GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  ·...

112
GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca microlepis CONSUMING FISH AND CRUSTACEAN PREY By ELIZABETH JOANNE BERENS A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2005

Transcript of GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  ·...

Page 1: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca microlepis CONSUMING FISH AND CRUSTACEAN PREY

By

ELIZABETH JOANNE BERENS

A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT

OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE

UNIVERSITY OF FLORIDA

2005

Page 2: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

Copyright 2005

by

Elizabeth Joanne Berens

Page 3: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

To my parents,

Charles and Joanne, For their support and encouragement

which has allowed me to pursue my interests in the marine environment

and the various organisms within.

Page 4: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

ACKNOWLEDGMENTS

I would like to thank Dr. Mark Luttenton for introducing me to fisheries science

many years ago at Grand Valley State University. This study would not have been

possible without the help of many people and I especially thank Mark Butler, Jackie

Debicella, Rick Kline, Steve Larsen, Eddie Leonard, and Doug Marcinek for assistance

with laboratory set-up and specimen collection, and Doug Colle and Larry Tolbert for

advice and assistance with certain aspects of this study. In particular, I thank Day Cherry

and DJ White for their assistance with invertebrate collections and for giving me

invaluable insight into a commercial fishing industry. I thank the American Fisheries

Society, Florida Chapter, Florida Sea Grant, and the University of Florida’s Fisheries and

Aquatic Sciences Department for financial and logistical assistance during this study.

I thank my committee, Dr. Debra Murie, Dr. Daryl Parkyn, Dr. William Lindberg,

University of Florida, Department of Fisheries and Aquatic Sciences, and Dr. Karen

Bjorndal, University of Florida, Department of Zoology, for offering advice and

assistance along the way.

iv

Page 5: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

TABLE OF CONTENTS page

ACKNOWLEDGMENTS ................................................................................................. iv

LIST OF TABLES............................................................................................................ vii

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

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

CHAPTER

1 INTRODUCTION ........................................................................................................1

2 METHODS.................................................................................................................19

Gag Collections and Maintenance ..............................................................................19 Experimental Feeding Trials for Gastric Evacuation Rates .......................................21 Caloric Analysis of Prey and Stomach Contents........................................................22 Gastric Evacuation Models.........................................................................................25 Energetic Models ........................................................................................................26 Indices of Digestive States..........................................................................................28 Models of Average Digestion Codes..........................................................................29

3 RESULTS...................................................................................................................32

Gag and Prey Collections ...........................................................................................32 Gastric Evacuation Models for Gag Consuming Baitfish Prey ..................................32 Gastric Evacuation Models for Gag Consuming Crab Prey.......................................34 Comparative Gastric Evacuation Models for Gag Consuming Baitfish versus Crab

Prey ........................................................................................................................36 Caloric Values of Baitfish and Crab Prey...................................................................36 Energy Values for Recovered Gag Stomach Contents ...............................................37 Indices of Digestion for Baitfish Consumed by Gag..................................................40 Indices of Digestion for Crab Prey Consumed by Gag ..............................................41

4 DISCUSSION.............................................................................................................69

Gastric Evacuation Models.........................................................................................69 Effects of Prey Type............................................................................................74

v

Page 6: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

Effects of Predator Size .......................................................................................77 Prey Composition .......................................................................................................79 Stomach Content Composition ...................................................................................83 Indices of Digestion....................................................................................................86 Consumption...............................................................................................................88 Conclusions.................................................................................................................90

LIST OF REFERENCES...................................................................................................92

BIOGRAPHICAL SKETCH .............................................................................................99

vi

Page 7: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

LIST OF TABLES

Table page 1 Indices of baitfish Harengula jaguana digestion by gag over post-prandial time

(hr). ...........................................................................................................................30

2 Indices of crab Portunus gibbesii digestion by gag over post-prandial time (hr). ...31

3 Regression parameters of the gastric evacuation wet weight data of gag consuming baitfish prey fit to each model, for small gag (n = 13), medium gag (n = 16), and large gag (n = 11)................................................................................55

4 Regression parameters of the gastric evacuation dry weight data of gag consuming baitfish prey fit to each model, for small gag (n = 13), medium gag (n = 16), and large gag (n = 11)................................................................................56

5 Regression parameters of the pooled gastric evacuation data (n=40) of gag consuming baitfish prey on a wet and dry weight basis fit to the expanded power exponential models with either gag weight or TL scaling exponents ......................57

6 Regression parameters of the gastric evacuation wet weight data of gag consuming crab prey fit to each model, for small gag (n=8), medium gag (n=10), and large gag (n=8) ..................................................................................................58

7 Regression parameters of the gastric evacuation dry weight data of gag consuming crab prey fit to each model, for small gag (n=8), medium gag (n=10), and large gag (n=8) ..................................................................................................59

8 Regression parameters of the pooled gastric evacuation data (n=26) of gag consuming crab prey on a wet and dry weight basis fit to the expanded power exponential models with either gag weight or TL scaling exponents ......................60

9 Composition of representative baitfish Harengula jaguana and crab Portunus gibbesii prey types used in gastric evacuation trials of gag, values are means (±S.E.). .....................................................................................................................61

10 Regression parameters for models describing the gross energy (kcal/g dry weight) of the stomach contents as a function of post-prandial time (PPT) by all gag consuming baitfish prey fit to the linear, exponential, and square root models, for small gag (n=12), medium gag (n=15), and large gag (n=11) ..............62

vii

Page 8: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

11 Regression parameters modeling the gross energy (kcal/g dry weight) present over PPT by all gag consuming crab prey fit to the linear, exponential, and square root models, small gag (n=7), medium gag (n=7), and large gag (n=5).......63

12 Regression parameters modeling the percent of gross energy digested over PPT by all gag consuming baitfish prey fit to each model, for small gag (n=11), medium gag (n=16), and large gag (n=8).................................................................64

13 Regression parameters modeling the percent of gross energy digested over PPT by all gag consuming crab prey fit to each model, small gag (n=7), medium gag (n=8), and large gag (n=7)........................................................................................65

14 Mean (±S.E.) and range of post-prandial times (PPT) in relation to digestion codes and % digestion for gag consuming baitfish Harengula jaguana versus crab prey Portunus gibbesi.......................................................................................66

15 Regression parameters of the average digestion code data of gag consuming baitfish prey fit to each model, for small gag (n=13), medium gag (n=16), and large gag (n=11) .......................................................................................................67

16 Regression parameters of the digestion code data of gag consuming crab prey fit to each model, for small gag (n=8), medium gag (n=11), and large gag (n=8) .......68

viii

Page 9: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

LIST OF FIGURES

Figure page 1 Commonly used gastric evacuation models depicting the digestion processes of

different fish species.................................................................................................17

2 The relationship of gag weight (W) as a function of total length (TL) for gag between 300 and 750 mm TL...................................................................................44

3 The power exponential model describing the gastric evacuation processes of small, medium, and large gag consuming baitfish prey (scaled sardines): (a) wet weight basis and (b) dry weight basis. .....................................................................45

4 The power exponential model expanded to include weight (W) or TL as scalers describing the combined gastric evacuation data of all gag consuming baitfish prey (scaled sardines): (a) wet weight basis and (b) dry weight basis. ...................46

5 The power exponential model describing the gastric evacuation processes of small, medium, and large gag consuming crab prey (Portunus gibbesii): (a) wet weight basis and (b) dry weight basis. .....................................................................47

6 The power exponential model expanded to include weight (W) or TL as scalers describing the combined gastric evacuation data of all gag consuming crab prey (Portunus gibbesii): (a) wet weight basis and (b) dry weight basis. .......................48

7 The expanded power exponential model describing the combined gastric evacuation wet weight data of all gag consuming both baitfish (scaled sardines) and crab (Portunus gibbesii) prey incorporated with: (a) weight (W) and (b) total length (TL) scalers. ..........................................................................................49

8 Models describing the gross energy of recovered stomach contents from small, medium, and large gag consuming: (a) baitfish prey (scaled sardines), fit to a square-root model and (b) crab prey (Portunus gibbesii), fit to a linear model.......50

9 The power exponential model describing the percentage of stomach content energy digested over elapsed time for small, medium, and large gag consuming: (a) baitfish prey (scaled sardines) and (b) crab prey (Portunus gibbesii). ...............51

ix

Page 10: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

10 The power exponential model: (a) expanded to include weight (W) or TL as scalers describing the combined stomach content energy digestion data for all gag consuming baitfish prey (scaled sardines) and (b) describing the combined stomach content energy digestion data for all gag consuming crab prey (Portunus gibbesii). ..................................................................................................52

11 The power exponential model describing the average digestion code values of gag consuming baitfish prey (scaled sardines) over elapsed time: (a) small, medium, and large gag and (b) all gag fit to the expanded model using weight (W) or TL as scalers. ................................................................................................53

12 The power exponential model describing the average digestion code values of gag consuming crab prey (Portunus gibbesii) over elapsed time: (a) small, medium, and large gag and (b) all gag fit to the expanded model using weight (W) or TL as scalers. ................................................................................................54

x

Page 11: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

Abstract of Thesis Presented to the Graduate School

of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science

GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca microlepis CONSUMING FISH AND CRUSTACEAN PREY

By

Elizabeth Joanne Berens

May 2005

Chair: Debra J. Murie Major Department: Fisheries and Aquatic Sciences

Gag Mycteroperca microlepis comprise one of the most valuable fisheries in the

Gulf of Mexico, especially off the west coast of Florida. Production of juvenile and pre-

reproductive female gag, as measured through growth, depends on the total amount of

surplus energy available to the fish for growth after losses due to metabolism and wastes.

Estimating consumption rates of gag, which feed primarily on fish prey and secondarily

on crustacean prey, requires prey-specific evacuation models.

To develop these models, the proportion of the gag meals, consisting of either

baitfish (scaled sardine Harengula jaguana) or crab (purple swimmer crab Portunus

gibbesii), that remained after a pre-determined post-prandial time (PPT) was fit to linear,

exponential, square root, logistic, and power exponential models, on a wet weight and dry

weight basis. The power exponential models were significant (p≤0.0008) and best fit the

wet weight and dry weight gastric evacuation data, the percentage of prey energy

digested over PPT, and the average digestion codes over PPT, regardless of prey type or

xi

Page 12: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

gag size (R2≥0.81). Gag weight (W) or total length (TL) exponential scalers were

incorporated into the power exponential model and fit to the wet weight, dry weight, and

average digestion code data across all gag sizes by prey type, resulting in significant

(p<0.0001 for each model) and highly predictive (R2≥0.87) gastric evacuation models.

The expanded power exponential models with W and TL exponential scalers fit to the

wet weight baitfish and crab gastric evacuation data differed significantly (Maximum

Likelihood Ratio: n=66, Χ2=88.26, df=3, p<0.0001; n=66, Χ2=88.40, df=3, p<0.0001,

respectively), with lag phases of up to 5.0 hrs PPT in crab digestion only. The power

exponential model, with W or TL scalers, proved significant (p<0.0001) and highly

predictive (R2=0.98) for gag consuming baitfish when fit to the percentage of baitfish

energy digested over PPT; however, power exponential models fit to the crab data for

each size of gag were coincident. Therefore, these data were pooled and fit to an

unexpanded power exponential model, which also yielded a significant (p<0.0001) and

predictive model (R2=0.91).

After correcting for energy from chitin, which is unavailable to the gag, total crab

energy density (2.22 kcal/g dry wt) was significantly lower than that of the baitfish prey

(2-tailed Satterthwaite t-test for unequal variance; n=48, t=-31.05, p <0.0001). In

general, digestion of crab prey was associated with a 5-6 hr lag period, low prey energy

densities, and digestion over a longer period of time relative to fish prey. Therefore, for

gag consuming a mixture of fish and crab prey it will be necessary to develop a

multiplicative or additive consumption model that also incorporates mixed-prey gastric

evacuation models.

xii

Page 13: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

CHAPTER 1 INTRODUCTION

Groupers (Pisces: Serranidae) support major commercial and recreational fisheries

throughout the world. Most species of groupers are demersal predators that generally

occupy a top niche in the food chain of tropical and subtropical marine waters and have

been recognized as slow-growing, late maturing reef fishes that consume a broad range of

prey, including fish, crustaceans, and cephalopods (Polovina & Ralston, 1987).

In recent years, grouper stocks have been overfished, in part due to their slow

growth and late maturity. Reef fish, such as groupers, tend to aggregate over specific

habitats characterized by a patchy distribution, thereby making the species vulnerable to

overfishing, and skewing traditional methods used to estimate their abundance (Federal

Register, 1998, Vol. 63 No. 208, p. 57660). One important grouper species in the Gulf of

Mexico, the gag Mycteroperca microlepis supports valuable recreational and commercial

fisheries. Gag range from New York to Brazil and through the Gulf of Mexico, but are

absent from most Caribbean waters (Smith, 1971). Commercial and recreational catches

in the southeastern U.S. have exceeded 2,268 metric tons (5 million lbs) annually, and in

2001, over 3,538 metric tons (7.8 million lbs) of gag were landed in the state of Florida

alone (Turner et al., 2001; Florida Fish and Wildlife Conservation Commission

[FFWCC], 2003).

Within the past decade, many studies have examined the life history of gag and the

effect of fishing on its populations (Hood & Schlieder, 1992; Ross & Moser, 1995;

Collins et al., 1998). Such studies are important for understanding the growth and

1

Page 14: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

2

production of the gag fishery, and for helping managers cope with a variety of issues,

such as setting catch limits, size restrictions, and analyzing exploitation issues. One of

these topics, growth, is an integral part of a fish’s energy budget. An energy budget

relates all the energy an organism acquires through ingestion to its energy used in

metabolic processes, lost as wastes through excretion and egestion, or synthesized into

new somatic or reproductive tissue (Adams & Breck, 1990). Energy budgets can be used

to evaluate the importance of many different factors controlling individual growth,

including diet and activity demands, subject to different environmental conditions

(Adams and Breck, 1990; Jobling, 1993). A generalized energy budget can be modeled

as (Winberg, 1956; Warren & Davis, 1967)

( ) ( ) ( )rsar GGUFSDAMMC ++++++= (1)

where C = rate of energy consumption, Mr = standard metabolic rate, Ma = metabolic rate

increase above the standard rate due to activity, SDA = metabolic rate increase due to

specific dynamic action, F + U = waste losses due to egestion (feces) and excretion

(urine) rates, Gs = somatic growth rate due to protein synthesis [and lipid deposition], and

Gr = growth rate due to gonad (reproductive) synthesis. This generalized energy budget

is a balanced equation where all energy consumed by the animal (C), equals the energy

lost to metabolism (Mr + Ma + SDA), wastes (F + U), and growth (Gs + Gr). Hence,

energy for growth is only available after all metabolic and waste demands have been

subtracted from the total amount of energy consumed:

( ) ( )[ ]UFSDAMMCGG arrs ++++−=+ (2)

Due to the equation being theoretically balanced, researchers have the advantage of

estimating one component of the energy budget by subtracting it from the other measured

Page 15: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

3

components, although this method tends to pool all error into that one component (Adams

& Breck, 1990). Total metabolism, waste, and growth all depend on the amount of

energy consumed, and hence accurate estimates of consumption rates are imperative.

Consumption rates, however, are often the least developed part of the energy budget,

even though they are one of the most direct “inter-links” between the trophic components

in an ecosystem (Klekowski & Duncan, 1975). Consumption rates are typically difficult

to estimate because of many complex factors that influence the amount of food that a fish

consumes, including: the fish’s size; the size, abundance and distribution of prey; the

types of prey consumed; the predator’s feeding history; and even physical parameters,

such as water temperature (Windell, 1978). In addition, these factors can affect the

number of meals consumed per day, the amount of food consumed in a single meal, and

the predator’s digestion rate, or the rate of gastric evacuation, which quantifies the rate at

which food passes out of the stomach (Adams & Breck, 1990; Bromley, 1994).

Consumption rates can partially affect gastric evacuation rates by causing an increase or

decrease in the amount of time food remains in the gut. For example, if a fish consumes

a large single meal of a certain prey type, that predator’s evacuation rate will tend to be

lower than if that fish had consumed a small single meal of the same food (Swenson &

Smith, 1973). Some of these complex factors are best studied in the field, others in the

lab, but some require the capture of large numbers of fish, which may or may not be

feasible depending on species. Considering the many different biotic and abiotic factors

that influence consumption and the fact that consumption rates are difficult to quantify

outside of the laboratory, most studies take into account only two or three of these

interrelated factors (Windell, 1978; Adams & Breck, 1990).

Page 16: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

4

Most consumption models require the calculation of gastric evacuation rates and,

additionally, some models require estimates of the original prey weight at time of

capture. For example, the Continuous Feeding Model involves sampling a population

over a certain amount of time, calculating the amount of food in the gut at the beginning

and end of the sampling period, and then incorporating the length of the sampling period

and the instantaneous rate of gastric evacuation into the model (Adams & Breck, 1990).

Elliott and Persson (1978) developed a commonly used Continuous Feeding Model for

modeling consumption rate in brown trout Salmo trutta, which directly incorporates the

rate of gastric evacuation and can be modeled as

∑= tCC (3)

where C = daily ration (% body weight/day), or the sum all Ct values in each time block,

and Ct = the amount of food ingested in a block of time, or the time between two

sampling periods:

kt

ktot

ekteSS

tC −

−=1

)( (4)

where So = amount of food present at the beginning of the sampling interval, St = amount

of food present at the end of the sampling interval, t = length of the sampling interval,

and k = instantaneous rate of gastric evacuation, calculated as =k loge ( ) 1−tt

oS

S .

Although consumption affects rates of gastric evacuation, gastric evacuation rates

can also affect the amount of food a fish consumes (Brett & Higgs, 1970; Elliott &

Persson, 1978; Grove et al., 1978). Grove et al. (1978) found that an increase in the rate

of gastric evacuation occurred with the consumption of low-energy food and was

correlated to a rapid return of appetite and a high frequency of consumption, or more

Page 17: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

5

meals consumed. Clearly, gastric evacuation rates impact consumption and therefore

they must be included in consumption models in order to calculate accurate estimates of

total energy intake.

The terms digestion rate and gastric evacuation rate are often used interchangeably

to denote the rate at which food passes from the stomach into the intestine even though

this is something of a misnomer (Windell, 1978). Digestion is the act of mechanical and

enzymatic breakdown in the fish’s stomach that converts food into soluble and diffusible

products capable of being absorbed, or assimilated, by cells in the fish’s stomach and

intestine (Knutsen & Salvanes, 1999). Materials that can’t be absorbed are simply passed

through the body, and upon defecation are often referred to as indigestible matter.

Hence, food that has been broken down and is no longer present in the stomach (i.e., food

that has been evacuated from the stomach) does not necessarily represent food that has

been assimilated completely in the pyloric cecae (possibly) or intestine and utilized by

the fish (Pandian, 1967). In fisheries, research using consumption models has depended

more heavily on rates of gastric evacuation than research in other fields, such as animal

nutrition, which often focus more on the rate of transit through the entire gastrointestinal

track (Dorcas et al., 1997; Roxburgh & Pinshow, 2002; Sponheimer et al., 2003; Butler

et al., 2004; Henriques et al., 2004). In general, the alimentary tracts of fish are much

simplier than the alimentary tracts of other animals, such as mammals (Stevens & Hume,

1995) and food consumption estimates are usually determined through direct gut content

analysis using lavage techniques or by sacrificing large numbers of animals (Adams &

Breck, 1990). Conversely, estimating food consumption in many animals requires

indirect methods, such as inferring feeding rates of wild animals from feeding rates of

Page 18: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

6

captive animals (Innes et al., 1987) or through bioenergetic modeling, because lavaging

certain types of animals or sacrificing large numbers of those animals may be logistically

difficult, prohibited, or simply undesirable (Winship et al., 2002).

There are several ways to estimate gastric evacuation rates, including using X-ray

and radioisotopic procedures and through serial slaughter (Beamish, 1972; Swenson &

Smith, 1973; Diana, 1979; Flowerdew & Grove, 1979). X-ray and radioisotopic

procedures have been used most often to monitor the movement of hard parts through the

entire digestion process (Jobling et al., 1977; Flowerdew & Grove, 1979). Serial

slaughter methods in situ are generally not feasible since it is very difficult to get large

groups of wild fish to simultaneously consume a measured amount of food, hold them in

pens, and at predetermined times collect their stomach contents (Windell, 1978).

Therefore, this method is usually used on captive fish where specific amounts of food can

be given to individual fish, the fish can be monitored, and their stomach contents easily

recovered by killing the fish or by pumping its stomach (Adams & Breck, 1990). The

fraction of original prey weight digested is plotted against hours post-feeding. The slope

of the resulting regression is then used to calculate the gastric evacuation rate and

determine the gastric evacuation model (Adams & Breck, 1990).

Linear, exponential, and square root gastric evacuation models have commonly

been used to quantify the gastric evacuation processes of fish species, and then are either

input directly into, or used to meet the assumptions of a consumption model (Figure 1).

A linear model often describes the gastric evacuation processes of top carnivores, or

piscivores, which tend to consume only a few fairly large prey items over a feeding cycle,

thereby causing relatively long digestion times compared to the length of their feeding

Page 19: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

7

period (Adams & Breck, 1990). Jobling (1987) attributes the linear model to large food

particles with lower surface-to-volume ratios, low fragmentation rates, and high dietary

energy densities. These types of prey items tend to be evacuated from the stomach at a

constant rate. Previous work with piscivores and linear digestion processes have

included studies on black and yellow rockfish Sebastes chrysomelas (Hopkins & Larson,

1990), plaice Pleuronectes platessa (Jobling, 1980b), and walleye Stizostedion vitreum

vitreum (Swenson & Smith, 1973).

