1 ESTABLISHING THE COAGULATION PROFILE OF THE FLORIDA MANATEE ( Trichechus manatus latirostris) AND IDENTIFYING COAGULOPATHIES IN THE PATHOPHYSIOLOGY COLD STRESS SYNDROME By ASHLEY BARRATCLOUGH A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2015
2 Â© 2015 Ashley Barratclough
3 To my parents
4 ACKNOWLEDGMENTS I would like to thank my advisor, Dr . Ruth Francis Floyd, for her unwavering support and guidance throughout completion of this project. Additionally I am very appreciative to my committee members Professor Roger Reep, Dr. Ray Ball and Dr. Bobbi Conner. I would particula rly like to thank Dr. Conner for patiently teaching me how to perform thromboelastography and assisting me with sampling. I am very grateful to Roger, Ruth and Bobbi for travelling to the United States Geological Survey (USGS) Crystal River manatee health assessments to help me with my field work and supporting me throughout the tricky sampling time frames. I also thank Dr. Robert Bonde , Dr. Mike Walsh, the FWC staff and the entire health assessment team for facilitating my sample collection in both Crystal River and Brevard County . I would like to particularly acknowledge Dr. Martine De Wit for her support and mentorship throughout this research project. I wish to express my thanks to Michelle Devlin whose veterinary technician support was invaluable throu ghout my entire project. Additionally I am grateful for her emotional support and time listening to me when I was having hurdles to cross. I would like to thank Virginia Edmonds and the staff at Lowry Park Zoo particularly the manatee team for being invest ed in this research project with me and supporting my hard work along the way. I also wish to tha nk Dr. Trevor Gerlach and Dr. Nicole Stacy for their ideas to get this project off the ground. Financially I would like to thank the Joy McCann Foundation for their support of my fellowship program and the University of Florida and Lowry Park Zoo for this incredible opportunity. I am grateful to Stephanie Stein for facilitating my participation in this program from the beginning .
5 Finally I would like to thank my family and friends for their unwavering support throughou t the last two years. This research was conducted under the University of Fl #201408623 and the Federal Fish and Wildlife Permit MA067116 2.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 10 ABSTRACT ................................ ................................ ................................ ................... 11 CHAPTER 1 COLD STRESS SYNDROME REVIEW ................................ ................................ .. 13 Introduction ................................ ................................ ................................ ............. 13 Biology ................................ ................................ ................................ .................... 13 Mortality ................................ ................................ ................................ .................. 16 Unusual Mortality Events ................................ ................................ ........................ 17 Demography ................................ ................................ ................................ ........... 18 Warm water refuges ................................ ................................ ................................ 20 Clinical Signs of Cold Stress Syndrome ................................ ................................ .. 22 Current Therapy ................................ ................................ ................................ ...... 22 Pathological Findings ................................ ................................ .............................. 24 Mitigation ................................ ................................ ................................ ................ 26 Coagulation ................................ ................................ ................................ ............. 27 Summary ................................ ................................ ................................ ................ 28 2 ESTABLISHING THE NORMAL COAGULATION FACTORS IN WILD FLORIDA MANATEES ( Trichechus manatus latirostris ) AS A BASIS FOR EXAMINING THE PATHOPHYSIOLOGY OF THEIR HEMOSTATIC DISORDERS. ................... 34 Introduction ................................ ................................ ................................ ............. 34 Methods ................................ ................................ ................................ .................. 39 Statistical Analysis ................................ ................................ ................................ .. 41 Results ................................ ................................ ................................ .................... 41 Discussion ................................ ................................ ................................ .............. 43 3 ESTABLISHING NORMAL REFERENCES FOR THROMBOELASTOGRAPHY IN WILD FLORIDA MANATEES ( Trichechus manatus latirostris ) .......................... 49 Introduction ................................ ................................ ................................ ............. 49 Methods ................................ ................................ ................................ .................. 51 Statistical Analysis ................................ ................................ ................................ .. 54 Results ................................ ................................ ................................ .................... 54
7 Discussion ................................ ................................ ................................ .............. 55 4 THE ROLE OF THROMBOEMBOLIC DISEASE IN THE PATHOPHYSIOLOGY OF COLD STRESS SYNDROME ................................ ................................ ........... 63 Introduction ................................ ................................ ................................ ............. 63 Methods ................................ ................................ ................................ .................. 67 Statistics ................................ ................................ ................................ ................. 70 Results ................................ ................................ ................................ .................... 70 Discussion ................................ ................................ ................................ .............. 71 5 CONCLUDING SUMMARY ................................ ................................ .................... 85 LIST OF REFERENCES ................................ ................................ ............................... 90 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 100
8 LIST OF TABLES Table page 1 1 Total mortality reports of th e Florida manatee from 1974 to 2015 provided by the FWC carcass recovery program. Total number of deaths per year and the greatest cause of mortality identified for each year. ................................ ........... 30 2 1 Com plete data set for coagulation panel results in wild healthy manatees. ........ 47 2 2 Reference values for coagulation factors in wild Florida manatees .................... 47 3 1 Statistical comparison between sex, time (comparing samples at 3hrs and post 3hrs) and size (comparing adult with calf) with the p value for each parameter with <0.05 classed as significant. This table demonstrates that there were no significant effects of time, sex or size on the results. ................... 60 3 2 Established thromboelastography reference intervals for wild Florida manatees showing mean and (standard de viation) results. Additional data of horse, cow, dog and human values are presented for relative comparison. ....... 61 3 3 Comparison of the wild manatee thromboelastography established refere nce intervals with captive healthy rehabilitated individuals to facilitate confirmation of their normal health status. ................................ ................................ .............. 61 4 1 Thromboelastography results from 10 cold stress syndrome ca ses compared with the wild manatee reference ranges. P value indicates statistical comparison, <0.05 is classed as being statistically significantly different. .......... 81 4 2 Coagulation panel results comparing cold stress syndrome cases with wild manatee reference ranges. P value indicates statistical comparison, <0.05 is classed as being statistically significantly different. ................................ ............ 81 4 3 Cold stress syndrome case results showing the coagulation panel consisting of PTT, PT, fibrinogen, antithrombin, D dimer and factors VII through XII. The mean and standard deviation are provided for each parameter for these 10 individuals. ................................ ................................ ................................ .......... 82 4 4 Cold stress syndrome coagulation factor results compared with wild normal values. P value indicates statistical comparison, <0.05 is classed as being statistically significantly differ ent. ................................ ................................ ........ 83
9 LIST OF FIGURES Figure page 1 1 Internal anatomy of the Florida manatee, note the horizontal diaphragm resulting in a large surface area of digestive organs. Adapted from The Florida Manatee Biology and Conservation by Roger Reep and Robert Bonde (Reep and Bonde 2006) ................................ ................................ .......... 31 1 2 A manatee calf admitted for rehabilitat cold stress syndrome. Note the classic epidermal lesions. Photograph taken by Ashley Barratclough. ................................ ................................ ...................... 31 1 3 Zoo demonstrating hyperkeratinisation which is frequently associated with cold stress syndrome epidermal lesions associated with cold stress syndrome. Photograph by Ashley Barratclough. ................................ ................................ .......................... 32 1 4 Weight loss particularly marked around the head and neck resulting in the ...................... 32 1 5 Location of blood sample collectio n in the intraosseous space between the ulna and the radius. Photograph courtesy of Dr Francis Floyd. .......................... 33 2 1 Graphical representation to indicate the increased variability of factor act ivity levels between locations and the lack of variation between locations when comparing sex. ................................ ................................ ................................ ... 48 3 1 Thromboelastogram of a wild healthy manatee demonstrating the shortened R=1.7 minu ................ 62 4 1 The right intercostal artery; demonstrating the dramatic reduction in diameter, between the large artery and the subsequ ent blood vessels showing similar branching structure as that of a broom. Anatomical photo credit Rommel et al. (Rommel and Caplan 2003) ................................ ................................ ........... 84 4 2 A thromboelastograph of a cold stress sy ndrome calf in white superimposed over a normal wild manatee TEG. Note the reduction in R with clot initiation commencing after just 1.1minutes. Also the lines begin to diverge demonstrating a degree of fibrinolysis occurring and 30minutes and increasing t o 8.7% at 60 minutes. ................................ ................................ ....... 84 5 1 Example of how diagnostic tests can be used to aid identification and diagnosis of CSS. ................................ ................................ ............................... 89
10 LIST OF ABBREVIATIONS AT CS S DIC FWC HMWK IUCN Angle, represents r ate of clot formation Antithrombin activity Cold stress syndrome Disseminated Intravascular Coagulation Florida Fish and Wildlife Conservation Commission High molecular weight kininogen The International Union for Con servation of Nature K LY30 MA PK PT Clot formation time Percentage of clot retraction in 30 minutes Maximum amplitude reflects the final clot strength P rekallikrein Prothrombin time PTT R Partial thromboplastin time Reaction time or precoagulation time TEG USGS Thromboelastography Unites States Geological Survey
11 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfilment of the Requirements for the Degree of Master of Science ESTA BLISHING THE COAGULATION PROFILE OF THE FLORIDA MANATEE ( Trichechus manatus latirostris) AND IDENTIFYING COAGULOPATHIES IN THE PATHOPHYSIOLOGY COLD STRESS SYNDROME By Ashley Barratclough December 2015 Chair: Ruth Francis Floyd Major: Veterinary Medical S ciences The exact pathophysiology of cold stress syndrome (CSS) in the Florida manatee ( Trichechus manatus latirostris) was previously unknown. The condition was hypothesized as a nutritional, immunological and metabolic disturbance caused by prolonged e xposure to water temperatures <20Â°C. Following extensive research into the coagulation system, we confirmed that CSS involves a severe hemostatic disorder. We established the following mean results for normal coagulation parameters in 40 wild manatees ; pro thrombin time (PT) 10.8 seconds , partial thromboplastin time (PTT) 9.2 seconds, fibrinogen level 132mg/dl , and D dimer level 82ng/ mL . We found CSS cases in comparison had statistically prolonged PT, PTT, increased D dimer and fibrinogen levels and a reduc ed platelet count, consistent with disseminated intravascular coagulation (DIC). Furthermore we characterized the normal clotting process, performing coagulation factor assays in 20 wild manatees and thromboelastography (TEG) in 29
12 wild manatees. Wild m an atees were relatively hypercoagulable compared to other species. T he following mean (SD) normal TEG parameters were determined : reaction maximum amplitude MA = 75mm (7.6) and clotting lysis LY30 = 0.41% (0.68). In comparison we found CSS cases to show increased coagulability supporting our hypothesis of thromboembolic disease playing a role in the pathophysiology of CSS . We established that increased PTT, PT, D dimer, and fibrinogen levels and thrombocytopenia have a negative prognostic value in assessment of CSS cases . We propose that prolonged hypothermia results in a coagulopathy which is a component of the syndrome and may ultimately contribute or lead to the clinical s igns associated with CSS including epidermal bleaching, enterocolitis and anorexia.
13 CHAPTER 1 COLD STRESS SYNDROME REVIEW Introduction The Florida manatee ( Trichechus manatus latirostris ) is an aquatic mammal found in the tropical waters around the coa st of Florida. They are currently protected under the E ndangered S pecies act of 1973 , the Marine Mammal Protection Act of 1972, and the M anatee S anctuary A ct of 1978. The Florida manatee is one of four living species in the order Sirenia with the other thr ee species comprising the West African manatee ( Trichechus senegalensis ), the Amazonian manatee ( Trichechus manatus ) and the dugong ( Dugong dugon ) (Reep and Bonde 2006) . T he West Indian manatee contains two subspecies , the Florida manatee ( Trichechus manatus latirostris ) and the Antillean manatee ( Trichechus manatus manatus ). Cold stress syndrome (CSS) has been defined as mortality resu lt ing from prolonged exposure to water temperatures <20Â°C (Bossart et al. 2002) . It is o ne of the main causes of natural mortality, accounting f or 18% of annual mortality rates. Despite the importance of the disorder , t he pathophysiology of CSS i s not well understood ( Bossart 1999 ) . Bossart et al. 2002 explain ed death attributed to cold exposure as multisystemic organ failure resulting from life threatening opportunistic infectious disease that was preceded by nutritional, immunological and metabolic di sturbance s. Biology The current Florida manatee population is thought to have originated from Caribbean stock over the last 12,000 years . The natural warm water sources available in Florida allowed manatee s to expand their range north from the C aribbean and ultimately the Floridian subspecies evolved . Currently the geographic range is
14 established as the entire coast of Florida , however individuals have been noted to stray as far north as Ca pe Cod and as far West as Texas (Reep and Bonde 2006) . Manatees have a very low natural metabolic rate, similar to their closest living non sirenian relative the elephant. Their low metabolism predis poses them to effects of temperature with their energetic requirements inhibiting their geographical distribution (Irvine 1983) . It has als o been shown that juvenile manatees have lower circulating thyroid levels , confirming an even lower basal metabolic rate , further predisposing them to the detrimental effects of hypothermia ( Worthy et al. 1999 ) . This low basal rate results in a low thermoneutral zone with the lowest water temperatures for survival being 20 23Â°C. This is significantly higher than other marine counterparts such as the bottlenose dolphin which can survive readily in temperatures 8 15Â°C (Costa and Williams 1999) . D espite their large size with average weight and length at 500kg and 3meters , the due to their reliance on warm water refuges for thermoregulation when water temperatures drop below 20Â°C . This dependence on warm water refuges results in a necessary seasonal migration to warmer water in the colder winter months. Historically the manatee s relied on natural thermal springs for warm water, however in recent decades warm water power plant effluents have provided a man made alternative. Unlike many other marine mammals, manatees are a primarily tropical species and therefore have not adapted to survive cold temperatures ( Bertram and Bertram 1973 ) . The closest living relatives to the Sirenia are the families of the rock hyrax (Hyracoidea) and the elephant (Proboscidae ) ( Domning 2001 ) . The latter is particularly impor tant when considering the lo w metabolic rate of the manatee. I t is likely that the
15 low basal metabolic rate and fairly high lower critical temperature contribute to morbidity and mortality following exposure to low temperatures (Irvine 1983, Worthy et al. 1999) . Further, the blubber anatomy is unlike that found in other ma rine mammal groups. It does not have a high insulative value and consequently may also predispose the species to hypothermia. Other marine mammals are more adapted to colder waters . For example, the harbor porpoise (Phocoena p hocoena) has a high lipid con tent within the blubber layer and therefore a low conductivity factor, facilitating retention of heat ( Worthy and Edwards 1990 ) . Despite the manatee dermal layer being relatively thick, it has relatively low lipid content, and therefore a high conductivity factor (Ames et al. 2002) . The high conductivity factor results in poor heat conservation, enabling heat to be easily lost in cold water. Adaptations that may enhance heat conservation include a counter current exchange system within the vasculature. There is a high ratio of veins to arteries at 2:1 facilitating each artery to be surrounded by veins. This close approximation enables a c oun ter current exchange system between the arterial and venous blood facilitating heat conservation ( Rommel and Caplan 2003 ) . In addition , the adaptation of arteriovenous anasto moses , are additional anatomical specializations that facilitate heat conservation , by the transfer of arterial heat back into the body via the venous system. Ultimately this facilitates reduced heat loss from the extremities. Despite these adaptat ions exp osure to prolonged cold is detrimental to manatee s . The manatee has unique anatomy with their pleural cavity compris ing the entire dorsal compartment . This creates a particularly large abdominal cavity facilitating a large digestive system , as illustrated in Figure 1 1 . Manatees survive on an herbivorous
16 diet , consuming a wide variety of both freshwater and saltwater plants. Their prehensile upper lip aids consumption of 10 % of their body weight in plant material per day (Walsh and Bossart 1999) . Manatees are hi ndgut fermenters like elephants , and require that a large volume of various sea grasses be consumed daily ( Reynolds and Rommel 1996 ) . Due to their slow metabolism food transit time is approximately 7 10 days , allowing a large generation of heat, a significant contributor to effective thermoregulation (Rommel et al. 2003) . Mortality The mai n anthropogenic cause of manatee mortality in Florida is trauma as a result of watercraft collisions as demonstrated in Table 1 1 . Brevetoxicosis and perinatal mortality are additional causes of mortality reported annually by FWC. Cold stress syndrome has been shown to be a major cause of natural mortality in the Florida manatee accounting for approximately 18% of annual mortality ( Bossart et al. 2004 ) . Morbidity and mortality caused by CSS can be highly variable due to the severity a nd length of cold weather events. Acute changes in temperature can result in exposure to extreme cold before a suit able migration to a warm water site can occur. In addition , prolonged cold weather periods can result in increased cold stress syndrome due to cold exposure when leaving the warm water refuges to facilitate feeding . The FWC operates a carcass salvage progr am, performing necropsies on all recovered carcasses either in the Marine Mammal P atho bio logy Laboratory or in the field if too remote to remove the carcass. The most commonly reported cause of mortality for each year is reported in Table 1 1. Over the for ty years that the carcass recovery program has been in operation , severe cold weather has resulted in multiple
