<%BANNER%>

Physiological and Life-History Responses to Patterns of Food Availability

Permanent Link: http://ufdc.ufl.edu/UFE0021002/00001

Material Information

Title: Physiological and Life-History Responses to Patterns of Food Availability
Physical Description: 1 online resource (159 p.)
Language: english
Creator: Roark, Alison M
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: antioxidant, dna, fecundity, food, growth, intake, longevity, rna
Zoology -- Dissertations, Academic -- UF
Genre: Zoology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Green turtles (Chelonia mydas) experience nutritional stochasticity as oceanic-stage juveniles and should therefore be capable of compensatory growth (CG) following periods of nutritional stress. The purpose of the first phase of my research was to test for the occurrence of CG in green turtles and determine its mechanism(s). Food-restricted turtles (R) grew more slowly, differed in cell size and body composition, and had proportionally smaller digestive organs than turtles feeding ad libitum (AL). After food conditions improved, previously food restricted turtles (R-AL) demonstrated CG. This growth pattern was elicited by enhanced food conversion efficiency rather than hyperphagia. The period of growth compensation may have ended when R-AL turtles attained a body composition similar to that of AL turtles. These findings indicate that growth rate, morphology, and body composition of juvenile green turtles are plastic in response to diet and that individuals can compensate for environmental variability to capitalize when conditions improve. However, CG was associated with altered antioxidant function. Activity of glutathione peroxidase and total antioxidant potential per liver cell were greater in AL turtles than in R and R-AL turtles, respectively, at the conclusion of the study. Therefore, impaired antioxidant capacity may be a cost of rapid growth in this species. To elucidate long-term responses to food availability, I tested the effects of food restriction (FR) imposed during several life stages in Indian stick insects (Carausius morosus). Intake pattern affected age and size at each life-history transition, with size decreasing and age increasing in response to FR. Early-onset FR increased lifespan, but this increase was negated by detrimental effects on reproductive output. Late-onset FR negatively affected both longevity and reproductive output. Carausius morosus appears to allocate incoming adult-derived nutrients to egg provisioning and stored nutrients to adult somatic maintenance and survival. Thus, I found no evidence for a trade-off between fecundity and longevity because nutrients allocated to reproduction and maintenance do not appear to be derived from a common resource pool. These results demonstrate that fluctuations in food availability can significantly alter the expression of life-history traits and that the magnitude of these effects depends on the developmental stage during which food availability changes.
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 Alison M Roark.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Bjorndal, Karen.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021002:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021002/00001

Material Information

Title: Physiological and Life-History Responses to Patterns of Food Availability
Physical Description: 1 online resource (159 p.)
Language: english
Creator: Roark, Alison M
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: antioxidant, dna, fecundity, food, growth, intake, longevity, rna
Zoology -- Dissertations, Academic -- UF
Genre: Zoology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Green turtles (Chelonia mydas) experience nutritional stochasticity as oceanic-stage juveniles and should therefore be capable of compensatory growth (CG) following periods of nutritional stress. The purpose of the first phase of my research was to test for the occurrence of CG in green turtles and determine its mechanism(s). Food-restricted turtles (R) grew more slowly, differed in cell size and body composition, and had proportionally smaller digestive organs than turtles feeding ad libitum (AL). After food conditions improved, previously food restricted turtles (R-AL) demonstrated CG. This growth pattern was elicited by enhanced food conversion efficiency rather than hyperphagia. The period of growth compensation may have ended when R-AL turtles attained a body composition similar to that of AL turtles. These findings indicate that growth rate, morphology, and body composition of juvenile green turtles are plastic in response to diet and that individuals can compensate for environmental variability to capitalize when conditions improve. However, CG was associated with altered antioxidant function. Activity of glutathione peroxidase and total antioxidant potential per liver cell were greater in AL turtles than in R and R-AL turtles, respectively, at the conclusion of the study. Therefore, impaired antioxidant capacity may be a cost of rapid growth in this species. To elucidate long-term responses to food availability, I tested the effects of food restriction (FR) imposed during several life stages in Indian stick insects (Carausius morosus). Intake pattern affected age and size at each life-history transition, with size decreasing and age increasing in response to FR. Early-onset FR increased lifespan, but this increase was negated by detrimental effects on reproductive output. Late-onset FR negatively affected both longevity and reproductive output. Carausius morosus appears to allocate incoming adult-derived nutrients to egg provisioning and stored nutrients to adult somatic maintenance and survival. Thus, I found no evidence for a trade-off between fecundity and longevity because nutrients allocated to reproduction and maintenance do not appear to be derived from a common resource pool. These results demonstrate that fluctuations in food availability can significantly alter the expression of life-history traits and that the magnitude of these effects depends on the developmental stage during which food availability changes.
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 Alison M Roark.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Bjorndal, Karen.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021002:00001


This item has the following downloads:


Full Text





PHYSIOLOGICAL AND LIFE-HISTORY RESPONSES TO PATTERNS OF
FOOD AVAILABILITY




















By

ALISON M. ROARK


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

2007





































2007 Alison M. Roark
































To my family, without whom this work would not have been possible









ACKNOWLEDGMENTS

My dissertation would not have been possible without the support, assistance, and

encouragement of many wonderful people. I would first like to thank my advisor and friend,

Karen Bjorndal, for guiding me through the adventures of graduate school. Karen's wisdom and

generosity have been truly inspirational. I also thank Alan Bolten for many years of emotional

and logistical support. My committee members Karen Bjorndal, Alan Bolten, Lou Guillette,

Dan Hahn, David Julian, and John Sivinski have been invaluable to my progress and I cannot

thank them enough for their contributions to my professional development.

I am indebted to many friends and colleagues who provided assistance with the evolution,

implementation, and/or analysis of my work. I am particularly grateful to my unbelievable

undergraduate assistants, including May Steward, Adam Sarnowski, Carie Reynolds, Justin

Emerson, Dana-Rachael La Kam, Kelly Johnson, Katherine Perez, Daphna Yasova, Ann Mazor,

and Sasha Strul, all of whom contributed significantly to the success of my research. I also thank

Richie Moretti for initiating the dialogue that led to fruitful collaborations between the Cayman

Turtle Farm and the Archie Carr Center for Sea Turtle Research, and I thank Ken Prestwich for

assistance with insect metabolic rate measurements. Debra Murie, Matthias Starck, Tony Zera,

Jamie Gillooly, and Scott Pletcher participated in thought-provoking discussions about the

conceptual framework and analysis of various components of my research. Lauren Chapman,

Greg Pryor, Mike McCoy, and Ben Bolker were invaluable statistical consultants who patiently

taught me how to use and interpret SPSS and R. Additional statistical consulting was provided

by Xueli Liu and Alex Trindade. Harvey Ramirez and Elliott Jacobson assisted with clinical

aspects of animal care, and Franco Giorgi graciously provided photocopies of embryonic staging

tables for Carausius morosus.









The remarkable graduate students, postdoctoral associates, and faculty members in the

Department of Zoology at the University of Florida were a source of constant support,

companionship, and friendship. I am especially appreciative for the contributions of my many

lab-mates, including Lindy Barrow, Sarah Bouchard, Peter Eliazar, Gabby Hrycyshyn, Kate

Moran, Greg Pryor, Kim Reich, Jeff Seminoff, Manjula Tiwari, Hannah Vander Zanden, and

Brian Riewald (who is remembered fondly). Nat Seavy, Kenney Krysko, Ryan McCleary, Krista

McCoy, and Jada-Simone White also contributed to various aspects of my work. I owe a special

debt of gratitude to Thea Edwards and Brandon Moore for allowing me to harvest English ivy

from their property at all hours of the day and night and for invaluable horticultural advice.

Additional members of the Department of Zoology worked tirelessly behind the scenes to

ensure that my progression through graduate school was as smooth as possible. I would like to

give special thanks to Karen Pallone, Vitrell Sherif, Pete Ryschkewitsch, Mike Gunter, Frank

Davis, Peggy Roberson, Diana Davis, and Cathy Moore for their logistical support over the

years.

Ginger Clark, Marty Cohn, Jason Curtis, Richard Fethiere, Paul Gulig, Christiaan

Leeuwenburgh, Frank Robbins, and Collette St. Mary graciously provided access to their labs

and assistance with many of the analyses I performed. Sharon Judge, Colin Selman, and other

members of the Leeuwenburgh laboratory helped with the antioxidant assays, and Beverly

Defense assisted with fluorescent spectroscopy.

I am grateful beyond words for the generosity and support of the Cayman Turtle Farm.

Without them, the work I report in Chapters 2-4 would have been impossible. I would like to

give special thanks and praise to Ken Hydes and Joe Parsons, who were integral members of my

Caymanian "family" and provided me with a wealth of technical, logistical, and emotional









support during what amounted to an unbelievably challenging and taxing semester. I do not think

that I will ever be able to repay them fully for their many gifts.

Conducting animal research requires the oversight of a number of permitting agencies,

particularly when this research entails working with endangered or restricted species. I am

especially grateful to the U.S. Fish and Wildlife Service, the Institutional Animal Care and Use

Committee at the University of Florida, and the U.S. Department of Agriculture. I owe enormous

thanks to Sanford Porter, Lim Nong, and Mike Thomas of the USDA in Gainesville for

assistance in constructing a quarantine facility and for obtaining the requisite plant pest permit

that enabled my work with Indian stick insects.

Funding for my dissertation was provided by the National Science Foundation through

both a Graduate Research Fellowship and a Doctoral Dissertation Improvement Grant.

Additional funds were obtained from the Society for Integrative and Comparative Biology,

Sigma Xi, the American Society of Ichthyologists and Herpetologists, the Brian Riewald

Memorial Grant, Sigma Delta Epsilon, two grants from the National Institute on Aging

(AG17994 and AG21042 to Christiaan Leeuwenburgh) and the University of Florida

Opportunity Fund. Numerous travel grants were provided by the University of Florida Graduate

Student Council, the Department of Zoology at the University of Florida, the Comparative

Nutrition Society, the Symposium on Sea Turtle Biology and Conservation, and the Society for

Integrative and Comparative Biology. Additionally, Benchmark Foliage donated English ivy,

and the Exploratorium in San Francisco, CA, donated the adult Indian stick insects I used in

Chapter 5. Charlie Carlson and Angela Armendariz of the Exploratorium were especially helpful

during my visit to their facility.









I owe an enormous debt to my family for the encouragement, love, and support they have

provided throughout my time in graduate school. It is impossible to put into words the gratitude I

feel for my parents and my sister. Their faith in me, their emotional support, and their belief in

the value of education have inspired and motivated me through thick and thin. I would like to

give special thanks to my father for his many hours in the garage helping me construct cages,

basking platforms, and lighting structures that I used during my feeding trials. I would also like

to thank my parents for traveling to the Cayman Islands to help me recover from a devastating

hurricane in November of 2001. They were wonderful field assistants during this time. Roger

and Annette Roark deserve many thanks, as well, for their encouragement through the years.

They have been unbelievably supportive, and I am incredibly lucky to be their daughter-in-law.

Lastly, I thank my husband, Andrew. He has experienced all of my "ups" and "downs" along the

way, and his patience and belief in me have been a constant source of comfort during the past

seven years.









TABLE OF CONTENTS



A C K N O W L E D G M E N T S ..............................................................................................................4

L IS T O F T A B L E S ............................................................................................... ..................... 10

LIST OF FIGURES ....................... ...................... .. ......... ............ ............... 12

A B S T R A C T .......................................................................................................... ..................... 14

CHAPTER

1 G EN ER A L IN TR O D U C TIO N ..............................................................................................16

2 COMPENSATORY GROWTH IN RESPONSE TO A CHANGE IN FOOD
AVAILABILITY IN JUVENILE GREEN TURTLES (Chelonia mydas) .........................23

In tro d u c tio n ............................................................................................................................. 2 3
M materials and M methods .............. .............................................................................. 26
Anim al Care..................................................... ........... .. ..................... 26
Gut M orphology and Body Composition.................................................... 27
Statistical A naly ses............... .. .................. .................. ............... ...... ... ... ............ 29
Results .................................................... ............................... 30
Discussion ...................................................... .................. 34

3 BIOCHEMICAL INDICES AS CORRELATES OF RECENT GROWTH IN
JUVENILE GREEN TURTLES (Chelonia mydas) ............... ................................... 55

Introduction ........................................................ .................. 55
M materials and M methods .............. .............................................................................. 56
Animal Care ............................................ .................................. 56
Tissue C collection ...................................................................................................... 57
B iochem ical A says ................................................................................................... 57
Statistical A analyses .................................................................................................... 58
Results ................................................... ............................... 60
Discussion ...................................................... .................. 62

4 COMPENSATORY GROWTH AND ANTIOXIDANT STATUS IN JUVENILE
GREEN TURTLES (Chelonia mydas)................................................................................78

Introduction ........................................................ .................. 78
M materials and M methods .................................................................................................... 80
A nim al C are ............................................................ ...................... ....................80
Tissue Collection and Homogenization........................................................................80
Glutathione Peroxidase Activity Assay........................................................................81
T o ta l A P A ssa y ............................................................... ................................................ 8 2


8









Statistical A naly ses............... .. .................. .................. ............ ......... ... ... ........... 82
Results .................................................... .............................. 83
Discussion ...................................................... .................. 85

5 TIMING OF DIETARY RESTRICTION ALTERS THE EXPRESSION OF LIFE-
HISTORY TRAITS IN A LONG-LIVED, PARTHENOGENETIC INSECT ...................... 95

Introduction ........................................................ .................. 95
M materials and M methods ...................................................... ............................................... 99
Anim al Husbandry and Feeding Treatm ents.............................................. ................ 99
Physiological and Life-History Response Variables...... ................... ................... 100
Statistical A naly ses............. .. .................. .................. ............ ........ .. .. ... ........... 102
Results ......................................................... .................. 104
Discussion ................................................ ............................. 109

6 SUMMARY AND CONCLUSIONS.........................................................136

L IST O F R E F E R E N C E S ....................................................... ................................................ 142

B IO G R A PH IC A L SK E T C H .................................................... ............................................. 159

































9









LIST OF TABLES


Table page

2-1. Kruskal-Wallis test results for nutrient content of biweekly food samples........................41

2-2. Repeated measures analyses of variance for weekly averages of daily intake and daily
m ass-specific intake. .............. .......................... ...................... ... ........ 42

2-3. Repeated measures analyses of variance for weekly body mass, straight carapace
length, and condition index .............................................................................. ................ 43

2-4. Repeated measures analyses of variance for weekly specific growth rates (SGR) for
body m ass (bm ) and carapace length (cl) ..................................................... ................ 44

2-5. Repeated measures analyses of variance for food conversion efficiencies (FCE) for
body m ass (bm ) and carapace length (cl) ..................................................... ................ 45

2-6. Omnibus F, x2, and p-values for analyses of variance of dissection data collected at
w e e k s 5 a n d 12 ................................................................................................................. .. 4 6

2-7. Organ masses (mean standard error) from turtles dissected at 0, 5, and 12 weeks
rep orted as in dices..................................................... ................................................ 4 7

2-8. Body composition (mean standard error) of turtles dissected at 0, 5, and 12 weeks
reported as percent of dry matter (% DM) and percent of organic matter (% OM). .........48

3-1. Omnibus F, x2, and p-values for comparisons of means among treatment groups for the
various morphometric and biochemical indices measured ...........................................69

3-2. Spearman's rank correlations (p) for morphometric (a) and biochemical indices for
liver (b), heart (c), and blood (d)......................................... ....................... ................ 70

3-3. Growth equation parameters for juvenile Chelonia mydas as determined by least
squares linear regression .... .................................................................. 71

3-4. Growth equation parameters for juvenile Chelonia mydas as determined by stepwise
m multiple linear regression ... ..................................................................... ... ............ 72

3-5. Coefficients of variation (C.V.) for RNA, DNA, and protein concentrations of
Chelonia mydas tissues ........................ ........................................73

4-1. Omnibus F, x2, and p-values for comparisons of means among treatment groups at five
w eeks (t5) and tw elve w eeks (t12)...................................................................... 89

4-2. Total protein concentrations of Chelonia mydas muscle and liver homogenates as
determined by Bradford assay expressed relative to wet mass of homogenized tissue.....90

4-3. Glutathione peroxidase (GPX) specific activity in Chelonia mydas muscle homogenate....91









4-4. Coefficients of variation (CV, %) for protein concentration, glutathione peroxidase
(GPX) activity, and antioxidant potential (AP) assays ................................ ................ 92

5-1. Cumulative intake expressed as the total dry mass consumed during each life-history
stage in each of five treatm ent groups. ...... .......... .......... .....................1... 19

5-2. Omnibus F, X2, and p-values for comparisons of body mass and age among five
treatment groups within each life-history stage. ...... ... ......................................... 120

5-3. Relative mass ( standard error) at the adult molt of insects in five treatment groups,
calculated as the ratio of actual to predicted body mass as determined by allometric
analysis ................................................................................................. .. 12 1

5-4. Adjusted mean fecundity ( standard error) of insects in each of five treatment groups
estimated using body mass at first oviposition as a covariate. .............. ...................122

5-5. Total number of ovarioles (mean standard error) in insects from each of five
treatment groups upon post-mortem dissection ..............................................123

5-6. Parameters for equations predicting realized fecundity and initial oviposition rate as
determined by stepwise multiple linear regression....... ... ...................................... 124

6-1. Summary of traits measured for Chelonia mydas (top half of table) and Carausius
morosus (bottom half of table) maintained on different schedules of ad libitum (ad
lib.) and restricted (rest.) intake. ................. ......................................................... 141









LIST OF FIGURES


Figure page

1-1. Hypothetical plot of size versus time for juvenile animals from the same cohort .............22

2-1. Average mass-specific daily intake (mean standard error) during each week of the
feeding trial .................................................................................................... ........ .. 4 9

2-2. Body mass (a) and straight carapace length (b) (mean standard error) at the midpoint
o f e a c h w e e k .................................................................................................................. ... 5 0

2-3. Condition index (mean standard error) in each week....................................................51

2-4. Specific growth rate (mean standard error) for body mass (BM, a) and straight
carapace length (CL, b) during each w eek.................................................... ............... 52

2-5. Food conversion efficiency (FCE, mean standard error) for body mass (BM, a) and
straight carapace length (CL, b) during each week....................................... ................ 53

2-6. Daily water temperatures (mean standard deviation) throughout the feeding trial............54

3-1. Morphometric indices and growth rates for turtles in each of three treatment groups..........74

3-2. Nucleic acid indices for turtles in each of three treatment groups. ..................................75

3-3. Liver protein and protein-based indices for turtles in each of three treatment groups .........77

4-1. Body mass of turtles at five weeks (ts) and twelve weeks (t12), when tissues were
sam p led .......................................................................................................... ........ .. 9 3

4-2. Glutathione peroxidase (GPX) specific activity and antioxidant potential (AP;
calculated as nmoles of copper reducing equivalents, CRE) in Chelonia mydas liver
homogenate at five weeks (ts) and twelve weeks (t12)................................................. 94

5-1. Experimental design for Carausius morosus feeding trial....................... ...................125

5-2. Mass-specific intake (g dry mass/g*day) consumed by insects in each of five treatment
groups on each day of the study..................................... ....................... ................ 126

5-3. Age and size at each life-history transition for insects in each of five treatment groups.... 127

5-4. Duration of (a) and specific growth rate during (b) each life-history stage for insects in
each of five treatm ent groups....................................... ......................... ................ 128

5-5. Event history diagram depicting periods of ad libitum intake, restricted intake, and
reproductive activity for individual stick insects maintained on five diet treatments. .... 129









5-6. Kaplan-Meier survivorship curves for the entire lifespan (a) and for the adult lifespan
(b) of insects maintained on five diet treatments. ......... ......................................131

5-7. Cumulative fecundity of insects in each of five treatment groups ................................133

5-8. Average egg mass (mean + standard error) for stick insects maintained on five diet
treatm ents...................................................................................................... ......... 134

5-9. Relationships between realized fecundity and cumulative intake during the
reproductive lifespan (a) and the duration of adult lifespan (b) for all insects that laid
e g g s ...................................................................................................... ........ . ....... 1 3 5









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

PHYSIOLOGICAL AND LIFE-HISTORY RESPONSES TO PATTERNS OF
FOOD AVAILABILITY

By

Alison M. Roark

August 2007

Chair: Karen A. Bjorndal
Major: Zoology

Green turtles (Chelonia mydas) experience nutritional stochasticity as oceanic-stage

juveniles and should therefore be capable of compensatory growth (CG) following periods of

nutritional stress. The purpose of the first phase of my research was to test for the occurrence of

CG in green turtles and determine its mechanismss. Food-restricted turtles (R) grew more

slowly, differed in cell size and body composition, and had proportionally smaller digestive

organs than turtles feeding ad libitum (AL). After food conditions improved, previously

food-restricted turtles (R-AL) demonstrated CG. This growth pattern was elicited by enhanced

food conversion efficiency rather than hyperphagia. The period of growth compensation may

have ended when R-AL turtles attained a body composition similar to that of AL turtles. These

findings indicate that growth rate, morphology, and body composition of juvenile green turtles

are plastic in response to diet and that individuals can compensate for environmental variability

to capitalize when conditions improve. However, CG was associated with altered antioxidant

function. Activity of glutathione peroxidase and total antioxidant potential per liver cell were

greater in AL turtles than in R and R-AL turtles, respectively, at the conclusion of the study.

Therefore, impaired antioxidant capacity may be a cost of rapid growth in this species.









To elucidate long-term responses to food availability, I tested the effects of food

restriction (FR) imposed during several life stages in Indian stick insects (Carausius morosus).

Intake pattern affected age and size at each life-history transition, with size decreasing and age

increasing in response to FR. Early-onset FR increased lifespan, but this increase was negated by

detrimental effects on reproductive output. Late-onset FR negatively affected both longevity and

reproductive output. Carausius morosus appears to allocate incoming adult-derived nutrients to

egg provisioning and stored nutrients to adult somatic maintenance and survival. Thus, I found

no evidence for a trade-off between fecundity and longevity because nutrients allocated to

reproduction and maintenance do not appear to be derived from a common resource pool. These

results demonstrate that fluctuations in food availability can significantly alter the expression of

life-history traits and that the magnitude of these effects depends on the developmental stage

during which food availability changes.









CHAPTER 1
GENERAL INTRODUCTION

Nearly every biological process depends to some extent on nutrition. Ingestion, digestion,

and nutrient absorption determine an animal's capacity for maintenance, growth, reproduction,

and survival. An individual's success thus depends on its capacity to extract and utilize nutrients

from its food. However, many organisms experience variation in food availability throughout

their lifetimes (Plotz et al. 1991, Boggs and Ross 1993, Carey et al. 2002a), and evolutionary

adaptation to these fluctuations is critical to species survival. Fluctuations in food availability can

result from climatic or seasonal dynamics of food sources and spatial or temporal heterogeneity

of nutrient distribution (Dagg 1977, Smith and Ballinger 1994, Forman 1995, Arnekleiv et al.

2006, Schradin and Pillay 2006). Periods of "feast" and "famine" result in correspondingly fast

and slow periods of growth, development, and reproduction (Ballinger 1977, Calbet and Alcaraz

1997, Kitaysky 1999, Morey and Reznick 2000). Such developmental plasticity is common to

many species (Stearns 1982, Smith-Gill 1983, Schew and Ricklefs 1998), especially those in

stochastic environments (Lochmiller et al. 2000).

One adaptation to cycles of low and high nutrient availability is compensatory growth

(CG). This phenomenon manifests itself as a period of growth faster than that demonstrated by

consistently well nourished conspecifics of the same age (Fig. 1-1) and can result in comparable

body sizes for individuals with very different dietary histories (Broekhuizen et al. 1994, Metcalfe

and Monaghan 2001). Compensatory growth presumably allows organisms to mitigate

size-specific mortality risks and developmental time constraints induced by periodic nutritional

stress (Arendt 1997). This growth pattern is typically effected by hyperphagia (increased feed

intake), improved food conversion efficiency (defined as growth per unit of food consumed), or

both (Miglavs and Jobling 1989, Ali et al. 2003). Compensatory growth has been documented in









many species (Wilson and Osbourn 1960, Ali et al. 2003, Jespersen and Toft 2003), particularly

those that commonly experience environmental stochasticity. However, compensatory growth is

certainly not a universal occurrence, especially in non-teleost species (e.g., Altwegg and Reyer

2003, Brzek and Konarzewski 2004).

For my dissertation, I studied the physiological and life-history responses of animals to

changes in food availability. The green turtle (Chelonia mydas) is an excellent model for

pursuing questions about fluctuating food availability because of the nutritional stochasticity

they are thought to encounter as juveniles. This stochasticity should select for growth patterns

that allow individuals to maximize their growth rates when conditions are good. Green turtles

consume a largely carnivorous diet during the juvenile oceanic stage (Reich et al. in review).

Food availability during this time is thought to be spatially heterogeneous. At a size of 20-35 cm

(carapace length), juvenile green turtles undergo an ontogenetic shift in habitat use and diet and

recruit to neritic habitats, where they continue development while consuming a primarily

herbivorous diet (Bjorndal 1997). Niche shifts are excellent opportunities for enhanced growth

because such shifts often correspond to improved nutritional conditions (Ali et al. 2003).

Although growth dynamics of oceanic-stage juveniles are unknown, post-recruitment growth

rates vary temporally in response to oceanographic conditions (Limpus and Chaloupka 1997) and

population density (Bjorndal et al. 2000). This variation in growth rates may have substantial

life-history effects, because body size is correlated with juvenile survivorship (Chaloupka and

Limpus 1998) and clutch size (Broderick et al. 2003) in this species. Because of the potential for

nutritional stochasticity and the potential effects of body size on fitness in C. mydas, I predicted

that green turtle juveniles should be capable of CG. Juvenile loggerheads with dietary









preferences similar to those of green turtles undergo CG in the wild (Bjorndal et al. 2003),

although the mechanism for this growth pattern is unknown.

The purpose of the first phase of my research was to manipulate food availability in

juvenile green turtles under controlled conditions to test whether previously food-restricted

turtles can undergo CG and to evaluate hyperphagia and enhanced food conversion efficiency as

possible mechanisms for this growth pattern. The results of Chapter 2 indicated that previously

food-restricted green turtle juveniles are indeed capable of transient growth compensation after a

return to ad libitum feeding. This finding suggests that growth rates of turtles under conditions of

continuously high food availability are optimal rather than maximal (Metcalfe and Monaghan

2001) and that trade-offs between growth and fitness probably exist.

Rapid growth in many species may be associated with a variety of costs, many of which

may not be paid until late in life (Einum and Fleming 2000 and references therein, Metcalfe and

Monaghan 2001, Altmann and Alberts 2005, Nagy and Holmes 2005). To pursue this hypothesis,

I examined antioxidant function of tissues from green turtles that had undergone CG.

Antioxidants prevent free-radical induced oxidative damage to nucleic acids and proteins by

converting reactive oxygen species (ROS) into less noxious compounds (Ji and Leichtweis 1997,

Gredilla and Barja 2005). Caloric restriction depresses the rate of production of ROS

(L6pez-Torres et al. 2002, Barj a 2004) and attenuates the accrual of irreparable damage to

cellular macromolecules (Hyun et al. 2006). As a result, dietary restriction slows aging and thus

extends lifespan relative to continuous ad libitum feeding in a diversity of species (Weindruch

and Walford 1988, Austad 1989, Mair et al. 2003, Vaupel et al. 2003, Hatle et al. 2006b).

Because ad libitum feeding is typically accompanied by accelerated aging, it is possible that

periods of rapid growth induced by high food availability after a period of nutritional stress may









be costly in terms of survival. I hypothesized that the short-term benefits of CG may be

countered by the rapid accumulation of oxidative damage, thereby imposing a cost to fast

growth. My results confirmed that turtles with a history of CG suffered decreased cellular

antioxidant potential relative to continuously ad libitum-fed turtles.

The diminished capacity for antioxidant defense exhibited by fast-growing turtles piqued

my interest in the more long-term consequences of changes in food availability on life-history

traits and trade-offs. Despite the tendency for caloric restriction to enhance longevity in the

laboratory, nutritional stress experienced early in life can have profound life-history

consequences (Metcalfe and Monaghan 2001). For example, growth rate, dominance status, and

age at first parturition in spotted hyenas are strongly correlated with food availability during

development (Hofer and East 2003). This "silver-spoon effect" (Grafen 1988, Madsen and Shine

2000) indicates that early nutritional conditions can have long-lasting effects. Resource

limitation experienced later in life can also influence life history by decreasing reproductive

output (Boggs and Ross 1993, Wheeler 1996, Olsson and Shine 1997, Carey et al. 2002b, Nagy

and Holmes 2005). On the other hand, food restriction has been shown to extend an animal's

reproductive lifespan (Hart and Turturro 1998), thereby partially mitigating the decline in

reproductive rate caused by food scarcity.

Most of the work cited above, however, provides incomplete information about the effects

of nutrition on fitness and life history because intake is not typically quantified throughout the

entire lifespan in studies of this kind (Zera and Harshman 2001), particularly for long-lived

species. As most animals experience fluctuations in food availability at some point in their

lifetimes, a more complete understanding of the effects of diet on fitness requires intake









manipulation and quantification under controlled conditions during both juvenile and adult

stages.

Furthermore, to evaluate the costs of reproduction and the effects of dietary restriction on

fitness in sexual species, females must be allowed to mate. However, mating is often prevented

(particularly in rodent studies) because experimental animals are maintained individually

(Vaupel et al. 2003). In those studies where mating is permitted, co-housing individuals can

complicate the quantification of individual intake and is known to influence longevity and

fecundity due to the effects of crowding (Joshi et al. 1998). To avoid such problems, studies of

food restriction and reproductive output are often undertaken using virgin or hermaphroditic

females of invertebrate species. This approach suffers from limitations, including the fact that

mating enhances fecundity (Chiang and Hodson 1950, Foster and Howard 1999, Chong and

Getting 2006) and that self-fertilization constrains reproductive output because of limits on

self-sperm production in protandrous hermaphroditic species like Caenorhabditis elegans (Cutter

2004, Partridge et al. 2005). For these reasons, the choice of an appropriate model organism for

the final phase of my dissertation research was of paramount importance.

Although the first phase of my research focused on green turtles, this species is not a

suitable animal model for investigating correlations between intake and life-history traits. To

evaluate the effects of food availability on development, longevity, and reproductive output, I

adopted a novel approach to life-history experimentation by using a parthenogenetic species as

my animal model. Using a parthenogen obviated the need for mating, and therefore allowed

females to be housed individually, while still permitting natural reproductive processes. Given

the paucity of parthenogenetic vertebrate species with a reasonable (i.e., less than two-year)

lifespan that are amenable to laboratory culture, I chose to work with an insect species for the









final phase of my dissertation research. The Indian stick insect (Carausius morosus (Br.))

(Phasmatodea, Lonchodinae) is a relatively long-lived species that reproduces via apomictic

parthenogenesis (Pijnacker 1966). This species is hemimetabolous and phytophagous, allowing

for life-long, quantitative dietary manipulations using the same food source throughout the entire

lifespan. My purpose was to determine the effects of differences in resource availability at

several developmental stages on life-history traits that have substantial influences on population

structure and dynamics, such as age and size at each developmental transition, longevity, and

fecundity.

The overall goal of my dissertation research was to elucidate the physiological and

life-history effects of variation in food availability by conducting feeding trials in a controlled

laboratory setting. In Chapter 2, I explore the capacity for and mechanisms of CG in green turtles

and evaluate the effects of intake and growth rates on body composition and digestive system

morphology. In Chapter 3, I examine the effects of diet on cell size and protein synthesis

capacity and assess the utility of a number of morphological and biochemical indices as potential

predictors of recent growth in green turtles. In Chapter 4, I compare the antioxidant potentials of

green turtles with known growth trajectories to establish a putative cost of rapid growth. In

Chapter 5, I determine the life-history consequences of changes in food availability imposed

either during development or after the attainment of sexual maturity in Indian stick insects. For

this final chapter, I pursued questions about the effects of variation in intake on developmental

transitions, longevity, and fecundity.




























Time


Figure 1-1. Hypothetical plot of size versus time for juvenile animals from the same cohort. The
solid line represents individuals feeding ad libitum, and the dashed line represents
individuals feeding at a restricted rate until the time indicated by the arrow, after
which food was offered ad libitum. The rapid growth demonstrated by the
previously-restricted individuals after the switch to ad libitum feeding represents
compensatory growth.









CHAPTER 2
COMPENSATORY GROWTH IN RESPONSE TO A CHANGE IN FOOD AVAILABILITY
IN JUVENILE GREEN TURTLES (Chelonia mydas)

Introduction

Many organisms experience varying levels of nutrient availability throughout their

lifetimes as a result of spatial or temporal heterogeneity of food distribution. These periods of

high and low food availability often lead to correspondingly fast and slow rates of growth,

development, and/or reproduction (Ballinger 1977, Calbet and Alcaraz 1997, Kitaysky 1999,

Morey and Reznick 2000). Given these life-history consequences, fluctuations in food

availability should select for adaptations that allow a previously food-limited individual to

maximize its ability to capitalize on better conditions when they are encountered.

One common response to alternating periods of high and low food availability is

compensatory growth (CG). This phenomenon manifests itself as a period of accelerated growth

during improved food conditions following a period of nutritional deprivation severe enough to

restrict growth rates (Wilson and Osbourn 1960, Reid and White 1977). As a result, growth

trajectories tend to converge, thereby minimizing the variance in body size among individuals of

a cohort (Atchley 1984). Compensatory growth presumably allows organisms to mitigate the

negative effects of slow growth on survival, development, and reproductive output. This growth

pattern has been documented in invertebrates (Jespersen and Toft 2003, Dmitriew and Rowe

2005), fish (Skalski et al. 2005), reptiles (Bjorndal et al. 2003, Caley and Schwarzkopf 2004),

birds (Kunz and Ekman 2000), and mammals (Lochmiller et al. 2000). However, compensatory

growth is certainly not a universal occurrence, especially in non-teleost species (e.g., Altwegg

and Reyer 2003, Brzek and Konarzewski 2004).

Although the degree of growth compensation can vary depending on the species in

question, the developmental stage of the organism at the times of restriction and improved food









availability, and the length and severity of the period of food restriction (Wilson and Osbourn

1960, Ali et al. 2003), the mechanisms for compensatory growth (when it is observed) are

relatively conserved. Hyperphagia, or increased food intake, is the most common proximate

cause of CG in a variety of animals (Ali et al. 2003). During hyperphagia, the rate of lipid

accumulation may direct the duration of the compensatory response, with intake and growth rates

returning to "normal" once adipose stores have been restored (Jobling and Johansen 1999,

Johansen et al. 2001). Improvements in food conversion efficiency (FCE, defined as growth per

unit of food consumed) can also allow for CG (Patterson et al. 1995, Boujard et al. 2000).

Reductions in the mass of energetically expensive viscera (especially gut and/or liver; Hornick et

al. 2000) during food restriction have been demonstrated in fish (Gaylord and Gatlin 2000),

mammals (Weindruch and Sohal 1997), and birds (Hume and Biebach 1996, Karasov and

Pinshow 1998). If decreased organ size persists for a period of time after a return to ad libitum

feeding, the resulting combination of lowered metabolic expenditure and high intake might allow

more nutrients to be allocated to growth. In this way, FCE would remain high, thus facilitating

CG during this period.

In addition to affecting visceral organ size, variation in intake and growth rates can cause a

variety of physiological changes. Body composition is one of the most plastic characteristics of

organisms undergoing food restriction and subsequent realimentation. Sequential mobilization of

reserves typically occurs when animals are food-limited, with lipid stores depleted preferentially

compared to protein stores (e.g., Cherel et al. 1993, Tian and Qin 2004). During realimentation

and CG, differential accretion of lipid and/or protein also occurs. Typically, the early phases of

compensatory growth are characterized by lean tissue deposition while later phases are

characterized by fat deposition. This differential accretion of lean tissue during the early stages









of realimentation may provide a mechanism for accelerated growth, as lean tissue deposition

requires less energy than fat deposition (Hornick et al. 2000).

In this study, I examined the capacity for and mechanisms of CG in juvenile green turtles

(Chelonia mydas). The green turtle leads an oceanic existence for the first several years of life

(Carr 1987) and consumes a largely carnivorous diet during this time (Reich et al. in review).

Intake of juvenile turtles in this stage most likely varies stochastically due to heterogeneous prey

distribution. At a size of approximately 20-25 cm carapace length (for Atlantic C. mydas) or 35

cm carapace length (for Pacific C. mydas), green turtles undergo an ontogenetic shift in habitat

use and diet by recruiting to neritic habitats, where they consume an herbivorous diet consisting

of various species of algae and seagrass (Bjorndal 1997). Although growth dynamics of juveniles

during the oceanic stage are unknown, post-recruitment growth rates are known to vary

temporally as a result of variation in oceanographic conditions (Limpus and Chaloupka 1997)

and population density (Bjorndal et al. 2000). This variation in juvenile growth rates may have

substantial effects on fitness, as body size is correlated with juvenile survivorship (Chaloupka

and Limpus 1998) and clutch size (Broderick et al. 2003) in C. mydas.

Because of the potential for stochasticity of the nutritional environment during the oceanic

and neritic stages of development and the effect of body size on survival and reproductive output

in C. mydas, I predicted that green turtle juveniles should be capable of CG. Loggerhead sea

turtles, which are also largely carnivorous as young juveniles (Bjorndal 1997), have been shown

to undergo CG in the wild (Bjorndal et al. 2003), although the proximate explanation for this

growth pattern is not known. By manipulating food availability under controlled conditions, I

tested whether previously food-restricted green turtles can undergo CG and evaluated

hyperphagia and enhanced FCE as potential mechanisms for this growth pattern.









Materials and Methods


Animal Care

All animal care components of this study were performed at the Cayman Turtle Farm in

Grand Cayman, British West Indies, in accordance with the guidelines of the Institutional

Animal Care and Use Committee at the University of Florida (permit #Z061). Chelonia mydas

hatchlings (n = 115) were housed individually in sea water in 68-liter tanks arranged within a

large outdoor concrete enclosure. Tanks were systematically arranged within the enclosure to

minimize position effects. Fresh sea water was continuously circulated within the enclosure (at a

depth of approximately 20-25 cm) to dampen the daily cycle of heating and cooling within each

of the tanks. Water temperature was monitored daily using five min/max thermometers

distributed throughout the array of tanks.

All turtles were maintained on an ad libitum diet for seven days prior to the beginning of

the study to establish average daily ad libitum intake. Turtles were then randomly assigned to

one of three treatment groups: ad libitum (AL), restricted (R), and restricted-ad libitum (R-AL).

Turtles in the AL group were offered food ad libitum for twelve weeks. Turtles in the R group

were offered approximately 50% of average initial mass-specific ad libitum daily intake for

twelve weeks. Because AL turtles increased their mass-specific intake after week 0, the actual

amount of food consumed by restricted turtles amounted to an average of 27% of the daily

mass-specific intake of AL turtles. This ration was sufficient to maintain restricted turtles on a

positive growth trajectory throughout the study. Turtles in the R-AL group were offered the

restricted diet for five weeks and were then offered food ad libitum for the remaining seven

weeks.

Turtles were provided 2.6-mm turtle pellets (Melick Aquafeed, Catawissa, PA) twice daily

and were allowed to feed for a total of seven to ten hours each day. Pellets remaining in each









tank were counted every afternoon, and approximate intake was calculated based on the average

mass per pellet (determined weekly), the known mass of pellets offered, and the number of

pellets remaining. Intake was estimated in this way for each turtle on each of six days a week

(weather permitting), with the seventh day reserved for tank cleaning. The water in each tank

was replaced daily. A subset of pellets was counted and weighed each week to determine average

pellet mass. Five food samples were weighed and dried every two weeks for nutrient analyses.

Body mass (BM, to the nearest 0.1 g) and minimum straight carapace length (CL, to the

nearest 0.05 mm) of each turtle were measured weekly. Daily intake measurements and weekly

body size measurements were used to calculate average daily intake (g/day), mass-specific daily

intake (g/g*day), condition index (CI, g/cm3) (Ricker 1975), food conversion efficiency (FCE,

g/g and mm/g), and specific growth rate (SGR, %/day) for each turtle for each week of the study.

FCE and SGR were calculated for both BM and CL using the following formulas:

FCE = (sizet+l sizet)/(average daily intake 7)
SGR= 100*[ln(sizet+i)-ln(sizet)]/7

where size is BM or CL in one week and size t+ is BM or CL in the following week. FCE and

SGR were calculated for both BM and CL because gut fill accounted for up to 15.4% of total wet

BM, and changes in CL are not affected by the mass of gut contents. Furthermore, straight CL is

considered to be the most reliable measure of growth in green turtles (Balazs 1995).

Gut Morphology and Body Composition

Turtles were sacrificed prior to, during, and after the study for an analysis of gut

morphology and body composition. Ten randomly chosen turtles (to AL) were euthanized prior

to the initiation of the study, at which time all turtles had been feeding ad libitum for at least one

week. At the conclusion of the fifth week of the experiment (immediately prior to switching

R-AL turtles to an ad libitum diet), ten AL (t5 AL), five R, and five R-AL turtles were









euthanized. The R-AL turtles had, until the end of week five, been maintained on the restricted

feeding treatment. Data for the five R and five R-AL turtles were therefore pooled into one group

(t5 R). Ten AL turtles (t12 AL), ten R turtles (t12 R), and ten R-AL turtles (t12 R-AL) were

euthanized at the conclusion of the twelve-week trial. Turtles were weighed to the nearest 0.1 g

and then euthanized using an intramuscular overdose injection of ketamine (Ketaset, 100 mg/kg

body mass).

The liver and digestive tract (from the esophageal-gastric junction to the termination of the

hindgut anterior to the cloaca) of each turtle were removed. The full gut was measured to the

nearest 0.05 mm. Measurements included straight stomach length (SSL) and total intestine length

(TIL). Small intestine and large intestine lengths could not be determined because the distinction

between midgut and hindgut was not easily discernible. Gut contents from each turtle were

gently removed from the excised gut using forceps. Wet masses of gut contents, liver, empty

stomach, and empty intestine were determined for each turtle. Organ mass and length indices

were calculated using the following formulas:

Mass Index = 100*Mor/(BM MGC)
Length Index = Lor/CL

where Mor is wet mass of each organ (liver, stomach, or intestine), MGC is wet mass of gut

contents, and Lor is length of stomach or intestine.

All tissues and carcasses were dried at 60 C for a minimum of seven days. Dried body

tissues (liver, stomach, intestines, and carcass) were recombined for each turtle (n = 10 to, 20 t5,

and 30 tl2 turtles) and ground for nutrient analyses. Each turtle was ground in a mill (C.W.

Brabender Instruments, Inc., South Hackensack, NJ) with dry ice. Dried food samples (collected

every two weeks during the study) were also ground in the mill (without dry ice). Dry matter

(DM) content of each turtle and each food sample was determined by drying subsamples at









105 C for 16 hours, and organic matter (OM) content was determined by combustion at 500 C

for three hours (AOAC 1960). Energy content of each turtle and food sample was determined by

bomb calorimetry (Parr 1960; Parr Instrument Co., Moline, IL). Lipid content was determined by

ether extraction using a soxhlet apparatus (AOAC 1984). Nitrogen content was determined using

a modified Kjeldahl procedure. Samples were digested for at least four hours at 375 C using a

modification of the aluminum block digestion procedure of Gallaher et al. (1975). Nitrogen in

the digestate was determined by semiautomated colorimetry using a Technicon Autoanalyzer

(Hambleton 1977; Pulse Instrumentation, Ltd., Saskatoon, SK, Canada). All nutrient analyses

were performed in duplicate unless relative error exceeded 2.0%, in which case additional

analyses were performed. The ratio of lipid to lean contents was calculated by dividing DM lipid

content by DM protein content (% nitrogen 6.25; Hambleton 1977).

Statistical Analyses

Data for food samples collected at two-week intervals were analyzed using Kruskal-Wallis

tests. Weekly data were analyzed using repeated measures ANOVA with Tukey's Honestly

Significant Difference (HSD) post hoc tests. Data for each week of the study and for midpoint

and endpoint samples were also tested for significance using one-way ANOVA with Tukey's

HSD post hoc tests. Bonferroni corrections were not used to account for multiple comparisons

among weeks with one-way ANOVA because of the risk of inflated Type II error (Perneger

1998). Data for t5 and t12 samples were tested for normality (using Shapiro-Wilk test) and

homogeneity of variances (using Levene's test) prior to parametric analysis. If both tests yielded

a significant result (p < 0.05), data were transformed using a natural log, reciprocal, square root,

square, reciprocal square, or reciprocal square root transformation. If transformation did not

improve normality or homoscedasticity, or if no post hoc test could be performed (e.g., for t5

samples), data were tested for statistical significance using a Kruskal-Wallis test and pairwise









Mann-Whitney U tests with a set at 0.017 (for t12 samples) to account for multiple comparisons.

Analysis of covariance (ANCOVA) could not be used to evaluate midpoint and endpoint data for

organ sizes because the assumption that covariate values have similar distributions and ranges

for all treatment groups (Quinn and Keough 2002) was violated.

All data were analyzed using SPSS for Windows (Release 11.0.0). Only turtles that

survived, displayed no external signs of illness, and continued to gain mass until the time of

tissue sampling were included in the analyses. Data are expressed as means + standard errors

with alpha set at 0.05 unless otherwise noted.

Results

OM, energy, nitrogen, and lipid contents of food samples (n = 7, collected at biweekly

intervals) were consistent throughout the experiment (Table 2-1). Although differences in energy

content of pellets among weeks approached significance, the relative difference between the

highest and lowest energy values measured was only 2.87% (for DM) and 2.73% (for OM).

R and R-AL turtles had comparable values for all repeated measures data collected during

weeks 0 through 5. Data for these two groups were analyzed and are presented separately for this

time period to demonstrate that there were no differences between groups prior to the switch to

an ad libitum diet for R-AL turtles.

Intake in week 0 (during which all turtles were feeding ad libitum) was similar for all three

treatment groups (ANOVA, F2,112= 0.946, p = 0.392). Repeated measures ANOVA of intake

during weeks 1 through 5 (n = 37 AL, 39 R-AL, and 39 R) and during weeks 6 through 12 (n =

17 AL, 29 R-AL, and 29 R) revealed significant time and treatment effects as well as significant

interactions between time and treatment group (Table 2-2). The pattern of AL intake was

significantly different from the patterns of R and R-AL intake during weeks 1 through 5

(Tukey's HSD post hoc test, p < 0.0001 for both comparisons), and intake patterns of all three









treatment groups were significantly different during weeks 6 through 12 (Tukey's HSD post hoc

test, p < 0.0001 for all comparisons). Because body size of turtles in each group differed after

week 0, intake was corrected for BM and re-analyzed as mass-specific intake.

The magnitude but not the overall pattern of mass-specific intake of R turtles differed from

that of AL turtles for the duration of the experiment, and the magnitude and overall pattern of

mass-specific intake of R-AL turtles was comparable to that of R turtles during weeks 1-5 and

comparable to that of AL turtles during weeks 6-12. Repeated measures ANOVA of mass-

specific intake (Table 2-2) revealed no significant linear interaction between time and treatment

group (p > 0.05 for time treatment interactions using tests of within-subjects contrasts in

repeated measures ANOVA) during either weeks 1 through 5 or weeks 6 through 12. However,

there was a significant linear effect of time on mass-specific intake during weeks 1 through 5 and

during weeks 6 through 12, as demonstrated by the decrease in mass-specific intake for turtles in

all treatment groups during these two time intervals (Fig. 2-1). The pattern of mass-specific

intake of AL turtles was significantly different from those of R and R-AL turtles during weeks 1

through 5 (Tukey's HSD post hoc test, p < 0.0001 for both comparisons), and the pattern of

mass-specific intake of R turtles was significantly different from those of AL and R-AL turtles

during weeks 6 through 12 (Tukey's HSD post hoc test, p < 0.0001 for both comparisons).

Body size (BM and CL) in week 0 (during which all turtles were feeding ad libitum) was

similar for all three treatment groups (ANOVA, F2,112 = 0.992, p = 0.374 for BM; F2,112 = 1.109,

p = 0.333 for CL) Repeated measures ANOVA of body size during weeks 0 through 5 and

during weeks 6 through 12 revealed significant time and treatment effects as well as significant

interactions between time and treatment group (Table 2-3). The pattern of body size (both BM

and CL) of AL turtles was significantly different from those of R and R-AL turtles during weeks









0 through 5 (Tukey's HSD post hoc tests, p < 0.0001 for both comparisons for BM and CL), and

the patterns of body size (both BM and CL) of all three treatment groups were significantly

different during weeks 6 through 12 (Tukey's HSD post hoc tests, p < 0.0001 for all comparisons

for BM and CL). Within individual weeks, body size of AL turtles was significantly greater than

those of R and R-AL turtles in weeks 1 through 6 (for BM) and in weeks 1 through 7 (for CL),

and body size of all three groups differed significantly in weeks 7 through 12 (for BM) and in

weeks 8 through 12 (for CL) (Fig. 2-2).

There were significant effects of time and treatment on condition index (CI) as well as a

significant time by treatment interaction during weeks 0 through 5. There was no significant

overall effect of time on CI during weeks 6 through 12, although there was a significant

interaction between time and treatment group (Table 2-3). The pattern of CI in AL turtles was

significantly different from those of R and R-AL turtles during weeks 0 through 5 (Tukey's HSD

post hoc test, p < 0.0001 for both comparisons), and the pattern of CI in R turtles was

significantly different from those of AL and R-AL turtles during weeks 6 through 12 (Tukey's

HSD post hoc test, p < 0.0001 for both comparisons). The difference between AL and R-AL

turtles approached significance (Tukey's HSD post hoc test, p = 0.076). Within individual

weeks, CI of AL turtles was significantly greater than CI of R and R-AL turtles during weeks 1

through 5, and CI of R-AL turtles was intermediate between those of R and AL turtles in week 6.

In weeks 7 through 12, CI of R turtles was significantly lower than CI of R and R-AL turtles, and

CI of R-AL turtles was consistently but not significantly lower than CI of AL turtles (Fig. 2-3).

Specific growth rates (SGR) for BM and CL also differed among treatment groups for all

weeks of the experiment. Repeated measures ANOVA of SGR revealed significant time and

treatment effects and interactions between time and treatment group during weeks 1 through 5









and during weeks 6 through 12 (Table 2-4). The SGR patterns (for both BM and CL) of AL

turtles were significantly different from those of R and R-AL turtles during weeks 1 through 5

(Tukey's HSD post hoc tests, p < 0.0001 for both comparisons for BM and CL), and the SGR

patterns (for both BM and CL) of R turtles were significantly different from those of AL and

R-AL turtles during weeks 6 through 12 (Tukey's HSD post hoc tests, p < 0.0001 for both

comparisons for BM and CL). Compensatory growth occurred during weeks 7 through 9, as

demonstrated by significantly larger SGR in R-AL turtles relative to AL turtles in weeks 7 and 8

(for SGRbm) and in weeks 8 and 9 (for SGRli) (Fig. 2-4).

Food conversion efficiencies (FCE) for BM and CL differed among treatment groups, but

the patterns depended on whether FCE was calculated as mass gain per unit of food consumed or

as carapace length gain per unit of food consumed. Repeated measures ANOVA of FCE revealed

significant time and treatment effects during weeks 1 through 5 and 6 through 12 and significant

interactions between time and treatment group during weeks 6 through 12 (Table 2-5). The

pattern of FCEbm of AL turtles was significantly different from those of R and R-AL turtles

during weeks 1 through 5 and during weeks 6 through 12 (Tukey's HSD post hoc tests, p < 0.01

for all comparisons). The pattern of FCEcl of AL turtles likewise differed from those of R and R-

AL turtles during weeks 1 through 5, but the pattern of FCEcl for all treatment groups differed

significantly for weeks 6 through 12 (Tukey's HSD post hoc tests, p < 0.0001 for all significant

comparisons). Specifically, the two groups feeding ad libitum after week 5 differed in conversion

efficiencies, with R-AL turtles demonstrating enhanced FCEbm in weeks 6 and 7 and enhanced

FCEci between weeks 6 and 11 (Fig. 2-5). Trends in conversion efficiencies were more consistent

among weeks when FCE was calculated as change in carapace length per unit of food consumed.









Visceral organ size expressed as a percentage of BM (minus MGC) or CL was smaller in

food-restricted turtles than in ad libitum-fed turtles, particularly at t12 (Tables 2-6 and 2-7). Liver

was lighter and stomach and intestine were shorter in R turtles than in AL turtles at t5. At the end

of the feeding trial, liver and stomach were lighter and stomach and intestine were shorter in R

turtles than in AL or R-AL turtles. Intestine mass was lighter in R turtles than in R-AL turtles at

t12, and the difference between R and AL turtles approached significance (p = 0.052). Turtles at 5

weeks and 12 weeks also differed significantly in body composition (Tables 2-6 and 2-8), with R

turtles having higher N and lower OM, lipid, energy, and lipid:lean contents than AL turtles at

both t5 and t12. Body composition of R-AL turtles at t12 was comparable to that of AL turtles.

Daily water temperatures dropped slightly as the experiment progressed into late autumn.

Occasional variation in temperatures was the result of rainfall from tropical weather systems,

including a hurricane that occurred during week 8 (Fig. 2-6).

Discussion

The data clearly demonstrate that green turtles are capable of CG within the first three

months after hatching. This growth pattern was not simply the result of rapid gut filling after the

diet switch, as both BM and CL increased proportionally faster in R-AL turtles than in AL

turtles. However, turtles undergoing CG achieved only partial compensation, because body size

of R-AL individuals did not catch up to that of AL individuals by the time the period of CG

ended. The fact that smaller, previously food-limited turtles grew faster than larger turtles of the

same age represents a strategy that decreases the phenotypic variance among individuals from

the same cohort (Wilson and Reale 2006). It appears that green turtles are able to evaluate their

growth patterns and adjust them, at least somewhat, toward a more "optimal" trajectory when

nutritional conditions improve. Because the risk of mortality from predation decreases as size

increases for juvenile sea turtles (Musick and Limpus 1997), this capacity for rapid growth when









food availability is high allows young green turtles to expedite progression through the

vulnerable hatchling stage.

The mechanism for CG in juvenile green turtles is enhanced FCE rather than hyperphagia.

Given that CG is often effected through increased intake, particularly in fish (e.g., Jobling and

Johansen 1999, Johansen et al. 2001, Nikki et al. 2004, Tian and Qin 2004), I expected

previously restricted green turtles to become hyperphagic when food was provided ad libitum.

The fact that green turtles were not hyperphagic suggests that intake rates of AL turtles were

already maximal and therefore could not be exceeded by R-AL turtles. McCauley and Bjorndal

(1999) previously demonstrated that juvenile loggerhead sea turtles also do not increase intake

beyond the maximal rate attained when feeding ad libitum. Animals that do not increase

consumption rates even when food is abundant may be intrinsically limited by constraints on

rates of digestion and/or passage (Speakman and Kr6ol 2005). In contrast, the hyperphagic

response of fish undergoing CG indicates that continuously ad libitum-fed fish feed at sub-

maximal rates.

It is possible that enhanced conversion efficiency was effected through improved

digestibility via either behavioral or morphological changes. Although I did not quantify

behavior throughout the study, the tanks of food-restricted turtles rarely contained feces because

these individuals practiced coprophagy. Reingestion of feces when food is limited has been

reported in rodents (Kenagy et al. 1999) and serves to increase digestive efficiency. The extent to

which R-AL turtles relied on nutrient recycling is unknown, but continued reingestion of feces

after the switch to an ad libitum diet could have allowed for the improved FCE and elevated

growth rates I observed. Digestibility could also have been increased by upregulation of

intestinal surface area. However, preliminary data on intestinal histology indicate that









food-restricted turtles had not only smaller guts than ad libitum turtles in terms of mass and

length but also decreased intestinal surface area (Roark and Bjorndal, unpublished data). Because

I did not evaluate density of epithelial transporters, I cannot rule out the possibility that uptake

rates may have been enhanced in food-restricted turtles despite a significant reduction in

intestinal mass, as has been shown in food-restricted birds (Brzek and Konarzewski 2001).

Alternatively, the enhanced FCE of R-AL turtles relative to AL turtles, especially in the

first few weeks after the switch to ad libitum feeding, may have resulted from decreased

metabolic expenditure. I did not quantify metabolic rates, but I did find that several major

visceral organs (e.g., liver, stomach, and intestine) in R turtles were smaller than those in AL

turtles. By down-sizing organs that would otherwise require disproportionate metabolic

expenditure to maintain them, R turtles may have been able to allocate less of their assimilated

food into maintenance metabolism and more into growth. Similar results have been obtained in

studies of fasted or food-restricted migratory birds (Lee et al. 2002, Karasov et al. 2004), which

had significantly smaller digestive and assimilatory organs compared to birds fed ad libitum. In

the intestines, this decrease in size was due largely to changes in the mucosal layer of the villi.

The observed decrease in organ mass in these birds was reversed by several days of feeding ad

libitum.

Because I sampled only at times to, t5, and t12, I was unable to determine the time course

over which the organ sizes of R and R-AL turtles changed. Upon a return to ad libitum

conditions, R-AL turtles may have experienced a delay in up-regulation of visceral organ size.

The switch to ad libitum feeding at a time when maintenance expenditure was minimized would

have allowed for the rapid growth I documented in R-AL turtles during weeks 7 through 9 (Ali et

al. 2003).









Although both BM and CL increased proportionally faster in R-AL turtles than in AL

turtles, these two morphometric measurements demonstrated different dynamics. During the first

week of ad libitum feeding, the SGRbm but not the SGRcl of R-AL turtles was comparable to that

of AL turtles. This decoupling of mass and length growth may result from differential allocation

of assimilated nutrients in the first week of ad libitum feeding. Less energy is required to convert

assimilated nutrients into reserve tissue than into more complex structural components such as

skeletal tissue (Broekhuizen et al. 1994). Growth efficiency would therefore be enhanced in

turtles that preferentially allocated nutrients to mobilizable tissue gain rather than to

unmobilizable tissue gain, at least in the initial stages of elevated growth. The time lag in the

increase of SGRcl but not in SGRbm may also result simply from a rapid increase in BM

immediately after the diet switch due to filling of the gut. At t12, gut contents accounted for an

average of 9.9%, 12.7%, and 13.5% of total wet BM for R, AL, and R-AL turtles, respectively.

Turtles on the restricted diet therefore carried proportionally less digesta than turtles feeding ad

libitum, meaning that gut filling probably accounted for a small percentage of the initial increase

in BM after the diet switch for R-AL turtles.

I used my measurements of BM and CL to assess body condition (as condition index, CI)

of turtles in each treatment group for each week of the study. Not surprisingly, CI of R turtles

decreased steadily until approximately the eighth week of the experiment, indicating that these

animals were becoming leaner as the study progressed. My body composition results support this

conclusion. Total body nitrogen content was higher and OM, lipid, and energy contents were

lower in R turtles than in AL turtles at both t5 and t12. Somatic growth in food-restricted turtles

therefore entailed either lower rates of lipogenesis and/or protein catabolism or higher rates of

protein deposition and/or lipid catabolism than in AL turtles. In other studies, food deprivation









has been linked to down-regulation of the activity of lipogenic enzymes (Bastrop et al. 1991,

Rosebrough and McMurtry 1993), and alterations in protein metabolism during food restriction

are also common (Dhahbi et al. 2001, Hagopian et al. 2003).

Turtles that experienced a switch from restricted to ad libitum feeding demonstrated a

rapid increase in CI after the diet switch, such that body condition of R-AL turtles was not

significantly different from that of AL turtles by week 7. The rapid growth of R-AL turtles was

accompanied by elevated lipid deposition between weeks 5 and 12. This increase in whole body

adiposity and concomitant decrease in nitrogen content allowed R-AL turtles to achieve a body

composition not significantly different from that of AL turtles by the end of the experiment.

Cessation of CG may have resulted from R-AL turtles attaining a tissue composition

similar to that of AL turtles. In other words, repletion of body stores may have served as a signal

regulating the duration of the compensatory response (Jobling and Johansen 1999, Ali et al.

2003). The fact that body composition was restored before full body size compensation was

achieved provides further evidence that CG may have been regulated by condition rather than

overall body size. In fish, several studies (e.g., Bull and Metcalfe 1997, Johansen et al. 2001)

have shown that the rate of repletion of lipid stores rather than the attainment of a certain body

size controlled the duration of the compensatory response. In contrast to my results, body

composition in these fish studies exerted its effects by altering appetite rather than FCE. Because

I found no evidence for hyperphagia, I conclude that juvenile green turtles, unlike many fish

species, do not adjust their intake in response to adiposity.

My finding that turtles experiencing consistently high food availability grew more slowly

than turtles undergoing compensatory growth implies that maximal growth rates may not always

be advantageous for green turtles. Despite the potential for increased body size to provide a









benefit to the individual in terms of fitness (Roff 1992, Steams 1992), fast growth may carry a

variety of costs (Arendt 1997, Blanckenhorn 2000, Metcalfe and Monaghan 2001). In other

animal models, these costs include delayed skeletal ossification (Arendt and Wilson 2000),

weakened musculature (Christiansen et al. 1992), reduced locomotor performance (Billerbeck et

al. 2001, Alvarez and Metcalfe 2005), accelerated telomere degradation (Jennings et al. 1999),

and decreased longevity (Olsson and Shine 2002). The proximate determinant of such costs may

be the accumulation of cellular damage during rapid growth, as modeled by Mangel and Munch

(2005). These detrimental effects of rapid growth may explain the sub-maximal growth rates

typically demonstrated by animals feeding ad libitum continuously (Mangel and Stamps 2001). I

have demonstrated that cellular antioxidant potential of R-AL turtles is decreased compared to

AL turtles, at least in mitotically active tissue (Chapter 4). If such costs place an upper limit on

growth in green turtles, they may further explain the transient and incomplete nature of the

compensatory response I observed.

This study is the first to document the existence of and mechanisms for CG in young green

turtles. The capacity to grow quickly, albeit only transiently, provides juveniles an opportunity to

mitigate some of the costs of being small. At the same time, however, the transitory nature of the

CG response suggests that the benefits of accelerated growth are countered by costs potentially

including decreased longevity and/or performance that may be mediated by altered antioxidant

function. The extent to which CG is possible at different ages and during different life stages is

unknown but deserves further study. For example, the ontogenetic shift in habitat use and diet

that green turtles undergo as juveniles may provide an opportunity for CG, as such niche shifts

often correspond to improved food availability (Ali et al. 2003). What is clear from this study is

that differences in food availability can induce plasticity in growth, morphology, and body









composition in young green turtles. This plasticity could substantially affect life-history

endpoints such as age and size at maturity, reproductive output, and longevity that directly

influence the viability of green turtle populations.









Table 2-1. Kruskal-Wallis test results for nutrient content of biweekly food samples (n = 2 at
each of 7 time points).
Source of Content 2
variation Mean SE /f x P
Per DM
OM (%) 91.95 + 0.04 6 7.371 0.288
Energy (kJ/g) 20.92 0.06 6 12.457 0.053
Nitrogen (%) 8.28 0.05 6 1.943 0.925
Lipids(%) 16.78 0.26 6 2.057 0.914
Per OM
Energy (kJ/g) 22.75 0.06 6 12.457 0.053
Nitrogen (%) 9.01 0.05 6 2.114 0.909
Lipids (%) 18.25 + 0.28 6 2.057 0.914
According to the Melick Aquafeed label, crude fiber was < 5% and phosphorus was > 1%.
Abbreviations: DM = dry matter, OM = organic matter.