The exponential model generally illustrates the digestion process of herbivores,

detritivores, planktivores, and omnivores that feed at lower trophic levels on diets

composed of many small food items, such as zooplankton, and eat more or less

continuously throughout the day (Adams & Breck, 1990). These food items would be

expected to have high surface-to-volume ratios, high fragmentation rates, and tend to be

low in energy densities (Jobling, 1987). In these fish, the digestion rate increases

exponentially with time until some point at which refractory materials slow the rate and

basically level off the digestion process. Problems with the exponential model occur

when the lower portion of the curve levels off and reaches its lower asymptote because it

leads to overestimates of the amount of food remaining in the stomach at later stages of

evacuation (Brodeur & Pearcy, 1987; Adams & Breck, 1990). This slowing of gastric

evacuation means that the fish’s motivation to feed would return more slowly, thereby

increasing the amount of time between feedings, and lowering the animal’s total food

consumption rate (Rindorf, 2002). Due to the fact that the exponential model levels off,

Brodeur and Pearcy (1987) considered the active part of the curve to be between 0 and

90% evacuation.

Page 20: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

8

The square root model has been used to describe digestion in predators such as

plaice, European perch Perca fluviatili, Atlantic cod Gadus morhua, and bluegill sunfish

Lepomis macrochirus (Jobling & Davies, 1979; Jobling, 1981). This volume-based

model assumes that the instantaneous rate of evacuation is dependent upon the amount of

food in the stomach; therefore, evacuation patterns of small meals correspond to the later

stages of large meal evacuation patterns and, theoretically, results in regression lines for

different meal sizes having the same slopes (Jobling, 1981).

Brodeur (1984) suggested that an alternate way to choose the most appropriate

evacuation model may be whether or not it is interpretable in terms of the inherent

biological processes that occur during digestion. The logistic model fits this idea

considering that it accounts for the lag phases often seen in the early stages of digestion

of many different fish species (Brodeur, 1984; Hopkins & Larson, 1990) (Figure 1). The

power exponential model has also been used to describe predators with and without lag

phases because it allows the shape of the evacuation curve to vary from sigmoidal to

concave, although it can not account for linear rates of digestion (Elashoff et al., 1982;

Hopkins & Larson, 1990; dos Santos & Jobling, 1992; Temming & Andersen, 1994)

(Figure 1). Recently, Temming and Andersen (1994) have developed a general gastric

evacuation model that integrates time after ingestion, the weight of the predator,

temperature, and meal size as predicting variables (R parameter) to determine values of a

shape parameter (B) as

BRSdtdS −=/ (5)

where S = residual stomach contents (g), t = elapsed time after ingestion, R = estimated

parameter(s), and B = shape parameter. This shape parameter describes the degree of

Page 21: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

9

curvilinearity, whether it is convex, linear, exponential, or any curve in between

(Temming & Herrmann, 2001). If B=1, the evacuation model is exponential, but B=0

indicates a linear evacuation process (Temming & Andersen, 1994). For example,

Temming and Herrmann (2001) determined that B=0.7 on a dry weight basis for horse

mackerel Trachurus trachurus thereby indicating that the model was more exponential

than linear. This general model of gastric evacuation is only valid within a limited

temperature range because it assumes that the evacuation constant (R) increases

exponentially with temperature, when in fact studies have shown that at high

temperatures the evacuation rate will decrease (Tyler, 1970; Temming & Andersen,

1994).

As post-prandial time (PPT) increases, an increasing proportion of the fish with

faster digestion are typically excluded from the sample distribution. Empty stomachs are

normally dropped from the distribution because it cannot be determined when 100%

digestion occurred (Olson & Mullen, 1986). In effect, the distribution includes both

faster and slower digesting fish throughout most of the distribution but only the slower

digesting fish at the later time periods, thereby resulting in biased evacuation rate

estimates with exaggerated curvilinearity (Olson & Mullen, 1986). In part, excluding all

of the faster digesting fish from the gastric evacuation distribution may have led many

studies to choose exponential gastric evacuation models rather than square root or linear

models (Olson & Mullen, 1986). Considering that exact times of 100% evacuation can

not be determined from empty stomachs, gastric evacuation models often include times to

90% or 95% digestion only, which truncates the sample distribution and reduces bias

Page 22: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

10

from fitting models to data that may include empty stomachs (Swenson & Smith, 1973;

Hopkins & Larson, 1990).

Many environmental, prey, and predator characteristics influence both consumption

and gastric evacuation rates by either speeding or slowing food digestion. Studies

dealing with these two different rates must take many different conditions into account,

including temperature, prey type, prey sizes, predator sizes, and meal sizes (Bromley,

1994). Temperature is among the most important environmental variables that influence

consumption and digestion rates in fish. Most fish are ectotherms, and, therefore, the

surrounding water temperature determines their body temperature (Hazel, 1993).

Metabolic rates of fish and their corresponding physical and chemical processes, such as

enzyme production and kinetics, are therefore directly related to water temperatures

(Diana, 1995). Previous studies have reported that consumption rates and digestion rates

increase with rising water temperatures (Brett & Higgs, 1970; Jobling, 1980b; He &

Wurtsbaugh, 1993). Body temperatures and digestion rates can therefore vary

considerably throughout the year, if temperature fluctuates seasonally.

Meal composition, or energy density, can vary greatly with prey type and prey size,

thereby having another very important influence on consumption and digestion rates.

Many studies on fish have found that meals high in energy result in an increase in time to

100% gastric evacuation (Flowerdew & Grove, 1979; Hopkins & Larson, 1990).

Additionally, diets with an added diluent, a non-digestible marker that lowers a meal’s

energy density, were evacuated more rapidly from fish stomachs than those with higher

energy content, thereby exhibiting an increased digestion rate (Flowerdew & Grove,

1979; Jobling, 1980a). High-energy fats tend to slow gastric emptying more than

Page 23: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

11

proteins or carbohydrates (Jobling, 1980a). Therefore, mature prey fish high in lipid

content will slow digestion compared to juvenile prey fish, which may be high in protein,

but low in lipids. While, generally, lipid-rich mature prey are physically larger than

immature prey of the same species, any increase in prey size decreases the surface area to

volume ratio available for enzymatic digestion (Swenson & Smith, 1973). Therefore, not

only do mature individuals contain more lipid, which slows gastric evacuation rates, but

they also have a lower surface area to volume ratio which slows gastric evacuation rates

even more. The least digestible, hard skeletal elements of prey are generally low in

energy and are often the last part of a meal to be emptied from the stomach (Flowerdew

& Grove, 1979; Hopkins & Larson, 1990). Crustacean exoskeletons, in particular, have

been shown to remain in the stomach for long periods of time compared to food with

fewer indigestible hard parts, such as fish prey (Hopkins & Larson, 1990). Additionally,

crustacean exoskeletons contain chitin, a carbohydrate (polysaccharide) that most fish

breakdown to N-acetyl-D-glucosamine (NAG) and D-glucosamine and pass out of the

body (Jackson et al., 1992). However, fish are not known to assimilate chitin, NAG or

D-glucosamine, and therefore, chitin contains energy that is unavailable to the predator

(Battle, 1935; MacDonald et al., 1982; Lindsay & Gooday, 1985; Medved, 1985).

Gastric evacuation studies using baitfish and invertebrate prey, including crustaceans,

commonly report the means or the ranges of prey energy densities on an ash-free basis

because these values exclude all inorganic elements, or ash, which is not a source of

energy for the predator (Brett & Higgs, 1970; Beamish, 1972). Clearly, unlike baitfish

prey, unavailable energy locked up in the crustacean’s exoskeleton, as well as inorganic

ash, must be taken into account when determining the energy density available from

Page 24: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

12

crustacean prey. In a form of self-regulation, fish often appear to maintain a relatively

constant energy intake for metabolic function and growth. For example, if food resources

are lower in energy then predators will tend to eat more, but if prey are high in energy

then they eat less and maintain similar energy consumption levels (Grove et al., 1978;

Jobling, 1980a). This idea originates from Optimal Foraging Theory, which states that

animals will maximize their food (energy) intake per unit of time or minimize the time

required to meet their energy requirements (Emlen, 1966; MacArthur & Pianka, 1966;

Schoener, 1971). Optimal Foraging Theory assumes that the rate of energy intake, or

foraging success, is a proxy for fitness (Krebs & Kacelnik, 1991). As the availability of a

gag’s food resources decreases, the gag’s dietary niche breadth should expand to include

lower energy prey and increase the chance of prey encounters.

Predator size is another factor that can affect consumption and digestion models.

Absolute gastric evacuation rates (grams of food leaving the stomach per hour) tend to

increase with increasing predator body size, but relative rates (per unit body weight)

either decrease with increasing predator size or stay the same (Flowerdew & Grove,

1979; Jobling, 1980b; Bromley, 1994). For example, a 500 g grouper fed a 15 g meal

would have a slower absolute gastric evacuation rate than a 1000 g grouper fed an

identical 15 g meal because the smaller grouper would be consuming a much larger meal

relative to its body size. On the other hand, a 500 g grouper fed a 10 g meal (meal = 2%

of grouper’s body weight) would have a faster relative gastric evacuation rate than a 1000

g grouper fed a 20 g meal (meal = 2% of grouper’s body weight). The effect of predator

size on gastric evacuation rates must be considered when using standardized meals to

prevent an underestimation of consumption for small fish and an overestimation of

Page 25: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

13

consumption for large fish (Adams & Breck, 1990). Studies have determined relative

gastric evacuation rates most often (Beamish, 1972; Swenson & Smith, 1973;

MacDonald et al., 1982; Ruggerone, 1989; dos Santos & Jobling, 1992). Both predator

weight and length has been used when quantifying the influences of body size on gastric

evacuation rates; length because it is easily measured and predator weight because it can

vary much more than length, depending on season and individual growth (Koed, 2001).

Many studies have shown positive correlations between gastric evacuation rates

and meal size, although some have shown negative correlations or no correlations at all

(Bromley, 1994). Differing definitions of gastric evacuation rates, including absolute

rates and rates relative to body size, expressed as

Absolute Rate = timeFoodofWeight

(6)

Relative to Body Size = 1−× timeWeightBodyFishFoodofWeight (7)

have led to these conflicting conclusions and have made comparisons of different

evacuation results problematic (Bromley, 1994). In general, it has been shown that an

increase in meal size leads to an increase in the rate of gastric emptying and an increase

in time to 100% evacuation (Jobling et al., 1977).

Like gastric evacuation estimates, calculating the original time of prey ingestion by

a wild predator and incorporating this required variable into a consumption model has

been difficult. One method involves calculating the original prey weights based on

vertebral column length, standard length, and maximum length regressions, and back-

calculating the time of prey ingestion based on its stage of digestion (Minton et al., 1981;

Adams et al., 1982; Lindberg et al., 2002):

Page 26: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

14

Stage of Digestion = 1 - Digested Weight of Prey (8) Original Weight of Prey

Another method involves creating a quantitative and qualitative visual index of digestion

stages using numerical codes and prey descriptions (MacDonald et al., 1982; Kao, 2000;

Lindberg et al., 2002). An index of digestion may be a faster method due to the fact that

a general time of prey ingestion can be estimated based on the prey’s appearance without

having to back-calculate its size through regression analysis, but it may not be as accurate

a method since it can only provide a general, rather than specific, time after feeding.

Both methods provide inaccurate estimates of ingestion times at later periods of digestion

due to the fact that the prey eventually become unrecognizable to species and tend to

have broken or missing vertebral columns.

Currently, gastric evacuation rates have only been estimated from field collections

of gag consuming baitfish from artificial reefs off the west coast of Florida during the

warmer months of the year (Lindberg et al., 2002). Through preliminary field estimates

based on back-calculated original prey weights and an assumed linear model, time to

90% evacuation has been estimated at 15 hours and 100% evacuation at 16 hours for

baitfish prey (Lindberg et al., 2002). Several factors affecting this assumption of linear

digestion include the type of prey gag consume and the amount of energy within that

prey. Gag are commonly considered to be highly piscivorous, with invertebrates

accounting for no more than 5% of the total food volume in their stomachs (Naughton &

Saloman, 1995). Weaver (1996), however, noted a 17.1% index of relative importance

for crabs in the stomachs of gag between 300 and 400 mm standard length (374-489 mm

TL). In addition, gag off the west coast of Florida were observed to have a significant

proportion of crabs in their diet (13-14%) during the warmer months of the year

Page 27: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

15

(Lindberg et al., 2002). These proportions contribute between 7% and 24% of the total

diet on a gross energy basis, with whole crab caloric densities estimated as 1.000 kcal/g

dry weight (Lindberg et al., 2002). The hard chitinous material found in crab

exoskeletons should have a strong influence on the gag’s digestion rate. Hard materials,

such as crab exoskeleton, have been shown to cause significant lag times in digestion,

followed by a rapid increase in evacuation rate and a subsequent leveling off as meal

remnants are retained in the stomach (Hopkins & Larson, 1990). As discussed earlier,

changes in digestion rates can lead to a slower return of appetite, thereby lowering the

frequency with which gag would consume meals. Generally, crab prey are less calorie

dense than baitfish prey because they commonly contain less lipid but also because the

chitin in their exoskeletons contain energy that is unavailable to most fish (Battle, 1935;

MacDonald et al., 1982; Lindsay & Gooday, 1985; Medved, 1985). To date, there have

been no studies that have attempted to correct for unavailable energy contained within the

chitin of crab prey exoskeletons. Considering that crabs contain unabsorbable chitin in

their exoskeletons, more ash, and less energy than baitfish prey, gag consuming higher

percentages of crab in their diet should consume prey less frequently and may have less

energy available for growth after their metabolic and waste removal energy needs have

been met. To maintain growth rates similar to gag consuming baitfish, gag consuming

crab prey must compensate for their reduction in rates of gastric evacuation and feeding

frequency by consuming more crabs per meal.

Interestingly, Lindberg et al. (2002) have shown that gag on artificial reefs

(Suwannee Regional Reef System) off the west coast of Florida are consuming baitfish

prey (round scad Decapterus punctatus, juvenile tomtate Haemulon aurolineatum, scaled

Page 28: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

16

sardine Harengula jaguana, Spanish sardine Sardinella aurita) that are abundant but not

necessarily high in energy (1.03-1.14 kcal/g wet weight, depending on species) compared

to relatively lipid-rich adult baitfish or other fish species that often are 2-3 times more

calorie dense. The organic composition of individual prey (in terms of energy content)

and the varying amounts of different prey items in the diet of wild gag during the warmer

months of the year may therefore influence digestion rates significantly. Lindberg et al.

(2002) suggested that estimates of consumption rates of wild gag may be improved with

additional knowledge of prey-specific evacuation models, especially for portunid crab

prey. However, evacuation models specific to prey type and fish size are currently

unavailable for gag.

The overall goal of this study was to develop gastric evacuation models for gag

consuming baitfish and crustacean prey. The specific objectives were: (1) experimentally

determine gastric evacuation rates for gag as a function of prey type, either baitfish or

crab prey, in relation to gag size; (2) compare the effect of prey type (baitfish versus

crab) on the gastric evacuation rates of gag; (3) create qualitative and quantitative indices

of prey digestive states in order to estimate consumption times of prey sampled from the

stomach contents of wild gag; and (4) model evacuation rates of gag and compare the

models for best fit by prey type and gag size.

Page 29: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

17

0

20

40

60

80

100

0 2 4 6 8 10 12 14 16 18

Elapsed Time Since Feeding

% R

emai

ning

in th

e St

omac

h

Linear Model BtAY −= Where,

Y = % prey remaining in the stomach

A = Y-intercept B = gastric evacuation rate t = elapsed time after ingestion

BtAeY −= Where,

Y = % prey remaining in the stomach

A = Y-intercept B = gastric evacuation rate t = elapsed time after ingestion

0

20

40

60

80

100

0 2 4 6 8 10 12 14 16 18

Elapsed Time Since Feeding

% R

emai

ning

in th

e St

omac

h

Exponential Model

BtAY −= Where,

Y = % prey remaining in the stomach

A = Y-intercept B = gastric evacuation rate t = elapsed time after ingestion 0

20

40

60

80

100

0 2 4 6 8 10 12 14 16 18

Elapsed Time Since Feeding

% R

emai

ning

in th

e St

omac

h

Square Root Model

Figure 1. Commonly used models depicting the gastric evacuation processes of different fish species.

Page 30: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

18

( )[ ]CtBeAY ++

−=1

100 Where,

Y = % prey remaining in the stomach

A = estimated parameter B = scale parameter C = x-ordinate of the point of inflection of the curve

t = elapsed time after ingestion

0

20

40

60

80

100

0 2 4 6 8 10 12 14 16 18

Elapsed Time Since Feeding

% R

emai

ning

in th

e S

tom

ach

Logistic Model

0

20

40

60

80

100

0 2 4 6 8 10 12 14 16 18

Elapsed Time Since Feeding

% R

emai

ning

in th

e St

omac

h

Power Exponential

Model

( )BAtY −= 2

Where, Y = % prey remaining in the

stomach A = half life of decaying prey B = shape coefficient t = elapsed time after ingestion and

dashed lines show the potential variability in the shape coefficient

Figure 1. (continued)

Page 31: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

CHAPTER 2 METHODS

Gag Collections and Maintenance

Gag were collected from the Suwannee Regional Reef System, a series of artificial

reefs off the west coast of Florida (28°59.06 N, 83°19.10 W to 29°20.91 N, 83°31.71 W),

at a depth of 13 m (40 feet) (Lindberg et al., 2002) between February 2003 and January

2004. Small (300-449 mm total length, TL), medium (450-599 mm TL), and large (600-

750 mm TL) gag were trapped by SCUBA divers or caught underwater on hook-and-line

using baited hooks. Grouper traps measured approximately 1.0 m by 0.9 m, were

constructed out of plastic-coated wire, and featured an opening approximately 0.46 m

high by 0.30 m wide. The gag were lifted to the surface in the traps, their air bladders

were vented (Lindberg et al., 2002), and they were transported 101 km to the University

of Florida’s (UF) Fisheries and Aquatic Science’s aquatic facility using aerated coolers

with 100% diffused oxygen supplied through carbon stones. While in transport to UF,

water changes, and ChlorAm-X (an ammonia, chloramines, and chlorine neutralizer)

were used as necessary to ease stress and bring the gag back in the best possible

condition.

Gag were held in 378 L (87-cm diameter) to 473 L (147-cm diameter) fiberglass

tanks in a recirculating saltwater system of approximately 4,163 L. Filtration equipment

included a 187 L sand filter, a 189.3 L bead filter, and a mesh bag filter covering the

tanks’ drainage pipe within the lower sump tank. Light regimes averaged 13.5 hrs of

light and 10.5 hrs of darkness per day, following the photoperiod of the eastern Gulf of

19

Page 32: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

20

Mexico during the warmer months of the year (~May to November). Small transitional

lights mimicked dusk and dawn periods, turning on or off 1 hr before or after the main

lights. Gag were maintained individually or in small groups of 2-3 individuals at a mean

water temperature of 28.0oC (+ 1.0 oC), which corresponded to the average temperature

during the warmer months of the year at 12.2 m (40 ft) depth where the gag were

collected (Bledsoe & Phlips, 2000; Phlips & Bledsoe, 2002). Pieces of PVC cut 45-cm

long with 20-cm diameters were placed in the tanks to provide shelter for the gag. All

gag were measured for maximum TL, fork length (FL), and weighed within 2 weeks of

capture. Fish exhibiting any health problems were anesthetized with Tricaine-S

(Methanesulfonate, or MS-222), their bodies scraped and their gills and fins clipped for

microscopic analysis and possible bacterial, fungal, or parasitic identification. Ammonia

(0.0-0.3 mg/L), nitrite (0.0-0.5 mg/L), pH (8.0-8.3), and salinity (30-35 ppt) parameters

were kept within acceptable limits for marine fish by weekly tests, while nitrate levels (0-

30 mg/L) were tested monthly (Stickney & Kohler, 1990). Monthly water changes

replaced between 10-15% of the system’s total water capacity. ChlorAm-X was used as

needed, along with supplemental water changes, to neutralize any ammonia spikes. Gag

were fed on maintenance rations of thawed, whole baitfish or crab prey at 3.0% body

weight every other day in the morning or evening, based on current estimates of baitfish

prey average daily consumption during the warmer months being between 1.2 and 1.8%

body weight (Lindberg et al., 2002). Gag that did not feed within 1 week of initial

capture were anesthetized with Tricaine-S and tube-fed a slurry of ground fish to

stimulate subsequent natural feeding. Only fish feeding voluntarily on the maintenance

diet were used in feeding experiments.

Page 33: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

21

Experimental Feeding Trials for Gastric Evacuation Rates

The model pelagic baitfish species and potential gag prey, the scaled sardine, were

caught by underwater cast nets towed by SCUBA divers over reef areas occupied by gag

during the Fall of 2002 and 2003. They were immediately placed on ice for transport to

UF, vacuum-sealed with a FoodSaver 550, and frozen until use. Portunid crab prey

Portunus gibbesii, also a potential gag prey item (Lindberg et al., 2002), were caught in

standard shrimp trawls at night off Horseshoe Beach or Keaton Beach, FL (29°25.10 N,

83°15.30 W to 29°30.50 N, 83°25.60 W), in the eastern Gulf of Mexico during the spring

of 2003 and 2004, with the help of commercial bait-shrimp fishermen. Crabs were

immediately placed on ice for transport, vacuum-sealed while being covered with bubble

wrap to prevent their spines from cutting the sealing bag, and frozen.

