<%BANNER%>

Active Adaptive Management for Native Fish Conservation in the Grand Canyon

University of Florida Institutional Repository
Permanent Link: http://ufdc.ufl.edu/UFE0021996/00001

Material Information

Title: Active Adaptive Management for Native Fish Conservation in the Grand Canyon Implementation and Evaluation
Physical Description: 1 online resource (173 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: adaptive, assessment, bioenergetics, canyon, chub, control, grand, humpback, management, mark, nonnative, rainbow, recapture, stock, trout
Fisheries and Aquatic Sciences -- Dissertations, Academic -- UF
Genre: Fisheries and Aquatic Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: My first objective was to evaluate the efficacy of a large scale non-native fish removal effort to benefit endemic fishes of the Colorado River within Grand Canyon. During 2003-2006, over 23,000 non-native fish, primarily rainbow trout Oncorhynchus mykiss, were removed from a 9.4 mile reach of the Colorado River. These removals resulted in a rapid shift in fish community composition from one dominated by cold water salmonids ( > 90%), to one dominated by native fishes and the non-native fathead minnow Pimephales promelas ( > 90%). Concurrent with the mechanical removal, data collected within a control reach of the river suggested a systemic decline in rainbow trout unrelated to the fish removal effort. Thus, the efficacy of the mechanical removal was aided by an external systemic decline, particularly in 2005-2006. My second objective was to improve current knowledge of humpback chub Gila cypha growth to aid in length-based age determination, and to provide a tool to evaluate temperature-dependent changes in growth rate. I estimated a temperature-dependent growth function for humpback chub by predicting more than 14,000 growth increments from a mark-recapture database. Results suggest that humpback chub growth is strongly dependent on temperature and that previous growth curves based on paired age-length data tend to over-estimate the age of small fish and under-estimate the age of large fish. My third objective was to update humpback chub stock assessment procedures following guidance from an external review panel. These recommendations were primarily to develop model selection procedures and to evaluate the effect of error in length-based age determination. I used both Pearson residual analysis and Akaike Information Criterion to evaluate candidate models ? leading to the conclusion that the most general assessment model was required to adequately model patterns in capture probability. I used the temperature-dependent growth relationship to estimate probabilistic relationships between age and length. These age-length relationships were then used in Monte Carlo simulations to capture the effect of ageing error on subsequent estimates of recruitment and adult abundance. The results indicate that the adult humpback chub population has likely increased between 20-25% since 2001. My fourth objective was to evaluate whether there was any evidence of effect from past adaptive management actions or uncontrollable factors on Grand Canyon fish populations, and to make recommendations for further adaptive management program development. These results are largely inconclusive except that the combined policy of mechanical removal and increased water temperatures is temporally correlated with increased native fish abundance in the mainstem Colorado River near the confluence of the Little Colorado River, a reach deemed critical habitat for humpback chub. I recommend that the adaptive management program invest additional effort in developing more explicit and measurable resource goals, particularly for focal Colorado River resources. I further recommend that additional investment in monitoring of juvenile native fish survival and growth in the mainstem is needed to adequately evaluate future adaptive management experiments. Finally, additional predictive capability is needed to both formalize a priori hypotheses about juvenile native fish survival and recruitment, and to screen future policy options.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Pine, William.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021996:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021996/00001

