Evaluation of the Nutrition of Florida Manatees (Trichechus Manatus Latirostris)

MISSING IMAGE

Material Information

Title:
Evaluation of the Nutrition of Florida Manatees (Trichechus Manatus Latirostris)
Physical Description:
1 online resource (192 p.)
Language:
english
Creator:
Harshaw, Lauren T
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Veterinary Medical Sciences, Veterinary Medicine
Committee Chair:
Larkin, Iske
Committee Co-Chair:
Hill, Richard C
Committee Members:
Reep, Roger L
Samuelson, Don A
Staples, Charles R
Worthy, Graham

Subjects

Subjects / Keywords:
digestibility -- manatees -- nutrition
Veterinary Medicine -- Dissertations, Academic -- UF
Genre:
Veterinary Medical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
Florida manatees (Trichechus manatus latirostris) are an endangered species with a slow reproductive rate.  Injured, sick, and/or orphaned manatees frequently need rescue and rehabilitation, but successful rehabilitation requires feeding a well-balanced diet.  The diet of captive manatees, however, consists primarily of romaine lettuce, which is lower in ash and plant fiber than the submerged aquatic vegetation consumed by free-ranging manatees.  Romaine lettuce may be supplemented with other foods, which contain more plant fiber, but the total amount of different nutrients and digestible energy consumed by captive manatees is unknown.  Manatees are able to gain weight and reproduce when fed mostly lettuce, but may consume too much energy, which can result in obesity.  Dietary fiber is vital for normal herbivore intestinal function, and inadequate amounts have been linked to disease in terrestrial herbivores.  Food and fecal samples were collected from healthy long-term and rehabilitating (coldstressed, orphaned, etc.) captive manatees, and free-ranging manatees to assess nutritional parameters.  A survey of the diet at three facilities indicated that manatees consume non-structural carbohydrates as 40% of their dry matter (DM) intake and have greater energy intakes than would be expected for manatees.  In vivo digestibility estimates indicated that manatees are thoroughly capable of digesting the diet offered in captivity (> 68% DM digestibility), whereas in vitro fermentations indicated that manatees ferment salt and fresh water grasses relatively well (33- 60% DM lost).  Fecal lactate and short-chain fatty acid concentrations were found to be different amongst manatees at different locations, and between manatees ingesting a natural vs. a captivity-based diet.  Lastly, body condition indices, assessed using morphological data, revealed differences in length-weight relationships between sexes and three geographic locations in Florida.  The results of this research suggest that the diet for captive Florida manatees may require additional plant fiber, but also needs further assessment paired with careful monitoring of body condition.  The information presented here will hopefully be useful for future management decisions involving the nutrition of captive and rehabilitating Florida manatees.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Lauren T Harshaw.
Thesis:
Thesis (Ph.D.)--University of Florida, 2012.
Local:
Adviser: Larkin, Iske.
Local:
Co-adviser: Hill, Richard C.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2014-08-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2012
System ID:
UFE0044549:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 EVALUATION OF THE NUTRITION OF FLORIDA MANATEES ( TRICHECHUS MANATUS LATIROSTRIS ) By LAUREN T. HARSHAW A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Lauren T. Harshaw

PAGE 3

3 To m y Grandparents, who showed me a love for all animals and encouraged me to follow my dreams

PAGE 4

4 ACKNOWLEDGMENTS I would like to begin by thanking my r esearch supervisory committee, who guided me through this process and were always available to discuss result s and navigate problems. My chair Dr. Iske Larkin and co chair, Dr. Richard Hill, provided direction throughout the duration of my project and pu shed me towards becoming a better research scientist. Dr. Don Samuelson offered encouragement and engaged me in discussions that contributed to my development of critical thinking. Dr. Roger Reep shared his vast knowledge of manatee biology with me and h elped me to better understand my research and how it could be applied to manatee conservation with which he usually conducts research and graciously opened up his nutrition laboratory to me. Dr. Graham Worthy was located at another university, but never felt distant, as he provided immense amounts of advice over the phone and traveled here for the important events of my PhD. Together, my committee formed a bridge between manatee biology and the science of nutrition, thus allowing me to successfully complete my research. I especially thank Dr. Karen Scott for mentoring me in my daily laboratory work and assisting me with various aspects of my research. She always let me bounce ideas off of her and discuss my research results. I would also like to thank my stellar undergraduate assistant, Leah Zigelsky, who accompanied me on trips to collect samples and spent countless hours in the laboratory helping me prep manate e food and fecal samples. Most of this work would not have been possible without Dr. Adegbola Adesogan and his amazing lab team, including Jan Kivipelto and Dr. Kathy Arriola. They allowed me to work in their lab space and helped me whenever I asked. I also owe thanks to Heather Daniel Maness and Patrick Thompson of the Aquatic Animal Health Program for their unwavering support and assistance.

PAGE 5

5 I received copious amounts of help and cooperation from the captive facilities that participated in this r esearch and thank all of them for their assistance. To start, I want to thank the Living Seas marine mammal training team and Dr. Andy Stamper for always making me feel at home when I was there and providing me with all of the information I needed to make this project possible. The staff members at Lowry Park Zoo and Homosassa Springs were always a pleasure to be around and provided me with the opportunity to collect samples from their animals, even though I was invading their space. Joe Gaspard and the staff of the manatee research program at Mote Marine Laboratory deserve special thanks for collecting the diet and fecal samples from their animals for the digestibility study, saving me lots of time and effort. I thank Marilyn Margold and the staff at Pa rker Aquarium for trying on more than one occasion to collect fecal samples from their beloved Snooty and allowing me to come meet him and try for myself as well. Lastly, I wish to thank Mark Campbell and the staff at Cincinnati Zoo for collecting last mi nute fecal samples from their animals and shipping them to me, because I could not travel to their facility and collect them myself. This work could also not have been done without the help and support of colleagues at USGS Sirenia P roject, particularly Dr Bob Bonde and Cathy Beck, and the Florida Fish and St. Pete pathology and necropsy lab. Also, I thank Joyce Kleen at the Crystal River Wildlife Refuge office for facilitating my access to the land portion of Three Sisters Springs. Lastly, I would have never made it through this without the support of my friends and family, who had to listen to me talk about manatee poo on a daily basis. I was even fortunate enough to have some brave friends (Michelle Davis, Susan Fogelson, and Elisa Livengood) snorkel with me in freezing cold conditions to try and collect fecal samples. I literally put my

PAGE 6

6 best friend, Jill Johnston, to sleep some nights with discussions of my rese arch and frustrations, but she always answered my phone calls anyway. I thank my friend Alexis Morris for encouraging me to keep going by helping me visualize the light at the end of the tunnel and always, always being there for me My parents, Barry and David Harshaw, always called to check on me and make sure I was taking proper care of myself. Robyn, my sister, moved to Switzerland, but still managed to support me from afar. I also thank my dog, Scamp, who had to forego longer walks and playtime on s ome days so that I could finish my work, but still provided believe in myself and I am forever grateful for that.

PAGE 7

7 TABLE OF CONTENTS Page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ......................... 10 LIST OF ABBREVIATIONS ................................ ................................ ................................ ........ 14 ABSTRACT ................................ ................................ ................................ ................................ ... 16 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .................. 18 Project Summary ................................ ................................ ................................ .................... 18 L iterature Review ................................ ................................ ................................ ................... 19 2 QUANTITATIVE SURVEY OF THE DIET OF FLORIDA MANATEES ( TRICHECHUS MANATUS LATIROSTRIS ) AT CAPTIVE FACILITIES IN FLORIDA .... 35 Background ................................ ................................ ................................ ............................. 35 Materials and Methods ................................ ................................ ................................ ........... 37 Manatees ................................ ................................ ................................ .......................... 37 Nutrient Intake Calculations ................................ ................................ ............................ 38 Statistics ................................ ................................ ................................ ........................... 39 Results ................................ ................................ ................................ ................................ ..... 40 Discu ssion ................................ ................................ ................................ ............................... 42 3 IN VIVO APPARENT DIGESTIBILITY TRIALS OF CAPTIVE FLORIDA MANATEES ( TRICHECHUS MANATUS LATIROSTRIS ) ................................ ................... 57 Background ................................ ................................ ................................ ............................. 57 Materials and Methods ................................ ................................ ................................ ........... 60 Manatees and Sample Collection ................................ ................................ .................... 60 Sample Processing ................................ ................................ ................................ ........... 61 Nutrient Analysis ................................ ................................ ................................ ............. 62 Statistical Analysis ................................ ................................ ................................ .......... 64 Resu lts ................................ ................................ ................................ ................................ ..... 65 Discussion ................................ ................................ ................................ ............................... 67 4 IN VITRO FERMENTATION ESTIMATES FOR FREE RANGING, CAPTIVE, AND REHABILITATING FLORIDA MANATEES ( TRICHECHU S MANATUS LATIROSTRIS ) ................................ ................................ ................................ ....................... 81 Background ................................ ................................ ................................ ............................. 81 Materials and Methods ................................ ................................ ................................ ........... 84

PAGE 8

8 Food Sample Collection, Processing, and In vitro Fermentation Method ...................... 84 Preliminary Study Using Horse Feces to Evaluate Methods of Preserving Fecal Microbes During Transport to the Laboratory ................................ ............................. 87 Study Comparing the Effect of Feces from Free Ranging and Captive Manatees on Fermentation of Various Foods ................................ ................................ ................... 88 Short chain a nd Branched chain Fatty Acid Analysis ................................ .................... 91 Calculations and Statistics ................................ ................................ ............................... 92 Results ................................ ................................ ................................ ................................ ..... 93 Discussion ................................ ................................ ................................ ............................... 95 5 FECAL SHORT CHAIN FATTY ACID AND LACTIC ACID CONCENTRATIONS OF FLORIDA MANATEES ( TRICHECHUS MANATUS LATIROSTRIS ) ........................ 110 Background ................................ ................................ ................................ ........................... 110 Materials and Methods ................................ ................................ ................................ ......... 113 Manatees and Sample Collection ................................ ................................ .................. 113 Sample Storage and Preparation ................................ ................................ .................... 114 HPLC Apparatus ................................ ................................ ................................ ........... 115 Data Processing ................................ ................................ ................................ ............. 116 Statistical Analysis ................................ ................................ ................................ ........ 117 Results ................................ ................................ ................................ ................................ ... 118 Discussion ................................ ................................ ................................ ............................. 119 6 MORPHOMETRIC BODY CONDITION INDICES OF FREE RANGING FLORIDA MANATEES ( TRICHECHUS MANATUS LATIROSTRIS ) ................................ ................. 142 Background ................................ ................................ ................................ ........................... 14 2 Materials and Methods ................................ ................................ ................................ ......... 144 Results ................................ ................................ ................................ ................................ ... 146 Discussion ................................ ................................ ................................ ............................. 147 7 CONCLUSIONS ................................ ................................ ................................ .................. 166 Studying Nutrition in the Florida Manatee ................................ ................................ ........... 166 Chapter Conclusions ................................ ................................ ................................ ............. 167 Chapter 2: Quantitative Survey of the Diet of Florida Manatees at Captive Facilities in Florida ................................ ................................ ................................ .................... 167 Chapter 3: In Vivo Apparent Digestibility Trials o f Captive Florida Manatees ............ 168 Chapter 4: In Vitro Fermentation Estimates for Free ranging, Captive, and Rehabilitating Florida Manatees ................................ ................................ ................ 170 Chapter 5: Fecal Short Chain Fatty Acid and Lactic Acid Concentrations of Florida Manatees ................................ ................................ ................................ .................... 172 Chapter 6: Morphometric Body Condition Indices of Free Ranging Florida Manatees ................................ ................................ ................................ .................... 173 General Conclusions and Future Directions ................................ ................................ ......... 174 LIST OF REFERENCES ................................ ................................ ................................ ............. 177

PAGE 9

9 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 192

PAGE 10

10 LIST OF TABLES Table page 1 1 The two categories of extant mammalian hindgut fermenters. ................................ .......... 34 2 1 Location, sex, seasons when food intake was documented, life stage, and body weight of manatees for which food intake was documented. ................................ ............ 49 2 2 Average perc ent of each dietary ingredient offered (as fed weight basis) by each facility. ................................ ................................ ................................ ............................... 50 2 3 Mean daily DM, CF, CP, Ash, NDF, ADF, NSC consumption DE intake, and DE intake/Body Weight for males and f emales of combined life stages. ................................ 51 3 1 Life stage, body weight, measurement period, and daily dry matter intake of manatees when food intake was documented and feces were collected ............................ 74 3 2 Crude protein (CP), neutral detergent fiber (NDF), and acid detergent fiber (ADF) composition of foods and complete diet fed to manatees at each facility. ........................ 75 3 3 Crude protein (CP), neutral detergent fiber (NDF), and acid detergent fiber (ADF) composition of combined fecal samples for each of the manatees in the study. ............... 76 3 4 Acid insoluble ash (AIA) and acid detergent lignin (ADL) concentrations in the diet and feces for each manatee and season. ................................ ................................ ............. 77 3 5 Apparent digestibility values for dry matter (DM), crude p rotein (CP), neutral detergent fiber (NDF), and acid detergent fiber (ADF). ................................ .................... 78 4 1 Manatees, location, and date of collection of feces used for in vitro fermentation assays. ................................ ................................ ................................ .............................. 103 4 2 Initial NDF concentration (dry matter basis) of each food used in the in vitro fermentation assays. ................................ ................................ ................................ ......... 104 4 3 Percent NDF lost from each foo d after in vitro fermentation with inoculums of feces collected from free ranging and captive manatees at three locations in Florida. ............ 105 4 4 Percent DM lost from each food after in vitro ferm entation with inoculums of feces collected from free ranging and captive manatees at three locations in Florida. ............ 106 4 5 Initial and post fermentation concentrations (mmol/L of in vitro fluid) of lactate, SCFA, BCFA ................................ ................................ ................................ ................... 107 5 1 Environmental Parameters and number of manatees sampled at each facility ................ 127 5 2 Health status, nature of injury, sex, location, and date of collection of captive manatees from which fecal samples were obtained. ................................ ........................ 128

PAGE 11

11 5 3 Lactic, short chain fatty acid concentrations in mmol/g DM feces, an d A:P ratio for free ranging manatees and manatees in captivity. ................................ ........................... 129 5 4 Dry matter content of feces from free ranging manatees and captive manatees located at Lowry Park Zoo, Homosassa Sprin gs, and Living Seas ................................ 130 5 5 Lactic and short chain fatty acid concentrations in mmol/L fecal water from free ranging and captive manatees at various locations in Florida. ................................ ........ 131 6 1 Body condition constants for free ranging manatees of each sex at three locations in Florida. ................................ ................................ ................................ ............................. 153 6 2 Intercepts and slopes of regression lin es of the logarithm of weight against logarithm of morphometric measurement of free ranging manatees of each sex for all locations. 154 6 3 Amended body condition constants of free ranging man atees of different sexes at three locations in Florida. ................................ ................................ ................................ 155

PAGE 12

12 LIST OF FIGURES Figure page 2 1 Percent of each nutrient on a DM basis for captivity ba sed diet surveyed compared to values reported previously for submerged aquatic vegetation (SAV) ............................... 52 2 2 Mean nutrient concentration per day, as a percent of total dry matter intake for each man atee studied during winter and spring ................................ ................................ ......... 53 2 3 Dry matter intake (grams per day) for each manatee studied during winter and spring .... 54 2 4 Linear regression of log of body weight vs. log of daily digestible energy intake of adult manatees ................................ ................................ ................................ .................... 55 2 5 Linear regression of log of body weight vs. log of digestible energy intake f or subadult manatees ................................ ................................ ................................ .............. 56 3 1 Percent acid insoluble ash (AIA) vs. total amount of collected feces ............................... 79 3 2 Linear regression of dry matter (DM) intake vs. % DM digestibility for manatees in the present study ................................ ................................ ................................ ................ 80 4 1 Percentage loss of NDF from two hays during in vitro fermentation using microbial inoculums produced from ho rse feces. ................................ ................................ ............ 108 4 2 Absolute increase in concentration of lactate, SCFA, BCFA, and total SCFA plus BCFA after in vitro fermentation of foods with inoculums of manatee feces ................. 109 5 1 Lactic acid concentration in mmol per grams of dry matter (DM) feces from free ranging and captive manatees at each location ................................ ................................ 132 5 2 Acetic acid concentration in mmol per grams of dry matter (DM) from free ranging and captive manatees at each location ................................ ................................ ............. 133 5 3 Propionic acid concentration in mmol per grams of dry matter (DM) feces from free ranging and captive manatees at each location. ................................ ............................... 134 5 4 Butyric acid concentration in mmol per grams of dry matter (DM) feces from free ranging and captive manatees at each location ................................ ................................ 135 5 5 Valeric ac id concentration in mmol per grams of dry matter (DM) feces from free ranging and captive manatees at each location ................................ ............................... 136 5 6 Isobutyric acid concentration in mmol per grams of dry matter (DM) feces from free ranging and captive manatees at each location ................................ ............................... 137 5 7 Isovaleric acid concentration in mmol per grams of dry matter ( DM) feces from free ranging and captive manatees at each location ................................ ............................... 138

PAGE 13

13 5 8 Total SCFA concentration in mmol per grams of dry matter (DM) feces from free ranging and captive ma natees at each location ................................ ............................... 139 5 9 Acetate to propionate (A:P) ratio from free ranging and captive manatees at each location ................................ ................................ ................................ ............................. 140 5 10 Individual and total SCFA concentrations in mmol per gram of dry matter (DM) feces for a single long term, healthy female ................................ ................................ .... 141 6 1 Regression of body condition constant, K SL (Body condition index 1), against straight length (SL) for all females ................................ ................................ .................. 156 6 2 Regression of body condition constant, K CL (Body condition index 2), against curvilinear length (CL) for all females ................................ ................................ ............ 157 6 3 Regression of body condition constant K SLMG (Body condition index 3), against straight length (SL) for all females ................................ ................................ .................. 158 6 4 Regression of body condition constant, K SL (Body condition index 1), against straight length (SL) for all males ................................ ................................ ..................... 159 6 5 Regression of body condition constant, K CL (Body condition index 2), against curvilinear length (CL) for all males ................................ ................................ ............... 160 6 6 Regression of body condition constant, K SLMG (Body condition index 3), against straight length (SL) for all males ................................ ................................ ..................... 161 6 7 Regression of the logarithm of weight against straight length for females (solid circles) and males (open circles) ................................ ................................ ...................... 162 6 8 Regression of the logarithm of maximum gi rth (MG) against logarithm of straight length (SL) for females (solid circles) and males (open circles) ................................ ..... 163 6 9 Condition constants (kg/m 3 ) of BCI 7 (W/SL 2.578 ) calculated from historical morphometric measurements for a long term, healthy, captive male .............................. 164 6 10 Condition constants (kg/m 3 ) of BCI 7 (W/SL 2.578 ) calculated from historical morphometric measurements for a second long term, healthy, captive male .................. 165

PAGE 14

14 LIST OF ABBREVIATION S 2N two normal A : P Acetate to propionate ACS American Chemi cal Society ADF Acid detergent fiber ADL Acid detergent lignin ADS Acid detergent solubles AIA Acid insoluble ash ANOVA Analysis of variance BC Brevard County BCFA Branched chain fatty acids BCI Body condition index BMR Basal metabolic rate CF Crude fat CL Curvilinear length CP Crude protein CR Crystal River CTAB Cetyl trimethylammonium bromide CV Coefficient of variation CZ Cincinnati Zoo DE Digestible energy DEi Digestible energy intake DM Dry matter FR Free ranging HCl Hydrochloric acid

PAGE 15

15 HPLC High perform ance liquid chromatography HS Homosassa Springs IACUC Institutional animal care and use committee IVDMD In vitro dry matter digestibility Kcal Kilocalories LPZ Lowry Park Zoo MG Maximum girth MRP Manatee Rehabilitation Program MRT Mean retention time NDF Neutral detergent fiber NDS Neutral detergent solubles NSC Non structural carbohydrates SAV Submerged aquatic vegetation SCFA Short chain fatty acids SL Straight length USGS United States Geological Survey W Body weight

PAGE 16

16 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EVALUATION OF THE NU TRITION OF FLORIDA M ANATEES ( TRICHECHUS MANATUS LATIROSTRIS ) By Lauren T. Harshaw Au gust 2012 Chair: Iske V. Larkin Cochair: Richard Hill Major: Veterinary Medical Sciences Florida manatees ( Trichechus manatus latirostris ) are an endangered species with a slow rep roductive rate. Injured, sick and /or orphaned manatees frequently need re scue and rehabilitation but s uccessful rehabilitation requires feeding a w ell balanced diet. The diet of captive manatees, however consists primarily of romaine lettuce, which is lower in ash and plant fiber than t he submerged aquatic vegetation consume d by free ranging manatees Romaine lettuce may be supplemented other foods which contain more plant fiber, but the total amount of different nutrients and digestible energy consumed by captive manatees is unknown. Manatees are able to gain weight and r eproduce when fed mostly lettuce, but may consume too much energy, which can result in obesity. D ietary fiber is vital for normal herbivore intestinal function, and i nadequate amounts have been linked to disease in terrestrial herbivores. Food and f ecal samples were collected from healthy long term and rehabilitating (cold stressed, orphaned, etc.) captive manatees and free ranging manatees to assess nutritional parameters A survey of the diet at three facilities indicated that manatees consume non str uctural carbohydrate s as 40% of their dry matter (DM) intake and have greater energy intakes than would be expected for manatees In vivo digestibility estimates indicated that manatees are thoroughly capable of

PAGE 17

17 d igesting the diet offered in cap tivity (> 68% DM digestibility), whereas in vitro fermentations indicated tha t manatees ferment salt and fresh water grasses relatively well (33 60% DM lost) Fecal lactate and short chain fatty acid concentrations were found to be different amongst manatees at di fferent locations and between manatees ingesting a natural vs. a captivity based diet Lastly, body condition indices, assessed using morphological data, revealed differences in length weight relationships between sexes and three geographic locations in Florida. The results of this research suggest that the diet for captive Florida manatees may require additional plant fiber but also needs further assessment paired with careful monitoring of body condition The information presented here will hopefully be useful for future management decisions involving the nutrition of captive and rehabilitating Florida manatees.

PAGE 18

18 CHAPTER 1 INTRODUCTION P roject S ummary The project presented here is divided into seven chapters Little is known about the nutrition and g astrointestinal function of Florida manatees, so the research in the present work quantified the food intake of captive manatees, investigated the digestibility of different diets, fermentative ability of fecal microbiota fecal short chain fatty acid conc entrations, and body condition indices to provide a foundation for future studies in manatee nutrition. An overall goal of the research was to determine whether low dietary fiber affects digestion and fermentation in captive manatees. Chapter 1, the intr oduction, provides background information on both the Florida manatee and herbivore nutr ition in general. Chapter 2 reports a survey of the diet at two different facilities that quantified the nutrient and energy intake of captive manatees Chapter 3 rep orts in vivo d igestibility values that were measured in long term, healthy, and recently reh abilitated captive manatees to determine how well individuals process the diet offered in captivity In Chapter 4, fermentation of forages was measured in vitro to find out how well fecal microbes from captive, free ranging, and rehabilitating manatees ferment other sources of plant fiber, including salt and fresh water grasses. Chapter 5 reports the c oncentration s of short chain fatty acids and lactate that were c ompared among fecal samples from healthy and rehabilitating manatees in captivity at different facilities and free ranging individuals In order to be able to evaluate the nutritional status of captive and free ranging manatees, m orphological measurements and body weights were compared among free ranging manatees from three geographic locations in Florida to establish a normal range of body condition indices in Chapter 6. The last chapter summarizes the results and findings of this project in the hope th at it will

PAGE 19

19 provide useful information to managers and caregivers responsible for the future care of captive and rehabilitating manatees. Literature Review The Florida manatee ( Trichechus manatus latirostris ) is a member of the o rder Sirenia, which is com prised of aquatic mammals including the dugong ( Dugong dugon ) and the three species of manatees the West Indian manatee ( Trichechus manatus ), the West African manatee ( Trichechus senegalensis ), the Amazonian manatee ( Trichechus inunguis ) and the now ext inct ( Hydrodamalis gigas ) There are two subspecies of West Indian manatee : the Antillean ( Trichechus manatus manatus ) and Florida manatees. Manatees a nd other s irenians are the only herbivorous marine mammals with the Florida manatee being a generalist herbivore Evolutionarily and genetically, they are more closely related to elephants ( family: Elephantidae ) and the hyrax (family: Hyracoidea ) than to other marine mammals [ Reep and Bo nde, 2006; Pardini et al., 2007] The Florida mana tee is protected under the Endangered Species Act of 1973, the Marine Mammal Protection Act of 1972 and the Florida Manatee Sanctuary Act of 1978. Nevertheless, these creatures are still vulnerable to anthropogenic effects that result from boat traffic, f ishing gear, pollution, and habitat destruction. Manatees are slow moving, so boat impact injuries are a significant problem for them, because they cannot get out of the way fast enough. Manatees are also susceptible to natural causes of disease and mort ality s uch as cold stress and red tide, and have a slow reproductive rate, so research on how to maintai n their population is critical for their recovery The aquatic nature of this mammal makes studying certain aspects of its biology difficult in its natu ral habitat Facilities that are part of the Manatee Rehabilitation P rogram (MRP), where injured or sick individuals are brought for care until they are deemed ready for release back into the wild provide an alternative location for study Specifically, these facilities provide

PAGE 20

20 a chance to gain a better understanding of health problems and causes of mortality in free ranging manatees such as cold stress and injuries in a captive setting In captivity, manatees are fed primarily romaine lettuce ( Lactuca s ativa L. var. longifolia ) w hich is sometimes mixed with raw vegetabl es, fruit, and monkey biscuits that contain more plant fiber, but the total amount of nutrients in the captive manatee diet is unknown. Manatees are able to gain weight and reproduce whe n fed this captivity based diet, but appropriate levels of dietary fiber are vital for normal herbivore intestinal function and colonic fermentation. Inadequate amounts of dietary fiber have been linked to dental disease, gastrointestinal stasis, colic, l aminitis insulin resistance, and obesity in terrestrial herbivore s such as the rabbit and horse. Free ranging manatees consume various forms of submerged aquatic vegetation (SAV) including grasses and algae that are higher in plant fiber and ash content than romaine lettuce [Siegal Willott et al., 2010] Thus, it is important to determine how these nutrient differences might affect the nutrition of captive manatees. Manatees, like other herbivores, require a complex digestive system to process the large quantity of plant material that th ey must ingest for subsistence [ Stevens and H ume, 1998] They can occupy fresh, brackish, or salt water habitats and often forage for about 6 8 hours a day on a wide variety of grasses and plants found in these habitats [ Best, 1981; Burn, 1986] and are estimated to consume up to 5 10% of their body weight each day as determined using mean chew rate per plant mass and regression equations to predict chew counts from body weight [ Etheridge et a l. 1985 ] By measuring natu rally occurring stable isotopes in both aquatic vegetation and manatee carcass ti ssue samples, Reich and Worthy [2006] found that Florida manatees consume nearly half of their vegetation from a marine or estuarine source, thus making these food sources sig nificant to their diet. Alves Stanley and Worthy [2009] expanded upon

PAGE 21

21 this knowledge and found carbon and nitrogen stable isotope turnover rates to be relatively slow in captive manatees using diet and living manatee skin samples thus making them useful for long term dietary intake studies. Diet modeling based on carbon and nitrogen stable isotopes of vegetation and skin samples from free ranging manate es found in freshwater and salt water habitats further suggested that sea grasses are a required componen t of the diet for both riverine and coastal Florida manatees [Alves Stanley et al., 2010] Based on mean feeding cycle lengths and handling ability for different a quatic plants, Marshall et al. [2000] noted that manatees could theoretically consume greate r amounts of Hydrilla in a single day than Thalassia but that the nutritional benefits of one plant versus the other are unknown. Manatees maintain a relatively high weight compared to other herbivores but no one has yet examined how the nutritional pro Herbivores must effectively utilize the insoluble fraction of fiber, also known as the structural carbohydrates or plant fiber contained within the cell walls of plants, to obtain the most benefit from th eir diet. Animal enzymes are unable to break down lignin, hemicellulose, and cellulose which are contained in these cell wall components [Van Soest, 1994] Most herbivores, therefore, possess a ferment ation vat to assist in the breakdown of th is plant f iber This vat, or chamber, usually located in ei ther the foregut or hindgut and contains bacteria protozoa, and anaerobic fungi maintained in an environment with a relatively neutral pH [ Mackie, 1997; Stevens and Hume, 199 8; Dehority, 2002; Mackie, 2002] Animals usually obtain these microbes from their environment when feeding and drinking or from the feces of conspecifics, usually their mother, shortly after birth [Troyer, 1984a] These gut microbes provide herbivores with energy and nutrients that t hey could not obtain from plants on their own, and have provided herbivores access to an assortment of ecological niches the se animals

PAGE 22

22 ble to occupy [Troyer, 1984a] In turn, the herbivores provide these microbes with an optimal env ironment in terms of moisture, temperature, and pH, so that these microbes can gr ow and reproduce [Troyer, 1984a] As a result, herbivores exhibit a wide ecological distribution and are found in eleven different orders, including the largest terrestrial s pecies, such as elephants and rhinoceroses ( family: Rhinocerotidae ) [Stevens and Hume, 1998] Foregut fermenters have a microbial fermentation vat located in the upper gast rointestinal tract (GIT). Whereas all foregut fermenters have some secondary fermen tation in the hindgut of the GIT, this secondary contribution is usually considered minor when com pared to that from the foregut [Hume, 2002] This secondary fermentation may exist because foregut fermentation evolved after hindgut fermentation, whereas r u mination evolved more recently [Chivers, 1989; Mackie, 2002] Ruminants, the most common type of foregut fermenters, are better able to digest fibrous foods than herbivorous non ruminants [ V an Soest, 1994; Jurgens, 2002] even though the rumen and hindgu t tend to have simi lar concentrations of microbes [Stevens and Hume, 1998] Ruminants regurgitate and chew their cud and then use microbial fermentations in both the foregut and hindgut to break down fiber, making them the most successful in digest ing hig h fiber foods. Ruminants include sheep and goats ( subfamily: Caprinae ), cattle and buffaloes ( subfamily: Bovinae ), deer ( family: Cervidae ), and giraffes ( family: Giraffidae ). The digestive strategy of ruminants is to maximize the extent versus the rate o f cell wall digestion, so these animals do not need to have high intakes [Hume, 2002] but can still ingest a large volume of low quality food Ruminants may have developed this strategy by living in ecosystems where food availability can be limited, eith er in quality or quantity, but many present day ruminants do not inhabit such environments [Hume, 2002] There are also non ruminant foregut fermenters, including the hoatzin ( Opisthocomus hoazin ), a bird, which contains a unique fermentation vat

PAGE 23

23 within i ts crop. Other non ruminant foregut fermenters include camels and llamas ( family: Camelidae ), kangaroos and wallabies ( family: Macropodidae ), hamsters and voles ( family: Cricetidae ), hippopotamuses ( family: Hippopotamidae ), colobine and langur monkeys ( su bfamily: Colobinae ), as well as the two toed and three toed sloths ( suborder: Folivora ) [Van Soest, 1994; Hume, 2002] All of these animals, with the exception of the hoatzin, have an expanded fore stomach that serves as the site of digesta retention, but without the extensive compartmentalization of the ruminant. Most of these animals employ a different type of digestive strategy, in which they tend to maximize flow or rate of cell wall digestion instead of the degree to which the food is digested [Hume, 2002] As a result, they are not as successful as ruminants in digesting cell wall components, but are able to partially compensate because of a greater rate of digestion and increased food intake. This digestive strategy probably resulted from living i n an environment where there is a lot of food to eat, but the food may not be of high nutritional quality [Hume, 2002] In this way, non ruminant foregut fermenters may serve as a n evolutionary bridge between the ruminant and the hindgut fermenter. Si renians including manatees, are hindgut fermenters. Hindgut fermenters possess other unique adaptations to increase the digestibility of their food, because they only have the one site of microbial fermentation within the entire GIT. They are able to ut ilize both the simple stomach and small intestine for minimal digestion of food, but most ingested fibrous material is passed to the large intestine for fermentation. Thus, the main site of fermentation occurs beyond the main site of absorption, and hindg ut fermenters may therefore not be able to absorb as many of the microbially synthesized proteins and vitamins as the foregut fermenters [Van Soest, 1994] However, hindgut ferme nters are still able to obtain energy from their food. The main products of hindgut fermentation are volatile fatty acids (VFAs), which include the short chain fatty acids

PAGE 24

24 (SCFA): acetate, butyrate propionate, and valerate; and the branched chain fatty acids (BCFA): isobutyrate and isovalerate Lactate another product of fermen tation, is usually produced from starch, a non structural carbohydrate [Macfarlane and Macfarlane, 2003] The SCFA are absorbed across the wall of the colon directly into the bloodstream as undissociated acids, via passive diffusion, which helps to mainta in a higher pH with in the envi ronment of the cecum and colon [Van Soest, 1994; Hume, 1997] These SCFA provide a significant portion of a hindgut digestive tract are usually re absorbed in the dist al colon with VFA [Van Soest, 1994; Hume, 1997] There are two categories of mammalian hindgut fer menters the colon ic fermenters and the cecal ferme nters (Table 1 1). Most colon ic fermenters are larger animals whereas the cecal fermenters tend to be smaller in size, with the exception of the capybara ( Hydrochoerus hydrochaeris ). Body size is a principle factor in determining the capacity for fermentation. Whereas the mass specific metabolic rate for an animal decreases with body size, the gut capacity of an animal tends to remain a co nstant fraction of body weight [ Parra, 1978; Mackie 1997; Clauss and Hummel, 2005]. This relationship leads to the Jarman Bell Principle suggests that larger animals should be ab le to tolerate a lower minimum die tary quality than smaller animals [Owen Smith, 1988; Mackie, 2002] Smaller herbivores have higher metabolic rates per unit of body mass than larger animals and must ingest diets that contain a hig her concentration of nu trients [Sakaguchi, 2003] Larger animals can retain their food for a longer period of time than small animals allowing for greater breakdown of fiber Mean retention time (MRT) of digesta was proposed to scale to body mass 0.25 in herbivores, but Clauss et al. [2007] concluded