Acclimation periods for fish in previous studies have varied from 2 to 6 weeks

(Brodeur & Pearcy, 1987; Hopkins & Larson, 1990). Gag were acclimated for a

minimum of 2 weeks or until they voluntarily fed on baitfish or crab prey. Only

completely whole scaled sardines or portunid crabs were used in the feeding trials (e.g.,

all skin and fins intact in the sardine prey and all legs attached in the crab prey). Prior to

consumption by the gag, sardine prey were thawed and then measured for TL, FL,

vertebral column length (VCL: measured from the atlas/axis to the hypural plate), and

weighed. Crab prey were measured for carapace length (CL), carapace width (CW), and

weighed. Food was withheld for 1-2 days before baitfish feeding trials and for 2 days

before crab trials to insure that all maintenance rations had been evacuated from the gag’s

stomach. Meals weighing approximately 1.5% of the grouper’s body weight on a wet

weight basis were fed to individual small, medium, and large gag. Only trials where all

of the prey were consumed were included in this study. Stomach contents were

Page 34: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

22

recovered via serial slaughter after various time intervals, ranging between 0.18 hrs (5

min) and 24 hrs. A serial slaughter method was necessary since previous lavage

techniques have only been between 65% and 77% complete, leaving whole crabs, pieces

of crab exoskeleton, large whole fish, and pieces of fish vertebrae and spines behind

(Lindberg et al., 2002). After a predetermined time interval, gag were sacrificed by

applying a brain-spinal pith (American Fisheries Society [AFS], 2004). The stomach

contents were recovered by removing the entire gastrointestinal tract from the esophagus

to the anus, opening up the stomach from the esophageal opening to the pylorus, and

gently scraping the contents out, without scraping so hard as to remove large amounts of

mucus off the stomach lining. Wet weight of the recovered stomach contents was taken

immediately by placing the contents on a damp sponge covered with a damp Kimwipe

(paper tissue) to remove excess water, transferring the stomach contents from the

Kimwipe to a preweighed weigh boat, and then weighing the contents to the nearest

0.0001 g. The % wet weight of the stomach contents remaining in the gag’s stomach was

calculated as

% Wet Weight = Wet Weight of Stomach Contents (g) x 100 (9) Remaining Wet Weight of Items Consumed (g)

Caloric Analysis of Prey and Stomach Contents

To determine the total caloric density of representative prey types, samples of

whole scaled sardines and portunid crabs were measured for TL, FL, VCL, and weighed

(g), or CL, CW, and weighed (g), respectively. Each individual prey was then chopped

up, freeze-dried to a constant weight, and % moisture content determined (methods

926.08 and 925.09, Association of Official Analytical Chemists [AOAC] 1990). Freeze

dried prey were then individually ground in a Braun coffee grinder until homogenized.

Page 35: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

23

Additionally, some crab prey had to be chopped or cut by hand using small dissecting

scissors to fully homogenize the remaining hard pieces of exoskeleton. Energy densities

in kcal/g dry weight of each individual prey type (0.70-0.99 g sample) were then

determined using an isoperibol bomb calorimeter (Parr 1261; Moline, IL). Individual

prey weighing <0.70 g dry weight were spiked using benzoic acid tablets to facilitate

burning and gross heat determinations. Standardized corrections to gross energy were

made for fuse wire burn (15.0 mm) and acid production (10.0 ml). To correct for

inorganic materials within each individual, the percentage of ash was determined by

ashing a subsample of individual freeze-dried baitfish and crabs (0.04-1.57 g dry weight

sample) in a muffle furnace for 12 hours at 450oC to determine their % ash-free dry

weight (g) (method 923.03, AOAC 1990):

% Ash-Free = (Tissue Dry Weight – Tissue Ash Dry Weight) x 100 (10) Dry Weight Dry Tissue Weight

Differences in % moisture content, mean energy densities, and % ash content between

baitfish and crab prey were determined using a 2-tailed Satterthwaite t-test for unequal

variances after a determination of homogeneity of variance was made using Levene’s

Test (α=0.05). The ash-free caloric densities of the initial whole baitfish and crab prey

were calculated in kcal/g dry weight by dividing the total available dry weight caloric

value by the ash-free dry weight:

Kcal/g Ash-Free = Total Energy kcal/g Dry Weight (11) Dry Weight Ash-Free Dry Weight

Due to the fact that portunid crabs contain chitin in their exoskeletons, energy

density determinations were used to correct for the unavailable energy contained within

the chitin component. For this correction, whole crab prey, throughout a size range, were

measured for CL and CW and weighed (g). Soft body tissue was removed by cutting the

Page 36: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

24

body and legs length-wise and placing the crab in a Pyrex beaker containing 15% KOH

solution. The KOH solution digested all protein completely through alkaline hydrolysis

while excluding chitin, a large beta-1,4-linked polysaccharide (Pandian, 1967). After

approximately 1 week, the residual chitinous exoskeleton from each crab was placed

against a light box (a lit background) to check for any remaining tissue. If any crab tissue

was remaining the crab was placed back in the 15% KOH solution for another 2-3 days or

until it was free of all tissue, after which the exoskeletons were rinsed with distilled

water. To determine the caloric density of the individual chitinous exoskeleton, the

exoskeleton was weighed for damp wet weight (as before), frozen (-80ºC), and freeze-

dried to a constant weight. The exoskeleton was then ground in a coffee grinder and

chopped or cut by hand using small dissecting scissors until homogenized. The energy

density of each exoskeleton was determined in the same manner as the whole baitfish and

crab prey. Initial whole crab wet weights (W), before the KOH treatments, were

regressed as a function of their exoskeleton’s total energy density (kcal/g dry weight) and

the resulting linear regression was used to estimate the amount of unavailable energy in

the chitin from the initial whole crab prey that had been used for % moisture and ash

determinations. The regression estimates of energy in the chitin exoskeleton of the initial

whole portunid crab were then subtracted from the crab’s total energy density, resulting

in the total available energy density in kcal/g dry weight of each initial individual whole

crab.

To determine the caloric densities of the individually recovered stomach contents,

the recovered stomach contents were freeze-dried to a constant weight. The stomach

contents were then homogenized and a subsample was analyzed for caloric density (as

Page 37: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

25

per previous samples). For energetic regression analyses, the average caloric content of

each meal type (i.e., fish or crab) (kcal/g dry weight) was multiplied by the estimated dry

weight of the meal fed to determine the original caloric energy content of each meal fed

to each gag. For stomach contents containing crab prey, the total available crab energy

density per meal in kcal/g dry weight was estimated by back-calculating and then

subtracting the average exoskeletal energy density of the entire meal (using the previous

regression of individual whole crab W plotted as a function of gross exoskeletal energy)

from that meal’s total energy density.

Gastric Evacuation Models

Linear, exponential, square-root, logistic, and power exponential evacuation

models were fit to the wet weight and dry weight gastric evacuation data (% remaining in

the stomach) separately for each prey type and gag size in order to model the percentage

of food remaining with PPT (Hopkins & Larson, 1990). In the power exponential model,

the percentage of food remaining in the stomach was divided by 100 in order to fit the

proportion of food remaining in the stomachs to the model and, therefore, all Y-intercepts

had to be multiplied by 100 to be comparable with the outputs of other models (Elashoff

et al., 1982; Hopkins & Larson, 1990). Only initial zeroes, or the first feeding trials

resulting in 0% remaining at PPT, were included in the data set for each gag size class in

order to prevent biases associated with excluding all zeros, and thereby, increasing the

proportion of fish with faster digestion that are eliminated from the distribution (Olson &

Mullen, 1986), as well as to prevent biases from including zeros past endpoints. A model

was considered adequate if it: (1) showed homoscedasticity of variances; (2) had y-

intercept values (estimates of the % prey remaining at time 0) between 95-105% prey

remaining; and (3) had lower asymptotes showing less than 5% prey remaining (Hopkins

Page 38: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

26

& Larson, 1990; Zar, 1999). For models that fit these initial selection criteria, r2 or R2

values were compared to determine which model best explained the gastric evacuation

data (Brodeur & Pearcy, 1987). All nonlinear R2 values were calculated as (Elashoff et

al., 1982; SAS v. 8.1, SAS Institute Inc., 1999)

( )TotalCSS

RSSR −= 12 (12)

where RSS = residual sums of squares and CSSTotal = corrected sums of squares total.

Differences among gastric evacuation models for small, medium, and large gag were

compared using Kimura’s Likelihood Ratio test (α=0.05) (Kimura, 1980; Haddon, 2001).

For curves that were not coincident (i.e., differed significantly),among gag size groups,

the data were pooled by prey type and modeled with either a gag weight (W) or TL

scaling exponent to create prey-specific models of gastric evacuation that account for

differences in gag size as

CWfunctionY ×= (13)

CTLfunctionY ×= (14)

where Y = % prey remaining in the stomach, function = evacuation model, and C = W or

TL scaling exponent.

Energetic Models

The linear, exponential, square-root, logistic, and power exponential models were

fit to the gross energy (kcal/g dry weight) of the stomach content data, as determined by

bomb calorimetry, for each size of gag consuming either baitfish or crab prey in order to

model the gross energy of the stomach contents as a function of PPT. The baitfish and

crab prey data sets were truncated as all recoverable stomach content energy values

remained constantly high and variation tended to increase as zero points, or empty

Page 39: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

27

stomachs, began to show up. A model was considered adequate if it showed

homogeneity of variances. Again, the highest r2 or R2 values determined which model

best explained the gross energy data (Brodeur & Pearcy, 1987) and all nonlinear R2

values were calculated using Eqn (12).

Next, each of the models fit to the gastric evacuation data (i.e., the linear,

exponential, square-root, logistic, and power exponential models) were fit to the

percentage of energy (kcal/g dry weight) digested as a function of PPT for each prey type

and gag size, based on the average energy density of each prey type. Again, only initial

zeroes were included in the data set for each gag size class in order to prevent biases

(Olson & Mullen, 1986). Like the gastric evacuation data, the percentage of energy

digested in the stomach was divided by 100 and all Y-intercepts multiplied by 100 to fit

the proportion of energy digested to the power exponential model and to allow

comparisons between Y-intercept estimates (Elashoff et al., 1982; Hopkins & Larson,

1990). Adequate models showed: (1) homoscedasticity of variances; (2) Y-intercept

values (estimates of the % energy digested at time 0) between -0.5 and 0.5% energy

digested; and (3) upper asymptotes greater than 85% energy digested (Hopkins & Larson,

1990; Zar, 1999). As per previous analyses, the model of best fit was determined to be

the model that fit all selection criteria and had the highest r2 or R2 value (Brodeur &

Pearcy, 1987), with non-linear R2 values calculated using Eqn (12). Kimura’s Likelihood

Ratio test was used to compare the energy digestion models among gag sizes (α=0.05)

(Kimura, 1980; Haddon, 2001). Again, W and TL exponential scalers were given to

models that differed with gag size to create prey-specific models of the percentage of

energy digested over time [as per Eqns (13) and (14)].

Page 40: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

28

Indices of Digestive States

Gag stomach contents recovered for evacuation rate models were immediately

analyzed for the state of prey digestion using the presence, absence, and appearance of

prey skin/carapace, eyes, muscle, and bones/exoskeleton. Numeric values, or codes,

were given to each individual prey item based on the approximate percentage of the prey

remaining (Tables 1 and 2) (modified from Lindberg et al., 2002). Digestion codes

correlated to prey descriptions, thereby creating both a quantitative (codes and

percentages of prey remaining) and qualitative (prey descriptions) index of prey digestive

states over time. For example, assigning a digestion index value of 0 to baitfish prey

would correlate to stomach contents that had complete prey items (intact eyes, skulls,

skin, and gut tracts) that were less than 5% digested (Lindberg et al., 2002). On the other

hand, an index value of 1 would indicate 5 to 10 % digestion and was correlated with

stomach contents that had prey with pieces of skin and muscle removed by digestion

(Lindberg et al., 2002). The maximum digestion index of 6 indicated well-digested prey,

or prey digested over 90%. Prey digestive states were evaluated for each prey item

recovered in the digestion rate feeding trials, then the mean digestion code per gag was

determined to create indices of scaled sardine and portunid crab digestive states after

PPT. These average code values for each size of gag consuming either baitfish or crab

prey were plotted as a function of PPT. Again, the first feeding trials in each gag size

class that resulted in 0% remaining at time were included in the model analyses but all

other 0 codes were excluded because it could not be determined exactly when 100%

evacuation occurred.

Page 41: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

29

Models of Average Digestion Codes

Linear, exponential, square-root, logistic, and power exponential models were fit to

the average digestion code data for each prey type by gag size in order to model the

average digestion code of food remaining with PPT after feeding (MacDonald et al.,

1982). As with the evacuation models, the power exponential model used the proportions

of average digestion codes and, therefore, all Y-intercepts were multiplied by 6 (a

maximum digestion code of 6 equated to prey being ≥90% digested at PPT) to facilitate

comparison with the other models (Elashoff et al., 1982; Hopkins & Larson, 1990). In

addition, only initial codes of 6 were included in the data set for each gag size class. A

model was considered adequate if it: (1) showed homoscedasticity of variances; (2) had

y-intercept values (estimates of the average digestion code at time 0) between -0.5 and

0.5; and (3) had upper asymptotes showing greater than 82.5% prey digested, or a

digestion code=5.5. For models that fit the initial selection criteria, the highest r2 or R2

value determined which model best fit the gastric evacuation data (Brodeur & Pearcy,

1987). All nonlinear R2 values were calculated using Eqn (12). As with evacuation rates,

differences among models for small, medium, and large gag gastric evacuation rates were

tested using the maximum likelihood ratio test (α=0.05) (Kimura, 1980; Haddon, 2001).

Page 42: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

30

Table 1. Indices of digestion for baitfish Harengula jaguana by gag over post-prandial time (hr).

Code Percent of Total

Fish Digested

Description

0 <5 Whole fish, complete VCL, most skin, head, skull, otoliths present, all meat, all guts, all bones present, most finrays, no chyme/digesta

1 5 -10 Mostly whole fish, complete VCL, most skin, head, skull, otoliths present, most meat but maybe bits missing, all guts and all bones present, some finrays maybe present, no chyme/digesta

2 10 - 25

Recognizable fish but maybe not complete, complete VCL, most skin but more missing than in code 1, complete or partial head, skull and otoliths present, most meat but more missing than code 1, most guts present, most bones present, most or all finrays gone, very little chyme/digesta

3 25 - 50 Mostly recognizable fish, complete VCL, some skin, partial head, complete or partial skull and otoliths present, some meat, some guts present, most bones present, no finrays, little chyme/digesta

4 50 - 75 May or may not be a recognizable fish, complete or incomplete VCL, little or no skin, no head, partial or no skull, otoliths present or absent, some meat, some guts present, bones present, no finrays, some chyme/digesta

5 75 - 90 Not a recognizable fish, incomplete VCL, bits of or no skin, no head, no skull, otoliths absent, little meat, no guts present, bones present, no finrays, more chyme/digesta than code 4

6 >90 Not a recognizable fish, incomplete VCL, no skin, no head, no skull, no otoliths, bits of or no meat, no guts, some bones present, no finrays, much chyme/digesta

Page 43: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

31

Table 2. Indices of digestion for crab Portunus gibbesii by gag over post-prandial time (hr).

Code Percent of Total

Crab Digested

Description

0 <5 Whole crab recognizable to species, complete and hard carapace, all spines, all meat, all guts, all legs, no chyme/digesta

1 5 -10 Whole crab recognizable to species, complete carapace but getting soft and folding, spines getting soft, all meat, all guts, most legs, no chyme/digesta

2 10 - 25 Partial crab, partial soft carapace, carapace usually folded in, spines soft if present, all meat, all guts, few or no legs, no chyme/digesta

3 25 - 50 Partial crab, possibly recognizable to species, partial soft carapace, carapace usually folded in, spines soft if present, most meat, most guts, few or no legs, no chyme/digesta

4 50 - 75 Partial crab, possibly recognizable to species, partial soft carapace, carapace folded in or top/bottom missing, no spines, some meat, some guts, no legs, little chyme/digesta

5 75 - 90 Partial crab, partial very soft carapace, carapace folded in or top/bottom missing, no spines, some meat present but exposed, some guts, no legs, more chyme/digesta than code 4

6 >90 Mostly still recognizable as a crab based on shell parts and color, partial very soft carapace, carapace anterior/posterior missing, no spines, little exposed meat present, few guts, no legs, more chyme/digesta than code 5

Page 44: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

CHAPTER 3 RESULTS

Gag and Prey Collections

A total of 66 gag were collected for gastric evacuation experiments; 21 small gag

(300-449 mm TL, 370-940 g), 26 medium gag (450-599 mm TL, 988-2295 g), and 19

large gag (600-750 mm TL, 2153-4702 g). The relationship between gag weight (W, in

g) as a function of gag maximum total length (TL, in mm) was given by (Figure 2):

W = 2E-05 x TL2.8865 R2 = 0.98 n=71 (15)

Baitfish fed to gag for use in the experimental feeding trials were on average 78.6 mm TL

(±0.31 mm S.E., range 64-93 mm, n=231) and weighed an average of 5.08 g (±0.06 g

S.E., range 2.9-8.5 g). Crabs fed to gag averaged 19.4 mm CL (±0.15 mm S.E., range

15.0-23.5 mm, n=111) and had an average mass of 5.93 g (±0.16 g S.E., range 2.77-11.68

g). Individual gag consumed meals of whole baitfish averaging 1.51% (±0.02% S.E.) of

their body weight, whereas meals fed to gag consuming meals of whole crabs averaged

1.45% (±0.02% S.E.) of their body weight. The small difference in average size of fish

and crab meals (0.06%) was attributed to gag being fed prey items that were completely

whole, and therefore individual prey items were not “pruned” in order to feed meals that

were exactly 1.5% of the fish’s body weight.

Gastric Evacuation Models for Gag Consuming Baitfish Prey

Baitfish feeding trials were completed on 40 gag, including 13 small gag, 16

medium gag, and 11 large gag. Based on the model selection criteria (homoscedasticity

of variances, y-intercepts between 95-105% prey remaining; and lower asymptotes

32

Page 45: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

33

showing less than 5% prey remaining), the power exponential models best described the

gastric evacuation processes of small, medium, and large gag consuming baitfish prey

when using recovered stomach contents on a wet weight basis (Table 3, Figure 3a). All

power exponential models fit to the wet weight gastric evacuation data for gag

consuming baitfish prey were significant (p<0.0001). Initial zero points from fish

digesting their meals faster were included in the wet weight analysis while all zero points

(n=3) following this were dropped to prevent model bias. Additionally, one baitfish point

was dropped as it was more than 2 standard deviations away from the mean. There was

no apparent lag phase in the digestion of baitfish by gag as 5-30 min trials showed

between 99.9-98.5% of the prey remaining, with 92.5% remaining 1.5 hr after feeding.

The active parts of the evacuation curves, or times to 5% remaining, on a wet weight

basis for small, medium, and large gag were calculated to be 14.7, 19.5, and 17.4 hrs

PPT, respectively. On a dry weight basis, the power exponential models also best

described the gastric evacuation processes of small, medium, and large gag consuming

baitfish prey (Table 4, Figure 3b). All power exponential models fit to the dry weight

gastric evacuation data were also significant (p<0.0001). Again, the three initial zero

points were dropped to prevent model bias. Times to 5% remaining for small, medium,

and large gag on a dry weight basis were calculated as 9.4, 16.5 and 12.5 hours PPT,

respectively. Among the three different size classes of gag consuming baitfish, the power

exponential evacuation models were not coincident on either a wet weight or dry weight

basis (Maximum Likelihood Ratio [ML]: n=40, Χ2=21.03, df=2, p<0.0001 and n=40,

Χ2=11.18, df=2, p=0.0040, respectively), and size-specific models were therefore

retained. Scaling factors for gag W or TL incorporated into the wet weight and dry

Page 46: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

34

weight power exponential models were significant (p<0.0001 for both expanded models

fit to the wet or dry weight data) and each met model selection criteria and had high R2

values (≥0.87) (Table 5, Figure 4). Although gag W and TL scaling exponents were

small (≤0.00134), the models were highly predictive (R2=0.87-0.97).

On a wet weight basis, the linear model met selection criteria for each size of gag,

was significant (p<0.0001), and explained between 93% and 95% of the variation in the

data (Table 3), as opposed to 90-96% of the wet weight variation explained by the power

exponential model. However, the linear model provided a much poorer fit (R2=0.76-

0.88) when using the gastric evacuation data on a dry weight basis. The square root and

logistic models were significant (p<0.0001) but did not meet selection criteria at every

gag size class using either the wet weight or dry weight data. Only small gag met criteria

when fit to the square root model using both the wet weight and dry weight data and was

significant (p<0.0001, R2=0.98 and 0.78, respectively). Although highly predictive and

significant (p<0.0001), the logistic models could only meet the selection criteria for small

gag when modeling the wet weight data (R2=0.99). The exponential models did not meet

selection criteria for any gag size using the wet weight or dry weight data.

