reduction in bone mechanical properties attributable to long periods of hibernation. Phelps et al. (2000) examined age-related changes in mechanical properties of the baboon femur, which is known to resemble human bone in microstructure. Although significant decreases in fracture toughness were noted for adult animals aged 6 years, they found no variation in bone density, mineral content, or porosity with age. Other Factors Influencing Bone Material Properties Material properties of cortical bone are determined in large part by mineral content, but also by porosity, histological architecture, and amount and quality of collagen (Martin et al., 1998). One deficit of this study is that we did not include a histological examination of our bone. This would have allowed us to measure porosity, and both quantitatively and qualitatively describe the type and arrangement of bone and collagen. Yan (2002) measured porosity from scanning electron micrographs for a few rib bone samples from one calf and four subadult manatees. Porosity varied from 3%, and was strongly correlated to flexural strength (R 2 = 0.88) and fracture toughness (R 2 = 0.89). However, porosity varied as much as 26% within the same rib, and was not significantly correlated to total body length. Although the very highest porosity (44%) was for the neonate MEC0033, some values for other animals were still somewhat greater than those reported by others, including 1% for both human femur and baboon femur, 4% for cow femur, and 2% for black bear tibia (Schaffler and Burr, 1988; Currey and Pond, 1989; Phelps et al., 2000; Harvey and Donahue, 2004). In a second set of measurements, Yan (2005) reported an average porosity of 7% for manatee rib and 4%

PAGE 59

47 for cow femur, values more in line with other published values. The average porosity for manatee rib reported by Domning and de Buffrenil (1991) was 10%. De Buffrenil and Schoevaert (1989) reported porosities of 16% for 1 subadult, to <1% for the adult. Increasingly, researchers are examining the influence of collagen on the mechanical properties of bone, as variations in structure, mineral content, and density are not sufficiently able to explain the decline in properties with age. Wang et al. (2001) found that collagen denaturation resulted in decreased strength and toughness of the human femur. They also found that modulus was constant irrespective of the degree of collagen degradation. In a related study, Wang et al. (2002) examined collagen content and quality in human femora aged 19 years. They found that age-related changes in collagen were associated with decreased strength, toughness, and work to fracture of bone. Moreover, the deterioration of collagen with age was significantly correlated to the post-yield deformation. Phelps et al. (2000) found that the organic fraction showed a significant, positive correlation with fracture toughness for baboon femora. The exact mechanism by which changes in collagen alter mechanical properties is not known. However, it has been speculated that changes in the amount of collagen or to either the interor intra-molecular stability of the collagen are responsible (Zioupos et al., 1999; Burr, 2002). Age-related changes in cross-linking have been shown to alter collagen structure, which in turn may alter its hydration (Lucchinetti, 2001; Wang et al., 2002). To date, studies on the effects of collagen cross linking have been contradictory. Regardless of the specific mechanism, there is growing evidence that collagen degradation with age is primarily responsible for the post-yield decrease in toughness and work of fracture, and that these declines are independent of mineral content or density (Lucchinetti, 2001; Burr,

PAGE 60

48 2002; Wang et al., 2001). In support of this, we found no relationship between work of fracture and mineral content for manatee rib bone (R 2 = 0.02). Bone type and degree of remodeling influence material properties. Mineral content for mature primary lamellar bone is 67% versus 63% for Haversian bone (Currey, 1969; Vincent, 1990). Bone remodeling replaces mature, mineralized bone with material that is initially less mineralized. Haversian bone has decreased strength and modulus compared to lamellar bone, but increased toughness. Remodeling may decrease strength by as much as 20% (Martin et al., 1998). That the mineral content of manatee bone is high relative to most lamellar bone may be indicative of low remodeling rates. Possible evidence for this may be from Fawcett (1942b), who conducted the only detailed examination of manatee bone histology to date. He described what is now known as plexiform bone, as well as endochondral (primary) bone that remained unremodeled, and some persistent calcified cartilage matrix. He did not quantify the extent of remodeling, only remarking that Haversian systems are predominantly located in the innermost third of the cross section, and their numbers diminish rapidly toward the outer edge. He also did not comment on whether the osteons were primary or secondary. Furthermore, the descriptive histology was based on one rib from one fresh adult animal, one dried humerus from a museum, and two preserved fetuses. Based on the limited information provided about the adult, its true age is ambiguous, and it was likely not skeletally mature. Marmontel (1993) examined histological sections of various manatee bones, including ribs. She estimated that bone resorption increased significantly with total body length. In addition, females had significantly greater numbers of Haversian systems than

PAGE 61

49 males. Her work suggested that there are age differences in manatee bone structure and composition that may directly influence its mechanical behavior. This pattern is different from what is seen in humans. It is generally thought that remodeling slows in older individuals, resulting in an increase in the amount of older, more mineralized bone. Hence, the modulus increases, and this contributes to the elevated susceptibility to fracture with age (Currey, 1979b). However, there is evidence to demonstrate that this is not true. Bone turnover has been found to increase in the elderly human population, compared to middle-aged adults (Gallagher et al., 1998; Gundberg, 2002). The scenario for manatees may be similar: an increase in remodeling with age may be partly responsible for the slight decline in strength seen in the largest individuals, as Haversian bone is weaker than lamellar. Similarly, sex differences in remodeling might result in significant variation in mechanical properties between males and females. In contrast to Marmontel (1993), we did not find any significant differences between the sexes for any mechanical variable. Histological examination of manatee bone will tell us whether or not remodeling differences exist. Manatee rib bone is composed of plexiform bone, which is a mix of primary lamellar and woven bone. Preliminary observations of undecalcified histological sections show that, in addition to plexiform bone, there appears to be a central area of woven bone. This may be what Fawcett (1942b) referred to as disorganized, primary endochondral bone. Woven bone, although less organized than lamellar, has the potential to become more mineralized (Currey, 1998; Martin et al., 1998). For manatees, the preponderance of woven bone, in conjunction with a low remodeling rate, likely accounts for the elevated bone density and high mineral content.

PAGE 62

50 Summary Manatees fill a unique aquatic niche. As the only herbivorous marine mammal, they inhabit shallow, coastal estuarine or riverine waters, feeding primarily on seagrasses and other aquatic vegetation. The pachyostotic long bones of the Florida manatee are hypothesized to regulate buoyancy in shallow water and serve as a means of attitude control (Domning and de Buffrenil, 1991). Because the ribs and long bones are not weight bearing in the aquatic environment, development of pachyostosis may have been evolutionarily advantageous. However, the arrival of the engine-driven watercraft to Florida waters has left the manatee ill-suited to cope with their encounters. Every year, 25% of all manatee deaths are due to injuries that result from bones broken in collisions with watercraft. To begin to understand the injuries from a bone mechanics perspective, we measured some material properties of manatee rib bone. We found that this bone is less strong and less tough than most other mammalian bone. We also found that, although the bone increases in static strength as the animals grow, the bone is not able to absorb more energy. In fact there is a strong indication that the ability to absorb energy (measured as work of fracture) declines with increasing size in animals over 265 cm total body length (i.e., large subadults). This decline is due in part to the bones high density and mineral content. Finally, we found no differences in mechanical properties between males and females. The pachyostotic nature of the bone, which makes the manatee well-adapted to its environment, also leaves it highly susceptible to fatal injuries from boats.

PAGE 63

51 Table 2-1. Mean material properties for 26 manatees Animal Sex TL Age N E WOF N KC K C Density Mineral (cm) class (MPa) (GPa) (MJ/m 3 ) (MPam 1/2 ) (g/cm 3 ) content MEC0033 F 102 C 5 62 4.0 3.1 5 1.4 MSTM0109 M 166 C 30 134 12.8 3.8 60 2.5 2.14 69 MSW0033 M 178 S 31 136 12.0 6.0 29 2.1 MSW0164 M 182 S 25 105 9.4 4.0 24 2.2 2.22 64 MNW0207 F 188 S 37 127 11.6 3.5 34 2.4 2.10 67 MSW0227 F 190 S 48 124 12.3 4.2 44 2.4 2.16 67 SWFTM0114 M 196 S 66 122 11.2 4.6 48 2.3 MSW0057 M 201 S 36 153 9.6 5.5 21 2.7 MSW0134 F 205 S 58 130 11.5 5.5 54 2.1 MSE0205 M 238 S 114 135 12.6 4.4 63 2.6 2.11 68 MSW0160 F 246 S 131 159 14.9 5.6 50 2.8 2.18 68 MSTM0102 F 265 S 169 157 15.4 4.3 155 2.6 MSW0165 F 267 S 157 152 14.5 4.6 59 2.7 2.15 68 MEC0220 M 275 S 139 144 14.4 3.6 59 2.6 2.11 69 MSW0251 M 276 A 142 160 17.7 3.6 114 2.8 2.20 71 MSW0225 F 282 A 160 151 16.0 3.8 114 2.7 2.18 69 MSW0223 F 294 A 184 153 15.5 4.0 60 2.8 2.20 69 MSW0208 F 299 A 175 154 16.0 4.2 60 2.8 2.22 70 MEC0213 F 314 A 143 143 16.0 3.2 60 2.6 2.16 68 MSW0226 M 315 A 156 151 16.6 3.2 60 2.7 2.18 69 MSW0239 M 321 A 181 150 16.4 3.7 60 2.8 2.21 70 MSW0331 F 326 A 173 150 17.2 3.2 118 2.6 2.19 68 MSW0253 M 327 A 189 149 15.7 3.8 119 2.6 2.19 69 SWFTM0414 F 335 A 189 158 17.4 3.6 60 2.9 2.20 69 MNW0208 M 343 A 178 159 18.0 3.0 60 2.9 2.22 70 MSW0245 M 344 A 307 143 15.6 3.1 93 2.3 2.16 68 Variables are: sex, male (M) or female (F); total body length (TL); age class, calf (C) subadult (S), or adult (A); sample size for strength, modulus, and work of fracture (N); flexural strength (); Youngs modulus (E); work of fracture (WOF); number of toughness samples (N KC ); fracture toughness (K C ); density; and percent mineral content. indicates no data for that variable.

PAGE 64

52 Figure 2-1. Sample sites for biomechanical analyses. A) Ribcage is divided into three regions: cranial, middle, and caudal. One rib from each region was used for bending tests. B) Sample sites for bending tests within each rib are indicated. Total body length (cm) 50100150200250300350400 Flexural strength (MPa) 6080100120140160180 Females Males R2=.79 Figure 2-2. Flexural strength vs. total length. Strength increases with body length in the subadults and plateaus in the adults. There was no significant difference in strength between sexes (ANCOVA, p = 0.93).

PAGE 65

53 Total body length (cm) 100200300400 Toughness (MPam1/2) 1.21.41.61.82.02.22.42.62.83.0 Females Males R2=.73 Figure 2-3. Fracture toughness vs. total length. Fracture toughness peaks in the subadults and thereafter plateaus in the adults. There was no significant difference between sexes (ANCOVA, p = 0.81) Total body length (cm) 100200300400 Work of Fracture (MJ/m3) 2.53.03.54.04.55.05.56.06.5 Females Males R2=.50 Figure 2-4. Work of fracture vs. total length. Work of fracture peaked in the subadults, and thereafter declined with increasing body length. There was no significant difference between sexes (ANCOVA, p = 0.96).

PAGE 66

54 Total body length (cm) 100200300400 Modulus (GPa) 2468101214161820 Females Males R2=.86 Figure 2-5. Elastic modulus vs. total length. Elastic modulus increased with increasing body length. There was no significant difference between sexes (ANCOVA, p = 0.70). Modulus (GPa) 2468101214161820 Flexural strength (MPa) 6080100120140160180 R2=.67 Figure 2-6. Flexural strength vs. modulus. Flexural strength is significantly correlated to the elastic modulus (p<0.0001).

PAGE 67

55 Displacement (mm) 0.00.51.01.52.02.53.03.5 Load (N) 020406080100 Calf Subadult Subadult Adult 81 N14 E6.1 W49 N11 E2.2 W 75 N17 E2.4 W 59 N13 E10.3 W Figure 2-7. Load-displacement curves for small sample specimens tested in 3-point flexure. Values given for each curve are: ultimate load, N (Newtons); modulus, E (GPa); and work of fracture, W (MJ/m 3 ). Three age classes are represented: calves (MNW0207); subadults (MSW0160); and adults (MNW0208). Two curves for the subadult MSW0160 are shown. The red curve is an example of the high work of fracture obtained for some subadult specimens. The black curve is a more typical curve for a large subadult.

PAGE 68

56 Total Body Length (cm) 150200250300350 Density (g/cm3) 2.082.102.122.142.162.182.202.222.24 R2=.19 Figure 2-8. Density vs. total body length. There was no significant relationship between bone tissue density and body size (p = 0.0517). Total Body Length (cm) 150200250300350 % Mineral Content 636465666768697071 R2=.27 Figure 2-9. Percent mineral content vs. total body length. There was a significant increase in percent mineral content with total body length (p = 0.0198), although the correlation was weak.

PAGE 69

57 Density (g/cm3) 2.082.102.122.142.162.182.202.222.24 Modulus (GPa) 8101214161820 R2=.64 Figure 2-10. Modulus vs. density. With one outlier (shown in red) removed, modulus is significantly correlated to density (p<0.0001). % Mineral Content 64666870 Modulus (GPa) 8101214161820 R2=.61 Figure 2-11. Modulus vs. percent mineral content. Modulus is significantly correlated to mineral content (p<0.0001).

PAGE 70

58 Total body length (cm) 100200300400 Crack size (m) 1.0e-41.5e-42.0e-42.5e-43.0e-43.5e-44.0e-44.5e-45.0e-4 Figure 2-12. Small sample crack size vs. total body length. The neonate calf MEC0033 was an outlier (shown in red). With the outlier removed, there was no change in crack size with total body length (R 2 = 0.0009, p = 0.8886).

PAGE 71

CHAPTER 3 STRUCTURAL PROPERTIES OF MANATEE RIBS TESTED IN IMPACT Introduction The endangered Florida manatee shares its nearshore, coastal habitat with an ever increasing number of boats. Each year, 25% of all manatee deaths are a result of collisions with watercraft. Boat speed zones are the primary tool for ameliorating the lethal interactions. However, existing limits are not based explicitly on information pertaining to the nature or severity of the injury, including bone fractures. In an effort to establish safer boat speeds for manatees, the Florida manatee recovery plan calls for research on the mechanical characteristics of manatee bone to better understand the effects of watercraft-related impacts (U.S. Fish and Wildlife Service, 2000). This study is the first to estimate mechanical properties of whole manatee bone. Quantitative fractography is a technique which uses principles of fracture mechanics and fractal geometry to analyze fracture surfaces of brittle materials such as ceramics (Mecholsky, 1993, 2001). The strength of a material is determined by the size of the fracture initiating crack, or flaw, and the fracture toughness of the material. Location of the crack can be determined by examining the markings on the fracture surface. In similar fashion, the topography of a fracture surface is also a record of fracture parameters and is fractal in nature. The fractal dimensional increment (D*) is a quantitative measure of the tortuosity of a brittle fracture surface, and is related to parameters such as the fracture toughness and elastic modulus West et al., 1999). 59

PAGE 72

60 Two features of manatee bone allowed us to use quantitative fractography to measure its mechanical parameters. First, all its ribs and long bones are constructed of solid cortical bone; there are no marrow cavities or spongy bone. This feature, unusual for mammalian bone, is thought to be an adaptation to aid in hydrostasis (Domning and de Buffrenil, 1989). Second, manatee bone often undergoes brittle fracture in static tests (Chapter 2; Yan et al., 2005). Bone typically behaves as a quasi-brittle solid (Zioupos, 1998). However, fully mineralized bone from adult manatees does not undergo any post-yield plastic deformation. The goal of this study was to estimate the stress at failure and the fracture toughness of whole manatee ribs fractured in impact. Fracture toughness is the ability to resist fracture, one measure of which is the critical stress intensity factor (K C ). This is an estimate of the amount of energy required to propagate a macrocrack that leads to fracture (Mecholsky, 1996). Toughness is an important property because a tough bone is resistant to fracture, and this is most important when bone is impacted (Wainwright et al., 1982; Turner and Burr, 1993). Since adult manatee bone fails in a brittle manner when loaded statically, it is likely that fracture resistance measured by K C is approximately similar to the work of fracture as a measure of toughness in manatee bone loaded in impact. We also addressed the hypothesis that forces generated by most watercraft under normal operation are sufficient to inflict fatal skeletal injuries to manatees. Materials and Methods Bone Sample Collection Ribs from 20 animals of both sexes, in three age classes (i.e., calf, subadult, and adult) were obtained from the Florida Fish and Wildlife Conservation Commissions Marine Mammal Pathobiology Laboratory. By convention, total body length was used as

PAGE 73

61 a proxy for age, and individuals were assigned to age classes based on established size definitions (OShea et al., 1985). Bones were obtained from fresh carcasses at the time of necropsy. One rib from the mid-thoracic region of each animal was collected and stored frozen until testing. Bones were cleaned of soft tissues and tested at room temperature. Whole Rib Impact Testing Whole ribs were impacted using a compressed air gun at the Structures Laboratory in the Department of Coastal and Civil Engineering, at the University of Florida (Fig. 3-1). The barrel is 20 in length with a 4 inner diameter. The gun is equipped with a compressor that allows projectiles to be fired at a range of speeds, and two photocells and an oscilloscope to record the speed of the projectile as it exits the barrel. Ribs were positioned vertically in a fixture that was mounted on a steel frame, so that the lateral (convex) side of the rib faced the gun, and the concave side faced the frame. They were impacted perpendicular to their long axis, in the middle of the lateral side with a 2 x 4 x 4pine projectile. The leading end of the projectile was equipped with a hemispherical nosepiece coated in fiberglass. The mass of the projectile was 13 kg. The distance between the end of the barrel and the rib was 2.3 m. The projectile was fired between 23-28 m/s. Kinetic energy (KE) of the projectile was calculated as KE = mv 2 (3-1) where m is the mass of the projectile (kg), and v is velocity (m/s). Quantitative fractography was used to identify the fracture origin and measure crack size for each rib (Figure 3-2). The failure stress was calculated as 2/1C)c(YK (3-2)

PAGE 74

62 where K C fracture toughness (MPam 1/2 ), measured as the critical stress intensity factor. Y is a geometric factor based on crack shape and loading conditions, and c is the crack size (m 2 ). Values for K C were taken from the middle region of the adjacent rib, which was used to determine bone material properties in Chapter 2. The value used was the mean for the adjacent section. Crack size was calculated as c = (ab) 1/2 (3-3) where a and b are crack depth and half-width (m), respectively, as measured from the fracture surface with digital calipers to the nearest 0.1 mm. Strain Gauge Testing To validate the failure stress calculations, three additional ribs were outfitted with strain gages to directly measure strain during impact. For each rib, a uniaxial gauge (CEA-06-250UN-350, Vishay Micro-Measurements, Raleigh, NC) was placed on the lateral (convex) surface, and a uniaxial gauge and a rosette gauge (CEA-06-125UR-350) were placed on the pleural (concave) side. Gauges were placed along the midline, parallel to the long axis, either above or below the center of the rib. The center element of the rosette was oriented parallel to the long axis of the bone. Signals were output to an 8 channel signal conditioner and sampled at 50 kHz. Raw data were converted to microstrain, which were used to calculate the principal strains using standard formulae. Failure stress was calculated as = E (3-4) where E is Youngs modulus (GPa) measured with ultrasound (Nuson, Inc., Boalsburg, PA), and is the maximum principal strain. As a check of the strain data, an additional rib was outfitted with gauges and loaded to failure in quasi-static 3 point bending at a load rate of 4.448 N/s. The signal

PAGE 75

63 was sampled at 20 Hz. Failure stress was calculated using the standard formula for a curved beam in bending. For all gauged ribs, the fracture origin was identified fractographically, and K C was determined using Equation 3-2. Toughness values calculated for whole ribs were compared to those from the materials test specimens to assess the validity of using the materials values from Chapter 2 in the failure stress calculations. Fractal Dimensional Increment The fractal dimensional increment (D*) was calculated using a modified slit island technique (Hill et al., 2001). The technique involves making an epoxy replica of the fracture surface (Figure 3-3). The replica is ground down to the plane of fracture to produce fracture surface contours, or islands. The island perimeter, represented by a line, is measured with rulers of different lengths. A log-log graph of perimeter length versus the ruler length produces a line whose slope is related to the fractal dimension, D (i.e., slope = 1-D). The non-integer, or fractional portion, of the slope is equal to D*, which is a measure of the fracture surface tortuosity. The higher the value of D*, the more tortuous (rough) the surface. For example, a fractal dimension (D) of 2.08 (D* = .08) would be a surface which is relatively smooth, and a D of 2.40 (D* = .40) would be a rough surface, with much tortuosity. For each rib, a negative impression mold of the fracture surface was made in dental putty and latex, and then double cast in epoxy. The cast was made such that the fracture origin was horizontal. The fracture surface of the positive cast was sputter coated (Denton Desk II cold sputter coater, Denton Vacuum, Moorestown, NJ) with gold-palladium for contrast, and the second cast was poured. Replicas were ground down to the fracture origin on a grinding wheel using increasingly finer grit sanding papers, then

PAGE 76

64 polished with 0.3m alumina slurry. Composite images of the fracture surface contours were made from high resolution images (200X) taken with a digital camera mounted on a light microscope (Figure 3-4). D* was calculated from the composite images using the Materials Pro Analyzer software add-on for the Image Pro-Plus software (Media Cybernetics, Inc., Silver Spring, MD). Ideally we would like to be able to calculate whole bone fracture toughness from the measured fractal dimension. The two parameters are related in the equation K C = Ea 0 1/2 D* 1/2 (3-5) where E (GPa) is the elastic modulus, a 0 ()is a proportionality constant, and D* is the fractal dimensional increment. The variable a 0 is a material constant that has been proposed as a measure of the average intermolecular strained bond length (Mecholsky, 2001). It is not directly measurable, but must be calculated from other, independently measured parameters or estimated using quantum mechanics models. Although a 0 has been established for a number of materials, there is no experimentally determined value for cortical bone. Using our measured D* values, K C values calculated from the strain gage data and fractographic measurements, and the ultrasonic modulus, we estimated a 0 for manatee cortical bone. Kinetic Energy Calculations The most common recreational watercraft in operation in Florida waters is a 17 vessel (Wright et al., 1995) which is typically equipped with a single outboard engine of 100 HP. A typical 17 vessel at idle speed (i.e., no wake) is traveling at 2 mph (depending on vessel configuration), and 10 mph corresponds to a boat traveling at or near hull speed (i.e., on a plane). Maximum speed is approximately 35 mph. Weights for a few common 17 vessels with the recommended engines were taken from the

PAGE 77

65 specifications posted on manufacturers websites. Kinetic energy was calculated for speeds of 2, 10, and 35 mph. We assumed one adult operator (200 lbs.) and one halftank of fuel. The soft tissues overlying bone have been shown to absorb and dissipate impact energy, thereby decreasing the energy transferred to the skeleton (Currey, 1968; Nikoli et al., 1975; Parkkari et al., 1994, 1995; Robinovitch et al., 1995). To estimate the energy damping ability of the tissues overlying the ribs, soft tissue thickness (e.g., skin, blubber, muscle) was measured approximately one third of the way down from the head of the ribs on some whole body sections. Robinovitch et al. (1995) measured the amount of impact energy transferred to the femoral trochanter with overlying soft tissues of varying thickness. They regressed tissue energy absorption against soft tissue thickness to determine the threshold of tissue thickness needed to prevent femur fracture during falls in the elderly. We used their regression to estimate the amount of energy dissipated by manatee soft tissues. Kinetic energy curves for the boats were visually compared to the kinetic energy of the projectile used to impact the bare rib, as well as to the estimate for ribs with overlying soft tissues. Statistical Analyses Failure stress and D* were compared between sexes and age classes in separate one-way ANOVAs (Zar, 1996). Fractal dimension was regressed against total body length (a proxy for age) to examine the relationship of D* to age. Analyses were performed with SAS (SAS Institute, Inc., 1989). Results Whole Rib Impact Tests Mechanical and fractal dimension data from the impact tests are presented in Table 3-1 and Fig 3-4. A total of 22 ribs from 20 animals were tested. Only one calf was

PAGE 78

66 tested, as the small size of the ribs made them difficult to test. Of the other 19, eight were subadults (i.e., juveniles) and 11 were adults. The sexes were represented equally, ten females and ten males. Mean failure stress was 47 MPa, varying from 37 MPa. Mean adjusted failure stress (see below) was 139 21 MPa (range 106 MPa), and mean D* was 0.13 0.04 (range 0.08.22). Means by sex and age class are reported in Table 3-2. We tested 4 ribs from animal MNW0208, which was a large adult. Three of those were tested in impact and one in quasi-static 3 point flexure. Values for D* were .09, .13, .13, and .20 (Tables 3-1 and 3-3). The variability in these numbers is similar to that for the entire dataset, and thus, not unexpected. This variability may represent natural variation in the ribs, or it may be an artifact of the replication or grinding processes. There were no statistically significant differences between the sexes for either failure stress (p = 0.51) or D* (p = 0.18). Similarly, no differences were found between age classes for either variable (p = 0.19 and p = 0.63 for failure stress and D*, respectively). The lone calf was omitted from the analysis by age class. The regressions of D* and failure stress against total body length indicated no relationship between either parameter to body size (Figures 3-5 and 3-6, respectively). These two independent measures both imply that the ability to resist bone fracture in impact does not increase as animals grow. Strain Gauge Tests Of the three ribs outfitted with strain gages for the impact tests, data were successfully collected for only one (Table 3-3). The rib tested in 3 point bending failed at a load of 4480 N. Peak strains for the impact and static tests were very similar (0.56% and 0.61% respectively). Fracture toughness calculated for the impacted rib was 7.7 MPam 1/2 essentially identical to the 7.8 MPam 1/2 for the rib from the static test. These values were much greater than those used for the failure stress calculations reported in

PAGE 79

67 Table 3-1. Our results indicate that it is not appropriate to use the small specimen fracture toughness values of Chapter 2 to calculate whole bone failure stress. By doing so, our stresses were most likely underestimates. To adjust for this, we used 8 MPam 1/2 to derive more appropriate estimates of failure stress, reported as W in Table 3-1 and Figure 3-6. The material constant a 0 calculated for the two gauged ribs is reported in Table 3-3. The value for rib 14 is twice that for rib 6, 13,000 and 6,700 respectively. Values for classes of ceramics are much lower, typically 1-100 but our values lie in the range reported for various polymers, from 2,700-14,000 (West et al., 1999; Mecholsky, 2001). Much more work needs to be done to determine whether it is possible to estimate a 0 for a complex material such as cortical bone. Kinetic Energy Calculations Projectile velocity, as measured exiting the barrel, was 23-28 m/s. To make conservative estimates of the threshold for fracture we used the maximum projectile velocity (28 m/s) for calculations. Vessel configurations and kinetic energies are presented in Tables 3-4 and 3-5, and Figure 3-7. Based on the configurations selected, rib fracture can occur at speeds of 7 mph, or just below hull speed. This is speed at which boats generate kinetic energy equal to that of the projectile, as determined by the point at which the regression curves cross the kinetic energy line for the projectile. This estimate is for bare bone, which of course is not a realistic scenario. Human impact studies indicate that for some body sites, soft tissues absorb and dissipate as much as 90% of impact energy (Nikoli et al., 1975). The maximum thickness of manatee soft tissues (i.e., skin, blubber, and muscle) was 58 mm. Robinovitch et al. (1995) measured impact energy absorption for soft tissues up to 43 mm thickness. Using their regression, we extrapolated the amount of energy dissipated by the manatee soft tissues to be 70%. That

PAGE 80

68 is, only 30% of impact energy is transferred to the bone. This projection is plotted on Figure 3-7. At this energy absorption estimate, boat speeds of 13 miles per hour are sufficient to fracture manatee ribs (i.e., where the regression curves cross the with skin kinetic energy estimate, indicated by the dashed lines on Figure 3-5). Discussion Fractals are geometrical objects with fractional dimensions (Mecholsky, 1986). Mandelbrot was the first to recognize that natural structures and phenomena are fractal, and this includes fracture surfaces (Mandelbrot et al., 1984; Mecholsky et al., 1996). Fractal geometry provides a means of measuring the roughness, or tortuosity of a fracture surface. The fractal dimension (D) is a quantitative measure of the dimensionality of an irregular, or tortuous, surface. The fractional dimensional increment (D*, the fractional portion of the fractal dimension) is related to fracture parameters such as the fracture toughness and elastic modulus. Techniques were first applied to metals, and were subsequently applied to ceramics and some composites (Mandelbrot et al., 1984; Mecholsky et al., 1989; West et al., 1999; Czarnecki et al., 2001). Fractal geometry is now being used for trabecular bone as a means to quantify changes in bone quality and quantity (Majumdar et al., 1993). We have utilized the fractal dimension to quantify fracture parameters of manatee cortical bone using the same methods applied to structural materials. The Fractal Dimensional Increment Although use of fractal analysis now can be found widely in the engineering literature, there are few examples of its application to biological composites. Fractal dimension is becoming a popular tool for quantifying changes in trabecular bone architecture related to disease conditions such as osteoporosis. In this body of literature,

PAGE 81

69 its proposed use is singly as a diagnostic tool. Only two studies of biological composites have related fracture energy to fractal geometry. Currey et al. (1995) measured fractal dimension for cortical bone for a variety of species. Fractal dimensional increment was correlated to the work of fracture, from .05 for a highly mineralized, brittle fin whale tympanic bulla to .25 for very tough, young deer antler. Various types of cortical bone had intermediate fractal dimensions, which roughly correlated to their calculated works of fracture. Hill (2001) measured fractal dimensional increments for the conch shell of Strombus gigas. Means ranged from .16.22 for specimens stored in different media. Our measured D* values for manatee cortical bone ranged from .08.22, which is in general agreement with the other two studies. However, it is difficult to make detailed comparisons among studies because, although the same theory was used to calculate fractal dimension, different measurement methods were used. Additionally, biological composites, like some other materials, may be multi-fractal, making results sensitive to measurement length scale. Currey et al. (1995) found this to be true for bone with high work of fracture, fractal dimension decreasing with increased length scale in these tough materials. A wide range of D* values exists for classes of materials (e.g., metals, polycrystalline ceramics, and glasses). However, since most of the data available in the literature are for essentially monolithic materials, we could only make a cursory comparison to our data. Aside from the two references to bone and shell, the most appropriate materials to compare to our data are the cement-epoxy composites. Czarnecki et al. (2001) measured fractal dimensional increments of .02.08 for seven epoxy concretes. However, they used the vertical profile technique to calculate fractal