PAGE 25

25 that MRT does not always increase with body mass and that t he combination of the three factors of gut capacity, food intake, and MRT is what d etermine s digestive efficiency. The main site of fermentation for the colon ic fermenters is the proximal colon, whic h may or may not be sacculated [Van Soest, 1994] The colon may also receive support from the cecum if one is present, with the digesta mixing vig orously between the two organs [Hume, 1997; Hume, 2002] For so me of these anima ls, such as p ig s ( family: Suidae ) a nd horse s ( family: Equidae ), the cecum is rather large, but still smaller than the colon, which is usually very long and enlarged in her bivores [Van Soest, 1994] Colon ic fermenters are usually able to meet most of their protein and vitamin requirements by absorbing hexoses, amino acids, and long chain fatty acids from the breakdown of digesta in the small intestine [ H ume, 2002] Colon ic fermenters are able to obtain greater digestibilit i es of more fibrous diets than cec al fermenters by increasing digesta retention times and thus, the extent of fiber digestion [Hume, 1997] One particular group of animals, consisting of the largest terrestrial herbivores, the megaherbivores, differs from the predicted norms of dig estive s trategies. Owen Smith [1988] provided a detailed review of each of the megaherbivores and their digestive strategies. A megaherbivore is any animal that has a minimum adult body mass of 1000 kg. Species discussed by Owen Smith [1988] include the two spe cies of elephants, the five species of rhinoceroses, the hippopotamuses, and the giraffe. Of these, only elephants and rhinoceroses are hindgut fermenters. The giraffe is a true ruminant that chews its cud, whereas hippopotamuses are forestomach fermente rs with no remastication or compartmentalization of the stomach. The non compartmentalized colon. Fermentation occurs in both the cecum and colon for elephants. In terms of diet, the African elephant ( genus: Loxodonta ) can either be a grazer, a browser, or a

PAGE 26

26 combination of the two depending on where the animal is located. Based on size, mean retention time for the elephant would be predicted to be long, but these animals have a high defecation rate, which correlates to a fast turnover of their food. A rapid passage rate usually equates to a lower digestive efficiency, but if an animal can eat more food per day, they can assimilate more nutrients per unit time to compensate. Elephants spend 60 75% of their 24 hour cycle foraging with a mean daily food intake ranging from 1 1.5% of body mass, depending on the sex and reproductive status of the animal. Digesta transit time has been recorded as ranging anywhere from 21 55 hours for captive Asian elephants ( Elephas maximus ) and African elephants [ Clauss et al., 2003; Dierenfeld, 2006] Elephants utilize their large body size to store and process a great amount of food all at once to compensate for this fair ly rapid t urnover rate. D igestibi lity estimates for elephants have been determined to be in the range of 22 60% for dry matter of foods consumed bot h in captivity and in the wild [Dierenfeld, 2006] which are lower than digestibility values calculated for horses fe d a similar diet [Clauss et al., 2003] This difference probably occurs because of the faster dige sta transit time in elephants. The anatomy of the GIT of the rhinoceroses is similar to that of equids, consisting of a simple stomach, a large, sacculated cecum, and a sacculated, compartmentalized colon. The principal site of fermentation for the rhinoceroses is in the cecum, with supplementary fermentation in the colon. The Indian rhinoceros ( Rhinoceros unicornis ) is a mixed feeder t hat browse s and graze s whereas the white rhinoceros ( Ceratotherium simum ) is a strict grazer and the black ( Diceros bicornis ), Javan ( Rhinoceros sondaicus ), and Sumatran ( Dicerorhinus sumatrensis ) rhinos are all browsers. R hinos also spend a significant portion (50 60%) of their 24 hour cycle foraging, with a more recent study by Clauss et al. [2005a] identifying that captive Indian rhinos have a daily intake rate of 0.5 1.1% of their body mass. To assist in food

PAGE 27

27 gathering, the rhinos also have a prehens ile lip expansion that increases their foraging efficiency by allowing them to select less fibrous forage [Clemens and Maloiy, 1983] Average mean retention times measured for the Indian, white, and black rhinos were 60 hours, 51 hours, and 38 hours, resp ectively [Clauss et al., 2005b] Bla ck rhinos are thought to have a MRT that is shorter than the other rhinos because they are browsers, and browse undergoes a faster fermentation than grasses do [Clauss et al., 2005b] It is not surprising that the Indi an and white rhinos have longer m ean digesta retention times, because the se rhinos ingest a lower quality diet with higher cell wall contents. Clemens and Maloiy [1983] determined dry matter digestibility throughout the GIT for the black rhino to range in values from 20.3 57.4%, whereas African elephants had much lower values of 7.4 31.6%. The results of this study suggest that the black rhino is more efficient than the African elephant, but the difference in dry matter digestibility values is probably du e more to the different foods con sumed by each of these species [Clemens and Maloiy, 1983] Though they utilize different strategies, both of these megaherbivore hindgut fermenters are highly efficient at utilizing plant material. While sirenians were not included in this initial discussion of the megaherbivores, at least the Florida manatee meet s the requirements to be considered as part of the group. Manatees are unique compared to the other megaherbivores, in that they combine hindgut fermentation [ Burn 1986; Reynolds and Rommel, 1996] with an extended digesta passage time of seven days on average [Larkin et al 2007] which is longer than would be expected based on scaling relationships determined for digesta passage time and body size [Clauss et al., 2007] Ana tomically, Reynolds and Rommel [1996] found that the Florida manatee possesses a large, C shaped sac like stomach that is bisected into superior and inferior portions, a small intestine that can exceed 20 meters in length, duodenal diverticulae and an extremely enlarged large

PAGE 28

28 intestine, or hindgut region. The colon alone can measure greater than 20 m in length and 15 cm in diameter for an adult manatee, and is the main site of absorption for the short chain fatty acids. Moreover, manatees and dugongs possess a unique accessory digestive organ the cardiac gland, which is attached to the stomach [ Reynolds and Rommel, 1996]. Langer [1988] first described like structure with a thin wall and a thin tunica muscularis about 2.4 mm thick. The author further described the lumen of the gland as being filled with submucosal folds that were covered by mucosa and having a single opening into the lumen of the stomach. The opening of the car diac gland is located about 32 cm away from and cranial to the esophageal opening within the stomach. Any food that enters into the stomach from the esophagus would be pushed upward by the muscular ridge that functionally bisects the stomach. As a result the digesta would be subjected to all of the enzymes and mucous being released from the cardiac gland. Although not discussed by Langer [1988] or Reynolds and Rommel [1996] it also seems that the location of the cardiac gland allows for gravity to ensu re that the secretions coat the digesta, because they would leak downwards onto anything in the stomach, even if the digesta was not forced upwards. Since the manatee eats a fibrous plant diet, the cardiac gland is likely separated from the stomach in a d istinct protuberance to protect the parietal and chief cells from the abrasive effects of both their diet and any sand or dirt the animal might ingest while foraging. Presence of the gland was noted for three fetuses and one calf, indicating that it devel ops early in life. Reynolds and Rommel [1996] described the gland similarly to Langer [1988] as a cylindrically shaped gland protruding out of the stomach in a cranial direction, situated between the left lobe of the liver and the left hemidiaphragm. The single opening into the lumen of the stomach is immediately surrounded by muscle fibers that form a small sphincter. It was hypothesized that this allows for the

PAGE 29

29 continual secretion of acid, enzymes, and mucous into the lumen of the stomach, which is imp T he manatee digestive strategy is unlike most others, but it is interesting to consider why rumination did not evolve in manatees. Clauss et al. [2003] demonstrated that a hindg ut fermenter weighing greater than 1,200 kg and subsisting on a grass diet is just as efficient as a ruminant, so there is no reason to restrain intake or digesta passage rate through a ruminant type system. Since most adult West Indian manatees weigh ove r 1 ,000 kilograms, they would no t need to develop such a system and strategy. Additionally, ruminants are unable to regulate the passage of digesta through their compartmentalized stomach, while non ruminants can regulate th is through the large intestine [Chivers, 1989] The ability to control digesta transit time may be extremely important for manatees, which can be forced to undergo longer periods of fasting during the winter months. The long digesta passage rate exhibited by manatees may be considered a pec uliarity, however, because Van Soest [1994] predicted that increased passage times past a threshold of 4 days in herbivores would result in a larger amount of methanogenes being produced, which are bacteria that are able to convert acetic acid to met hane and carbon dioxide, causing great energy losses for the anima l. Clauss et al. [2003] hypothesized that this was the reason why elephants have such a comparatively fast digesta passage rate. With such a long passage rate, manatees could have a potent ial problem with methanogenes, but this has not yet been investigated for these animals. Evolutionarily speaking, however it seems unlikely that manatees would maintain such a long passage rate if it were causing major decreases in energy availability. Moreover, methanogenes may not have access to the acetate produced in the hindgut if the surface: volume ratio is large enough to permit a higher rate of absorptio n of the fermentation products [Clauss and Hummel, 2005] which is true of the manatee Long er

PAGE 30

30 retention times may also help maintain a larger population of microbes in the hindgut [ Stevens and Hume, 1998] which may be beneficial for manatees because they are generally solitary and would not have opportunities to ingest fecal material from cons pecifics to supplement their microbia l population if necessary Low metabolic rates for animals may also facilitate a longer digesta retention time [ Barboza, 199 3] Along with this, since manatees have a critical need to thermoregulate, having a fermenta tion vat in the hindgut that is constantly digesting and turning over food may provide an additional source of heat for them. When compared to other mammals, manatees also have a lower metabolic rate. The first study on manatee m etabolism was performed b y Scholander and Irving [1941] who determined that manatees have a basal metabolic rate slightly above 50% of predicted values determined by Kleibe law that scales basal metabolic rate ( BMR ) to body weight ( W) with an exponent of 0.75 These re searchers, however, had a very rudimentary experimental design that involved restraining the manatees and the metabolic rate determined by them has since been deemed an ov erestimate. Gallivan and Best [1980] found metabolic rate estimates for the Amazonia n manatee to be approximately 36%, or one third of predicted values and therefore, much less than the rates deter mined by Scholander and Irving [1941] More recent estimates of metabolic rate for Florida manatees by Irvine [1983] were reported as 17 22% o f expected values third of the rates deter mined by Scholander and Irving [1941] This information is important because below a certain lower critical temperature, animals must elevate their metabolic rate. These low m etabolic rate s suggest that manatee s are potentially very susceptible to cold, with a lowe r critical temperature around 20 o C (G.A.J Worthy, pers.comm.).

PAGE 31

31 Manatees occupy tropic al and subtropic al habitats [Reep and Bonde, 2006] and Florida is considered to b e at the northern edge of their range. Adult manatees are generally able to cope with falling temperatures during the winter months by seeking out warm water refuges, but juveniles and calves tend to be more susceptible to cold water because of lower body fat mass [Ortiz and Worthy, 2004] l ack of experience and an apparent inability to increase their metabolic rates in response to cold exposure (G.A.J Worthy and T.A.M. Worthy, unpubl. data) J uveniles and calves are often rescued and brought into rehabi litation facilities for treatment of cold stress syndrome. Cold stress syndrome (CSS) is associated with a reduction in food intake leading to emaciation, fat store depletion, constipation and dehydration [Bossart et al., 2003] Providing adequate nutriti on is therefore essential for successful rehabilitation, but little is known about manatee nutrition and the role that it may play in their overall health, especially in cold stresse d manatees and orphaned calves. Recovery from cold stress, other illnesses and injury could potentially be further complicated because the diet of captive rehabilitating manatees differs from that of f ree ranging manatees Concentrations of insoluble carbohydrates found in Thalassia testudinum Syringodium filiforme and Halod ule wrightii three species of sea grasses commonly consumed by manatees, ranged from 34 41%, 32 42%, and 34 46%, respectively, and were most different between the winter and spring seasons [Dawes and Lawrence, 1980] In a more recent study, it was determi ned that romaine lettuce contains much less soluble (neutral detergent fiber) and insoluble (acid detergent fiber and lignin) fiber when compared to four species of sea grasses commonly con sumed by free ranging manatees [Siegal Willott et al., 2010] Romai ne lettuce is also higher in crude fat, crude protein, carbohydrates, and digestible energy on a dry matter basis than these sea grasses [ Siegal Willott et al., 2010] Nevertheless, the amount of fiber in the diet

PAGE 32

32 of rehabilitating captive manatees is unkn own because the amount of each dietary component offered and amount ingested has not been documented. Suggested guidelines for feeding of manatees stipulate that their diet be composed of 70 85% leafy green vegetables, 10 20% dried forage, and 5% other ve getables and fruits [Walsh and Bossart, 1999] but how closely each facility conforms to these guidelines is unknown. The relative amount of fiber, fat, and protein consumed can only be determined by documenting how much of each dietary component is ingest ed by each animal. The digestibility of a diet is a measure of how well an animal is able to break down and absorb the nutrients that are available to them for energy and growth. Digestibility can be measured as true, which accounts for any endogenous los ses from the feces, or as apparent Apparent digestibility, which measures nutrients in the food and feces as a means to determine efficiency, is common in nutritional research. Total collection of food and feces is the preferred method for determination of apparent digestibility, but is not appropriate for an animal such as the manatee because it would involve sequestering each of the individuals in separate tanks and no facility is equipped with enough space to employ this method Alternatively, a mark er can be used to allow food and feces with digestibility determined using the ratio of the marker in the food and feces Types of markers include internal markers, which are already present in the food and feces, and externa l markers, which have to be added to the food. All markers should be completely indigestible and fully recovered in the feces. I nternal markers include silica, acid insoluble ash (AIA), indigestib le NDF, lignin, and chromogens. E xternal markers include metal oxides, such as iron oxide, chromic oxide, and titanium oxide, and rare earth metals such as ytterbium oxide, dysprosium chloride, and lanthanum oxide. Another form of external marker is chromium

PAGE 33

33 mordanted fiber, in which the chromium is chemically bound to the fiber source that is to be fed to the animal. No single marker is perfect, and each one has flaws, as well as advantages. Lignin is the most commonly used internal marker and chromic oxide is the most commonly used external marker. As determ ined in previous research, t he manatee has a unique physiology and biology and their digestive physiology proves to be no different. Their distinct adaptations are predicted to allow manatees to acquire a proper amount of energy fr om a low nutrient natur al diet and to continue to thrive as aquatic herbivores. While food resources for the manatee are not currently limited, habitat destruction is becoming a threat for manatees If food availability should decrease, it may be necessary for manatees t o alte r their behavior or evolve a different digestive strategy Understanding the basic nutrition of this species is therefore essential.

PAGE 34

34 Table 1 1. The two categories of extant mammalian hindgut fermenters. Colon ic Fermenters Cecal fermenters Artiodactyla P igs and boars ( Suidae ) Perissodactyla Horses ( Equidae ) Tapirs ( Tapiridae ) Rhinoceroses ( Rhinocerotidae ) Proboscidea Elephants ( Elephantidae ) Diprotodontia Wombats ( Vombatidae ) Sirenia Dugongs ( Dugongidae ) Manatees ( Trichechidae ) Primata Bab oons, guenons, m acaques ( Cercopithecidae ) Gibbons ( Hylobatidae ) Great apes humans ( Hominidae ) Rodentia Squirrels ( Sciuridae ) Beavers ( Castoridae ) Gophers ( Geomyidae ) Rats and mice ( Muridae ) Guinea pigs and capybaras ( Caviidae ) Porcupines ( Erethizontida e ) Lagomorpha Rabbits and hares ( Leporidae ) Pikas ( Ochotonidae ) Hyracoidea Hyraxes ( Hyracoidea ) Various marsupial families Bandicoots ( Peramelidae ) Bilbies ( Thylacomyidae ) Ringtail possums and g reater gliders ( Pseudocheiridae ) Pygmy possums ( Burramyid ae ) Brushtail possums and c uscuses ( Phalangeridae ) Koalas ( Phascolarctos cinereus ) Primata Lemurs ( Lemuridae ) Indris and s ifakas (I ndriidae ) Loris and p ottos ( Lorisidae ) Bush babies ( Galagidae ) Howler monkeys ( Atelidae ) Capuchins ( Cebidae ) Marmosets and t amarins ( Callitrichidae ) Adapted from [ Hume, 2002 ]

PAGE 35

35 CHAPTER 2 QUANTITATIVE SURVEY OF THE DIET OF FLORI DA MANATEES ( TRICHECHUS MANATUS LATIROSTRIS ) AT CAPTIVE FACILIT IES IN FLORIDA Background The Florida manatee ( Trichechus manatus latirostris ), a mem ber of the O rder Sirenia, is an endangered marine mammal protected by both state and federal acts including the Endangered Species Act of 1973, the Marine Mammal Protection Act of 1972, and the Florida Manatee Sanctuary Act of 1978. Free ranging manatees that become injured or ill are rehabilitated in various facilities in Florida as part of the Manate e Rehabilitation Program (MRP). S irenians are the only herbivorous marine mammals and have a unique digestive strategy that combines hindgut fermentation [B urn, 1986; Reynolds and Rommel, 1996] with an extended digesta transit time of seven days on average [Larkin et al., 2007] to efficiently utilize the plant fiber portion of their diet. Free ranging manatees generally consume a low quality diet of salt and fresh water grasses that are hi gh in fiber and low in protein [Reep and Bonde, 2006], but is appropriate for a fiber fermenting herbivore. In captivity, manatees usually are fed a diet composed mo stly of romaine lettuce sometimes supplemented with monkey biscuits, apples, carrots, and other vegetables. Siegal Willott et al. [2010] found that submerged aquatic vegetation (SAV) and sea grasses commonly consumed by manatees contain ed more ash, lignin, and structural carbohydrates including, neutral detergent fiber (NDF) and acid detergent fiber (ADF) than romaine lettuce Romaine lettuce contains more crude fat (CF) crude protein (CP) non structural carbohydrates (NSC) and digestible energy (DE) on a dry matter basis [Siegal Willott et al., 2010]. Manate es probably obtain most of their energy from the fermentation of structural carbohydrates, also known as plant fiber or the plant cell wall components to short chain fatty potent ially be come compromised if the diet offered in captivity does not contain enough of this plant fiber.

PAGE 36

36 Guidelines for feeding manatees specify that their diet should be composed of 70 85% leafy green vegetables, 10 20% dried forage, and 5% other veget ables and fruits [Walsh and Bossart, 1999]. Neve rtheless, the amount of each ingredient and nutrient particularly plant fiber, in the diet of captive manatees is unknown, because food intake of captive manatees has not been quantified yet It is also un known how much energy ( kilocalories (kcal) /day) manatees normally consume to m aintain themselves. The resting energy needs (BMR basal metabolic rate) of mammals can 1) relating BMR to body weight ( W) [Kleiber, 1975]. BM R (kcal/day) = 70 x ( W ) 0.75 (2 1) Adult daily requirements can be estimated as two times the BMR, whereas energy needs for growing animals can be estimated as three times the BMR [Popovich and Dier enfeld, 1997]. Basal metabolic rates of manatees have been reported to be 25 30% of the predicted values of other mammals [Gallivan and Best, 198 0; Irvine, 1983; Worthy, 2001], but maintenance energy intakes have been determined for only two captive dug ongs [Goto et al., 2008] and have not yet been documented for manatees. Some knowledge of the DE requirements is important if manatees are to be maintained at suitable weights and body conditions in captivity. Thus, t o investigat e whether changes to the captivity based diet should be made and to ascertain the energy needs of manatees in captivity, a quantitative survey of the captive diet of manatees in facilities in Florida was conducted to determine how much plant fiber, protein, ash, fat, non structura l carbohydrates, and digestible energy captive manatees were actually consuming. It was predicted that there would be sex, location, and age class differences for nutrient and

PAGE 37

37 digestible energy consumption. Moreover, it was also hypothesized that facilit ies were feeding diets that differed with respect to both nutrient and ingredient composition. Materials and Methods The following work was conducted with approval from the University of Florida Institutiona l Animal Care and Use Committee ( IACUC ) (#2009027 62) and under a permit obtained from US Fish and Wildlife Service (#MA038448 3) Manatees All manatees in the present study were considered to be stable and healthy at the time of intake documentation. I ntake of offered food was documented for fourteen ma natees nine males and five females, maintained at the Living Seas at Epcot (n = 6 males, 5 females ) Disney W orld, Orlando, Mote Marine Laboratory (n = 2 males ) Sarasota, and Parker Marine Aquarium (n = 1 male ) Bradenton, all in Florida. These are the only facilities that currently actively monitor food intake of individuals or that were able to do so for the duration of the study. Each facility was assigned a letter for identification purposes. Historical records of intake were used for the five fe males and two males, and current intake was monitored for the other seven males. Reported body weights ( W) are those measured most recently before food intake was documented and were usually obtained within 1 2 months of the intake measurement period. Un fortunately W measured before and after the intake period were not always available for each individual to determine whether or not the manatee was growing. Life stage was determined using length age class parameters established by United States Geolog (USGS) Sirenia Project (pers. comm.) and/or length of time in captivity if known. Manatees with a straight length of greater than 266 cm were considered to be adults, whereas manatees in the length range of 236 265 cm were considered suba dults and to be growing. All manatees had been in captivity for at least two years and were considered healthy at the time of data

PAGE 38

38 collection. Intake measurements were attempted for each animal during both winter and spring seasons. Nutrient Intake Calcu lations A representative sample of each dietary ingredient offered to the manatees at each facility was weighed. An intern or staff member, responsible for monitoring the manatees recorded intake of r omaine lettuce by each animal. Heads of lettuce were offered intermittently left. Number of heads of lettuce consumed were counted and then multiplied by the weight of a single head of lettuce to determine tota l weight of lettuce ingested. Unconsumed pieces of lettuce found in the filter or floating in the pool after each meal were minute in size and quantity and were considered insignificant compared to the amount ingested. Intake of supplementary diet items, including other vegetables and fruits, was recorded when they were offered as part of husbandry practices. Each was consumed completely. Intake of each dietary ingredient over three weeks was averaged for each manatee to obtain mean daily intake of each ingredient Percent of each dietary ingredient consumed was calculated as the average intake of each ingr e dient divided by the total consumed amount, on an as fed weight basis. These percentages of individual ingredients were then averaged for all of the manatees at a particular facility to obtain an overall diet composition for each location. The amount of each nutrien t CF CP ash, NDF and ADF as well as dry matter (DM), of each dietary component was ascertained from Schmidt et al. [2005b]. Percen t of hemicellulose was determined as the difference between %NDF and %ADF and percent non structural carbohydrates (NSC) was calculated using Equation 2 2. NSC = 100 %NDF %CF %Ash %CP (2 2)

PAGE 39

39 The nutrient composition of beets was not reported by Schm idt et al. [2005b], so percentages of CF, CP, and ash reported in the USDA National Nutrient Database [USDA, 2011] were used Neutral detergent fiber and ADF values fo r beets were determined using a f iber analyzer ( ANKOM200, ANKOM Technology Corp., Fairpo rt, NY) Digestible energy (DE), on a dry matter basis, was calculated for each animal using Equation 2 3 developed for horse feeds [Pagan, 1998]. DE (kcal/kg DM) = 2,118 + 12.18 (%CP) 9.37 (%ADF) 3.83 (%hemicellulose) + 47.18 (%CF) + 20.35 (%NSC) 26.3 (%ash) (2 3) Digestible energy intake (DEi) was calculated as the product of DE and DM intake per day (kg). Percent of each nutrient consumed was averaged across all fourteen manatees from the three different facilities for comparison to published me an values for submerged aquatic vegetation (SAV) from Siegal Willott et al. [2010] to indicate how the captivity based diet in the present study differed from a natural diet. Statistics Results are reported as means +/ one standard deviation. Data were a nalyzed using SAS (SAS for Windows Version 9.3, SAS Institute Inc., Cary, NC ) and tested for normality and equality of variances. Any data that were not normally distributed were log transformed and re assessed for normality before analysis. Absolute individual nutrient, DM, and DE intake, as well as DE intake per W were analyzed using a general linear mixed model procedure with sex, life stage, and seasons as factors that allowed for the variances of the two sexes to be different and adjusted for re peated seasonal measures. The use of a mixed model procedure also allowed for missing data points (i.e. if an individual only had intake documented for one of the seasons). Interactions between life stage and season and between sex and season were also c onsidered and

PAGE 40

40 when a significant interaction was found, Tukey The log of DEi was regressed against the log of body weight for adults and subadults to determine if there was a correlation between digestible energy intake and body weight [Evans and Miller, 1968; Allen et al., 1995]. A probability of type 1 error > 0.05 was considered significant. Results Of the fourteen manatees in the present study, five individuals were subadults ranging in body weight from 220 276 kg and nine were adults ranging in body weight from 450 786 kg Food intake was not documented during both spring and winter seasons for some individuals because these manatees were either released from captivity back into the wild or t ransferred to a different facility (Table 2 1). Romaine lettuce was the major dietary ingredient (> 85%) of each tly throughout the day, but comprised the greater proportion of the diet (> 99%) at facility C. Manate es generally ate the entire head of lettuce, including the stalk. Supplemental fruits and vegetables offered to the manatees included carrots, apples, sweet potatoes, bananas, beets, kale, cabbage, and broccoli (Table 2 2), but specific ingredients varied among the facilities. A detailed nutrient composition of monkey biscuits (Zupreem Premium Nutritional Products, Inc. Shawnee, KS ) was not available from the manufacturer or previous studies and unfortunately, could not be completed in the present stu dy because only very little amounts of this ingredient were collected. However, m onkey biscuits comprised only 0.2% of the diet at facility B and were found to contain 57 59% DM in a separate study ( L. Harshaw unpubl. data, Chapter 3). Thus, these monke y biscuits would have only contributed a maximum of 0.6% to the total DM intake for the manatees at facility B and were considered an insignificant portion of the diet in the present study.

PAGE 41

41 Concentration of each nutrient in the diet of captive manatees fe d at all three facilities was fairly consistent with the ranges as follows: 15.8 18.1% CP, 6.3 8.7% CF, 15.9 16.5% NDF, 12.8 14.1% ADF, 2.2 3.2% hemicellulose, 8.2 9.0% ash and 48.5 52.0% NSC, all on a DM basis (Figure 2 2 ). Percent of each nutrient cons umed by captive manatees appeared to be different compared to published mean values for SAV and sea grasses from Siegal Willott et al. [2010] (Figure 2 1) (Mean percent ADF for SAV was not provided, and thus was estimated to be the midpoint of the range g iven). The diet of captive manatees surveyed here contained higher mean percentages of CF CP and NSC than SAV, but less NDF, ADF, and ash than have been reported in SAV (Figure 2 1 ). Manatees in the present study consumed 3 20% of their body weight o n an as fed basis, with an average of 7%. Total daily intake of individual nutrients, DM, and DE were not different for adults and subadults (p > 0.1 for all). Females ingested significantly less grams of DM ( 1447 g/d vs. 2345 g/d, p = 0.003), CF ( 123 g/ d vs. 183 g/d, p = 0.003), CP ( 241 g/d vs. 408 g/d, p = 0.001), ash (125 g/d vs. 207 g/d, p < 0.001), NDF ( 233 g/d vs. 382 g/d, p < 0.001), ADF ( 196 g/d vs. 323 g/d, p < 0.001), NSC ( 726 g/d vs. 1165 g/d, p = 0.0002), and DE ( 4890 g/d vs. 7832 g/d, p = 0.0 04) on a daily basis than males (Table 2 3), but there was no evidence for a difference between winter and spring for either sex or life stage (p > 0.1 for all). No interactions were found to be significant among the individual nutrients, DM, and DE intak e. Digestible energy intake relative to body weight was different between life stages ( adults = 25 kcal/kg/day, subadults = 12 kcal/kg/day, p < 0.001) and sexes ( females = 9 kcal/ kg/ day, males = 14 kcal/ kg/ day, p = 0.001). There was an interaction found between life stage and season for DEi relative to body weight (p = 0.007), and Tukey significantly more DE relative to their body weights in the spring which is not surprising

PAGE 42

42 because these individuals were younger in the spring seasons of their documented intake There was no evidence for a relationship between DE intake and body weight in adults for either the spring (p = 0.91, R 2 = 0.0) or winter (p = 0.95, R 2 = 0.0) (Figure 2 4). There was also no evide nce for a relationship between DE intake and body weight in subadults for either the spring (p = 0.91, R 2 = 0.0) or winter (p = 0.99, R 2 = 0.0) (Figure 2 5). Discussion E ach of the facilities in the present survey fed sim ilar diets, but Facility A and B p rovided a greater amount of supplemental fruits and vegetables (Table 2 2). None of the facilities surveyed were feeding any type of dried forage, even though it is recommended that the diet should contain 10 20% of this ingredient [Walsh and Bossart, 199 9]. Overall, t he percent age of each nutrient in the diet appeared relatively similar across all three facilities, even though facility C fed greater than 10% more lettuce than the other two (Figure 2 2) How e ver, the proportion of NSC consumed by manatee s at facility B appeared slightly higher and may be attributed to beets and kale having greater concentrations of NSC In the present study, the amount and proportion of individual ingredients did not matter provided the nutrient intake met the energetic needs of the manatee Mean DEi of the captive manatee diet was similar for the adult (6745 + 1819 kcal) and subadult (7148 + 2171 kca l) manatees in this study. These energy intake values are equivalent to the resting energy expenditure predicted for terr estrial animals of similar body weight (Equation 2 1). However, manatees have a much lower basal metabolic rate than most other mammals, which is probably due to their herbivory and slow lifestyle [McNab, 1980]. Using oxygen consumption rates, Gallivan a nd Best [1980] related metabolic rate to W in Amazonian manatees ( Trichechus inunguis ) (Equation 2 predicted values.

PAGE 43

43 BMR = 30* W 0.68 (2 4) Irvine [1983] measured oxygen consumption in Florida manat ees and found measured values to be 15 22% of predicted values. Therefore, assuming that Amazonian and Florida manatees have similar energetic requirements; basal metabolic needs of adults calculated using would range from 1,900 2,800 kcal per day, whereas basal metabolic needs of the subadults would range from 1,200 1,700 kcal per day. Assuming adult maintenance is 2 x BMR, the adults would need to consume 3,800 5,600 kcal per day. The subadult manatees here would requir e 3,600 5,100 kcal per day, assuming growing requirements are 3 x BMR. Both age classes therefore had mean DE intakes that were higher than their expected needs relative to resting metabolic rate but obtaining a diet history and measuring actual intake tend to result in more accurate estimates of energy needs when an animal is maintaining weight which was found to be true for companion animals [Hill, 2006] Moreover, calculated needs are based on metabolizable energy, which also accounts for urinary a nd gaseous losses and thus, is less than digestible energy The results of the present study suggest that manatees may require more energy than previously thought based on BMR restimated and equation Digestible energy estimations in this study were based on an equation developed for horses. Horses are a fellow hindgut fermenter and can serve as a model for energy calculations in exotic species [Clauss et al., 2009], but are different from manatees in some aspects of their digestive strategy. Captive manatees have higher digestibility values ( L. Harshaw unpubl. data, Chapter 3) and longer passage rates than horses, so the equation used here may actually underestimate the DE of the captiv e manatee diet. An equation has not been developed yet for

PAGE 44

44 manatees or other exotic hindgut fermenters, because gross energy and digestibility of a ll nutrients have not been measured in manatees Development of a DE equation for manatees would increase the accuracy of any future assessments of the energy needs of manatees and might help t o prevent obesity in healthy, long term individuals B ody con dition index should be calculated regularly for captive manatees and intake should be adjusted accordingly if manatees are found to have condition constant values that are outside the normal range ( L. Harshaw unpubl. data, Chapter 6). Activity level, rep roductive state, thermoregulation, and growth all contribute to an weight than adults. Subadult manatees are non reproductive individuals that are completely in dependent from their mothers, but still growing. The mean DEi per body weight of subadults (25.2 + 10.7 kcal/kg/day) was twice the mean of the adults (11.7 + 4.3 kcal/kg/day), which suggests that they are probably allocating a greater proportion of dietar y energy for growth. Mean DEi per body weight values for captive gorillas also ingesting a produce based diet ranged from 24 67 kcal/kg/day, with a juvenile gorilla that was confirmed to be growing having the highest values [Remis and Dierenfeld, 2004]. The other components of the energy budget would probably not explain the difference in energy intake relative to body weight, because both adult and subadult manatees in this study were mainta ined in the same facilities, would have had similar activity le vels, were all reproductively inactive, and lived in environments with relatively constant water temperatures. Moreover, manatees do not have to use energy to resist the force of gravity, because the water supports their weight. Two adult captive dugongs including one male and one female, that were fed a diet of eelgrass had mean digestible energy intake per unit body weight of 20 kcal/kg/day and 30 kcal/kg/day, respectively [Goto et al., 2008 ], similar to subadult

PAGE 45

45 manatees in the present study. A dult d ugongs are smaller than adult manatees and weigh about the sam e as the subadults Digestible energy intake scales allometrically to body weight in adult mammals because of the energy needed to maintain cells and tissues as they relate to increasing organ size [Evans and Miller, 1968; Allen et al., 1995; Stahl et al., 2011], but this relationship was not found here for either adult (Figure 2 4), or subadult manatees (Figure 2 5). DE intake may not increase with increasing body weight for the manatees in th e present study because they have lower energy budgets in captivity than they would in the wild. Aside from the reasons mentioned before, captive manatees do not have to expend any energy actively searching, or foraging, for food. The positive associati on between digestible energy intake and body weight also may not have been found here, because these manatees are communally housed and thus somewhat limited in how much food they can consume. It is most likely however t hat the scaling relationship was not be evident in the present study because of the small range in body size of manatees and large individual variability. M ales consumed more nutrients and had higher total DM and DE intakes than females in this study, with the exception of male manatee nu mber 4 consuming less in the spring than females Information about the food intake of females in this study came from historical records, however, so it is unclear whether these females were offered less food than current manatees were or how carefully o bservations of consumption were made. Goto et al. [2008] found that a female dugong had higher average DM and DE intakes of eelgrass than a male dugong over the course of a year, but potential reasons for this were not discussed. The lower intake of fema les in this study may also be related to their estrous cycle, even though they were not actively reproducing. A captive female Indian rhinoceros had lower food intake when she came into heat

PAGE 46

46 [Clauss et al., 2005 a ], whereas reindeer voluntarily consumed le ss food with the onset of sexual maturation [McEwan and Whitehead, 1970]. The hormonal status was unknown for the female manatees in this study, but the females here were either subadults or small adults and were very close to the period of sexual maturat ion [Marmontel, 1995]. All of the manatees in the present study had similar intakes of nutrient s DM, and DE in the spring and winter seasons but this is not surprising as the housing provided for captive manatees offered protection against any seasonal changes in temperature Protein, fat, and NSC are generally easily digestible and when consumed in excess quantities, could contribute to obesity and other nutritional disorders. Sugars and starches are metabolized to glucose and require higher insulin c oncentrations in the body to move the glucose into cells [Schmidt et al., 2001]. High insulin concentrations have been linked to diabetes and metabolic disorders in other species, including humans. Glucose utilization is also a more efficient means of ob taining energy than fiber fermentation, but this higher efficiency can lead to excess weight gain [Schmidt et al., 2001]. Despite the differences in the ingredients of the diets at each of the facilities (Table 2 2), individual manatees appeared to consum e similar amounts of the nutrients. Herbivores, including hindgut fermenters, require adequate amounts of plant fiber [Schmidt et al., 2001], so it is assumed that manatees do as well. Both NDF and ADF were consumed in lower proportions than NSC or CP (F igure 2 2) by the manatees in this study. Low plant fiber diets have been linked to ulcerative colitis in captive apes [Scott and Keymer, 1975], which also typically receive fruits and vegetables as the main components of their diet [Popovich and Dierenfe ld, 1997]. Captive tap irs with high DE intakes of low plant fiber, low roughage diets were found to tend toward obesity [Clauss et al., 2009]. These latter authors further suggested that adding more roughage and/or browse to the captive diet could help p revent

PAGE 47

47 digestive disease, such as colitis. Colitis has been found in deceased free ranging and rehabilitating captive manatees, but the direct cause for it is unknown [M. DeWit, pers. comm.]. Leafy greens, other than romaine lettuce, could be investigated as an additional source of plant fiber for manatees. Kale, for instance, contains approximately 2% more NDF than romaine [Schmidt et al., 2005b]. More traditional forms of roughage for zoo animals include hays and dried grasses. A facility not included in the present study has fed alfalfa hay in the past, but caregivers were concerned about the possibility of colic due to the hay being too fibrous (R. Bonde, pers. comm.). A standardized diet across all captive manatee fa cilities may not be feasible, be cause each facility has different husbandry needs and protocols, but effort should be made towards providing adequate percentages of dietary nutrients. Free ranging manatees consume a variety of grasses and plant material that are not easy to harvest or o btain for feeding to captive manatees The manatee rehabilitation program is successful in returning a large percentage of sick manatees back to their natural habitat but the diet manatees are fed in captivity is very different from the free ranging diet in terms of nutrient and energy content. Captive manatees are usually communally housed in pools containing two or more manatees so monitoring individual intake is time consuming, but should still be done occasionally because m ore dominant manatees may over consume, while less dominant individuals may not get enough to eat in communal housing situations. The results of this study indicated that captive manatee facilities are feeding diets with dissimilar ingredients but all are higher in NSC and CP tha n the natural diet Further research should be done to confirm whether or not this is true for other captive manatees and effort should be made to partner intake data with more frequent body weight measurements and assessment of body condition. However, information presented here includes the first accu rate description of the diet offered in captivity

PAGE 48

48 and digestible energy intake of Florida manatees, and therefore adds to the knowledge of manatee nutrition.