Material Information

Title: Active Adaptive Management for Native Fish Conservation in the Grand Canyon Implementation and Evaluation
Physical Description: 1 online resource (173 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: adaptive, assessment, bioenergetics, canyon, chub, control, grand, humpback, management, mark, nonnative, rainbow, recapture, stock, trout
Fisheries and Aquatic Sciences -- Dissertations, Academic -- UF
Genre: Fisheries and Aquatic Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: My first objective was to evaluate the efficacy of a large scale non-native fish removal effort to benefit endemic fishes of the Colorado River within Grand Canyon. During 2003-2006, over 23,000 non-native fish, primarily rainbow trout Oncorhynchus mykiss, were removed from a 9.4 mile reach of the Colorado River. These removals resulted in a rapid shift in fish community composition from one dominated by cold water salmonids ( > 90%), to one dominated by native fishes and the non-native fathead minnow Pimephales promelas ( > 90%). Concurrent with the mechanical removal, data collected within a control reach of the river suggested a systemic decline in rainbow trout unrelated to the fish removal effort. Thus, the efficacy of the mechanical removal was aided by an external systemic decline, particularly in 2005-2006. My second objective was to improve current knowledge of humpback chub Gila cypha growth to aid in length-based age determination, and to provide a tool to evaluate temperature-dependent changes in growth rate. I estimated a temperature-dependent growth function for humpback chub by predicting more than 14,000 growth increments from a mark-recapture database. Results suggest that humpback chub growth is strongly dependent on temperature and that previous growth curves based on paired age-length data tend to over-estimate the age of small fish and under-estimate the age of large fish. My third objective was to update humpback chub stock assessment procedures following guidance from an external review panel. These recommendations were primarily to develop model selection procedures and to evaluate the effect of error in length-based age determination. I used both Pearson residual analysis and Akaike Information Criterion to evaluate candidate models ? leading to the conclusion that the most general assessment model was required to adequately model patterns in capture probability. I used the temperature-dependent growth relationship to estimate probabilistic relationships between age and length. These age-length relationships were then used in Monte Carlo simulations to capture the effect of ageing error on subsequent estimates of recruitment and adult abundance. The results indicate that the adult humpback chub population has likely increased between 20-25% since 2001. My fourth objective was to evaluate whether there was any evidence of effect from past adaptive management actions or uncontrollable factors on Grand Canyon fish populations, and to make recommendations for further adaptive management program development. These results are largely inconclusive except that the combined policy of mechanical removal and increased water temperatures is temporally correlated with increased native fish abundance in the mainstem Colorado River near the confluence of the Little Colorado River, a reach deemed critical habitat for humpback chub. I recommend that the adaptive management program invest additional effort in developing more explicit and measurable resource goals, particularly for focal Colorado River resources. I further recommend that additional investment in monitoring of juvenile native fish survival and growth in the mainstem is needed to adequately evaluate future adaptive management experiments. Finally, additional predictive capability is needed to both formalize a priori hypotheses about juvenile native fish survival and recruitment, and to screen future policy options.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Pine, William.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021996:00001


This item has the following downloads:


Full Text





ACTIVE ADAPTIVE MANAGEMENT FOR NATIVE FISH CONSERVATION IN
THE GRAND CANYON: IMPLEMENTATION AND EVALUATION





















By

LEWIS GEORGE COGGINS, JR.


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

UNIVERSITY OF FLORIDA

2008



































2008 Lewis G. Coggins, Jr.




































For Dad









ACKNOWLEDGMENTS

This research was funded by the U.S. Geological Survey through the Grand Canyon

Monitoring and Research Center. I thank Dr. Ted Melis, Dr Barbara Ralston, and Dr. Denny

Fenn for supporting the agreement that allowed me to return to school. I also thank John Hamill

and Matthew Andersen for their continued support. I thank my friend and colleague Dr. Mike

Yard for his collaboration in the design and implementation of the mechanical removal project. I

thank Clay Nelson and his staff with the Arizona Game and Fish Department for data collection

during the 2005-2006 mechanical removal efforts. I also thank Carol Fritzinger, Brian Dierker,

Peter Weiss, Brent Berger, Steve Jones, Stewart Reeder, Danny Martinez, Yael Bernstein,

Courtney Giaque, Emily Thompson, Dave Doring, Dave Baker, Melanie Caron, Scotty Davis,

Ted Kennedy, Ally Martinez, Scott Perry, Lynn Rhoder, Park Stefensen, John Taylor, Todd and

Erica Tietjen, and Josh Winiecki for outstanding field and logistical support. I also would like to

thank field biologists and technicians from ASU, Bio West, USFWS, AGFD, and SWCA for

their efforts to collect data on Grand Canyon fishes for the past two decades.

I thank my advisor, Dr. William Pine, and the rest of my supervisory committee (Dr. Carl

Walters, Dr. Micheal Allen, Dr. Tom Frazer, and Dr. Christina Staudhammer) for their service

and assistance with my research. I especially thank Dr. Pine for encouraging me to return to

school and for his assistance and guidance in making it a rewarding and fruitful experience. I

also owe a tremendous debt to my mentor and trusted friend, Dr. Carl Walters, who has assisted

and guided much of my research activity for the last 8 years. I thank Dr. Steve Martell for his

assistance in evaluating humpback chub stock assessment models.