PAGE 82

70 dimension, which is known to produce values consistently lower than those of the slit island technique (Hill et al., 2001). Fractal Dimensional Increment and Fracture Toughness Because of the large amount of variation in fractal dimension both within and among classes of materials, comparing D* values alone is not informative. It is more instructive to examine the relationship between the fractal dimension and fracture toughness. For most materials that fail in a brittle manner, there is a positive correlation between fracture toughness (K C ) and D* 1/2 and the slit island technique has been shown to produce a D* that is a more accurate predictor of K C than other techniques (Russ, 1994; Hill et al., 2001; Mecholsky, 2001). Over the range of D* values for manatee bone, fracture toughness only varied by 1 MPam 1/2 which resulted in no significant relationship between fractal dimension and fracture toughness (R 2 = 0.02, p = 0.5535, Figure 3-8). There are a few possible explanations for the lack of correlation. It could be that there is no relationship between the parameters for cortical bone, or that the range of fracture toughness values was too narrow to detect a trend. A third possibility is that error was introduced into our technique. Calculation of the fractal dimensional increment using the slit island technique is sensitive to error in the polishing angle. Della Bona et al. (2001) showed that variation of 5 or greater from the fracture plane significantly underestimates D*. There may have been angle errors introduced in the production and grinding of the replicas that would account for some of the low values reported here. However, our methodology was consistent throughout, so this would not explain the non-significant relationship between fractal dimension and fracture toughness. We hypothesize that the lack of correlation between fractal dimension and fracture toughness reflects the true relationship between these properties for manatee

PAGE 83

71 bone. Critical crack size for the whole ribs was similar for all animals regardless of total body length (R 2 = 0.10, p = 0.1538, Figure 3-9) or cross sectional area of ribs. In Chapter 2 we reported that density and mineral content also did not vary significantly over the range of body sizes. If these values accurately reflect the material properties of rib bone, then our finding of non-significant variation in critical crack size and failure stress among animals is not anomalous. Despite the increase in rib size (340 mm 2 cross-sectional area) with increasing body size in manatees, there is no concomitant increase in ability to resist crack formation because the material is essentially the same. R-curve Behavior and Viscoelasticity in Bone The greater fracture toughness values calculated from the strain data relative to the small specimen values of Chapter 2 may be indicative of R-curve behavior in manatee bone. A rising R-curve describes an increase in fracture toughness with the increase in crack extension (Vashishth, 2004). This behavior is characteristic of quasi-brittle materials such as some ceramics, and has now been documented in human, bovine, equine, and deer antler bone (Les et al., 2002; Malik et al., 2003; Nalla et al., 2004; Vashishth et al., 1997). If manatee bone exhibits R-curve behavior, then it is not appropriate to use toughness values from small sample material properties tests to calculate failure stresses of whole bones. Those values, measured from machined specimens tested in 3-point bending, ranged from 2.0.0 MPam 1/2 True brittle materials exhibit flat R-curve behavior (Vashishth et al., 1997; Zioupos, 1998). The material tests in Chapter 2 showed that adult manatee bone tested in static 3 point bending undergoes very little or no plastic deformation; it typically fails catastrophically near the yield point. This seemingly contradicts the toughness data presented here that suggests R-curve behavior. However, Vashishth et al. (1997) point

PAGE 84

72 out that some ceramics do produce rising R-Curves. Like all bone, manatee bone is a ceramic-polymer composite of complex, hierarchical structure. Bone microstructure, mineral density, and collagen content all affect the R-curve behavior (Les et al., 2002; Malik et al., 2003; Nalla et al., 2004; Vashishth et al., 1997). Malik et al. (2003) reported lower R-curves in bone of increased stiffness, tested in tension. Nalla et al. (2004) reported an age-related decrease in both crack initiation and crack growth resistance. The high mineral content of manatee bone (mean 69%) likely has a similar effect. Appropriate testing needs to be done to characterize R-curve behavior in manatee bone. The R-curve behavior of bone is due to its viscoelastic nature; its collagen component is partly responsible for the toughening mechanism of crack bridging. A larger sample allows for a greater crack size, and more collagen is available to contribute to the toughening, resulting in a larger fracture toughness. Therefore, absolute size most likely explains the greater fracture toughness of the whole ribs over the machined specimens. For some materials, including most bone, fracture toughness is strain rate dependent. As strain rate increases, so does strength, modulus, and K C ; a feature also attributable to viscoelasticity (Zioupos, 1998). However, note that work of fracture (WOF) will decrease with increased strain rate, also due to the viscoelastic nature of bone. The different behavior of K C versus WOF with strain rate is due to the method of measurement of both. Because we tested the whole bones and the small sample specimens at different rates, the effects of strain rate and specimen size were confounded in this study. However, in ceramics, fracture toughness is not strain rate dependent, as long as it is not strongly sensitive to stress corrosion processes. The impact tests provide

PAGE 85

73 additional evidence that manatee bone behaves as a ceramic; the whole rib tested in quasi-static 3 point flexure had a crack size identical to the other ribs tested in impact. Impact Energy Absorption of Soft Tissues Impact testing of human bone is often done to determine the loads at which bones, such as the femur, break. The whole rib manatee tested in static 3 point bending failed at a load of 4480 N, which is only moderately greater than the critical trochanteric impact load of 4170 N, the average load required to fracture in impact the proximal femur in the elderly (Parkkari et al., 1995). Soft tissues overlying the bone absorb impact energy, thereby reducing the load transferred to the bone. Energy absorbed by soft tissues has been shown to be as high as 90%. In femoral impact tests conducted by Robinovitch et al. (1995), energy absorption varied from 6%. As tissue thickness increases, energy absorption increases, and the peak load transferred to the bone decreases. The authors found that for each millimeter of soft tissue thickness, peak load transferred to the bone decreased by 70 N, and soft tissue energy absorption increased by 1.7 J. In a different study, Parkkari et al. (1995) impacted artificial hips protected by soft tissues at high and low impact rates. They found that, regardless of impact rate, 80% of the total load was transferred to the bone. In these cited two studies, impact energies of 132 J and 140 J were sufficient to generate loads capable of creating fractures. Our projectile generated 5 kJ, so we likely could have created fractures in manatee ribs with much less force. Taking the soft tissues into consideration, our calculations indicated that a kinetic energy of 17 kJ is required to create rib fractures in manatees. Based on the regression of Robinovitch et al., (1995), 12 kJ of the 17 kJ (70%) would be dissipated by the soft tissues, and the remaining 5 kJ (30%) transferred to the skeleton. For a typical 17 vessel, this translates to speeds of 13 mph. We needed to make a few assumptions to generate this estimate.

PAGE 86

74 While our weights for the vessels are accurate, the total weight would likely be greater, as we did not include items such as the battery, safety equipment, coolers, other equipment, or additional passengers. Greater vessel weights would shift the curves up, effectively lowering the vessel speed able to create fractures. More importantly, it may not be valid to use the energysoft tissue thickness regression based on human tissues for manatees. The dermis of manatee skin is different from that of humans, and from that other marine mammals (Kipps et al., 2002), and likely has different energy-damping capabilities. Additionally, manatees have a blubber layer. Blubber is different from fat in terrestrial animals. To more accurately determine the position of the with skin line in Figure 3-7, the compliance of manatee skin and blubber needs to be determined, and impact testing of ribs with the associated soft tissues in place needs to be performed. Summary The goal of this study was twofold: to estimate the stress at failure and the fracture toughness of whole manatee ribs fractured in impact, and to determine if the typical watercraft is able to generate enough force to break manatee ribs upon impact. The unique construction of manatee ribs enabled us to apply fractal geometry techniques to measure some fracture mechanics parameters. Adult manatee bone behaves more like a ceramic than other types of bone. Because of this, we were able to see many of the features observed for brittle fracture in ceramics. We were able to identify crack origins, and make quantitative measurements of crack size. Failure stress was constant across body size, indicating that the amount of force needed to break ribs does not change as the animals grow, despite the increase in rib size. There was no change in the amount of kinetic energy needed to fracture whole bone with total body length, measured as the fractal dimension (D*). The whole rib toughness and measured D* independently

PAGE 87

75 corroborate the finding that flaw size is the same for all animals, regardless of body size. Just as for the material properties tests of Chapter 2, we found no differences between the sexes in their ability to resist fracture. Finally, we found that the boats typically found in Florida waters can easily generate enough impact force to break manatee ribs.

PAGE 88

76 Table 3-1. Mechanical and fractal data for 20 manatee ribs fractured in impact Animal Sex TL (cm) Age D* c (m 2 ) K C (MPam 1/2 ) M (MPa) W (MPa) MSTM0109 M 166 C .13 1.9E-03 2.5 46 142 MNW0207 F 188 S .09 1.9E-03 2.4 45 145 MSW0227 F 190 S .09 2.5E-03 2.3 37 125 MSTM0113 M 196 S .17 1.5E-03 2.0 41 159 MSW0057 M 201 S .14 1.5E-03 2.2 46 163 MSE0205 M 238 S .22 1.8E-03 2.5 48 149 MSW0160 F 246 S .17 1.8E-03 2.7 52 148 MSTM0102 F 265 S .13 1.5E-03 2.5 51 159 MSW0165 F 267 S .11 2.7E-03 2.5 45 120 MSW0251 M 276 A .12 2.4E-03 3.0 50 128 MSW0225 F 282 A .12 2.4E-03 3.0 49 127 MSW0223 F 294 A .14 2.2E-03 2.7 46 133 MSW0208 F 299 A .17 2.7E-03 2.8 43 120 MEC0213 F 314 A .08 1.5E-03 2.7 61 176 MSW0226 M 315 A .11 2.8E-03 2.7 41 118 MSW0239 M 321 A .17 1.6E-03 2.5 56 175 MSW0253 M 327 A .11 2.3E-03 2.7 46 130 MSTM0414 F 335 A .12 2.0E-03 2.8 50 138 MNW0208#6 M 343 A .20 2.3E-03 2.9 48 129 MNW0208#9 M 343 A .09 1.5E-03 2.9 67 179 MNW0208#10 M 343 A .13 3.5E-03 2.9 40 106 MSW0245 M 344 A .15 2.2E-03 2.4 41 133 Variables are: sex, male (M) or female (F); total body length (TL); age classes: calf (C), subadult (S), or adult (A); fractal dimensional increment (D*); crack size (c); fracture toughness from small specimens (K C ); and failure stresses ( M calculated with fracture toughness data from machined, small specimen tests of Chapter 2, and W calculated from estimated whole bone fracture toughness). Three ribs were tested for animal MNW0208 (rib number indicated). Table 3-2. Mean mechanical and fractal data by sex and age clas s Group M (MPa) W (MPa) D* Sexes (ages pooled) Females (10) 48 6 139 18 .12 .03 Males (12) 46 8 143 23 .15 .04 Ages classes (sexes pooled) Subadults (8) 48 5 146 16 .14 .05 Adults (13) 49 8 137 24 .13 .03 Variables are failure stress ( M ), adjusted failure stress ( W ), and fractal dimensional increment (D*). Values are mean S.D. Sample sizes are reported in parentheses for each group. No groups were significantly different (p>0.05).

PAGE 89

77 Table 3-3. Mechanical data for strain gauged ribs Peak Modulus Stress c Toughness a 0 Rib # microstrain (GPa) (MPa) (m 2 ) (MPam 1/2 ) D* () 6 6075 21 128 2.3E-03 7.7 .20 0.7E-06 14 5562 19 106 3.5E-03 7.8 .13 1.3E-06 Ribs were from animal MNW0208. Rib #6 was tested in impact, rib #14 in quasi-static 3 point bending. Youngs modulus was measured with ultrasound. Failure stress was calculated as = E. Crack size, c (m 2 ), as measured from the fracture surface. Fracture toughness calculated from Equation 3-3. Constant a 0 was calculated from Equation 3-5. Table 3-4. Boat data for kinetic energy calculations Weight Tank Fuel Engine Boater Total Make Length (kg) (gal) (kg) HP (kg) (kg) (kg) Maverick 1 17'0" 684 40 53 115 162 91 989 Mako 2 17'3" 567 37 49 115 162 91 869 Key West 3 17'6" 454 40 53 100 161 91 758 Boat weights were taken from manufacturers specifications posted on their official websites. Fuel weight is for one half tank. Total weight is equal to the sum of weights for the vessel, fuel, engine, and one passenger. 1Maverick Boat Company, Inc, 2004. Model: 17 Master Angler 2Mako Marine International, Inc. 2005. Model: 171 Center Console 3Key West Boats, Inc., 2005. Model: 1760 Stealth Table 3-5. Kinetic energy calculations for vessel configurations and impact projectile Kinetic Kinetic Energy (kJ) Manatee Energy (kJ) Boat 2mph 10mph 35mph Bare bone 5 Maverick 0.40 10 121 With skin 17 Mako 0.35 9 106 Key West 0.30 8 93 Energy for bare bone is the kinetic energy of the projectile. With skin is the minimum amount of kinetic energy needed to create a fracture based on the calculated 70% energy absorption by manatee soft tissues.

PAGE 90

78 Pressure gauge Trigger Compressor Photo cells Barrel A B C ribs. A) Compressed air gun the e. Rosette gauge Uniaxial gauge Fixture Steel frame Uniaxial gau ge D Figure 3-1. Setup for impact tests of whole manatee assembly. The barrel is a 20 PVC pipe with a 4 diameter. The end ofbarrel is equipped with 2 photocells to measure the velocity of the projectileB) The gun is equipped with a compressor, pressure gauge, and trigger. C) Manatee rib positioned in test fixture, mounted on frame for impact testing.The rib is outfitted with a uniaxial strain gauge on the convex (lateral) surfacD) View of the concave (pleural, or medial) side. This side is outfitted with a uniaxial and a rosette strain gauge. The intended area of impact is between these two gauges.

PAGE 91

79 A B Figure 3-2. Fracture surface of a manatee rib fractured by collision with a boat. A) Examination of fracture surface indicated that the crack that led to failure (dashed line) originated at a groove on the medial surface (arrow). B) Schematic of fracture surface features used for fractography. Crack size (c) for a semi-elliptical surface crack is calculated by measuring crack depth, a, and half the crack width, b (i.e., (2b)).

PAGE 92

80 A B C E D Figure 3-3. Procedure for making epoxy replicas of manatee rib fracture surfaces. A) A negative impression of the fracture surface is made in dental putty and latex. B) A positive mold is made from epoxy using a PVC ring. C) The fracture surface of the positive epoxy cast is sputter coated with gold-palladium for contrast. The second layer of epoxy is then poured. D) The completed cast is ground down to the area of interest on a grinding wheel using sanding papers of increasingly fine grits. The cast is polished with an alumina slurry. E) The fracture contours, or islands (black arrows), are photographed at high magnification (200X). The composite photographs are used for fractal analysis.

PAGE 93

81 Figure 3-4. Example of a fracture surface contour used for calculation of fractal dimensional increment (D*). Image is a composite of photos taken at 200X. The computer traces the island contours (red line), and measures its length with different size rulers to calculate D*.

PAGE 94

82 significant differences between sexes or age classes. Figure 3-6was no relationship between adjusted failure stress (W) (R = 0.012, p = 0.5535). Total body length (cm) 50100150200250300350400 D* 0.060.080.100.120.140.160.180.200.220.240.26 Male calf Female subadult Male subadult Female adult Male adult Figure 3-5. Fractal dimensional increment (D*) vs. total body length. There was no relationship between D* and animal size (R2 = 0.002). Thus, the ability to resist bone fracture does not increase as animals grow. There were no Failure stress vs. total body length. There 2 Total body length (cm) 50100150200250300350400 Failure stress (MPa) 80100120140160180200

PAGE 95

83 Kinetic energy as a function of boat speed. Three ves Figure 3-7.sel configurations are presented. Calculated kinetic energy of the projectile is plotted, along with an energy estimate with overlying soft tissues in place. With an estimated 70% of the total kinetic energy being dissipated by the soft tissues, minimum boat speeds capable of producing rib fracture are 13 mph (dashed lines). Boat speed (mph) 0510152025303540 Kinetic energy (kJ) 020406080100120140 Projectile With skin Key West Mako Maverick

PAGE 96

84 Figure 3-8. Fracture toughness (KC) vs. fractal dimensional increment (D*). There was no lationship between the two variables (R2 = 0.0226, p = 0.5047). Figure 3-9.e ze. D*1/2 0.250.300.350.400.450.50 Kc (MPam1/2) 1.82.02.22.42.62.83.03.2 re Whole rib crack size vs. total body length. There was no relationship between flaw size and total body length (R2 = 0.10, p = 0.1538). Flaw size is the samregardless of rib cross sectional area, which increases with total body si Total body length (cm) 50100150200250300350400 Crack size (m) 5.0e-41.0e-31.5e-32.0e-32.5e-33.0e-33.5e-34.0e-3

PAGE 97

CHAPTER 4 CONCLUSIONS AND FUTURE RESEARCH Summary and Conclusions Based on the work presented here, the following conclusions can be drawn: Measured as a material property, the strength of manatee bone is generally lower than that of other mammalian bone. This is likely due to three factors: its high mineral content (67%) and density (mean 2.17 g/cm3); its microstructure; and its porosity. Strength is known to increase with increasing mineral content between 63% and 68%. may be the case for manatee bone. Most other mammalian bone does not reach these large values of mineral content. We did not address microstructure and porosity in this study. Manatee bone, as a material, is less tough than other mammalian bone. We calculated two measures of toughness, fracture toughness (KC) and work of fracture (WOF). The first, KC, is calculated from the ultimate strength and crack size. Because KC mirrored the flexural strength curve so closely, this implies that crack size varied little over the range of animals tested here. This is significant because for many ceramics, strength is usually crack size dependent. However, KC is a measure of the behavior in the elastic, or pre-yield region only. Because of this limitation, more than one measure is propaga 1 When mineral content is above 68%, static strength decreases; we contend that this 2 needed to fully characterize toughness. Work of fracture is the energy needed to te a crack, measured as the area under the entire stress-strain curve. Because 85

PAGE 98

86 toughness is largely determined by the post-yield behavior, the decline in work of fracture seen in the adults may be due to age-related changes of the collagen matrix. 3. The age-related patterns in mechanical varilated to the amount of bone material prerowth is by deposition of plexiform bof woven bone is laid down firs bone 4. 5. ole ribs (i.e., structural tests) indicated that adult manatees are not able to a as e rib ables appear to be re sent. Manatees grow quickly; appositional bone g ne. In general, a scaffolding o t, then later the spaces are filled in by lamellar bone. So in young animals the is porous, but it becomes less so as they grow. This likely accounts for the ontogenetic patterns observed here, as an increase in bone volume results in an increase in strength. There were no significant differences between the sexes for any mechanical variable. Marmontel (1993) described possible differences in bone resorption between the sexes, which we would expect to see reflected in the mechanical properties. However,we found no differences between the sexes. Tests of wh better absorb and dissipate impact energy to prevent bone fractures than smaller animals. There was no change in the energy needed to fracture whole bone with total body length, measured as the fractal dimension (D*). Additionally, there was no relationship between the energy needed to initiate a crack (K C ) and D*. This last result is reasonable, since we found no significant change in mineral content or density with body size. Despite the fact that the ribs increase in cross sectional arethey grow, the tissue properties remain relatively constant. Regardless of rib size, it takes about the same amount of energy to create and propagate a crack. The whol

PAGE 99

87 toughness and measured D* independently corroborate the finding that flaw size isthe same for all animals, regardless of body size. re 7. ittle nts the first attempt to quantify the biomechanical effects of boat man momannt the biology of the animal to devise anabonip of mt the casacro prowith calcium content of 230 mg/g or higher, bending strength may be either high or low. 6. The typical watercraft found in Florida waters is able to generate enough kinetic energy to create fatal rib fractures in manatees. Vessel speeds of 13 mph weestimated to be sufficient to create fractures. We successfully applied fractographic techniques to calculate fracture toughness for manatee rib bone. We were able to see many of the features observed for brfracture in ceramics. Our study represe strikes on manatees; this information contributes significantly to our understanding of atee-boat interactions. We think these data will be useful in shifting the focus to are objective approach for revising boat speed regulatory zones. It is our hope that agers and policy makers will take into accou speed zones that are adequate to minimize the chance of fatal impacts. Further Research It became clear in the course of this research that a detailed histomorphometric lysis of manatee bone is needed to elucidate the relationship between manatee rib e architecture and its material properties. In most mammalian bone, the relationshaterial properties to age is largely dictated by bone mineral content. This was noe for manatee bone, as mineral content and bone tissue density varied only slightly ss the range of animals tested, from a newborn calf to a 25 year old animal. Materialperties can vary widely at high mineral contents. Currey (1999) found that in bone

PAGE 100

88 Other factors that contribute to material properties to a lesser degree in other mals may have greater significance for manatees mam. For example, modulus (i.e., Curvarin bending, he found that porosity had one typr) differ in mechanical properties, but even the orientation of fibers wye ry having a region of un-remodeled endochondral bone. the manner in which plexiform bone is he stiffness) is influenced by porosity and bone architecture as well as mineral content. rey (1999) found that porosity and calcium content together explained 60% of the iability in modulus. However, for bone tested no observable effect on strength (Currey, 1998). Bone architecture is known to affect mechanical properties. Not only do the bes (e.g. woven, lamella ithin lamellae will affect the mechanical competency of the whole tissue (Martin et al., 1998). Fawcett (1942b) described eccentric rings, or layers, visible to the naked eon the lateral surface of the ribs. We too have observed these markings in our preliminahistological observations. Microscopic examination revealed that these rings mark theboundaries between layers among which the osteons are oriented obliquely. Fawcett (1942b) also described what he considered to be adult manatee bone as This area, along with deposited, may result in a degree of porosity that persists regardless of age. It islikely that manatee bone architecture and porosity affect its mechanical behavior in a unique way; histology will bear this out. The issue of manatee bone toughness needs to be examined further. We measured fracture toughness (K C ), both for the small sample specimens and the whole ribs from timpact tests. Yan (2005) also calculated fracture toughness for small sample specimens, but his samples were notched. As a result, his crack sizes were intermediate to ours. Correspondingly, his fracture toughness values of 4.1.5 MPam1/2 were intermediate to

PAGE 101

89 our val8 is is ield regions. Further analysis of our data mi the leton. If we are correct, our kinetic energy threshoe ues for the small samples (1.4.9 MPam1/2) and the estimate for whole ribs (MPam1/2). These data indicate that fracture toughness is dependent on crack size; thevidence for R-curve behavior in manatee bone. Our investigation of fracture needs to be re-examined to determine the relative contributions of the preand post-yield regions to the total work. The factors influencingmechanical properties differ in the preand post-y ght shed light on the relative importance of these factors for manatee bone. Yan (2005) calculated the J integral for manatee bone. The J integral is a measure of the energy used to propagate a crack; it is similar to the work of fracture. He found that, on average, 4 to 5 times more energy was consumed in the post-yield region than in the pre-yield region. This finding indicates that there are factors in addition to mineral content as having an important role in the mechanical behavior of manatee bone. The whole bone impact tests of Chapter 3 were only the first step in modelingeffects of boat impacts on manatees. This study originally proposed measuring the compliance of the skin and soft tissues overlying the ribcage, as well as impacting large body sections to directly determine the amount of kinetic energy needed to fracture ribs. We believe that the composition of manatee skin allows a significant portion of the impact energy to be transferred to the ske ld is an overestimate. It is likely that vessel speeds below hull speed are sufficient to create fatal bone fractures. Unfortunately, time constraints did not allow for these tests to be completed. This information is vital to accurately determining how much impact energy is transferred to the manatee, and how much of that is in turn transferred to thskeleton.

PAGE 102

90 Additional Avenues of Research In the course of this research, three additional avenues of research possibly relevant to manatee bone have emerged: the influ ence of LRP5, VEGF, and cathepsin K on manto or, o under the same appliedtain in o submergence of air breathing animals. Sickenberg (1931) proposed a mechanism for development of pachyostosis, hypothesizing that it arose as a result of atee bone. The purpose here is not to examine these topics in depth, but simply put forth the ideas. As mentioned in Chapter 1, the gene for LRP5, a low density lipoprotein receptplays a role in bone development. The LRP5 pathway is thought to play a role in mechanosensation in bone (Johnson et al., 2002). A single point mutation is known tresult in high bone density in humans (Johnson et al., 2002). The mutation results in the hyperfunctioning of bone building cells, or osteoblasts. The high bone mass mutation in LRP5 is thought to produce bone that is overadapted to the loads being applied. Bone with this mutation experiences strains 40% higher than wild-type bone load. In other words, it takes less of a load to elicit the same response in animals with this mutation. This may be the mechanism by which manatees are able to mainhigh bone mass in non weight bearing bones. If this is the case with manatee bone, the exact mutation of the LRP5 gene, and the mechanism of altered function needs to be determined. In contrast, the sclerosing conditions, which also superficially resemble manatee bone, are all a result of osteoclast malfunctions. Again, it is important to determine which cell type has altered function if we are to elucidate the mechanism by which manatees are able to maintain high bone mass and density in the aquatic environment. Nopsca (1923) speculated that pachyostosis is a neotenous condition that arose response t

PAGE 103

91 thyroid and/or pituitary dysentry inf that ar endothelial growth factor, or VEGF, may provide a mechanism by which t e re as f manatee bone. Cathepsin K deficiency may be responsible for the mechan function caused by chronic oxygen deficiency as a result of to the aquatic environment. This idea had led to the generally accepted beliemanatees are hypothyroid, although there is no direct evidence for this. It may be possiblethat hypoxia can modify bone formation directly, without the influence of an altered metabolism. Vascul his can occur. VEGF is expressed by many types of cells, including osteoblasts. VEGF production is stimulated by hypoxic conditions that arise during development, and as a result of bone fracture (Steinbrech et al., 1999; Warren et al., 2001). Maes et al. (2005) showed that overexpression of one isoform resulted in a thickened perichondriumand excessive bone formation within the diaphysis of the long bones. Cathepsin K is a cysteine protease that plays a role in osteoclast-mediated collagen degradation (Meier et al., 2005). In mice, deficiency of Cathepsin K results in pycnodysostosis, a form of osteosclerosis, a condition which manatee bone resembles, at least superficially (Li et al., 2005a). In Cathepsin K deficient mice, femoral cortical bonearea is increased by 30%, although mechanical properties such as ultimate load to failurand stiffness are not different from wild type animals (Li et al., 2005b). However, the bones were 60% more brittle as measured by post-yield deflection, and work to fractuwas reduced by 40% overall. These bones had an altered matrix construction that wpresumed to be the cause of the altered mechanical properties. Viewed under polarized light, the architecture resembles that observed in some preliminary histological observations o ical properties of manatee rib bone.

PAGE 104

APPENDIX A DATA TABLES Table A-1. Mortality information for animals used in this study Field ID Sex Size (cm) Date County Region Probable cause of death MEC0033 F 102 05/06/2000 Brevard Banana River Perinatal MSTM0109 M 166 02/04/2002 Monroe Lake Surprise Watercraft MSW0033 M 178 03/11/2000 Lee Caloosahatchee Undetermined, MSW0164 M 182 05/18/2001 Collier Naples Bay Cold stress MSW0227 F 190 02/20/2002 Lee Orange River Cold stress MSW0057 M 201 05/06/2000 Charlotte Gasparilla SounMSW0134 F 205 02/12/2001 Collier Faka Union Canal River decomposed MNW0207 F 188 03/03/2002 Pinellas Boca Ciega Bay Cold stress SWFTM0114 M 196 05/13/2001 Brevard Barge Canal Watercraft d Watercraft Cold stress MSE02ft MEC0220 M 275 03/17/2002 Indian Indian River Watercraft MSW0251 M 276 03/24/2002 Sarasota Forked Creek Natural MSW0223 F 294 02/10/2002 Lee Pine Island Sound Watercraft SWFTM0414b F 335 05/26/2004 St. Johns Matanzas River Natural l MSW0245 M 344 03/22/2002 Sarasota Blackburn Bay Natural 05 M 238 01/26/2002 Palm Beach Loxahatchee River WatercraMSW0160 F 246 05/14/2001 Collier North Rookery Bay Channel Watercraft MSTM0102 F 265 02/12/2001 Collier Halfway Creek Watercraft MSW0165 F 267 05/19/2001 DeSoto Peace River Human, other River MSW0225 F 282 02/18/2002 Lee Ten Mile Canal Natural MSW0208 F 299 01/15/2002 Lee Orange River Natural MEC0213 F 314 02/03/2002 Indian River Indian River Watercraft MSW0226 M 315 02/20/2002 Lee Caloosahatchee River Watercraft MSW0239 M 321 03/15/2002 Charlotte Stump Pass Natural MSW0331 F 326 02/27/2003 Collier Faka Union Canal Natural MSW0253 M 327 03/25/2002 Sarasota Lemon Bay Natural MNW0208 M 343 03/10/2002 Citrus Kings Bay Natura Source: Florida Fish and Wildlife Institute, 2005. 92

PAGE 105

93 Table A-2. Means for four mechanical properties by body region Animal Site Mechanical variable Cranial Rib position Central Caudal MEC0033 Proximal Middle S (MPa) E (GPa) W (MJm) WK1) 1, 1, 1 87 (n/a, 1) 4.8 (n/a, 1) 4.0 (n/a, 1) (n/a, 1) 1) ) 59 (n/a, 1) 3.1 (n/a, 1) 2.2 (n/a, 1) 1.2 (n/a, 1) ) -3 K C (MPam1/2) 1.6 SE C 66 (n/a, 3.6 (n/a ) 3.1 (n/a, 2.3 (n/a (n/a ) 4.0 (n/a, 1 1.3 ) 1.4 (n/a, 1) 66 (n/a, 1) 32 (n/a, 1) 1) 1.1 (n/a, 1.3 (n/a, 1 0.9 (n/a, 1) MSTM0109 Pma Middle SEWKSEWK9, 88.7, 88 12) ) 10) ) ) 0) roxi l C C 130 (1 ) 138 (15 12.6 (1.7, 2 (0 ) 12.9 (0.7, 12)2 3. ) 4.5 (1.4, 1 2.2 (0.3, ) 2.5 (0.2, 132 (13, 10 ) 12.9 (0.7, 10 3.6 (0.7, 10 2.3 (0.2, 1 MSW0164 Pima Middle SKCSEWK8, 9) 9, 9.2, 9 ) ) 11) 10) ) rox l EW C 96 (1 8.6 (0.9.0 ) 9.8 (1.1,, 3.2 (1 2.1 (0 ) ) 4.2 (1.0 2.3 (0.2, 111 (13, 11 11 106 (17, 5 9.7 (1.4, 5) 5) 4.9 (1.5, 2.1 (0.2, 5) MSW0033 Pma Middle SEWKSEW 10, 13, 1, 9 ) 12) ) ) 9) 8) roxi l C K C 125 (9, ) 11.7 (0.7 0) 4.9 (1 1.9 (0.3 0) ) 132 (15, 12) 11.9 (1.3, 12 5.3 (1.3 2.1 (0.2, 12) 153 (24, 912.6 (1.0, 9 8.0 (3.4, 2.3 (0.4, Strength, S; Youngs modulus, E; work of fracture, W ; and fracture toughness, KC. Standard deviation and sample size are reported for each value in parentheses, respectively. indicates no samples from that body region.