17 deaths as a result of CSS. At the start of the carcass recovery program, cold stress syndrome was still poorly defined and it is not clear that a ll possible cases would have been correctly categorized. In recent years the numbers of CSS mor t alities have increased from only 14 deaths confirmed in 2000, to 50 deaths in 2004, 282 deaths in 2010 and 114 deaths in 2011 (FWC 2015) . The winter of 2010 2011 was particularly cold which accounted for the increased mortality (Barlas et al. 2011) . This increase in reported mortality for C SS may be due to a more specific definition of the condition, and improved recognition by recovery personnel. However increased deaths were also a result of extreme, prolonged cold temperatures. Only 26 deaths were confirmed as a result of CSS in 2014 and 17 in 2015 (November) indicating the winter of 2014 2015 was relatively mild. Unusual Mortality Events An unusual mortality event is described unde r the Marine Mammal Protection A ct as off of any marine mammal population; and demands im mediate Cold stress syndrome was dete rmined to be responsible for two recent unusual mortality event s (UME s ) in winter of 2009 2010 and winter of 2010 2011 where 252 manatees deaths were attributed to cold exposure (Barlas et al. 2011) . Further, an additional 228 carcasses wer e recovered during this period , which were too decomposed to make a definitive determination about the cause of death . There were two phases to the 2010 2011 UME which lasted a total of 89 days . The first phase consisted of acute mortality resulting from b rief exposure to extreme cold <15Â° with 10 carcasses being recovered on average per day for the first 22 days. The second phase consisted of chronic mortality following a period of prolonged cold exposure which resulted in a total of 181 carcasses over a t wo month period . The start of the UME
18 correlated with the coldest water temperatures recorded in this period of 10Â°C.This extremely low temperature resulted in the large numbers of acute deaths due to severe hypothermia (Barlas et al. 2011) . Spatial and temporal UME data analysis corrobo rated with the geographical locations of warm water refuges where lower levels of mortality were reported (Barlas et al. 2011) . All carcasses confirmed as cold stress mortality in th e s e UME s demonstrated skin lesions. These ranged from mild epidermal lesions such as bleaching of the epi dermis (Figure 1 2) to large ulcerations and abscesses (Barlas et al. 2011) . The cobblestone skin (Figure 1 3) appearance was frequently observed with mild to severe hyperkeratinisation of the epidermis. Thickening of the epidermis was particularly prevalent at the extremities . Bossart e t al. (2002) suggested that this may be an adaptive response to the reduction in temperature. Serous fat atrophy and chronic emaciation was understandably less apparent in the acute mortality phase but was consistently present in the chronic cold stress ca ses. The post mortem figures supported the demographic of the manatees affected with a greater representation by the smaller manatees with 58% of the carcasses being juvenile manatees (Barlas et al. 2011) . Demography The first abundance estimate conducted on the Florida manat ee was con ducted in 2011 and 2012 and estimate d 6350 individuals ( Martin et al. 2015 ) . Previously performed synoptic surveys counted 3802 manatees in 2009. Despite an apparent population increase the manatee is still classed as endangered ( Martin et al. 2015 ) . This consistent classification is partially due to the stochastic nature of the mor tality patterns with large population die offs occurring periodically due to harmful algal blooms and cold
19 stress syndrome as demonstrated in Table 1 1 . The demographic most affected by cold stress syndrome are the juveniles and sub adults ( O'Shea et al. 1985 ) , which was confirmed in the 2010 UME data . The reasons for this are multifactorial. The c urrent guidelines to indicate ag e categories during physical examination are length of >260cm adult, 236 260cm juvenile, and <236cm calves (Bonde et al. 2012) . In smaller animals there is a reduced surface area to volume ratio resulting in increased heat loss , which can challeng e homeo stasis in cold environments ( Bossart et al. 2004 ) . Young animals may be experiencing cold wat er independently for the first time. T his inexperience may inhibit suitable migration to warm water sites when water temperatures fall. Social factors , such as learning to survive without the guidance of the mother , may also be influential. The minute numb er of perinatal animals affected is primarily due to the fact that they are still nursing and are subsequently guided to appropriate warm water refuges by the dam. The lack of perinatal deaths by cold stress syndrome and the high preponderance of juveniles supports this hypothesis ( O'Shea et al. 1985 ) . As previously mentioned, juveniles have reduced body mass and fewer fat reser ves . Thus, subsequent serous fat atrophy a common result of cold stress syndrome will have an even greater detrimental effect on thermoregulation. Inadequate ac c ess to forage may contribute to weight loss, exacerbating the condition. Early separation f rom the mother or the inability to locate sufficient food sources will increase susceptibility of young animals to CSS . As gut digestion can provide a large proportion of heat generation, the potential of the younger animals having a lower food intake coul d therefore reduce the fermentation furnace efficiency of the gastrointestinal system. Incomplete weaning would also result in lower forage intake and reduced heat
20 production. The reduced ability of juveniles to adjust their metabolic rate for cold exposur e is a major factor in their susceptibility to cold stress ( Worthy et al. 1999 ) . In conclusion , a reduction in nutritional status compounded with a small surface area exacerbates the metabolic drain on these individuals (Irvine 1983) . Warm water refuges Previous studies into the winter habitat preferences of the Florida manatee have shown how manatees rely on one of three types of warm water sources ( Laist et al. 2013 ) . These are natural springs, passive thermal basins or warm water effluent from power plants . The ideal wa rm water refuge should have a short transit time to access foraging sites and minimize cold exposure. Cold exposure coupled with lack of forage has been shown to have subclinical effects on surviving manatees in addition to those which succumb to CSS ( Wilson et al. 2011 ) . S ubclinical consequ ences in survivors includ e increased vulnerability to other stressors , and reduced reproductive performance ( Wingfield and Sapolsky 2003 ) . The very large numbers of manatees aggregating at warm water sites du ring cold weather increases the d emand on the available food sources . Longer commutes may be needed to reach forage areas, potentially increasing cold water exposure. The geographical range of cold stress mortality has been widespread throughout Florida. The temperatures within a warm water area can fluctuate , providing varying degrees of protection against cold temperatures . Warm water springs are the most consistent in protecting manatees from declining water temperatures in the wi n ter. Pow er plants may be u nable to elevate discharge temperatures enough to raise ambient water temperatures during significant cold spells ( Laist et al. 2013 ) . As a result , the levels of mortality between locations can vary significantly depending on weather
21 condition s and the quality and availability of a refuge site ( Langtimm and Beck 2003 ) . The largest mor talities attributed to cold stress were recorded in the record breaking winter of 2009 2010 (Stith et al. 2012) . As all power plants currently used by manatees in Florida are over 35 years old , the current population has adapted to relying on their warm water supply. The site fideli ty shown by the manatees to a particular warm water location has resulted in the formation of four somewhat discrete subpopulations of manatees within the state of Florida ( Laist et al. 2013 ) . In a recent winter survey it was found that almost 50% of the current population were residing at power plants ( Laist et al. 2013 ) . There are several proposed power plant closures in the near future which could have dramatic effects on these populations and their offspring , who have learned to migrate to the se locations (Laist and Reynolds 2005b) . Due to the longevity of the operation of power plants , manatees habituated to these sites may be unaware of an alternative warm water source if this refuge is withdrawn. The ris k of succumbing to cold stress syndrome is linked directly to the duration and intensity of the cold weather and the quality of forage within , or adjacent to, the water tempera tures drop to 18 19Â°C (Campbell 1981) . If manatees have to venture out of the warm water refug e to feed they experience increased exposure to hypothermia. Remaining in warm water without feeding may result in starvation , causing a cascade of nutritional deficiencies, decreased energy reserves, and a reduction in the thermogenic activity of gut mic roflora (Rommel et al. 2003) .
22 Clinical Signs of Cold Stress Syndrome Obvious clinical symptoms commonly include bleached epidermis with varying degrees of severity of ulceration a nd abscessation (Figure 1 2) , predominantly confined to the extremities ( Buergelt et al . 1984 ) . In addition to the epidermal findings , behavioral signs observed in the live manatee include shivering followed by a reduction in activity and floating at the surface of the water ( Walsh and Bossart 1999 ) . D ecreased , or a complete cessation of feeding activity may be observed early in the disease proces s, therefore chronically affected individuals are oft en underweight . Extreme weight loss is most evident in the head and neck region , and may be descri (Figure 1 4 ) . On physical exam a reduction in heart rate and respiratory rate can be seen with oral body temperatures as low as 25Â°C. Normal parameters in the manatee include a heart rate of 50 60bpm , respiratory rate of 1 breath per minute and core body temperature of 36Â°C (Wong et al. 2012) . Dyspnea indicated by shallow , rapid breaths (>10 per 5 minutes) is often seen when the animal is severely compromised (Barlas et al. 2011) . Marked lethargy is invariably present , however in rescued manatees neurological deficits as a result of extreme hypothermia are rarely seen, most likely because these animals die prior to rescue (Barlas et al. 2011) . Current Therapy Historical l y, the pathophysiology of cold stress was poorly understood and treatment largely based on providing supporti ve care ( Bossart et al. 2002 ) . M ost manatees admitted with signs of cold stress syndrome will be nutritionally compromised . G radual return to f ood is advised to minimize colic and to allow the gut microflora to re establish. This can be achieved by passing a gastric tube through the nostril to allow
23 administration of water, electrolytes, and oral medications. The amount of food offered initially will be strictly controlled to prevent excessive initial consumption. A blood sample is obtained from the capillary bed located in the interosseous space between the ulna and radius from either pectoral flipper as demonstrated in Figure 1 5 (Walsh and Bossart 19 99) . Following sterile preparation using betadine and alcohol gauze , a 20 gauge 1.5 inch needle is inserted and attached to a wing tipped blood collection set. Standard collection involves serum and EDTA tubes for biochemistry and hematology analysis wit h the addition of citrated plasma to allow coagulation analysis (Harvey et al. 2007, Harvey et al. 2009) . Laboratory tests that can be immediately performed include packed cell volume, total protein and blood gluco se. Hypoglycemia is particularly important to identify in cold stress cases. Blood work frequentl y demonstrates low numbers of heterophils and lymphocytes . H owever , white blood cell values in manatees appear to show similar trends to the bovine species wh ere leukocyte counts will only elevate during very severe disease (Harr et al. 2006, Reep and Bonde 2006) . As a result white blood cells are not a sensitive indic ator of infection or inflammation. Alternative tests such as serum amyloid A are now being measured in the manatee as they have improved specificity and sensitivity for inflammation (Harr et al. 2006, Cray et al. 2013) . Increased serum amyloid A exceeding >1200Âµg/ mL was demonstrated to carry a poor prognostic indic ation which can be used to aid clinical decision making (Harr et al. 2006) . Pneumonia is a common sequelae to cold stress syndrome . N asal discharge may indicate an upper respiratory tract infection and antibiotics are f requently indicated. The respiratory system was affected in approximately 60% of 188 CSS cases examined
24 in 2001 (Bossart et al. 2002) . B ronchopn eumonia is common in acutely affected individuals that may be early in the course of the disease , showing little to no change in the epidermis. Epidermal bleaching and subsequent hyperkeratinization and abscessation are more chronic changes (Barlas et al. 2011) . On arrival at the rehabil itation centers CSS manatees will usually receive broad spectrum antibiotics to treat both local abscesses and systemic infections ( Bossart 1999 ) . In cases w i th extensive epidermal sloughing, t opical treatments such as dimethyl sulfoxide may reduce cellulitis and minimize further infection. Non steroidal anti inflammatories can be used to provide both a reductio n in inflammation in the epidermis and some pain relief for the abscessation . Pathological Findings The p athological features of cold stress syndrome were first identified and described in 2002 ( Bossart et al. 2002 ) . Following chronic exposure to cold water <20Â°C several pathological conditions were observed that supported a clinical diagnosis of CSS . In the winter of 2000 2001, 188 manatees were necropsied and diagnosed with CSS (Bossart et al. 2002) . Consistent necropsy findings in these 188 cases were poor body condition, frequently d emonstrating emaciation, with internally depleted fat stores and serous fat atrophy. These pathological findings are consistent with insufficient food stores within the warm water refuge sites as previously described ( Flamm et al. 2013 ) . Internal pathology consistently associated with col d stress primarily involves the gastrointestinal system with a recurrent finding being little to no ingesta throughout the length of the gastrointestinal tract ( O'Shea et al. 1985 ) . This is consistent with the cessation of feeding attributed to both hypothermia and decreased access to forage .
25 The opposite was apparent however in the closely related dugong ( Dugong dugon) with CSS where large amounts of sea grasses were found in the stomach at post mortem. The explanation for this is likely due to the fact that sea grasses are present in the dugong ' s refuge warm water zones whereas manatees usually must leave the warm water refuge to forage ( Owen et al. 2013 ) . The lack of food in combination with hypothermia in the Florida manatee appears to predispose them to the cascade of events frequently observed during CSS such as immunosuppression, epidermal lesions and metabolic abnormalities. Interestingly the dugong appears to have improved adaptation to the colder water temperatures, wit h temperatures as low as 18Â°C being required before movement to warmer areas is initiated and subsequently appear to be less predisposed to CSS (Owen et al. 2013) . Immunosuppression with reduced lymphocyte numbers has been documented in cold stress manatees and ma y predispose affected animals to systemic bacterial infections ( Walsh et al. 2005 ) . Lymphoid depletion of the mucosal associated lymphoid tissue (MALT) has been reported in CSS and was moderate to severe in most cases. Depletion of lymphoid tissue was consistent ly observed in all peripheral lymph nodes and the spleen ( Bossart et al. 2002 ) . Both the cell mediate d and humoral immune systems are affected by the lymphoid depletion , predisposing to secondary infections (Halvorsen and Keith 2008) . Enterocolitis, often with dried black f ecal material observed in the gastrointestinal tract, is also reported in manatees suspected of succumbing to CSS. The feces are hemacult positive indicating prior bleeding during passage through the intestines. This is significant when considering the pos sibility of a coagulopathy in the pathophysiology of
26 CSS. Although thromboembolisms can be challenging to diagnose grossly at necropsy they are a frequent finding in histopathology (Marder et al. 2012) . Myocardial degeneration has also been a consistent finding in cold stress syndrome. This was characterized histologically as attenuated myofibres and extensive cytoplasmic vacuolation ( Buergelt et al. 1984 ) . Occasionally mild interstitial fibrosis was also present. A possible etiology for the cardiac changes could be the link to the nutrit ional disturbance l eading to cachexia from the advanced emaciation ( Bossart et al. 2004 ) . Th e presence of cardiorespiratory, gastrointestinal and lymphoid pathology supports the likelihood of a systemic inflammatory response and subsequent multi systemic organ failure. Exposure to cold has resulted in a similar effect in other species such as col d stunning in sea turtles ( Anderson et al. 2011 ) . Pathology r eports of cold stunned sea turtles between 2001 2006 found that hypothermia resulted in a multitude of pathologies (Innis et al. 2009) . The most frequently observe d pathologies included pneumonia and necrotizing enterocolitis. Additional contributing factors were described as respiratory, metabolic and electrolyte derangements as a result of hypothermia with drowning frequently being the ultimate cause of death (Innis et al. 2009, Keller et al. 2012) . Mitigation Due to cold stress syndrome being a natural phenomenon, mitigation is particularly challenging. Manatees have adapted to reliance on power plant refuges as a source of warm water during cold weather for more than fifty years . It may require a similar time frame rks of warm water springs and thermal basins as they have in the historic context ( Laist et al. 20 13 ) . Disease attributed to hypothermia is linked to the duration as well as the intensity of
27 cold water exposure , as well as access to quality forage ( Flamm et al. 2013 ) . In the more northern sites , where the cold can be more extreme , it is even more important that forage is accessible without having to spend a large amount of time away from the warm water source ( Zoodsma 1991 ) . Coagulation Before the possible presence of an underlying thromboembolic component to cold stress syndrome could be explored it wa s necessary to describe the coagulation parameters in normal, healthy animals (Chapter 2). T hromboelastography (TEG) is a tool that allows assessment of the rate of clot formation and the strength of the clot formed. Measurement of D dimer concentrations a llows detection of prior thromboses by measuring fibrin fibrinogen degradation products (FDPs) ( Stokol et al. 2000b ) .High levels of D dimer concentration >500ng/ mL are indicative of prior thromboses. Assessment of the more routine coagulation tests prothrombin (PT) and partial thromboplastin time (PTT) are important however they lack the specif icity and sensitivity of D dimer concentrations and thromboelastography (Massignon et al. 1996) . Fibrinogen is the final parameter to complete the coagulation panel. This is a soluble protein wh ich is broken down by the enzyme thrombin into fibrin during the clotting cascade. Fibrinogen is the primary protein responsible for platelet aggregation ( Marder et al. 2012 ) . Very l ow levels of fibrinog en are indicative of DIC and high levels are often associated with increased clot formation. High levels must be interpreted with other coagulation tests as fibrinogen is also an acute phase protein therefore high levels can result from inflammation or inf ection ( Marder et al. 2012 ) . Coagulation factors VII through XII activity levels will also be determined to complete the cascade analysis. By assessing all these coagulation parameters in cold stress syn drome cases and
28 comparing these results to previously established normal reference ranges in wild manatees, improved understanding of the pathophysiology of cold stress syndrome will be obtained. TEG enables evaluation of the entire clot formation pro cess rather than a specific pathway such as PT and PTT. In veterinary medicine reference intervals for TEG have been established in dogs, horses and cats ( Wiinberg et al. 2005 , Epstein et al. 2009 , Hall et al. 2012 ) . Recently, TEG has been used to diagnose hypercoagulability, hypocoagulability and disseminated intravascular coagulation during various disease processes including parvovirus, immune mediated hemolytic anemia, and sepsis ( Otto et al. 2000 , Wiinberg et al. 2008 , Sinnott and Otto 2009 , Bentley et al. 2013 ) . TEG could identify possible coagulopathies occurring during CSS that routine coagulation panel tests could be too insensitive to detect. Summary M anatees are susceptible to death from cold stress syndrome when water temperatures remain below 20Â°C f or a prolonged period of time. This condition is recognized as a leading natural cause of mortality in the Florida manatee ( Laist et al. 2013 ) , accounting for as much as 18% of annual mortality. A clinical diagnosis of cold stress syndrome is confirmed when there is a combination of advanced emaciation, characteristic skin lesions, and a correlated prolonged cold weather period ( Bossart et al. 2004 ) . At post mortem, enterocolitis, lymphoid depletion, serous fat atrophy and myocardial degeneration supp ort the diagnosis of cold stress syndrome ( Bossart et al. 2004 ) . Due to the decreased availa bility of warm water refuge during winter months in Florida, the incidence of cold stress will likely increase in the future ( Bossart 1999 ) .