Finally, I thank my wife Jennifer and our daughters Ellie and Annie for their love, support,

and understanding of the long hours we were forced to endure apart.









TABLE OF CONTENTS

page

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

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

LIST O F FIG U RE S ........................................

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

CHAPTER

1 G EN ER A L IN TR O D U C TIO N ........................................................................................ 14

2 NON-NATIVE FISH CONTROL IN THE COLORADO RIVER IN GRAND
CANYON, ARIZONA: AN EFFECTIVE PROGRAM OR SERENDIPITOUS
T IM IN G ? ...........................................................................16

Adaptive M management in Grand Canyon .................. ........................................... 18
Fish Com m unity Background................................................. 19
O bj ectiv e ................................................................................ 2 0
M e th o d s ..................................................................................................................................2 1
Mechanical Removal Reach: Study Areas and Field Protocols ...................................21
Control Reach: Study Areas and Field Protocols .....................................................23
Mechanical Removal Reach: Data Analysis .................................. ...............23
C control R each : D ata A naly sis ................................................................................... 27
R results ................... .. ........ .............................9
M echanical Removal Reach ................................. .......................... .. ....... 29
C control R each ........................................ .......... ...................31
Comparison of Mechanical Removal and Control Reaches ............... ............ 32
D iscu ssion ........................................ ................................................................................. 33
M mechanical Rem oval: Effective Program ? ........................... ........................................... 33
Serendipitous Timing: What Led to the Decline of Rainbow Trout in the Control
R e a c h ? ...........................................................................................3 4
Other Species......... ........................................................... ... .. ....... 36
Bias in Capture Probability and Abundance Estimates .........................................36
Recommendations for Future Mechanical Removal Operations ...............................37
Adaptive Management in Grand Canyon: The Future .............................................. 38

3 DEVELOPMENT OF A TEMPERATURE-DEPENDENT GROWTH MODEL FOR
THE ENDANGERED HUMPBACK CHUB USING MARK-RECAPTURE DATA .........57

M methods ................................................................... ........ .......... 59
Results ................... ......................................... 67
Discussion ................... .......................... ..................70









4 ABUNDANCE TRENDS AND STATUS OF THE LITTLE COLORADO RIVER
POPULATION OF HUMPBACK CHUB: AN UPDATE CONSIDERING DATA
1 9 8 9 -2 0 0 6 ................... ........................................................... ................ 8 3

M methods .................... ................. .................................... .......... ..... 86
Index -B ased M etrics........... ..... ......................................................................... .. .... 86
T agging-B asked M etrics ......................................................................... ....................87
Evaluating M odel Fit ................... .......... ...................... .. ........ .............. ... 90
Incorporation of Ageing Error in ASMR Assessments................................................92
R esu lts ................. ....... .............. ............................. ............................ 9 5
Index-B asked A ssessm ents ....................................................................... ..................95
T agging-B asked A ssessm ents ........................................ ............................................96
C closed P population M odels...................................................................... ...................97
ASM R W without Tag Cohort Specific Data.................................... ....... ............... 97
M odel Evaluation and Selection.............. ............................................. ............... 98
ASMR with Tag Cohort Specific Data................................... ...............100
M odel Evaluation and Selection...................................... ...........................100
Incorporation of Ageing Error in ASMR Assessments ..........................................100
R results Sum m ary ......... .. ...... .... .. ...... .................................................... .. 102
D iscu ssion ................. ..................................... ............................102

5 LINKING TEMPORAL PATTERNS IN FISHERY RESOURCES WITH ADAPTIVE
MANAGEMENT: WHAT HAVE WE LEARNED AND ARE WE MANAGING
A D A P T IV E L Y ? ................................................................................................... .... 13 4