PAGE 49

49 Table 2 1. L ocation, sex, seasons when food i ntake was documented life stage, and body weight of manatees for which food intake was documented. Animal ID Facility Sex Season Body Weight (kg) Life Stage 1 A M Spring and Winter 562 Adult 2 B M Spring and Winter 514 Adult 3 B M Spring and Winter 78 6 Adult 4 C M Spring and Winter 646, 626 Adult 5 C M Winter 450 Adult 6 C M Spring 590 Adult 7 C F Spring and Winter 590, 540 Adult 8 C F Spring and Winter 641 Adult 9 C F Winter 455 Adult 10 C M Spring and Winter 275 356 Subadult 11 C M Spring an d Winter 220 376 Subadult 12 C M Spring and Winter 240, 313 Subadult 13 C F Winter 375 Subadult 14 C F Spring 239 Subadult Two body weights are listed for manatees when they were documented for both seasons.

PAGE 50

50 Table 2 2. Average percent of each dieta ry ingredient offered (as fed weight basis) by each facility. Facility A Facility B Facility C Historical Facility C Present Romaine Lettuce 84.3% 85.8% 98.2% 99.5% Apples 0.6% 2.7% 0.6% 0.2% Bananas 0.1% Beets 4.3% Broccoli 2.5% Cabbage 5.1% Carrots 1.6% 1.1% 0.7% 0.1% Kale 4.5% 5.8% Monkey Biscuits 0.2% Sweet Potatoes 1.5% 0.3% 0.2% (Zupreem Premium Nutritional Products, Inc. Shawnee, KS )

PAGE 51

51 Table 2 3. Mean daily DM, CF, CP, Ash, NDF, ADF, NSC cons umption DE intake, and DE intake/Body Weight for males and females of combined life stages Males Females P value Dry Matter (g) 2345 + 104 1447 + 147 0.003 Crude Fat (g) 183 + 12 123 + 17 0.0 03 Crude Protein (g) 408 + 19 241 + 26 0.001 Ash (g) 207 + 10 125 + 14 < 0.001 Neutral Detergent Fiber (g) 382 + 17 233 + 24 < 0.001 Acid Detergent Fiber (g) 323 + 15 196 + 21 < 0.001 Non structural Carbohydrates (g) 1165 + 52 726 + 73 0.00 2 Digestible E nergy Intake (kcal) 7832 + 348 4890 + 492 0.00 4 DE intak e/Body Weight (kcal/kg) 14 + 2 9 + 2 0.001 Results are presented as means +/ one standard error. DM = dry matter, CF = crude fat, CP = crude protein, NDF = neutral detergent fiber, ADF = acid detergent fiber, NSC = non structural carbohydrates, DE = dige stible energy.

PAGE 52

52 Figure 2 1. Percent of each nutrient on a DM ba sis for captivity based diet surveyed ( averaged across the fourteen individuals from the three facilities in this study ) compared to values reported previously for submerged aquatic vegetat ion (SAV) by Siegal Willott et al. [2010]. The captive manatee diet appears to contain less neutral detergent (NDF) and acid detergent fiber (ADF) and ash than the SAV, but more fat (CF), protein (CP), and non structural carbohydrates (NSC). Columns rep resent means. 0.0 10.0 20.0 30.0 40.0 50.0 60.0 %NDF %ADF %CP %CF %NSC %Ash SAV Captive Diet

PAGE 53

53 Figure 2 2. Mean nutrient concentration per day, as a percent of total dry matter intake for each manatee studied during winter and spring. Manatees consumed greater proportions of NSC and CP than other individual nutrients measu red 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Winter Spring Winter Spring Winter Spring Winter Spring Winter Spring Winter Spring Winter Spring Winter Winter Spring Winter Spring Winter Spring Winter Spring 1 1 2 2 3 3 4 4 5 6 7 7 8 8 9 10 10 11 11 12 12 13 14 Nutrient Intake (as % of total) Animal ID and Season NSC ADF NDF Ash CP CF

PAGE 54

54 Figure 2 3. D ry matter intake (grams per day) for each manatee studied during winter and spring. Adults ( individuals 1 9) and subadults ( individuals 10 14) have similar dry matter intakes, but males (blue bars) have higher total intakes than f emales (red bars). Columns represent daily means of intake documented over a three week period 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Winter Spring Winter Spring Winter Spring Winter Spring Winter Spring Winter Spring Winter Spring Winter Winter Spring Winter Spring Winter Spring Winter Spring 1 1 2 2 3 3 4 4 5 6 7 7 8 8 9 10 10 11 11 12 12 13 14 Dry Matter Intake (kg/day) Animal and Season

PAGE 55

55 Figure 2 4. Linear regre ssion of log of body weight vs. log of daily digestible energy intake of adult manatees. Open circles represent winter data and cl osed c ircles represent spring data

PAGE 56

56 Figure 2 5. Linear regre ssion of log of body weight vs log of digestible energy intake for subadult manatees. Open circles represent winter data and closed circles represent spring data.

PAGE 57

57 CHAPTER 3 IN VIVO APP ARENT DIGESTIBILITY TRIALS OF CAPTIVE FL ORIDA MANATEES ( TRICHECHUS MANATUS L ATIROSTRIS ) Background Florida manatees ( Trichechus manatus latirostris ) are an endangered species that are subject to both anthropogenic and natural causes of disease, injury, and mortality, despite being protected by the Endangered Species Act of 1973, the Marine Mammal Protection Act of 1972, and the Florida Manatee Sanctuary Act of 1978. When a free ranging animal becomes injured or ill, they are often rescued and transferred t o a rehabilitation center. Rehabilitating manatees a re usually maintained on a diet comprised primarily of romaine lettuce. Romaine lettuce, however, contains significantly less ash and structural carbohydrates than the grasses manatees would typically c onsume in the wild [Siegal Willott et al., 2010]. Manatees are herbivores that utilize hindgut fermentation to break down the structural carbohydrates, or plant cell wall components also known as plant fiber This fiber cannot be degraded by animal enzym es, and includ es he micellulose, cellulose, and lignin, a phenolic compound. The fiber content of food can be determined using either a proximate analysis or a Van Soest analysis. In proximate analysis, crude fiber ( hydroxide insoluble lignin and cellulos e) is measured and nitrogen free extract (hydroxide soluble lignin, hemicelluloses, pectins and other digestible carbohydrates ) is determined by difference Crude fiber, therefore, does not separately measure all of the fiber types in the diet [ Van Soest 1978]. In the Van Soest analysis, neutral detergent and acid detergent solutions are used to separate the fiber portion of the feed into neutral detergent solubles (NDS), neutral detergent fiber (NDF), acid detergent solubles (ADS), and acid detergent f iber (ADF). In this way, NDF contains all of the cell wall components and ADF contains the cellulose and lignin.

PAGE 58

58 The digestibility of a diet is a measure of how well an animal is able to break down and absorb the nutrients that are available to them for energy and growth It is measured by comparing food intake to fecal excretion of nutrients. To avoid the need to collect all feces excreted over several days, apparent digestibility is often measured using a marker. Internal markers, such as a cid insolu ble ash (AIA) or acid detergent lignin (ADL) have the advantage in endangered and exotic species, in that they do not require a change in diet, which would interfere Acid insoluble ash has yielded similar estimates o f apparent digestibility when compared to the total collection method in other hindgut fermenters, such as the horse and elephant [Miraglia et al., 1999; Bergero et al., 2004; Pendlebury et al., 2005]. Lignin can be measured either using the Klason or the acid detergent assay, with the latter being the preferred method for forages [Jung et al., 1999]. Both yield similar predictions of digestibility [Jung et al., 1997], but Klason lignin can become contaminated with the protein of forage material [Jung et al., 1999]. Acid detergent lignin has also been used to successfully determine digestibility with elephants [Clauss et al., 2003]. Understanding how manatees process their diet is vital to their manage ment and care. T here is little information about the digestibility of foods consumed by free ranging or captive Florida manatees, but apparent digestibility of eelgrass ( Zostera marina ) was measured in two healthy, captive dugongs ( Dugong dugon ) another sirenian using a total collection method in which the weight of all feces excreted was compared with the weight of all food offered, less that of all unconsumed food [Goto et al., 2004b]. Apparent digestibility values of DM [Goto et al., 2004b], organic matter, ash, crude protein (CP) crude fat (CF) crude fiber, nitrogen free extracts, and NDF were all reported [Goto et al., 2008]. Eelgrass appeared to be a suita ble diet for captive dugongs because it was representative of the natural diet, and highly d igestible [Goto

PAGE 59

59 et al., 2008], but m ost captive mana tees do not receive natural aquatic vegetation as part of their diet, so it is important to understand how well manatees digest the diet they do receive. Additionally, seasonal variation in digestibility was noted for long term, captive dugongs [Aketa et al., 2003], as well as for terrestrial herbivores based on quantity and quality of food [Aagnes et al., 1995; 1996]. Seasonal differences have also been noted in the seagrasses most commonly consumed by free ranging manatees, with nutrient levels being mo st different in December January and April May [Dawes and Lawrence, 1980]. It is important, therefore to assess seasonal variation in apparent digestibility of captive manatees consuming the diet that is normally offered to them in captivity. The studies performed thus far with manatees have focused on the digestibility of dry matter, organic matter, or energy, and have not assessed the digestibility of specific portions of the diet. Lomolino and Ewel [1984] reported the dry matter apparent digestibility of a single adult female Florida manatee to be about 83% for water hyacinth and 91% for lettuce, but the methods used were not clearly explained. Using permanganate solubilized lignin as a marker, and assuming that the stomach and rectal contents of mana tee carcasses were representative of the diet consumed and feces excreted respectively, Burn [1986] calculated that the apparent digestibility of org anic matter, nitrogen, CF and cellulose for West Indian manatees were 71%, 61%, 77%, and 80%, respectivel y. Permanganate lignin has been found to underestimate digestibility [Fahey and Jung, 1983], however and it is unknown how death may affect the hindgut microbes and digestion in manatees. I t is important therefore, to conduct digestibility studies with living manatees as well and explore the use of other markers. Thus, t his study sought to compare the digestibility of a captive manatee diet that included more ingredients and plant fiber to a less varied diet fed to captive Florida manatees.

PAGE 60

60 Addition ally, we measured digestibility during the spring and winter to determine if seasonal differences were present. We also hoped to identify a useful internal marker for digestibility studies with manatees. It was hypothesized that type of diet and season w ould affect apparent digestibility values. Moreover, it was also predicted that AIA would be a suitable internal marker. Materials and Methods All of the following work was performed under a permit obtained from US Fish and Wildlife Service (# MA038448 3) and with approval from University of Florida IACUC (#200902762). Manatees and Sample C ollection Food and fecal samples were collected from two groups of healthy manatees (Table 3 1): two adult male manate es born at Miami Se aquarium, which remai n in captivity at Mote Marine Laboratory in Sarasota, FL; and three recently rehabilitated male manatees receiving secondary care at the Living Seas of Disney World, in Orlando, FL. These recently rehabilitated individuals were rescued from the wild follo wing boat injuries and cold stress, and had previously received critical care at primary rehabilitation facilities. Life stage of the manatees was determined using length age class parameters established by the United States Geologi cal enia P roject (pers. comm.) or by known amount of time in captivity Both facilities maintain a water temperature of 25 26 o C and salinity of 30 34 ppt. Representative samples of the diet offered at each facility within a collection period were collected f or three weeks, and included several different dietary ingredients Facilities were given the option to collect smaller sampl es every day, or larger samples every three days, as it suited their needs and staff availability. Fo od samples were placed in ga llon sized plastic bags (Ziploc SC Johnson and Son, Inc., Racine, WI) and frozen at 8 o C or 20 o C until the end of the

PAGE 61

61 collection period. Then, food samples were transported back to the laboratory on wet ice and kept at 20 o C until analysis. Intake was also measured during these three weeks by visually observing each of the dietary ingredients being ingested. All of the foods offered to the manatees were weighed before feeding to determine offered amount. Heads of romaine lettuce were observed to be c onsumed whole by each of the manatees, and any remnants were c onsidered to be negligible in quantity Supplemental fruits and vegetables were hand fed as part of husbandry practices and were wholly consumed Fecal samples were collected over a period of five to seven days following the diet collection period, with the goal of obtaining a total of at least 100 grams of dried feces. Fecal samples were collected non invasively from the holding tank within five minutes of defecation using a pool net then p laced in either polypropylene tubes (Nunc 50 mL, Thermo Fisher Scientific, Waltham, MA) or quart sized plastic bags (Ziploc SC Johnson and Son, Inc., Racine, WI) and frozen at 8 o C or 20 o C. Fecal samples were only collected from the water column, beca use the holding tanks had sand on the bottom, which could contaminate the samples. Samples were then transported back to the lab and stored at 80 o C until analysis. Sample Processing Individual samples of all dietary ingredients and fecal samples were dried to a stable weight over seven to ten days in a laboratory oven (Model 05015 58, Cole Parmer, Niles, IL ) at 55 o C. Samples were weighed before and after drying to determine percent dry matter. Dried food and fecal samples were then individually groun d using a Wiley mill (Thomas Scientific, USA, 3383 L10 Series, Swed esboro, NJ) to pass through a 1 mm screen. S amples from each individual were then combined with an effort to include equal amounts of dried feces from each day, to form a single sample fo r each collection period. Daily samples of each dietary ingredient

PAGE 62

62 were combined to form a single sample for each facility and for each ingredient for each collection period. These composites were then secured inside sealed plastic bags (Ziploc SC John son and Son, Inc., Racine, WI) until analysis. Nutrient Analysis Individual dietary ingredients and fecal samples were analyzed for crude protein (CP) [Hansen, 1989], acid detergent fiber (ADF) and neutral detergent fiber (NDF) [Van Soe st, 1978]. O verall values for the varied diet of facility A were calculated by dividing average intake of the nutrient (i.e. CP, ADF, NDF) per day as determined by the sum of average grams per day of each nutrient in each of the individual ingredients, b y the average amount of total food intake per day. Crude protein was determined by measuring elemental nitrogen using a modified Dumas combustion method [Hansen, 1989] in an Elementar rapid N cube ( Model Vario Max CN, Elementar Americas, Inc Mt. Laurel, NJ). Protein s amples were analyzed in triplicate, and CP concentrations were determined by multiplying nitrogen values by 6. 25. Acid detergent fiber and NDF were determined using a Fiber analyzer ( ANKOM 200, ANKOM Technology Corp., Fairport, NY). Neutr al detergent fiber, ADF, and ADL were performed sequentially on all diet and fecal samples and run in duplicate. Approximately 0.5 grams of sample material was weighed into 3 x 5 cm bags made of polyester filter material ( Unibond HK 250 N Midwest Filtrat ion, Cincinnati, OH) and heat sealed. Two blank bags were included in each run, with weights before and after used to calculate blank bag correction factor (C1) and to determine if there was any particle loss from the filter bags. Filter b ags were insert ed into the twenty four slots of the bag suspender and this entire apparatus was then placed inside the fiber analyzer vessel and held down by a small weight to keep it submerged in 1900 200 0 mL of neut ral detergent solution (30.0 g s odiu m dodecyl sulfate, USP; 18.61g e thylenediaminetetraacetic

PAGE 63

63 di sodium salt, dihydrate; 6.81 g sodium borate; 4.56 g s odium phosphate d ibasic, anhydrous; and 10.0 ml t riethylene glycol, in 1 L distilled H 2 O), to which twenty grams of sodium sulfite (Na 2 SO 3 anhydrous, Fisher Sc ientific, Hampton, NH ) and four mL of heat stable alpha amylase ( 17 ,400 Liquefon Units/mL, ANKOM Technology Corp., Fairport, NY ) had been added. Samples were exposed to agitation and heat for 75 minutes, followed by three five minute rinses with 70 90 o C w ater. Four mL of alpha amylase were included in the first two rinses. The filter bags were then removed from the vessel, gently squeezed and soaked in acetone (C 3 H 6 O Certified ACS, Fisher Scientific, Hampton, NH ) for five minutes to remove any excess wa ter. After air drying to evaporate the acetone, filter bags were dried in a forced air oven (Model 0 5015 58, Cole Parmer, Niles, IL ) at 102 o C for four hours and then reweighed to determine post NDF dry weight and % NDF was calculated. The NDF extracted f ilter bags were then once again inserted into the bag suspender and submerged in 1900 2000 mL of acid detergent solution (twenty grams cetyl trimethylammonium bromide (CTAB) in 1 L 1.00N sulfuric acid H 2 SO 4 ) within the fiber analyzer vessel and exposed to agitation and heat for sixty minutes. Bags were then rinsed for three five minute cycles with 70 90 o C water, and soaked in acetone for five minutes, dried, and weighed as described above, to determine % ADF by difference. Acid insoluble ash was measured a t an analytical laboratory (Dairy One Cooperative Inc., Ithaca, NY), using the two normal hydrochloric acid (2N HCl) method [Van Keulen and Young, 1977]. Five grams of sample was weighed into porcelain crucibles and dried in an oven (Isotemp Model 501, Th ermo Fisher Scientific, Waltham, MA) at 135 o C for two hours to determine dry weight. Samples were then ashed overnight in these same crucibles at 450 o C in a muffle furnace (Thermolyne Model 30400, Thermo Fisher Scientific, Waltham, MA). The ash residue w as allowed to cool and then transferred to a 600 mL Berzelius beaker, with 2N HCl

PAGE 64

64 used to rinse the ash into the beaker up to a total volume of 100 mL. The beaker was then placed on a crude fiber apparatus and allowed to boil for five minutes. The hot so lution was then filtered through ashless filter paper (Whatman grade 41, Thermo Fisher Scientific, Waltham, MA) and rinsed three times with boiling water (85 100 o C). Any remaining ash and the filter paper were then transferred back to the crucible and ash ed overnight once again. Once samp les were cooled in a desiccator, they were weighed to obtain a final weight. Acid detergent lignin was determined using a rotary incubator ( Daisy II, ANKOM Technology Corp., Fairport, NY). The sample filter bags previo usly used for NDF and ADF determination were first dried in an oven (Model 05015 58, Cole Parmer, Niles, IL ) at 102 o C for two hours and then allowed to cool to ambient temperature in a desiccator (MoistureStop weigh pouch, model F39, ANKOM Technology Corp. Fairport, NY). Next, they were randomly sorted and placed into three of the jars in the rotary incubator with 500 ml of 72% sulfuric acid (H 2 SO 4 ) to cover the bags. Jars were allowed to rotate for three hours, after which the sulfuric acid was decanted Then bags were rinsed using distilled water until a neutral pH was reached, soaked in acetone to remove excess water, dried in a 102 o C oven (Model 05015 58, Cole Parmer, Niles, IL ) for four hours, and weighed. Percent AIA and % ADL were calculated for the entire diet by determining how many grams of AIA and ADL were present in each dietary ingredient and then dividing their sum by the total amount of food ingested all on a DM basis Apparent digestibility for each nutrient was then calculated using eq uation 3 2. Apparent Digestibility = 100 [100 x (% indicator in feed/% indicator in feces) x (% nutrient in feces/% nutrient in feed)] (3 2) Statistical Analysis Results are reported as means +/ one standard deviation. Coefficient of variation (CV) c alculated as the ratio of the standard deviation to the mean, was determined for each nutrient to

PAGE 65

65 assess whether or not values could be averaged across replicates of each dietary ingredient for a collection period and combined fecal samples Digestibility was measured during only one season for one of the manatee s, so apparent digestibility values from this individual were not included in seasonal comparisons. Statistical analyses were performed using Sigmaplot ( Version 11.2, Systat Software Inc., San Jo se, CA). Apparent digestibility values were compared between seasons using paired t tests. Digestibilities were not compared among locations or diets because sample sizes were small. A linear regression was also performed to determine if there was a cor relation between low AIA concentrations and amount of feces collected. In a post hoc analysis, % DM digestibility was regressed against DM intake to determine if apparent digestibility changed with increasing intake. A probability of type 1 error > 0.05 was considered significant. Results Two of the m anatees in the present study were subadults and three were adults at the time of the digestibility trials and weighed between 275 and 786 kg. Winter and spring samples were collected for both of the manate es at facility A and two of the manatees at facility B (Table 3 1). A winter collection could only be done for animal numb er four at facility B, because this individual was released back into its natural habitat soon afterwards. Romaine lettuce was the only type of dietary ingredient collected and analyzed for facility B, which offered a less varied diet. Other dietary items were sporadically offered at facility B as part of normal husbandry practices, but quantities offered were negligible, because the y comprised less than 1% of the total diet on an as fed basis ( L. Harshaw unpubl. data, Chapter 2) Diet samples from facility A, which offered a more varied diet, included romaine lettuce, kale,

PAGE 66

66 apples, carrots, beets, and monk ey biscuits (Zupreem Pre mium Nutritional Products, Inc. Shawnee, KS ). Mean DM intake varied across the five manatees and across seasons ranging from 1797 3925 grams per day (Table 3 1). Values obtained from the two replicates of NDF, ADF, and ADL for each of the foods and combined fecal samples were all 10% or less ranging from 0.1 10% for NDF, 0.02 10% for ADF, and 0.6 10 % for ADL. The three replicates of CP concentrations were also averaged for the foods and combined fecal samples beca were all under 4%. Percent CP, NDF, and ADF appeared to vary for each dietary ingredient but appear ed similar between seasons at each facility (Table 3 2). Fecal concentrations of each of the nutrients appeared to vary more between seasons for each of the five manatees and ranged from 18.8 24.6% CP, 22.0 33.1% NDF, and 16.0 21.9% ADF (Table 3 3). Unexpectedly, AIA concentrations were lower in the feces than in the diet of all manatees but on e. This implied that AIA may have been absorbed at some point in the digestive tract or that larger amounts of endogenous substances, such as mucous, were secreted into the digesta and decreased the concentration of AIA which precluded using AIA as a marker to measure digestibility. There was no evid ence of a correlation between AIA concentrations and amount of collected feces (p = 0.107, R 2 = 0.292) (Figure 3 1). Acid detergent lignin concentrations were consistently higher in the feces than in the diet, as expected of an undigested marker, and were therefore used to calculate digestibility (Table 3 4). There was no detectable difference between spring and winter seasons for any of the nutrient digestibility values (DM: p = 0.852, CP: p = 0.815, NDF: p = 0.952, ADF: p = 0.374) (Table 3 5). Average apparent digestibility values for facility A were 69.8%, 72.0%, 70.0%, and 70.5%, whereas average apparent digestibility values for facility B were 78.9%, 80.5%, 73.7%, and 75.6% for DM, CP,

PAGE 67

67 NDF, and ADF, respectively. Percent DM digestibility decreased w ith increasing DM intake (p = 0.001, R 2 = 0.8) (Figure 3 2). Discussion Herbivores have symbiotic relationships with microorganisms to digest the plant fiber portion of their diet, which is usually un degradable by animal enzymes. Foregut fermenters are generally considered to be more efficient than hindgut fermenters [Van Soest, 1994], but the high apparent digestibility values for DM, CP, NDF, and ADF for manatees in this study are more comparable to those of foregut fermenters, rather than other hindgu t fermenters, which suggests that manatees are capable of thoroughly digesting the diet offered to them in captivity Plant f iber concentrations in the captive manatee diet are much lower, however, than those found in the diet of other hindgut fermenters, such as the horse [National Research Council, 2007] and captive elephant [Clauss et al., 2003; Hatt and Clauss, 2006]. In general, diets with less plant fiber and more NSC are more digestible for herbivorous species Manatees have an extended digesta p assage time of 6 10 days 7 days on average [Larkin et al., 2007]. This is similar to the digesta passage time of 6 7 days for the dugong [Lanyon and Marsh, 1995], but longer than for horses or elephants, which have average digesta passage times of 40 hours [ Van Weyenberg et al., 2006] and 24 48 hours [Dierenfeld, 2008], respectively. H aving an extended passage time w ould allow for a greater amount of fermentation, a nd could thus contribute to the high digestibility values found for manatees i n this study. Dry matter digestibility for the captive manatees in this study ranged from 68 82%, which is higher than for most other hindgut fermenters. Dry matter digestibility values for horses, the domestic model of hind gut fermenters, usually averages a round 40 60%, depending on the forage being tested [Miraglia et al., 1999; Bergero et al., 2004]. Dry matter digestibility by African elephants averaged 35 38%[Pendlebury et al., 2005], whereas DM digestibility by

PAGE 68

68 Asian elephants ranged from 25 41% [Clauss et al., 2003]. Koalas had slightly lower DM digestibility values than the ones found here, ranging from 54 70% [Ullrey et al., 1981], but also have an extended digesta passage time of 99 213 hours (4 8 days) on average [Caroline et al., 2003]. Captive Gala pagos tortoises achieved DM digestibility values of 49 80% being fed a diet with amounts of ADF and ADL similar to the diet in the current study [Hatt et al., 2005]. The authors attributed slower ingesta passage rate s. Dugongs, however, had an even higher average DM digestibility of about 90% for eelgrass [Goto et al., 2008]. Dry matter digestibility values are affected greatly by the proportion of NSC in the diet, which were higher in the captive manatee diet in th e present study than in the diet of the aforementioned other species. Thus, fiber digestibility values may provide more valuable information. The fiber digestibility values determined for captive manatees in this study were almost twice the digestibility values reported previously for horses and elephants but the diet of these animals also tends to have higher concentrations of NDF and ADF Neutral detergent fiber and ADF digestibility values averaged abou t 40 5 0% for horses [Miraglia et al., 1999]. Crude fiber and ADF digestibility values for captive Asian elephants were similarly lower and ranged from 18 39% and 13 36%, respectively [Clauss et al., 2003]. Even lower plant fiber digestibility values have been linked to rapid digesta passage rate s in both Giant pandas [Sims e t al., 2007] and Black lemurs [ Schmidt et al., 2005a]. Even for the koala, NDF, ADF, and CP digestibility values were determined to be lower than the ones in the present study for manatees, ranging from 22 57%, 8 54%, and 31 49%, respectively, for a diet with similar NDF and ADF values, but lower CP values [Ullrey et al., 1981]. Manatees thus may be capable of digesting the diet offered to them in captivity more thoroughly than koalas being fed a diet with similar conce ntrations of NDF and ADF even though koalas have a ver y similar digestive strategy.

PAGE 69

69 Manatees and koalas both have very long and large hindguts relative to body size, with the koala considered to have the largest fermentation chamber proportional to body size of any herbivore [Tyndale Biscoe, 2005] As mentioned before, the sea grasses and SAV typically ingested by free ranging manatees are higher in plant fiber than romaine lettuce [Siegal Willott et al., 2010] and the other dietary ingredients fed to cap tive manatees Dugongs being fed eelgrass, which is more similar to the natural diet of manatees had average crude fiber and NDF digestibility values of 98% and 95%, respectively [Goto et al., 2008]. Manatees may therefore be capable of thoroughly d iges ting captivity based diets that are higher in plant fiber, but more research is needed to determi ne whether or not this is true. The manatees consuming the diet containing more ingredients at facility A appear ed to have lower mean apparent digestibility va lu es ( 68 75 % ) than manatees eating primarily romaine lettuce at facility B ( 72 84 % ) (Table 3 5) Th is is not surprising, however because overall NDF concentrations for the facility A diet appear slightly higher than those for facility B, and DM diges tibility for monogastric animals tends to decrease as the NDF concentration of the diet increases [Schmidt et al., 2005a]. Intake can also affect the DM digestibility of a diet composed of several ingredients [Edwards and Ullrey, 1999], in that higher int ake reduces digestibility because of a reduced residence time in the GIT It was predicted that the longer digesta passage times of manatees might compensate for greater intakes but DM digestibility decreased with increasing DM intake in the present stu dy (Figure 3 2) Adult manatees at facility A consumed more DM per kg of W than the one adult at facility B (Table 3 1), which could contribute to their lower digestibility values The two juveniles consumed more DM per kg of W than adults at facility B but this wa s expected because they were growing. Giant pandas had higher DM digestibility values when consuming a more varied diet than when they

PAGE 70

70 ate only bamboo, but the authors suggested that the digestibility o f the bamboo itself could be reduced when supplemented with other dietary ingredients [Sims et al., 2007]. Supplemental ingredients that are greater in NSC could be fermented more quickly, resulting in a pH change of the intestinal environment and potentially, microbial population changes. I t m ay be that the supplemental ingredients offered to manatees at facility A could de crease the digestibility of individual ingredients le ading to the lower digestibilities for facility A manatees and the higher digestibility values of facility B manatees e ven when NDF was higher. Higher digestibility values are not nece ssarily a good thing, however, because they can result from a diet being easily fermenta ble [Edwards and Ullrey, 1999], and thus result in changes in fecal pH and gut microbiota, or weight g ain and ultimately, obesity, if not monitored carefully. There was no evidence that manatees digested their diets differently between spring and winter seasons. These manatees were maintained in a controlled environment with consistent water temperature a nd salinity year round. V ariation in DM digestibility between seasons was significantly different in two long term captive dugongs [Aketa et al., 2003], but these dugongs were fed a diet of eelgrass, compared to the manatees in the present study being fed a produce based diet. The temporal variation in seagrasses is more pronounced than in produce, and Aketa et al. [2003] noted that lower apparent DM digestibility correlated with higher lignin content in the eelgrass for both dugongs Lignin has previous ly been identified as the cell wall component that most influences the digestibility of plant material [Van Soest 1994; Thayer et al., 1984], because lignin often chemically bonds with the cellulose and hemicellulose, creating a complex that cannot be eas ily digested [Bjorndal, 1980]. Acid detergent lignin content of the foods in this study did not appear to vary greatly (Table 3 4) between seasons, so this might help account for the lack of a seasonal difference in apparent digestibility Seasonal diffe rences in digestibility

PAGE 71

71 may also result from a shift in diet types for free ranging animals [Remis et al., 2001]. Manatees in their natural habitat may alter which grasses they ingest across seasons but in captivity, manatees are not given this choice. Percent CP, NDF, a nd ADF composition values for the dietary ingredients collected (Table 3 2) were generally similar to those previously reported for apples, carrots, and kale [Schmidt et al., 2005b]. Crud e protein concentrations of Romaine lettuce were similar to those reported by Siegal Willott et al. [2010], but about 6 9% higher than the value of 18 % from Schmidt et al. [2005b]. Plant f iber concentrations of romaine lettuce from both facilities were higher than values published by Schmidt et al. [ 2005b] and Siegal Willott et al. [2010] with NDF 3 12 % higher and ADF 1 but these differences may be explained by differences in growing locations and conditions such as fertilization, moisture, soil, an d light [Van Soest, 1978]. Composition concentrations of NDF, ADF, and lignin of foods in this study were determined using the same assay and equipment (ANKOM A200, ANKOM Technology Corp. Fairport NY) as Siegal Willott et al. [2010], so the difference i n concentrations was not due to a difference between assays. A potential source of error may have resulted from samples in this study being oven dried at 55 o C, instead of being lyophilized, as they were in the other two studies. Heat can sometimes alter the fiber and protein structures [Parissi et al., 2005], but a freeze drier was not available. Lignin values for romaine lettuce were similar to those reported by Siegal Willott et al. [2010]. Acid insoluble ash was not a reliable marker for measuring th e digestibility of nutrients by the Florida manatee, whereas ADL was found to be useful. A marker must be nearly or completely recovered in the feces for it to be suitable for measuring digestibility. Acid insoluble ash values were lower in the feces tha n in the diet. Unless there is a net secretion of DM into the

PAGE 72

72 gut, then the only explanation for this observation is that AIA is not being fully recovered in the feces of manatees (Table 3 5). The total DM amount of feces collected was less than eighty g rams for all but one animal, so it was theorized that insufficient fecal material was collected for the com bined sample to be representative [Guevara et al., 2008]. However, as shown in Figure 3 1, the highest concentrations of AIA were found in the small est samples of feces and there was no detectable linear correlation between amount of feces collected and concentrations of AIA detected in the feces. Fecal collections were only permitted during business hours, typically from 9 am to 5 pm, so diurnal and daily variation were also considered as potential factors leading to low AIA. These variations have been considered insignificant, however in poultry, sheep, cattle, pigs, and horses [Sales and Janssens, 2003]. Poor recovery of AIA in the feces has mos tly been attributed to low values of AIA in the diet causing errors in the analytical precision [Guevara et al., 2008; Sales and Janssens, 2003], which may be the case here. Most of the grasses and natural plant species that herbivores usually consume an d are typically used in AIA studies have greater concentrations of AIA than romaine lettuce, the main ingredient of the captive manatee diet in the present study Digestibility t rials with manatees consuming di ets with higher concentrations of AIA should be conducted to further investigate this. Acid insoluble ash and ADL have been used successfully with both captive African and Asian elephants, which are th s irenian relatives. The African elephants in the AIA study were maintained on a Bermuda grass hay based diet [Pendlebury et al. 2005], whereas the Asian elephants in the ADL study were fed mainly meadow hays [Clauss et al., 2003]. M anatees retain their digesta for longer periods of time than elephants, however so it may be possible that manatees are somehow physiologically using the AIA portion of the diet during transit, but a physiological use of AIA has not yet been noted in any other species.