29 Examining the migration patterns d uring winter months and behavio rs of the Florida manatee is essential to monitoring the predisposition to the condition ( Laist et al. 2013 ) . Understanding the pathophysiology and fac ilitating effective treatment is therefore paramount to reduce this cause of mortality. Current treatment is mainly by supportive care involving oral nutritional and fluid therapy as well as appropriate antibiotic therapy via culture and sensitivity result s when practical ( Dierauf and Gulland 2001 ) . Consistent surveillance of post mortem findings of cold stress syndrome will help to convey the message from the dead to the living, to improve our understanding of the complex pathophysiology. Increased understanding of cold stress syndrome, will aid with conservation of this endangered species to ensure their long term survival, which is our ultimate goal.
30 Table 1 1. Total mortality reports of the Florida manatee from 1974 to 2015 provided by the FWC carcass recovery program . Total number of deaths per year and the greatest cause of mortality identified for each year. Year Total overall m ortality count Greatest cause of mortality: Year Total overa ll mortality count Greatest cause of mortality: 1974 7 Watercraft 1995 201 Perinatal 1975 29 Perinatal 1996 415 Red Tide 1976 62 Perinatal 1997 242 Perinatal 1977 114 Watercraft 1998 232 Watercraft 1978 84 Watercraft 1999 269 Watercraft 1979 77 Water craft 2000 272 Watercraft 1980 63 Watercraft 2001 325 Watercraft 1981 116 Watercraft 2002 305 Red Tide 1982 114 Watercraft 2003 380 Red Tide 1983 81 Watercraft 2004 276 Perinatal 1984 128 Watercraft 2005 396 Red Tide 1985 119 Watercraft 2006 417 Wate rcraft 1986 122 Watercraft 2007 317 Red Tide 1987 114 Watercraft 2008 337 Perinatal 1988 133 Watercraft 2009 429 Cold Stress 1989 168 Watercraft 2010 766 Cold Stress 1990 206 Watercraft 2011 453 Cold Stress 1991 174 Perinatal 2012 392 Watercraft 199 2 163 Perinatal 2013 830 Red Tide 1993 145 Perinatal 2014 371 Perinatal 1994 193 Watercraft 2015 (Oct) 344 Perinatal
31 Figure 1 1. Internal anatomy of the Florida manatee, note the horizontal diaphragm resulting in a large surface area of digestive o rgans. Adapted from The Florida Manatee Biology and Conservation by Roger Reep and Robert Bonde (Reep and Bonde 2006) Figure 1 2 . cold stress syndrome. Note the classic epidermal lesions. Photograph taken by Ashley Barratclough.
32 Figure 1 3 . Lowry Park Zoo demonstrating hyp erkeratinisation which is frequently associated with cold stress syndrome . Photograph by Ashley Barratclough . Figure 1 4. Weight loss particularly marked around the head and neck resulting in the Photograph by Ashley Barratcl ough.
33 Figure 1 5. Location of blood sample collection in the intraosseous space between the ulna and the radius. Photograph courtesy of Dr. Francis Floyd.
34 CHAPTER 2 ESTABLISHING THE NORMAL COAGULATION FACTORS IN WILD FLORIDA MANATEES ( Trichechus manatus latirostris ) AS A BASIS FOR EXAMINING THE PATHOPHYSIOLOGY OF THEIR HEMOSTATIC DISORDERS. Determination of normal coagulation factor concentrations in the Florida manatee ( Trichechus manatus latirostris) is important to facilitate further underst anding of potential hemostatic disorders. During annual wild manatee health assessments organized by U.S. Geological Survey (USGS) in November and December 2014 citrated blood samples were collected for coagulation analysis. Twenty samples were collected i n total, with 10 animals from the east coast and 10 from the west coast of Florida. The following analyses were performed: prothrombin (PT), partial thromboplastin time (PTT), and concentrations of fibrinogen, D dimers and factors VII, VIII, IX, X, XI and XII. Manatees were found to have relatively high levels of factor VIII activity (134%) and relatively low factor VII activity (106%). Shortened PT results (mean ( s.d. )) of 9.2 (1.6) seconds and PTT 10.8 (0.5) seconds indicate a lack of hypo coagulability. D Dimer results of 82 (65) ng/ mL confirmed a lack of prior thromboembolic disease. Further research is required to determine how hemostatic abnormalities of the manatee coagulation system co uld influence disease processes. Introduction Recent studies perfor thromboembolic disease in the pathophysiology of cold stress syndrome in the Florida manatee. Prior to assessing a potentially abnormal state in diseased animals it is necessary to establish normal coagulation animals. An effective coagulation system requires a delicate balance between clot breakdown and clot formation . This highly complex system is the ultimate physiological
35 defense against traumatic injury of blood vesse ls and prevents excessive hemorrhage. The basic mechanisms of initiation of clot formation, post endothelial damage, are comparable across species ( Gentry 2004 ) . The level of variation in quantitative differences between species in regards to thrombin generation and fibrin f ormation supports the demand for species specific analysis of coagulation systems. Studies of the blood coagulation system in humans are extensive, primarily due to the identification and treatment of congenital and acquired hematologic disorders. In vete rinary medicine, detailed information on the coagulation system is only available for a few domesticated species ( Hawkey 1975 ) . An important reason that these analyses are not pursued mo re frequently is that plasma factors are often species specific and therefore require individualized tests to measure them. This is both expensive and time consuming, which limits further investigation. Previous work has found that with the exception of fa ctor XII in Cetaceans ( Robinson et al. 1969 ) and the lack of factor XI in the Greater Kudu ( Hawkey 1975 ) , the clotting factors which have bee n identified in humans can be found in other mammals. The level of activity of these factors varies between species , making quantitative comparison complicated. Increased factor activity in non human mammals results in faster in vitro clot formation. This can be readily identified using tests such as thromboelastography and whole blood clotting time. Interestingly , there is no evidence that a relative increase in coagulability predisposes wild mammals to spontaneous thromboembolic diseases which cause high rates of morbidity in humans ( Finlayson 1965 ) . The Florida manatee ( Trichechus manatus latirostris) is an obligate aquatic mammal of the order Sirenia ( Reep and Bonde 2006 ) . Manatees reside in the warm
36 coastal waters of Florida and can survive in both salt and fresh water. Studies of the hematology of marine mammals are sporadic and are particularly rare in manatees. The large erythrocytes in man atees were first described in 1976 by White et al. and are comparable to those found in the close relative, the African elephant ( Loxodonta Africana) ( Simon 1961 , White et al. 1976 ) . The large erythrocyte size was origin ally thought to be linked to the size of the overall animal. However the presence of large erythrocytes in manatees and the rock hyrax has confirmed that this is likely due to origin from the same ancestor rather than having a direct relation to gross bod y size ( Lewis 1996 ) . Factors II, V, VII, VIII, IX, X, XI, and XII were measured in 10 long term captive manate es previously by Medway in 1982. In this study healthy normal canine plasma was used as a reagent control and as an established standard for comparison to the manatee as there were no previous Sirenia coagulation studies for comparison. It was important fo r the basis of our study to use wild, rather than captive, manatees from various locations to facilitate establishing the normal reference values for the Florida manatee. Medway et al. 1982 found that all known human clotting factors (identified at that t ime) were present in the manatee, unlike cetacean s which lack Factor XII ( Robinson et al. 1969 ) . Most marine mammals are carnivorous, however the s irenians are herbivorous and subsequently are perhaps more suita bly compared to the elephant. Medway et al. (1982) found that manatee factor VIII was nine times higher (889Â± 387%) than in normal dogs, with factor IX seven times higher (712 Â± 195%), and factor X activity to be equivalent to a normal dog. Increased fact or XI and XII activity was also
37 present indicating rapid activation of the intrinsic clotting cascade enabling rapid initiation of clot formation. Factor XII was found to be 6 times higher than in dogs, which is particularly interesting conside ring the lac k of factor XII in c etaceans. This could be linked to the deep diving physiology adaptations in cetaceans. From an evolutionary standpoint, Factors XI and XII are derived from ancestral proteins that originally were related to digestive enzymes ( Patthy 1990 ) . It is interesting to consider that the great variation in digestive physiology may be ultimately responsible for some of the variation in hemostatic proteins we are discovering in different species ( Gentry 2004 ) . Elephant coagulation studies have shown that blood coagulation in vitro is more rapid in Indian ( Elephas maximus) than in African elephants ( Hawkey 1975 ) . Like manatees, elephants have been reported to have similarly high levels of factor XII ( Gentry et al. 1996 ) . The fact that both elephants and manatees have large erythrocytes, and can live to a considerable age with no evidence of age related thromboembolic problems, indicates an appropriately balanced hemostatic system. Further studies are required to investigate the exact m echanisms in vivo , because laboratory studies so far have found that elephant fibrin clots formed in vitro do not lyse ( Lewis 1996 ) . In vitro hemostatic studies are clearly insufficient for fully describing the coagulation and fibrinolytic activity in these species and further studies are needed to better understand these systems in health and disease. The most fr equently performed coagulation tests in veterinary medicine are PT (prothrombin time) and PTT (partial thromboplastin time) with more emphasis being placed on D dimers and fibrinogen in recent years ( Stokol et al. 2000a ) . Measuring PT enables evaluation of factor VII and the traditionally referred to extrinsic coagulation
38 pathway, w hereas PTT evaluates the intrinsic pathway , assessing the activity of factors VIII, IX, XI, and XII. The function of factors I, II, V and X are assessed by both PT and PTT and is traditionally referred to as the common pathway. While PT and PTT have been u tilized in a wide variety of species they are actually the least sensitive and specific tests when attempting to identify a mild coagulopathy, with changes only occurring when severe coagulopathies are present (Massignon et al. 1996) . Prolonged PT and PTT are often indicative of hypocoagulation however the results are very insensitive in detection of hypercoagulation (Donahue and Otto 2005) . Consequently , when interested in identifying hypercoagulation it is more beneficial to perform these tests as part of a coagulation panel and interpret the results along with D dim er concentration, fibrinogen levels and platelet count to get a more comprehensive analysis of the coagulopathy (Fenty et al. 2011) . Thrombin cleaves fibrinogen to form fibrin which functions in clot stabilization. D dimers measure the cross linked fibrin degradation products giving an indication tha t clot formation has occurred. Increased D dimer levels, increased fibrinogen concentration and reduced platelet count give a more accurate indication of the presence of thromboembolic disease than relying on PT and PTT alone (Donahue and Otto 2005) . Individual coagulation factor activity concentrations are often overlooked. The major limitation in the availability of measuring coagulation factor activity levels lies in the lack of individual reagent for th at species. T he assumption is that there will be significant structural homology between the species being measured , and human proteins is relied on for the test to function. This enables assessment of any hemostatic abnormality or factor deficiency relative to human factor deficient plasma. Unfortunately
39 this makes the results unreliable when trying to determine the absolute activity for an alternative species ( Gentry 2004 ) . By determining the m anatee coagulation factor concentrations against a pooled plasma sample of normal healthy manatees the relative normal factor concentrations for this species can be establish ed . A more recent study into the hematology of the manatee confirmed hematologica l variation between clinically normal wild and captive manatees. Captive manatees had lower MCV, MCH and eosinophil counts but higher heterophil, nucleated red blood cells, and fibrinogen concentrations ( Ha rvey et al. 2009 ) . Florida manatees are kn own to have heterophils with pink to red granules rather than neutrophils with nearly colorless granules. Despite hematological studies in manatees there is a lack of coagulation data on wild manatees, which prom pted th e present investigation. By establishing the normal concentrations of coagulation factors in wild manatees, we will then be able to apply this tool in the clinical investigation of debilitated animals, incl uding cold stress syndrome and brevetoxicos is cases. Method s All wild manatees were selected via opportunistic random sampling as part of the ongoing USGS health assessments survey ( Bonde et al. 2012 ) . Two sample locations were utilized to give a broader population representation, Brevard County on the East coast of Florida and Crystal River (Citrus County) on the West coast. In November 2014 10 animals were sampled in Crystal River and in December 2014 10 animals were sampled in Brevard County. Wild manatees were deemed healthy according to phys ical exam, activity level and results of biochemistry and hematology analysis ( Bonde et al. 2012 ) .
40 Blood samples were obtained via standard manatee phlebotomy techniques from the brachial plexus between the radius and ulna from either pectoral ( Walsh and Bossa rt 1999 ) . Following standard sterile preparation using alternating alcohol and betadine gauze swabs , a 20 gauge 1.5 inch needle was inserted and attached to a wing tipped blood collection set. An initial non additive tube was obtained to prevent tissue co ntamination of blood, followed by two anticoagulated 0.32% sodium citrate tubes to obtain plasma. Sampling processing was consistent with standardized procedure throughout considering collection site, preparation of site, tube collection order and processi ng. Following centrifugation remaining plasma was harvested and maintained at 80Â°C. All samples were sent to The Comparative Coagulation Laboratory at Cornell University for analysis within 6 weeks of collection. Assays were performed for coagulation fac tors VII through XII as well for PTT, PT and concentrations of fibrinogen, antithrombin and D dimers. A coagulation panel consisting of PTT, PT, and fibrinogen was performed using an auto mated clot detection instrument ( STA Compact, Diagnostica Stago, Pars ippany, NJ ) commercial reagents (Dade Actin FS, Dade Behring, Newark, DE, Thromboplastin LI, Helena Diagnostics, Beaumont, TX, Fibrinogen, Diagnostica Stago, Parsippany, NJ) and reaction conditions as previously described ( Stokol et al. 2000a ) . Antithrombin activity (AT) was measured in a functional assay configured to measure thromb in inhibition (anti IIa assay) using a commercial chromogenic kit ( Stachrom ATIII, Diagnostica Stago, Parsippany, NJ ) and ( STA Compact, Diagnostica Stago, Parsippany, NJ ) . The standard curves for determination of clott able fibrinogen ( Clauss 1957 ) and antithrombin activities in the test plasmas were derived from a c alibrated human plasma
41 standard (STA Unicalibrator, Diagnostica Stago, Parsippany, NJ). D dimer concentration in ng/ mL was measured using a quantitative, immuno turbidometric method as previously described, ( Delgado et al. 2009 ) using a commercial kit and the manufac dimer standards ( HemosIL, D dimer Calibrator, Instrumentation Laboratory, Bedford, MA ). For coagulant activity assays, a standard curve was generated from serial dilutions of pooled manatee plasma, previously prepared from 8 wild healthy manatees. The coagulant activities of Factors VII and X (FVII:C and FX:C) were measured in one stage PT and activities of Factors VIII, IX, XI, and XII (FVIII:C, FIX:C, FXI:C, FXII:C) were measured in one stage PTT assays ( Triplett and Harms 1981 ) configured with a series of human substrate deficient plasmas ( George King Biomedical, Overland Park, KS ) and commercial PT ( Thromboplastin LI, Helen a Diagnostics, Beaumont, TX ) or PTT ( Fibrinogen, Diagnostica Stago, Parsippany, NJ ) reagents. Statistical Analysis Statistical analysis was performed using the software program R (http://www.R project.org). Histograms were mapped to identify any outliers in data. The extreme studentized deviate method confirmed the outliers statistically at p < 0.05. A Kolmogorov Smirnov test was used to check for normality. As not all the data were normally distributed, a Wilcoxon signed rank test was used to compare the data between locations and between sex and size. Statistical differences were confirmed when p < 0.05. Results Twenty wild manatees were included in this study. There were 11 males and 9 females with body length ranging from 219cm to 324cm and weight rang ing from 220kg
42 727kg. Current length guidelines to indicate age categories are: >260cm adult, 236 260cm juvenile, and <236cm calves. All three age categories were represented in this study with 9 a dults, 6 juveniles and 5 calves. The results of factor analyses were reported as the percentage activity of the manatee standard which had an assigned value of 100%. The antithrombin activities of the test samples were reported as a percentage of the human calibrated standard. Mean values Â± standard deviation (SD), were calculated for each parameter. Reference ranges were established within 95% confidence intervals. Our results were generally very consistent , with low levels of variation between the 20 wild manatees. Out of 220 results only 2 were deemed as ou tliers and removed from the formation of reference intervals. The mean, SD, and 95% confidence interval for the coagulation panel are listed in Table 2 1 and coagulation factors in Table 2 2. These values provide the standard reference ranges for wild mana tee coagulation factor activity. Wilcoxon rank sum tests were performed to compare location, sex and size. There were some statistically significant differences between the two locations. Factors VIII, IX and XII were significantly higher in Brevard Count y compared to Crystal River with p values of 0.007, 0.005 and 0.007 respectively. PT and antithrombin were significantly higher in Crystal River than in Brevard County with p values 0.005 and 0.01 respectively. There were no statistically significant diffe rences between sex and size irrespective of location. Figure 2 1 represents the differences between location and the similarities between sex es graphically, clearly demonstrating the differences present in locations.