Table 2-2. Repeated measures analyses of variance for weekly averages of daily intake and daily
mass-specific intake.
Source of variation df SS F p
Intake (g/day), weeks 1-5
Between subjects effects
Group 2 451.274 168.22 < 0.0001
Error 112 150.225
Within subjects effects
Time 4 22.823 57.65 < 0.0001
Group Time 8 44.743 56.51 < 0.0001
Error (Time) 448 44.343
Within subjects linear contrasts
Time 1 22.108 71.99 < 0.0001
Group Time 2 43.336 70.56 < 0.0001
Error (Time) 112 34.395
Intake (g/day), weeks 6-12
Between subjects effects
Group 2 1.747 x 103 103.56 < 0.0001
Error 72 607.367
Within subjects effects
Time 6 248.000 168.37 < 0.0001
Group Time 12 139.612 47.39 < 0.0001
Error (Time) 432 106.055
Within subjects linear contrasts
Time 1 236.124 296.96 < 0.0001
Group Time 2 129.989 81.74 < 0.0001
Error (Time) 72 57.250
Mass-specific intake (g/g*day), weeks 1-5
Between subjects effects
Group 2 5.497 x 10-2 359.41 < 0.0001
Error 112 8.566 x 10-3
Within subjects effects
Time 4 3.863 x 10-4 17.02 < 0.0001
Group Time 8 1.834 x 104 4.04 < 0.01
Error (Time) 448 2.543 x 10-3
Within subjects linear contrasts
Time 1 3.044 x 104 19.64 < 0.0001
Group Time 2 7.915 x 10-0 0.26 0.775
Error (Time) 112 1.736 x 10-3
Mass-specific intake (g/g*day), weeks 6-12
Between subjects effects
Group 2 4.330 x 10-2 387.27 < 0.0001
Error 72 4.025 x 10-3
Within subjects effects
Time 6 7.750 x 104 15.86 < 0.0001
Group Time 12 3.655 x 104 3.74 < 0.001
Error (Time) 432 3.518 x 10-3
Within subjects linear contrasts
Time 1 3.465 x 104 18.04 < 0.0001
Group Time 2 3.775 x 105 0.98 0.379
Error (Time) 72 1.383 x 10-3
Three groups were tested: ad libitum for 12 weeks, food-restricted for 12 weeks, and
food-restricted for 5 weeks followed by ad libitum for 7 weeks. When Mauchley's test indicated
that the compound symmetry assumption was violated, Greenhouse-Geisserp-values are
presented. Significant p-values are in bold.











Table 2-3. Repeated measures analyses of variance for weekly
length, and condition index.


body mass, straight carapace


Source of variation df SS F p
Body Mass (g), weeks 0-5
Between subjects effects
Group 2 7.591 x 104 75.97 < 0.0001
Error 112 5.596 x 104
Within subjects effects
Time 5 1.022 x 105 600.23 < 0.0001
Group Time 10 4.433 x 104 130.23 < 0.0001
Error (Time) 560 1.906 x 104
Body Mass (g), weeks 6-12
Between subjects effects
Group 2 1.411 x 106 68.24 < 0.0001
Error 72 7.442 x 105
Within subjects effects
Time 6 5.885 x 105 395.15 < 0.0001
Group Time 12 2.320 x 105 77.90 < 0.0001
Error (Time) 432 1.072 x 105
Carapace Length (mm), weeks 0-5
Between subjects effects
Group 2 6.012 x 103 50.02 < 0.0001
Error 112 6.731 x 103
Within subjects effects
Time 5 2.599 x 104 2711.94 < 0.0001
Group Time 10 3.180 x 103 165.88 < 0.0001
Error (Time) 560 1.073 x 103
Carapace Length (mm), weeks 6-12
Between subjects effects
Group 2 5.701 x 104 80.43 < 0.0001
Error 72 2.552 x 104
Within subjects effects
Time 6 2.891 x 104 1041.43 < 0.0001
Group Time 12 6.613 x 103 119.13 < 0.0001
Error (Time) 432 1.998 x 103
Condition Index (g/cm3), weeks 0-5
Between subjects effects
Group 2 1.937 x 10-2 22.75 < 0.0001
Error 112 4.767 x 10-2
Within subjects effects
Time 5 4.218 x 10-2 657.58 < 0.0001
Group Time 10 5.556 x 10-3 43.30 < 0.0001
Error (Time) 560 7.185 x 10-3
Condition Index (g/cm3), weeks 6-12
Between subjects effects
Group 2 3.872 x 10-2 33.42 < 0.0001
Error 72 4.171 x 10-2
Within subjects effects
Time 6 1.351 x 10-4 2.14 0.103
Group Time 12 7.117 x 10-4 5.63 < 0.0001
Error (Time) 432 4.553 x 10-3
Three groups were tested: ad libitum for 12 weeks, food-restricted for 12 weeks, and
food-restricted for 5 weeks followed by ad libitum for 7 weeks. When Mauchley's test indicated
that the compound symmetry assumption was violated, Greenhouse-Geisserp-values are
presented. Significant p-values are in bold.










Table 2-4. Repeated measures analyses of variance for weekly specific growth rates (SGR) for
body mass (bm) and carapace length (cl).
Source of variation df SS F p
SGRbm (%/day), weeks 1-5
Between subjects effects
Group 2 336.783 306.07 < 0.0001
Error 112 61.620
Within subjects effects
Time 4 115.394 205.55 < 0.0001
Group Time 8 9.937 8.85 < 0.0001
Error (Time) 448 62.875
SGRbm (%/day), weeks 6-12
Between subjects effects
Group 2 310.715 218.38 < 0.0001
Error 72 51.222
Within subjects effects
Time 6 18.779 18.27 < 0.0001
Group Time 12 7.770 3.78 < 0.0001
Error (Time) 432 74.002
SGRcI (%/day), weeks 1-5
Between subjects effects
Group 2 24.174 202.20 < 0.0001
Error 112 6.695
Within subjects effects
Time 4 27.471 562.96 < 0.0001
Group Time 8 1.494 15.31 < 0.0001
Error (Time) 448 5.465
SGRlI (%/day), weeks 6-12
Between subjects effects
Group 2 25.945 180.39 < 0.0001
Error 72 5.178
Within subjects effects
Time 6 2.029 38.20 < 0.0001
Group Time 12 1.851 17.43 < 0.0001
Error (Time) 432 3.824
Three groups were tested: ad libitum for 12 weeks, food-restricted for 12 weeks, and
food-restricted for 5 weeks followed by ad libitum for 7 weeks. When Mauchley's test indicated
that the compound symmetry assumption was violated, Greenhouse-Geisserp-values are
presented. Significant p-values are in bold.










Table 2-5. Repeated measures analyses of variance for food conversion efficiencies (FCE) for
body mass (bm) and carapace length (cl).
Source of variation df SS F p
FCEbm (g/g), weeks 1-5
Between subjects effects
Group 2 7.755 18.546 < 0.0001
Error 112 23.418
Within subjects effects
Time 4 13.142 36.504 < 0.0001
Group Time 8 1.278 1.775 0.086
Error (Time) 448 40.323
FCEbm (g/g), weeks 6-12
Between subjects effects
Group 2 1.932 5.975 0.004
Error 72 11.640
Within subjects effects
Time 6 3.261 7.224 < 0.0001
Group Time 12 4.082 4.522 < 0.0001
Error (Time) 432 32.498
FCEcl (mm/g), weeks 1-5
Between subjects effects
Group 2 24.260 198.288 < 0.0001
Error 112 6.851
Within subjects effects
Time 4 18.711 232.633 < 0.0001
Group Time 8 0.123 0.767 0.574
Error (Time) 448 9.008
FCEcl (mm/g), weeks 6-12
Between subjects effects
Group 2 6.792 130.635 < 0.0001
Error 72 1.872
Within subjects effects
Time 6 1.800 72.206 < 0.0001
Group Time 12 0.294 5.891 < 0.0001
Error (Time) 432 1.795
Three groups were tested: ad libitum for 12 weeks, food-restricted for 12 weeks, and
food-restricted for 5 weeks followed by ad libitum for 7 weeks. When Mauchley's test indicated
that the compound symmetry assumption was violated, Greenhouse-Geisserp-values are
presented. Significant p-values are in bold.










Table 2-6. Omnibus F, x2, and p-values for
weeks 5 and 12.


Omnibus F and /i


analyses of variance of dissection data collected at


Identity of Groups Tested in Pairwise Comparisons
AL and R-AL AL and R R-AL and R


Week 5
BM (g)
CL (mm)
LM Index (%)
SM Index (%)
TIM Index (%)
SSL Index
TIL Index
% OM
N (% DM)
N (% OM)
Lipid (% DM)
Lipid (% OM)
Energy (kJ/g DM)
Energy (kJ/g OM)
Lipid:Lean
Week 12
BM (g)
CL (mm)
LM Index (%)
SM Index (%)
TIM Index (%)
SSL Index
TIL Index
% OM
N (% DM)
N (% OM)
Lipid (% DM)
Lipid (% OM)
Energy (kJ/g DM)
Energy (kJ/g OM)
Lipid:Lean


F1,18
F1,18
F1,18
F1,18
F1,18
F1,18
F1,18
Fl,18





F1,18
F;2:s


30.06,p < 0.0001
21.98, p< 0.001
80.05,p < 0.0001
1.60,p = 0.222
1.03,p = 0.324
19.28,p < 0.001
8.03,p = 0.011
20.22, p < 0.001
12.10,p = 0.001
14.35,p < 0.001
13.72,p < 0.001
11.84, p < 0.001
31.33,p < 0.0001
14.29,p < 0.001
14.29,p < 0.001

227.3, p < 0.0001
73.08,p < 0.0001
107.8,p < 0.0001
14.55,p < 0.0001
4.91,p = 0.015
17.86, p < 0.0001
18.67,p < 0.0001
64.16, p < 0.0001
11.05,p < 0.001
19.94,p < 0.0001
63.75, p < 0.0001
51.73,p < 0.0001
69.97, p < 0.0001
44.34,p < 0.0001
73.56, p < 0.0001


0.050
0.061
0.765
0.544
0.892
0.150
0.697
0.454
0.099
0.280
0.562
0.468
0.467
0.172
0.696


< 0.0001
< 0.0001
< 0.0001
< 0.0001
0.052
< 0.0001
0.0001
< 0.0001
< 0.001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001


< 0.0001
< 0.0001
< 0.0001
0.001
0.019
0.001
< 0.0001
< 0.0001
0.043
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001


When F values are reported, data were analyzed using analysis of variance with Tukey's HSD
post hoc test. When Z2 values are reported, data were analyzed using a Kruskal-Wallis test and
pairwise Mann-Whitney U tests with a set at 0.017 to account for multiple comparisons among
t12 groups. Statistically significant p-values are indicated in bold. Abbreviations: AL = ad libitum
for 12 weeks, R = food-restricted for 12 weeks, R-AL = food-restricted for 5 weeks and ad
libitum for 7 weeks, BM = body mass, CL = carapace length, LM = liver mass, SM = stomach
mass, TIM = total intestine mass, SSL = stomach straight length, TIL = total intestine length,
OM = organic matter, DM = dry matter.










Table 2-7. Organ masses (mean + standard error) from turtles dissected at 0, 5, and 12 weeks reported as indices (organ mass or
length as a proportion of body mass or carapace length, respectively).
Group Week n BM (g) CL (mm) LM Index (%) SM Index (%) TIM Index (%) SSL Index TIL Index
AL 0 10 34.2 0.6 61.7 0.6 3.61 0.15 0.92 0.04 2.97 0.12 0.67 0.02 8.22 0.18
AL 5 10 92.6 9.6a 88.0 2.6a 4.04 0.17a 1.40 0.09a 3.89 0.21a 0.75 0.03a 9.08 0.22a
R 5 10 51.6 1.9b 75.2 1.0b 2.38 0.07b 1.24 0.09a 3.62 0.16a 0.61 0.02b 8.03 0.29'
AL 12 10 227.5 25.8a 117.4 4.3al 5.29 0.27a 1.58 0.11a 4.17 0.15ab2 0.82 0.01a 10.76 0.26a
R-AL 12 10 150.0 9.2b 105.7 1.7al 4.97 0.20a 1.45 0.07a 4.27 0.17a 0.78 0.02a 11.07 0.34a
R 12 10 71.7 1.0 85.9 0.9' 2.58 0.10b 0.97 0.05b 3.61 0.16b 2 0.69 0.02' 8.91 0.18b

iDifference between AL and R-AL approaches significance (p = 0.061).
2Difference between AL and R approaches significance (p = 0.052).
Within columns, values with different letter superscripts are significantly different within time periods (Kruskal-Wallis test or analysis
of variance followed by pairwise Mann-Whitney U tests or Tukey's HSD post hoc test, p < 0.05). Abbreviations are the same as in
Table 2-6.










Table 2-8. Body composition (mean + standard error) of turtles dissected at 0, 5, and 12 weeks reported as percent of dry matter (%
DM) and percent of organic matter (% OM).
Group Week n % OM Nitrogen Content Lipid Content Energy Content Lipid:Lean
% DM % OM % DM % OM kJ/g DM kJ/g OM
AL 0 10 85.5 0.3 11.96 0.09 13.99 0.08 20.3 0.4 23.8 0.4 20.9 0.1 24.4 0.1 0.272 0.006
AL 5 10 82.2 0.3a 12.20 0.21a 14.85 0.28a 15.5 1.4a 18.8 1.7a 19.7 0.3a 23.9 0.4a 0.206 0.023a
R 5 10 80.3 0.3b 13.00 0.06b 16.19 0.04b 10.0 0.1b 12.5 0.2b 17.8 0.1b 22.2 0.1b 0.123 0.002b
AL 12 10 81.4 0.3a 11.68 0.17a 14.35 0.25a 18.6 1.4a 22.8 1.6a 20.5 0.4a 25.1 0.4a 0.258 0.023a
R-AL 12 10 82.0 + 0.3a 12.07 0.12a 14.72 0.15a 16.6 0.8a 20.2 1.0a 20.0 + 0.3a 24.3 0.4a 0.221 0.013a
R 12 10 76.6 0.5b 12.54 0.10b 16.38 0.06b 9.4 0.3b 12.3 0.4b 16.7 0.1b 21.8 0.1b 0.120 0.004b

Within columns, values with different letter superscripts are significantly different within time periods. Treatment groups and data
analysis are the same as in Table 2-7.












0.04 -*-AL

a ----- R-AL

0.03 -a a a
a
:aa a a *aaa

| 0.02 a a




bb b
v-b b --b ...... Ob .......b ... .b .b .... .............. -...... ......
b b b b b b

0.00
0 1 2 3 4 5 6 7 8 9 10 11 12

Time (weeks)

Figure 2-1. Average mass-specific daily intake (mean + standard error) during each week of the
feeding trial. Different letters indicate values that are significantly different within
weeks (analysis of variance, Tukey's HSD post hoc test, p < 0.05). Turtles in the
R-AL group were switched from a restricted diet to an ad libitum diet at the
beginning of week 6. Sample sizes in weeks 0 through 5: n = 37 AL, 39 R-AL, and 39
R. Sample sizes in weeks 6 through 12: n = 17 AL, 29 R-AL, and 29 R.
Abbreviations: AL = ad libitum for 12 weeks, R = food-restricted for 12 weeks, R-AL
= food-restricted for 5 weeks and ad libitum for 7 weeks.















-*- AL
---0--- R-AL
---+-- R


350


300


250


) 200


? 150
CO
100


50


0


0 1 2 3 4(weeks)


Time (weeks)


7 8 9 10 11 12


--- AL
--...--- R-AL
---c-- R


b
b ..- -"] "


a b .ED'"
a b [.. .0 ....... ------ .... .

a. bb b b

S b bbbbb bb


0 1 2 3 4 5 6


7 8 9 10 11 12


Time (weeks)

Figure 2-2. Body mass (a) and straight carapace length (b) (mean + standard error) at the
midpoint of each week. Different letters indicate values that are significantly different
within weeks (analysis of variance, Tukey's HSD post hoc test, p < 0.05). The arrow
indicates the time at which turtles in the R-AL group were switched from a restricted
diet to an ad libitum diet. Sample sizes and abbreviations are the same as in Figure
2-1.


a a b
a b ,o--
ab
b 42
b
a b -... "c
. .. . ..- ------

bb b b b C


E
125

C
.J

c 100


U)

f 75

Co












U. If a -- AL
..." .. R-AL
0.16 ---+--- R

E a a a a
& a a I
i0.15 b b*

,b '.b """ .. .... . . . ..
b a a
.2 0.14 b ,aa ...

b
0 b j K-
1 ... .. -. -..... ... ...{ ... {
b b b b

0.12
0 1 2 3 4 5 T 6 7 8 9 10 11 12

Time (weeks)

Figure 2-3. Condition index (mean + standard error) in each week calculated as BMICL3, where
BM = body mass (g) and CL = carapace length (cm). Different letters indicate values
that are significantly different within weeks (analysis of variance, Tukey's HSD post
hoc test, p < 0.05). The arrow indicates the time at which turtles in the R-AL group
were switched from a restricted diet to an ad libitum diet. Sample sizes and
abbreviations are the same as in Figure 2-1.

















3.5

3.0


2.5


-*- AL
S...- R-AL

---+-- R


--- a
--S--Ff


~b ^ a .'" T 8 a

1.5 -
b
b", b

b b. -- ... b

0.5 .... --- '" ".. .


2 3 4 5 6 7

Time (weeks)


8 9 10 11 12


-.- AL
---a--- R-AL

---+-- R


1.2 aa





b b
0.8 "'""-,
,b b
0.6 b,

IN- b .. b
0.4- b b "---


0.2


b "'"*I-......0 ----- ..... -------.
c b c c -------....

b b
b


1 2 3 4 5 t 6 7 8 9 10 11 12

Time (weeks)


Figure 2-4. Specific growth rate (mean + standard error) for body mass (BM, a) and straight
carapace length (CL, b) during each week calculated as 100*(ln[sizet +]-ln[sizet])/7
where size = BM (a) or CL (b) and t = time (weeks). Different letters indicate values
that are significantly different within weeks (analysis of variance, Tukey's HSD post
hoc test, p < 0.05). The arrow indicates the time at which turtles in the R-AL group
were switched from a restricted diet to an ad libitum diet. Sample sizes and
abbreviations are the same as in Figure 2-1.


Z!
V




a


.2


CD
(U



U)
0:



0.
(C)


U
-I0

a)

0
.

0
.2
a:




U)
0

0.
(C)
















-*- AL
S...--- R-AL

---*-- R


2 3 4 5 6 7

Time (weeks)


8 9 10 11 12


a -*- AL

S------..... R-AL


a'.
a-e- R


a a a




-----------------5-.-----------------------------b
Sa a a

b" b b a''"*- a
b b .. .. b. b b .
b 2" -- ....... -. b


c c c


1 2 3 4 5 6


7 8 9 10 11 12


Time (weeks)


Figure 2-5. Food conversion efficiency (FCE, mean + standard error) for body mass (BM, a) and
straight carapace length (CL, b) during each week calculated as size change per unit
of food consumed, where size = BM (a) or CL (b). Different letters indicate values
that are significantly different within weeks (analysis of variance, Tukey's HSD post
hoc test, p < 0.05). The arrow indicates the time at which turtles in the R-AL group
were switched from a restricted diet to an ad libitum diet. Sample sizes and
abbreviations are the same as in Figure 2-1.


1.6

C 1.4

1.2
U
C
A) 1.0


.
LU 0.8

o 0.6

o 0.4
'a-
o 0.2
UL
oL


E 1.2
-I

1.0
00
U
r 0.8
._-
WJ 0.6
C
0
0.4
C
0
0.2

U-


















r33 a^r ^S






Max
Min


0 1 2 3 4 5 6 7 8 9 10 11 12

Week

Figure 2-6. Daily water temperatures (mean + standard deviation) throughout the feeding trial.
Water temperature was monitored using five min/max thermometers in tanks without
turtles. Water temperatures in weeks 8 and 9 fluctuated as a result of a hurricane.


35



'25-
25
2L 20
E

. 10


15 -

0









CHAPTER 3
BIOCHEMICAL INDICES AS CORRELATES OF RECENT GROWTH IN JUVENILE
GREEN TURTLES (Chelonia mydas)

Introduction

The green turtle, Chelonia mydas, is an endangered marine herbivore with a circumglobal

distribution (Seminoff 2002). Overexploitation of this species by humans during the last several

centuries has caused drastic population declines (Jackson et al. 2001). Development of effective

management plans for this species requires knowledge of demographic parameters such as

somatic growth rates. However, assessing growth rates for long-lived and far-ranging green

turtles typically requires time-consuming mark and recapture programs in which recapture

probabilities can be quite low (Limpus 1992). Furthermore, growth rates calculated using

morphometrics represent long-term, cumulative changes and often do not correlate well with

biochemical indices of short-term growth due to differences in the latency of these responses to

environmental influences (Ferron and Leggett 1994, Gilliers et al. 2004). Establishment of

alternative techniques for estimating recent growth rates of individual turtles upon first capture

would substantially improve our ability to evaluate the instantaneous status of C. mydas

populations and therefore to assess progress toward recovery goals for this endangered species.

Macromolecular indices (RNA concentrations, RNA:DNA ratios, RNA:protein ratios,

and/or protein:DNA ratios) are frequently measured as indicators of protein synthesis potential

and growth in marine fish and invertebrates (Bulow 1970, Carter et al. 1998, Buckley et al.

1999, Dahlhoff 2004, Caldarone 2005, Mercaldo-Allen et al. 2006, Vidal et al. 2006). These

indices are particularly useful for evaluating recent environmental conditions, as they reflect

differences in growth rates over a period of several days (Rooker and Holt 1996, Buckley et al.

1999, Vrede et al. 2002). The use of these indices depends on the assumption that total RNA

content of a cell (including messenger RNA, transfer RNA, and ribosomal RNA) should increase









as the cellular demand for protein synthesis and growth increases (Buckley et al. 1999), while

DNA content per cell should be relatively constant (Wallace 1992). The RNA:DNA ratio is

therefore an index of cellular protein synthesis capacity. Because nucleic acid concentrations and

the ratios between them respond rapidly to fluctuations in food availability, they are considered

reliable indices of instantaneous condition and growth (Rooker et al. 1997, Okumura et al. 2002,

Islam and Tanaka 2005, Vidal et al. 2006). Despite widespread use as sensitive measures of

recent growth rates in marine fish and invertebrates, nucleic acid ratios have not been validated

for application to studies of reptile growth.

The purpose of this study was to evaluate the use of morphometric and biochemical indices

as predictors of recent growth rates in green turtles maintained under controlled feeding

conditions. Because analyzing biochemical indices of growth typically requires homogenization

of tissues extracted from euthanized individuals, I also examined the potential for measuring

nucleic acid concentrations in whole blood, a tissue that is not typically tested in studies of this

kind. Validation of a physiological growth index that can be assessed using minimally invasive

sampling techniques and without sacrificing the animals would substantially enhance our ability

to monitor short-term responses of green turtles to environmental perturbations.

Materials and Methods

Animal Care

A twelve-week feeding trial was conducted at the Cayman Turtle Farm in Grand Cayman,

British West Indies, in accordance with the policies of the Institutional Animal Care and Use

Committee at the University of Florida (permit #Z061). Details of the animal care aspects of this

study can be found in Chapter 2. Briefly, Chelonia mydas hatchlings were housed individually in

sea water in 68-liter tanks. Turtles were fed turtle pellets (Melick Aquafeed, Catawissa, PA)

twice daily.









Prior to the beginning of the study, turtles were all fed ad libitum to establish average daily

intake. During the study, turtles in the ad libitum group (AL) were offered an excess of food each

day for 12 weeks, turtles in the restricted group (R) were fed 50% of average initial AL intake

each day for 12 weeks, and turtles in the restricted-ad libitum group (R-AL) were fed the

restricted amount of food for five weeks and then were fed ad libitum for seven weeks. The

amount of food offered during food restriction was sufficient to maintain turtles on a positive

growth trajectory. Turtles were weighed and measured (straight carapace length) each week.

Tissue Collection

At the conclusion of the twelve-week study, seven AL turtles, ten R turtles, and ten R-AL

turtles were weighed to the nearest 0.1 g and euthanized with an intramuscular overdose

injection of ketamine (Ketaset, 100 mg/kg body mass) in the right pectoral muscle. When each

turtle failed to respond to a pain stimulus, it was decapitated.

A blood sample was collected from the decapitation site (as in Storey et al. 1993 and

Packard et al. 1997). The heart and a portion of the right lobe of the liver were excised, and

blood, heart, and liver samples were snap-frozen in liquid nitrogen no more than three minutes

after decapitation. Tissues were maintained at -80 C until they were homogenized as described

below.

Biochemical Assays

Tissues from each individual turtle were analyzed for DNA and RNA concentrations.

Subsamples of frozen whole blood, heart, and liver were weighed, and DNA was isolated with

DNeasy kits (Qiagen Inc., Valencia, CA) using the manufacturer's protocol for animal tissues.

To isolate RNA, subsamples of frozen heart and liver tissue (different from those used for DNA

isolation) were weighed and then ground in liquid nitrogen using mortar and pestle. Subsamples

of blood (different from those used for DNA isolation) were weighed but not ground in liquid









nitrogen because of their tendency to thaw quickly. Frozen subsamples of blood and ground

heart and liver tissue were homogenized using QIAshredder spin columns (Qiagen Inc.). RNA

was then isolated with RNeasy Mini kits (Qiagen Inc.) using the manufacturer's protocol for

isolation of total RNA from animal tissues. DNA and RNA were isolated separately from a

minimum of three subsamples of each tissue from each turtle for a total of at least 18 subsamples

for each of 27 turtles unless tissue mass was insufficient.

The concentrations of DNA and RNA in each subsample of blood, heart, and liver tissue

were determined using a PicoGreen dsDNA quantitation kit (Invitrogen Corporation, Carlsbad,

CA) and a RiboGreen RNA-specific quantitation kit with DNase I (Invitrogen Corporation) by

measuring fluorescence at standard fluorescein wavelength settings (485 nm excitation, 528 nm

emission) using a fluorescent microplate reader. Data were collected using KCJuniorTM data

analysis software.

In addition, protein concentration of liver (but not heart or blood) was quantified for a

separate project (Chapter 4), so I included those data in the analysis of this study. Hepatic protein

concentrations were determined by Bradford assay (details, Chapter 4).

Statistical Analyses

Data for body mass (BM) and carapace length (CL) were used to calculate specific growth

rates (SGR) for each turtle during the final 10-11 days of the study according to the following

equations:

SGRbm = (ln[BMf] ln[BMi])*100/t
SGRcl = (ln[CLf] ln[CLi])*100/t

where BMi and CLi represent body size 10 or 11 days prior to tissue sampling, BMf and CLf

represent body size on the day of tissue sampling, and t represents time (10 or 11 days).

Condition index (CI) was calculated as Fulton's K (CI = BMf/CLf3, Ricker 1975). Data for BMf,









CLf, CI, SGRbm, SGRci, RNA concentration of each tissue, DNA concentration of each tissue,

RNA:DNA ratio of each tissue, liver protein concentration, liver protein:DNA ratio, and liver

RNA:protein ratio were compared for the three feeding treatments using analysis of variance

(ANOVA). All ratios were calculated as the quotient of the average protein, DNA, and/or RNA

concentration for a particular tissue. RNA:protein ratios should reflect RNA content per cell, but

only if protein:DNA ratios (as a measure of cellular protein content) are consistent among

treatment groups.

All data were tested for normality (using Shapiro-Wilk test) and homogeneity of variances

(using Levene's test) prior to parametric analysis and transformed, if necessary, using a natural

log, reciprocal, square root, square, reciprocal square, or reciprocal square root transformation. If

transformation did not improve normality, data were tested using a Kruskal-Wallis test followed

by pairwise Mann-Whitney tests with a Bonferroni adjustment for multiple comparisons. If

ANOVA revealed a significant difference, pairwise comparisons were evaluated using Tukey's

Honestly Significant Difference post hoc test (if variances were equal) or Tamhane's T2 post hoc

test (if variances were unequal). To evaluate repeatability of biochemical assays, coefficients of

variation (C.V.) were determined for RNA and DNA concentrations in liver, heart, and blood

and for protein concentration in liver.

Spearman's rank correlation test was used to test the strength of the relationships among

BMf, CLf, CI, SGRbm, SGRci, [RNA] in each tissue, [DNA] in each tissue, [RNA]: [DNA] in each

tissue, liver [protein], liver [protein]: [DNA], and liver [RNA]: [protein]. Regression models for

SGRbm and SGRcl were then developed using CI and biochemical indices as independent

variables. Although body length has been correlated with RNA:DNA ratios in fish (e.g., Rooker

et al. 1997), I did not include treatment group or any measure of total body size as independent









variables in my linear regression models. I chose not to include BM or CL as variables because

body size was strongly affected by diet treatment (Fig. 3-1), and the goal of this study was to

assess the applicability of RNA:DNA measurements in estimating recent growth rates of wild

turtles with unknown dietary histories.

Regression equations for SGRbm and SGRcI versus each biochemical index were

determined using least squares linear regression. Data were natural log-transformed, if necessary,

to linearize them and to decrease heteroscedasticity. I verified the assumptions of linear

regression by visually inspecting plots of Studentized deleted residuals versus standardized

predicted values. To construct comprehensive growth models for predicting SGR, data were

analyzed using stepwise multiple linear regression. The same transformations used for linear

regressions were used for stepwise multiple linear regressions. Condition index and all

biochemical indices measured for a particular tissue (liver, heart, or blood) were included in

separate models. A growth model incorporating condition index and all biochemical indices

measured for all tissues was also constructed. To enter a model, variables had to meet a 0.05

significance level. All statistical tests were performed using SPSS for Windows (Release 11.0.0).

Means are reported + standard errors with alpha set at 0.05.

Results

When fed to satiation, green turtle juveniles in the final 10-11 days of the 12-week trial

grew at an average SGRbm of 1.84% and 2.01% per day and an average SGRcI of 0.68% and

0.64% per day for AL and R-AL individuals, respectively. Food-restricted turtles grew much

more slowly at an average SGRbm of 0.34% per day and an average SGRcI of 0.15% per day (Fig.

3-1 and Table 3-1).

Intake and growth patterns significantly affected all morphometric measurements of body

size (Fig. 3-1 and Table 3-1). At the time of tissue sampling, AL turtles were significantly









heavier and longer than both R-AL and R turtles, and R turtles were significantly lighter and

shorter than both AL and R-AL turtles. Despite differences in body size between AL and R-AL

turtles, condition indices for these two groups were comparable at the time of tissue sampling,

and CI of R-AL turtles was significantly greater than CI of R turtles. The difference in CI

between AL and R turtles approached significance (p = 0.057). Although R-AL turtles were

food-restricted for the first five weeks of the study, they grew more rapidly than AL turtles

during weeks seven through nine after the switch to ad libitum feeding, but this period of

compensatory growth ended prior to tissue sampling (Chapter 2). As a result, R-AL turtles were

growing at comparable rates to AL turtles during the last 10-11 days of the study, and R turtles

were growing significantly slower than both AL and R-AL turtles. Significant differences among

treatment groups also existed for many of the biochemical indices I measured (Figs. 3-2 and 3-3

and Table 3-1). The patterns exhibited by the various biochemical indices varied depending on

the tissue analyzed, particularly for [RNA] and [RNA]:[DNA] ratios.

Many of the morphometric and biochemical indices I measured demonstrated significant

positive or negative correlations (Table 3-2). For most indices, significant correlations with

growth rates were stronger when growth was expressed as SGRbm rather than as SGRli. In some

cases (e.g., [protein]liver, [RNA]heart, [RNA]: [DNA]heart, and [RNA]: [DNA]blood), the correlations

between indices and growth were significant only for SGRbm. Heart yielded the lowest and liver

yielded the highest number of significant correlations between morphometric and biochemical

indices and growth.

When SGRbm and SGRcl were regressed against each index independently, all indices

except [RNA]blood, [DNA]heart, and [protein]: [DNA]iiver yielded significant relationships (Table









3-3). The R2 values for significant relationships ranged from 0.161 to 0.659, with the best fits

achieved by regressing SGRbm and SGRI against [RNA]iiver.

Stepwise multiple linear regression analyses for each individual tissue yielded a series of

nine significant growth models (Table 3-4). SGRbm was the dependent variable for five models,

with two models (1-2) based on liver, one model (3) based on heart, and two models (4-5) based

on blood. SGRcI was the dependent variable for the final four models, with one model (6) based

on liver, one model (7) based on heart, and two models (8-9) based on blood. The significant

independent variables predicting growth rate in each of these equations are listed in the table in

the order in which they were selected by the models.

When condition index and all biochemical indices for all tissues were combined and

analyzed using stepwise multiple linear regression, the resulting models were identical to models

1 and 2 (for SGRbm) and model 6 (for SGRci). The growth equation that best estimated recent

growth rate was Model 2. Despite the strong coefficient of determination for several SGR

models, coefficients of variation for RNA and DNA concentrations in liver, heart, and blood and

for protein concentration in liver (Table 3-5) were fairly substantial, indicating a high degree of

interassay variation.

Discussion

The purpose of this study was to evaluate the use of morphometric and biochemical indices

for predicting recent growth rates in juvenile green turtles. Validation of assays with substantial

predictive power for estimating growth would provide a less intensive alternative to tag and

recapture programs and facilitate population monitoring in this endangered species. Nucleic acid

concentrations and ratios hold promise as potential biomarkers of recent growth, as RNA content

of tissues increases with feeding and growth in many marine organisms including krill (Shin et

al. 2003), cephalopods (Melzner et al. 2005, Vidal et al. 2006), tuna (Carter et al. 1998),









haddock (Caldarone 2005), flounder, and tautog (Kuropat et al. 2002). Given the applicability of

tissue nucleic acid content to growth studies in these organisms, I expected to find strong

positive correlations between growth, RNA and/or protein concentrations, and ratios among

nucleic acids and protein concentrations in green turtles.

Contrary to my expectations, the biochemical indices I measured were neither consistently,

nor always positively, correlated with feeding treatment and growth rates. Perhaps most

surprisingly, liver RNA concentration was inversely correlated with SGR. I therefore infer that

slow-growing R turtles had more total RNA, and consequently higher putative protein synthesis

capacity, per unit of liver wet mass than fast-growing AL or R-AL turtles. Conversely, heart

RNA concentrations in this study were positively correlated with SGRbm (but not with SGRcl) as

expected, although this relationship was not strong. Growth rate had no apparent correlation with

blood RNA content.

The pattern between DNA and growth rate was quite different from that between RNA and

growth rate. Concentrations of DNA in blood and liver (but not in heart) were both negatively

correlated with SGR, a trend that has also been noted in fish (Mercaldo-Allen et al. 2006).

Because DNA concentration is a measure of the density of nuclei and therefore correlates with

cell number, I conclude that total blood cell count increases in response to food restriction. It is

unclear which of the six predominant types of nucleated blood cells in green turtles (Wood and

Ebanks 1984) accounts for this increase in blood cell number. The typical hematological

response to caloric restriction is either no change (Lochmiller et al. 1993) or a decrease

(Maxwell et al. 1990b, Walford et al. 1992) in total leukocyte count, although the number of

circulating basophils and thrombocytes has been shown to increase in food-restricted birds

(Maxwell et al. 1990b, Maxwell et al. 1992). My DNA results could also reflect differences in









red blood cell counts. Hematocrit may be correlated with body size in green turtles (Wood and

Ebanks 1984, but see also Bolten and Bjorndal 1992), but this relationship (if it exists) should

result in elevated DNA concentrations in larger, rather than smaller, turtles. It is therefore

unlikely that my results for DNA concentration reflect a body size-dependence for this

parameter. Hematocrit does not normally increase during food restriction (e.g., Maxwell et al.

1990a, Lochmiller et al. 1993). However, Maxwell et al. (1990a) demonstrated that enhanced

erythropoiesis with concomitant microcytosis occurs in food-restricted birds, suggesting that my

DNA results may reflect differences in blood cell size between slow- and fast-growing turtles.

My DNA results also indicate that liver growth results from hypertrophy more than

hyperplasia in fast-growing green turtles. In reptiles, feeding has been shown to increase the size

of lipid droplets and glycogen deposits in hepatocytes, thus leading to hypertrophic growth of

liver cells (Starck and Beese 2002). Although I did not examine livers from my animals

histologically, farm-raised marine turtles fed ad libitum are known to have hepatocytes

dominated by large lipid droplets (Solomon and Tippett 1991). I surmise that a process of lipid

and glycogen deposition similar to that observed by Starck and Beese (2002) occurs to a greater

extent in green turtles feeding ad libitum than in food-restricted turtles, therefore leading to more

extensive hepatocyte hypertrophy in the former.

Increased lipid deposition in hepatocytes of fast-growing green turtles may also explain the

negative correlation I observed between hepatic protein concentration and SGR. In other studies,

however, overall protein content as well as protein content per cell was strongly and positively

correlated with growth rate (Carter et al. 1998, Caldarone 2005). To explore the mechanistic

basis for the discrepancy between my nucleic acid and protein concentrations and those found in

comparable studies using fish, I assessed cellular protein synthesis capacity by calculating ratios









of RNA:protein (for liver only) and RNA:DNA (for all tissues). These ratios should both provide

information about the protein synthesis capacity per cell, but the former index is only valid as a

measure of cellular RNA content if the protein:DNA ratio (as a measure of protein content per

cell) is unaffected by intake and growth. Because hepatic cellular protein content was influenced

by treatment, only RNA:DNA ratio is an appropriate index of cellular RNA content for this

study.

Many authors have demonstrated significant positive relationships between RNA:DNA

ratio (of muscle, liver, or whole body) and growth rate, particularly in fish (Westerman and Holt

1994, Carter et al. 1998, Caldarone 2005, Mercaldo-Allen et al. 2006). Given this common

result, I expected to find similar trends in my turtle tissues. Indeed, heart and blood RNA:DNA

ratios did correlate positively with growth, but they explained only a small percentage (16-28%)

of the variance in SGR. On the contrary, hepatic RNA:DNA ratios were inversely correlated with

SGR and explained 29-34% of the variance in SGR. I suggest several possible explanations for

this discrepancy among tissues.

The liver is a mitotically active tissue, and elevated rates of cellular proliferation can lead

to over-estimation of cell number (Darzynkiewicz et al. 1980). It is possible, therefore, that the

RNA:DNA ratios I calculated for liver of fast-growing turtles were under-estimates of the true

cellular RNA content in fast-growing turtles. However, the difference in these ratios between

fast-growing turtles in groups AL and R-AL and slow-growing turtles in group R is likely too

substantial to result from differences in rates of DNA synthesis alone. Instead, I suggest that my

RNA:DNA ratios reflect real, tissue-specific differences in cellular ribosomal RNA content.

Typically, RNA:DNA ratio declines as ribosomes are degraded during periods of food

deprivation (Clemmesen 1994). In my study, however, slow growth was induced by food









restriction rather than starvation, and food-restricted turtles were never in negative energy

balance. Studies in rodents have demonstrated that protein turnover rates increase in response to

caloric restriction and that enzymes involved in gluconeogenesis are upregulated in the liver

(Spindler 2001, Hagopian et al. 2003). A similar upregulation of metabolic enzyme production

may have occurred in the livers of my food-restricted turtles. Thus, the effect of intake and

growth rates on RNA:DNA ratios may differ depending on whether the individual is in positive

or negative energy balance and the physiological role of the tissue studied.

To expand the predictive power of the various indices I measured, I incorporated condition

index and all biochemical indices measured for a particular tissue (liver, heart, or blood) into a

series of models using stepwise multiple linear regression. The resulting predictive equations

explained a maximum of 68% of the variance in growth rate. This maximal predictive power was

achieved by model 2, in which SGRbm is estimated using liver RNA content and CI. The

remaining indices, including nucleic acid concentrations and ratios for heart and blood, did not

explain any additional variance in growth rate. Although a model for juvenile green turtle growth

in the Caribbean has previously been developed (Bjorndal et al. 2000), this model uses body size

to predict recent growth and therefore does not allow for discrimination of growth rates among

individuals of similar size that may have experienced different nutritional conditions.

Furthermore, the coefficient of determination for my Model 2 was greater than that of the earlier

model and therefore indicates that the combined use of morphometric and biochemical indices

holds promise for applications to studies of growth in wild populations. Specifically, using

nucleic acid content to predict growth rates (for body mass) increased the coefficient of

determination from 38% using CI alone to 55% using blood DNA content or to 68% using liver









RNA content. Predictive power is therefore substantially improved by incorporating biochemical

indices into growth models.

In the various growth models I tested, CI was repeatedly selected as an independent

variable with significant predictive power. Bjorndal et al. (2000) found a similar positive

correlation between condition index and recent growth rates in wild green turtles. These findings

are particularly interesting in light of criticisms of the use of ratio-based indices (Hayes and

Shonkwiler 2001) and suggest that, at least for green turtles, the use of "body condition" as

measured using Fulton's K (Ricker 1975) for predictive purposes is meaningful and appropriate.

The growth model I developed fails to explain 32% of the variance in growth rates. A

portion of this unexplained variability probably results from fairly large coefficients of variation

for the biochemical assays I performed. This variation could potentially have been improved by

measuring DNA and RNA concentrations from the same subsamples of tissue, but the nucleic

acid isolation kits I used precluded me from doing so. Additionally, a number of nucleic acid

quantification techniques are available (Caldarone et al. 2006), and it is possible that one of these

techniques might have allowed for improved precision in measuring DNA and RNA content. The

remaining unexplained variability in growth rate may result from a mismatch in the time scales

over which the various indices in the model accurately detect changes in growth. As condition

index relies on measurement of body mass (a result of tissue accretion) and body length (a result

of bony growth), it most likely provides information about longer term growth processes than

nucleic acid and protein concentrations, which presumably fluctuate over shorter time scales

(Ferron and Leggett 1994).

Because sacrificing wild green turtles to collect liver samples for measuring nucleic acid

concentrations is not an acceptable practice, the multivariate model that best predicted recent









growth (model 2) has limited applicability in studies of wild turtle demography. However, the

fact that several biochemical indices for blood (including DNA concentration and RNA:DNA

ratio) were significantly correlated with growth suggests that further calibration of these assays

for application to growth estimation is warranted. Indeed, 55% of the variance in body mass

growth was predicted using only CI and concentration of DNA in the blood. This coefficient of

determination represents a loss of only 13% of the maximal predictive power achieved by the

best model I developed. Both CI and blood DNA content are easily measured with limited

disturbance to the animal. In combination with morphometric measurements, the blood cells of

C. mydas may therefore allow for the development of minimally invasive techniques for

estimating recent growth rates in this endangered species.









Table 3-1. Omnibus F, x2, andp-values for comparisons of means among treatment groups for
the various morphometric and biochemical indices measured (n = 27 for each
variable).
Groups Tested in Pairwise Comparisons
Omnibus F and X2 AL and R-AL AL and R R and R-AL
Body Mass F2,24 = 99.476, p < 0.0001 < 0.0001 < 0.0001 < 0.0001
Carapace Length 2 = 22.013, p < 0.0001 < 0.001 < 0.001 < 0.0001
Condition Index F2,24 = 6.918,p = 0.004 0.684 0.057 0.004
SGRbm F2,24 = 28.863,p < 0.0001 0.616 0.002 < 0.001
SGRol F2,24 = 62.995,p < 0.0001 0.737 < 0.0001 < 0.0001
[RNA]iiver F2,24 = 28.953, p < 0.0001 0.946 < 0.0001 < 0.0001
[RNA]heart F2,24 = 6.076, p = 0.007 0.021 0.990 0.015
[RNA]blood F2,24= 1.946,p = 0.165 N/A N/A N/A
[DNA]iiver F2,24 = 12.349,p < 0.001 0.154 < 0.001 0.010
[DNA]heart F2,24 = 2.111, p = 0.143 N/A N/A N/A
[DNA]blood F2,24 = 5.278,p = 0.013 0.312 0.010 0.162
[RNA]:[DNA]iiver F2,24 = 6.546, p = 0.005 0.191 0.319 0.004
[RNA]:[DNA]heart F2,24 = 4.678, p = 0.019 0.240 0.524 0.015
[RNA]:[DNA]blood F2,24 = 5.224,p = 0.013 0.303 0.010 0.172
[Protein]liver F2,24 = 3.545, p = 0.045 0.320 0.630 0.036
[Protein]:[DNA]iiver F2,24 = 7.655,p = 0.003 0.012 0.003 0.793
[RNA]:[Protein]liver F2,24 = 13.438, p < 0.001 0.888 0.003 0.003
When F values are reported, data were analyzed using analysis of variance and pairwise
comparisons were evaluated using Tukey's Honestly Significant Difference or Tamhane's T2
post hoc tests. When 2 values are reported, data were analyzed using a Kruskal-Wallis test with
pairwise Mann-Whitney U tests. Statistically significant p-values are indicated in bold.
Abbreviations: SGR = specific growth rate, bm = body mass, cl = carapace length, [ ] =
concentration, AL = ad libitum for twelve weeks, R-AL = food-restricted for five weeks and ad
libitum for seven weeks, R = food-restricted for twelve weeks.









Table 3-2. Spearman's rank correlations (p) for morphometric (a) and biochemical indices for


liver (b), heart (c), and blood (d) (n


27 for each variable).


(a) Morphometrics
Variable CL CI SGRbm SGRcl
BM 0.988** 0.463* 0.693** 0.797*
CL 0.391* 0.649** 0.785**
CI 0.499** 0.465*
SGRbm 0.839


(b) Liver
Variable
BM
CL
CI
SGRbm
SGRcl
[RNA]iiver
[DNA]iiver
R:Diiver
[Protein]liver
P:Diiver
(c) Heart


[RNA]iiver
-0.675"
-0.654"
-0.521"
-0.734"
-0.733"


[DNA]iiver
-0.673"
-0.635"
-0.534"
-0.519"
-0.442"
0.609"


R:Diiver
-0.412*
-0.423*
-0.168
-0.618"
-0.601"
0.778
0.066


[Protein]liver
-0.287
-0.240
-0.135
-0.561"
-0.332
0.433*
0.423*
0.249


Variable [RNA]heart [DNA]heart R:Dheart
BM 0.031 -0.394 0.247
CL -0.033 -0.369 0.195
CI 0.429* -0.454* 0.631**
SGRbm 0.435* -0.292 0.523**
SGRlI 0.294 -0.217 0.374
[RNA]heart -0.062 0.800**
[DNA]heart -0.594**
(d) Blood
Variable [RNA]blood [DNA]blood R:Dblood
BM 0.357 -0.563 0.549
CL 0.339 -0.520 0.527**
CI 0.543** -0.237 0.441*
SGRbm 0.352 -0.549* 0.440*
SGRlI 0.284 -0.446 0.352
[RNA]blood -0.253 0.816**
[DNA]blood -0.680**
Significant correlations are indicated in bold. Asterisks indicate level of significance (*p < 0.05,
** p< 0.01). Abbreviations: BM = body mass, CL = carapace length, CI= condition index, SGR
= specific growth rate, [ ] = concentration, R:D = RNA:DNA ratio, P:D = protein:DNA ratio,
R:P = RNA:protein ratio.


P:Dliver
0.428*
0.426*
0.427*
0.068
0.248
-0.292
-0.615**
0.134
0.332


R:Pliver
-0.673
-0.675
-0.487
-0.536
-0.690
0.846
0.459
0.703
-0.051
-0.545









Table 3-3. Growth equation parameters for juvenile Chelonia mydas as determined by least
squares linear regression.
y x Intercept slope Adjusted R2 F p
Body Mass
SGRbm Ln[RNA]liver 8.040 -1.132 0.629 45.011 < 0.0001
SGRbm Ln[RNA]hean -2.133 0.955 0.161 5.990 0.022
SGRbm [RNA]blood 0.196 0.010 0.109 4.197 0.051
SGRbm Ln[DNA]liver 8.574 -1.276 0.306 12.482 0.002
SGRbm Ln[DNA]heart 6.808 -0.967 0.089 3.536 0.072
SGRbm Ln[DNA]blood 12.305 -1.828 0.298 12.016 0.002
SGRbm Ln(R:D)liver 1.610 -1.046 0.294 11.821 0.002
SGRbm Ln(R:D)heart 3.054 0.852 0.253 9.804 0.004
SGRbm Ln(R:D)blood 2.720 1.061 0.277 10.937 0.003
SGRbm [Protein]liver 2.878 -7.203 0.196 7.353 0.012
SGRbm Ln(P:D)liver 1.511 0.470 0.003 1.083 0.308
SGRbm Ln(R:P)liver 1.934 -0.987 0.366 15.993 < 0.001
SGRbm CI -3.940 38.475 0.379 16.897 < 0.001
Carapace Length
SGRol Ln[RNA]liver 2.549 -0.353 0.659 51.159 < 0.0001
SGRol Ln[RNA]heart -0.368 0.229 0.084 3.382 0.078
SGRcl [RNA]blood 0.171 0.003 0.065 2.811 0.106
SGRcl Ln[DNA]liver 2.610 -0.379 0.289 11.554 0.002
SGRcl Ln[DNA]heart 1.984 -0.269 0.067 2.882 0.102
SGRcl Ln[DNA]blood 3.486 -0.504 0.236 9.052 0.006
SGRcl Ln(R:D)liver 0.550 -0.339 0.337 14.228 0.001
SGRcl Ln(R:D)heart 0.899 0.217 0.164 6.107 0.021
SGRcl Ln(R:D)blood 0.846 0.294 0.222 8.426 0.008
SGRcl [Protein]liver 0.802 -1.584 0.083 3.357 0.079
SGRcl Ln(P:D)liver 0.535 0.203 0.047 2.283 0.143
SGRcl Ln(R:P)liver 0.669 -0.343 0.490 25.954 < 0.0001
SGRcl CI -0.903 9.951 0.262 10.245 0.004
Specific growth rate for body mass or carapace length was regressed independently against each
index (n = 27 for each variable). Significant p-values are indicated in bold. Abbreviations are the
same as in Table 3-2.









Table 3-4. Growth equation parameters for juvenile Chelonia mydas as determined by stepwise multiple linear regression.
Tissue Model # y x1 x2 intercept 01 02 R2 F p-value
Liver 1 SGRbm Ln[RNA] 8.040 -1.132 0.629 45.011 < 0.0001
Liver 2 SGRbm Ln[RNA] CI 4.316 -0.913 17.689 0.680 28.567 < 0.0001
Heart 3 SGRbm CI -3.940 38.475 0.379 16.897 < 0.001
Blood 4 SGRbm CI -3.940 38.475 0.379 16.897 < 0.001
Blood 5 SGRbm CI Ln[DNA] 5.384 31.770 -1.402 0.547 16.705 < 0.0001
Liver 6 SGRol Ln[RNA] 2.549 -0.353 0.659 51.159 < 0.0001
Heart 7 SGRcl CI -0.903 9.951 0.262 10.245 0.004
Blood 8 SGRcl CI -0.903 9.951 0.262 10.245 0.004
Blood 9 SGRcl CI Ln[DNA] 1.730 8.058 -0.396 0.398 9.950 0.001
Specific growth rate for body mass or carapace length (the dependent variables) was regressed against condition index and
biochemical indices for a particular tissue (liver, heart, or blood) (n = 27 for each variable). When specific growth rate was regressed
against condition index and biochemical indices for all tissues together, the resulting models were identical to models 1 and 2 (for
-7j SGRbm) and model 6 (for SGRli). Variables are listed in the order in which they were selected by the models. Significant p-values are
indicated in bold. Abbreviations are the same as in Table 3-2.









Table 3-5. Coefficients of variation (C.V.) for RNA, DNA, and protein concentrations of
Chelonia mydas tissues.
Tissue Assay C.V. (%)
Liver [RNA] 33.2
Heart [RNA] 32.2
Blood [RNA] 20.0
Liver [DNA] 32.4
Heart [DNA] 35.2
Blood [DNA] 38.9
Liver [Protein] 11.0
Values represent averages of C.V.s for 27 individual turtles for each assay. A minimum of three
replicates for each individual for each assay was performed unless sample mass was insufficient.












600


m 400
U)


" 200
0


0


a



31
b

C







a

b
a" C











a b


-3.0
cu

0
M 2.0



1.0
0

0 0.0



1.0

" 0.8
C-
0.6
_j
0 0.4

W 0.2

O0.0


AL R-AL R

Figure 3-1. Morphometric indices and growth rates for turtles in each of three treatment groups.
Turtles in the AL group (n = 7) were fed adlibitum for 12 weeks. Turtles in the R-AL
group (n = 10) were fed 50% of initial mass-specific AL intake for 5 weeks and then
fed ad libitum for 7 weeks. Turtles in the R group (n = 10) were fed 50% of initial
mass-specific AL intake for 12 weeks. Each point represents mean + standard error.
Means were evaluated using analysis of variance with Tukey's Honestly Significant
Difference or Tamhane's T2 post hoc tests or using a Kruskal-Wallis test and
pairwise Mann-Whitney U tests with a Bonferroni correction for multiple
comparisons. Means that are significantly different atp < 0.05 are indicated by
different letters. Abbreviations: SGR = specific growth rate, BM = body mass, CL =
carapace length.


a
a r





b







a a





b
A,--

I I I
AL R-AL R


0.25
E
- 0.20
x
S 0.15

a 0.10

0.05
o
0n nn
L)nn









Figure 3-2. Nucleic acid indices for turtles in each of three treatment groups. Each point
represents mean + standard error. Means that are significantly different atp < 0.05 are
indicated by different letters. Treatments, data analysis, sample sizes, and
abbreviations are the same as in Figure 3-1.










BLOOD


AL R-AL R


b




a
a
a I


1000


750


500 -


250


0


b


a a











a

a a I










a

ab -

b





I I IR
AL R-AL R


200


150


100


50


0


600



400



200



0


0.6



0.4



0.2



0.0


a a

Sa


I I I
AL R-AL R


600



400


z200
3.


0


3.0


2.0



1.0


a
a
I:


ab
a
I


a
S ab

{ b


*


LIVER


HEART














0.4


E 0.3
3)
E
--0.2

o
0.1


0.0


1.5
')
')
E 1.0

z
'E'0.5
0
i=.


AL R-AL R

Figure 3-3. Liver protein and protein-based indices for turtles in each of three treatment groups.
Each point represents mean + standard error. Means that are significantly different at
p < 0.05 are indicated by different letters. Treatments, data analysis, sample sizes, and
abbreviations are the same as in Figure 3-1.


LIVER




ab b
aba










a

I
b b









b




a
a




SI I









CHAPTER 4
COMPENSATORY GROWTH AND ANTIOXIDANT STATUS IN JUVENILE GREEN
TURTLES (Chelonia mydas)

Introduction

Poor early nutrition and resulting periods of depressed growth can have profound

life-history consequences, some of which extend into subsequent generations (as reviewed by

Metcalfe and Monaghan 2001). Slow growth that results in small size at a particular age or

developmental stage can increase vulnerability to predation (Arendt 1997, Janzen et al. 2000),

weaken dominance status (Richner et al. 1989), impede the establishment of feeding and/or

breeding territories (Einum and Fleming 2000 and references therein), alter patterns of sexual

dimorphism, or delay maturation and the onset of reproductive competence (Altmann and

Alberts 2005). The nutritional environment an animal experiences early in its life can also

adversely affect adult body size (Madsen and Shine 2000), survival (McDonald et al. 2005),

fecundity (Nagy and Holmes 2005), and offspring size (Reznick et al. 1996).

Given the negative effects of slow growth and small size on performance, survival, and

ultimately fitness, selection should favor compensatory strategies that allow individuals to grow

rapidly when conditions improve (Metcalfe and Monaghan 2001). Such a period of 'catch up' or

compensatory growth (CG) has been demonstrated in a number of organisms (Wilson and

Osbourn 1960, Ali et al. 2003, Bjorndal et al. 2003, Jespersen and Toft 2003). Compensatory

growth is characterized by growth rates greater than those of consistently well nourished

conspecifics of the same age and can result in comparable body sizes for individuals with

drastically different dietary histories (Metcalfe and Monaghan 2001). The benefits of a rapid

increase in size include improved short-term survival due to decreased size-specific mortality

(e.g., due to predation) (Arendt 1997, Metcalfe and Monaghan 2003) and enhanced reproductive









output, especially for organisms (e.g., ectotherms) in which fecundity is proportional to body

size (Roff 1992).

Despite the potential fitness benefits of accelerated growth, the occurrence of CG suggests

that maximal growth rates are not always optimal under conditions of ample food availability.

Sub-maximal growth rates, even when food availability is high, presumably reflect a balance

between the benefits and costs of rapid growth. Structures formed during periods of fast growth

may be prone to weakness, as is the case for bird primary feathers (Dawson et al. 2000) and leg

bones (Leterrier and Nys 1992) as well as fish scales (Arendt et al. 2001). Animals that have

undergone a period of CG can incur costs including increased muscle protein degradation

(Therkildsen 2005), decreased muscle mass (Belanger et al. 2002), impaired locomotor

performance (Alvarez and Metcalfe 2005), accelerated telomere shortening, and decreased

longevity (Jennings et al. 1999). At the cellular level, animals feeding ad libitum (and therefore

growing rapidly) typically produce more mitochondrial free radicals and thus experience more

oxidative damage than calorie-restricted animals (Gredilla and Barja 2005). It has been

suggested (although not tested, to my knowledge) that the detrimental effects of CG on

performance and longevity may result from transient elevated rates of free radical-induced

cellular damage (Mangel and Munch 2005).

In the present study, I manipulated growth trajectories of green turtle (Chelonia mydas)

juveniles by controlling intake. Before and after a demonstrated period of CG, I evaluated

muscle and liver glutathione peroxidase (GPX) activity and hepatic antioxidant potential (AP).

Glutathione peroxidase is the major cytosolic and mitochondrial enzyme that catalyzes the

reduction of hydroperoxides into water (Li et al. 2000), thereby protecting cells from oxidative

damage to protein, lipids, and DNA (Barja 2004). Total AP reflects activity of antioxidant









enzymes such as GPX in addition to reducing capacity of non-enzymatic antioxidants (Sies

1997). Quantifying GPX activity and AP permits me to assess whether diminished capacity to

combat cellular oxidative damage may be a cost of CG in this species. To my knowledge, this

study is the first to test the effects of CG on these parameters.

Materials and Methods

Animal Care

All animal care components of this study were performed at the Cayman Turtle Farm in

Grand Cayman, British West Indies. Chelonia mydas hatchlings were housed individually in

68-L bins of sea water. Turtles were fed turtle pellets (Melick Aquafeed, Catawissa, PA) twice

daily.

Turtles in the ad libitum group (AL) were fed ad libitum for twelve weeks. Turtles in the

restricted group (R) were fed approximately 50% of the average initial AL intake (on a

mass-specific basis) for twelve weeks. Turtles in the restricted-ad libitum group (R-AL) were fed

the restricted diet for five weeks and were then fed ad libitum for the remaining seven weeks.

Turtles were weighed and measured weekly. Additional details regarding animal husbandry can

be found in Chapter 2.

Tissue Collection and Homogenization

At the conclusion of the fifth week of the experiment (immediately prior to switching

R-AL turtles to an ad libitum diet), ten AL (t5 AL), five R, and five R-AL turtles were sacrificed.

All R-AL turtles had, until the end of week five, been maintained on the restricted diet. Data for

the five R and five R-AL turtles were therefore pooled into one group (t5 R). The remaining

turtles seven AL turtles (t12 AL), ten R turtles (t12 R), and ten R-AL turtles (t12 R-AL) were

sacrificed at the conclusion of the twelve-week trial. Turtles were euthanized with an

intramuscular injection of ketamine (Ketaset, 100 mg/kg body mass).









After each turtle had been injected with ketamine and was no longer responsive to a pain

stimulus, it was decapitated. Portions of the right lobe of the liver and the left pectoral muscle

were removed with forceps and snap-frozen in liquid nitrogen no more than three minutes after

decapitation. Tissues were maintained at -80 C until they were homogenized as described

below.

Subsamples of liver and pectoral muscle were homogenized in 1.0 ml of Sigma T-6789

buffer (0.05 M Tris, 0.138 M NaCl, 2.7 mM KC1, pH 8.0 with 1% bovine serum albumin)

yielding a 10% weight:volume tissue solution for each sample. Liver and muscle homogenates

were further diluted to 10% and 33%, respectively, to insure that enzyme activities would be

within the range of the standard curves used. Total protein concentration of each tissue solution

was evaluated using a Bradford assay with standard curves constructed using dilutions of bovine

serum albumin (BSA, 2 mg/ml undiluted). Concentrations of standards (in duplicate) and tissue

solutions (in triplicate) were determined by absorbance at 595 nm with a 695 nm reference

wavelength using a microplate reader.

Glutathione Peroxidase Activity Assay

Glutathione peroxidase activity was evaluated using a total GPX assay modified from

Nakamura et al. (1974). Diluted muscle and liver homogenate solutions (n = 46) were incubated

for three minutes at 25 C in a reaction cocktail containing 0.297 U/ml glutathione reductase,

1.25 mM glutathione, and 0.188 mM NADPH in a 100 mM potassium phosphate buffer with 10

mM EDTA (pH 7.4). T-butyl hydroperoxide (12 mM) was then added to the reaction mixture

and the absorbance of the resulting solution at 340 nm was recorded every minute for four

minutes using a microplate reader. A blank consisting of 100 mM potassium phosphate buffer

with 10 mM EDTA (pH 7.4) was also assayed to evaluate glutathione-independent reaction rates.









Samples and blanks were analyzed in duplicate (for muscle, because no significant relationships

were found) or triplicate (for liver).

Total GPX activity of each sample was calculated by determining the rate of change in

absorbance of NADPH (AA340/min, calculating using only linear data) and dividing this value by

the extinction coefficient for NADPH (6.22). This quotient was doubled to account for

stoichiometry and then multiplied by the final dilution factor. Activity of blanks was likewise

calculated and subtracted from each sample's activity to yield total GPX activity. Total GPX

activity was then normalized to total protein concentration of each sample as determined by

Bradford assay.

Total AP Assay

Overall, nonspecific AP of homogenized liver samples was evaluated using the Bioxytech

AOP-490 assay (OxisResearch, Portland, OR). This assay evaluates the total activity of cellular

antioxidants including enzymes (e.g., superoxide dismutase, GPX, and catalase), small molecules

(e.g., ascorbic acid), large molecules (e.g., albumin), and hormones (e.g., estrogen)

(OxisResearch Bioxytech Assay Systems label, 2002).

Samples from nine t5 AL, nine t5 R, seven t12 AL, eight t12 R, and eight t12 R-AL turtles and

standards were analyzed in duplicate. A standard curve of uric acid (an antioxidant) was

constructed and used to calculate AP of diluted liver homogenate samples as concentration (pM)

of copper reducing equivalents (CRE). Total AP was then normalized to total protein

concentration of each sample as determined by Bradford assay.

Statistical Analyses

Data for GPX activity and total AP were normalized to total protein content as described

above. In addition, I also calculated these parameters per Ptg of DNA in t12 turtles using nucleic

acid contents from Chapter 3 for liver only (DNA contents of muscle and t12 liver were not









evaluated for that study). Presumably, expressing GPX activity and AP per mg of protein reflects

the proportion of total proteins functioning as antioxidants, whereas expressing these parameters

per [g of DNA reflects antioxidant capacity per cell.

Data were tested for normality (Shapiro-Wilk test) and homogeneity of variances

(Levene's test) prior to parametric analysis. If either test yielded a significant result (p < 0.05),

data were transformed using a natural log, reciprocal, square root, square, reciprocal square, or

reciprocal square root transformation. If transformation did not improve homoscedasticity and

Tamhane's T2 post hoc test could not be used, data were tested for statistical significance using a

Kruskal-Wallis test. Otherwise, data within each sampling period were tested for statistical

significance using analysis of variance (ANOVA). When statistically significant differences

among t12 treatment groups were found, pairwise comparisons were evaluated using Tukey's

Honestly Significant Difference (HSD) post hoc test (if variances were equal) or Tamhane's T2

post hoc test (if variances were not equal).

Data were analyzed using SPSS for Windows (Release 11.0.0). For all reported analyses,

data are expressed as means + standard errors (unless otherwise noted) with alpha set at 0.05.

Results

Turtles in the AL group grew significantly faster than those in the R group during each

week of the study. After the switch from a restricted to an ad libitum diet (at the end of five

weeks), R-AL turtles grew significantly faster than those in the AL group during weeks 7

through 9. This period of growth compensation ceased prior to the end of the study such that AL

and R-AL turtles were growing at comparable rates by the time t12 samples were collected (as

reported in Chapter 2). Different growth rates yielded significantly different body masses for

turtles in each treatment sampled at the conclusion of week five and at the conclusion of week

twelve (Fig. 4-1 and Table 4-1).