PAGE 73

73 Hackenberger [1987] suggested that Asian elephants may retain their dige sta longer and feed more on grass than browse, so these elephants would be more similar to the manatee than their African counterparts. All of the manatees in this study were males. It is possible that digestibility is different in females, but Aketa et al. [2001], as cited in Aketa et al. [2003], found no support for differences in sea grass digestibility between a male and female dugong. There were only two dugongs in the aforementioned study, however, so it is unclear how accurate this finding may be. Moreover, it would be difficult to separate differences in location from sex based differences, because sexually mature male and female Florida manatees are not permitted to share holding tanks. The Florida manatees studied here were all capable of thorou ghly digesting a captivity based diet of romaine lettuce and supplemental fruits and vegetables The current captive manatee diet may be too digestible, however so supplemental sources of plant fiber should be explored as a means to prevent gastrointesti nal disease or obesity. Manatees consuming a more varied diet, with less romaine lettuce, appeared to have lower digestibility values, which may have result ed from there being greater concentrations of NDF and ADF in these diets No evidence of seasonal differenc es in digestibility was found, but produce items have less seasonal variation than the grasses manatees might typically consume in the wild. Future studies should thus examine the digestibility of a more natu ral diet. Acid insoluble ash proved n ot to be a reliable marker, whereas ADL was for digestibility studies with manatees.

PAGE 74

74 Table 3 1. L ife stage, body weight, measurement period, and daily dry matter intake of manatees when food intake was documented and feces were collected to assess apparen t digestibility Animal Life Stage Body Weight (kg) Facility Season Collection Month/Year DM Intake (g/day) DM Intake/ W (g/kg) 1 Adult 786 A Spring April May 2010 3925 5.0 1 786 A Winter December 2009 January 2010 3664 4.7 2 Adult 514 A Spring April Ma y 2010 3804 7.4 2 514 A Winter December 2009 January 2010 3654 7.1 3 Juvenile 275 B Spring April 2010 2320 8.4 3 356 B Winter December 2010 January 2011 2688 7.6 4 Juvenile 376 B Winter December 2009 January 2010 1951 5.2 5 Adult 626 B Spring April 2010 1797 2.9 5 646 B Winter 1 December 2009 January 2010 1816 2.8 5 655 B Winter 2 December 2010 January 2011 2431 3.7 Dry matter (DM) intake was averaged over three weeks of observation.

PAGE 75

75 Table 3 2. Crude protein (CP), neutral detergent fiber (N DF), and acid detergent fiber (ADF) composition of foods and complete diet fed to manatees at each facility. Diet Component Facility Season % CP % NDF % ADF Romaine Lettuce A Spring 23.9 + 0 .2 28.2 + 1.1 20.9 + 0 .6 Romaine Lettuce A Winter 23.9 + 0 .2 25. 1 + 0 .7 18.8 + 0 .2 Monkey Biscuit A Spring 22.5 + 0 .1 10.1 + 0 .3 2.9 + 0 .2 Monkey Biscuit A Winter 23.1 + 0 .3 8.6 + 0 .9 2.7 + 0 .3 Kale A Spring 27.9 + 0 .1 18.9 + 0 .0 14.5 + 0 .0 Kale A Winter 25.9 + 0 .2 16.9 + 1.2 11.9 + 0 .1 Beet A Spring 13.0 + 0 .4 16 .9 + 0 .3 9.6 + 0 .0 Beet A Winter 9.5 + 0 .1 15.3 + 0 .0 8.1 + 0 .2 Baby Carrot A Spring 7.6 + 0 .2 15.2 + 0 .2 11.5 + 0 .2 Baby Carrot A Winter 8.7 + 0 .1 13.8 + 0 .0 10.7 + 0 .0 Apple A Spring 2.6 + 0 .1 14.5 + 0 .0 10.6 + 0 .2 Apple A Winter 2.3 + 0 .1 11.0 + 0 2 8.0 + 0 .0 Overall Diet A Spring 22.0 25.7 18.8 Overall Diet A Winter 22.8 23.1 17.1 Romaine Lettuce B Spring 25.4 + 0 .2 21.0 + 0 .4 16.9 + 0 .2 Romaine Lettuce B Winter 1 21.6 + 0 .2 19.2 + 0 .1 15.3 + 0 .0 Romaine Lettuce B Winter 2 26.5 + 0 .1 25.9 + 0 4 19.8 + 0 .0 Nutrient values are expressed on a percent dry matter (DM) basis and reported as mean (average of replicates for each ingredient ) + / the standard deviation.

PAGE 76

76 Table 3 3. Crude protein (CP), neutral detergent fiber (NDF), and acid detergent f iber (ADF) composition of combined fecal samples for each of the manatees in the study. Animal Facility Season % CP % NDF % ADF 1 A Spring 19.4 + 0.7 23.4 + 0.1 17.9 + 0.2 1 A Winter 22.3 + 0.2 27.1 + 0.8 18.4 + 0.3 2 A Spring 18.8 + 0.2 25.0 + 1.5 1 8.2 + 0.3 2 A Winter 22.9 + 0.1 22.0 + 0.3 16.0 + 0.1 3 B Spring 24.6 + 0.1 29.6 + 0.9 21.9 + 0.7 3 B Winter 21.0 + 0.1 33.1 + 1.9 21.7 + 0.9 4 B Spring 22.7 + 0.1 28.4 + 1.5 21.0 + 0.8 4 B Winter 1 19.5 + 0.1 24.0 + 0.0 18.4 + 0.5 4 B Winter 2 23.6 + 0.0 27.2 + 0.1 19.7 + 0.5 5 B Winter 23.8 + 0.0 22.8 + 1.1 17.6 + 0.6 All nutrient percentages are expressed on a dry matter (DM) basis as means +/ one standard deviation Spring collection from animal number five was not possible, because it was r eleased back into the wild soon after the winter collection.

PAGE 77

77 Table 3 4. Acid insoluble ash (AIA) and acid detergent lignin (ADL) concentrations in the diet and feces for each manatee and season. Animal Season % AIA Diet % AIA Feces % ADL Diet % ADL Fec es 1 Spring 0.48 0.49 3.69 11.41 1 Winter 0.36 0.16 3.16 12.42 2 Spring 0.48 0.36 3.68 11.42 2 Winter 0.36 0.13 3.16 10.29 3 Spring 0.38 0.22 2.42 12.59 3 Winter 0.9 0.19 2.83 13.53 4 Spring 0.38 0.16 2.42 13.22 4 Winter 1 0.16 0.06 2.60 11.63 4 W inter 2 0.9 0.08 2.83 12.00 5 Winter 0.16 0.11 2.60 11.58 Values are expressed as a percentage of dry matter (DM).

PAGE 78

78 Table 3 5. Apparent digestibility values for dry matter (DM), crude protein (CP), neutral detergent fiber (NDF), and acid detergent fiber (ADF). Apparent Digestibility (%) Animal Facility Season DM CP NDF ADF 1 A Spring 67.6 71.5 70.6 69.3 1 A Winter 74.5 75.0 70.2 72.6 2 A Spring 67.8 72.5 68.6 68.9 2 A Winter 69.3 69.1 70.7 71.2 3 B Spring 80.8 81.4 73.0 75.2 3 B Winter 79.1 83.5 73.3 77.1 4 B Spring 81.7 83.7 75.3 77.2 4 B Winter 1 77.6 79.9 72.0 73.2 4 B Winter 2 76.4 79.1 75.2 76.5 5 B Winter 77.6 75.3 73.4 74.2 Values are expressed as a percentage of dry matter (DM).

PAGE 79

79 Figure 3 1. Percent acid insoluble ash (AIA) v s. total amount of collected feces. The line represents the linear correlation between the two variables. There was no evidence of a correlation (p = 0.107, R 2 = 0.292).

PAGE 80

80 Figure 3 2 Linear regre ssion of dry matter (DM) intake vs % DM digestibility for manatees in the present study Percent DM digestibility decreased with increasing dry matter intake. y = 91.2 5.614x R 2 = 0.8

PAGE 81

81 CHAPTER 4 IN VITRO FERMENTATION ESTIMATES FOR FREE RANGING, CAPTIVE, AN D REHABILITATING FLORI DA MANATEES ( TRICHECHUS MANATUS L ATIROSTRIS ) Backgroun d The Florida manatee ( Trichechus manatus latirostris ) is an herbivorous, endangered marine mammal that consumes both fresh and salt water species of vegetation in its natural habitat, while foraging for up to 6 8 hours per day [Hartman, 1979; Best, 1981; Etheridge et al., 1985; Burn, 1986]. Manatees are hindgut fermenters that rely on microbes in their colon to break down the large amounts of pl ant fiber they ingest. M icrobial population s and their abili ty to degrade carbohydrates have been known to chan ge depending on the diet being fed, particularly the amount of fiber ingested in pigs [Imoto and Namioka, 1978], humans [Fleming and Rodriguez, 1983], and dogs and cats [Sunvold et al., 1995d] M icrobial diversity and its ability to degrade different sub strates can also be affected by disease state [Krisztina et al., 2011] as found for horses predisposed to laminitis [Murray et al., 2009]. The diet fed to captive manatees has significantly less plant fiber and ash than the vegetation manatees would cons ume in their natural habitat [Siegal Willott et al., 2010 L. Harshaw, unpubl. data, Chapter 2 ], but the impact microbial function is unknown In the present study, romaine lettuce represent ed the captive manatee diet, bec ause it is the main dietary ingredient at most captive facilities. Natural diets were typified with Syringodium filiforme (manatee grass) representing the marine diet, Vallisneria americana (eel grass/tapegrass) for freshwater vegetation and Ruppia marit ima (widgeongrass) as a brackish water plant Free ranging Florida manatees are known to consume all of these plant species in their natural habitat [Hartman, 1979; Best, 1981; Bengtson, 1983; Provancha an d Hall, 1991]. Alfalfa hay was also analyzed, bec ause it has been considered a potential supplemental source of plant fiber in the diet fed to captive manatees.

PAGE 82

82 able to process and break down food ingredients to ob tain available nutrients. During fermentation, microbes metabolize carbohydrates and proteins into short chain fatty acids (SCFA) and branched chain fatty acids (BCFA). The SCFA include acetate, propionate, butyrate, and valerate whereas the BCFA includ e isobutyrate and isovalerate Conventional in vivo digestibility and fermentation trials are time consuming, expensive, often require manipulating both the animal and its diet, and are not always p ossible with endangered species In vitro fermentation e xperiments have been successful in predicting in vivo fermentation and in comparing the relative fermentation of different feeds for several species including horses [Abdouli and Attia, 2007; Denek et al., 2008], dogs and cats [Sunvold et al., 1995a; Sunvo ld et al., 1995b], and ruminants [Tilley and Terry, 1963; Meyer et al., 1971; Adesogan, 2002]. These in vitro studies simulate the fermentative processes in the digestive tract by exposing feeds to a mixture of buffer fluid and a source of microbial inocu lum in an enclosed anaerobic environment The traditional source of microbial inoculum is ruminal fluid, but this requires access to a cannulated individual of the ruminant being studied Feces have been shown to be a reliable source of microbes for such an inoculum in a variety of mammalian species, including other hindgut fermenters [Stark and Madar, 1993; Sunvold et al., 1995c; Sunvold et al., 1995d; Lowman et al., 1999; Ferguson and Jones, 2000; Fernandes et al., 2000; Warren and Kivipelto, 2005; Latt imer et al., 2007; Murray et al., 2009; Earing et al., 2010]. Additionally, Hastie et al. [2008] used PCR and deter m ined that hindgut and fecal microbial species were similar in the horse. Feces have recently been used to evaluate in vitro fermentation i n another Sirenian mammal, the dugong ( Dugong dugon ) [Goto et al., 2004a]. Digesta can only be obtained by

PAGE 83

83 surgery in live manatees or by immediate necropsy following the death of an ani mal. Neither of these options wa s feasible, so feces were the best a vailable source of microbes for this purpose. The in vitro dry matter digestibility (IVDMD) method developed by Tilley and Terry [1963], has proved an accurate in vitro method of predicting digestibility of forages in ruminant animals, but is laborious and time consuming. The Daisy II rotary incubator apparatus (ANKOM Technology Corp, Fairport, NY) provides a more convenient alternative, because it allows for the incubation of several foods or feeds together using a single inoculum. This rotary incubat or system has provided accurate in vitro digestibility values for ruminants including sheep and cows [Holden, 1999; Vogel et al., 1999; Mabjeesh et al., 2000; Wilman and Adesogan, 2000; Adesogan, 2002 ], and fermentation values for horses [Warren and Kivipe lto, 2005; Lattimer et al., 2007; Earing et al., 2010] and therefore, may also be able to evaluate the fermentation of different foods consumed by manatees using fecal microbes. The objectives of the present study were to compare the use of fecal microbe s from captive and free ranging manatees to evaluate the fermentation of the diet offered to captive manatees possible alternative sou rces of fiber, and marine, brackish, and fresh water grasses, by measuring % dry matter (DM) lost, % neutral detergent fi ber (NDF) lost and the production of lactate, SCFA and BCFA This study also sought to deter mine whether the Daisy II incubator was suitable for determining fermentative ability using manatee feces as the microbial inoculum. It was hypothesized that mi crobes from free ranging manatees would ferment the foods to a greater degree than the microbes from captive manatees with % DM lost and % NDF lost from each of the foods greater for free ranging manatees It was also predicted that the Daisy II rotary i ncubator would be suitable to study manatee fermentative capacity in vitro A logistical problem associated with using microbes from the feces of manatees for in vitro fermentation

PAGE 84

84 assays is that the facilities housing manatees and free livi ng manatees ca n be two or more hours away from the laboratory where analyses are run and it is unknown whether fecal microbes from manatees would remain viable during transport Published results indicate that r uminal fluid fermentation and gas prod uction were unaffect ed when ruminal microbes were chilled for up to six hours under anaerobic conditions [Hervas et al., 2005]. Robinson et al. [1999] found that delaying the use of ruminal fluid preserved under anaerobic conditions at a temperature of 39 o C for up to six and half hours after collection did not affect digestion values after a 48 hour fermentation period in an incubation system such as the one used in the study reported here. Thus, it was predicted that manatee fecal microbes could be preserved similarly and w e sought to determine the best preservation method for fecal microbes and predicted that using a blanket of gas or vacuum packing the feces would preserve the anaerobic environment. Materials and Methods All of the following work was performed under a permit from US Fish and Wildlife Service (#MA038448 3) and with approval from University of Florida IACUC (#200902762). Foo d Sample Collection, Processing and In vitro Fermentation M ethod Romaine lettuce was purchased from a local grocery store in Septemb er 2010. All aquatic grasses wer e collected under permit # 48008225 from the Florida Department of Agriculture and Consumer Services, Division of Plant Industry, and under a saltwater fishing license (#494266489) from Florida Fish and Wildlife Conservation Commission. Only fresher grass pieces were collected for all thr ee species, as determined by their hue of green color, along with rhizome parts. Beach cast Syringodium was collected in February 2011 from the sh oreline of Seahorse Key, FL (N 29.06275, W 83.06234) Vallisneria was collected from a sho reline in Crystal River, FL (N 28.891111, W 82.597222) in October 2010, and Ruppia was collected at Fort Island Gulf Beac h, FL (N 28.904952, W 82.689507) in May 2011. Alfalfa hay was

PAGE 85

85 purchased in bulk from Id aho and has been use d as a labora tory standard Coastal b ermuda hay was purchased in 2008 from Roland Vaughn Farm in Alach u a, FL Grasses (leaves and rhizomes) hays, and romaine lettuce (leaves and core) were dried to a constant weight in a forced ai r oven (Model 05015 58, Cole Parmer, Niles, IL ) at 55 o C and then ground using a Wiley Mill (Thomas Scientific, USA, 3383 L10 Series, Swedesboro, NJ) to pass through a 1 mm screen. The dried and ground food samples were kept i n sealed bags (Ziploc SC Joh nson and Son, Inc., Racine, WI) until they were used to assess fermentation To evaluate fermentation, a pproximately 0.5 g of dried food was measured into 3 x 5 cm bags made of polyester filter material (Unibond HK 250 N, Midwest Filtration, Cincinnati, O H). Bags were weighed again after processing through a fiber analyzer (ANKOM 200, Technology Corp., Fairport, NY) to determine the NDF concentration of each food (Table 4 2), as detailed in Chapter 2. The fermentative capability of each fecal sample w as assessed in vitro using a rotary incubator ( ANKOM Daisy II ANKOM Technology Corp., Fairport, NY) the appropriate in vitro dry matter digestibility procedure ( ANKOM method 3, ANKOM Tech nology Corp., Fairport, NY ), and using a method adapted to use feca l material as the inoculum instead of ruminal fluid. This incubator contains four jars, which are rotated on slow turning rollers to agitate the jar contents during each incubation Three bags of each food type and two blank bags for a total of seventee n bags, were added to each jar in the incubator B lank bags were used to calculate a correction factor to account for any potential particle loss from filter bags during an incubation This rotary incubator maintains a constant cabinet temperature of 39 5 o C using a thermostat ically controlled heating bulb located in the lower right corner.

PAGE 86

86 Buffer A (10.0 g of KH 2 PO 4 0.5 g of MgSO 4 *7H 2 O, 0.5 g of NaCl, 0.1 g of CaCl 2 *H 2 O, and 0.5 g of Urea dissolved in 1 L distilled water) and b uffer B (15.0 g of Na 2 CO 3 and 1.0 g of Na 2 S*9H 2 O dissolved in 1 L of distilled water) were first made as separate chemical solutions. Buffers A and B were then kept warm in separate flasks at 39 o C in a large incubator chamber (Model B10620H, Lunaire Environmental Inc, Williamspor t, PA) and were not mixed together until a fecal sample had been collected, because the mixture was unstable Urea, the least stable component, was not added to buffer A until immediately before mixing with buffer B Buffers A and B were mixed in a ratio of 5:1.5 on a warm stir plate in a large glass bottle that had also been kept in the same incubator chamber and the pH was adjusted to 7.0 using additional buffer B if necessary. The pH of the mixed buffer was adjusted to 7.0, instead of 6.8 as recommend ed for ruminal fluid (ANKOM Metho d 3) Earing et al., [ 2010 ] had previously noted a significant drop in pH when adding fecal inoculum to the buffer mixture with an initial pH of 6.8 and a more acidic environment could adversely affect microbial fermentat ion This is not surprising, however, because the normal pH of the horse intestine is usually around 6.9 [Hassel et al., 2009]. A total of 2000 mL of buffer was mixed for each jar of the rotary incubator, of which 1600 mL was immediately added to a jar a nd 400 mL was retained in the large incubator chamber for making an inoculum later. The buffer and filter bags in each jar were allowed to equilibrate while being subjected to heat and agitation in the rotary incubator for at least one hour before fecal i noculum was added to the jar To prepare an inoculum, a 20 g aliquot of wet fecal sample from each animal was added to a stomacher bag with the 400 mL of the buffer mixture previously set aside and kept warm. Carbon dioxide (CO 2 ) gas was then blanketed o ver the feces/buffer mix for thirty seconds to establish an an a erobic environment and the whole bag was then blended in a large laboratory stomacher (Model #STO 400, Tekmar Company, Cincinnati,

PAGE 87

87 OH) for thirty seconds. The blended fece s buffer inoculum was then added to one of the jars of the rotary incubator and the whole jar was swirled to ensure proper initial mixing. Next, 5 mL of fluid was with drawn from the mixed contents of the jar using a plastic pipette, the contents of the jar were blanketed with CO 2 gas for thirty seconds and the lid was sealed After 48 hours in the rotary incubator, another 5 mL sa mple was withdrawn from the fluid in each jar Each fluid sample was frozen at 80 o C for lactic acid SCFA and BCFA analysis. Consistent rotatio n of the jars was confirmed visually every few hours. At the end of the incubation period, j ars were drained and filter bags were rinsed with distilled water until water was relatively clear. Filter bags were then dried to a stable weight in a 102 o C forc ed air oven (Model 05015 58, Cole Parmer, Niles, IL ) for at least four hours to determine % DM lost and weighed again after processing with a fiber analyzer (ANKOM 200) to determine post fermentation NDF loss Preliminary Study Using Horse Feces to Evalua te Methods of Preserving Fecal Microbes During Transport to the Laboratory Microbial preservation methods were tested using fec es caught during defecation from a single male thoroughbred horse stabled at the Univ m e dicine in Gainesville, Florida, once in the morning an d then three hours later. Feces were processing in a p re warmed, sealed thermos jug The first sample of fe ces collected was mixed by hand and separated into three twenty gram aliquots that were preserved using one of three different methods of excluding air: a blanket of nitrogen gas; a blanket of carbon dioxide gas; or by vacuum packing the sample. The first two of these fecal aliquots were p laced in polypropylene conical tubes (Nunc 50 mL Thermo Fisher Scientific, Waltham, MA) covered with a blanket of the selected gas, and sealed with a screw top lid The third aliquot was placed

PAGE 88

88 in a quart sized vacuum pack bag (Ziploc Handi Vac SC Johnson and Son, Inc., Racine, WI) and the air was then removed using the accompanying device provided for that purpose ( Ziploc Handi Vac SC Johnson and Son, Inc., Racine, WI) Then, both the tubes and the bag were kept warm in a thermos jug containing water at a temperature of approximately 39 o C. After three hours, the three fecal aliquots from the first sample and an aliquot from the second sample were each mixed with 400 mL of buffer solution in a stomacher for thirt y seconds to create an inoculum. Each inoculum was then added to one of the four jars of the in vitro fermentation system, so that each jar represented a different treatment. Each of the four j ars contained ten filter bags (four containing alfalfa hay, f our containing coastal bermuda hay, an d 2 blanks) and 2 L of buffer solution. Jars were mo ved relative to each other and the stationary heat bulb in a clockwise manner every six hours to eliminate any position influenced temperature differences [Adesogan, 2002; Warren and Kivipelto, 2005]. Sample processing p ost incubation was as described previously. Study Comparing the Effect of Feces f rom Free Ranging a nd Captive Manatees on Fermentation o f Various Foods To eliminate the need to rotate the position of the jars relative to each other every six hours as in the preliminary trial the entire incubator was covered with a thick blanket and temperature data loggers (High Resolution T hermochron iButton Model DS1921H; Maxim Integrated Products, Dallas Semicon ductor, San Jose, CA) housed in waterproof cases (i Button capsules, Model DS9107; Maxim Integrated Products, Dallas Semiconductor, San Jose, CA) were used to measure exact temperature differences between each of the four jar positions in the rotary incu bator. These modifications maintained fluid temperature in the jars in the first three positions within + 1 degree of 39.5 o C, but the temperature data loggers showed that fluid in the jar in the fourth position was still maintained at a temperature of at least one degree Celsius

PAGE 89

89 warmer than the jars in the other three positions, so only jars in the other three positions were used for each incubation when evaluating manatee fecal inoculums These temperature loggers were used in each test incubation and se t to record temperature every hour over the course of the incubation. For each test incubation, two of the jars were inoculated with fresh feces from manatees, whereas the jar at position three was always inoculated with an aliquot of digesta from a single animal for quality control to demonstrate consistency of operation of the rotary incubator across separate incubations To obtain a large sample of digesta to use as inoculum for quality control, digesta was collected from the hindgut of a manatee carcas s (ID number LPZ102900) necropsie d in the manatee pathology labor atory in St. Petersburg, FL within 24 hours after the manatee was euthanized on March 21, 2011 for an acute watercraft injury and p yothorax. The digesta was collec ted, homogenized in a gall on sized plastic bag (Ziploc SC Johnson and Son, Inc., Racine, WI), partitioned into 20 22 gram aliquots, and then frozen at 20 o C. When a test sample was collected from a manatee, an aliquot of this control digesta was warmed to 39 o C in a thermos jug o f warm water in pr eparation for making an inoculum when the test sample arrived at the laboratory. In addition, to show repeatability of the loss of DM and NDF during an incubation and to determine a normal range of these measures for comparison during te st incubations, two additional incubations were performed in which inoculums prepared from the control digesta were added to all three of the jars of the rotary incubator that maintained the correct temperature. Only incubations in which there was close a greement betwee n the values for DM and NDF lost from control samples and this normal range were used in the analysis. Fermentations were conducted using only a single rotary incubator, because only two fecal samples were processed on any one day, but two incubators were used on one occasion to

PAGE 90

90 evaluate four samples at once. Condition parameters and quality control values were similar for the two different incubators. A minimum of 20 g of wet feces was collected from each of five female manatees maintaine d in a semi captive environment at Homosassa Springs (HS) Wildlife State Park in Homosassa, FL six free ranging manatees located in Crystal River (CR) FL (N 28.891111, W 82.597222) and two rehabilitating manatees from Lowry Park Zoo (LPZ) in Tampa, FL ( Table 4 1). Individual fecal samples from each manatee were used to create single inoculums. All three locations are fresh water environments. Sex was only confirmed for two of the six free ranging manatees because these fecal samples were obtained duri ng annual manatee hea lth assessments at CR samples were obtained from free swimming manatees in Three Sisters Spr ings located within CR under a special use permit (# 41516 12002) iss ued by US Fish and Wildlife Service. Feces were collected from the water column while snorkeling with manatees shortly after excretion, but as the animal was leaving the area. Manatees were not pursued to determine sex, due to the need to immediately pr ocess the samples. Fecal samples from the HS manatees were obtained during annual health assessments, or from their holding tank shortly following excretion. Samples from the rehabilitating manatees at LPZ were collected from their holding tanks with a p ool net. The H S and CR manatees were considered healthy at the time of fecal collection, while the two rehabilitating manatees were considered stable. Th e first LPZ animal had been rehabilitated for cold stress syndrome, and had been in captivity for app roximately one year. The other rehabilitating animal had been admitted to LPZ for showing emaciation and had been in captivity for approximately four months. The manatees at HS were being offered a diet of mostly romaine lettuce and some cabbage, CR mana tees were assumed to be ingesting aquatic vegetation, and

PAGE 91

91 the manatees at LPZ were being offered a diet of mostly r omaine lettuce, with some endive/escarole, and Hydrilla verticillata a freshwater plant As soon as fecal samples were collected, they were placed in separate quart sized vacuum pack bags (Ziploc Handi Vac ) and the a ir was removed using the accompanying device (Ziploc Handi Vac ) T he entire bag was then placed in a warm thermos jug containing water at a temperature of approximately 39 o C for transport back to the lab oratory within three hours. Short chain and Branched chain Fatty Acid Analysis Supernatant fluid samples from each jar were allowed to thaw on the laboratory bench and then filtered through a micro concentrator tube (Amicon Ultra 4 3K; 3,000 MWCO, Millipore Billerica, MA ) in a high speed centrifuge (SM24 rotor, Sorval l RC5C centrifuge Newtown, CT ) at 7,600 rpm (7,000 x g) for 90 min at 2 o C. Approximately 180 L of the filtered fluid was transferred into 200 L inserts (Pro duct #C4012 465, National Scientific, Rockwood, TN) of a 2.0 mL screw thread glass vial (Product #033918, Fisher Scientific, Hampton, NH) and then analyzed for SCFA using high performance liquid chromatography (HPLC). The HPLC apparatus included a pump (Se ries 200 LC pump Perkin Elmer, Waltham, MA ), an automatic sample injector (Series 200 Auto Sampler Perkin Elmer, Waltham, MA ), a SCFA chromatographic column (ORH 801, 30x0.65 cm internal diameter ChromTech, Inc. Apple Valley, MN ) with guard column ( GC 801, ChromTech, Inc. Apple Valley, MN ), a column heater set at 55 o C, and a detector (Flexar UV/VIS detector Perkin Elmer, Waltham, MA ) set at a wavelength of 210 nm. The solvent used was 25.0 mM H 2 SO 4 (HPLC Grade, Sigma Aldrich, St. Louis, MO) at a flow rate of 0.7 mL/min for 30 minutes. The sample injection volume was 20 L. Each chromatogram was evaluated using a computer chromatographic analysis program

PAGE 92

92 ( TotalChrom Workstation softwa re, Version 6.2.1, 2003, Perkin Elmer, Waltham, MA ), and co ncentra tions were recorded as mmol /L (mM) of in vitro fluid. Calculations and Statistics Statistical analysis for the preliminary study was performed in SAS (SAS for Windows Version 9.2, SAS Institute Inc., Cary, NC ) Percent NDF lost was compared amongst ty pes of hay and preservation method s using a two way ANOVA, and an interaction was also considered. Percent DM lost during fermentation was calculated using Equation 4 1 and % NDF lost was calculated using Equation 4 2 where initial grams of NDF was equal to the initial sample weight (DM basis) of the food times the initial % NDF concentration of the food and the final grams of NDF was equal to the sample weight (DM basis) left after the fermentation and NDF analysis % DM Lost = 100 [(Initial sample d ry weight Po st fermentation dry weight / I nitial sample dry weight)] (4 1) % Pre fermentation NDF Lost = 100 [( Initial grams of NDF Final grams of NDF ) / Initial grams of NDF ] (4 2) S tatistical analyses of manatee data were performed using Sigmap lot (Systat Software Inc., version 11.2, San Jose, CA). The average % DM lost and % NDF lost from replicates of foods and percent coefficient of variation (CV) among replicates were calculated. These averages of replicates were log transformed prior to analysis or compared using non parametric tests when replicate averages were not normally distributed, as determined using the Shapiro Wilk test or when variances were unequal. Only values from the CR and HS manatees were compared, because samples were ob tained from only two manatees at LPZ. Percent NDF lost was compared using a general linear model procedure with food type as a repeated factor and

PAGE 93

93 location as a non repeated factor. An interaction between food type and location was also considered and po st hoc pairwise multiple comparisons were made using the Holm Sidak method Percent DM lost was compared among foods for the CR manatees using a one way repeated measures ANOVA Percent DM lost was compared among foods for the HS manatees using a Friedma n repeated measures ANOVA on ranks. Post hoc pairwise multiple comparisons were made among foods within each group of manatees s Percent DM lost of each food was then co mpared between locations using S tests. A Bonferroni co rrection was used to adjust the experiment wise probability of error to 0.05. Initial baseline post fermentation, and absolute increase in concentrations of acetate propionate, butyrate, isobutyrate, valerate, and total SCFA plus BCFA were co mpared between locations using S tests, whereas the same concentrations of lactate and isovalerate were compared between locations usin g Mann Whitney rank sum tests. A Bonferroni correction was also used here to adjust the experiment wise probability o f error to 0.05. The ratio of absolute increase in concentrations of acetate to propionate (A:P) was also compared between locations test. The CV for both % DM lost and % NDF lost was calculated in between each incubation of the contr ol digesta and for each of the all control incubations A probability of type 1 error > 0.05 was considered significant. Results In the preliminary study, mean NDF lost was greater (P < 0. 0 0 1) for alfalfa hay (38.3 + 2.9) than for bermuda hay (23.2 + 1.3 ) but no evidence of digestibility difference s among the methods of preservation when compared to the sample that was p repared immediately (P = 0.1 ) (Figure 4 1). For the manatee study, f eces were collected from sixteen individuals but only thirteen in vitro fermentations were successfully completed because three sample fermentations were not completed due to laboratory equipment failure All five of the HS manatees were

PAGE 94

94 adults, the two LPZ manatees were suba dults, one CR manatee was an adult and the o ther was a subadult Life stage was not determined for the other four CR manatees because they were free swimming and it was thus impossible to obtain morphometric measurements from these individuals Mean % NDF lost from each of the foods was not diffe rent between locations CR and HS (p = 0.8), but was different among the food types for both locations (p < 0.001) (Table 4 3) Percent NDF lost from romaine lettuce was greatest among the foods at 78 83% whereas % NDF lost from alfalfa and Ruppia was t he least among the foods at 42 44%. There was no interaction between location and food type for % NDF lost (p = 0.2). Mean % DM lost was different among the different food types for manatees at CR (p < 0.001) and HS (p = 0.001), but there was no eviden ce of a difference between locations for any of the food types (Table 4 4 ). Percent DM lost from romaine lettuce was greatest among the foods at 82 85% and % DM lost from Ruppia was the least at 36 38%. A large volume of buffer fluid was required to cover the filter bags in each jar of the rotary incubator, but this resulted in a substantial dilution of the feces. Many of the initial baseline SCFA concentrations of fermentation media were below the limits of detection because of this dilution, so pe rcent increase could not be calculated for most of the individual SCFA, BCFA and lactate There was no evidence of a difference between m ean i nitial baseline concentrations of lactate individual SCFA, individual BCFA, and total SCFA plus BCFA between lo cations, but mean post fermentation concentrations of lactate, acetate, propionate and total SCFA plus BCFA were differ ent between locations (Table 4 5 ). Post fermentation lactate was greater in HS (0.1 + 0.1) than CR manatees (0.01 + 0.01), whereas post fermentation acetate, propionate, and total concentrations were greater in CR manatees (11.5 + 1.9, 3.7 + 0.7, 16.7 + 3.0, respectively) than HS manatees ( 8.6 + 1.4, 2.4 + 0.3, 12.6 + 2.3, respectively) Mean absolute increase in