PAGE 106

94 Table A-2. Continued Mechanical Animal Site variable Cranial Rib position Central Caudal MNW0207 133, 9) 12.3 (9) 4.9 (9) 2.4 (9) 1 Proximal Middle S (MPa) E (GPa) W (MJm-3) K C (MPam1/2) S E W K C (13 0.8, 0.9, 0.3, 123 (33, 4) 11.5 (1.0, 4) 4.7 (2.9, 4) 2.3 (0.5,4) 128 (12, 12) 1.5 (1.1, 12) 4.8 (1.0, 12) 2.5 (0.2, 10) 139 (17, 3) 12.9 (0.5, 3) 6.1 (1.0, 3) 2.3 (0.1, 2) 118 (11, 9) 10.5 (1.0, 9) 3.9 (0.7, 9) 2.3 (0.2, 9) MSW0227 Proximal Middle S E W K C S E W K C 116 (24, 9) 12.1 (1.7, 9) 2.9 (1.3, 9) 2.3 (0.2, 9) 145 (16, 12) 14.0 (1.4, 12) 5.1 (1.7, 12) 2.6 (0.2, 10) 111 12) 11.1 (2, 12) 3.2 (1, 12) 2.4 (0.4, 10) 11210) 11.4 (0.7, 10) 3.1 (0.5, 10) 2.3 (0.2, 10) (25, .7.1 145 (21, 5) 13.2 (0.6, 5) 8.3 (2.2, 5) 2.6 (0.3, 5) (7, SWFTM011 4 1 1 106 23) 9.4 (1.0, 23) 4.5 (1, 23) 2.1 (0, 14) Proximal Middle S E W K C S E W K C 20 (24, 13) 1.9 (1.9, 13) 3.3 (1.3, 13) 2.2 (0.3, 10) 140 (18, 19) 12.7 (1.7, 19) 5.7 (1.8, 19) 2.5 (0.3, 19) (10, .0 .3 125 (14, 11) 11.6 (1.2, 11) 4.6 (1.1, 11) 2.4 (0.3, 10) MSW0057 Proximal Middle S E W K C S E W K C 155 9.8 (.8, 9) 4.4 (.0, 9) 2.9 (.3, 5) 149 6) 9.7 (6) 5.4 (6) 2.4 (4) 151, 3) 9.1 (.2, 3) 5.9 (.9, 3) 2.3 (0.2, 3) (27, 9) 11 0 154 (30, 10) 9.6 (1.7, 10) 5.7 (1.7, 10) 2.8 (0.5, 5) 162 (16, 6) 10.0 (1.4, 6) 7.1 (1.5, 6) 2.9 (0.2, 3) (12, 1.0, 1.1, 0.3, 117 (4, 2) 7.9 (0.1, 2) 4.0 (2.4, 2) 2.3 (0.0, 1) (10 11

PAGE 107

95 Table A-2. Continued Mechanical Animal Site variable Cranial Rib position Central Caudal MSW0134 Proximal Middle S (MPa) E (GPa) W (MJm-3) K C (MPam1/2) S E W K C 125 (22, 14) 12.0 (1.7,14) 4.5 (1.9, 14) 2.1 (0.5, 14) 141 (24, 12) 12.5 (0.9, 12) 5.6 (2.1, 12) 2.1 (0.3, 11) 145 (18, 6) 12.2 (0.5, 6) 6.9 (2.9, 6) 2.4 (0.4, 4) 125 (14, 26) 10.6 (0.9, 26) 5.7 (1.8, 26) 2.0 (0.3, 25) MSE0205 Proximal Middle S E W K C S E W K C 123 14) 12.2 (2.1, 14) 3.5 (1.3, 14) 2.4 (0.2, 10) 151 (19, 14) 14.9 (0.8, 14) 4.5 (1.2, 14) 2.9 (0.2, 10) (27, 148 (18, 19) 13.5 (1.0, 19) 5.2 (1.5, 19) 2.8 (0.2, 10) 122 (12, 26) 11.2 (0.9, 26) 3.8 (1.2, 26) 2.5 (0.3, 13) 140 (13, 18) 12.7 (0.6, 18) 5.3 (1.9, 18) 2.6 (0.2, 10) 131 (11, 23) 12.0 (0.8, 23) 4.3 (1.3, 23) 2.5 (0.2, 10) MSW0160 Proximal Middle 16034) 15.1 (1.0, 34) 5.3 (2.1, 34) 2.8 (0.3, 10) S E W K C S E W K C 179 (20, 19) 16.5 (1.2, 19) 5.7 (1.5, 19) 3.0 (0.3, 10) (14, 167 (13, 23) 15.7 (1.0, 23) 5.6 (2.2, 23) 2.9 (0.2, 10) 153 (13, 36) 13.8 (1.1, 36) 6.4 (2.2, 36) 2.7 (0.2, 10) 142 (20, 19) 14.0 (1.0, 19) 4.4 (2.5, 19) 2.6 (0.2, 10) MSTM0102 1 1 Proximal Middle S E W K C S E W K C 155 (20, 32) 15.5 (2.0, 32) 4.2 (1.0, 32) 2.4 (0.5, 31) 155 (22, 32) 15.1 (1.5, 32) 4.2 (1.5, 32) 2.8 (0.3, 26) 164 (15, 24) 15.8 (1.2, 24) 5.3 (1.4, 24) 2.5 (0.3, 24) 153 (22, 28) 4.8 (1.0, 28) 3.7 (0.8, 28) 2.6 (0.2, 23) 61 (16, 15) 16.2 (1.2, 15) 5.2 (1.3, 15) 2.7 (0.4, 14) 155 (16, 38) 15.3 (1.3, 38) 4.0 (0.7, 38) 2.6 (0.3, 37)

PAGE 108

96 Table A-2. Continued Mechanical Animal Site variable Cranial Rib position Central Caudal MSW0165 Proximal Middle S (MPa) E (GPa) W (MJm-3) K C (MPam1/2) S E W K C 125 (38, 21) 12.6 (3.3, 21) 3.4 (1.3, 21) 2.7 (0.2, 10) 160 31) 14.9 (2.4, 31) 4.5 (1.2, 31) 2.9 (0.3, 10) 164 27) 14.7 (1.0, 27) 5.4 (0, 27) 2.9 (0, 10) (26, 156 (23, 32) 15.7 (1.9, 32) 4.5 (1.8, 32) 2.7 (0.4, 10) (13, .8 .2 163 (20, 20) 16.0 (1.7, 20) 5.2 (1.8, 20) 2.7 (0.3, 9) 137 (21, 26) 13.0 (2.4, 26) 4.5 (1.1, 26) 2.5 (0.6, 10) MEC0220 Proximal Middle S E W K C S E W K C 132 (52, 20) 13.3 (4.9, 20) 3.2 (1.5, 20) 2.6 (0.3, 10) 137 (37, 22) 14.5 (3.4, 22) 3.2 (1.2, 22) 2.8 (0.3, 10) 148 (15, 23) 14.4 (1.2, 23) 3.9 (1.1, 23) 2.7 (0.2, 10) 142 (13, 28) 14.4 (0.9, 28) 3.3 (0.7, 28) 2.5 (0.3, 10) 153 (24, 22) 15.0 (2.0, 22) 4.4 (1.6, 22) 2.6 (0.3, 10) 134 (14, 28) 13.5 (0.9, 28) 3.1 (0.9, 28) 2.4 (0.2, 9) MSW0251 Proximal Middle S E W K C S E W K C 153 27) 17.1 (127) 3.4 (127) 2.7 (0 26) 145 (25, 24) 17.0 (1.5, 24) 2.7 (0.7, 24) 2.5 (0.3, 10) 162 (22, 28) 18.0 (1.1, 28) 3.7 (1.2, 28) 2.8 (0.3, 28) 158 (26, 21) 17.1 (1.8, 21) 4.0 (1.2, 21) 2.8 (0.4, 10) 177 (21, 30) 19.0 (1.3, 30) 4.0 (1.2, 30) 3.1 (0.3, 30) 157 (21, 12) 16.8 (1.5, 12) 3.7 (0.8, 12) 2.8 (0.4, 10) (17, .0, .2, .4, MSW0225 Proximal Middle S E W K C S E W K C 142 (23, 16) 15.9 (1.9, 16) 3.0 (0.9, 16) 2.5 (0.3, 10) 142 (17, 26) 15.6 (1.3, 26) 3.4 (0.8, 26) 2.5 (0.3, 26) 157 (19, 32) 16.5 (1.4, 32) 4.0 (1.3, 32) 2.7 (0.3, 10) 158 (18, 32) 16.2 (2.0, 32) 4.3 (0.9, 32) 3.0 (0.3, 30) 147 (32, 24) 15.9 (1.2, 24) 3.9 (1.0, 24) 2.6 (0.3, 10) 149 (13, 30) 15.8 (1.0, 30) 3.8 (0.7, 30) 2.8 (0.3, 28)

PAGE 109

97 Table A-2. Continued Mechanical Animal Site variable Cranial Rib position Central Caudal MSW0223 Proximal Middle S (MPa) E (GPa) W (MJm-3) K C (MPam1/2) S E W K C 152 (18, 25) 15.6 (1.2, 25) 3.8 (0.6, 25) 2.7 (0.3, 10) 150 (16, 30) 15.7 (0.9, 30) 3.7 (1.2, 30) 2.7 (0.2, 10) 157 (23, 32) 16.0 (1.4, 32) 4.1 (1.2, 32) 2.8 (0.3, 10) 151 (13, 33) 15.2 (1.1, 33) 3.9 (0.8, 33) 2.8 (0.2, 10) 156 (17, 32) 15.3 (1.3, 32) 4.3 (1.0, 32) 3.0 (0.4, 10) 150 (12, 32) 15.3 (0.6, 32) 4.2 (1.1, 32) 2.7 (0.2, 10) MSW0208 Proximal Middle S E W K C S E W K C 140 (24, 26) 15.0 (2.1, 26) 3.5 (1.0, 26) 2.6 (0.3, 10) 152 (15, 32) 16.1 (1.3, 32) 3.9 (0.7, 32) 2.8 (0.2, 10) 156 (16, 28) 16.4 (1.2, 28) 4.3 (0.9, 28) 2.9 (0.2, 10) 156 (17, 32) 15.7 (1.1, 32) 4.2 (1.3, 32) 2.8 (0.2, 10) 160 (16, 21) 16.8 (1.2, 21) 5.0 (1.2, 21) 2.8 (0.2, 10) 156 (20, 36) 16.2 (1.5, 36) 4.3 (1.4, 36) 2.7 (0.2, 10) MEC0213 Proximal Middle S E W K C S E W K C 124 (18, 26) 14.8 (1.4, 26) 2.1 (0.7, 26) 2.4 (0.2, 10) 136 (16, 22) 15.1 (1.4, 22) 2.7 (0.5, 22) 2.6 (0.2, 10) 142 (21, 17) 15.4 (1.6, 17) 3.5 (0.7, 17) 2.6 (0.2, 10) 154 (19, 33) 16.1 (1.2, 33) 3.2 (0.6, 33) 2.7 (0.4, 10) 149 (25, 18) 15.7 (1.6, 18) 3.7 (1.3, 18) 2.8 (0.3, 10) 151 (17, 27) 15.8 (1.1, 27) 4.0 (1.2, 27) 2.8 (0.3, 10) MSW0226 Proximal Middle S E W K C S E W K C 129 (28, 20) 15.9 (1.7, 20) 2.3 (0.6, 20) 2.5 (0.3, 10) 139 (19, 25) 15.5 (1.7, 25) 2.5 (0.6, 25) 2.5 (0.2, 10) 159 (22, 26) 17.3 (1.6, 26) 3.6 (1.0, 26) 2.8 (0.2, 10) 154 (14, 32) 16.8 (0.9, 32) 3.1 (0.6, 32) 2.7 (0.3, 10) 164 (22, 24) 17.4 (1.3, 24) 4.1 (1.4, 24) 2.8 (0.2, 10) 154 (16, 29) 16.3 (1.4, 29) 3.2 (0.6, 29) 2.8 (0.2, 10)

PAGE 110

98 Table A-2. Continued Animal Site Mechanical variable Cranial Rib position Central Caudal MSW0239 Proximal Middle S (MPa) E (GPa) W (MJm-3) K C (MPam1/2) S E W K C 136 (24, 30) 15.6 (1.9, 30) 2.7 (0.8, 30) 2.6 (0.2, 10) 154 (20, 28) 16.5 (1.8, 28) 3.6 (0.9, 28) 2.8 (0.3, 10) 149 (21, 31) 16.8 (1.3, 31) 3.7 (1.1, 31) 2.9 (0.3, 10) 143 (24, 32) 15.9 (1.6, 32) 3.3 (1.0, 32) 2.5 (0.5, 10) 165 (16, 28) 17.1 (1.1, 28) 5.2 (1.7, 28) 3.0 (0.1, 10) 153 (18, 32) 16.4 (1.2, 32) 3.7 (0.8, 32) 2.9 (0.3, 10) MSW0331 Proximal Middle S E W K C S E W K C 153 (20, 27) 17.4 (2.2, 27) 2.7 (0.7, 27) 2.8 (0.4, 9) 148 (19, 30) 16.7 (1.5, 30) 3.1 (0.7, 30) 2.6 (0.3, 29) 148 (25, 27) 17.3 (1.7, 27) 3.0 (0.9, 27) 2.6 (0.4, 10) 154 (17, 32) 17.1 (0.8, 32) 3.4 (1.0, 32) 2.5 (0.3, 30) 151 (22, 25) 17.4 (1.2, 25) 3.4 (1.2, 25) 2.7 (0.3, 10) 151 (21, 30) 17.3 (2.4, 30) 3.4 (1.0, 30) 2.5 (0.4, 29) MSW0253 Proximal Middle S E W K C S E W K C 145 (29, 31) 16.1 (1.8, 31) 3.3 (1.2, 31) 2.6 (0.5, 10) 155 (21, 33) 15.9 (1.7, 33) 4.3 (1.2, 33) 2.8 (0.4, 29) 156 (22, 32) 16.5 (1.5, 32) 4.0 (1.2, 32) 2.8 (0.3, 10) 152 (18, 36) 15.6 (1.9, 36) 3.9 (1.0, 36) 2.7 (0.3, 31) 152 (19, 24) 15.4 (1.6, 24) 4.2 (1.6, 24) 2.7 (0.2, 10) 137 (22, 32) 14.7 (1.3, 32) 3.3 (0.9, 32) 2.3 (0.4, 29) SWFTM041 4 Proximal Middle S E W K C S E W K C 151 (18, 33) 17.3 (1.0, 33) 2.9 (1.0, 33) 2.8 (0.3, 10) 150 (22, 32) 16.6 (1.6, 32) 3.6 (1.4, 32) 2.7 (0.3, 10) 171 (18, 32) 18.6 (1.4, 32) 3.9 (0.9, 32) 3.2 (0.2, 10) 156 (18, 28) 17.2 (0.7, 28) 3.4 (0.9, 28) 2.9 (0.2, 10) 165 (19, 32) 18.0 (0.5, 32) 4.4 (1.5, 32) 3.0 (0.3, 10) 152 (17, 32) 17.0 (1.1, 32) 3.6 (0.9, 32) 2.9 (0.3, 10)

PAGE 111

99 Table A-2. Continued Animal Site Mechanical variable Cranial Rib position Central Caudal MNW0208 Proximal Middle S (MPa) E (GPa) W (MJm-3) K C (MPam1/2) S E W K C 153 (16, 31) 18.6 (1.3, 31) 2.4 (0.6, 31) 2.9 (0.3, 10) 157 (21, 33) 17.8 (1.7, 33) 3.0 (0.9, 33) 2.8 (0.3, 10) 159 (24, 27) 18.1 (1.6, 27) 2.9 (0.8, 27) 2.9 (0.3, 10) 162 (19, 27) 17.7 (1.1, 27) 3.1 (0.8, 27) 2.9 (0.2, 10) 173 (17, 24) 19.0 (1.3, 24) 3.5 (0.8, 24) 3.1 (0.2, 10) 154 (20, 30) 17.5 (1.0, 30) 2.9 (0.7, 30) 2.9 (0.3, 10) MSW0245 Proximal Middle S E W K C S E W K C 89 (28, 31) 9.3 (3.3, 31) 2.0 (0.9, 31) 1.9 (0.3, 15) 141 (20, 52) 15.9 (1.8, 52) 3.2 (0.6, 52) 2.4 (0.3, 25) 154 (19, 32) 16.8 (1.1, 32) 3.2 (0.9, 32) 2.5 (0.4, 15) 150 (18, 87) 16.6 (1.4, 87) 3.2 (0.7, 87) 2.4 (0.3, 22) 149 (21, 35) 16.2 (2.0, 35) 3.4 (1.1, 35) 2.5 (0.3, 8) 149 (17, 70) 16.2 (1.5, 70) 3.3 (0.8, 70) 2.3 (0.3, 8)

PAGE 112

100 Table A-3. Critical crac k measurements for whole manatee ribs fractured in impact Anial KCb m A c c Y SW0.00190 FTM0109 2.5 0.00195 0.001925 0.043873 1.24 46 MN 0.00210 00165 861 5 1.24 0.00230 0.00272 13 0.00160 0.0019MSW0057 2.2 0001MSE0205 2.5 012MSW0160 2.7 0.00210 0.00150 0.001775 0.042129 1.24 52 SWFTM0102 0.00170 0.00143MSW0165 2.9 0290 0.0023MEC0220 2.4 050 0.0022MSW0251 3.0 0250 0.00220.00270 0.00220 0280 0.0014MSW0208 2.8 070 0.0020MEC0213 2.7 025 0.0011MSW0226 2.7 0.00280 0.00285 MSW0239 2.5 0140 0.0015 MSW0253 2.7 0220 0.002SWFTM0414 2.8 0.00140 0.0039MNW0208 #6 2.9 0.00220 0.0025NW0208 #9 2.9 0.00150 0.00150 0.001500 0.038730 1.12 67 0208 #10 2.9 0.00400 0.00300 0.003464 0.058857 1.24 40 MNW0208 #14 2.9 0.00350 0.00350 0.003500 0.059161 1.24 40 MSW0245 2.4 0.00280 0.00175 0.002214 0.047049 1.24 41 W0207 2.4 0. 0.001 0.04314 45 MSW0227 SWFTM01 2.3 2.0 0 0.00249 0.049920 1.24 37 50 0.00154 0.039360 1.24 41 .00190 0. 15 0.00147 8 0.038447 1.24 46 .00230 0.00 35 0.00176 0.041977 1.24 48 2.5 0 0.00154 0.039278 1.24 51 .00 50 0.00269 0.051890 1.24 45 .002 75 0.00262 0.051206 1.24 38 .00 25 0.00237 0.048700 1.24 50 MSW0225 3.0 MSW0223 0.002437 0.049368 1.24 49 2.7 .00 75 0.00221 0.047049 1.24 46 .00201 70 0.00270 0.051962 1.24 43 .0 90 0.001540.002825 0.0392570.053150 1.12 611.24 41 .00 75 0.00156 0.039563 1.12 56 .00 40 0.002298 0.047936 1.24 45 00 0.00204 0.045270 1.24 50 50 0.00234 0.048427 1.24 48 MMNW a and b are crack depth and half-width (m). Crack size was calculated as c = (ab)1/2. Y is a geometric factor based on crack shape. is flexural strength (MPa).

PAGE 113

APPEND IXRNKTfine the the area under the stress-strain curve ) he ultimtrtr. Wctcad f er the lois c thte stress and selations e type odiestn,ate B DE IVATIO OF WOR OF FRAC URE We de work of fracture (WOF) as uOF 0 d W (B-1 where is t u ate s ain, is s ess (MPa) ork of fra ure can be lculate rom the area und ad d placement urve with e appropria train r applied to th f loa ng. For sp cimens te ed in tensio it is calcul d as A 2 WWOF ) areait ) and A is the apparent area re ceaeniat ce w r flexure. Stress in 3 point flexure is calculated as (B-2 where W is the under the load-d splacemen curve (N m (m) of the fractu 2 surfa This equ tion is oft inappropr ely used to alculat ork of fracture for specimens tested in flexure. To correctly calculate work of fracture, the load-displacement curve must be converted to the stress-strain curve fo 2BD2FL3 (B-3) where F is the peak load (Newtons) at failure, l is the distance (m) between the lower supports of the bending fixture, B is the specimen width (m), and D is the specimen depth (m). Strain is calculated as LLL (B-4) where is the displacement (m). 101

PAGE 114

102 If we substitute Equations B-3 and B-4 into Equation B-1 we get LdBD2FL3WOFu02 (B-5) which is equivalent to 2BD2 W3WOF (B -6) where W is the area under the load-displacement c urve (N m).

PAGE 115

LIST OF REFERENCES Ackerman, B.B., S.D. Wright, R.K. Bonde, D.K. Odeand patterns in mortality of manatees in Florida, 1974-1992. Pages 223-258 T.J. OShea. B.B. Ackerman, and H.F. Percival, editors. Population biology of the Florida ology Report 1. Springfield, VA: National Technical Information Service. 289 p. Aksel, C., and P.D. Warren. 2003. Work of fracture and fracture surface energy of manesia-spinel composites. Composites Science and Technology 63:1433-1440. ASTM Annual Book of Standards. 2001. ASTM D790M-92. Standard test methods for flexural properties of unreinforced plastics and electrical insulating materials. ASTM International. Batson, E.L., G.C. Reilly, J.D. Currey, and D.S. Balderson. 2000. Postexercise and positional variation in mechanical properties of the radius in young horses. Equine Veterinary Journal 32(2):95-100. Bellairs, A.DA., and C.R. Jenkin. 1960. The skeleton of birds. Pages 241-300 in A.J.Marshall, editor. Biology and comparative physiology of birds. Volume 1. NewYork, NY: Academic Press. 518 p. Brear, K., and J.D. Currey. 1993. The mechanical design of the tusk of the narwhal (Monodon monoceros: Cetacea). Journal of Zoology, London 230:411-423. Budker, P. 1971. The life of sharks. New York, NY: Columbia University Press. 222 p. Burr, D.B. 2002. The contribution of the organic matrix to bones material properties. Bone 31(1):8-11. Carter, D.R., and G.S. Beaupr. 2001. Skeletal Function and Form. Cambridge: Cambridge University Press. 318 p. Compagno, L.J.V. 1999. Endoskeleton. Pages 69-92 in W.C Hamlett, editor. Sharks, skates, and rays: the biology of elasmobranch fishes. Baltimore,MD: Johns Hopkins University Press. 515 p. Currey, J.D. 1960. Differences in the blood supply of bone of different histological types. Quarterly Journal of Microscopical Science 101(3):351-370. ll, and D.J. Banowetz. 1995. Trends in manatee. National Biological Service Information and Techn g 103

PAGE 116

104 Currey, J.D. 1968. The effect of protection on the impact strength of rabbits bones. Acta Anatomica 71:87-93. Currey, J.D. 1969. The mechanical consequences of variation in the mineral content of bone. Journal of Biomechanics 2:1-11. Currey, J.D. 1975. The effects o and mineral content on some mechanical properties of bovine bone. Journal of Biomechanics 8:81-86. CuCu with age. Journal of Biomechanics 12:459-469. CuUniversity Press. 294 p. Cue. Journal of Biomechanics 21:439-441. al of Currey, J.D. 1998. Mechanical properties of vertebrate hard tissues. Proceedings of the Currey, J.D. 1999. What determines the bending strength of compact bone? Journal of Currey, J.D. 2001. Ontogenetic changes in compact bone material properties. Pages 19.1ty 487-Currey, J.D. 2004b. Tensile yield in compact bone is determined by strain, post-yield f strain rate, reconstruction rrey, J.D. 1979a. Mechanical properties of bone tissues with greatly differing functions. Journal of Biomechanics 12:313-319. rrey, J.D. 1979b. Changes in the impact energy absorption of bone rrey, J.D. 1984. The Mechanical Adaptations of Bones. Princeton, NJ: Princeton rrey, J.D. 1988a. The effects of drying and re-wetting on some mechanical properties of cortical bon Currey, J.D. 1988b. Strain rate and mineral content in fracture models of bone. JournOrthopaedic Research 6:32-38. Institution of Mechanical Engineers Part H. Journal of Engineering in Medicine212(H6):399-411. Experimental Biology 202:2495-2503. 16 in S.C. Cowin, ed. Bone mechanics handbook. Boca Raton, FL: CRC Press. 980 p. Currey, J.D. 2002. Bones: structures and mechanics. Princeton, NJ: Princeton UniversiPress. 436 p. Currey, J.D. 2003. The many adaptations of bone. Journal of Biomechanics 36:11495. Currey, J.D. 2004a. Incompatible mechanical properties in compact bone. Journal of Theoretical Biology 231:569-580. behaviour by mineral content. Journal of Biomechanics 37:549-556.

PAGE 117

105 Currey, J.D., and K. Brear. 1990. Hardness, Youngs modulus and yield stress in mammalian mineralize d tissues. Journal of Materials Science: Materials in Medicine 1:14-20. Currey, J.D., K. Brear, and P. Zioupos. 1994. Dependence of mechanical properties on fibre angle in narwhal tusk, a highly oriented biological composite. Journal of Currey, J.D., K. Brear, and P. Zioupos. 1996. The effects of ageing and changes in echanics rgery 57A(6):810-814. axis ns. Journal of Zoology, London 218:59-67. ss. p. 117-144 in: Sarikaya, M. and Aksay, I.A. eds. Biomimetics: design and processing of materials. Woodbury, NY: De Buffrenil, V., A. de Ricqles, C.E. Ray, and D.P. Domning. 1990. Bone histology of y Della Bona, A., T.J. Hill, and J. J. Mecholsky, Jr. 2001. The effect of contour angle on 07 thed whales and sirenians. Special Issue 3, Reports of the International Whaling Commission, Cambridge, UK. 229 p. Biomechanics 27:885-897. mineral content in degrading the toughness of human femora. Journal of Biom29(2):257-260. Currey, J.D., and G. Butler. 1975. The mechanical properties of bone tissue in children. Journal of Bone and Joint Su Currey, J.D., and C.M. Pond. 1989. Mechanical properties of very young bone in thedeer (Axis axis) and huma Currey, J.D., P. Zioupos, and A. Sedman. 1995. Microstructure-property relations in vertebrate bony hard tissues: microdamage and toughne AIP Press. 285 p. Dayoub, S., H. Devlin, and P. Sloan. 2003. Evidence for the formation of metaplastic bone from pericytes in calcifying fibroblastic granuloma. Journal of Oral Pathology and Medicine 32:232-236. the ribs of the archaeocetes (Mammalia: Cetacea). Journal of Vertebrate Paleontolog10:455-466. De Buffrenil, V., and D. Schoevaert. 1989. Donnees quantitatives et observations histologiques sur la pachyostose du squelette du dugong, Dugong dugon (Muller) (Sirenia, Dugongidae). Canadian Journal of Zoology 67:2107-2119. fractal dimension measurement for brittle materials. Journal of Materials Science 36:2645-2650. Domning, D.P., and V. de Buffrenil. 1991. Hydrostasis in the Sirenia: quantitative data and functional interpretations. Marine Mammal Science 7(4):331-368. Domning, D.P., and A.C. Myrick, Jr. 1980. Tetracycline marking and the possible layering rate of bone in an Amazonian manatee (Trichechus inunguis). Pages 203-2in W.F. Perrin and A.C. Myrick, Jr., editors. Age determination of too

PAGE 118

106 Eberhardt, L.L., and T. J. OShea. 1995. Integration of manatee life history data anpopulation modeling. Pages 269-279 in T.J. OShea. B.B. Ackerman, and H.F. Percival, d editors. Population biology of the Florida manatee. National Biological Service Information and Technology Report 1. Springfield, VA: National Technical Enlow, D.H., and S.O. Brown. 1956. A comparative histological study of fossil and recent bone tissues. Part I. Texas Journal of Science 8(4):405-443. Enes. Part II. Texas Journal of Science 9(2):186-214. 10(2):187-230. olume fractions, and in a bone analogue. Journal of Materials Science: Materials in Medicine Fabryology 196:349-362. Fawcett, D.W. 1942b. The amedullary bones of the Florida manatee (Trichechus Fe., L.C. Drickamer, S.H. Vessey, and J.F. Merritt. 1999. Mammalogy: adaptation, diversity, and ecology. Dubuque: McGraw-Hill. 563 p. Fes f 03-134. Zoomorphology 110:339-345. Fl Information Service. 289 p. low, D.H., and S.O. Brown. 1957. A comparative histological study of fossil and recent bone tissu Enlow, D.H., and S.O. Brown. 1958. A comparative histological study of fossil and recent bone tissues. Part III. Texas Journal of Science Evans, G.P., J.C. Behiri, J.D. Currey, and W. Bonfield. 1990. Microhardness and Youngs modulus in cortical bone exhibiting a wide range of mineral v 1:38-43. ng, J., and B.K. Hall. 1997. Chondrogenic cell differentiation from membrane bone periostea. Anatomy and Em Fawcett, D.W. 1942a. On the amedullary bones of the Florida manatee (Trichechus latirostris). Anatomical Record 82:410-411. latirostris). American Journal of Anatomy 71:271-309. ldhamer, G.A lts, W.J.L., and F.A. Spurrell. 1965. Structural orientation and density in cetacean humeri. American Journal of Anatomy 116:171-203. Felts, W.J.L., and F.A. Spurrell. 1966. Some structural and developmental characteristicof cetacean (Odontocete) radii. A study adaptive osteogenesis. American Journal oAnatomy 118:1 Fish, F.E., and B.R. Stein. 1991. Functional correlates of differences in bone density among terrestrial and aquatic genera in the family Mustelidae (Mammalia). orida Department of Natural Resouces. 1989. Recommendations to improve boating safety and manatee protection for Florida waterways. Final report to the Governor and Cabinet, State of Florida. 20 p.

PAGE 119

107 Florida Fish and Wildlife Conservation Commission. 2005. Manatee Mortality StatistFlorida Marine Research Institute, St. Petersburg, FL. http://floridamarine.org/features/category_sub.asp?id=2241 (last accessed: 8/2005 ics. ). Foote, J.S. 1916. A contribution to the comparative histology of the femur. Smithsonian Contributions to Knowledge 35:1-242. Ga Dalisky, and S.S. Sherman. 1998. Caliotropic hormones and bone markers in the elderly. Journal of Goodrich, E.S. 1986. Studies on the structure and development of vertebrates. Chicago, Gundberg, C.M., A.C. Looker, S.D. Nieman, and M.S. Calvo. 2002. Patterns of Hall, B.K. 1984. Development processes underlying the evolution of cartilage and bone. Pages 155-176 in M.J.W. Ferguson, editor. The structure, development and evolution Hall, B.K. 2005. Bones and cartilage: developmental and evolutionary skeletal biology. Harvey, K.B., and S.W. Donahue. 2004. Bending properties, porosity, and ash fraction of espite 13-1520. nual periods of hibernation. Journal of Biomechanics 38:2143-2150. Haf cortical and trabecular bone: Implications for assessment of fracture risk. Pages 69-111 in Basic orthopaedic Hildebrand, M. 1988. Analysis of vertebrate structure. Third Ed. New York, NY: John Wiley & Sons. 701 p. Hi. PhD ida, Gainesville, Florida. 100 p. llagher, J.C., H.K. Kinyamu, S.E. Fowler, B. Dawson-Hughes, G.P. Bone and Mineral research 13(3):475-482. IL: University of Chicago Press. 837 p. osteocalcin and bone specific alkaline phosphatase by age, gender, and race or ethnicity. Bone 31(6):703-708. of reptiles. London: Zoological Society of London. 697 p. London, UK: Elsevier Academic Press. 760 p. black bear (Ursus americanus) cortical bone are not compromised with aging dannual periods of disuse. Journal of Biomechanics 37:15 Harvey, K.B., and S.W. Donahue. 2005. The tensile strength of black bear (Ursus americanus) cortical bone is not compromised with aging despite an yes, W.C., and M.L. Bouxsein. 1997. Biomechanics o biomechanics. Second Ed. Philadelphia, PA: Lippincott-Raven Publishers. 514 p. ll, T.J. 2001. Quantitative fracture analysis of a biological ceramic compositeDissertation, University of Flor Hill, T.J., A. Della Bona, and J.J. Mecholsky, Jr. 2001 Establishing a protocol for measurements of fractal dimensions in brittle materials. Journal of Materials Science 36:2651-2657.