Diet treatments affected protein content of muscle (for t5 turtles) and liver (for t12 turtles)

expressed as concentration of protein per wet mass of tissue (Tables 4-1 and 4-2). After five

weeks of food restriction, protein content of muscle was 59% greater in R turtles than in AL

turtles, but this difference decreased after week 5 such that muscle protein content did not differ

among treatment groups by week 12 of the study. Hepatic protein content only differed at t12,

with liver of R turtles containing 36% more protein than liver of R-AL turtles.

Specific activity of GPX in pectoral muscle did not differ significantly among treatment

groups at t5 or at t12 (Tables 4-1 and 4-3). Furthermore, interassay variation was quite high for

measurements of muscle GPX activity (Table 4-4). However, differences in GPX activity of liver

expressed per mg of protein approached significance (p = 0.090) for t12 turtles, with AL turtles

demonstrating 20-24% greater hepatic GPX activity than R and R-AL turtles (Fig. 4-2 and Table

4-1). Because muscle GPX activity was not affected by intake and growth rates, I restricted my

analysis of AP to liver.

Total hepatic AP per mg of protein differed significantly among treatment groups for both

t5 and t12 turtles (Fig. 4-2 and Table 4-1). After five weeks of food restriction, R turtles had a

higher hepatic AP per mg of protein than AL turtles. After twelve weeks of food restriction,

hepatic AP per mg of protein of R-AL turtles was significantly lower than that of R turtles, and

the difference between AL and R-AL turtles approached significance (p = 0.098).

The aforementioned values for GPX activity and AP were calculated by correcting for total

protein content of samples. Using DNA content per mg of liver from Chapter 3 as a correlate of

cell number in t12 turtles, I also compared putative GPX activity and total AP per tg of DNA and

found differences among treatment groups (Fig. 4-2 and Table 4-1). When normalized to tissue

DNA content in this way, both GPX activity and total AP per cell were significantly higher in t12









AL turtles than in R and R-AL turtles, despite the fact that AL and R-AL turtles were growing at

comparable rates at this time.

Discussion

The objective of this study was to evaluate decreased antioxidant capacity as a possible

cost of growth compensation in juvenile green turtles. By manipulating food intake during a

controlled, twelve-week feeding trial, I elicited a CG response from previously food-restricted

turtles after a switch to ad libitum feeding. Because compensating turtles grew faster than

continuously ad libitum turtles, I conclude that growth rates in juveniles of this species are

sub-maximal when individuals have continuous access to unlimited food. This finding suggests

that the benefits of fast growth are countered by one or more costs, potentially including the

accrual of cellular oxidative damage. Although Mangel and Munch (2005) incorporated

oxidative stress into their CG model, the present study is the first to provide empirical evidence

linking growth compensation with effects on antioxidant function.

To assess antioxidant capacity, I measured the activity of GPX in mitotic (liver) and

post-mitotic (skeletal muscle) tissues. Specific GPX activity of liver but not muscle responded to

diet, so I restricted my subsequent analysis of total, non-specific antioxidant potential to liver.

Although the parameters I measured are typically expressed relative to protein concentration, I

found differences in hepatocyte protein content among treatment groups (Chapter 3). As a result,

differences in protein content per liver cell may have confounded my measurements of GPX

activity and AP. In Chapter 3, hepatic DNA content of t12 turtles was measured for the same t12

turtles I examined here, allowing me to correct my measurements for total DNA content as a

putative correlate of cell number. Doing so allowed me to compare the levels of antioxidant

function per cell rather than assessing the proportion of total proteins functioning as antioxidants.









In this study, turtles with a dietary history of continuous ad libitum feeding for twelve

weeks demonstrated higher hepatic GPX activity relative to turtles experiencing a continuous

food restriction. Qualitatively, I observed the same pattern regardless of whether hepatic GPX

activity was calculated relative to protein or DNA content. Glutathione peroxidase serves as one

of several important enzymes in the cellular antioxidant defense system, and upregulation of its

activity likely reflects increased endogenous production of organic hydroperoxides (Judge et al.

2005). My results therefore suggest that high intake and growth rates cause elevated oxidative

stress in mitotically active tissues of juvenile green turtles.

The liver plays an important role in metabolism and detoxification and is a major source of

peroxides via autoxidation reactions. It is therefore not surprising that I detected an upregulation

of a component of the antioxidant defense system in liver of fast-growing AL animals. My GPX

results parallel the finding that food restriction depressed hepatic GPX gene expression in young

mice relative to ad libitum controls (Mura et al. 1996). Additionally, activity of antioxidant

enzymes such as GPX typically increases with age due to elevated oxidative stress

(Leeuwenburgh et al. 1994, Phaneuf and Leeuwenburgh 2002), but this increase is often

attenuated by food restriction (Luhtala et al. 1994).

A counterintuitive result from the present study is the finding that R-AL turtles had GPX

activities comparable to those of continuously food-restricted turtles at the conclusion of the

study. Therefore, turtles in the R-AL group either produced fewer hydroperoxides and

experienced less oxidative stress than continuously AL turtles during the final weeks of the study

or experienced comparable levels of oxidative stress but demonstrated a reduced capacity to

upregulate GPX activity. Because R-AL turtles underwent CG several weeks before the end of

the experiment and had growth rates similar to AL turtles when tissues were sampled, I suggest









the latter scenario occurred. If my hypothesis is correct, then depressed antioxidant enzyme

activity paired with elevated growth rates would exacerbate oxidative damage and therefore be a

cost of CG.

Although my GPX results appear to support my initial predictions, specific components of

the antioxidant defense system often respond differently to oxidative stress depending on the

species and tissues measured. In addition, conflicting effects of dietary restriction on individual

antioxidant enzyme activities are common. For example, GPX activities have been shown to

increase (Agarwal et al. 2005), decrease (Grattagliano et al. 2004), or remain the same (Wu et al.

2003) in food-restricted animals relative to ad libitum-fed controls. Furthermore, individual

enzymes in the antioxidant defense system do not function in isolation from each other. Instead,

combating oxidative stress requires the concerted involvement of a variety of enzymatic and

non-enzymatic molecules that scavenge or neutralize reactive oxygen species (ROS). Many of

these molecules function synergistically (Niki et al. 1995, Bohm et al. 1997). For this reason, I

measured total, non-specific hepatic antioxidant potential in t5 and t12 turtles and again found

differences among treatment groups.

When calculated relative to protein content, hepatic AP of R turtles was significantly

higher than hepatic AP of AL turtles after five weeks of food restriction. Kalani et al. (2006)

found similar results in ad libitum-fed versus calorie-restricted rats. After seven additional

weeks, the difference between AL and R turtles was diminished, but R turtles had higher AP than

R-AL turtles and AP of AL turtles was marginally higher than AP of R-AL turtles. These results

indicate that diet history affects the proportion of total proteins that function as antioxidants

within the liver.









However, when I corrected my values of t12 hepatic AP for DNA content, fast-growing AL

turtles had higher AP per cell than both slow-growing R turtles and fast-growing R-AL turtles.

As a result, turtles that underwent an earlier period of CG had decreased cellular antioxidant

function compared to age-matched AL turtles feeding and growing at the same rate. Because I

did not measure hepatic DNA concentration for t5 turtles, I could not evaluate putative AP

content per cell for these individuals.

The results of this study imply that individuals typically grow at rates that optimize their

ability to prevent oxidative damage to lipids, nucleic acids, and proteins. Oxidative stress results

from an imbalance between the rate of ROS production and the availability of antioxidants to

scavenge these ROS within a cell (Agarwal et al. 2005). Given the assumption that ROS

production increases with intake rate (L6pez-Torres et al. 2002, Barja 2004), my finding that ad

libitum-fed turtles compensating for a prior food restriction also had diminished antioxidant

function implies a cost of CG. It is unclear whether elevated oxidative stress during early

development in this long-lived species would adversely affect longevity or performance.

However, this study provides evidence of cellular stresses coincident with growth compensation

and suggests that sub-maximal growth protects individuals from the detrimental effects of

impaired antioxidant defense.









Table 4-1. Omnibus F, 2, and p-values for comparisons of means among treatment groups at five weeks (t5) and twelve weeks (t2).
t12 Groups Tested in Pairwise Comparisons
Omnibus F and x2 AL and R AL and R-AL R and R-AL
Body Mass
t5 F1,1s = 99.517,p < 0.0001
tF2 F2,24 = 99.476,p < 0.0001 < 0.0001 < 0.0001 < 0.0001
Protein Content, Muscle
t5 F1,1s = 5.952,p = 0.025
tF2 F2,24 = 0.698,p = 0.507 N/A N/A N/A
Protein Content, Liver
t5 Z2 = 2.286,p= 0.131
tF2 F2,24 = 3.545,p = 0.045 0.630 0.320 0.036
GPX Specific Activity, Muscle
t5 F1,18 = 0.507,p = 0.486
t12 F2,24 = 0.226,p = 0.800 N/A N/A N/A
GPX Specific Activity, Liver
St5 F1,18 = 0.811,p = 0.380
tF2 F2,24 = 2.660, p = 0.090 N/A N/A N/A
Total Antioxidant Potential, Liver
t5 F1,F6 = 10.443,p = 0.005
tF2 F2,20 = 5.135,p = 0.016 0.697 0.098 0.015
GPX Activity Per ag DNA, Liver
tF2 F2,24 = 9.617, p < 0.001 0.001 0.003 0.937
Antioxidant Potential Per atg DNA, Liver
t12 F2,20 = 6.416,p = 0.007 0.029 0.008 0.808
When F values are reported, data were analyzed using analysis of variance, and pairwise comparisons of t12 samples were evaluated
using Tukey's Honestly Significant Difference post hoc tests. When 2 values are reported, data were analyzed using a Kruskal-Wallis
test. Statistically significant p-values are indicated in bold. Abbreviations: AL = ad libitum for twelve weeks, R = food-restricted for
twelve weeks, R-AL = food-restricted for five weeks and ad libitum for seven weeks, GPX = glutathione peroxidase.









Table 4-2. Total protein concentrations of Chelonia mydas muscle and liver homogenates as
determined by Bradford assay expressed relative to wet mass of homogenized tissue.
Treatment Week n Protein Concentration, Muscle Protein Concentration, Liver
Group (mg/mg tissue) (mg/mg tissue)
AL 5 10 0.198 + 0.023a 0.178 0.021a
R 5 10 0.314 0.040b 0.241 0.021a
AL 12 7 0.173 0.018x 0.218 0.024xy
R 12 10 0.187 + 0.026x 0.242 0.014x
R-AL 12 10 0.211 0.021x 0.178 0.018y
Values represent means standard errors. Treatment groups and data analysis are the same as in
Table 4-1. Letters (a and b for week 5, x and y for week 12) indicate statistically significant
differences (p < 0.05) among treatment groups within sampling periods.









Table 4-3. Glutathione peroxidase (GPX) specific activity in Chelonia mydas muscle
homogenate.
Treatment Week n GPX Activity, Muscle
Group (nmol/min*mg protein)
AL 5 10 2.78 + 0.16
R 5 10 3.07 0.37
AL 12 7 2.68 + 0.12
R 12 10 2.66 0.11
R-AL 12 10 2.78 0.18
Values represent means standard errors. Treatment groups and data analysis are the same as in
Table 4-1. No significant differences in GPX activity of muscle tissue were detected among
treatment groups in either sampling period.









Table 4-4. Coefficients of variation (CV, %) for protein concentration, glutathione peroxidase
(GPX) activity, and antioxidant potential (AP) assays.
Treatment Week Protein Content GPX Activity AP
Group Liver Muscle Liver Muscle Liver
AL 5 8.0 9.2 6.7 26.9 19.1
R 5 10.8 9.1 6.3 38.2 19.2
AL 12 9.9 7.7 6.3 27.7 12.2
R 12 13.8 10.8 8.0 24.9 12.3
R-AL 12 9.0 7.7 6.4 24.0 24.3
Each value represents the average of individual CVs for each turtle and each assay. Treatment
groups are the same as in Table 4-1. Protein content and hepatic GPX activity assays were
performed in triplicate, whereas muscle GPX activity and hepatic AP assays were performed in
duplicate.













AL
DR
400 AR-AL


300


S200


100


Figure 4-1. Body mass of turtles at five weeks (ts) and twelve weeks (t12), when tissues were
sampled. Each point represents mean + standard error. Sample sizes: n = 10 for all
groups except t12 AL (n = 7). Treatment groups and data analysis are the same as in
Table 4-1. Letters (a, b, and c) indicate statistically significant differences (p < 0.05)
among treatment groups within sampling periods.











































ab

p = 0.098









t12


40

Z

0
O3
-,x


E 30
U)

>




X
0
E


Z20

x
(L


*AL
DR
A R-AL


z
0
M- 30

U)

0
E
- 20
a.


Figure 4-2. Glutathione peroxidase (GPX) specific activity and antioxidant potential (AP;
calculated as nmoles of copper reducing equivalents, CRE) in Chelonia mydas liver
homogenate at five weeks (ts) and twelve weeks (t12). For graphs on the left, GPX
activity and total AP were normalized to total protein concentration as determined by
Bradford assay. For graphs on the right, GPX activity and total AP were normalized
to tissue DNA content as reported in Chapter 3 for t12 turtles. Each point represents
mean + standard error. Treatment groups and data analysis are the same as in Table
4-1. Sample sizes for GPX activities are the same as in Figure 4-1. Sample sizes for
AP are n = 9 for t5 turtles and n = 7, 8, and 8 for t12 AL, R, and R-AL turtles,
respectively. Letters (a and b) indicate statistically significant differences (p < 0.05)
among treatment groups within sampling periods.









CHAPTER 5
TIMING OF DIETARY RESTRICTION ALTERS THE EXPRESSION OF LIFE-HISTORY
TRAITS IN A LONG-LIVED, PARTHENOGENETIC INSECT

Introduction

Life-history theory seeks to explain how natural selection optimizes life cycles to

maximize fitness. Central to this body of theory is the assumption that developmental trajectories

and life histories should demonstrate plasticity in response to environmental variation including

food availability (Roff 1992, Steams 1992). Because of extrinsic or intrinsic upper limits on rates

of resource acquisition (Speakman and Kr6ol 2005), most animals cannot simultaneously

maximize the allocation of nutrients to all traits that influence fitness. As a result, resources are

differentially allocated to various reproductive and somatic functions according to priority rules

(Boggs 1992, Zera and Harshman 2001).

These priority rules predict that survival should be favored over reproduction in times of

resource limitation. One of the most pervasive findings in life-history studies is that food

restriction (FR) leads to increased lifespan in a diversity of organisms including worms, spiders,

insects, rodents, and primates (Weindruch and Walford 1988, Austad 1989, Turturro and Hart

1992, Mair et al. 2003, Hatle et al. 2006b). Coincident with lifespan extension, adult FR usually

decreases or inhibits oogenesis and egg production (Chippindale et al. 1993, Wheeler 1996).

Presumably, the suppression of reproductive activity during times of food scarcity allows

food-restricted individuals to divert available resources into maintenance and storage, thereby

increasing starvation resistance and the probability of survival until conditions more conducive

to reproduction are encountered (Holliday 1989, Masoro and Austad 1996, Simmons and

Bradley 1997).

This negative correlation between longevity and fecundity is interpreted as evidence for a

cost of reproduction (Stearns 1992). A reproductive cost can also be expressed as a trade-off









between current and future reproduction (Calow 1979, Reznick 1985 and 1992). For such a cost

to exist, the decreased survival demonstrated by very fecund individuals feeding at high rates as

adults must be attributable to the process of egg production. However, preventing oogenesis or

vitellogenesis in ad libitum-fed adult Drosophila did not decrease mortality rates (Mair et al.

2004). Additionally, Kaeberlein et al. (2006) found that FR in C. elegans extended lifespan even

when the food restriction was imposed after the cessation of reproductive activity. Therefore, the

enhanced longevity demonstrated by food-restricted adults may not result simply from a

re-allocation of nutrients away from egg production and toward somatic maintenance and

survival.

Although it is clear that the nutritional environment experienced during adulthood can

directly alter reproductive output, juvenile food restriction can also influence fecundity indirectly

through its effects on body size. High food availability during juvenile stages favors rapid

growth and larger sizes at developmental transitions such that adult size is maximized. In insects,

adult body mass is often strongly and directly correlated with fecundity (Hon6k 1993, Tammaru

et al. 1996; although see also Leather 1988), meaning that the conditions conducive to rapid

juvenile growth and large adult size tend to maximize reproductive output. In contrast, when the

nutritional environment during juvenile stages leads to slow growth rates, size at developmental

transitions may shift downward to reduce the demographic costs of extended development time

(Rowe and Ludwig 1991, Berrigan and Koella 1994, Leips and Travis 1994, Bradshaw and

Johnson 1995, Day and Rowe 2002). The response to reduced intake in juvenile stages in insects

is therefore a reduction in adult body size with a potential for concomitant decreased fecundity

(Hon6k 1993).









Whereas juvenile and adult food restriction can have similar effects on lifetime fecundity,

their effects on longevity differ. In insects, the effects of juvenile FR on life-history traits and

trade-offs are less clear than the effects of adult FR, primarily because few authors have assessed

the responses of longevity to juvenile food availability. Those who have done so have reported

that juvenile FR either has no effect (Tu and Tatar 2003) or a negative effect (Boggs and

Freeman 2005) on lifespan. The pattern of resource acquisition early in an individual's lifetime

therefore has the potential to alter later allocation patterns and the expression of life-history traits

and trade-offs. Furthermore, allocation patterns may also change with age (models reviewed by

Perrin and Sibly 1993), depending on the diets experienced in different life stages and the needs

of the individual at the time of allocation. For example, the expression of a trade-off between

current and future reproduction may depend on whether an organism relies on adult-derived

nutrients or stored larval-derived nutrients for egg production (Boggs 1992).

It is clear, then, that resource acquisition patterns can profoundly affect the expression of

life-history traits and trade-offs. Because fluctuations in food availability almost certainly occur

for most animals at some point in their lifetimes (Boggs and Ross 1993, Carey et al. 2002a), a

complete understanding of the effects of intake during different life stages requires data

collection throughout an individual's entire lifespan. However, such experiments are rare,

particularly for long-lived species. In addition, studies of FR and life-history trade-offs in insects

often suffer from one or more fundamental limitations. Firstly, FR in insects typically entails ad

libitum consumption of lower quality food rather than a quantitative reduction in the amount of

food offered (Partridge et al. 2005). Furthermore, intake of these lower quality diets is typically

not quantified, despite the need for such information when assessing trade-offs (Zera and

Harshman 2001). In those studies that do impose a quantitative food restriction in which absolute









intake is limited, the restriction usually occurs only during adulthood and longevity is often not

enhanced (Boggs and Ross 1993, Carey et al. 2002b, Cooper et al. 2004), contradicting the

nearly universal finding of increased lifespan under quantitative FR in many taxa. In addition, to

evaluate the costs of reproduction and the effects of dietary restriction on fitness in sexual

species, females must be allowed to mate. However, co-housing individuals complicates the

quantification of individual intake and can influence longevity due to the effects of crowding

(Joshi et al. 1998). To avoid such problems, reproductive output of virgin female insects is often

studied as a proxy for fitness, yet mating has been shown to enhance egg production in several

species where unmated females would otherwise lay infertile eggs (De Clercq and Degheele

1997, Foster and Howard 1999). Furthermore, the males of many species can alter the

physiology and behavior of females via sex peptides (Wolfner 1997, Gillott 2003, Carvalho et al.

2006) or nutritious nuptial gifts (Voigt et al. 2006). Lastly, sexual species incur a number of

costs associated with reproductive behaviors including courtship, repulsion of unwanted mates,

intrasexual competition, and locomotory costs of carrying mates during copulation (Watson et al.

1998).

To overcome these obstacles, I adopted a novel approach to life-history experimentation by

using a parthenogenetic species as my animal model. Carausius morosus (Br.) (Phasmatodea,

Lonchodinae) is a relatively long-lived species that reproduces via obligate apomictic

parthenogenesis (Pijnacker 1966). Using a parthenogen as my animal model obviated the need

for mating while still permitting natural reproductive processes. This species is hemimetabolous

and phytophagous, allowing for life-long, quantitative dietary manipulations. Additionally, C.

morosus consumes the same food throughout its lifetime, enabling me to test dietary treatments

that spanned both juvenile and adult stages. My purpose was to determine the effects of









differences in resource availability at several developmental stages on life-history traits that have

substantial influences on population structure and dynamics, such as age and size at each

developmental transition, longevity, and fecundity.

Materials and Methods

Animal Husbandry and Feeding Treatments

This study was conducted in a USDA-approved quarantine facility within the Department

of Zoology at the University of Florida (permit # PPQ 69292). Lights were maintained on a

12:12 light:dark cycle. Room temperature averaged 22.5-24.5 C, and relative humidity averaged

45-55% throughout the trial. Twenty adult Indian stick insects (Carausius morosus) were

obtained from the Exploratorium in San Francisco, California. Eggs laid by these females were

individually incubated until hatching. The resulting offspring (n = 86) were systematically

assigned to one of six treatment groups such that all experimental insects produced by a

particular mother were evenly distributed among groups.

These insects were maintained individually for their entire lifetimes in plastic cages

(29.5 cm x 19 cm x 19 cm) with locking vented lids lined with fine-mesh screening. Each cage

was misted daily with deionized water to provide drinking water. Insects were fed discs cut from

leaves of English ivy (Hedera helix) daily. Biopsy punches (Miltex Instrument Co., Inc.) were

used to create discs of multiple diameters: 2 mm for first instar insects, 3 mm for second instar

insects, 4 mm for third instar insects, 5 mm for fourth instar insects, 6 mm for fifth instar insects,

and 8 mm for sixth instar insects and adults. When cutting leaf discs, care was taken to avoid

major leaf veins such that discs contained as little vascular tissue as possible.

Insects were offered either more leaf discs than they could consume within 24 hours (ad

libitum, AL) or a restricted number of discs (R) equal to 60% of the average daily mass-specific

intake of AL-fed insects in the same life-history stage. Life-history stages were categorized as









each of six instars, the adult stage prior to first oviposition, and the adult stage after first

oviposition. Because mass-specific intake of AL insects declined after first oviposition, the

amount of food offered to food-restricted adults after first oviposition was decreased

proportionally to match this decline.

Discs were offered according to five treatment schedules (Fig. 5-1). Insects in the AL group

(n = 15) were offered food ad libitum for the duration of their lifetimes. This group served as the

control group from which intake data were used to determine the appropriate number of discs to

offer to restricted insects. Insects in the R group (n = 28) were offered the restricted amount of

food for the duration of their lifetimes. Individuals in the AL-R groups were initially fed ad

libitum and were switched to the restricted diet at the beginning of the fifth instar (AL-R at 5th, n

= 15) or at first oviposition (AL-R at Ov, n = 14). Insects in the R-AL at 5th group (n = 14) were

initially fed the restricted diet and were switched to an ad libitum diet at the beginning of the

fifth instar. Food-restricted insects generally ate all of the food offered each day, although

food-restricted adults occasionally failed to consume all discs. I initially planned a diet switch

from restricted to ad libitum at first oviposition (R-AL at Ov) but was unable to test this treatment

schedule because survival to oviposition was extremely low for insects maintained throughout

the juvenile stages on the restricted diet. To ensure a sufficient sample size in the R group, all

insects that were food-restricted throughout all six instars and successfully oviposited (n = 7)

were maintained on the restricted diet throughout adulthood. The sample sizes indicated in each

table and figure (except Figs. 5-5 and 5-6) include only those individuals that survived through

the end of the sixth instar.

Physiological and Life-History Response Variables

Daily intake of each insect was estimated by determining the number of discs remaining

each day and subtracting this quantity from the number of discs offered the previous day. Whole









discs were counted, and partial discs were pressed between microscope slides and scanned

(Visioneer OneTouch scanner). Surface area of each disc fragment as a proportion of uneaten

leaf disc surface area was determined using ImageJ (1.37v). A sample of each size of leaf discs

was dried daily to constant mass at 60 C and weighed. The approximate daily dry matter intake

for each insect was then calculated as number of discs consumed estimated dry mass per disc.

Daily mass-specific intake was calculated using estimates of daily body mass computed from

periodic body mass measurements, as described below.

Each insect was weighed weekly, at the end of each instar (defined as the day when no

food was eaten in preparation for ecdysis), at first oviposition, and at death. Insects were also

photographed at these times (Nikon Coolpix 3200), and body lengths at the end of each life-

history stage were then determined using ImageJ. Body length was measured as the distance

between the base of the antennal socket and the end of the tergum on the terminal abdominal

segment. Measurements of body size at the end of each instar for AL insects were then fitted to

the allometric equation ln(y) = ln(a) + bln(x), where y = body mass and x = body length. Relative

body mass (as an index of body condition) of insects in all treatment groups at the adult molt was

assessed as the ratio between measured body mass and body mass predicted by the AL allometric

equation (Perrin et al. 1990). Specific growth rate (SGR) of each insect in each life-history stage

was calculated as:

SGR = 100*(lnBMf lnBMi)/t

where BMf is body mass at the end of a stage, BMi is body mass at the beginning of a stage, and

t is the time in that stage.

During each day of adulthood, all eggs laid by each female were collected and individually

weighed. Average egg mass was calculated as the mean mass of individual eggs for each female.









Reproductive lifespan was calculated as the time between first and last eggs laid. Clutch size was

not quantified because oviposition in C. morosus occurs continuously throughout the

reproductive lifespan rather than in discrete clutches. After death, each female was dissected, and

the number of ovarioles in both ovaries was determined. The numbers of fully chorionated eggs

and non-chorionated eggs in terminal follicles and in the oviducts were counted. Although I

intended to evaluate hatch success, egg viability was inexplicably low in this study, particularly

compared to the expected hatchability of nearly 100% (Brock 2000).

Statistical Analyses

Analysis of variance (ANOVA) was used to test for differences among treatment groups.

Data were first tested for normality (Shapiro-Wilk test) and homogeneity of variances (Levene's

test) and transformed, if necessary, using a natural log, reciprocal, square root, square, reciprocal

square, or reciprocal square root transformation. Pairwise comparisons were evaluated using

Tukey's Honestly Significant Difference post hoc test (if variances were homogeneous) or

Tamhane's T2 post hoc test (if variances were not homogeneous). If transformation did not

normalize data, they were analyzed using a Kruskal-Wallis test, and pairwise comparisons were

evaluated using Mann-Whitney U tests with a set at 0.005 to account for the number of

comparisons tested (ten). Reproductive output was analyzed using ANOVA as described above

and also using analysis of covariance with body mass at first oviposition as the covariate to test

for size-independent differences in allocation to reproduction. Other covariates (including length

at first oviposition, body mass at adult molt, and length at adult molt) could not be used because

of significant interactions between these variables and treatment group.

Stepwise linear regression was used to determine the factors that best explained variance in

reproductive output. For this analysis, the dependent variable was the cumulative fecundity of

each female. The independent variables tested were mass-specific intake and growth rates during









the final instar; duration of the final instar; age, body mass, and relative body mass at the adult

molt; duration of the pre-oviposition adult stage; mass-specific intake and growth rates during

the pre-oviposition adult stage; age and body mass at first oviposition; cumulative intake during

instars 1-4, 5-6, and 1-6; and cumulative intake during the pre-oviposition adult stage and during

the reproductive lifespan. Cumulative intake and body condition were tested as potential

determinants of reproductive output because insects have been shown to require a threshold level

of food consumption or body stores to initiate reproductive processes (Juliano et al. 2004, Hatle

et al. 2006a). Variables had to meet a 0.05 significance level to enter a model. Duration of the

reproductive lifespan was dropped from this analysis because of collinearity with the first

variable selected by the model.

Stepwise linear regression was also used to determine the factors that best explained

variance in initial oviposition rate. For this analysis, the dependent variable was the cumulative

number of eggs laid by each female on day 6 of the reproductive lifespan, and the independent

variables tested were the same as those used above. Specific growth rate during adulthood prior

to first oviposition was dropped from this analysis due to collinearity with other variables already

selected by the models. To test for a longevity cost of reproduction in this species, least squares

linear regression was used to examine the relationship between cumulative fecundity and

lifespan (both total lifespan and adult lifespan). Least squares linear regression was also used to

evaluate the strength of the relationship between fecundity and cumulative intake during the

reproductive lifespan.

Kaplan-Meier survivorship curves were constructed for the entire lifespan (n = 86) and for

adult lifespan (n = 70, including only those individuals that successfully molted to the adult

stage). Pairwise comparisons among treatment groups were evaluated using log-rank tests with a









set at 0.005 to account for the number of comparisons tested (ten). Data were analyzed using

SPSS for Windows (Release 11.0.0), and S-Plus (Version 7.0) was used for graphing smoothing

functions and Kaplan-Meier curves.

Results

The diets I imposed yielded different mass-specific intake trajectories for insects in each

treatment group (Fig. 5-2). Total dry mass of food consumed during each life-history stage

except the first instar differed significantly among treatment groups (Table 5-1). In the second

instar, food-restricted insects consumed more total food than insects feeding ad libitum despite

being significantly smaller (Fig. 5-3 and Table 5-2) and receiving proportionally less food on a

daily basis. This discrepancy is explained by the significantly longer instar duration in

food-restricted groups relative to groups feeding ad libitum (Figs. 5-3 and 5-4a). The pattern of

cumulative intake changed after the second instar and was dependent on diet history. Although

cumulative intake prior to the adult molt differed among treatment groups (F4,65 = 28.872, p <

0.0001), the total amount of food consumed between hatching and first oviposition did not differ

among treatment groups (F4,53 = 1.321, p = 0.274), hinting at a potential intake threshold for

induction of reproductive activity (Juliano et al. 2004).

Size and age at each life-history transition differed significantly among treatment groups

(Fig. 5-3 and Table 5-2). Insects did not differ in body mass among treatment groups at hatching

(F4,65 = 1.00, p = 0.414). At the end of instars 1-4, body mass was greater and molting occurred

at younger ages in initially ad libitum insects (groups AL, AL-R at 5th, and AL-R at Ov) than in

initially food-restricted insects (groups R and R-AL at 5th). From the fifth instar until first

oviposition, all treatment groups except AL and AL-R at Ov differed significantly in body mass

and age at the end of each stage (Table 5-2). At death, body mass of AL-R at Ov insects was not

significantly different from body mass of AL and R-AL at 5th insects, but all other pairwise









comparisons of size were significantly different. Age at death differed significantly among all

groups except R and R-AL at 5th

Relative body mass also differed among treatment groups at the adult molt as determined

by allometric analysis (Table 5-3). Least squares regression of body mass (y) and length (x) for

AL insects at the end of each instar yielded the equation ln(y) = 2.7112*ln(x) 12.018 (F1,76 =

14077.66, p < 0.0001, R2 = 0.995). This equation was used to calculate predicted body masses at

actual body lengths for each insect at the adult molt (Table 5-3). The ratio of actual to predicted

body mass was lower for insects feeding at a restricted rate during the final two instars (groups R

and AL-R at 5th), indicating that insects in these groups had proportionally lower body masses for

a given body length than insects in the other three groups.

The duration of each life-history stage differed among treatment groups (Fig. 5-4a).

Food-restricted insects generally progressed more slowly through each stage than insects feeding

ad libitum. Previous diet history affected the duration of the fifth and sixth instars and the

pre-oviposition adult stage for insects in groups AL-R at 5th and R-AL at 5th, as individuals in

these groups progressed through these stages more rapidly than continuously food-restricted

individuals (group R) but more slowly than continuously ad libitum individuals (group AL).

Insects experiencing food restriction during adulthood prior to first oviposition laid their first

eggs later in the adult stage than insects feeding ad libitum during this time. However, duration

of adulthood after first oviposition was significantly shorter for insects that were food-restricted

than for insects that were feeding ad libitum during this time, regardless of when the food

restriction was imposed.

Food-restricted insects also grew more slowly than insects feeding ad libitum (Fig. 5-4b),

although diet history affected the magnitude of this difference. After a switch from restricted to









ad libitum feeding at the beginning of the fifth instar (group R-AL at 5th), specific growth rates

through the final two instars were comparable to those of continuously ad libitum insects (group

AL). I therefore found no evidence of growth compensation in R-AL at 5th insects. However,

insects that experienced a switch from ad libitum to restricted feeding grew significantly faster in

both the fifth and sixth instars than insects that were continuously food restricted. All insects

gained body mass between the adult molt and first oviposition, with growth of R-AL at 5th insects

slower than that of AL and AL-R at Ov insects but greater than that of R and AL-R at 5th insects.

All insects lost body mass between first oviposition and death. AL-R at Ov insects lost

proportionally more body mass than AL insects during this time, but all other pairwise

comparisons of adult growth rates after first oviposition were not significant.

An event history diagram depicting the lifespan of each insect in the study (Carey et al.

1998) demonstrates the variation in life histories and survivorship induced by diet treatments

(Fig. 5-5). Pairwise log-rank tests of survival indicated that all groups except R and R-AL at 5th

differed significantly in total lifespan (Fig. 5-6a). This result parallels the ANOVA results for

age at death (Table 5-2). Pairwise log-rank tests of adult survivorship (Fig. 5-6b) indicated that

longevity was greater for AL and R-AL at 5th insects than for all insects feeding at a restricted

rate during adulthood, suggesting that food restriction experienced during reproductive activity

negatively affected adult lifespan regardless of dietary history.

Treatment groups differed significantly in realized fecundity (F4,53 = 50.31, p < 0.0001,

Fig. 5-7). These differences appear to result both from differences in reproductive lifespan (F4,53

= 41.70, p < 0.0001) and from differences in reproductive rate. The high initial slopes in Figure

5-7 for groups AL, AL-R at Ov, and R-AL at 5th corresponded to higher early oviposition rates

(calculated as eggs laid per day in the first six days of the reproductive lifespan) in these groups









compared to groups R and AL-R at 5th (F4,53 = 15.576,p < 0.0001). These differences suggest that

egg output is enhanced early in the reproductive lifespan by ad libitum feeding during adulthood

prior to first oviposition. The low reproductive output of R insects was compounded by low

survival to first oviposition, such that egg production was severely diminished by lifelong FR.

Differences in egg production did not simply result from differences in body size, as analysis of

covariance revealed significant differences in adjusted mean fecundity when body mass at first

oviposition was used as a covariate (Table 5-4). Adjusted mean fecundity also differed among

groups when relative body mass at the adult molt was used as a covariate (F4,53 = 42.872, p <

0.0001; data not shown). Egg production was also altered by diet history through effects on

average egg mass (F4,53 = 8.195,p < 0.0001, Fig. 5-8), with insects experiencing a diet switch

from ad libitum to restricted feeding producing significantly smaller eggs than continuously ad

libitum insects. Differences in reproductive output do not appear to result from differences in

ovarian morphology among groups (Table 5-5). However, although Tukey's HSD post hoc test

did not reveal significant differences in ovariole number among treatment groups, a less

conservative post hoc test (the Least Significant Difference test) indicated that initially restricted

insects (groups R and R-AL at 5th) had significantly fewer ovarioles than initially ad libitum

insects (groups AL, AL-R at 5th, and AL-R at Ov) (p < 0.05 for all significant comparisons).

Diet history did affect the number of eggs remaining in the ovaries at death (unfulfilled

reproductive potential, F4,53 = 5.286, p = 0.001), with R insects having more eggs remaining in

the ovaries at death than AL-R at 5th insects. All other pairwise comparisons of unfulfilled

reproductive potential were not significant. Groups also differed in potential fecundity (F4,53

71.62, p < 0.0001, calculated as unfulfilled reproductive potential plus realized fecundity) and

total reproductive investment (F4,53 = 49.57, p < 0.0001, calculated as the summed mass of all









eggs laid by each female). The patterns for potential fecundity and total reproductive investment

(data not shown) were identical to that demonstrated by realized fecundity.

I used stepwise multiple linear regression to identify the most significant determinants of

realized fecundity (Table 5-6). Cumulative intake during the reproductive lifespan was the

primary variable selected by the model, explaining 82.8% of the variance in fecundity (Fig.

5-9a). When potential fecundity (number of eggs laid + number of eggs remaining in the ovaries

at death) was regressed against cumulative intake during the reproductive lifespan, the same

relationship existed with an adjusted R2 value of 0.847 (data not shown). In addition, growth rate

during adulthood prior to first oviposition and cumulative intake during all juvenile stages were

also selected as variables in a model that explained 92.8% of the variance in realized fecundity

(model 3, F3,54 = 245.46, p < 0.0001). Stepwise multiple linear regression identified body mass at

first oviposition, age at the adult molt, mass-specific intake during adulthood prior to first

oviposition, and cumulative intake during adulthood prior to first oviposition as significant

independent variables in a model that explained 70.6% of the variance in initial oviposition rate

(model 7, F4,53 = 35.20, p < 0.0001).

The data do not support the contention that decreased longevity is a cost of reproduction, at

least in C. morosus. On the contrary, fecundity was significantly and positively related to adult

lifespan when data for all treatments were combined (F1,56 = 25.67, p < 0.0001, R = 0.302, Fig.

5-9b). When potential fecundity (number of eggs laid + number of eggs remaining in the ovaries

at death) was regressed against adult lifespan, the same relationship existed with an adjusted R2

value of 0.343 (data not shown). However, analysis of covariance indicated that adult lifespan

did not have a significant effect on reproductive output (F1,52 = 3.284, p = 0.076) whereas

treatment did have a significant effect (F4,52 = 32.463, p < 0.0001). This result was confirmed by









individual regressions of fecundity versus adult lifespan for each treatment group, none of which

was significant (p > 0.2 in all cases). Unlike adult lifespan, total lifespan was not significantly

related to realized or potential fecundity (p > 0.5). I also found no evidence for a trade-off

between number and size of eggs laid. There was neither a significant interaction between

average egg mass and fecundity (p > 0.2) nor an effect of fecundity on average egg mass (p >

0.3). Individual regressions of average egg mass versus fecundity indicated that there were no

significant relationships (p > 0.3) except in the case of R-AL@5th insects, for which there was a

positive relationship between egg size and number (F1,1o = 5.161, p = 0.046, R2 = 0.274).

Discussion

Developmental plasticity in response to food availability is nearly universal (Juliano et al.

2004 and references therein), and this study was no exception. In C. morosus, both size and age

at each life-history transition depended on diet history. As is common in studies of this kind

(e.g., Gebhardt and Steams 1988 and 1993), insects that experienced FR prior to the onset of

reproductive activity progressed through juvenile stages more slowly and were smaller at each

molt than individuals feeding at a consistently high rate. Decreasing size and increasing age at

developmental transitions represent a compromise between the need to maximize body size

(because of its potential effects on fitness) and the need to minimize the demographic costs of

extended development time (Rowe and Ludwig 1991).

This plasticity in development rate corresponded to substantial differences in survival

trajectories among treatment groups. One of my most salient results was the finding that

longevity enhancement is not a ubiquitous outcome of dietary restriction. Although individuals

that experienced early-onset FR (R and R-AL at 5th) survived longer than initially ad libitum-fed

individuals, this increased longevity was due solely to extended development time rather than to

extended adult lifespan. Conversely, FR during adulthood decreased the duration of the adult









stage, such that AL-R at Ov insects had shortened lifespans compared to AL insects. A diet switch

from ad libitum to restricted feeding during development extended the duration of the fifth and

sixth instars relative to continuously ad libitum insects, but this difference was not sufficient to

mitigate the negative effect of FR on adult lifespan.

As a result of decreased growth rates, insects that experienced FR at any point during

development were smaller at the adult molt than ad libitum insects. Although subsequent

reproductive output of food-restricted insects was significantly diminished, mean fecundities

differed significantly among treatment groups even when corrected for body mass at first

oviposition. Plasticity in adult size alone therefore does not explain the drastic differences in

fecundity I observed among treatment groups.

Reproductive output may have been mildly constrained by ovarian morphology. Although

I detected no significant differences in total ovariole number among treatment groups when a

conservative post hoc test was used, I did find significant differences in ovariole number

between initially restricted and initially ad libitum insects when a more liberal post hoc test was

used. This result suggests that ovarian development in C. morosus is somewhat plastic in

response to diet. In Drosophila, ovariole number responds strongly to larval diet (Tu and Tatar

2003) and is correlated with fecundity (David 1970). Although ovariole number in C. morosus

appears to be much less plastic than in D. melanogaster, it is possible that decreased fecundity in

food-restricted insects in this study is partially explained by differences in ovary size but only for

insects that were food-restricted during early development.

The primary determinant of fecundity in this study was adult intake, with approximately

83% of the variance in reproductive output explained by the total amount of food consumed

during the reproductive lifespan. Because of this strong, positive correlation between fecundity









and cumulative intake during the reproductive lifespan, I conclude that Indian stick insects use

an "income" breeding strategy (sensu Steams 1992, Jonsson 1997), in which the resources

allocated to reproduction are acquired primarily during the reproductive period. Cumulative

intake between the adult molt and the end of the reproductive lifespan was less strongly

correlated with reproductive output, suggesting that the food acquired prior to first oviposition

was allocated to some degree of pre-reproductive somatic growth rather than being allocated

exclusively to egg production. The putative level of body stores accumulated by the time of the

adult molt does not appear to dictate reproductive success, as demonstrated by significant

differences among groups in realized fecundity when corrected for relative body mass.

An income breeding strategy is appropriate for an organism like C. morosus, in which

oogenesis and vitellogenesis are non-cyclic and continuous (Bradley et al. 1995) throughout a

comparatively long reproductive lifespan. Additionally, species that rely heavily on incoming

resources for reproduction should have ovaries containing primarily immature oocytes

immediately after the adult molt (Jervis et al. 2005), as is the case for C. morosus (Bradley et al.

1995). Although ovaries are present in juvenile stages, mature eggs are not present at the adult

molt. Given this breeding strategy, it is not surprising that both mass-specific intake and age-

specific fecundity in this study declined after first oviposition for adults feeding ad libitum. This

decrease in consumption and production with time typifies insects that are income breeders

(Kindlmann et al. 2001, Dixon and Agarwala 2002).

One might expect food-restricted insects that employ an income breeding tactic to extend

the duration of reproductive activity and thereby to mitigate (at least partially) the effects of

decreased daily intake on oviposition rate. Given that food-restricted flies respond in this way to

FR (Carey et al. 2002a), I expected to see a similar pattern in this study. Surprisingly, females









could not simply compensate for adult FR by increasing the length of the reproductive lifespan.

The very low reproductive output of R and AL-R at 5th females therefore resulted from the

combined effects of decreased daily intake and shortened reproductive lifespan.

The proximate cause of shortened reproductive lifespan may relate to the level of body

stores accumulated prior to adulthood. Insects feeding at a restricted rate late in development

were lighter for their length than insects feeding ad libitum immediately prior to the adult molt.

Although fat body mass and storage proteins were not quantified, these results imply that these

individuals may have accumulated proportionally fewer body stores by the beginning of

adulthood compared to insects feeding ad libitum during the final two instars. Therefore, food

restriction late in development appears to have shifted allocation away from the accumulation of

mobilizable reserves. I suggest that these reserves serve as the source of nutrients that are

allocated to somatic maintenance after the onset of reproductive activity. Because R and AL-R at

5th insects were smaller in both absolute and relative body mass at the adult molt, they probably

depleted their limited stores more rapidly after the onset of reproductive activity than insects that

were feeding ad libitum as young adults. Conversely, R-AL at 5th insects were feeding ad libitum

as pre-oviposition adults, although they were doing so at lower mass-specific rates than AL

insects. Although R-AL at 5th insects had relative body masses similar to those of AL and AL-R at

Ov insects at the adult molt, they were nearly twice as old as AL and AL-R at Ov insects at first

oviposition. It is therefore possible that reproductive lifespans of R-AL at 5th insects were shorter

than those of continuously ad libitum-fed adults simply because of age-specific declines in

physiological function. It does not appear that shortened reproductive lifespans resulted from

exhaustion of available oocytes, as almost all individuals in the study had chorionated eggs

remaining in the ovaries at death.









In addition to sustaining the soma during adulthood, body stores present during early

adulthood may also serve as a signal that coordinates rates of vitellogenesis and oviposition at

the onset of reproductive activity (Moehrlin and Juliano 1998, Hatle et al. 2004, Juliano et al.

2004). In this way, body composition may function as an index of food availability that entrains

subsequent reproductive function (Rowe et al. 1994). If this hypothesis is correct, AL, AL-R at

Ov, and R-AL at 5th insects were committed to a high early oviposition rate in the first several

days of the reproductive lifespan because these individuals had accumulated proportionally more

body mass (with potentially higher levels of body stores) prior to first oviposition than R and AL-

R at 5th insects. However, the mismatch between intake and pre-determined oviposition rate in

AL-R at Ov insects after the switch to a restricted diet may have forced these individuals to

supplement incoming resources by withdrawing nutrients from mobilizable stores that would

otherwise have been allocated to somatic functions such as maintenance and survival. Data for

all AL-R at Ov individuals were located above the regression line correlating fecundity with

cumulative intake during the reproductive lifespan (Fig. 5-9a), suggesting that these individuals

produced more eggs than would have been predicted by the amount of the food they consumed

as reproductively active adults. If body stores were in fact used to supplement incoming nutrients

for oogenesis, the exhaustion of these stores would then explain the shortened adult lifespan of

AL-R at Ov insects relative to AL individuals.

The hypothesis that body stores determine adult lifespan and establish initial oviposition

rate is supported by my data for insects that experienced a switch to the restricted diet late in

development. Despite feeding ad libitum in the first four instars, AL-R at 5th insects had similar

reproductive lifespans and fecundities as continuously food-restricted insects. Because growth is

typically exponential in juvenile insects, body composition at the adult molt is largely









determined by food availability during the final instar(s) (Scriber and Slansky 1981). As a result,

ad libitum feeding early in life does not appear to provide a substantial fecundity benefit for

individuals that subsequently experience a decline in juvenile food availability. However, the

marginal dependence of ovariole number on food availability during the first few instars does

indicate that early nutritional conditions could potentially affect subsequent reproductive

function.

Conversely, ad libitum feeding later in development and during adulthood provides

nutrients necessary for somatic maintenance. However, mass-specific intake declined after first

oviposition for both groups feeding ad libitum as reproductively active adults, suggesting either

that oogenesis and vitellogenesis are less costly than somatic growth during juvenile stages or

that digestive and reproductive functions decline with time due to senescence (Kindlmann et al.

2001, Carey et al. 2002a, Dixon and Agarwala 2002). Evidence for the latter hypothesis was

demonstrated by the lower mass-specific intake of R-AL at 5th insects, which matured at older

ages, compared to AL insects after first oviposition. I therefore conclude that adult survival and

reproductive decisions (such as age at first oviposition and initial oviposition rate) are based on

the extent of reserves accumulated prior to maturity, whereas fecundity depends on food

consumed during the reproductive lifespan in C. morosus. Perhaps not surprisingly, "capital"

breeders (sensu Stearns 1992) tend to demonstrate the opposite strategy, allocating stored

reserves to egg production and using adult-acquired nutrients to increase survival (e.g., Tammaru

et al. 1996).

The apparent dependence of initial oviposition rate on accumulated body stores may result

from differences in hormone signaling induced by diet. In insects, the fat body serves as the main

depot for stored lipids and is responsible for synthesizing yolk proteins (e.g., vitellogenin) and









lipids for incorporation into developing follicles (Chapman 1998). Fat body mass and

hemolymph vitellogenin titers decrease in response to adult FR in grasshoppers (Hatle et al.

2006a), suggesting that my early fecundity results may be mediated by differences in the size of

this storage organ. Additionally, the synthesis of yolk compounds by the adult fat body is

controlled by hormones including juvenile hormone (JH) and ecdysone (Klowden 2002), both of

which have been shown to respond to feeding rates (Hatle et al. 2003, Tu and Tatar 2003).

Specifically, JH stimulates vitellogenesis in most adult insects (Chapman 1998) but not in C.

morosus (Bradley et al. 1995), so diet-induced differences in JH synthesis were probably not

responsible for the decreased fecundity I observed in this study. However, JH does facilitate the

uptake of vitellin by developing follicles in C. morosus (Bradley et al. 1995), suggesting that my

results for egg size may reflect differences in JH signaling.

Unlike lipids, proteins are thought to be stored primarily in hemolymph (Chapman 1998).

These hemolymph storage proteins are critical to egg production (Wheeler et al. 2000) and are

responsive to diet (Hatle et al. 2004). Because egg production is a protein-limited process for

phytophagous insects like C. morosus (Nijhout 1994, Chapman 1998), the quantity of

hemolymph storage proteins present during adulthood may therefore serve as a nutrient sensor

regulating reproductive output.

Contrary to my expectations, fecundity was significantly and positively correlated with

adult lifespan and not correlated with total lifespan. My results therefore indicate that insects

feeding ad libitum as adults did not incur mortality costs simply because they reproduced more

than food-restricted insects. Consequently, decreased longevity is not a cost of reproduction in

this species. This result contradicts the assumption that FR elicits a shift in allocation and

therefore a trade-off between reproduction and survival (Stearns 1992). Furthermore, I detected









no evidence of a trade-off between current and future reproduction or between early fecundity

and adult lifespan (Reznick 1985 and 1992). In fact, insects feeding ad libitum as pre-oviposition

adults (groups AL, AL-R at Ov, and R-AL at 5th) exhibited correspondingly high initial

oviposition rates and also higher cumulative fecundities than R and AL-R at 5th insects.

The longevity costs of reproduction may differ among species depending on the relative

timing of resource acquisition and allocation to reproduction (Boggs 1992). The occurrence of a

trade-off between longevity and fecundity requires that the resources allocated to reproduction or

somatic maintenance are derived from a common resource pool and that the utilization of

resources from this pool for egg production necessarily decreases the availability of resources for

subsequent egg production or survival (van Noordwijk and de Jong 1986, Zera and Harshman

2001). The lack of a negative correlation between longevity and fecundity in C. morosus

therefore implies that the processes of survival and reproduction may not compete for resources

from a common pool. Instead, I suggest that females of this species allocate existing stores

primarily to maintenance and divert incoming resources to egg production.

Although the diet treatments I imposed elicited clear and significant differences in egg

production rates, these data alone do not establish that diet directly affects fitness. There is

evidence that maternal nutritional environment can substantially alter offspring phenotypes

(Wayne et al. 2006) and survival (Prasad et al. 2003) such that fitness depends on more than

simply total egg output of a female. In addition, host plant quality has been shown to alter

fertility of phytophagous insects with no effect on fecundity (Moreau et al. 2006). Furthermore,

fertility can decline as age-specific intake declines over time in income breeders (Dixon and

Agarwala 2002), thereby further uncoupling fecundity and fitness.









Consequently, my fecundity data in isolation are not sufficient to draw conclusions about

the effects of diet history on fitness. Unfortunately, my ability to assess these effects was

compromised by egg inviability. Compared to the expected hatch success of nearly 100% for

captive C. morosus (Brock 2000), egg hatchability in this study was disappointingly low. I

suggest two possibilities for this occurrence. The English ivy I fed to experimental animals may

have been deficient in one or more limiting nutrients, thereby largely preventing embryogenesis.

Alternatively, my incubation protocol may not have been appropriate for this species. However,

my reproductive output results do indicate that food restriction imposed late in development and

during reproductive activity has profound negative implications for fitness. Furthermore, high

mortality prior to the onset of reproductive activity compounded the negative effects of lifelong

FR on egg production and indicates that reproductive output is severely reduced by continuous,

quantitative FR.

In summary, food availability strongly influences the expression of life-history traits and

trade-offs in C. morosus. To my knowledge, this study was the first to evaluate the effects of

quantitative dietary manipulations throughout life in a long-lived, hemimetabolous insect. My

methodology allowed for accurate measurement of daily food consumption rates, a critical but

often neglected component in studies of life history (Zera and Harshman 2001). In addition,

using a parthenogenetic species as my animal model was a novel approach that obviated the need

for mating while still permitting oviposition of fertile eggs. My data demonstrated that the life-

history responses to differences in intake depended to a large extent on the timing of nutritional

stress. Food restriction experienced at any point during life led to decreased fecundity. This

decrease resulted primarily from differences in the quantity of reserves accumulated prior to the

onset of reproductive activity and resulting differences in reproductive rate and adult survival. In









contrast, the effect of food restriction on overall lifespan depended on when the restriction was

first imposed. As such, lifespan was maximized when food consumption was limited early in

life, whereas reproductive output was maximized when food consumption throughout life was

maximized. In effect, food restriction extended development but shortened adult lifespan, with

negative consequences for final body size, reproductive lifespan, reproductive output, and, quite

possibly, fitness. In C. morosus, it appears that storage reserves acquired early in life are

essential for determining adult survival and for entraining the timing and rate of reproductive

processes, but adult income is essential for egg provisioning. Putative fitness is therefore dictated

both by past and current nutritional conditions.










Table 5-1. Cumulative intake expressed as the total dry mass consumed during each life-history stage in each of five treatment
groups.
Cumulative Intake (g dry matter)
Treatment First Instar Second Instar Third Instar Fourth Instar Fifth Instar Sixth Instar Pre-Ov Adult Post-Ov Adult

AL 0.0082 0.0002a 0.0197 0.0005a 0.0411 0.0009a 0.0795 0.0016a 0.1558 0.0025a 0.3473 0.0054'b 0.5757 0.3280a 2.6454 0.1859a
AL-R at5th 0.0080 0.0002a 0.0188 0.0004a 0.0403 0.0008a 0.0802 0.0013a 0.1438 0.0018b 0.3912 0.0098c 0.6500 0.0293'" 0.4289 0.0478b
AL-R at Ov 0.0079 0.0002a 0.0196 0.0004' 0.0410 0.0007' 0.0772 0.0012' 0.1550 0.0016' 0.3514 0.0053b 0.5747 0.0135a 0.9439 0.0328c
R 0.0088 0.0003 0.0226 0.0004b 0.0358 0.0010b 0.0636 0.0009b 0.1289 0.0031c 0.3162 0.0078ad 0.7312 0.0614b 0.2287 0.0242d
R-AL at 5th 0.0084 0.0004a 0.0223 0.0005b 0.0352 0.0011b 0.0606 0.0021b 0.1061 0.0035d 0.2792 0.0122d 0.7529 0.0491b 1.5177 0.1963c

Life-history stages were categorized as each of six instars, the adult stage prior to first oviposition (pre-ov adult), and the adult stage
between first oviposition and death (post-ov adult). Values represent means + standard errors. Abbreviations: AL = ad libitum, R =
restricted. See Figure 5-1 for a description of diet treatments. Sample sizes are the same as in Figure 5-2. Values with different
superscripts are significantly different among treatment groups within life-history stages. See text for statistical analyses.












Table 5-2. Omnibus F, x2, and p-values for comparisons of body mass and age among five treatment groups within each life-history
stage.
Identity of Groups Tested in Pairwise Comparisons
Omnibus F and X2 1&2 1&3 1&4 1&5 2&3 2&4 2&5 3&4 3&5 4&5


Body Mass
Hatch
End of 1st Instar
End of 2nd Instar
End of 3rd Instar
End of 4th Instar
End of 5th Instar
End of 6th Instar
First Oviposition
Death

Age
End of 1st Instar
Is3 End of 2nd Instar
0
End of 3rd Instar
End of 4th Instar
End of 5th Instar
End of 6th Instar
First Oviposition
Death


F4,65- 1.00,p 0.414 0.780 0.780 0.999 1.000 1.000 0.560 0.804 0.560 0.804 0.999
F4,65 21.67, p < 0.0001 1.000 0.999 < 0.001 < 0.001 1.000 < 0.001 < 0.001 < 0.0001 < 0.0001 0.998
F4,65 108.34, p < 0.0001 0.827 0.999 < 0.0001 < 0.0001 0.926 < 0.0001 < 0.0001 < 0.0001 < 0.0001 1.000
F4,65 115.37, p < 0.0001 0.998 1.000 < 0.0001 < 0.0001 0.997 < 0.0001 < 0.0001 < 0.0001 < 0.0001 1.000
F4,65 269.35, p < 0.0001 0.996 0.970 < 0.0001 < 0.0001 0.854 < 0.0001 < 0.0001 < 0.0001 < 0.0001 0.999
F4,65 260.84, p < 0.0001 < 0.001 0.998 < 0.0001 < 0.0001 < 0.001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 0.002
/ 62.92, p < 0.0001 < 0.0001 0.801 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001
F4,53 = 59.01, p < 0.0001 < 0.0001 0.993 < 0.0001 0.010 < 0.0001 < 0.001 < 0.0001 < 0.0001 0.031 < 0.0001
F4,53 33.26, p < 0.0001 < 0.0001 0.366 < 0.0001 0.003 < 0.0001 0.001 0.019 < 0.0001 0.243 < 0.0001



/ 51.66,p < 0.0001 0.880 0.840 < 0.001 < 0.001 0.960 < 0.0001 < 0.0001 < 0.0001 < 0.0001 0.795
F4,65 686.21,p < 0.0001 1.000 1.000 < 0.0001 < 0.0001 1.000 < 0.0001 < 0.0001 < 0.0001 < 0.0001 0.975
F4,65 -626.17, p < 0.0001 0.980 1.000 < 0.0001 < 0.0001 0.991 < 0.0001 < 0.0001 < 0.0001 < 0.0001 0.989
F4,65 -518.30, p < 0.0001 1.000 0.994 < 0.0001 < 0.0001 0.995 < 0.0001 < 0.0001 < 0.0001 < 0.0001 0.997
F4,65 -591.77, p < 0.0001 < 0.0001 0.942 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.001
F4,65 888.90, p < 0.0001 < 0.0001 0.952 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001
/- 50.46, p< 0.0001 < 0.0001 0.840 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001
F4,53= 95.65, p< 0.0001 0.031 < 0.0001 < 0.0001 < 0.001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 0.360


Identity of groups tested in pairwise comparisons: 1 = AL, 2 = AL-R at 5th, 3 = AL-R at Ov, 4 = R, and 5 = R-AL at 5th (where AL = ad
libitum and R = restricted). See Figure 5-1 for a description of diet treatments. Sample sizes are the same as in Figure 5-2. When F
values are reported, parametric tests were used. When /2 values are reported, nonparametric tests were used. See text for statistical
analyses. Statistically significant p-values are indicated in bold.









Table 5-3. Relative mass ( standard error) at the adult molt of insects in five treatment groups,
calculated as the ratio of actual to predicted body mass as determined by allometric
analysis (see text for statistical analyses).
Treatment Actual Mass/Predicted Mass
AL 1.087 0.016a
AL-R at 5th 0.977 0.016b
AL-R at Ov 1.093 + 0.020a
R 0.995 0.014b
R-AL at 5th 1.148 + 0.013a
Abbreviations: AL = ad libitum, R = restricted. See Figure 5-1 for a description of diet
treatments. Sample sizes are the same as in Figure 5-2 for juveniles. Values with different
superscripts are significantly different among treatment groups.









Table 5-4. Adjusted mean fecundity ( standard error) of insects in each of five treatment
groups estimated using body mass at first oviposition as a covariate.
Treatment Adjusted Mean Fecundity
AL 68.061 4.386a
AL-R at 5th 22.204 4.524b
AL-R at Ov 36.649 4.190bc
R 26.929 6.815bc
R-AL at 5th 42.719 + 3.579c
Abbreviations: AL = ad libitum, R = restricted. See Figure 5-1 for a description of diet
treatments. Sample sizes are the same as in Figure 5-2 for adults. Values with different
superscripts are significantly different among treatment groups according to analysis of
covariance with a Bonferroni correction for multiple comparisons.









Table 5-5. Total number of ovarioles (mean standard error) in insects from each of five
treatment groups upon post-mortem dissection.
Treatment Number of Ovarioles
AL 53.38 0.488
AL-R at 5th 53.00 + 0.453
AL-R at Ov 53.23 + 0.556
R 50.86 1.262
R-AL at 5th 51.33 + 0.449
Abbreviations: AL = ad libitum, R = restricted. See Figure 5-1 for a description of diet
treatments. Sample sizes are the same as in Figure 5-2 for adults. Although the omnibus F-value
was significant (F4,53 = 3.387, p = 0.015), Tukey's Honestly Significant Difference post hoc test
found no significant differences among treatment groups. Results of a less conservative post hoc
test (the Least Significant Difference test) indicated that initially restricted insects (groups R and
R-AL at 5th) had significantly fewer ovarioles than initially ad libitum insects (groups AL, AL-R
at 5th, and AL-R at Ov) (p < 0.05 for all significant comparisons).










Table 5-6. Parameters for equations predicting realized fecundity and initial oviposition rate as determined by stepwise multiple linear
regression.
Model y Xi x2 X3 X4 intercept pi p2 p3 B4 R2

1 Fecundity Cum. Int. RL 9.038 28.053 0.828
2 Fecundity Cum. Int. RL Pre-Ov. SGR -4.363 19.085 11.278 0.904
3 Fecundity Cum. Int. RL Pre-Ov. SGR Cum. Int. Juv. -40.833 18.140 11.973 57.654 0.928

4 In. Ov. Rate BM at Ov. -7.231 22.529 0.546
5 In. Ov. Rate BM at Ov. Age at Ad. Molt -16.711 29.005 0.034 0.620
6 In. Ov. Rate BM at Ov. Age at Ad. Molt MS Int. Pre-Ov. -20.081 23.482 0.038 186.022 0.659
7 In. Ov. Rate BM at Ov. Age at Ad. Molt MS Int. Pre-Ov. Cum. Int. Pre-Ov. -23.776 14.926 0.016 322.323 12.407 0.706

See methods for a list of independent variables tested. Significant independent variables are listed in the order in which they were
selected. For all models, n = 58. All models are significant atp < 0.0001. Abbreviations: In. Ov. Rate = initial oviposition rate (total
number of eggs laid during the first 6 days of the reproductive lifespan), Cum. Int. RL = cumulative intake during the reproductive
lifespan, BM at Ov. = body mass at first oviposition, Pre-Ov. SGR = specific growth rate (per day) during adult stage prior to first
S oviposition, Age at Ad. Molt = age at the adult molt, Cum. Int. Juv. = cumulative intake during all juvenile stages, MS int. Pre-Ov. =
average mass-specific intake during adult stage prior to first oviposition, Cum. Int. Pre-Ov. = cumulative intake during adult stage
prior to first oviposition.









Juvenile Adult


SAL
I I I I I I I I AL

............AL-R at 5th

I I I I I I I AL-RatOv



liiiiii iliiiiii iiiiiiiii iiiiiiil I I I R -A L at 5th

iiiiiiiiiiii I R-AL at Ov

I +
Hatching Death

Figure 5-1. Experimental design for Carausius morosus feeding trial. Lifespans are represented
by horizontal bars divided into six instars and an adult stage. Time is not to scale, and
differences in timing of life-history transitions between groups are not graphically
presented. Vertical lines in juvenile stages denote ecdyses. White bars represent life
stages when food was offered ad libitum (AL); shaded bars represent life stages when
food was restricted (R) to 60% of the amount of food consumed by insects in group
AL on a percent body mass basis. Insects in groups AL and R were maintained for the
duration of their lifespans on ad libitum and restricted diets, respectively. Insects in
the AL-R at 5th and R-AL at 5th groups experienced a diet switch on the first day of the
fifth instar. Insects in the AL-R at Ov group experienced a diet switch at first
oviposition. Because survival to first oviposition was extremely low for insects that
were food-restricted for the duration of juvenile development, I was unable to test the
effects of a diet switch from R to AL at first oviposition.
















0125 -
0100 -
0075 -
0050 -
0 025 -
0000 -


0 100 200 300 400


0 100 200 300 400


1- O0 AL-R at Ov
n = 13




0 100 200 300 400


R
n = 19 juv., n = 7 ad.




) 100 200 300


0 100 200 300 400

Time (days)


Time (days)


Figure 5-2. Mass-specific intake (g dry mass/g*day) consumed by insects in each of five
treatment groups on each day of the study. Curves were constructed by scaling the
duration of each stage for each insect to the average duration of that stage for each
treatment group and fitting a loess smoothing function to these data. Points where
mass-specific intake declined to zero correspond to ecdyses. The first six resulting
time intervals for each group represent juvenile stages and the seventh time interval
for each group represents the adult stage. Arrowheads denote the average age at first
oviposition for each group. Abbreviations: AL = ad libitum, R = restricted, juv. =
juveniles, ad. = adults. See Figure 5-1 for a description of diet treatments. Food
restriction was imposed by offering restricted individuals approximately 60% of the
mass-specific intake of insects in group AL on a stage-specific basis. Because
mass-specific intake of AL insects declined after first oviposition, the amount of food
offered to food-restricted adults after first oviposition was decreased proportionally to
match this decline.

