Description of Adaptive M management Actions ........................ ................. ...................136
1996 Experimental High Flow ........... .. ........ ...................... 136
2000 Low Summer Steady Flow ....................................... 136
2004 Experimental High Flow ..................................... ...................................... 137
2003-2005 Non-native Fish Suppression Flows.................................137
2003-2006 Mechanical Removal of Non-native Fish .............................................138
Description of Uncontrolled Factors ................................................ ...................... 138
Paria, Little Colorado, and Colorado River Hydrology .............................................139
Release Water Temperature from Glen Canyon Dam....................................140
Juvenile Native Fish Production in the Little Colorado River .................................... 140
How Are Fish Populations Affected by Adaptive Management Actions and
U controlled F actors? .................. ........... ..... ..... .......... .......... ............... 142
Has Increased Turbidity Affected Fish Populations in Grand Canyon? .....................142
Has Reduced Non-native Fish Abundance and Increased Temperature Affected
Native Fish Populations in Grand Canyon?...........................................................143
Conclusions and Recom m endations ......................................................... ............... 145

L IST O F R E F E R E N C E S ..................................................................................... ..................159

B IO G R A PH IC A L SK E T C H ......................................................................... ... ..................... 173









LIST OF TABLES


Table page

2-1 Electrofishing catch by species in the mechanical removal reach for each month,
2 0 0 3 -2 0 0 6 ............................................................................. 4 1

2-2 Estimated abundance and density of rainbow trout in the mechanical removal reach
at the beginning of each m month, 2003-2006................................... ........................ 42

2-3 Electrofishing catch by species in the control reach for each month, 2003-2006. ...........43

2-4 Estimated abundance and density of rainbow trout in the control reach at the
beginning of each m onth, 2003-2006. ............................................................................44

3-1 General growth m odel results. ...... ........................... .......................................... 74

3-2 Parameter correlation matrix for the temperature-independent growth model ..................75

3-3 Parameter correlation matrix for the temperature-dependent growth model .....................76

4-1 AIC model evaluation results among ASMR models fit to data pooled among tag
cohort ...........................................................................................108

4-2 AIC model evaluation results among ASMR models fit to data stratified by tag
cohort ...........................................................................................109









LIST OF FIGURES


Figure page

2-1 Map of the mechanical removal reach of the Colorado River within Grand Canyon,
A riz o n a ................... ............................... .......................... ................ 4 5

2-2 Map of the control reach of the Colorado River within Grand Canyon, Arizona. ............46

2-3 Percent composition (A) and number of fish (B) by species captured with
electrofishing in the mechanical removal reach among months, 2003-2006...................47

2-4 Estimated catch rate (A), abundance (B), and capture probability (C) for rainbow
trout in both the upstream and downstream strata in the mechanical removal reach
am ong m months, 2003-2006............................................ ................... ............... 48

2-5 Net immigration rate into the upstream (A) and downstream (B) strata of the
mechanical removal reach within time intervals between January 2003 and August
2 006 .............. ....................... ................................................ ...... 4 9

2-6 Probability plots for coefficient values influencing capture probability among 10
trips (January 2003-July 2003; January 2004-September 2004). ................ ..............50

2-7 Estimated catch rate (A), abundance (B), and capture probability (C) for rainbow
trout in the control reach among months, 2003-2006 ........................................... ........... 51

2-8 Estimated monthly survival rate of rainbow trout in the control reach during 2003-
2 006 .............. ....................... ................................................ ...... 52

2-9 Estimated rainbow trout abundance in both the mechanical removal and control
reaches at the beginning of each trip during 2003-2006.................................................53

2-10 Estimated total length and 95% confidence intervals of rainbow trout captured in
both the mechanical removal and control reaches during 2003-2006. ...........................54

2-11 Length frequency distributions of rainbow trout captured using electrofishing in the
Colorado River from river mile -15 to river mile 56. .............................. ......... ...... .55

2-12 Daily mean water temperatures observed in the Colorado River at approximately
river m ile 6 1, 1990-2006........ ................................................ ................ .. .... ...... 56

3-1 Predicted and observed growth rate (dL/dt) as a function of total length (TL) at the
start of the tim e interval ................................................... .. ............................... 77

3-2 Observed and predicted monthly Little Colorado River water temperature......................78









3-3 Observed and predicted humpback chub growth rate (dL/dt) from the temperature-
independent growth model and the temperature-dependent growth model during
sum m er and w inter ............... ........................... ............................. 79

3-4 Observed and predicted humpback chub growth rate (dL/dt) from the temperature-
dependent growth model during summer and winter. ................................................. 80