Gastric Evacuation Models for Gag Consuming Crab Prey

Crab feeding trials were completed on 26 gag, including 8 small gag, 10 medium

gag, and 8 large gag. As with gag consuming baitfish, the power exponential models best

described the gastric evacuation processes of small, medium, and large gag consuming

crab prey on a wet weight basis (Table 6, Figure 5a). All power exponential models fit to

the crab wet weight gastric evacuation data were significant (p<0.0001). Again, initial

zero points from the faster digesting fish were included in the model analyses, however

one zero point following these fish was dropped to prevent model bias. The crab prey

Page 47: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

35

caused long lag phases in digestion with noticeable breakdown only starting

approximately 6 hrs after ingestion. The active parts of the curves, or times to 5% prey

remaining, for small, medium, and large gag were calculated as 21.0, 19.6, and 24.5

hours PPT, respectively. On a dry weight basis, the power exponential models were also

the most adequate models to describe the gastric evacuation processes of small, medium,

and large gag consuming crab prey, although lag phases were approximately ≤3.0 hrs

PPT (Table 7, Figure 5b). As with the crab wet weight data, all power exponential

models fit to the crab dry weight gastric evacuation data were significant (p≤0.0007).

Again, one initial zero point was dropped to prevent model bias. Times to 95% gastric

evacuation were calculated as 20.1, 17.2, and 26.2 hrs PPT for small, medium, and large

gag consuming crab prey on dry weight basis, respectively. Models of gastric evacuation

for crab prey among the three different gag size classes differed significantly for both wet

weight and dry weight crab data (ML: n=26, Χ2=7.36, df=2, p=0.025 and n=26, Χ2=8.48,

df=2, p=0.014, respectively). Because gag size had a significant effect on the gastric

evacuation rate, predator size-specific models of crab evacuation rates were retained.

Power exponential models with scaling factors for gag W or TL incorporated were

significant (p<0.0001 for both expanded models fit to the wet weight or dry weight data)

and met selection criteria using both the wet weight and dry weight data (Table 8, Figure

6). As with the baitfish gastric evacuation data, gag W and TL scaling exponents were

small (≤0.00123) but the models were highly predictive (R2≥0.94).

The logistic models also met selection criteria, were significant (p≤0.0003), and fit

the crab prey wet weight data well (R2=0.96) (Table 6). However, on a dry weight basis,

the logistic model could only meet selection criteria using the medium gag data (R2=0.96)

Page 48: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

36

(Table 7). In addition, the linear model adequately met selection criteria and was

significant (p≤0.0001) when using the medium gag dry weight data (R2=0.90), although

the model could not meet criteria when data were expressed in wet weight. The

exponential and square root models did not meet the selection criteria for any gag size

class.

Comparative Gastric Evacuation Models for Gag Consuming Baitfish versus Crab Prey

The expanded power exponential model with either W or TL exponential scalers fit

to the baitfish or crab wet weight data, pooled across all gag size classes by prey type,

differed significantly from one another (ML: n=66, Χ2=88.26, df=3, p<0.0001; n=66,

Χ2=88.40, df=3, p<0.0001, respectively) (Figure 7). Gastric evacuation of fish prey in

gag occurred earlier than crab prey on both a wet weight and dry weight basis, with no

lag period obvious with fish prey and at least a 5 hr lag period evident with crab prey.

Caloric Values of Baitfish and Crab Prey

Scaled sardine used in the prey composition analysis ranged from 67 to 111 mm TL

and had an average mass of 6.09 g (Table 9). Crabs ranged from 11.9 to 32.4 mm

maximum CL and had a mean mass of 7.03 g. At a mean of 73.47 % moisture, the

baitfish prey had a significantly greater moisture content than the mean for crab prey at

69.59% moisture (Levene: p=0.001, 2-tailed Satterthwaite t-test for unequal variance;

n=49, t=-4.77, p≤0.0001). As expected, the ash content of crabs (49.6%) was

significantly higher than the ash content of the baitfish prey (24.2%) (Levene: p<0.0001,

2-tailed Satterthwaite t-test for unequal variance; n=38, t=23.98, p<0.0001). The mean

caloric energy density of the baitfish prey was significantly greater than that of the crab

prey (Levene: p=0.009, 2-tailed Satterthwaite t-test for unequal variance; n=48, t=-31.05,

Page 49: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

37

p≤0.0001), being approximately double (4.24 kcal/g dry weight versus 2.22 kcal/g dry

weight, respectively). Baitfish mean ash-free caloric energy density (5.60 kcal/g dry wt)

was also greater than the mean ash-free caloric energy density of the crab prey (4.55

kcal/g dry wt).

A representative size range of crabs (15.1–28.0 mm CL, mean mass=6.71 g) from

both 2003 and 2004 were soaked in 15% potassium hydroxide (KOH) to dissolve away

all soft tissues and proteins (Table 9). The remaining chitinous exoskeletons had an

average mass of 2.57 g and were used to correct for energy unavailable to the gag for

assimilation. The mean caloric energy density of the remaining chitinous crab

exoskeletons was 0.76 kcal/g dry weight. When the mean energy density of the crab

exoskeletons was subtracted from the mean total energy density of the initial whole crabs,

the mean caloric energy density available to the gag for assimilation was 2.03 kcal/g dry

weight. The mean available crab energy density was significantly lower than the mean

energy density of the baitfish prey (Levene: p=0.006, 2-tailed Satterthwaite t-test for

unequal variance; n=48, t=-40.50, p≤0.0001).

Energy Values for Recovered Gag Stomach Contents

The square-root model best described the total gross energy (kcal/g dry weight) of

the recovered baitfish stomach contents from each size of gag when plotted as a function

of PPT (Table X, Figure 8a). The square-root and exponential models fit to the baitfish

gross energy data were significant, for each size of gag (p<0.0001). However, the square

root model’s R2 values were identical to the exponential model’s R2 values for each size

of gag (R2=0.01-0.29) and both had very low R2 values, indicating that the square-root

and exponential models fit to the baitfish data were descriptive rather than predictive.

The baitfish data set was truncated at 16.5 hrs PPT because stomach content energy

Page 50: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

38

densities remained constantly high (2.42-3.53 kcal/g dry weight, respectively) resulting in

the exclusion of three baitfish data points. Additionally, one baitfish data point was

dropped because it had a gross energy of zero at 16.5 hrs PPT. Y-intercept energy

densities for the stomach contents from baitfish trials were 4.41, 4.21, and 4.59 kcal/g dry

weight for small, medium, and large gag at time zero and remained fairly constant,

ending at 4.31, 4.71, and 4.95 kcal/g dry weight at 16.5 hrs PPT, respectively (Figure 8a).

The linear model met the selection criterion of homogeneity of variances but was

insignificant at each gag size (p=0.7962, 0.0627, and 0.0907 for small, medium, and large

gag, respectively). Conversely, the linear model best fit the total gross energy (kcal/g dry

weight) of the crab prey stomach contents consumed by each size of gag when plotted as

a function of PPT due to its simplicity (Table 11, Figure 8b). The linear, exponential, and

square-root models were significant (p≤0.0159) regardless of gag size and R2 values were

indistinguishable (R2=0.89-0.92, for each model, regardless of gag size). Model R2

values only differed by 0.01-0.02 when small gag were fed crab (R2=0.92, 0.90, and 0.91

for the linear, exponential, and square-root models, respectively. Crab data sets for total

gross energy recovered were truncated at 20.0 hrs PPT, because again, stomach content

energy densities remained constantly high (3.53-5.24 kcal/g dry weight), resulting in the

exclusion six crab data points. Additionally, one crab data point was dropped at 20.0 hrs

PPT because it had a gross energy of zero. Energy values for the stomach contents from

crab trials using the linear model were 2.29, 2.51, and 2.24 kcal/g dry weight for small,

medium, and large gag at PPT=0 and increased to 3.58, 3.38, and 3.39 kcal/g dry weight

at 20.0 hrs PPT, respectively (Figure 8b). The logistic and power exponential models

Page 51: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

39

could not converge on parameter estimates on either the baitfish or the crab gross energy

data, and these models were therefore excluded.

When the percentage of gross energy digested (energy passed out of stomach) was

regressed against PPT, the power exponential models best fit the evacuation processes of

each size of gag consuming either baitfish or crab prey (Tables 12 and 13, Figure 9). All

power exponential models were significant (p<0.0001) at each size of gag, regardless of

prey type. Six gag fed baitfish and three gag fed crab were excluded from the analysis

due to analytical problems. While initial points representing 100% gross energy digested

at PPT were included in the regression analyses, three gag fed baitfish and one gag fed

crab were dropped to prevent model bias. No other models met both the Y-intercept

selection criteria of ≤5% energy digested and an upper asymptote of ≥85% energy

digested for small, medium, and large gag consuming baitfish prey. Both the power

exponential and linear models met model criteria and were significant (p<0.0003) when

gag were consuming crab prey, however r2 values for the linear model were lower (Table

13). Both the baitfish and crab prey energy digested data exhibited lag phases. Baitfish

prey energy did not begin to digest until 2-3 hrs PPT, while crab prey energy did not

begin to digest until approximately 4.5 hrs PPT.

The power exponential models describing the percentage of baitfish energy density

digested over time for each gag size differed significantly (ML: n=35, Χ2=13.22, df=2,

p=0.0013). Exponential scaling factors incorporating gag W and TL were added to the

original power exponential model to retain the size-specific models of baitfish energy

digestion (Figure 10a). Both expanded models with W and TL scaling factors met the

selection criteria and were significant (p<0.0001). The W and TL scaling exponents were

Page 52: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

40

small (C=0.0180 and 0.0403, respectively) but highly predictive (R2=0.91 and 0.92,

respectively). The power exponential models describing the percentage of crab energy

digested over time for each gag size were coincident (ML: n=22, Χ2=2.23, df=2, p=0.33),

therefore, the data were pooled and fit to the original model (Figure 10b). The power

exponential model fit to the combined percentage data of crab energy digestion fit the

selection criteria, was significant (p<0.0001), and was highly predictive (R2=0.91,).

Indices of Digestion for Baitfish Consumed by Gag

All gag consumed the baitfish and crab prey whole with very minimal, if any, loss

of scales and skin of fish prey, or the cracking of carapaces and limb removal of crab

prey. Digestion was observed to be most rapid in prey located in the pylorus of the

stomach. Baitfish prey digested continuously after their initial consumption (Table 14).

At 9 hrs PPT, incomplete baitfish vertebral columns began to appear, and heads, skulls,

otoliths, guts, skin, and fin rays were often absent (code>4) . At 12 hrs PPT the baitfish

were unrecognizable with only incomplete vertebral columns and very small bits of meat

present (code=6). Between 16 and 18 hrs after consumption, the baitfish consisted only

of chyme, loose bones, and digesta (code=6). Due to the fact that the baitfish became

unrecognizable and indistinguishable from one another after being 90% digested, all

recovered stomach contents ≥90% digested were given a digestion code=6.

The power exponential model best described the relationship between the average

digestion code for each gag and PPT for small, medium, and large gag consuming

baitfish prey (R2≥0.91) (Table 15, Figure 11a). All power exponential models fit to the

average digestion code data for gag consuming baitfish prey were significant (p<0.0001).

Initial points with digestion codes of 6 were included in the analyses, however, three data

points with a code of 6 were dropped to prevent model bias. There were slight lag phases

Page 53: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

41

detected as 2.5 hr trials resulted in average digestion codes of 1.12 or less. For each size

of gag, the power exponential model’s upper asymptote equaled or exceeded an average

digestion code of 5.5, which equates to approximately 82.5% digestion. While the

logistic models met selection criteria, were significant (p<0.0001), and had high R2

values for each size of gag (all R2≥0.93), the model’s upper asymptotes did not reach an

average digestion code of 5.5 for large gag. The linear, exponential, and square root

models could not meet selection criteria for any gag size (Y-intercepts≥0.63). The

average digestion code data differed significantly between the three different gag size

classes (ML: n=40, Χ2=7.54, df=2, p=0.0231) when fit to the power exponential model.

Because predator size had a significant effect on the gag’s average digestion code values

over PPT, the gastric evacuation data for gag consuming baitfish prey were pooled and

scaling factors for gag W or TL incorporated (Figure 11b). The expanded models scaled

for W or TL met selection criteria and were significant (p<0.0001), and, although the

scaling exponents were small (C=0.0202 and 0.0355, respectively), the expanded models

were highly predictive (both R2=0.91).

Indices of Digestion for Crab Prey Consumed by Gag

Crab prey stomach content analyses showed that several legs were commonly

detached from the crab carapace located toward the pyloric end of the stomach after the

lag phase, approximately 6-8 hrs PPT. However, the exoskeleton was not yet softened

nor had it been noticeably digested at 6-8 hrs PPT. Subsequently, the carapace began to

soften and fold in on itself while the carapace spines began to erode. Digestion continued

with all appendages detaching from the carapace(s) and with further softening. The

carapace softened, folded, and eroded until holes formed and allowed digestive enzymes

access to the body cavity around 12 hrs PPT. This pattern continued until around 15 hrs

Page 54: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

42

PPT when the crab prey was recognizable only by scattered pigmented carapace piece(s)

and small bits of partial carapace, void of any meat, gut, or appendages. From 18 to 24

hrs PPT, the crabs were often unrecognizable because only tiny bits of carapace were

present in the digesta.

The power exponential model best described the average digestion code values as a

function of PPT for small, medium, and large gag consuming crab prey (R2≥0.95) (Table

16, Figure 12a). As with the baitfish data, all power exponential models fit to the average

digestion code data for small, medium, and large gag consuming crab were significant

(p≤0.0008). Again, initial points with a digestion code of 6 were included in the

regression analyses but one point with a code of 6 was dropped to prevent model bias.

The average digestion code data clearly showed a lag in crab digestion for each size of

gag as codes at 6 hrs PPT were still <1.0. For all sizes of gag, the power exponential

models’ average digestion code estimates equaled or exceeded code 5.5 at 24 hrs PPT,

which equates to approximately 82.5% of the total prey digested at time. Average

digestion codes for each size of gag consuming crab prey differed significantly (ML:

n=26, Χ2=10.5601, df=2, p=0.0051). Because gag size had a significant effect on the

average digestion code values of crab prey over PPT, size-specific models were retained

and scaling factors for either W (C=0.0188) or TL (C=0.0258) were added (Figure 12b).

The resulting expanded power exponential models met selection criteria, were significant

(p<0.0001), and were highly predictive (R2=0.95).

The square root models were significant (p<0.0001) and had high R2 values (0.86-

0.92) for each size of gag consuming crab prey, but slightly overestimated the average

digestion code at the earliest time intervals (Y-intercepts = 0.12 -0.41 hrs PPT). The

Page 55: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

43

linear model met selection criteria and was significant (p=0.0002) when fit to the large

gag data, however this model projected the Y-intercept of the small gag size class at -2.04

hrs and medium gag at -0.80 hrs PPT, both of which are well below acceptable criteria.

In addition, the logistic models were significant (p≤0.0008), fit the small, medium, and

large gag size classes well, estimating Y-intercepts at 0.01, 0.06, and 0.15, respectively,

and had R2 values at 0.94, 0.96, and 0.99, respectively. However, the logistic models

could not adequately estimate the average digestion code at time of large gag consuming

baitfish, as the upper asymptote estimate could not reach an average code of 5.5 at 24 hrs.

Again, the exponential model could not meet selection criteria for any size of gag

consuming crab prey.

Page 56: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

44

y = 2E-05x2.8865

R2 = 0.9878

0

1000

2000

3000

4000

5000

300 400 500 600 700 800

Total Length (TL)

Wei

ght (

g)

Figure 2. The relationship of gag weight (W) as a function of total length (TL) for gag between 300 and 750 mm TL.

Page 57: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

45

0 5 10 15 20 25

Post-Prandial Time (PPT) (hr)

0

20

40

60

80

100

120Pe

rcen

t Foo

d R

emai

ning

in S

tom

ach

L

L

L

L

L

(a)

L

M

M

M

M

M

M

S

SS

S

SS

S

S

S

S

S

S

S

L

L

L

L

L

MM

M

M

M

M

M

M

M

M

…. Small Gag (S) ( ) 68.116.62 tY −=

____ Medium Gag (M) ( ) 52.1

44.72 tY −=

---- Large Gag (L) ( ) 89.101.82 tY −=

0 5 10 15 20 25

Post-Prandial Time (PPT) (hr)

0

20

40

60

80

100

120

Perc

ent F

ood

Rem

aini

ng in

Sto

mac

h

(b)

LLM

M

M

M

M

M

S

SS

S

S S

S

S

S

S

S

S

S

L

L

L

L

L

L

L

L

M

M

M

M

M

M

M

M

M

M

L

…. Small Gag (S) ( ) 64.326.62 tY −=

____ Medium Gag (M) ( ) 34.1

54.52 tY −=

---- Large Gag (L) ( ) 85.247.72 tY −=

Figure 3. The power exponential model describing the gastric evacuation processes of

small, medium, and large gag consuming baitfish prey (scaled sardines) on a: (a) wet weight basis (see Table 3) and (b) dry weight basis (see Table 4).

Page 58: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

46

(a)

0 5 10 15 20 25

Time Elapsed

0

20

40

60

80

100

120

Per

cent

Foo

d R

emai

ning

in S

tom

ach

…. Weight ( ) 0013.064.1

12.72 WY t×= −

____ TL ( ) 0008.065.1

16.72 TLY t×= −

Post-Prandial Time (PPT) (hr)

(b)

0 5 10 15 20 25

Time Elapsed

0

20

40

60

80

100

120

Per

cent

Foo

d R

emai

ning

in S

tom

ach

…. Weight ( ) 0024.085.140.62 −− ×= WY t

____

TL ( ) 0044.088.146.62 −− ×= TLY t

Post-Prandial Time (PPT) (hr)

Figure 4. The power exponential model expanded to include weight (W) or TL as scalers

describing the combined gastric evacuation data of all gag consuming baitfish prey (scaled sardines) on a: (a) wet weight basis and (b) dry weight basis (see Table 5).

Page 59: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

47

0 5 10 15 20 25

Post-Prandial Time (PPT) (hr)

0

20

40

60

80

100

120P

erce

nt F

ood

Rem

aini

ng in

Sto

mac

h L

L

L

L

L

L

L

M

M

M

M

M M

MM

M

M

M

(a)

S

S

S

S

S

SS

S

…. Small Gag (S) ( ) 85.249.122 tY −=

____ Medium Gag (M) ( ) 07.3

17.122 tY −=

---- Large Gag (L) ( ) 28.291.122 tY −=

0 5 10 15 20 25

Post-Prandial Time (PPT) (hr)

0

20

40

60

80

100

120

Per

cent

Foo

d R

emai

ning

in S

tom

ach

L

(b)

L

L

LM

M

S

S

S

S

S

SS

S

L

L

L

L

M

MM M

MM

M

M

…. Small Gag (S) ( ) 80.192.82 tY −=

____ Medium Gag (M) ( ) 32.2

19.92 tY −=

---- Large Gag (L) ( ) 23.100.82 tY −=

Figure 5. The power exponential model describing the gastric evacuation processes of

small, medium, and large gag consuming crab prey (Portunus gibbesii) on a: (a) wet weight basis (see Table 6) and (b) dry weight basis (see Table 7).

Page 60: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

48

(a)

0 5 10 15 20 25

Time Elapsed

0

20

40

60

80

100

120

Per

cent

Foo

d R

emai

ning

in S

tom

ach

…. Weight ( ) 0008.065.247.122 WY t

×= − ____

TL ( ) 0012.064.246.122 TLY t

×= −

Post-Prandial Time (PPT) (hr)

(b)

0 5 10 15 20 25

Time Elapsed

0

20

40

60

80

100

120

Per

cent

Foo

d R

emai

ning

in S

tom

ach

…. Weight ( ) 0037.088.130.92 −− ×= WY t

____

TL ( ) 0037.086.126.92 −− ×= TLY t

Post-Prandial Time (PPT) (hr)

Figure 6. The power exponential model expanded to include weight (W) or TL as scalers

describing the combined gastric evacuation data of all gag consuming crab prey (Portunus gibbesii) on a: (a) wet weight basis and (b) dry weight basis (see Table 8).

Page 61: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

49

(a)

0 5 10 15 20 25

Time Elapsed

0

20

40

60

80

100

120

Per

cent

Foo

d R

emai

ning

in S

tom

ach C

F

F

F

F

F

F

F

F

F

F

F

FF

F

F

F

F

F

F

F

F

F

F

FF

F

F

F

FF

F

FF

F

F

F

F

F

F

F C

C

C

C

C

C

C

C

C

C

C C

CC

C

C

C

C

C

C

C

C

CC

C

____ Baitfish (F) ( ) 0013.064.1

12.72 WY t×= −

…. Crab (C) ( ) 0008.065.247.122 WY t

×= −

Post-Prandial Time (PPT) (hr)

(b)

0 5 10 15 20 25

Time Elapsed

0

20

40

60

80

100

120

Per

cent

Foo

d R

emai

ning

in S

tom

ach C

F

F

F

F

F

F

F

F

F

F

F

FF

F

F

F

F

F

F

F

F

F

F

FF

F

F

F

FF

F

FF

F

F

F

F

F

F

F C

C

C

C

C

C

C

C

C

C

C C

CC

C

C

C

C

C

C

C

C

CC

C

____ Baitfish (F) ( ) 0008.065.1

16.72 TLY t×= −

…. Crab (C) ( ) 0012.064.246.122 TLY t

×= −

Post-Prandial Time (PPT) (hr)

Figure 7. The expanded power exponential model describing the combined gastric

evacuation wet weight data of all gag consuming both baitfish (scaled sardines) and crab (Portunus gibbesii) prey incorporated with: (a) weight (W) and (b) total length (TL) scalers (see Tables 5 and 8).