PAGE 120

108 Jackson, A.P. 1992. Bone, nacre, and other ceramics. Pages 33-56 in Vincent, J.F.V., editor. Biomechanicsmaterials: a practical approach. NUniversity Press. 247 p. ewYork, NY: Oxford d. Baltimore, MD: Urban & Schwarzenberg. 1158 p. Jo 7:366. Kaufman, J.J., and R.S. Siffert. 2001. Nonivasive measurement of bone integrity. Pages Kawakami, T., T. Kawai, A. Kimura, H. Hasegawa, H. Tsujigiwa, M. Gunduz, H. ohistochemical and in situ hybridization examinations. Journal of International Medical Research 29:480-487. KeKipps, E.K., W.A McLellan, S.A. Rommel, and D.A Pabst. 2002. Skin density and its rine keleton. Trends in Molecular Medicine 11(3):129-137. ng avior in equine compact bone material. Journal of Orthopaedic Research 20:607-614. Liragility in mice. Journal of Bone and Mineral Research 20(1):S295. Jee, W.S.S. 1988. The skeletal tissues. Pages 213-254 in L. Weiss, Editor. Cell and tissuebiology: a textbook of histology. Sixth E hnson, M.L., J.L.Picconi, and R.R. Recker. 2002. The gene for high bone mass. TheEndocrinologist 12(5):445-453. Kaiser, H.E. 1965. The phylogenetic adaptation of aquatic mammals to their environment. American Zoologist 5:229. Kaiser, H.E. 1967. Pachyostotic bone conditions in certain regions of the skull ofOdobenus rosmarus L., in relation to weight distribution. Anatomical Record 15 34.1-34.25 in S.C. Cowin, editor. Bone mechanics handbook. Boca Raton, FL: CRC Press. 980 p. Nagatsuka, and N. Nagai. 2001. Characteristics of bone morphogenetic protein-induced chondriod bone: histochemical, immun nt, G.C., and L. Miller. 1997. Comparative anatomy of the vertebrates. Eighth ed. Dubuque, IA: McGraw Hill. 487 p. Key West Boats, Inc. 2005. 1760 Stealth. Ridgeville, SC: Key West Boats, Inc. http://www.keywestboatsinc.com/boats/1760stealth.html. (last accessed 12/2005). influence on buoyancy in the manatee (Trichechus manatus latirostris), harbor porpoise (Phocoena phocoena), and bottlenose dolphin (Tursiops truncatus). MaMammal Science 18(3):765-778. Koay, M.A., and M.A. Brown. 2005. Genetic disorders of the LRP5-Wnt signalling pathway affecting the s Les, C.M., S.M. Stover, J.H. Keyak, K.T. Taylor, and A.J. Kaneps. 2002. Stiff and strocompressive properties are associated with brittle post-yield beh C.Y., B.D. Gelb, K.J. Jepsen, and M.B. Schaffler. 2005a. Cathepsin K-deficiency results in bone f

PAGE 121

109 Li, Y.P., Y. Abe, R Isoda, R. Moroi, S. Yang, J.Z. Shao, E. Li, and W. Chen. 2005b. Cathepsin K-deficient mice exhibit pycnodysostosis features due to loss of cathK-mediated osteoclast se epsin nescence for bone homeostasis. Journal of Bone and Mineral Research 20(1):S253. Lormation of cartilage in the heart of the Spanish Terrapin, Mauremys leprosa (Reptilia, Chelonia). Journal of Morphology 258(1):97-105. Lucchinetti, E. 2001. Dense bone tissue as a molecular composite. Pages 13.1-15 25 in S.C. Cowin, editor. Bone mechanics handbook. Boca Raton, FL: CRC Press. 980 p. Mouillon, G. Carmeleir, and A. Nagy. 2005. Aberrent bone formation in mice over-expressing specific VEGF isoforms in cartilage. Majumdar, S., R.S. Weinstein, and R.R. Prasad. 1993. Application of fractal geometry Mako Marine International, Inc. 2005. 171 Center Console. Springfield, MO: Mako Marine International, Inc. http://www.mako-Mits rising R-curve fracture mechanics. Journal of Biomechanics 36:191-198. Mandelbort, B.B., D.E. Passoja, and A.J. Paullay. 1984. Fractal character of fracture surfaces of metals. Nature 308(19):721-722. MBenjamin/Cummings. 1040 p. Me. Maverick Boat Company, Inc. 2004. Maverick 17 Master Angler. Ft. Pierce, FL: 12/2005). pez, D., A.C. Duran, A.V. de Andrs, A. Guerrero, M. Blasco, and V. Sans-Coma. 2003. Fo aes, C., J.J. Haigh, A. Chan, K. Haigh, R. B Journal of Bone and Mineral Research 20(1):S39. techniques to the study of trabecular bone. Medical Physics 20(6):1611-1619. boats.com/boat/index.cfm?fuseaction=boat.specs&nav=193&boat=1678. (last accessed 12/2005). alik, C.L., S.M. Stover, R.B. Martin, J.C. Gibeling. 2003. Equine cortical bone exhib arieb, E.N.1995. Human anatomy and physiology. Third Ed. Redwood City, CA: armontel, M. 1993. Age determination and population biology of the Florida manatee, Trichechus manatus latirostris. Ph.D. dissertation, University of Florida, Gainesville. 408 p. Martin, R.B, and D.B. Burr. 1989. Structure, function, and adaptation of compact bonNew York, NY: Raven Press. 275 p. Martin, R.B., D.B. Burr, and N.A. Sharkey. 1998. Skeletal tissue mechanics. New York, NY: Springer-Verlag. 392 p. Maverick Boat Company, Inc. http://www.maverickboats.com/maverick/boat_specs/17ma.htm. (last accessed

PAGE 122

110 Mecholsky, J.J., Jr. 1986. Fractalsfact or fiction? Earth and Mineral Sciences 55(3)33. echolsky, J.J., Jr. 1993. Quantitativ :29-Me fracture surface analysis of glass materials. Pages 483-521 in C.J. Simmons and O.H. El-Bayoumi. editors. Experimental techniques of M fractal geometry: an integration. Pages 385-393 in J.R. Varner, V.D. Frechette, and G.D. Quinn, editors. n is eramic Society 72(1):60-65. M. 2002. Fractal dimension as a zine A 82 (17/18):3163-3176. ty in women with postmenopausal osteoporosis and patients with Pagets disease. Journal of Bone Meister, W. 1962. Histological structure of the long bones of penguins. Anatomical Moss, M.L. 1961. Osteogenesis of acellular teleost fish bone. American Journal of Murray, P.D.F. 1936. Bones: a study of the development and structure of the vertebrate skeleton. London: Cambridge University Press. 203 p. Nah, H.D., M. Pacifici, L.C. Gerstenfeld, S.L. Adams, and T. Kirsch. 2000. Transient chondrogenic phase in the intramembranous pathway during normal skeletal Nalla, R.K., J.J. Kruzic, J.H. Kinney, and R.O. Ritchie. 2004. Effect of aging on the Nihe palmar soft tissues. Anatomy and Embryology 148:215-221. glass science. Westerville: American Ceramic Society. 524 p. echolsky, J.J., Jr. 1996. Fractography, fracture mechanics and Ceramic Transations. Fractography of glasses and ceramics III. Westerville: AmericaCeramic Society. 547 p. Mecholsky, J.J., Jr. 2001. Fractography: brittle materials.Pages 3257-3265 in K.H.J. Buchow et al., eds. Encyclopedia of Materials: Science and Technology. Oxford, UK: Pergamon Press. 10,000 p. Mecholsky, J.J., Jr., D.E. Passoja, and K.S. Feinberg-Ringel. 1989. Quantitative Analysof Brittle Fracture Surfaces Using Fractal Geometry. Journal of the American C echolsky, J.J., Jr., J.K. West, and D.E. Passojacharacterization of free volume created during fracture in brittle materials. Philosophical Maga Meier, C., U. Meinhardt, M.E. Kraenzlin, J.R. Greenfield, T.V. Nguyen, C.R. Dunstan, and M.J. Seibel. 2005. Serum cathepsin K levels reflect osteoclastic activi and Mineral Research 20(1):S222. Record 143:377-387. Anatomy 108:99-110. development. Journal of Bone and Mineral Research. 15(3):522-533. toughness of human cortical bone: evaluation by R-curves. Bone 35:1240-1246.koli, V., J. Hanevi, M. Hudec, and B. Banovi. 1975. Absorption of the ipact energy in t

PAGE 123

111 Nopcsa, F. von. 1923. Vorlufige notiz ber die Pachyostose und Osteosclerose eininger mari ner Wirbeltiere. Anatomischer Anzeiger 56(15/16):353-359. Parkkari, J., P. Kannus, J. Heikkila, J. Poutala, H. Sievanen, and I. Vuori. 1995. Energyne and Mineral Research 10(10):1437-1442. c Research 1:405-411. l terials Reynolds, J.E. III, and D.K. Odell. 1991. Manatees and Dugongs. New York, NY: Facts Rhodin, A.G.J., J.A. Ogden, and G.J.Conlogue. 1981. Chondro-osseus morphology of e earch : OShea, T.J., C.A. Beck, R.K. Bonde, H.I. Kochman, and D.K. Odell. 1985. An analysis of manatee mortality patterns in Florida, 1976-81. Journal of Wildlife Management 49(1):1-11. Parkkari, J., P. Kannus, J. Poutala, and I. Vuori. 1995. Force attenuation properties of varioustrochanteric padding materials under typical falling conditions of the elderly. Journal of Bone and Mineral Research 9(9):1391-1396. shunting external hip protector attenuates the peak femoral impact force below thetheoretical fracture threshold: an in vitro biomechanical study under falling conditions of the elderly. Journal of Bo Pelker, R.R., G.E. Friedlaender, T.C. Markham, M.M. Panjabi, and C.J. Moen. 1984. Effects of freezing and freeze-drying on the biomechanical properties of rat bone. Journal of Orthopaedi Phelps, J.B., G.B. Hubbard, X. Wang, and C.M. Agrawal. 2000. Microstructuraheterogeneity and the fracture toughness of bone. Journal of Biomedical MaResearch 51:735-741. Reilly, G.C., and J.D. Currey. 1999. The development of microcracking and failure in bone depends on the loading mode to which it is adapted. Journal of Experimental Biology 202:543-552. On File, Inc. 192 p. Dermochelys coriacea, a marine reptile with mammalian skeletal features. Natur290:244-246. Riley, W.F., L.D. Sturges, and D.H. Morris. 1999. Mechanics of Materials, 5th Ed. New York, NY: John Wiley and Sons. 1999 p. Robinovitch, S.N., T.A. McMahon, and W.C. Hayes. 1995. Force attenuation in trochanteric soft tissues during impact from a fall. Journal of Orthopaedic Res13:956-962. Romer, A.S., and T.S. Parsons. 1977. The vertebrate body. Fifth ed. Philadelphia, PASaunders College Publishing. 624 p. Russ, J.C. 1994. Fractal surfaces. New York, NY: Plenum Press. 309 p.

PAGE 124

112 SAS Institute Inc. 1989. SAS/STAT User's Guide, Version 6, Fourth Edition, Volume 2. Cary, NC:SAS Institute Inc. 846 p. Savage, R.J.G. 1976. Review of early Sirenia. Systematic Zoology 25:344-351. Schaffler, M.B., and D.B. Burr. 1988. Stiffness of compact bone: effects of porosity and density. Journal of Biomechanics 21(1):13-16. Sc Shhopaedics e Paleobiologica (Wien) 4(6/7):405-444. Sin in -476. sts in vitro. Plastic and Reconstructive Surgery 104(3):738-Swartz, S.M., and P. Watts. 1999. Structural design of finger bones of bats. American Zoologist 39(5):101A. :2839-2849. d States Coast (last accessed 11/05). hryver, F. 1978. Bending properties of cortical bone of the horse. American Journal ofVeterinary Research 39(1):25-28. Sedlin, E.D., and C. Hirsch. 1966. Factors affecting the determination of the physical properties of femoral cortical bone. Acta orthop. Scandinavia 37:29-48. apiro, F. 1993. OsteopetrosisCurrent clinical considerations. Clinical Ortand Related Research 294:34-44. Sickenberg, O. 1931. Morphologie und stammesgeschichte der Sirenen I. Die einflssdes Wasserlebens auf die innere secretion und formgestaltung der Sirenen. omin, V., R. Willinsky, P. Shannon, and A.Guha1. 2003. Metaplastic bone formatioa low grade conus glioma: case report and review of the literature. Journal of Neuro-Oncology 62: 275. Stein, B.R. 1989. Bone density and adaptation in semiaquatic mammals. Journal of Mammalogy 70(3):467 Steinbrech, D.S, B.J. Mehrara, P.B. Saadeh, G. Chin, M.E. Dudziak, R.P. Gerrets, G.K. Gittes, and M.T. Longaker. 1999. Hypoxia regulates VEGF expression and cellular proliferation by osteobla 747. Thompson, D.W. 1942. On growth and form: the revised edition. Mineola, NY: Dover Publications, Inc. 1116 p. Tolar, J., S.L. Teitelbaum, P.J. Orchard. 2004. Osteopetrosis. New England Journal of Medicine 351 Turner, C.H., and D.B. Burr. 1993. Basic biomechanical measurements of bone: a tutorial. Bone 14:595-608. U.S.Coast Guard. 2005. Boating statistics-2004. Washington, DC: UniteGuard. 47p. http://www.uscgboating.org/statistics/Boating_Statistics_2004.pdf

PAGE 125

113 U.S. Fish and Wildlife Service. 1996. Florida manatee (Trichechus manatus latirostris) recovery plan. U.S. Fish and Wildli fe Service, Atlanta, Ga. 160 p. recovery Plan, (Trichechus manatus latirostris), Third Revision. Atlanta, GA: U.S. Fish and Vashishth, D. 2004. Rising crack-growth-resistance behavior in cortical bone: Vashishth, D., J.C. Behiri, and W. Bonfield. 1997. Crack growth resistance in cortical Vincent, J.F.V. 1990. Structural Biomaterials. Princeton, NJ: Princeton University Press. Wainwright, S.A., W.D. Biggs, J.D. Currey, and J.M. Gosline. 1982. Mechanical Design Whigh limb-bone density and aquatic habits in recent mammals. Journal of Paleontology 57(2):197-203. Wang, X., R.A. Bank, J.M. TeKoppele, and C.M. Agrawal. 2001. The role of collagen in nical properties. Journal of Orthopaedic Research 19:1021-1026. Wang, X., X. Shen, X. Li, and C. M. Agrawal. 2002. Age-related changes in the collagen network and toughness of bone. Bone 31(1):1-7. Wr, ssion. Journal of Surgical Research 99:147-155. ittle fracture. Journal of Non-Crystalline Solids 260:99-108. Salmon (Salmo salar). International Journal of Developmental Biology 46:719-730. W003. Seasonal changes in the lower jaw skeleton in male Atlantic salmon (Salmo salar L.): remodelling and regression of the kype after U.S. Fish and Wildlife Service. 2000. Technical/Agency Draft, Florida Manatee Wildlife Service. 145 p. implications for toughness measurements. Journal of Biomechanics 37:943-946. bone: concept of microcrack toughening. Journal of Biomechanics 30(8): 763-769. 244 p. in Organisms. Princeton, NJ: Princeton University Press. 423 p. all, W.P. 1983. The correlation between determining bone mecha arren, S.M., D.S. Steinbrech, B.J. Mehrara, P.B. Saadeh, J.A. Greenwald, J.A. SpectoP.J. Bouletreau, and M.T. Longaker. 2001. Hypoxia regulates osteoblast gene expre West, J.K., J.J. Mecholsky, Jr., and L.L. Hench. 1999. The application of fractal and quantum geometry to br Witten, P.E., and B.K. Hall. 2002. Differentiation and growth of kype skeletal tissues inanadromous male Atlantic itten, P.E., and B.K. Hall. 2 spawning. Journal of Anatomy 203:435-450.

PAGE 126

114 Wright, S.D., B.B. Ackerman, R.K. Bonde, C.A. Beck, and D.J. Banowetz. 1995. Analysis of watercraft-related mortality of manatees in Florida, 197 9-1991. Pages 259-268 in T.J. OShea, B.B. Ackerman, and H.F. Percival, editors. Population biology of Yan, J. 2002. Biomechanical properties of manatee rib bone and analytical study using rida. e rib and bovine femur using a Chevron-notched beam test. Journal of Ye composition and Zioupos, P. 1998. Recent developments in the study of failure of solid biomaterials and Zi 1997. Mechanical properties Mesoplodon densirostris the Florida manatee. National Biological Service Information and Technology Report 1. Springfield, VA: National Technical Information Service. 289 p. finite element analysis. Master thesis, University of Florida, Gainesville, Flo115 p. Yan, J. 2005. Elastic-plastic fracture mechanics of compact bone. PhD Dissertation, University of Florida, Gainesville, Florida. 110 p. Yan, J., K.B. Clifton, R.L. Reep, and J.J. Mecholsky, Jr. 2005. Fracture toughness of manate Biomechanics. Accepted February 15, 2005. ni, Y.N., C.U. Brown, and T.L. Norman. 1998. Influence of bone apparent density on fracture toughness of the human femur and tibia. Bone 22(1):79-84. Zar, J.H. 1996. Biostatistical Analysis. Third Edition. Upper Saddle River: Prentice Hall. 662 p. bone: fracture and pre-fracture toughness. Materials Science and Engineering C 6:33-40. Zioupos, P., and J.D. Currey. 1998. Changes in the stiffness, strength, and toughness of human cortical bone with age. Bone 22(1):57-66. oupos, P., J.D. Currey, A. Casinos, and V. de Buffrenil. of the rostrum of the whale a remarkably dense bony tissue. Journal of Zoology, London 241:725-737. Zioupos, P., J.D. Currey, and A.J. Hamer. 1999. The role of collagen in the declining mechanical properties of aging human cortical bone. Journal of biomedical Materials Research 45(2):108-116.

PAGE 127

BIOGRAPHICAL SKETCH Kari Beth Clifton was born in Chicago, Illinois, and grew up in the near west suburb of Oak Park. She attended Valparaiso University where she majored in biolog y and French, and minored in chemistry. While there she spent a semester studying in Paago. She also played field hockey, served as an editor for the schools literary publication, and lear graduation Kari moved to Tampa, Florida to attend graduate school at the University of South Florida. Her research topic was entitled Ecomorphological relationships of five Master of Science in zoology in 1993. In January 1992, Kari began working for the work on the population biology of the Florida my, captures, rescues, and necropsies. mechanics that eventually led her to Gainesville. In 1998, Kari resigned her position to enter the PhD program at the University of Floridas College of Veterinary Medicine, to continue the bone research full time. Kari has accepted a postdoctoral position at the Mayo Clinic, where she will continue to do bone research. ris, and another term working with marine mammals at Brookfield Zoo in Chic rned to SCUBA dive. Kari graduated in 1988 with a Bachelor of Science degree. Afte species of labrids (Labridae, Teleostei) in a coral reef community. She graduated with a (former) Florida Department of Natural Resources tagging redfish and mullet. After 2 months, she moved to the Endangered and Threatened Species program, where she begananatee. During her 7 years with the state, she flew aerial surveys of manatees and right whales, and participated in radiotelemetrIn 1996 she undertook a study on manatee bone 115


Permanent Link: http://ufdc.ufl.edu/UFE0013069/00001

Material Information

Title: Skeletal Biomechanics of the Florida Manatee (Trichechus manatus latirostris)
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0013069:00001

Permanent Link: http://ufdc.ufl.edu/UFE0013069/00001

Material Information

Title: Skeletal Biomechanics of the Florida Manatee (Trichechus manatus latirostris)
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0013069:00001


This item has the following downloads:


Full Text












SKELETAL BIOMECHANICS OF THE FLORIDA MANATEE
(Trichechus manatus latirostris)

















By

KARI BETH CLIFTON


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2005





























Copyright 2005

by

Kari Beth Clifton

































to all those-people and animal alike-who have loved, supported, encouraged,
instructed, amazed, inspired, and empowered me















ACKNOWLEDGMENTS

This project was an enormous undertaking-one that I could not have completed

alone. First I would like to acknowledge my supervisory committee chair (Roger Reep)

and co-chair (Jack Mecholsky). I thank Roger for his unqualified generosity,

encouragement, and guidance; and for feeding the spirit with food and wine. I benefited

greatly from being allowed to wander off the path in pursuit of new ideas. I thank Jack

for his patience in teaching a biologist to think like an engineer, and for the tailgates. My

supervisory committee members (MaryBeth Horodyski, Donna Wheeler, and Tom

Wronski) provided guidance, instruction, and support for which I am grateful.

Many individuals contributed to the success of this project. While I was working

at the Florida Department of Environmental Protection, Dr. Sentiel Rommel suggested I

take a look at manatee bone. He and Dr. Tom Koob at the Shriners Hospital for Children

introduced me to the field of bone mechanics. The Save the Manatee Club funded a pilot

study that led to the inception of this project. I am indebted to Maggie Stoll for her

invaluable technical expertise and administrative support. To my fellow graduate students

in the fracture mechanics group, I am grateful for the weekly discussions of ceramics and

other things. I learned a great deal about fractography and bone mechanics from Jiahau

Yan.

From the College of Engineering, Chuck Broward provided access to the air

cannon, as well as his expert engineering skills. Billy Schultz performed the strain gauge

testing. Bryan Conrad from the Department of Orthopaedics taught me to operate the









biomechanical testing equipment, and provided invaluable technical support. Joan

Yonchek lent her amazing expertise in hard and soft tissue histology. I thank Allyson

Barrett and Ben Lee from the Department of Dental Biomaterials for their assistance with

technical issues and data collection. This project made great strides forward with the help

of veterinary students Julia Golden and Beth Carson; and undergraduate volunteers

Johnattan, Rebecca, Jessica, Stephanie, Jaime, Travis, and Chris.

Tissue samples were provided by the FWC's Marine Mammal Pathobiology Lab

in St. Petersburg, and by Sea World of Florida. I extend my thanks and appreciation to

Dr. Sentiel Rommel, Mr. Tom Pitchford, Mr. Ken Arrison and the staff of the MMPL,

and to Dr. Beth Chittick of Sea World.

I wish to thank the college of Veterinary Medicine, the Marine Animal Health

program, and in particular Dr. Charles Courtney for his interest and enthusiasm for my

project. This project was funded by a College of Veterinary Medicine Consolidated

Faculty Research award and a Marine Conservation Trust Fund award. Additional

funding was provided by grants from the Sigma Xi Scientific Research Society, a College

of Veterinary Medicine Graduate Student Seed Grant, and a Merck-Merial Veterinary

Scholars Research Grant. The University's Graduate Student Council and Graduate

School provided travel awards so that I could attend conferences.

Last, but not at all least, I would like to express my gratitude and appreciation for

the love and support of my family throughout my many years of schooling. I am indebted

to my nephews David, Adam, and Brain, for never, ever letting me get a lick of work

done while at their house. There are, after all, more important things in life.
















TABLE OF CONTENTS



A C K N O W L E D G M E N T S .................................................................... ......... .............. iv

L IST O F T A B L E S ........ ........................................................... ...... .. ....... .. viii

LIST OF FIGURES ......... ......................... ...... ........ ............ ix

ABSTRACT .............. .................. .......... .............. xi

CHAPTER

1 IN TR OD U CTION ............................................... .. ......................... ..

S k eleto g en e sis ............................................................................... 5
Bone Shape and M acrostructure........................................................ ............... 7
M icroscopic Structure of B one............................................................... ....................8
Mechanical Properties of Bone Related to Microscopic Structure.............................11
Variations in the Microscopic Structure of Vertebrate Bone ...............................14
M anatees in P erspectiv e ..................................................................... ..................24

2 MATERIAL PROPERTIES OF MANATEE RIB BONE .............. .....................29

Introdu action ................ ...... ......... ......... ................ ................. 2 9
M materials and M methods ....................................................................... ..................32
Bone Sample Collection .............. ..... ................................. 32
Specimen Preparation ........... .. .................................... 32
M material Properties T testing ................................................. ............... .... 33
Bone Density and Mineral Content Determination ......................... ..........35
Statistical A n aly ses........... ................................................ .......... ..... .... ... 36
Results ............. .................. ..... ...................... ... 36
W ithin-A nim al Com parisons......................................... .......................... 37
M material P properties T ests ........ ........ ................. .......... .............................37
Bone Density and Mineral Content..... ................. ...............38
D iscu ssion ....... ............... ... .. ...... ........................... ................... ............... 39
The Material Properties of Manatee Bone...................................... .................39
The Effect of Mineral Content on the Material Properties of Bone ..................45
Other Factors Influencing Bone Material Properties.............. ............ 46
S u m m a ry ................................................... ................. ................ 5 0









3 STRUCTURAL PROPERTIES OF MANATEE RIBS TESTED IN IMPACT........59

In tro d u ctio n ............. ...... ...... ... ................. ................................ 5 9
M materials and M methods ....................................................................... ..................60
Bone Sample Collection ............... ......... .................... 60
W hole R ib Im pact Testing........................................................ ............... 61
Strain G auge T testing .......... .................................. .............. ..... 62
Fractal D im ensional Increm ent ........................................ ....................... 63
Kinetic Energy Calculations.....................................................64
Statistical A n aly ses........... ................................................ .......... ..... .... .... 6 5
Results ............... ...... ............ ............. ...............65
W hole R ib Im pact T ests ........................................................... .....................65
Strain Gauge Tests................... ......................... ......66
Kinetic Energy Calculations.....................................................67
D discussion ......... ..... ..... ........ ........................ ...................... 68
The Fractal D im ensional Increm ent .............................................................. 68
Fractal Dimensional Increment and Fracture Toughness................ ........ 70
R-curve Behavior and Viscoelasticity in Bone ................................................71
Impact Energy Absorption of Soft Tissues ............................... ............... 73
Su m m ary ...................................... ............................... ................ 74

4 CONCLUSIONS AND FUTURE RESEARCH .....................................................85

Sum m ary and C onclu sions .............................................................. .....................85
Further Research ................................................................ .. ... ..... 87
A additional Avenues of Research ................................................... ............... ... 90

APPENDIX

A DATA TABLES .......................................................... 92

B DERIVATION OF WORK OF FRACTURE .................................. ...............101

L IST O F R E F E R E N C E S ...................................................................... ..................... 103

BIOGRAPHICAL SKETCH ........................................................................115
















LIST OF TABLES


Tablege

2-1 M ean m material properties for 26 m anatees .......................................... ................51

3-1 Mechanical and fractal data for 20 manatee ribs fractured in impact....................76

3-2 Mean mechanical and fractal data by sex and age class........................................76

3-3 M mechanical data for strain gauged ribs ........................................ .....................77

3-4 Boat data for kinetic energy calculations ...................................... ............... 77

3-5 Kinetic energy calculations for vessel configurations and impact projectile...........77

A-i Mortality information for animals used in this study .............................................92

A-2 Means for four mechanical properties by body region ........................................93

A-3 Critical crack measurements for whole manatee ribs fractured in impact............. 100
















LIST OF FIGURES


Figurege

2-1 Sample sites for biomechanical analyses ...................................... ............... 52

2-2 Flexural strength vs. total length .......................................................... ..... ......... 52

2-3 Fracture toughness vs. total length................................ ........ ................. 53

2-4 W ork of fracture vs. total length .............. ...... ........... ... .....................53

2-5 Elastic modulus vs. total length ............................ .......................... .. 54

2-6 Flexural strength vs. m odulus ............................................................................ 54

2-7 Load-displacement curves for small sample specimens tested in 3 point flexure....55

2-8 D ensity vs. total body length ........................................................................ .. .... 56

2-9 Percent mineral content vs. total body length .................................. ............... 56

2-10 M odulus vs. density .......... .............................................. .... .... .. .... .. 57

2-11 M odulus vs. percent m ineral content ............................................ ............... 57

2-12 Small sample crack size vs. total body length............. .. ............. ............... 58

3-1 Setup for impact tests of whole manatee ribs............................ .. ...............78

3-2 Fracture surface of a manatee rib fractured by collision with a boat....................79

3-3 Procedure for making epoxy replicas of manatee rib fracture surfaces .................80

3-4 Example of a fracture surface contour used for calculation of fractal dimensional
in crem en t (D *) ..................................................... ................ 8 1

3-5 Fractal dimensional increment (D*) vs. total body length .............. .................. 82

3-7 K inetic energy as a function of boat speed ........................................ ...................83









3-8 Fracture toughness (Kc) vs. fractal dimensional increment (D*) ..........................84

3-9 W hole rib crack size vs. total body length .................................... ............... 84















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

SKELETAL BIOMECHANICS OF THE FLORIDA MANATEE
(Trichechus manatus latirostris)

By

Kari Beth Clifton

December 2005

Chair: Roger L. Reep
Cochair: John J. Mecholsky, Jr.
Major Department: Veterinary Medicine

Watercraft-related mortality of the Florida manatee (Trichechus manatus

latirostris) accounts for 25% of all deaths from 1974-2004, comprising 85% of

anthropogenic-related deaths. Of those, the proportion of deaths due to impact accounts

for 67% of all watercraft fatalities. Reducing this preventable source of mortality is a

high priority in state and federal manatee recovery efforts, which focus primarily on

regulating boating activities. However, with more boats using the waterways, the

potential threat posed by watercraft continues to increase. To establish safe boat speeds

for manatee protection, an estimate of the forces required to fracture their bone is needed.

This study is the first to quantify the biomechanical effects of boat strikes on manatees.

Our goal was to estimate material and structural properties of manatee rib bone. To

measure bone material properties, machined specimens were tested in 3-point flexure.

Strength, toughness, modulus, and work of fracture were determined for three size

classes, by sex. Mean flexural strengths ranged from 62-160 MPa, fracture toughness









from 1.4-2.9 MPa-m1/2, elastic moduli from 4-18 GPa, and work of fracture 3-6 MJ-m-3,

indicating that manatee bone is less strong and tough than other mammalian bone,

including human. There were no differences in mechanical variables between the sexes.

Density and mineral content increased with body size, suggesting that material properties

are likely correlated bone tissue quality, as well as to microstructure. For whole bone

structural tests, ribs were impacted to estimate failure stress and kinetic energy needed to

cause fracture. Fractographic analysis was used to calculate the stress at failure. Strengths

ranged from 106-179 MPa. Strain gauges were used to validate these calculations.