43 Discussion By d etermining the quantita tive measures of manatee coagulation factors against a pooled plasma sample from normal manatees , we have established the relative normal values for this species. These results are incomparable to other species, but will facilitate the identification of un derlying abnormalities during pathological processes such as cold stress syndrome or brevetoxicosis in manatees. The reference intervals reported showed no variation according to sex or age class. This suggests that these values will be clinically relevan t for any manatee >219cm. Very small calves were not included in this study ; therefore interpretation of results in an individual less than 6 months old or < 219cm should be taken with caution. The significance of the variation in results according to loca tion is not fully understood. A larger sample size may negate the degree of variation observed between locations. Further, the human literature suggests that coagulation factor activity may be influenced by both genetics and diet ( Mennen et al. 1998 ) . We hypothesize that there could be variation as a result of being exposed to different diets on the opposite coasts and being linked to different gene pools as these are recognized subpopulations (Laist and Reynolds 2005a) . These results support the work of Medway et al. in the first coagulation study performed in captive manatees ( Med way et al. 1982 ) . The levels of activity reported here confirm that manatees have a very active intrinsic clotting system; however the significance of this is not currently understood. Our results are similar to the coagulation profile of the elephant a lthough not directly comparable in the relatively high levels of plasma FVIII:C activity. Interestingly the elephant was shown to have a relatively long PTT with a mean of 65.6 seconds ( Gentry et al. 1996 ) whereas the manatee was 9.2
44 seconds. This is unusual in the elephant in that PTT is thought to be primarily influenced by FVIII:C activity, with high levels of FVIII:C activity resulting in a shorter PTT as supported by our results. Other species with the same combination of increased FVIII:C activity and shortened PTT include the camel and llama ( Morin et al . 1995 , Hussein et al. 2010 ) . A possible reason for the variation between species could be due to potential presence of an inhibitor of activated FXII that prevents the contact pathway in the elephant plasma ( Gentry et al. 1996 ) . It is still unclear whether high FVIII:C levels in some species actually represent a greater amount of circulating FVIII or indeed a greater sensitivity to activation by thromb in. This study also confirmed the high levels of FXII:C activity in manatees consistent with elephants and unlike several other aquatic mammals such as whales, dolphins and porpoises ( Robinson et al. 1969 , Lewis 1996 ) . The shortened PT and PTT found in manatees indicates rapid contact activation, enabling fibrin formation via the intrinsic cascade ( Maas and Renn Ã© 2012 ) . The contact activation pathway involves high molecular weight kininogen (HMWK) forming a primary complex on collagen along with three serine protein zymogens: prekallikrein (PK), FXI and FXII. The evolution of the contact system is particularly r elevant to manatees. HMWK is present in all vertebrates however PK, FXI and FXII only evolved later in tetrapods. Birds lost the gene for FXII over 300 million years ago as did most marine mammals. FXI and PK however have continued to be present in all mam mals withstanding evolution ( Ponczek et al. 2008 ) . The high level of FXI (124%) in manatees, and the fast rate of generation of FXII accounts for the fast PTT. The similarities in the coagulation factors of manatees to desert animals such as the camel and sand gazelle presents a curious finding. As manatees are one of the few
45 marine mammals that can survive in both salt and fresh water, the blood similarities to desert animals who can survive long periods of dehydration may suggest an evolutionary adaptation to environmental stressors. The results all demonstrate increased factor VIII and fibrinogen and shortened PTT. The exact significance for thi s is unknown but could be linked to the fact that manatees also undergo periods of dehydration when in salt water (Ortiz 2001) . Potentially, the location and habitat of the manatee will therefore influence these concentrations. The normal circulating levels of D Dimer s and fibrinogen established here in wild dimer levels should be <250ng/ mL which is indicative of no prior thromboembolic disturbances (Stokol et al. 2000b) . The 95% confidence interval in the 20 wild manatees sampled ranged from 51 113ng/ mL with a mean of 82 and a standard deviation of 6 5. These results are consistent with other species and confirm that normal manatees should have low circulating levels of D dimers. This provides a solid reference range to establish a prognostic indicator for elevated results in debilitated manatees. Simi larly all fibrinogen results were within an expected low range with a mean of 367 and standard deviation of 52, this provided 95% confidence intervals of 343 392mg/dl. In other species fibrinogen levels <400mg/dl are classed as normal. This is indicative of no active inflammation or infection in addition to a normal coagulation status (Stokol et al. 2000b) . The present findings are likely to improve our understanding of the pathophysiology of two important causes of mortality in manat ees cold stress syndrome and brevetoxicosis. By providing a baseline of coagulation factors in wild manatees that appear to be in good health, comparison can be made to apply the
46 information to the clinical management of debilitated animals, some of which appear to have abnormal coagulation profiles. Animals in a rehabilit ative setting with symptoms of either CS S or brevetoxicosis have been observed to have evidence of coagulation abnormalities. These abnormalities have been observed pre mortem in the clinical setting as well as on post mortem examination. Brevetoxicosis often results in catarrhal rhinitis clinic ally with hemorrhage and haemosiderosis at necropsy ( Bossart et al. 1998 ) . CSS has demonstrated the opposite with thromboembolic disease apparent in epidermal lesions clinically and at necropsy ( Bossart et al. 2002 ) . Having identified the normal coagulation activity levels , this will allow identification of any abnormalities that could explain the pathophysiology of these conditions. Further studies will help us to expand our knowledge on the relevance of these results and their implication in determining disease pathophysiology, and as prognostic indicators in rehabilitation. Products and References: a. STA Compact, Diagnostica Stago, Parsippany, NJ b. Dade Actin FS, Dade Behring, Newark, DE c. Throm boplastin LI, Helena Diagnostics, Beaumont, TX d. Fibrinogen, Diagnostica Stago, Parsippany, NJ e. Stachrom ATIII, Diagnostica Stago, Parsippany, NJ f. STA Unicalibrator, Diagnostica Stago, Parsippany, NJ g. HemosIL, D dimer Calibrator, Instrumentation Laboratory, Bed ford, MA h. George King Biomedical, Overland Park, KS i. Actin, Dade Behring, Newark, DE).
47 Table 2 1. Complete data set for coagulation panel results in wild healthy manatees. ID PTT PT Fibrinogen Anti thrombin D dimer (sec) (sec) (mg/dL) (%) (ng/mL) Me an 9.2 10.8 367.1 131.5 82 SD 1.6 0.5 52.3 11 65 95% CI 8.4 9.9 10.5 11.0 343 392 126 137 51 113 Table 2 2. Reference values for coagulation factors in wild Florida manatees ID FVII FVIII FIX FX FXI FXII (%) (%) (%) (%) (%) (%) Mean 103. 5 134 126 114 124 113 SD 12 35 32 19 23 25.6 95% CI 98 109 118 151 111 141 105 123 114 135 101 125
48 Figure 2 1. Graphical representation to indicate the increased variability of factor activity levels between locations and the lack of variation between locations when comparing sex . 0 20 40 60 80 100 120 140 160 180 FVII:C FVIII:C FIX:C FX:C FXI:C FXII:C Percentage of Activity % Coagulant Factor Location Comparison Brevard County Crystal River 0 20 40 60 80 100 120 140 160 FVII FVIII FIX FX FXI FXII Percentage of Activity % Coagulant Factor Sex Comparison Female Male
49 CHAPTER 3 ESTABLISHING NORMAL REFERENCES FOR THROMBOELASTOGRAPHY IN WILD FLORIDA MANATEES ( Trichechus manatus latirostris ) Thromboelastography (TEG) provides a comprehensive evaluation of blood clot forma tion. This test can be used to identify many abnormalities in coagulation by assessing multiple aspects of the clotting cascade from speed of clot initiation and formation, clot strength, and ultimately fibrinolysis. Thromboembolic disease has been hypothe sized to play a role in the pathophysiology of cold stress syndrome (CSS), an important cause of natural mortality in the Florida manatee ( Trichechus manatus latirostris ). The objective of this study was to establish thromboelastography reference ranges us ing the TEG 5000 with citrated whole blood samples and kaolin activation in manatees. In December 2014 and January 2015 twenty nine wild manatees were sampled as part of the annual wild manatee health assessments organized by U .S. Geological Survey. The sa mples were obtained from Crystal River, Citrus county and used in this study to identify the following mean (SD) normal TEG parameters: R = reaction time 2.1 min (0.8), K = MA = maximum amplitude 75mm (7. 6) and LY30 = clot lysis 0.41% (0.68). No significant differences were found between manatee size, sex or time after sampling to running the test. Manatee TEG parameters demonstrate a relatively hypercoagulable condition when compared to other mammals. This information will facilitate detection of changes in hemostasis during injury and disease. Introduction Thromboelastography (TEG) is a novel diagnostic test, rarely applied to wild animals. However it has been used extensively in human medicine for several decades, particularly in Europe ( De Nicola and Mazzetti 1955 ) . In veterinary medicine, reference
50 intervals for TEG have been established in dogs, horses and cats ( Wiinberg et al. 2005 , Epstein et al. 2009 , Hall et al. 2012 ) . Recently, TEG has been used to diagnose hypercoagulability, hypocoagulability and disseminated intravascular coagulation during various disease processes including parvovirus, immune mediated hemolytic anemia, and sepsis ( Otto et al. 2000 , Wiinberg et al. 2008 , Sinnott and Otto 2009 , Bentley et al. 2013 ) . Conventional methods to assess coagulation in veterinary medicine have relied on PT (prothrombin time), PTT (partial thromboplastin time) and platelet count. Unfortunately these tes ts are insensitive when trying to identify a hypercoagulable state, a nd are nonspecific for determining risk of hemorrhage ( Fenty et al. 2011 ) . The TEG apparatus generates a thromboelastogram tracing which represents the mechanics of clot formation. The thromboelastogram consists of a precoagulation phase represented by a flat line, a coagulation phase where the line diverges into two, and a fibrinolysis phase where the lines converge. Four standard values are obtained from a TEG tracing; R = Reaction Time or precoagulation time. R is the time take n for clot formation to initiate ( Donahue and Otto 2005 ) . It is measured as the distance in mm from the start of the tracing t o the point where the lines diverge 1mm. The R value can be interpreted to evaluate the intrinsic pathway and can therefore be influenced by changes in factor VIII, IX, XI and XII. K = c lot formation time. K represents the time in minutes from clot initi ation (R) until the distance between the two diverging lines reaches 20mm. The rapidity of the clot formation is influenced by factor II, VIII, platelet count, platelet function, hematocrit , fibrinogen concentrations and fibrin precipitation. = the angle between the midline and the tangent to the curve at the point where the lines diverge. This gives an additional
51 indication of the rate of clot formation and is therefore affected by the same factors that affect K. can therefore provide an i ndication of whether the blood is hypercoagulable or hypocoagulable. MA = the maximal amplitude , represented by the greatest distance between the two diverging branches. This reflects the final clot strength because at the greatest distance between the two lines the clot will be completely formed ( Jackson et al. 2009 ) . Clot lysis is reported as the LY30, which is the percentage of clot retraction 30 minutes after MA is reached. Figure 3 1 illustrates a thromboelas togram from a healthy wild manatee during this study . The Florida manatee ( Trichechus manatus latirostris ) is a marine mammal distributed along the warm waters of the southeastern United States ( Garcia Rodriguez et al. 1998 ) . Thromboembolic disease is suspected of playing a role in cold stress syndrome (CSS), a poorly understood but extremely important cause of natural mortality in Florida manatees. CSS can account for as much as 18% of annual mortalit y and in the 2010 2011 UME more than 480 deaths were at tributed to the condition (Bossart et al. 2004, Barlas et al. 2011) . Prior to investigation of the potential role of thromboembolic di sease in the pathophysiology of CSS, it is essential to define normal reference ranges in clinically healthy wild animals. The primary objective of this study was to validate thromboelastography in manatees using a TEG 5000 and achieve a standardized set of reference intervals. An additional objective was to investigate the effect of time on sample stability allowing this test to be available in the field and in institutions without a TEG machine. Method s All wild manatees were selected via opportunistic random sampling as part of the ongoing United States Geological Survey (USGS) health assessments survey
52 ( Bonde et al. 2012 ) . In Dece mber 2014 and January 2015 USGS captured 29 wild manatees for routine health assessment in Crystal River, Florida . Each animal was dry docked to facilitate veterinary assessment and biological sampling. A normal healthy manatee at the time of sampling was determined by veterinary physical exam, weight, fat ultrasound measurements, morphometrics and level of activity. Follow ing bloodwork analysis , selection criteria included all manatees that had routine biochemistry and hematology performed as part of the health assessment. All individuals were classed as normal according to published biochemistry and hematology reference ra nges ( Harvey et al. 2007 ) . In addition to the 29 wild manatees described above, 7 captive manatees were tested. These were animals that were deemed health y and ready for rele ase following a period of rehabilitation from CSS . These animals were not included in the formation of the reference ranges. Samples from these animals allowed comparison between healthy captive individuals and their wild counterparts. Blood samples fro m all manatees were obtained via standard manatee phlebotomy techniques from the capillary bed between the radius and ulna from either pectoral ( Walsh and Bossart 1999 ) . Following sterile preparation using betadine and alcohol gauze , a 20 gauge 1.5 inch needle w as inserted and attached to a wing tipped blood collection set ( BD Vacutainer Safety Lok blood collection set (Ref 367281), Becton, Dickinson and Co, Franklin Lakes, NJ). An initial non additive tube (serum separator) ( BD Vacutainer 3.0 mL serum blood coll ection tube (Ref 366668), Becton, Dicki nson, and Co, Franklin Lakes, NJ) was obtained to prevent tissue contamination of blood, followed by two anti coagulated 0.32% sodium citrate tubes ( BD Vacutainer 1.8 -
53 mL buffered sodium citrate 3.2% (Ref 366392), Bect on, Dickin son, and Co, Franklin Lakes, NJ) to obtain plasma. The citrated whole blood was stored on ice immediately after collection . Samples were warmed and gently rocked for 30 minutes prior to using the citrated whole blood for thromboelastography. The TEG assays were performed using a commercially available TEGÂ® 5000 analyzer ( TEG 5000 thromboelastograph hemostasis analyzer, Hae moscope Corp, Niles, IL) at 37Â°C using a thromboelastography kit provided by the manufacturer ( Donahue and Otto 2005 ) . These kits were stored at 4Â°C in a refrigerator until 30 minutes prior to use. Machine balance and e test was confirmed prior to run ning each batch of samples. 1 mL of Citrated whole blood was added to a proprietary blood tube with kaolin present ( Kaolin, Haemoscope Corp, Niles, IL ) . A small volume, 340 ÂµL of the kaolin activated blood, was added to the cup along with 20 ÂµL of calcium ch loride ( Calcium chloride 0 .2M, Haemoscope Corp, Niles, IL) and the pin was then lowered into the blood ( Donahue and Otto 2005 ) . The apparatus oscillated the cup for ten seconds at an angle. Once a fibrin clot starts to form the pin started to move within the cup and the resistance against movement was transferred up the wire to an electrical transducer which quantifies the criti cal parameters of the clot ( Palmer and Martin 2014 ) . TEG output was recorded on a laptop computer using commercial software (TEG V4 4.2.97). The program displayed the thr omboelastogram which was visually inspected for asymmetry or lack of activation prior to accepting the results into analysis. Most samples were processed within 3 hours of collection; however, duplicate samples from 10 individuals were tested beyond the t hree hour sampling time. As a result we ran 12 additional samples on those individuals that had been processed in
54 <3hrs at 4, 6, 9, 12 and 16 hours post sampling to address whether time was influential on the results. Four individuals had 3 additional samp les performed and the results compared below and above the 3hr threshold to assess the influence of time. Statistical Analysis TEG data is reported as mean and standard deviation. The Shapiro Wilk test was performed to establish whether data was normall y distributed. Wilcoxon rank sum tests were performed to analyze the effect of time, sex and age class upon the results. 95% reference intervals were calculated as mean +/ two standard deviations. Histograms were performed on all parameters to highlight potential outliers prior to determining reference intervals ( Friedri chs et al. 2012 ) interquartile fences were used to establish whether these outliers should be included. John Tukey defined data points as outliers if they are 1.5*IQR above the third quartile or below the first quartile. Results Twe nty nine wild manatees (17 male and 12 female) were sampled to establish normal TEG reference values. All individuals were classed as healthy having normal biochemistry and hematology according to published reference ranges ( Harvey et al. 2007 , Harvey et al. 2009 ) data points from the TEG analysis 6 were classified as outliers and removed from further analysis. The following mean (SD) normal TEG paramete rs were calculated: R = reaction time 2. 1 min (0.77) , K= 83.1Â° (2), maximum amplitude 75.1mm (7.6) and LY30 0.4% (0.68 ).