Page 62: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

50

0 5 10 15 20 25

Post-Prandial Time (PPT) (hr)

0

1.0

2.0

3.0

4.0

5.0

6.0G

ross

Ene

rgy

(kca

l/g d

ry w

t)

L LLL L

LL

L

L LL

MMM

M

M

M

MMM

M

MMMM

M SS

S

S

SS

S

SS

S

S

S

(a)

…. Small Gag (S) tY 0015.04139.4 +=____

Medium Gag (M) tY 0073.02053.4 −=

---- Large Gag (L) tY 0051.05871.4 −=

0 5 10 15 20 25

Post-Prandial Time (PPT) (hr)

0

0.7

1.4

2.1

2.8

3.5

4.2

Gro

ss E

nerg

y (k

cal/g

dry

wt)

(b)

L

MS

S

S

S

S

SL

L

LM

M

M

ML

…. Small Gag (S) tY 0645.02935.2 += ____

Medium Gag (M) tY 0436.05073.2 +=

---- Large Gag (L) tY 0578.02388.2 +=

Figure 8. Models describing the gross energy of recovered stomach contents from small,

medium, and large gag consuming: (a) baitfish prey (scaled sardines), fit to a square-root model (see Table 10) and (b) crab prey (Portunus gibbesii), fit to a linear model (see Table 11).

Page 63: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

51

0 5 10 15 20 25

Post-Prandial Time (PPT) (hr)

0

20

40

60

80

100P

erce

nt G

ross

Ene

rgy

Dig

este

d (k

cal/g

dry

wt)

L

L

L

L

L

L

L

L

L

LM

M

M

M

M

M

M

M

M

M

M

M

MM

MM

M

M

M

S

SS

S

S

SS

S

S

S

S

(a)

…. Small Gag (S) ( ) 98.164.52

−−=tY

____ Medium Gag (M) ( ) 46.1

33.62−−=

tY

---- Large Gag (L) ( ) 89.161.72

−−=tY

0 5 10 15 20 25

Post-Prandial Time (PPT) (hr)

0

20

40

60

80

100

Per

cent

Gro

ss E

nerg

y D

iges

ted

(kca

l/g d

ry w

t)

L

(b)

L

M

M

S

S

S

S

S

S

S

L

L

L

L

M

MM M

M

M

L

…. Small Gag (S) ( ) 73.107.92

−−=tY

____ Medium Gag (M) ( ) 78.2

64.92−−=

tY

---- Large Gag (L) ( ) 98.123.102

−−=tY

Figure 9. The power exponential model describing the percentage of stomach content

energy digested over elapsed time for small, medium, and large gag consuming: (a) baitfish prey (scaled sardines) (see Table 12) and (b) crab prey (Portunus gibbesii) (see Table 13).

Page 64: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

52

(b)

____ All Gag ( ) 51.2

21.102−−=

tY

0 5 10 15 20 25

Post-Prandial Time (PPT) (hr)

0

20

40

60

80

100

Per

cent

Gro

ss E

nerg

y D

iges

ted

(kca

l/g d

ry w

t)

0 5 10 15 20 25

Post-Prandial Time (PPT) (hr)

0

20

40

60

80

100

Per

cent

Gro

ss E

nerg

y D

iges

ted

(kca

l/g d

ry w

t)

…. Weight ( ) 0180.032.139.72 WY t

×=−−

____ TL ( ) 0403.012.1

69.82 TLY t×=

−−

(a) Figure 10. The power exponential model: (a) expanded to include W or TL as scalers

describing the combined stomach content energy digestion data for all gag consuming baitfish prey (scaled sardines) and (b) describing the combined stomach content energy digestion data for all gag consuming crab prey (Portunus gibbesii).

Page 65: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

53

0 5 10 15 20 25

Post-Prandial Time (PPT) (hr)

0

1

2

3

4

5

6Av

erag

e D

iges

tion

Cod

e

(a)

L

MM

M

M

S

SS

S

S

S

S

S

S

S

S

S

S

L

L

L

L

L

L

L

L

L

L

MM

M

M

M

M

M

M

M

M

M

M

…. Small Gag (S) ( ) 66.118.42

−−=tY

____ Medium Gag (M) ( ) 84.1

56.42−−=

tY

---- Large Gag (L) ( ) 97.179.52

−−=tY

0 5 10 15 20 25

Post-Prandial Time (PPT) (hr)

0

1

2

3

4

5

6

Ave

rage

Dig

estio

n C

ode

(b)

…. Weight ( ) 0202.030.171.52 WY t

×=−−

____ TL ( ) 0355.030.1

33.62 TLY t×=

−−

Figure 11. The power exponential model describing the average digestion code values of

gag consuming baitfish prey (scaled sardines) over elapsed time: (a) small, medium, and large gag (see Table 15) and (b) all gag fit to the expanded model using W or TL as scalers.

Page 66: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

54

0 5 10 15 20 25

Post-Prandial Time (PPT) (hr)

0

1

2

3

4

5

6Av

erag

e D

iges

tion

Cod

e

(a)

L

LLM S

S

S

S

S

SS

S

L

L

L

L

L

M

M

M

M

M

M

MM

M

M

…. Small Gag (S) ( ) 80.417.112

−−=tY

____ Medium Gag (M) ( ) 47.3

29.112−−=

tY

---- Large Gag (L) ( ) 38.272.92

−−=tY

0 5 10 15 20 25

Post-Prandial Time (PPT) (hr)

0

1

2

3

4

5

6

Ave

rage

Dig

estio

n C

ode

(b)

…. Weight ( ) 0188.059.297.112 WY t

×=−−

____TL ( ) 0258.047.2

19.122 TLY t×=

−−

Figure 12. The power exponential model describing the average digestion code values of

gag consuming crab prey (Portunus gibbesii) over elapsed time: (a) small, medium, and large gag (see Table 16) and (b) all gag fit to the expanded model using W or TL as scalers.

Page 67: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

Table 3. Regression parameters of the gastric evacuation wet weight data of gag consuming baitfish prey fit to each model, for small gag (n=13), medium gag (n=16), and large gag (n=11). Models meeting selection criteria are highlighted in bold; assumptions met for heterogeneity of variance and characteristics of the lower asymptote are given by Y (yes) or N (no). All models were significant at the α=0.05 level (p<0.0001) (see Figure 3).

55

Model Gag Size A B C Y-intercept MSE R2 Homogeneity of Variance

Lower Asymptote

<5%

Linear Small 95.50 6.32 95.50 85.81 0.95 Y

Medium

99.85 6.03 99.85 52.15 0.95 Y

Large 103.19 6.18 103.19 83.44 0.93 YExponential Small 106.50 0.14 106.50 68.49 0.96 N N

Medium 114.00 0.12 114.00 70.54 0.93 N N

Large 115.70 0.11 115.70 120.70 0.90 N YSquare Root Small 103.80 0.52 103.80 29.42 0.98 Y Y

Medium 108.40 0.45 108.40 44.09 0.96 Y Y

Large 111.30 0.44 111.30 77.63 0.94 Y YLogistic Small 92.89 -0.54 -5.93 96.32 11.47 0.99 Y Y

Medium 91.23 -0.38 -7.02 93.90 69.43 0.94 Y N

Large 86.84 -0.53 -7.20 98.16 40.14 0.97 Y N

Small 6.16 1.68 100.00a 0.0012 0.99 Y YPower Exponentiala

Medium 7.44 1.52 100.00a 0.0015 0.96 Y Y

Large 8.01 1.89 100.00a 0.0047 0.96 Y Y

a Due to model specifications, the Y-intercept of the power exponential model is fixed at Y=1 (100%) and not uniquely estimated.

Page 68: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

Table 4. Regression parameters of the gastric evacuation dry weight data of gag consuming baitfish prey fit to each model, for small gag (n=13), medium gag (n=16), and large gag (n=11). Models meeting selection criteria are highlighted in bold; assumptions met for homogeneity of variance and characteristics of the lower asymptote are given by Y (yes) or N (no). All models were significant at the α=0.05 level (p<0.0001) (see Figure 3).

56

Model Gag Size A B C Y-intercept MSE R2 Homogeneity of Variance

Lower Asymptote

<5%

Small 90.21 6.11 90.21 460.60 0.76 YMedium

90.59 6.05 90.59 133.16 0.88 Y

Linear

Large 106.44 6.84 106.44 206.65 0.88 YSmall 99.38 0.14 99.38 489.00 0.74 Y NMedium 114.60 0.16 114.60 50.47 0.96 Y N

Exponential

Large 124.50 0.13 124.50 200.50 0.88 Y NSmall 97.86 0.52 97.86 425.20 0.78 Y YMedium 106.30 0.57 106.30 51.63 0.96 Y Y

Square Root

Large 119.70 0.54 119.70 141.00 0.92 Y YSmall 90.23 -24.88 -5.98 100.00 303.00 0.84 N NMedium 91.00 -0.49 -5.27 93.46 78.40 0.94 Y N

Logistic

Large 91.41 -0.66 -7.16 99.21 50.86 0.97 Y N

Small 6.26 3.64 100.00a 0.0368 0.81 Y Y

Medium 5.54 1.34 100.00a 0.0050 0.95 Y Y

Power Exponentiala

Large 7.47 2.85 100.00a 0.0060 0.96 Y Y

a Due to model specifications, the Y-intercept of the power exponential model is fixed at Y=1 (100%) and not uniquely estimated.

Page 69: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

Table 5. Regression parameters of the pooled gastric evacuation data (n=40) of gag consuming baitfish prey on a wet and dry weight

basis fit to the expanded power exponential models with either gag weight or TL scaling exponents. Assumptions met for constant variance and characteristics of the lower asymptote are given by Y (yes) or N (no). All models were significant at the α=0.05 level (p<0.0001) (see Figure 4).

57

Stomach Content Conditiona

Scaler for Gag A B C Y-intercept MSE R2 Homogeneity

of Variance Lower

Asymptote <5%

Wet Weight Weight 7.12 1.64 0.00134 100.00a 0.00537 0.96 Y Y

Wet Weight TL 7.16 1.65 0.00081 100.00a 0.00538

0.96 Y Y

Dry Weight Weight 6.40 1.85 -0.00240 100.00a 0.01900 0.87 Y Y

Dry Weight TL 6.46 1.88 -0.00444 100.00a 0.01900 0.87 Y Y

a Due to model specifications, the Y-intercept of the power exponential model is fixed at Y=1 (100%) and not uniquely estimated.

Page 70: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

Table 6. Regression parameters of the gastric evacuation wet weight data of gag consuming crab prey fit to each model, for small gag (n=8), medium gag (n=10), and large gag (n=8). Models meeting selection criteria are highlighted in bold; assumptions met for homogeneity of variance and characteristics of the lower asymptote are given by Y (yes) or N (no). All models were significant at the α=0.05 level (p<0.0001) (see Figure 5).

58

Model Gag Size A B C Y-intercept MSE R2 Homoogeneity

of Variance Lower

Asymptote <5%

Small 125.75 5.88 125.75 113.32 0.93 YMedium

112.41 5.04 112.41 114.04 0.94 Y

Linear

Large 111.17 4.81 111.17 56.94 0.96 YSmall 173.70 0.11 173.70 171.30 0.88 N Y

Medium 130.40 0.09 130.40 301.80 0.63 N N

Exponential

Large 118.80 0.08 118.80 207.50 0.87 N NSmall 149.50 -0.43 149.50 102.70 0.93 Y N

Medium 124.90 -0.38 124.90 155.20 0.93 Y Y

Square Root

Large 115.90 -0.31 115.90 124.30 0.92 Y YSmall 87.76 -0.49 -11.42 99.69 54.39 0.96 Y Y

Medium 101.30 -0.37 -12.31 98.96 14.82 0.99 Y Y

Logistic

Large 129.90 -0.19 -16.23 94.70 77.68 0.96 Y YSmall 12.49 2.85 100.00a 0.0066 0.96 Y Y

Medium 12.17 3.07 100.00a 0.0015 0.99 Y Y

Power Exponentiala

Large 12.91 2.28 100.00a 0.0050 0.96 Y Y

a Due to model specifications, the Y-intercept of the power exponential model is fixed at Y=1 (100%) and not uniquely estimated.

Page 71: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

Table 7. Regression parameters of the gastric evacuation dry weight data of gag consuming crab prey fit to each model, for small gag (n=8), medium gag (n=10), and large gag (n=8). Models meeting selection criteria are highlighted in bold; assumptions met for homogeneity of variance and characteristics of the lower asymptote are given by Y (yes) or N (no). All models were significant at the α=0.05 level (p≤0.0007) (see Figure 5).

59

Model Gag Size A B C Y-intercept MSE R2 Homogeneity of Variance

Lower Asymptote

≤5%

Small 90.47 4.46 90.47 117.22 0.87 YMedium

102.60 4.85 102.60 212.60 0.90 Y

Linear

Large 89.47 4.16 89.47 38.49 0.97 YSmall 137.10 0.13 137.10 128.60 0.86 Y NMedium 134.80 0.12 134.80 181.20 0.91 Y Y

Exponential

Large 103.10 0.10 103.10 63.90 0.95 Y NSmall 114.00 0.42 114.00 88.74 0.90 N YMedium 125.60 0.46 125.60 89.84 0.96 N Y

Square Root

Large 97.24 0.34 97.24 36.07 0.97 N YSmall 98.91 -0.29 -9.07 93.40 76.90 0.92 Y YMedium 96.98 -0.45 -8.88 98.18 84.31 0.96 Y Y

Logistic

Large 105.00 -0.19 -9.77 86.04 64.25 0.96 Y Y

Small 8.92 1.80 100.00a 0.0081 0.91 Y Y

Medium 9.19 2.32 100.00a 0.0058 0.97 Y Y

Power Exponentiala

Large 8.00 1.23 100.00a 0.0052 0.96 Y Y

a Due to model specifications, the Y-intercept of the power exponential model is fixed at Y=1 (100%) and not uniquely estimated.

Page 72: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

Table 8. Regression parameters of the pooled gastric evacuation data (n=26) of gag consuming crab prey on a wet and dry weight basis fit to the expanded power exponential models with either gag weight or TL scaling exponents. Assumptions met for constant variance and characteristics of the lower asymptote are given by Y (yes) or N (no). All models were significant at the α=0.05 level (p<0.0001) (see Figure 6).

60

Stomach Content Conditiona

Scaler A B C Y-intercept MSE R2 Homogeneity of Variance

Lower Asymptote

<5%

Wet Weight Weight 12.47 2.65 0.00084 100a 0.00528 0.97 Y Y

Wet Weight TL 12.46 2.64 0.00123 100a 0.00528

0.97 Y Y

Dry Weight Weight 9.30 1.88 -0.00370 100a 0.00752 0.94 Y Y

Dry Weight TL 9.26 1.86 -0.00371 100a 0.00755 0.94 Y Y

a Due to model specifications, the Y-intercept of the power exponential model is fixed at Y=1 (100%) and not uniquely estimated.

Page 73: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

Table 9. Composition of representative baitfish Harengula jaguana and crab Portunus gibbesii prey types used in gastric evacuation trials of gag, values are means (±S.E.).

61

Parameters Baitfish Crab ExoskeletonCrab

Number 25 24 23

TL or Carapace Length (mm) 85.04 (1.99) 19.76 (0.93) 19.96 (0.78)

TL or Carapace Length (mm) Range 67 - 111 11.9 - 32.4 15.1 - 28.0

Mass (g) 6.09 (0.36) 7.03 (1.22) 2.57 (0.36)

Mass Range (g) 3.2 - 10.7 1.0 - 29.3 1.9 - 19.0

Exoskeleton Mass Range After KOH Treatment (g) 0.7 - 7.2

% Moisture 73.47 (0.18) 69.62 (0.79)

% Ash 24.22 (0.27)) 49.61 (1.03)

Caloric Density (kcal/g dry weight) 4.24 (0.02) 2.22 (0.06) 0.76 (0.03)

Caloric Ash-Free Energy Density (kcal/g ash-free dry weight) 5.60 (0.03) 4.55 (0.07)

Caloric Energy Density Available for Assimilation (kcal/g dry weight) 4.24 (0.02) 2.03 (0.00)

Page 74: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

Table 10. Regression parameters for models describing the gross energy (kcal/g dry weight) of the stomach contents as a function of post-prandial time (PPT) by all gag consuming baitfish prey fit to the linear, exponential, and square root models, for small gag (n=12), medium gag (n=15), and large gag (n=11). All models met selection criterion; assumptions met for homogeneity of variance are given by Y (yes) or N (no). Exponential and square-root models were significant at the α=0.05 level (p<0.0001) (see Figure 8).

62

Model Gag Size A B Y-intercept MSE R2 Homogeneity of Variance

Linear Small 4.4140 -0.0062 4.4140 0.18757 0.01 Y

Medium

4.2058 0.0303 4.2058 0.06630 0.24 Y

Large 4.5871 0.0221 4.5871 0.03790 0.29 Y

Exponential Small 4.4137 0.0014 4.4137 0.18758 0.01 Y

Medium 4.2048 -0.0070 4.2048 0.06575 0.25 Y

Large 4.5871 -0.0047 4.5871 0.03769 0.29 Y

Square Root Small 4.4139 -0.0015 4.4139 0.18757 0.01 Y

Medium 4.2053 0.0073 4.2053 0.06603 0.25 Y

Large 4.5871 0.0051 4.5871 0.03780 0.29 Y

Page 75: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

Table 11. Regression parameters modeling the gross energy (kcal/g dry weight) present over PPT by all gag consuming crab prey fit to the linear, exponential, and square root models, small gag (n=7), medium gag (n=7), and large gag (n=5). All models met selection criterion; assumptions met for homogeneity of variance are given by Y (yes) or N (no). All models were significant at the α=0.05 level (p≤0.0159) (see Figure 8).

63

Model Gag Size A B Y-intercept MSE R2 Homogeneity of Variance

Linear Small 2.2935 0.0645 2.2935 0.01265 0.92 Y

Medium

2.5073 0.0436 2.5073 0.00894 0.91 Y

Large 2.2388 0.0578 2.2388 0.01602 0.89 Y

Exponential Small 2.3852 -0.0205 2.3852 0.01598 0.90 Y

Medium 2.5274 -0.0149 2.5274 0.00899 0.91 Y

Large 2.2908 -0.0202 2.2908 0.01560 0.89 Y

Square Root Small 2.3423 0.0182 2.3423 0.01428 0.91 Y

Medium 2.5176 0.0127 2.5176 0.00895 0.91 Y

Large 2.5654 0.0171 2.5654 0.01571 0.89 Y

Page 76: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

Table 12. Regression parameters modeling the percent of gross energy digested over PPT by all gag consuming baitfish prey fit to each model, for small gag (n=11), medium gag (n=16), and large gag (n=8). Models meeting selection criteria are highlighted in bold; assumptions met for homogeneity of variance and characteristics of the upper asymptote are given by Y (yes) or N (no). All models were significant at the α=0.05 level (p<0.0001) (see Figure 9).

64

Model Gag Size A B C Y-intercept MSE R2 Homogeneity of Variance

Upper Asymptote >85%

Small 12.38 5.70 12.38 97.75 0.92 Y

Medium

3.37 5.88 3.37 114.24 0.89 Y

Linear

Large -0.68 5.90 -0.68 62.73 0.93 Y

Small 28.62 -0.08 28.62 207.70 0.82 Y Y

Medium 20.07 -0.10 20.07 223.10 0.79 Y Y

Exponential

Large 22.20 -0.09 22.20 149.00 0.83 Y Y

Small 22.03 0.35 22.03 150.60 0.87 Y Y

Medium 14.02 0.39 14.02 169.30 0.84 Y Y

Square Root

Large 14.60 0.37 14.60 108.00 0.88 Y Y

Small 94.75 -0.46 -5.88 6.01 24.21 0.98 Y Y

Medium 86.05 -0.46 -6.08 4.83 104.30 0.91 Y Y

Logistic

Large 88.24 -0.42 -7.32 3.79 47.37 0.96 Y Y

Small 5.64 -1.98 0.00a 0.0057 0.95 Y Y

Medium 6.33 -1.46 0.00a 0.0094 0.91 Y Y

Power Exponentiala

Large 7.61 -1.89 0.00a 0.0033 0.96 Y Y

a Due to model specifications, the Y-intercept of the power exponential model is fixed at Y=0 (0%) and not uniquely estimated.

Page 77: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

Table 13. Regression parameters modeling the percent of gross energy digested over PPT by all gag consuming crab prey fit to each model, small gag (n=7), medium gag (n=8), and large gag (n=7). Models meeting selection criteria are highlighted in bold; assumptions met for homogeneity of variance and characteristics of the upper asymptote are given by Y (yes) or N (no). All models were significant at the α=0.05 level (p≤0.0003) (see Figure 9).