0.8


U) 0.6
U)


"o 0.4
o

0.2


0.0


-o-AL
-o- AL-R at 5th
---A--- AL-R at Ov

I --*---R-AL at5th
IK /"^ -
^./


0 50 100 150 200 250 300 350

Age (d)

Figure 5-3. Age and size at each life-history transition for insects in each of five treatment
groups. Each point represents mean + standard error at the end of an instar (first six
points for each line), at first oviposition (seventh point for each line), or at death (last
point for each line). Abbreviations: AL = ad libitum, R = restricted. See Figure 5-1
for a description of diet treatments. Sample sizes are the same as in Figure 5-2. See
Table 5-2 forp-values for size and age at each point.














AL
---- AL-R at 5th
125 --A--- AL-R at Ov
-- -R
--+-- R-AL at 5th


a
a a a


25 a
25
bn-


1st Instar 2nd Instar 3rd Instar 4th Instar 5th Instar 6th Instar Pre-Ov Adult Post-Ov Adult


10.0


-o-AL
---- AL-Rat 5th
----- AL-R at Ov
--- R
---- R-AL at 5th


-2.0


1st Instar 2nd Instar 3rd Instar 4th Instar 5th Instar 6th Instar Pre-Ov Adult Post-Ov Adult


Figure 5-4. Duration of (a) and specific growth rate during (b) each life-history stage for insects
in each of five treatment groups. Each point represents mean + standard error. Stages
were categorized as each of six instars, adult prior to first oviposition (pre-ov adult),
and adult after first oviposition (post-ov adult). AL-R at Ov insects lost proportionally
more body mass than AL insects after first oviposition, but all other pairwise
comparisons of post-oviposition adult growth rates were not significant (Mann-
Whitney U tests, p > 0.005). Abbreviations: AL = ad libitum, R = restricted. See
Figure 5-1 for a description of diet treatments. Sample sizes are the same as in Figure
5-2. Means with different letters are significantly different among treatment groups
within life-history stages. See text for data analysis.









Figure 5-5. Event history diagram depicting periods of ad libitum intake, restricted intake, and
reproductive activity for individual stick insects maintained on five diet treatments.
Each horizontal line represents the lifespan of one individual, with insects in each
group arranged in order (top to bottom within a treatment group) from shortest to
longest lifespan. Abbreviations: AL = ad libitum, R = restricted, Repr. Life. =
reproductive lifespan. See Figure 5-1 for a description of diet treatments. Data for
insects that died during the juvenile stages (i.e., insects represented by the shortest 1
or 2 bars for each treatment group) were not included in any analyses except for
survivorship curves (Fig. 5-6).














. . . . . . . i ....... . .





. . . . . . . . ..1
.................x ................. .


...............7.7.7.7..

.............................s
............ .......
.................................7.........
.........


Ad Libitum


* Adult Molt


Restricted

* Repr. Life. Ad Libitum

o Repr. Life. Restricted


AL, n = 15






AL-R at 5th, n =15





AL-RatOv, n =14


R, n = 28


I *


SSSflflS1
SSSSJdS
FSSSiVfl


...............
................ R -A L at 5th n 14
........................
. . . . . ..
..................
..........
..............
............
. . . . .
............


200

Age (days)


100


150


250


300


350


400










Figure 5-6. Kaplan-Meier survivorship curves for the entire lifespan (a) and for the adult
lifespan (b) of insects maintained on five diet treatments, including insects that died
prior to the adult molt. Abbreviations: AL = ad libitum, R = restricted. See Figure 5-1
for a description of diet treatments. For graph b, only insects that survived to
adulthood are included. Only 7 of the insects in group R laid eggs, but all 19
individuals in this treatment group are included in the survivorship curve. For graph
a, pairwise log-rank tests with a set at 0.005 to account for multiple comparisons
indicated that all groups except R and R-AL at 5th differed significantly in longevity.
For graph b, AL and R-AL at 5th insects had significantly enhanced adult longevity
compared to AL-R at 5th, AL-R at Ov, and R insects.


































0 100 200 300


Lifespan (days)


AL, n = 13, n = 13
AL-R at 5th, n = 13
AL-R at Ov, n= 13
R, n = 19
R-AL at 5th, n = 12


0 50


100 150
Adult Lifespan (days)













AL


a, w


R-AL at 5th

T ----db, x


0 20 40 60 80 100 120


Reproductive Lifespan (days)


Figure 5-7. Cumulative fecundity of insects in each of five treatment groups. The x-axis
represents days of the reproductive lifespan, with day 0 representing the first day of
oviposition. Each curve terminates at a point corresponding to the mean duration (
standard error) of reproductive activity (x) and the mean fecundity ( standard error)
for each group (y). Curves were constructed by scaling the reproductive lifespan of
each insect to the mean reproductive lifespan for that group, averaging the cumulative
fecundity of all insects in that group on each day of the reproductive lifespan, and
fitting a smooth spline (df= 7) to the resulting averages. Abbreviations: AL = ad
libitum, R = restricted. See Figure 5-1 for a description of diet treatments. Sample
sizes are the same as in Figure 5-2 for adults. Different letters to the right of each
endpoint indicate significantly different means for fecundity (a, b, and c) and
reproductive lifespan (w, x, y, and z) among treatment groups. See text for data
analysis.











0.008


0.007
a
-ac

Sabc

0.006 C
u
a)
Sb


0.005





0.004
AL AL-R at 5th AL-R at Ov R R-AL at 5th


Figure 5-8. Average egg mass (mean + standard error) for stick insects maintained on five diet
treatments. Abbreviations: AL = ad libitum, R = restricted. See Figure 5-1 for a
description of diet treatments. Sample sizes are the same as in Figure 5-2 for adults.
Means with different letters are significantly different among treatment groups. See
text for data analysis.














120


100 y = 28.053x + 9.0378 0 0

U 80

.4-
0
60 -
6 L AL
SO AL-R at 5th
z 40 A AL-R at Ov

0R
20 R-AL at 5th


0
0 i ---------------------------------

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Cumulative Intake During Reproductive Lifespan (g)

(b)


120 -
O AL
o AL-R at 5th
100 A AL-R at Ov
"R
= 80 R-AL at 5th

.4-
60

0 40 A A y = 0.4184x 7.9985
z 400

20 4


0
50 70 90 110 130 150 170 190 210 230

Adult Lifespan (days)

Figure 5-9. Relationships between realized fecundity and cumulative intake during the
reproductive lifespan (a) and the duration of adult lifespan (b) for all insects that laid
eggs (n = 58) as determined by least squares linear regression. The regression in a is
significant (F1,56 = 276.00, p < 0.0001) with an adjusted R2 value of 0.828. The
regression in b is also significant (F1,56 = 25.67, p < 0.0001) with an adjusted R2 value
of 0.302. Abbreviations: AL = ad libitum, R = restricted. See Figure 5-1 for a
description of diet treatments.









CHAPTER 6
SUMMARY AND CONCLUSIONS

Food availability is arguably one of the most fundamental and often-cited modulators of

phenotypic and life-history plasticity. For my dissertation, I addressed questions about the effects

of changes in food availability during different life stages in two taxa. In Chapters 2-4, I

evaluated the physiological and morphological responses to short-term (e.g., 12-week)

differences in food availability in a species (the green turtle, Chelonia mydas) that experiences

nutritional stochasticity during the juvenile stage in the wild. To elucidate long-term responses to

differences in food availability, I conducted a lifespan study using a more tractable animal model

(the Indian stick insect, Carausius morosus). A summary of the major findings of my work can

be found in Table 6-1.

Animals living in nutritionally stochastic environments demonstrate a variety of

adaptations, including the capacity for compensatory growth (CG) (Wilson and Osbourn 1960,

Reid and White 1977), that enable them to capitalize when conditions are favorable for growth

and reproduction. Although CG has been documented in turtles and lizards (Bjorndal et al. 2003,

Caley and Schwarzkopf 2004), my work is the first to assess the mechanistic basis for this

growth pattern in reptiles. One of the most salient findings to emerge from my work on C. mydas

was the fact that CG is effected via enhanced food conversion efficiency (FCE) rather than

hyperphagia. This result stands in direct contrast to most CG studies in fish (reviewed by Ali et

al. 2003).

Additionally, working at the Cayman Turtle Farm afforded me a unique opportunity to

investigate growth dynamics in green turtles without sampling animals taken from the wild. As a

result, I was able to elucidate the effects of intake and growth rates on a number of parameters

(e.g., body composition, digestive tract morphology, nucleic acid content, and antioxidant









function) that cannot typically be studied in healthy individuals of this endangered species.

Green turtles responded to food restriction by mobilizing lipid reserves, conserving protein

reserves, and down-regulating the size of visceral organs. Assuming that digestive organs are

energetically expensive to maintain (Hornick et al. 2000), a decrease in organ size and

concomitant decreased metabolic rate may explain the improved FCE in food-restricted turtles.

Turtles undergoing CG not only grew faster than continuously ad libitum-fed turtles but also

adjusted their rates of tissue deposition such that body composition and organ morphology were

restored after seven weeks of improved food conditions. Clearly, young green turtles have the

capacity to adjust to fluctuations in food availability.

Given this flexibility in the response to changes in nutritional condition and the

previously demonstrated effects of climate (Limpus and Chaloupka 1997) and population density

(Bjorndal et al. 2000) on growth rates, a comprehensive assessment of green turtle population

health requires a better understanding of short-term growth dynamics. I therefore explored the

potential for measuring a number of biochemical indices as predictors of recent growth in

juvenile green turtles. By analyzing RNA, DNA, and protein content of tissues from the same

turtles I examined in Chapter 2, I was able to correlate these indices and ratios among them with

known short-term growth rates (e.g., during the preceding 10-11 days). The models I developed

predicted 55-68% of the variance in recent growth rates (Chapter 3). Specific growth rate for

body mass was best explained (R2 = 0.68) by RNA content of the liver and condition index

(Fulton's K, Ricker 1975), and specific growth rate for carapace length was best explained (R2 =

0.66) by only RNA content of the liver. Because these analyses rely on destructive tissue

extraction, they are not widely applicable to studies of growth in wild turtles. However, specific

growth rate for body mass was also explained moderately well (R2 = 0.55) by condition index









and DNA content of blood. Both of these parameters are easily quantified with minimal

disturbance to the animal, suggesting that biochemical indices hold promise as potential

indicators of recent growth in wild turtles.

After demonstrating the substantial physiological and morphological plasticity of green

turtles exposed to different nutritional environments over short time scales, I became interested

in the long-term effects of food availability. A conspicuous feature of green turtle growth is the

transient nature of CG after a switch from restricted to ad libitum feeding. In addition, the

occurrence of CG indicates that "normal" growth rates in this species are sub-maximal,

suggesting that rapid growth may be associated with one or more costs (Metcalfe and Monaghan

2001). Mangel and Munch (2005) posited that these costs could include elevated levels of

oxidative damage incurred during CG. The results I presented in Chapter 4 provide the first

empirical evidence supporting this hypothesis. Although antioxidant function of muscle (a post-

mitotic tissue) was unaffected by diet, the activity of glutathione peroxidase (an antioxidant

enzyme) and total antioxidant potential per cell in the liver (a mitotically active tissue) were

approximately two-fold greater in continuously ad libitum-fed turtles than in continuously food-

restricted turtles or fast-growing turtles that had undergone growth compensation. An impaired

antioxidant defense system is therefore a cost of CG in green turtles. However, the duration of

this impairment is unknown, as are its life-history consequences.

The long lifespan and large body size of green turtles were not conducive to evaluating

long-term responses to fluctuations in food availability. Instead, I took a novel approach by using

a parthenogenetic insect as my animal model for Chapter 5, and this tactic proved to be a fruitful

one for investigating questions about life history. In contrast to my work on C. mydas, I did not

find any evidence of CG in C. morosus. It is possible that an herbivorous diet precludes growth









compensation because digestive efficiency may be maximized in animals consuming a

low-quality diet. Although I was unable to quantify the potential life-history costs of CG, my

methodology allowed me to assess the effects of different diet treatments on traits such as

development rate, longevity, and fecundity that could not be measured in C. mydas.

Not surprisingly, insects that experienced food restriction prior to the onset of reproductive

activity progressed through juvenile stages more slowly and were smaller at each molt than

individuals feeding at a consistently high rate. These results provide support for the model of

Day and Rowe (2002) and suggest that development rate in response to food availability

represents a compromise between selection for maximized body size (because of its fitness

benefits) and selection against extended development time (because of its demographic costs)

(Rowe and Ludwig 1991).

Although my results for age and size at developmental transitions are typical in studies of

this kind, the quantification of lifespan and cumulative fecundity in individuals with drastically

different developmental trajectories is a novel contribution of my research. My results indicate

that quantitative food restriction experienced early in development extended lifespan, as is

common in other animal models (e.g., Weindruch and Walford 1988, Austad 1989, Mair et al.

2003, Vaupel et al. 2003, Hatle et al. 2006b). However, this longevity enhancement resulted

from extended development time rather than enhanced adult survival. Conversely, food

restriction experienced later in development or at maturity significantly decreased total lifespan.

In contrast to my results for longevity, food restriction imposed at any point during the

lifespan decreased fecundity. Putative fitness was therefore maximized when daily intake was

also maximized throughout life. These findings indicate that the beneficial effects of early-onset

food restriction on lifespan were negated by the detrimental effects on reproductive output.









Cumulative intake during the reproductive lifespan explained 83% of the variance in fecundity,

indicating that C. morosus primarily allocates incoming adult-derived resources to egg

provisioning. In contrast, body stores appear to be the source of nutrients allocated to somatic

maintenance and survival during adulthood, with proportionally heavier females living longer as

adults than smaller females. Given this breeding strategy, it is not surprising that I found no

evidence for a trade-off between longevity and fecundity, as nutrients allocated to reproduction

and maintenance do not appear to be derived from a common resource pool. These results

indicate that fluctuations in food availability can significantly alter the expression of life-history

traits and that the magnitude of these effects depends on the developmental stage during which

food availability changes and on the timing of resource acquisition relative to allocation.

In conclusion, this dissertation provides new insights into the short- and long-term

consequences of quantitative food restriction and has wide-reaching implications for studies of

food availability in both vertebrates and invertebrates. Furthermore, the successful use of a

parthenogenetic animal model underscores the importance of natural reproductive processes in

studies of this kind, because the true fitness effects of diet can only be evaluated if reproductive

potential is not constrained by methodological limitations. The results of my work highlight the

need for further research into the proximate mechanisms underlying differences in life histories

within and among taxa.











Table 6-1. Summary of traits measured for Chelonia mydas (top half of table) and Carausius
morosus (bottom half of table) maintained on different schedules of ad libitum (ad
lib.) and restricted (rest.) intake.
Continuously Continuously Rest. -- Ad Lib. Ad Lib. -- Rest. Ad Lib. -- Rest.
Trait Measured Ad ib. Rest. During During at First
Development Development Oviposition
C. mydas
Body Size Large Small Intermediate

Growth Rate Fast Slow Fast, with CG

Body Condition High Low High
Conversion Low High Intermediate
Efficiency
Digestive Organ Size Large Small Large
OM, Lipid, and High Low High
Energy Content
Nitrogen Content Low High Low

Cell Size Large Small Intermediate/
Large
RNA:DNA, Liver Intermediate High Low

RNA:DNA, Heart Intermediate Low High

RNA:DNA, Blood High Low Intermediate
Protein Content, Intermediate High Low
Liver
Antioxidant Function High Low Low

C. morosus
Body Size Highest Lowest High Low Highest
at First Ovivosition
Age Youngest Oldest Old Young Youngest
at First Oviposition
Growth Rate Fast Slow Fast, with No CG Intermediate Fast
(Juveniles)
Body Condition High Low High Low High
at Adult Molt
Ovary Size No Difference in Total Number of Ovarioles Among Treatment Groups

Total Lifespan Long Longest Longest Short Shortest
Reproductive Highest Lowest High Lowest Low
Lifespan
Adult Lifespan Long Short Long Short Short

Lifetime Fecundity High Low Intermediate Low Intermediate
Mass-Corrected High Intermediate/ Intermediate Low Intermediate/
Lifetime Fecundity Low Low
Egg Success High Low Intermediate/ Low Intermediate
High
Abbreviations: CG = compensatory growth, OM = organic matter, Rest. = food-restricted, Ad
Lib. = ad libitum-fed. Diet switches from ad libitum to restricted feeding were not tested for C.
mydas.









LIST OF REFERENCES


Agarwal, S., S. Sharma, V. Agrawal, and N. Roy. 2005. Caloric restriction augments ROS
defense in S. cerevisiae, by a Sir2p independent mechanism. Free Radical Research 39:55-
62.

Ali, M., A. Nicieza, and R. J. Wootton. 2003. Compensatory growth in fishes: a response to
growth depression. Fish and Fisheries 4:147-190.

Altmann, J., and S. C. Alberts. 2005. Growth rates in a wild primate population: ecological
influences and maternal effects. Behavioral Ecology and Sociobiology 57:490-501.

Altwegg, R., and H.-U. Reyer. 2003. Patterns of natural selection on size at metamorphosis in
water frogs. Evolution 57:872-882.

Alvarez, D., and N. B. Metcalfe. 2005. Catch-up growth and swimming performance in
threespine sticklebacks (Gasterosteus aculeatus): seasonal changes in the cost of
compensation. Canadian Journal of Fisheries and Aquatic Sciences 62:2169-2176.

Arendt, J. D. 1997. Adaptive intrinsic growth rates: an integration across taxa. Quarterly Review
of Biology 72:149-177.

Arendt, J. D., and D. S. Wilson. 2000. Population differences in the onset of cranial ossification
in pumpkinseed (Lepomis gibbosus), a potential cost of rapid growth. Canadian Journal of
Fisheries and Aquatic Sciences 57:351-356.

Arendt, J. D., D. S. Wilson, and E. Stark. 2001. Scale strength as a cost of rapid growth in
sunfish. Oikos 93:95-100.

Arnekleiv, J. V., A. G. Finstad, and L. Running. 2006. Temporal and spatial variation in growth
of juvenile Atlantic salmon. Journal of Fish Biology 68:1062-1076.

Association of Official Agricultural Chemists (AOAC) 1960. Official methods of analysis of the
Association of Official Agricultural Chemists, 9th edition. Association of Official
Agricultural Chemists, Washington, D. C.

Association of Official Agricultural Chemists (AOAC) 1984. Official methods of analysis of the
Association of Official Agricultural Chemists, 14th edition. Association of Official
Agricultural Chemists, Washington, D. C.

Atchley, W. R. 1984. Ontogeny, timing of development, and genetic variance-covariance
structure. American Naturalist 123:519-540.

Austad, S. N. 1989. Life extension by dietary restriction in the bowl and doily spider, Frontinella
pyramitela. Experimental Gerontology 24:83-92.









Balazs, G. H. 1995. Growth rates of immature green turtles in the Hawaiian Archipelago. Pages
117-125 in K.A. Bjorndal, editor. Biology and conservation of sea turtles, revised edition.
Smithsonian Institution Press, Washington, DC.

Ballinger, R. E. 1977. Reproductive strategies: food availability as a source of proximal variation
in a lizard. Ecology 58:628-635.

Barja, G. 2004. Aging in vertebrates, and the effect of caloric restriction: a mitochondrial free
radical production-DNA damage mechanism? Biological Reviews 79:235-251.

Bastrop, R., R. Spangenberg, and K. Jirss. 1991. Biochemical adaptation of juvenile carp
(Cyprinus carpio L.) to food deprivation. Comparative Biochemistry and Physiology A
98:143-149.

Belanger, F., P. U. Blier, and J.-D. Dutil. 2002. Digestive capacity and compensatory growth in
Atlantic cod (Gadus morhua). Fish Physiology and Biochemistry 26:121-128.

Berrigan, D., and J. C. Koella. 1994. The evolution of reaction norms: simple models for age and
size at maturity. Journal of Evolutionary Biology 7:549-566.

Billerbeck, J. M., T. E. Lankford, Jr., and D. 0. Conover. 2001. Evolution of intrinsic growth
and energy acquisition rates. I. Trade-offs with swimming performance in Menidia
menidia. Evolution 55:1863-1872.

Bjorndal, K. A. 1997. Feeding ecology and nutrition in sea turtles. Pages 199-231 in P.L. Lutz
and J.A. Musick, editors. The biology of sea turtles. CRC Press, Boca Raton, FL.

Bjorndal, K. A., A. B. Bolten, and M. Y. Chaloupka. 2000. Green turtle somatic growth model:
evidence for density dependence. Ecological Applications 10:269-282.

Bjorndal, K. A., A. B. Bolten, T. Dellinger, C. Delgado, and H. R. Martins. 2003. Compensatory
growth in oceanic loggerhead sea turtles: response to a stochastic environment. Ecology
84:1237-1249.

Blanckenhorn, W. U. 2000. The evolution of body size: what keeps organisms small? Quarterly
Review of Biology 75:385-407.

Boggs, C. L. 1992. Resource allocation: exploring connections between foraging and life history.
Functional Ecology 6:508-518.

Boggs, C. L., and K. D. Freeman. 2005. Larval food limitation in butterflies: effects on adult
resource allocation and fitness. Oecologia 144:353-361.

Boggs, C. L., and C. L. Ross. 1993. The effect of adult food limitation on life history traits in
Speyeria mormonia (Lepidoptera: Nymphalidae). Ecology 74:433-441.

Bohm, F., R. Edge, E. Land, D. J. McGarvey, and T. G. Truscott. 1997. Carotenoids enhance
vitamin E antioxidant efficiency. Journal of the American Chemical Society 119:621-622.









Bolten, A. B., and K. A. Bjorndal. 1992. Blood profiles for a wild population of green turtles
(Chelonia mydas) in the southern Bahamas: size-specific and sex-specific relationships.
Journal of Wildlife Diseases 28:407-413.

Boujard, T., C. Burel, F. Medale, G. Haylor, and A. Moisan. 2000. Effect of past nutritional
history and fasting on feed intake and growth in rainbow trout Oncorhynchus mykiss.
Aquatic Living Resources 13:129-137.

Bradley, J. T., M. Masetti, A. Cecchettini, and F. Giorgi. 1995. Vitellogenesis in the
allatectomized stick insect Carausius morosus (Br.) (Phasmatodea: Lonchodinae).
Comparative Biochemistry and Physiology 110B:255-266.

Bradshaw, W. E., and K. Johnson. 1995. Initiation of metamorphosis in the pitcher-plant
mosquito: effects of larval growth history. Ecology 76:2055-2065.

Brock, P. D. 2000. A complete guide to breeding stick and leaf insects. Kingdom Books, Havant,
UK.

Broderick, A. C., F. Glen, B. J. Godley, and G. C. Hays. 2003. Variation in reproductive output
of marine turtles. Journal of Experimental Marine Biology and Ecology 288:95-109.

Broekhuizen, N., W. S. C. Gurney, A. Jones, and A. D. Bryant. 1994. Modelling compensatory
growth. Functional Ecology 8:770-782.

Brzek, P., and M. Konarzewski. 2001. Effect of food shortage on the physiology and competitive
abilities of sand martin (Riparia riparia) nestlings. Journal of Experimental Biology
204:3065-3074.

Brzek, P., and M. Konarzewski. 2004. Effect of refeeding on growth, development, and behavior
of undernourished bank swallow (Riparia riparia) nestlings. The Auk 121:1187-1198.

Buckley, L., E. Caldarone, and T.-L. Ong. 1999. RNA-DNA ratio and other nucleic acid-based
indicators for growth and condition of marine fishes. Hydrobiologia 401:265-277.

Bull, C. D., and N. B. Metcalfe. 1997. Regulation of hyperphagia in response to varying energy
deficits in overwintering juvenile Atlantic salmon. Journal of Fish Biology 50:498-510.

Bulow, F. J. 1970. RNA-DNA ratios as indicators of recent growth rates of a fish. Journal of the
Fisheries Research Board of Canada 27:2343-2349.

Calbet, A., and M. Alcaraz. 1997. Growth and survival rates of early developmental stages of
Acartia grani (Copepoda: Calanoida) in relation to food concentration and fluctuations in
food supply. Marine Ecology Progress Series 147:181-186.

Caldarone, E. M. 2005. Estimating growth in haddock larvae Melanogrammus aeglefinus from
RNA:DNA ratios and water temperature. Marine Ecology Progress Series 293:241-252.









Caldarone, E. M., C. M. Clemmesen, E. Berdalet, T. J. Miller, A. Folkvord, G. J. Holt, M. P.
Olivar, and I. M. Suthers. 2006. Intercalibration of four spectrofluorometric protocols for
measuring RNA/DNA ratios in larval and juvenile fish. Limnology and Oceanography:
Methods 4:153-163.

Caley, M. J., and L. Schwarzkopf. 2004. Complex growth rate evolution in a latitudinally
widespread species. Evolution 58:862-869.

Calow, P. 1979. The cost of reproduction a physiological approach. Biological Reviews of the
Cambridge Philosophical Society 54:23-40.

Carey, J. R., P. Liedo, L. Harshman, X. Liu, H.-G. Muller, L. Partridge, and J.-L. Wang. 2002a.
Food pulses increase longevity and induce cyclical egg production in Mediterranean fruit
flies. Functional Ecology 16:313-325.

Carey, J. R., P. Liedo, L. Harshman, Y. Zhang, H.-G. Muller, L. Partridge, and J.-L. Wang.
2002b. Life history response of Mediterranean fruit flies to dietary restriction. Aging Cell
1:140-148.

Carey, J. R., P. Liedo, H.-G. Muller, J.-L. Wang, and J. W. Vaupel. 1998. A simple graphical
technique for displaying individual fertility data and cohort survival: case study of 1000
Mediterranean Fruit Fly females. Functional Ecology 12:359-363.

Carr, A. 1987. New perspectives on the pelagic stage of sea turtle development. Conservation
Biology 1:103-121.

Carter, C. G., G. S. Seeto, A. Smart, S. Clarke, and R. J. van Barneveld. 1998. Correlates of
growth in farmed juvenile southern bluefin tuna Thunnus maccoyii (Castelnau).
Aquaculture 161:107-119.

Carvalho, G. B., P. Kapahi, D. J. Anderson, and S. Benzer. 2006. Allocrine modulation of
feeding behavior by the sex peptide of Drosophila. Current Biology 16:692-696.

Chaloupka, M., and C. Limpus. 1998. Modeling green turtle survivorship rates. Pages 24-26 in S.
P. Epperly and J. Braun, editors. Proceedings of the 17th annual sea turtle symposium.
NOAA technical memorandum NMFS-SEFSC-415, Miami, FL.

Chapman, R. F. 1998. The insects: structure and function. Cambridge University Press, New
York, NY.

Cherel, Y., F. Freby, J. Gilles, and J.-P. Robin. 1993. Comparative fuel metabolism in Gentoo
and King penguins: adaptation to brief versus prolonged fasting. Polar Biology 13:263-
269.

Chiang, H. C., and A. C. Hodson. 1950. The relation of copulation to fecundity and population
growth in Drosophila melanogaster. Ecology 31:255-259.









Chippindale, A. K., A. M. Leroi, S. B. Kim, and M. R. Rose. 1993. Phenotypic plasticity and
selection in Drosophila life-history evolution. I. Nutrition and the cost of reproduction.
Journal of Evolutionary Biology 6:171-193.

Chong, J.-H., and R. D. Getting. 2006. Influence of temperature and mating status on the
development and fecundity of the mealybug parasitoid, Anagyrus sp. nov. nr. sinope Noyes
and Menezes (Hymenoptera: Encyrtidae). Environmental Entomology 35:1188-1197.

Christiansen, J. S., I. Martinez, M. Jobling, and A. B. Amin. 1992. Rapid somatic growth and
muscle damage in a salmonid fish. Basic and Applied Myology 2:235-239.

Clemmesen, C. 1994. The effect of food availability, age or size on the RNA/DNA ratio of
individually measured herring larvae: laboratory calibration. Marine Biology 118:377-382.

Cooper, T. M., R. J. Mockett, B. H. Sohal, R. S. Sohal, and W. C. Orr. 2004. Effect of caloric
restriction on life span of the housefly, Musca domestic. FASEB Journal 18:1591-1593.

Cutter, A. D. 2004. Sperm-limited fecundity in nematodes: how many sperm are enough?
Evolution 58:651-655.

Dagg, M. 1977. Some effects of patchy food environments on copepods. Limnology and
Oceanography 22:99-107.

Dahlhoff, E. P. 2004. Biochemical indicators of stress and metabolism: applications for marine
ecological studies. Annual Review of Physiology 66:183-207.

Darzynkiewicz, Z., T. Sharpless, L. Staiano-Coico, and M. R. Melamed. 1980. Subcompartments
of the Gi phase of cell cycle detected by flow cytometry. Proceedings of the National
Academy of Sciences USA 77:6696-6699.

David, J. R. 1970. Le nombre d'ovarioles chez Drosophila melanogaster: relation avec la
fecundite et le valeur adaptative. Archives de Zoologie Experimentale et Generale
111:357-370.

Dawson, A., S. A. Hinsley, P. N. Ferns, R. H. C. Bonser, and L. Eccleston. 2000. Rate of moult
affects feather quality: a mechanism linking current reproductive effort to future survival.
Proceedings of the Royal Society of London B 267:2093-2098.

Day, T., and L. Rowe. 2002. Developmental thresholds and the evolution of reaction norms for
age and size at life-history transitions. American Naturalist 159:338-350.

De Clercq, P., and D. Degheele. 1997. Effects of mating status on body weight, oviposition, egg
load, and predation in the predatory stinkbug Podisus maculiventris (Heteroptera:
Pentatomidae). Annals of the Entomological Society of America 90:121-127.

Dhahbi, J. M., P. L. Mote, J. Wingo, B. C. Rowley, S. X. Cao, R. L. Walford, and S. R. Spindler.
2001. Caloric restriction alters the feeding response of key metabolic enzyme genes.
Mechanisms of Ageing and Development 122:1033-1048.









Dixon, A. F. G., and B. K. Agarwala. 2002. Triangular fecundity function and ageing in ladybird
beetles. Ecological Entomology 27:433-440.

Dmitriew, C., and L. Rowe. 2005. Resource limitation, predation risk and compensatory growth
in a damselfly. Oecologia 142:150-154.

Einum, S. and I. A. Fleming. 2000. Selection against late emergence and small offspring in
Atlantic salmon (Salmo salar). Evolution 54:628-639.

Ferron, A., and W. C. Leggett. 1994. An appraisal of condition measures for marine fish larvae.
Advances in Marine Biology 30:217-303.

Forman, R. T. T. 1995. Land mosaics: the ecology of landscapes and regions. Cambridge
University Press, Cambridge, UK.

Foster, S. P., and A. J. Howard. 1999. The effects of mating, age at mating, and plant stimuli, on
the lifetime fecundity and fertility of the generalist herbivore Epiphyaspostvittana.
Entomologia Experimentalis et Applicata 91:287-295.

Gallaher, R. N., C. 0. Weldon, and J. G. Futral. 1975. An aluminum block digester for plant and
soil analysis. Soil Science Society of America Proceedings 39:803-806.

Gaylord, T. G., and D. M. Gatlin, III. 2000. Assessment of compensatory growth in channel
catfish Ictaluruspunctatus R. and associated changes in body condition indices. Journal of
the World Aquaculture Society 31:326-336.

Gebhardt, M. D., and S. C. Stearns. 1988. Reaction norms for developmental time and weight at
eclosion in Drosophila mercatorum. Journal of Evolutionary Biology 1:335-354.

Gebhardt, M. D., and S. C. Stearns. 1993. Phenotypic plasticity for life history traits in
Drosophila melanogaster. I. Effect on phenotypic and environmental correlations. Journal
of Evolutionary Biology 6:1-16.

Gilliers, C., R. Amara, J.-P. Bergeron, and 0. Le Pape. 2004. Comparison of growth and
condition indices of juvenile flatfish in different coastal nursery grounds. Environmental
Biology of Fishes 71:189-198.

Gillott, C. 2003. Male accessory gland secretions: modulators of female reproductive physiology
and behavior. Annual Review of Entomology 48:163-184.

Grafen, A. 1988. On the uses of data on lifetime reproductive success. Pages 454-471 in T. H.
Clutton-Brock, editor. Reproductive success: studies of individual variation in contrasting
breeding systems. University of Chicago Press, Chicago, IL.

Grattagliano, I., P. Portincasa, T. Cocco, A. Moschetta, M. Di Paola, V. 0. Palmieri, and G.
Palasciano. 2004. Effect of dietary restriction and N-acetylcysteine supplementation on
intestinal mucosa and liver mitochondrial redox status and function in aged rats.
Experimental Gerontology 39:1323-1332.









Gredilla, R., and G. Barja. 2005. Minireview: the role of oxidative stress in relation to caloric
restriction and longevity. Endocrinology 146:3713-3717.

Hagopian, K., J. J. Ramsey, and R. Weindruch. 2003. Caloric restriction increases gluconeogenic
and transaminase enzyme activities in mouse liver. Experimental Gerontology 38:267-278.

Hambleton, L. G. 1977. Semiautomated method for simultaneous determination of phosphorus,
calcium and crude protein in animal feeds. Journal of the Association of Official Analytical
Chemists 60:845-852.

Hart, R. W., and A. Turturro. 1998. Evolution and dietary restriction. Experimental Gerontology
33:53-60.

Hatle, J. D., D. W. Borst, and S. A. Juliano. 2004. Plasticity and canalization in the control of
reproduction in the lubber grasshopper. Integrative and Comparative Biology 43:635-645.

Hatle, J. D., S. A. Juliano, and D. W. Borst. 2003. Hemolymph ecdysteroids do not affect
vitellogenesis in the lubber grasshopper. Archives of Insect Biochemistry and Physiology
52:45-57.

Hatle, J. D., T. Waskey, Jr., and S. A. Juliano. 2006a. Plasticity of grasshopper vitellogenin
production in response to diet is primarily a result of changes in fat body mass. Journal of
Comparative Physiology B 176:27-34.

Hatle, J. D., S. M. Wells, L. E. Fuller, I. C. Allen, L. J. Gordy, S. Melnyk, and J. Quattrochi.
2006b. Calorie restriction and late-onset calorie restriction extend lifespan but do not alter
protein storage in female grasshoppers. Mechanisms of Ageing and Development 127:883-
891.

Hayes, J. P., and J. S. Shonkwiler. 2001. Morphometric indicators of body condition: worthwhile
or wishful thinking? Pages 8-38 in J. R. Speakman, editor. Body composition analysis of
animals: a handbook of non-destructive methods. Cambridge University Press, Cambridge,
UK.

Hofer, H., and M. L. East. 2003. Behavioral processes and costs of co-existence in female
spotted hyenas: a life history perspective. Evolutionary Ecology 17:315-331.

Holliday, R. 1989. Food, reproduction and longevity: is the extended lifespan of calorie-
restricted animals an evolutionary adaptation? Bioessays 10:125-127.

Honek, A. 1993. Intraspecific variation in body size and fecundity in insects: a general
relationship. Oikos 66:483-492.

Hornick, J. L., C. Van Eenaeme, 0. Gerard, I. Dufrasne, and L. Istasse. 2000. Mechanisms of
reduced and compensatory growth. Domestic Animal Endocrinology 19:121-132.

Hume, I. D., and H. Biebach. 1996. Digestive tract function in the long-distance migratory
garden warbler, Sylvia born. Journal of Comparative Physiology 166B:388-395.









Hyun, D.-H., S. S. Emerson, D.-G. Jo, M. P. Mattson, and R. de Cabo. 2006. Calorie restriction
up-regulates the plasma membrane redox system in brain cells and suppresses oxidative
stress during aging. Proceedings of the National Academy of Sciences 103:19908-19912.

Islam, M. S., and M. Tanaka. 2005. Nutritional condition, starvation status and growth of early
juvenile Japanese sea bass (Lateolabraxjaponicus) related to prey distribution and feeding
in the nursery ground. Journal of Experimental Marine Biology and Ecology 323:172-183.

Jackson, J. B. C., M. X. Kirby, W. H. Berger, K. A. Bjorndal, and 15 others. 2001. Historical
overfishing and the recent collapse of coastal ecosystems. Science 293:629-638.

Janzen, F. J., J. K. Tucker, and G. L. Paukstis. 2000. Experimental analysis of an early life-
history stage: selection on size of hatchling turtles. Ecology 81:2290-2304.

Jennings, B. J., S. E. Ozanne, M. W. Dorling, and C. N. Hales. 1999. Early growth determines
longevity in male rats and may be related to telomere shortening in the kidney. FEBS
Letters 448:4-8.

Jervis, M. A., C. L. Boggs, and P. N. Ferns. 2005. Egg maturation strategy and its associated
trade-offs: a synthesis focusing on Lepidoptera. Ecological Entomology 30:359-375.

Jespersen, L. B., and S. Toft. 2003. Compensatory growth following early nutritional stress in
the wolf spider Pardosaprativaga. Functional Ecology 17:737-746.

Ji, L. L., and S. Leichtweis. 1997. Exercise and oxidative stress: Sources of free radicals and
their impact on antioxidant systems. Age 20:91-106.

Jobling, M., and S. J. S. Johansen. 1999. The lipostat, hyperphagia and catch-up growth.
Aquaculture Research 30:473-478.

Johansen, S. J. S., M. Ekli, B. Stangnes, and M. Jobling. 2001. Weight gain and lipid deposition
in Atlantic salmon, Salmo salar, during compensatory growth: evidence for lipostatic
regulation? Aquaculture Research 32:963-974.

Jonsson, K. I. 1997. Capital and income breeding as alternative tactics of resource use in
reproduction. Oikos 78:57-66.

Joshi, A., W.-P. Wu, and L. D. Mueller. 1998. Density-dependent natural selection in
Drosophila: adaptation to adult crowding. Evolutionary Ecology 12:363-376.

Judge, S., Y. M. Jang, A. Smith, T. Hagen, and C. Leeuwenburgh. 2005. Age-associated
increases in oxidative stress and antioxidant enzyme activities in cardiac interfibrillar
mitochondria: implications for the mitochondrial theory of aging. FASEB Journal 19:419-
421.

Juliano, S. A., J. R. Olson, E. G. Murrell, and J. D. Hatle. 2004. Plasticity and canalization of
insect reproduction: testing alternative models of life history transitions. Ecology 85:2986-
2996.









Kaeberlein, T. L., E. D. Smith, M. Tsuchiya, K. L. Welton, J. H. Thomas, S. Fields, B. K.
Kennedy, and M. Kaeberlein. 2006. Lifespan extension in Caenorhabditis elegans by
complete removal of food. Aging Cell 5:487-494.

Kalani, R., S. Judge, C. Carter, M. Pahor, and C. Leeuwenburgh. 2006. Effects of caloric
restriction and exercise on age-related, chronic inflammation assessed by C-reactive
protein and interleukin-6. Journal of Gerontology 61A:211-217.

Karasov, W. H., and B. Pinshow. 1998. Changes in lean mass and in organs of nutrient
assimilation in a long-distance passerine migrant at a springtime stopover site.
Physiological Zoology 71:435-448.

Karasov, W. H., B. Pinshow, J. M. Starck, and D. Afik. 2004. Anatomical and histological
changes in the alimentary tract of migrating blackcaps (Sylvia atricapilla): a comparison
among fed, fasted, food-restricted, and refed birds. Physiological and Biochemical
Zoology 77:149-160.

Kenagy, G. J., C. Veloso, and F. Bozinovic. 1999. Daily rhythms of food intake and feces
reingestion in the degu, an herbivorous Chilean rodent: optimizing digestion through
coprophagy. Physiological and Biochemical Zoology 72:78-86.

Kindlmann, P., A. F. G. Dixon, and I. Dostalkova. 2001. Role of ageing and temperature in
shaping reaction norms and fecundity functions in insects. Journal of Evolutionary Biology
14:835-840.

Kitaysky, A. S. 1999. Metabolic and developmental responses of alcid chicks to experimental
variation in food intake. Physiological and Biochemical Zoology 72:462-473.

Klowden, M. J. 2002. Physiological systems in insects. Academic Press, New York, NY.

Kunz, C., and J. Ekman. 2000. Genetic and environmental components of growth in nestling blue
tits (Parus caeruleus). Journal of Evolutionary Biology 13:199-212.

Kuropat, C., R. Mercaldo-Allen, E. Caldarone, R. Goldberg, B. Phelan, and F. Thurberg. 2002.
Evaluation of RNA concentration as an indicator of growth in young-of-the-year winter
flounder Pseudopleuronectes americanus and tautog Tautoga onitis. Marine Ecology
Progress Series 230:265-274.

Leather, S. R. 1988. Size, reproductive potential and fecundity in insects: things aren't as simple
as they seem. Oikos 51:386-389.

Lee, K. A., W. H. Karasov, and E. Caviedes-Vidal. 2002. Digestive response to restricted
feeding in migratory yellow-rumped warblers. Physiological and Biochemical Zoology
75:314-323.









Leeuwenburgh, C., R. Fiebig, R. Chandwaney, and L. L. Ji. 1994. Aging and exercise training in
skeletal muscle: responses of glutathione and antioxidant enzyme systems. American
Journal of Physiology Regulatory, Integrative, and Comparative Physiology 267:R439-
R445.

Leips, J., and J. Travis. 1994. Metamorphic responses to changing food levels in two species of
hylid frogs. Ecology 75:1345-1356.

Leterrier, C., and Y. Nys. 1992. Composition, cortical structure and mechanical properties of
chicken tibiotarsi: effect of growth rate. British Poultry Science 33:925-939.

Li, S., T. Yan, J.-Q. Yang, T. D. Oberley, and L. W. Oberley. 2000. The role of cellular
glutathione peroxidase redox regulation in the suppression of tumor cell growth by
manganese superoxide dismutase. Cancer Research 60:3927-3939.

Limpus, C. J. 1992. Estimation of tag loss in marine turtle research. Wildlife Research 19:457-
469.

Limpus, C., and M. Chaloupka. 1997. Nonparametric regression modelling of green sea turtle
growth rates (southern Great Barrier Reef). Marine Ecology Progress Series 149:23-34.

Lochmiller, R. L., S. S. Ditchkoff, and J. A. Sinclair. 2000. Developmental plasticity of
postweanling cotton rats (Sigmodon hispidus) as an adaptation to nutritionally stochastic
environments. Evolutionary Ecology 14:127-142.

Lochmiller, R. L., M. R. Vestey, and S. T. McMurry. 1993. Selected immune responses of adult
cotton rats (Sigmodon hispidus) to dietary restriction. Comparative Biochemistry and
Physiology 104A:593-599.

L6pez-Torres, M., R. Gredilla, A. Sanz, and G. Barja. 2002. Influence of aging and long-term
caloric restriction on oxygen radical generation and oxidative DNA damage in rat liver
mitochondria. Free Radical Biology and Medicine 32:882-889.

Luhtala, T. A., E. B. Roecker, T. Pugh, R. J. Feuers, and R. Weindruch. 1994. Dietary restriction
attenuates age-related increases in rat skeletal muscle antioxidant enzyme activities.
Journals of Gerontology 49:B231-B238.

Madsen, T. and R. Shine. 2000. Silver spoons and snake body sizes: prey availability early in life
influences long-term growth rates of free-ranging pythons. Journal of Animal Ecology
69:952-958.

Mangel, M., and S. B. Munch. 2005. A life-history perspective on short- and long-term
consequences of compensatory growth. American Naturalist 166:E 155-E 176.

Mangel, M., and J. Stamps. 2001. Trade-offs between growth and mortality and the maintenance
of individual variation in growth. Evolutionary Ecology Research 3:583-593.









Mair, W., P. Goymer, S. D. Pletcher, and L. Partridge. 2003. Demography of dietary restriction
and death in Drosophila. Science 301:1731-1733.

Mair, W., C. M. Sgro, A. P. Johnson, T. Chapman, and L. Partridge. 2004. Lifespan extension by
dietary restriction in female Drosophila melanogaster is not caused by a reduction in
vitellogenesis or ovarian activity. Experimental Gerontology 39:1011-1019.

Masoro, E. J., and S. N. Austad. 1996. The evolution of the antiaging action of dietary
restriction: a hypothesis. Journal of Gerontology 51A:B387-B391.

Maxwell, M. H., P. M. Hocking, and G. W. Robertson. 1992. Differential leucocyte responses to
various degrees of food restriction in broilers, turkeys and ducks. British Poultry Science
33:177-187.

Maxwell, M. H., G. W. Robertson, S. Spence, and C. C. McCorquodale. 1990a. Comparison of
haemotological values in restricted- and ad libitum-fed domestic fowls: red blood cell
characteristics. British Poultry Science 31:407-413.

Maxwell, M. H., G. W. Robertson, S. Spence, and C. C. McCorquodale. 1990b. Comparison of
haemotological values in restricted- and ad libitum-fed domestic fowls: white blood cells
and thrombocytes. British Poultry Science 31:399-405.

McCauley, S. J., and K. A. Bjorndal. 1999. Conservation implications of dietary dilution from
debris ingestion: sublethal effects in post-hatchling loggerhead sea turtles. Conservation
Biology 13:925-929.

McDonald, P. G., P. D. Olsen, and A. Cockburn. 2005. Sex allocation and nestling survival in a
dimorphic raptor: does size matter? Behavioral Ecology 16:922-930.

Melzner, F., J. W. Forsythe, P. G. Lee, J. B. Wood, U. Piatkowski, and C. Clemmesen. 2005.
Estimating recent growth in the cuttlefish Sepia officinalis: are nucleic acid-based
indicators for growth and condition the method of choice? Journal of Experimental Marine
Biology and Ecology 317:37-51.

Mercaldo-Allen, R., C. Kuropat, and E. M. Caldarone. 2006. A model to estimate growth in
young-of-the-year tautog, Tautoga onitis, based on RNA/DNA ratio and seawater
temperature. 2006. Journal of Experimental Marine Biology and Ecology 329:187-195.

Metcalfe, N. B., and P. Monaghan. 2001. Compensation for a bad start: grow now, pay later?
TRENDS in Ecology and Evolution 16:254-260.

Metcalfe, N. B., and P. Monaghan. 2003. Growth versus lifespan: perspectives from evolutionary
ecology. Experimental Gerontology 38:935-940.

Miglavs, I., and M. Jobling. 1989. Effects of feeding regime on food consumption, growth rates
and tissue nucleic acids in juvenile Arctic charr, Salvelinus alpinus, with particular respect
to compensatory growth. Journal of Fish Biology 34:947-957.









Moehrlin, G. S., and S. A. Juliano. 1998. Plasticity of insect reproduction: testing models of
flexible and fixed development in response to different growth rates. Oecologia 115:492-
500.

Moreau, J., B. Benrey, and D. Thiery. 2006. Grape variety affects larval performance and also
female reproductive performance of the European grapevine moth Lobesia botrana
(Lepidoptera: Tortricidae). Bulletin of Entomological Research 96:205-212.

Morey, S., and D. Reznick. 2000. A comparative analysis of plasticity in larval development in
three species of spadefoot toads. Ecology 81:1736-1749.

Mura, C. V., X. Gong, A. Taylor, R. Villalobos-Molina, and M. M. Scrofano. 1996. Effects of
calorie restriction and aging on the expression of antioxidant enzymes and ubiquitin in the
liver of Emory mice. Mechanisms of Ageing and Development 91:115-129.

Musick, J. A., and C. J. Limpus. 1997. Habitat utilization and migration in juvenile sea turtles.
Pages 137-163 in P.L. Lutz and J.A. Musick, editors. The biology of sea turtles. CRC
Press, Boca Raton, FL.

Nagy, L. R., and R. T. Holmes. 2005. Food limits annual fecundity of a migratory songbird: an
experimental study. Ecology 86:675-681.

Nakamura, W., S. Hosoda, and K. Hayashi. 1974. Purification and properties of rat liver
glutathione peroxidase. Biochimica et Biophysica Acta. 358:251-261.

Nijhout, H. F. 1994. Insect hormones. Princeton University Press, Princeton, NJ.

Niki, E., N. Noguchi, H. Tsuchihashi, and N. Gotoh. 1995. Interaction among vitamin C, vitamin
E, and 3-carotene. American Journal of Clinical Nutrition 62 (suppl):1322S-1326S.

Nikki, J., J. Pirhonen, M. Jobling, and J. Karjalainen. 2004. Compensatory growth in juvenile
rainbow trout, Oncorhynchus mykiss (Walbaum), held individually. Aquaculture 235:285-
296.

Okumura, T., T. Nagasawa, I. Hayashi, and Y. Sato. 2002. Effects of starvation on RNA:DNA
ratio, glycogen content, and C:N ratio in columellar muscle of the Japanese turban shell
Turbo (Batillus) cornutus (Gastropoda). Fisheries Science 68:306-312.

Olsson, M., and R. Shine. 1997. The limits to reproductive output: offspring size versus number
in the sand lizard (Lacerta agilis). American Naturalist 149:179-188.

Olsson, M., and R. Shine. 2002. Growth to death in lizards. Evolution 56:1867-1870.

Packard, G. C., J. K. Tucker, D. Nicholson, and M. J. Packard. 1997. Cold tolerance in hatchling
slider turtles (Trachemys scripta). Copeia 1997:339-345.

Parr Instrument Company. 1960. Oxygen bomb calorimetry and combustion methods. Technical
manual 130:1-56.









Partridge, L, D. Gems, and D. J. Withers. 2005. Sex and death: what is the connection? Cell
120:461-472.

Partridge, L, M. D. W. Piper, and W. Mair. 2005. Dietary restriction in Drosophila. Mechanisms
of Ageing and Development 126:938-950.

Patterson, D. C., R. W. J. Steen, and D. J. Kilpatrick. 1995. Growth and development in beef
cattle. 1. Direct and residual effects of plane of nutrition during early life on components of
gain and food efficiency. Journal of Agricultural Science 124:91-100.

Perneger, T. V. 1998. What's wrong with Bonferroni adjustments? British Medical Journal
316:1236-1238.

Perrin, N., M. C. Bradley, and P. Calow. 1990. Plasticity of storage allocation in Daphnia
magna. Oikos 59:70-74.

Perrin, N., and R. M. Sibly. 1993. Dynamic models of energy allocation and investment. Annual
Review of Ecology and Systematics 24:379-410.

Phaneuf, S., and C. Leeuwenburgh. 2002. Cytochrome c release from mitochondria in the aging
heart: a possible mechanism for apoptosis with age. American Journal of Physiology -
Regulatory, Integrative, and Comparative Physiology 282:R423-R430.

Pijnacker, L. P. 1966. The maturation divisions of the parthenogenetic stick insect Carausius
morosus Br. (Orthoptera, Phasmidae). Chromosoma 19:99-112.

Plotz, J., W. Ekau, and P. J. H. Reijnders. 1991. Diet of Weddell seals Leptonychotes weddellii at
Vestkapp, Eastern Weddell Sea (Antarctica), in relation to local food supply. Marine
Mammal Science 7:136-144.

Prasad, N. G., M. Shakarad, M. Rajamani, and A. Joshi. 2003. Interaction between the effects of
maternal and larval levels of nutrition on pre-adult survival in Drosophila melanogaster.
Evolutionary Ecology Research 5:903-911.

Quinn, G. P., and M. J. Keough. 2002. Experimental design and data analysis for biologists.
Cambridge University Press, New York, NY.

Reich, K. J., K. A. Bjorndal, and A. B. Bolten. In review. Testing Carr's conjecture: green turtles
shift from oceanic to neritic habitats during ontogeny. Proceedings of the National
Academy of Sciences.

Reid, J. T., and 0. D. White. 1977. The phenomenon of compensatory growth. Pages 16-27 in
Proceedings of the Cornell Nutrition Conference for Feed Manufacturers 21-23 June 1977,
Cornell University, Ithaca, NY.

Reznick, D. 1985. Costs of reproduction: an evaluation of the empirical evidence. Oikos 44:257-
267.









Reznick, D. 1992. Measuring the costs of reproduction. Trends in Ecology and Evolution 7:42-
45.

Reznick, D., H. Callahan, and R. Llauredo. 1996. Maternal effects on offspring quality in
poeciliid fishes. American Zoologist 36:147-156.

Richner, H., P. Schneiter, and H. Stirnimann. 1989. Life-history consequences of growth rate
depression: an experimental study on carrion crows (Corvus corone corone L.). Functional
Ecology 3:617-624.

Ricker, W. E. 1975. Computation and interpretation of biological statistics of fish populations.
Bulletin of the Fisheries Research Board of Canada 191:1-382.

Roff, D. A. 1992. The evolution of life histories: theory and analysis. Chapman and Hall, New
York, NY.

Rooker, J. R., and G. J. Holt. 1996. Application of RNA:DNA ratios to evaluate the condition
and growth of larval and juvenile red drum (Sciaenops ocellatus). Marine and Freshwater
Research 47:283-290.

Rooker, J. R., G. J. Holt, and S. A. Holt. 1997. Condition of larval and juvenile red drum
(Sciaenops ocellatus) from estuarine nursery habitats. Marine Biology 127:387-394.

Rosebrough, R. W., and J. P. McMurtry. 1993. Energy repletion and lipid metabolism during
compensatory gain in broiler chickens. Growth, Development and Aging 57:73-83.

Rowe, L., and D. Ludwig. 1991. Size and timing of metamorphosis in complex life cycles: time
constraints and variation. Ecology 72:413-427.

Rowe, L., D. Ludwig, and D. Schluter. 1994. Time, condition, and the seasonal decline of avian
clutch size. American Naturalist 143:698-772.

Schew, W. A., and R. E. Ricklefs. 1998. Developmental plasticity. Pages 288-304 in J. M. Starck
and R. E. Ricklefs, editors. Avian growth and development: evolution within the altricial-
precocial spectrum. Oxford University Press, New York, NY.

Schradin, C., and N. Pillay. 2006. Female striped mice (Rhabdomyspumilio) change their home
ranges in response to seasonal variation in food availability. Behavioral Ecology 17:452-
458.

Scriber, J. M., and F. Slansky, Jr. 1981. The nutritional ecology of immature insects. Annual
Review of Entomology 26:183-211.

Seminoff, J. A. 2002. IUCN Red List global status assessment: green turtle (Chelonia mydas).
IUCN/SSC Marine Turtle Specialist Group, Gland, Switzerland.









Shin, H.-C., S. Nicol, and R. A. King. 2003. Nucleic acid content as a potential growth rate
estimator of Antarctic krill; results from field-caught krill from the Indian sector of the
Southern Ocean. Marine and Freshwater Behaviour and Physiology 36:295-305.

Sies, H. 1997. Oxidative stress: oxidants and antioxidants. Experimental Physiology 82:291-295.

Simmons, F. H., and T. J. Bradley. 1997. An analysis of resource allocation in response to
dietary yeast in Drosophila melanogaster. Journal of Insect Physiology 43:779-788.

Skalski, G. T., M. E. Picha, J. F. Gilliam, and R. J. Borski. 2005. Variable intake, compensatory
growth, and increased growth efficiency in fish: models and mechanisms. Ecology
86:1452-1462.

Smith, G. R., and R. E. Ballinger. 1994. Temporal and spatial variation in individual growth in
the spiny lizard, Sceloporusjarrovi. Copeia 1994:1007-1013.

Smith-Gill, S. J. 1983. Developmental plasticity: developmental conversion versus phenotypic
modulation. American Zoologist 23:47-55.

Solomon, S. E., and R. Tippett. 1991. Lipid inclusion in the livers of captive reared marine
turtles. Animal Technology 42:77-81.

Speakman, J. R., and E. Kr6ol. 2005. Limits to sustained energy intake IX: a review of
hypotheses. Journal of Comparative Physiology 175B:375-394.

Spindler, S. R. 2001. Caloric restriction enhances the expression of key metabolic enzymes
associated with protein renewal during aging. Annals of the New York Academy of
Sciences 928:296-304.

Starck, J. M., and K. Beese. 2002. Structural flexibility of the small intestine and liver of garter
snakes in response to feeding and fasting. Journal of Experimental Biology 205:1377-
1388.

Stearns, S. C. 1982. The role of development in the evolution of life histories. Pages 237-258 in
J. T. Bonner, editor. Evolution and development. Springer-Verlag Publishers, New York,
NY.

Stearns, S. C. 1992. The evolution of life histories. Oxford University Press, New York, NY.

Storey, K. B., J. R. Layne, Jr., M. M. Cutwa, T. A. Churchill, and J. M. Storey. 1993. Freezing
survival and metabolism of box turtles, Terrapene carolina. Copeia 1993:628-634.

Tammaru, T., P. Kaitaniemi, and K. Ruohomaki. 1996. Realized fecundity in Epirrita autumnata
(Lepidoptera: Geometridae): relation to body size and consequences to population
dynamics. Oikos 77:407-416.

Therkildsen, M. 2005. Muscle protein degradation in bull calves with compensatory growth.
Livestock Production Science 98:205-218.









Tian, X., and J. G. Qin. 2004. Effects of previous ration restriction on compensatory growth in
barramundi Lates calcarifer. Aquaculture 235:273-283.

Tu, M.-P., and M. Tatar. 2003. Juvenile diet restriction and the aging and reproduction of adult
Drosophila melanogaster. Aging Cell 2:327-333.

Turturro, A. and R. Hart. 1992. Dietary alteration in the rates of cancer and aging. Experimental
Gerontology 27:583-592.

van Noordwijk, A. J., and G. de Jong. 1986. Acquisition and allocation of resources: their
influence on variation in life history tactics. American Naturalist 128:137-142.

Vaupel, J. W., J. R. Carey, and K. Christensen. 2003. It's never too late. Science 301:1679-1681.

Vidal, E. A. G., P. DiMarco, and P. Lee. 2006. Effects of starvation and recovery on the survival,
growth and RNA/DNA ratio in loliginid squid paralarvae. Aquaculture 260:94-105.

Voigt, C. C., G. U. C. Lehmann, R. H. Michener, and M. M. Joachimski. 2006. Nuptial feeding
is reflected in tissue nitrogen isotope ratios of female katydids. Functional Ecology 20:656-
661.

Vrede, T., J. Persson, and G. Aronsen. 2002. The influence of food quality (P:C ratio) on
RNA:DNA ratio and somatic growth rate of Daphnia. Limnology and Oceanography
47:487-494.

Walford, R. L., S. B. Harris, and M. W. Gunion. 1992. The calorically restricted low-fat nutrient-
dense diet in Biosphere 2 significantly lowers blood glucose, total leukocyte count,
cholesterol, and blood pressure in humans. Proceedings of the National Academy of
Sciences USA 89:11533-11537.

Wallace, B. 1992. The search for the gene. Cornell University Press, Ithaca, NY.

Watson, P. J., G. Arnqvist, and R. R. Stallmann. 1998. Sexual conflict and the energetic costs of
mating and mate choice in water striders. American Naturalist 151:46-58.

Wayne, M. L., U. Soundararajan, and L. G. Harshman. 2006. Environmental stress and
reproduction in Drosophila melanogaster: starvation resistance, ovariole numbers and
early age egg production. BMC Evolutionary Biology 6:Article 57.

Weindruch, R., and R. S. Sohal. 1997. Caloric intake and aging. New England Journal of
Medicine 337:986-994.

Weindruch, R., and R. L. Walford. 1988. The retardation of aging and disease by dietary
restriction. Charles C. Thomas, Springfield, IL.

Westerman, M., and G. J. Holt. 1994. RNA:DNA ratio during the critical period and early larval
growth of the red drum Sciaenops ocellatus. Marine Biology 121:1-9.









Wheeler, D. 1996. The role of nourishment in oogenesis. Annual Review of Entomology 41:407-
431.

Wheeler, D. E., I. Tuchinskaya, N. A. Buck, and B. E. Tabashnik. 2000. Hexameric storage
proteins during metamorphosis and egg production in the diamondback moth, Plutella
xylostella (Lepidoptera). Journal of Insect Physiology 46:951-958.

Wilbur, H. M., and J. P. Collins. 1973. Ecological aspects of amphibian metamorphosis. Science
182:1305-1314.

Wilson, P. N., and D. F. Osbourn. 1960. Compensatory growth after undernutrition in mammals
and birds. Biological Reviews of the Cambridge Philosophical Society 35:324-363.

Wilson, A. J., and D. Reale. 2006. Ontogeny of additive and maternal genetic effects: lessons
from domestic mammals. American Naturalist 167:E23-E38.

Wolfner, M. F. 1997. Tokens of love: functions and regulation of Drosophila male accessory
gland products. Insect Biochemistry and Molecular Biology 27:179-192.

Wood, F. E., and G. K. Ebanks. 1984. Blood cytology and hematology of the green sea turtle,
Chelonia mydas. Herpetologica 40:331-336.

Wu, A., X. Sun, F. Wan, and Y. Liu. 2003. Modulations by dietary restriction on antioxidant
enzymes and lipid peroxidation in developing mice. Journal of Applied Physiology
94:947-952.

Zera, A. J., and L. G. Harshman. 2001. The physiology of life history trade-offs in animals.
Annual Review of Ecology and Systematics 32:95-126.









BIOGRAPHICAL SKETCH

Alison M. Roark was born in Norfolk, Virginia on October 26th, 1978. She attended Mills

Godwin High School in Richmond, Virginia, where she was a leader in the marching and concert

bands. After graduating in 1996, Alison attended the University of Virginia (Charlottesville) and

obtained her Bachelor of Science in chemistry with specialization in biochemistry in 2000. She

also fulfilled the requirements for the Distinguished Majors Program in biology with highest

distinction and served as President of the Biology Society. Alison participated in undergraduate

research in two different laboratories at the University of Virginia and also completed two

Research Experience for Undergraduates programs, one at the University of Texas (Austin) and

one at the Long-Term Ecological Research station in Oyster, Virginia. After her second year, she

spent three weeks in San Salvador, The Bahamas, for a class in coral reef ecology and credits

this course with steering her toward a career in academia.

In 2000, Alison joined the Department of Zoology at the University of Florida under the

direction of Karen Bjorndal. In the spring of 2003, she completed her master's bypass. While at

the University of Florida, Alison taught a number of classes, including functional vertebrate

anatomy laboratory and the discussion sections and laboratories for both semesters of

introductory biology. In her final year as a graduate student, she taught her own non-majors

biology course (Cells, Organisms, and Genetics). In the summer of 2007, Alison begins a

postdoctoral position in the laboratory of Lou Guillette through the Howard Hughes Medical

Institute's Group Advantaged Training of Research (G.A.T.O.R.) program.