3-5 Predicted humpback chub length-at-age from the U.S. Fish and Wildlife Service
(USFWS) growth curve, the temperature-independent growth model, the
temperature-dependent growth model for the Little Colorado River (LCR) humpback
chub population, and the temperature-dependent growth model for humpback chub
living in the mainstem Colorado River under a constant temperature of 10C. ................81

3-6 Predicted monthly growth rate from the temperature-dependent growth model for the
Little Colorado River (LCR) population of humpback chub and for humpback chub
living in the mainstem Colorado River under a constant temperature of 10C. ................82

4-1 Relative abundance indices of sub-adult (150-199 mm total length; TL) and adult
(>200 mm TL) humpback chub based on hoop net catch rate (fish/hour) in the lower
1,200 m section of the Little Colorado River (A) and trammel net catch rate
(fish/hour/100 m) of adult humpback chub in the Little Colorado River inflow reach
of the C olorado R iver (B ). ............................................................................... ..... 110

4-2 Numbers of humpback chub marked (A) and recaptured (B) by age and year ..............111

4-3a Numbers offish marked by age in years 1989 (A), 1990 (B), 1991 (C), and 1992 (D)
indicated by dark circles and subsequently recaptured (light circles) by age and
y ears ................... ........................................................................... 1 12

4-3b Numbers offish marked by age in years 1993 (A), 1994 (B), 1995 (C), and 1996 (D)
indicated by dark circles and subsequently recaptured (light circles) by age and
y ears ................... ........................................................................... 1 13

4-3c Numbers offish marked by age in years 1997 (A), 1998 (B), 1999 (C), and 2000 (D)
indicated by dark circles and subsequently recaptured (light circles) by age and
y ears ................... ........................................................................... 1 14

4-3d Numbers of fish marked by age in years 2001 (A), 2002 (B), 2003 (C), and 2004 (D)
indicated by dark circles and subsequently recaptured (light circles) by age and
y ears ................... ........................................................................... 1 15

4-3e Numbers offish marked by age in years 2005 (A) and 2006 (B) indicated by dark
circles and subsequently recaptured (light circles) by age and years. ...........................116

4-4 Mark-recapture closed population model estimates of humpback chub abundance >
150 mm total length in the Little Colorado River ...................................................... 117









4-5 Humpback chub adult abundance (age-4+) estimates from the ASMR 1 (A), ASMR
2 (B), and ASMR 3 (C) models using data pooled among tag cohorts. ..........................118

4-6 Humpback chub recruit abundance (age-2) estimates from the ASMR 1 (A), ASMR
2 (B), and ASMR 3 (C) models using data pooled among tag cohorts. ..........................119

4-7 Pearson residual plots for model ASMR 1 using data pooled among tag cohorts...........20

4-8 Pearson residual plots for model ASMR 2 using data pooled among tag cohorts...........21

4-9 Pearson residual plots for model ASMR 3 using data pooled among tag cohorts...........22

4-10 Capture probability by age and year estimated from model ASMR 3 using data
pooled am ong tag cohorts. ..... ........................... ........................................... 123

4-11 Humpback chub adult abundance (age-4+) estimates from the ASMR 1 (A), ASMR
2 (B), and ASMR 3 (C) models using data stratified by tag cohort ..............................124

4-12 Humpback chub recruit abundance (age-2) estimates from the ASMR 1 (A), ASMR
2 (B), and ASMR 3 (C) models using data stratified by tag cohort ..............................125

4-13 Pearson residual plots for model ASMR 1 using data stratified by tag cohort...............126

4-14 Pearson residual plots for model ASMR 2 using data stratified by tag cohort............127

4-15 Pearson residual plots for model ASMR 3 using data stratified by tag cohort............128

4-16 Capture probability by age and year estimated from model ASMR 3 using data
stratified by tag cohort. ................................................ .... .................. 129

4-17 Seasonal probability surfaces of age for a particular length bin................ ................. 130

4-18 Estimated adult abundance (age-4+) from ASMR 3 incorporating uncertainty in
assig n m en t o f ag e ...................................... ............. ............... ................ 13 1

4-19 Estimated recruit abundance (age-1) from ASMR 3 incorporating uncertainty in
assign ent of age .................... .. ....... ................. .... .......... ........... 132