PAGE 95

95 concentration of lactate was greater in HS manatees (0.1 + 0.1) than CR manatees (0.002 + 0.004), whereas mean absolute increase in concentration of acetate, propionate, and total SCFA plus BCFA were greater for CR manatees ( 9.0 + 2.0, 3.7 + 0.7, 13.9 + 3.1, respectively) than HS manatees (6.0 + 1.6, 2.4 + 0.3, 9.6 + 2.3, respectively) (Figure 4 4). There was no evidence of a difference between CR (2.5 + 0.5) and HS (2.5 + 0.2) for t he ratio of absolute increase in concentration of A:P (p = 0.7). Coefficients of variation for the filter bag replicates of a food within a single jar for both % DM lost and % NDF lost, were all less than 10%, and most were below 5%. The CV for % DM lost from foods using the control digesta inoculum ranged from 4.0 17.4% between incubations with a ll foods except Vallisneria and Ruppia The CV of % NDF lost ranged from 5.9 23.3% between incubations with that of Vallisneria Alfalfa, and Ruppia all over 10%. Jar tempe ratures ranged from 39.1 39.4 o C, with mean jar temperat ure being 39.3 + 0.1 o C, with a CV of 0.3% (n = 20) Inter jar CV of average jar temperature within a single incubation ranged from 0.04 1.2%. Discussion There was no evidence that % NDF lost (Table 4 3) and % DM lost (Table 4 4) for HS manatees were different from CR manatees across foods Lowry park zoo manatees appeared to be different from the other two groups, but samples from more rehabilitating manatees at this location are needed to make a statistical comparison. The diet offered to the LPZ m anatees contained more dietary ingredients than the diet offered to H S manatees and included ingredients with higher plant fiber content. Some studies have found that in vitro digestibility or fermentation is affected by the type of diet consumed by the inoculum donor animal [Bezeau, 1965; Cherney et al., 1993; Holden, 1999]. Sunvold et al. [1995d] found that organic matter disappearance increased when fecal inoculum taken from dogs and cats fed more fermentable

PAGE 96

96 fiber was used. Dogs and cats can not util ize fiber t o the same extent as herbivores, so it is possible that their fermentative capability could be more easily in fluenced by diet Diet of inoculum donor did not seem to affect DM fermentation estimates in studies with horses after incubation perio ds of 48 or 72 hours, but differences were noted at a shorter incubation period of 30 ho urs [Earing et al., 2010]. I n vitro fermentations in the present study w ere carried out for 48 hours, which is the normal fermentation period for the Daisy II rotary i ncubator and was used because manatees have a longer digesta passage time [Larkin et al., 2007]. However it is possible that any potential diet influe nces may have been masked by this longer incubation period. Future studies should analyze nutrient disa ppearance at different fermentation time points for manatees consuming different diets to examine this. Percent DM lost during fermentations with feces from all groups of manatees was greatest for romaine lettuce and lowest for Ruppia with Syringodium V allisneria and a lfalfa being similar to each other ( Table 4 4 ). Lettuce similarly h ad the highest % NDF loss whereas Ruppia and a lfalfa had the lowest % NDF loss ( Table 4 3). This corresponds to lettuce containing the least concentration of NDF initial ly and Ruppia containing the highest initial % NDF, whereas the three other species were similar to each other and intermediate in value (Table 4 2). Alfalfa hay appeared to have a lower % initial NDF than Syringodium and Vallisneria but % NDF lost was l ower. Lignin has been shown to reduce digestibility [ Thayer et al., 1984 ; Van Soest, 1994 ] and alfal fa hay contains more lignin (5 %) than Vallisneria (3 %) on a dry matter basis [Linn et al., 1975], and more than the highest amount of lignin (3 %) reported in Syringodium plant parts [Siegal Willott et al., 2010]. P ercent DM lost from alfalfa hay after a 48 hour fermentation in the Daisy incubator with manatee feces as a source of the inoculums, was slightly higher in the present study (51 52%) than has be en previously reported with horse feces as the inoculum (34

PAGE 97

97 44%) [Earing et al., 2010]. Both manatees and horses have % DM lost values that were less than those that have been previously reported with ruminal fluid as the inoculum source 59 60 % for o ne study [Holden, 1999] and 54% in another [Mabjeesh et al., 200 0], but % DM lost values obtained here for romaine lettuce are similar to in vivo estimates obtained from captive manatees (L. Harshaw, unpub l data, Chapter 3). Thus, t he fermentation method used here may be useful for predicting digestibility estimates of dietary ingredients with lower concentrations of plant fiber in manatees. In vivo digestibility of fresh or salt water grasses by manatees is not available for comparison, but in vitro DM di gestibility was found to be 47 % for eelgrass after a 36 hour incubation with a fecal inoculum from a single dugong [Goto et al., 2004a], whereas in vivo DM digestibility of eelgrass in the same female dugong was over 90% [Goto et al., 2008]. The DM loss during in vitro fermentation using feces from this dugong was slightly less than in vitro DM loss from all of the different grasses except Ruppia in the study reported here It is therefore possible that the shorter incubation period in the dugong stud y could partly account for the lower DM loss Similarly, overall values for DM and NDF lost from the different aquatic grasses and alfalfa were less than those for romaine lettuce, but could be explained by the 48 hour incubation period, which is relativel y short compared to the digesta passage time of 6 10 days for manatees [Larkin et al., 2007]. Greater losses of NDF might be observed with a longer incubation period, because the microbes would have more time to break down the plant fiber. Earing et al. [2010] found that most digestion was completed within a 48 hour incubation in a similar rotary incubator using horse feces as the inoculum, but horses are known to have a shorter digesta passage time than manatees and actually had lower mean % NDF digestib ility of alfalfa after 48 hours (24 29%) than the manatees in the present study (42 4 6%) The alfalfa hay used in the

PAGE 98

98 Earing et al. [2010] study, however, contained a greater concentration of NDF ( 51% ) and these authors noted that diets with higher fi ber content may require longer incubation periods Other researchers have suggested that a longer incubation time may be necessary with horse feces because fecal microbes may require a longer lag time compared to ruminal microbes before they can begin fe rmenting substrates [Sunvold et al., 1995c]. Spanghero et al. [2010] estimated NDF digesti bility of alfalfa hay to be 36 % after a similar 48 hour rotary incubation with cow ruminal fluid as the inocul um source, but initial concentration of NDF of the alfa lfa was higher ( 58% ) All of the foods tested in the present study contained greater concentrations of plant fiber, with the exception of romaine lettuce. In another study, it was suggested that % NDF lost c an decrease when using 0.5 g of substrate, inste ad of 0.25 g [Lattimer et al., 2 007]. Using less than 0.5 g for the lettuce would probably not have been possible in this study, however, because almost all of the DM disappeared from lettuce over the course of the 48 hour incubation. Mean absolute incr ease in acetate, propionate, and total SCFA plus BCFA concentrations was greater for CR manatees than HS manatees whereas mean absolute increase in lactate was greater in HS manatees than CR manatees These differences in SCFA, BCFA, and lactate concent rations could result from CR and HS manatees having different microbial populations as a result of diet or environment. The HS manatees are maint ained in a natural spring head that can be acc essed by free ranging manatees, but t he two sets of individuals are not permitted to interact and are separated by a chain fence. However, this arrangement would allow for feces, and hence, microbes from free ranging manatees to be present in the More studies are needed with captive manatees that do not have access to free ranging microbes to confirm whether or not this is the case. Additionally distribution of species within microbial

PAGE 99

99 populations should be determined in digesta or fecal samples obtained from free ranging and captive manate es. The captivity based diet co ntains significantly less plant fiber and more non structural carbohydrates [Siegal Willott et al., 2010] than aquatic vegetation and it has been shown that inoculum from donor animals being fed diets with a greater concent ration of plant fiber produces greater total amounts of SCFA in horses [McDaniel et al., 1993] and dogs and cats [ Sunvold et al., 1995d]. Concentration of lactate in the large intestine was higher in dogs fed a meat diet than in dogs fed a cereal diet wit h a greater proportion of structural carbohydrates [Banta et al., 1979] and may indicate a difference in microbial populations because of the different diets consumed Lower concentrations of lactate at the end of a fermentation may result from there bei ng a greater proportion of lactate utilizing microbes in the inoculum sample [Sunvold et al., 1995c ; Anguita et al., 2006 ]. The p ercent ages of DM lost and NDF lost were not different between CR and HS even though the concentrations of some SCFA and tota l SCFA plus BCFA were different between the two locations but these differences could be related to the by products of fermentation, such as methane Sunvold et al., [1995c] suggested that cattle ruminal microflora may have produced less SCFA than cat fe cal microflora, because the cattle ruminal microflora may have actively produced methane. Therefore, it is possible that HS fecal microflora may have produced less SCFA than CR fecal microflora because HS fecal microflora produced methane. Methanogenic bacteria are not found in the hindgut of all individuals [Coles et al., 2005], so methane production could be affected by differences in hindgut microflora between the two locations. Methane is created from the hydrogen and carbon dioxide produced during carbohydrate fermentation, which can interfere with the metabolism of the microbes, so methanogenesis is

PAGE 100

100 considered somewhat beneficial for the animal [Hook et al., 2010]. However, if methane is produced in large quantities, the animal could incur signifi cant energy losses. Acetate, propionate, and butyrate were found in the greatest initial concentrations and had the greatest absolute increases in concentration (Figure 4 2) but this is not surprising as these are the three main SCFA produced by microb ia l fermentation [Coles et al., 2005]. Acetate to propionate ratios were not different between locations, so there did not seem to be a preference for production of propionate over acetate in either CR or HS manatees Goto et al. [2004a] found these thre e SCFA were produced in the greatest proportions after an in vitro fermentation with a female dugong, but final concentrations were greater with an average acetate concentration of 59 mmol/L, approximately 5 6 times the mean final concentrations found in t his study (9 12 mmol/L) (Table 4 5) Lowman et al. [1999] also reported greater total mean SCFA production of 43 mmol/L, but after a 140 hour long incubation. In vitro fermentation assays in b oth of these studies, however, were performed in tubes or bo ttles, so the degree of dilution by buffer would be less than in the present study. Percent CV between runs for foods incubated with the control inoculum samples were generally lower than 10% but higher than desired for % DM lost and % NDF lost from Va llisneria and Ruppia Percent NDF lost was also more variable for alfalfa. This rotary incubator is usually used to compare digestibility of different feeds or foods with an inoculum prepared from a single animal Any studies comparing inocula from dif ferent animals are usually completed within a single incubation and the author is unaware of any previous literature describing variation in measurements among separate incubations. Within incu bations, CV for % DM and % NDF lost from the foods using b oth test sample inoculums and control digesta inoculums, were all under 10%, except for % NDF lost from Vallisneria and Ruppia

PAGE 101

101 Thus, variation between runs and within incubations was similar. One potential source of variation is that each aliquot of con trol digesta/feces may not have contained the same number of microbes or the same species of microbes even though the whole sample was mixed thoroughly before separating out aliquots. Some microbes digest fiber better than others. S ome filter bags may also have contained a greater proportion of leaf or rhizome parts than others, even though care was taken to thoroughly mix the dried ground samples within their storage bags It may be better to separate these two fractions of the plant in future studie s to decrease potential variability Percent NDF lost was more variable than % DM lost, but this could be due to inconsistency in NDF analysis and from assuming initial NDF concentration of each food was the same in all filter bags. Increases in concent ration of SCFA and BCFA within the jars confirm ed some degree of microb ial preservation and activity for all experimental manatee fecal samples and control digesta samples frozen at 20 o C Frozen feces have also been used successfully as the microbial ino culum for in vitro fermentation experiments with horse feces [Murray et al., 2009] Nevertheless, freezing feces may not preserve all microbes. Jar position within the rotary incubator has been shown to affect the temperature of the jar contents and appa rent digestibility values [Adesogan, 200 2; Warren and Kivipelto, 2005], so previous researchers have resorted to moving the positio n of each jar every six hours. In this study, insulating the entire Daisy incubator with a thick blanket maintained temperat ure close to the desired 39.5 o C in all jars and provided a more convenient alternative to moving jars However, the jar in the lower right hand corner (fourth position) could not be used using this method, because it was still warmer than the other three and only two fecal experimental samples could be incubated at once Low variation

PAGE 102

102 among replicates of food within a jar also suggested that the jars were consistently mixed as they rotated and that microbes had equal access to each filter bag. Thi s stu dy indicated differences among food s for both % DM lost and % NDF lost and differences in lactate and SCFA production between two different groups of manatees These differences may result from consumption of different diets and/ or different hindgut microbial populations Results from this study also suggest ed that in vitro fermentation, with a rotary incubator is a useful way to evaluate fermentation of different foods consumed by manatee s a nd differences among fecal microbial populations among gro ups of manatees Managers and clinicians could therefore test potential diet changes for captive manatees without having to For example the results reported here suggest that alfalfa hay may be a suitable source of plant fib er for captive manatees but managers have been concerned previously about alfalfa hay causing colic in captive manatees Utilizing longer incubation times and smaller amounts of substrate might provide better estimates of digestibilit y for food types with greater amounts of plant fiber, but is not recommended for diets containing less plant fiber, such as romaine lettuce. More work needs to be done with other types of manatees including long term fully captive individuals and rehabili tating manatees of varying condition to determine how fermentative ability of fecal microbes might change with disease state and among manatees without access to free ranging microbes

PAGE 103

103 Table 4 1. Manatees, location, and date of collection of feces used for in vitro fermentation assays Animal Location Animal Status Date of Collection of Feces 1 HS Long term captive 8 July 2011 2 HS Long term captive 8 July 2011 3 HS Long term captive 22 September 2011 4 HS Long term captive 3 February 2012 5 HS Long term captive 3 February 2012 6 CR Free ranging 24 January 2012 7 CR Free ranging 24 January 2012 8 CR Free ranging 16 February 2012 9 CR Free ranging 16 February 2012 10 CR Free ranging 16 February 2012 11 CR Free ranging 16 February 2012 12 LPZ Rehabilitating captive 18 January 2012 13 LPZ Rehabilitating captive 18 January 2012 HS = Homosassa springs FR = free ranging, and LPZ = Lowry Park Zoo

PAGE 104

104 Table 4 2. Initial NDF concentration (dry matter basis) of each food used in the in vi tro fermentation assays. Food Init ial NDF (%)* Alfalfa hay 37.0 + 1.1 Romaine l ettuce 23.3 + 2.1 Ruppia maritima 53.9 + 0.1 Syringodium filiforme 40.0 + 1.7 Vallisneria americana 41.0 + 0.9 Values are m ean s +/ one standard deviation obtained from 3 replicates

PAGE 105

105 Ta ble 4 3 Percent NDF l ost from each food after in vitro fermentation with inoculums of feces collected from free ranging and captive manatees at three locations in Florida. Crystal River Homosassa Springs Lowry Park Zoo Alfalfa hay 42. 3 a + 7.1 42.3 a + 7.4 46.0 + 7.7 Romaine lettuce 83.0 b + 4.4 78.1 b + 6.7 87.5 + 1.0 Ruppia maritim a 43.4 a + 6.0 44.2 a + 6.0 45.6 + 7.6 Syringodium filiforme 55.0 c + 8.7 53.3 c + 4.9 70.6 + 2.6 Vallisneria americana 47.0 c + 7.6 53.5 c + 4.4 67.8 + 0. 3 Man atees were free ranging at Crystal River, healthy captives at Homosassa Springs, and rehabilitating captives at Lowry Park Zoo. Values are mean s +/ one standard deviation. Means with different s uperscripts (a, b, c) within each column are significantly different among foods within a location

PAGE 106

106 Table 4 4. Percent DM lost from each food after in vitro fermentation with inoculums of feces collected from free ranging and captive manatees at three locations in Florida. Crystal River Homosassa Springs Lowry Park Zoo Alfalfa hay 51.3 a + 1.9 51.7 a + 2.4 52.0 + 3.1 Romaine lettuce 84.9 b + 1.0 81.5 b + 2.4 84.2 + 4.1 Ruppia maritima 36.2 c + 8.3 37.8 c + 10.3 32.9 + 7.8 Syringodium filiforme 51.8 a + 4.8 51.0 a + 1.5 59.2 + 5.1 Vallisneria americana 54.3 a + 3.8 5 7.2 a + 3.2 60.1 + 0.3 Manatees were free ranging at Crystal River, healthy captives at Homosassa Springs, and rehabilitating captives at Lowry Park Zoo. Values are means +/ one standard deviation. Means with different superscripts (a, b, c) within each column are significantly different among foods within a location

PAGE 107

107 Table 4 5. Initial and post fermentation concentration s (mmol/L of in vitro fluid) of lactate, acetate, propionate, butyrate, valerate, isobutyrate, isovalerate, and total short chain (SCF A) plus branched chain fatty acids (BCFA) in buffer solution used for fermenting foods with inoculums of feces from free ranging and captive manatees at three locations in Florida Duration of fermentation (h) Crystal River Homosassa Springs Lowry Park Z oo Lactate 0 0.0 + 0.01 0.01 + 0.02 0.0 + 0.0 Lactate 48 a 0.01 b + 0.01 0.1 c + 0.1 0.0 + 0.0 Acetate 0 2.5 + 0.3 2.6 + 0.6 4.0 + 0.9 Acetate 48 a 11.5 b + 1.9 8.6 c + 1.4 11.2 + 1.8 Propionate 0 0.1 + 0.0 0.1 + 0.0 0.3 + 0.2 Propionate 48 a 3.7 b + 0.7 2 .4 c + 0.3 3.8 + 0.9 Butyrate 0 0.0 + 0.1 0.1 + 0.1 0.1 + 0.2 Butyrate 48 0.8 + 0.4 1.0 + 0.5 0.9 + 0.3 Valerate 0 0.0 + 0.0 0.0 + 0.1 0.0 Valerate 48 0.4 + 0.1 0.3 + 0.2 1.2 + 0.2 Isobutyrate 0 0.0 + 0.0 0.0 + 0.0 0.03 + 0.04 Isobutyrate 48 0.1 + 0.0 0.1 + 0.1 0.2 + 0.1 Isovalerate 0 0.1 + 0.0 0.1 + 0.0 0.3 + 0.1 Isovalerate 48 0.2 + 0.1 0.2 + 0.1 0.4 + 0.1 Total SCFA + BCFA 0 2.8 + 0.3 3.0 + 0.7 4.6 + 1.3 Total SCFA + BCFA 48 a 16.7 b + 3.0 12.6 c + 2.3 17.6 + 3.3 Values are mean s +/ one standa rd deviation. A superscript (a) within the column duration of fermentation indicate s significantly different concentrations between locations for that time point Means within a row with different superscript letters (b,c) are significantly different betw een Crystal River and Homosassa Springs.

PAGE 108

108 Figure 4 1. Percentage loss of NDF from two hays during in vitro fermentation using microbial inoculums produced from horse feces immediately after collection or three hours after collection preserved u sing three different methods either using a blanket of nitrogen (N 2 ) or carbon dioxide (CO 2 ) gas, or vacuum packing the feces Letters ( a, b ) indicate differences between the two hays. No difference was evident among preservation methods. Means are pre sented as colum n s; error bars represent +/ one standard deviation. 0 10 20 30 40 50 60 Immediate N2 Gas CO2 Gas Vacuum-Pack %NDF Loss Preservation Method Alfalfa Coastal Bermuda a a a a b b b b

PAGE 109

109 Figure 4 2 Absolute increase in concentration of lactate, SCFA, BCFA, and total SCFA plus BCFA after in vitro fermentation of foods with inoculums of feces collected from free ranging manatees at Crystal River (CR) and healthy, long term captive manatees at Homosassa Springs (HS) Stars indicate a significant difference between locations (p < 0.03 for all) Means are presented as columns; error bars represent +/ one standard deviation. 0 2 4 6 8 10 12 14 16 18 Absolute Increase in Concentration (mM) CR HS

PAGE 110

110 CHAPTER 5 FECAL SHORT CHAIN FATTY ACID AND LACTIC ACID CONC ENTRATIONS OF FLORIDA MANATEES ( TRICHECHUS MANATUS L ATIROSTRIS ) Background Florida manatees ( Trichechus manatus latirostris ) are endangered aquatic mammals that consume a variety of s alt and fresh water grasses in their natural habitat. Manatees, as hindgut fermenters, possess a highly specialized microbial environment in the large intestine to break down this plant material. These microbes are primarily responsible for the fermentat ion of carbohydrates to short chain fatty acids (SCFA), including acetic (C 2 ), propionic (C 3 ), butyric (C 4 ) and valeric (C 5 ) acids. These SCFA are the preferred oxidative fuel for the colonic mucosa in herbivorous species [Settle, 1988] and considered to physiology. The branched c hain fatty acids (BCFA) include isobutyric (iC 4 ) and isovaleric (iC 5 ) acids and are fermented from the amino acids produced by the breakdown of protein [Zarling and Ruchim, 1987; Sato, 2009]. These SCFA and BCFA provide herbivores with most of their dietary energy [Stevens and Hume, 1998] and are assumed to provide manatees with most of theirs as well. As with previous publications and for the purposes of this introduction BCFA are discusse d within the context of SCFA, unless otherwise noted. Free ranging manatees that are ill or injured are often rescued and transported to a critical care facility for rehabilitation until they can be released. Such manatees may not have a properly func tioning digestive system, because illness or injury may result in changes to the intestinal population of bacteria [Krisztina et al., 2011] For instance, systemic antibiotics given to animals upon their arrival could reduce the number of microbes in the hindgut. C hanges in pH of the intestinal environment can also alter the microbial population, because different types of microbes favor acidic and basic environments [Krisztina et al., 2011]. Manatees like horses are hindgut fermenters and d isease stat e has been shown to influence microbial diversity in horses,

PAGE 111

111 and to affect the intestinal microbes ability to break down certain substrates [Merritt and Smith, 1980; Murray et al., 2009] The amount of SCFA produced and absorbed may also be related to t he diet consumed [McNab, 2002], particularly the form and amount of fiber [Imoto and Namioka, 1978; Fleming and Rodriguez, 1983; Sunvold et al., 1995d]. Sea grasses and submerged aquatic vegetation (SAV) that are commonly consumed by free ranging manatees have more structural carbohydrates (plant fiber) and ash, and less protein, fat, and non structural carbohydrates (NSC) than the captive diet [Siegal Willott et al., 2010]. When an herbivore consumes more structural carbohydrates, the plant fiber can crea te a bulking effect that results in a dilution of the gut microbes, which could decrease microbial activity [Rowland et al., 1985]. Alternatively, if an herbivore consumes a greater proportion of more fermentable nutrients, such as NSC, the pH of the inte stinal environment could change and result in changes in number or type of microbes [Rowland et al., 1985]. It is therefore possible that the different diets consumed by manatees may result in different concentrations and types of SCFA produced and absorb ed Thus, SCFA concentrations in the feces of manatees might provide a measure of large intestinal health. Ideally, SCFA concentrations are measured in the intestinal dige sta of herbivorous animals but obtaining these samples requires surgery or removi ng digesta shortly after death. Microbial digestion and SCFA production can continue for long periods of time following death [Banta et al., 1979] however, so SCFA concentrations are more commonly measured in the feces of live animals to obtain a better representation of what the intestinal microbiota is capable of producing. Fecal concentrations of SCFA have been measured in humans [Chen and Lifschitz, 1989; Ogawa et al., 1992] and other animals, including, the horse [Merritt and Smith, 1980], pig [Urri ola and Stein, 2010], and gopher tortoise ( Gopherus

PAGE 112

112 polyphemus ) [Bjorndal, 1987], but the author is unaware of any measurements of fecal SCFA concentrations in Florida manatees. Short chain fatty acid concentrations have been measured previously in d ige sta samples taken from the gastrointestinal tract of other marine herbivorous hindgut fermenters including the green turtle ( Chelonia mydas ), the dugong ( Dugong dugon ), and the West Indian manatee ( Trichechus manatus ) [Murray et al., 1977; Bjorndal, 1979; Burn and Odell, 1987]. The greatest concentrations of SCFA were found in digesta located in the cecum and colon of these animals, as would be expected for hindgut fermenters. The lowest concentration of SCFA noted was found in the rectum of the green tur tle, but may be because most SCFA are absorbed within the large intestine before reaching this point [Bjorndal, 1979]. Thus, the rectal and consequently, the fecal, concentration of SCFA could be less than that of the cecum or colon, but could also be gre ater if a lot of water is absorbed in the colon, and the SCFA become more concentrated in drier feces. Thus, t his study sought to compare fecal SCFA concentrations of healt hy and rehabilitating manatees in captivity and manatees consuming different diets offered in captivity that contain less dietary fiber and ash and increased non structural carbohydrates with fecal SCFA concentrations of healthy, free ranging manatees consuming natural diets containing increased dietary fiber and ash and less non struct ural carbohydrate It was hypothesized that SCFA concentrations would be different among different populations and locations of healthy and rehabilitating manatees and between manatees consuming a captive manatee diet vs. a natural diet

PAGE 113

113 Materials and Me thods All of the following work was performed under a permit obtained from US Fish and Wildlife Service (#MA038448 3) and with approval from University of Florida IACUC (#200902762). Manatees and Sample C ollection Samples were collected from captive manate es housed at facilities in Florida throughout 2010 2012 including Lowry Park Zoo (LPZ) in Tampa, Parker Aquarium (PA) in Bradenton, Homosassa Springs (HS) State Park in Homosassa, and the Living Seas (LS) at Disney W orld in Orlando. Samples were also coll ected from two manatees housed at Cincinnati Zoo (CZ) in Ohio. Environmental parameters for each of the facilities are p resented in Table 5 1 Samples were only collected from manatees known to be consuming a plant based diet including an independent cal f subadults, and adult s as determined by length age class parameters established in the U nited Sta roject field guide (pers. comm.) Formula fed and fully dependent calves were not included. Fecal samples were col lected non invasively from the holding pool with a net within five minutes of defecation in cooperation with the staff at each location. Free ranging manatee fecal samples were obtained during winter health assessments in the months of November, December and January 2010 2012 hosted by USGS at two locations in Florida Crystal River (CR) (N 28.891111, W 82.597222), and the Indian River in Brevard County (BC) (N 28.483333, W 80.766667). Manatees were netted and brought onto shore for a complete biologic al evaluation during these events. S amples were also collected from free ranging manatees present in Three Sisters Springs, located in Kings Bay of Crystal River, Florida (N 28.888725, W 82.589191) during the winter months under a special use permit (#415 16 12002) issued by the US Fish and Wildlife Service. These samples were obtained by snorkeling with free swimming manatees and collecting the feces by hand from the

PAGE 114

114 water column shortly following excretion. The captive manatees at LPZ were being rehabil itated after injury or illness and were not considered to be healthy. The other captive manatees were either long term captives or had been recently rehabilitated and were con sidered to be healthy, because they were receiving secondary care having recover ed from illness or injury (Table 5 2). All free ranging manatees were considered to be healthy. Captive animals were being fed a diet comprised mostly of romaine lettuce, whereas free ranging manatees were assumed to be eating a diet of submerged aquatic vegetation (SAV) and sea grasses Sample Storage and P reparation Following collection, feces were placed in a quart sized plastic bag (Ziploc SC Johnson and Son, Inc., Racine, WI) gently mixed by hand within the bag to ensure homogenization of the SCF A, and sealed with an attempt to expel as much air as possible from the bag. Samples were immediately frozen on site either by placing the sample on dry ice, or in a 8 o C, 20 o C, or 80 o C freezer, depending on availability Samples were then transported back to the laboratory on dry ice at the end of the collection period, and kept frozen at 80 o C until analysis. For analysis, frozen fecal samples were thawed out on the laboratory bench and prepared using the high performance liquid chromatography (HPLC) method outlined by Chen and Lifschitz [1989], with minor modifications. Briefly, one to five grams of thawed feces were weighed i nto a plastic stomacher bag and between 2.5 and 7.5 mL of 0.15 mM H 2 SO 4 (Product #339741, Sigma Aldrich, St. Louis, MO) solut ion was then added to the bag. The amount of acid added to the feces varied with the consistency of the feces: less acid was added to more fluid samples, more acid was added to drier ones. The feces/solution mix was then homogenized in a laboratory stoma cher blender (Model Stomacher 80, Seward laboratory sy stems, Port Saint Lucie, FL ) for two minutes at medium speed. T hree to five mL of the liquid/slurry portion of the sample was

PAGE 115

115 then centrifuged (SM24 rotor, Sorval l RC5C centrifuge Buckinghamshire, En gland ) at 10,000 rpm (9,000 x g) for twenty minutes at 2 o C. Next, one to four mL of the supernatant from the previous step was filtered through a microconcentrator tube (Amicon Ultra 4 3K; 3,000 MWCO, Millipore Billerica, MA ) in the centrifuge (SM24 rot or, Sorval l RC5C centrifuge Buckinghamshire, England ) at 7,600 rpm (7,000 x g) for sixty min at 2 o C. One hundred and eighty L of the filtered fluid was then pipetted into a 200 L glass insert (Product #C4012 465, National Scientific, Rockwood, TN) of a 2.0 mL screw thread glass vial ( Product #033918, Fisher Scientific Hampton, NH ), and sealed with a cap to avoid loss of SCFA. These glass vials containing sample fluid were then placed in the injector tray of the HPLC autosampler. HPLC Apparatus The HPL C apparatus included a pump (Series 200 LC pump Perkin Elmer, Waltham, MA ), an automatic sample injector (Series 200 Auto Sampler Perkin Elmer, Waltham, MA ), a SCFA chromatographic column (ORH 801, 30x0.65 cm (i.d.), ChromTech, Inc. Apple Valley, MN ) wi th guard column (ChromTech, Inc. Apple Valley, MN ), a column heater (Timberline Instruments Inc., Boulder, CO) and a UV/VIS detector (Flexar FX 10 Perkin Elmer, Waltham, MA ) set at a wavelength of 210 nm. Samples were analyzed with the column heater se t at first 55 o C and then 65 o C to optimize peak separation The mobile phase was 25.0 mM H 2 SO 4 (Product #339741, Sigma Aldrich, St. Louis, MO) delivered at a flow rate of 0.7 mL/min. Run time was thirty minutes per sample, with an injection volume of 20 L. Between each analytic run, a 20 L sample of water was passed through the column to rinse off any residue Each sample was analyzed in duplicate and e ach chromatogram was evaluated using a computer chromatographic analysis program ( TotalChrom Worksta tion softwa re, Version 6.2.1, 2003, Perkin Elmer, Waltham, MA ).

PAGE 116

116 Data Processing A set of external standards was analyzed before each batch of samples to establish a standard curve The set of standards consisted of a stock solution and 75, 50, 25, 10, and 5% dilutions of that solution. The stock solution consisted of 50 L each of lactic (Product #69785 Sigma Aldrich, St. Louis, MO), acetic (Product #BP2401 212, Fisher Scientific, Hampton, NH) propionic (Product #P 1336, Sigma Aldrich, St. Louis, MO) i sobutyric (Product #I 1754, Sigma Aldrich, St. Louis, MO) butyric (Product #B 2503, Sigma Aldrich, St. Louis, MO) isovaleric (Product #I 7128, Sigma Aldrich, St. Louis, MO) and valeric (Product #V 0125, Sigma Aldrich, St. Louis, MO) acids mixed with pur i fied water to a volume of 10 mL. Standard curves were established by regressing the mM concentration o f each acid against peak height. Standard curves with an R 2 of 0.97 or greater were considered acceptable. To determine how much fecal SFCA concentrat ion had been diluted by the addition of acid to each fecal sample, it was necessary to establish the water content of the feces before dilution. To this end, the moisture content of each fecal sample was measured in duplicate by drying. Two small (0.5 2 grams) test samples o f each fecal sample were dried to stable weight overnight at 105 o C in a forced air laboratory oven (Model 05015 58, Cole Parmer, Niles, IL ) to determine percent dry matter (DM). The water volume of the sample was then established by m ultiplying the wet weight of the fecal sample by [(100 % DM )/100] The dilution factor was then calculated by dividing the test sample water volume by the sum of the test sample water volume and the volume of 0.15 mM H 2 SO 4 (HPLC Grade, Fisher Scientific Hampton, NH) solution added to the test sample assuming a water density of 1g/mL. The undiluted concentration (mM) of SCFA wa s then calculated by dividing the dilute concentration by the dilution factor. To obtain the amount of SCFA in mmol, the mM und iluted concentration was multiplied by the water content in L. This mmo l amount was then divided by the DM weight of the test sample of feces to

PAGE 117

117 obtain the mmol/g DM feces. Concentrations of individual SCFA and BCFA were added together to obtain the tota l concentration of SCFA Statistical Analysis Statistical analyses were performed using Sigmaplot ( version 11.2, Systat Software Inc., San Jose, CA). Values that were not normally distributed or that had unequal variances following log transformation w ere compared using non parametric tests. E ach of the individual SCFA and BCFA, lactic a cid, and the total SCFA plus BCFA concentrations in mmol/g DM of feces, as well as the acetate to propionate ratio (A:P) were compared among locations using a Kruskall Wallis one way analysis of variance on rank s to determine whether or not any of the locations/ groups of manatees were different from each other Post hoc comparisons among locations were performed using e multiple comparison procedure Lacti c acid, individual SCFA and BCFA, and total SCFA plus BCFA concentrations for free ranging CR and BC manatees were combined to represent those consuming a natural diet (n = 29) and the same concentrations for the LS, HS, PA, and CZ manatees were combined to represent those ingesting a diet offered in captivity (n = 12) L actic acid, SCFA and BCFA concentrations were then compared between these two groups of healthy manatees using Mann Whitney rank sum tests to determine if there was a difference be tween manatees consuming a natural diet and manatees consuming a diet offered to them in captivity Percent DM of feces was compared between manatees consuming a natural diet (n = 29) and manatees consuming a captivity based diet (n = 12) te st, after passing normality and equ al variance tests. Percent DM of feces was also compared among the different groups/locations of manatees using a Kruskal l Wallis one w ay analysis of variance on ranks A probability of type 1 error > 0.05 was considere d significant.