55 Effects of time, sex and size were assessed via a Wilcoxon rank sum test. This determined statistically significant differences if p <0.05. The test could not be performed on the K parameter as the standard deviation was zero as all results were identical. Sex and size class did not influence TEG values (Table 3 1 ). In comparing size classes 136cm was the t hreshold cut off to compare adults with calves. Time from blood sample collection t o TEG processing (compared as < 3 hr and > 3 hr) did not affect TEG parameters (Table 3 1 ) . Table 3 2 provides the mean, (SD), 95% CI, for the Florida manatee as establishe d in this study and horse, cow, dog and human references for comparison (no other marine mammal data is available). There was also no difference in TEG parameters between t he wild Crystal River manatees and the 7 captive individuals being prepared for rel ease (Table 3 3 ). Mann Whitney U analysis between these two data sets showed no significant difference. Discussion Thromboelastography is a clinical tool that can be used to assess hypercoagulation, hypocoagulation, and to investigate thromboembolic diseas e. Various methodologies exist in both human and veterinary medicine for the preparation of a blood sample prior to thromboelastography ( Zambruni et al. 2004 ) . The major advantage of using thromboelastography as a diagnostic tool is that TEG allows a single assay to a ssess both the kinetic and mechanical clot properties in whole blood. Currently the most readily available coagulation tests are PT and PTT. These tests are insensitive in that they only analyze small components of the clotting cascade rather than apprecia ting the entire process. PT analyses the extrinsic and common pathways whereas PTT looks at the intrinsic and common pathways. They both provide
56 information about hypocoagulability as a result of a defect in secondary hemostasis but fail to provide reliabl e information on hypercoagulability ( Donahue and Otto 2005 ) They also do not account for the crucial role that platelets and other cells play in hemostasis in vivo . To detect abnormalities in these tests factor levels need to have decreased by over 30 40% of normal, deeming these tests as insensitive ( Marder et al. 2012 ) . Wh en sufficiently robust statistics are utilized , as few as 20 individuals can be used to establish reference intervals in endangered species ( Friedri chs et al. 2012 ) Therefore the twenty nine wild individuals presented here provide sufficient data to enable establishment of TEG reference values for this species. Several data sets (R, K, MA, LY30) were not normally distributed due to the extensive over lap of values. This is however beneficial in establishing normal reference intervals, particularly in a wild endangered species where sample sizes are limited ( Horn and Pesce 2003 ) . The high consistency in values was clearly represented in the small standard deviation across all variables (Table 3 2 ). Running a TEG analysis o n wildlife under field conditions does not permit the immediate processing time that is possible when a patient is hospitalized. In addition to the challenge of transport time from field sites to the laboratory where the machine is located, there are furth er delays caused by multiple samples. It can take up to two hours to process a single sample, therefore it was essential to develop a protocol that allowed for accurate results while permitting a delay between sample collection and TEG processing. In the c urrent study, there were samples for 10 individuals that could not be run within the desired 3 hour time limit. Time delays between blood collection and processing TEG sample have been shown to affect test results in dogs ( Wiinberg et al.
57 2005 ) however samples stored for up to 48hours in cattle were unaffected ( Sommerey et al. 2014 ) .Conse quently, it was essential to assess the effect of delayed processing time on TEG samples taken from wild manatees. To address this concern, w e performed replicate tests on blood samples after the 3 hour threshold and compared these results to those process ed within the desired time frame . In the 12 repeated samples performed, we found no statistically significant differences in TEG results run at >3hours compared to those run within the 3 hour time frame. TEG results for wild manatees demonstrate that cl inically normal manatees show a 3 2 ). The consistency in K with every manatee producing a result of 0.8 minutes provided a very solid reference value. Data presented in Table 3 3 sh ows minimal variation between healthy individuals. Th ese data suggest that in a clinical setting a very minor alteration may have large clinical implications. Minor variations that might be overlooked in other species may be clinically relevant in manatee s . This relative hypercoagulation, which appears to be normal for manatees, suggests that manatees may be predisposed to thromboembolic disease. Use of citrated kaolin samples may be particularly important because of their natural hypercoagulable state. I n a pilot study with hospitalized animals we found that when native samples were used the blood clotted before the TEG began recording data, Interpretation of the TEG usually starts with MA which i s the most frequently evaluated TEG parameter in veterinary medicine ( Hall et al. 2012 ) . MA is reflective of final clot strength. R is arguably the second most reliable parameter as it reflects time
58 taken for clo t formation to initiate demonstrating changes in hypercoagulation or hypocoagulation. R is influenced by changes in the coagulation factors, which are more important for initiation of clot formation but have little effect on the clot strength. The greatest variation in the manatee TEG when compared to other species was noted in R and MA, indicating that manatees form strong clots quickly (Table 3 2) . The possible adaptive advantage of this capacity is not fully understood. The repeatability of these resul ts in healthy captive manatees demonstrates the appropriate and practical methodology established for thromboelastography in manatees. The protocol we have established does differ in the time taken to sampling from what is typically used in other veterinar y species , but this was to account for the clinical scenario experienced in wild mammals. The similarity in TEG parameters between wild apparently healthy animals and captive animals being prepared for release further supports the clinical assessment of th ese rehabilitated manatees and provides additional criteria that they are appropriate candidates for release. Manatee age categories are defined according to their length. Current guidelines to indicate age categories >260cm adult, 236 260cm juvenile, and <236cm calves ( Bonde et al. 2012 ) . All three categories were represented in this study. We did not include any animals less than 210cm (neonatal) in length in this study so all animals in our study are presumed to be over 6 months ( Bonde et al. 2012 ) . Lack of data on neonatal manatees, or young calves, makes it impossible to assess whether similar variation occurs in this species. However we can confirm that very little variation was observed in manatees older than 6 months of age. The effects of age on TEG have been reported in neonates less than 6 months old ( Edwards et al. 2 008 ) . TEG studies in
59 humans have reported some variation in results based on age and sex. Donahue and Otto (2005) reported slight variation in TEG parameters in cattle between bulls and steers . Concerns with this study were that they did not disting uish variation related to age from that related to sex . Additional investigation would be needed to clarify these observations. TEG facilitates in depth analysis of the interplay of protein and cellular elements of coagulation in addition to fibrinolysis, ther eby focusing on the bigger picture of hemostasis. This study enabled us to establish a TEG protocol for manatees and determine normal reference intervals in wild healthy individuals of this endangered species. TEG analysis has also allowed us to better und erstand the manatee coagulation process, and results support previous work that suggests that manatees are relatively hypercoagulable compared to other species.
60 Table 3 1. Statistical comparison between sex , time (comparing samples at 3hrs and post 3hr s) and size (comparing adult with calf) with the p value for each parameter with <0.05 classed as significant. This table demonstrates that there were no significant effects of time, sex or size on the results. R MA LY30 Sex p= 0.65 p=0.15 p=0.7 p=0.14 Time p=0.15 p=0.2 p=0.77 p=0.098 Size p=0.32 p=0.76 p=0.12 p=0.45
61 Table 3 2 . Established t hromboelastography reference intervals for wild Florida manatees showing mean and (standard deviation) results. Addition al data of horse, cow, dog and human values are presented for relative comparison. Sample Set R (min) K (min) AngleÂ° MA (mm) LY30(%) Wild Manatee 2.0 (0.5 ) 0.8 (0.0) 83.1 (2.0) 77 (3.4 ) 0.3 (0.4 ) 95% CI 1.8 2. 2 0.8 82.3 83 . 9 75.7 78.3 0.14 0.45 Hor se comparison ( Eps tein et al. 2009 ) 17.0 (3.0) 5.8 (2.3) 42(14) 60.3 (5.7) 0.6 (0.6) Cow comparison ( Sommerey et al. 2014 ) 2.2 6.2 0.8 2.0 58.2 to 81.8 64.3 89.2 Dog com parison ( Don ahue and Otto 2005 ) 7.8 11.1 4.4 5.3 57.8 62.8 57.8 62.8 Human intervals ( Antony et al. 2015 ) 5.3 9.3 7.7 (1.7) 1.4 3.5 48.8 72.2 61.9 (6.4) 55.3 69.3 61.9 (4.2) 0.9 (0.8) Table 3 3 . Comparison of the wild manatee thromboelastography established reference intervals with captive healthy rehabilitated individu als to facilitate confirmation of their normal health status . Sample Set R (min) K (min) AngleÂ° MA (mm) LY30(%) Wild (29 ) 2.0 (0.5 ) 0.8 (0.0) 83.1 (2.0) 77 (3.4) 0.3 (0.4) Captive (7) 1.9 (0.5) 0.8 (0.8) 83.4 (1.4) 79.8 (2.3) 0.4 (0.55) p value 0.5 1.0 0.9 0.1 1.0
62 Figure 3 1. Thromboelastogram of a wild healthy manatee demonstrating the shortened
63 CHAPTER 4 THE ROLE OF THROMBOEMBOLIC DISEASE IN THE PATHOPHYSIOLOGY OF COLD STRES S SYNDROME During the winter of 2014 2015, ten Florida manatees with clinical signs of cold stress syndrome were presented to Lowry Park Zoo. Thromboelastography and coagulation panels were performed at the time of entry. In addition , coagulation panel d ata from 23 retrospective cases were added to the clinical data base to identify average CSS results. Having established normal manatee reference ranges in coagulation parameters and thromboelastography , comparison could be pe rformed with coagulation resul ts in CSS cases. The results demonstrated that severe coagulation abnormalities are associated with cold stress syndrome in the manatee . The p values MA (0.5) and LY30 (0. 048). With p=<0.05 classed as statistically significant, R, K and LY30 showed significant differences in the CSS cases. These results indicate that manatees are relatively hypercoagulable and ultimately at increased risk of t hromboembolic disease Introduc tion The Florida manatee ( Trichechus manatus latirostris) is currently classed as endangered with the most recent aerial survey of the manatee population estimated at 6350 individuals ( Martin et al. 2015 ) . There are several important causes of mortality, the majority of which are anthropogenic , with the highest number being attributed to watercraft collisions as demonstrated in Table 1 1 (Barlas et al. 2011) . Environmental factors such as exposure to red tide dinoflagellate blooms, habitat defragmentation, and
64 loss of artificial warm water refuges also contribut e to manatee morbidity and mortality ( Ackerman et al. 1995 ) . Cold stress syndrome is recognized as one of the leading natural causes of mortality in the Florida manatee ( Laist et al. 2013 ) , accounting for approximately 18% of annual manatee deaths ( Bossart et al. 2004 ) . The pathophysiology of this condition is not well understood . It has been described as a series of ph ysical and behavioral changes that can progress to multisystemic organ failure and death in severe cases ( Bos sart 1999 ) . The condition has been defined as mortality caused by prolonged exposure to water temperatures <20Â°C ( Dierauf and Gulland 2001 ) . Improved u nderstanding of the pathophysiology of cold stress syndrome is needed. The risk of succumbing to cold stress syndrome is linked directly to the duration an d intensity of the cold weather and the quality of forage within the warm water site (Campbell 1981) . Manatee exposure to cold water is expected to increase as older power plants are closed in coming years. These have provided refuge from dropping water temperatures in winter months for manatees for the past 35 years , and the animals do not disting uish between these temporary , man made refuges and natural refuges ( i.e. artisanal springs and warm water basins) that are available to them ( Bossart 1999 ) . As some animals will seek refuge from cold water in areas that will no longer be set up to provide that warm water effluent the incidence of cold stress related disease will likely increase in future winters . The geographical range of cold stress mortality has been widespread throughout Florida. Natural artesian springs provide optimal warm water refuge security for the manatees. There is minimal fluctuation in the water temperature near the spring head,
65 eve n when the temperature of surrounding water falls shar p ly. In contrast, artificial refuge provided by power plants in recent decades is neither permanent, nor stable. Warm w ater effluent from power plants may not be sufficient to maintain stable water temp erature s in t he event of extreme cold (Laist and Reynolds 2005a) . As a result , mortality can vary significantly in populations that seek refuge from natural or artificial warm water locations ( Langtimm and Beck 2003 ) . T he largest mortalit y event attributed to cold temperature was in the UME in 2010 2011 as described in Chapter 1. The most frequently observed clinical sign in CSS are epidermal lesions as described in Chapter 1 . These can vary from mild epidermal bleaching to abscessation as shown in Figure 1 2 . The pattern and distribution of epider mal lesions associated with cold stress syndrome gave rise to the hypothesis that thromboembolic disease . Interestingly , on personal experience some of the more severely affected individuals actually showed evidence of bleeding post blood sampling, rather than clotting. We hypothesized this was likely due to the presence of disseminated intravascular coagulation resulting in a consumptive coagulopathy in the more severely affected individuals as demonstrated in humans ( Bick 2003 ) . From personal experience there were occasional incidences of prolonged bleeding at the site of blood collection in cold stress syndrome cases . To increase our understanding of this phenomenon , an d further our understanding of the pathophysiology of cold stress syndrome , the coagulation cascade was examined (Chapter 2). Initially the normal coagulation parameters needed to be identified. This was performed on 40 wild manatees to establish reference ranges for PT, PTT, D dimer concentrations and fibrinogen levels. To understand the clot
66 formation physiology in the manatee thromboelastography was performed as pre viously described in Chapter 3 . Anatomically , manatees are likely to be predisp osed to a higher incidence of thromboembolic disease due to their complex subdivision of large arteries into vast numbers of parallel small arteries ( Rommel and Caplan 2003 ) . In most mammals this occurs in a patter n similar to a tree with each subsequent vessel branching off becoming smaller. In manatees the reduction from a larger vessel to multiple small vessels occurs abruptly rather like the head of a broom as illustrated in Figure 4 1 . In theory this could be an adaptation to facilitate counter current heat exchange and reduce the risk of a hypothermic insult in the extremities (Rommel and Cap lan 2003) . Ironically this anatomy could predispose the manatee to thromboembolic disease and explain why many of the lesions see n in cold stress syndrome are confined to the extremities. The narrower lumen and thinner walls of the vessels in juveniles co mpared to adults could also impact the prevalence of cold stress lesions in the extremities of the younger demographic ( Rommel and Caplan 2003 ) . The normal reference ranges for coagulation factor activity; coagulation panel and TEG were previously established in wild healthy manatees also sampled in December 2014 . By assessing all these coagulation parameters in cold stress syndrome cases and comparing these results to previo usly established normal reference ranges in wild manatees, improved understanding of the pathophysiology of cold stres s syndrome will be obtained. The normal thromboelastography results for the manatee (Chapter 3) reveal that compared to TEGs established in other species, clinically normal manatees show relative hypercoagulation with a short R, K and
67 (Table 3 2) . There was a high consistency in results of wild individuals indicating minimal variation in healthy individuals (Table 3 3) . As a result, what appears as a very minor alteration could actually have a large clinical implication. Minor variations would often be overlooked in other species however the lack of variation in manatees and the high level of consistency usually demonstra ted, indicates that even minor variation might be interpreted with significance. Method s From December 5 th 2014 to March 3 rd 2015, 10 opportunistic cold stress Z oo for rehabilitation. A diagnosis of cold stress syndrome was made using previously described clinical criteria (Bossart et al. 2002) . Blood was collected on admission and obtained via standard manatee phlebotomy techniques ( Walsh and Bossart 1999 ) . For TEG and coagulation factor analysis, a n initial non additive tube ( BD Vacutainer 3.0 mL serum blood collection tube (Ref 366668), Becton, Dickins on, and Co, Franklin Lakes, NJ ) was obtain ed to prevent tissue contamination of blood, followed by two anti coagulated 0.32% sodium citrate tubes ( BD Vacutainer 1.8 mL buffered sodium citrate 3.2% (Ref 366392), Becton, Dickin son, and Co, Franklin Lakes, NJ ) , to obtain plasma. The citrated whole bl ood was stored on ice immediately after collection . Samples were warmed and gently rocked for 30 minutes prior to using the citrated whole blood for thromboelastography (TEG). Immediately following TEG , the samples were centrifuged at 1000xg for 15 minutes to separate the plasma. This remaining plasma was harvested and maintained at 80Â°C for batch analysis of PT, PTT, fibrinogen and D Dimers and coagulation factor activity levels. These samples were sent to The Comparative Coagulation Laboratory at Cornell