Fractal geometry was used to calculate the fractal dimensional increment (D*). Values

for manatee bone (.08-.22) were at the lower end of the range for ceramics, suggesting

that a low amount of energy is needed to cause fracture. There was no relationship

between D* and age. The ability to resist bone fracture does not appear to increase as

animals grow. Impact energy calculations suggest that boats may inflict fatal injuries to

manatees at speeds of 13-15 mph.














CHAPTER 1
INTRODUCTION

The West Indian manatee (Trichechus manatus) is one of four extant species

belonging to the order Sirenia. It occurs throughout the Caribbean, from the southeastern

United States to northeastern Brazil. Manatees inhabit shallow, coastal estuarine or

riverine waters, feeding primarily on seagrasses and other aquatic vegetation. Their

reliance on shallow-water habitats has resulted in two distinct, geographically isolated

subspecies: the Florida manatee (Trichechus manatus latirostris) is found primarily in

Florida, whereas the Antillean manatee (Trichechus manatus manatus) is found in

disjunct populations in eastern Central America, the Greater Antilles, and northeastern

coastal South America (Marmontel, 1993).

The Florida manatee is listed as endangered by the U.S. Department of the

Interior, and is considered one of the most endangered marine mammals found in U.S.

waters (U.S. Fish and Wildlife Service, 1996). Manatees are protected by state and

federal legislation. Florida provided the first protection in 1893, when a law was passed

prohibiting the killing of manatees (Reynolds and Odell, 1991). The Florida Manatee

Sanctuary Act of 1978 established the entire state of Florida as a manatee refuge, and

mandated the creation of boat speed regulatory zones (Reynolds and Odell, 1991). The

Marine Mammal Protection Act of 1972 established federal guidelines for maintaining

marine ecosystems, including marine mammals (Reynolds and Odell, 1991). The

Endangered Species Act of 1973 mandates the development and implementation of

recovery plans for wildlife identified as threatened or endangered with extinction. The









main objective of the Florida manatee recovery plan, first released by the U.S. Fish and

Wildlife Service in 1989, is to downlist, and to ultimately delist the species when the

population has reached maximum sustainable levels, and threats to the habitat and

mortality are controlled (Eberhardt and O'Shea, 1995). The plan also identifies the

actions needed to meet those criteria. However, the rapid growth of the human population

in Florida, and the continued decline in manatee habitat and increasing numbers of deaths

makes it unlikely that the species will be downlisted.

Sources of manatee mortality have been documented in Florida since 1974

(O'Shea et al., 1985; Ackerman et al., 1995; Wright et al., 1995). Watercraft-related

mortality, caused by propeller wounds or by impact with a boat's hull, accounts for 25%

of all deaths from 1974 through 2004, and comprises 85% of anthropogenic-related

deaths (Florida Fish and Wildlife Conservation Commission, 2005). Wright et al. (1995)

examined necropsy records for 406 manatees killed by boats from 1979 through 1991.

Deaths were broken down into three sub-categories: deaths from propeller wounds; from

impact-related injuries; or from a combination of propeller and impact injuries. The

proportion of impact-related mortality has increased over time, and now accounts for

more than half (66%) of all watercraft-related deaths (Florida Fish and Wildlife

Conservation Commission, 2005). Regardless of cause of death, most manatees bore

scars of healed propeller wounds. Fatal propeller wounds were usually more severe than

non-fatal wounds, and animals killed by propeller cuts rarely suffered skeletal damage.

Conversely, approximately 66% of the carcasses attributed to impact injuries had

fractured and/or luxated ribs. In addition, healed fractures were rarely observed in any









carcasses. These data suggest that while many propeller wounds are sub-lethal, impact

injuries sufficient to inflict skeletal damage are usually fatal.

The number of watercraft-related deaths has increased at a rate of 7% annually

from 1992 to 2004 (Florida Fish and Wildlife Conservation Commission, 2005).

Reducing watercraft-related mortality is identified as a high priority in the manatee

recovery plan. To date, recovery efforts have focused primarily on regulation of boating

activities by establishing speed zones in areas where manatees and boats coexist. The

Florida Fish and Wildlife Conservation Commission (FWC) is responsible for

establishing boat speed regulatory zones in any body of water in Florida where manatee

protection is warranted (Florida DNR, 1989). The Florida Manatee Sanctuary Act of

1978 and subsequent amendments give the FWC the authority to regulate boat speeds and

to limit boating activities in areas essential to manatees. The 1989 "Recommendations to

Improve Boating Safety and Manatee Protection for Florida Waterways" identified 13

key counties and gave them the option to develop site specific speed zones, adopt 300

foot buffer slow speed zones within all inshore/coastal waters, or adopt 30 mph in

channels and 20 mph outside the channels. Predominantly, the counties chose to develop

their own zones, and most were adopted and approved by the state of Florida by the early

1990s.

To designate boat speed zones, the FWC primarily uses manatee aerial survey

data, mortality records, satellite telemetry, and boat use data. Based on these datasets,

speed zones are delimited by visual estimation of the areas used by manatees. Maximum

speed limits are typically established by examining the number and pattern of manatees

that utilize the area and the nature and extent of watercraft usage (e.g., a nearby boat









channel indicates frequent, high speed use) if known. The proposed speed zones are then

reviewed and modified in a series of public and legal hearings. The speed zones that are

ultimately implemented by law are a compromise of the rules originally proposed by the

FWC and the affected parties (e.g., commercial boating industries, recreational boaters).

Thus, creation of boat speed zones is a highly subjective process, not based on any

information pertaining to the biological effects of boat strikes on manatees.

Although boat speed regulatory zones have been in existence since the inception

of the Florida Manatee Sanctuary Act, the increasing human population along the Florida

coastline has resulted in more boats utilizing the waterways. From 1996 through 2004,

the number of watercraft on Florida waters has increased by 3% per year (U.S. Coast

Guard, 2005). Florida is the fastest-growing recreational boating state in the U.S., ranking

number 1 in the number of registered boats. In 2004, there were over 928,000 registered

boats in the state (Florida Department of Motor Vehicles, unpublished data).

Furthermore, new boat and engine designs allow boaters to utilize shallow areas once

inaccessible (Wright et al., 1995). Thus, the potential threat to manatees posed by

watercraft continues to increase, and regulatory efforts to date have not been successful in

reducing the number of watercraft-related deaths. It is of considerable interest, to

conservationists and boaters alike, to identify ways to reduce this major source of

manatee mortality.

In an effort to improve the efficacy of boat speed zones, the second revision of the

U.S. Fish and Wildlife Service's manatee recovery plan calls for research on mechanical

characteristics of manatee bone (U.S. Fish and Wildlife Service, 2000). As discussed

above, two-thirds of manatee deaths attributed to boats are a result of impact with some









part of the vessel. Ribs broken by the impact damage the surrounding soft tissues

severely enough to result in death. Information on the amount of force needed to break

manatee ribs would be valuable in refining boat speed zones to reduce lethal impacts.

Currently, there is no information on the amount of energy needed to break manatee ribs

in impact.

Manatee bone is unique in the animal kingdom. The ribs are solid cortical bone.

They are pachyostotic, osteosclerotic, and appear to have a unique microstructure. It is

these features that likely make manatee bone susceptible to fracture when struck by boats.

The bony tetrapod endoskeleton first appeared some 400 million years ago (Carter and

Beaupre, 2001). Throughout evolution, the basic form was constrained by genetics.

However, nature has allowed a great deal of latitude in the details. Genetic mutations and

also epigenetic and environmental factors have altered the mechanical environment of the

skeleton (Carter and Beaupre, 2001). The result is immense variation in skeletal form and

function, at both the gross structural level and the microscopic level. The remainder of

this chapter presents a brief overview of the tremendous variety in bone tissues found

among the vertebrates. It provides the framework for understanding where the manatee

falls in the spectrum, and why we find it worthy of study.

Skeletogenesis

The bones of the skeleton are formed one of two ways. Intramembranous bone

forms directly within the mesenchyme (Hall, 2005). There are a few exceptions, but

generally the craniofacial bones and cartilages, as well as dentine and alveolar bone, are

all formed intramembranously, and are derived from neural crest. In contrast,

endochondral bone forms by ossification of a pre-existing cartilaginous model. The bones









of the limbs, vertebral column, and the ribs are endochondral in origin. Endochondral

bones are derived from mesoderm.

Dermal bones are derived from the dermis of the skin, and are present in some

members of every vertebrate class except birds (Kent and Miller, 1997). They are

membrane bones, although they are derived from ectoderm or endoderm, unlike

intramembranous bone which originates from mesoderm (Hall, 2005). Meckel's cartilage

and some bones of the chondrocranium are dermal bones. The pectoral girdle in most

vertebrates is comprised of both dermal and endochondral bone (Romer and Parsons,

1977). Other dermal bones in modem vertebrates take the form of scales or plates.

Dermal bones in fishes are comprised of various combinations of lamellar bone (cellular

or acellular), dentin, and an enamel-like substance; all of which may be covered with

epidermis (Kent and Miller, 1997). A few fishes possess bony plates. Some tetrapods

have minute bony scales called osteoderms (Kent and Miller, 1997). Caecilians and some

tropical toads have them. Crocodilians have osteoderms below cornified epidermis

(Romer and Parsons, 1977). Some lizards have osteoderms that fuse with the skull shortly

after birth, and lizards and crocodilians both possess abdominal ribs that are remnants of

ventral bony scales (Romer and Parsons, 1977). The carapace and plastron of turtles are

comprised of bony plates with immobile sutures (Kent and Miller, 1997). Armadillos and

some whales are the only mammals that normally form dermal bone (Romer and Parsons,

1977; Kent and Miller, 1997). The armor of armadillos consists of polygonal bones with

immobile sutures that are covered with epidermal scales.

Heterotopic bones are not part of the endoskeleton proper. They typically form in

fibrous connective tissues and are endochondral, intramembranous, or dermal in origin









(Romer and Parsons, 1977; Kent and Miller, 1997). In mammals, sesamoids are

heterotopic bones that form in tendons at points of friction, such as where they pass over

joints. Sesamoid bones are also common in reptiles (Hall, 1984). Tendons in the legs of

turkeys and pheasants often ossify (Currey, 1984; Kent and Miller, 1997). Kangaroos

have ossified tendons on the mid-ventral surface of the tail (Murray, 1936). The baculum

is a heterotopic bone found in insectivores, bats, rodents, carnivores, and in nonhuman

primates (Romer and Parsons, 1977). Crocodilians have a bony plate in the eyelids, called

the os palpebrae. The os cordis forms in the heart of deer and bovids. Rostral bones are

formed in the snout of some mammals, such as the pig (Romer and Parsons, 1977; Kent

and Miller, 1997). The os falciforme functions as a supernumerary digit in the digging

hand of moles (Romer and Parsons, 1977). In birds, heterotopic bones are found in the

gizzard of some doves and in the syrinx of others. Some lizards possess a cloacal bone in

the ventral wall of the cloaca. One species of bat has a bone in its tongue, and bone is

found in the diaphragm of camels (Kent and Miller, 1997). Some heterotopic elements

are cartilaginous. Cartilage and bone have been reported in the hearts and aortae of

snakes, turtles, crocodiles, alligators, birds, rats, and hamsters (Hall, 1984; Lopez et al.,

2003). Hyaline cartilage is normally found in the aorta of rats, and fibrocartilage is found

in the hearts of dogs and humans (Hall, 1984).

Bone Shape and Macrostructure

Bones are classified into four groups, based on their shape (Marieb, 1995). Long

bones are roughly cylindrical in shape. They are generally longer than they are wide. The

ends of the bones epiphysess) are more expanded than the shaft diaphysiss). Bones of the

limbs are long bones (such as the femur and humerus; as well as smaller bones like the

metacarpals, metatarsals, and phalanges). Short bones such as carpals and tarsals are









somewhat cuboidal in shape. Bones with complicated shapes (such as vertebrae and

bones of the pelvic girdle) are classified as irregular. Flat bones, such as the frontal and

parietal, are comprised of cancellous bone sandwiched between two layers of compact

bone.

At the gross structural level bone is classified into two types based on its density

(Carter and Hayes, 1976; Hayes and Bouxsein, 1997). Compact, or cortical bone ranges

in density from 1.6-2.0 g/cm3, and has a porosity of 5-30%. Cortical bone forms a shell

around the outside of bones, encasing the inner cancellous bone, and constitutes the bulk

of the shaft of long bones (Martin and Burr, 1989). Cancellous, or trabecular bone is

much more porous (30-90% porosity), with density from 0.07-1.0 g/cm3. Cancellous

bone is found in the interior of flat, short, and irregular bones and in the ends of long

bones. It takes the form of needles or plates called trabeculae (Jee, 1988). The pores are

generally filled with bone marrow.

Microscopic Structure of Bone

Bone, whether compact or cancellous, is comprised of two types of bone tissue:

primary bone, of which there are two types- woven and lamellar; and secondary bone

(Jee, 1988). Woven bone is found in sites of initial bone formation, during early stages of

development. It forms directly in mesenchyme (intramembranous bone), or by

ossification of a cartilaginous model (endochondral bone). Woven bone forms rapidly,

but as a result it is structurally inferior to secondary bone (Martin and Burr, 1989).

Collagen fibers are randomly oriented within the matrix, and osteocytes are distributed

irregularly. It is generally low in mineral content, although it has the potential to become

more highly mineralized than lamellar bone (Martin and Burr, 1989). Woven bone can be

formed de novo, with no previous bone or cartilage model (Martin and Burr, 1989). It









forms the callus during fracture repair, and is formed in disease processes that involve the

skeleton. It is also formed on the periosteal surfaces of bones that have been subjected to

excessive strains (Martin et al., 1998).

Primary lamellar bone, whether cortical or cancellous, must be laid down on a

pre-existing surface of either cartilage or bone (Martin and Burr, 1989). Trabeculae are

thickened by deposition of lamellae on woven bone. Compact bone is formed by

continued deposition until the intertrabecular spaces are filled in (Jee, 1988). There are

three major patterns (Jee, 1988) of lamellar arrangement in cortical bone: 1)

Circumferential lamellae surround the internal and external surfaces of the bone (inner

and outer circumferential lamellae, respectively), 2) Haversian systems, and 3) Interstitial

lamellae, fragments of previous Haversian systems and circumferential lamellae that have

been partially resorbed, fill in the gaps between secondary Haversian systems. Primary

Haversian systems, or osteons, are formed by compaction of woven bone, or when blood

vessels on the periosteal surface become incorporated into new bone as a result of

appositional growth. Bone is deposited in concentric rings around the vessel, forming a

vascular channel.

Plexiform bone is another type of primary bone that contains both woven and

primary lamellar bone (Currey, 1960; Enlow and Brown, 1956, 1958; Martin and Burr,

1989; Martin at el., 1998). It forms on either periosteal or endocortical surfaces.

Subperiosteal or subendosteal bone forms buds perpendicular to the bone surface that

grow for a short time, then join laterally, creating vascular spaces. The vascular spaces

are then filled in with primary lamellar bone. The resulting tissue has the appearance of a









brick wall. Plexiform bone is most commonly found on the periosteal surface of the

diaphysis of long bones (Martin and Burr, 1989).

Over time, bone must adjust to variable loading conditions and it must repair

fatigue damage, a result of repetitive loading (Martin at al., 1998). Secondary lamellar

bone is a result of bone remodeling, the replacement of existing bone with new bone. The

secondary Haversian systems of remodeled bone are formed by tandem processes of bone

removal and replacement (Jee, 1988). In cortical bone, groups of cells collectively

referred to as a bone multicellular unit (BMU) first remove existing bone; then new bone

is deposited in concentric lamellae, leaving a canal in the center. The lamellae are

separated from the surrounding interstitial bone by the cement line, a layer of mineralized

matrix devoid of collagen fibers (Jee, 1988; Martin and Burr, 1989). The resulting

structure is a secondary Haversian system. So like primary lamellar bone, secondary

lamellar bone is comprised of parallel sheets of lamellae. However, this bone is formed

more slowly and is more organized than primary bone (Martin et al., 1998). Osteocytes

are more regularly arranged. Within each lamella, the collagen fibers are all parallel, but

the fibers in adjacent lamellae are oriented at 90 angles (other patterns of collagen fiber

arrangement have also been documented). Primary and secondary Haversian systems can

be distinguished from one another by their size and structure (Martin and Burr, 1989).

Primary osteons have a smaller canal, fewer concentric lamellae, and no cement line. The

trabeculae of cancellous bone are also lamellar in structure. In contrast to cortical bone,

trabeculae are remodeled on the surface. The remodeled trabecular packets have a

lamellar construction similar to secondary osteons (Jee, 1988).









Mechanical Properties of Bone Related to Microscopic Structure

Three mechanical properties are typically measured in bone: elastic modulus,

strength, and toughness. The elastic modulus, or the Young's modulus, is a measure of

the stiffness of the material; it is a measure of the relationship between the external load

applied, expressed in terms of stress; and the deformation response, in terms of strain.

The strength of a material is the load capacity at fracture per unit area. For this

calculation, knowledge of the type of loading and geometry of the structure is critical to

proper comparison of strengths among different investigators. Toughness is a measure of

the energy absorbed before fracture and released during fracture. With isotropic,

monolithic materials, there is a direct relationship between the elastic modulus and

toughness. Because of the complex, composite nature of bone, the relationships among

elastic modulus, toughness, and strength may not be the same as with monolithic

materials. Additionally, because bone is a hierarchical structure, these properties may

vary as a function of scale of measurement. This section gives a brief overview of the

mechanical properties of bone as related to its organization.

The mechanical properties of bone are affected by its porosity and degree of

mineralization (Martin and Burr, 1998). Porosity affects bone strength: the more porous

the bone, the weaker it is. Hence cancellous bone is weaker than compact bone. Bone

strength and stiffness increase with increased mineral content. Currey (1969) found that a

small increase in mineralization results in a large change in modulus, a result of the

organization of the mineral crystals. He showed that the optimal mineral content is about

66-67%. Above this, toughness decreases with increased mineralization.

The organization of both cortical and cancellous bone influences its strength,

stiffness, and toughness (Martin et al., 1998). Woven bone is weaker than lamellar bone,









because of its randomly arranged collagen fibers, osteocytes, and mineral crystals (Martin

and Burr, 1998). However, the random arrangement makes it behave as an isotropic

material. This makes it an ideal tissue for structures that must be equally strong in all

directions, such as rapidly growing bones that will soon be remodeled (Wainwright et al.,

1982).

The parallel collagen fibers of lamellar bone make it stronger than woven bone,

but strength is affected by the orientation of the fibers in relation to the direction of the

applied load (Martin and Burr, 1989). Generally, fibers parallel to the applied load are

strongest in tension, and fibers perpendicular to the load are strongest in compression

(Martin and Burr, 1989). Thus, anisotropic bone is stronger and stiffer in the direction of

the predominant orientation of the collagen fibers. In most long bones, the fibers,

lamellae, and osteons are parallel to the long axis. Hence they are stronger in that

direction, but may be significantly weaker when loaded in other directions (Wainwright

et al., 1982). In the transverse plane, plexiform bone is weaker when loaded radially

(perpendicular to lamella) than tangentially (in the plane of the lamella). In the tangential

plane plexiform bone fails near the interlamellar blood vessels. However, Haversian bone

is equally strong in all orientations in the transverse plane, due to the concentric

arrangement of the lamellae. Therefore, osteonal bone is transversely isotropic (Reilly

and Burstein, 1974).

Plexiform bone confers two mechanical advantages on structures (Martin and

Burr, 1989). The rapid increase of the surface area for primary lamellar bone deposition

greatly increases the strength of the bone in a short period of time. Second, adding this









bone on the periosteal surface quickly increases the diameter of the structure; which

increases its rigidity, thereby increasing its resistance to bending.

Remodeled bone (i.e., that containing Haversian systems) is weaker than purely

lamellar bone in compression, tension, shear, and bending (Martin and Burr, 1989). There

are three possible reasons for this (Currey, 1984). First, Haversian bone is younger than

the bone it replaced: therefore, it is less mineralized. Second, the creation of Haversian

canals increases the porosity of the bone. Third, the cement lines surrounding Haversian

systems do not contain collagen, so they are lower in strength than the lamellae. The fact

that cement lines fracture easily imparts another important property to remodeled bone-

toughness (the ability to resist failure). The tougher a material, the more energy required

to break it (Gordon, 1978). The low-strength cement lines dissipate energy and stop

cracks from propagating (Currey, 1984; Martin et al., 1998).

The strength of cancellous bone is affected by the structure of the individual

trabeculae, and by the arrangement of the trabeculae (Martin et al., 1998). Primary

trabeculae may be stronger than remodeled ones, as the cement lines associated with

trabecular packets of remodeled bone decrease its strength. Orientation of the trabeculae

is also important: they are stronger when oriented parallel to applied loads (Martin et al.,

1998). The trabeculae in the ends of long bones are oriented either along the lines of

stress or in support of those that are (Thompson, 1942; Hildebrand, 1988; Martin et al.,

1998). The porosity and trabecular structure of cancellous bone that make it weaker than

cortical bone also make it more compliant, thus making it a good shock absorber (Martin

and Burr, 1989).









Variations in the Microscopic Structure of Vertebrate Bone

Most vertebrates do not possess highly remodeled Haversian bone structure

(Foote, 1916; Enlow and Brown, 1956, 1957, 1958; Currey, 1960). Foote (1916) made a

histological comparison of the diaphysis of 440 femurs (294 nonhuman) encompassing

all tetrapod vertebrate classes. He found that primary lamellar bone was the predominant

type in the 39 species of amphibians examined. A small number had some plexiform

bone, and a larger number had secondary Haversian systems. In general, the cortex was

thin relative to the size of the bone. Among the 34 reptiles examined, primary lamellar

bone predominated in the lizards, while Haversian bone was found in the turtles.

Plexiform bone was found in the alligator and in some turtles. Most turtles follow the

typical vertebrate pattern of long bone development. However, all turtles, whether

terrestrial or aquatic, had medullary cavities densely filled with dense cancellous bone

(Foote, 1916; Rhodin et al., 1981). In contrast, there was no cancellous bone in the

lizards.

Dinosaur bone was more similar to mammalian bone than to reptilian bone (Reid,

1984). Primary osteonal or plexiform bone was the most common types present.

Secondary remodeling did occur, although the number and distribution of Haversian

systems was highly variable. The degree of remodeling varied from none at all to

extensive. Haversian distribution patterns ranged from randomly scattered to discreet

regional concentrations. Areas with high densities of Haversian systems were associated

with sites of muscular attachment, which is corroborated by the presence of Sharpey's

fibers near these regions (Reid, 1984). Areas where tendons insert on bones are subject to

heavy local stresses (Hildebrand, 1988). The diaphyseal medullary cavity was either

empty or contained cancellous bone (Reid, 1984). Some of the trabeculae of primary









cancellous bone were not remodeled, but rather lamellar bone was deposited onto the

surfaces of primary endochondral trabeculae, and the cartilage core remained (Reid,

1984).

Considerable variation in bone structure was noted among the 40 birds examined

by Foote (1916). Although primary lamellar and Haversian bone were present in

numerous species, he considered plexiform bone to be the most characteristic of the

group. Cortical bone thickness varied considerably among species. He calculated that

birds with "empty" bones (i.e., containing air sacs, no marrow) had thinner cortices

relative to bone diameter than those with marrow, or those that had marrow cavities filled

with cancellous bone, of similar size. There is relatively little variation in bone structure

among birds compared to other vertebrate groups (Bellairs and Jenkin, 1960).

The 188 mammals (nonhuman) examined by Foote (1916) were characterized by

a great deal of structural diversity. Overall, secondary Haversian systems were present in

a greater proportion of mammals than in any other class. However, some mammals were

extensively remodeled, while others such as bats experienced little or none. Similarly,

plexiform bone was found more often in mammals than in other classes.

Among mammals, plexiform bone is found most commonly in the artiodactyls

(Foote, 1916; Enlow and Brown, 1958; Martin and Burr, 1989). Manatees and their close

relatives the elephants both have plexiform bone (Foote, 1916; Enlow and Brown, 1958;

Marmontel, 1993). Deer antlers are plexiform, as are some long bones of large-breed

canines, cats, and bears (Enlow and Brown, 1958; Currey, 1960). Plexiform bone has also

been identified in the human mandible, and in the long bones of children who are under

some degree of dietary stress (Martin and Burr, 1989). Foote (1916) found plexiform









bone in many human fetal and juvenile femora. Currey (1960) reported a thin layer of

plexiform bone on the periosteal surface of an adult human femur. Currey generalized

that in many mammals larger than rats, some bones grow by deposition of plexiform

bone at some stage of development.

The extent to which primary bone is replaced by Haversian systems varies

enormously from species to species, as well as with age, sex, and health, both within and

among individuals (Foote, 1916; Currey, 1960). Haversian systems are almost completely

absent from rodents and bats. Most bats possess primary lamellar bone except for the

large Pteropus species, which do have some Haversian systems (Foote, 1916). Cats do

not begin to remodel their bone until after reaching sexual maturity (Currey, 1960).

Foote (1916) indicated the presence of cancellous bone in the femora of a number

of large terrestrial and semi-aquatic mammals, such as the elephant, hippo, white rhino,

tapir, otter, polar bear, bearded seal, and a number of artiodactyls. Wall (1983) also noted

the presence of cancellous bone in the medullary cavities of the white rhino and the

hippo. Bones in the bat wing have little or no medullary cavity and are poorly

mineralized, making them less dense and stiff than most mammalian bone (Swartz and

Watts, 1999).

In contrast to terrestrial mammals, cetaceans have a reduced bone mass and

density (de Buffrenil et al. 1990). The outer cortical layer of the humerus and radius were

found to vary in thickness, but were relatively thin compared to terrestrial vertebrates,

and the medullary cavities were filled with porous spongy bone (Felts and Spurrell, 1965,

1966). The forelimb is not a weight-bearing structure, nor is it used for propulsion, but

rather it functions for maneuvering and attitude control (Felts and Spurrell, 1965, 1966;









Domning and de Buffrenil, 1991). As such, it is exposed to significant tensile and

compressive loads in the medio-lateral plane. The bones are correspondingly broad and

flattened medio-laterally, modified from the tubular form typical of most animals.

Foote (1916) observed that while all bones consist of the same structural units,

they are combined in many different combinations and proportions. Because of this, he

suggested that the lamella be considered the structural unit of bone. Other authors concur

that bone tissues and their organizational patterns form a continuum rather than discreet

categories (Currey, 1960; Moss, 1961; Martin et al., 1998). Moss (1961) stated that

confusion arises when we try to "classify skeletal tissue types in terms applicable to

mammals alone." Many variations from what is considered the "typical" vertebrate bone

structure exemplified by humans have been examined in detail. Some modifications

characterize an entire class, while others are shared by species from more than one group.

Some of these variations are described below.

Three significant deviations from the classical vertebrate bone structure are found

among the fishes. Elasmobranchs (sharks, skates, and rays) have a cartilaginous skeleton

that is not replaced by endochondral bone (Budker, 1971; Compagno, 1999). In most

elasmobranchs, however, the skeleton is strengthened by calcification of the cartilage.

Calcified polygons of hydroxyapatite cover the surface of bones, and the vertebral centra

are partially calcified in the interior (Compagno, 1999). There are three general patterns

of vertebral calcification, and these patterns are so consistent that they have been used to

classify sharks taxonomically (Goodrich, 1930; Budker, 1971). Additionally, the degree

of calcification somewhat reflects age and habitat preference (Compagno, 1999).

Skeletons of older animals are generally more calcified than younger ones. Inshore









demersal sharks possess the most heavily calcified skeletons, while deep-water and some

oceanic and inshore sharks are variably less calcified.

The teleosts, or bony fishes, are unique in that their bone is largely acellular

(Moss, 1961). Acellular bone is distinguished from cellular bone by the absence of

osteocytes. It can be formed by two different mechanisms, through modified osteogenesis

or chondroidal osteogenesis (Moss, 1961). During endochondral ossification, osteoblasts

secrete matrix around themselves and become embedded. However, soon afterward the

nuclei become pycnotic and die. The lacunae are subsequently filled in with bone. In

appositional bone deposition the osteoblasts at the periosteal surface secrete matrix, but

they continually "withdraw" from the surface. As a result, they never become embedded

in the matrix, thereby producing consecutive layers of acellular bone (Moss, 1961). The

process of acellular bone formation is not well understood, but it appears that there is a

buildup of calcium salts inside the pycnotic cells. This "intracytoplasmic mineralization"

leads to the death of the embedded cells (Moss, 1961).The mechanical properties of

acellular bone are unknown (Wainwright et al., 1982; Currey, 1984).

Chondroid bone, produced by a modified form of intramembranous bone

formation, has characteristics of bone and cartilage (Fang and Hall, 1997; Kawakami et

al., 2001; Witten and Hall, 2002). Chondroidal osteogenesis is a process of bone

formation found primarily in fish (Moss, 1961). Cellular and acellular bone are formed

from chondroid. The kype that appears in the Atlantic salmon during breeding season is

composed of chondroid bone (Witten and Hall, 2003). This structure develops rapidly,

and is structurally similar to bone formed under pathological conditions in humans such









as periosteal osteosarcomas. Additionally, there is evidence that bone morphogenetic

proteins are involved in the production of chondroid bone (Kawakami et al., 2001).

Adventitious (secondary) cartilage forms on membrane bone in response to

mechanical stimulation (Fang and Hall, 1997). This type of cartilage is found in skulls of

birds and mammals, however, it does not form in amphibians and it is unclear whether or

not it forms in reptiles. Additionally, amphibians do not form a cartilaginous callus in

response to fracture, and whether or not reptiles do is unknown (Hall, 1984). It is also

unclear whether these groups lack adventitial cartilage because their periostea have a

reduced osteogenic potential compared to birds and mammals, or because the required

mechanical stimulus for its formation is lacking (Hall, 1984).

Reptiles commonly form metaplastic bone, in which chondrocytes are

transformed into osteoblasts and cartilaginous matrix becomes bone (Hall, 1984).

Additionally, metaplastic bone is also formed by chondrification and subsequent

ossification of tendons and ligaments (Reid, 1984). It is possible that reptiles form

metaplastic bone to compensate for the reduced bone- and cartilage- forming capabilities

of the periosteum (Hall, 1984). Metaplastic bone similar to that in reptiles was also found

in dinosaurs, typically in ossified tendons, but occasionally in the periosteal regions of

limb bones (Reid, 1984). Osseous metaplasias have also been documented in a variety of

soft tissues in humans, where the condition is always pathological (Dayoub et al., 2003;

Siomin et al., 2003).

A skeletal adaptation in birds is pneumatization, the inclusion of air sacs in the

bones. Many bones of the skull and postcranial skeleton are hollow and contain air sacs

lined with epithelium, which are continuous with the nasal sacs, tympanic cavities, or the









bronchial system (Bellairs and Jenkin, 1960). The post-cranial skeleton is pneumatized

by extensions of the respiratory air sacs. Typically, an extension of an air sac enters a

bone through a pneumatic foramen. The sac extends into the marrow cavity, and as it

grows around the degenerating trabeculae, it becomes broken up into a network of

branching tubes. The trabeculae and marrow disappear from the bone shaft, but remain in

the epiphyses in small quantities. The branches of the air sacs reunite to form a single sac

that occupies the entire bone cavity (Bellairs and Jenkin, 1960).