PAGE 1

PHYSIOLOGICAL AND LIFE-HISTORY RESPONSES TO PATTERNS OF FOOD AVAILABILITY By ALISON M. ROARK A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007 1

PAGE 2

2007 Alison M. Roark 2

PAGE 3

To my family, without whom this work would not have been possible 3

PAGE 4

ACKNOWLEDGMENTS My dissertation would not have been possi ble without the support, assistance, and encouragement of many wonderful people. I woul d first like to thank my advisor and friend, Karen Bjorndal, for guiding me through the adventur es of graduate school. Karens wisdom and generosity have been truly inspirational. I al so thank Alan Bolten for many years of emotional and logistical support. My committee members Karen Bjorndal, Alan Bolten, Lou Guillette, Dan Hahn, David Julian, and John Si vinski have been invaluable to my progress and I cannot thank them enough for their contributions to my professional development. I am indebted to many friends and colleague s who provided assistance with the evolution, implementation, and/or analysis of my work. I am particularly gratef ul to my unbelievable undergraduate assistants, including May Stewar d, Adam Sarnowski, Carie Reynolds, Justin Emerson, Dana-Rachael La Kam, Kelly Johnson, Ka therine Perez, Daphna Yasova, Ann Mazor, and Sasha Strul, all of whom cont ributed significantly to the succe ss of my research. I also thank Richie Moretti for initiating the dialogue that le d to fruitful collaborations between the Cayman Turtle Farm and the Archie Carr Center for Sea Turtle Research, and I thank Ken Prestwich for assistance with insect metabolic rate measurem ents. Debra Murie, Matthias Starck, Tony Zera, Jamie Gillooly, and Scott Pletcher participated in thought-provoking discussions about the conceptual framework and analysis of various components of my rese arch. Lauren Chapman, Greg Pryor, Mike McCoy, and Ben Bolker were i nvaluable statistical consultants who patiently taught me how to use and interpret SPSS and R. Additional statistical consulting was provided by Xueli Liu and Alex Trindade. Harvey Ramirez and Elliott Jacobson assisted with clinical aspects of animal care, and Fran co Giorgi graciously provided photocopies of embryonic staging tables for Carausius morosus 4

PAGE 5

The remarkable graduate students, postdocto ral associates, and faculty members in the Department of Zoology at the University of Florida were a source of constant support, companionship, and friendship. I am especially appreciative for the cont ributions of my many lab-mates, including Lindy Barrow, Sarah Bouc hard, Peter Eliazar, Gabby Hrycyshyn, Kate Moran, Greg Pryor, Kim Reich, Jeff Seminoff, Manjula Tiwari, Hannah Vander Zanden, and Brian Riewald (who is remembered fondly). Na t Seavy, Kenney Krysko, Ryan McCleary, Krista McCoy, and Jada-Simone White also contributed to various aspects of my work. I owe a special debt of gratitude to Thea Edwards and Brandon Moore for allowing me to harvest English ivy from their property at all hours of the day and night and for invaluable horticultural advice. Additional members of the Department of Zool ogy worked tirelessly behind the scenes to ensure that my progression thr ough graduate school was as smoot h as possible. I would like to give special thanks to Karen Pallone, Vitrell Sherif, Pete Ry schkewitsch, Mike Gunter, Frank Davis, Peggy Roberson, Diana Davis, and Cathy Moore for their logis tical support over the years. Ginger Clark, Marty Cohn, Jason Curtis, Rich ard Fethiere, Paul Gulig, Christiaan Leeuwenburgh, Frank Robbins, and Collette St. Mary graciously provided access to their labs and assistance with many of the analyses I pe rformed. Sharon Judge, Colin Selman, and other members of the Leeuwenburgh laboratory helped with the antioxidant assays, and Beverly Deffense assisted with fluorescent spectroscopy. I am grateful beyond words for the generosity and support of the Cayman Turtle Farm. Without them, the work I report in Chapters 24 would have been impossible. I would like to give special thanks and praise to Ken Hydes and Joe Parsons, who were integral members of my Caymanian family and provided me with a wealth of techni cal, logistical, and emotional 5

PAGE 6

support during what amounted to an unbelievably ch allenging and taxing semester. I do not think that I will ever be able to repa y them fully for their many gifts. Conducting animal research requires the ove rsight of a number of permitting agencies, particularly when this research entails worki ng with endangered or restricted species. I am especially grateful to the U.S. Fish and Wildlif e Service, the Institutional Animal Care and Use Committee at the University of Florida, and the U. S. Department of Agriculture. I owe enormous thanks to Sanford Porter, Li m Nong, and Mike Thomas of the USDA in Gainesville for assistance in constructing a quarantine facility and for obtaining the requisite plant pest permit that enabled my work with Indian stick insects. Funding for my dissertation was provided by the National Science Foundation through both a Graduate Research Fellowship and a Doctoral Dissertation Improvement Grant. Additional funds were obtained from the So ciety for Integrative and Comparative Biology, Sigma Xi, the American Society of Ichthyologi sts and Herpetologists, the Brian Riewald Memorial Grant, Sigma Delta Epsilon, two grants from the National Institute on Aging (AG17994 and AG21042 to Christiaan Leeuwe nburgh) and the University of Florida Opportunity Fund. Numerous travel grants were provided by the Un iversity of Florida Graduate Student Council, the Department of Zoology at the University of Florida, the Comparative Nutrition Society, the Sy mposium on Sea Turtle Biology and C onservation, and the Society for Integrative and Comparative Biology. Additiona lly, Benchmark Foliage donated English ivy, and the Exploratorium in San Francisco, CA, donat ed the adult Indian stick insects I used in Chapter 5. Charlie Carlson and Angela Armendariz of the Exploratorium were especially helpful during my visit to their facility. 6

PAGE 7

I owe an enormous debt to my family for th e encouragement, love, and support they have provided throughout my time in gradua te school. It is impossible to put into words the gratitude I feel for my parents and my sister Their faith in me, their emotiona l support, and their belief in the value of education have insp ired and motivated me through thick and thin. I would like to give special thanks to my father for his many hours in the garage helpin g me construct cages, basking platforms, and lighting structures that I used during my feeding trials. I would also like to thank my parents for traveling to the Cayman Islands to help me recover from a devastating hurricane in November of 2001. They were wonderf ul field assistants du ring this time. Roger and Annette Roark deserve many thanks, as well, for their encouragement through the years. They have been unbelievably supportive, and I am incredibly lucky to be their daughter-in-law. Lastly, I thank my husband, Andrew. He has experienced all of my ups and downs along the way, and his patience and belief in me have been a constant source of comfort during the past seven years. 7

PAGE 8

TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ........10 LIST OF FIGURES.......................................................................................................................12 ABSTRACT...................................................................................................................................14 CHAPTER 1 GENERAL INTRODUCTION..............................................................................................16 2 COMPENSATORY GROWTH IN RESPONSE TO A CHANGE IN FOOD AVAILABILITY IN JUVENILE GREEN TURTLES (C helonia mydas)............................23 Introduction................................................................................................................... ..........23 Materials and Methods...........................................................................................................26 Animal Care.....................................................................................................................26 Gut Morphology and Body Composition........................................................................27 Statistical Analyses..........................................................................................................2 9 Results.....................................................................................................................................30 Discussion...............................................................................................................................34 3 BIOCHEMICAL INDICES AS CORRELATES OF RECENT GROWTH IN JUVENILE GREEN TURTLES (C helonia mydas)...............................................................55 Introduction................................................................................................................... ..........55 Materials and Methods...........................................................................................................56 Animal Care.....................................................................................................................56 Tissue Collection.............................................................................................................57 Biochemical Assays.........................................................................................................57 Statistical Analyses..........................................................................................................5 8 Results.....................................................................................................................................60 Discussion...............................................................................................................................62 4 COMPENSATORY GROWTH AND ANTIOXIDANT STATUS IN JUVENILE GREEN TURTLES (C helonia mydas)...................................................................................78 Introduction................................................................................................................... ..........78 Materials and Methods...........................................................................................................80 Animal Care.....................................................................................................................80 Tissue Collection and Homogenization...........................................................................80 Glutathione Peroxida se Activity Assay...........................................................................81 Total AP Assay................................................................................................................82 8

PAGE 9

Statistical Analyses..........................................................................................................8 2 Results.....................................................................................................................................83 Discussion...............................................................................................................................85 5 TIMING OF DIETARY RESTRICTION AL TERS THE EXPRESSION OF LIFEHISTORY TRAITS IN A LONG-LIV ED, PARTHENOGENETIC INSECT......................95 Introduction................................................................................................................... ..........95 Materials and Methods...........................................................................................................99 Animal Husbandry and Feeding Treatments...................................................................99 Physiological and Life-History Response Variables.....................................................100 Statistical Analyses........................................................................................................102 Results...................................................................................................................................104 Discussion.............................................................................................................................109 6 SUMMARY AND CONCLUSIONS...................................................................................136 LIST OF REFERENCES.............................................................................................................142 BIOGRAPHICAL SKETCH.......................................................................................................159 9

PAGE 10

LIST OF TABLES Table page 2-1. Kruskal-Wallis test results for nutr ient content of biweekly food samples...........................41 2-2. Repeated measures analyses of variance fo r weekly averages of daily intake and daily mass-specific intake........................................................................................................... 42 2-3. Repeated measures analyses of vari ance for weekly body mass, straight carapace length, and condition index................................................................................................43 2-4. Repeated measures analyses of variance for weekly specific growth rates (SGR) for body mass (bm) and carapace length (cl)..........................................................................44 2-5. Repeated measures analyses of varian ce for food conversion efficiencies (FCE) for body mass (bm) and carapace length (cl)..........................................................................45 2-6. Omnibus F 2, and p-values for analyses of variance of dissection data collected at weeks 5 and 12...................................................................................................................46 2-7. Organ masses (mean standard error) from turtles dissected at 0, 5, and 12 weeks reported as indices............................................................................................................ ..47 2-8. Body composition (mean standard error) of turtles dissected at 0, 5, and 12 weeks reported as percent of dry matter (% DM) and percent of organic matter (% OM)..........48 3-1. Omnibus F 2, and p-values for comparisons of means among treatment groups for the various morphometric and bioc hemical indices measured................................................69 3-2. Spearmans rank correlations ( ) for morphometric (a) a nd biochemical indices for liver (b), heart (c), and blood (d)........................................................................................70 3-3. Growth equation parameters for juvenile Chelonia mydas as determined by least squares linear regression....................................................................................................71 3-4. Growth equation parameters for juvenile Chelonia mydas as determined by stepwise multiple linear regression...................................................................................................72 3-5. Coefficients of variation (C.V.) for RNA, DNA, and protein concentrations of Chelonia mydas tissues......................................................................................................73 4-1. Omnibus F 2, and p-values for comparisons of mean s among treatment groups at five weeks (t5) and twelve weeks (t12).......................................................................................89 4-2. Total protein concentrations of Chelonia mydas muscle and liver homogenates as determined by Bradford assay expressed rela tive to wet mass of homogenized tissue.....90 4-3. Glutathione peroxidase (GPX) specific activity in Chelonia mydas muscle homogenate....91 10

PAGE 11

4-4. Coefficients of variation (CV, %) for protein concentration, gl utathione peroxidase (GPX) activity, and antioxidant potential (AP) assays......................................................92 5-1. Cumulative intake expressed as the tota l dry mass consumed during each life-history stage in each of five treatment groups.............................................................................119 5-2. Omnibus F 2, and p-values for comparisons of body mass and age among five treatment groups within each life-history stage...............................................................120 5-3. Relative mass ( standard error) at the adult molt of insects in five treatment groups, calculated as the ratio of actual to predicted body mass as determined by allometric analysis.............................................................................................................................121 5-4. Adjusted mean fecundity ( standard erro r) of insects in each of five treatment groups estimated using body mass at first oviposition as a covariate.........................................122 5-5. Total number of ovarioles (mean sta ndard error) in insects from each of five treatment groups upon post-mortem dissection...............................................................123 5-6. Parameters for equations predicting realiz ed fecundity and initial oviposition rate as determined by stepwise mu ltiple linear regression..........................................................124 6-1. Summary of traits measured for Chelonia mydas (top half of table) and Carausius morosus (bottom half of table) mainta ined on different schedules of ad libitum (ad lib.) and restricted (rest.) intake.......................................................................................141 11

PAGE 12

LIST OF FIGURES Figure page 1-1. Hypothetical plot of size versus time fo r juvenile animals from the same cohort................22 2-1. Average mass-specific daily intake (mean standard error) during each week of the feeding trial.................................................................................................................. ......49 2-2. Body mass (a) and straight carapace length (b) (mean standard error) at the midpoint of each week......................................................................................................................50 2-3. Condition index (mean sta ndard error) in each week.........................................................51 2-4. Specific growth rate (mean standard error) for body mass (BM, a) and straight carapace length (CL, b) during each week.........................................................................52 2-5. Food conversion efficiency (FCE, mean standard error) for body mass (BM, a) and straight carapace length (CL, b) during each week............................................................53 2-6. Daily water temperatures (mean standa rd deviation) throughout the feeding trial............54 3-1. Morphometric indices and gr owth rates for turtles in each of three treatment groups..........74 3-2. Nucleic acid indices for turtles in each of three treatment groups........................................75 3-3. Liver protein and proteinbased indices for turtles in each of three treatment groups..........77 4-1. Body mass of turtles at five weeks (t5) and twelve weeks (t12), when tissues were sampled..............................................................................................................................93 4-2. Glutathione peroxidase (GPX) specific activity and antioxidant potential (AP; calculated as nmoles of copper reducing equivalents, CRE) in Chelonia mydas liver homogenate at five weeks (t5) and twelve weeks (t12).......................................................94 5-1. Experimental design for Carausius morosus feeding trial..................................................125 5-2. Mass-specific intake (g dry mass/g*day) co nsumed by insects in each of five treatment groups on each day of the study.......................................................................................126 5-3. Age and size at each life-history transition for insects in each of five treatment groups....127 5-4. Duration of (a) and specific growth rate during (b) each life-history stage for insects in each of five treatment groups...........................................................................................128 5-5. Event history diagra m depicting periods of ad libitum intake, restricted intake, and reproductive activity for individu al stick insects maintained on five diet treatments.....129 12

PAGE 13

5-6. Kaplan-Meier survivorship curves for the entire lifespan (a) and for the adult lifespan (b) of insects maintained on five diet treatments.............................................................131 5-7. Cumulative fecundity of insects in each of five treatment groups......................................133 5-8. Average egg mass (mean standard error) for stick insects maintained on five diet treatments..................................................................................................................... ....134 5-9. Relationships between realized fecundity and cumulative intake during the reproductive lifespan (a) and the duration of adult lifespan (b) for all insects that laid eggs..................................................................................................................................135 13

PAGE 14

Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PHYSIOLOGICAL AND LIFE-HISTORY RESPONSES TO PATTERNS OF FOOD AVAILABILITY By Alison M. Roark August 2007 Chair: Karen A. Bjorndal Major: Zoology Green turtles (Chelonia mydas) experience nutritional stochasticity as oceanic-stage juveniles and should therefore be capable of co mpensatory growth (CG) following periods of nutritional stress. The purpose of the first phase of my research wa s to test for the occurrence of CG in green turtles and determine its mechanis m(s). Food-restricted turtles (R) grew more slowly, differed in cell size and body compositi on, and had proportionally smaller digestive organs than turtles feeding ad libitum (AL). After food conditions improved, previously food-restricted turtles (R-AL) demonstrated CG This growth pattern was elicited by enhanced food conversion efficiency rather than hyperpha gia. The period of growth compensation may have ended when R-AL turtles attained a body com position similar to that of AL turtles. These findings indicate that growth rate, morphology, and body composition of juvenile green turtles are plastic in response to diet and that individuals can compensa te for environmental variability to capitalize when conditions improve. However, CG was associated with altered antioxidant function. Activity of glutathione peroxidase and total antioxidant potential per liver cell were greater in AL turtles than in R and R-AL turt les, respectively, at the conclusion of the study. Therefore, impaired antioxidant capacity may be a cost of rapid growth in this species. 14

PAGE 15

To elucidate long-term responses to food av ailability, I tested the effects of food restriction (FR) imposed during several life stages in Indian stick insects ( Carausius morosus ). Intake pattern affected age and size at each lif e-history transition, with size decreasing and age increasing in response to FR. Earl y-onset FR increased lifespan, but this increase was negated by detrimental effects on reproductive output. Lateonset FR negatively affected both longevity and reproductive output. Carausius morosus appears to allocate incomi ng adult-derived nutrients to egg provisioning and stored nutrients to adult somatic maintenance and survival. Thus, I found no evidence for a trade-off between fecundity an d longevity because nutrients allocated to reproduction and maintenance do not appear to be derived from a common resource pool. These results demonstrate that fluctuations in food avai lability can significantly alter the expression of life-history traits and that th e magnitude of these effects de pends on the developmental stage during which food availability changes. 15

PAGE 16

CHAPTER 1 GENERAL INTRODUCTION Nearly every biological process depends to some extent on nutriti on. Ingestion, digestion, and nutrient absorption determine an animals capacity for maintenance, growth, reproduction, and survival. An individuals success thus depends on its capacity to extract and utilize nutrients from its food. However, many organisms experi ence variation in food availability throughout their lifetimes (Pltz et al 1991, Boggs and Ross 1993, Carey et al 2002a), and evolutionary adaptation to these fluctuations is critical to sp ecies survival. Fluctuations in food availability can result from climatic or seasonal dynamics of food sources and spatial or temporal heterogeneity of nutrient distribution (Dagg 1977, Smith and Ballinger 1994, Forman 1995, Arnekleiv et al 2006, Schradin and Pillay 2006). Periods of feast and famine result in correspondingly fast and slow periods of growth, development, and reproduction (Ballinger 1977, Calbet and Alcaraz 1997, Kitaysky 1999, Morey and Reznick 2000). Such developmental plasti city is common to many species (Stearns 1982, Smith-Gill 1983, Schew and Ricklefs 1998), especially those in stochastic environments (Lochmiller et al 2000). One adaptation to cycles of low and high nutri ent availability is compensatory growth (CG). This phenomenon manifests itself as a period of growth faster than that demonstrated by consistently well nourished conspecifics of the same age (Fig. 1-1) and can result in comparable body sizes for individuals with very di fferent dietary histories (Broekhuizen et al 1994, Metcalfe and Monaghan 2001). Compensatory growth pr esumably allows organisms to mitigate size-specific mortality risks and developmental time constr aints induced by pe riodic nutritional stress (Arendt 1997). This growth pattern is typically effected by hyperphagia (increased feed intake), improved food conversion efficiency (defin ed as growth per unit of food consumed), or both (Miglavs and Jobling 1989, Ali et al 2003). Compensatory growth has been documented in 16

PAGE 17

many species (Wilson and Osbourn 1960, Ali et al 2003, Jespersen and Toft 2003), particularly those that commonly experience e nvironmental stochasticity. Howeve r, compensatory growth is certainly not a universal occurr ence, especially in non-teleost species (e.g., Altwegg and Reyer 2003, Brz k and Konarzewski 2004). For my dissertation, I studied the physiological a nd life-history responses of animals to changes in food availability. The green turtle ( Chelonia mydas ) is an excellent model for pursuing questions about fluctuating food availab ility because of the nutritional stochasticity they are thought to encounter as juveniles. This stochasticity should select for growth patterns that allow individuals to maximize their grow th rates when conditions are good. Green turtles consume a largely carnivorous diet duri ng the juvenile oceanic stage (Reich et al. in review). Food availability during this time is thought to be spatially heterogeneous. At a size of 20-35 cm (carapace length), juvenile green turtles undergo an ontogenetic shift in habitat use and diet and recruit to neritic habitats, where they con tinue development while consuming a primarily herbivorous diet (Bjorndal 1997). Niche shifts are excellent oppor tunities for enhanced growth because such shifts often correspond to improved nutritional conditions (Ali et al. 2003). Although growth dynamics of oceanic-stage juveniles are unknown, post-recruitment growth rates vary temporally in response to oceanogr aphic conditions (Limpus and Chaloupka 1997) and population density (Bjorndal et al 2000). This variation in grow th rates may have substantial life-history effects, because body size is correlat ed with juvenile survivorship (Chaloupka and Limpus 1998) and clutch size (Broderick et al 2003) in this species. B ecause of the potential for nutritional stochasticity and the poten tial effects of body size on fitness in C. mydas I predicted that green turtle juveniles should be capable of CG. Juve nile loggerheads with dietary 17

PAGE 18

preferences similar to thos e of green turtles undergo CG in the wild (Bjorndal et al 2003), although the mechanism for this growth pattern is unknown. The purpose of the first phase of my resear ch was to manipulate food availability in juvenile green turtles under controlled conditions to test whether previously food-restricted turtles can undergo CG and to evaluate hyperpha gia and enhanced food conversion efficiency as possible mechanisms for this growth pattern. The results of Chapter 2 indicated that previously food-restricted green turtle juve niles are indeed capable of transient growth compensation after a return to ad libitum feeding. This finding suggests that growth rates of turtles under conditions of continuously high food availability are optimal rather than maximal (Metcalfe and Monaghan 2001) and that trade-offs between growth and fitness probably exist. Rapid growth in many species may be associat ed with a variety of costs, many of which may not be paid until late in life (Einum and Fleming 2000 and references therein, Metcalfe and Monaghan 2001, Altmann and Alberts 2005, Nagy a nd Holmes 2005). To pursue this hypothesis, I examined antioxidant function of tissues from green turtles that had undergone CG. Antioxidants prevent free-radical induced oxidati ve damage to nucleic acids and proteins by converting reactive oxygen species (ROS) into less noxious compounds (Ji and Leichtweis 1997, Gredilla and Barja 2005). Caloric restricti on depresses the rate of production of ROS (Lpez-Torres et al 2002, Barja 2004) and attenuates the accrual of irreparable damage to cellular macromolecules (Hyun et al 2006). As a result, dietary rest riction slows aging and thus extends lifespan rela tive to continuous ad libitum feeding in a diversity of species (Weindruch and Walford 1988, Austad 1989, Mair et al 2003, Vaupel et al 2003, Hatle et al 2006b). Because ad libitum feeding is typically accompanied by acc elerated aging, it is possible that periods of rapid growth induced by high food ava ilability after a period of nutritional stress may 18

PAGE 19

be costly in terms of survival I hypothesized that the shortterm benefits of CG may be countered by the rapid accumulation of oxidativ e damage, thereby imposing a cost to fast growth. My results confirmed that turtles with a history of CG suffe red decreased cellular antioxidant potential relative to continuously ad libitum -fed turtles. The diminished capacity for antioxidant defens e exhibited by fast-growing turtles piqued my interest in the more long-term consequences of changes in food availability on life-history traits and trade-offs. Despite the tendency for caloric restriction to enhance longevity in the laboratory, nutritional stress experienced ear ly in life can have profound life-history consequences (Metcalfe and Monaghan 2001). For example, growth rate, dominance status, and age at first parturition in spotted hyenas are strongly correlat ed with food availability during development (Hofer and East 2003). This silver-spoon effect (Grafen 1988, Madsen and Shine 2000) indicates that early nutri tional conditions can have longlasting effects. Resource limitation experienced later in life can also in fluence life history by decreasing reproductive output (Boggs and Ross 1993, Wheeler 1996, Olsson and Shine 1997, Carey et al 2002b, Nagy and Holmes 2005). On the other hand, food restri ction has been shown to extend an animals reproductive lifespan (Hart and Turturro 1998), th ereby partially mitigating the decline in reproductive rate caused by food scarcity. Most of the work cited above, however, provide s incomplete information about the effects of nutrition on fitness and life history because in take is not typically quantified throughout the entire lifespan in studies of this kind (Zera and Harshman 2001), particularly for long-lived species. As most animals experience fluctuations in food availability at some point in their lifetimes, a more complete understanding of the effects of diet on f itness requires intake 19

PAGE 20

manipulation and quantification under controll ed conditions during bot h juvenile and adult stages. Furthermore, to evaluate the costs of reproduc tion and the effects of dietary restriction on fitness in sexual species, females must be allo wed to mate. However, mating is often prevented (particularly in rodent studies) because experimental animals are maintained individually (Vaupel et al 2003). In those studies where mating is permitted, co-housing individuals can complicate the quantification of individual intake and is known to influence longevity and fecundity due to the eff ects of crowding (Joshi et al 1998). To avoid such problems, studies of food restriction and reproductive output are often undertaken us ing virgin or hermaphroditic females of invertebrate species. This approach suffers from lim itations, including the fact that mating enhances fecundity (Chiang and Hodson 1950, Foster and Howard 1999, Chong and Oetting 2006) and that self-fertilization constrains reproducti ve output because of limits on self-sperm production in protandrous hermaphroditic species like Caenorhabditis elegans (Cutter 2004, Partridge et al 2005). For these reasons, the choice of an appropriate model organism for the final phase of my dissertation re search was of paramount importance. Although the first phase of my research focuse d on green turtles, this species is not a suitable animal model for investigating correlati ons between intake and life-history traits. To evaluate the effects of food av ailability on development, longev ity, and reproductive output, I adopted a novel approach to life-history experi mentation by using a part henogenetic species as my animal model. Using a parthenogen obviated the need for mating, and therefore allowed females to be housed individually, while still permitting natural reproductive processes. Given the paucity of parthenogenetic vertebrate species with a reas onable (i.e., less than two-year) lifespan that are amenable to laboratory culture, I chose to work with an insect species for the 20

PAGE 21

final phase of my dissertation rese arch. The Indian stick insect ( Carausius morosus (Br.)) (Phasmatodea, Lonchodinae) is a relatively longlived species that reproduces via apomictic parthenogenesis (Pijnacker 1966). This species is hemimetabolous and phytophagous, allowing for life-long, quantitative dietary manipulations using the same food source throughout the entire lifespan. My purpose was to determine the effects of differences in resource availability at several developmental stages on life-history traits that have substantial influences on population structure and dynamics, such as age and size at each developmental transition, longevity, and fecundity. The overall goal of my dissertation research was to elucidate the physiological and life-history effects of va riation in food availability by conduc ting feeding trials in a controlled laboratory setting. In Chapter 2, I explore the capacity for and mechanisms of CG in green turtles and evaluate the effects of intake and grow th rates on body composition and digestive system morphology. In Chapter 3, I examine the effects of diet on cell size and protein synthesis capacity and assess the utility of a number of morphologi cal and biochemical indices as potential predictors of recent growth in green turtles. In Chapter 4, I compare the antioxidant potentials of green turtles with known growth trajectories to establish a putative cost of rapid growth. In Chapter 5, I determine the life-history consequen ces of changes in food availability imposed either during development or after the attainment of sexual maturity in Indian stick insects. For this final chapter, I pursued questions about the effects of vari ation in intake on developmental transitions, longevity, and fecundity. 21

PAGE 22

Time Size Figure 1-1. Hypothetical pl ot of size versus time for juvenile animals from the same cohort. The solid line represents individuals feeding ad libitum and the dashed line represents individuals feeding at a restricted rate unt il the time indicate d by the arrow, after which food was offered ad libitum The rapid growth demonstrated by the previously-restricted individuals after the switch to ad libitum feeding represents compensatory growth. 22

PAGE 23

CHAPTER 2 COMPENSATORY GROWTH IN RESPONSE TO A CHANGE IN FOOD AVAILABILITY IN JUVENILE GREEN TURTLES (C helonia mydas) Introduction Many organisms experience varying levels of nutrient availabili ty throughout their lifetimes as a result of spatia l or temporal heterogeneity of food distribution. These periods of high and low food availability often lead to correspondingly fast and slow rates of growth, development, and/or reproduction (Ballinge r 1977, Calbet and Alcaraz 1997, Kitaysky 1999, Morey and Reznick 2000). Given these life-his tory consequences, fluctuations in food availability should select for adaptations that allow a previously f ood-limited individual to maximize its ability to capitalize on bette r conditions when they are encountered. One common response to alternating periods of high and low food availability is compensatory growth (CG). This phenomenon manife sts itself as a period of accelerated growth during improved food conditions following a peri od of nutritional depriv ation severe enough to restrict growth rates (Wilson and Osbourn 1960, Re id and White 1977). As a result, growth trajectories tend to converge, thereby minimizi ng the variance in body size among individuals of a cohort (Atchley 1984). Compensatory growth presumably allows organisms to mitigate the negative effects of slow growth on survival, deve lopment, and reproductive output. This growth pattern has been documented in invertebrates (Jespersen and Toft 2003, Dmitriew and Rowe 2005), fish (Skalski et al 2005), reptiles (Bjorndal et al 2003, Caley and Schwarzkopf 2004), birds (Kunz and Ekman 2000), and mammals (Lochmiller et al 2000). However, compensatory growth is certainly not a univers al occurrence, especially in non-teleost species (e.g., Altwegg and Reyer 2003, Brz k and Konarzewski 2004). Although the degree of growth compensati on can vary depending on the species in question, the developmental stage of the organism at the times of restriction and improved food 23

PAGE 24

availability, and the length a nd severity of the period of food restriction (Wilson and Osbourn 1960, Ali et al 2003), the mechanisms for compensatory growth (when it is observed) are relatively conserved. Hyperphagi a, or increased food intake, is the most common proximate cause of CG in a variety of animals (Ali et al 2003). During hyperphagia, the rate of lipid accumulation may direct the duration of the compensatory response, with intake and growth rates returning to normal once adipose stores ha ve been restored (Job ling and Johansen 1999, Johansen et al 2001). Improvements in food conversion effi ciency (FCE, defined as growth per unit of food consumed) can al so allow for CG (Patterson et al 1995, Boujard et al 2000). Reductions in the mass of energetically expensive viscera (especially gut and/or liver; Hornick et al 2000) during food restriction have been demons trated in fish (Gaylord and Gatlin 2000), mammals (Weindruch and Sohal 1997), and birds (Hume and Biebach 1996, Karasov and Pinshow 1998). If decreased organ size persists for a period of time after a return to ad libitum feeding, the resulting combination of lowered me tabolic expenditure and high intake might allow more nutrients to be allocated to growth. In this way, FCE would remain high, thus facilitating CG during this period. In addition to affecting visceral organ size, va riation in intake and growth rates can cause a variety of physiological changes. Body composition is one of the most plastic characteristics of organisms undergoing food restriction and subsequent realimentation. Sequential mobilization of reserves typically occurs when an imals are food-limited, with lipid stores depleted preferentially compared to protein stores (e.g., Cherel et al. 1993, Tian and Qin 2004). During realimentation and CG, differential accretion of lipid and/or protein also occurs Typically, the early phases of compensatory growth are characterized by lean tissue deposition while later phases are characterized by fat deposition. This differential accretion of lean tissue during the early stages 24

PAGE 25

of realimentation may provide a mechanism for accelerated growth, as lean tissue deposition requires less energy than fat deposition (Hornick et al 2000). In this study, I examined the capacity for and mechanisms of CG in juvenile green turtles ( Chelonia mydas ). The green turtle leads an oceanic exis tence for the first se veral years of life (Carr 1987) and consumes a largely carni vorous diet during this time (Reich et al. in review). Intake of juvenile turtles in this stage most likel y varies stochastically due to heterogeneous prey distribution. At a size of a pproximately 20-25 cm carapace length (for Atlantic C. mydas ) or 35 cm carapace length (for Pacific C. mydas ), green turtles undergo an ontogenetic shift in habitat use and diet by recruiting to ne ritic habitats, where they consum e an herbivorous diet consisting of various species of algae and seagrass (Bjorndal 1997). Although growth dynamics of juveniles during the oceanic stage are unknown, post-recr uitment growth rates are known to vary temporally as a result of va riation in oceanographic conditions (Limpus and Chaloupka 1997) and population density (Bjorndal et al 2000). This variation in juve nile growth rates may have substantial effects on fitness, as body size is correlated with ju venile survivorship (Chaloupka and Limpus 1998) and clutch size (Broderick et al 2003) in C. mydas Because of the potential for stochasticity of the nutritional environment during the oceanic and neritic stages of development and the effect of body size on survival and reproductive output in C. mydas I predicted that green turtle juvenile s should be capable of CG. Loggerhead sea turtles, which are also largely carnivorous as young juveniles (Bjorndal 1997), have been shown to undergo CG in th e wild (Bjorndal et al 2003), although the proximate explanation for this growth pattern is not known. By manipulating food availability under controlled conditions, I tested whether previously food-restricted green turtles can undergo CG and evaluated hyperphagia and enhanced FCE as potential mechanisms for this growth pattern. 25

PAGE 26

Materials and Methods Animal Care All animal care components of this study were performed at the Cayman Turtle Farm in Grand Cayman, British West Indies, in accordan ce with the guidelines of the Institutional Animal Care and Use Committee at the University of Florida (permit #Z061). Chelonia mydas hatchlings ( n = 115) were housed individually in sea wate r in 68-liter tanks arranged within a large outdoor concrete enclosure. Tanks were sy stematically arranged within the enclosure to minimize position effects. Fresh sea water was con tinuously circulated within the enclosure (at a depth of approximately 20-25 cm) to dampen the daily cycle of heating an d cooling within each of the tanks. Water temperature was monitore d daily using five min/max thermometers distributed throughout the array of tanks. All turtles were maintained on an ad libitum diet for seven days pr ior to the beginning of the study to estab lish average daily ad libitum intake. Turtles were then randomly assigned to one of three treatment groups: ad libitum (AL), restricted (R ), and restrictedad libitum (R-AL). Turtles in the AL group were offered food ad libitum for twelve weeks. Turtles in the R group were offered approximately 50% of average initial mass-specific ad libitum daily intake for twelve weeks. Because AL turtles increased their mass-specific intake after week 0, the actual amount of food consumed by restri cted turtles amounted to an average of 27% of the daily mass-specific intake of AL turtles. This ration wa s sufficient to maintain restricted turtles on a positive growth trajectory throughout the study. Turtles in the R-AL group were offered the restricted diet for five weeks and were then offered food ad libitum for the remaining seven weeks. Turtles were provided 2.6-mm tu rtle pellets (Melick Aquafeed, Catawissa, PA) twice daily and were allowed to feed for a total of seven to ten hours each day. Pellets remaining in each 26

PAGE 27

tank were counted every afternoon, and approximate intake was calculated based on the average mass per pellet (determined weekly), the known mass of pellets offered, and the number of pellets remaining. Intake was estimated in this way for each turtle on each of six days a week (weather permitting), with the seventh day rese rved for tank cleaning. The water in each tank was replaced daily. A subset of pellets was count ed and weighed each week to determine average pellet mass. Five food samples were weighed and dried every two weeks for nutrient analyses. Body mass (BM, to the nearest 0.1 g) and minimum straight carapace length (CL, to the nearest 0.05 mm) of each turtle we re measured weekly. Daily intake measurements and weekly body size measurements were used to calculate av erage daily intake (g/day), mass-specific daily intake (g/g*day), condition index (CI, g/cm3) (Ricker 1975), food conversion efficiency (FCE, g/g and mm/g), and specific growth rate (SGR, %/ day) for each turtle for each week of the study. FCE and SGR were calculated for both BM and CL using the following formulas: FCE = (sizet+1 sizet)/(average daily intake 7) SGR = 100*[ln(sizet+1)-ln(sizet)]/7 where sizet is BM or CL in one week and sizet+1 is BM or CL in the following week. FCE and SGR were calculated for both BM and CL because gut fill accounted for up to 15.4% of total wet BM, and changes in CL are not affected by the mass of gut contents. Furthermore, straight CL is considered to be the most reliable measure of growth in green turtles (Balazs 1995). Gut Morphology and Body Composition Turtles were sacrificed prio r to, during, and after the st udy for an analysis of gut morphology and body composition. Ten randomly chosen turtles (t0 AL) were euthanized prior to the initiation of the study, at which time all tu rtles had been feeding ad libitum for at least one week. At the conclusion of the fifth week of th e experiment (immediately prior to switching R-AL turtles to an ad libitum diet), ten AL (t5 AL), five R, and five R-AL turtles were 27

PAGE 28

euthanized. The R-AL turtles had, until the end of week five, been maintained on the restricted feeding treatment. Data for the five R and five R-AL turtles were theref ore pooled into one group (t5 R). Ten AL turtles (t12 AL), ten R turtles (t12 R), and ten R-AL turtles (t12 R-AL) were euthanized at the conclusion of the twelve-week trial. Turtles were wei ghed to the nearest 0.1 g and then euthanized using an intramuscular over dose injection of ketamine (Ketaset, 100 mg/kg body mass). The liver and digestive tract (from the esophageal -gastric junction to th e termination of the hindgut anterior to the cloaca) of each turtle were removed. The full gut was measured to the nearest 0.05 mm. Measurements included straight stomach length (SSL) and total intestine length (TIL). Small intestine and large intestine lengths could not be determined because the distinction between midgut and hindgut was not easily discer nible. Gut contents from each turtle were gently removed from the excised gut using forceps. Wet masses of gut contents, liver, empty stomach, and empty intestine were determined for each turtle. Organ mass and length indices were calculated using the following formulas: Mass Index = 100*MOr/(BM MGC) Length Index = LOr/CL where MOr is wet mass of each organ (liver, stomach, or intestine), MGC is wet mass of gut contents, and LOr is length of stomach or intestine. All tissues and carcasses were dried at 60 C for a minimum of seven days. Dried body tissues (liver, stomach, intestines, and carcass) were recombined for each turtle ( n = 10 t0, 20 t5, and 30 t12 turtles) and ground for nutrient analyses. Each turtle was ground in a mill (C.W. Brabender Instruments, Inc., South Hackensack, NJ ) with dry ice. Dried food samples (collected every two weeks during the study) were also ground in the mill (without dry ice). Dry matter (DM) content of each turtle and each food sample was determined by drying subsamples at 28

PAGE 29

105 C for 16 hours, and organic matter (OM) co ntent was determined by combustion at 500 C for three hours (AOAC 1960). Energy content of each turtle and food sample was determined by bomb calorimetry (Parr 1960; Parr Instrument Co., Moline, IL). Lipid content was determined by ether extraction using a soxhlet apparatus (AOAC 1984). Nitrogen content was determined using a modified Kjeldahl procedure. Samples were digested for at least four hours at 375 C using a modification of the aluminum block digestion procedure of Gallaher et al (1975). Nitrogen in the digestate was determined by semiautomated colorimetry using a Technicon Autoanalyzer (Hambleton 1977; Pulse Instrumentation, Ltd., Sa skatoon, SK, Canada). All nutrient analyses were performed in duplicate unless relative e rror exceeded 2.0%, in which case additional analyses were performed. The ratio of lipid to lean contents was calcula ted by dividing DM lipid content by DM protein content (% nitrogen 6.25; Hambleton 1977). Statistical Analyses Data for food samples collected at two-week intervals were analyzed using Kruskal-Wallis tests. Weekly data were analyzed using repeated measures ANOVA with Tukeys Honestly Significant Difference (HSD) post hoc tests. Data for each week of the study and for midpoint and endpoint samples were also tested for significance using one-way ANOVA with Tukeys HSD post hoc tests. Bonferroni corrections were not used to account for multiple comparisons among weeks with one-way ANOVA because of the risk of inflated Ty pe II error (Perneger 1998). Data for t5 and t12 samples were tested for normality (using Shapiro-Wilk test) and homogeneity of variances (using Levenes test) prio r to parametric analysis. If both tests yielded a significant result (p < 0.05), data were transformed using a natural log, reciprocal, square root, square, reciprocal square, or re ciprocal square root transfor mation. If transformation did not improve normality or homoscedasticity, or if no post hoc test could be performed (e.g., for t5 samples), data were tested for statistical significance using a Kruskal-Wallis test and pairwise 29

PAGE 30

Mann-Whitney U tests with set at 0.017 (for t12 samples) to account for multiple comparisons. Analysis of covariance (ANCOVA) could not be us ed to evaluate midpoint and endpoint data for organ sizes because the assumption that covariat e values have similar distributions and ranges for all treatment groups (Quinn and Keough 2002) was violated. All data were analyzed using SPSS for Windows (Release 11.0.0). Only turtles that survived, displayed no external signs of illness, and continued to gain mass until the time of tissue sampling were included in the analyses. Data are expressed as means standard errors with alpha set at 0.05 unless otherwise noted. Results OM, energy, nitrogen, and lipid contents of food samples ( n = 7, collected at biweekly intervals) were consistent throughout the experi ment (Table 2-1). Although differences in energy content of pellets among weeks approached sign ificance, the relative difference between the highest and lowest energy values measured was only 2.87% (for DM) and 2.73% (for OM). R and R-AL turtles had comparable values for all repeated measures data collected during weeks 0 through 5. Data for these two groups were an alyzed and are presented separately for this time period to demonstrate that there were no di fferences between groups prior to the switch to an ad libitum diet for R-AL turtles. Intake in week 0 (during wh ich all turtles were feeding ad libitum ) was similar for all three treatment groups (ANOVA, F2,112 = 0.946, p = 0.392). Repeated measures ANOVA of intake during weeks 1 through 5 ( n = 37 AL, 39 R-AL, and 39 R) and during weeks 6 through 12 ( n = 17 AL, 29 R-AL, and 29 R) revealed significant time and treatment effects as well as significant interactions between time and treatment group (Table 2-2). The pattern of AL intake was significantly different from the patterns of R and R-AL intake during weeks 1 through 5 (Tukeys HSD post hoc test, p < 0.0001 for both comparisons), and intake patterns of all three 30

PAGE 31

treatment groups were significantly different during weeks 6 through 12 (Tukeys HSD post hoc test, p < 0.0001 for all comparisons). Because body si ze of turtles in each group differed after week 0, intake was corrected for BM and re-analyzed as mass-specific intake. The magnitude but not the overall pattern of mass-specific intake of R turtles differed from that of AL turtles for the duration of the expe riment, and the magnitude and overall pattern of mass-specific intake of R-AL turtles was comparab le to that of R tur tles during weeks 1-5 and comparable to that of AL turtles during weeks 6-12. Repeated measures ANOVA of massspecific intake (Table 2-2) reve aled no significant linear inter action between time and treatment group ( p > 0.05 for time treatment in teractions using tests of w ithin-subjects contrasts in repeated measures ANOVA) during either week s 1 through 5 or weeks 6 through 12. However, there was a significant linear eff ect of time on mass-specific inta ke during weeks 1 through 5 and during weeks 6 through 12, as demonstrated by the d ecrease in mass-specific intake for turtles in all treatment groups during thes e two time intervals (Fig. 2-1) The pattern of mass-specific intake of AL turtles was signifi cantly different from those of R and R-AL turtles during weeks 1 through 5 (Tukeys HSD post hoc test, p < 0.0001 for both comparisons ), and the pattern of mass-specific intake of R turtles was significantly different from those of AL and R-AL turtles during weeks 6 through 12 (T ukeys HSD post hoc test, p < 0.0001 for both comparisons). Body size (BM and CL) in week 0 (dur ing which all turt les were feeding ad libitum ) was similar for all three treatment groups (ANOVA, F2,112 = 0.992, p = 0.374 for BM; F2,112 = 1.109, p = 0.333 for CL) Repeated measures ANOVA of body size during weeks 0 through 5 and during weeks 6 through 12 revealed significant time and treatment effects as well as significant interactions between time and treatment group (T able 2-3). The pattern of body size (both BM and CL) of AL turtles was signifi cantly different from those of R and R-AL turtles during weeks 31

PAGE 32

0 through 5 (Tukeys HSD post hoc tests, p < 0.0001 for both comparisons for BM and CL), and the patterns of body size (both BM and CL) of all three treatment groups were significantly different during weeks 6 through 12 (Tukeys HSD post hoc tests, p < 0.0001 for all comparisons for BM and CL). Within individual weeks, body size of AL turtles was significantly greater than those of R and R-AL turtles in weeks 1 through 6 (for BM) and in week s 1 through 7 (for CL), and body size of all three groups differed signif icantly in weeks 7 through 12 (for BM) and in weeks 8 through 12 (for CL) (Fig. 2-2). There were significant effects of time and treatment on condition index (CI) as well as a significant time by treatment interaction duri ng weeks 0 through 5. There was no significant overall effect of time on CI during weeks 6 through 12, although there was a significant interaction between time and treatment group (Table 2-3). The pattern of CI in AL turtles was significantly different from thos e of R and R-AL turtles during weeks 0 through 5 (Tukeys HSD post hoc test, p < 0.0001 for both comparisons), and the pattern of CI in R turtles was significantly different from those of AL and R-AL turtles during weeks 6 through 12 (Tukeys HSD post hoc test, p < 0.0001 for both comparisons). The difference between AL and R-AL turtles approached significance (Tukeys HSD post hoc test, p = 0.076). Within individual weeks, CI of AL turtles was significantly greater than CI of R and R-AL turtles during weeks 1 through 5, and CI of R-AL turtles was intermediate between those of R and AL turtles in week 6. In weeks 7 through 12, CI of R turtles was significan tly lower than CI of R and R-AL turtles, and CI of R-AL turtles was consistently but not signif icantly lower than CI of AL turtles (Fig. 2-3). Specific growth rates (SGR) for BM and CL also differed among treatment groups for all weeks of the experiment. Repeated measures ANOVA of SGR revealed significant time and treatment effects and interac tions between time and treatmen t group during weeks 1 through 5 32

PAGE 33

and during weeks 6 through 12 (Table 2-4). The SGR patterns (for both BM and CL) of AL turtles were significantly differe nt from those of R and R-AL turtles during weeks 1 through 5 (Tukeys HSD post hoc tests, p < 0.0001 for both comparisons for BM and CL), and the SGR patterns (for both BM and CL) of R turtles were significantly different from those of AL and R-AL turtles during weeks 6 through 12 (Tukeys HSD post hoc tests, p < 0.0001 for both comparisons for BM and CL). Compensatory growth occurred during weeks 7 through 9, as demonstrated by significantly larg er SGR in R-AL turtles relative to AL turtles in weeks 7 and 8 (for SGRbm) and in weeks 8 and 9 (for SGRcl) (Fig. 2-4). Food conversion efficiencies (FCE) for BM and CL differed among treatment groups, but the patterns depended on whether FCE was calculate d as mass gain per unit of food consumed or as carapace length gain per unit of food consume d. Repeated measures ANOVA of FCE revealed significant time and treatment effects during week s 1 through 5 and 6 through 12 and significant interactions between time and treatment gr oup during weeks 6 through 12 (Table 2-5). The pattern of FCEbm of AL turtles was significantly differe nt from those of R and R-AL turtles during weeks 1 through 5 and during weeks 6 through 12 (Tukeys HSD post hoc tests, p < 0.01 for all comparisons). The pattern of FCEcl of AL turtles likewise diffe red from those of R and RAL turtles during weeks 1 thr ough 5, but the pattern of FCEcl for all treatment groups differed significantly for weeks 6 through 12 (Tukeys HSD post hoc tests, p < 0.0001 for all significant comparisons). Specifically, the two groups feeding ad libitum after week 5 differed in conversion efficiencies, with R-AL turtle s demonstrating enhanced FCEbm in weeks 6 and 7 and enhanced FCEcl between weeks 6 and 11 (Fig. 2-5). Trends in conversion efficiencies were more consistent among weeks when FCE was calculate d as change in carapace length per unit of food consumed. 33

PAGE 34

Visceral organ size expressed as a percentage of BM (minus MGC) or CL was smaller in food-restricted turtles than in ad libitum -fed turtles, particularly at t12 (Tables 2-6 and 2-7). Liver was lighter and stomach and intestine were shor ter in R turtles than in AL turtles at t5. At the end of the feeding trial, liver and stomach were lighter and stomach and intestine were shorter in R turtles than in AL or R-AL turtles. Intestine mass was lighter in R turtles than in R-AL turtles at t12, and the difference between R and AL turtles approached significance ( p = 0.052). Turtles at 5 weeks and 12 weeks also differed significantly in body composition (Tables 2-6 and 2-8), with R turtles having higher N and lower OM, lipid, energy, and lipid:lean contents than AL turtles at both t5 and t12. Body composition of R-AL turtles at t12 was comparable to that of AL turtles. Daily water temperatures dropped slightly as the experiment progressed into late autumn. Occasional variation in temperatur es was the result of rainfall from tropical weather systems, including a hurricane that occu rred during week 8 (Fig. 2-6). Discussion The data clearly demonstrate that green turtle s are capable of CG within the first three months after hatching. This growth pattern was not simply the result of ra pid gut filling after the diet switch, as both BM and CL increased proportionally faster in R-AL turtles than in AL turtles. However, turtles undergoing CG achie ved only partial compen sation, because body size of R-AL individuals did not catch up to that of AL individuals by the time the period of CG ended. The fact that smaller, prev iously food-limited turtles grew fast er than larger turtles of the same age represents a strategy that decrease s the phenotypic variance among individuals from the same cohort (Wilson and Rale 2006). It appears that green turtles are ab le to evaluate their growth patterns and adjust them, at least some what, toward a more optimal trajectory when nutritional conditions improve. Because the risk of mortality from predation decreases as size increases for juvenile sea turtles (Musick and Limpus 1997), this capacity for rapid growth when 34

PAGE 35

food availability is high allo ws young green turtles to expedite progression through the vulnerable hatchling stage. The mechanism for CG in juvenile green turtle s is enhanced FCE rather than hyperphagia. Given that CG is often effected through increased intake, particularly in fish (e.g., Jobling and Johansen 1999, Johansen et al 2001, Nikki et al 2004, Tian and Qin 2004), I expected previously restricted green turtles to become hyperphagic when food was provided ad libitum The fact that green turtles were not hyperphagic suggest s that intake rates of AL turtles were already maximal and therefore could not be exc eeded by R-AL turtles. McCauley and Bjorndal (1999) previously demonstrated that juvenile logg erhead sea turtles also do not increase intake beyond the maximal rate attained when feeding ad libitum Animals that do not increase consumption rates even when food is abundant may be intrinsically limited by constraints on rates of digestion and/or passage (Speakma n and Krl 2005). In c ontrast, the hyperphagic response of fish undergoing CG indicates that continuously ad libitum -fed fish feed at submaximal rates. It is possible that enhanced conversion efficiency was effected through improved digestibility via either behavioral or mo rphological changes. Although I did not quantify behavior throughout the study, the ta nks of food-restricted turtles rarely contained feces because these individuals practiced c oprophagy. Reingestion of feces when food is limited has been reported in rodents (Kenagy et al 1999) and serves to increase digestive efficiency. The extent to which R-AL turtles relied on nutrient recycling is unknown, but continued reingestion of feces after the switch to an ad libitum diet could have allowed for the improved FCE and elevated growth rates I observed. Diges tibility could also have been increased by upregulation of intestinal surface area. However, preliminary data on intestinal histology indicate that 35

PAGE 36

food-restricted turtles had not only smaller guts than ad libitum turtles in terms of mass and length but also decreased intestinal surface area (Roark and Bjorndal, unpublished data). Because I did not evaluate density of ep ithelial transporters, I cannot rule out the possibility that uptake rates may have been enhanced in food-restrict ed turtles despite a significant reduction in intestinal mass, as has been s hown in food-restricted birds (Brz k and Konarzewski 2001). Alternatively, the enhanced FCE of R-AL turtle s relative to AL turtles, especially in the first few weeks after the switch to ad libitum feeding, may have resulted from decreased metabolic expenditure. I did not quantify metabolic rates, but I did fi nd that several major visceral organs (e.g., liver, stomach, and intestine) in R turtles were smaller than those in AL turtles. By down-sizing organs that would otherwise require disproportionate metabolic expenditure to maintain them, R turtles may have been able to allocate less of their assimilated food into maintenance metabolism and more into growth. Similar results have been obtained in studies of fasted or food-restricted migratory birds (Lee et al 2002, Karasov et al 2004), which had significantly smaller digestive and assi milatory organs compared to birds fed ad libitum In the intestines, this decrease in size was due largely to changes in the mucosal layer of the villi. The observed decrease in organ mass in these birds was reversed by several days of feeding ad libitum Because I sampled only at times t0, t5, and t12, I was unable to determine the time course over which the organ sizes of R and R-AL turtles changed. Upon a return to ad libitum conditions, R-AL turtles may have experienced a delay in up-regul ation of visceral organ size. The switch to ad libitum feeding at a time when maintenan ce expenditure was minimized would have allowed for the rapid growth I documente d in R-AL turtles duri ng weeks 7 through 9 (Ali et al 2003). 36

PAGE 37

Although both BM and CL increased proportionally faster in R-AL turtles than in AL turtles, these two morphometric measurements de monstrated different dynamics. During the first week of ad libitum feeding, the SGRbm but not the SGRcl of R-AL turtles was comparable to that of AL turtles. This decoupling of mass and length growth may result from differential allocation of assimilated nutrients in the first week of ad libitum feeding. Less energy is required to convert assimilated nutrients into reserve tissue than in to more complex structural components such as skeletal tissue (Broekhuizen et al 1994). Growth efficiency woul d therefore be enhanced in turtles that preferentially allocated nutrient s to mobilizable tissue gain rather than to unmobilizable tissue gain, at least in the initial stages of elevated growth. The time lag in the increase of SGRcl but not in SGRbm may also result simply from a rapid increase in BM immediately after the diet switch due to filling of the gut. At t12, gut contents accounted for an average of 9.9%, 12.7%, and 13.5% of total wet BM for R, AL, and R-AL turtles, respectively. Turtles on the restricted diet th erefore carried proportionally le ss digesta than turtles feeding ad libitum meaning that gut filling probably accounted for a small percentage of the initial increase in BM after the diet switch for R-AL turtles. I used my measurements of BM and CL to assess body condition (as condition index, CI) of turtles in each treatment gr oup for each week of the study. Not surprisingly, CI of R turtles decreased steadily until approximately the eighth week of th e experiment, indicating that these animals were becoming leaner as the study progr essed. My body composition results support this conclusion. Total body nitrogen content was high er and OM, lipid, and energy contents were lower in R turtles than in AL turtles at both t5 and t12. Somatic growth in food-restricted turtles therefore entailed either lower ra tes of lipogenesis and/or protei n catabolism or higher rates of protein deposition and/or lipid cat abolism than in AL turtles. In other studies, food deprivation 37

PAGE 38

has been linked to down-regulation of the activity of lipogenic enzymes (Bastrop et al 1991, Rosebrough and McMurtry 1993), and alterations in protein metabolism during food restriction are also common (Dhahbi et al 2001, Hagopian et al 2003). Turtles that experienced a switch from restricted to ad libitum feeding demonstrated a rapid increase in CI after the diet switch, su ch that body condition of R-AL turtles was not significantly different from that of AL turtles by week 7. The rapi d growth of R-AL turtles was accompanied by elevated lipid deposition between weeks 5 and 12. This increase in whole body adiposity and concomitant decrease in nitrogen content allowed R-AL turtles to achieve a body composition not significantly different from that of AL turtles by the e nd of the experiment. Cessation of CG may have resulted from R-AL turtles attaining a tissue composition similar to that of AL turtles. In other words, repletion of body stores may have served as a signal regulating the duration of th e compensatory response (Jobling and Johansen 1999, Ali et al 2003). The fact that body composition was restored before full body size compensation was achieved provides further evidence that CG may have been regulated by condition rather than overall body size. In fish, several studies (e.g., Bull and Metcalfe 1997, Johansen et al 2001) have shown that the rate of repl etion of lipid stores rather than the attainment of a certain body size controlled the duration of the compensatory response. In contrast to my results, body composition in these fish studies exerted its effect s by altering appetite rather than FCE. Because I found no evidence for hyperphagia, I conclude th at juvenile green turt les, unlike many fish species, do not adjust their inta ke in response to adiposity. My finding that turtles experien cing consistently high food ava ilability grew more slowly than turtles undergoing compensatory growth imp lies that maximal growth rates may not always be advantageous for green turtles. Despite th e potential for increased body size to provide a 38

PAGE 39

benefit to the individual in te rms of fitness (Roff 1992, Stearns 1992), fast growth may carry a variety of costs (Arendt 1997, Blanckenhorn 2000, Metcalfe and Monaghan 2001). In other animal models, these costs include delayed skeletal ossification (Arendt and Wilson 2000), weakened musculature (Christiansen et al 1992), reduced locomotor performance (Billerbeck et al 2001, lvarez and Metcalfe 2005), accelerated telomere degradation (Jennings et al 1999), and decreased longevity (Olsson and Shine 2002). The proximate determinant of such costs may be the accumulation of cellular damage during rapid growth, as modeled by Mangel and Munch (2005). These detrimental effects of rapid grow th may explain the sub-maximal growth rates typically demonstrated by animals feeding ad libitum continuously (Mangel and Stamps 2001). I have demonstrated that cellular antioxidant poten tial of R-AL turtles is decreased compared to AL turtles, at least in mitotically active tissu e (Chapter 4). If such costs place an upper limit on growth in green turtles, they may further expl ain the transient and incomplete nature of the compensatory response I observed. This study is the first to document the exis tence of and mechanisms for CG in young green turtles. The capacity to grow quickly, albeit only transiently, provides juven iles an opportunity to mitigate some of the costs of being small. At the same time, however, the transitory nature of the CG response suggests that the benefits of acceler ated growth are countered by costs potentially including decreased longevity and/or performance that may be me diated by altered antioxidant function. The extent to which CG is possible at different ages a nd during different life stages is unknown but deserves further study. For example, th e ontogenetic shift in habitat use and diet that green turtles undergo as juveniles may provide an opportunity for CG, as such niche shifts often correspond to improved food availability (Ali et al 2003). What is clear from this study is that differences in food availability can i nduce plasticity in gr owth, morphology, and body 39

PAGE 40

composition in young green turtles. This plasticity could substa ntially affect life-history endpoints such as age and size at maturity, repr oductive output, and longe vity that directly influence the viability of green turtle populations. 40

PAGE 41

Table 2-1. Kruskal-Wallis test results for nutrient content of biweekly food samples ( n = 2 at each of 7 time points). Source of variation Content Mean SE df 2 p Per ) 91.95 0.04 6 07.371 0.288 J/g) ) (%) 18.25 0.28 02.057 0.914 DM OM (% Energy (k 20.92 0.06 6 12.457 0.053 Nitrogen (%) Lipids (%) 08.28 0.05 16.78 0.26 6 6 01.943 0.925 02.057 0.914 Per OM Energy (kJ/g Nitrogen (%) 22.75 0.06 09.01 0.05 6 6 12.457 02.114 0.909 0.053 Lipids 6 According to the Meli bel, cde fib 5% and phosphorus was 1%. M = = organic m ck Aquafeed la ru er was Abbreviations: D dry matter, OM atter. 41

PAGE 42

Table 2-2. Repeated measures anal yses of variance for weekly averages of daily intake and daily mass-specific intake. Source of variation df SS F p Intake (g/day), weeks 1-5 Between subjects effects Group 2 451.274 168.22 < 0.0001 112 ects effects 057.65 < 0.0001 044.743 056.51 < 0.0001 448 044.343 ubjects linear contrasts 1 ffects 168.37 < 0.0001 047.39 < 0.0001 432 106.055 trasts 1 236.124 296.96 < 0.0001 Time Time) ntake (g/g*day), weeks 1-5 Error Within subj 150.225 Time 4 022.823 Group Time Error (Time) 8 Within s Time Group Time 1 2 022.108 043.336 071.99 070.56 < 0.0001 < 0.0001 Error (Time) 112 034.395 Intake (g/day), weeks 6-12 Between subjects effects Group 2 .747 x 103 103.56 < 0.0001 Error Within subjects e 72 607.367 Time 6 12 248.000 139.612 Group Time Error (Time) Within subjects linear con Time Group 2 129.989 081.74 < 0.0001 Error ( Mass-specific i 72 057.250 Between subjects effects Group Error 5-2 112 8.566 x 10 ffects 3-4 017.02 < 0.0001 1 00 < 0.0100 448 2.543 x 10-3 ubjects linear contrasts 3 7 112 1.736 x 10 ke (g/g*day), weeks 6-12 4.330 x 10-2 387.27 < 0.0001 bjects effects 015.86 < 0.0001 432 inear contrasts 018.04 < 0.0001 2 .497 x 10-3 359.41 < 0.0001 Within subjects e Time Group Time 4 8 .863 x 10 .834 x 10-4 4.04 Error (Time) Within s Time 1 .044 x 10-4 019.64 < 0.0001 Group Time 2 .915 x 10-6 -3 000.26 0.775 Error (Time) Mass-specific inta Between subjects effects Group 2 Error u 72 4.025 x 10-3 Within s 7.750 x 10-4 Time Time 6 Group 12 3.655 x 10-4 003.74 < 0.0010 Error (Time) l 3.518 x 10-3 Within subjects 3.465 x 10-4 Time Time 1 Group 2 3.775 x 10-5 000.98 0.379 Error (Time) 72 1.383 x 10-3 Three groups w ere tested: ad libitum for 12 weeks, food-restricted for 12 weeks, and by ad libitum for 7 weeks. Wen Mauchleys test indicated metry assumpti on was violated, Greenhouse-Geisser lues are Significant p-values are in bold food-restricted for 5 weeks followed h that the compound sym p-va presented 42

PAGE 43

Table 23. Repeated measures analyses oa riance for weekly body mass, straight carapace th, and condition index. f v leng Source of variatio n df SS F p Body Mas s (g), weeks 0-5 Between subjects effects 00 < 0.0001 Group 2 1.411 x 106 0068.24 < 0.0001 7.442 x 105 Within subjects effects 5.885 x 105 0395.15 < 0.0001 12 2.320 x 105 0077.90 < 0.0001 432 ength (mm), weeks 0-5 75.97 Group 2 7.591 x 104 4 Error 112 5.596 x 10 Within subjects effects Time 5 1.022 x 105 0600.23 < 0.0001 Group Time 10 4.433 x 104 0130.23 < 0.0001 Error (Time) 560 1.906 x 104 Body Mass (g), weeks 6-12 Between subjects effects Error 72 Time 6 Group Time ime) Error (T 1.072 x 105 Carapace L Between subjects effects Group 2 6.012 x 103 0050.02 < 0.0001 Error e 112 6.731 x 103 Within subjects ffects 2711.94 < 0.0001 10 3.180 x 103 0165.88 < 0.0001 560 1.073 x 103 ength (mm), weeks 6-12 subjects effects 5.701 x 104 0080.43 < 0.0001 ffects 1041.43 < 0.0001 12 6.613 x 103 0119.13 < 0.0001 432 1.998 x 103 eks 0-5 2.599 x 104 Time Group Time 5 Error (Time) Carapace L Between Group 2 Error 72 2.552 x 104 Within subjects e Time 6 2.891 x 104 Group Time Error (Time) Index (g/cm3), we Condition Between subjects effects 0022.75 < 0.0001 112 ffects 0043.30 < 0.0001 560 7.185 x 10-3 eks 6-12 subjects effects 0033.42 < 0.0001 72 4.171 x 10 ubjects effects 000 63 < 0.0001 432 4.553 x 10 2 1.937 x 10-2 Group Error 4.767 x 10-2 Within subjects e 0657.58 < 0.0001 Time 5 10 4.218 x 10-2 5.556 x 10-3 Group Time Error (Time) Condition Index (g/cm3), we Between Group 2 3.872 x 10-2 -2 Error Within s Time 6 1.351 x 10-4 0002.14 0.103 Group Time 12 7.117 x 10-4 -3 5. Error (Time) Three groups were tes ted: ad libitum for 12 weeks, food-restricted for 12 weeks, and ricted for 5 weeks followed by aibitu eeksMa test indicated compound symmetry assumpti on was violated, Greenhouse-Geisser lues are t p-values are in bold. food-rest d l m for 7 w When uchleys that the p-va presented. Significan 43

PAGE 44

Table 2-4. Re peated measures analyses ri aekly spcific growth rates (SGR) for d carapace lngth (cl).d SS of va nce for we e body mass (bm) an e Source of variation f F p SGR (%/ bm day), weeks 1-5 Between s ubjects effects < 0 112 effects 310.715 218.38 < 0.0001 Error 72 0 51.222 Within subjects effects Time 6 0 18.779 0 18.27 < 0.0001 Group Time 12 00 7.770 00 3.78 < 0.0001 Error (Time) 432 0 74.002 SGRcl (%/day), weeks 1-5 Group 2 33683 0 61.620 .7 3067 .0 .01 00 Error Within subjects Time 4 115.394 205 < 0.01 Group Time 8 00 9.937 00 8.85 < 0.0001 Error (Time) 448 0 62.875 SGRbm (%/day), weeks 6-12 Between subjects effects Group 2 5.5 00 Between subjects effects Group 2 0 24.174 202.20 < 0.0001 Error 112 00 6.695 Within subjects effects Time 4 0 27.471 562.96 < 0.0001 Group Time 8 00 1.494 0 15.31 < 0.0001 Error (Time) 448 00 5.465 SGRcl (%/day), weeks 6-12 Between subjects effects Group 2 0 25.945 180.39 < 0.0001 Error 72 00 5.178 Within subjects effects Time 6 00 2.029 0 38.20 < 0.0001 Group Time 12 00 1.851 0 17.43 < 0.0001 Error (Time) 432 00 3.824 Three groups were tested: ad libitum for 12 weeks, food-restricted for 12 weeks, and food-restricted for 5 weeks followed by ad libitum for 7 weeks. When Mauchleys test indicated that the compound symmetry assumpti on was violated, Greenhouse-Geisser p-values are presented. Significant p-values are in bold. 44

PAGE 45

Table 2-5. Repeated measures analyses of va riance for food conversion efficiencies (FCE) for body mass (bm) and carapace length (cl). Source of variation df SS F p FCEbm (g/g), weeks 1-5 Between subjects effects Group 2 .5 00 0 7.755 23.418 0 0 0 1.932 00 575 0.4 ects effects 00 724 < 0.01 eks 1-5 1846 < 0.01 Error 1 12 Within subj ects effects Time 4 13.142 0 1.278 3604 00 1.775 .5 < 0.01 < 0.086 0 00 Group Time Error (Time) 8 4 48 40.323 FCE (g/g), week bms 6-12 Between subjects effects Group 2 .9 00 Error 72 11.640 Within subj 0 3.261 Time Time 6 .2 00 Group 12 0 4.082 00 4.522 < 0.0001 Error (Time) 432 32.498 FCEcl (mm/g), we Between subjects effects Group 2 2 24.260 198.88 < 0.0001 112 bjects effects 23233 < 0.01 0 0.123 00 067 < 0.4 0 448 ks 6-12 0 62 13035 < 0.01 72 0 1.872 0 00 591 < 0.01 432 Error 0 6.851 Within su Time 4 18.711 .6 00 Group Time 8 .7 57 Error ( Time) FCE (mm/g), wee 0 9.008 cl Between subjec ts effects Group 2 .79 .6 00 Error bjects effects Within su 7206 < 0.01 Time 6 12 0 1.800 0 0.294 .2 00 Group Time Time) .8 00 Error ( 0 1.795 Three groups we re tested: ad libitum for 12 weeks, food-restricted for 12 weeks, and for 5 weeks followed by itum eeks. Wen Mauchleys test indicated are food-restricted ad lib for 7 w h that the compound symmetry assumpti on was violated, Greenhouse-Geisser p-values presented. Significant p-values are in bold. 45

PAGE 46

46 Identf Groups Tested in Pairwise Comparisons Table 2-6. Omnibus F 2, and p-values for analyses of variance of dissection data collected at weeks 5 and 12. ity o Omnibus F and 2 AL anAL nd R R-AL and R d RAL a Week 5 F 1,18 = 30.06, p < 0.000 F1,18 = 21.98, p < 0.001 F1,18 = 80.05, p < 0.000 ex (%) F1,18 = 0 1.60, p = 0.222 F1,18 = 0 1.03, p = 0.324 F1,18 = 19.28, p < 0.001 F1,18 = 0 8.03, p = 0.011 1,18 = 20.22, p < 0.001 ) F 2 = 12.10, p = 0.001 M) F 2 = 14.35, p < 0.001 F 2 = 13.72, p < 0.001 OM) F 2 = 11.84, p < 0.001 ) F1,18 = 31.33, p < 0.000 F 2 = 14.29, p < 0.001 F 2 = 14.29, p < 0.001 F2,27 = 227.3, p < 0.000 01 0001 F2,27 = 73.08, p < 0.000 .0001 < 0.0001 F2,27 = 107.8, p < 0.000 01.0001 F2,27 = 14.55, p < 0.000 01 0.001 0 F2,27 = 4.91, p = 0.015 0.052 0 < 0.019 0 2,27 = 17.86, p < 0.000 00 .0001 < 0.001 0 2,27 = 18.67, p < 0.000 0.7 0.0001 < 0.0001 2,27 = 64.16, p < 0.000 01.0001 ) F2,27 = 11.05, p < 0.001 01 0 0.043 0 F 2 = 19.94, p < 0.000 00 0.0001 < 0.0001 DM) F2,27 = 63.75, p < 0.000 01.0001 F2,27 = 51.73, p < 0.000 01.0001 M) F = 69.97, p < 0.00 0.0001 < 0.0001 01 BM (g) ) 1 CL (mm 0 LM Index (%) 1 SM Ind 0 TIM Index (%) 0 SSL Index 0 TIL Index 0 % OM F 0 N (% DM 0 N (% O 0 Lipid (% DM) 0 Lipid (% 0 Energy (kJ/g DM 1 Energy (kJ/g O M) Lipid:Lean 0 0 Week 12 < 0.00 BM (g) 1 0.050 < 0. CL (mm) ex (%) 1 0.61 0 < 0 LM Ind 1 0. 765 < 0.00 < 0 SM Index (%) 1 0.544 < 0.00 < TIM Index (%) 0 0.892 < SSL Index F 1 .15 < 0 TIL Index F F 1 69 < % OM M 1 0.454 < 0.00 0.0 < 0 < N (% D 0 0.099 < N (% OM) 1 .28 < Lipid (% 1 0.562 < 0.00 < 0 Lipid (% OM) 1 0.468 < 0.00 < 0 Energy (kJ/g D 2,27 Energy (kJ/g OM) F2,27 = 44.34, p < 0.0001 0.172 < 0.0001 < 0.00 Lipid:Lean F2,27 = 73.56, p < 0.0001 0.696 < 0.0001 < 0.0001 01 0.467 < When F values are reported, data we re analyzed using analysis of variance with Tukeys H post hoc test. When 2 values are reported, d SD ata were analyzed using a Kruskal-Wallis test and pairwise Mann-Whitney U tests with set at 0.017 to account for multiple comparisons among t12 groups. Statistically significant p -values are indicated in bold. Abbreviations: AL = ad libitum for 12 weeks, R = food-restricted for 12 week s, R-AL = food-restricted for 5 weeks and ad libitum for 7 weeks, BM = body mass, CL = carapace length, LM = liver mass, SM = stomach mass, TIM = total intestine mass, SSL = stomach straight length, TIL = total intestine length, OM = organic matter, DM = dry matter.