68 University within 6 weeks of sampling for analysis to ensure comparability. Assays were performed for coagulation factors VII through XII as well as activated partial thromboplastin time (PTT), prothrombin time (PT), and concentrations of fibrinogen, anti thrombin and D dimers. A coagulation panel consisting of PTT, PT, and fibrinogen was performed using an auto mated clot detection instrument ( STA Compact, D iagnostica Stago, Parsippany, NJ) , commercial reagents ( Dade Actin FS, Dade Behring, Newark, DE , Thro mboplastin LI, Helena Diagnostics, Beaumont, TX , Fibrinogen, Diagnostica Stago, Parsippany, NJ ) , and reaction conditions as previously described ( Stokol et al. 2000a ) . Antithrombin activity (AT) was measured in a functional assay configured to measure thrombin inhibition (anti IIa assay) using a commercial chromogenic kit ( Stachrom A TIII, Diagnostica Stago, Parsippany, NJ ) and the ( STA Compact, Diagnostica Stago, Parsippany, NJ ) . The standard curves for determination of clottable fibrinogen ( Clauss 1957 ) and antithrombin activities in the test plasmas we re derived from a calibrated human plasma standard ( STA Unicalibrator, D iagnostica Stago, Parsippany, NJ). D dimer concentration in ng/mL was measured using a quantitative, immunoturbidometric method as previously described, ( Delgado et al. 2009 ) using a commercial kit and the dimer standards (STA Unicalib rator, Diagnostica Stago, Parsippany, NJ ) . The coagulant activities of Factors VII and X (FVII:C and FX:C) were measured in one stage PT and activities of Factors VIII, IX, XI, and XII (FVIII:C, FIX:C, FXI:C, FXII:C) were measured in one stage aPTT assays ( Triplett and Harms 1981 ) configured with a series of human substrate deficient plasmas ( George King Biomedical, Overland Park, KS ) and commercial PT (Thromboplastin LI, Helena
69 Diagnostics, Beaumont, TX) or PTT (Fibrinogen, Diagnostica Stago, Parsippany, NJ reagents ) . The TEG assays were performed using a commercially available TEGÂ® 5000 analyzer (Hemonetics, Braintree, MA) Hemostasis Analyzer ( TEG 5000 thromboelastograp h hemostasis analyzer, Haemoscope, Niles, IL) at 37Â°C using a thromboelastography kit provided by the manufacturer ( Donahue an d Otto 2005 ) . These kits were stored at 4Â°C in a refrigerator until 30 minutes prior to use. Machine balance and e test was confirmed prior to running each batch of samples. 1 mL of c itrated whole blood was added to a proprietary blood tube with kaolin pre sent ( Kao lin, Haemoscope Corp, Niles, IL). A small volume, 340 ÂµL of the kaolin activated blood, was added to the cup along with 20 ÂµL of calcium chloride ( Calcium chloride 0 .2M, Haemoscope Corp, Niles, IL), and the pin was then lowered into the blood ( Donahue and Otto 2005 ) . The apparatus oscillated the cup for ten seconds at an angle. Once a fibrin clot starts to form the pin s tarted to move within the cup and the resistance against movement was transferred up the wire to an electrical transducer which quantifies the critical parameters of the clot ( Palmer and Martin 2014 ) . TEG output was recorded on a laptop computer using commercial software (TEG V4 4.2.97). The program displays the thromboelastogram which was visually inspected for asymmetry or lack of activation prior to accepting the result s into analysis. Retrospective analysis of an additional 23 cold stress syndrome cases admitted to Lowry Park Zoo between March 2009 and December 2014 were included to increase the sample population . These manatees had also been rescued and exhibited clin ical evidence of cold stress pathology . C i trated blood sample s had been drawn at the time
70 of admission and PT, PTT, D dimer concentration and fibrinogen levels were documented . No thromboelastography or factor activity results were available for this cohor t due to the retrospective analysis . Statistics Statistical analysis was performed to compare the results of cold stress syndrome cases to wild manatee reference ranges using the statistical software program R ( http://www.R project.org .) for data analysis . Data distribution was examined for normality by the Kolmogorov Smirnov test. A Mann Whitney U test was performed to compare the thromboelastography results from cold stress syndrome cases with the wild manatee reference ranges. Statistical significance w as defined as a p value of <0.05. TEG data is reported as mean and standard deviation. An unpaired T test was performed to compare coagulation panel results of PT, PTT, D dimer and fibrinogen concentrations to established normal reference ranges. Results Ten cold stress cases were included in the thromboelastography st udy. All ten individuals were rescued with signs of overt cold str ess syndrome . These consisted of f our male s and six females with sizes ranging from 59kg to 405kg in weight and 145cm to 275c m in length . Thromboelastography results from the cold stress syndrome cases revealed several statistically significant differences when compared to the wild manatee reference range . The results of thromboelast ography comparison are presented in Table 4 1 . The p values MA (0.5) and LY30 (0.048). With p=<0.05 classed as sta ti stically significant, R, K and LY30 showed significant differences in the CSS cases.
71 The following coagulation panel results , were obtai ned for the mean and standard deviation in the 33 cold stress cases: PT 12.6 ( 16.1) seconds, PTT 25.9, (38.6), D dimer 924 (814) and fibrinogen 424 (226). Statistical comparison of these results with the wild manatee reference ranges revealed statisticall y significant increases in both PTT ( p = 0.0156 ) and D Dimers (p=0.0001) for CSS manatees . These results are presented in Table 4 2. The results of the coagulation factor activity levels in 10 CSS manatees are displayed in Table 4 3 . Comparison can be made with Table 2 2 which contains the reference values for coagulation factors in wild manatees and is summarized in Table 4 4 . The obvious difference in factor level activity is present in Factor XI where there was a significant difference (p= 0.0012) in act ivity level in CSS cases with increased activity compared to wild manatees . Anti thrombin also showed a statistically lower level of activity in CSS compared to normal manatee activity levels. Discussion The objective of this study was to compare coagula tion profiles of manatees affected with cold stress syndrome to previously established reference values for wil d healthy animals (chapters 2 and 3) . The aim was to identify coagulopathies associated with CSS with a view to improving the understanding of th e pathophysiology of this complex syndrome. Thromboelastography is one of the best diagnostic test ing options for the assessment of hypercoagulation ( Fenty et al. 2011 ) . In comparing the TEG results from wild healthy manatees (Chapter 3 ) with cold stress cases we found subtle but statistically signi ficant differences . Low variation in the healthy population (Chapter 3) allowed detection of changes in the coagulation profile of CSS affected manatees
72 despite the relatively small sample size form clinical cases received during the winter of 2014 2015 . O ur findings support the hypothesis that cold stress manatees are hypercoagulable compared to normal wild manatees. The time taken for the clot to form (R) was the most statistically significantly different variable (p = 0.019) indicating that manatees with cold stress syndrome form blood clots at a faster rate. K was also statistically significant ly higher at p = 0.02 . This is particularly remarkable considering the standard deviation in the wild subset was zero i.e. no wild animals had a result different f rom 0.8. As K reflects the time taken to reach 20mm of divergence , it provides an indication of speed of clot formation. The refore the significance in the variation of K indicates CSS affected manatees have more rapid clot development confirming the hyperc oagulable state. K can be influenced by factors II and VIII as well as hematocrit , platelet count, thrombin formation and fibrin precipitation ( Donahue and Otto 2005 ) . In combination , the R and K values reflect the coagulation time to formation of predetermined clot strength. The changes seen in the K values in manatee were anticipated due to the hypercoagulable nature of their coagulation system (Chapter 2 and 3) . Similar findings of shortened R and K parameters were discovered in d ogs with septic peritonitis which were also fou nd to have hypercoagulable TEGS with shortened R and K parameters ( Bentley et al. 2013 ) . An example of a hypercoagulable TEG is illustrated in Figure 4 2 and is superimposed on a normal TEG for comparison. Neither MA or compared to wild healthy manatees . This is t o be expected as MA represents the final clot strength , which would be the same as in formation. LY30 was significantly higher in CSS cases than in the normal wild
73 individuals. As LY30 measures the rate of amplitude reduction thirty minutes after the maximum amplitude was reached . I t can be interpreted as an indication of the stability of the clot. This suggests that although the CSS manatees are reaching the same clot strength as normal m anatees there i s a small degree of fibrinolysis occurring in some cases. The largest value of LY30 in CSS case was 4.3%. This is still class ified as a normal percentage of clot breakdown in other species (<7.5% at 30 minutes ( Kelley et al. 2015 ) ) however the difference was statistically significant in CSS cases d ue to the minimal occurrence of fibrinolysis in normal wild manatees (95% CI 0.18 0.70 %). In other veterinary species TEG can be used as a prognostic survival indicator and has been found to be more accurate than routine coagulation tests such as PT and P TT ( Bentley et al. 2013 ) . Having established the normal manatee parameters and observed statistically significant differences in CSS cases , TEG may be valuable as a prognostic indicator in manatee cases. T hromboela stographs with the greatest variation from normal would have a poorer prognosis. In addition to TEG we compared the coagulation panel consisting of PT, PTT, and D dimer and fibrinogen concentrations in the 10 CSS cases acquired during this study period to normal references. In addition , data from 23 retrospective cases admitted for treatment of CSS , since March 2009 were added to the analysis to increase the sample population . Inclusion criteria involved a coagulation panel of PT, PTT, platelet count, D Di mers and fibrinogen , performed at the time of hospital admission . The following values were obtained for the mean and standard deviation in the 33 cold stress cases: PT 12.6 (16.1) seconds, PTT 25.9, (38.6). In wild manatees the mean and standard deviatio n were PT 9.2 (1.5) and PTT 10.8 (0.5). Statistical
74 comparison between CSS and wild reference range of PT revealed p = 0.1876 and PTT p = 0.0156. Although changes in PT w ere not statistically significant , normal PT is <10 seconds so there was a trend for C SS cases to have a prolonged PT . PTT was statistically significant but this is most likely due to the large standard deviation caused by v ery prolonged values such as 180 seconds. By increasing our sample size rather than just using the 10 cases obtained during this study period we can reduce the bias of extreme outliers and provide a more accurate representation of CSS . The extreme results of 180 seconds are however important to include as these represen t t he more severely affected individuals likely und ergoing disseminated intravascular coagulation. All of the 10 CSS manatees which had TEG performed survived. Therefore from our current analysis we cannot comment on the use of TEG as an accurate prognostic indicator. Further analysis would need to be perf ormed on more severe CSS cases to identify TEG changes in fatal cases. As 2014 2015 was a relatively mild winter it is likely more severe abnormalities would be detected via TEG during colder conditions. Normal D Dimer levels are classed as < 250ng/ mL ( Stokol et al. 2000b ) . Wild manatees (n=40) were all normal with mean (SD) of 132 (126)ng/ m L . Cold stress manatees showed a wide variation in D Dimer concentrations with a range of 250 3309ng/ mL . The 33 cases produced a mean and SD of 924 (814) ng/ mL . This is clearly very significant with a p value of 0.0001. This is the most clinically relev ant value produced in this report as it can be used as an accurate prognostic indicator. As D dimers only result from the degradation of a stabilized clot (cross linked fibrin) they provide a sensitive and specific quantification of active coagulation and fibrinolysis ( Nelson 2005 ) . The use of the sensitive D dimer assay enables informed clinical
75 assessment and prognostic indication as to the extent of thromboembolic disease. Continual monitor ing can also indicate whether the patient is stable, improving or deteriorating as to whether the D dimer levels are the same, decreasing or increasing ( Stokol et al. 2000b ) . From comparing D dimer levels in the 33 cases with survival outcome, any value > 500ng/ mL is classed as abnormal with values >1500ng/ mL carrying a grave prognosis. Incre asing values throughout rehabilitation are also indicative of a poor survival outcome. Fibrinogen is highly variable as a prognostic indicator due to the fact that it is also an a cute phase protein therefore levels can be increased due to inflammation and infectio n as well as active coagulation ( Sato et al. 1995 ) . Wild manatee s were found to have a mean ( SD ) of 379 (79) with CSS having a mean of 424 (226). There was no statistical ly significant difference in the CSS cases with p = 0.24. This is likely due to the very wide variation in values amongst the CSS cases from 1 5 860mg/dl resulting in a large standard deviation. Increasing the number of cold stress cases would likely improve the fibrinogen significance. The wide variation can be explained by fibrinogen being raised due to systemic inflammation as an acute phase protein in addition to changes in levels as a result of coagulation ( Wada et al. 2003 ) . Its role in the coagulation cascade can result in depleted levels due to excess clot formation. As a result low values of 15mg /dl can also indicate a consumptive coagulopathy and could actually be of greater clinical significance in extreme cold stress cases. Consequently those individuals undergoing more extreme CSS are actually more likely to have greatly reduced fibrinogen lev els. The wide range of values suggests that plasma fibrinogen is not a sensitive marker for CSS or DIC and results should be interpreted on a case by
76 case basis . Although higher levels are often associated with a poorer outcome as there is a correlation wi th elevated fibrinogen and reduced activation of secondary fibrinolysis which can be linked to organ failure and a poorer prognosis ( Wada et al. 2003 ) . Comparison of antithrombin levels between CSS and normal samp les revealed p =0.0014 demonstrating a statistically significant difference with CSS manatees having lower values. Antithrombin is a plasma proteinase inhibitor that functions to inactivate thrombin and the enzymes responsible for the generation of thrombin . As a result it is a very effective anticoagulant and can be enhanced by heparin which encourages further antithrombin binding to factors II and X ( Marder et al. 2012 ) . A deficiency in antithrombin resu lts in clinical thrombosis. Acquired deficiencies are associated with severe sepsis, trauma, neoplasia, burns and surgery ; all of which result in consumptive coagulopathy and acute inflammation. Previous studies have found that a reduction in plasma antith rombin below 70% of normal reflects the crucial threshold in the pathogene sis of normal venous thrombosis ( Marder et al. 2012 ) . As a result low levels of antithrombin are often found in DIC due to rapid consumption ( Fourrier et al. 1992 ) . Low levels in humans have been positively correlated with a poor prognosis, as a result treatment with antithrombin has been instrumental in preventing DIC and death ( Fourrier et al. 1992 ) . Three of the CSS cases (103223, 103224, 103230) showed antithrombin levels <70% indicating active DIC and a consumptive coagulopathy. This was supported by these individuals also s howing extremely prolonged PTT at 180 seconds and prolonged PT at 90 seconds. This is in stark contrast to the wild sample set where the lowest antithrombin value was 112% compared to the lowest CSS value of 41%.
77 With regard to the coagulation factors , th e only factor which showed statistical difference in activity level was factor XI (p=0.0012) in addition to antithrombin (p = 0.0014) . Prior work by Medway established that Factor XI in manatees has double the activity of Factor XI in normal dogs. Interest ingly these results showed further increased Factor XI levels in CSS cases. In humans increased Factor XI levels have been associated with increased risk for venous thrombosis ( Meijers et al. 2000 , Siegemund et al. 2004 ) . This supports the hypothesis that CSS individuals are likely to have a great occurrence of thromboembolisms . Although not significantly different , factor VIII did show reduced levels in CSS which could be explained by the thrombin induced generation of protein C resulting in proteolysis of the activated factor VIII ( Marder et al. 2012 ) . This is further support of a coagulopathy in asso ciation with CSS. Factors VII, IX and X showed little variation between the two groups, however the consistently high activity levels were comparable to the previous work by Medway ( Medway et al. 1982 ) . Finally factor XII showed increased activity in CSS although p =0.135 therefore was not deemed statistically significant. The rate of generation of FXII accounts of the increased variability in PTT seen in the CSS cases . The cold stress samples used in this study for TEG and factor level analysis were opportunistically obtained within the time frame of the project over winter 2014. This was a mild winter in Florida; however the coastal Gulf of Mexico temperatures did drop below 20Â°C ; therefore warm water migration was required for thermoregulation. Cold stress is defined as chronic exposure to low temperatures. Due to the short time frame of a few days of cold weather this winter , ess
78 with mild epidermal bleaching but normal core temperatures. Consequently the coagulopathies detected were likely less extreme than in severe cold stress cases seen historically. 100% of CSS in this sample set survived despite three individuals demonstr ating severe coagulation abnormalities. The fact that these CSS cases all presented mildly clinically yet demonstrated dramatic coagulopathies internally indicates the potential extent of subclinical abnormalities. In a more severe winter it is likely t hat the survival outcome would be less favorable as demonstrated in the archived 23 cases where there was a 30% mortality rate. Although manatees can give birth year round they have been described to engage in a diffusely seasonal reproductive pattern with s uppression of activity occurring in the winter months ( Reep and Bonde 2006 ) . Due to the 12 14 month gestation period and mating frequ ently occurring post winter, most calves are several months old before enduring their first winter ( Koelsch 2001 ) . The majority of cold stress cases are usually 2 nd year calves having to survive their first winter alone ( Bossart et al. 2002 ) .If they are estimated to be over 6 months the standard reference intervals established here could feasibly be used in these cases to try to establish hyp ercoagulation alterations in cold stress syndrome. This study successfully measured changes in coagulation in Florida manatees undergoing CSS. By comparing these results to the previously established normal coagulation parameters (Chapters 2 and 3), coagul opathy occurring in manatees with CSS could be identified and clinical tools were defined which may have value as prognostic indicator s . Cold stress syndrome has had a profound effect on this species,
79 therefore a thorough understanding of the predisposing factors to this multifactorial condition is important for future management and conservation .