Pneumatization lightens the skeleton, and therefore is widely believed to be an

adaptation for flight (Bellairs and Jenkin, 1960). The extent of pneumatization varies

widely among species, but appears to be related to body size and lifestyle. Small birds,

and aquatic birds such as penguins, are poorly pneumatized (Goodrich, 1930; Bellairs and

Jenkin, 1960). In smaller species, the ratio of air sac volume to body surface area is small,

so extensive pneumatization may not be economical. Diving in aquatic, flightless birds

may be hindered by pneumatization. For larger, flighted birds the benefit of

pneumatization may be significant, and this is especially beneficial because their skeleton

is subjected to greater stress during flight (Bellairs and Jenkin, 1960). However, the

cortical bone layer is about the same thickness in non-pneumatized bones of similar size.

To compensate for this, pneumatized bones have a larger diameter than non-pneumatized

bones of similar length, which increases its resistance to bending. Additionally, long

bones of some large birds have extra internal support in the form of struts.

Pneumatization has been documented in animals other than birds. The centra of

some dinosaurs exhibit an unusual remodeled structure in which a few thin sheets of

compact lamellae formed the external shell and a series of internal longitudinal partitions









(Reid, 1984). The internal vertebral spaces were filled with air (Goodrich, 1930). This

structure resulted in a substantial lightening of the centrum (Reid, 1984). Similarly,

elephants have air-filled sinuses in the skull to lighten it (Feldhamer et al., 1999).

In fishes, amphibians, reptiles, and birds, chondrocytes of the epiphyseal

cartilages do not organize into vertical columns as seen in mammals (Bellairs and Jenkin,

1960; Moss, 1961; Rhodin et al., 1981; Hall, 1984). In reptiles and birds the epiphyses

ossify differently from mammals (Bellairs and Jenkin, 1960; Hall, 1984). In small lizards,

an area of epiphyseal cartilage never ossifies. In other lizards, there is no secondary

center of ossification at all in the epiphysis, so ossification proceeds from the diaphysis.

In these animals the articular cartilage functions as the growth plate. Most lizards do have

secondary centers of ossification. However, ossification proceeds laterally, from the

perichondreal surface inward (Hall, 1984). In contrast, the ossification patterns of large

varanid lizards (e.g., Komodo dragon) are nearly identical to mammals. Unlike mammals,

varanids retain the cartilage canals surrounding blood vessels of the epiphyses. Lizards

that retain some cartilage in the epiphyses are capable of indeterminate growth (Hall,

1984). Like some other reptiles, turtles never develop secondary ossification centers in

the epiphyses. As a result these animals are capable of continual endochondral growth

throughout adulthood (Rhodin et al., 1981; Hall, 1984). Birds only have secondary

centers of ossification at the proximal end of the tibia, which may be related to the

invasion of the air sacs (Bellairs and Jenkin, 1960). Presence of secondary ossification

centers might interfere with penetration of the sacs, and blood vessels associated with the

ossification centers influence the arrangement of the chondrocytes.









Pachyostosis is defined as an increase in cross-sectional area of bones by

increasing the thickness of the periosteal cortices, and the absence of a free medullary

cavity (Domning and de Buffrenil, 1991). Osteosclerosis refers to the replacement of

cancellous bone with cortical bone (de Buffrenil et al., 1990; Domning and de Buffrenil,

1991). Increased bone density, mineralization, and reduced porosity are characteristic of

this condition. Domning and de Buffrenil (1991) pointed out that the term pachyostosis is

often, though erroneously, used to refer to both conditions simultaneously. The bones of

some sirenians are, in fact, osteosclerotic without being pachyostotic.

Fawcett (1942a, b) described the compact bone of manatee ribs as structurally less

differentiated than that of other mammals; largely non-lamellar, periosteal bone laid

down in an eccentric pattern, and the periosteum of the ribs continues to lay down new

bone well after the animal has stopped growing. Large amounts of endochondral

(primary) bone remain unabsorbed, and calcified cartilage matrix persists. The non-

lamellar bone described by Fawcett (1942b) was later classified as plexiform bone by

Enlow and Brown (1958). De Buffrenil and Schoevaert (1989) quantified pachyostosis in

the dugong. They found that rather than being uniform, density and mineralization were

greatest in the skull, anterior (cranial-most) vertebrae and ribs, and decreased caudad. A

similar gradient existed in the forelimb, with greatest degree of pachyostosis in the

scapula and humerus, and a lesser degree in the forearm and in the phalanges. In a review

of sirenian phylogeny, Savage (1976) identified pachyostosis as the only truly

autapomorph characteristic among the Sirenia. All sirenians, which first appeared in the

middle Eocene, exhibited pachyostosis to some degree.









Pachyostosis and osteosclerosis are not unique to sirenians. Pachyostosis is found

in many species that are at least partially adapted to the aquatic environment (Wall,

1983). Nopsca (1923) identified the condition in dinosaurs. Pachyostotic bones have been

described in pelicans and penguins (Foote, 1916; Meister, 1962). Walruses have a very

dense rostrum, while the rest of the skull is comprised of spongy bone. This is

hypothesized to aid the animals in diving for food (Kaiser 1965, 1967). De Buffrenil et al.

(1990) described pachyostotic bone in two archaeocetes. Wall (1983) found that limb

bone density was significantly greater in aquatic and semi-aquatic mammals than in

terrestrial mammals. Stein (1989) compared limb bone densities among 15 species of

marsupials and rodents. Fish and Stein (1991) compared the limb bone densities of

several mustelids. Both studies reported that densities of limb bones of semi-aquatic

species were significantly greater than those of terrestrial species. When Fish and Stein

(1991) combined their data with that of Wall (1983), they found that some semi-aquatic

mammals exhibited increased limb bone densities similar to fully aquatic mammals while

others did not differ from terrestrial animals (Fish and Stein, 1991). Pachyostotic bone

has also been found in a wide variety of fossil artiodactyls (Morales et al., 1993).

Pachyostosis is believed to be an adaptation to promote reduced buoyancy in

aquatic and semi-aquatic animals (Nopsca, 1923; Meister, 1962; Kaiser 1965, 1967; Felts

and Spurrell, 1965, 1966; Rhodin et al., 1981; Wall, 1983; de Buffrenil and Schoevaert,

1989; Domning and de Buffrenil, 1991). Bone density in semi-aquatic animals appears to

be related to the extent to which the animals are adapted to the aquatic environment

(Wall, 1983; Stein, 1989; Fish and Stein, 1991). Wall (1983) postulated that selection for

increased limb bone density occurred when total body submergence became a habitual









part of the animals' lifestyle. However, some fully aquatic mammals such as the

cetaceans likely underwent a secondary reduction in bone density as they developed other

methods of buoyancy control related to deep diving.

The occurrence of pachyostotic bone in manatees has been likened to pathological

conditions observed in humans. Fawcett (1942b) described it as resembling Albers-

Schonberg disease, or marble bone disease, an autosomal dominant form of osteopetrosis.

Increased bone density is a primary feature of the sclerosing diseases. However, many

forms of osteopetrosis and osteosclerosis are accompanied by skeletal architectural

deformities and a suite of soft tissue and physiological defects (Shapiro, 1993; Tolar et

al., 2004) that are not present in manatees or other species exhibiting pachyostosis.

Autosomal dominant osteopetrosis lacks the pathologies of other forms. Another

candidate for the cause of pachyostosis is the recently discovered role of the LRP5 gene

in skeletal mineralization. Some mutations of the gene have been shown to result in high

bone mass, accompanied by an increase in bone volume and reduced marrow spaces, but

without any other evident pathologies (Johnson et al., 2002; Koay and Brown, 2005).

Manatees in Perspective

Manatees share a number of developmental features with other marine animals

that are not found in terrestrial vertebrates. Manatees (Fawcett, 1942a, b), cetaceans

(Felts and Spurrell, 1965, 1966), penguins (Meister, 1962), and the marine turtle

Dermochelys (Rhodin et al., 1981) all develop amedullary long bones that retain some

endochondral bone that is not remodeled. In all these species except manatees the

medullary cavity is filled with coarse-cancellous bone; manatees fill the cavity with

compact bone. Additionally, these aquatic animals vascularize their epiphyses both

perichondrally and transphyseally. Some pinnipeds (Versaggi, 1977) and extinct









amphibians (Rhodin et al., 1981) also show similar patterns of development. Manatees

and penguins both have an unusually thick periosteum that adheres tightly to the bone

surface (Fawcett, 1942b; Meister, 1962). In manatees the periosteum of the ribs continues

to lay down new bone well after the animal has stopped growing (Fawcett, 1942b).

Felts and Spurrell (1965, 1966) identified some common structural and

developmental characteristics of cetacean limb bones. They found that although

cetaceans retain the typical mammalian pattern of fusion of ossification centers, the

process is delayed. The zones of cartilage maturation are not as distinct or as organized as

they are in terrestrial mammals. At no time during development are regular columns of

chondrocytes observed near the growth plates, a feature shared with fishes, amphibians,

reptiles, and birds. Secondary ossification of all cetacean limb elements begins after birth.

Furthermore, there is an extended lag time between cartilage calcification and

endochondral ossification. All of these features have been described in manatees

(Fawcett, 1942b).

Sirenians (e.g., manatees and dugongs) and cetaceans (e.g., whales) are the only

fully aquatic marine mammals. As a result of their aquatic lifestyle, they are free of the

mechanical constraints that influence limb bone architecture in animals that spend at least

part of their time on land. In contrast to the cetaceans, manatees and dugongs are the only

herbivorous, fully aquatic mammals (Domning and de Buffrenil, 1990). As herbivores,

they are tied to the shallow waters of the euphotic zone, which has likely hindered them

from taking advantage of the same hydrodynamic mechanisms of buoyancy control

utilized by cetaceans (Domning and de Buffrenil, 1990). Domning and de Buffrenil

(1991) offered reasonable evidence that pachyostosis and osteosclerosis in the manatee









are adaptations to increase skeletal ballast for buoyancy control. Increased bone diameter,

density, mineralization, and compactness all serve to increase the mass of the skeleton to

offset the positive buoyancy of the lungs.

Although increased mineralization will increase the mass of the bones, Currey

(1969) demonstrated that as mineral content rises beyond a certain point, bone becomes

more brittle in static and impact loading. Increasing bone diameter may offset the

reduction in strength somewhat, by increasing flexural stiffness. In manatee ribs,

pachyostosis is accomplished by the addition of plexiform bone, largely to the lateral

surface. Mechanically, this may be advantageous compared to purely lamellar bone.

When a load is applied to plexiform bone, the woven bone will experience less stress than

the lamellar bone due to its lower modulus (Wainwright et al., 1982). Additionally,

remodeling rates appear to be low in manatee ribs, as indicated by the low number of

Haversian systems. This may be because manatees, being aquatic animals, do not

experience loads that influence remodeling to the same degree as terrestrial animals. The

crack-stopping ability of osteons increases bone toughness (Martin et al., 1998), so the

low number of osteons in manatee bone may contribute to its low toughness compared to

other bone.

Most developmental and structural traits of manatee bone are shared with other

animals, a result of phylogenetic relatedness and common evolutionary problems

associated with the return to the marine environment. Other features such as long bones

constructed of solid compact bone likely result from the manatee's unique ecological

niche. Although the material properties of bone types have been studied, manatees

provide a unique combination of structure and function. There are two articles that









qualitatively describe the histology of manatee bone (Fawcett, 1942b; Domning and

Myrick, 1980). Marmontel (1993) qualitatively assessed remodeling rates in a number of

bones. Domning and de Buffrenil, 1991 measured density, mineral content, and porosity

for manatee rib bone for a range of animal sizes. Yan (2002) and Yan et al. (2005)

calculated fracture toughness for rib bone. There is no information on the mechanical

properties of whole bone, or of manatee bone tested in impact. This study is the first to

comprehensively measure mechanical properties of manatee bone.

The overall goal of this project was to quantify the biomechanical effects of boat

strikes on manatees. This was accomplished by integrating basic research data on bone

biomechanics with data from impact tests, which have a more practical application. We

conducted two kinds of tests to measure both material properties and structural properties

of manatee bone. Material properties are those of the tissue itself, independent of whole

bone size or shape. Structural properties are those properties conferred by the size and

shape of whole bones, in addition to the properties of the material. For example, a block-

shaped vertebra will behave differently from a long, slender rib when tested in

compression, even if they are constructed out of the same material. In Chapter 2, the

objective was to conduct 3 point flexure tests using standard test protocols to measure

material properties of manatee rib compact bone. Two hypotheses were tested: 1)

manatee bone is less strong and tough compared to other species when loaded statically,

and 2) the biomechanical properties of manatee bone vary between sexes and among age

classes. We measured strength, Young's modulus, fracture toughness, and work of

fracture for both sexes, in three age classes. Chapter 3 addresses the hypothesis that

forces generated by most watercraft under normal operation are sufficient to inflict fatal






28


skeletal injuries to manatees. Impact tests were used to estimate the energy needed to

fracture whole ribs and the stress at failure. The work presented here provides the first

detailed information about material and structural properties of manatee bone. Data

generated by the studies herein will contribute to our understanding of manatee-boat

interactions, and will be instrumental in shifting the focus to a more objective approach

for establishing regulatory boat speed zones adequate to reduce watercraft-related

mortality.














CHAPTER 2
MATERIAL PROPERTIES OF MANATEE RIB BONE

Introduction

The Florida manatee (Trichechus manatus latirostris) is listed as endangered by

the U.S. Department of the Interior, and is considered one of the most endangered marine

mammals found in U.S. waters (U.S. Fish and Wildlife Service, 1996). Sources of

manatee mortality have been documented in Florida since 1974 (O'Shea et al., 1985;

Ackerman et al., 1995; Wright et al., 1995). Watercraft-related mortality, caused by

propeller wounds or by impact with a boat's hull, accounted for 25% of all deaths from

1974 through 2004, and comprised 85% of human-related deaths. The proportion of

deaths due to impact accounts for more than half (55%) of all watercraft-related deaths

(Wright et al., 1995). Animals killed by propeller wounds rarely suffer skeletal damage.

Conversely, approximately 66% of the carcasses attributed to hull impact injuries have

fractured and/or luxated ribs (Wright et al., 1995). Furthermore, healed fractures are

observed infrequently. This suggests that while many propeller wounds are sub-lethal,

hull impact injuries sufficient to inflict skeletal damage are usually fatal.

The number of documented watercraft-related deaths increased at a rate of 7%

annually from 1992 through 2004 (Florida Fish and Wildlife Conservation Commission,

2003). Attempts at reducing watercraft-related mortality have focused primarily on

regulation of boating activities, such as creating speed zones in areas where manatees and

boats coexist. However, the increasing human population along the Florida coastline has

resulted in more boats utilizing the waterways. From 1996 through 2004, the number of









watercraft on Florida waters increased by 3% per year (U.S. Coast Guard, 2005). In 2004,

with over 928,000 registered boats in the state; Florida moved ahead of California and

Michigan to be ranked the number one recreational boating state in the U.S. (Florida

Department of Highway Safety and Motor Vehicles, unpublished data; U.S. Coast Guard,

2005). Additionally, new boat and engine designs allow boaters to utilize shallow areas

once inaccessible. Thus, the potential threat to manatees posed by watercraft continues to

increase. Creation of boat speed zones is a highly subjective process, not based on any

information pertaining to the biological effects of boat strikes on manatees. A major goal

of the U.S. Fish and Wildlife Service's manatee recovery plan is to refine the boat speed

zones rules to reduce boat strike deaths (U.S. Fish and Wildlife Service, 2000). Toward

this goal, the plan calls for research on mechanical characteristics of manatee bone. The

data from this study will be useful in shifting the focus to a more objective approach for

establishing regulatory boat speed zones adequate to reduce watercraft-related mortality.

Manatee bone is unusual compared to that of other marine mammals. Compared

to terrestrial mammals, the general trend in marine mammals has been a reduction of

bone mass and density (de Buffrenil et al., 1990). In contrast to most marine mammals,

the manatee skeleton exhibits pachyostosis, characterized by thickening of bone tissue,

replacement of cancellous bone with compact bone, and absence of a free medullary

cavity (Fawcett, 1942b; de Buffrenil et al., 1990; Domning and de Buffrenil, 1991). The

mechanical properties of bone have been well studied for humans and some domestic

animals (Wainwright et al., 1982; Currey, 2002); however, mechanical studies on the

bones of marine mammals have been few. Mechanical property data are available for fin

whale (Balaenoptera physalus), dense-beaked whale (Mesoplodon densirostris), narwhal









(Monodon monocceros), polar bear (Ursus maritimus), walrus (Odobenus rosmarus), and

dugong (Dugong dugon) (Currey, 1979a; Currey and Brear, 1990; Evans et al., 1990;

Brear et al., 1993; Currey et al., 1994; Zioupos et al., 1997; Currey, 1999).

Some information on the composition of manatee bone exists (Fawcett, 1942a, b;

Domning and Myrick, 1980; Domning and de Buffrenil, 1991; Marmontel, 1993).

Fawcett (1942b) and Domning and Myrick (1980) provide qualitative reports on the

microstructure of rib bone. Domining and de Buffrenil (1991) measured density, mineral

content, and porosity (measured as its inverse, compactness). They found that mineral

content and density increased with body size through the subadults, and leveled off in

adults. Additionally, de Buffrenil and Schoevaert (1989) examined bone density in the

dugong, a close relative of the manatee. They found a bone density gradient that

increased from craniad to caudad, and from dorsad to ventrad. In developing an aging

technique for manatees, Marmontel (1993) examined histological sections of various

bones. She reported differences in the number of osteons in males and females of similar

age, and inferred that there may be differences in remodeling rate between the sexes. We

recently reported on the use of fractography and Chevron notch beam tests to measure

fracture toughness (Yan, 2002; Yan et al., 2005). The present study is the first

comprehensive measure of the mechanical properties of manatee bone.

The first step in quantifying the biomechanical effects of boat strikes on manatees

was to measure some material properties of manatee rib compact bone. Based on the

limited information on manatee bone available in the literature, we hypothesized that

manatee bone is less strong and tough compared to other species when loaded statically,

and that the material properties of manatee bone vary between sexes and among age









classes. The objectives were to calculate flexural strength, toughness, Young's modulus,

and work of fracture in bending for rib bone, and to measure density and mineral content

for manatees of both sexes in three age classes.

Materials and Methods

Bone Sample Collection

Bone tissues were obtained from the Florida Fish and Wildlife Conservation

Commission's Marine Mammal Pathobiology Laboratory. Collection and use of tissues

for research was conducted under U.S. Fish and Wildlife Service permit #MA067116-0

issued to the University of Florida. Animals of both sexes in three age classes (calf,

subadult, adult) were tested. Because it is difficult to age manatees, there is little age data

available. By convention, total body length (cm) is used as a proxy for age (O'Shea et al,

1985). Calves were defined as < 175 cm total body length, subadults were 176-275 cm

total body length, and adults > 275 cm total body length (O'Shea et al., 1985). Three ribs

were obtained from fresh carcasses at the time of necropsy. One rib was taken from the

cranial, middle, and caudal thoracic regions (Figure 2-1). Ribs were taken from the same

positions whenever possible, and stored frozen in plastic bags until machining. Freezing

and thawing of bone does not significantly alter the mechanical properties of bone

(Sedlin and Hirsch, 1966; Currey, 1988b; Pelker et al., 1984). They were brought to room

temperature for machining and testing.

Specimen Preparation

A band saw was used to rough cut 60 mm segments from the proximal and middle

1/3 sections of the ribs under a constant spray of water, to keep the bone from heating and

to keep it wet. Specimens with dimensions of 50 x 3 x 3 mm were machined with a

diamond-edged band on a wet saw (Exakt Technologies, Hamburg, Germany), with the









long axis of the test pieces parallel to the long axis of the rib. Dimensions were verified

with digital calipers (Mitutoyo, Tokyo, Japan) to the nearest 0.01 mm. Machined

specimens were stored in physiological buffered saline until testing, and they were tested

wet.

Material Properties Testing

Three point bending tests were performed with a Mini-Bionix materials testing

machine following ASTM standard D790M-92 for flexural testing of plastics (MTS

Systems, Eden Prairie, MN; ASTM, 2001). For each specimen, a 10-cycle, 10-Newton

preload was applied, then loaded to failure using a standard 3-point bending fixture at a

displacement rate of Imm/min. The load was applied to the surface that corresponded to

the lateral surface of the rib (i.e., tested in the transverse orientation). Load and

displacement were recorded at 5 Hz, and output to a personal computer for analysis.

Mechanical variables measured were flexural strength, Young's modulus (i.e.,

elastic modulus) in bending, fracture toughness, and work of fracture. The flexural

strength of a material is the load required to break a piece of the material in bending

(Wainwright et al., 1982). Flexural strength in 3-point bending was calculated as

3FL
S-= (2-1)
2BD2

where F is peak load (Newtons) at failure, 1 is the distance (meters) between the supports

of the bending fixture, B is the specimen width (m), and D is the specimen depth (m). A

correction factor was applied to the strength calculation if the specimen fractured

somewhere other than opposite the point at which the load was applied (Riley et al.,

1999).The correction factor was calculated as









2((L/2)- x) (2-2)
Ocorr = a (2-2)
L

where Gcorr is the corrected flexural strength, L is the original length of the test specimen,

and x is the length of the shorter of the two halves after testing, and a is the flexural

strength from Eq. 2-1. Young's modulus is an indication of how easily the material

strains elastically under a given stress; it is a measure of the stiffness of the material

(Jackson, 1992). Young's modulus in bending was calculated as


E (F/x) (2-3)
481

where F/x is the slope of the load-displacement curve, 1 is the distance between the

supports of the bending fixture (m), and I is the moment of inertia, calculated as

BD3
I= (2-4)
12

Yan et al. (2005) showed that, unlike typical bone, manatee bone behaves more as

a ceramic. Therefore, fractographic analysis and fracture mechanics can be used to

calculate fracture toughness. Fracture toughness is the ability to resist fracture, measured

as the critical stress intensity factor (Kc). This is an estimate of the amount of energy

required to propagate a macrocrack that leads to fracture (Mecholsky, 1996); it is a

measure of the behavior in the pre-yield region. The critical stress intensity factor was

calculated as

Kc = Yo(c)1/2 (2-5)

where Y is a geometric factor based on crack shape and loading conditions, a is flexural

strength, and c is the crack size (m2). Crack size is calculated as

c = (a.b)1/2 (2-6)









where a and b are crack depth and half-width (m), respectively, as measured from the

fracture surface with a light microscope. A correction factor was applied if the failure did

not originate from the surface of the specimen (Mecholsky, 1993). Scanning electron

micrographs were taken of some fracture surfaces for detailed examination and validation

of light microscope measurements. The work of fracture is another measure of energy.

Like fracture toughness, work of fracture is that energy required to propagate a crack

through a specimen (Aksel and Warren, 2003). It is calculated from the area under the

stress-strain curve; it is a measure of the behavior in both the pre- and post-yield regions.

For materials that have little or no post-yield deformation, work of fracture and fracture

essentially measure the same property, i.e., toughness. However, for materials that have

post-yield deformation, or exhibit some inelasticity in the pre-yield region, the two

variables may be different. For those materials, work of fracture gives a more complete

picture of the behavior. Work of fracture was calculated [cf Appendix B] for specimens

tested in 3-point flexure as

3W
WOF = (2-7)
2BD2

where W is the area under the load displacement curve (N m).

Bone Density and Mineral Content Determination

Extra bone from the middle sampling region of each rib was used to measure

Young's modulus with ultrasound to compare to calculated values, and to measure

density and mineral content. Since calculation of Young's modulus from the load-

displacement curve is prone to error, ultrasound was used to as a check for the calculated

modulus (Vincent, 1990). Specimens were defatted with ammonia, dried at 950 for 48 h,

and weighed to the nearest 0.001 g. Density was determined using an Accupyc 1330









helium pycnometer (Micromeritics Instrument Corp., Norcross, GA). Ultrasound

measurements were made with a Nuson Ultrasan (Nuson, Boalsburg, PA). Samples were

then incinerated at 8000C for 10 h, and weighed to the nearest 0.001 g. Mineral content

was expressed as a percent of ash weight over dry weight (Kaufman and Siffert, 2001).

Statistical Analyses

Statistical analyses were conducted both within and among animals. For each

individual, specimens within each sample site were compared (Figure 2-1B) to look for

differences within sample sites. Next, specimens for each sample site were pooled to

compare among sample sites within individuals. To examine ontogenetic changes in

material properties, specimens were pooled across sites to calculate means by animal. For

each variable, animal means were regressed against total body length to examine

ontogenetic patterns. Analysis of covariance (ANCOVA) was performed, with length as

the covariate, to look for differences between the sexes (Zar, 1996). For all tests, results

were considered significant if p<0.05. Analyses were performed with Statistical Analysis

Systems using Proc GLM (SAS Institute, Inc., 1989).

Results

Six males and six females were tested for the subadult and adult age classes

(Table 2-1). Tissues were obtained from only two calves. Owing to their small size, it is

difficult to recover calf carcasses in fresh condition. The ribs of the smallest calf (total

body length 102 cm, a newborn calf) were too small to obtain more than one or two

specimens each. Additionally, some ribs of the smaller subadults were too small to obtain

samples from both the proximal and middle regions. For these animals, only the proximal

region was taken.









Within-Animal Comparisons

A total of 3,220 specimens from 26 animals were tested in 3-point bending.

Fracture toughness was determined for 1,650 of those. At each sample site, no significant

differences were found among samples, indicating that the bone at each sample site was

homogeneous. Specimens for each sample site were pooled Analyses comparing

sampling locations within animals did not reveal any consistent patterns, either cranio-

caudal or dorso-ventral, for any variable. Therefore, specimens were pooled for each

animal for analyses by sex and total body length (a proxy for age). Mean values for each

variable, by animal, are reported in Table 2-1.

Material Properties Tests

Figures 2-2 to 2-4 and 2-6 show the relationship of each mechanical variable to

total body length. For each mechanical variable, ANCOVA indicated no difference

between the sexes. Therefore, sexes were pooled for regression analyses of variables

against total body length. For all variables, linear regression did not adequately describe

the relationships between variables and total body length. Rather, the relationships were

curvilinear; a second order polynomial regression was the best fit, and highly significant,

in all cases. Flexural strength increased with body length up to 267 cm, then reached a

plateau (R2 = 0.79, p<0.0001, Figure 2-2). Fracture toughness exhibited a similar pattern

(R2 = 0.73, p<0.0001, Figure 2-3).

Work of fracture increased with body size through the subadult size class, just as

flexural strength and toughness (R2 = 0.50, p = 0.0004, Figure 2-4). Unlike the other

variables, work of fracture showed a strong decline with increasing body length for

animals 265 cm total length or greater. Although there was a large amount of scatter

among the subadults, the value for the smallest calf MEC0033 was much less than the









others. The regression was run minus this animal to determine if it was responsible for

the parabolic shape; it was not. Fracture energy, as measured by work of fracture, was

greatest in animals of intermediate body length, about 200-267cm total length (i.e, the

large subadults).

Young's modulus showed a different pattern, increasing with body length over the

entire range of body sizes (R2 = 0.86, p < 0.0001, Figure 2-5). Ultrasound was used to

measure modulus of 13 bone samples from 11 animals. On average, the ultrasound

modulus was 1 GPa greater than that calculated from the load-displacement curves. The

good agreement between the ultrasound results and our calculated values indicates that

our modulus values are reasonably accurate. Currey (2004) has shown that, in general,

Young's modulus and flexural strength are highly correlated (R2 = 0.86). Our data

reflected the same positive correlation, although the relationship was not as close (R2

0.67, Figure 2-6). Figure 2-7 shows some example load-displacement curves for the small

sample specimens tested here. Representative curves for each size class are given. To

facilitate comparison among age classes, the values for ultimate load, Young's modulus,

and work of fracture are given for each curve.

Bone Density and Mineral Content

Extra bone was available to measure density and mineral content for 20 animals

(Table 2-1). Mean density by animal was 2.17 g/cm3 ( 0.04 SD, range 2.10-2.22 g/cm3),

and mean mineral content was 69% ( 2% SD, range 64-71%). Density showed a weak

positive correlation with total length, however, the correlation was not significant (R2

0.19, p = 0.0517, Figure 2-8). Percent mineral content also correlated weakly to total

body length, although the relationship was significant (R2 = 0.27, p = 0.0198, Figure 2-9)

In contrast, density showed a strong, positive correlation with elastic modulus with the









removal of one outlier (R2 = 0.64, p<0.0001, Figure 2-10). Percent mineral content also

showed a strong correlation with modulus (R2 = 0.61, p<0.0001, Figure 2-11). Our results

show that while modulus was highly correlated to both bone tissue density and mineral

content, only mineral content increased with total body length.

Discussion

Reducing watercraft-related manatee mortality is identified as a Level 1 high

priority in the Florida Manatee Recovery Plan of the U.S. Fish and Wildlife Service (U.S.

Fish and Wildlife Service, 2000). The recovery plan calls for research on mechanical

characteristics of bone to obtain a better understanding of the effects of watercraft-related

impacts To meet this objective, the goal of this study was to measure some material

properties of manatee rib compact bone. Based on the limited information available in the

literature, we hypothesized that manatee bone is less strong and tough compared to other

species when loaded statically, and that the material properties of manatee bone vary

between sexes and with the size of the animal.

The Material Properties of Manatee Bone

The main finding of this study is that manatee bone is on average less strong and

tough than other mammals. Mean flexural strength by animal ranged from 62-160 MPa.

In comparison, typical flexural strengths for human and bovine bone specimens tested in

3-point bending are 209 MPa and 224 MPa, respectively (Martin et al., 1998). Flexural

strengths have been reported for horse radius (154-249 MPa), black bear tibia (211-328

MPa), and ox femur (227-264 MPa) (Schryver, 1978; Wainwright et al., 1982; Reilly and

Currey, 1999; Batson et al., 2000; Harvey and Donahue, 2004). We would expect

normally manatee bone to have a bending strength comparable to that of other animals

with similar organization. However, we predicted that out values would be lower than









values for other species, based on the high mineral content reported by others. We did

find higher than average mineral content, is likely the reason for the low strengths (see

below).

The Young's modulus, or the modulus of elasticity, is the measure of the stiffness

of a material. Mean modulus ranged from 4-18 GPa, which was similar to that for equine

metacarpals (14 GPa), horse radius (14-21 GPa), human femur (15 GPa), cow femur (20

GPa), and black bear tibia (16-30 GPa) (Martin et al., 1998; Zioupos and Currey, 1998;

Currey, 2001; Les et al., 2002; Harvey and Donahue, 2004).

Fracture toughness is the ability to resist fracture, one measure of which is the

critical stress intensity factor (Kc). Toughness is an important property because a tough

bone is more resistant to fracture, and this is most important when bone is impacted

(Wainwright et al., 1982; Martin et al., 1998). Because bone is viscoelastic, it will behave

differently when loaded at different rates (Martin et al., 1998). Generally, the more

rapidly bone is loaded, the greater the modulus, and the less energy it is able to absorb.