PAGE 47

47Table 2-7. O ass o r lGroup We TIL Index rgan masses (mean st ength as a proportion of ek n BM (g) and bod ar y d e maC rro ss L ( r) or mm fro car) m ap tur acL tle e leM s d ngInd iss th,ex ec re(% ted sp) at ectS 0, iveM 5, ly)Ind and ex ( 1%) 2 w eeTI ksM I rend porex ( te%) d a s i n diSSL ces In (de orgx an m AL 0 0.92 aa 1 .18 AL 5 6a 1 4.04 1.40 0.0 0.21ab a 1 9.08 R 5 0 .2 1.0b 1 2.38 0.07b 1.24 0.03.62 0.16ab 0.61 0.02b 1 8.03 AL 1a 117.4 4.3a 1 5.29 0.27a 1.58 0.11a a 4.17 0.15ab 2 0.82 0.01a 10.76 R-AL 1 5 11.07 R 1 7 1 8 10 0 34.2 0.6 0 10 0 92.6 9.6a0 10 0 51.6 1.9b2 10 227.5 25.8 2 10 150.0 9.2b2 10 0 71.7 1.0c0 a 0 0 61 88 075 .7 .0 0. 2. 6a 1 3.6 1 0.1 0.1 5a 7a 0.0 4a 9a 9a 2. 3. 97 89 0 .12 0 .67 0.75 0 0 .0 .03 2a 8.22 0 a 0.22a 0.29b 0.26a 0.34a 0.18b 0 1 05 085 .7 .9 1. 0 7a 1 .9b 1 4.9 2.5 7 8 0.2 0. 0a 10b 1.4 0.9 0.0 0.05 7a b 4. i 3. 27 61 0 0 .17 .16 ab ba 2 0 .78 0.69 0 .02 0.0 a 2b .91 1Difference 2Difference Within colu er est analy of variance f t h as in Table 2-6. between AL and R-AL a between AL and R appro mns, values with differen ollowed by pairwise Ma ppr ac t le nnoa hes tte W ch si r s hit es gn up ne sig ific ers y U nif an cri te ica ce pts sts nc ( p ar o e ( = e s r T p = 0.0 ig uk 0 52 nifi ey .06 ). ca s 1). ntl HS y d D iff pos ent oc wit tes hi t, p n ti < me 0. p 05 eri ). A od bb s ( re Kr via usk tio alns W ar all e t is t he s or me a sis

PAGE 48

(% Co Lipi Table 2-8. Body composition (mean standard error) of turtles dissected at 0, 5, an d 12 weeks reported as percent of dry matter DM) and percent of organic matter (% OM). n % O tent Lip Group Week M Nitrogen Con id Content Energy ntent d:Lean % DM % OM % DM % OM kJ/g DM kJ/g OM AL 0 10 85.5 0.3a .96 0.09a .99 0.08a 203 0.4a 23.8 0.4a 20.9 0.1a 24.4 0.1a 0.272 0.006a AL 5 10 82.2 0.3 12.20 0.2114.85 0.28 15.5 1.418.8 1.7a 19.7 0.3a 23.9 0.4 0.206 0.023 R 5 10 i80.3 0.3b 13.00 0.06b 16.19 0.04b 10.0 0.1b 12.5 0.2b 17.8 0.1b 22.2 0.1b 0.123 0.002b 81 .6 5 0. 1.4 1.6a a 0.25 R-AL 12 10 82.0 0.3a 12.07 0.12a 14.72 0.15a 16.6 0.8a 20.2 1.0a 20.0 0.3a 24.3 0.4a 0.221 0.013a i76. .54 38 0. 0.4b 0.12 11a 13a .a a a a AL 12 10 .4 0.3a 11 8 0.17a 14.3 25a 18.6 a 22.8 20.5 0.4 25.1 0.4a 8 0.023a R 12 10 6 0.5b 12 0.10b 16. 0.06b 0 9.4 3b 12.3 16.7 0.1b 21.8 0.1b 0 0.004b 48 Within columns, values with different le tter superscripts are significan tly different within time periods. Treatment groups and data is are saas i analys e th me n Table 2-7.

PAGE 49

0.00 0.0 1 02 0. 0. 0.04 0123456789101112Tiwe) 03 me ( eks AL R-AL Average Mass-Specific Daily Intake (g/g) Ra b b a b b a b a b b a a b a a b a a b a a b a a b a a b a a a ur1. Average mass-specific daily intake (m ean standard error) du ring each week of the feeding tr Difrent letters indica val that ignificantly different within weeks (analysis of variance, Tukeys HSD post hoc test, p < 0.05). Turtles in the R gr weritd fr ad libitum diet at the beginning of week 6. Sample sizes in weeks 0 through 5: n = 37 AL, 39 R-AL, and 39 R. Sample sizes in weeks 6 through 12: n = 17 AL, 29 R-AL, and 29 R. Abbreviations: AL = abitu r e R =d-restricted for 12 weeks, R-AL = food-restricted for 5 weeks and aditu or 7ks. b a b b a a bare s foo wee Fig e 2ial. oup fe e sw te u es -AL che om a restricted diet to an d li m fo 12 w lib eks, m f 49

PAGE 50

(a ) 0 50 100 150 200 250 300 350 0123456789101112Time (weeks)Body Mass (g) AL R-AL R a b b a b b a b b a b b a b b b a b a b c a b c a b c a b c a b c a b c a a a (b) 50 75 100 125 150 0123456789101112Time (weeks)Straight Carapace Length (mm) AL R-AL R a b b a b b a b b a b b a b b b a b b b b c b c b c b c b c a a a Figure 2-2. Body mass (a) and straight carapace length (b) (mean standard error) at the midpoint of each week. Different letters indicate values that are significantly different within weeks (analysis of variance, Tukeys HSD post hoc test, p < 0.05). The arrow indicates the time at which turtles in the R-AL group were switched from a restricted diet to an ad libitum diet. Sample sizes and abbreviations are the same as in Figure 2-1. a a a a a a 50

PAGE 51

0.12 0.13 0.14 0.15 0.16 0.17 0123456789101112Time (weeks)Condition Index (g/cm3) AL R-AL R a b b a b b a b b a b b a b b b a c a a b a a b a a b a a b a a b a b a a a a Figu re 2-3. Condition index (mean standard error) in each week calculated as BM/CL3, where BM = body mass (g) and CL = carapace length (cm). Different lette rs indicate values that are significantly differe nt within weeks (analysis of variance, Tukeys HSD post hoc test, p < 0.05). The arrow indicates the time at which turtles in the R-AL group were switched from a restricted diet to an ad libitum diet. Sample sizes and abbreviations are the same as in Figure 2-1. 51

PAGE 52

( a) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 123456789101112Time (weeks)Specific Growth Rate for BM (%/day) AL R-AL R a b c a a b a c b a a a a a a a a a a a a a b b b b b b b b b bb b b b (b) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 123456789101112Time (weeks)Specific Growth Rate for CL (%/day) AL R-AL Ra a b a a b b c c a a b a a a a a a b bb b b b b b c b b b a a a a b b Figure 2-4. Specific growth rate (mean standard error) for body mass (BM, a) and straight carapace length (CL, b) during ea ch week calculated as 100*(ln[sizet+1]-ln[sizet])/7 where size = BM (a) or CL (b) and t = time (weeks). Different letters indicate values that are significantly differe nt within weeks (analysis of variance, Tukeys HSD post hoc test, p < 0.05). The arrow indicates the time at which turtles in the R-AL group were switched from a restricted diet to an ad libitum diet. Sample sizes and abbreviations are the same as in Figure 2-1. 52

PAGE 53

(a) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 123456789101112Time (weeks)Food Conversion Efficiency for BM (g/g) AL R-AL R a ab b a a a b a b a a a a a a b a a a a a a a a a a a b a b a b b a a a (b) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 123456789101112Time (weeks)Food Conversion Efficiency for CL (mm/g) AL R-AL R a c b a a b b c c a a b b a a b a c b a a b a b a b a a a a a b c b c b Figure 2-5. Food conversion effici ency (FCE, mean standard error) for body mass (BM, a) straight carapace length (CL, b) during each week calculated as size change per unit of food consumed, where size = BM (a) or CL (b). Different letters indicate values that are significantly differe nt within weeks (analysis of variance, Tukeys HSD post hoc test, p < 0.05). The arrow indicates the time at which turtles in the R-AL group were switched from a restricted diet to an ad libitum diet. Sample sizes and abbreviations are the same as in Figure 2-1. and 53

PAGE 54

0 5 10 15 20 25 30 35 40 0123456789101112 WeekWater Temperature (oC) Max Min Figu re 2-6. Daily water temperatures (mean sta ndard deviation) throughout the feeding trial. Water temperature was monitored using five min/max thermometers in tanks without turtles. Water temperatures in weeks 8 a nd 9 fluctuated as a result of a hurricane. 54

PAGE 55

CHAPTER 3 BIOCHEMICAL INDICES AS CORRELATES OF RECENT GROWTH IN JUVENILE GREEN TURTLES (C helonia mydas) Introduction The green turtle, Chelonia mydas is an endangered marine herbivore with a circumglobal distribution (Seminoff 2002). Over exploitation of this species by humans during the last several centuries has caused drastic population declines (Jackson et al 2001). Development of effective management plans for this species requires knowledge of demographic parameters such as somatic growth rates. However, assessing grow th rates for long-lived and far-ranging green turtles typically requires time-consuming mark and recapture programs in which recapture probabilities can be quite low (Limpus 1992). Fu rthermore, growth ra tes calculated using biochemica to nvironmental influences (Ferron and Leggett 1994, Gilliers et al. 2004). Establishment of alternative techniques for estimating recent growth rates of individual turtles upon first capture would substantially improve our ability to evaluate th e instantaneous status of C. mydas populations and therefore to asse ss progress toward recovery goals for this endangered species. Macromolecular indices (RNA concentrations RNA:DNA ratios, RNA:protein ratios, and/or protein:DNA ratios) are frequently measured as indicators of prot ein synthesis potential and growth in marine fish and invertebrates (Bulow 1970, Carter et al. 1998, Buckley et al 1999, Dahlhoff 2004, Caldarone 2005, Mercaldo-Allen et al. 2006, Vidal et al. 2006). These indices are particularly useful for evaluating recent environmental conditions, as they reflect differences in growth rates over a period of several days (Rooker and Holt 1996, Buckley et al. 1999, Vrede et al. 2002). The use of these indices depends on the assumption that total RNA content of a cell (including messe nger RNA, transfer RNA, and ri bosomal RNA) should increase morphometrics represent long-term, cumulative changes and often do not correlate well with l indices of short-term growth due to differences in the latency of these responses e 55

PAGE 56

as the cellular demand for protein synt hesis and growth increases (Buckley et al. 1999), while DNA content per cell should he RNA:DNA ratio is therefore an index of cellular pr otein synthesis capacity. Because nucleic acid concentrations and the ra ted dices n his an, his be relatively co nstant (Wallace 1992). T tios between them respond rapidly to fluctua tions in food availability, they are considered reliable indices of instantane ous condition and growth (Rooker et al. 1997, Okumura et al. 2002, Islam and Tanaka 2005, Vidal et al. 2006). Despite widespread use as sensitive measures of recent growth rates in marine fish and invertebra tes, nucleic acid ratios have not been valida for application to studies of reptile growth. The purpose of this study was to evaluate the use of morphometric and biochemical in as predictors of recent growth rates in green turtles maintained under controlled feeding conditions. Because analyzing bi ochemical indices of growth typically requires homogenizatio of tissues extracted from euth anized individuals, I also exam ined the potential for measuring nucleic acid concentrations in whole blood, a tissue that is not typically tested in studies of t kind. Validation of a physiological growth index th at can be assessed using minimally invasive sampling techniques and without sa crificing the animals would substantially enhance our ability to monitor short-term responses of green turtles to environmental perturbations. Materials and Methods Animal Care A twelve-week feeding trial was conducted at the Cayman Turtle Farm in Grand Caym British West Indies, in accordance with the policies of the Institutional Animal Care and Use Committee at the University of Florida (permit #Z061) Details of the animal care aspects of t study can be found in Chapter 2. Briefly, Chelonia mydas hatchlings were housed individually in sea water in 68-liter tanks. Turtles were fed turtle pellets (Melick Aquafeed, Catawissa, PA) twice daily. 56

PAGE 57

Prior to the beginning of th e study, turtles were all fed ad libitum to establish average dai intake. During the study, turtles in the ad libitum group (AL) were offered an excess of food day for 12 weeks, turtles in the restricted group (R) were fed 50% of average initial AL intake each day for 12 weeks, and turtles in the restricted-ad libitum group (R-AL) were fed the restricted amount of food for fi ve weeks and then were fed ad libitum for seven weeks. The amount of food offered during food restriction wa s sufficient to maintain turtles on a positive growth trajectory. Turtles were weighed and me asured (straight carapace length) each week. Tissue Collection ly each ed as described below. ssays les erent from those used for DNA is olation) were weighed but not ground in liquid At the conclusion of the twelve -week study, seven AL turtles, ten R turtles, and ten R-AL turtles were weighed to the nearest 0.1 g and euthanized with an intramuscular overdose injection of ketamine (Ketaset, 100 mg/kg body ma ss) in the right pectoral muscle. When each turtle failed to respond to a pa in stimulus, it was decapitated. A blood sample was collected from th e decapitation site (as in Storey et al 1993 and Packard et al 1997). The heart and a portion of the righ t lobe of the liver were excised, and blood, heart, and liver samples were snap-frozen in liquid nitrogen no more than three minutes after decapitation. Tissues were maintained at C until they were homogeniz Biochemical A Tissues from each individual turtle were analyzed for DNA and RNA concentrations. Subsamples of frozen whole blood, heart, and liver were weighed, and DNA was isolated with DNeasy kits (Qiagen Inc., Valencia, CA) using the manufacturers protocol for animal tissues. To isolate RNA, subsamples of frozen heart a nd liver tissue (different from those used for DNA isolation) were weighed and then ground in liqui d nitrogen using mortar and pestle. Subsamp of blood (diff 57

PAGE 58

nitrog ples ing a PicoGreen dsDNA quantit ation kit (Invitrogen Corporation, Carlsbad, CA) a 28 nm otein ata for body mass (BM) and cara pace length (CL) were used to calculate specific growth urtle during the final 10-11 days of the study according to the following equat SGR = (ln[CL ] ln[CL ])*100/t where BM and CL represent body size 10 or 11 days prior to tissue sampling, BM and CL represent body size on the day of tissue sampling, and t represents time (10 or 11 days). Condition index (CI) was calcula ted as Fultons K (CI = BM /CL3, Ricker 1975). Data for BMf, en because of their te ndency to thaw quickly. Frozen subsamples of blood and ground heart and liver tissue were homogenized using QI Ashredder spin columns (Qiagen Inc.). RNA was then isolated with RNeasy Mini kits (Qia gen Inc.) using the manufacturers protocol for isolation of total RNA from animal tissues. DNA and RNA were isolat ed separately from a minimum of three subsamples of each tissue from each turtle for a total of at least 18 subsam for each of 27 turtles unless tissue mass was insufficient. The concentrations of DNA and RNA in each su bsample of blood, heart, and liver tissue were determined us nd a RiboGreen RNA-specific quantitation kit with DNase I (Invitrogen Corporation) by measuring fluorescence at standard fluorescein wavelength settings (485 nm excitation, 5 emission) using a fluorescent microplate read er. Data were collected using KCJuniorTM data analysis software. In addition, protein concentra tion of liver (but not heart or blood) was quantified for a separate project (Chapter 4), so I included those data in the analys is of this study. Hepatic pr concentrations were determined by Br adford assay (details, Chapter 4). Statistical Analyses D rates (SGR) for each t ions: SGRbm = (ln[BMf] ln[BMi])*100/t cl f i ii ff ff 58

PAGE 59

CLf, CI, SGRbm, SGRcl, RNA concentration of each tissue, DNA concentration of each tissue RNA:DNA ratio of each tissue, liver protein concentration, liv er protein:DNA ratio, and liver RNA:protein ratio were compared for the three feeding treatment s using analysis of variance (ANOVA). All ratios were calculated as the quo tient of the average protein, DNA, and/or RN concentration for a particular tis sue. RNA:protein ratios should reflect RNA content per cell, but only if protein:DNA ratios (as a measure of cel lular protei A n content) are consistent among treatm ed hitney tests with a Bonferroni adjustment for multiple comparisons. If ANO s l assays, coefficients of determined for RNA and DNA concentrations in liver, heart, and blood and fo I, SGRbm, SGRcl, [RNA] in each tissue, [DNA] in each tissue, [RNA]:[DNA] in each tissue, liver [protein], liver [pr in]. Regression models for SGR and SGR were then developed using CI and biochemical indices as independent variables. Although body length has been correlat ed with RNA:DNA ratios in fish (e.g., Rooker et al 1997), I did not include treat ment group or any measure of total body size as independent ent groups. All data were tested for normality (using Shapiro-Wilk test) and homogeneity of variances (using Levenes test) prior to parametric analys is and transformed, if n ecessary, using a natural log, reciprocal, square root, square reciprocal square, or reciprocal square r oot transformation. If transformation did not improve normality, data were tested using a Kruskal-Wallis test follow by pairwise Mann-W VA revealed a significant diff erence, pairwise comparisons were evaluated using Tukey Honestly Significant Difference post hoc test (if va riances were equal) or Tamhanes T2 post hoc test (if variances were unequal). To evaluate repeatabili ty of biochemica variation (C.V.) were r protein con centration in liver. Spearmans rank correlation test was used to te st the strength of th e relationships among BMf, CLf, C otein]:[DNA], and liver [RNA]:[prote bm cl 59

PAGE 60

variables in my linear regression models. I chose not to include BM or CL as variables because body size was strongly affected by diet treatment (Fig. 3-1), and the goal of this study was to assess the applicability of RNA:DNA measurements in estimating recent growth rates of wild turtles with unknown dietary histories. Regression equations for SGRbm and SGRcl versus each biochemical index were determined using least squares linear regression. Data were natural log-transformed, if nece to linearize them an ssary, d to decrease heteroscedas ticity. I verified the assumptions of linear regres Results y (Fig. size (Fig. 3-1 and Table 3-1). At the time of tissue sampling, AL turtles were significantly sion by visually inspecting plots of Stude ntized deleted residuals versus standardized predicted values. To construct comprehensive growth models for predicting SGR, data were analyzed using stepwise multiple linear regressi on. The same transformations used for linear regressions were used for stepwise multiple linear regressions. Condition index and all biochemical indices measured for a particular tiss ue (liver, heart, or blood) were included in separate models. A growth model incorporatin g condition index and al l biochemical indices measured for all tissues was also constructed. To enter a model, variables had to meet a 0.05 significance level. All statistica l tests were performed using SPSS for Windows (Release 11.0.0). Means are reported standard e rrors with alpha set at 0.05. When fed to satiation, green tu rtle juveniles in th e final 10-11 days of the 12-week trial grew at an average SGRbm of 1.84% and 2.01% per day and an average SGRcl of 0.68% and 0.64% per day for AL and R-AL individuals, resp ectively. Food-restricted turtles grew much more slowly at an average SGRbm of 0.34% per day and an average SGRcl of 0.15% per da 3-1 and Table 3-1). Intake and growth patterns significantly affected all morphometric measurements of body 60

PAGE 61

heavier and longer than both R-AL and R turtles, and R turtles were significantly lighter and shorter than both AL and R-AL turtles. Despite differences in body size between AL and R-A turtles, condition indices for these two groups we re comparable at the time of tissue sampling, and CI of R-AL turtles was significantly L greater than CI of R turtles. The difference in CI betwe ere ere s and 3-3 n t Rbm rather than as SGRcl. In some cases (e.g., [protein]liver, [RNA]heart, [RNA]:[t, and [RNA]:[DNA]blood), the correlations betwe r l NA]heart, and [protein]:[DNA]liver yielded significant relationships (Table en AL and R turtles approached significance (p = 0.057). Although R-AL turtles w food-restricted for the first five weeks of the study, they grew more ra pidly than AL turtles during weeks seven through ni ne after the switch to ad libitum feeding, but this period of compensatory growth ended prior to tissue sampling (Chapter 2). As a result, R-AL turtles w growing at comparable rates to AL turtles durin g the last 10-11 days of the study, and R turtle were growing significantly slower than both AL and R-AL turtle s. Significant differences among treatment groups also existed for many of the biochemical indices I measured (Figs. 3-2 and Table 3-1). The patterns exhibited by the various biochemical indi ces varied depending o the tissue analyzed, particularly for [RNA] and [RNA]:[DNA] ratios. Many of the morphometric and biochemical i ndices I measured demonstrated significan positive or negative correlations (Table 3-2). For most indices, significant correlations with growth rates were stronger when growth was expressed as SG DNA]hear en indices and growth were significant only for SGRbm. Heart yielded the lowest and live yielded the highest number of significant correl ations between morphometric and biochemica indices and growth. When SGRbm and SGRcl were regressed against each index independent ly, all indices except [RNA]blood, [D 61

PAGE 62

3-3). The R2 values for significant relationships ra nged from 0.161 to 0.659, with the best fits achieved by regressing SGRbm and SGRcl against [RNA]liver. Stepwise multiple linear regression analyses fo r each individual tissue yielded a series of nine significant growth models (Table 3-4). SGRbm was the dependent variable for five mo with two models (1-2) based on liver, one model (3) based on heart, and two models (4-5) ba on blood. SGRcl was the dependent variable for the final four models, with one model (6) ba on liver, one model (7) based on heart, and tw o models (8-9) based on blood. The significa independent variables predicting grow th rate in each of these equa tions are listed in the table in the order in which they were selected by the models. When condition index and all biochemical i ndices for all tissues were combined and analyzed using stepwise multiple linear regression, the resulting models were identical to models 1 and 2 (for SGRbm) and model 6 (for SGRcl). The growth equation that best estimated recent growth rate was Model 2. Desp ite the strong coeffi cient of determinati dels, sed sed nt on for several SGR mode of stimating growth would pr ovide a less intensive alternative to tag and recap acid t of tissues increases with feeding and growth in many marine organisms including krill (Shin et al 2003), cephalopods (Melzner et al 2005, Vidal et al. 2006), tuna (Carter et al 1998), ls, coefficients of variation for RNA and DNA concentrations in liver, heart, and blood and for protein concentration in liver (Table 3-5) were fairly substantial, indicating a high degree interassay variation. Discussion The purpose of this study was to evaluate the use of morphometric and biochemical indices for predicting recent growth rates in juvenile gree n turtles. Validation of assays with substantial predictive power for e ture programs and facilitate population monito ring in this endangered species. Nucleic concentrations and ratios hold promise as potentia l biomarkers of recent growth, as RNA conten 62

PAGE 63

haddock (Caldarone 2005), flounder, and tautog (Kuropat et al 2002). Given the applicability tissue nucleic acid content to gr owth studies in these organism of s, I expected to find strong positi t AL or R-AL turtles. Conversely, heart RNAcl) as A and tration is a measure of the dens ity of nuclei and ther efore correlates with cell number, I conclude that tota l blood cell count increases in re sponse to food restriction. It is uncle ve correlations between growth, RNA and/or protein concentrations, and ratios among nucleic acids and protein concentrations in green turtles. Contrary to my expectations, the biochemical indices I measur ed were neither consistently, nor always positively, correlated with feeding treatment and growth rates. Perhaps most surprisingly, liver RNA concentrat ion was inversely correlated with SGR. I therefore infer tha slow-growing R turtles had more total RNA, and consequently higher pu tative protein synthesis capacity, per unit of liver wet mass than fast-g rowing concentrations in this study we re positively correlated with SGRbm (but not with SGR expected, although this relationshi p was not strong. Growth rate ha d no apparent correlation with blood RNA content. The pattern between DNA and growth rate was quite different from that between RN growth rate. Concentrations of DNA in blood and liver (but not in heart) were both negatively correlated with SGR, a trend that has al so been noted in fish (Mercaldo-Allen et al. 2006). Because DNA concen ar which of the six predominant types of nucleated blood cells in green turtles (Wood and Ebanks 1984) accounts for this increase in blood cell number. The typical hematological response to caloric restriction is either no change (Lochmiller et al. 1993) or a decrease (Maxwell et al. 1990b, Walford et al. 1992) in total leukocyte count, although the number of circulating basophils and thrombocytes has been shown to increase in food-restricted birds (Maxwell et al. 1990b, Maxwell et al. 1992). My DNA results could al so reflect differences in 63

PAGE 64

red blood cell counts. Hematocrit may be correlat ed with body size in green turtles (Wood and Ebanks 1984, but see also Bolten and Bjorndal 1992) but this relationshi p (if it exists) sho result in elevated DNA concentrations in larger, rather than smaller, tu rtles. It is therefore unlikely that my results for DNA concentra tion reflect a bo uld dy size-d ependence for this param nced Beese 2002). Although I did not examine livers from my animals histol ater e tudies, itively d in eter. Hematocrit does not normally incr ease during food rest riction (e.g., Maxwell et al. 1990a, Lochmiller et al. 1993). However, Maxwell et al. (1990a) demonstrated that enha erythropoiesis with concomitant mi crocytosis occurs in food-restricted birds, suggesting that my DNA results may reflect differences in blood cell size between slowand fast-growing turtles. My DNA results also indicate that liver gr owth results from hypertrophy more than hyperplasia in fast-growing green turtles. In rept iles, feeding has been shown to increase the size of lipid droplets and glycogen de posits in hepatocytes, thus lead ing to hypertrophic growth of liver cells (Starck and ogically, farm-rais ed marine turtles fed ad libitum are known to have hepatocytes dominated by large lipid droplet s (Solomon and Tippett 1991). I surm ise that a process of lipid and glycogen deposition similar to that observed by Starck and Beese (2002) occurs to a gre extent in green turtles feeding ad libitum than in food-restricted turtles, therefore leading to more extensive hepatocyte hype rtrophy in the former. Increased lipid deposition in hepatocytes of fa st-growing green turtles may also explain th negative correlation I observed be tween hepatic protein concentrati on and SGR. In other s however, overall protein content as well as prot ein content per cell was strongly and pos correlated with growth rate (Carter et al 1998, Caldarone 2005). To explore the mechanistic basis for the discrepancy between my nucleic aci d and protein concentrat ions and those foun comparable studies using fish, I assessed cellula r protein synthesis capacity by calculating ratios 64

PAGE 65

of RNA:protein (for liver only) and RNA:DNA (for all tissues). These ratios should both provid information about the protein synt hesis capacity per cell, but the former index is only valid as a measure of cellular RNA content if the protein:DNA ratio (as a measure of protein content cell) is unaffected by intake and growth. Because hepatic cellular protein content was in by treatment, only RNA:DNA ratio is an appropria te index of cellular RNA content for this study. Many authors have demonstrated significan t positive relationships between RNA:DNA ratio (of muscle, liver, or whole body) and growth rate, particular ly in fish (Westerman and Hol e per fluenced t 1994, ) ted with ons for wing turtles were under-estimates of the true cellul Carter et al 1998, Caldarone 2005, Mercaldo-Allen et al 2006). Given this common result, I expected to find simila r trends in my turtle tissues Indeed, heart and blood RNA:DNA ratios did correlate positively with growth, but they explained only a small percentage (16-28% of the variance in SGR. On the cont rary, hepatic RNA: DNA ratios were inversely correla SGR and explained 29-34% of the variance in SG R. I suggest several possible explanati this discrepancy among tissues. The liver is a mitotically active tissue, and elev ated rates of cellular proliferation can lead to over-estimation of cell number (Darzynkiewicz et al. 1980). It is possible, therefore, that the RNA:DNA ratios I calculated for liver of fast-gro ar RNA content in fast-growing turtles. However, the difference in these ratios between fast-growing turtles in groups AL and R-AL and slow-growing turtles in group R is likely too substantial to result from differences in rates of DNA synthesis alone. In stead, I suggest that my RNA:DNA ratios refl ect real, tissue-specific differences in cellular ribosomal RNA content. Typically, RNA:DNA ratio declines as ribos omes are degraded during periods of food deprivation (Clemmesen 1994). In my study, ho wever, slow growth was induced by food 65

PAGE 66

restriction rather than starva tion, and food-restricted turtles were never in negative energy balance. Studies in rodents have demonstrated that protein turnove r rates increase in response to caloric restriction and that en zymes involved in gluconeogenesis are upregulated in the liver (Spindler 2001, Hagopian et al 2003). A similar upregulation of metabolic enzyme production may have occurred in the livers of my food-restri cted turtles. Thus, the effect of intake and growth rates on RNA:DNA ratios may differ depe nding on whether the individual is in positive or neg a n growth rate. Although a model for juvenile green turtle growth in the er ng liver ative energy balance and the phys iological role of the tissue studied. To expand the predictive power of the various indices I measured, I incorporated condition index and all biochemical indices measured for a pa rticular tissue (liver, h eart, or blood) into series of models using stepwise multiple linea r regression. The resulting predictive equations explained a maximum of 68% of the variance in gr owth rate. This maximal predictive power was achieved by model 2, in which SGRbm is estimated using liver RNA content and CI. The remaining indices, including nucle ic acid concentrations and ra tios for heart and blood, did not explain any additional variance i Caribbean has previously been developed (Bjorndal et al 2000), this model uses body size to predict recent growth and th erefore does not allow for discrimination of growth rates among individuals of similar size that may have experienced different nutritional conditions. Furthermore, the coefficient of determination for my Model 2 was greater than that of the earli model and therefore indicates that the combined use of morphometric and biochemical indices holds promise for applications to studies of growth in wild populat ions. Specifically, using nucleic acid content to predict growth rates (for body mass) increased the coefficient of determination from 38% using CI alone to 55% using blood DNA content or to 68% usi 66

PAGE 67

RNA content. Predictive power is therefore subs tantially improved by incorporating bioche indices into growth models. In the various growth models I tested, CI was repeatedly selected as an independent variable with significant pr edictive power. Bjorndal et al (2000) found a similar positive correlation between condition index and recent growth rates in wild green turtles. These find are particularly interesting in light of criticis ms of the use of ratio-b ased indices (Hayes and Shonkwiler 2001) and suggest that, at least for green turtles, the use of bod mical ings y condition as measu n leic ndition concentrations is not an acceptable practice, the multivariate model that best predicted recent red using Fultons K (Ricker 1975) for predictive purposes is meaningful and appropriate. The growth model I developed fails to explai n 32% of the variance in growth rates. A portion of this unexplained variability probably results from fairly large coefficients of variatio for the biochemical assays I performed. This va riation could potentially have been improved by measuring DNA and RNA concentrations from the same subsamples of ti ssue, but the nuc acid isolation kits I used precluded me from doing so. Additionally, a number of nucleic acid quantification tec hniques are available (Caldarone et al. 2006), and it is possible that one of these techniques might have allowed for improved pr ecision in measuring DNA and RNA content. The remaining unexplained variability in growth rate may result from a mismatch in the time scales over which the various indices in the model accura tely detect changes in growth. As co index relies on measurement of bod y mass (a result of tis sue accretion) and body length (a result of bony growth), it most likely provides informa tion about longer term growth processes than nucleic acid and protein concentr ations, which presumably fluctuate over shorter time scales (Ferron and Leggett 1994). Because sacrificing wild green turtles to coll ect liver samples for measuring nucleic acid 67

PAGE 68

growth (model 2) has limited applicability in studies of wild turtle demography. However, the fact that several biochemical indices for blood (including DNA concentration and RNA:DNA ratio) says ss C. my were significantly correlate d with growth suggests that furt her calibration of these as for application to growth estimation is warranted. Indeed, 55% of th e variance in body ma growth was predicted using only CI and concentrat ion of DNA in the blood. This coefficient of determination represents a lo ss of only 13% of the maximal predictive power achieved by the best model I developed. Both CI and blood DNA content are easily measured with limited disturbance to the animal. In combination with morphometric measurements, the blood cells of das may therefore allow for the development of minimally invasive techniques for estimating recent growth rates in this endangered species. 68

PAGE 69

Table 3-1. Omnibus F 2, and p-values for comparisons of means among treatment groups for variable). the various morphometric and bi ochemical indices measured (n = 27 for each Groups Tested in Pairwise Comparisons Omnibus F and 2 AL and R-AL AL and R R and R-AL Body Mass F2,24 = 99.476, p < 0.0001 < 0.0001 < 0.0001 < 0.0001 Carapace Length 2 = 22.013, p < 0.0001 < 0.001 0 < 0.001 0 < 0.000 Condition Index F2,24 = 6.918, p = 0.004 0 < 0.6840 < 0.0570 < 0.0040 SGRbm F2,24 = 28.863, p < 0.0001 < 0.6160 < 0.0020 < 0.001 0 SGRcl F2,24 = 62.995, p < 0.0001 < 0.7370 < 0.0001 < 0.0001 [RNA]liver F2,24 = 28.953, p < 0.0001 < 0.9460 < 0.0001 < 0.0001 [RNA]heart F2,24 = 6.076, p = 0.007 0 < 0.0210 < 0.9900 < 0.015 [RNA]blood F2,24 = 1.946, p = 0.165 0 N/A N/A N/A [DNA]liver F2,24 = 12.349, p < 0.001 0 < 0.1540 < 0.001 0 < 0.0100 [DNA]heart F2,24 = 2.111, p = 0.143 0 N/A N/A N/A [DNA]blood F2,24 = 5.278, p = 0.013 0 < 0.3120 < 0.0100 < 0.162 1 0 0 [ RNA ] : [ DNA ] live r F2 24 = 6.546 p = 0.005 0 < 0.1910< 0.3190 < 0. 0040 [ RNA ] : [ DNA ] hear t F2 24 = 4.678 p = 0.019 0 < 0.240 [RNA]:[DNA]blood F2,24 = 5.224, p = 0.013 0 < 0.3 0< 0.5240 < 0.0150 030 < 0.0100 < 0.1720 [Protein]liver F2,24 = 3.545, p = 0.045 0 < 0.3200 < 0.6300 < 0.0360 [Protein]:[DNA]liver F2,24 = 7.655, p = 0.003 0 < 0.0120 < 0.0030 < 0.7930 [RNA]:[Protein]liver F2,24 = 13.438, p < 0.001 0 < 0.8880 < 0.0030 < 0.0030 When F values are reported, data were analyzed using analysis of variance and pairwise comparisons were evaluated using Tukeys Honestly Significant Difference or Tamhanes T2 post hoc tests. When 2 values are reported, data were analyzed using a Kruskal-Wallis test with pairwise Mann-Whitney U tests. Statistically significant p-values are indicated in bold. Abbreviations: SGR = specific growth rate, bm = body mass, cl = carapace length, [ ] = concentration, AL = ad libitum for twelve weeks, R-AL = food-restricted for five weeks and ad libitum for seven weeks, R = food-re stricted for twelve weeks. 69

PAGE 70

Table 3-2. Spearmans rank correlations ( ) for morphometric (a) and biochemical indices for liver (b), heart (c), and blood (d) ( n = 27 for each variable). ometrics (a) Morph Variable CL CI SGRbm SGRcl BM .0.988** 0.463* 0.693 0.797** CL 0.391* 0.649 ** ** 9** ** 0.785 CI 0.49 0.465 SGRbm 0.839 (b) Li Var[RNAver ]liver ver iable ]liver [DNA]liver R:Dli [Protein P:Dliver R:Pliver BM .-0.6 2* ** 75** .-0.673** .-0.41 -0.287 0.428* a -0.673** CL .-0.6 3* ** .-0.5 8** 35** a .-0 S .-0.7 8** ** .-0.7 1** 32** ** .-0 [R 8** 6** 3* .** 54** .-0.635** .-0.42 -0.240 0.426* a -0.675** CI 21** .-0.534** -0.16 -0.1 0.427 .487* a GRbm 34** .-0.519** .-0.61 -0.561 0.068** -0.536** SGRcl 33** .-0.442* .-0.60 -0.3 0.248 .690** NA] liver[DNA] .0.609** .0.77 0.433 -0.292** 0.846** liverR:D 0.06 0.42 -0.615** 0.459* a liver[Protein] 0.24 9 0.134** 0.703** liverP:D 0.332** -0.051** liver -0.545** (c) Heart Variable [RN A ] [DNA] R:D heart heart heart BM 0.031** -0.394* 0.247** CL -0.033** -0.369** 0.195** CI 0.429* -0.454* 0.631** SGRbm 0.435* -0.292** 0.523** SGRcl 0.294** -0.217** 0.374** [RNA]heart -0.062** 0.800** [DNA]heart -0.594** (d) Blood Variable [RNA]blood [DNA]blood R:Dblood BM 0.357** -0.563** 0.549** CL 0.339** -0.520** 0.527** CI 0.543** -0.237** 0.441* SGRbm 0.352** -0.549** 0.440* SGRcl 0.284** -0.446* 0.352** [RNA]blood -0.253** 0.816** [DNA]blood -0.680** Significant correlations are indicated in bold. Asterisks indica te level of significance (* p < 0.05, ** p < 0.01). Abbreviations: BM = body mass, CL = carapace length, CI = condition index, SGR = specific growth rate, [ ] = concentration, R:D = RNA:DNA ratio, P: D = protein:DNA ratio, R:P = RNA:protein ratio. 70

PAGE 71

71 x Intercept slope Adjusted R2 F p Table 3-3. Growth equation parameters for juvenile Chelonia mydas as determined by least squares linear regression. y Bo bm 40 -1.132 45.011 < 0.0001bm NA]heart 161 05.990 < 0.0220bm [RNA]bl96 0.010 04.197 < 0.0510bm Ln[DNA] 8.574 -1.276 12.482 < 0.0020 Sbm Ln[DNA]heart 6.808 -0.967 03.536 < 0.0720bm Ln[DNA]blood 12.305 -1.828 12.016 < 0.0020 Ln(R:D)liver 1.610 -1.046 11.821 < 0.0020bm Ln(R:D)heart 3.054 0.852 0.253 09.804 < 0.0040 R:D)b22.06 .0030bm0bm ) 1 3 0bm ) 387 0 0 dy Mass SGR SGR Ln[RNA] Ln[R liver 8.0 -2.133 0.629 0. 0.955 SGR SGR ood 0.1liver 0.109 0.306 GR SGR 0.089 0.298 SGRbm 0.294 SGR SGRbm Ln( lood .7 0 1 1 0.277 10.937 7.353 < 0 < SGR S [Protein] liver 2.878 -7.203 0.196 0 1.08 0.012 .308 <. GR SGR Ln(P:D Ln(R liver 1.5 1 0.470 4 0.003 0.366 0 15.993 < 0 0001 <. :P liver 1.9 -0.9 40 38.475 0.379 16.897 SGR bm CI -3.9 0 001 Carength L[RNA 4353 1 Ln[RNA]t -0.329 3.382 0 A]blood 0.171 0.003 1 0 Ln[DNA]liver 2.610 -0.379 0.289 0 Ln[DNA]heart 1.984 -0.269 0.067 2.882 .1020cl Ln[DNA]blood 3.486 -0.504 0.236 09.052 0 ) 5339 0.337 14.228 < 0.0010cl ) 9 0.164 06.107 < 0.0210cl )b84 0.222 08.426 < 0.0080cl n] 0-1.584 0.083 03.357 < 0.0790 R )li53 0.047 02.283 < 0.1430l )li66343 0.490 25.954 < 0.0001 CI -0.903 9.951 0.262 10.245 < 0.0040 apace L SGRcl n ]liver 2.5 9 -0. 0.659 51.159 <.000 0 < .078 SGRcl hear 68 0.2 0.084 0 0 < .106 SGRcl [RN 0.065 02.81 0 SGRcl 11.554 < 0.002 SGRcl 0 < 0 < 0.006 SGR SGRcl Ln(R:D liver 0.5 0 -0. SGR Ln(R:D heart 0.8 9 0.217 SGR Ln(R:D lood 0. 6 0.294 2 SGR [Protei liver 0.8 SGcl R Ln(P:D ver 0. 5 0.203 SGcR Ln(R:P ver 0. 9 -0 SGcl S th rate for body mass or carapace lengt ressed independently against each 27 for each variable). Significant re indicated in bold. Abbreviations are the sa pecific grow h was reg index ( n = p -values a me as in Table 3-2.

PAGE 72

parameters for juvenile Chelonia mydas as determined by stepwise multiple linear re Tissue Model # y x1 x2 intercept 1 2 R2 F l gression. p-va ue Liver 1 SGRbm Ln[RNA] 8.040 -1.132 0.629 45.0 Liver 2 SGRbm Ln[RNA] CI 4.316 -0.913 17.689 0.680 28.5 Heart 3 SGRbm CI -3.940 38.475 0.379 16.8 Blood 4 SGRbm CI -3.940 38.475 0.379 16.8 01 0 Blood 5 SGRbm CI Ln[DNA] 5.384 31.770 -1.402 0.547 16.7 001 11 < 0.0 67 < 0.0 97 < 0.0 97 < 0.0 05 < 0.0 001 001 01 0 Liver 6 SGRcl Ln[RNA] 2.549 -0.353 0.659 51.1 001 Heart 7 SGRcl CI -0.903 9.951 0.262 10.245 04 Blood 8 SGRcl CI -0.903 9.951 0.262 10.245 04 Blood 9 SGRcl CI Ln[DNA] 1.730 8.058 -0.396 0.398 09.950 01 59 < 0.0 0.0 0.0 0.0 Specific growth rate for body mass or carap ace length (the dependent variables) wa s regressed against condition ind biochemical indices for a particular tissue (liver, heart, or blood) ( n = 27 for each variable). When specific growth raressed against condition index and biochemical indice s for all tissues together, the resulting m odels were identical to model SGRbm) and model 6 (for SGRcl). Variables are listed in the order in which they were selected by the models. Signifi a indicated in bold. Abbreviations are the same as in Table 3-2. ex and te was reg s 1 and 2 cant p -val (for ues re 72Table 3-4. Growth equation

PAGE 73

Table 3-5. Coefficients of va riation (C.V.) for RNA, DNA, and protein concentrations of Chelonia mydas tissues. Tissue Assay .V. (%) C Liver [RNA] 33.2 Heart Blood [RNA] [RNA32.2 20.0 Li[DNA] Heart [DNA] 35.2 Blo[DNA] Liver [Protein] 11.0 ] ver 32.4 od 38.9 Valuent aver.s fori du for eay. Am replicates for each individual for each assay was performed unless sample mass was insufficient. es repres ages of C.V 27 indiv al turtles ach ass minimu of three 73

PAGE 74

0 2 400 6Body Mass (g) a b 00 00 c 0 100 1 200 Length (mm) a b c 50Carapace 50 0.S 0 0 2.0 0G BM payb 1.R for 3.) (% er d a a 0.0 6 8 0SG ey) 0.2 0.4R for CL b 0. (% p 0.r da 1. a a 0.00 0.05 0.10 0.15 0.20 0.25Condition Index (g/cm3)ab a b Figure 3-1. Morphometric indices and growth rates for turtles in each of three treatment groups. Turtles in the AL group ( n = 7) were fed ad libitum for 12 weeks. Turtles in the R-AL group ( n = 10) were fed 50% of initial mass-specific AL intake for 5 weeks and then fed ad libitum for 7 weeks. Turtles in the R group ( n = 10) were fed 50% of initial mass-specific AL intake for 12 weeks. Each point represents mean standard error. Means were evaluated using analysis of variance with Tukeys Honestly Significant Difference or Tamhanes T2 post hoc tests or using a Kruskal-Wallis test and pairwise Mann-Whitney U tests with a Bonferroni correction for multiple comparisons. Means that are significantly different at p < 0.05 are indicated by different letters. Abbreviations: SGR = sp ecific growth rate, BM = body mass, CL = carapace length. AL R ALR AL R ALR 74

PAGE 75

Figure 3-2. Nucleic acid indice s for turtles in each of three treatment groups. Each point or. tly different at p < 0.05 are f rea ta analysis, sample sizes, and F re in ab pre dic bre sent ated viat s m by ion ean dif s ar st eren e the and t le sam ard tter e err s. T as in Me ans that are significan tme nts, da igure 3-1. 75

PAGE 76

0 250 500 750 1000[RNA] (ng/mg)a a b 0 200 400 600[DNA] (ng/mg)a a b 0.0 1.0 2.0 3.0[RNA]:[DNA]ab a b Figure 2. 0 20 40 60 80 a b a 0 200 400 600 a a a 0.0 0.1 0.2 0.3 ab a b 0 50 100 150 200 a a 0 200 400 600 a a ab b 0.0 0.2 0.4 0.6 a b R BLOOD ab R A L AL ALR ALR ALR ALR HEART LIVER 76

PAGE 77

77 0.0 0.1 0.2 0.3 0.4[Protein] (mg/mg)ab a b 0.0 0.5 1.0 1.5[Protein]:[DNA] (mg/ g)a b b 0 1 2 3 4NA]:[Protein] ( g/mg)a a b [R tein and protein-based indices for turtles in each of three treatment groups. Each point represents mean standard error. Means that are significantly different at p < 0.05 are indicated by differe nt letters. Treatments, data analysis, sample sizes, and abbreviations are the same as in Figure 3-1. Figure 3-3. Liver pro AL R AL R LIVER

PAGE 78

CHAPTER 4 NSATORY GROWTH AND ANTIOXIDANT STATUS IN JUVENILE GREEN TURTLES (C helonia mydas) COMPE Introduction of depressed growth can have profound Poor early nutrition and resulting periods fe-history consequences, some of which extend into subsequent generations (as reviewed by Metcalfe and Monaghan 2001). Slow growth that re sults in small size at weaken dominance status (Richner et al 1989), impede the establishm breeding territories (Einum and Fleming 2000 and references therein), al dimorphism, or delay maturation and the ons et of reproductive competence (Altm Alberts 2005). The nutritional envi ronment an animal experiences early in fecundity (Nagy and Holmes 2005), and offspring size (Reznick et al 1996). Given the negative effects of slow growth and small size on performa ultimately fitness, selection should favor compensato ry strategies that allo rapidly when conditions improve (Metcalfe and M onaghan 2001). Such a period of catch up or compensatory growth (CG) has been demons trated in a number of organisms ( Osbourn 1960, Ali et al 2003, Bjorndal et al 2003, Jespersen and Toft 2003). Compensatory growth is characterized by grow th rates greater than those of consistently well nourished conspecifics of the same age and can result in comparable body sizes for individuals with drastically different dietary histories (Metcalfe and Monaghan 2001). The benefits of a rapid increase in size include improved short-term survival due to decreased size-specific mortality (e.g., due to predation) (Arendt 1997, Metcalfe and Monaghan 2003) and enhanced reproductive li a particular age or developmental stage can increase vulnera bility to predation (Arendt 1997, Janzen et al 2000), ent of feeding and/or ter patterns of sexual ann and its life can also adversely affect adult body size (Madsen and Shine 2000), survival (McD onald et al 2005), nce, survival, and w individuals to grow Wilson and 78

PAGE 79

79 output, especially for organisms (e.g., ectotherm s) in which fecundity is proportional to body erated growth, the occu rrence of CG suggests ys optim al under conditions of ample food availability. bility is high, presumably reflect a balance ween the benefits and costs of rapid growth. Structures formed during periods of fast growth may be prone to weakness, as is the case for bird primary feathers (Dawson et al 2000) and leg well as fish scales (Arendt et al 2001). Animals that ha undergone a period of CG can incur costs incl uding increased muscle protein degradation (Therkildsen 2005), decreased muscle mass (Blanger et al 2002), impaired locomotor performance (lvarez and Metcalfe 2005), accele rated telomere shortening, and decreased longevity (Jennings et al 1999). At the cellular level, animals feeding ad libitum (and therefore itochondrial free ra dicals and thus experience m oxidative damage than calorie-restricted anim als (Gredilla and Barj a 2005). It has been suggested (although not tested, to my knowledge ) that the detrimental effects of CG on performance and longevity may result from tran sient elevated rates of free radical-induced cellular damage (Mangel and Munch 2005). In the present study, I manipulated growth trajectories of green turtle ( Chelonia mydas ) juveniles by controlling intake. Before and afte r a demonstrated period of CG, I evaluated scle and liver glutathione peroxidase (GPX) activity and hepatic antioxidant potential (AP). Glutathione peroxidase is the major cytosolic and mitochondrial enzyme that catalyzes the reduction of hydroperoxides into water (Li et al 2000), thereby protecting cells from oxidative damage to protein, lipids, and DNA (Barja 2004) Total AP reflects activity of antioxidant size (Roff 1992). Despite the potential fitness benefits of accel that maximal growth rates are not alwa Sub-maximal growth rates, even when food availa bet bones (Leterrier and Nys 1992) as ve gro wing rapidly) typica lly produce more m ore mu

PAGE 80

e nzymes such as GPX in addition to reduci ng capacity of non-enzymatic antioxidants (Sies 1997). Quantifying GPX activity and AP permits me to assess whether diminished capacity to combat cellular oxidative damage may be a cost of CG in this species. To my knowledge, this study is the first to test the e ffects of CG on these parameters. Materials and Methods Animal Care All animal care components of this study were performed at the Cayman Turtle Farm in Grand Cayman, British West Indies. Chelonia mydas hatchlings were housed individually in 68-L bins of sea water. Turtles were fed turtle pellets (Melick Aquaf eed, Catawissa, PA) twice daily. Turtles in the ad libitum group (AL) were fed ad libitum for twelve weeks. Turtles in the restricted group (R) were fed approximately 50% of the average initial AL intake (on a mass-specific basis) for twelve weeks. Turtles in the restrictedad libitum group (R-AL) were fed the restricted diet for five weeks and were then fed ad libitum for the remaining seven weeks. Turtles were weighed and measured weekly. A dditional details regarding animal husbandry can be found in Chapter 2. Tissue Collection and Homogenization At the conclusion of the fifth week of the experiment (immediately prior to switching All R-AL t the five R and five R-AL turtles we re therefore pooled into one group (t5 R). The remaining rtles seven AL turtles (t12 AL), ten R turtles (t12 R) and ten R-AL turtles (t12 R-AL) were crificed at the conclusion of the twelve-week trial. Turtle s were euthanized with an intramuscular injection of ketamine (Ketaset, 100 mg/kg body mass). R-AL turtles to an ad libitum diet), ten AL (t5 AL), five R, and five RAL turtles were sacrificed. urtles had, until the end of week five, b een maintained on the restricted diet. Data for tu sa 80

PAGE 81

After each turtle had been inje cted w nd was no longer responsive to a pain stimulus, it was decapitated. Porti r and the left pectoral muscle were removed with forceps and snap-froz ogen no more th an three minutes after decap ed 9 es tion ovine ue m reference wave bated ith 10 re ith ketamine a ons of the right lobe of the live en in liquid nitr itation. Tissues were maintained at C until they were homogenized as describ below. Subsamples of liver and pectoral muscle were homogenized in 1.0 ml of Sigma T-678 buffer (0.05 M Tris, 0.138 M NaCl, 2.7 mM KCl, pH 8.0 with 1% bovine serum albumin) yielding a 10% weight:volume tissue solution fo r each sample. Liver and muscle homogenat were further diluted to 10% and 33%, respectivel y, to insure that enzyme activities would be within the range of the standard curves used. Total protein concentration of each tissue solu was evaluated using a Bradford assay with standa rd curves constructed us ing dilutions of b serum albumin (BSA, 2 mg/ml undiluted). Concentr ations of standards (i n duplicate) and tiss solutions (in triplicate) were determined by absorbance at 595 nm with a 695 n length using a microplate reader. Glutathione Peroxidase Activity Assay Glutathione peroxidase activity was evaluated using a total GPX assay modified from Nakamura et al (1974). Diluted muscle and liver homogenate solutions ( n = 46) were incu for three minutes at 25 C in a reaction cock tail containing 0.297 U/ml glutathione reductase, 1.25 mM glutathione, and 0.188 mM NADPH in a 100 mM potassium phosphate buffer w mM EDTA (pH 7.4). T-butyl hydrop eroxide (12 mM) was then a dded to the reaction mixtu and the absorbance of the resulting solution at 340 nm was recorded every minute for four minutes using a microplate reader. A blank consisting of 100 mM potassium phosphate buffer with 10 mM EDTA (pH 7.4) was also assayed to evaluate glutathione-indep endent reaction rates. 81

PAGE 82

Samples and blanks were analyzed in duplicate (f or muscle, because no significant relationship were found) or trip s licate (for liver). y tech s s and ard curve of uric acid (an antioxidant) was const ) leic acid contents from Chapter 3 for live r only (DNA contents of muscle and t12 liver were not Total GPX activity of each sample was calculated by determining the rate of change in absorbance of NADPH ( A340/min, calculating using only linear data) and dividing this value b the extinction coefficient for NADPH (6.22) This quotient was doubled to account for stoichiometry and then multiplied by the final di lution factor. Activity of blanks was likewise calculated and subtracted from each samples activity to yield total GPX activity. Total GPX activity was then normalized to total protein c oncentration of each samp le as determined by Bradford assay. Total AP Assay Overall, nonspecific AP of homogenized liver samples was evaluated using the Bioxy AOP-490 assay (OxisResearch, Portland, OR). This assay evaluates the tota l activity of cellular antioxidants including enzymes (e.g., superoxide di smutase, GPX, and catalase), small molecule (e.g., ascorbic acid), larg e molecules (e.g., albumin), and hormones (e.g., estrogen) (OxisResearch Bioxytech Assay Systems label, 2002). Samples from nine t5 AL, nine t5 R, seven t12 AL, eight t12 R, and eight t12 R-AL turtle standards were analyzed in duplicate. A stand ructed and used to calcula te AP of diluted liver homogenate samples as concentration ( M of copper reducing equivalents (CRE). Total AP was then normalized to total protein concentration of each sample as determined by Bradford assay. Statistical Analyses Data for GPX activity and total AP were norma lized to total protein content as described above. In addition, I also calculated these parameters per g of DNA in t12 turtles using nuc 82

PAGE 83

evaluated for that study). Presumably, expressing GPX activity and AP per mg of protein ref the proportion of total proteins functioning as an tioxidants, whereas expr essing these parameter per g of DNA reflects antioxidant capacity per cell. Data were tested for normality (Shapiro-Wilk test) and h lects s omogeneity of variances (Levenes test) prior to parametric a d a significant result ( p < 0.05), formed using a na tural log, reciprocal, square root square, reciprocal square, or recipr g a ance using analysis of variance (ANOVA). When statis tically significant differences amon nes T2 Results cantly fast er than those in the R group during each week of week nalysis. If either test yielde data were trans ocal square root transformation. If tran sformation did not improve homoscedasticity and Tamhanes T2 post hoc test could not be used, data were tested for statis tical significance usin Kruskal-Wallis test. Otherwise, data within ea ch sampling period were tested for statistical signific g t12 treatment groups were found, pairwise co mparisons were evaluated using Tukeys Honestly Significant Difference (HSD) post hoc test (if variances were equal) or Tamha post hoc test (if variances were not equal). Data were analyzed using SPSS for Windows (Release 11.0.0). For al l reported analyses data are expressed as means standard errors (unless otherwise noted ) with alpha set at 0.05. Turtles in the AL group grew signifi of the study. After the swit ch from a restricted to an ad libitum diet (at the end of five weeks), R-AL turtles grew significantly faster than those in the AL group during weeks 7 through 9. This period of growth compensation ceased prior to the end of the study such that AL and R-AL turtles were growing at comparable rates by the time t12 samples were collected (as reported in Chapter 2). Different growth rate s yielded significantly different body masses for turtles in each treatment sampled at the conclusion of week five and at the conclusion twelve (Fig. 4-1 and Table 4-1). 83

PAGE 84

Diet treatments affected protei n content of muscle (for t5 turtles) and liver (for t12 turtles) expressed as concentration of protein per wet mass of tissue (Tables 4-1 and 4-2). After five weeks of food restriction, protein content of muscle was 59% great er in R turtles than in AL turtles, but this difference decreased after week 5 such that musc le protein content did not di among t ffer reatment groups by week 12 of the study. Hepatic protein conten t only differed at t12, with l ent ver in differe d significantly among treatment groups for both fter five weeks of f ood restriction, R turtles had a highe f and iver of R turtles cont aining 36% more protein than liver of R-AL turtles. Specific activity of GPX in pectoral muscle did not differ significantly among treatm groups at t5 or at t12 (Tables 4-1 and 4-3). Furthermore, interassay variation was quite high for measurements of muscle GPX activity (Table 4-4) However, differences in GPX activity of li expressed per mg of protein approached significance ( p = 0.090) for t12 turtles, with AL turtles demonstrating 20-24% greater hepa tic GPX activity than R and R-AL turtles (Fig. 4-2 and Table 4-1). Because muscle GPX activity was not affected by intake and growth rates, I restricted my analysis of AP to liver. Total hepatic AP per mg of prote t5 and t12 turtles (Fig. 4-2 and Table 4-1). A r hepatic AP per mg of protein than AL tu rtles. After twelve w eeks of food restriction, hepatic AP per mg of protein of R-AL turtles was significantly lower than that of R turtles, and the difference between AL and R-AL turtles approached significance ( p = 0.098). The aforementioned values for GPX activity and AP were calculated by correcting for total protein content of samples. Using DNA content per mg of liver from Chapter 3 as a correlate o cell number in t12 turtles, I also compared puta tive GPX activity and total AP per g of DNA found differences among treatment groups (Fig. 4-2 and Table 4-1). When normalized to tissue DNA content in this way, both GPX activity and to tal AP per cell were si gnificantly higher in t12 84

PAGE 85

AL turtles than in R and R-AL turtles, despite th e fact that AL and R-AL turtles were growing at comparable rates at this time. Discussion The objective of this study wa s to evaluate decreased antioxidant capacity as a possible cost of growth compensation in juvenile green turtles. By manipulat ing food intake duri controlled, twelve-week feed ing trial, I elicited a CG respons e from previously food-restricted turtles after a switch to ad libitum feeding. Because compensating turtles grew faster than continuously ad libitum turtles, I conclude that growth ra tes in juveniles of this species are sub-maximal wh ng a en individuals have continuous access to unlimited food. This finding suggests f fast growth are countered by one or more costs, potentially including the accru r) and vity of liver but not muscle responded to diet, s ion, I GPX was measured for the same t12 e, allowing me to correct my measurements for total DNA content as a putati that the benefits o al of cellular oxidative damage. Alt hough Mangel and Munch (2005) incorporated oxidative stress into their CG m odel, the present study is the firs t to provide empirical evidence linking growth compensation with effects on antioxidant function. To assess antioxidant capacity, I measured the activity of GPX in mitotic (live post-mitotic (skeletal muscle) tissues. Specific GP X acti o I restricted my subsequent analysis of total, non-specific anti oxidant potential to liver. Although the parameters I measured are typically expressed relative to protein concentrat found differences in hepatocyte protein content among treatment groups (Cha pter 3). As a result, differences in protein content per liver cell may have confounded my measurements of activity and AP. In Chapter 3, hepatic DNA content of t12 turtles turtles I examined her ve correlate of cell number. Doing so allo wed me to compare the levels of antioxidant function per cell rather th an assessing the proportion of total pr oteins functioning as antioxidants 85

PAGE 86

In this study, turtles with a dietary history of continuous ad libitum feeding for twelve weeks demonstrated higher hepatic GPX activity relative to turtles ex periencing a continuous food restriction. Qualitatively, I observed the same pa ttern regardless of whether hepatic GPX activi s as one ce of ion X t al 1996). Additionally, activity of antioxidant enzym attenuated by food restriction (Luhtala et al f ty was calculated rela tive to protein or DNA content. Glut athione peroxidase serve of several important enzymes in the cellular anti oxidant defense system, and upregulation of its activity likely reflects increased endogenous production of organic hydroperoxides (Judge et al. 2005). My results therefore suggest that high intake and growth rates cause elevated oxidative stress in mitotically active tissues of juvenile green turtles. The liver plays an important role in metabolis m and detoxification and is a major sour peroxides via autoxidation reactions It is therefore not surprising that I detected an upregulat of a component of the antioxidant defense system in liver of fast-growing AL animals. My GP results parallel the finding that food restriction depressed hepa tic GPX gene expression in young mice relative to ad libitum controls (Mura e es such as GPX typically increases with age due to elevated oxidative stress (Leeuwenburgh et al 1994, Phaneuf and Leeuwenburgh 2002) but this increase is often 1994). A counterintuitive result from the present study is the finding that R-AL turtles had GPX activities comparable to those of continuously food-restricted turt les at the conclusion of the study. Therefore, turtles in the R-AL gr oup either produced fewer hydroperoxides and experienced less oxidative stress than continuously AL turtles during the final weeks of the study or experienced comparable levels of oxidative stress but demonstrated a reduced capacity to upregulate GPX activity. Because R-AL turtles underwent CG seve ral weeks before the end o the experiment and had growth rates similar to AL turtles when tissues were sampled, I suggest 86

PAGE 87

the la e a on individual antiox l synergistically (Niki et al 1995, Bhm et al 1997). For this reason, I measu itional week tter scenario occurred. If my hypothesis is correct, then depressed antioxidant enzyme activity paired with elevated gr owth rates would exacerbate oxidat ive damage and therefore b cost of CG. Although my GPX results appear to support my initial predictions, specific components of the antioxidant defense system often respond di fferently to oxidative stress depending on the species and tissues measured. In addition, conflict ing effects of dietary restriction idant enzyme activities are common. For example, GPX ac tivities have been shown to increase (Agarwal et al 2005), decrease (Grattagliano et al 2004), or remain the same (Wu et a 2003) in food-restricted animals relative to ad libitum -fed controls. Furthermore, individual enzymes in the antioxidant defense system do not function in isolation from each other. Instead, combating oxidative stress requires the concerte d involvement of a variety of enzymatic and non-enzymatic molecules that scavenge or neutralize reactive oxygen sp ecies (ROS). Many of these molecules function red total, non-specific hepa tic antioxidant potential in t5 and t12 turtles and again found differences among treatment groups. When calculated relative to protein content, hepatic AP of R turtles was significantly higher than hepatic AP of AL turtles after five weeks of food restriction. Kalani et al (2006) found similar results in ad libitum -fed versus calorie-restricted rats. After seven add s, the difference between AL and R turtles wa s diminished, but R turtles had higher AP than R-AL turtles and AP of AL turtles was marginally higher than AP of R-AL turtles. These results indicate that diet history affects the proportion of total prot eins that function as antioxidants within the liver. 87

PAGE 88

88 rowing R turtles and fast-growing R-AL turtles. As a result, turtles that underwent an earlier sed cellular antioxidant functi sults ad developm However, when I corrected my values of t12 hepatic AP for DNA content, fast-growing AL turtles had higher AP per cell than both slow-g period of CG had decrea on compared to age-matched AL turtles f eeding and growing at the same rate. Because I did not measure hepatic DNA concentration for t5 turtles, I could not evaluate putative AP content per cell for these individuals. The results of this study imply that individuals typically grow at rates that optimize their ability to prevent oxidative damage to lipids, nu cleic acids, and proteins. Oxidative stress re from an imbalance between the rate of ROS pr oduction and the availabili ty of antioxidants to scavenge these ROS within a cell (Agarwal et al 2005). Given the assumption that ROS production increases with in take rate (Lpez-Torres et al 2002, Barja 2004), my finding that libitum -fed turtles compensating for a prior food re striction also had di minished antioxidant function implies a cost of CG. It is unclear whether elevated oxidative stress during early ent in this long-lived species would adversely affect longevity or performance. However, this study provides evidence of cellula r stresses coincident with growth compensation and suggests that sub-maximal growth protects individuals from the de trimental effects of impaired antioxidant defense.