PAGE 118

118 Results Fecal s amples were collected from eleven rehabilitating manatees seven long term healthy captive manatees and five recently rehabilitated manatees (Table 5 1 ). One of the long term healthy manatees was originally housed at HS bu t transferred to CZ in November 2011. Samples were collected from this animal at bot h locations Fecal samples were collected from twenty nine free ranging manatees; of which sex and life stage were known for twenty one individuals: 11 males, of which 7 were subadults and 4 were adults; and 10 females, of which 3 were subadults and 7 were adults In general, rehabilitating manatees at LPZ appeared to have the widest range and some of the greatest fecal SCFA concentrations, whereas the LS manatees appeare d to have the lowest mean total SCFA concentrations. Median fecal concentrations of lactic acid differed among manatees from different locations (p = 0.01), but specific pairwise differences were not detected (Figure 5 1). Median fecal a cetic acid concen trations were greater (p=0.002) in free ranging manatees at CR ( 0.2 mmol/g DM feces ) and BC (0.3 mmol/g DM feces ) than in manatees at LS ( 0. 07 mmol/g DM feces ) but similar to fecal concentrations in manatees at all other locations (Figure 5 2). The media n propionic acid concentration in free ranging BC manatees (0.091 mmol/g DM feces) was greater (p = 0.04) than that in LS manatees (0.040 mmol/g DM feces), but there were no other differences evident among locations (Figure 5 3). Free ranging B C manatees also had greater (p = 0.02) median concentrations of isobutyric acid (0.019 mmol/g DM feces) th an LS manatees (0.005 mmol/g DM feces) (Figure 5 6 ). Median concentrations of both butyric and iso valeric acid differed among locations (p = 0.03 and 0.01 resp ectively ), but specific p airwise differences were not detected (Figures 5 4 an d 5 7 ). Valeric acid concentrations were not different among locations (p = 0.29) (Figure 5 5 ) The median concentration of total SCFA was greater for fre e ranging Brevard Coun ty manatees (0.521 mmol/g DM feces) than for Living Seas manatees (0.152 mmol/g DM feces) (p = 0.01),

PAGE 119

119 but no other location differences were detected (Figure 5 8). The ratio of acetate to propionate was different among the locations (p = 0.01), but specif ic differences were not detected (Figure 5 9). Manatees consuming a natural diet had greater acetic, butyric, and total SCFA plus BCFA concentrations c ompared to manatees consuming a diet offered in captivity (Table 5 3), but had concentrations of propion ic, valeric, isobutyric, and isovaleric acids that wer e similar to the captive manatees (Table 5 3) There was a trend towards manatees consuming a natural diet having lower concentrations of lactic acid than manatees consuming a diet offered to them in captivity (p = 0.1) Manatees consuming a natural diet had higher A:P ratios compared to manatees consuming a captivity based diet (p < 0.001). There was no evidence of a difference in the median % DM of feces among different groups/locations of manatees (p = 0.9) or in the mean % DM of feces between manatees consuming a natural diet (23.9 + 5.9) and manatees consuming a captivity based diet (22.9 + 6.5) (p = 0.6 ) (Table 5 4). Discussion Fecal SCFA and BCFA concentrations in the present study were found to be different among different locations/groups of manatees and between manatees consuming a natural diet versus manatees consuming a diet offered to them in captivity Fecal SCFA concentrations represent the net result of productio n and absorption for i ndividuals and also include any endogenous sources of SCFA such as intestinal cells that have been shed and degraded [Bugaut, 1987]. Greater concentrations of SCFA were seen in the feces of some of the reha bilitating manatees (Figure 5 8), and could repr esent increased production of SCFA or reduced absorption either from a change in the amount of substrate available for fermentation, a change in microbial flora, or a change in the intestinal mucosa. Rehabilitating manatees may have an incr eased appetite because they had limited access to f ood, or were not able to forage before being rescued

PAGE 120

120 and may over consume food offered to them upon arrival at a critical care facility. This sudden increase in food could exceed their digestive capacity, with the exces s nutrients undergoing rapid fermentation and then being over released in their feces, as was found with human preterm infants [Szylit et al., 1998]. Conversely, these manatees may also be producing a normal amount of SCFA and not absorbing them efficient ly. Some of the rehabilitating manatees were observed to produce looser feces than the healthier ones in this study, even though there was no evidence of a difference in % DM of feces between the different groups/locations. Looser feces may have result ed from reduced absorption of water. Absorption of SCFA assists in both sodium and water absorption, so if SCFA are not absorbed, water and sodium absorption could be compromised [Stevens and Hume, 1998] The malabsorption of carbohydrates has been linked previously to diarrhea in humans [Soergel, 1994]. Ad d itionally, fecal microflora from captive rehabilitating animals at LPZ and healthy, free ranging CR manatees appeared to produce similar amounts of SCFA from typical substrates during 48 hour in vitro f ermentation experiments (L. Harshaw, unpubl. data, Chapter 4). This latter speculation that rehabilitating manatees may not be absorbing SCFA efficiently is therefore more likely and warrants further investigation into the digestive health of rehabilitati ng manatees The three main SCFA found in the highest concentrations in the feces of manatees were acetate, propionate, and butyrate with acetate being present at higher concentrations than propionate, which was in turn higher in concentration than butyrate The relative order of concentrations of these three SCFA was similar to those reported in the digesta from the rectum of the Green sea turtle [Bjorndal, 1979], and from the feces of land and marine iguanas [Mackie et al., 2004], which are both h erbivorous hindgut fermenters. Butyrate concentration was higher than propio nate in all parts of the digestive tract in the Green turtle except the rectum,

PAGE 121

121 suggesting that most of the butyrate was absorbed in the colon [Bjorndal, 1979]. Butyrate is abso rbed more quickly than propionate and acetate, because it is more lipid soluble and therefore, passes more easily through intestinal cell membranes [Stevens and Hume, 1998]. More propionate than butyrate being found in the feces of healthy manatees in the present study suggested that most butyrate is probably absorbed in the colon for healthy manatees as well. The greater concentrations of butyrate found in the feces of reha bilitating manatees (Figure 5 4 ), however, may also support the hypothesis that un healthy manatees may not be absorbing SCFA efficiently, particularly butyrate. B utyrate can be converted to acetate but is inhibited when there is already a large amount of acetate present [Bjo rndal, 1979]. Thus, this prevention of conversion may have a lso contributed to the greater concentrations of but yrate found in the feces of rehabilitating manatees and individuals consuming a natural diet in the present study which both had greater concentration s of acetate Lactate concentrations were genera lly low for all individuals b ut captive manatees appeared to have great er concentrations than their free ranging counterparts even though no statistical significance could be detected (Table 5 3) (Figure 5 1). Starch, a non structural carbohydrate, is e asily fermentable and results in increased lactic acid concentrations in the hindgut when excess amounts of starch are passed to the large intestine of the horse [Kentucky Equine Research, 2007]. The captive manatees in this study were fed mainly romaine lettuce, so it is not surprising that their fecal concentrations of lactic acid appeared greater because romaine lettuce contains more NSC than the natural diet [Siegal Willott et al., 2010] and contains 16 % starch on a DM basis [Schmidt et al., 2005b] The greatest concentr ations of lactic acid found in the feces of LS manatees could indicate a need for more plant fiber in their diet to offset the larger quantities of soluble carbohydrates. Rehabilitating manatees also appeared to have greater

PAGE 122

122 concentra tions, however, which may reflect their illness or injury leading to malabsorption as discussed earlier or overconsumption of foods with a lower concentration of plant fiber It has been postulated that m etabolic acidosis and colitis can both occur if to o much lactate is absorbed in the hind gut of dairy calves [Shimomura and Sato, 2006; Sato, 2009], so fecal lactate concentration may provide a measure of gut health. Fecal lactate concentrations decreased in young calves recovering from diarrhea, indicati ng improvement in both gut flora and fermentation [Sato, 2009]. The status of hindgut function is usually unknown when a manatee is brought into rehabilitation, so monitoring fecal lactate concentrations may similarly be useful in tracking gut health over the course of their stay. Free ranging manatees had greater concentrations of acet ate (Figure 5 2) butyrate (Figure 5 4) total SCFA plus BCFA (Figure 5 8) and greater A:P values t han manatees consuming a captivity based diet (Table 5 3) (Figure 5 9), which is not surprising because the natural diet contains a greater amount of structural carbohydrates, or plant cell wall components [Siegal Willott et al., 2010]. A lower A:P ratio in ruminants and equids has been associated with these animals consuming a diet containing less forage and more grain and as a result, more soluble carbohydrates There is one major outlier in the free ranging CR group with an A:P ratio of 12 This high value is attributed to the animal having such a low concentration of pr opionate. This fecal sample was collected from one of the free swimming manatees, so it is not known what the animal was eating or if it had any internal digestive abnormalities. It is possible that this particular manatee produced less propionate becaus e it was consuming a diet that had a low concentration of NSC Overall, the manatees in this study had lower A:P ratios than those of elephants, their closest living land relatives. Studies with captive Asian elephants and free ranging African elephants fou nd A:P ratios to range from 4.1 8 .0 [Van Hoven et al., 1981;

PAGE 123

123 Clemens and Maloiy, 1983; Clauss et al., 2003] in fecal samples. Nevertheless, a ll of the se elephants were ingesti ng a diet with a greater concentration of plant fiber which likely explain s the greater values. Acetate to propionate ratios were even higher for gopher tortoises (14) being fed a diet containing a high concentration of plant fiber [Bjorndal, 1987]. The LS manatees had the lowest A:P values, but also consume a diet with the le ast plant fiber out of all of the groups being manag ed in captive facilities. Diets lower in structural carbohydrates resulted in a greater production rate of acetate in the large intestine of pigs, but significantly low acetate excretion in the feces, le ading the authors to suggest that the absorption rate of acetate in the large intestine was also higher for pigs consuming a diet with a low concentration of plant fiber [Imoto and Namioka, 1978]. This greater absorption rate could also potentially contri bute to the low A:P values seen here for LS manatees Homosassa manatees which appeared to have a slightly higher A:P ratio than other healthy, captive manatees consume a mixed diet of romaine lettuce and cabbage, which has slightly more NDF [Schmidt et al., 2005b]. The rehabilitating manatees were offered several dietary ingredients including romaine lettuce, endive/escarole, an d H ydrilla verticillata a fresh water grass The amount of each of these dietary ingredients that individual manatees consum ed was not documented, however, and could account for some of the variability found in the present study A differential absorption rate, in which acetate is absorbed more quickly than butyrate, which is absorbed more quickly than propionate, was proposed for sea turtles based on the proportions of propionate and butyrate increasing relative to acetate in successive sections of the colon [Bjorndal, 1979] Argenzio et al. [1974] bathed large intestine tissue from the horse in ringer solution containing ace tate, propionate, and butyrate and measured the concentration of these three SCFA in the solution recovered from the blood side of the tissue at different times during the exposure to obtain SCFA transport rates and found that

PAGE 124

124 acetate was absorbed more qui ckly than propionate and butyrate. A similar differential absorption rate may also explain the lower A:P values reported in the present study for manatees Isobutyrate (Figure 5 6 ) and isovalerate (Figure 5 7 ) concentrations appear ed to be greater in cap tive and free ranging BC manatees than in the C R manatees It is unclear why BC manatees might have had greater concentrations of isobutyrate and isovalerate than CR manatees, but it is likely that free ranging manatees ingest different foods at different locations depending on availability Isobutyrate is produced from valine, whereas isovalerate is produced from leucine [Zarling and Ruchim, 1987; Rasmussen et al., 1988]. Valine and leucine are both essential amino acids, so they must be consumed in th e diet. Greater concentrations of isobutyrate and isovalerate in the feces can indicate improved or greater amounts of fermentation [Sato, 2009], or result from consuming a diet that is higher in protein [Sato, 2010]. It is possible that the concentratio n of protein in the grasses at each of the locations studied here varied because of the amino acid content of the grasses as was found for some species of Mediterranean seagrasses [Pirc and Wollenweber, 1988] or protein concentration could have be en diff erent because of incidental ingestion of epibiota In this way BC manatees might have ingest ed more protein if they consume d greater quantities of broad leaf sea grasses, such as Thalassia because more epibiota tend to attach to these sea grasses [Parker et al ., 2001]. There was no evidence of a difference in isobutyric or isovaleric concentrations between manatees consuming a natural diet and those individuals consuming a captivity based diet but some of the captive animals appeared to have greater conc entrations of these BCFA (Table 5 3). Romaine lettuce, t he main component of the diet offered in captivity is higher in protein than the aquatic vegetation free ranging manatees ingest [Siegal Willott et al., 2010], so this might explain why some of the captive manatees particularly the males from PA and CZ, appear ed to have greater

PAGE 125

125 concentration s These two manatees were receiving a greater variety of dietary ingredients every day in captivity, including cabbage kale, and other fruits and vegetables These supplemental ingredients contain more ami no acids than romaine lettuce. Some support for this hypothesis is provided by the apparent increase in conce ntration of isobutyrate when the female moved from Homosassa Springs to Cincinnati Zoo, and started consuming more diet ary ingredients (Figure 5 10). It would be beneficial to determine whether SCFA concentrations change similarly for other animals that may begin receiving more supplementary diet ary ingredients. It is difficult to make direct compari sons of fecal SCFA concentrations across spec ies and studies, as researchers use different methods to analyze and report SCFA concentrations [Millet et al., 2010]. Comparable numbers cannot be obtained without specific information that most do not report within the methods of their papers. Clauss et al. [2003] analyzed fecal SCFA using methods similar to those used in this study but reported their concentrations of acetate, propionate, and butyrate in mmol/L. When converted to mmol/g DM feces using info rmation provided by the author (M. Clauss, pers. comm.), a cetate and propionate fecal concentrations in captive Asian elephants ranged from 0.149 0.195 mmol/g DM feces and 0.035 0.048 mmol/g DM feces respectively [C lauss at el., 2003], which fall wit hin the range of free ranging CR manatees from this study. Butyrate concentrations averaged 0.009 0.014 mmol/g DM feces, which were generally lower than concentrations reported in the present study for manatees. ng relative, but exhibit a different digestive strategy. Manatees are known to have an extended digesta passage rate of six to ten days [Larkin et al., 2007], whereas elephants have digesta passage rates of around 21 55 hours [Clauss et al., 2003; Diere nfeld, 2006]. E lephants also typically ingest captive and natural diet s that have a great concentration of plant fiber than manatees, so it is surprising that fecal SCFA are similar for

PAGE 126

126 these animals. These results suggest that manatees and elephants may be similar in the net result of production and absorption of SCFA In conclusion, the results of this study show that free ranging and captive manatees exhibit differences in fecal short chain fatty acid concentrations, which may result from captive manat ees consuming less plant fiber and/or changes in microbial populations Additional sources of plant fiber may be considered when feeding primarily romaine lettuce. Rehabilitating manatees that were either injured or ill appeared to have greater SCFA conc entrations, which may be indicative of digestive malfunction and/or absorption problems The differences found amongst different groups of manatees in the present study suggest that fecal SCFA concentrations may provide managers and caretakers with inform ation about the digestive function of captive and rehabilitating manatees Determining fecal SCFA concentrations for manatees experiencing diet changes could be useful for manatees fed natural vegetation before their release back into the natural habitat and should be considered for future research For additional studies, it would also be beneficial to obtain samples from free ranging manatees in other geographic locations and include more facilities to better determine specific diet and/or microbial pop ulation differences Collecting samples from more rehabilitating manatees may also help to determine how different injuries and illnesses affect SCFA concentrations.

PAGE 127

127 Table 5 1 Environmental Parameters an d number of manatees sampled at each facility Fac ility Name Number of Manatees Salinity Water Temperature ( o C) Lowry Park Zoo 11 0 22.2 26.7 Parker Aquarium 1 0 31.1 Homosassa Springs 6* 0 23.3 Living Seas 4 33.9 ppt 25.8 Cincinnati Zoo 2* 0 26.7 Water temperature represents an average for the ye ar for each facility. *One of the long term, healthy manatees was originally housed at Homosassa Springs in Florida and then moved to Cincinnati Zoo. Samples were collected from her at both locations.

PAGE 128

128 Table 5 2. Health status, nature of injury, sex, lo cation, and date of collection of captive manatees from which fecal samples were obtained. Health Status Facility Sex Life Stage Injury/Illness Long term captive PA M Adult Long term captive HS F Adult Long term captive HS F Adult Long term captive HS F Adult Long term captive HS F Adult Long term captive HS F Adult Long term captive** HS F Adult Long term captive** CZ F Adult Recen tly rehabilitated CZ M Subadult CSS Recently rehabilitated LS M Subadult Orphaned Recently rehabilitated LS M Subadult CSS Recently rehabilitated LS M Adult WC Recently rehabilitated LS M Adult WC Rehabilitating LPZ M Independent calf Calf Rehabilitating LPZ M Subadult WC Rehabilitating LPZ M Subadult CSS Rehabilitating LPZ F Subadult CSS Rehabilitatin g LPZ F Subadult CSS Rehabilitating LPZ F Subadult Emaciation Rehabilitating LPZ F Adult WC Rehabilitating LPZ F Adult WC Rehabilitating LPZ F Adult CSS Rehabilitating LPZ F Adult WC Rehabilitating LPZ F Adult WC Injury or illness for which rehabili tating manatees were currently undergoing treatment or for which recently rehabilitated manatees were originally brought into captivity. LPZ = Lowry Park Zoo, HS = Homosassa Springs state park, PA = Parker Aquarium, LS = Living Seas, CZ = Cincinnati Zoo. CSS = cold stress syndrome, WC = watercraft injury. *This animal was weaned upon its arrival at the rehabilitation facility, due to its mother being unable to nurse and was observed eating plant material. **This female was originally housed at HS and the n transferred to CZ.

PAGE 129

129 Table 5 3. Lactic short chain fatty acid concentrations in mmol/g DM feces, and A:P ratio for free ranging manatees and manatees in captivity Free Ranging In Captivity P value Lactic Acid 0.001 + 0.001 0.002 + 0.002 0.110 Acet ic Acid 0.223 + 0.077 0.124 + 0.096 < 0.001 Propionic Acid 0.061 + 0.026 0.053 + 0.030 0.234 Butyric Acid 0.038 + 0.016 0.031 + 0.026 0.030 Isobutyric Acid 0.011 + 0.005 0.012 + 0.010 0.874 Isovaleric Acid 0.021 + 0.010 0.028 + 0.018 0.367 Valeric A cid 0.005 + 0.003 0.007 + 0.006 0.576 Total short chain fatty acids 0.359 + 0.115 0.259 + 0.172 0.003 A:P 4.1 + 2.0 2.4 + 1.0 < 0.001 Values are mean s +/ one standard deviation.

PAGE 130

130 Table 5 4. D ry matter content of feces from free ranging manatees and ca ptive manatees located at Lowry Park Zoo, Homosassa Springs, and Living Seas Number of Manatees % Dry Matter of Feces Free r anging (Crystal River) 25 23.8 + 6.1 Free r anging (Brevard County) 4 25.0 + 5.7 Rehabilitating (Lowry Park Zoo) 11 25.6 + 9.7 Long term healthy (Homosassa Springs) 6 22.5 + 3.0 Recently r ehabilitated (Living Seas) 4 27.5 + 8.8 Natural diet 29 23.9 + 5.9 Captiv ity based diet 12 22.9 + 6.5 Values are mean s +/ one standard deviation. There was no evidence of a difference in % dry matter of feces among different locations/groups and between manatees consuming different diets.

PAGE 131

131 Table 5 5. Lactic and short chain fatty acid concentrations in mmol/L fecal water from free ranging and captive manatees at various locations in Florida FR BC n = 4 FR CR n = 2 5 HS n = 6 LPZ n = 11 LS n = 4 PA n = 1 CZ n =1 Lactic 0.3 + 0.3 0.4 + 0.3 0.6 + 0.7 1.1 + 0.7 0.6 + 0.4 1.1 0 Acetic 85.6 + 8.4 59.9 + 20.2 49.8 + 28.1 48.3 + 20.5 17.4 + 5.7 26.6 22 Propionic 27.1 + 4.1 16.4 + 7.5 17.1 + 8.5 20.9 + 10.1 13.0 + 8.9 16.8 10.7 Butyric 13.9 + 1.2 10.0 + 3.9 11.9 + 7.4 17.6 + 10.6 4.9 + 3.1 12 5.1 Valeric 2.0 + 1.3 1.5 + 1.2 1.8 + 1.2 2.8 + 2.2 1.8 + 2.7 2.9 7.1 Isobutyric 5.1 + 0.4 2.7 + 1.3 3.6 + 2.1 3.3 + 1.5 1.8 + 0.8 3.9 6.2 Isovaleric 10. 1 + 2.0 5.6 + 2.4 8.0 + 5.7 11.0 + 5.3 8.3 + 1.0 8.6 3.3 Total short chain fatty acids 144.2 + 14.1 96.6 + 30.9 92.7 + 50.2 105.0 + 40.5 47.9 + 15.3 71.8 54.4 Values are mean s +/ one standard deviation. FR = free ranging, BC = Brevard County, CR = Crys tal River, HS = Homosassa Springs, LPZ = Lowry Park Zoo, LS = Living Seas, PA = Parker Aquarium, CZ = Cincinnati Zoo, SCFA = short chain fatty acids, BCFA = branched chain fatty acids.

PAGE 132

132 Figure 5 1. Lactic acid concentration in mmol per grams of dry mat ter (DM) feces from free ranging and captive manatees at each location The bottom of the box represents the 25 th percentile, the top of the box represents the 75 th percentile, and the whiskers extend to the minimum and the maximum. Dots represent the me dian values. The asterisk represents data from the single animal at Parker Aquarium, while the plus sign indicates data from the male at Cincinnati Zoo. FR = free ranging, CR = Crystal River, BC = Brevard County, HS = Homosassa Springs, LS = Living Seas. 0.000 0.002 0.004 0.006 0.008 0.010 0.012 FR-CR FR-BC LPZ HS LS Concentration (mmol/g DM Feces) Location +

PAGE 133

133 Figure 5 2. Acetic acid concentration in mmol per grams of dry matter (DM) from free ranging and captive manatees at each location The bottom of the box represents the 25 th percentile, the top of the box represents the 75 th percentile, and the whis kers extend to the minimum and the maximum. Dots represent the median values. Both free ranging (FR) groups had higher values than the Living Seas (LS) manatees The asterisk represents data from the single animal at Parker Aquarium, while the plus sign indicates data from the male at Cincinnati Zoo. CR = Crystal River, BC = Brevard County, HS = Homosassa Springs. 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 FR-CR FR-BC LPZ HS LS Concentration (mmol/g DM Feces) Location +

PAGE 134

134 Figure 5 3. Propionic acid concentration in mmol per grams of dry matter (DM) feces from free ranging and captive manatees at each loca tion The bottom of the box represents the 25 th percentile, the top of the box represents the 75 th percentile, and the whiskers extend to the minimum and the maximum. Dots represent the median values. The asterisk represents data from the single animal a t Parker Aquarium, while the plus sign denotes data from the male at Cincinnati Zoo. CR = Crystal River, BC = Brevard County, HS = Homosassa Springs. 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 FR-CR FR-BC LPZ HS LS Concentration (mmol/g DM Feces) Location +

PAGE 135

135 Figure 5 4 Butyric acid concentration in mmol per grams of dry matter (DM) feces from free ranging and captive manatees at each location The bottom of the box represents the 25 th percentile, the top of the box represents the 75 th percentile, and the whiskers extend to the minimum and the maximum. Dots represent the median values. The asterisk represe nts data from the single animal at Parker Aquarium, while the plus represents the male from Cincinnati Zoo. FR = free ranging, CR = Crystal River, BC = Brevard County, HS = Homosassa Springs, LS = Living Seas. 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 FR-CR FR-BC LPZ HS LS Concentration (mmol/g DM Feces) Location +

PAGE 136

136 Figure 5 5 Valeric acid concentration i n mmol per grams of dry matter (DM) feces from free ranging and captive manatees at each location The bottom of the box represents the 25 th percentile, the top of the box represents the 75 th percentile, and the whiskers extend to the minimum and the maxi mum. Dots represent the median values. The asterisk represents data from the single animal at Parker Aquarium, while the plus represents the male from Cincinnati Zoo. FR = free ranging, CR = Crystal River, BC = Brevard County, HS = Homosassa Springs, LS = Living Seas. 0.00 0.01 0.02 0.03 0.04 0.05 0.06 FR-CR FR-BC LPZ HS LS Concentration (mmol/g DM Feces) Location +

PAGE 137

137 Figure 5 6 Isobutyric acid concentration in mmol per grams of dry matter (DM) feces from free ranging and captive manatees at each location The bottom of the box represents the 25 th percentile, the top of the box represents the 75 th percentile, and the whiskers extend to the minimum and the maximum. Dots represent the median values. The asterisk represents data from the single animal at Parker Aquarium, while the plus sign represents the male from Cincinnati Zoo. CR = Crystal River, BC = Brevard County, HS = Homosassa Springs. 0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 FR-CR FR-BC LPZ HS LS Concentration (mmol/g DM Feces) Location +

PAGE 138

138 Figure 5 7 Isovaleric acid concentration in mmol per grams of dry matter (DM) feces from free ranging and captive manatees at each location Locations were found to be different from each other, but spe cific differences were not determined. The bottom of the box represents the 25 th percentile, the top of the box represents the 75 th percentile, and the whiskers extend to the minimum and the maximum. Dots represent the median values. The asterisk represen ts data from the single animal at Parker Aquarium, while the plus represents the male from Cincinnati Zoo. FR = free ranging, CR = Crystal River, BC = Brevard County, HS = Homosassa Springs, LS = Living Seas. 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 FR-CR FR-BC LPZ HS LS Concentration (mmol/g DM Feces) Location +

PAGE 139

139 Figure 5 8 Total SCFA concentration in mmol per grams of dry matter (DM) feces from free ranging and captive manatees at each location The bottom of the box represents the 25 th percentile, the top of the box represents the 75 th percentile, and the whiskers extend to the minimum and the maximu m. Dots represent the median values. The asterisk represents data from the single animal at Parker Aquarium, while the plus represents the male from Cincinnati Zoo. CR = Crystal River, HS = Homosassa Springs. 0.0 0.2 0.4 0.6 0.8 1.0 1.2 FR-CR FR-BC LPZ HS LS Concentration (mmol/g DM Feces) Location +

PAGE 140

140 Figure 5 9. Acetate to propionate (A: P) ratio from free ranging and captive manatees at each location This is an outlier in the free ranging (FR) Crystal River (CR) group. The bottom of the box represents the 25 th percentile, the top of the box represents the 75 th percentile, and the whi skers extend to the minimum and the maximum. Dots represent the median values. The asterisk represents data from the single animal at Parker Aquarium, while the plus represents the male from Cincinnati Zoo. BC = Brevard County, HS = Homosassa Springs, LS = Living Seas. 0 2 4 6 8 10 12 FR-CR FR-BC LPZ HS LS A: P Ratio Location +

PAGE 141

141 Figure 5 10. Indi vidual and total SCFA concentrations in mmol per gram of dry matte r (DM) feces for a single long term, healthy female managed first at Homosassa Springs and then transferred to Cincinnati Zoo. 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Concentration (mmol/g DM Feces) Homosassa Springs Cincinnati Zoo

PAGE 142

142 CHAPTER 6 MORPHOMETRIC BO DY CONDITION INDICES OF FREE RANGING FLORIDA MANATEES ( TRICHECHUS MANATUS L ATIROSTRIS ) Background The Florida manatee is a federa lly listed endangered species that it is protected under the Endangered Species Act of 1973, the Marine Mammal Protection Act of 1972 and the Florida Manatee Sanctuary Act of 1978. Manatees are susceptible to both anthropogenic and natural causes of injury, disease, and death. Among free ranging manatees, the inability to find sufficient food can result in weight loss, metabol ic compromise, illness and death, and could result in population decline. To prevent this from happening, injured and sick manatees are brought to facilities to receive care until they are deemed ready for release back into their natural habitat through t he manatee rehabilitation program (MRP). Manatees may be underweight when first brought to these facilities, but can become overweight over time because they are fed a diet composed mainly of romaine lettuce, other vegetables, and fruits, which are often greater in soluble sugars and lower in plant fiber than the natural diet [Siegal Willot t et al., 2010 ]. This excess energy intake could potentially lead to obesity and/or other health problems. The more dominant manatees in group housing may also be able t o consume a disproportionate amount of the available food and become overfed at the expense of less dominant ones. Currently, however, there is no objective method of assessing whether rehabilitating or long term captive manatees are over or underweight. One method of objectively assessing nutritional status is to compare morphometric measurements such as length (L) or height (H) with body weight (W) in healthy populations of a species to obtain a normal range of body condition index values [ Muchlisin et al., 2010] The relationship between length and weight is typic ally represented by Equation 6 1, where a

PAGE 143

143 represents the p roportionality constant and b represents the growth exponent, either isometric or allometric [ Biswas et al. 201 1]. W = a L b (6 1) The most familiar length weight relationship is the body mass index (BMI) used to assess the degree of obesity in human beings (Equation 6 2). BMI = W/H 2 (6 2) In fish species, a variant of Equation 6 1 has been used to obtain the condition index of K (Equa tion 6 has been well established to determine the state or well being of fish and is often used in the commerci al industry to evaluate growth [ Anderson and Neumann 1996; Stevenson and Woods 2006; Bisw as et al. 2011] K = W/L 3 (6 3) This length weight condition index has also been useful in determining differences in body condition during physiological time periods such as spawning and growth in several species of fish [Godinho, 1997; Yildirim et al. 2008; Iqbal and Suzuki, 2009; Percin and Akyol, 2009; Muchlisin et al., 2010; Biswas et al., 2011]. Others have used length girth relationships to estimate optimal net mesh size for minimizing catch of smaller fish [Stergiou and Karpouzi, 2003; Mendes e t al., 2006; Santos et al., 2006; Jawad et al., 2009]. The volume of an ellipsoid is proportional to length times maximum girth squared, so mass may be related to body volume estimated from girth and length in this way, assuming the body approximates an e llipsoid with a maximum girth at the umbilicus of mammals [Castellini and Calkins, 1993; Amaral et al., 2010]. Length weight relationships have also been evaluated in other species of aquatic animals includ ing sharks [Kohler et al., 1995] and marine mammal s [ Ridgway and Fenner, 1982; McBain, 2001; Perrin et al., 200 5], but there have been only a few studies of sirenians. Growth

PAGE 144

144 curves and length weight relationships have been determined for the Amazonian manatee ( Trichechus inunguis ) [Amaral et al., 2010; Vergara Parente et al., 2010], but a length weight relationship in the Florida manatee ( Trichechus manatus latirostris ) has only been reported for carcasses of dead manatees [Odell et al., 1978]. Fat, or blubber, measurements have also been used in conjun ction with length and weight measurements to develop body condition indices for Florida manatees [Ward Geiger, 1997], but obtaining fat measurements is not always possible. To date, the relationship between length and weight has not been used to establish a body condition index for live, healthy, free ranging Florida manatees. Thus, the purpose of this study was to compare length, weight, and girth measurements of free ranging manatees in order to define a normal range of body condition for the Florida ma natee. This study also sought to determine whether o r not condition indices remain constant as manatees increase in size and whether there were any differences between sexes or among locations. It was predicted that condition indices would not change wit h size, but that there would be differences between males and females, and among the three geographic locations. Materials and Methods Measurements were obtained from 146 free ranging manatees of all ages during annual winter health assessments at thr ee sites in Florida: in Crystal River ( N 28.891111, W 82.597222) from 2007 2011 (n = 22 females, 48 males); and, in the Indian River near Port St. John, Brevard County (N 28.483333 W 80.766667 ) from 2009 2010 (n = 8 females, 12 males), both hosted by the United States Geological Survey (USGS) Sirenia P roject; and off, Apollo Beach, Tampa Bay (N 27.79346, W 82.41871) (n = 33 females, 23 males), hosted by the Florida Fish and Wildlife Conservation Commission. All lengths were measured in meters using an ope n reel tape measure. Straight length (SL) and curvilinear length (CL) were measured over the entire length of the animal, from the tip of the snout to the end of the tail paddle. The tape measure was laid

PAGE 145

145 flat over the animal for CL, while it was held an for SL, with t he middle of the tape measure l measurements were taken at the umbilicus for a maximum (MG) by wrapping the tape around the circumference of the a nim al at this site. Body weight ( W) was measured (in kg) using a large hanging scale (Crystal River: Model EDjunior, Dillon 2500lbs/1000kg, Data Weighing Systems, Inc., Elk Grove, IL; Brevard and Tampa: Model MSI 7200, Dyna link 2000lbs/1000kg, Measuremen t Systems International Inc., Seattle, WA). Sex and animal condition were also noted. The head veterinarian present at the health assessments determined the condition of the animal on a scale of 1 5, with a score of one representing an emaciated animal a nd five representing an obese animal. Manatees were also assigned a qualitative descriptor ranging from poor to excellent. Manatees were only included in the analysis if they had a visual condition score of three or four, representing normal to fat and a qualitative descriptor of fair, good, or excellent. Any manatees with abnormal health parameters or noted as being potentially or actually pregnant were not included. All free ranging manatees were presumed to be consuming a natural diet of both fresh a nd salt water aquatic plants. Statistical comparisons were performed using a computerized program (SAS for W indows 9.3, SAS Institute, Inc., Cary, NC). Figures were prepared using SigmaPlot (Version 12.2, Systat Software, San Jose, CA ). Data distribu tions were first assessed for normality both visually and using the Shapiro Wilk test. Any data sets that were not normally distributed were log transformed before analysis. Three different constants were considered as potential objective measures of bod y condition (Equations 6 4, 6 5, 6 6). BCI 1 = K SL =W/SL 3 (6 4) BCI 2 = K CL =W/CL 3 (6 5)

PAGE 146

146 BCI 3 = K SLMG =W/(SLxMG 2 ) (6 6) Indices were compar ed using a general linear model procedure with sex and location as factors, with an interaction between the two also considered. To establish whether these body condition indices changed with age (with length being an indicator of age), each index was regressed against SL, CL, and SL, respectively. Where body condition changed with length, the logarithm of W was regressed against the logarithm of SL and CL, and the logarithm of W/SL was regressed against the logarithm of MG to establish an exponent that would produce body condition indices that would accommodate changes in body density and shape during growth. Th e slopes of these regression lines were compared between sexes. Post hoc, the logarithm of SL was also regressed against the logarithm of MG and the slopes compared between sexes using an ANCOVA, to determine whether males were more streamlined than femal es. Lastly, amended indices of body condition using the new exponents were compared among the different locations using another general linear model procedure. A probability of error less than 5% was considered significant when rejecting the null hypothe sis. Results Straight le ngths of the manatees in the present study ranged from 1.47 to 3.23 m, whereas curvilinear le ngths ranged from 1.62 to 3.49 m. Umbilical girths of individuals ranged from 1.13 to 2.85 m, and body weights ranged from 77 to 751 kg. All three of the body condition indices, K SL K CL and K SL MG were greater in females than males (p < 0.0001, p < 0.000 3, and p = 0.01, respectively) (Table 6 1). T here was no difference evident among locations (Tab le 6 1) but there was a trend towards a location effect for K SL (p = 0.1), and t here was no inter action between sex and location Body condition indices, K SL and K CL did not change with length in females (Figures 6 1 and 6 2), but decreased with increasing length in males (p < 0.0001) (Figure s 6 4 and 6 5). The third body condition index, K SL MG decreased with increasing length

PAGE 147

147 in both females and males (p = 0.002 and p = 0.004, respectively) (Figures 6 3 and 6 6). Slopes of the regression for logarithm of weight versus logarithm of length w ere different between males and females for both SL and CL (p = 0.0004 and p = 0.0006), but there was no difference between sexes in the slopes of the regressions for the logarithm of W/SL against MG (p = 0.11) (Table 6 2). The slopes of the regression l ines for the first two condition indices are not significantly different from 3 for females, because the 95% confidence limits of the slope encompass 3.0 for both SL and CL (Table 6 2) (Figure 6 7). However, the slope is significantly less than 3.0 for ma les for the first two condition indices, because the 95% confidence limits of the slope do not encompass three (Table 6 2) (Figure 6 7). Maximum girth increases less with increasing length in males than in females (p < 0.001) (Figure 6 8). Slopes of the regression equations were thus used to produce amended body condition indices to account for changing shape and density in growing male manatees (Table 6 3). Crystal River females had greater values for both BCI 4 and BCI 5 than Brevard (p = 0.03) and the re was a trend for them to have greater BCI 4 and 5 values than for Tampa females (p = 0.05). No other location differences were noted for either sex. Discussion Morphometric data from free ranging, healthy manatees at three different locations was use d to establish normal ranges of condition indices for the different sexes of free ranging Florida manatees and at three different locations in Florida. These indices can be used to help evaluate the nutritional status of rehabilitating and long term capti ve manatees. It would be appropriate to use the original K SL and K CL as indice s of body condition for all sizes of females, but not for all sizes of males. Body weight increased proportionately with the cube of body length in female manatees, but increas ed with an exponent of length less than 3 in males. In this

PAGE 148

148 way, females are more similar to other aquatic species, such as fish, and males are more similar to terrestrial animals, whose body weight typically scales to length with an exponent of less than three, due to the scaling of skeletal mass (M skel ) to body mass (M b ) The skeletal mass of aquatic species usually scales proportionately to body mass, with an exponent of o ne, because the water supports their weight, reducing the effects of gravity [Rey nolds, 1977; Schmidt Nielsen, 1984]. Skeletal mass of terrestrial animals, however, tends to scale out of proportion w ith body mass (Equation 6 7), because their skeleton and limbs must support their weight [Schmidt Nielsen, 1984]. M skel = 0.0608*M b 1.0 83 (6 7) M skel = 0.0207*M b 1.130 (6 8) Domning and DeBuffrenil [1991] found that skeletal mass increases positively in proportion to body mass for Florida manatees (Equation 6 8), making them similar to terrestrial animals. This is an unusual finding for an aquatic species, but may be due to the fact that thirty of the fifty manatees analyzed were males, which were found to have an exponent of less than three for the length weight relationship in this study. Moreover, most of the manatees included in the analysis were calves and juveniles, and having a greater proportion of younger animals can bias the scaling of skeletal mass to body mass. When the immature specimens were excluded from the analysis of scaling skeletal mass to body mass for six species o f whales, the exponent decreased from 1.107 to 1.024 [Schmidt Nielsen, 1984]. Females in this study were found to maintain a constant body shape as length and weight increase, whereas males become slimmer as length increases. A potential reason for why males becomes slimmer is that free ranging males have been shown to have significantly thinner blubber than females [Ward Geiger, 1997]. Males are also probably more streamlined and/or

PAGE 149

149 hydrodynamic than females because they need to chase and outcompete ot her males for a chance to mate, whereas females may be wider as they grow longer because of the maternal reproductive investment [Bonnet et al., 2010] K CL (BCI 2) and K SL (BCI 1) may be the best indices to use for females, because they have the lowest va riance and a re easy to calculate The normal range of the first condition index would be from 17.1 25.9 kg/m 3 whereas the normal range for the second condition index for females would be from 14.6 19.9 kg/m 3 (Table 6 1). The amended second index (B CI 8) may be the best to use for males (Table 6 3) because it has the lowest variance of the three, producing a normal range of 19.5 25.6 kg/m 3 The original third body condition constant, K SL MG is less ideal to use for both females and males, becaus e the slope of the regression line for MG is less than two and has high variance. However, the amended indices could be used with the adjust ed exponents, because they have the lowest coefficient of variations of the amended indices (Table 6 3). In a stud y similar to the present one, g irth was found to not be a reliable predictor of body condition in Steller sea lion pups, with the weight and length based condition index as the only one able to distinguish between pups of differing health condition starv ing vs. non starving [Trites and Jonker, 2000]. In this way, the length weight body condition indices established in the present study for manatees may also reveal differences between underweight and healthy individuals. Both Steller sea lions and manate es have thinner layers of blubber than other marine mammals, which may account for why girth is not useful for estimating body condition. As these animals lose body weight, the girth would not proportionally change [Trites and Jonker, 2000]. Girth may al so fluctuate with water intake and/or gut contents [Willemsen and Hailey, 2002], which would cause variability in these condition indices.