65

Model Gag Size A B C Y-intercept MSE R2 Homogeneity of Variance

Upper Asymptote >85%

Small 2.05 4.55 2.05 44.10 0.95 Y

Medium

-3.44 4.71 -3.44 125.83 0.90 Y

Linear

Large -2.53 4.57 -2.53 31.62 0.97 Y

Small 24.35 -0.07 24.35 121.30 0.87 Y Y

Medium 26.02 -0.06 26.02 257.30 0.80 Y Y

Exponential

Large 18.66 -0.08 18.66 52.19 0.95 Y Y

Small 16.13 0.28 16.13 78.86 0.91 Y Y

Medium 16.22 0.27 16.22 191.30 0.85 Y Y

Square Root

Large 11.32 0.31 11.32 37.27 0.96 Y Y

Small 97.42 -0.27 -10.32 5.80 15.21 0.99 Y Y

Medium 101.60 -0.32 -11.00 3.05 36.74 0.98 Y Y

Logistic

Large 152.30 -0.14 -17.35 13.17 53.51 0.96 Y Y

Small 9.07 -1.73 0.00a 0.0089 0.90 Y Y

Medium 9.64 -2.78 0.00a 0.0053 0.96 Y Y

Power Exponentiala

Large 10.23 -1.98 0.00a 0.0138 0.87 Y Y

a Due to model specifications, the Y-intercept of the power exponential model is fixed at Y=0 (0%) and not uniquely estimated.

Page 78: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

Table 14. Mean (±S.E.) and range of post-prandial times (PPT) in relation to digestion codes and % digestion for gag consuming baitfish Harengula jaguana (n=44) versus crab prey Portunus gibbesii(n=27). Digestion code descriptions are given in Tables 1 and 2.

PPT (hr)

Baitfish Prey Crab Prey

Digestion Code % Digestion Mean(+S.E.) Range Mean(+S.E.) Range

0 <5 0.2 (0.09) 0.08 – 0.5 3.6 (1.08) 0.8 – 8.0

1 5 - 10 2.36 (0.53) 0.5 - 4.5 7.3 (0.75) 6.5 - 8.0

2 10 - 25 4.5 (1.50) 3.0 – 6.0 10.0 (0.00) 10.0

3 25 - 50 4.0 (0.50) 3.0 – 4.5 11.0 (1.00) 9.0 – 12.0

4 50 - 75 8.0 (0.67) 6.0 – 12.0 15.0 (1.73) 12.0 – 18.0

5 75 - 90 10.5 (2.03) 7.5 - 16.5 16.0 (0.00) 16.0

6 >90 14.8 (0.71) 9.0 – 18.0 20.9 (0.84) 16.0 – 24.0

66

Page 79: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

Table 15. Regression parameters of the average digestion code data of gag consuming baitfish prey fit to each model, for small gag (n=13), medium gag (n=16), and large gag (n=11). Models meeting selection criteria are highlighted in bold; assumptions met for homogeneity of variance and characteristics of the upper asymptote are given by Y (yes) or N (no). All models were significant at the α=0.05 level (p<0.0001) (see Figure 11).

67

Model Gag Size A B C Y-intercept MSE R2 Homogeneity of Variance

Upper Asymptote > Code 5.5

Small 0.74 0.36 0.74 0.90 0.86 Y

Medium

1.05 0.32 1.05 0.82 0.81 Y

Linear

Large 0.63 0.35 0.63 0.66 0.84 Y

Small 1.63 -0.09 1.63 1.67 0.75 Y

Medium 1.84 -0.08 1.84 1.33 0.70 Y

Exponential

Large 1.62 -0.08 1.62 1.22 0.72 Y

Small 1.22 0.10 1.22 1.25 0.81 Y

Medium 1.47 0.09 1.47 1.08 0.76 Y

Square Root

Large 1.23 0.09 1.23 0.94 0.79 Y

Small 6.02 -0.46 -5.32 0.47 0.52 0.93 Y YMedium 5.62 -0.57 -4.77 0.34 0.34 0.93 Y Y

Logistic

Large 5.43 -0.50 -5.66 0.30 0.28 0.94 Y N

Small 4.18 -1.66 0.00a 0.02 0.91 Y Y

Medium 4.56 -1.84 0.00a 0.01 0.93 Y Y

Power Exponentiala

Large 5.79 -1.97 0.00a 0.01 0.92 Y Y

a Due to model specifications, the Y-intercept of the power exponential model is fixed at Y=0 (0%) and not uniquely estimated.

Page 80: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

Table 16. Regression parameters of the digestion code data of gag consuming crab prey fit to each model, for small gag (n=8), medium gag (n=11), and large gag (n=8). Models meeting selection criteria are highlighted in bold; assumptions met for homogeneity of variance and characteristics of the upper asymptote are given by Y (yes) or N (no). All models were significant at the α=0.05 level (p≤0.0008) (see Figure 12).

68

Model Gag Size A B C Y-intercept MSE R2 Homogeneity of Variance

Upper Asymptote > Code 5.5

Small -2.04 0.42 -2.04 0.74 0.91 Y

Medium

-0.80 0.31 -0.80 0.48 0.93 Y

Linear

Large -0.43 0.30 -0.43 0.36 0.95 Y

Small 0.78 -0.10 0.78 1.66 0.79 Y

Medium 1.03 -0.08 1.03 1.35 0.81 Y

Exponential

Large 0.83 -0.09 0.83 0.87 0.88 Y

Small 0.12 0.11 0.12 1.13 0.86 YMedium 0.40 0.08 0.40 0.83 0.89 Y

Square Root

Large 0.41 0.09 0.41 0.60 0.92 YSmall 6.01 -0.56 -11.72 0.01 0.28 0.96 Y YMedium 6.22 -0.37 -12.55 0.06 0.07 0.99 Y Y

Logistic

Large 5.61 -0.36 -9.99 0.15 0.59 0.94 Y Y

Small 11.18 -4.80 0.00a 0.01 0.97 Y Y

Medium 11.29 -3.47 0.00a 0.01 0.96 Y Y

Power Exponentiala

Large 9.72 -2.38 0.00a 0.01 0.95 Y Y

a Due to model specifications, the Y-intercept of the power exponential model is fixed at Y=0 (0%) and not uniquely estimated.

Page 81: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

CHAPTER 4 DISCUSSION

Gastric Evacuation Models

The power exponential model (Elashoff et al., 1982) provided the best description

of the relationship of evacuated digesta as a function of PPT for gag, regardless of prey

type or method of measuring stomach contents, based on consuming meals at 1.5% of the

gag’s body weight (Tables 3, 4, 6 and 7, Figures 3 and 5). In all cases, R2 values were

very high (0.81-0.99), explaining >81% of the variation in both the baitfish and crab

digestion data. The gastric evacuation rates of several piscivorous fish species have

previously been described using the power exponential model, including black and

yellow rockfish consuming whitebait smelt Allosmerus elongatusus and purple shore crab

Hemigrapsus nudus (Hopkins & Larson, 1990), and Atlantic cod consuming capelin

Mallotus villosus and Atlantic herring Clupea harengus (dos Santos & Jobling, 1992).

Other studies have fit gastric evacuation data using the linear model (Swenson & Smith,

1973; Jones, 1974; Olson & Boggs, 1986), the exponential model (Brett & Higgs, 1970;

Tyler, 1970; Persson, 1979; MacDonald et al., 1982), the square root model (Jobling,

1981), the logistic model (Hopkins & Larson, 1990; Nelson & Ross, 1995), and with

expanded power exponential models incorporating time after ingestion, predator weight,

temperature, and meal size as predicting variables (Temming & Andersen, 1994;

Andersen, 1999; Koed, 2001). In particular, temperature has one of the strongest

influences on rates of gastric evacuation in fish, increasing gastric evacuation rates as

temperature increases (Bromley, 1994). In the present study, gastric evacuation rates

69

Page 82: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

70

were quantified at a mean temperature found on artificial reefs in the eastern Gulf of

Mexico during the warmer months of the year (mean=28°C from May-November)

because baitfish are only present on these reefs during these warmer months.

Goodness of fit evaluations among different gastric evacuation models has often involved

using a combination of Y-intercept values, residual plots, standard errors, residual mean

square values, and/or the coefficient of determination (R2) (Swenson & Smith, 1973;

Eleshoff, 1982; Brodeur & Pearcy, 1987; Ruggerone, 1989). Using the R2 value, which

represents the total variability in y that can be explained by the fitted regression (Zar,

1999), may not by itself reliably evaluate regression models (Draper, 1984; Healy, 1984;

Rao, 1998). Specifically, small R2 values can be statistically significant while large R2

values can be insignificant (Rao, 1998). In addition, equal values of the residual sums of

squares can result in different R2 values depending on the steepness of the regression,

with steeper regressions resulting in higher R2 values (Barrett, 1974; Rao, 1998). Using

R2 values to compare among different models can also be problematic if the models use

different numbers of predictor variables, since R2 will increase with an increasing number

of predictor variables (Healy, 1984). While in the present study goodness of fit was

determined by multiple critieria, including Y-intercepts, homogeneity of variance,

upper/lower asymptote criteria, and R2 values, using adjusted R2 values to determine

model fit may evaluate regression models more reliably (Healy, 1984; Rao, 1998).

Adjusted R2 values comprise the proportion of variance accounted for in the data and

takes into account both the number of predictor variables and the sample size, and is

calculated as (Healy, 1984)

Total

ResidualMS

MSadjR −= 12

(16)

Page 83: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

71

where MSResidual = residual mean squares and MSTotal = total mean squares. Since

adjusted R2 values are not influenced by the number of predictor variables within each

model, they should better evaluate fit between models with differing numbers of

variables, such as the logistic and power exponential models in this study.

Wet weight measurement error had the potential to affect this study due to the fact

that prey items recovered at later time periods could not be easily blotted prior to

weighing, and so water, enzymes, or body fluids may have remained and contributed to

an overestimation of the amount of prey remaining over PPT (Hopkins & Larson, 1990).

Because of this measurement error, gastric evacuation rate estimates and the shape of the

evacuation curve have often depended on whether the stomach contents were measured

by wet weight, dry weight, or volume (Brodeur, 1984; Olson & Boggs, 1986; Hopkins &

Larson, 1990; dos Santos & Jobling, 1992). Theoretically, dry weight measures of prey

remaining over PPT should have corrected for measurement error but it also introduced

new error because the original baitfish and crab dry weights had to be back-calculated

based on predictive regressions for wet weight versus dry weight of whole representative

prey. In this study, model selection was only minimally influenced by the way in which

the stomach contents were measured considering that the power exponential model best

fit both baitfish and crab wet weight and dry weight data, although this likely reflects the

curvilinear flexibility of the power exponential model. Because wet weight

measurements of the amount of food remaining in the stomach are the most common

measurement taken in the field (Bromley, 1994), the wet weight gastric evacuation curves

generated from gag feeding trials in this study may be the most useful.

Page 84: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

72

Both baitfish and crab prey used in all gastric evacuation feeding trials were

vacuum-sealed fresh, frozen immediately, and stored frozen because live baitfish were

only present on artificial reefs in the eastern Gulf of Mexico in the warmer months

(capturing and maintaining scaled sardines was not feasible) and collections of crab prey

were only successful during February and March. Previous work has shown that prey

(whole Cape anchovy Engraulis capensis, pieces of squid Loligo vulgaris, pieces of hake

Merluccius sp. and detached rock lobster tails Jasus lalandii) that had been frozen and

then thawed before use in in vitro digestion rate experiments digested faster in pepsin

than controls of fresh prey (Jackson et al., 1987). Using previously frozen prey in gastric

evacuation feeding trials may therefore result in underestimating the amount of food

remaining in the stomach at time and consequently lead to overestimating the gastric

evacuation rate. MacDonald et al. (1982) tested the effects of frozen versus live prey on

digestion indices and found that Atlantic cod and ocean pout Macrozoarces americanus

had significantly lower indices of digestion when consuming pellets of fresh bivalves

Yoldia sapotilla or polychaete worms Nephtys incise at 5 hrs and 20 hrs PPT,

respectively, as compared to thawed pellets. Conversely, MacDonald et al. (1982) also

found that ocean pout had lower indices of digestion at 20 hrs PPT when fed pellets of

thawed Y. sapotilla. Additionally, Atlantic cod and ocean pout had lower digestion

indices when fed thawed versus fresh polychaete worms N. incise at 5 and 20 hrs PPT,

respectively. In contrast, field and laboratory experiments with tagged, free-swimming

black and yellow rockfish indicated that the evacuation rates of thawed purple shore crab

determined in the laboratory gave acceptable approximations of evacuation rates in situ

using freshly fed purple shore crabs (Hopkins & Larson, 1990). Results from the present

Page 85: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

73

study indicate that thawed baitfish prey are 95% evacuated from the stomachs of gag

(300-750 mm TL) between 14.7–19.5 hrs PPT. These results correspond very closely

with preliminary field estimates on the stomach contents of wild gag, with times to 90%

gastric evacuation being approximately 15 hrs PPT (Lindberg et al., 2002).

Most gastric evacuation experiments have described predator digestion patterns

based on single-meals of a single prey type, even though multiple prey types may be

incorporated into a multivariate model (Tyler, 1970; MacDonald et al., 1982; Brodeur,

1984; Hopkins & Larson, 1990; He & Wurtsbaugh, 1993; Temming & Andersen, 1994;

Nelson & Ross, 1995; Singh-Renton & Bromley, 1996; Andersen, 1999; Koed, 2001).

Applying single-meal models to sequential-meal experimental conditions assumes that

stomach contents are homogenous, and considering this may not be the case, models may

tend to underestimate the evacuation rates of early meals and overestimate the rates of

later meals (Persson, 1984; Ruggerone, 1989; dos Santos & Jobling, 1992). Even so,

evidence suggests that the extrapolation of single-meal models can still yield reasonable

estimates of total daily ration (Persson, 1984; dos Santos & Jobling, 1992). Singh-

Renton and Bromley (1996) determined that the gastric evacuation rates of Atlantic cod

and whiting Merlangius merlangus fed mixed-prey meals of brown shrimp Crangon

vulgaris and Atlantic herring, or meals of brown shrimp and lugworm Arenicola marina

were not significantly different than the gastric evacuation rates of single-prey meals.

Further work is necessary to determine if the single-meal gastric evacuation rates

determined in this study are biased when applied to sequential-meal, mixed-meal, and

sequential-meal/mixed-meal situations, and whether or not their extrapolation will yield

acceptable estimates of total daily ration. Evacuation trials for meals of combined fish

Page 86: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

74

and crab prey were not addressed in the present study but would be useful to develop for

gag, since some stomach contents recovered in the field contain mixed prey types

(personal observation).

Effects of Prey Type

For all sizes of gag consuming baitfish or crab prey, on both a wet weight and dry

weight basis, the power exponential model best fit the gastric evacuation data (Tables 3,

4, 6, and 7, Figures 3 and 5). Other studies have used the power exponential model,

including those on black and yellow rockfish consuming whitebait smelt (Hopkins &

Larson, 1990), and on Atlantic cod consuming capelin, herring, prawn Pandalus borealis,

and krill Meganyctiphanes norvegica (dos Santos & Jobling, 1992). The power

exponential model in the present study accounted for the lack of a lag phase seen in the

gastric evacuation of scaled sardine prey based on wet weight and with the slight lag

phase observed in the dry weight data (0-3 hrs PPT) (Figure 3). The power exponential

model best fit black and yellow rockfish gastric evacuation data based on the wet weight

and dry weight measurements of recovered whitebait smelt, both measurement types

showed lag phases in digestion, and 80-90% prey remaining at 2-5 hrs PPT, likely due to

the lower temperature (14.5°C) (Hopkins & Larson, 1990). Likewise, dos Santos and

Jobling (1992) used the power exponential model and found that on a dry weight basis

the gastric evacuation rates of Atlantic cod fed capelin, Atlantic herring, prawn, or krill

were initially faster that rates of gastric evacuation considered on a wet weight basis.

As with most fish predators, gag capture and swallow their prey whole and

therefore, along with mechanically macerating the prey, digestive enzymes must work

through the prey’s scales and skin or carapace before the muscle, gut, and other tissues

can be easily digested (Diana, 1995). The scaled sardine prey chosen for this study were

Page 87: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

75

immature and therefore, relatively small and low in energy density compared to mature

sardines. However, they were of a similar size to prey found in the stomach contents of

gag during warmer months on reefs in the eastern Gulf of Mexico (Lindberg et al., 2002).

Because of their size, the sardine prey had fairly small scales and thin skin, therefore

these sardines were considered highly friable compared to mature sardines of the same

species. The gag’s digestive enzymes were likely better able to quickly and continuously

break down these comparatively low-energy, immature sardines. In fact, the linear model

described the continuous gastric evacuation of baitfish prey on a wet weight basis for

each gag size class relatively well (r2>0.93) (Table 3). Although low in energy compared

to mature sardines, these immature sardines are high-energy prey when compared to

other prey types such as portunid crabs (Table 9). Therefore, the fact that a linear model

adequately describes the gastric evacuation process of baitfish prey is still consistent with

the continuous digestion of high energy prey types over PPT as described by Jobling

(1987).

The power exponential model also provided the best fit for both the wet weight and

dry weight gastric evacuation data for all sizes of gag consuming crab prey (Tables 6 and

7, Figure 5). Similarly, the wet weight and volume gastric evacuation data from black

and yellow rockfish fed purple shore crabs were modeled using a power exponential

model, although the dry weight gastric evacuation data was better fit using a linear model

(Hopkins & Larson, 1990). Few studies have modeled the gastric evacuation of crab prey

fed to piscivorous fish species. Most studies have concentrated on shrimp, krill,

amphipods, or polychaete prey, which all tend to possess thinner exoskeletons and

therefore less chitin material than most crab exoskeletons (Tyler, 1970; MacDonald et al.,

Page 88: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

76

1982; dos Santos & Jobling, 1992; Temming & Andersen, 1994; Andersen, 1999). The

power exponential model was able to capture the large digestive lag phases seen in the

gastric evacuation patterns of the crab prey (Figure 5). Generally, when gag consumed

crab prey on a wet weight basis, lag phases of approximately 0-6.5 hrs with little or no

digestion occurred, followed by a rapid increase in digestion for approximately the next

12-15 hrs. Digestion was then slowed again as residual hard parts persisted in the

stomach, accounting for the model’s lower asymptote estimates. At these later time

intervals, digestion-resistant exoskeleton material and other residual hard parts are known

to cause evacuation curves to level off and form the lower asymptotes of various gastric

evacuation models (Battle, 1935; Windell, 1966; dos Santos & Jobling, 1992). Basing

models on the dry weight data tended to decrease the lag phases (0-3 hrs), thereby

appearing more similar to the dry weight digestion curves of gag consuming baitfish

prey. This trend was likely caused by measurement error because the crabs may have

more fluid within their exoskeleton normally, as compared to baitfish prey, and in

addition, this fluid retention may vary greatly with crab molt stage. The crabs could not

be easily damp-blotted for weighing and therefore water and body fluids may have

remained within the crabs’ exoskeletons, thereby causing the large lag phases seen in the

wet weight data (Figure 5). Similarly, Hopkins and Larson (1990) found that modeling

the gastric evacuation data from black and yellow rockfish consuming the purple shore

crab on a dry weight basis reduced the digestive lag phase, as compared to modeling the

data on a wet weight basis.

Lags in digestion after prey consumption have been attributed to impaired enzyme

secretion, force-feeding, and starvation, and have the potential to cause bias by

Page 89: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

77

overestimating the percentages of prey remaining at PPT (Swenson & Smith, 1973;

Jones, 1974; MacDonald et al., 1982; Jackson et al., 1987). In the present study,

however, food was withheld from gag for <48 hr prior to testing and the gag voluntarily

consumed thawed baitfish or crab prey. In addition, most gag either gained a very slight

amount of weight or maintained their weight while in captivity. Therefore, the observed

lag phase was not due to force-feeding or starvation. More likely, the lag was due to the

gastric softening and digestive break down of whole crustaceans, through the secretion of

hydrochloric acid to decalcify the calcium carbonate material found in their exoskeletons,

plus chitinase, chitobiase, and mechanical peristalsis necessary to breakup the chitinous

material in the softened exoskeleton (Pandian, 1967; Lindsay, 1984; Hopkins & Larson,

1990; Lovell, 1998). In addition, the pyloric sphincter may limit the size of items that

pass into the intestine, thereby playing a role in the retention of exoskeletal material and

the formation of digestive lag phases.

Effects of Predator Size

Gag size had a significant effect on their rates of gastric evacuation, whether they

were consuming baitfish prey or crab prey, and whether stomach contents were measured

on a wet weight or dry weight basis. Similarly, studies on haddock Melanogrannus

aeglefinus, Atlantic cod, and whiting fed saithe Pollachius virens (Jones, 1974), turbot

Scophthalmus maximus fed processed pellets (Flowerdew & Grove, 1979), and Atlantic

cod fed capelin, Atlantic herring, prawn, or krill (dos Santos & Jobling, 1995), found that

predator size had a significant effect on the rate of gastric evacuation, with smaller

predators having faster rates of gastric evacuation. In contrast, other studies on Atlantic

cod that were fed shrimp Pandalus montagui (Tyler, 1970), dab Limanda limanda fed an

artificial paste diet (Jobling et al., 1977), and plaice Pleuronectes platessa fed fish-paste

Page 90: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

78

(Jobling, 1980a) did not observe size-specific evacuation rates. Previous work has shown

that feeding predators meal sizes relative to their body weight accounts for the variability

associated with gastric evacuation rates in different sized predators (Swenson & Smith,

1973). Swenson and Smith (1973) found that large and small fish evacuate meals at

approximately the same rate when meals where fed relative to their body weight, whereas

their evacuation rates differed when they were fed equal-sized meals (Swenson & Smith,

1973). In contrast, the present study shows that gag size significantly affects rates of

gastric evacuation despite feeding meals of baitfish or crab relative to each gag’s body

weight. Meal size was not included as a predicting factor in the gastric evacuation

models because initial meal sizes cannot be estimated in wild gag caught in the field.