PAGE 89

F a p g tmt groups iveet5)d tveek12). 12 GroupesnwCas nibus 2nd -values for comparisons of means amon treaen at fe wks ( anwel wes (t t s Tted i Pairise omprison Onib F 2 A and mus and L R AL and R-AL R and R-AL BM F =1 <0 F =7 <0 010101 Protein Content, Muscle F =5 =2 F =9 =0 N = 8 F =4 =4 0 0 0 0 0 GPX Specifici le F =8 F =0 N GPX Specifi F =8 F =9 N TAxidant Potentl, Liver F =0 F =1 0 0 0 GPX Activit ody ass t5 t12 t5 t12 Protein Content, Liver t5 t12 Actvity,Musc t5 t12 c Activity, Liver t5 t12 otal ntio ia t5 t12 1,18 99.57, p 0.001 2,24 99.46, p 0.001 <.000 1,18 5.92, p 0.05 0 2,24 0.68, p 0.57 0 N/A 2 2.26, p = 0.131 0 2,24 3.55, p 0.05 0 .630 1,18 = 0.507, p 0.46 0 2,24 = 0.226, p 0.80 0 N/A 1,18 = 0.811, p 0.30 0 2,24 = 2.660, p 0.00 0 N/A 1,16 = 10.443, p 0.05 0 2,20 = 5.135, p 0.06 0 0.697 < .000 < .000 N/A /A .320 0.036 N/A /A N/A /A .098 0.015 y Per g DNA Live r 2 F t12 24 9.67 =1 p < 01 0 0. Axi Pte D, r F 0 0 0 0 0 0.0 0 0.001 2,20 = 6.416, p = 0.007 0 .029 .003 937 .008 0.808 ntiodantotenial Pr g NALive t12 Whe values are reported, data we yzed usanisva pars 2 pleaed usingkHtga ifnce posc n aa, data re ana a laricated bold. Abbreviations: AL = lim for tw s= -rico sews,X = gluoera isonof t1sames wre evluat welyzed using a Kruskal-Wllis elveweek, R foodestrted fr tathine poxidse. n F er anal Tueys onesly Sinificnt Dfere test. Statistically significant p-vaues e ind twelve weeks, R-AL = food-restricted for five weeks and ing alys of rian ce, and pairwise com t hotestsWhe 2 vlues re reported in ad bitu ad libitum for ven eek GP 89Table 4-1. Om

PAGE 90

Table 4-2. Total protein concentrations of Chelonia mydas muscle and liver homogenates as Treatment Week n Protein Concentration, Muscle Protein Concentration, Liver determined by Bradford assay expressed rela tive to wet mass of homogenized tissue. Group (mg/mg tissue) (mg/mg tissue) AL 5 10 0.198 0.023a 0.178 0.021ay Rx xy 5 10 0.314 0.040b 0.241 0.021ay AL 12 7 0.173 0.018 0.218 0.024 R 12 10 0.187 0.026x 0.242 0.014x y R-AL 12 10 0.211 0.021x 0.178 0.018y x Values represent means standard errors. Treatmen t groups and data analysis are the same as Table 4-1. Letters (a and b for week 5, x and y for week 12) indicate st atistically significant differences ( p < 0.05) among treatment groups within sampling periods. in 90

PAGE 91

Table homogenate. Group (nmol/min*mg protein) 4-3. Glutathione peroxidase (GPX) specific activity in Chelonia mydas muscle Treatment Week n GPX Activity, Muscle AL 5 10 2.78 0.16 R 5 10 3.07 0.37 AL 12 7 2.68 0.12 R 12 10 2.66 0.11 R-AL 12 10 2.78 0.18 Value Table s represent means standard errors. Treatmen t groups and data analysis are the same as in 4-1. No significant differences in GPX activity of muscle tissue were detected among treatment groups in either sampling period. 91

PAGE 92

Tab le 4-4. Coefficients of vari ation (CV, %) for protein concen tration, glutathione peroxidase (GPX) activity, and antioxidant potential (AP) assays. Protein Content GPX Activity AP Tre a Gro tment up Week Liver Muscle Liver MusclLiver e A R AL R R-A L 5 5 08 .0 6.7 26.9 1 10.8 9.1 06.3 38.2 19.2 12 09.9 07.7 06.3 27.7 12.2 13.8 .8 08.0 24.9 12.3 6.4 24.0 2 09.2 0 0 9.1 L 12 1 10 0 2 09 .0 7.7 0 4.3 Eac gro per dup h up form lic valu s ar e repnts the averag individu al CVs for eaurtle a eatment same as in Table 4-1. Protein tent and hepatic GPX act ere ed in triplicate, whereas muscle GPX activ ity and hepatic AP assays were performed in rese e of ch t nd ea ch assay. Tr ivity assays w e the con ate. 92

PAGE 93

0 3 4t5 t12Body Mass (g) 100 200 00 00 500 AL R R-AL a b c a b Figure 4-1. Body mass of tu rtles at five weeks (t5) and twelve weeks (t12), when tissues were sampled. Each point represents mean standard error. Sample sizes: n = 10 for all groups except t12 AL ( n = 7). Treatment groups and data analysis are the same as in Table 4-1. Letters (a, b, and c) indicate statistically significant differences ( p < 0.05) among treatment groups within sampling periods. t5 t12 93

PAGE 94

20GP 22 24 30X activity (nmoles/mi 26 28n*mg protein) 10 20 30 40 50t5 t12AP (nmoles CRE/mg protein)a b a b ab p = 0.098 10 20 30les/mi AL 40n* A) GPX activi ty (nmo g DN R R-ALa b b 10 20 30 40AP (nmoles CRE/ g DNA)a b b Figure 4-2. Glutathione peroxidase (GPX) specific activit y and antioxidant potential (AP; calculated as nmoles of copper reducing e quivalents, CRE) in Chelonia mydas liver homogenate at five weeks (t5) and twelve weeks (t12). For graphs on the left, GPX activity and total AP were normalized to total protein concentration as determined by Bradford assay. For graphs on the right, GPX activity and total AP were normalized to tissue DNA content as reported in Chapter 3 for t12 turtles. Each point represents mean standard error. Treatment groups a nd data analysis are the same as in Table 4-1. Sample sizes for GPX activities are the same as in Figure 4-1. Sample sizes for AP are n = 9 for t5 turtles and n = 7, 8, and 8 for t12 AL, R, and R-AL turtles, respectively. Letters (a and b) indicate statistically signi ficant differences (p < 0.05) among treatment groups within sampling periods. t5 t12 t12 94

PAGE 95

CHAPTER 5 OF DIETARY RESTRICTION ALTERS THE EXPRE TIMING SSION OF LIFE-HISTORY TRAITS IN A LONG-LIVED, PENOGEIC INSECT Introductio -historyeory to e howturaltion optimizes life cycles to maximitness. Central to th body of theory is the assumption that developmental trajectories and lifistories ld dtr ate plasticity in response to environmental variation including foodbility f 19arn ). Buse oinsi trinsic upper limits on rates differentially allocated to various reproductive and somatic functi ons according to priority rules (Boggs 1992, Zera and Harshman 2001). These priority rules predict th at survival should be favored over reproduction in times of resource limitation. One of the most pervasive findings in lifehistory studies is that food restriction (FR) leads to increa sed lifespan in a diversity of organisms including worms, spiders, insects, rodents, and primates (Weindruch and Walford 1988, Austad 1989, Turturro and Hart 1992, Mair et al 2003, Hatle et al 2006b). Coincident with lifespa n extension, adult FR usually decreases or inhibits oogenesis and egg production (Chippindale et al 1993, Wheeler 1996). Presumably, the suppression of reproductive activ ity during times of f ood scarcity allows food-restricted individuals to di vert available resources into maintenance and storage, thereby increasing starvation resistance and the probabil ity of survival until conditions more conducive to reproduction are encountered (Holliday 1989, Masoro and Austad 1996, Simmons and Bradley 1997). This negative correlation between longevity a nd fecundity is interpreted as evidence for a cost of reproduction (Stearns 1992). A reproductive cost can also be expressed as a trade-off ARTH NET n Life th seeks xplain na selec ize f is e h shou emons availa (Rof 92, Ste s 1992 eca f extr c or in of resource acquisition (Speakman and Krl 2005), most animals cannot simultaneously maximize the allocation of nutrients to all traits th at influence fitness. As a result, resources are 95

PAGE 96

96 to exist, the decreased survival demo eeding at high rates as adults must be attributable to the process of vitellogenesis in ad libitum -fed adult Drosophila et al 2004). Additionally, Kaeberlein et al (2006) found that FR in extended lifespan even when the food restriction was im pos erefore, the enhanced longevity demonstrated by food-restri re-allocation of nutrients away from egg pr survival. y through its growth and larger sizes at devel opmental transitio zed. In insects, adult body mass is often strongly and dire k 1993, Tammaru et al 1996; although see also Leather 1988), meani juvenile growth and large adult size tend to ma xim nutritional environment during juven ile stages lea ental transitions may shift downward to reduce the de ent time (Rowe and Ludwig 1991, Berrigan and Koella 1994, Leips and Travis 1994, Bradshaw and Johnson 1995, Day and Rowe 2002). The response to redu ced intake in juvenile stages in insects is therefore a reduction in adu lt body size with a potential for concomitant decreased fecundity (Hon k 1993). between current and future reproduction (Cal ow 1979, Reznick 1985 and 1992). For such a cost nstrated by very fecund indi viduals f egg production. However, preventing oogenesis or did not decrease mortality rates (Mair C. elegans ed after the cessation of reproductive activity. Th cted adults may not result simply from a oduction and toward somatic maintenance and Although it is clear that the nutritional e nvironment experienced during adulthood can er reproductive output, ju venile food restriction can also influence fecundity indirectl effects on body size. High food availa bility during juvenile stages favors rapid directly alt ns such that adult size is maximi ctly correlated with fecundity (Hon ng that the conditions conducive to rapid ize reproductive output. In contrast, when the ds to slow growth rates, size at developm mographic costs of extended developm

PAGE 97

Whereas juvenile and adult food restriction can have simila r effects on lifetime fecundity, their effects on longevity differ. In insects, the effects of juven ile FR on life-history traits and trade-offs are less clear than the effects of adu lt FR, primarily because few authors have assessed the responses of longevity to j uvenile food availability. Those who have done so have reported that juvenile FR either has no effect (Tu a nd Tatar 2003) or a negative effect (Boggs and Freeman 2005) on lifespan. The pattern of resource acquisition early in an individuals lifetime therefore has the potential to alte r later allocation patterns and the expression of life-history traits and trade-offs. Furthermore, al location patterns may also change with age (models reviewed by Perrin and Sibly 1993), depending on the diets experi enced in different life stages and the needs of the individual at the time of allocation. For example, the ex pression of a trade-off between current and future reproduction may depend on whether an organism relies on adult-derived nutrients or stored larval-derived nutrients for egg production (Boggs 1992). It is clear, then, that reso urce acquisition patterns can prof oundly affect the expression of life-history traits and trade-offs. Because fluctuations in food avai lability almost certainly occur for most animals at some point in their lifetimes (Boggs and Ross 1993, Carey et al 2002a), a complete understanding of the e ffects of intake during different life stages requires data collection throughout an individual s entire lifespan. However, such experiments are rare, particularly dition, studs of FR and li ory trade-offs in insects s ad libitum con f food offere not quantif Harshman bsolute for long-lived species. In ad ie f e-hist often suffer from one or more fundamental limita tions. Firstly, FR in in sects typically entail sumption of lower quality food rather than a quantitativ e reduction in the amount o d (Partridge et al 2005). Furthermore, intake of thes e lower quality diets is typically ied, despite the need for such info rmation when assessing trade-offs (Zera and 2001). In those studies that do impose a quantitative food restric tion in which a 97

PAGE 98

intake is limited, the re striction usually occurs only during adulthood and longevity is often not enhanced (Bogg cting the nearly universal finding of incr eased lifes titative FR in many taxa. In addition, to evalu l veral peptides (Wolfner 1997, Gillott 2003, Carvalho et al. 2006) ates, C. es the same food throughout its lifetime, enabling me to test dietary treatments that s s and Ross 1993, Carey et al 2002b, Cooper et al 2004), contradi pan under qua n ate the costs of reproducti on and the effects of dietary re striction on fitness in sexua species, females must be allowed to mate. Ho wever, co-housing individuals complicates the quantification of individu al intake and can influence longevi ty due to the effects of crowding (Joshi et al. 1998). To avoid such problems, reproductive ou tput of virgin female insects is often studied as a proxy for fitness, yet mating has be en shown to enhance egg production in se species where unmated females would otherwise lay infertile eggs (D e Clercq and Degheele 1997, Foster and Howard 1999). Furthermore, the males of many sp ecies can alter the physiology and behavior of females via sex or nutritious nuptial gifts (Voigt et al. 2006). Lastly, sexual species incur a number of costs associated with reproducti ve behaviors including courtship, repulsion of unwanted m intrasexual competition, and locomotory costs of carrying mates during copulation (Watson et al. 1998). To overcome these obstacles, I adopted a novel approach to li fe-history experimentation by using a parthenogenetic species as my animal model. Carausius morosus (Br.) (Phasmatodea, Lonchodinae) is a relatively long-lived speci es that reproduces via obligate apomictic parthenogenesis (Pijnacker 1966). Using a parthe nogen as my animal model obviated the need for mating while still permitting natural reproductive processes. This species is hemimetabolous and phytophagous, allowing for life-long, quantita tive dietary manipulations. Additionally morosus consum panned both juvenile and adult stages. My purpose was to determine the effects of 98

PAGE 99

differences in resource availability at several de velopmental stages on life-hi story traits that have substantial influences on population structure and dynamics, such as age and size at each developmental transition, longevity, and fecundity. Materials and Methods Animal Husbandry and Feeding Treatments This study was conducted in a USDA-approved quarantine facility within the Department of Zoology at the University of Florida (per mit # PPQ 69292). Lights were maintained on 12:12 light:dark cycle. Room temperature averaged 22.5-24.5 oC, and relative humidity av 45-55% th a eraged roughout the tria l. Twenty adult Indian stick insects ( Carausius morosus ) were obtain e e ects, re offered either more leaf disc s than they could co nsume within 24 hours ( ad libitum AL) or a restricted number of discs (R) equa l to 60% of the average daily mass-specific intake of AL-fed insects in the same life-history stage. Life-history stag es were categorized as ed from the Exploratorium in San Francisc o, California. Eggs laid by these females wer individually incubated until hatching. The resulting offspring (n = 86) were systematically assigned to one of six treatment groups such that all experimental insects produced by a particular mother were even ly distributed among groups. These insects were maintained individually fo r their entire lifetimes in plastic cages (29.5 cm x 19 cm x 19 cm) with locking vented li ds lined with fine-mesh screening. Each cag was misted daily with deionized water to provide drinking water. Insects were fed discs cut from leaves of English ivy ( Hedera helix ) daily. Biopsy punches (Miltex Instrument Co., Inc.) were used to create discs of multiple diameters: 2 mm for first instar insects, 3 mm for second instar insects, 4 mm for third instar insect s, 5 mm for fourth instar insects, 6 mm for fifth instar ins and 8 mm for sixth instar insects and adults. When cutting leaf discs, care was taken to avoid major leaf veins such that discs containe d as little vascular tissue as possible. Insects we 99

PAGE 100

each o L group n re eginning of the fifth i nt hout all each day and subtracting this quantity from the nu mber of discs offered the previous day. Whole f six instars, the adult st age prior to first oviposition, and the adult st age after first oviposition. Because massspecific intake of AL insects declined after first oviposition, the amount of food offered to food-restricted a dults after first ovi position was decreased proportionally to match this decline. Discs were offered according to five treatm ent schedules (Fig. 5-1). Insects in the A ( n = 15) were offered food ad libitum for the duration of their lifetimes. This group served as the control group from which intake data were used to determine the appropriate number of discs to offer to restricted insects. Insects in the R group ( n = 28) were offered the restricted amount of food for the duration of their lifetimes. Individuals in the AL-R groups were initially fed ad libitum and were switched to the restricted diet at the beginning of the fifth instar ( AL-R at 5th, = 15) or at first oviposition ( AL-R at Ov n = 14). Insects in the R-AL at 5th group ( n = 14) we initially fed the restricted diet and were switched to an ad libitum diet at the b nstar. Food-restricted insects generally ate all of the food offered each day, although food-restricted adults occasionally failed to consume all discs. I initially planned a diet switch from restricted to ad libitum at first oviposition ( R-AL at Ov ) but was unable to te st this treatme schedule because survival to ov iposition was extremely low for insects maintained throug the juvenile stages on the restricted diet. To ensure a sufficient sample size in the R group, insects that were food-restricted throughout all six instars and su ccessfully oviposited ( n = 7) were maintained on the restricted diet throughou t adulthood. The sample sizes indicated in each table and figure (except Figs. 55 and 5-6) include only those individuals that survived through the end of the sixth instar. Physiological and Life-History Response Variables Daily intake of each insect was estimated by determining the number of discs remaining 100

PAGE 101

discs were counted, and partial discs were pressed between mi croscope slides and scanned (Visioneer OneTouch scanner). Surface area of each disc fragment as a proportion of uneate leaf disc surface area was determined using Imag eJ (1.37v). A sample of each size of leaf discs was dried daily to constant mass at 60 C and weighed. The approximate daily dry matter in for each insect was then calculated as number of discs consumed estimated dry mass per dis Daily mass-specific intake was calculated usi ng estimates of daily body mass computed from periodic body mass measurements, as described below. Each insect was weighed weekl y, at the end of each instar (defined as the day when no food was eaten in preparation for ecdysis), at firs t oviposition, and at death. Insects were also photographed at these times (Nikon Coolpix 3200) and body lengths at the end of each history stage were then determined using Im ageJ. Body length was measured as the distance between the base of the antenna l socket and the end of the tergum on the terminal abdominal segment. Measurements of body size at the end of each instar for AL insects were then fitted to the allometric equation ln( y ) = ln(a ) + bln( x ), where y = body mass and x = body length. Relative body m n take c. lifeass (as an index of body condition) of insects in all treatment groups at the adult molt was assess e female. ed as the ratio between measured body mass and body mass predicted by the AL allometric equation (Perrin et al 1990). Specific growth rate (SGR) of eac h insect in each life-history stag was calculated as: SGR = 100*(lnBMf lnBMi)/t where BMf is body mass at the end of a stage, BMi is body mass at the beginning of a stage, and t is the time in that stage. During each day of adulthood, all eggs laid by each female were collected and individually weighed. Average egg mass was calculated as the m ean mass of individual eggs for each 101

PAGE 102

Reproductive lifespan was calcu lated as the time between first and last eggs laid. Clutch size was not quantified because oviposition in C. morosus occurs continuously throughout the reproductive lifespan rather than in discrete clut ches. After death, each female was dissected, and the number of ovarioles in both ova umbers of fully chorionated eggs d in the oviducts were counted. Although I intend ups. rocal ing ariances were homogeneous) or Tamh e in g ries was dete rmined. The n and non-chorionated eggs in terminal follicles an ed to evaluate hatch success, egg viability was inexplicably low in this study, particularly compared to the expected hatchab ility of nearly 100% (Brock 2000). Statistical Analyses Analysis of variance (ANOVA) was used to test for differences among treatment gro Data were first tested for normality (Shapiro-Wilk test) and homogeneity of variances (Levenes test) and transformed, if necessar y, using a natural log, reciprocal, square root, square, recip square, or reciprocal square root transformation. Pairwise co mparisons were evaluated us Tukeys Honestly Significant Difference post hoc test (if v anes T2 post hoc test (i f variances were not homogeneous). If transformation did not normalize data, they were analyzed using a Kruskal-Wallis test, and pairwise comparisons were evaluated using Mann-Whitney U tests with set at 0.005 to account for the number of comparisons tested (ten). Reproductive output was analyzed using ANO VA as described above and also using analysis of cova riance with body mass at first ovipos ition as the covariate to test for size-independent differences in allocation to reproduction. Other covari ates (including length at first oviposition, body mass at adult molt, and le ngth at adult molt) could not be used becaus of significant interactions between these variables and treatment group. Stepwise linear regression was used to determine the factors that best explained variance reproductive output. For this analys is, the dependent variable was the cumulative fecundity of each female. The independent variables tested we re mass-specific intake and growth rates durin 102

PAGE 103

the final instar; duration of the final instar; age, body mass, and relative body mass at the molt; duration of the pre-oviposition adult stage; mass-specific intake and growth rates durin the pre-oviposition adult stage; age and body mass at first oviposition; cumulative inta instars 1-4, 5-6, and 1-6; and cu mulat adult g ke during ive intake during the pre-ovi position adult stage and during the re e ent o orship curves were constructed for the entire lifespan ( n = 86) and for that successfully molted to the adult stage) productive lifespan. Cumulative intake and body condition were tested as potential determinants of reproductive output because insect s have been shown to require a threshold level of food consumption or body stores to in itiate reproductive processes (Juliano et al. 2004, Hatle et al. 2006a). Variables had to meet a 0.05 significan ce level to enter a model. Duration of the reproductive lifespan was dropped from this analys is because of collinea rity with the first variable selected by the model. Stepwise linear regression was also used to determine the factors that best explained variance in initial ovipos ition rate. For this analysis, the dependent variable was the cumulativ number of eggs laid by each female on day 6 of the reproductive lifes pan, and the independ variables tested were the same as those used above. Specific growth rate during adulthood prior to first oviposition was dropped from this analysis due to collinearity with other variables already selected by the models. To test for a longevity cost of reproduction in this species, least squares linear regression was used to examine the relationship between cumulative fecundity and lifespan (both total lifespan and adult lifespan). L east squares linear regression was also used t evaluate the strength of the relationship between fecundity an d cumulative intake during the reproductive lifespan. Kaplan-Meier surviv adult lifespan ( n = 70, including only those individuals Pairwise comparisons among treatment gr oups were evaluated using log-rank tests with 103

PAGE 104

set at 0.005 to account for the number of comparisons tested (ten). Data were analyzed using SPSS for Windows (Release 11.0.0), a nd S-Plus (Version 7.0) was used for graphing smoothin functions and Kaplan-Meier curves. Results The diets I imposed yielded different mass-speci fic intake trajectories for insects in each treatment group (Fig. 5-2). Total dry mass of fo od consumed during ea ch life-history stage except the first instar differed significantly among treatm g ent gr oups (Table 5-1). In the second instar e f 5-2). Ins ects did not differ in body mass am ong treatment groups at hatching ( F4,65 = 1.00, p = 0.414). At the e greater and molting occurred ects (groups R and R-AL at 5th). From the fifth instar until first ovipo significantly different from body mass of AL and R-AL at 5th insects, but all other pairwise food-restricted insects consumed more total food th an insects feeding ad libitum despit being significantly smaller (Fi g. 5-3 and Table 5-2) and recei ving proportionally less food on a daily basis. This discrepancy is explained by the significantly longe r instar duration in food-restricted groups relative to groups feeding ad libitum (Figs. 5-3 and 5-4a). The pattern o cumulative intake changed after the second instar and was dependent on diet history. Although cumulative intake prior to the adult molt differed among treatment groups ( F4,65 = 28.872, p < 0.0001), the total amount of food consumed between hatching and first oviposition did not differ among treatment groups ( F4,53 = 1.321, p = 0.274), hinting at a poten tial intake threshold for induction of reproductive activity (Juliano et al 2004). Size and age at each life-history transition differed significantly among treatment groups (Fig. 5-3 and Table nd of instars 1-4, body mass was at younger ages in initially ad libitum insects (groups AL, AL-R at 5th, and AL-R at Ov ) than in initially food-restricted ins sition, all treatment groups except AL and AL-R at Ov differed significantly in body mass and age at the end of each stage (Table 5-2). At death, body mass of AL-R at Ov insects was not 104

PAGE 105

comparisons of size were significantly different Age at death differed significantly among all groups except R and R-AL at 5th. Relative body mass also differed among treatment groups at th e adult molt as determined by allometric analysis (Table 5-3). Least squares regr ession of body mass (y ) and length (x ) for AL insects at the end of each in star yielded the equation ln( y ) = 2.7112*ln( x ) 12.018 (F1,76 = 14077.66, p < 0.0001, R2 = 0.995). This equation was used to calculate predicted body masses at actual body lengths for each insect at the adult molt (Table 5-3). The r atio of actua l to predicted for insects feeding at a restri cted rate during the fi nal two instars (groups R and A r ing L). body mass was lower L-R at 5th), indicating that insects in these gr oups had proportionally lower body masses fo a given body length than insects in the other three groups. The duration of each life-history stage di ffered among treatment groups (Fig. 5-4a). Food-restricted insects generally progressed more slowly through each stag e than insects feed ad libitum Previous diet history affected the dura tion of the fifth and sixth instars and the pre-oviposition adult stag e for insects in groups AL-R at 5th and R-AL at 5th, as individuals in these groups progressed through these stages more rapidly than continuously food-restricted individuals (group R ) but more slowly than continuously ad libitum individuals (group A Insects experiencing food restric tion during adulthood prior to firs t oviposition laid their first eggs later in the adult stage than insects feeding ad libitum during this time. However, duration of adulthood after first oviposit ion was significantly shorter for in sects that were food-restricted than for insects that were feeding ad libitum during this time, regardless of when the food restriction was imposed. Food-restricted insects also grew more slowly than insects feeding ad libitum (Fig. 5-4b), although diet history affected the magnitude of this difference. After a switch from restricted to 105

PAGE 106

ad libitum feeding at the beginning of the fifth instar (group R-AL at 5th), specific growth rates through the final two instars were co mparable to those of continuously ad libitum insects (group AL). I therefore found no evidence of growth compensation in R-AL at 5th insects. However, insects that experienced a switch from ad libitum to restricted feeding grew significantly faster in both the fifth and sixth instars th an insects that were continuous ly food restricted. All inse gained body mass between th e adult molt and first ovipos ition, with growth of R-AL at 5th insects slower than that of AL and AL-R at Ov insects but greater than that of R and AL-R at 5th insects. All insects lost body mass between first oviposition and death. AL-R at Ov insects lost proportionally more body mass than AL insects during this time, but all other pairwise comparisons of adult growth rat cts es after first oviposition we re not significant. l. ivity ces appear to result both from differences in reproductive lifespan (F4,53 = 41.7 s An event history diagram depicting the lifespan of each insect in the study (Carey et a 1998) demonstrates the variation in life histories and survivorsh ip induced by diet treatments (Fig. 5-5). Pairwise log-ra nk tests of survival indicat ed that all groups except R and R-AL at 5th differed significantly in total lifespan (Fig. 56a). This result parallels the ANOVA results for age at death (Table 5-2). Pairwise log-rank tests of adult survivor ship (Fig. 5-6b) indicated that longevity was greater for AL and R-AL at 5th insects than for all insect s feeding at a restricted rate during adulthood, suggesting that food restriction experien ced during reproductive act negatively affected adult lifespan regardless of dietary history. Treatment groups differed significan tly in realized fecundity ( F4,53 = 50.31, p < 0.0001, Fig. 5-7). These differen 0, p < 0.0001) and from differences in reproductiv e rate. The high initial slopes in Figure 5-7 for groups AL, AL-R at Ov and R-AL at 5th corresponded to higher early oviposition rate (calculated as eggs laid per day in the first si x days of the reproductive lifespan) in these groups 106

PAGE 107

compared to groups R and AL-R at 5th ( F4,53 = 15.576, p < 0.0001). These differences suggest egg output is enhanced early in the reproductive lifespan by ad libitum feeding during adulthood prior to first oviposition. The low rep that roductive output of R insects was compounded by low survival to first oviposition, such that egg pr inished by lifelong FR. Diffe rst switch ted r ovarioles than initially ad libitum insect,53 = oduction was severely dim rences in egg production did not simply result from differences in body size, as analysis of covariance revealed significant differences in ad justed mean fecundity when body mass at fi oviposition was used as a covari ate (Table 5-4). Adjusted mean fecundity also differed among groups when relative body mass at the adult molt was used as a covariate ( F4,53 = 42.872, p < 0.0001; data not shown). Egg production was also altered by diet hist ory through effects on average egg mass (F4,53 = 8.195, p < 0.0001, Fig. 5-8), with insect s experiencing a diet from ad libitum to restricted feeding producing signifi cantly smaller eggs than continuously ad libitum insects. Differences in reproductive output do not appear to result from differences in ovarian morphology among groups (Table 5-5). However, although Tukeys HSD post hoc test did not reveal significant di fferences in ovariole number among treatment groups, a less conservative post hoc test (the Least Significant Difference test) indi cated that initially restric insects (groups R and R-AL at 5th) had significantly fewe s (groups AL, AL-R at 5th, and AL-R at Ov ) (p < 0.05 for all significant comparisons). Diet history did affect the number of eggs re maining in the ovaries at death (unfulfilled reproductive potential, F4,53 = 5.286, p = 0.001), with R insects having more eggs remaining in the ovaries at death than AL-R at 5th insects. All other pairwise comparisons of unfulfilled reproductive potential were not significant. Grou ps also differed in potential fecundity ( F471.62, p < 0.0001, calculated as unfulfilled reproductive potential plus realized fecundity) and total reproductive investment ( F4,53 = 49.57, p < 0.0001, calculated as the summed mass of all 107

PAGE 108

eggs laid by each female). The patterns for poten tial fecundity and total reproductive investment (data not shown) were iden tical to that demonstrated by realized fecundity. the variance in realized fecundity (mod ass at te n, at t d R2 hown). However, analysis of covariance indicated that adult lifespan did no I used stepwise multiple linear regression to id entify the most significant determinants of realized fecundity (Table 5-6). Cumulative intake during the reproductive lifespan was the primary variable selected by the model, explai ning 82.8% of the variance in fecundity (Fig. 5-9a). When potential fecundity (number of eggs laid + number of eggs remaining in the ovaries at death) was regressed against cumulative in take during the reproducti ve lifespan, the same relationship existed with an adjusted R2 value of 0.847 (data not shown) In addition, growth rate during adulthood prior to first oviposition and cumulative intake during all juvenile stages were also selected as variables in a model that explained 92.8% of el 3, F3,54 = 245.46, p < 0.0001). Stepwise multiple linear re gression identified body m first oviposition, age at the adult molt, mass-sp ecific intake during adulthood prior to first oviposition, and cumulative intake during adulthood prior to firs t oviposition as significant independent variables in a model that explained 70.6% of the variance in initial oviposition ra (model 7, F4,53 = 35.20, p < 0.0001). The data do not support the conten tion that decreased longevity is a cost of reproductio least in C. morosus On the contrary, fecundity was signifi cantly and positively related to adul lifespan when data for all treatments were combined ( F1,56 = 25.67, p < 0.0001, R2 = 0.302, Fig. 5-9b). When potential fecundity (n umber of eggs laid + number of eggs remaining in the ovaries at death) was regressed against adult lifespan, the same relationship existed with an adjuste value of 0.343 (data not s t have a significant e ffect on reproductive output (F1,52 = 3.284, p = 0.076) whereas treatment did have a significant effect ( F4,52 = 32.463, p < 0.0001). This result was confirmed by 108

PAGE 109

individual regressions of fecund ity versus adult lifespan for each treatment group, none of whic was significant (p > 0.2 in all cases). Unlike adult lifespan, total lifespa n was not significantly related to realized or potential fecundity ( p > 0.5). I also found no evidence for a trade-off between number and size of eggs laid. There was neither a significant interaction between average egg mass and fecundity (p > 0.2) nor an effect of fecundity on average egg mass ( p > 0.3). Individual regressions of average egg mass versus fecundity indicated that there were no significant relationships ( p > 0.3) except in the case of R-AL@5th insects, for which there was a positive relationship between egg size and number ( F1,10 = 5.161, p = 0.046, R2 = 0.274 Discussion Developmental plasticity in response to food availability is nearly unive h ). rsal (Juliano et al 2004 e d extended adult lifespan. Conversely, FR during adulthood decreased the duration of the adult and references therein), and this study was no exception. In C. morosus both size and ag at each life-history transition depended on diet history. As is common in studies of this kind (e.g., Gebhardt and Stearns 1988 and 1993), insects th at experienced FR prior to the onset of reproductive activity progressed thro ugh juvenile stages more slowly and were smaller at each molt than individuals feeding at a consistently high rate. Decreasing size and increasing age at developmental transitions represent a compro mise between the need to maximize body size (because of its potential effects on fitness) and the need to mini mize the demographic costs of extended development time (Rowe and Ludwig 1991). This plasticity in developmen t rate corresponded to substant ial differences in survival trajectories among treatment groups. One of my most salient results was the finding that longevity enhancement is not a ubiquitous outcome of dietary restricti on. Although individuals that experienced early-onset FR ( R and R-AL at 5th) survived longer than initially ad libitum -fe individuals, this increased longevity was due solely to extended de velopment time rather than to 109

PAGE 110

stage, such that AL-R at Ov insects had shortened lifespans compared to AL insects. A diet switch from ad libitum to restricted feeding during developmen t extended the duration of the fifth and sixth instars relativ e to continuously ad libitum insects, but this differe nce was not sufficient mitigate the negative effect of FR on adult lifespan. As a result of decreased growth rates, inse cts that experienced FR at any point during development were smaller at the adult molt than ad libitum insects. Although subsequent reproductive output of food-restricted insects wa s significantly diminish ed, mean fecundities differed significantly among treatment groups even when corrected for body mass at first oviposition. Plasticity in adult size alone therefore does not expl ain the drastic differences in fecundity I observed among treatment groups. Reproductive output may have been mildly constrained by ovarian morphology. Although I detected no significant differences in total ovariole number among treatment groups when a conservative post hoc test was used, I did find significant differences in ovariole number between initially restricted and initially ad libitum insects when a more liberal post hoc te used. This result suggests that ovarian development in C. morosus is somewhat plastic in response to diet. In Drosophila ovariole number responds strongly to larval diet (Tu and Tat 2003) and is correlated with fecundity (David 1970). A lthough ovariole number in C. moros to st was ar us appea in r rs to be much less plastic than in D. melanogaster it is possible that decreased fecundity food-restricted insects in this study is partially explained by differences in ovary size but only fo insects that were food-restrict ed during early development. The primary determinant of fecundity in this study was adult intake, with approximately 83% of the variance in reproduc tive output explaine d by the total amount of food consumed during the reproductive lifespan. Because of this strong, positive correlation between fecundity 110

PAGE 111

and cumulative intake during the reproductive lifesp an, I conclude that In dian stick insects use an income breeding strategy (sensu Stearns 1992, Jnsson 1997), in which the resources alloca n e ng juvenile stages, mature eggs are not present at the adult molt. ay to ted to reproduction are acquired primarily during the reproductive period. Cumulative intake between the adult molt and the end of the reproductive lifes pan was less strongly correlated with reproductive output suggesting that the food acquire d prior to first ovipositio was allocated to some degree of pre-reproductive somatic growth rather than being allocated exclusively to egg production. The putative leve l of body stores accumulated by the time of th adult molt does not appear to dictate reproduc tive success, as demonstrated by significant differences among groups in re alized fecundity when corrected for relative body mass. An income breeding strategy is appropriate for an organism like C. morosus in which oogenesis and vitellogenesis are noncyclic and continuous (Bradley et al 1995) throughout a comparatively long reproductive lifespan. Additiona lly, species that rely heavily on incomi resources for reproduction should have ovaries containing primarily immature oocytes immediately after the adult molt (Jervis et al 2005), as is the case for C. morosus (Bradley et al 1995). Although ovaries are present in Given this breeding strate gy, it is not surprising that both mass-specific intake and agespecific fecundity in this study declined after first ovi position for adults feeding ad libitum This decrease in consumption and production with ti me typifies insects that are income breeders (Kindlmann et al 2001, Dixon and Agarwala 2002). One might expect food-restricted insects that employ an income breeding tactic to extend the duration of reproductive activ ity and thereby to mitigate (at least partially) the effects of decreased daily intake on oviposit ion rate. Given that food-restrict ed flies respond in this w FR (Carey et al 2002a), I expected to see a similar pa ttern in this study. Surprisingly, females 111

PAGE 112

could not simply compensate for adult FR by increasing the length of the reproductive lifespan. The very low reproductive output of R and AL-R at 5th females therefore resulted from the combined effects of decreased daily intake and shortened reproductive lifespan. The proximate cause of shortened reproductive lifespan may relate to the level of bod stores accumulated prior to adulthood. Insects feed ing at a restricted rate late in development were lighter for their length than insects feeding ad libitum immediately prior to the adult molt. Although fat body mass and storage proteins were not quantified, these results imply that these individuals may have accumulated proportionally fewer body stores by the beginning of y adulthood compared to insects feeding ad the final two instars. Therefore, food restric at ly t at d as AL and AL-R at Ov insects at first ovipo rter n libitum during tion late in development appears to have shifted allocation away from the accumulation of mobilizable reserves. I s uggest that these reserves serve as the source of nutrients that are allocated to somatic maintenance after th e onset of reproducti ve activity. Because R and AL-R 5th insects were smaller in both absolute and re lative body mass at the ad ult molt, they probab depleted their limited stores more rapidly after th e onset of reproductive ac tivity than insects tha were feeding ad libitum as young adults. Conversely, R-AL at 5th insects were feeding ad libitum as pre-oviposition adults, alt hough they were doing so at lo wer mass-specific rates than AL insects. Although R-AL at 5th insects had relative body ma sses similar to those of AL and AL-R Ov insects at the adult molt, they were nearly twice as ol sition. It is therefore possibl e that reproductive lifespans of R-AL at 5th insects were sho than those of continuously ad libitum -fed adults simply because of age-specific declines i physiological function. It does not appear that shortened repro ductive lifespans resulted from exhaustion of available oocytes, as almost all individuals in the st udy had chorionated eggs remaining in the ovaries at death. 112

PAGE 113

In addition to sustaining the soma during adulthood, body stores present during early adulthood may also serve as a signal that coor dinates rates of vitell ogenesis and oviposition at the onset of reproductive activity (Moehrlin and Juliano 1998, Hatle et al 2004, Juliano et al 2004). In this way, body composition may function as an index of food availability that entrains subse at ral ore nd ALd diet may have fo rced these individuals to suppl th med s itch to the re stricted diet late in devel h is quent reproducti ve function (Rowe et al 1994). If this h ypothesis is correct, AL, AL-R Ov and R-AL at 5th insects were committed to a high early oviposition rate in the first seve days of the reproductive lifespan because these individuals had accumulated proportionally m body mass (with potentially higher levels of body stores) prior to firs t oviposition than R a R at 5th insects. However, the mismatch between in take and pre-determined oviposition rate in AL-R at Ov insects after the switch to a restricte ement incoming resources by withdrawing nut rients from mobilizable stores that would otherwise have been allocated to somatic functions such as main tenance and survival. Data for all AL-R at Ov individuals were located above the regression line correlating fecundity wi cumulative intake during the repr oductive lifespan (Fig. 5-9a), suggesting that these individuals produced more eggs than would have been pred icted by the amount of the food they consu as reproductively active a dults. If body stores were in fact us ed to supplement incoming nutrient for oogenesis, the exhaustion of these stores woul d then explain the shortened adult lifespan of AL-R at Ov insects relative to AL individuals. The hypothesis that body stores determine adu lt lifespan and establ ish initial oviposition rate is supported by my data for insects that expe rienced a sw opment. Despite feeding ad libitum in the first four instars, AL-R at 5th insects had similar reproductive lifespans and fecundities as continuously food-restricted insects. Because growt typically exponential in juvenile insects, body composition at the adult molt is largely 113

PAGE 114

determined by food availability during the final in star(s) (Scriber and Slansky 1981). As a resul ad libitum feeding early in life does not appear to provide a substantial fecundity benefit for individuals that subsequently ex perience a decline in juvenile food availability. However, the marginal dependence of ovariole number on food availability during the first few instars indicate that early nu tritional conditions could potentiall y affect subsequent reproductive function. Conversely, ad libitum feeding later in developmen t and during adulthood provides nutrients necessary for somatic maintenance. However, mass-specific intake declined after oviposition for both groups feeding ad libitum as reproductively active ad ults, suggesting t, does first either that o r older trients to increase survival (e.g., Tammaru et al sult ogenesis and vitellogenesis ar e less costly than somatic growth during juvenile stages o that digestive and reproductive functions d ecline with time due to senescence (Kindlmann et al 2001, Carey et al 2002a, Dixon and Agarwala 2002). Evidence for the latter hypothesis was demonstrated by the lower mass-specific intake of R-AL at 5th insects, which matured at ages, compared to AL insects after first oviposi tion. I therefore conclude that adult survival and reproductive decisions (such as age at first ovi position and initial oviposition rate) are based on the extent of reserves accumulated prior to maturity, whereas fecundity depends on food consumed during the reproductive lifespan in C. morosus Perhaps not surprisingly, capital breeders (sensu Stearns 1992) tend to demonstr ate the opposite strategy, allocating stored reserves to egg production and us ing adult-acquired nu 1996). The apparent dependence of initial oviposit ion rate on accumulated body stores may re from differences in hormone signaling induced by di et. In insects, the fat body serves as the main depot for stored lipids and is responsible for synthesizing yolk proteins (e.g., vitellogenin) and 114

PAGE 115

lipids for incorporation into developing follicles (Chapman 1998). Fat body mass and hemolymph vitellogenin titers decr ease in response to adult FR in grasshoppers (Hatle et al 2006a), suggesting that my early fecundity results may be mediated by difference s in the size of this st of not or ore duction and survival (Stearns 1992). Furthermore, I detected orage organ. Additionall y, the synthesis of yolk compo unds by the adult fat body is controlled by hormones includi ng juvenile hormone (JH) and ecdysone (Klowden 2002), both which have been shown to respond to feeding rates (Hatle et al 2003, Tu and Tatar 2003). Specifically, JH stimulates vitellogenesis in most adult insects (Chapman 1998) but not in C. morosus (Bradley et al 1995), so diet-induced differences in JH synthesis were probably responsible for the decreased fec undity I observed in this study. Ho wever, JH does facilitate the uptake of vitellin by developing follicles in C. morosus (Bradley et al 1995), suggesting that my results for egg size may reflect differences in JH signaling. Unlike lipids, proteins are thought to be stor ed primarily in hemolymph (Chapman 1998). These hemolymph storage proteins are critical to egg production (Wheeler et al. 2000) and are responsive to diet (Hatle et al 2004). Because egg production is a protein-limited process for phytophagous insects like C. morosus (Nijhout 1994, Chapman 1998), the quantity of hemolymph storage proteins pres ent during adulthood may therefore serve as a nutrient sens regulating reproductive output. Contrary to my expectations, fecundity wa s significantly and positively correlated with adult lifespan and not correlated with total lifespa n. My results therefore indicate that insects feeding ad libitum as adults did not incur mortality costs simply because they reproduced m than food-restricted insects. C onsequently, decreased longevity is not a cost of reproduction in this species. This result contradicts the assump tion that FR elicits a shift in allocation and therefore a trade-off between re pro 115

PAGE 116

no ev ty n a r s depends on more than simpl e, idence of a trade-off between current and future reproduction or between early fecundi and adult lifespan (Reznick 1985 and 1992). In fact, insects feeding ad libitum as pre-ovipositio adults (groups AL, AL-R at Ov and R-AL at 5th) exhibited correspond ingly high initial oviposition rates and also higher cumulative fecundities than R and AL-R at 5th insects. The longevity costs of reproduction may di ffer among species depending on the relative timing of resource acquisition and allocation to reproduction (Boggs 1992). The occurrence of trade-off between longevity and fecundity requires that the resources allo cated to reproduction or somatic maintenance are derived from a common resource pool and th at the utilization of resources from this pool for egg production necessari ly decreases the availability of resources fo subsequent egg production or survival (van Noordwijk and de Jong 1986, Zera and Harshman 2001). The lack of a negative correlati on between longevity and fecundity in C. morosus therefore implies that the pro cesses of survival and reproducti on may not compete for resources from a common pool. Instead, I suggest that female s of this species allo cate existing stores primarily to maintenance and divert incoming resources to egg production. Although the diet treatments I imposed elicit ed clear and significant differences in egg production rates, these data alone do not establish that diet directly affects fitness. There is evidence that maternal nutritional environmen t can substantially alter offspring phenotypes (Wayne et al 2006) and survival (Prasad et al 2003) such that fitnes y total egg output of a female. In addition, host plant quality has been shown to alter fertility of phytophagous insects w ith no effect on fecundity (Moreau et al 2006). Furthermor fertility can decline as age-specific intake de clines over time in income breeders (Dixon and Agarwala 2002), thereby further un coupling fecundity and fitness. 116

PAGE 117

Consequently, my fecundity data in isolation are not sufficien t to draw conclusions about the effects of diet history on fitness. Unfortunately, my ability to assess these effects was compromised by egg inviability. Compared to the expected hatch success of nearly 100% for captive C. morosus (Brock 2000), egg hatchability in th is study was disappointingly low. I suggest two possibilities for this occurrence. The English ivy I fed to experimental animal have been s may deficient in one or more limiting nut rients, thereby largely preventing embryogenesis. Altern ever, l but e need nses to differences in intake depended to a large extent on the timi ng of nutritional stress n atively, my incubation protocol may not have been appropriate for this species. How my reproductive output results do indicate that f ood restriction imposed la te in development and during reproductive activity has profound negative implications for fitness. Furthermore, high mortality prior to the onset of reproductive activity compounded the negative effects of lifelong FR on egg production and indicate s that reproductive output is severely reduced by continuous, quantitative FR. In summary, food availability strongly influe nces the expression of life-history traits and trade-offs in C. morosus To my knowledge, this study was the first to evaluate the effects of quantitative dietary manipulati ons throughout life in a long-lived, hemimetabolous insect. My methodology allowed for accurate measurement of daily food consumption rates, a critica often neglected component in studies of life history (Zera and Harshman 2001). In addition, using a parthenogenetic species as my animal mode l was a novel approach that obviated th for mating while still permitting oviposition of fer tile eggs. My data demonstrated that the lifehistory respo Food restriction experienced at any point during life led to decreased fecundity. This decrease resulted primarily from differences in th e quantity of reserves accumulated prior to the onset of reproductive activity and resulting differen ces in reproductive rate and adult survival. I 117

PAGE 118

118 contrast, the effect of food re striction on overall lifespan depe nded on when the restriction was first imposed. As such, lifespan was maximized when food consumption was limited early in life, whereas reproductive output was maximized when food c onsumption throughout life was maximized. In effect, food restri ction extended development but shortened adult lifespan, with negative consequences for final body size, repr oductive lifespan, reproductive output, and, quite possibly, fitness. In C. morosus it appears that storage reserv es acquired early in life are essential for determining adult survival and fo r entraining the timing and rate of reproductive processes, but adult income is essential for egg provisioning. Putativ e fitness is therefore dictated both by past and curren t nutritional conditions.

PAGE 119

119Table 5-1. Cumlative intake expr se thtal dry s cu ach of fireen groups. Ctikey ) u esd ase to masonsmed during ea ch life-history stage in eumulave Inta (g drmatter ve tatmt Treatment Fita nr Third Instar s ftr Sixth Instar PrA rst Insr Secod Insta Fourth In tar Fih Insta e-Ov dult Post-Ov Adult AL 00 7 00050 9a 0.0795. 500 4ab0.5757. 18 AL-R at 5th 00 8 00040 8a 0.0802. 38010 8c0.6500. 04 AL-R at Ov 00 6 00040 7a 0.0772. 500. 3b b 0.5747. 03 R 00 6 000. 0b 0.0636.0 800 8ad0.7312.0 02 R-AL at 5th 00 3 000. 1b 0.0606.0 61030. 2d d 0.7529.0 19 0.0082 0.02a 0.019 0.a .04110.000 0.0080 0.02a 0.018 0.a .04030.000 0.0079 0.02a 0.019 0.a .04100.000 0.0088 0.03a 0.022 0.04b 03580.001 0.0084 0.04a 0.022 0.05b 03520.001 00016a 0.158 0.025a .34730.005 0 00013a 0.14 0.08b .39120.009c 0 00012a 0.150 0.016a 35140.005 0 0009b 0.129 0.031c .31620.007 0 0021b 0.10 0.05d 27920.012 0 3280a a 2.6454 0.59a 0293ab 0.4289 0.78b 0135a a0.9439 0.28c 614b b 0.2287 0.42d 491b b 1.5177 0.63c Life-historget x are t e r poed ah ulg t-ov adu)upn starobis lim = restricted. ee Figure 5-1 for a description f tmnts. Sample t sams g5-awdi n superscripe fl psheo aSx stines y staes wre caegorized as eachof siinst between first oviposition and death (pos lt. Val S odietreat ts arsigniicanty different among treatment s, thadulstagprioto first oviositin (pr-ov ault), es rereset means andd errrs. Abrevation: AL ee sizs arehe e ain Fiure 2. Vlues grou witin lif-hist ry stges. ee tet fo rtatiscal a nd te adt stae = ad bitu R ith ffer et alys.

PAGE 120

F a p-values for comparisons of body ma ong hiche-ory ityouteiroisons nibus 2nd stage. ss and age am five treatment groups w itn ea lifhistIdent of Grps Tesd in Pawise Cmpar Omnibus F an2 1 2 1 & 3& 4 d & 1 5 & 5 3 & 1 &2 & 3 2 4 2 & & 4 3 5 4 & 5 Body M Hatch 4,651.0 0. 78 0. 000 .5 4 80 9 End of star4,651.6 0. 00 0.< 0.001 .0 1 < 00 8 End of st4,658.3 0. 82 0.< 0.000 .0 0001 < 00 0 End of star4,655.3 0. 99 0.< 0.000 .0 0001 < 00 0 End of star4,659.3 0. 99 0.< 0.000 .0 0001 < 00 9 End of star4,650.8 0. 00 0.< 0.000< 0 .0 0001 < 00 2 End of star 2.9 0. 00 0.< 0.000<1 .0 0001 < 00 01 First Oio4,539.0 0. 00 0. 0.010 <1 .0 0001 < 03 01 Death 4,533.2 0. 00 0. 0.003 <1 .0 9 < 24 01 Age End of star .6 0. 88 0.< 0.001 .0 0001 < 00 5 End of st4,656.2 0. 00 0.< 0.000 .0 0001 < 00 5 End of star4,656.1 0. 98 0.< 0.000 .0 0001 < 00 9 End of star4,658.3 0. 00 0.< 0.000 .0 0001 < 00 7 End of star4,651.7 0. 00 0.< 0.000<1 .0 0001 < 00 1 0 End of star4,658.9 0. 00 0.< 0.000<1 .0 0001 < 00 01 First Oio .4 0. 00 0.< 0.000<1 .0 0001 < 00 01 Death 4,535.6 0. 03 1 0.< 0.001<1 .0 0001 < 00 0 ass F = 0, p =414 0 0.0 0.780999 1st In F = 27, p <0001 1.0 0.999 < 0010 2nd Inar F = 104, p <0001 0.7 0.999 < 0001 3rd In F = 117, p <0001 0.8 1.000 < 0001 4th In F = 265, p <0001 0.6 0.970 < 0001 5th In F = 264, p <0001 < 0.10 0.998 < 0001 6th In 2 = 62, p <0001 < 0.01 0.801 < 0001 vipositn F = 51, p <0001 < 0.01 0.993 < 0001 F = 36, p <0001 < 0.01 0.366 < 0001 1st In 2 = 516, p < 0001 0.0 0.840 < 0010 2nd Inar F = 681, p <0001 1.0 1.000 < 0001 3rd In F = 627, p <0001 0.0 1.000 < 0001 4th In F = 510, p <0001 1.0 0.994 < 0001 5th In F = 597, p <0001 < 0.01 0.942 < 0001 6th In F = 880, p <0001 < 0.01 0.952 < 0001 vipositn 2 = 506, p < 0001 < 0.01 0.840 < 0001 F = 95, p <0001 0.1 < 0.000< 0001 1. 1.000 060 0.800.560 0.4 0.99 0 1.000 < 0010 < 0.00 0 0.0001< 0.01 0.99 1 0.926 < 0001 < 0.0.0001< 0.01 1.00 1 0.997 < 0001 < 0.0.0001< 0.01 1.00 1 0.854 < 0001 < 0.0.0001< 0.01 0.99 1 0.001 < 0001 < 0.0.0001< 0.01 0.00 1 0.000< 0001 < 0.0.0001< 0.01 < 0.00 0.000< 001 0 < 0.0.0001 0.1 < 0.00 0.000 001 0.010.0001 0.3 < 0.00 0 0.960 < 0001 < 0.0.0001< 0.01 0.79 1 1.000 < 0001 < 0.0.0001< 0.01 0.97 1 0.991 < 0001 < 0.0.0001< 0.01 0.98 1 0.995 < 0001 < 0.0.0001< 0.01 0.99 1 0.000< 0001 < 0.0.0001< 0.01 < 0.00 1 0.000< 0001 < 0.0.0001< 0.01 < 0.00 1 0.000< 0001 < 0.0.0001< 0.01 < 0.00 0 0.000< 0001 < 0.0.0001< 0.01 0.36 Identity of groups tested in pairse cmparisons: 1 = 2 ALm values are reported, paramtr ic tests were used. When valus are analyses. Statistically significant p -values are indicated in bo. wio AL = R at 5thAL-R at Ov R -at we = libitu and R = restricted). See Figure 5-1 for a description of diet treatm g-h e2e reportoaicts e .torisl ld 3 = 4 = and5 = RAL 5th (herAL ad ents. Sample sizes are the same asin Fiure 52. Wen F ed, nnparmetr tes werused See ext f stattica 120Table 5-2. Om

PAGE 121

Table 5-3. Relative mass ( standard error) at th e adult molt of insects in five treatment groups, analysis (see text for statistical analyses). Treatment Actual Mass/Predicted Mass calculated as the ratio of actual to predicted body mass as determined by allometric AL 1.087 0.016 a AL-R at 5th 0.977 0.016b a R 0.995 0.014b th a AL-R at Ov 1.093 0.020 R-AL at 5 1.148 0.013 Abbreviations: AL = ad libitum R = restricted. See Figure 5-1 for a description of diet superscripts are significantly different among treatment groups. treatments. Sample sizes are the same as in Fi gure 5-2 for juveniles. Values with different 121

PAGE 122

Table 5-4. A ndity ( standard error) of insects in each of five treatment ro udy mass at fi rst oviposition as a covariate. Treant teecundity d jus ups ted es m tim A ea at d n f ed jus ecu g tme sing bo d Mean F A 886aa L 6 .061 4.3 AL-R at 5th 22b AL-RO 6 R 6 R-AL at 5t2c .204 4.524b.649 4.190bc .929 6.815bc .719 3.579c at v h 3 2 4 Abb trea sup cov revias: u ted. See Figure 5-1 for a description of diet tments. Sample sizes are t Figure 5-2 for adults. Values with different erscripts ar d ng treatment groups according to analysis of ariancit i r multiple comparisons. tion AL = a d li bit m R = restric he same as in ifferent am o correction fo e si h a gn Bo ifi nf can err tly on e w 122

PAGE 123

Table 5-5. Total num 123 ber of ovarioles (mean st andard error) in ins each of five orte Treatmnt iole ects from treatm e ent groups upon post-m m s dis se cti on. N u mb er of Ov ar AL 0.488 53 .3 8 AL-R at 5th AL-R Ov .6 R 50.86 1.262 R-AL at 5th 53.00 0.453 53 at .2 3 0 55 51.33 0.449 Abb treatm was significant ( fou test (the Least Significa R-AL at 5 at 5 revians: A ents. Sample sizes are the same as in Figure 5-2 for adults. Ae omnibus F -value Tu i ndnifi reat nt Difference test) indicated that in itially restricts (groups R and thhad iole u AL, AL-R th, and L-R tio L = ad lib itu m R = restricted. S ee Figure 5-1 for a description of diet lthough th cant Difference post hoc test conservative post hoc ted insec m insects (groups F4,can 53 t d = 3 iff .38 ere 7, nc p es = 0 am .0 on 15) g t key me s nt H gr oups. Results of a less on est ly S ig nif no sig ) A sig at ni Ov fic ) ( ant p < 0.05 for all significant com ly few er ov ar s th an in itia lly parisons). ad li bit

PAGE 124

eters for equations pred icting realized fecundity a nd initial oviposition termined by stepwise multi ple lin regression. Model y x1 x2 x3 x4 intercept 2 R rate as de1 ear 2 3 4 1 Fecundity Cum. Int. RL -0 9.038 2 Fecundity Cum. Int. RL Pre-Ov. SGR 0 -4.363 278 3 Fecundity Cum. Int. RL Pre-Ov. SGR Cum. Int. Juv. -40.833 11.973 57.654 4 In. Ov. Rate BM at Ov. 0 -7.231 5 In. Ov. Rate BM at Ov. Age at Ad. Molt -16.711 0 0.034 6 In. Ov. Rate BM at Ov. Age at Ad. Molt MS Int. Pre-Ov. -20.081 0 0.038 186.022 7 In. Ov. Rate BM at Ov. Age at Ad. Molt MS Int. Pre-Ov. Cum. Int. Pre-Ov. -23.776 28.053 19.085 18.140 22.529 29.005 23.482 14.926 0.828 0.904 0.928 0.546 0.620 0.659 0.706 11. 0 407 0 0 .016 3 22.32 3 12. See methods for a list of independent variab les tested. Significant independent variab les i selected. For all models, n = 58. All models are significant at p < 0.0001. Abbreviations: number of eggs laid during the first 6 days of the reproductive lifespan), Cum. Int. RL = pr lifespan, BM at Ov. = body mass at first oviposition, Pre-Ov. SGR = specific growth rat rior oviposition, Age at Ad. Molt = age at the adult molt, Cum. Int. J uv. = cumulative intake juvenile stages, MS int. average mass-specific intake duri ng adult stage prior to first ov iposition, Cum. Int. Pre-O ulative intake during adu prior to first oviposition. are listed In. Ov. R cumulativ e (per day) during a ll v. = cum n th ate = e int dur e or init ake ing der i ial o duri adul n w vip ng t t sta hich ositi he re ge p they w on rate oduc to Pr e lt st ere (total tive first -Ov. = age 124Table 5-6. Param

PAGE 125

Juvenile Adult AL AL-R at Ov R Ov R-AL at 5 Figure 5-1. Experimental design for Carausius morosus feeding trial. Lifespans are represented by horizontal bars divided into six instars and an adult stage. Time is not to scale, and differences in timing of life-history tran sitions between groups are not graphically presented. Vertical lines in juvenile stages denote ecdyses. White bars represent life stages when food was offered ad libitum (AL); shaded bars repr esent life stages when food was restricted (R) to 60% of the am ount of food consumed by insects in group AL on a percent body mass basi s. Insects in groups AL and R were maintained for the duration of their lifespans on ad libitum and restricted diets, respectively. Insects in the AL-R at 5th and R-AL at 5th groups experienced a diet sw itch on the first day of the fifth instar. Insects in the AL-R at Ov group experienced a diet switch at first oviposition. Because survival to first ovipos ition was extremely low for insects that were food-restricted for the duration of juve nile development, I was unable to test the effects of a diet switch from R to AL at first oviposition. Hatching Death R-AL at th AL-R at 5 th 125

PAGE 126

Figure 5-2. Mass-specific intake (g dry mass/g*day) consumed by insects in each of five treatment groups on each day of the study. Cu rves were constructed by scaling the duration of each stage for each insect to th e average duration of that stage for each treatment group and fitting a loess smoothing function to these data. Points where mass-specific intake declined to zero co rrespond to ecdyses. The first six resulting time intervals for each group represent juvenile stages and the seventh time interval for each group represents the adult stage. Arrowheads denote the average age at first oviposition for each group. Abbreviations: AL = ad libitum R = restricted, juv. = juveniles, ad. = adults. See Figure 5-1 fo r a description of diet treatments. Food restriction was imposed by offering restrict ed individuals approximately 60% of the mass-specific intake of insects in group AL on a stage-specific basis. Because mass-specific intake of AL insects declined after first ovipositi on, the amount of food offered to food-restricted adults after firs t oviposition was decreas ed proportionally to match this decline. 0 100200 0100200300400 0100200300400 300400 0100200300400 0100200300400 0.0 0 0.000 0.025 0.050 0.075 0.100 0.125 0.000 0.025 0.050 0.075 0.100g/g*day) 00 5 0.02 0.050 0.075 0.100 .125 0.000 0.025 0.050 0.075 0.100 0.125 0.000 0.025 0.050 0.075 0.100 0.125 0.125Mass-Specific Intake ( Ti me (days) Time (days) A L n = 13 AL-R at 5 th n = 13 v AL-R at O n = 13 R n = 19 juv., n = 7 ad. R-AL at 5 th n = 12 126

PAGE 127

0.0 0.2 0.4 0.6 0 050100 250300350 .8 1.0 AL AL-R at 5th AL-R at Ov R Body Mass (g) 150200Age (d) R-AL at 5thA ure 5-3. Age and size at each liist sects in each of five treatment groups. Each point represents ror at the end of an instar (first six points foeach line), at first ov point for each line), or at death (last point for each line). viat um R = restricted. See Figure 5-1 are the same as in Figure 5-2. See Table 5-2 for p-valuize int. Fig fe-h ory transition for in mean standard er iposition (sev enth ions: AL = ad libit ents. Sample sizes and age at each po r Ab es f bre or s for a description of diet treatm 127

PAGE 128

128 (a) 0 25 50 75 100 125 150 1st Instar2nd Instar3rd Instar4th Instar5th Instar6th InstarPre-Ov AdultPost-Ov AdultLength of Stage (days) AL AL-R at 5th AL-R at Ov R R-AL at 5tha a aa a a a a b b b b bb b bb b b b a b c dd a b c d d a b c d d a a b c c (b) -2.0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 1st Instar2nd Instar3rd Instar4th Instar5th Instar6th InstarPre-Ov AdultPost-Ov AdultSpecific Growth Rate (per day) AL AL-R at 5th AL-R at Ov R R-AL at 5thb b a a a a a a a a a a a a a a a b c a a a b c a a b b b b b b b c c Figure 5-4. Duration of (a) and specific growth ra te during (b) each life-history stage for insects in each of five treatment groups. Each point represents mean standard error. Stages were categorized as each of six instars, adul t prior to first oviposition (pre-ov adult), and adult after first ov iposition (post-ov adult). AL-R at Ov insects lost proportionally more body mass than AL insects after first ovipos ition, but all other pairwise comparisons of post-oviposition adult gr owth rates were not significant (MannWhitney U tests, p > 0.005). Abbreviations: AL = ad libitum R = restricted. See Figure 5-1 for a description of diet treatment s. Sample sizes are the same as in Figure 5-2. Means with different le tters are significantly diffe rent among treatment groups within life-history stages. See text for data analysis.