4-20 Retrospective analysis of adult abundance (A) and mortality rate (B) considering
datasets beginning in 1989 and ending in the year indicated in the figure legend..........133

5-1 Discharge in ft3/s (cfs) for the Colorado River at Lees Ferry (A), the Paria River at
Lees Ferry (B), and the Little Colorado River at Cameron, AZ (C), 1990-2006. ...........150

5-2 Daily mean water temperatures observed in the Colorado River at approximately
river m ile 6 1, 1990-2006........... ........................................................ .. .... .... ... ..151









5-3 Monthly electrofishing catch rate (fish/hour) in the Colorado River between river
mile (RM) 56.3 and RM 65.7 for rainbow trout (A), brown trout (B), common carp
(C), and fathead m innow (D ). ............................................... .............................. 152

5-4 Monthly electrofishing catch rate (fish/hour) in the Colorado River between river
mile (RM) 56.3 and RM 65.7 for humpback chub (A), flannelmouth sucker (B),
bluehead sucker (C), and speckled dace (D).................................................... 153

5-5 Monthly hoop net catch rate (fish/hour) in the Colorado River between river mile
(RM) 63.7 and RM 64.2 for humpback chub (A), flannelmouth sucker (B), bluehead
sucker (C), and speckled dace (D). ............................................................................154

5-6 Monthly average total length (TL; mm) observed in electrofishing sampling in the
Colorado River between river mile (RM) 56.3 and RM 65.7 for rainbow trout (A),
brown trout (B), common carp (C), and fathead minnow (D) .....................................155

5-7 Monthly average total length (TL; mm) observed in electrofishing sampling in the
Colorado River between river mile (RM) 56.3 and RM 65.7 for humpback chub (A),
flannelmouth sucker (B), bluehead sucker (C), and speckled dace (D)...........................156

5-8 Monthly average total length (TL; mm) observed in hoop net sampling in the
Colorado River between river mile (RM) 63.7 and RM 64.2 for humpback chub (A),
flannelmouth sucker (B), bluehead sucker (C), and speckled dace (D)...........................157

5-9 Smoothed kernel density plot of the total length of flannelmouth sucker captured
with electrofishing in the mechanical removal reach during 2003 and 2006 ...............158









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

ACTIVE ADAPTIVE MANAGEMENT FOR NATIVE FISH CONSERVATION IN
THE GRAND CANYON: IMPLEMENTATION AND EVALUATION

By

Lewis George Coggins, Jr.

May 2008

Chair: William E. Pine, III
Major: Fisheries and Aquatic Sciences

My first objective was to evaluate the efficacy of a large scale non-native fish removal

effort to benefit endemic fishes of the Colorado River within Grand Canyon. During 2003-2006,

over 23,000 non-native fish, primarily rainbow trout Oncorhynchus mykiss, were removed from

a 9.4 mile reach of the Colorado River. These removals resulted in a rapid shift in fish

community composition from one dominated by cold water salmonids (>90%), to one dominated

by native fishes and the non-native fathead minnow Pimephalespromelas (>90%). Concurrent

with the mechanical removal, data collected within a control reach of the river suggested a

systemic decline in rainbow trout unrelated to the fish removal effort. Thus, the efficacy of the

mechanical removal was aided by an external systemic decline, particularly in 2005-2006.

My second objective was to improve current knowledge of humpback chub Gila cypha

growth to aid in length-based age determination, and to provide a tool to evaluate temperature-

dependent changes in growth rate. I estimated a temperature-dependent growth function for

humpback chub by predicting more than 14,000 growth increments from a mark-recapture

database. Results suggest that humpback chub growth is strongly dependent on temperature and

that previous growth curves based on paired age-length data tend to over-estimate the age of

small fish and under-estimate the age of large fish.









My third objective was to update humpback chub stock assessment procedures following

guidance from an external review panel. These recommendations were primarily to develop

model selection procedures and to evaluate the effect of error in length-based age determination.

I used both Pearson residual analysis and Akaike Information Criterion to evaluate candidate

models leading to the conclusion that the most general assessment model was required to

adequately model patterns in capture probability. I used the temperature-dependent growth

relationship to estimate probabilistic relationships between age and length. These age-length

relationships were then used in Monte Carlo simulations to capture the effect of ageing error on

subsequent estimates of recruitment and adult abundance. The results indicate that the adult

humpback chub population has likely increased between 20-25% since 2001.