Scaling factors were incorporated into the power exponential models for both prey

types and measures of stomach contents (wet weights and dry weights) to account for

significant differences in either gag weight or total length, based on the assumption that

the effects of predator size on gastric evacuation times are multiplicative (Andersen,

1999) (Tables 5 and 8; Figures 4 and 6). Previous studies have incorporated scaling

factors such as temperature, meal size, prey size, prey energy density, predator weight

(W), and predator total length (TL) based on this assumption (dos Santos & Jobling,

1992; Temming & Andersen, 1994; Andersen, 1999; Andersen, 2001; Koed, 2001). In

the present study, all W and TL coefficients were ≤0.00134. Therefore, while gag size

significantly affected the gastric evacuation rates of both baitfish and crab prey, values

for W and TL raised to a power approaching 0.0 created multiplicative scaling factors for

the evacuation models that were very close to 1.0 (e.g., W=1). These small but

significant scaling factors are reasonable considering that the gag were fed meals on a

Page 91: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

79

relative weight basis, i.e., a percentage of gag body weight, because using the percentage

of meal recovered has been known to take out much of the variation in the weight of

meals recovered from different sized predators (Bromley, 1994). Incorporating W or TL

scaling factors into the power exponential model with data on gag <200 mm TL and >750

mm TL should further improve the R2 values of the power exponential model, better

determine the extent to which the W and TL exponential scalers account for gag size, and

better predict the percentages of prey remaining in the stomachs of all sizes of gag.

Prey Composition

Many fish secrete chitinase but it has been commonly accepted that piscivorous

fish, especially marine fish, are not efficient at converting carbohydrates, such as chitin,

into an absorbable form and therefore can not utilize chitin as an energy source because

the β-linkages of glucose within the chitin molecule cannot be broken by amylase

enzymes (which only breaks α-linkages) (Battle, 1935; MacDonald et al., 1982; Lindsay

& Gooday, 1985, Medved, 1985, Vollhardt & Schore, 1987). However, the role of

chitinase in the gastric processes of fish has been somewhat controversial. Lindsay

(1984) hypothesized that the primary function of chitinase in the gut of fish may be to

chemically disrupt the outer chitinous material of prey, such as crabs. Lower levels of

chitinase activity were found in fish that mechanically disrupted prey before ingestion

compared to fish that swallowed prey whole and likely required more chitinase to break-

up chitinous material (Lindsay, 1984). Conversely, other work has shown that high

levels of gastric N-acetyl-D-glucosamine (NAG) were caused by high levels of

chitinolytic enzymes (chitinase and chitobiase) and resulted in significantly lower growth

rates in fish fed diets containing 4, 10, and 25% chitin, as compared to diets containing

starch (Lindsay et al., 1984). Jackson et al. (1992) suggested that absorbing NAG and D-

Page 92: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

80

glucosamine (Gln) (i.e., the products of chitin digestion) may actually be inhibiting

growth. Kohn et al. (1962) reported that mammals metabolize NAG and Gln more

slowly than D-glucose, which possibly inhibited individual growth because NAG and Gln

could not be utilized as quickly as D-glucose for metabolic, waste removal, and growth

processes (Jackson et al., 1992). Jackson et al. (1992) hypothesized that, although chitin

may be available to assimilate, fish that consume chitin may have adapted to inefficiently

absorb chitin end-products because of the possible costs associated with reduced growth.

One advantage of chitinolysis seems to be the increase in gastric evacuation rates

associated with the increase in mechanical breakdown of the exoskeleton, thereby

allowing easier access to more readily digestible tissues and allowing proteolytic

enzymes better access to cuticle proteins (Jackson et al., 1992). It has been commonly

accepted that crustaceans with their chitinous exoskeletons allow piscivorous predators to

assimilate only between 70–80% of the total crustacean energy consumed, versus

approximately 89-96% of the total energy when fish prey are consumed (Diana, 1995).

In the present study, the mean energy densities of both baitfish and crab prey were

determined, along with the mean energy density of the crab shells by themselves, to

better estimate the mean energy density available for assimilation by individual gag

consuming crab prey. However, estimations of exoskeletal energy densities may have

been biased. Potassium hydroxide treatments for tissue removal of the crab prey did not

decalcify the exoskeleton or extract the cuticle’s pigments (Pandian, 1967). Compere et

al. (2002) has shown, however, that many proteins can be extracted from the exoskeleton

without decalcification and that many proteins in the exoskeleton have relatively high

apparent molecular weights. The presence of calcium carbonate, and potentially,

Page 93: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

81

pigments, remaining in the crab exoskeletons may have lowered the estimates of

exoskeletal energy density.

All representative baitfish and crab prey used in the prey analyses reflected size

(TL and CL) ranges that mirrored those normally found in the stomach contents of wild

gag (Murie, unpublished data). The representative baitfish prey had lower mean mass,

higher mean % moisture, lower mean % ash, higher mean available caloric energy

density, and higher mean available ash-free caloric energy density estimates than the crab

prey (Table 9). Quantifying baitfish energy densities in terms of their ash-free dry weight

corrects for the inorganic materials of their bones and other parts, such as scales. The

crab prey used in the regression analysis to correct for the known unavailable energy in

crab fed to gag, specifically unabsorbable chitin energy, were within the size range (CL)

and mass (g) of the whole crabs used for the prey analyses. The composition of food

prey, in terms of energy density and specifically, fat and ash content, is known to affect

rates of gastric emptying (Elliott & Persson, 1978; Jobling, 1980a). Diets containing

unabsorbable materials tend to lower a meal’s energy density and increase rates of gastric

evacuation (Flowerdew & Grove, 1979; Jobling, 1980a). On the other hand, digestion-

resistant materials have been shown to slow rates of gastric evacuation due to the

relatively large amounts of residual matter and ash that must be digested and passed from

the body (Battle, 1935; Windell, 1966; Hopkins & Larson, 1990; dos Santos & Jobling,

1992). Therefore, prey with high amounts of digestion-resistant materials, such as crabs

with exoskeletal coverings, would be expected to be lower in energy density and have

slower gastric evacuation rates (Flowerdew & Grove, 1979; Jobling, 1980a). In fact, in

the present study, times to 95% evacuation at 28.0oC for small, medium, and large gag

Page 94: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

82

consuming crab prey on a wet weight basis took 4.3 hrs longer than gag consuming

baitfish (Figure 7). In a previous study, black and yellow rockfish evacuated whitebait

smelt to 95% in 27 hrs, while purple shore crabs were not evacuated to 95% until 49.5 hrs

PPT, a 22.5 hr difference (Hopkins & Larson, 1990). Similarly, there was no lag phase in

digestion for black and yellow rockfish fed whitebait smelt, as 46.8% of the meal was

remaining at 10 hrs PPT, compared to a large lag phase seen when rockfish were fed the

purple shore crab and 99.7% of the meal was remaining 10 hrs PPT, a 52.9% difference

(Hopkins & Larson, 1990). While these earlier findings are corroborated by the present

study, it is important to recognize that the whitebait smelt used in the black and yellow

rockfish feeding trials were thin-skinned and almost scaleless, while the purple shore

crabs were smaller, relative to the scaled sardine and portunid crab prey used in the

present study.

Most digestion and consumption models do not account for fluctuating prey energy

densities between species and seasons (Koed, 2001). In the present study, immature

scaled sardines were collected during late October or November of 2002 and 2003, and

all sardines collected or observed on artificial reefs in the eastern Gulf of Mexico

(Lindberg et al., 2002) were of a similar size range. Due to the fact that all baitfish,

including immature scaled sardines, are only present on artificial reefs in the eastern Gulf

of Mexico in the warmer months, fluctuating baitfish prey energy densities between

seasons did not need to be determined. Crab prey, however, were collected in February

and March of 2003 and 2004. Spring samples of crab could have biased the rates of

gastric evacuation and estimates of energy density determined in this study. Crab

samples collected in the spring may have contained higher percentages of pre-molting

Page 95: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

83

and post-molted individuals that have no gonad development, resulting in greatly reduced

energy densities. However, considering that all crabs used in the prey composition

determinations were similar in energy density (mean=2.22, S.E.=0.06) and more than

81% of the variation in the gastric evacuation data was explained by the power

exponential model, any biases caused by pre-molt or post-molt individuals were likely

small and within error. Comparing the energy of crabs sampled in the present study to

crabs sampled from the stomachs of wild gag could determine the extent to which pre-

molt and post-molt crabs potentially biased results from the present study. Determining

the extent of fluctuating crab prey energy densities by season may be necessary

depending on when they begin to show up in the gags’ diet, or if they are continuously

found in their diets, and their seasonal molting patterns. Considering that prey

composition and energy densities greatly affect both gastric evacuation and consumption

rates, it is important to know the true energy densities of the prey consumed by season.

Stomach Content Composition

Both the scaled sardines and portunid crabs used in the gastric evacuation and

average digestive code regressions were within the size range of the subsampled prey

used in the whole prey regression analyses, although the measurements of mean mass for

both prey types was slightly smaller (Tables 9 and 10). The baitfish stomach contents

had higher mean % moisture and mean available caloric energy density estimates but

lower mean % ash and mean available ash-free caloric energy density estimates than the

representative baitfish prey. All values are very close to those of the representative

baitfish prey and likely reflect the natural variation seen in the scaled sardine population

residing on artificial reefs in the eastern Gulf of Mexico. The crab prey stomach contents

had higher mean % moisture, mean available caloric energy density, and mean available

Page 96: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

84

ash-free caloric energy density estimates but a lower mean % ash estimate than the

representative crab prey. Again, all values are very close to those of the representative

crab prey, most likely reflecting the natural variation seen in the crab population. The

fact that mean % moisture differed significantly between whole baitfish and crab prey but

did not differ between baitfish and crab stomach contents may reflect wet weight

measurement error ( i.e., the ability to damp-blot excess water off recovered prey), or the

addition of secreted digestive juices (hydrochloric acid, pepsin, and mucous).

The square-root and linear models provided adequate fits for the gross energy

(kcal/g dry wt) of baitfish and crab prey stomach contents, respectively, regressed as a

function of PPT (Tables 11 and 12, Figure 8). However, the square-root model fit to the

baitfish data was more descriptive than predictive, likely reflecting the friable nature of

the baitfish prey that was easily broken down through mechanical and enzymatic action

in the gag’s stomach. This likely resulted in a homogenous mixture of baitfish with a

relatively consistent energy density in the gag’s stomach. In contrast, the crab prey may

not comprise such a homogenous mixture because the crabs contain higher amounts of

indigestible matter, such as chitin, that may preferentially remain in the stomach for

longer periods. Few studies have looked at stomach content gross energy as a function of

PPT. Jobling (1980a) fed plaice meals of fish paste with kaolin (inert material) and

regressed the square-root of the meal’s energy density by time with a linear model and

found that energy decreased over PPT. A linear model assumes that energy turnover is

maintained at a constant level, therefore, non-nutritive bulk , such as chitin, should be

instantaneously evacuated from the stomach (Jobling, 1980a). Jobling (1980a)

hypothesized that as the energy density of meals increased, muscular activity and rates of

Page 97: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

85

gastric peristalsis may restrain energy turnover. One study by Beamish (1972) on

largemouth bass Micropterus salmoides fed emerald shiners Notropis atherinoides noted

that the caloric energy density of the stomach contents increased over time when

expressed as cal/g dry weight, and that protein (i.e., nitrogen) digested more quickly than

lipid, but did not fit a quantitative function to the data. Conversely, Bromley (1988)

found that the gross energy density (kcal/g dry weight) of sandeel Ammodytes sp. in the

stomach contents of whiting declined over time and attributed the decline to changes in

the ratio of readily digestible soft tissue to skeletal material but again, the function was

not modeled. In the present study, the caloric energy density of the stomach contents for

both prey types increased over PPT for each size of gag, except small gag consuming

baitfish (Figure 8). These results most likely reflect energy:weight ratio changes over

PPT, that is, a larger portion of low energy unabsorbable skeletal or exoskeletal material

remains in the stomach longer compared to proteins or lipids but, at the same time, large

portions of digesta and heavy macromolecules are evacuated from the gut. Relatively

heavy macromolecules, such as calcium carbonate and some proteins within the

exoskeleton, may be extracted from the exoskeleton with the addition of digestive

enzymes (including HCl) more quickly then lighter macromolecules. A greater weight

change in the energy:weight ratio compared to the energy change due to low energy

skeletal or exoskeletal material remaining in the gut may be causing the increase in the

gross caloric energy densities of the recovered gag stomach contents over time.

The power exponential model best described the percentage of stomach content

energy digested over PPT for both prey types and in all sizes of gag, with predictive

models explaining over 87% of the variation (R2 = 0.87-0.96) (Tables 13 and 14, Figure

Page 98: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

86

9). Previous work, such as that of Beamish (1972) and Bromley (1988), have determined

gross energy densities over time, others have estimated the amount of energy consumed,

absorbed, converted, and residual energy densities per day as a function fish body weight

for estimations of absorption efficiency (Pandian, 1967). In the present study, the power

exponential model fit the percentage of digested baitfish energy in a pattern most similar

to the gastric evacuation of the wet weight data for baitfish (A ranging from 5.64 to 7.61

and from 6.16 to 8.01, and B ranging from -1.46 to -1.98 and from 1.52 to 1.89, for the

digested energy data and wet weight gastric evacuation data, respectively) (Tables 3 and

8, Figures 3 and 9). However, the lag phases present in the baitfish energy digested data

(2-3 hrs PPT) were more similar to the lag phases seen in the dry weight gastric

evacuation data of small and large gag (2-2.5 hrs PPT). Conversely, the power

exponential model fit the percentage of digested crab energy from the gag stomach

contents in a pattern most similar to the gastric evacuation of the dry weight crab prey

data, with A ranging from 9.07 to 10.23 and from 8.00 to 9.19, and B ranging from -1.73

to -2.78 and from 1.23 to 2.32, for the digested energy data and dry weight gastric

evacuation data, respectively (Tables 7 and 14, Figures 5 and 9). Lag phases were more

similar to the wet weight gastric evacuation crab data, with the lag in the percentage of

digested crab prey energy being approximately 5 hrs PPT and the wet weight gastric

evacuation lag phase being 6 hrs PPT.

Indices of Digestion

Visual indices of prey digestive stages over PPT are one way that researchers can

estimate times of prey consumption in the field based on the average digestion codes

given to recovered stomach contents instead of quantitative measurements (i.e., back-

calculation of original prey sizes and weights). When each size class of gag consumed

Page 99: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

87

either baitfish or crab prey the power exponential model best described their average

digestive codes over PPT (Tables 15 and 16, Figures 11 and 12). Again, little or no lag

phase was seen in the average digestion code data when gag consumed baitfish prey due

to the fact that the gag’s digestive enzymes and mechanical peristalsis of the stomach

could easily breakup the individual prey items. Lag phases of up to 6.5 hrs in digestion

of crab were seen in both the average digestion code data and the wet weight gastric

evacuation data, as the crab’s exoskeleton acts as a barrier to prevent quick enzymatic

and mechanical digestion of the crab tissues (MacDonald et al., 1982; Hopkins & Larson,

1990; dos Santos & Jobling, 1992; Bromley, 1994). These visual indices of scaled

sardine and portunid crab digestive stages clearly support gastric evacuation results from

the present study and the contention that prey containing higher percentages of chitin will

remain recognizable for longer periods in the gut (MacDonald et al., 1982). Evacuation

studies on Atlantic cod, ocean pout, winter flounder Pseudopleuronectes americanus, and

American plaice Hippoglossoides platessoide fed bivalves Y. sapotilla, amphipods

Tmetonyx cicada, and polychaetes N. incisa have also found that visual indices of prey

digestion are correlated with quantitative differences in evacuation rates between prey

types (MacDonald et al., 1982). Using the visual indices of prey digestion qualified in

this study, along with their corresponding models of average digestion codes at time, will

allow researchers to approximate the wet weight and dry weight percentages of prey

remaining in a gag’s stomach and estimate an approximate time of prey consumption for

wild gag, based on preliminary average daily consumption estimates of between 1.2-1.8%

body weight per day (Lindberg et al., 2002). Therefore, only average digestion codes of

prey at time will need to be estimated, thereby saving measurement effort and processing

Page 100: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

88

time for stomach contents collected in the field. Most importantly, these back-calculated

times of prey consumption can be used to determine the frequency with which gag feed

and their diurnal pattern of feeding. Both feeding frequency and the diurnal feeding

pattern (synchronous or asynchronous) of gag will affect which type of consumption

model is most appropriate for wild gag (Adams & Breck, 1990).

Consumption

Based on results from this study, gag will likely require a multiplicative

consumption model, such as that of Atlantic cod feeding on Atlantic herring in the

Barents Sea (Johansen et al., 2004) or an additive model, with a delay for gag consuming

crab. The multiplicative or additive model chosen must incorporate gastric evacuation

models explicit to the prey types consumed by gag, i.e., both fish and crustacean prey.

Such models can account for the differing gastric evacuation rates of various

representative prey items, such as baitfish, crab, squid, and possibly larger pelagic prey.

Water temperature influences on gag gastric evacuation rates must be determined on a

species-specific basis. Similarly, each species’ seasonal energy density fluctuations

should be incorporated into a comprehensive gastric evacuation model for gag, taking

into account the seasonality of prey quality and quantity

Predators consuming prey that contain less available energy and higher amounts of

indigestible materials that slow rates of gastric evacuation, such as crab prey, will need to

increase their consumption of that prey to grow at rates similar to predators consuming

more energy-rich prey, such as baitfish. If baitfish and crab energy densities were equal,

the rate of crab prey exploitation would likely better approximate a Type II Functional

Response curve (Holling, 1959) compared to baitfish prey as gag became satiated.

Specifically, handling time, which includes the pursuit, capture, handling, and gastric

Page 101: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

89

evacuation of prey, is taken into account, resulting in the number of prey consumed per

predator initially rising as prey density increases and then leveling off with further

increases in prey density (Holling, 1959). Results from the present study clearly show

that crab prey require more handling time (at a minimum, more gastric evacuation time)

than baitfish prey, with times to 95% gastric evacuation of baitfish and crab prey between

14.7-19.5 hrs and 19.6-24.5 hrs PPT, respectively . Based on the Type II Functional

Response curve, the consumption of crab prey by gag will level off at a maximum rate

that is lower than the maximum rate of baitfish consumption because crabs require more

handling time in digesting food. Therefore, even when gag are fed to satiation on crab,

energy available for gag growth will be limited to the time required for individual gag to

handle the crab prey. Gag fed to satiation on friable baitfish prey should be limited less

by prey handling times, resulting in more energy available for individual gag growth.

Based on preliminary analyses of gag stomach contents, it is unclear why some gag may

be consuming significant amounts of energy-poor crab prey (Lindberg et al., 2002).

Different prey have different vulnerabilities, in terms of encounter and capture success

rates, therefore, differences in consumption rates of baitfish and crab prey may be a

function of their differing vulnerabilities (Sih & Christensen, 2001). Optimal Foraging

Theory predicts that prey are detected and consumed based on their energy density, in

addition to the time, effort and risk involved in capture of prey (Emlen, 1966; MacArthur

and Pianka, 1966; Gill, 2003), i.e., if the time and effort needed to capture baitfish and

crab prey were similar, a gag should preferentially choose energy-rich baitfish prey over

lower energy crab prey. However, density-dependent effects may reduce a gag’s

successful foraging episodes due to increased competition with a reduction in prey

Page 102: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

90

numbers, thereby causing an increase in the gag’s optimal diet breadth to include lower

energy prey, such as crabs (Emlen, 1966; MacArthur and Pianka, 1966). Interestingly,

some gag seemed to naturally take to crab prey much more readily in the laboratory than

others (personal observation). This feeding disparity may be evidence of differing prey

recognition capabilities, previous experience, individual food preferences in gag,

previous density-dependant effects on foraging successes, or differing portunid crab and

baitfish abundances in the environment affecting predator-prey encounter rates.

Conclusions

In conclusion, the power exponential model best fit both the wet weight and dry

weight gastric evacuation data irrespective of gag size and prey type. Gag gastric

evacuation rates as a function of PPT were significantly affected by baitfish and crab prey

composition, in terms of the curve’s shape and total prey retention times, most probably a

result of the crab exoskeleton acting as a barrier to digestion, differing levels of

unabsorbable material, and energy density. Gag size (300-750 mm TL) significantly

affected both the scaled sardine and portunid crab gastric evacuation models when using

either the wet weight or dry weight data. This range in gag size reflects the stage-specific

size range of gag normally encountered on coastal reefs in the eastern Gulf of Mexico

(Lindberg et al., 2002). Results from the present study independently verify preliminary

gastric evacuation rate estimates of wild gag consuming baitfish prey on artificial reefs

off the west coast of Florida (Lindberg et al. 2002). Based on field estimates, wild gag

evacuated 90% of a baitfish meal in 15 hrs and 100% in 16 hrs, compared to 95% gastric

evacuation between 14.7-17.4 hrs PPT determined from experimental feeding trials of

captured gag in this study. The flexibility of the power exponential model also best fit

the percentage of stomach content energy digested over time and the average scaled

Page 103: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

91

sardine and portunid crab digestion code data for each size of gag. Additional work on

quantifying the gastric evacuation rates of gag < 300 mm TL and >750 mm TL, at

temperatures <28°C and >28°C, with different relative meal sizes, mixed meals,

sequential meals, and differing prey types will improve the ability of the best-fit power

exponential model to predict the percentages of prey remaining over PPT, the

percentages of prey energy digested over PPT, and the average digestion codes over PPT

of wild gag. The power exponential gastric evacuation models for gag determined in the

present study have shown that prey type and gag size are not only significant parameters

when determining rates of gastric evacuation but they, along with back-calculated times

of prey consumption by wild gag, are variables that must be included in any future

consumption model for wild gag.