PAGE 129

Figure 5-5. Event history di 129 agram depicting periods of ad libitum intake, restricted intake, and reproductive activity for individu al stick insects maintained on five diet treatments. Each horizontal line represents the lifespan of one individual, with insects in each group arranged in order (top to bottom within a treatment group) from shortest to longest lifespan. Abbreviations: AL = ad libitum R = restricted, Repr. Life. = reproductive lifespan. See Figure 5-1 for a desc ription of diet treatments. Data for insects that died during the juvenile stages (i.e., insects represen ted by the shortest 1 or 2 bars for each treatment group) were not included in any analyses except for survivorship curves (Fig. 5-6).

PAGE 130

0 50100150 250300350400 ays) 200 Age (d Ad Libitum Restricted Repr. Life. Ad Libitum Repr. Life. Restricted AL, n = 15 AL-R at 5th, n = 15 AL-R at Ov, n = 14 R, n = 28 R-AL at 5th, n = 14 Adult Mo lt 130

PAGE 131

Figure 5-6. Kaplan-Meier survivorship curves for the entire lifespan (a) and for the adult lifespan (b) of insects maintain ed on five diet treatments including insects that died prior to the adult molt. Abbreviations: AL = ad libitum R = restricted. See Figure 5-1 for a description of diet treatments. For graph b, only insects that survived to adulthood are included. Only 7 of the insects in group R laid eggs, but all 19 individuals in this treatmen t group are included in the su rvivorship curve. For graph a, pairwise log-rank tests with set at 0.005 to account for multiple comparisons indicated that all groups except R and R-AL at 5th differed significantly in longevity. For graph b, AL and R-AL at 5th insects had significantly enhanced adult longevity compared to AL-R at 5th, AL-R at Ov and R insects. 131

PAGE 132

(a) 0 100 200 300 400 0.00.20.40.60.81.0 AL AL-R at 5th AL-R at Ov R R-AL at 5th 100 80 60 40 20 0 0 i 100 200 300 400Lifespan (days)Survivorship (%) AL, n = 15AL-R at 5th, n = 15AL-R at Ov, n = 14R, n = 28R-AL at 5th, n = 14 0 100 200 300 400 0.00.20.40.60.81.0 AL AL-R at 5th AL-R at Ov R R-AL at 5th 100 80 60 40 20 0 0 i 100 200 300 400Lifespan (days)Survivorship (%) AL, n = 15AL-R at 5th, n = 15AL-R at Ov, n = 14R, n = 28R-AL at 5th, n = 14 (b) 0 50 100 150 200 250 0.00.20.40.60.8 1.0 AL AL-R at 5th AL-R at Ov R R-AL at 5th AL, n = 13, n = 13AL-R at 5th, n = 13AL-R at Ov, n = 13R, n = 19R-AL at 5th, n = 12 100 80 60 40 20 0Adult Lifespan (days)Survivorship (%) 0 ii 50 i 100 i 150 i 200 250 i 0 50 100 150 200 250 0.00.20.40.60.8 1.0 AL AL-R at 5th AL-R at Ov R R-AL at 5th 100 AL, n = 13, n = 13AL-R at 5th, n = 13AL-R at Ov, n = 13R, n = 19R-AL at 5th, n = 12 80 60 40 20 0Adult Lifespan (days)Survivorship (%) 0 ii 50 i 100 i 150 i 200 250 i 132

PAGE 133

0 15 30 45 60 75 90Cumulative Fecundity (number of eggs laid) 020406080100120140 Reproductive Lifespan (days) a, w b, y b, x c, z c, zAL AL-R at Ov R-AL at 5thR AL-R at 5th Figure 5-7. Cumulative fecundity of insects in each of five treatment groups. The x-axis represents days of the reproductive lifespa n, with day 0 representing the first day of oviposition. Each curve terminates at a poi nt corresponding to the mean duration ( standard error) of reproductive activity (x) and the mean fecundity ( standard error) for each group (y). Curves were constructe d by scaling the reproductive lifespan of each insect to the mean reproductive lifes pan for that group, averaging the cumulative fecundity of all insects in that group on each day of the reproductive lifespan, and fitting a smooth spline ( df = 7) to the resulting av erages. Abbreviations: AL = ad libitum R = restricted. See Figure 5-1 for a de scription of diet treatments. Sample sizes are the same as in Figure 5-2 for adul ts. Different letters to the right of each endpoint indicate significantly different means for fecundity (a, b, and c) and reproductive lifespan (w, x, y, and z) among treatment groups. See text for data analysis. 133

PAGE 134

0.004 0.005 0.006 0.007 0.008 a c abc ac ALAL-R at 5thAL-R at Ov R R-AL at 5thAverage Egg Mass (g)b Figure 5-8. Average egg mass (mean standard erro r) for stick insects maintained on five diet treatments. Abbreviations: AL = ad libitum R = restricted. See Figure 5-1 for a description of diet treatments. Sample sizes are the same as in Figure 5-2 for adults. Means with different letter s are significantly differen t among treatment groups. See text for data analysis. 134

PAGE 135

(a) y = 28.053x + 9.0378 0 20 40 60 80 100 120 0.00.51.01.52.02.53.03.54.0 Cumulative Intake During Reproductive Lifespan (g)Total Number of Eggs Laid AL AL-R at 5th AL-R at Ov R R-AL at 5th (b) y = 0.4184x 7.9985 0 20 40 60 80 507090110130150170190210230 Adult Lifespan (days)Total Number of Eggs L 100 120aid AL AL-R at 5th AL-R at Ov R R-AL at 5th Figure 5-9. Relationships between realized fecundity and cumulative intake during the reproductive lifespan (a) and the duration of adult lifespan (b) for all insects that laid eggs ( n = 58) as determined by least square s linear regression. The regression in a is significant ( F1,56 = 276.00, p < 0.0001) with an adjusted R2 value of 0.828. The regression in b is also significant ( F1,56 = 25.67, p < 0.0001) with an adjusted R2 value of 0.302. Abbreviations: AL = ad libitum R = restricted. See Figure 5-1 for a description of diet treatments. 135

PAGE 136

CHAPTER 6 SUMMARY AND CONCLUSIONS Food availability is arguably one of the most fundamental and often-cited modulators of phenotypic and life-history plasticit y. For my dissertation, I addresse d questions about the effects of changes in food availability during different life stages in two taxa. In Chapters 2-4, I evaluated the physiological and morphologica l responses to short-term (e.g., 12-week) differences in food availability in a species (the green turtle, Chelonia mydas ) that experiences nutritional stochasticity during the juvenile stage in the wild. To elucidate long-term responses to differences in food availability, I conducted a lifespan study using a more tractable animal model (the Indian stick insect, Carausius morosus ). A summary of the major findings of my work can be found in Table 6-1. Animals living in nutritionally stochastic environments demonstrate a variety of adaptations, including the cap acity for compensatory growth (CG) (Wilson and Osbourn 1960, Reid and White 1977), that enable them to cap italize when conditions are favorable for growth Caley and growth pat was the fac hyperphagi al 2003). Additionally, working at the Cayman Turtle Farm afforded me a unique opportunity to investigate growth dynamics in gr een turtles without sampling animals taken from the wild. As a result, I was able to elucidate th e effects of intake and growth rates on a number of parameters (e.g., body composition, digestive tract morphology, nucleic acid content, and antioxidant and reproduction. Although CG has been docum ented in turtles and lizards (Bjorndal et al 2003 Schwarzkopf 2004), my work is the fi rst to assess the mechanistic basis for this tern in reptiles. One of the most salient findings to emerge from my work on C. mydas t that CG is effect ed via enhanced food conversion efficiency (FCE) rather than a. This result stands in direct contrast to most CG studies in fish (reviewed by Ali et 136

PAGE 137

function) that cannot ty pically be studied in healthy indi viduals of this endangered species. Bjorndal et al 2000) on growth rates, a comprehensive assessment of green turtle population potential fo juvenile gr g RNA, DNA, and protein content of tissues from the same turtles I examined in Chapter 2, I was able to correlate these i ndices and ratios among them with known short-term growth rates (e.g., during the preceding 10-11 days). The models I developed predicted 55-68% of the variance in recent growth rates (Chapter 3). Specific growth rate for body mass was best explained (R2 = 0.68) by RNA content of the liver and condition index (Fultons K, Ricker 1975), and specific growth rate for carapace length was best explained ( R2 = 0.66) by only RNA content of the liver. Because these analyses rely on destructive tissue extraction, they are not widely app licable to studies of growth in wild turtles. However, specific growth rate for body mass was also explained moderately well ( R2 = 0.55) by condition index Green turtles responded to food restriction by mobilizing lipid reserves, conserving protein reserves, and down-regulating the size of visceral organs. Assu ming that digestive organs are energetically expensive to maintain (Hornick et al 2000), a decrease in organ size and concomitant decreased metabolic rate may explain the improved FCE in food-restricted turtles. Turtles undergoing CG not only gr ew faster than continuously ad libitum -fed turtles but also adjusted their rates of tissue deposition such that body composition and organ morphology were restored after seven weeks of improved food conditions. Clearly, young green turtles have the capacity to adjust to fluctuations in food availability. Given this flexibility in the response to changes in nutritional condition and the previously demonstrated effects of climate (Limpus and Chaloupka 1997) and population density ( health requires a better understa nding of short-term growth dynamics. I therefore explored the r measuring a number of biochemical indices as pr edictors of recent growth in een turtles. By analyzin 137

PAGE 138

and DNA content of blood. Both of these parameters are easily quan tified with minimal disturbance to the animal, suggesting that bi ochemical indices hold promise as potential indicators of recent growth in wild turtles. After demonstrating the substantial physiol ogical and morphological plasticity of green turtles exposed to different nutritional environmen ts over short time scales, I became interested in the long-term effects of food availability. A co nspicuous feature of green turtle growth is the transient nature of CG after a switch from restricted to ad libitum feeding. In addition, the occurrence of CG indicates that normal grow th rates in this species are sub-maximal, suggesting that rapid growth may be associated with one or more costs (Metcalfe and Monaghan 2001). Mangel and Munch (2005) po sited that these costs could include elevated levels of ative damage incurred during CG. The result s I presented in Chapter 4 provide the first empirical evidence supporting this hypothesis. Although antioxidant function of muscle (a postmitotic tissue) was unaffected by diet, the activit y of glutathione peroxidase (an antioxidant enzyme) and total antioxidant potential per cell in the liver (a mitotical ly active tissue) were approximately two-fold greater in continuously ad libitum -fed turtles than in continuously foodrestricted turtles or fast-growing turtles that had undergone growth compensation. An impaired antioxidant defense system is therefore a cost of CG in green turtles. However, the duration of this impairment is unknown, as ar e its life-history consequences. The long lifespan and large body size of green turtles were not conducive to evaluating long-term responses to fluctuations in food avai lability. Instead, I took a novel approach by using one for inv find any ev oxid a parthenogenetic insect as my animal model for Ch apter 5, and this tactic proved to be a fruitful estigating questi ons about life history. In contrast to my work on C. mydas I did not idence of CG in C. morosus It is possible that an herb ivorous diet precludes growth 138

PAGE 139

compensation because digestive efficiency ized in animals consuming a low-quality diet. Although I was unable to quantify the potential life-history costs of CG, my methodology allowed me to assess the effects of different diet treatments on traits such as development rate, longevity, and fecund ity that could not be measured in C. mydas Not surprisingly, insects that experienced food re striction prior to the onset of reproductive activity progressed through juvenile stages more slowly and were smaller at each molt than individuals feeding at a consistently high rate. These results provide support for the model of Day and Rowe (2002) and suggest that develo pment rate in response to food availability represents a compromise between selection for maximized body size (because of its fitness benefits) and selection against extended develo pment time (because of its demographic costs) (Rowe and Ludwig 1991). Although my results for age and size at develo pmental transitions are typical in studies of this kind, the quantificati on of lifespan and cumulative fecundity in individuals with drastically different developmental trajectorie s is a novel contributi on of my research. My results indicate that quantitative food restricti on experienced early in development extended lifespan, as is common in other animal m odels (e.g., Weindruch and Walford 1988, Austad 1989, Mair et al 2003, Vaupel et al 2003, Hatle et al 2006b). However, this longev ity enhancement resulted from extended development time rather than enhanced adult survival. Conversely, food restriction experienced la ter in development or at maturity si gnificantly decreased total lifespan. In contrast to my results for longevity, f ood restriction imposed at any point during the lifespan decreased fecundity. Putative fitness wa s therefore maximized when daily intake was also maximized throughout life. These findings indi cate that the beneficial effects of early-onset food restriction on lifespan were negated by th e detrimental effects on reproductive output. may be maxim 139

PAGE 140

Cumulative intake during the reproductive lifespa n explained 83% of the variance in fecund indicating that C. morosus primarily allocates incoming a dult-derived resources to egg provisioning. In contrast, body stor es appear to be the source of nutrients allocated to somatic maintenance and survival during adulthood, with proportionally heavier females living adults than smaller females. Given this breed ing strategy, it is not su rprising that I found no evidence for a trade-off between longevity and f ecundity, as nutrients al located to reproductio and maintenance do not appear to be derived from a common resource pool. These results indicate that fluctuations in f ood availability can significantly alte r the expression of life-histor traits and that the magnitude of these effects depends o ity, longer as n y n the developmental stage during which food s in availability changes and on the timing of resource acquisition relative to allocation In conclusion, this dissertation provides new insights into the shortand long-term consequences of quantitative food restriction and has wide-reaching implications for studies of food availability in both vertebrates and invertebrates. Fu rthermore, the successful use of a parthenogenetic animal model unde rscores the importance of natu ral reproductive processe studies of this kind, beca use the true fitness effects of diet can only be evaluated if reproductive potential is not constrained by methodological limitations. The result s of my work highlight the need for further research into the proximate mechanisms underlyi ng differences in life histories within and among taxa. 140

PAGE 141

Table 6-1. Summary of traits measured for Chelonia mydas (top half of table) and Carau lib.) and restricted (rest.) intake. Trait Measured Continuously Ad Lib. Continuously Rest. sius morosus (bottom half of table) mainta ined on different schedules of ad libitum (ad Rest. Ad Lib. During Development Ad Lib. Rest. During Development Ad Lib. Rest. at First Oviposition C. mydas Body Size Large Small Intermediate Conversion Growth Rate Fast Slow Fast, with CG Body Condition High Low High Efficienc y Low High Intermediate OM, Lipid, and Digestive Organ Size Large Small Large Ener gy Conten t Intermediate/ High Low High Nitrogen Content Low High Low Cell Size Large Small Lar g e RNA:DNA, Heart Intermediate Low High RNA:DNA, Blood High Low Intermediate Live RNA:DNA, Liver Intermediate High Low Protein Content, r Antioxidant Function High Low Low Intermediate High Low C. morosus Body Size at First Ovi p osition Highest Lowest High Low Highest at First Ovi Age p osition Growth Rate Youngest Oldest Old Young Youngest ( Juveniles ) Fast Slow Fast, with No CG Intermediate Fast Body Condition at Adult Mol t High Low High Low High Ovary Size No Difference in Total Number of Ovarioles Among Treatment Groups Total Lifespan Long Longest Longest Short Shortest Re Lifes productive p an Highest Lowest High Lowest Low Adult Lifespan Long Short Long Short Short Lifetime Fecundity High Low Intermediate Low Intermediate Mass-Corrected Lifetime Fecundit y Intermediate/ Low Intermediate/ Low Egg Success High Low Intermediate/ Hi High Intermediate Low g h Low Intermediate Abbreviations: CG = compensatory growth, OM = organic matter, Rest. = food-restricted, Ad Lib. = ad libitum -fed. Diet switches from ad libitum to restricted feeding were not tested for C. mydas 141

PAGE 142

LIST OF REFERENCES Agarwal, S., S. Sharma, V. Agrawal, and N. Roy. 2005. Caloric restriction augments ROS 62. Ali, M., A. Nicieza, and R. J. Wootton. 2003. Co mpensatory growth in fishes: a resp growth depression. Fish and Fisheries 4:147-190. defense in S. cerevisiae, by a Sir2p independent mechan ism. Free Radical Research 39 :55onse to Altma w 72 Arendt, J. D., and D. S. Wilson. 2000. Population di fferences in the onset of cranial ossification Fisheries and Aquatic Sciences 57:351-356. Arendt, J. D., D. S. Wilson, and E. Stark. 2001. S cale strength as a cost of rapid growth in wth of juvenile Atlantic salm on. Journal of Fish Biology 68:1062-1076. is of the Atchl nn, J., and S. C. Alberts. 2005. Growth rates in a wild primate population: ecological influences and maternal effects. Behavioral Ecology and Sociobiology 57 :490-501. Altwegg, R., and H.-U. Reyer. 2003. Patterns of na tural selection on size at metamorphosis in water frogs. Evolution 57 :872-882. lvarez, D., and N. B. Metcalfe. 2005. Catc h-up growth and swimming performance in threespine sticklebacks ( Gasterosteus aculeatus ): seasonal changes in the cost of compensation. Canadian Journal of Fisheries and Aquatic Sciences 62:2169-2176. Arendt, J. D. 1997. Adaptive intrinsic growth rates: an integration across taxa. Quarterly Revie of Biology :149-177. in pumpkinseed ( Lepomis gibbosus ), a potential cost of rapid growth. Canadian Journal of sunfish. Oikos 93:95-100. Arnekleiv, J. V., A. G. Finsta d, and L. Rnning. 2006. Temporal and spatial variati on in gro Association of Official Agricult ural Chemists (AOAC) 1960. Official methods of analysis of the Association of Official Agricultural Chemists, 9th edition. Association of Official Agricultural Chemists, Washington, D. C. Association of Official Agricult ural Chemists (AOAC) 1984. Official methods of analys Association of Official Agricultural Chemists, 14th edition. Association of Official Agricultural Chemists, Washington, D. C. ey, W. R. 1984. Ontogeny, timing of develo pment, and genetic variance-covariance structure. American Naturalist 123 :519-540. Austad, S. N. 1989. Life extension by dietar y restriction in the bowl and doily spider, Frontinella pyramitela Experimental Gerontology 24:83-92. 142

PAGE 143

Balazs, G. H. 1995. Growth rates of immature gr een turtles in the Hawaiian Archipelago. Pages Smithsonian Institution Press, Washington, DC. Ballinger, R. E. 1977. Reproductive strategies: food availability as a source of proximal variati Barja, G. 2004. Aging in vertebrates, and the eff ect of caloric restrictio n: a mitochondrial free radical production-DNA damage mechanism? Biological Reviews 79:235-251. Bastrop, R., R. Spangenberg, and K. Jrss. 1991. Biochemical adaptation of juvenile carp ( Cyprinus carpio L.) to food deprivation. Comparat ive Biochemistry and Physiology A 98:143-149. Blanger, F., P. U. Blier, and J.-D. Dutil. 2002. Digestive capacity and compensatory growth in Atlantic cod (Gadus morhua). Fish Physiology and Biochemistry 26 :121-128. Berrigan, D., and J. C. Koella. 1994. The evolution of reaction norms: simple models for age an size at maturity. Journal of Evolutionary Biology 7:549-566. 117-125 in K.A. Bjorndal, editor. Biology and conser vation of sea turtles, revised edition. on in a lizard. Ecology 58:628-635. d Billerbeck, J menidia Evolution 55 :1863-1872. Bjorndal, K. A. 1997. Feeding ecology and nutrition in sea turtles. Pages 199-231 in P.L. Lutz evidence for density dependence. Ecological Applications 10 :269-282. Bjorndal, K. A., A. B. Bolten, T. Dellinger, C. Delgado, and H. R. Martins. 2003. Compensatory 84:1237-1249. 75:385-407. Boggs, C. L. 1992. Resource allocation: exploring connections between foraging and life history. Functional Ecology 6:508-518. Boggs, C. L., and K. D. Freeman. 2005. Larval food limitation in butterflies: effects on adult resource allocation and fitness. Oecologia 144 :353-361. Boggs, C. L., and C. L. Ross. 1993. The effect of adult food limitation on life history traits in Speyeria mormonia (Lepidoptera: Nymphalidae). Ecology 74:433-441. Bhm, F., R. Edge, E. Land, D. J. McGarvey, and T. G. Truscott. 1997. Carotenoids enhance vitamin E antioxidant efficiency. Journal of the American Chemical Society 119:621-622. M., T. E. Lankford, Jr., and D. O. Conover. 2001. Evolution of intrinsic growth and energy acquisition rates. I. Trad e-offs with swimming performance in Menidia and J.A. Musick, editors. The biology of sea turtles. CRC Press, Boca Raton, FL. Bjorndal, K. A., A. B. Bolten, and M. Y. Ch aloupka. 2000. Green turtle somatic growth model: growth in oceanic loggerhead sea turtles: re sponse to a stochastic environment. Ecology Blanckenhorn, W. U. 2000. The evolution of body si ze: what keeps organisms small? Quarterly Review of Biology 143

PAGE 144

Bolten, A. B., and K. A. Bjorndal. 1992. Blood pr ofiles for a wild populat ion of green turtles lonia mydas ) in the southern Bahamas: size-speci fic and sex-specific relationships. al of Wildlife Diseases 28:407-41 ( Che Journ 3. Boujard, T., C. Burel, F. Mdale, G. H and A. 000. st nut history and fasting on feed intake and growthtrou hus m tic Livingesources 13:129-137. Bradley, J. T., M. Masetti, A. Cecchettini, and F. Giorgi. 1995. Vitellogenesis in the ized stick insect Carausius morosus matodea: Lonchodinae). e Bioc 110B Bradshaw, W. E., and K. Johnson. 1995. Initiati on ohosi in the pitcher-pant ffects oal growthry. Ecolo Brock, P. D. 2000. A complete guide to breeding stick and leaf insects. Kingdom Books, Havant, Broderick, A. C., F. Glen, B. J. Godley, and G. C. Hays. 2003. Variation in reproductive output rtle f Expetal Marine Biology and Ecology 288 :95-109. Broekhuizen, N., W. S. C. Gurney, A. Jones, and A. D. Bryant. 1994. Modelling compensatory ctionaogy 8:770-782. Br Ko 001. Effect of food shortage on the physiology and competitive f sand m Riparia riparia) nestlings. Journal of Experimental Biology 65-3074. Brz e feeding on growth, development, and behavior ourishedwallow ia riparia lings. The A :1187-1 Buckley, L., E. Caldarone, and T.-L. Ong. 1999. RNA-DNA ratio and othe for growd conditi on of marine obio 65-277. Bull, C. D., and N. B. Metcalfe. 1997. Regulation overwintering 510. Bulow, F. J. 1970. RNA-DNA ratios as indicators of recent growth rates of a fish. Journal of the esearc of Can 2343-234 Calbet, A., and M. Alcaraz. 1997. Growth and surviv al rates of early developmental stages of grani (Copepoda: Calanoida) in relation to food concentra and fluc Marin 1 Cald. 2005. Eating grow haddo lanogras aeglef aylor, Moisan. 2 in rainbow Effect of pa t Oncorhync ritional ykiss. Aqua R allatectom (Br.) (Phas Comparativ hemistry and Physiology :255-266. f metamorp s l mosquito: e f larv histo gy 76:2055-2065. UK. of marine tu s. J ournal o rimen growth. Fun l Ecol z k, P., and M. narzewski. 2 abilities o 204:30 artin ( k, P., and M. Konarzewski. 2004. Effect of r of undern bank s ( Ripar ) nest uk 121 198. r nucleic acid-based logia 401 :2 indicators th an fishes. Hydr of hyperphagia in response to varying energy juvenile Atlantic salmon. Journal of Fish Biology 50:498deficits in Fisheries R h Board ada 27: 9. Acartia tion tuations in food supply e Ecology Progress Series 47 :181-186 arone, E. M stim th in ck larvae Me mmu inus from RNA:DNA ratios and water temperatur e. Marine Ecology Progress Series 293:241-252. 144

PAGE 145

Caldarone, E. M., C. M. Clemmese A. Folkvord, G. J. Holt, M. P. Olivar, and I. M. Suthers. 2006. Intercalibrati on of four spectrofluorometric protocols for y: Caley, M. J., and L. Schwarzkopf. 2004. Complex growth rate evolution in a latitudinally Calow, P. 1979. The cost of reproduction a phy siological approach. Biological Reviews of the Carey, J. R., P. Liedo, L. Harshman, X. Liu, H.-G. Mller, L. Partridge, and J.-L. Wang. 2002a. 5. g. ng Cell indivi dual fertility data and cohor t survival: case study of 1000 Carr, f outhern bluefin tuna Thunnus maccoyii (Castelnau). Carva modulation of Chalo -26 in S. gs of the 17th annual sea turtle symposium. Chap ew too prolonged fasting. Polar Biology 13:263Chian opulation to f ecundity and population growth in Drosophila melanogaster. Ecology 31:255-259. n, E. Berdalet, T. J. Miller, measuring RNA/DNA ratios in larval and juvenile fish. Limnology and Oceanograph Methods 4:153-163. widespread species. Evolution 58 :862-869. Cambridge Philosophical Society 54 :23-40. Food pulses increase longevity and induce cycl ical egg production in Mediterranean fruit flies. Functional Ecology 16:313-32 Carey, J. R., P. Liedo, L. Harshman, Y. Zha ng, H.-G. Mller, L. Partridge, and J.-L. Wan 2002b. Life history response of Mediterranean fruit flies to di etary restriction. Agi 1:140-148. Carey, J. R., P. Liedo, H.-G. Mller, J.-L. Wa ng, and J. W. Vaupel. 1998. A simple graphical technique for displaying Mediterranean Fruit Fly females. Functional Ecology 12:359-363. A. 1987. New perspectives on the pelagic st age of sea turtle development. Conservation Biology 1:103-121. Carter, C. G., G. S. Seeto, A. Smart, S. Clarke, and R. J. van Barneveld. 1998. Correlates o growth in farmed juvenile s Aquaculture 161:107-119. lho, G. B., P. Kapahi, D. J. Anderson, and S. Benzer. 2006. Allocrine feeding behavior by the sex peptide of Drosophila Current Biology 16:692-696. upka, M., and C. Limpus. 1998 Modeling green turtle surv ivorship rates. Pages 24 P. Epperly and J. Braun, editors. Proceedin NOAA technical memorandum NMFS-SEFSC-415, Miami, FL. man, R. F. 1998. The insects: structure a nd function. Cambridge University Press, N York, NY. Cherel, Y., F. Frby, J. Gilles, and J.-P. Robin. 1993. Comparative fuel metabolism in Gen and King penguins: adaptation to brief versus 269. g, H. C., and A. C. Hodson. 1950. The relati on of c 145

PAGE 146

Chippindale, A. K., A. M. Leroi, S. B. Kim, and M. R. Rose. 1993. Phenotypic plasticity and selection in Drosophila life-history evolution. I. Nutr ition and the cost of reproduction. Journal of Evolutionary Biology 6:171-193. Chong, J.-H., and R. D. Oetting. 2006. Influence of temperature and mating status on the development and fecundity of the mealybug parasitoid, Anagyrus sp. nov. nr. sinope Noyes and Menezes (Hymenoptera: Encyrt idae). Environmental Entomology 35:1188-1197. Chris th and muscle damage in a salmonid fish. Basic and Applied Myology 2:235-239. Clem easured herring larvae: laboratory calibration. Marine Biology 118:377-382. 91-1593. Dahlhoff, E. P. 2004. Biochemical indicators of stress and metabolism: applications for marine Darzynkiewicz, Z., T. Sharpless, L. StaianoCoico, and M. R. Melamed. 1980. Subcompartments sm linking curre nt reproductive effort to future survival. Proceedings of the Royal Society of London B 267 :2093-2098. Day, T., and L. Rowe. 2002. Developmental thresholds and the evolution of reaction norms for age and size at life-history transitions. American Naturalist 159 :338-350. De Cl on body weight, oviposition, egg load, and predation in the predatory stinkbug Podisus maculiventris (Heteroptera: Dhahbi, J. M., P. L. Mote, J. Wingo, B. C. Rowle y, S. X. Cao, R. L. Walford, and S. R. Spindler. tiansen, J. S., I. Martin ez, M. Jobling, and A. B. Amin. 1992. Rapid somatic grow mesen, C. 1994. The effect of food availa bility, age or size on the RNA/DNA ratio of individually m Cooper, T. M., R. J. Mockett, B. H. Sohal, R. S. Sohal, and W. C. Orr. 2004. Effect of caloric restriction on life span of the housefly, Musca domestica. FASEB Journal 18:15 Cutter, A. D. 2004. Sperm-limited fecundity in nematodes: how many sperm are enough? Evolution 58 :651-655. Dagg, M. 1977. Some effects of patchy food environments on copepods. Limnology and Oceanography 22:99-107. ecological studies. Annual Review of Physiology 66:183-207. of the G1 phase of cell cycle detected by flow cytometry. Proceedings of the National Academy of Sciences USA 77:6696-6699. David, J. R. 1970. Le nombre dovarioles chez Drosophila melanogaster: relation avec la fcundit et le valeur adaptative. Archives de Zoologie Experi mentale et Gnerale 111:357-370. Dawson, A., S. A. Hinsley, P. N. Ferns, R. H. C. Bonser, and L. Eccleston. 2000. Rate of moult affects feather quality: a mechani ercq, P., and D. Degheele. 1997. Effects of mating status Pentatomidae). Annals of the Entomological Society of America 90:121-127. 2001. Caloric restriction alters the feeding response of ke y metabolic enzyme genes. Mechanisms of Ageing and Development 122:1033-1048. 146

PAGE 147

Dixon, A. F. G., and B. K. Agarwala. 2002. Triangul ar fecundity function a nd ageing in ladyb beetles. Ecological Entomology 27:433-440. ird Dmitriew, C., and L. Rowe. 2005. Resource limitati on, predation risk and compensatory growth Einum gai nst late emergence and small offspring in Atlantic salmon ( Salmo salar ). Evolution 54:628-639. Ferro arvae. etime fecundity and fertility of the generalist herbivore Epiphyas postvittana Entomologia Experimentalis et Applicata 91:287-295. Galla t and soil analysis. Soil Science Society of America Proceedings 39 :803-806. Gaylord, T. G., and D. M. Gatlin, III. 2000. Asse ssment of compensatory growth in channel catfish Ictalurus punctatus R. and associated changes in body condition indices. Journal of Gebhardt, M. D., and S. C. Stearns. 1988. Reactio n norms for developmental time and weight at eclosion in Drosophila mercatorum Journal of Evolutionary Biology 1:335-354. Gebh Drosophila melanogaster I. Effect on phenotypic and environmental correlations. Journal Gilliers, C., R. Amara, J.-P. Bergeron, a nd O. Le Pape. 2004. Comparison of growth and -184. intestinal mucosa and liver mitochondrial redox status and function in aged rats. Experimental Gerontology 39:1323-1332. in a damselfly. Oecologia 142:150-154. S. and I. A. Fleming. 2000. Selection a n, A., and W. C. Leggett. 1994. An appraisal of condition measures for marine fish l Advances in Marine Biology 30:217-303. Forman, R. T. T. 1995. Land mosaics: the eco logy of landscapes and regions. Cambridge University Press, Cambridge, UK. Foster, S. P., and A. J. Howard. 1999. The effect s of mating, age at mating, and plant stimuli, on the lif her, R. N., C. O. Weldon, and J. G. Futral. 1975. An aluminum block digester for plan the World Aquaculture Society 31 :326-336. ardt, M. D., and S. C. Stearns. 1993. Phe notypic plasticity for lif e history traits in of Evolutionary Biology 6:1-16. condition indices of juvenile flatfish in different coastal nursery grounds. Environmental Biology of Fishes 71:189-198. Gillott, C. 2003. Male accessory gland secretions : modulators of female reproductive physiology and behavior. Annual Review of Entomology 48:163 Grafen, A. 1988. On the uses of data on lifetime reproductive success. Pages 454-471 in T. H. Clutton-Brock, editor. Reproductive success: stud ies of individual variation in contrasting breeding systems. University of Chicago Press, Chicago, IL. Grattagliano, I., P. Portincasa, T. Cocco, A. Mo schetta, M. Di Paola, V. O. Palmieri, and G. Palasciano. 2004. Effect of diet ary restriction and N-acetylcysteine supplementation on 147

PAGE 148

Gredilla, R., and G. Barja. 2005. Minireview: the role of oxidative stress in relation to caloric restriction and l ongevity. Endocrinology 146:3713-3717. pian, K., J. J. Ramsey, Hago and R. Weindruch. 2003. Caloric restriction increases gluconeogenic and transaminase enzyme activities in mouse liver. Experimental Gerontology 38:267-278. Hamb r simultaneous determination of phosphorus, calcium and crude protein in animal feeds. Journal of the Association of Official Analytical Hart, R. W., and A. Turturro. 1998. Evolution and dietary restriction. Expe rimental Gerontology Hatle 004. Plasticity and canalizat ion in the control of reproduction in the lubber grasshopper. Integrative and Comparative Biology 43:635-645. Hatle is in the l ubber grasshopper. Archives of Insect Biochemistry and Physiology 52:45-57. Hatle body mass. Journal of Comparative Physiology B 176:27-34. Hatle E. Fuller, I. C. Allen, L. J. Gordy, S. Melnyk, and J. Quattrochi. 2006b. Calorie restriction and late-onset calorie restriction ex tend lifespan but do not alter 83Hayes, J. P., and J. S. Shonkwiler. 2001. Morphometric indicators of bod y condition: worthwhile mbridge, UK. Hofer :315-331. imals an evolu tionary adaptation? Bioessays 10:125-127. Hornick, J. L., C. Van Eenaeme, O. Grard, I. Dufrasne, and L. Istasse. 2000. Mechanisms of Hume, I. D., and H. Biebach. 1996. Digestive tract function in the l ong-distance migratory garden warbler, Sylvia borin. Journal of Comparative Physiology 166B :388-395. leton, L. G. 1977. Semiautomated method fo Chemists 60 :845-852. 33:53-60. J. D., D. W. Borst, and S. A. Juliano. 2 J. D., S. A. Juliano, and D. W. Bors t. 2003. Hemolymph ecdysteroids do not affect vitellogenes J. D., T. Waskey, Jr., and S. A. Juliano. 2006a. Plasticity of grasshopper vitellogenin production in response to diet is primarily a re sult of changes in fat J. D., S. M. Wells, L protein storage in female grasshoppers. Mechanisms of Ageing and Development 127:8 891. or wishful thinking? Pages 8-38 in J. R. Speakman, editor. Body composition analysis of animals: a handbook of non-destructive methods. Cambridge University Press, Ca H., and M. L. East. 2003. Behavioral pr ocesses and costs of co -existence in female spotted hyenas: a life history perspective. Evolutionary Ecology 17 Holliday, R. 1989. Food, reproduction and longevity : is the extended lifespan of calorierestricted an Hon k, A. 1993. Intraspecific variation in body si ze and fecundity in insects: a general relationship. Oikos 66:483-492. reduced and compensatory growt h. Domestic Animal Endocrinology 19:121-132. 148

PAGE 149

Hyun, D.-H., S. S. Emerson, D.-G. Jo, M. P. Mattson, and R. de Cabo. 2006. Calorie restrictio up-regulates the plasma membrane redox system in brain cells and suppresses oxidative stress during aging. Proceedings of the Natio n nal Academy of Sciences 103 :19908-19912. early 83. 38. Letters 448:4-8. Jervis, M. A., C. L. Boggs, and P. N. Ferns. 2005. Egg maturation strategy and its associated trade-offs: a synthesis focusing on Lepidoptera. Ecological Entomology 30:359-375. Jespersen, L. B., and S. Toft. 2003. Compensatory growth following early nutritional stress in the wolf spider Pardosa prativaga Functional Ecology 17:737-746. Ji, L. ce s of free radicals and their impact on antioxidant systems. Age 20:91-106. Joblin ition ic uaculture Research 32 :963-974. Joshi, A., W.-P. Wu, and L. D. Mueller. 1998. Density-dependent na tural selection in Judge, S., Y. M. Jang, A. Smith, T. Hagen, and C. Leeuwenburgh. 2005. Age-associated :419986Islam, M. S., and M. Tanaka. 2005. Nutritional condition, starvation status and growth of juvenile Japanese sea bass ( Lateolabrax japonicus ) related to prey di stribution and feeding in the nursery ground. Journal of Experi mental Marine Biology and Ecology 323:172-1 Jackson, J. B. C., M. X. Kirby, W. H. Berger, K. A. Bjorndal, and 15 others. 2001. Historical overfishing and the recent collapse of coastal ecosystems. Science 293:629-6 Janzen, F. J., J. K. Tucker, and G. L. Paukstis 2000. Experimental anal ysis of an early lifehistory stage: selection on si ze of hatchling turtles. Ecology 81 :2290-2304. Jennings, B. J., S. E. Ozanne, M. W. Dorling, and C. N. Hales. 1999. Early growth determines longevity in male rats and may be related to telomere shortening in the kidney. FEBS L., and S. Leichtweis. 1997. Exercise and oxidative stress: Sour g, M., and S. J. S. Johansen. 1999. The lipostat, hyperphagia and catch-up growth. Aquaculture Research 30 :473-478. Johansen, S. J. S., M. Ekli, B. Stangnes, and M. Jobling. 2001. Weight ga in and lipid depos in Atlantic salmon, Salmo salar during compensatory growth: evidence for lipostat regulation? Aq Jnsson, K. I. 1997. Capital and income breeding as alternative tactics of resource use in reproduction. Oikos 78:57-66. Drosophila : adaptation to adult crowding. Evolutionary Ecology 12: 363-376. increases in oxidative stress and antioxidant enzyme activities in cardiac interfibrillar mitochondria: implications for the mitoc hondrial theory of aging. FASEB Journal 19 421. Juliano, S. A., J. R. Olson, E. G. Murrell, and J. D. Hatle. 2004. Plasticity and canalization of insect reproduction: testing alternative models of life histor y transitions. Ecology 85:2 2996. 149

PAGE 150

Kaeberlein, T. L., E. D. Smith, M. Tsuchiya, K. L. Welton, J. H. Thomas, S. Fields, B. K. Kennedy, and M. Kaeberlein. 2006. Lifespan extension in Caenorhabditis elegans by complete removal of food. Aging Cell 5 :487-494. Kalan Leeuwenburgh. 2006. Effects of caloric restriction and exercise on age-related, chronic inflammation assessed by C-reactive Karasov, W. H., and B. Pinshow. 1998. Changes in lean mass and in organs of nutrient igrating blackcaps ( Sylvia atricapilla ): a comparison Kena s of food intake and feces Kunz Evolutionary Biology 13:199-212. 002. ericanus and tautog Tautoga onitis Marine Ecology Leath inse cts: things arent as simple as they seem. Oikos 51:386-389. Lee, K nd Biochemical Zoology i, R., S. Judge, C. Carter, M. Pahor and C protein and interleukin-6. Journal of Gerontology 61A :211-217. assimilation in a long-dis tance passerine migrant at a springtime stopover site. Physiological Zoology 71 :435-448. Karasov, W. H., B. Pinshow, J. M. Starck, and D. Afik. 2004. Anatomi cal and histological changes in the alimentary tr act of m among fed, fasted, food-restricted, and refe d birds. Physiological and Biochemical Zoology 77:149-160. gy, G. J., C. Veloso, and F. Bozinovic. 1999. Daily rhythm reingestion in the degu, an herbivorous Chilean rodent: optimizing digestion through coprophagy. Physiological and Biochemical Zoology 72:78-86. Kindlmann, P., A. F. G. Dixon, and I. Dostl kov. 2001. Role of ageing and temperature in shaping reaction norms and fecundity functions in insects. Jo urnal of Evolutionary Biology 14:835-840. Kitaysky, A. S. 1999. Metabolic and developmental responses of alcid chicks to experimental variation in food intake. Physiological and Biochemical Zoology 72:462-473. Klowden, M. J. 2002. Physiological systems in insects. Academic Press, New York, NY. C., and J. Ekman. 2000. Genetic and environm ental components of growth in nestling blue tits ( Parus caeruleus ). Journal of Kuropat, C., R. Mercaldo-Allen, E. Caldarone, R. Goldberg, B. Phelan, and F. Thurberg. 2 Evaluation of RNA concentration as an indicator of growth in young-of-the-year winter flounder Pseudopleuronectes am Progress Series 230:265-274. er, S. R. 1988. Size, reproductiv e potential and fecundity in A., W. H. Karasov, and E. Caviedes -Vidal. 2002. Digestive re sponse to restricted feeding in migratory yellow-rumped warble rs. Physiological a 75:314-323. 150

PAGE 151

Leeuwenburgh, C., R. Fiebig, R. Chandwaney, and L. L. Ji. 1994. Aging and exercise training skeletal muscle: responses of glutathione and antioxidant e in nzyme systems. American Journal of Physiology Regulatory, Inte grative, and Comp arative Physiology 267:R439Leips, J., and J. Travis. 1994. Metamorphic respons es to changing food levels in two species of Leterrier, C., and Y. Nys. 1992. Composition, cor tical structure and mechanical properties of chicken tibiotarsi: effect of grow th rate. British Poultry Science 33:925-939. Li, S., T. Yan, J.-Q. Yang, T. D. Oberley, and L. W. Oberley. 2000. The role of cellular glutathione peroxida se redox regulation in the suppression of tumor cell growth by Limpus, C. J. 1992. Estimation of tag loss in marine turtle research. Wildlife Research 19:457Limpus, C., and M. Chaloupka. 1997. Nonparametric regression modelling of green sea turtle growth rates (southern Great Barrier Reef). Marine Ecology Progress Series 149:23-34. Lochm idus ) as an adaptation to nutritionally stochastic environments. Evolutionary Ecology 14:127-142. Lochm ology 104A :593-599. Luhtala, T. A., E. B. Roecker, T. Pugh, R. J. Feuers, and R. Weindruc h. 1994. Dietary restriction ly in life nimal Ecology 69:952-958. Mangel, M., and S. B. Munch. 2005. A life-hist ory perspective on shortand long-term consequences of compensatory growth. American Naturalist 166 :E155-E176. Mang tenance of individual variation in growt h. Evolutionary Ecology Research 3:583-593. R445. hylid frogs. Ecology 75:1345-1356. manganese superoxide dismutase. Cancer Research 60:3927-3939. 469. iller, R. L., S. S. Ditchkoff, and J. A. Sinclair. 2000. Developmental plasticity of postweanling cotton rats (Sigmodon hisp iller, R. L., M. R. Vestey, and S. T. McMurry. 1993. Selected immune responses of adult cotton rats ( Sigmodon hispidus ) to dietary restriction. Co mparative Biochemistry and Physi Lpez-Torres, M., R. Gredilla, A. Sanz, and G. Barja. 2002. Influence of aging and long-term caloric restriction on oxygen radical generation and oxidati ve DNA damage in rat liver mitochondria. Free Radical Biology and Medicine 32:882-889. attenuates age-related increases in rat skel etal muscle antioxidant enzyme activities. Journals of Gerontology 49:B231-B238. Madsen, T. and R. Shine. 2000. Silver spoons and sn ake body sizes: prey availa bility ear influences long-term growth rates of fr ee-ranging pythons. Journal of A el, M., and J. Stamps. 2001. Trade-offs between growth and mortality and the main 151

PAGE 152

Mair, W., P. Goymer, S. D. Pletcher, and L. Partridge. 2003. Demography of dietary restriction and death in Drosophila Science 301 :1731-1733. Mair, W., C. M. Sgr, A. P. Johnson, T. Chapma n, and L. Partridge. 2004. Lifespan extension by Masoro, E. J., and S. N. Austad. 1996. The e volution of the antiaging action of dietary Maxwell, M. H., P. M. Hocking, and G. W. Ro bertson. 1992. Differential leucocyte responses to ce ritish Poultry Science 31:407-413. f lls and thrombocytes. British Poultry Science 31 :399-405. McCa tary dilution from debris ingestion: sublethal e ffects in post-hatchling loggerh ead sea turtles. Conservation McDonald, P. G., P. D. Olsen, and A. Cockbur n. 2005. Sex allocation and nestling survival in a Melzner, F., J. W. Forsythe, P. G. Lee, J. B. Wood, U. Piatkowski, and C. Clemmesen. 2005. rine in ga onitis based on RNA/DNA ratio and seawater Metca y later? TRENDS in Ecology and Evolution 16:254-260. Metca nary Miglavs, I., and M. Jobling. 1989. Effects of feeding regime on food consumption, growth rates ct dietary restriction in female Drosophila melanogaster is not caused by a reduction in vitellogenesis or ovarian activ ity. Experimental Gerontology 39 :1011-1019. restriction: a hypothesis. Journal of Gerontology 51A:B387-B391. various degrees of food restrict ion in broilers, turkeys and du cks. British Poultry Scien 33:177-187. Maxwell, M. H., G. W. Robertson, S. Spence, and C. C. McCorquodale. 1990a. Comparison of haemotological values in restrictedand ad libitum -fed domestic fowl s: red blood cell characteristics. B Maxwell, M. H., G. W. Robertson, S. Spen ce, and C. C. McCorquodale. 1990b. Comparison o haemotological values in restrictedand ad libitum -fed domestic fowls: white blood ce uley, S. J., and K. A. Bjorndal. 1999. Conser vation implications of die Biology 13:925-929. dimorphic raptor: does size matter? Behavioral Ecology 16:922-930. Estimating recent growth in the cuttlefish Sepia officinalis : are nucleic acid-based indicators for growth and condition the method of choice? Journal of Experimental Ma Biology and Ecology 317 :37-51. Mercaldo-Allen, R., C. Kuropat, and E. M. Caldarone. 2006. A model to estimate growth young-of-the-year tautog, Tauto temperature. 2006. Journal of Experimental Marine Biology and Ecology 329:187-195. lfe, N. B., and P. Monaghan. 2001. Compensa tion for a bad start: grow now, pa lfe, N. B., and P. Monaghan. 2003. Growth vers us lifespan: perspectives from evolutio ecology. Experimental Gerontology 38:935-940. and tissue nucleic acids in juvenile Arctic charr, Salvelinus alpinus with particular respe to compensatory growth. Journal of Fish Biology 34:947-957. 152

PAGE 153

Moehrlin, G. S., and S. A. Juliano. 1998. Plastici ty of insect reproduction: testing models o flexible and fixed development in respons e to different growth rates. Oecologia 115:49 500. f 2also 6:205-212. ent in Mura, C. V., X. Gong, A. Taylor, R. Villalobos-Molina, and M. M. Scrofano. 1996. Effects of the Music pus. 1997. Habitat ut ilization and migration in juvenile sea turtles. Pages 137-163 in P.L. Lutz and J.A. Musick, editors. The biology of sea turtles. CRC Nagy mi gratory songbird: an experimental study. Ecology 86:675-681. Naka roxidase. Bi ochimica et Biophysica Acta. 358:251-261. Niki, E., N. Noguchi, H. Tsuchihashi, and N. Gotoh. 1995. Interaction among vitamin C, vitamin quaculture 235 :285296. Okum ll tropoda). Fisheries Science 68 :306-312. American Naturalist 149 :179-188. Packard, G. C., J. K. Tuc Trachemys scripta ). Copeia 1997:339-345. Parr Instrument Company. 1960. Oxygen bomb calor imetry and combustion methods. Technical manual 130:1-56. Moreau, J., B. Benrey, and D. Thiry. 2006. Grape variety affects larval performance and female reproductive performance of the European grapevine moth Lobesia botrana (Lepidoptera: Tortricidae). Bulle tin of Entomological Research 9 Morey, S., and D. Reznick. 2000. A comparative analys is of plasticity in larval developm three species of spadefoot toads. Ecology 81:1736-1749. calorie restriction and aging on the expression of antioxidant enzymes and ubiquitin in liver of Emory mice. Mechanisms of Ageing and Development 91:115-129. k, J. A., and C. J. Lim Press, Boca Raton, FL. L. R., and R. T. Holmes. 2005. Food limits annual fecundity of a mura, W., S. Hosoda, and K. Hayashi. 1974. Purification and properties of rat liver glutathione pe Nijhout, H. F. 1994. Insect hormones. Prin ceton University Press, Princeton, NJ. E, and -carotene. American Journa l of Clinical Nutrition 62 (suppl) :1322S-1326S. Nikki, J., J. Pirhonen, M. Jobling, and J. Karj alainen. 2004. Compensatory growth in juvenile rainbow trout, Oncorhynchus mykiss (Walbaum), held individually. A ura, T., T. Nagasawa, I. Hayashi, and Y. Sato. 2002. Effects of starvation on RNA:DNA ratio, glycogen content, and C:N ratio in colu mellar muscle of the Japanese turban she Turbo ( Batillus) cornutus (Gas Olsson, M., and R. Shine. 1997. The limits to repr oductive output: offspring size versus number in the sand lizard (Lacerta agilis). Olsson, M., and R. Shine. 2002. Growth to death in lizards. Evolution 56 :1867-1870. ker, D. Nicholson, and M. J. Packard. 1997. Cold tolerance in hatchling slider turtles ( 153

PAGE 154

Partridge, L, D. Gems, and D. J. Withers. 2005. Sex and death: what is the connection? Cell 120:461-472. Partridge, L, M. D. W. Piper, and W. Mair. 2005. Dietary restriction in Drosophila Mechanisms of Ageing and Development 126:938-950. Patterson, D. C., R. W. J. Steen, and D. J. K ilpatrick. 1995. Growth and development in beef cattle. 1. Direct and residual e ffects of plane of nutrition duri ng early life on components of Perneger, T. V. 1998. Whats wrong with Bonfe rroni adjustments? Br itish Medical Journal Perrin tion in Daphnia magna. Oikos 59:70-74. Perrin, N., and R. M. Sibly. 1993. Dynamic models of energy allocation and investment. Annual Review of Ecology and Systematics 24:379-410. Phane g heart: a possible mechanism for apoptosis wi th age. American Journal of Physiology Pijnac rthenogenetic stick insect Carausius morosus Br. (Orthoptera, Phasmidae). Chromosoma 19:99-112. Pltz, llii at ea (Antarctica), in relation to local food supply. Marine Mammal Science 7 :136-144. Prasa of rosophila melanogaster Evolutionary Ecology Research 5:903-911. Quinn Y. ciences. -27 in June 1977, Cornell University, Ithaca, NY. Rezni kos 44:257267. gain and food efficiency. Jour nal of Agricultural Science 124 :91-100. 316:1236-1238. N., M. C. Bradley, and P. Calow. 1990. Plasticity of storage alloca uf, S., and C. Leeuwenburgh. 2002. Cytochrome c release from mitochondria in the agin Regulatory, Integrative, and Comparative Physiology 282:R423-R430. ker, L. P. 1966. The maturation division s of the pa J., W. Ekau, and P. J. H. Reijnders. 1991. Diet of Weddell seals Leptonychotes wedde Vestkapp, Eastern Weddell S d, N. G., M. Shakarad, M. Rajamani, and A. Joshi. 2003. Interaction between the effects maternal and larval levels of nutrition on pre-adu lt survival in D G. P., and M. J. Keough. 2002. Experimental design and data anal ysis for biologists Cambridge University Press, New York, N Reich, K. J., K. A. Bjorndal, and A. B. Bolten. In review. Testing Carrs conjecture: green turtles shift from oceanic to ner itic habitats during ontogeny. Proceedings of the National Academy of S Reid, J. T., and O. D. White. 1977. The phenom enon of compensatory growth. Pages 16 Proceedings of the Cornell Nutrition Conf erence for Feed Manufacturers 21-23 ck, D. 1985. Costs of reproduction: an ev aluation of the empirical evidence. Oi 154

PAGE 155

Reznick, D. 1992. Measuring the costs of reproduction. Trends in Ec ology and Evolution 7:4245. Reznick, D., H. Callahan, and R. Llauredo. 1996. Maternal effects on offspring quality in poeciliid fishes. American Zoologist 36:147-156. Richner, H., P. Schneiter, and H. Stirnimann. 198 9. Life-history consequences of growth rate unctional Roff, D. A. 1992. The evolution of life historie s: theory and analysis. Chapman and Hall, New Rook NA ratios to evaluate the condition and growth of larval and juvenile red drum ( Sciaenops ocellatus). Marine and Freshwater Rook rval and juvenile red drum ( Sciaenops ocellatus ) from estuarine nursery habitats. Marine Biology 127 :387-394. Roseb broiler chickens. Growth, Development and Aging 57:73-83. n Schew and R. E. Ricklefs, editors. Avian growth a nd development: evolution within the altricialSchra e Scriber, J. M., and F. Slansky, Jr. 1981. The nutritional ecology of immature insects. Annual Seminoff, J. A. 2002. IUCN Red List gl ent: green turtle ( Chelonia mydas). depression: an experiment al study on carrion crows (Corvus corone corone L.). F Ecology 3:617-624. Ricker, W. E. 1975. Computation a nd interpretation of biological statistics of fish populations. Bulletin of the Fisheries Research Board of Canada 191:1-382. York, NY. er, J. R., and G. J. Holt. 1996. Applicati on of RNA:D Research 47 :283-290. er, J. R., G. J. Holt, and S. A. Holt. 1997. Condition of la rough, R. W., and J. P. McMurtry. 1993. En ergy repletion and lipid metabolism during compensatory gain in Rowe, L., and D. Ludwig. 1991. Size and timing of metamorphosis in co mplex life cycles: time constraints and variation. Ecology 72:413-427. Rowe, L., D. Ludwig, and D. Schluter. 1994. Time condition, and the seas onal decline of avia clutch size. American Naturalist 143 :698-772. W. A., and R. E. Ricklefs. 1998. Developmental plasticity. Pages 288-304 in J. M. Starck precocial spectrum. Oxford University Press, New York, NY. din, C., and N. Pillay. 2006. Female striped mice ( Rhabdomys pumilio ) change their hom ranges in response to seasonal variation in food availabilit y. Behavioral Ecology 17:452458. Review of Entomology 26 :183-211. obal status assessm IUCN/SSC Marine Turtle Sp ecialist Group, Gland, Switzerland. 155

PAGE 156

Shin, H.-C., S. Nicol, and R. A. King. 2003. Nucl eic acid content as a po tential growth rate estimator of Antarctic krill; results from fiel d-caught krill from the Indian sector of the South ern Ocean. Marine and Fres hwater Behaviour and Physiology 36:295-305. 5. Simm nse to dietary yeast in Drosophila melanogaster. Journal of Insect Physiology 43 :779-788. Skals Variable intake, compensatory growth, and increased growth efficiency in fish: models and mechanisms. Ecology Smith l growth in the spiny lizard, Sceloporus jarrovi Copeia 1994 :1007-1013. Smith ic oologist 23 :47-55. enzymes associated with protein renewal during ag ing. Annals of the New York Academy of Starck, J. M., and K. Beese. 2002. Structural flexib ility of the small intes tine and liver of garter Stearns, S. C. 1982. The role of development in the evolution of life histories. Pages 237-258 in Stearns, S. C. 1992. The evolution of life histories. Oxford University Press, New York, NY. Store ey. 1993. Freezing survival and metabolism of box turtles, Terrapene carolina. Copeia 1993 :628-634. Tammaru, T., P. Kaitaniemi, and K. Ruohomki. 1996. Realized fecundity in Epirrita autumnata on Science 98:205-218. Sies, H. 1997. Oxidative stress: oxidants a nd antioxidants. Experimental Physiology 82 :291-29 ons, F. H., and T. J. Bradley. 1997. An analysis of resource allocation in respo ki, G. T., M. E. Picha, J. F. Gilliam, a nd R. J. Borski. 2005 86:1452-1462. G. R., and R. E. Ballinger. 1994. Temporal and spatial variation in individua -Gill, S. J. 1983. Developmental plasticity: developmental conversion versus phenotyp modulation. American Z Solomon, S. E., and R. Tippett. 1991. Lipid inclusion in the liver s of captive reared marine turtles. Animal Technology 42:77-81. Speakman, J. R., and E. Krl. 2005. Limits to sustained energy intake IX: a review of hypotheses. Journal of Comparative Physiology 175B :375-394. Spindler, S. R. 2001. Caloric re striction enhances the expre ssion of key metabolic Sciences 928 :296-304. snakes in response to feeding and fas ting. Journal of Expe rimental Biology 205:13771388. J. T. Bonner, editor. Evolution and developm ent. Springer-Verlag Publishers, New York, NY. y, K. B., J. R. Layne, Jr., M. M. Cutwa, T. A. Churchill, and J. M. Stor (Lepidoptera: Geometridae): relation to body size and consequences to population dynamics. Oikos 77:407-416. Therkildsen, M. 2005. Muscle protein degradatio n in bull calves with compensatory growth. Livestock Producti 156

PAGE 157

Tian, X., and J. G. Qin. 2004. Effects of previous ration restriction on compensatory growth i barramundi La n tes calcarifer. Aquaculture 235:273-283. -333. al van Noordwijk, A. J., and G. de Jong. 1986. Acqui sition and allocation of resources: their Vaupel, J. W., J. R. Carey, and K. Chri stensen. 2003. Its never too late. Science 301 :1679-1681. Vidal, A. G., P. DiMarco, and P. Lee. 2006. E ffects of starvation and r ecovery on the survival, growth and RNA/DNA ratio in loliginid squid paralarvae. Aquaculture 260 :94-105. Voigt M. Joachimski. 2006. Nuptial feeding is reflected in tissue nitroge n isotope ratios of female katydids. Functional Ecology 20 :656Vrede C ratio) on RNA:DNA ratio and somatic growth rate of Daphnia. Limnology and Oceanography Walford, R. L., S. B. Harris, and M. W. Guni on. 1992. The calorically restricted low-fat nutrient7. Watso 98. Sexual conflict and th e energetic costs of mating and mate choice in water striders. American Naturalist 151:46-58. Wayn ental stress and reproduction in Drosophila melanogaster: starvation resistance, ovariole numbers and Weindruch, R., and R. S. Sohal. 1997. Caloric intake and aging. New England Journal of Medicine 337 :986-994. Wein pringfield, IL. h of the red drum Sciaenops ocellatus. Marine Biology 121:1-9. Tu, M.-P., and M. Tatar. 2003. Juvenile diet rest riction and the aging a nd reproduction of adult Drosophila melanogaster Aging Cell 2:327 Turturro, A. and R. Hart. 1992. Di etary alteration in the rates of cancer and aging. Experiment Gerontology 27:583-592. influence on variation in life hist ory tactics. American Naturalist 128:137-142. C. C., G. U. C. Lehmann, R. H. Michener, and M 661. T., J. Persson, and G. Aronsen. 2002. Th e influence of food quality (P: 47:487-494. dense diet in Biosphere 2 significantly lo wers blood glucose, total leukocyte count, cholesterol, and blood pressure in humans. Proceedings of the National Academy of Sciences USA 89:11533-1153 Wallace, B. 1992. The search for the gene. Cornell University Press, Ithaca, NY. n, P. J., G. Arnqvist, and R. R. Stallmann. 19 e, M. L., U. Soundararajan, and L. G. Harshman. 2006. Environm early age egg production. BMC Evolutionary Biology 6:Article 57. druch, R., and R. L. Walford. 1988. The re tardation of aging and disease by dietary restriction. Charles C. Thomas, S Westerman, M., and G. J. Holt. 1994. RNA:DNA ratio during the crit ical period and early larval growt 157

PAGE 158

Wheeler, D. 1996. The role of nourishment in oogenesis. Annual Review of Entomology 41:407 431. e in the diamondback moth, Plutella xylostella (Lepidoptera). Journal of Insect Physiology 46:951-958. Wilbu ociety 35 :324-363. ammals. American Naturalist 167 :E23-E38. Wood, F. E., and G. K. Ebanks. 1984. Blood cytol ogy and hematology of the green sea turtle, Wu, A., X. Sun, F. Wan, and Y. Liu. 2003. Modul ations by dietary rest riction on antioxidant 5-126. Wheeler, D. E., I. Tuchinskaya, N. A. Buck, and B. E. Tabashnik. 2000. Hexameric storag proteins during metamorphosis and e gg production r, H. M., and J. P. Collins. 1973. Ecological aspects of amphibian metamorphosis. Science 182:1305-1314. Wilson, P. N., and D. F. Osbourn. 1960. Compensato ry growth after undernutrition in mammals and birds. Biological Reviews of the Cambridge Philosophical S Wilson, A. J., and D. Rale. 2006. Ontogeny of a dditive and maternal genetic effects: lessons from domestic m Wolfner, M. F. 1997. Tokens of l ove: functions and regulation of Drosophila male accessory gland products. Insect Biochemistry and Molecular Biology 27 :179-192. Chelonia mydas Herpetologica 40 :331-336. enzymes and lipid peroxidation in developing mice. Journal of Applied Physiology 94:947-952. Zera, A. J., and L. G. Harshman. 2001. The physiology of life history trade-offs in animals. Annual Review of Ecology and Systematics 32:9 158

PAGE 159

BIOGRAPHICAL SKETCH Alison M. Roark was born in Norfolk, Virginia on October 26th, 1978. She atten ded Mills ) and also f est resear d also completed two r, she spent three weeks in San Salva as, for a class in coral reef ecology and credits this c the direction of Karen Bjorndal. In the spring of 2003, she completed her masters bypass. While at anatoratory and the discussion secti ons and laboratories for both semesters of biology course (Cells, Organisms, and Gene tics). In the summer of 2007, Alison begins a Godwin High School in Richmond, Virginia, where she was a leader in the marching and concert bands. After graduating in 1996, Alison attended the University of Virginia (Charlottesville obtained her Bachelor of Science in chemistry wi th specialization in biochemistry in 2000. She ulfilled the requirements for the Disti nguished Majors Program in biology with high distinction and served as Presid ent of the Biology Society. Alison participated in undergraduate ch in two different laborat ories at the University of Virginia an Research Experience for Undergraduates programs one at the University of Texas (Austin) and one at the Long-Term Ecological Re search station in Oyster, Virgin ia. After her second yea dor, The Baham ourse with steering her toward a career in academia. In 2000, Alison joined the Department of Z oology at the University of Florida under the University of Florida, Alison taught a numbe r of classes, includi ng functional vertebrate my labo introductory biology. In her final year as a graduate student, she taught her own non-majors postdoctoral position in the laboratory of Lou Guillette through the Howard Hughes Medical Institutes Group Advantaged Training of Research (G.A.T.O.R.) program. 159