PAGE 150

150 An equation to describe the length weight relationship in Amazonian manatees was calculated using mass in kg and len gth in cm (Equation 6 9) [Amaral et al., 2010]. W = 1.070*10 5 x SL 3.122 (6 9) Amazonian ma natees are the smallest of the s irenians, with a maximum reported weight of 379 kg [Amaral et al., 2010]. The higher exponent, or slope of the regression, foun d i n that study compared to the present study may have resulted from the fact that the majority of the data used was collected fro m captive individuals. The diet offered to captive Amazonian manatees is similar to that for captive Florida manatees, which is nutritionally different from a natural diet, and may therefore potentially influence body mass and condition Amazonian manatee males and females were also not found to be significantly different from each other, whereas Florida manatees were found to be different in this study. Amazonian manatees are not known to exhibit sexual dimorphism [Rosas, 1994], but Florida manatee females are generally regarded as being larger in size and heavier than males [Reep and Bonde, 2006]. Among fish, higher values of condition indices for one habitat compared to another can sometimes indicate that the former habitat is a more favorable environment than another [Muchlisin et al. 2010]. It is possible that CR provides a better habitat than BC or TB because BCI 4 and B CI 5 were greater for females in this area. Crystal River is also further north than TB and the Indian River in BC The Crystal River population generally consists of residents that either do not leave the area throughout the year, or individuals that tra vel further north during the summer. These manatees may compensate for cooler water temperatures by increasing body fat stores. Alternatively, a greater body condition index value could indicate that an animal is in poor health, such as when an animal is obese or has trapped air in their abdominal cavity from an injury or illness [Willemsen and Hailey, 2002]. No abnormal findings were noted during health

PAGE 151

151 assessments, however so the manatees reported here were considered healthy. Moreover, obesity is le ss likely to occur in the free ranging manatees and trapped air would only affect the condition index if girth was being used. Some of the larger CR females in the present study may also have been pregnant, even though every effort was made to exclude suc h individuals from the data set based on health assessment parameters. A potential limitation of the study presented here is that weights and lengths were measured only during the winter season. During the winter, food sources may be limited, so manatees may weigh less or the temperature may be cooler. Seasonal differences in condition have been noted previously for fish species [Yildirim, 2008; Iqbal and Suzuki, 2009]. Unfortunately, manatees are usually more scattered during the non winter mo nths and therefore more difficult to catch for health assessments. The condition indices for free ranging, healthy Florida manatees established here may provide a normal range for manatees that are part of the rehabilitation program in order to obtain qua ntitative estimate s of the degree to which individuals ma y be depleted of energy stores, or can be used to examine data for long term captive manatees to monitor how their condition changes over time. Figures 6 9 and 6 10 provide examples of using BCI 7 f or two long term, healthy captive males, but more data would need to be obtained regarding diet and energy intake at the time of these measurements to discuss why the first manatee appeared underweight initially, but has since remained in the normal range, and why the second manatee increased above normal range, but then decreased back just within normal range. These length weight relationship values may additionally be useful as baseline data to observe potential changes in the overall population of free ranging Florida manatees in years to come. For instance, if resources become more limited, an overall decline in body condition

PAGE 152

152 could be observed. Long term data from ringed seals ( Phoca hispida ) revealed that correlated with more severe ice years and lower food availability [Harwood et al., 2000]. Female polar bears ( Ursus maritimus ) similarly had poor er body condition when the ice they use as habitat broke up earlier than usual in the reproductive seaso n [Stirling et al., 1999]. The prese nt study found that three separate geographical populations of Florida manatees differ in values of body condition indices, with males and females also found to be different. Thus, this study provides baseline information that can be used by ecologists an d managers to quantitatively assess condition of rehabilitating and captive manatees

PAGE 153

153 Table 6 1. Body condition constants for free ranging manatees of each sex at three locations in Florida Females Males p value K SL 21.45 (17.1 25.9) 19.94 (15. 6 24.2) < 0.0001 K CL 17.23 (14.6 19.9) 16.41 (13.7 19.2) < 0.0003 K SL MG 37.83 (30.5 45.2) 36.44 (32.7 40.2) 0.01 K SL K CL and K SL MG represent the ratio in kg/m 3 of body weight to the cube of straight length, cube of curvilinear length, and straight length times the square of the maximum girth, respectively. Crystal River, Brevard County, and Tampa Bay were combined, because there was no evidence of a difference among locations (p = 0.1). Values are means with the normal range, defined as +/ two times the standard deviation from the mean, in parentheses

PAGE 154

154 Table 6 2. Intercepts and slopes of regression lines of the logarithm of weight against logarithm of morphometric measurement of free ranging manatees of each sex for all locations. F emales Males Measurement SL CL MG SL CL MG Intercept of Log W 3.140 2.886 3.811 3.392 3.113 3.703 Antilog of Intercept 23.10 17.93 45.21 29.73 22.48 50.55 Slope 2.915 2.957 1.717 2.578 2.689 1.835 Lower CI of slope 2.768 2.845 1.596 2. 460 2.587 1.760 Upper CI of slope 3.063 3.070 1.838 2.696 2.792 1.910 The measurements SL, CL and MG are the straight length, curvilinear length and the maximum girth of manatees, respectively. Intercept and slope describe the equation of the regression line for each of the measurements with log measurement on the x axis and log weight on the y axis. the 95% confidence intervals for each slope. As an example the equation of the regression line for SL for females would be: log wt = 3.14 + 2.915 x log SL or wt = 23.1 x SL 2.915

PAGE 155

155 Table 6 3. Amended body condition constants of free ranging manatees of different sexes at three locations in Florida CV (%) Females BCI 4 = W/SL 2.915 CR: 24.2 BC: 22.1 TB: 22.9 Combined: 23.2 (19 29.4) (17.1 27.1) (18.9 26.9) (18.5 27.9) CR: 10.8 BC: 11.5 TB: 8.7 Combined: 10.2 BCI 5 = W/CL 2.957 CR: 23.2 BC: 21.2 TB: 22.0 Combined: 22.3 (18.2 28.2) (16.4 26) (18.1 25.9) (17.7 26.9) CR: 10.7 BC: 11.3 TB: 8.8 Combined: 10.2 BCI 6 = W/(SLxMG 1.717 ) 45.4 (37.8 53) 8.4 Males BCI 7 = W/SL 2.578 29.8 (24.9 34.7) 8.1 BCI 8 = W/CL 2.689 22.5 (19.4 25.6) 6.8 BCI 9 = W/(SLxMG 1.835 ) 40.6 (36.9 44 .3) 4.5 Amended body condition constants are ratios of body weight (W) to either straight length (SL), curvilinear length (CL), or SL multiplied by maximum girth (MG) with amended exponents for each measurement and sex. Values are for means with normal r anges, representing +/ two times the standard deviation from the mean, in parentheses. The coefficient of variation (CV) for each ratio is also shown. Mean body condition constants using SL and CL were different (p < 0.05) for female manatees from Cryst al River (CR) and Brevard County (BC), and there was a statistical trend (p = 0.05) for these constants from manatees at Tampa Bay (TB) to be different from those from manatees at CR. The values for all three locations for other constants are combined bec ause there was no evidence of a difference among locations.

PAGE 156

156 Figure 6 1. Regression of body condition constant, K SL (Body condition index 1), against straight length (SL) for all females. There was no evidence that BCI 1 changes with length (p = 0.31), suggesting that it is a suitable condition index for females of all lengths.

PAGE 157

157 Figure 6 2. Regression of body condition constant, K CL (Body condition index 2), against curvilinear length ( CL ) for all females. BCI 2 does not change with length (p = 0. 54), suggesting that it is a suitable condition index for females of all lengths.

PAGE 158

158 Figure 6 3. Regression of body condition constant, K SLMG (Body condition index 3), against straight length ( SL ) for all females. BCI 3 decreases with length (p = 0.002 ), suggesting that it is not a suitable condition index for females of all lengths.

PAGE 159

159 Figure 6 4. Regression of body condition constant, K SL (Body condition index 1), against straight length (SL) for all males. K SL decreases with length (p < 0.0001), suggesting that it is not a suitable condition constant for males of all lengths.

PAGE 160

160 Figure 6 5. Regression of body condition constant, K C L (Body condition index 2), against curvilinear length ( CL ) for all males. BCI 2 decreases with length (p < 0.000 1), suggesting that it is not a suitable condition index for males of all lengths.

PAGE 161

1 61 Figure 6 6. Regression of body condition constant, K SLMG (Body condition index 3), against straight length ( SL ) for all males. BCI 3 decreases with length (p = 0.004), suggesting that it is not a suitable condition index for males of all lengths.

PAGE 162

162 Figure 6 7. Regression of th e logarithm of weight against straight length for females (solid circles) and males (open circles). The slope of the line for females is highe r for females (2.915) than for males (2.578) (p < 0.001)

PAGE 163

163 Figure 6 8. Regression of the logarithm of maximum girth ( MG ) against logarithm of straight length ( SL ) for females (solid circles) and males (open circles). Slope for males was significantly less than that for the females (p < 0.001). Regression equation for females : MG = 1.0658*SL 0.1497, R 2 = 0.90. Regression equation for males : MG = 0.844*SL 0.067, R 2 = 0.89.

PAGE 164

164 Figure 6 9. Condition constants (kg/m 3 ) of BCI 7 (W/SL 2.578 ) calcula ted from historical morphometric measurements for a long term, healthy, captive male. At the time of the first measurement, this manatee was a little less than three years old. Generally, condition constants for this male stay ed within the normal range. 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 Condition constant (kg/m3) Month and Year

PAGE 165

165 Figure 6 10. Condition constants (kg/m 3 ) of BCI 7 (W/SL 2.578 ) calculated from historical morphometric measurements for a second long term, healthy, captive male. At the time of the first measurement, this manatee was a little less than four years ol d. Condition constants for this male fluctuated more with time, but it is unknown why his values were above the normal range for several years. 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0 50.0 Condition constant (kg/m3) Month and Year

PAGE 166

166 CHAPTER 7 CONCLUSIONS Stu dying Nutrition in the Florida M anatee The Florida manatee is a charismatic marine mammal that has garnered attention from its endangered status and interaction with humans Determining conservation needs is vital for the continued survival of a threatened or enda ngered species The biology and physiology of Florida manatees is rather unique and warrants further investigation to enable the agencies that regulate policies regarding manatees to make more informed decisions about their management. Manatees have slow reproductive rates and generally only produce one offspring at a time, ev ery 2 4 years, so parental investment in terms of time and energy spent on the calf, is high. They are also subject to population decline from habitat destruction, disease, and injury, so it is important that research persists. Nutrition is one aspect of manatee biology that not been greatly studied Most nutrition al research has been performed using domesticated and farm animals, which requires relatively unrestricted access to animals and the ability to manipulate living conditions and diet These conditions do not apply to an endangered, aquatic mammal and therefore, studying such a Working with captive individuals that are part of the Manatee Rehabilitation Program (MRP) provided access to manatees in a more c ontrolled setting and allowed for adequate sample collection periods. This research was time consuming, however and not all captive facilities were able to participate or provide data because of limited staff availability In the winter, free ranging manatees in Florida congregate around warm water sources, such as springs. The winter season is the only time when manatees are seen in large groups that facilitate fecal sample collection Unfortunately, manatees eat less during the winter, because food sources are not necessarily located near the warm water sources and thus, produce

PAGE 167

167 less feces. As a result, the number of samples obtained was less than desired and resulted in some of the studies in the present project having a low statistical power to d etect differences among groups. However, the results and conclusions still provide novel information about Florida manatee nutrition. The primary objective for this entire project was to evaluate intestinal function and how it relates to nutrition in Flo rida manatees to provide information about potential deficiencies and suggest impr ovements for the current captivity based diet. The general conclusions of each chapter are thus summarized and presented here to show how these goals were accomplished. Chap ter Conclusions Chapter 2: Quantitative Survey of the Diet of Florida Manatees at Captive Facilities in Florida This chapter describes the first quantitative documentation of the foods fed to captive Florida manatees and provides an estimate of their maint enance digestible energy requirements. The foods fed to a total of fourteen animals residing at three different facilities were documented. Nutrient and dry matter concentrations of the foods were determined using published values. Digestible energy of the diets offered to captive manatees was estimated using an equation designed for horses, but this equation may not have provided the best estimates of true values Digestible energy intake in kcal/day was similar for adult and subadult manatees but mea n digestible energy intake per kg of body weight of subadults was twice the mean of a dults. This is not surprising because subadults are considered to be growing whereas adults are not The estimated caloric energy requirements of the manatees surveyed here were not very different among individuals despite the manatees having a wide range of body weights. B oth age classes were consuming more digestible energy than would be expected from calculated needs, but measuring intake and having a detailed diet history provide better estimates of energy needs than

PAGE 168

168 calculations. However, c onsuming more energy than is necessary could lead to obesity if not monitored carefully. Males consumed more nutrients and energy than females, but there was no evidence of sea sonal differences in nutrient or energy consumption. It was not clear why males consumed more, but lower consumption coul d be related to hormonal cycles in the females such as the onset of sexual maturation or coming into heat Each of the captive manate e facilities fed a slightly different diet, with variation in the amount and type of each dietary ingredient offered All of the captivity based diets, however, resulted in manatees con suming a greater proportion of NSC and CP which are the more digestib l e portions of the diet, and lower proportions of NDF and ADF Most herbivores rely on the fermentation of plant fiber for the majority of their energy needs, so the same is assumed for manatees. If manatees are able to obtain most of their d ietary energ y from NSC, the need for fermentation may be less. Foods containing more plant fiber may need to be added to the diet of captive manatees, however, to reduce energy consumption that could lead to obesity and insulin resistance and to increase fermentation that could contribute to a more healthy gut. Chapter 3: In Vivo Apparent Digestibi lity Trials of Captive Florida M anatees T his study presents the first available data on the digestibility of different nutrients in the captivity based diet offered to liv ing Florida manatees in different facilities Food and fecal samples were collected from five different manatees during four week digestibility trials. Sample collection from a sixth long term healthy animal at a third facility was attempted several time s, but never completed due to the animal having infrequent and irregular bowel movements. In general, the captive manatees in this study were found to be more irregular than predicted, so the amount of fecal material collected was less than planned. It w as therefore not possible to run all of the nutrient analyses tha t were originally proposed.

PAGE 169

169 Manate es that consumed a d iet containing more ingredients appeared to have lower mean digestibility values than those eating a diet comprised of almost only r omain e lettuce. Lower digestibility could result from the greater amount of plant fiber in the diet containing more ingredients, but may likely be due to greater DM intakes. Manatee s at both facilities, however, had greater apparent digestibility of DM, NDF, ADF, and CP in their diets than fellow hindgut fermenters, including the horse and elephant These greater digestibility values may result from longer digesta passage times but are likely due to the consumption of different diets whereby h orses and elephants ingest diets that are higher in NDF and ADF There was no evidence found for seasonal differences in digestibility for any of the manatees in this study, but this may be indicative of the consistent environmental parameters maintained at the facilities and less seasonal variation in the produce captive manatees consumed than would be found in the natural diet Acid insoluble ash was found to be an unreliable internal marker for measuring apparent digestibility of nutrients by Florida m anatees, but ADL was useful as an internal marker. It was unclear why AIA was found in lower quantities in the feces than in the food, but may result from manatees somehow absorbing it within the digestive tract or a dilution of AIA by endogenous secretio ns into the digestive tract However, t his phenomenon has not been noted in any other species The overall conclusion of this chapter is that Florida manatees are capable of thoroughly digesting a captivity based diet of romaine lettuce, other vegetables, and fruits. High digestibility values are not always desirable, however and can result from individuals consuming a diet that is higher in non structural carbohydrates and lower in plant fiber, which was confirmed in chapter 2 of this work. However, mo re digestible nutrients may be beneficial for rehabilitating animals that are underweight upon their arrival to primary care facilities.

PAGE 170

170 Chapter 4: In Vitro Fermentation Estimates for Free ranging, Captive, and Rehabilitating Florida Manatees This chapte r provides the first description of the use of a rotary incubator to evaluate how well manatees ferment different types of food in vitro without having to alter manatee feeding regimes in vivo It was assumed that fecal microbes were representative of hi ndgut microbes in manatees, and feces have been used successfully with many other species for in vitro fermentation experiments. Manatees were all at least two hours away from where the incubator was located, so d ifferent methods of preserving fecal micro bes were tested in a preliminary trial using horse feces. An inoculum prepared from a fecal sample preserved by vacuum packing and keeping the fecal sample warm for three hours resulted in similar loss of NDF during fermentation in vitro compared to an in oculum prepared from feces immediately after collection. In vitro fermentation assays were successfully completed with the feces of thirteen different manatees P ercent DM lost and % NDF lost were not different b etween captive Homosassa Springs (HS) man atees and Crystal River (CR) free ranging manatees for any of the foods tested including romaine lettuce, alfalfa hay, Vallisneria americana Syringodium filiforme and Ruppia maritima but were different among the foods Rehabilitating manatees from Lowr y Park Zoo (LPZ) appeared to have higher values of % DM lost and % NDF lost which may be due to a greater amount of plant fiber in their diet, but statistical comparisons were not performed because samples were only collected from two manatees The HS ma natees are maintained in a natural spring head, however so it is possible that they and free ranging manatees derive similar microbia l populations from the environment. The most fermentable food was romaine lettuce, followed by Vallisneria Syringodium a nd alfalfa. Ruppia was the least fermentable The percent loss corresponds to the differing amounts of NDF in the foods however with Ruppi a having the greatest concentration of NDF

PAGE 171

171 and lettuce the least Overall values of % DM lost and % NDF lost for the grasses and alfalfa were low, but this may be due to the relatively short incubation period of 48 hours compared to the extended digesta passage rate of manatees Thus, m anatees may be capable of further fermenting these grasses and alfalfa in vivo Short chain fatty acids (SCFA) primarily acetate, propionate, and butyrate, and branched chain fatty acids (BCFA) were produced during each of the fermentation trials confirming microbia l activity. The CR manatees had significantly greater absolute inc rease s of acetate, propionate and total SCFA plus BCFA concentrations and a smaller increase in lactic acid concentrations than the HS manatees Greater production of SCFA could result from differences in the hindgut microbial population that result from the increased plant fiber concentration of the diet of free ranging manatees Variation between incubations of control digesta samples was generally low (CV < 10%), but greater variation for two of the grasses between incubations was likely due to incons istency of sample s in the filter bags. The results found in this chapter, however reveal that in vitro incubations are useful for comparing the fermentation of di fferent foods by fecal microbes obtained from manatees Moreover, t he DM and % NDF digestib ility values of romaine lettuce found in this chapter using in vitro fermentation assays are similar to the in vivo digestibility values found in chapter 3. Future experiments should assess the fermentation of more potential captivity based diet ingredien ts and other types of natural vegetation if possible. Different lengths of incubation time with fecal inoculums, both shorter and longer than 48 hours, should also be assessed to determine whether or not DM and NDF loss change over time within the rotary incubator system. Based on a study using horse feces as the inoculum for in vitro fermentations with the same rotary incubator [Earing et al., 2010] to detect a difference among three different

PAGE 172

172 incubation time periods for DM loss from a type of food usin g an ANOVA with power of 0.95 there would need to be inoculums from at least three manatees for each time period. Chapter 5: Fecal Short Chain Fatty Acid and Lactic Acid Concentrations of Florida Manatees This chapter provides the first available data on fecal short chain (SCFA) and branched chain fatty acid (BCFA) concentrations in living Florida manatees. SCFA and BCFA concentrations were previously measured in digesta samples taken from the digestive tract of manatee carcasses, but it is unknown how de ath may affect SCFA concentrations. Manatees likely obtain most o f their dietary energy from the SCFA and BCFA that are produced from the fermentation of carbohydrates and protein by mi crobes located in their hindgut Disease state and diet have both bee n found to affect the type and amount of SCFA and B CFA produced, so measuring the physiological concentrations of SCFA c health and nutrition. Fecal samples were collected from eleven rehabilitating manatees seven long term healthy captive manatees five recently rehabilitated manatees and twenty nine free ranging manatees Differences in SCFA concent rations were detected among different groups of free ranging and captive manatees and between manatees consuming a natural vs. captivity based diet Rehabilitating manatees appeared to have the widest range and highest total SCFA concentrations, and Living Seas (LS) manatees had some of the lowest SCFA concentrations. The greater total SCFA concentrations in the reh abilitating manatees most likely resulted from malabsorption of nutrients, whereas the lower total SCFA concentrations found for LS manatees may be related to these individuals consuming less plant fiber in their diet Lactate concentrations appeared to b e greater whereas acetate to propionate ratios were lower in captive manatees These findings may result from the greater proportion of non structural c arbohydrates

PAGE 173

173 found in the captivity based diet as identified in chapter 2. Greater concentrations of BCFA in the feces of captive manatees may result from greater protein and amino acid concentrations in the produce consumed by these manatees Manatees and elephants have simil ar fecal SCFA concentrations as determined using amended data fr om Clauss et al. [2003], which wa s surprising because they have different digestive strategies. Fecal samples were collected from a single long term, healthy female captive manatee at two different locations, where she consumed two different diets the more recent di et contained more dietary ingredients These two fecal samples appeared to have different SCFA and BCFA concentrations, and this finding provides support for the hypothesis that diet may affect fecal concentrations. However, more research needs to be don e to determine whether or not this is true. A future experiment measuring SCFA concentrations in the feces of manatees switching diets would require at least 6 individuals to detect a difference at a statistical power of 0.95 using a paired t test Chapt er 6: Morphometric Body Condition Indices of Free Ranging Florida Manatees Length, weight, and maximum girth measurements were used to determine body condition indices for manatees in this last chapter. Morphometric measurements were obtained for 146 free ranging manatees at three different geographic locations in Florida : Crystal River (CR), Brevard County (BC), and Tampa Bay (TB) Three indices were considered: one based on straight length (K SL =W/SL 3 ), one based on curvilinear length (K CL =W/CL 3 ), and th e last one using maximum girth and straight length [K SLMG =W/(SLxMG 2 )] to relate mass to body volume. F emales had greater body condition index values than males and there was a trend towards a location effect. The first two condition indices did not chang e with increasing length in females, but decreased with increasing length in males. The third body condition also de creased with increasing length in both males and females.

PAGE 174

174 Thus, r egression analyses were performed to determine alternative exponents for amended condition indices that accounted for changing shape and density as manatees increase in size T he first two original indices are appropriate for all sizes of females, whereas amended indices are appropriate for all sizes of males Males become sl immer as length increases, but this may result from them needing to be more streamlined than females, because they have to pursue female manatees for a chance to mate. Using the amended indices, it was found that CR females weighed more relative to length than BC manatees and the re was a trend for them to be heavier than TB manatees in this respect It is unclear why there are geographic differences, but these differences may have result ed from habitat and climate differen ces among the three locations T he condition indices that were established in this chapter can now be used to determine quantitative estimates of body condition for rehabilitating and long term captive manatees. Manatees that fall below the normal ranges established here are likely to b e depleted of energy stores, whereas manatees that have condition index values above the normal range may be considered overweight, or even obese. Moreover, using these index ranges as baseline data may also be helpful in identifying potential population level changes in the body condition of free ranging Florida manatees in the future. General Conclusions and Future Directions The results of this dissertation provided information about the basic nutrition of Florida manatees. Plant f iber is important for herbivores and thus, would be assumed to be important for manatees as well, but may not b e offered in adequate amounts to captive manatees The quality of the diet being offered can be assessed using the body condition indices determined in the sixth ch apter. In this way, a diet would be considered appropriate, regardless of ingredients values while consuming said diet.

PAGE 175

175 It is hoped that the results found here w ill encourage captive manatee facilities to more closely monitor food intake and perform more frequent morphometric measurements when possible. It would be particularly useful to weigh younger manatees before and after specific periods of intake to confir m growth and more accurately estimate growing energy needs. More in vivo digestibility trials, paired with measurement of digestible energy consumption, are necessary to learn more about the energy needs of captive manatees and to attempt to develop a dig estible energy equation for manatees. However, distinguishing differences among individuals consuming different diets will continue to prove a challenge, because there are only a couple of manatees maintained at each facility and it will be difficult to t ell whether or not differences result from the diet offered or from facility differences. Feeding trials should also be performed with other sources of plant fiber to test for palatability. Additionally, in vitro incubation experiments could be run to det ermine the fermentation of th ese additional plant fiber sources Obtaining adequate amounts of the natural vegetation manatees consume is not p ossible, so some potential dietary ingredients that could provide more plant fiber include other types of leafy greens, high fiber primate biscuits, and hays. Managers have previously had concerns regarding the use of dried hays in the captive manatee diet, but manatee fecal microbes were able to ferment alfalfa hay similarly to grasses consumed by free ranging man atees in vitro Alternatively, some zoos have designed high fiber gel matrices for their herbivorous animals that are more palatable than some of the other options. This may be possible for manatees as well, and should therefore be invest igated. Feeding a diet that has a greater concentration of plant fiber could also be more cost e ffective for some facilities, because dry matter intake decreases with increasing plant fiber content. However, increasing plant fiber content in the diet could also result i n manatees requiring more water.

PAGE 176

176 Florida manatees continue to persevere even though they face both anthropogenic and natural threats. Rehabilitating these manatees in captive facilities has contributed to their success and will continue to be important f or their survival. Manatees have be en successfully returned to their natural habitat following rehabilitation, but it is difficult to monitor their nutritional status after their release. In conclusion, t he research of this dissertation provides informat ion about the nutrition of Florida manatees, and will hopefully contribute to the overall success of the rehabilitation program.

PAGE 177

177 LIST OF REFERENCES Aagnes TH, Blix AS, Mathiesen SD. 1996. Food intake, digestibility and rumen fermentation in reindeer fed b aled timothy silage in summer and winter. Journal of Agricultural Science 127:517 523. Aagnes TH, Sormo W, Mathiesen SD. 1995. Ruminal microbial digestion in free living, in captive lichen fed, and in starved reindeer (Rangifer tarandus) in winter. Applied and Environmental Microbiology 61(2):583 591. Abdouli H, Ben Attia S. 2007. Evaluation of a two stage in vitro technique for estimating digestibility of equine feeds using horse faeces as the source of microbial inoculum. Animal Feed Science and Technolog y 132(1 2):155 162. Adesogan AT. 2002. What are feeds worth?: A critical evaluation of selected nutritive value methods. 13th Annual Florida Ruminant Nutrition Symposium. Gainesville, FL. p 33 47. Aketa K, Asano S, Wakai Y, Kawamura A. 2001. Apparent diges tibility of eelgrass in dugongs (Dugong dugon). Honyurui Kagaku 41(1):23 34. Aketa K, Asano S, Wakai Y, Kawamura A. 2003. Apparent digestibility of eelgrass (Zostera marina) by captive dugongs Dugong dugon in relation to the nutritional content of eelgrass and dugong feeding parameters. Mammal Study 28(1):23 30. Allen ME, Oftedal OT, Earle KE, Seidensticker J, Vilarin L. 1995. Do maintenance energy requirements of felids reflect their feeding strategies? 1st Annual conference of the AZA nutrition advisory g roup. Toronto, Ontario, Canada. p 97 103. Amaral RS, da Silva VMF, Rosas FCW. 2010. Body weight/length relationship and mass estimation using morphometric measurements in Amazonian manatees Trichechus inunguis (Mammalia: Sirenia). Marine Biodiversity Recor ds 3:Unpaginated. Anderson RO, Neumann RM. 1996. Length, weight, and associated structural indices. In: Murphy BR, Willis DW, editors. Fisheries techniques, Second edition: American Fisheries Society {a}, Suite 110, 5410 Grosvenor Lane, Bethesda, Maryland 20814 2199, USA. p 447 482. Anguita M, Canibe N, Perez JF, Jensen BB. 2006. Influence of the amount of dietary fiber on the available energy from hindgut fermentation in growing pigs: Use of cannulated pigs and in vitro fermentationi. Journal of Animal Sci ence 84(10):2766 2778. Argenzio RA, Southwor.M, Stevens CE. 1974. Sites of organic acid production and absorption in equine gastrointestinal tract. American Journal of Physiology 226(5):1043 1050. Banta CA, Clemens ET, Krinsky MM, Sheffy BE. 1979. Sites of organic acid production and patterns of digesta movement in the gastrointestinal tract of dogs. Journal of Nutrition 109(9):1592 1600.

PAGE 178

178 Barboza PS. 1993. Digestive strategies of the wombats feed intake, fiber digestion, and digesta passage in two grazing marsupials with hindgut fermentation. Physiological Zoology 66(6):983 999. Bengtson JL. 1983. Estimating food consumption of free ranging manatees in Florida. Journal of Wildlife Management 47(4):1186 1192. Bergero D, Miraglia N, Abba C, Polidori M. 2004. Apparent digestibility of Mediterranean forages determined by total collection of faeces and acid insoluble ash as internal marker. Livestock Production Science 85(2 3):235 238. Best RC. 1981. Foods and feeding habits of wild and captive Sirenia. Mammal R eview 11(1):3 29. Bezeau LM. 1965. Effect of source of inoculum on digestibility of substrate in in vitro digestion trials. Anim Sci 24:823 825. Biswas G, Sundaray J, Thirunavukkarasu A, Kailasam M. 2011. Length weight relationship and variation in conditi on of Chanos chanos (Forsskal, 1775) from tide fed brackishwater ponds of the Sunderbans India. Indian Journal of Geo Marine Sciences 40(3):386 390. Bjorndal KA. 1979. Cellulose digestion and volatile fatty acid production in the Green turtle, Chelonia m ydas. Comparative Biochemistry and Physiology a Physiology 63(1):127 133. Bjorndal KA. 1980. Nutrition and grazing behavior of the Green turtle Chelonia mydas. Marine Biology 56(2):147 154. Bjorndal KA. 1987. Digestive efficiency in a temperate herbivorous reptile, Gopherus polyphemus. Copeia(3):714 720. Bonnet X, Delmas V, El Mouden H, Slimani T, Sterijovski B, Kuchling G. 2010. Is sexual body shape dimorphism consistent in aquatic and terrestrial chelonians? Zoology 113(4):213 220. Bossart GD, Meisner RA, Rommel SA, Ghim S j, Jensen AB. 2003. Pathological features of the Florida manatee cold stress syndrome. Aquatic Mammals 29(1):9 17. Bugaut M. 1987. Occurrence, absorption and metabolism of short chain fatty acids in the digestive tract of mammals. Compar ative Biochemistry and Physiology B Biochemistry & Molecular Biology 86(3):439 472. Burn DM. 1986. The digestive strategy and efficiency of the West Indian manatee, Trichechus manatus. Comparative Biochemistry and Physiology a Physiology 85(1):139 142 Burn DM, Odell DK. 1987. Volatile fatty acid concentrations in the digestive tract of the West Indian manatee, Trichechus manatus. Comparative Biochemistry and Physiology B Biochemistry & Molecular Biology 88(1):47 49.