Page 104: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

LIST OF REFERENCES

Adams, S. M. & Breck, J. E. (1990). Bioenergetics. In Methods for Fish Biology (Schreck, C. B. & Moyle, P. B., eds), pp. 389-409. Bethesda, Maryland: American Fisheries Society.

Adams, S. M., McLean, R. B., & Huffman, M. M. (1982). Structuring of a predator population through temperature-mediated effects on prey availability. Canadian Journal of Fisheries and Aquatic Science 39, 1175-1185.

American Fisheries Society (AFS). (2004). Guidelines for the use of fishes in research. Online.http://www.fisheries.org/html/Public_Affairs/Sound_Science/Guidelines2004.shtml. Last accessed, March 25, 2005.

Andersen, N. G. (1999). The effects of predator size, temperature, and prey characteristics on gastric evacuation in whiting. Journal of Fish Biology 54, 287-301.

Andersen, N. G. (2001). A gastric evacuation model for three predatory gadoids and implications of using pooled field data of stomach contents to estimate food rations. Journal of Fish Biology 59, 1198-1217.

Association of Official Analytical Chemists (AOAC). (1990). Method 925.09. In Official Methods of Analysis, 15th edn. Arlington, Virginia: Association of Official Analytical Chemists.

Association of Official Analytical Chemists (AOAC). (1990). Method 926.08. In Official Methods of Analysis, 15th edn. Arlington, Virginia: Association of Official Analytical Chemists.

Association of Official Analytical Chemists (AOAC). (1990). Method 923.03. In Official Methods of Analysis, 15th edn. Arlington, Virginia: Association of Official Analytical Chemists.

Battle, H. I. (1935). Digestion and digestive enzymes in the herring (Clupea herrengus L.). Journal of the Biological Board of Canada 1, 145-157.

Beamish, F. W. H. (1972). Ration size and digestion in largemouth bass, Micropterus salmoides Lacepede. Canadian Journal of Zoology 50, 153-164.

92

Page 105: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

93

Bledsoe, E. L. & Phlips, E. J. (2000). Relationships between phytoplankton standing crop and physical, chemical and biological gradients in the Suwannee River and plume region, U.S.A. Estuaries 23, 458-473.

Brett, J. R. & Higgs, D. A. (1970). Effect of temperature on the rate of gastric digestion in fingerling sockeye salmon, Oncorhynchus nerka. Journal of the Fisheries Research Board of Canada 27, 1767-1779.

Brodeur, R.D. (1984). Gastric evacuation rates for two foods in the black rockfish, Sebastes melanops Girard. Journal of Fish Biology 24, 287-298.

Brodeur, R. D. & Pearcy, W. G. (1987). Diel feeding chronology, gastric evacuation and estimated daily ration of juvenile coho salmon, Oncorhynchus kisutch (Walbaum), in the coastal marine environment. Journal of Fish Biology 31, 465-477.

Bromley, P. J. (1988). Gastric digestion and evacuation in whiting, Merlangius merlangus (L.). Journal of Fish Biology 33, 331-338.

Bromley, P. J. (1994). The role of gastric evacuation experiments in quantifying the feeding rates of predatory fish. Reviews in Fish Biology and Fisheries 4, 36-66.

Butler, D. A., Palmer, W. E., & Dowell, S. D. (2004). Passage of arthropod-diagnostic fragments in Northern Bobwhite chicks. Journal of Field Ornithology 75, 372-375.

Collins, L. A., Johnson, A. G., Koenig, C. C., & Baker, M. S. (1998). Reproductive patterns, sex ratio, and fecundity in gag, Mycteroperca microlepis (Serranidae), and protogynous grouper from the northeastern Gulf of Mexico. Fishery Bulletin 96, 415-427.

Compere, P., Jaspar-Versali, M., F., & Goffinet, G. (2002). Glycoproteins from the cuticle of the Atlantic shore crab, Carcinus maenas: I. Electrophoresis and western-blot analysis by use of lectins. The Biological Bulletin 202, 61-73.

Diana, J. S. (1979). The feeding pattern and daily ration of a top carnivore, the northern pike (Esox lucius). Canadian Journal of Zoology 57, 2121-2127.

Diana, J. S. (1995). Biology & Ecology of Fishes. Carmel, Indiana: Cooper Publishing Group LLC.

Dorcas, M. E., Peterson, C. R., & Flint, M. E. T. (1997). The thermal biology of digestion in rubber boas (Charina bottae): Physiology, behavior, and environmental constraints. Physiological Zoology 70, 292-300.

dos Santos, J. & Jobling, M. (1992). A model to describe gastric evacuation in cod (Gadus morhua L.), fed single-meals of natural prey. ICES Journal of Marine Science 49, 145-154.

Page 106: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

94

dos Santos, J. & Jobling, M. (1995). Test of a food consumption model for the Atlantic cod. ICES Journal of Marine Science 52, 209-219.

Draper, N. R. (1984). The Box-Wetz criterion versus R2. Journal of the Royal Statistical Society 147, 100-103.

Elashoff, J. D., Reedy, T. J., & Meyer, J. H. (1982). Analysis of gastric emptying data. Gastroenterology 83, 1306-1312.

Elliott, J. M. & Persson, L. (1978). The estimation of daily rates of food consumption for fish. Journal of Animal Ecology 47, 977-991.

Emlen, J. M. (1966). The role of time and energy in food preference. American Naturalist 100, 611-617.

Florida Fish and Wildlife Conservation Commission (FFWCC). (2003). Gag (Mycteroperca microlepis). Online. Http://research.myfwc.com/engine/ download_redirection_process.asp?file=revgag_1004.pdf&objid=5131&dltype=article. Last accessed, March 25, 2005.

Flowerdew, M. W. & Grove, D. J. (1979). Some observations of the effects of body weight, temperature, meal size and quality on gastric emptying time in the turbot, Scophthalmus maximus (L.) using radiography. Journal of Fish Biology 14, 229-238.

Gill, A. B. (2003). The dynamics of prey choice in fish: the importance of prey size and satiation. Journal of Fish Biology 63, 105-114.

Grove, D. J., Loizides, L. G., & Nott, J. (1978). Satiation amount, frequency of feeding and gastric emptying rate in Salmo gairdneri. Journal of Fish Biology 12, 507-516.

Haddon, M. (2001). Modeling and Quantitative Methods in Fisheries. Boca Raton, Florida: Chapman and Hall/CRC.

Hazel, J. R. (1993). Thermal biology. In The Physiology of Fishes (Evans, D. H., ed.), pp. 427-467. Boca Raton, Florida: CRC Press, Inc.

He, E. & Wurtsbaugh, W. A. (1993). An empirical model of gastric evacuation rates for fish and an analysis of digestion in piscivorous brown trout. Transactions of the American Fisheries Society 122, 717-730.

Healy, M. J. R. (1984). The use of R2 as a measure of goodness of fit. Journal of the Royal Statistical Society 147, 608-609.

Henriques, L. T., da Silva, J. F. C., & Vasquez, H. M. (2004). Effect of acipin on the degradability and rate of passage of elephant-grass and corn silages in Holstein x Zebu cattle. Arquivo Brasileiro de Medicina Veterinaria e Zootecnia 56, 757-763.

Page 107: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

95

Holling, C. C. (1959). Some characteristics of simple types of predation and parasitism. Canadian Entomologist 91, 385-398.

Hood, P. B. & Schlieder, R. A. (1992). Age, growth, and reproduction of gag, Mycteroperca microlepis (Pisces:Serranidae), in the eastern Gulf of Mexico. Bulletin of Marine Science 51, 337-352.

Hopkins, T.E. & Larson, R. J. (1990). Gastric evacuation of three food types in the black and yellow rockfish Sebastes chrysomelas (Jordan and Gilbert). Journal of Fish Biology 36, 673-681.

Innes, S., Lavigne, D. M., Earle, W. M., & Kovacs, K. M. (1987). Feeding rates of seals and whales. Journal of Animal Ecology 56, 115-130.

Jackson, S., Duffy, D. C., & Jenkins, J. F. G. (1987). Gastric digestion in marine vertebrate predators: in vitro standards. Functional Ecology 1, 287-291.

Jackson, S., Place, A. R., & Seiderer, L. J. (1992). Chitin digestion and assimilation by seabirds. The Auk 109, 758-770.

Jobling, M. (1980a). Gastric evacuation in plaice, Pleuronectes platessa L.: effects of dietary energy level and food composition. Journal of Fish Biology 17, 187-196.

Jobling, M. (1980b). Gastric evacuation in plaice, Pleuronectes platessa L.: effects of temperature and fish size. Journal of Fish Biology 17, 547-551.

Jobling, M. (1981). Mathematical models of gastric emptying and the estimation of daily rates of food consumption for fish. Journal of Fish Biology 19, 245-257.

Jobling, M. (1987). Influences of food particle size and dietary energy content on patterns of gastric evacuation in fish: test of a physiological model of gastric emptying. Journal of Fish Biology 30, 299-314.

Jobling, M. (1993). Bioenergetics: feed intake and energy partitioning. In Fish Ecophysiology (Rankin, J. C. & Jensen, F. B, eds), pp. 1-44. London, England: Chapman & Hall.

Jobling , M., Gwyther, D., & Grove, D. J. (1977). Some effects of temperature, meal size and body weight on gastric evacuation time in the dab Limanda limanda (L). Journal of Fish Biology 10, 291-298.

Jobling, M. & Davies, P. S. (1979). Gastric evacuation in plaice, Pleuronestes platessa L.: effects of temperature and meal size. Journal of Fish Biology 14, 539-546.

Johansen, G. O., Bogstad, B., Mehl, S., & Ulltang, O. (2004). Consumption of juvenile herring (Clupea harengula) by cod (Gadus morhua) in the Barents Sea: a new approach to estimating consumption in piscivorous fish. Canadian Journal of Fisheries and Aquatic Science 61, 343-359.

Page 108: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

96

Jones, R. (1974). The rate of elimination of food from the stomachs of haddock Melanogrammus aeglefinus, cod Gadus morhua, and whiting Merlangius merlangus. Journal du Conseil international pour l’Exploration de la Mer 35, 225-243.

Kao, J. S. (2000). Diet, daily ration and gastric evacuation of the leopard shark (Triakis semifasciata). MS Thesis, Moss Landing Marine Laboratories, California State University Hayward.

Kimura, D. K. (1980). Likelihood methods for the von Bertalanffy growth curve. Fishery Bulletin 77, 765-776.

Klekowski, R. Z. & Duncan, A. (1975). Feeding and nutrition. In Methods for Ecological Bioenergetics (Grodzinski, W., Klekowski, R. Z., & Duncan, A., eds.), pp. 227-257. Oxford, England: Blackwell Scientific Publications.

Knutsen, I. & Salvanes, A. G. (1999). Temperature-dependent digestion handling time in juvenile cod and possible consequences for prey choice. Marine Ecological Progress Series 181, 61-79.

Koed, A. (2001). The effects of meal size, body size and temperature on gastric evacuation in pikeperch. Journal of Fish Biology 58, 281-290.

Kohn, P., Winsler, R. J., & Hoffmann, R. C. (1962). Metabolism of D-glucosamine and N-acetyl-D-glucosamine in the intact rat. Journal of Biological Chemistry 237, 304-308.

Krebs, J. R., & Kacelnik, A. (1991). Decision-making. In Behavioural Ecology: An Evolutionary Approach (Krebs, J. R., & Davies, N. B., eds.), pp.105-136. Oxford, England: Blackwell Scientific Publications.

Lindberg, W., Mason, D, & Murie, D. (2002). Habitat-mediated predator-prey interactions: implications for sustainable production of gag in the eastern Gulf of Mexico. Final Project Report, Florida Sea Grant Program R/LR-B-49.

Lindsay, G. J. H. (1984). Distribution and function of digestive tract chitinolytic enzymes in fish. Journal of Fish Biology 24, 529-536.

Lindsay, G. J. H., Walton, M. J., Adron, J. W., Fletcher, T., C., Cho, C. Y., & Cowey, C. B. (1984). The growth of rainbow trout (Salmo gairdneri) given diets containing chitin, and its relationship to chitinolytic enzymes and chitin digestibility. Aquaculture 37, 315-334.

Lindsay, G. J. H. & Gooday, G. W. (1985). Chitinolytic enzymes and the bacterial microflora in the digestive tract of cod, Gadus morhua. Journal of Fish Biology 26, 255-266.

Page 109: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

97

Lovell, T. (1998). Nutrition and Feeding of Fish. Boston, Massachusetts: Kluwer Academic Publishers.

MacArthur, R. H., & Pianka, E. R. (1966). On optimal use of a patchy environment. American Naturalist 100, 603-609.

MacDonald, J. S., Waiwood, K. G., & Green, R. H. (1982). Rates of digestion of different prey in Atlantic cod (Gadus morhua), ocean pout (Macrozoarces americanus), winter flounder (Pseudopleuronectes americanus) and American plaice (Hippoglossoides platessa). Canadian Journal of Fish and Aquatic Science 39, 651-659.

Medved, R. J. (1985). Gastric evacuation in the sandbar shark, Charcharhinus plumbes. Journal of Fish Biology 26, 239-253.

Minton, J. W., McLean, R. B., & Singley, P. T. (1981). Bioenergetics of suger in Watts Bar Reservoir. Oak Ridge National Laboratory, Oak Ridge, Tennessee, USA ORNL/TM-7630.

Naughton, S. P. & Saloman, C. H. (1995). Food of gag (Mycteroperca microlepis) from North Carolina and three areas of Florida. NOAA Technical Memorandum NMFS-SEFC-160.

Nelson, G. A. & Ross, M. R. (1995). Gastric evacuation in little skate. Journal of Fish Biology 46, 977-986.

Olson, R. J. & Boggs, C. H. (1986). Apex predation by yellowfin tuna (Thunnus albacareas): independent estimates from gastric evacuation and stomach contents, bioenergetics, and cesium concentrations. Canadian Journal of Fisheries and Aquatic Science 43, 1760-1775.

Olson, R. J. & Mullen, A. J. (1986). Recent developments for making gastric evacuation and daily ration determinations. Environmental Biology of Fishes 16, 183-191.

Pandian, T. J. (1967). Intake, digestion, absorption and conversion of food in the fishes Megalops cyprinoids and Ophiocephalus striatus. Marine Biology 1, 16-32.

Persson, L. (1979). The effects of temperature and different food organisms on the rate of gastric evacuation in perch (Perca fluviatilis). Freshwater Biology 9, 99-104.

Persson, L. (1984). Food evacuation and models for multiple meals in fishes. Environmental Biology of Fishes 10, 305-309.

Phlips, E. J. & Bledsoe, E. L. (2002). The consequences of Suwannee River eutrophication for the dynamics of algae in the river and associated estuary. Report submitted to the Suwannee River Water Management District. Live Oak, Florida.

Page 110: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

98

Polovina, J. J. & Ralston, S. (1987). Tropical Snappers and Groupers: Biology and Fisheries Management. Boulder, Colorado: Westview Press.

Rao, P. V. (1998). Statistical Methods in the Life Sciences. Pacific Grove, California: Doxbury.

Rindorf, A. (2002). The effect of stomach fullness on food intake of whiting in the North Sea. Journal of Fish Biology 61, 579-593.

Ross, S. W. & Moser, M. L. (1995). Life history of juvenile gag, Mycteroperca microlepis, in North Carolina estuaries. Bulletin of Marine Science 56, 222-237.

Roxburgh, L., & Pinshow, B. (2002). Digestion of nectar and insects by Palestine sunbirds. Physiological and Biochemical Zoology 75, 583-589.

Ruggerone, G. T. (1989). Gastric evacuation rates and daily ration of piscivorous coho salmon, Oncorhynchus kisutch Walbaum. Journal of Fish Biology 34, 451-463.

SAS Institute Inc. (1999). SAS Version 8.1. Cary, North Carolina: SAS Institute Inc.

Schoener, T. W. (1971). Theory of feeding strategies. Annual Review of Ecology and Systematics 2, 369-404.

Sih, A. & Christensen, B. (2001). Optimal diet theory: when does it work, and when and why does it fail? Animal Behavior 61, 379-390.

Singh-Renton, A. & Bromley, P. J. (1996). Effects of temperature, prey type and prey size on gastric evacuation in small cod and whiting. Journal of Fish Biology 49, 702-713.

Smith, C. L. (1971). A revision of the American groupers: Epinephelus and allied genera. Bulletin of the American Museum of Natural History 146, 67-241.

Sponheimer, M., Robinson, T., & Roeder, B. (2003). Digestion and passage rates of grass hays by llamas, alpacas, goats, rabbits, and horses. Small Ruminant Research 48, 149-154.

Stevens, C. E. & Hume, I. D. (1995). Comparative Physiology of the Vertebrate Digestive System, 2nd edn. Cambridge, England: Cambridge University Press.

Stickney, R. R. & Kohler, C. C. (1990). Maintaining fishes for research and teaching. In Methods for Fish Biology (Schreck, C. B. & Moyle, P. B., eds), pp. 633-663. Bethesda, Maryland: American Fisheries Society.

Swenson, W. A. & Smith, L. L. (1973). Gastric digestion, food conversion, feeding periodicity, and food conversion efficiency in walleye (Stizostedion vitreum). Journal of the Fisheries Research Board of Canada 30, 1327-1336.

Page 111: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

99

Temming, A. & Andersen, N. G. (1994). Modeling gastric evacuation without meal size as a variable. A model applicable for the estimation of daily ration of cod (Gadus morhua L.) in the field. ICES Journal of Marine Science 51, 429-438.

Temming, A. & Herrmann, J. P. (2001). Gastric evacuation in horse mackerel. I. The effects of meal size, temperature and predator weight. Journal of Fish Biology 58, 1230-1245.

Turner, S.C., Porch, C. E., Heinemann, D., Scott, G. P. & Ortiz, M. (2001). Status of gag in the Gulf of Mexico, Assessment 3.0. Online. Http://www.sefsc.noaa.gov/PDFdocs/GulfMexicoGag Assessment86_2000.pdf. Last accessed, April 11, 2005.

Tyler A. V. (1970) Rate of gastric emptying in young cod. Journal of the Fisheries Research Board of Canada 27, 1177-1189.

Vollhardt, K. P. C. & Schore, N. E. (1987). Organic Chemistry, 2nd edn. New York: W. H. Freeman and Company.

Warren, C. E. & Davis, G. E. (1967). Laboratory studies on the feeding, bioenergetics, and growth of fish. In The Biological Basis of Freshwater Fish Production (Gerking, S. D., ed), pp. 175-214. Oxford, England: Blackwell Scientific Publications.

Weaver, D. C. (1996). Feeding ecology and ecomorphology of three sea basses (Pisces:Serranidae) in the northeastern Gulf of Mexico. MS Thesis, Department of Zoology, University of Florida.

Winberg, G. G. (1956). Rate of metabolism and food requirements of fishes. Fisheries Research Board of Canada. Ser. No. 194.

Windell, J. T. (1966). Rate of digestion in the bluegill sunfish. Investigations of Indiana Lakes and Streams 7, 185-214.

Windell, J. T. (1978). Estimating food consumption rates of fish populations. In Methods for Assessment of Fish Production in Fresh Waters (Bagenal, T., ed.), pp. 227-254. Oxford, England: Blackwell Scientific Publications.

Winship, A. J., Trites, A. W., & Rosen, D. A. S. (2002). A bioenergetic model for estimating the food requirements of Stellar sea lions Eumetopias jubatus in Alaska, USA. Marine Ecology Progress Series 229, 291-312.

Zar, J.H. (1999). Biostatistical Analysis. Upper Saddle River, New Jersey: Prentice-Hall, Inc.

Page 112: GASTRIC EVACUATION AND DIGESTION STATE …nsgl.gso.uri.edu/flsgp/flsgpx05008.pdf ·  · 2007-03-07GASTRIC EVACUATION AND DIGESTION STATE INDICES FOR GAG Mycteroperca ... of the stomach

BIOGRAPHICAL SKETCH

Born on April 15, 1975, in Grand Rapids, Michigan, the author showed a great

curiosity for various aquatic organisms and their environments at an early age. Her

curiosity developed into an interest in fisheries science at Grand Valley State University

in Allendale, Michigan, and resulted in a Bachelor of Science in biology with an aquatic

emphasis and minors in Russian studies and studio art.

After working as a Lake Management Assistant on Kiawah Island, South Carolina,

the author decided to pursue a graduate-level education to further her interest in marine

fisheries. While pursuing a Master of Science in biology at the University of Florida, she

served as Treasurer for Students United in the Research of Fisheries (SURF). During

September of 2004, she began work at Mote Marine Laboratory in Sarasota, Florida, as a

staff biologist with the Sarasota Dolphin Research Program. She plans to continue work

on commercially important marine species, species of special concern, predator-prey

relationships, and the various human influences affecting these relationships.

99