My fourth objective was to evaluate whether there was any evidence of effect from past

adaptive management actions or uncontrollable factors on Grand Canyon fish populations, and to

make recommendations for further adaptive management program development. These results

are largely inconclusive except that the combined policy of mechanical removal and increased

water temperatures is temporally correlated with increased native fish abundance in the

mainstem Colorado River near the confluence of the Little Colorado River, a reach deemed

critical habitat for humpback chub. I recommend that the adaptive management program invest

additional effort in developing more explicit and measurable resource goals, particularly for

focal Colorado River resources. I further recommend that additional investment in monitoring of

juvenile native fish survival and growth in the mainstem is needed to adequately evaluate future

adaptive management experiments. Finally, additional predictive capability is needed to both

formalize apriori hypotheses about juvenile native fish survival and recruitment, and to screen

future policy options.









CHAPTER 1
GENERAL INTRODUCTION

Modifications to river ecosystems to serve human interests are a ubiquitous feature of

human occupied landscapes. Postel et al. (1996) estimated that 54% of global annual runoff was

appropriated for human use in 1996, and forecast that this figure might approach 70% by the

year 2025. A recent review by Nilsson et al. (2005) documented that 50% of the Earth's large

river systems are fragmented by dams. Thus, anthropogenic modifications are a major, and

frequently detrimental, influence on riverine ecosystems on a global basis.

Alterations to riverine ecosystems in the U.S. have led to an increase in river restoration

projects and an active dialogue between scientists and policy makers (Poff et al. 2003). This has

led to increased research in both river restoration science and the appropriate measures of river

restoration success (Palmer et al. 2005). However, scientists frequently are unable to predict

with great certainty the outcome of management actions designed to achieve restoration, and this

uncertainty can lead to skepticism and mistrust on the part of policy makers. Given this

uncertainty, adaptive management (Holling 1978, Walters 1986) has been widely advocated as a

strategy to guide restoration programs (Poff et al. 2003). Adaptive management recognizes that

predictions of system response to management actions are uncertain, and seeks to use thoughtful

application of management actions to learn about system behavior and hence, how to achieve

resource management goals.

The 1,470 mile course of the Colorado River begins at high altitude in the Rocky

Mountains and terminates at the northern extent of the Gulf of California. The Colorado River

drainage area encompasses seven U.S. states (AZ, CA, CO, NM, NV, UT, WY) and a small

portion of northwestern Mexico (Benke and Cushing 2005). Described as the "Life-blood" of

the southwestern U.S. (Reisner 1993) and one of the most highly regulated rivers in the world









(Nilsson et al. 2005), this river has enormous social, economic, recreational, and political

importance. Additionally, the Grand Canyon of the Colorado River is widely recognized as one

of the 7 natural wonders of the world and a national treasure of the United States. Following

recognition of degraded conditions in the Grand Canyon reach of the Colorado River

downstream of Glen Canyon Dam (NRC 1987), the Glen Canyon Dam Adaptive Management

Program has attempted to use adaptive management for river restoration since its formation in

1996.

A focal resource of the Glen Canyon Dam Adaptive Management program is the native

fishes endemic to this basin, particularly the federally listed endangered humpback chub Gila

cypha. This dissertation is focused on evaluating the efficacy of the implementation of a specific

adaptive management experiment and developing improved monitoring capability for humpback

chub. In Chapter 2, I describe and evaluate an adaptive management experiment to remove non-

native fish from a large section of the Colorado River heavily used and deemed critical for

humpback chub and other native fishes. In Chapter 3, I used mark-recapture information to

develop a temperature-dependent humpback chub growth model in support of improved

monitoring and stock assessment. In Chapter 4, I used the growth model along with all available

monitoring data through 2006 to provide an updated evaluation of recent humpback chub

population dynamics. In Chapter 5, I provide a synthesis of available fish monitoring data to

evaluate the effect of past adaptive management experiments, as well as to make

recommendations for future adaptive management experimentation, monitoring, and research

priorities.