The bone mineral content and density previously described in manatees led us to predict

that manatee bone may be on average less tough than other bone (Domning and de

Buffrenil, 1991). As predicted, data reported here indicate that manatee bone (1.4-2.9

MPa-m1/2) is generally less tough than other bone such as human tibia (2.2-5.7

MPa-m1/2), bovine femur and tibia (2.2-5.5 MPa-m1/2), and equine metacarpus (4.4-7.5

MPa-m1/2) (Martin et al., 1998; Malik et al., 2003). The ontogenetic pattern in fracture

toughness mirrored the pattern seen in flexural strength. Since fracture toughness is

proportional to both the flexural strength and the flaw size (Equation 2-5), our results

suggest that the flaw size is similar over the range of body sizes tested here. A plot of









mean crack size by animal vs. total body length indicated that there was no difference in

crack size in relation to body size (R2 = 0.0009, p = 0.8886, Figure 2-12).

For both strength and fracture toughness, the regressions may indicate a slight

decline for the largest animals. It appeared as though the largest animal, MSW0245, may

be responsible for this. This animal had been aged at 25, and many of the test specimens

had a qualitatively different look from those of other adults. It may be that this possible

decline in strength and toughness in the largest animals is a result of age-related

decreases in material properties similar to those well-documented in humans. However,

without more ages available for the other larger animals, and without more animals of

advanced age, this cannot be determined conclusively.

Another measure of toughness is work of fracture. The work of fracture for

manatee bone varied from 3.0-6.0 MJ/m3, which is considerably lower than the 28-46

MJ/m3 for human femur, but well within the range of values reported for other mammals

(Currey and Pond, 1989). Harvey and Donahue (2004) recorded values of 3.2-9.4 MJ/m3

for black bear tibia. Currey (2002) reported works of fracture between 0.3-15.6 MJ/m3

for a wide variety of terrestrial and marine animals. Typically, young bone has a low

work of fracture that first increases with age, then declines. Manatee bone followed the

same pattern.

Mean manatee rib bone density of 2.17 g/cm3 was greater than that reported for

humans, which is typically 1.7-1.9 g/cm3, but equal to that of cow femur (2.1 g/cm3) and

baboon femur (1.90-2.17 g/cm3) (Yeni et al., 1998; Phelps et al., 2000; Currey, 2002). De

Buffrenil and Schoevaert (1989) reported densities of 1.79-2.10 for ribs from three

dugongs (2 subadults and 1 adult), a close relative of manatees who also exhibit









pachyostosis of the skeleton. In similar fashion, mineral content for the manatee bone

averaged 69%, greater than what is commonly reported for other species. With one

exception, all values were between 68-71%. Even greater values of up to 75% mineral

content (range 67-75%) were reported for dugong ribs (de Buffrenil and Schoevaert,

1989). According to Martin et al. (1998), mineral content is typically 65 + 3% for cortical

bone. Currey and Butler (1975) reported mineral contents of 60-66% for human femora

aged 2-48 years, and 64-67% for black bear tibias from animals 2-14 years (Harvey and

Donahue, 2004).

It is important to clearly define the terms density and mineral content. Density can

be measured two ways, as the weight per unit volume of only the constituent material, or

of the volume inclusive of the spaces, or voids (Martin et al., 1998). The latter is termed

apparent density. Thus, bone that is highly porous will have low apparent density. If,

however, just the bone tissue is considered, the true tissue density of the porous sample

would be just as high as a volume completely filled with the same material. The

pycnometer used for this study forces helium gas, under pressure, into the sample

chamber to determine volume. In concept, this fills the sample voids if the permeability is

adequate, thereby measuring true density. In similar fashion, mineral content can be

construed as either the amount of mineral per unit sample volume (i.e., volumetric

mineralization), or per volume of bone matrix, exclusive of voids (i.e., specific

mineralization). Ashing bone, as done here, is a measure of specific mineralization. A

more complete picture of manatee bone density and mineralization patterns may have

been drawn if we had also calculated apparent density and volumetric mineralization.









We can summarize our findings about manatees bone as follows: flexural strength

was less than for other species; both stiffness and fracture toughness were roughly equal

to or less than other species; and work of fracture was similar to values reported for a

number of species. Material properties of manatee cortical bone increased with total body

length up to about 267 cm, then leveled off, or declined as for work of fracture, with

increasing body size. We expected this pattern to be related to an increase in density

and/or mineral content with increasing body length. This appears to be what happened, as

density and mineral content increased with body size. However, the relationship between

density and body size was not significant. This is likely due to two things: that the

variation in density was slight, and that comparatively our values were at the high end of

the spectrum of values reported in the literature. Mineral content showed greater

variation, and increased significantly with body size. Importantly, mineral content was

higher than that seen in most mammalian bone. The high mineral contents, coupled with

densities at the high end of published values may explain the patterns of the mechanical

variables to total body length in animals greater than 265 cm. To better understand the

material behavior of manatee rib bone, additional factors need to be evaluated, such as its

plexiform structure, collagen content/quality, and its reportedly low remodeling rate

(Marmontel, 1993).

We can look to Domning and de Buffrenil (1991) to gain insight into our findings.

The authors determined density, mineral content, and compactness of manatee rib for 12

animals. Compactness is the percent of the total cross sectional area that is occupied by

bone; it is the inverse of porosity. Total body lengths were missing for the three largest

animals. Of the nine animals of known body length, only one was an adult (324 cm TL);









the others were calves and subadults (141-254 cm TL). It is important to note that the

authors calculated apparent density, whereas we calculated the material density. They

reported densities of 1.59-2.07 g/cm3 (mean 1.85 g/cm3), lower than our densities.

However, mineral content for those animals was 67-70%, essentially identical to our

data. There was a strong correlation between mineral content and total length, and

apparent density and total body length. However, that was related to the fact that those

animals were primarily calves and subadults (i.e, less than 275 cm TL); we saw a similar

pattern in our data over the same size range. The authors noted that there seemed to be no

relation of mineral content and density to body size greater than 275 cm. Compactness

varied from 84-94%, which translates to 6-16% porosity. Porosity declined with

increasing body length (R2 = 0.53). The authors concluded that mineralization was not

dependent on size, and that the increase in apparent density with size was due to the

decrease in porosity. Considering both datasets together, we can draw some conclusions

about the relationship between mineral content and material properties. Manatee bone

tissue density, consistently high, varies little over a wide range of body sizes. The

increase in apparent density and mineral content with body size, along with a decrease in

porosity, indicates an increase in bone tissue volume. Furthermore, there appears to be

little change in the quality of the material. As the animals grow, the interstitial spaces are

filled in with more bone. Some additional mineral is added to the existing bone.

However, because the existing bone is already so highly mineralized, the resulting bone

has reduced strength and toughness. Thus, the ontogenetic patterns in mechanical

variables appears to be related to the amount of material present, but the overall low

values for strength and toughness are related to the quality of the material.









The Effect of Mineral Content on the Material Properties of Bone

Bending strength of compact bone is primarily determined by the Young's

modulus (Currey, 1999). In turn, the primary determinant of modulus is mineral content

and to a lesser extent, porosity and bone architecture (Schaffler and Burr, 1988; Currey,

1998, 2003, 2004). The effect of mineral content on bone material properties has been

well studied, both in terms of changes with age (Currey and Butler, 1975; Currey, 1979b)

and interspecific variation (Currey, 1969, 1988, 1999, 2002, 2004a, b). Dramatic changes

in material properties result from a relatively small change in mineral content. Generally,

for mineral contents of 63-68%, static strength, toughness, and modulus increase as

mineral is deposited in the collagen. However, above 68%, strength may actually

decrease as mineral content rises, due to the fact that the hydroxyapatite crystals coalesce

and are therefore unable to stop cracks from propagating (Currey, 1969). Additionally,

the impact strength of highly mineralized bone is less than its static strength, and impact

strength drops more precipitously than static strength as mineralization increases (Currey,

1969). A decrease in toughness is generally associated with an increase in mineral

content (Currey, 1969; Malik et al., 2003). Specifically, toughness, as measured by work

of fracture, increases with mineral content up to 66-67%, and then declines precipitously

as mineral content continues to rise (Currey, 1969). Our mineral content ranged from 64-

71% (mean 69%). This may be why we saw the decrease in strength and toughness for

the largest manatees.

Degree of bone mineralization is known to increase with age in humans, and there

is a corresponding increase in strength and stiffness, and a decrease in absorption of

impact energy (Currey, 1979b, 1984, 2002; Currey et al., 1996). Although this pattern is

well documented in humans, this is not the case for all animals. Harvey and Donahue









(2004, 2005) found that although mineral content, hence modulus, increased in black bear

tibia from animals aged 2-14 years, there was no decline in strength, toughness, or

porosity with age. Furthermore, they found no reduction in bone mechanical properties

attributable to long periods of hibernation. Phelps et al. (2000) examined age-related

changes in mechanical properties of the baboon femur, which is known to resemble

human bone in microstructure. Although significant decreases in fracture toughness were

noted for adult animals aged 6-27 years, they found no variation in bone density, mineral

content, or porosity with age.

Other Factors Influencing Bone Material Properties

Material properties of cortical bone are determined in large part by mineral

content, but also by porosity, histological architecture, and amount and quality of

collagen (Martin et al., 1998). One deficit of this study is that we did not include a

histological examination of our bone. This would have allowed us to measure porosity,

and both quantitatively and qualitatively describe the type and arrangement of bone and

collagen. Yan (2002) measured porosity from scanning electron micrographs for a few

rib bone samples from one calf and four subadult manatees. Porosity varied from 3-44%,

and was strongly correlated to flexural strength (R2 = 0.88) and fracture toughness (R2

0.89). However, porosity varied as much as 26% within the same rib, and was not

significantly correlated to total body length. Although the very highest porosity (44%)

was for the neonate MEC0033, some values for other animals were still somewhat greater

than those reported by others, including 1-10% for both human femur and baboon femur,

4% for cow femur, and 2-7% for black bear tibia (Schaffler and Burr, 1988; Currey and

Pond, 1989; Phelps et al., 2000; Harvey and Donahue, 2004). In a second set of

measurements, Yan (2005) reported an average porosity of 7% for manatee rib and 4%









for cow femur, values more in line with other published values. The average porosity for

manatee rib reported by Domning and de Buffrenil (1991) was 10%. De Buffrenil and

Schoevaert (1989) reported porosities of 16% for 1 subadult, to <1% for the adult.

Increasingly, researchers are examining the influence of collagen on the

mechanical properties of bone, as variations in structure, mineral content, and density are

not sufficiently able to explain the decline in properties with age. Wang et al. (2001)

found that collagen denaturation resulted in decreased strength and toughness of the

human femur. They also found that modulus was constant irrespective of the degree of

collagen degradation. In a related study, Wang et al. (2002) examined collagen content

and quality in human femora aged 19-89 years. They found that age-related changes in

collagen were associated with decreased strength, toughness, and work to fracture of

bone. Moreover, the deterioration of collagen with age was significantly correlated to the

post-yield deformation. Phelps et al. (2000) found that the organic fraction showed a

significant, positive correlation with fracture toughness for baboon femora. The exact

mechanism by which changes in collagen alter mechanical properties is not known.

However, it has been speculated that changes in the amount of collagen or to either the

inter- or intra-molecular stability of the collagen are responsible (Zioupos et al., 1999;

Burr, 2002). Age-related changes in cross-linking have been shown to alter collagen

structure, which in turn may alter its hydration (Lucchinetti, 2001; Wang et al., 2002). To

date, studies on the effects of collagen cross linking have been contradictory. Regardless

of the specific mechanism, there is growing evidence that collagen degradation with age

is primarily responsible for the post-yield decrease in toughness and work of fracture, and

that these declines are independent of mineral content or density (Lucchinetti, 2001; Burr,









2002; Wang et al., 2001). In support of this, we found no relationship between work of

fracture and mineral content for manatee rib bone (R2 = 0.02).

Bone type and degree of remodeling influence material properties. Mineral

content for mature primary lamellar bone is 67% versus 63% for Haversian bone (Currey,

1969; Vincent, 1990). Bone remodeling replaces mature, mineralized bone with material

that is initially less mineralized. Haversian bone has decreased strength and modulus

compared to lamellar bone, but increased toughness. Remodeling may decrease strength

by as much as 20% (Martin et al., 1998). That the mineral content of manatee bone is

high relative to most lamellar bone may be indicative of low remodeling rates. Possible

evidence for this may be from Fawcett (1942b), who conducted the only detailed

examination of manatee bone histology to date. He described what is now known as

plexiform bone, as well as endochondral (primary) bone that remained unremodeled, and

some persistent calcified cartilage matrix. He did not quantify the extent of remodeling,

only remarking that Haversian systems are predominantly located in the innermost third

of the cross section, and their numbers diminish rapidly toward the outer edge. He also

did not comment on whether the osteons were primary or secondary. Furthermore, the

descriptive histology was based on one rib from one fresh adult animal, one dried

humerus from a museum, and two preserved fetuses. Based on the limited information

provided about the adult, its true age is ambiguous, and it was likely not skeletally

mature.

Marmontel (1993) examined histological sections of various manatee bones,

including ribs. She estimated that bone resorption increased significantly with total body

length. In addition, females had significantly greater numbers of Haversian systems than









males. Her work suggested that there are age differences in manatee bone structure and

composition that may directly influence its mechanical behavior. This pattern is different

from what is seen in humans. It is generally thought that remodeling slows in older

individuals, resulting in an increase in the amount of older, more mineralized bone.

Hence, the modulus increases, and this contributes to the elevated susceptibility to

fracture with age (Currey, 1979b). However, there is evidence to demonstrate that this is

not true. Bone turnover has been found to increase in the elderly human population,

compared to middle-aged adults (Gallagher et al., 1998; Gundberg, 2002). The scenario

for manatees may be similar: an increase in remodeling with age may be partly

responsible for the slight decline in strength seen in the largest individuals, as Haversian

bone is weaker than lamellar. Similarly, sex differences in remodeling might result in

significant variation in mechanical properties between males and females. In contrast to

Marmontel (1993), we did not find any significant differences between the sexes for any

mechanical variable. Histological examination of manatee bone will tell us whether or

not remodeling differences exist.

Manatee rib bone is composed of plexiform bone, which is a mix of primary

lamellar and woven bone. Preliminary observations of undecalcified histological sections

show that, in addition to plexiform bone, there appears to be a central area of woven

bone. This may be what Fawcett (1942b) referred to as disorganized, primary

endochondral bone. Woven bone, although less organized than lamellar, has the potential

to become more mineralized (Currey, 1998; Martin et al., 1998). For manatees, the

preponderance of woven bone, in conjunction with a low remodeling rate, likely accounts

for the elevated bone density and high mineral content.









Summary

Manatees fill a unique aquatic niche. As the only herbivorous marine mammal,

they inhabit shallow, coastal estuarine or riverine waters, feeding primarily on seagrasses

and other aquatic vegetation. The pachyostotic long bones of the Florida manatee are

hypothesized to regulate buoyancy in shallow water and serve as a means of attitude

control (Domning and de Buffrenil, 1991). Because the ribs and long bones are not

weight bearing in the aquatic environment, development of pachyostosis may have been

evolutionarily advantageous. However, the arrival of the engine-driven watercraft to

Florida waters has left the manatee ill-suited to cope with their encounters. Every year,

25-30% of all manatee deaths are due to injuries that result from bones broken in

collisions with watercraft. To begin to understand the injuries from a bone mechanics

perspective, we measured some material properties of manatee rib bone. We found that

this bone is less strong and less tough than most other mammalian bone. We also found

that, although the bone increases in static strength as the animals grow, the bone is not

able to absorb more energy. In fact there is a strong indication that the ability to absorb

energy (measured as work of fracture) declines with increasing size in animals over 265

cm total body length (i.e., large subadults). This decline is due in part to the bone's high

density and mineral content. Finally, we found no differences in mechanical properties

between males and females. The pachyostotic nature of the bone, which makes the

manatee well-adapted to its environment, also leaves it highly susceptible to fatal injuries

from boats.









Table 2-1. Mean material properties for 26 manatees
Animal Sex TL Age N o E WOF NKC Kc Density Mineral
(cm) class (MPa) (GPa) (MJ/m3) (MPam1/2) (g/cm3) content


MEC0033 F 102 C
MSTM0109 M 166 C
MSW0033 M 178 S
MSW0164 M 182 S
MNW0207 F 188 S
MSW0227 F 190 S
SWFTM0114 M 196 S
MSW0057 M 201 S
MSW0134 F 205 S
MSE0205 M 238 S
MSW0160 F 246 S
MSTM0102 F 265 S
MSW0165 F 267 S
MEC0220 M 275 S
MSW0251 M 276 A
MSW0225 F 282 A
MSW0223 F 294 A
MSW0208 F 299 A
MEC0213 F 314 A
MSW0226 M 315 A
MSW0239 M 321 A
MSW0331 F 326 A
MSW0253 M 327 A
SWFTM0414 F 335 A
MNW0208 M 343 A
MSW0245 M 344 A


5 62 4.0 3.1
30 134 12.8 3.8
31 136 12.0 6.0
25 105 9.4 4.0
37 127 11.6 3.5
48 124 12.3 4.2
66 122 11.2 4.6
36 153 9.6 5.5
58 130 11.5 5.5
114 135 12.6 4.4
131 159 14.9 5.6
169 157 15.4 4.3
157 152 14.5 4.6
139 144 14.4 3.6
142 160 17.7 3.6
160 151 16.0 3.8
184 153 15.5 4.0
175 154 16.0 4.2
143 143 16.0 3.2
156 151 16.6 3.2
181 150 16.4 3.7
173 150 17.2 3.2
189 149 15.7 3.8
189 158 17.4 3.6
178 159 18.0 3.0
307 143 15.6 3.1


2.14 69


2.22
2.10
2.16




2.11
2.18

2.15
2.11
2.20
2.18
2.20
2.22
2.16
2.18
2.21
2.19
2.19
2.20
2.22
2.16


Variables are: sex, male (M) or female (F); total body length (TL); age class, calf (C)
subadult (S), or adult (A); sample size for strength, modulus, and work of fracture (N);
flexural strength (c); Young's modulus (E); work of fracture (WOF); number of
toughness samples (NKc); fracture toughness (Kc); density; and percent mineral content.
- indicates no data for that variable.























Cranial Ribs Central Ribs Caudal Ribs Bending Test
1-6 7-12 13-17 Specimens


Figure 2-1. Sample sites for biomechanical analyses. A) Ribcage is divided into three
regions: cranial, middle, and caudal. One rib from each region was used for
bending tests. B) Sample sites for bending tests within each rib are indicated.


180


S160


S140
-D
C
) 120-


- 100-
X
x ,
U


50 100 150 200 250 300 350


Total body length (cm)


Figure 2-2. Flexural strength vs. total length. Strength increases with body length in the
subadults and plateaus in the adults. There was no significant difference in
strength between sexes (ANCOVA, p = 0.93).


Females
Males


R2=.79











3.0

2.8 -

S2.6 -

C 2.4-
0-
2.2-

C 2.0-

--
1.8 -

1.6-


Total body length (cm)


Figure 2-3. Fracture toughness vs. total length. Fracture toughness peaks in the subadults
and thereafter plateaus in the adults. There was no significant difference
between sexes (ANCOVA, p = 0.81)


E 5.5

5.0



ILL
"5 4.5-

4.0-
4-
0
S3.5.


100 200 300


Total body length (cm)
Figure 2-4. Work of fracture vs. total length. Work of fracture peaked in the subadults,
and thereafter declined with increasing body length. There was no significant
difference between sexes (ANCOVA, p = 0.96).


o (0 #
O Females
SMales 0 +




S=73



R2=.73

O


O Females
* Males


R2=.50






























200 300


Total body length (cm)
Figure 2-5. Elastic modulus vs. total length. Elastic modulus increased with increasing
body length. There was no significant difference between sexes (ANCOVA,
p = 0.70).


160


140


120


100


80


2 4 6 8 10 12 14 16 18 20

Modulus (GPa)

Figure 2-6. Flexural strength vs. modulus. Flexural strength is significantly correlated to
the elastic modulus (p<0.0001).


Females
Males


R2=.86


R2=.67












100
-- Calf
Subadult
80 Subadult
81 N Adult
14E
60 75 N 6.1 W
17 E
Z 2.4 W
0 40 59 N
CU 13E
0 49 N 10.3 W
11E
20 2.2 W


0 -



0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5

Displacement (mm)

Figure 2-7. Load-displacement curves for small sample specimens tested in 3-point
flexure. Values given for each curve are: ultimate load, N (Newtons);
modulus, E (GPa); and work of fracture, W (MJ/m3). Three age classes are
represented: calves (MNW0207); subadults (MSW0160); and adults
(MNW0208). Two curves for the subadult MSW0160 are shown. The red
curve is an example of the high work of fracture obtained for some subadult
specimens. The black curve is a more typical curve for a large subadult.










2.24

2.22

2.20

E 2.18
C)
2.16

c 2.14
2.12
2.12


2.10

2.08


Total Body Length (cm)

Figure 2-8. Density vs. total body length. There was no significant relationship between
bone tissue density and body size (p = 0.0517).


0 0



0 0 0
0 0

0
0
2
O






R2=.27


0


Total Body Length (cm)

Figure 2-9. Percent mineral content vs. total body length. There was a significant increase
in percent mineral content with total body length (p = 0.0198), although the
correlation was weak.


00 00

0 0 oO



0 0



0 R2=.19
0
O
O


























8 -
2.08


2.10 2.12


2.14 2.16 2.18 2.20 2.22 2.24


Density (g/cm3)

Figure 2-10. Modulus vs. density. With one outlier (shown in red) removed, modulus is
significantly correlated to density (p<0.0001).


64 66 68 70


% Mineral Content
Figure 2-11. Modulus vs. percent mineral content. Modulus is significantly correlated to
mineral content (p<0.0001).


0 O0

8 o
O
O



0
O
O


O


00


Po 0


R2=.61


0O O












5.0e-4

4.5e-4 -

4.0e-4

3.5e-4
CU
N
'FJ 3.0e-4
SO
g 0
C 2.5e-4 -

0
2.0e-4 O "O '

1.5e-4

1.0e-4
100 200 300 400

Total body length (cm)

Figure 2-12. Small sample crack size vs. total body length. The neonate calfMEC0033
was an outlier (shown in red). With the outlier removed, there was no
change in crack size with total body length (R2 = 0.0009, p = 0.8886).














CHAPTER 3
STRUCTURAL PROPERTIES OF MANATEE RIBS TESTED IN IMPACT

Introduction

The endangered Florida manatee shares its nearshore, coastal habitat with an ever

increasing number of boats. Each year, 25-30% of all manatee deaths are a result of

collisions with watercraft. Boat speed zones are the primary tool for ameliorating the

lethal interactions. However, existing limits are not based explicitly on information

pertaining to the nature or severity of the injury, including bone fractures. In an effort to

establish safer boat speeds for manatees, the Florida manatee recovery plan calls for

research on the mechanical characteristics of manatee bone to better understand the

effects of watercraft-related impacts (U.S. Fish and Wildlife Service, 2000). This study is

the first to estimate mechanical properties of whole manatee bone.

Quantitative fractography is a technique which uses principles of fracture

mechanics and fractal geometry to analyze fracture surfaces of brittle materials such as

ceramics (Mecholsky, 1993, 2001). The strength of a material is determined by the size

of the fracture initiating crack, or flaw, and the fracture toughness of the material.

Location of the crack can be determined by examining the markings on the fracture

surface. In similar fashion, the topography of a fracture surface is also a record of fracture

parameters and is fractal in nature. The fractal dimensional increment (D*) is a

quantitative measure of the tortuosity of a brittle fracture surface, and is related to

parameters such as the fracture toughness and elastic modulus West et al., 1999).









Two features of manatee bone allowed us to use quantitative fractography to

measure its mechanical parameters. First, all its ribs and long bones are constructed of

solid cortical bone; there are no marrow cavities or spongy bone. This feature, unusual

for mammalian bone, is thought to be an adaptation to aid in hydrostasis (Domning and

de Buffrenil, 1989). Second, manatee bone often undergoes brittle fracture in static tests

(Chapter 2; Yan et al., 2005). Bone typically behaves as a quasi-brittle solid (Zioupos,

1998). However, fully mineralized bone from adult manatees does not undergo any post-

yield plastic deformation.

The goal of this study was to estimate the stress at failure and the fracture

toughness of whole manatee ribs fractured in impact. Fracture toughness is the ability to

resist fracture, one measure of which is the critical stress intensity factor (Kc). This is an

estimate of the amount of energy required to propagate a macrocrack that leads to

fracture (Mecholsky, 1996). Toughness is an important property because a tough bone is

resistant to fracture, and this is most important when bone is impacted (Wainwright et al.,

1982; Turner and Burr, 1993). Since adult manatee bone fails in a brittle manner when

loaded statically, it is likely that fracture resistance measured by Kc is approximately

similar to the work of fracture as a measure of toughness in manatee bone loaded in

impact. We also addressed the hypothesis that forces generated by most watercraft under

normal operation are sufficient to inflict fatal skeletal injuries to manatees.

Materials and Methods

Bone Sample Collection

Ribs from 20 animals of both sexes, in three age classes (i.e., calf, subadult, and

adult) were obtained from the Florida Fish and Wildlife Conservation Commission's

Marine Mammal Pathobiology Laboratory. By convention, total body length was used as









a proxy for age, and individuals were assigned to age classes based on established size

definitions (O'Shea et al., 1985). Bones were obtained from fresh carcasses at the time of

necropsy. One rib from the mid-thoracic region of each animal was collected and stored

frozen until testing. Bones were cleaned of soft tissues and tested at room temperature.

Whole Rib Impact Testing

Whole ribs were impacted using a compressed air gun at the Structures

Laboratory in the Department of Coastal and Civil Engineering, at the University of

Florida (Fig. 3-1). The barrel is 20' in length with a 4" inner diameter. The gun is

equipped with a compressor that allows projectiles to be fired at a range of speeds, and

two photocells and an oscilloscope to record the speed of the projectile as it exits the

barrel. Ribs were positioned vertically in a fixture that was mounted on a steel frame, so

that the lateral (convex) side of the rib faced the gun, and the concave side faced the

frame. They were impacted perpendicular to their long axis, in the middle of the lateral

side with a 2" x 4" x 4'pine projectile. The leading end of the projectile was equipped

with a hemispherical nosepiece coated in fiberglass. The mass of the projectile was 13 kg.

The distance between the end of the barrel and the rib was 2.3 m. The projectile was fired

between 23-28 m/s. Kinetic energy (KE) of the projectile was calculated as

KE = 2mv2 (3-1)

where m is the mass of the projectile (kg), and v is velocity (m/s).

Quantitative fractography was used to identify the fracture origin and measure

crack size for each rib (Figure 3-2). The failure stress was calculated as

K
o = (3-2)
Y(c)1/2









where Kc fracture toughness (MPa-ml/2), measured as the critical stress intensity factor.

Y is a geometric factor based on crack shape and loading conditions, and c is the crack

size (m2). Values for Kc were taken from the middle region of the adjacent rib, which

was used to determine bone material properties in Chapter 2. The value used was the

mean for the adjacent section. Crack size was calculated as

c = (a-b)1/2 (3-3)

where a and b are crack depth and half-width (m), respectively, as measured from the

fracture surface with digital calipers to the nearest 0.1 mm.

Strain Gauge Testing

To validate the failure stress calculations, three additional ribs were outfitted with

strain gages to directly measure strain during impact. For each rib, a uniaxial gauge

(CEA-06-250UN-350, Vishay Micro-Measurements, Raleigh, NC) was placed on the

lateral (convex) surface, and a uniaxial gauge and a rosette gauge (CEA-06-125UR-350)

were placed on the pleural (concave) side. Gauges were placed along the midline, parallel

to the long axis, either above or below the center of the rib. The center element of the

rosette was oriented parallel to the long axis of the bone. Signals were output to an 8

channel signal conditioner and sampled at 50 kHz. Raw data were converted to

microstrain, which were used to calculate the principal strains using standard formulae.

Failure stress was calculated as

a = EE (3-4)

where E is Young's modulus (GPa) measured with ultrasound (Nuson, Inc., Boalsburg,

PA), and F is the maximum principal strain.

As a check of the strain data, an additional rib was outfitted with gauges and

loaded to failure in quasi-static 3 point bending at a load rate of 4.448 N/s. The signal









was sampled at 20 Hz. Failure stress was calculated using the standard formula for a

curved beam in bending. For all gauged ribs, the fracture origin was identified

fractographically, and Kc was determined using Equation 3-2. Toughness values

calculated for whole ribs were compared to those from the materials test specimens to

assess the validity of using the materials values from Chapter 2 in the failure stress

calculations.

Fractal Dimensional Increment

The fractal dimensional increment (D*) was calculated using a modified slit

island technique (Hill et al., 2001). The technique involves making an epoxy replica of

the fracture surface (Figure 3-3). The replica is ground down to the plane of fracture to

produce fracture surface contours, or islands. The island perimeter, represented by a line,

is measured with rulers of different lengths. A log-log graph of perimeter length versus

the ruler length produces a line whose slope is related to the fractal dimension, D (i.e.,

slope = 1-D). The non-integer, or fractional portion, of the slope is equal to D*, which is

a measure of the fracture surface tortuosity. The higher the value of D*, the more

tortuous (rough) the surface. For example, a fractal dimension (D) of 2.08 (D* = .08)

would be a surface which is relatively smooth, and a D of 2.40 (D* = .40) would be a

"rough" surface, with much tortuosity.

For each rib, a negative impression mold of the fracture surface was made in

dental putty and latex, and then double cast in epoxy. The cast was made such that the

fracture origin was horizontal. The fracture surface of the positive cast was sputter coated

(Denton Desk II cold sputter coater, Denton Vacuum, Moorestown, NJ) with gold-

palladium for contrast, and the second cast was poured. Replicas were ground down to

the fracture origin on a grinding wheel using increasingly finer grit sanding papers, then









polished with 0.3 Lm alumina slurry. Composite images of the fracture surface contours

were made from high resolution images (200X) taken with a digital camera mounted on a

light microscope (Figure 3-4). D* was calculated from the composite images using the

Materials Pro Analyzer software add-on for the Image Pro-Plus software (Media

Cybernetics, Inc., Silver Spring, MD).

Ideally we would like to be able to calculate whole bone fracture toughness from

the measured fractal dimension. The two parameters are related in the equation

Kc = Ea01/2D*1/2 (3-5)

where E (GPa) is the elastic modulus, ao (A)is a proportionality constant, and D* is the

fractal dimensional increment. The variable ao is a material constant that has been

proposed as a measure of the average intermolecular strained bond length (Mecholsky,

2001). It is not directly measurable, but must be calculated from other, independently

measured parameters or estimated using quantum mechanics models. Although ao has

been established for a number of materials, there is no experimentally determined value

for cortical bone. Using our measured D* values, Kc values calculated from the strain

gage data and fractographic measurements, and the ultrasonic modulus, we estimated ao

for manatee cortical bone.