80 Products and References: a. BD Vacutainer Safety Lok blood collection set (Ref 367281), Becton, Dickinson and Co, Franklin Lakes, NJ. b. BD Vacutainer 3.0 mL serum b lood collection tube (Ref 366668), Becton, Dickinson, and Co, Franklin Lakes, NJ. c. BD Vacutainer 1.8 mL buffered sodium citrate 3.2% (Ref 366392), Becton, Dickinson, and Co, Franklin Lakes, NJ. d. STA Compact, Diagnostica Stago, Parsippany, NJ e. Dade Actin FS, Dade Behring, Newark, DE f. Thromboplastin LI, Helena Diagnostics, Beaumont, TX g. Fibrinogen, Diagnostica Stago, Parsippany, NJ h. Stachrom ATIII, Diagnostica Stago, Parsippany, NJ i. STA Unicalibrator, Diagnostica Stago, Parsippany, NJ j. HemosIL, D dimer Calibrator, I nstrumentation Laboratory, Bedford, MA k. George King Biomedical, Overland Park, KS l. TEG 5000 thromboelastograph hemo stasis analyzer, Haemoscope , Niles, IL. m. Kaolin, Haemoscope Corp, Niles, IL. n. Calcium chloride 0.2M, Haemoscope Corp, Niles, IL.
81 Table 4 1. Th romboelastography results from 10 cold stress syndrome cases compared with the wild manatee reference ranges. P value indicates statistical comparison , <0.05 is classe d as being statistically significantly different. Table 4 2. Coagulation panel results comparing cold stress syndro me cases with wild manatee reference ranges. P value indicates statistical comparison , <0.05 is classe d as being statistically significantly different. Sample Set PT (sec) PTT (sec) D Dimer (ng/ mL ) Fibrinogen (mg/dl) Wild (n = 40) 9.2 (1.5) 10.8 (0.5) 8 2 (65 ) 3 67 (52 ) CSS (n=33) 12.6 (16.1) 25.9 (38.6) 924 (814) 424 (226) P value 0.1876 0.0156 0.0001 0.126 Sample Set R (min) K (min) AngleÂ° MA ( mm) LY30(%) Wild (n = 29) 2.1 (0.75) 0.8 (0.0) 83.3 (2.0) 75.1 (7.8) 0.4 (0.69) CSS(n=10) 1.48 (0.34) 0.90 (0.23) 82.6 (3.2) 76.6 (12.7) 1.2 (1.62) P value 0.019 0.02 0.60 0.50 0.048
82 Table 4 3. Cold stress syndrome case results showing the coagulation panel consisting of PTT, PT, fibrinogen, antithrombin, D dimer and factors VII through XII. The mean and standard deviation are provided for each parameter for these 10 individuals. Animal ID PTT PT Fibrinogen AT D dimer FVII:C FVIII:C FIX:C FX:C FXI:C FXII:C (sec) (sec) (mg/dL) (%) (ng/mL) (%) (%) (%) (%) (%) (%) 103215 12.1 12.3 398 125 79 105 135 98 117 125 112 103223 180 90 15 68 134 107 19 96 < 10 338 64 103224 180 90 15 41 544 96 11 112 < 10 163 73 103226 19.8 27.6 436 116 647 104 121 108 92 132 103 103229 11.3 11.2 344 123 221 113 135 97 137 168 120 103231 6.8 11.7 429 123 174 128 172 168 102 187 148 103230 180 90 15 60 470 64 14 33 44 158 195 103232 10.8 12.3 625 130 1743 143 189 155 151 265 181 103234 10 11.8 490 124 253 135 169 186 136 150 147 103235 6 11.4 413 129 216 143 182 130 149 132 180 Mean 61.68 36.8 3 318 103.9 448.1 113.8 114.7 118.3 116 181.8 132.3 SD 77.54 35.12 210.14 31.95 466.86 23.18 68.70 41.62 33.75 64.54 43.18
83 Table 4 4 . Cold stress syndrome coagulation facto r results compared with wild normal values. P value indicates statistical comp arison , <0.05 is classe d as being statistically significantly different. Antithrombin (%) FVII:C (%) FVIII:C (%) FIX:C (%) FX:C (%) FXI:C (%) FXII:C (%) Mean CSS (10) 103.9 113.8 114.7 118.3 116.0 181.8 132.3 SD CSS 32.0 23.2 68.7 41.6 33.8 64.5 43.2 Mean wild (20) 131. 5 103.5 134.3 126.4 113.9 124.5 113.4 SD wild 10.7 1 2.0 35 .0 31. 9 19.0 22.6 25. 6 P Value 0.0014 0.2840 0.3000 0.5500 0.7500 0.0012 0.1350
84 Figure 4 1. The right intercostal artery; demonstrating the dramatic reduction in diameter, between the large artery and the subsequent blood vessels showing similar branching structure as that of a broom. Anatomical photo credit Rommel et al. (Rommel and Caplan 2003) Figure 4 2 . A thromboelastograph of a cold stress syndrome calf in white superimposed over a normal wild manatee TEG. Note the reduction in R with clot initiation commencing after just 1.1minutes. Also the lines begin to diverge demonstrating a degree of fibrinolysis occurring and 30minutes and increasing to 8.7% at 60 minutes.
85 CHAPTER 5 CONCLUDING SUMMARY C old stress syndrome in manatees has been defined as morbidity and mortality that result from prolonged expos ure to water temperatures <20Â°C. The pathophysiology of this condition is not well understood and has previously been described as nutritional, immunological and metabolic disturbance s that contribute to multisystemic organ failure, and may result in life threatening opportunistic infectious disease ( Bossart et al. 2002 ) . Further investigation of the pathophysiology of CSS was warranted to improve understanding and enhance mitigation of one of the leading natural causes of mortality in this endangered species. Federal and State protection of the Florida manatee is provided by the E ndangered S pecies A ct of 1973 , the Marine Mammal Protection Act of 1972, and the M anatee S anctuary A ct of 1978 . State and Federal recovery effort success is reflected in an apparent increase in Florida manatee populations in recent years. Annual pop ulation counts were estimated by synoptic survey as 2639 in 1996 ( Bossart 1999 ) and estimated 6063 by abundance survey in 2011 2012 ( Martin et al. 2015 ) . Unfortunately despite th is success, stochastic events can result in mass manatee mortality ; events such as red tides ( brevetoxicosis ) and prolonged cold weather (resulting in CSS ) can result in hundreds of manatees d ying within a short time frame (Ackerman et al. 1995, Barlas et al. 2011) . The impact that these unpredictable but natural events can have on the manatee population in Florida play a role in its continued endangered stat us despite the apparent improvement in population size. The two greatest threats to the manatee are loss of warm water refuge and high mortality as a result of water craft injuries (Runge et al. 2007) . On an average year, cold stress
86 syndrome accounts for up to 18% of annual mortality , or approximately 124 manatees despite the morbidity being much higher (Bossart et al. 2002) . However , during particularly cold periods this figure can increase dramatically. For example during the winter of 2010 there were 480 manatee deaths attributed to CSS ; almost 4x higher than average . At the time thi s represented almost 10% of the documented population (Barlas et al. 2011) . Consequently, from the rehabilitation perspective there is the possibility of receiving a large number of manatees affected by cold stress syndrome in a very short period of time. The rehabilitation capacity of t he state can be rapidly overwhelmed when such events occur. This concern is exacerbated by the loss of artificial warm water refuges which is projected to occur in coming years as aging power plants are closed. The problem is further exacerbated by contin ued habitat defragmentation and lack of access to natural warm water springs ( Laist et al. 2013 ) . This study aimed to establish the normal reference ranges in wild manatees for thromboelastography, coagulation factor activity levels, PT, PTT, D dim er concentration and fibrinogen levels. Forty wild manatees were sampled for PT, PTT, D dimer concentration and fibrinogen levels also referred to as the coagulation panel. Twenty nine of these manatees had thromboelastography performed. TEG analysis was l imited by the requirement that samples be processed within 3 hours of blood collection . Further, financial constraints limited coagulation factor analysis to only twenty of the forty wild manatees that were sampled for coagulation factor activity levels. Nonetheless, sample size was sufficient to statistically establish the reference range in this endangered species ( Friedri chs et al. 2012 ) .
87 Once the normal reference levels were established for wild healthy Florida manatees, comparison with cold stress syndrome cases could be performed. Within this study period there were ten opportunistic cold stress syndrome cases admitted to y Park Zoo. In addition to this, 23 archived cases had limited coagulation factor analysis performed at hospital admission for CSS . Data f rom these cases was added to that available for the 10 active CSS cases seen during the study period . T hromboelastography and coagulation factor level analysis were not available for the archived CSS cases . The results of this study support our hypothesis that cold stress syndrome results in coagulopathy in the Florida manatee. This is indicated by the pr esence of elevated D dimer levels in CSS cases indicating prior thromboembolic disease. D dimers are degradation products of cross linked fibrin and can be used clinically to indicate prior venous thromboembolisms. Levels <250ng/ mL are classed as normal ( Eichinger et al. 2003 ) . This was reinforced with the mean (SD) result in the healthy wild population of 132 (126)ng/ mL compared to the mean CSS D dimer value of 924 (814)ng/ mL with the highest indiv idual at 3309 ng/ mL . High D dimer results are a sensitive and specific indicator of prior thromboembolism and can be used as a prognostic indicator in CSS ( Stokol et al. 2000b ) . Other coagulation panel abnormalities included elongated PTT and PT. It is likely these animals are undergoing disseminated intravascular coagulation with increased bl ood clotting occurring as a result of the hypothermia followed by prolonged bleeding due to a consumptive coagulopathy. By performing a complete coagulation panel and assessing PT, PTT, platelet count and fibrinogen levels in addition to D dimer
88 concentrat ion, more information can be gathered regarding the extent of prior thromboembolisms and also the presence of disseminated intravascular coagulation. By sampling at the time of admission and repeating throughout rehabilitation , this can give a prognostic i ndication as to whether the patient is improving internally. Thromboelastography is a novel test in veterinary medicine , particularly in marine mammals. We were able to demonstrate a protocol for successfully performing TEG in manatees that may be useful for assessing hemostatic abnormalities in debilitated manatees as well as aid in the evaluation of the pathophysiology of various disease processes. The high level of consistency in the results from wild manatees makes this test of particular interest for allowing assessment of abnormal findings. We demonstrated that in general manatees are relatively hypercoagulable and form clots faster than any of the domestic species previously analyzed. What was particularly significant in this case was that despite manatees being naturally relatively hypercoagulable , the cold stress syndrome cases demonstrated further increase in coagulability. This demonstrates that even a marginal change in TEG results in CSS cases can be significant. Due to the minimal variation seen in the wild manatees the slightest change was actually clinically relevant. Improv ed understanding of the pathophysiology of cold stress syndrome has enabled greater insight into the underlying cause s of cold stress syndrome and related multisystemic organ failure. A lthough it is still most likely that cold stress syndrome is a multifactorial condition, a component of the underlying pathophysiology is now understood. The clinical tests described can be used to effectively diagnose cold stress syndrome in addition to relying on the classical clinical presentation of advanced
89 emaciation, characteristic skin lesions, and a correlated prolonged cold weather period. Measurement of the coagulation panel can give a prognostic indication during rehabilitation which may influence treatment protocols and ultimately improve clinical management and enhance survival of this condition. Figure 5 1. Example of how diagnostic tests can be used to aid identification and diagnosis of CSS. Cold Weather Period Severe epidermal bleaching, abscessation and emaciation Mild epidermal bleaching, robust body condition Normal biochemistry, hematology, TEG and coagulation panel Diagnosis = mild exposure to hypothermia but not CSS Diagnosis of CSS Hypercoagulable TEG Increased D Dimers Increased WBC Thrombocytopenia
90 LIST OF REFERENCES Ackerman B, Wright S, Bonde R, Odell D, Banowetz D. 1995. Trends and patterns in mortality of manatees in Florida, 1974 1992. Population biology of the Florida manatee. National Biological Service Informatio n and Technology Report 1: 223 258. Ames AL, Van Vleet ES, Reynolds JE. 2002. Comparison of lipids in selected tissues of the Florida manatee (Order Sirenia) and bottlenose dolphin (Order Cetacea; Suborder Odontoceti). Comparative Biochemistry and Physiolo gy Part B: Biochemistry and Molecular Biology 132: 625 634. Anderson ET, Harms CA, Stringer EM, Cluse WM. 2011. Evaluation of hematology and serum biochemistry of cold stunned green sea turtles (Chelonia mydas) in North Carolina, USA. Journal of Zoo and Wi ldlife Medicine 42: 247 255. Antony KM, Mansouri R, Arndt M, Rocky Hui SK, Jariwala P, Mcmullen VM, Teruya J, Aagaard K. 2015. Establishing Thromboelastography with Platelet Function Analyzer Reference Ranges and Other Measures in Healthy Term Pregnant Wom en. American Journal Perinatology 60: 20 29. Barlas ME, C. J. Deutsch, M. De Wit A, Ward Geiger. LI. 2011. Florida manatee cold related unusual mortality event, January April 2010. Final report to USFWS. Florida Fish and Wildlife Conservation Commission, St. Petersburg, FL. 138pp. Bentley AM, Mayhew PD, Culp WT, Otto CM. 2013. Alterations in the hemostatic profiles of dogs with naturally occurring septic peritonitis. Journal of Veterinary Emergency and Critical Care 23: 14 22. Bertram G, Bertram CR. 1973. The modern Sirenia: their distribution and status. Biological Journal of the Linnean Society 5: 297 338. Bick RL. 2003. Disseminated intravascular coagulation current concepts of etiology, pathophysiology, diagnosis, and treatment. Hematology Oncol Clinic s of North America 17: 149 176. Bonde RK, Garrett A, Belanger M, Askin N, Tan L, Wittnich C. 2012. Biomedical health assessments of the Florida manatee in Crystal River providing opportunities for training during the capture, handling, and processing of th is endangered aquatic mammal. Journal of Marine Animal Ecology 5: 17 28.
91 Bossart GD. 1999. The Florida manatee: on the verge of extinction? Journal American Veterinary Medical Association 214: 1178 1183. Bossart GD, Baden DG, Ewing RY, Roberts B, Wright SD . 1998. Brevetoxicosis in manatees (Trichechus manatus latirostris) from the 1996 epizootic: gross, histologic, and immunohistochemical features. Toxicologic Pathology 26: 276 282. Bossart GD, Meisner RA, Rommel S, Ghim S, Jenson AB. 2002. Pathological fea tures of the Florida manatee cold stress syndrome. Aquatic Mammals 29: 9 17. Bossart GD, Meisner RA, Rommel SA, Lightsey JD, Varela RA, Defran R. 2004. Pathologic Findings in Florida Manatees(Trichechus manatus latirostris). Aquatic Mammals 30: 434 440. Bu ergelt C, Bonde R, Beck C, O'S hea T. 1984. Pathologic findings in manatees in Florida. Journal of the American Veterinary Medical Association 185: 1331 1334. Campbell HW, Irvine, A.B. 1981. Manatee mortality dueing the unusually cold winter of 1976 1977. . Brownell RL Jr The West Indian manatee in Florida Proceedings of a workshop held in Orlando Florida 27 29 March 1978: 86 91. Clauss A. 1957. Rapid physiological coagulation method in determination of fibrinogen. Acta Haematologica 17: 237 246. Costa D, Wi lliams T. 1999. Marine mammal energetics. Biology of marine mammals. Smithsonian Institution Press, Washington, DC : 176 217. Cray C, Rodriguez M, Dickey M, Brewer LB, Arheart KL. 2013. Assessment of serum amyloid A levels in the rehabilitation setting in t he Florida Manatee (Trichechus manatus latirostris). Journal of Zoo and Wildlife Medicine 44: 911 917. De Nicola P, Mazzetti GM. 1955. Evaluation of thrombelastography. American Journal of Clinical Pathology 23: 447 452. Delgado M, Monreal L, Armengou L, R Ãos J, Segura D. 2009. Peritoneal d Dimer Concentration for Assessing Peritoneal Fibrinolytic Activity in Horses with Colic. Journal of Veterinary Internal Medicine 23: 882 889.