PAGE 179

179 Caroline J, Gross MS, Spillman D M. 2003. Fiber digestion in mammals. Pakistan Journal of Biological Sciences 6(17):1564 1573. Castellini MA, Calkins DG. 1993. Mass estimates using body morphology in Steller sea lions. Marine Mammal Science 9(1):48 54. Chen HM, Lifschitz CH. 1989. Preparation of fecal samples for assay of volatile fa tty aci d s by gas liquid chromatography and high performance liquid chromatography. Clinical Chemistry 35(1):74 76. Cherney DJR, Sicilianojones J, Pell AN. 1993. Forage in vitro dry matter digestibility as influenced by fiber source in the donor cow diet. Journal of Animal Science 71(5):1335 1338. Chivers DJ. 1989. Adaptations of digestive systems in non ruminant herbivores. Proceedings of the Nutrition Society 48(1):59 67. Clauss M, Hummel J. 2005. The digestive perform ance of mammalian herbivores: why big may not be that much better. Mammal Review 35(2):174 187. Clauss M, Loehlein W, Kienzle E, Wiesner H. 2003. Studies on feed digestibilities in captive Asian elephants (Elephas maximus). Journal of Animal Physiology and Animal Nutrition:160 173. Clauss M, Polster C, Kienzle E, Wiesner H, Baumgartner K, von Houwald F, Streich WJ, Dierenfeld E. 2005. Energy and mineral nutrition and water intake in the captive Indian rhinoceros (Rhinoceros unicornis). Zoo Biology 24(1):1 1 4. Clauss M, Schwarm A, Ortmann S, Streich WJ, Hummel J. 2007. A case of non scaling in mammalian physiology? Body size, digestive capacity, food intake, and ingesta passage in mammalian herbivores. Comparative Biochemistry and Physiology a Molecular & Int egrative Physiology 148(2):249 265. Clauss M, Wilkins T, Hartley A, Hatt JM. 2009. Diet composition, food intake, body condition, and fecal consistency in captive Tapirs (Tapirus spp.) in UK collections. Zoo Biology 28(4):279 291. Clemens ET, Maloiy GMO. 1 983. Nutrient digestibility and gastrointestinal electrolyte flux in the elephant and rhinoceros. Comparative Biochemistry and Physiology a Physiology 75(4):653 658. Coles LT, Moughan PJ, Darragh AJ. 2005. In vitro digestion and fermentation methods, inclu ding gas production techniques, as applied to nutritive evaluation of foods in the hindgut of humans and other simple stomached animals. Animal Feed Science and Technology 123:421 444.

PAGE 180

180 Dawes CJ, Lawrence JM. 1980. Seasonal changes in the proximate constitu ents of the seagrasses Thalassia testudinum, Halodule wrightii, and Syringodium filiforme. Aquatic Botany 8(4):371 380. Dehority BA. 2002. Gastrointestinal tracts of herbivores, particularly the ruminant: Anatomy, physiology and microbial digestion of plan ts. Journal of Applied Animal Research 21(2):145 160. Denek N, Polat E, Koncagul S, Can A. 2008. The determination of incubation time and amount of faecal content of horse faeces as an inoculum source for digestibility determination of forages with in vitr o procedure. Journal of Animal and Veterinary Advances 7(6):698 702. Dierenfeld ES. 2006. Nutrition. In: Fowler, ME, Mikota, SK, editor s Biology, Medicine, and Surgery of Elephants. Oxford, UK: Blackwell Publishing Ltd. Domning DP, Debuffrenil V. 1991. Hy drostasis in the Sirenia quantitative data and functional interpretations. Marine Mammal Science 7(4):331 368. Earing JE, Cassill BD, Hayes SH, Vanzant ES, Lawrence LM. 2010. Comparison of in vitro digestibility estimates using the Daisy(II) incubator wi th in vivo digestibility estimates in horses. Journal of Animal Science 88(12):3954 3963. Edwards MS, Ullrey DE. 1999. Effect of dietary fiber concentration on apparent digestibility and digesta passage in non human primates. II. Hindgut and foregut ferme nting folivores. Zoo Biology 18(6):537 549. Etheridge K, Rathbun GB, Powell JA, Kochman HI. 1985. Consumption of aquatic plants by the West Indian manatee. Journal of Aquatic Plant Management 23(JAN):21 25. Evans E, Miller DS. 1968. Comparative nutrition, growth and longevity. Proceedings of the Nutrition Society 27(02):121 129. Fahey GC, Jung HG. 1983. Lignin as a marker in digestion studies A review. Journal of Animal Science 57(1):220 225. Ferguson MJ, Jones GP. 2000. Production of short chain fatty ac ids following in vitro fermentation of saccharides, saccharide esters, fructo oligosaccharides, starches, modified starches and non starch polysaccharides. Journal of the Science of Food and Agriculture 80(1):166 170. Fernandes J, Rao AV, Wolever TMS. 2000 Different substrates and methane producing status affect short chain fatty acid profiles produced by in vitro fermentation of human feces. Journal of Nutrition 130(8):1932 1936. Fleming SE, Rodriguez MA. 1983. Influence of dietary fiber on fecal excretio n of volatile fatty acids by human adults. Journal of Nutrition 113(8):1613 1625.

PAGE 181

181 Fowler ME, Mikota SK. 2006. Biology, medicine, and surgery of elephants / [edited by] Murray E. Fowler, Susan K. Mikota. 1st ed. ed. Ames, Iowa: Blackwell Pub. Gallivan GJ, B est RC. 1980. Metabolism and respiration of the Amazonian manatee (Trichechus inunguis). Physiological Zoology 53(3):245 253. Godinho A. 1997. Weight length relationship and condition of the characiform Triportheus guentheri. Environmental Biology of Fishe s 50(3):319 330. Goto M, Ito C, Yahaya MS, Wakai Y, Asano S, Oka Y, Ogawa S, Fruta M, Kataoka T. 2004a. Characteristics of microbial fermentation and potential digestibility of fiber in the hindgut of dugongs (Dugong dugon). Marine and Freshwater Behaviour and Physiology 37(2):99 107. Goto M, Ito C, Yahaya MS, Wakamura K, Asano S, Wakai Y, Oka Y, Furuta M, Kataoka T. 2004b. Effects of age, body size and season on food consumption and digestion of captive dugongs (Dugong dugon). Marine and Freshwater Behavio ur and Physiology 37(2):89 97. Goto M, Watanabe A, Karita S, Tokita N, Yamamoto Y, Wakaki Y, Asano S, Oka Y, Furuta M. 2008. Nutrient and energy consumption of captive mature dugong (Dugong dugon) consuming eelgrass at the Toba Aquarium. Marine and Freshwa ter Behaviour and Physiology 41(3):169 177. Guevara P, Claeys T, Janssens GPJ. 2008. Apparent digestibility in meat type guinea pigs as determined by total collection or by internal marker. Veterinarni Medicina 53(4):203 206. Guilloteau P, Martin L, Eeckha ut V, Ducatelle R, Zabielski R, Van Immerseel F. 2010. From the gut to the peripheral tissues: the multiple effects of butyrate. Nutrition Research Reviews 23(2):366 384. Hackenberger MK. 1987. Diet digestibilities and ingesta transit times of captive Asia n and African elephants. Canada: University of Guelph. Hansen B. 1989. Determination of nitrogen as Elementary N, an alternative to Kjeldahl. Acta Agriculturae Scandinavica 39(2):113 118. Hartman DS. 1979. Ecology and behavior of the manatee (Trichechus ma natus) in Florida. Special Publication American Society of Mammalogists:1 153. Harwood LA, Smith TG, Melling H. 2000. Variation in reproduction and body condition of the ringed seal (Phoca hispida) in western Prince Albert Sound, NT, Canada, as assessed th rough a harvest based sampling program. Arctic 53(4):422 431. Hassel DM, Spier SJ, Aldridge BM, Watnick M, Argenzio RA, Snyder JR. 2009. Influence of diet and water supply on mineral content and pH within the large intestine of horses with enterolithiasis. Veterinary Journal 182(1):44 49.

PAGE 182

182 Hastie PM, Mitchell K, Murray J. 2008. Semi quantitative analysis of Ruminococcus flavefaciens, Fibrobacter succinogenes and Streptococcus bovis in the equine large intestine using real time polymerase chain reaction. Brit ish Journal of Nutrition 100(3):561 568. Hatt JM, Clauss M. 2006. Feeding Asian and African elephants Elephas maximus and Loxodonta africana in captivity. International Zoo Yearbook 40:88 95. Hatt JM, Clauss M, Gisler R, Liesegang A, Wanner M. 2005. Fiber digestibility in juvenile Galapagos tortoises (Geochelone nigra) and implications for the development of captive animals. Zoo Biology 24(2):185 191. Hervas G, Frutos P, Giraldez FJ, Mora MJ, Fernandez B, Mantecon AR. 2005. Effect of preservation on ferment ative activity of rumen fluid inoculum for in vitro gas production techniques. Animal Feed Science and Technology 123:107 118. Hill RC. 2006. Challenges in measuring energy expenditure in companion animals: A clinician's perspective. Journal of Nutrition 1 36(7):1967S 1972S. Holden LA. 1999. Comparison of methods of in vitro dry matter digestibility for ten feeds. Journal of Dairy Science 82(8):1791 1794. Hook SE, Wright AG, McBride BW. 2010. Methanogens: Methane Producers of the Rumen and Mitigation Strateg ies. Archaea 2010:11. Hume ID. 1997. Fermentation in the hindgut of mammals. In: Mackie RI, White BA, editors. Gastrointestinal Microbiology: Gastrointestinal ecosystems and fermentations. New York, NY: Chapman and Hall. p 84 115. Hume ID. 2002. Digestive strategies of mammals. Acta Zoologica Sinica 48(1):1 19. Imoto S, Namioka S. 1978. VFA production in pig large intestine. Journal of Animal Science 47(2):467 478. Iqbal K, Suzuki H. 2009. Length weight relationships and condition factor of the Japanese sil ver biddy, Gerres equulus, in the Yatsushiro Sea, western Kyushu, Japan. Journal of Applied Ichthyology 25(2):219 222. Irvine AB. 1983. Manatee metabolism and its influence on distribution in Florida. Biological Conservation 25(4):315 334. Jawad LA, McKenz ie A, Al Noor SS. 2009. Relationship between opercular girth, maximum girth and total length of fishes caught in gillnets in the estuarine and lower river sections of Shatt al Arab River (Basrah Province, Iraq). Journal of Applied Ichthyology 25(4):470 473

PAGE 183

183 Jung HG, Mertens DR, Payne AJ. 1997. Correlation of acid detergent lignin and Klason lignin with digestibility of forage dry matter and neutral detergent fiber. Journal of Dairy Science 80(8):1622 1628. Jung HJG, Varel VH, Weimer PJ, Ralph J. 1999. Accur acy of Klason lignin and acid detergent lignin methods as assessed by bomb calorimetry. Journal of Agricultural and Food Chemistry 47(5):2005 2008. Jurgens MH. 2002. Animal feeding and nutrition. 9th ed. ed. Dubuque, Iowa: Kendall/Hunt Pub. Kentucky Equine Research. 2007 Subclinical Acidosis: Is Your Horse At Risk? Equinews:6 9. Kleiber M. 1975. Metabolic turnover rate physiological meaning of metabolic rate per unit of body weight. Journal of Theoretical Biology 53(1):199 204. Kohler NE, Casey JG, Turner PA. 1995. Length weight relationships for 13 species of sharks from the western North Atlantic. Fishery Bulletin 93(2):412 418. Kristensen NB, Harmon DL. 2004. Effect of increasing ruminal butyrate absorption on splanchnic metabolism of volatile fatty aci ds absorbed from the washed reticulorumen of steers. Journal of Animal Science 82(12):3549 3559. Krisztina M, Kristf K, Tulassay E, Ivnyi Z, Erdei A, Kirly A, Gl J, Bajtay Z. 2011. Mucosal immunity and the intestinal microbiome in the development of cr itical illness. ISRN Immunology. p 12. Langer P. 1988. The mammalian herbivore stomach: comparative anatomy function and evolution. Stuttgart: Gustav Fischer. Lanyon JM, Marsh H. 1995. Digesta passage times in the dugong. Australian Journal of Zoology 43(2 ):119 127. Larkin ILV, Fowler VF, Reep RL. 2007. Digesta passage rates in the Florida manatee (Trichechus manatus latirostris). Zoo Biology 26(6):503 515. Lattimer JM, Cooper SR, Freeman DW, Lalman DL. 2007. Effect of yeast culture on in vitro fermentation of a high concentrate or high fiber diet using equine fecal inoculum in a Daisy II incubator. Journal of Animal Science 85(10):2484 2491. Linn JG, Staba EJ, Goodrich RD, Meiske JC, Otterby DE. 1975. Nutritive value of dried or ensiled aquatic plants. 1. c hemical composition. Journal of Animal Science 41(2):601 609. Lomolino MV, Ewel KC. 1984. Digestive efficiencies of the West Indian manatee Trichechus manatus. Florida Scientist 47(3):176 179.

PAGE 184

184 Lowman RS, Theodorou MK, Hyslop JJ, Dhanoa MS, Cuddeford D. 199 9. Evaluation of an in vitro batch culture technique for estimating the in vivo digestibility and digestible energy content of equine feeds using equine faeces as the source of microbial inoculum. Animal Feed Science and Technology 80(1):11 27. Mabjeesh SJ Cohen M, Arieli A. 2000. In vitro methods for measuring the dry matter digestibility of ruminant feedstuffs: Comparison of methods and inoculum source. Journal of Dairy Science 83(10):2289 2294. Mackie RI. 1997. Gut environment and evolution of mutualist ic fermentative digestion. In: Mackie RI, White BA, editors. Gastrointestinal Microbiology: Gastrointestinal ecosystems and fermentations. New York, NY: Chapman and Hall. p 13 38. Mackie RI. 2002. Mutualistic fermentative digestion in the gastrointestinal tract: Diversity and evolution. Integrative and Comparative Biology 42(2):319 326. Mackie RI, Rycyk M, Ruemmler RL, Aminov RI, Wikelski M. 2004. Biochemical and microbiological evidence for fermentative digestion in free living land iguanas (Conolophus pal lidus) and marine iguanas (Amblyrhynchus cristatus) on the Galapagos archipelago. Physiological and Biochemical Zoology 77(1):127 138. Marmontel M. 1995. Age and reproduction in female Florida manatees. In: O'Shea TJ, Ackerman BB, Percival HF, editors. Pop ulation Biology of the Florida manatee. Washington, D.C.: U.S. Department of the Interior, National Biological Service. p 98 119. Marshall CD, Kubilis PS, Huth GD, Edmonds VM, Halin DL, Reep RL. 2000. Food handling ability and feeding cycle length of manat ees feeding on several species of aquatic plants. Journal of Mammalogy 81(3):649 658. McBain JF. 2001. Cetacean medicine. CRC handbook of marine mammal medicine Second edition: CRC Press. p 895 906. McDaniel AL, Martin SA, McCann JS, Parks AH. 1993. Effect s of Aspergillus oryzae fermentation extract on in vitro equine cecal fermentation. Journal of Animal Science 71(8):2164 2172. McEwan EH, Whitehead PE. 1970. Seasonal changes in energy and nitrogen intake in reindeer and caribou. Canadian Journal of Zoolog y 48(5):905 907. McNab BK. 1980. Food habits, energetics, and the population biology of mammals. American Naturalist 116(1):106 124. McNab BK. 2002. The physiological ecology of vertebrates: A view from energetics. Ithaca, NY: Comstock Publishing Associate s. Mendes B, Fonseca P, Campos A. 2006. Relationships between opercula girth, maximum girth and total length of fish species caught in gillnet and trammel net selectivity surveys off the Portuguese coast. Journal of Applied Ichthyology 22(3):209 213.

PAGE 185

185 Merri tt AM, Smith DA. 1980. Osmolarity and volatile fatty acid content of feces from horses with chronic diarrhea. American Journal of Veterinary Research 41(6):928 931. Meyer RM, Bartley EE, Julius F, Fina LR. 1971. Comparison of 4 in vitro methods for predict ing in vivo digestibility of forages. Journal of Animal Science 32(5):1030 1037. Millet S, Van Oeckel MJ, Aluwe M, Delezie E, De Brabander DL. 2010. Prediction of in vivo short chain fatty acid production in hindgut fermenting mammals: problems and pitfall s. Critical Reviews in Food Science and Nutrition 50(7):605 619. Miraglia N, Bergero D, Bassano B, Tarantola M, Ladetto G. 1999. Studies of apparent digestibility in horses and the use of internal markers. Livestock Production Science 60(1):21 25. Mountfor t DO, Campbell J, Clements KD. 2002. Hindgut fermentation in three species of marine herbivorous fish. Applied and Environmental Microbiology 68(3):1374 1380. Muchlisin Z, Musman M, Azizah M. 2010. Length weight relationships and condition factors of two t hreatened fishes, Rasbora tawarensis and Poropuntius tawarensis, endemic to Lake Laut Tawar, Aceh Province, Indonesia. Journal of Applied Ichthyology 26(6):949 953. Murray J, Scott B, Hastie PM. 2009. Fermentative capacity of equine faecal inocula obtained from clinically normal horses and those predisposed to laminitis. Animal Feed Science and Technology 151(3 4):306 311. Murray RM, Marsh H, Heinsohn GE, Spain AV. 1977. Role of midgut cecum and larg e intestine in digestion of sea grasses by dugong (Mammalia Sirenia). Comparative Biochemistry and Physiology a Physiology 56(1):7 10. National Research Council Committee. 2007. Nutrient requirements of horses. 6th rev. ed. ed. Washington, D.C: National Academies Press. Nijboer J, Dierenfeld ES. 1996. Comparison o f diets fed to southeast Asian colobines in North American and European zoos, with emphasis on temperate browse composition. Zoo Biology 15(5):499 507. Odell DK, Forrester DJ, Asper ED. A preliminary analysis of organ weights and sexual maturity in the Wes t Indian manatee (Trichechus manatus); 1978; Orlando. Florida Department of Natural Resources. p 52 65. Ogawa K, Ben RA, Pons S, Depaolo MIL, Fernandez LB. 1992. Volatile fatty acids, lactic acid, and pH in the stools of breast fed and bottle fed infants. Journal of Pediatric Gastroenterology and Nutrition 15(3):248 252. Ortiz RM, Worthy GAJ. 2004. Could lower body fat mass contribute to cold water susceptibility in calves of the West Indian manatee (Trichechus manatus)? Marine Mammal Science 20(1):176 183.

PAGE 186

186 Oshio S, Tahata I, Minato H. 1987. Effect of diets differing in ratios of roughage to concentrate on microflora in the rumen of heifers. Journal of General and Applied Microbiology 33(2):99 111. Owen Smith RN. 1988. Megaherbivores : the influence of very large body size on ecology. Cambridge ; New York: Cambridge University Press. Pagan JD. 1998. Measuring the digestible energy content of horse feeds. In: Pagan JD, editor. Advances in equine nutrition. Nottingham, UK: Nottingham University Press. p 71 76. Pardini AT, O'Brien PCM, Fu B, Bonde RK, Elder FFB, Ferguson Smith MA, Yang F, Robinson TJ. 2007. Chromosome painting among Proboscidea, Hyracoidea and Sirenia: Support for Paenungulata (Afrotheria, Mammalia) but not Tethytheria. Proceedings of the Royal S ociety B Biological Sciences 274(1615):1333 1340. Parissi ZM, Papachristou TG, Nastis AS. 2005. Effect of drying method on estimated nutritive value of browse species using an in vitro gas production technique. Animal Feed Science and Technology 123:119 12 8. Parker JD, Duffy JE, Orth RJ. 2001. Plant species diversity and composition: experimental effects on marine epifaunal assemblages. Marine Ecology Progress Series 224:55 67. Parra R. 1978. Comparison of foregut and hindgut fermentation in herbivores. In: Montgomery GC, editor. The Ecology of Arboreal Folivores. Washington, D.C.: Smithsonian Institution Press. p 205 229. Pendlebury C, Odongo N, Renjifo A, Naelitz J, Valdes E, McBride B. 2005. Acid insoluble ash as a measure of dry matter digestibility in c aptive African elephants (Loxodonta africana). Zoo Biology:261 265. Percin F, Akyol O. 2009. Length weight and length length relationships of the bluefin tuna, Thunnus thynnus L., in the Turkish part of the eastern Mediterranean Sea. Journal of Applied Ich thyology 25(6):782 784. Perrin WF, Dolar MLL, Chan CM, Chivers SJ. 2005. Length weight relationships in the spinner dolphin (Stenella longirostris). Marine Mammal Science 21(4):765 778. Popovich DG, Dierenfeld ES. 1997. Gorilla nutrition. In: Ogden J, Whar ton D, editors. Management of Gorillas in Captivity: Husbandry Manual, Gorilla Species Survival Plan: American Association of Zoos and Aquariums. Provancha JA, Hall CR. 1991. Observations of associations between seagrass beds and manatees in East Central F lorida USA. Florida Scientist 54(2):87 98. Rasmussen HS, Holtug K, Mortensen PB. 1988. Degradation of amino acids to short chain fatty acids in humans an in vitro study. Scandinavian Journal of Gastroenterology 23(2):178 182.

PAGE 187

187 Reep R, Bonde RK. 2006. The Florida manatee : biology and conservation. Gainesville: University Press of Florida. Reich KJ, Worthy GAJ. 2006. An isotopic assessment of the feeding habits of free ranging manatees. Marine Ecology Progress Series 322:303 309. Remis MJ, Dierenfeld ES. 20 04. Digesta passage, digestibility and behavior in captive gorillas under two dietary regimens. International Journal of Primatology 25(4):825 845. Remis MJ, Dierenfeld ES, Mowry CB, Carroll RW. 2001. Nutritional aspects of western lowland gorilla (Gorilla gorilla) diet during seasons of fruit scarcity at Bai Hokou, Central African Republic. International Journal of Primatology 22(5):807 836. Reynolds JE, Rommel SA. 1996. Structure and function of the gastrointestinal tract of the Florida manatee, Trichechu s manatus latirostris. Anatomical Record 245(3):539 558. Reynolds WW. 1977. Skeleton weight allometry in aquatic and terrestrial vertebrates. Hydrobiologia 56(1):35 37. Ridgway SH, Fenner CA. 1982. Weight length relationships of wild caught and captive Atl antic Bottlenose dolphins. Journal of the American Veterinary Medical Association 181(11):1310 1315. Robinson PH, Mathews MC, Fadel JG. 1999. Influence of storage time and temperature on in vitro digestion of neutral detergent fibre at 48 h, and comparison to 48 h in sacco neutral detergent fibre digestion. Animal Feed Science and Technology 80(3 4):257 266. Rosas FCW. 1994. Biology, conservation and status of the Amazonian manatee Trichechus inunguis. Mammal Review 24(2):49 59. Rowland IR, Mallett AK, Wise A. 1985. The effect of diet on the mammalian gut flora and its metabolic activities. Crc Critical Reviews in Toxicology 16(1):31 103. Rubsamen K, Hume ID, Vonengelhardt W. 1982. Physiology of the Rock hyrax. Comparative Biochemistry and Physiology a Physi ology 72(2):271 277. Sakaguchi E. 2003. Digestive strategies of small hindgut fermenters. Animal Science Journal 74(5):327 337. Sales J, Janssens GPJ. 2003. Acid insoluble ash as a marker in digestibility studies: a review. Journal of Animal and Feed Scien ces 12(3):383 401. Santos MN, Canas A, Lino PG, Monteiro CC. 2006. Length girth relationships for 30 marine fish species. Fisheries Research 78(2 3):368 373. Sato H. 2009. Increased fecal lactate and decreased volatile fatty acid (VFA), particularly n Buty rate concentrations in diarrheic young calves. Journal of Veterinary Medical Science 71(1):117 119.

PAGE 188

188 Sato H. 2010. Fecal acetone and isopropanol levels of wild sika deer (Cervus nippon yesoensis) in spring and winter. Grassland Science 56(1):56 58. Schmidt D, Kerley M, Porter J, Dempsey J. 2005a. Structural and nonstructural carbohydrate, fat, and protein composition of commercially available, whole produce. Zoo Biology:359 373. Schmidt DA, Dempsey JL, Kerley MS, Porton IJ. 2001. Fiber in ape diets: a review The apes: challenges for the 21st century Conference proceedings:177 179. Schmidt DA, Kerley MS, Porton IJ, Porter JH, Dempsey JL, Griffin ME, Ellersieck MR, Sadler WC. 2005b. Fiber digestibility by black lemurs (Eulemur macaco macaco). Journal of Zoo an d Wildlife Medicine 36(2):204 211. Schmidt Nielsen K. 1984. Scaling. Why is animal size so important? New York: Cambridge University Press. Scholander PF, Irving L. 1941. Experimental investigations on the respiration and diving of the Florida manatee. Jou rnal of Cellular and Comparative Physiology 17(2):169 191. Schulte Hostedde AI, Zinner B, Millar JS, Hickling GJ. 2005. Restitution of mass size residuals: validating body condition indices. Ecology 86(1):155 163. Scott GBD, Keymer IF. 1975. Ulcerative col itis in apes: a comparison with the human disease. Journal of Pathology 115(4):241 244. Settle RG. 1988. Short chain fatty acids and their potential role in nutritional support. Journal of Parenteral and Enteral Nutrition 12(6):S104 S107. Shimomura Y, Sato H. 2006. Fecal D and L lactate, succinate, and volatile fatty acid levels in young dairy calves. Journal of Veterinary Medical Science 68(9):973 977. Siegal Willott JL, Harr K, Hayek LAC, Scott KC, Gerlach T, Sirois P, Reuter M, Crewz DW, Hill RC. 2010. Proximate nutrient analyses of four species of submerged aquatic vegetation consumed by Florida manatee (Trichechus manatus latirostris) compared to Romaine lettuce (Lactuca sativa var. longifolia). Journal of Zoo and Wildlife Medicine 41(4):594 602. Sims JA, Parsons JL, Bissell HA, Sikes RS, Ouellette JR, Rude BJ. 2007. Determination of bamboo diet digestibility and fecal output by giant pandas. Ursus 18(1):38 45. Soergel KH. 1994. Colonic fermentation metabolic and clinical implications. Clinical Invest igator 72(10):742 748. Spanghero M, Berzaghi P, Fortina R, Masoero F, Rapetti L, Zanfi C, Tassone S, Gallo A, Colombini S, Ferlito JC. 2010. Technical note: Precision and accuracy of in vitro digestion of neutral detergent fiber and predicted net energy of lactation content of fibrous feeds. Journal of Dairy Science 93(10):4855 4859.

PAGE 189

189 Stahl M, Osmann C, Ortmann S, Kreuzer M, Hatt JM, Clauss M. 2011. Energy intake for maintenance in a mammal with a low basal metabolism, the giant anteater (Myrmecophaga tridac tyla). Journal of Animal Physiology and Animal Nutrition. doi: 10.1111/j.1439 0396.2011.01226.x. Stark AH, Madar Z. 1993. In vitro production of short chain fatty acids by bacterial fermentation of dietary fiber compared with effects of those fibers on hep atic sterol synthesis in rats. Journal of Nutrition 123(12):2166 2173. Stergiou KI, Karpouzi VS. 2003. Length girth relationships for several marine fishes. Fisheries Research 60(1):161 168. Stevens CE, Hume ID. 1998. Contributions of microbes in vertebrat e gastrointestinal tract to production and conservation of nutrients. Physiological Reviews 78(2):393 427. Stevenson RD, Woods WA, Jr. 2006. Condition indices for conservation: new uses for evolving tools. Integrative and Comparative Biology 46(6):1169 119 0. Stirling I, Lunn NJ, Iacozza J. 1999. Long term trends in the population ecology of polar bears in western Hudson Bay in relation to climatic change. Arctic 52(3):294 306. Sunvold GD, Fahey GC, Merchen NR, Bourquin LD, Titgemeyer EC, Bauer LL, Reinhart GA. 1995a. Dietary fiber for cats in vitro fermentation of selected fiber sources by cat fecal inoculum and in vivo utilization of diets containing selected fiber sources and their blends. Journal of Animal Science 73(8):2329 2339. Sunvold GD, Fahey GC, Merchen NR, Reinhart GA. 1995b. In vitro fermentation of selected fibrous substrates by dog and cat fecal inoculum influence of diet composition on substrate organic matter disappearance and short chain fatty acid production. Journal of Animal Science 73 (4):1110 1122. Sunvold GD, Fahey GC, Merchen NR, Titgemeyer EC, Bourquin LD, Bauer LL, Reinhart GA. 1995c. Dietary fiber for dogs .4. In vitro fermentation of selected fiber sources by dog fecal inoculum and in vivo digestion and metabolism of fiber supple mented diets. Journal of Animal Science 73(4):1099 1109. Sunvold GD, Hussein HS, Fahey GC, Merchen NR, Reinhart GA. 1995d. In vitro fermentation of cellulose, beet pulp, citrus pulp, and citrus pectin using fecal inoculum from cats, dogs, horses, humans, a nd pigs and ruminal fluid from cattle. Journal of Animal Science 73(12):3639 3648. Szylit O, Maurage C, Gasqui P, Popot F, Favre A, Gold F, Borderon JC. 1998. Fecal short chain fatty acids predict digestive disorders in premature infants. Journal of Parent eral and Enteral Nutrition 22(3):136 141. Thayer GW, Bjorndal KA, Ogden JC, Williams SL, Zieman JC. 1984. Role of larger herbivores in seagrass communities. Estuaries 7(4A):351 376.

PAGE 190

190 Tilley JMA, Terry RA. 1963. A two stage technique for the in vitro digesti on of forage crops. Jour Brit Grassland Soc 18((2)):104 111. Trites AW, Jonker RAH. 2000. Morphometric measurements and body condition of healthy and starveling Steller sea lion pups (Eumetopias jubatus). Aquatic Mammals 26(2):151 157. Troyer K. 1984a. Beh avioral acquisition of the hindgut fermentation system by hatchling Iguana iguana. Behavioral Ecology and Sociobiology 14(3):189 193. Troyer K. 1984b. Microbes, herbivory and the evolution of social behavior. Journal of Theoretical Biology 106(2):157 169. Tyndale Biscoe H. 2005. Life of marsupials. Second edition ed. Australia: Csiro Publishing. U.S. Department of Agriculture ARS. 2011. USDA National Nutrient Database for Standard Reference, Release 24. Nutrient Data Laboratory Home Page. Ullrey DE, Robinson PT, Whetter PA. 1981. Eucalyptus digestibility and digestible energy requirements of adult male koalas, Phascolarctos cinereus (Marsupialia). Australian Journal of Zoology 29(6):847 852. Urriola PE, Stein HH. 2010. Effects of distillers dried grains with solubles on amino acid, energy, and fiber digestibility and on hindgut fermentation of dietary fiber in a corn soybean meal diet fed to growing pigs. Journal of Animal Science 88(4):1454 1462. Van Keu len J, Young BA. 1977. Evaluation of acid insoluble ash as a natural marker in ruminant digestibility studies. Journal of Animal Science 44(2):282 287. Van Soest PJ. 1978. Dietary fibers: their definition and nutritional properties. The American journal of clinical nutrition 31(10 Suppl):S12 S20. Van Soest PJ. 1994. Nutritional ecology of the ruminant. Second ed. Ithaca: Comstock Publishing Associates. Van Weyenberg S, Sales J, Janssens GPJ. 2006. Passage rate of digesta through the equine gastrointestinal tract: A review. Livestock Science 99(1):3 12. Vanhoven W, Prins RA, Lankhorst A. 1981. Fermentative digestion in the African elephant. South African Journal of Wildlife Research 11(3):78 86. Vergara Parente JE, Parente CL, Marmontel M, Ramos Silva JC, Sa FB. 2010. Growth curve of free ranging Trichechus inunguis. Biota Neotropica 10(3):89 92. Vogel KP, Pedersen JF, Masterson SD, Toy JJ. 1999. Evaluation of a filter bag system for NDF, ADF, and IVDMD forage analysis. Crop Science 39(1):276 279. Walsh MT, Bo ssart GD. 1999. Manatee medicine. In: Fowler ME, Miller RE, editors. Zoo and Wild Animal Medicine. Philadelphia: W.B. Saunders. p 507 516.

PAGE 191

191 Ward Geiger LL. 1997. Blubber depth and body condition indices in the Florida manatee (Trichechus manatus latirostris ). Tampa, FL: University of South Florida. Warren LK, Kivipelto J. 2005. Comparison of cecal or fecal inocula used in the determination of in vitro fiber digestibility. Equine Science Society 19th Symposium. Tucson, Arizona. p 314 315. Willemsen RE, Hailey A. 2002. Body mass condition in Greek tortoises: regional and interspecific variation. Herpetological Journal 12(3):105 114. Wilman D, Adesogan A. 2000. A comparison of filter bag methods with conventional tube methods of determining the in vitro digestib ility of forages. Animal Feed Science and Technology 84(1 2):33 47. Worthy GAJ. 2001. Nutrition and Energetics. In: Dierauf LA, Gulland FMD, editors. CRC Handbook of Marine Mammal Medicine. 2nd ed. Boca Raton, FL: CRC Press. p 791 827. Yildirim A, Arslan M Bektas S. 2008. Length weight relationship and seasonal condition in Capoeta sieboldii in the upper Coruh River, Turkey. Journal of Applied Ichthyology 24(6):711 712. Zarling EJ, Ruchim MA. 1987. Protein origin of the volatile fatty acids isobutyrate and isovalerate in human stool. Journal of Laboratory and Clinical Medicine 109(5):566 570.

PAGE 192

192 BIOGRAPHICAL SKETCH Lauren T. Harshaw was born near Seoul, South Korea, but only spent three short months there. On April 15, tax day, Lauren flew over to the United States, where her wonderful parents adopted her. Lauren grew up in a small town outside of Philadelphia, Pennsylvania before heading off to Davidson College in North Carolina. She ea rned her Bachelor of Science in b iology at Davidson and published her f irst set of papers on aquatic insect ecology from her research conducted there. After spending a few months learning Florida manatee photo identification during an internship at Mote Marine Laboratory in Sarasota, Florida, Lauren returned to Philadelphia, where she was employed as a research technician at Monell Chemical Senses Center. Even though she enjoyed her time working on taste research in mice, Lauren was ready to continue her ege of Veterinary Medicine in the Aquatic Animal Health Program. She soon found her niche in the realm of manatee nutrition and considers herself to be a self proclaimed fecal ologist. Lauren hopes to continue with a career dedicated to conservation and looks forward to working with species that are just as amazing as the Florida manat ee.