Kinetic Energy Calculations

The most common recreational watercraft in operation in Florida waters is a 17'

vessel (Wright et al., 1995) which is typically equipped with a single outboard engine of

100-115 HP. A typical 17' vessel at idle speed (i.e., no wake) is traveling at 2-3 mph

(depending on vessel configuration), and 10-12 mph corresponds to a boat traveling at or

near hull speed (i.e., on a plane). Maximum speed is approximately 35-40 mph. Weights

for a few common 17' vessels with the recommended engines were taken from the









specifications posted on manufacturers' websites. Kinetic energy was calculated for

speeds of 2, 10, and 35 mph. We assumed one adult operator (200 lbs.) and one half-tank

of fuel. The soft tissues overlying bone have been shown to absorb and dissipate impact

energy, thereby decreasing the energy transferred to the skeleton (Currey, 1968; Nikolic

et al., 1975; Parkkari et al., 1994, 1995; Robinovitch et al., 1995). To estimate the energy

damping ability of the tissues overlying the ribs, soft tissue thickness (e.g., skin, blubber,

muscle) was measured approximately one third of the way down from the head of the ribs

on some whole body sections. Robinovitch et al. (1995) measured the amount of impact

energy transferred to the femoral trochanter with overlying soft tissues of varying

thickness. They regressed tissue energy absorption against soft tissue thickness to

determine the threshold of tissue thickness needed to prevent femur fracture during falls

in the elderly. We used their regression to estimate the amount of energy dissipated by

manatee soft tissues. Kinetic energy curves for the boats were visually compared to the

kinetic energy of the projectile used to impact the bare rib, as well as to the estimate for

ribs with overlying soft tissues.

Statistical Analyses

Failure stress and D* were compared between sexes and age classes in separate

one-way ANOVAs (Zar, 1996). Fractal dimension was regressed against total body

length (a proxy for age) to examine the relationship of D* to age. Analyses were

performed with SAS (SAS Institute, Inc., 1989).

Results

Whole Rib Impact Tests

Mechanical and fractal dimension data from the impact tests are presented in

Table 3-1 and Fig 3-4. A total of 22 ribs from 20 animals were tested. Only one calf was









tested, as the small size of the ribs made them difficult to test. Of the other 19, eight were

subadults (i.e., juveniles) and 11 were adults. The sexes were represented equally, ten

females and ten males. Mean failure stress was 47 MPa, varying from 37-67 MPa. Mean

adjusted failure stress (see below) was 139 21 MPa (range 106-179 MPa), and mean

D* was 0.13 0.04 (range 0.08-0.22). Means by sex and age class are reported in Table

3-2. We tested 4 ribs from animal MNW0208, which was a large adult. Three of those

were tested in impact and one in quasi-static 3 point flexure. Values for D* were .09, .13,

.13, and .20 (Tables 3-1 and 3-3). The variability in these numbers is similar to that for

the entire dataset, and thus, not unexpected. This variability may represent natural

variation in the ribs, or it may be an artifact of the replication or grinding processes.

There were no statistically significant differences between the sexes for either failure

stress (p = 0.51) or D* (p = 0.18). Similarly, no differences were found between age

classes for either variable (p = 0.19 and p = 0.63 for failure stress and D*, respectively).

The lone calf was omitted from the analysis by age class. The regressions of D* and

failure stress against total body length indicated no relationship between either parameter

to body size (Figures 3-5 and 3-6, respectively). These two independent measures both

imply that the ability to resist bone fracture in impact does not increase as animals grow.

Strain Gauge Tests

Of the three ribs outfitted with strain gages for the impact tests, data were

successfully collected for only one (Table 3-3). The rib tested in 3 point bending failed at

a load of 4480 N. Peak strains for the impact and static tests were very similar (0.56%

and 0.61% respectively). Fracture toughness calculated for the impacted rib was 7.7

MPa-m1/2, essentially identical to the 7.8 MPa-m1/2 for the rib from the static test. These

values were much greater than those used for the failure stress calculations reported in









Table 3-1. Our results indicate that it is not appropriate to use the small specimen fracture

toughness values of Chapter 2 to calculate whole bone failure stress. By doing so, our

stresses were most likely underestimates. To adjust for this, we used 8 MPa-m1/2 to derive

more appropriate estimates of failure stress, reported as ow in Table 3-1 and Figure 3-6.

The material constant ao calculated for the two gauged ribs is reported in Table 3-

3. The value for rib 14 is twice that for rib 6, 13,000 and 6,700 A, respectively. Values

for classes of ceramics are much lower, typically 1-100 A, but our values lie in the range

reported for various polymers, from 2,700-14,000 A (West et al., 1999; Mecholsky,

2001). Much more work needs to be done to determine whether it is possible to estimate

ao for a complex material such as cortical bone.

Kinetic Energy Calculations

Projectile velocity, as measured exiting the barrel, was 23-28 m/s. To make

conservative estimates of the threshold for fracture we used the maximum projectile

velocity (28 m/s) for calculations. Vessel configurations and kinetic energies are

presented in Tables 3-4 and 3-5, and Figure 3-7. Based on the configurations selected, rib

fracture can occur at speeds of 7-8 mph, or just below hull speed. This is speed at which

boats generate kinetic energy equal to that of the projectile, as determined by the point at

which the regression curves cross the kinetic energy line for the projectile. This estimate

is for bare bone, which of course is not a realistic scenario. Human impact studies

indicate that for some body sites, soft tissues absorb and dissipate as much as 90% of

impact energy (Nikolic et al., 1975). The maximum thickness of manatee soft tissues

(i.e., skin, blubber, and muscle) was 58 mm. Robinovitch et al. (1995) measured impact

energy absorption for soft tissues up to 43 mm thickness. Using their regression, we

extrapolated the amount of energy dissipated by the manatee soft tissues to be 70%. That









is, only 30% of impact energy is transferred to the bone. This projection is plotted on

Figure 3-7. At this energy absorption estimate, boat speeds of 13-15 miles per hour are

sufficient to fracture manatee ribs (i.e., where the regression curves cross the "with skin"

kinetic energy estimate, indicated by the dashed lines on Figure 3-5).

Discussion

Fractals are geometrical objects with fractional dimensions (Mecholsky, 1986).

Mandelbrot was the first to recognize that natural structures and phenomena are fractal,

and this includes fracture surfaces (Mandelbrot et al., 1984; Mecholsky et al., 1996).

Fractal geometry provides a means of measuring the roughness, or tortuosity of a fracture

surface. The fractal dimension (D) is a quantitative measure of the dimensionality of an

irregular, or tortuous, surface. The fractional dimensional increment (D*, the fractional

portion of the fractal dimension) is related to fracture parameters such as the fracture

toughness and elastic modulus. Techniques were first applied to metals, and were

subsequently applied to ceramics and some composites (Mandelbrot et al., 1984;

Mecholsky et al., 1989; West et al., 1999; Czarnecki et al., 2001). Fractal geometry is

now being used for trabecular bone as a means to quantify changes in bone quality and

quantity (Majumdar et al., 1993). We have utilized the fractal dimension to quantify

fracture parameters of manatee cortical bone using the same methods applied to structural

materials.

The Fractal Dimensional Increment

Although use of fractal analysis now can be found widely in the engineering

literature, there are few examples of its application to biological composites. Fractal

dimension is becoming a popular tool for quantifying changes in trabecular bone

architecture related to disease conditions such as osteoporosis. In this body of literature,









its proposed use is singly as a diagnostic tool. Only two studies of biological composites

have related fracture energy to fractal geometry. Currey et al. (1995) measured fractal

dimension for cortical bone for a variety of species. Fractal dimensional increment was

correlated to the work of fracture, from .05 for a highly mineralized, brittle fin whale

tympanic bulla to .25 for very tough, young deer antler. Various types of cortical bone

had intermediate fractal dimensions, which roughly correlated to their calculated works

of fracture. Hill (2001) measured fractal dimensional increments for the conch shell of

Strombus gigas. Means ranged from .16-.22 for specimens stored in different media. Our

measured D* values for manatee cortical bone ranged from .08-.22, which is in general

agreement with the other two studies. However, it is difficult to make detailed

comparisons among studies because, although the same theory was used to calculate

fractal dimension, different measurement methods were used. Additionally, biological

composites, like some other materials, may be multi-fractal, making results sensitive to

measurement length scale. Currey et al. (1995) found this to be true for bone with high

work of fracture, fractal dimension decreasing with increased length scale in these tough

materials.

A wide range of D* values exists for classes of materials (e.g., metals,

polycrystalline ceramics, and glasses). However, since most of the data available in the

literature are for essentially monolithic materials, we could only make a cursory

comparison to our data. Aside from the two references to bone and shell, the most

appropriate materials to compare to our data are the cement-epoxy composites. Czarnecki

et al. (2001) measured fractal dimensional increments of .02-.08 for seven epoxy

concretes. However, they used the vertical profile technique to calculate fractal









dimension, which is known to produce values consistently lower than those of the slit

island technique (Hill et al., 2001).

Fractal Dimensional Increment and Fracture Toughness

Because of the large amount of variation in fractal dimension both within and

among classes of materials, comparing D* values alone is not informative. It is more

instructive to examine the relationship between the fractal dimension and fracture

toughness. For most materials that fail in a brittle manner, there is a positive correlation

between fracture toughness (Kc) and D*1/2, and the slit island technique has been shown

to produce a D* that is a more accurate predictor of Kc than other techniques (Russ,

1994; Hill et al., 2001; Mecholsky, 2001). Over the range of D* values for manatee bone,

fracture toughness only varied by 1 MPa-ml/2, which resulted in no significant

relationship between fractal dimension and fracture toughness (R2 = 0.02, p = 0.5535,

Figure 3-8). There are a few possible explanations for the lack of correlation. It could be

that there is no relationship between the parameters for cortical bone, or that the range of

fracture toughness values was too narrow to detect a trend. A third possibility is that error

was introduced into our technique. Calculation of the fractal dimensional increment using

the slit island technique is sensitive to error in the polishing angle. Della Bona et al.

(2001) showed that variation of 50 or greater from the fracture plane significantly

underestimates D*. There may have been angle errors introduced in the production and

grinding of the replicas that would account for some of the low values reported here.

However, our methodology was consistent throughout, so this would not explain the non-

significant relationship between fractal dimension and fracture toughness.

We hypothesize that the lack of correlation between fractal dimension and

fracture toughness reflects the true relationship between these properties for manatee









bone. Critical crack size for the whole ribs was similar for all animals regardless of total

body length (R2 = 0.10, p = 0.1538, Figure 3-9) or cross sectional area of ribs. In Chapter

2 we reported that density and mineral content also did not vary significantly over the

range of body sizes. If these values accurately reflect the material properties of rib bone,

then our finding of non-significant variation in critical crack size and failure stress among

animals is not anomalous. Despite the increase in rib size (340-1926 mm2 cross-sectional

area) with increasing body size in manatees, there is no concomitant increase in ability to

resist crack formation because the material is essentially the same.

R-curve Behavior and Viscoelasticity in Bone

The greater fracture toughness values calculated from the strain data relative to

the small specimen values of Chapter 2 may be indicative of R-curve behavior in manatee

bone. A rising R-curve describes an increase in fracture toughness with the increase in

crack extension (Vashishth, 2004). This behavior is characteristic of quasi-brittle

materials such as some ceramics, and has now been documented in human, bovine,

equine, and deer antler bone (Les et al., 2002; Malik et al., 2003; Nalla et al., 2004;

Vashishth et al., 1997). If manatee bone exhibits R-curve behavior, then it is not

appropriate to use toughness values from small sample material properties tests to

calculate failure stresses of whole bones. Those values, measured from machined

specimens tested in 3-point bending, ranged from 2.0-3.0 MPa-ml2.

True brittle materials exhibit flat R-curve behavior (Vashishth et al., 1997;

Zioupos, 1998). The material tests in Chapter 2 showed that adult manatee bone tested in

static 3 point bending undergoes very little or no plastic deformation; it typically fails

catastrophically near the yield point. This seemingly contradicts the toughness data

presented here that suggests R-curve behavior. However, Vashishth et al. (1997) point









out that some ceramics do produce rising R-Curves. Like all bone, manatee bone is a

ceramic-polymer composite of complex, hierarchical structure. Bone microstructure,

mineral density, and collagen content all affect the R-curve behavior (Les et al., 2002;

Malik et al., 2003; Nalla et al., 2004; Vashishth et al., 1997). Malik et al. (2003) reported

lower R-curves in bone of increased stiffness, tested in tension. Nalla et al. (2004)

reported an age-related decrease in both crack initiation and crack growth resistance. The

high mineral content of manatee bone (mean 69%) likely has a similar effect. Appropriate

testing needs to be done to characterize R-curve behavior in manatee bone.

The R-curve behavior of bone is due to its viscoelastic nature; its collagen

component is partly responsible for the toughening mechanism of crack bridging. A

larger sample allows for a greater crack size, and more collagen is available to contribute

to the toughening, resulting in a larger fracture toughness. Therefore, absolute size most

likely explains the greater fracture toughness of the whole ribs over the machined

specimens. For some materials, including most bone, fracture toughness is strain rate

dependent. As strain rate increases, so does strength, modulus, and Kc; a feature also

attributable to viscoelasticity (Zioupos, 1998). However, note that work of fracture

(WOF) will decrease with increased strain rate, also due to the viscoelastic nature of

bone. The different behavior of Kc versus WOF with strain rate is due to the method of

measurement of both. Because we tested the whole bones and the small sample

specimens at different rates, the effects of strain rate and specimen size were confounded

in this study. However, in ceramics, fracture toughness is not strain rate dependent, as

long as it is not strongly sensitive to stress corrosion processes. The impact tests provide









additional evidence that manatee bone behaves as a ceramic; the whole rib tested in

quasi-static 3 point flexure had a crack size identical to the other ribs tested in impact.

Impact Energy Absorption of Soft Tissues

Impact testing of human bone is often done to determine the loads at which bones,

such as the femur, break. The whole rib manatee tested in static 3 point bending failed at

a load of 4480 N, which is only moderately greater than the critical trochanteric impact

load of 4170 N, the average load required to fracture in impact the proximal femur in the

elderly (Parkkari et al., 1995). Soft tissues overlying the bone absorb impact energy,

thereby reducing the load transferred to the bone. Energy absorbed by soft tissues has

been shown to be as high as 90%. In femoral impact tests conducted by Robinovitch et al.

(1995), energy absorption varied from 6-58%. As tissue thickness increases, energy

absorption increases, and the peak load transferred to the bone decreases. The authors

found that for each millimeter of soft tissue thickness, peak load transferred to the bone

decreased by 70 N, and soft tissue energy absorption increased by 1.7 J. In a different

study, Parkkari et al. (1995) impacted artificial hips protected by soft tissues at high and

low impact rates. They found that, regardless of impact rate, 80% of the total load was

transferred to the bone. In these cited two studies, impact energies of 132 J and 140 J

were sufficient to generate loads capable of creating fractures. Our projectile generated 5

kJ, so we likely could have created fractures in manatee ribs with much less force. Taking

the soft tissues into consideration, our calculations indicated that a kinetic energy of 17 kJ

is required to create rib fractures in manatees. Based on the regression of Robinovitch et

al., (1995), 12 kJ of the 17 kJ (70%) would be dissipated by the soft tissues, and the

remaining 5 kJ (30%) transferred to the skeleton. For a typical 17' vessel, this translates

to speeds of 13-15 mph. We needed to make a few assumptions to generate this estimate.









While our weights for the vessels are accurate, the total weight would likely be greater, as

we did not include items such as the battery, safety equipment, coolers, other equipment,

or additional passengers. Greater vessel weights would shift the curves up, effectively

lowering the vessel speed able to create fractures. More importantly, it may not be valid

to use the energy-soft tissue thickness regression based on human tissues for manatees.

The dermis of manatee skin is different from that of humans, and from that other marine

mammals (Kipps et al., 2002), and likely has different energy-damping capabilities.

Additionally, manatees have a blubber layer. Blubber is different from fat in terrestrial

animals. To more accurately determine the position of the "with skin" line in Figure 3-7,

the compliance of manatee skin and blubber needs to be determined, and impact testing

of ribs with the associated soft tissues in place needs to be performed.

Summary

The goal of this study was twofold: to estimate the stress at failure and the fracture

toughness of whole manatee ribs fractured in impact, and to determine if the typical

watercraft is able to generate enough force to break manatee ribs upon impact. The

unique construction of manatee ribs enabled us to apply fractal geometry techniques to

measure some fracture mechanics parameters. Adult manatee bone behaves more like a

ceramic than other types of bone. Because of this, we were able to see many of the

features observed for brittle fracture in ceramics. We were able to identify crack origins,

and make quantitative measurements of crack size. Failure stress was constant across

body size, indicating that the amount of force needed to break ribs does not change as the

animals grow, despite the increase in rib size. There was no change in the amount of

kinetic energy needed to fracture whole bone with total body length, measured as the

fractal dimension (D*). The whole rib toughness and measured D* independently






75


corroborate the finding that flaw size is the same for all animals, regardless of body size.

Just as for the material properties tests of Chapter 2, we found no differences between the

sexes in their ability to resist fracture. Finally, we found that the boats typically found in

Florida waters can easily generate enough impact force to break manatee ribs.









Table 3-1. Mechanical and fractal data for 20 manatee ribs fractured in impact
Animal Sex TL (cm) Age D* c (m2) Kc (MPa-m1 2) GM (MPa) w (MPa)
MSTM0109 M 166 C .13 1.9E-03 2.5 46 142
MNW0207 F 188 S .09 1.9E-03 2.4 45 145
MSW0227 F 190 S .09 2.5E-03 2.3 37 125
MSTM0113 M 196 S .17 1.5E-03 2.0 41 159
MSW0057 M 201 S .14 1.5E-03 2.2 46 163
MSE0205 M 238 S .22 1.8E-03 2.5 48 149
MSW0160 F 246 S .17 1.8E-03 2.7 52 148
MSTM0102 F 265 S .13 1.5E-03 2.5 51 159
MSW0165 F 267 S .11 2.7E-03 2.5 45 120
MSW0251 M 276 A .12 2.4E-03 3.0 50 128
MSW0225 F 282 A .12 2.4E-03 3.0 49 127
MSW0223 F 294 A .14 2.2E-03 2.7 46 133
MSW0208 F 299 A .17 2.7E-03 2.8 43 120
MEC0213 F 314 A .08 1.5E-03 2.7 61 176
MSW0226 M 315 A .11 2.8E-03 2.7 41 118
MSW0239 M 321 A .17 1.6E-03 2.5 56 175
MSW0253 M 327 A .11 2.3E-03 2.7 46 130
MSTM0414 F 335 A .12 2.0E-03 2.8 50 138
MNW0208#6 M 343 A .20 2.3E-03 2.9 48 129
MNW0208#9 M 343 A .09 1.5E-03 2.9 67 179
MNW0208#10 M 343 A .13 3.5E-03 2.9 40 106
MSW0245 M 344 A .15 2.2E-03 2.4 41 133
Variables are: sex, male (M) or female (F); total body length (TL); age classes: calf (C),
subadult (S), or adult (A); fractal dimensional increment (D*); crack size (c); fracture
toughness from small specimens (Kc); and failure stresses (cM calculated with fracture
toughness data from machined, small specimen tests of Chapter 2, and aw calculated
from estimated whole bone fracture toughness). Three ribs were tested for animal
MNW0208 (rib number indicated).


Table 3-2. Mean mechanical and fractal data by sex and age class


Group ao (MPa) ow (MPa) D*
Sexes (ages pooled)
Females (10) 48 6 139 18 .12 .03
Males (12) 46 8 143 23 .15 .04
Ages classes (sexes pooled)
Subadults (8) 48 5 146 + 16 .14 .05
Adults (13) 49 8 137 24 .13 .03


Variables are failure stress (cM), adjusted failure stress (ow), and fractal dimensional
increment (D*). Values are mean S.D. Sample sizes are reported in parentheses for
each group. No groups were significantly different (p>0.05).









Table 3-3. Mechanical data for strain gauged ribs
Peak Modulus Stress c Toughness ao
Rib # microstrain (GPa) (MPa) (m2) (MPa-ml12) D* (A)
6 6075 21 128 2.3E-03 7.7 .20 0.7E-06
14 5562 19 106 3.5E-03 7.8 .13 1.3E-06
Ribs were from animal MNW0208. Rib #6 was tested in impact, rib #14 in quasi-static 3
point bending. Young's modulus was measured with ultrasound. Failure stress was
calculated as a = EE. Crack size, c (m2), as measured from the fracture surface. Fracture
toughness calculated from Equation 3-3. Constant ao was calculated from Equation 3-5.


Table 3-4. Boat data for kinetic energy calculations
Weight Tank Fuel Engine Boater Total
Make Length (kg) (gal) (kg) HP (kg) (kg) (kg)
Maverick1 17'0" 684 40 53 115 162 91 989
Mako2 17'3" 567 37 49 115 162 91 869
Key West3 17'6" 454 40 53 100 161 91 758
Boat weights were taken from manufacturers' specifications posted on their official
websites. Fuel weight is for one half tank. Total weight is equal to the sum of weights for
the vessel, fuel, engine, and one passenger.
Maverick Boat Company, Inc, 2004. Model: 17' Master Angler
2Mako Marine International, Inc. 2005. Model: 171 Center Console
3Key West Boats, Inc., 2005. Model: 1760 Stealth


Table 3-5. Kinetic energy calculations for vessel configurations and impact projectile
Kinetic Kinetic Energy (kJ)
Manatee Energy (kJ) Boat 2mph 10mph 35mph
Bare bone 5 Maverick 0.40 10 121
With skin 17 Mako 0.35 9 106
Key West 0.30 8 93
Energy for bare bone is the kinetic energy of the projectile. With skin is the minimum
amount of kinetic energy needed to create a fracture based on the calculated 70% energy
absorption by manatee soft tissues.





































D
Figure 3-1. Setup for impact tests of whole manatee ribs. A) Compressed air gun
assembly. The barrel is a 20' PVC pipe with a 4" diameter. The end of the
barrel is equipped with 2 photocells to measure the velocity of the projectile.
B) The gun is equipped with a compressor, pressure gauge, and trigger. C)
Manatee rib positioned in test fixture, mounted on frame for impact testing.
The rib is outfitted with a uniaxial strain gauge on the convex (lateral) surface.
D) View of the concave (pleural, or medial) side. This side is outfitted with a
uniaxial and a rosette strain gauge. The intended area of impact is between
these two gauges.






















A











I--^--I

B

Figure 3-2. Fracture surface of a manatee rib fractured by collision with a boat. A)
Examination of fracture surface indicated that the crack that led to failure
(dashed line) originated at a groove on the medial surface (arrow). B)
Schematic of fracture surface features used for fractography. Crack size (c) for
a semi-elliptical surface crack is calculated by measuring crack depth, a, and
half the crack width, b (i.e., /2(2b)).













































Figure 3-3. Procedure for making epoxy replicas of manatee rib fracture surfaces. A) A
negative impression of the fracture surface is made in dental putty and latex.
B) A positive mold is made from epoxy using a PVC ring. C) The fracture
surface of the positive epoxy cast is sputter coated with gold-palladium for
contrast. The second layer of epoxy is then poured. D) The completed cast is
ground down to the area of interest on a grinding wheel using sanding papers
of increasingly fine grits. The cast is polished with an alumina slurry. E) The
fracture contours, or "islands" (black arrows), are photographed at high
magnification (200X). The composite photographs are used for fractal
analysis.

























Figure 3-4. Example of a fracture surface contour used for calculation of fractal
dimensional increment (D*). Image is a composite of photos taken at 200X.
The computer traces the "island" contours (red line), and measures its length
with different size rulers to calculate D*.











0.26
0.24
0.22
0.20
0.18

Q 0.16
0.14
0.12
0.10
0.08
0.06


50 100 150 200 250 300 350 400

Total body length (cm)
Figure 3-5. Fractal dimensional increment (D*) vs. total body length. There was no
relationship between D* and animal size (R2 = 0.002). Thus, the ability to
resist bone fracture does not increase as animals grow. There were no
significant differences between sexes or age classes.


200

180

160

140

120


100 -

80 .
50 100 150 200 250 300 350 400

Total body length (cm)

Figure 3-6. Failure stress vs. total body length. There was no relationship between
adjusted failure stress (cw) (R2 = 0.012, p = 0.5535).


0 Male calf
0 Female subadult
O Male subadult
* Female adult
* Male adult

O o o*

0 -
0 o


Y O0

0 CO
O0
0 o8


o







83



140

120 Projectile
With skin
Key West
100 Mako
S- Maverick
2, 80


o 60

40

20 -
I I
0
0 5 10 15 20 25 30 35 40

Boat speed (mph)

Figure 3-7. Kinetic energy as a function of boat speed. Three vessel configurations are
presented. Calculated kinetic energy of the projectile is plotted, along with an
energy estimate with overlying soft tissues in place. With an estimated 70% of
the total kinetic energy being dissipated by the soft tissues, minimum boat
speeds capable of producing rib fracture are 13-15 mph (dashed lines).



















E 2.6
0_
S2.4
C-)
2.2

2.0


1.8 --
0.25


0.30


0.35


0.40


0.45


0.50


Figure 3-8. Fracture toughness (Kc) vs. fractal dimensional increment (D*). There was no
relationship between the two variables (R2 = 0.0226, p = 0.5047).


4.0e-3

3.5e-3

3.0e-3

2.5e-3

2.0e-3

1.5e-3


1.0e-3 -

5.0e-4 -
50 100 150 200 250 300 350


Total body length (cm)
Figure 3-9. Whole rib crack size vs. total body length. There was no relationship between
flaw size and total body length (R2 = 0.10, p = 0.1538). Flaw size is the same
regardless of rib cross sectional area, which increases with total body size.


0
0 0 0 0
0 0
0 0 0 0

0 0 0
0 0
0
0

0
O


0



0 0



0 O














CHAPTER 4
CONCLUSIONS AND FUTURE RESEARCH

Summary and Conclusions

Based on the work presented here, the following conclusions can be drawn:

1. Measured as a material property, the strength of manatee bone is generally lower than

that of other mammalian bone. This is likely due to three factors: its high mineral

content (67-71%) and density (mean 2.17 g/cm3); its microstructure; and its porosity.

Strength is known to increase with increasing mineral content between 63% and 68%.

When mineral content is above 68%, static strength decreases; we contend that this

may be the case for manatee bone. Most other mammalian bone does not reach these

large values of mineral content. We did not address microstructure and porosity in

this study.

2. Manatee bone, as a material, is less tough than other mammalian bone. We calculated

two measures of toughness, fracture toughness (Kc) and work of fracture (WOF). The

first, Kc, is calculated from the ultimate strength and crack size. Because Kc mirrored

the flexural strength curve so closely, this implies that crack size varied little over the

range of animals tested here. This is significant because for many ceramics, strength

is usually crack size dependent. However, Kc is a measure of the behavior in the

elastic, or pre-yield region only. Because of this limitation, more than one measure is

needed to fully characterize toughness. Work of fracture is the energy needed to

propagate a crack, measured as the area under the entire stress-strain curve. Because









toughness is largely determined by the post-yield behavior, the decline in work of

fracture seen in the adults may be due to age-related changes of the collagen matrix.

3. The age-related patterns in mechanical variables appear to be related to the amount of

bone material present. Manatees grow quickly; appositional bone growth is by

deposition of plexiform bone. In general, a scaffolding of woven bone is laid down

first, then later the spaces are filled in by lamellar bone. So in young animals the bone

is porous, but it becomes less so as they grow. This likely accounts for the

ontogenetic patterns observed here, as an increase in bone volume results in an

increase in strength.

4. There were no significant differences between the sexes for any mechanical variable.

Marmontel (1993) described possible differences in bone resorption between the

sexes, which we would expect to see reflected in the mechanical properties. However,

we found no differences between the sexes.

5. Tests of whole ribs (i.e., structural tests) indicated that adult manatees are not able to

better absorb and dissipate impact energy to prevent bone fractures than smaller

animals. There was no change in the energy needed to fracture whole bone with total

body length, measured as the fractal dimension (D*). Additionally, there was no

relationship between the energy needed to initiate a crack (Kc) and D*. This last

result is reasonable, since we found no significant change in mineral content or

density with body size. Despite the fact that the ribs increase in cross sectional area as

they grow, the tissue properties remain relatively constant. Regardless of rib size, it

takes about the same amount of energy to create and propagate a crack. The whole rib









toughness and measured D* independently corroborate the finding that flaw size is

the same for all animals, regardless of body size.

6. The typical watercraft found in Florida waters is able to generate enough kinetic

energy to create fatal rib fractures in manatees. Vessel speeds of 13-15 mph were

estimated to be sufficient to create fractures.

7. We successfully applied fractographic techniques to calculate fracture toughness for

manatee rib bone. We were able to see many of the features observed for brittle

fracture in ceramics.

Our study represents the first attempt to quantify the biomechanical effects of boat

strikes on manatees; this information contributes significantly to our understanding of

manatee-boat interactions. We think these data will be useful in shifting the focus to a

more objective approach for revising boat speed regulatory zones. It is our hope that

managers and policy makers will take into account the biology of the animal to devise

speed zones that are adequate to minimize the chance of fatal impacts.

Further Research

It became clear in the course of this research that a detailed histomorphometric

analysis of manatee bone is needed to elucidate the relationship between manatee rib

bone architecture and its material properties. In most mammalian bone, the relationship

of material properties to age is largely dictated by bone mineral content. This was not the

case for manatee bone, as mineral content and bone tissue density varied only slightly

across the range of animals tested, from a newborn calf to a 25 year old animal. Material

properties can vary widely at high mineral contents. Currey (1999) found that in bone

with calcium content of 230 mg/g or higher, bending strength may be either high or low.









Other factors that contribute to material properties to a lesser degree in other

mammals may have greater significance for manatees. For example, modulus (i.e.,

stiffness) is influenced by porosity and bone architecture as well as mineral content.

Currey (1999) found that porosity and calcium content together explained 60% of the

variability in modulus. However, for bone tested in bending, he found that porosity had

no observable effect on strength (Currey, 1998).

Bone architecture is known to affect mechanical properties. Not only do the bone

types (e.g. woven, lamellar) differ in mechanical properties, but even the orientation of

fibers within lamellae will affect the mechanical competency of the whole tissue (Martin

et al., 1998). Fawcett (1942b) described eccentric rings, or layers, visible to the naked eye

on the lateral surface of the ribs. We too have observed these markings in our preliminary

histological observations. Microscopic examination revealed that these rings mark the

boundaries between layers among which the osteons are oriented obliquely. Fawcett

(1942b) also described what he considered to be adult manatee bone as having a region of

un-remodeled endochondral bone. This area, along with the manner in which plexiform

bone is deposited, may result in a degree of porosity that persists regardless of age. It is

likely that manatee bone architecture and porosity affect its mechanical behavior in a

unique way; histology will bear this out.

The issue of manatee bone toughness needs to be examined further. We measured

fracture toughness (Kc), both for the small sample specimens and the whole ribs from the

impact tests. Yan (2005) also calculated fracture toughness for small sample specimens,

but his samples were notched. As a result, his crack sizes were intermediate to ours.

Correspondingly, his fracture toughness values of 4.1-5.5 MPa-m1/2 were intermediate to