92 Dierauf L, Gulland FM. 2001. CRC handbook of marine mammal medicine: health, d isease, and rehabilitation. CRC press. Domning DP. 2001. The earliest known fully quadrupedal sirenian. Nature 413: 625 627. Donahue SM, Otto CM. 2005. Thromboelastography: a tool for measuring hypercoagulability, hypocoagulability, and fibrinolysis. Journ al of Veterinary Emergency and Critical Care 15: 9 16. Edwards RM, Naik Mathuria BJ, Gay AN, Olutoye OO, Teruya J. 2008. Parameters of thromboelastography in healthy newborns. American Journal of Clinical Pathology 130: 99 102. Eichinger S, Minar E, Bialon czyk C, Hirschl M, Quehenberger P, Schneider B, Weltermann A, Wagner O, Kyrle PA. 2003. D dimer levels and risk of recurrent venous thromboembolism. Journal of American Medical Association 290: 1071 1074. Epstein KL, Brainard BM, Lopes MA, Barton MH, Moore JN. 2009. Thrombelastography in 26 healthy horses with and without activation by recombinant human tissue factor. Journal of Veterinary Emergency and Critical Care 19: 96 101. Fenty RK, Shaw SE, O'toole TE. 2011. Identification of hypercoagulability in do gs with primary immune mediated hemolytic anemia by means of thromboelastography. Journal of the American Veterinary Medical Association 238: 463 467. Finlayson R. 1965. Spontaneous arterial disease in exotic animals. In Proceedings: Proceedings of the Zool ogical Society of London. Wiley Online Library. pp. 239 343. Flamm RO, Reynolds III JE, Harmak C. 2013. Improving Conservation of Florida Manatees (Trichechus manatus latirostris): Conceptualization and Contributions Toward a Regional Warm Water Network Ma nagement Strategy for Sustainable Winter Habitat. Environmental M anagement 51: 154 166. Fourrier F, Chopin C, Goudemand J, Hendrycx S, Caron C, Rime A, Marey A, Lestavel P. 1992. Septic shock, multiple organ failure, and disseminated intravascular coagulat ion. Compared patterns of antithrombin III, protein C, and protein S deficiencies. CHEST Journal 101: 816 823.
93 Friedrichs KR, Harr KE, Freeman KP, Szladovits B, Walton RM, Barnhart KF, Blanco Chavez J. 2012. ASVCP reference interval guidelines: determination of de novo reference intervals in veterinary species and other related topics. Veterinary Clinical Pathology 41: 441 453. Fwc. 2015. http://myfwc.com/research/manatee/rescue mortality response/mortality statistics/yearly/ . Internet Retreived on October 31 st 2015. Garcia Rodriguez A, Bowen B, Domning D, Mignucci Giannoni A, Marmontel M, M ontoya Ospina R, Morales Vela B, Rudin M, Bonde R, Mcguire P. 1998. Phylogeography of the West Indian manatee (Trichechus manatus): how many populations and how many taxa? Molecular Ecology 7: 1137 1149. Gentry PA. 2004. Comparative aspects of blood coagul ation. The Veterinary Journal 168: 238 251. Gentry PA, Ross ML, Yamada M. 1996. Blood coagulation profile of the Asian elephant (Elephas maximus). Zoo Biology 15: 413 423. Hall DJ, Rush JE, Delaforcade AM, Shaw SP. 2012. Kaolin activated thromboelastograph y in echocardiographically normal cats. American Journal of Veterinary Research 73: 775 778. Halvorsen KM, Keith EO. 2008. Immunosuppression cascade in the Florida manatee (Trichechus manatus latirostris). Aquatic Mammals 34: 412 419. Harr K, Harvey J, Bon de R, Murphy D, Lowe M, Menchaca M, Haubold E, Francis Floyd R. 2006. Comparison of methods used to diagnose generalized inflammatory disease in manatees (Trichechus manatus latirostris). Journal of Zoo and Wildlife Medicine 37: 151 159. Harvey JW, Harr KE , Murphy D, Walsh MT, Chittick EJ, Bonde RK, Pate MG, Deutsch CJ, Edwards HH, Haubold EM. 2007. Clinical biochemistry in healthy manatees (Trichechus manatus latirostris). Journal of Zoo and Wildlife Medicine 38: 269 279. Harvey JW, Harr KE, Murphy D, Wals h MT, Nolan EC, Bonde RK, Pate MG, Deutsch CJ, Edwards HH, Clapp WL. 2009. Hematology of healthy Florida manatees (Trichechus manatus). Veterinary Clinical Pathology 38: 183 193.
94 Hawkey C. 1975. Comparative mammalian hematology: cellular components and blo od coagulation of captive wild mammals. William Heinemann Educational Books Ltd, London, UK. ISBN 10 433133902: 324. Horn PS, Pesce AJ. 2003. Reference intervals: an update. Clin Chim Acta 334: 5 23. Hussein M, Aljumaah R, Homeida A, Alhaidary A, Alshaikh M, Elnabi AG, Mohammed O, Omer SA, Macasero W. 2010. Coagulation profile and platelet parameters of the Arabian sand gazelle (Gazella subgutturosa marica): comparison with humans and camels. Journal of Wildlife Diseases 46: 1165 1171. Innis C, Nyaoke AC, W illiams Iii CR, Dunnigan B, Merigo C, Woodward DL, Weber ES, Frasca Jr S. 2009. Pathologic and parasitologic findings of cold stunned Kemp's ridley sea turtles (Lepidochelys kempii) stranded on Cape Cod, Massachusetts, 2001 2006. Journal of Wildlife Diseas es 45: 594 610. Irvine B. 1983. Manatee metabolism and its influence on distribution in Florida. Biological Conservation 25: 315 334. Jackson G, Ashpole K, Yentis S. 2009. The TEGÂ® vs the ROTEMÂ® thromboelastography/thromboelastometry systems. Anaesthesia 6 4: 212 215. Keller KA, Innis CJ, Tlusty MF, Kennedy AE, Bean SB, Cavin JM, Merigo C. 2012. Metabolic and respiratory derangements associated with death in cold stunned Kemp's ridley turtles (Lepidochelys kempii): 32 cases (2005 2009). Journal of American V eterinary Medical Association 240: 317 323. Kelley D, Lester C, Shaw S, De Laforcade A, Webster CR. 2015. Thromboelastographic Evaluation of Dogs with Acute Liver Disease. Journal of Veterinary Internal Medicine 29: 1053 1062. Koelsch JK. 2001. Reproductio n in female manatees observed in Sarasota Bay, Florida. Marine Mammal Science 17: 331 342. Laist DW, Reynolds J E . 2005a. Influence of power plants and other warm water refuges on florida manatees. Marine Mammal Science 21: 739 764. Laist DW, Reynolds JE. 2 005b. Florida manatees, warm water refuges, and an uncertain future. Coastal Management 33: 279 295.
95 Laist DW, Taylor C, Reynolds JE, 3rd. 2013. Winter habitat preferences for Florida manatees and vulnerability to cold. PLoS One 8: e58978. Langtimm CA, Bec k CA. 2003. Lower survival probabilites for adult Florida manatees in years with intense coastal storms. Ecological Applications 13: 257 268. Lewis JH. 1996. Comparative Hemostasis. In Comparative Hemostasis in Vertebrates. Springer. pp. 325 359. Maas C, RennÃ© T. 2012. Regulatory mechanisms of the plasma contact system. Thrombosis Research 129: S73 S76. Marder VJ, Aird WC, Bennett JS, Schulman S, White GC. 2012. Hemostasis and Thrombosis: Basic Principles and Clinical Practice. Wolters Kluwer Healt h. Martin J, Edwards HH, Fonnesbeck CJ, Koslovsky SM, Harmak CW, Dane TM. 2015. Combining information for monitoring at large spatial scales: First statewide abundance estimate of the Florida manatee. Biological Conservation 186: 44 51. Massignon D, Moulsm a M, Bondon P, Debize G, Abidi H, Buttin T, Bon C, Pillonchery G, Coeur P. 1996. Prothrombin time sensitivity and specificity to mild clotting factor deficiencies of the extrinsic pathway: evaluation of eight commercial thromboplastins. Thrombosis and H aem ostasis 75: 590 594. Medway W, Dodds W, Moynihan A, Bonde R. 1982. Blood coagulation of the West Indian manatee (Trichechus manatus). The Cornell V eterinarian 72: 120 127. Meijers JC, Tekelenburg WL, Bouma BN, Bertina RM, Rosendaal FR. 2000. High levels of coagulation factor XI as a risk factor for venous thrombosis. New England Journal of Medicine 342: 696 701. Mennen LI, De Maat MP, Schouten EG, Kluft C, Witteman JC, Hofman A, Grobbee DE. 1998. Dietary effects on coagulation factor VII vary across genotyp es of the R/Q353 polymorphism in elderly people. J ournal of Nutr ition 128: 870 874. Morin D, Yamada M, Gentry P. 1995. Procoagulant, anticoagulant and fibrinolytic activities in llama plasma. Comparative Haematology International 5: 125 129.
96 Nelson OL. 200 5. Use of the D dimer assay for diagnosing thromboembolic disease in the dog. Journal of the American Animal Hospital Association 41: 145 149. O'S hea TJ, Beck CA, Bonde RK, Kochman HI, Odell DK. 1985. An analysis of manatee mortality patterns in Florida, 1 976 81. The Journal of Wildlife M anagement : 1 11. Ortiz RM. 2001. Osmoregulation in marine mammals. Journal of Experimental Biology 204: 1831 1844. Otto CM, Rieser TM, Brooks MB, Russell MW. 2000. Evidence of hypercoagulability in dogs with parvoviral ente ritis. Journal of the American Veterinary Medical Association 217: 1500 1504. Owen HC, Flint M, Limpus CJ, Palmieri C, Mills PC. 2013. Evidence of sirenian cold stress syndrome in dugongs Dugong dugon from southeast Queensland, Australia. Dis orders of Aqua t atic Organ isms 103: 1 7. Palmer L, Martin L. 2014. Traumatic coagulopathy Part 1: Pathophysiology and diagnosis. Journal of Veterinary Emergency and Critical Care 24: 63 74. Patthy L. 1990. Evolution of blood coagulation and fibrinolysis. Blood C o agulation & Fibrinolysis: An International Journal in Haemostasis and T hrombosis 1 : 153 166. Ponczek MB, Gailani D, Doolittle RF. 2008. Evolution of the contact phase of vertebrate blood coagulation. Journal of Thrombosis and Haemostasis 6: 1876 1883. Reep R, Bonde RK. 2006. The Florida manatee: biology and conservation. University Pres s of Florida. Reynolds JE, Rommel SA. 1996. Structure and function of the gastrointestinal tract of the Florida manatee, Trichechus manatus latirostris. The Anatomical Record 245: 539 558. Robinson AJ, Kropatkin M, Aggeler PM. 1969. Hageman factor (factor XII) deficiency in marine mammals. Science 166: 1420 1422.
97 Rommel S, Reynolds III J, Lynch H. 2003. Adaptations of the herbivorous marine mammals. In Proceedings: Matching Herbivore Nutrition to Ecosystems Biodiversity. VIth International Symposium on the Nutrition of Herbivores. Proceedings of an International Symposium held in Merida, MÃ©xico. pp. 19 24. Rommel SA, Caplan H. 2003. Vascular adaptations for heat conservation in the tail of Florida manatees (Trichechus manatus latirostris). Journal of A natom y 202: 343 353. Runge MC, Sanders Reed CA, Langtimm CA, Fonnesbeck CJ. 2007. A Quantitative Threats Analysis for the Florida Manatee (Trichechus manatus latirostris). US Geological Survey . Sato N, Takahashi H, Shibata A. 1995. Fibrinogen/fibrin degradation products and D dimer in clinical practice: interpretation of discrepant results. American Journal Hematology 48: 168 174. Siegemund A, Petros S, Siegemund T, Scholz U, Seyfarth H J, Engelmann L. 2004. The endogenous thrombin potential and high levels of c oagulation factor VIII, factor IX and factor XI. Blood Coagulation and Fibrinolysis 15: 241 244. Simon K. 1961. Haematological studies on elephants. Indian Veterinary Journal 38: 241 245. Sinnott VB, Otto CM. 2009. Use of thromboelastography in dogs with i mmune mediated hemolytic anemia: 39 cases (2000 2008). Journal of Veterinary Emergency and Critical Care 19: 484 488. Sommerey C C, Williams TL, Mccrone I, Ruiz Ferreras A, Freeman D, Archer J. 2014. Thromboelastography in healthy dairy cows. Journal of Dairy S cience 97: 5474 5480. Stith B, Slone D, De Wit M, Edwards H, Langtimm C, Swain E, Soderqvist L, Reid J. 2012. Passive thermal refugia provided warm water for Florida manatees during the severe winter of 2009 2010. Marine Ecology Progress Series 462: 287 30 1. Stokol T, Brooks MB, Erb HN. 2000a. Effect of citrate concentration on coagulation test results in dogs. Journal of the American Veterinary Medical Association 217: 1672 1677.
98 Stokol T, Brooks MB, Erb HN, Mauldin GE. 2000b. D dimer concentrations in hea lthy dogs and dogs with disseminated intravascular coagulation. American Journal of Veterinary Research 61: 393 398. Triplett D, Harms C. 1981. Factor assays. Procedures for the coagulation laboratory. Chicago: American Society of Clinical Pathologists Pre ss. p : 38 57. Wada H, Mori Y, Okabayashi K, Gabazza EC, Kushiya F, Watanabe M, Nishikawa M, Shiku H, Nobori T. 2003. High plasma fibrinogen level is associated with poor clinical outcome in DIC patients. American Journal of Hematology 72: 1 7. Walsh CJ, Lu er CA, Noyes DR. 2005. Effects of environmental stressors on lymphocyte proliferation in Florida manatees, (Trichechus manatus latirostris). Veterinary Immunology and Immunopathology 103: 247 256. Walsh M, Bossart G. 1999. Manatee medicine. Zoo & Wild Anim al Medicine. Current Therapy 4: 507 516. White J, Harkness D, Isaacks R, Duffield D. 1976. Some studies on blood of the Florida manatee, Trichechus manatus latirostris. Comparative Biochemistry and Physiology Part A: Physiology 55: 413 417. Wiinberg B, Jen sen AL, Johansson P, Rozanski E, Tranholm M, Kristensen AT. 2008. Thromboelastographic evaluation of hemostatic function in dogs with disseminated intravascular coagulation. Journal of Veterinary Internal Medicine 22: 357 365. Wiinberg B, Jensen AL, Rojkja er R, Johansson P, Kjelgaard Hansen M, Kristensen AT. 2005. Validation of human recombinant tissue factor activated thromboelastography on citrated whole blood from clinically healthy dogs. Veterinary Clinical Pathology 34: 389 393. Wilson R, Reynolds III J, Wetzel D, Schweirzke Wade L, Bonde R, Breuel K, Roudebush W. 2011. Secretion of anti Mullerian hormone in the West Indian manatee (Trichechus manatus) with implications for assessing conservation status. Endangered Species Research 10: 3354. Wingfield J, Sapolsky R. 2003. Reproduc tion and resistance to stress: when and how. Journal of N euroendocrinology 15: 711 724.
99 Wong AW, Bonde RK, Siegal Willott J, Stamper MA, Colee J, Powell JA, Reid JP, Deutsch CJ, Harr KE. 2012. Monitoring oral temperature, heart rate, and respiration rate o f West Indian manatees (Trichechus manatus) during capture and handling in the field. Aquatic Mammals 38: 1. Worthy G, Miculka T, Wright S. 1999. Manatee response to cold: how cold is too cold. In Proceedings: Florida Manatees and Warm water: Proceedings of the Warm Water Workshop, Jupiter, Florida. pp. 1 6. Worthy GA, Edwards EF. 1990. Morphometric and biochemical factors affecting heat loss in a small temperate cetacean (Phocoena phocoena) and a small tropical cetacean (Stenella attenuata). Physiological Z oology : 432 442. Zambruni A, Thalheimer U, Leandro G, Perry D, Burroughs AK. 2004. Thromboelastography with citrated blood: comparability with native blood, stability of citrate storage and effect of repeated sampling. Blood Coagulation and Fibrinolysis 15 : 103 107. Zoodsma BJ. 1991. Distribution and behavioral ecology of manatees in southeastern Georgia. Biological and Physical aspects of Dredging Kings Bay, Georgia: 15 17.
100 BIOGRAPHICAL SKETCH Ashley Barratclough was born in the North East of England and attended the Royal Veterinary School in London, graduating with her veterinary degree in 2009. Following three years in mix ed practice Ashley undertook a m aster s in Wild Animal Health at the Zo ological Society of London graduating with Merit in September 2013. Ashley was always interested in manatees and in January 2014 commenced a Veterinary Fellowship at the University of Florida to pursue her Master of Science in Veterinary Medical Science wi th emphasis on m anatee h ealth whilst working as a