CHAPTER 2
NON-NATIVE FISH CONTROL IN THE COLORADO RIVER IN GRAND CANYON,
ARIZONA: AN EFFECTIVE PROGRAM OR SERENDIPITOUS TIMING?

Harvest, species introductions, and large-scale habitat alterations have resulted in dramatic

changes in the structure and function of ecosystems on a global basis (Vitousek et al. 1997).

Humans have modified all types of ecosystems through various interventions, yet despite more

than a century of focused ecological research, there remains much uncertainty as to how

ecosystems will respond to anthropogenic interventions (Holling 1973; Walters and Holling

1990). Partly as a result of this uncertainty, efforts to manage human activities using prescriptive

science-based policies to achieve basic goals as they relate to ecosystems (e.g., sustainability or

species conservation) have been widely unsuccessful (Christensen et al. 1996; Mangel et al.

1996). In response to these failures, adaptive environmental assessment and management or

adaptive management (AM) has been proposed as a strategy to link scientific inquiry and natural

resource management (Holling 1978; Walters 1986; Walters and Holling 1990).

Adaptive management assumes that successful management of natural resources can occur

only if objectives are clearly defined and future management policy choices are informed and

directed by past policy performance. Holling and Walters (1990) describe three classes of AM

implementation: (1) "trial and error", (2) passive adaptive, and (3) active adaptive. "Trial and

error" is a structure where initial policy choices are completely uninformed, and later policy

choices are selected from the set of best performing initial choices. A passive adaptive structure

chooses polices based on historic data informing a single predictive model. A passive adaptive

structure differs from trial and error by the use of a predictive model to screen policy choices.

This model can be either conceptual or quantitative and is presumed to be accurate until proven

otherwise. Finally, active adaptive management recognizes that there are usually multiple

predictive models that can explain the historic data equally well, and seeks to implement specific









policies that can both optimize short-term system performance and provide insight into which

model provides the best predictions. The predictive models then become hypotheses of system

behavior under different management policies. If it is possible to quantify the likelihood of each

hypothesis apriori, then policy choices are further selected considering tradeoffs in the future

value of increased understanding of system behavior. Most basically, the concept of AM

embraces three related ideas:

(1) Predictive models of complex systems can never be fully trusted in their ability to
structure management policies that unambiguously attain specific system objectives,

(2) Detailed research into the processes that define system complexity (e.g., resilience,
feedbacks, thresholds, or alternative stable states) can never fully resolve prediction
ambiguity, and

(3) Predictive models contain key uncertainties that may only be resolved (if ever) by
observing the response of the system to particular interventions.

Though AM has been widely adopted as a conceptual strategy for the management of

natural resources (Williams et al. 2007) including: waterfowl (e.g., Nichols et al. 1995), forests

(e.g., Sit and Taylor 1998), wildlife (e.g., Pascual and Hilborn 1995), fisheries (e.g., Sainsbury

1991), large river systems (e.g., NRC 1999), wetlands (e.g., Walters et al. 1992), and others,

critics argue that many management programs that supposedly operate within an adaptive

framework have embraced this term as a "buzz word", but fail to apply the strategy as originally

proposed (Gunderson 1999; Lee 1999). A frequent failure in AM programs has been that

following the rigorous knowledge assessment and modeling that characterizes the initial steps of

the process, the subsequent implementation and monitoring of candidate policies is not

completed (Walters 1997; Gunderson 1999; Ladson and Argent 2002; Schreiber et al. 2004). I

have extensively reviewed the primary literature and found few true empirical tests of AM as a

strategy for management of ecosystems (e.g., Allan and Curtis 2005). This study documents the









implementation of an ecosystem-scale adaptive management experiment in the Colorado River

within Grand Canyon, Arizona.

Adaptive Management in Grand Canyon

Following the Final Record of Decision from the Environmental Impact Statement on the

operation of Glen Canyon Dam (USDOI 1995), the Glen Canyon Dam Adaptive Management

Program was formed and charged with managing the Colorado River ecosystem (CRE) within

Grand Canyon, Arizona. This program consists of a multi-stakeholder federal advisory

committee that defines objectives for the CRE and makes recommendations to the U.S. Secretary

of Interior regarding the operation of Glen Canyon Dam (GCD) and other management actions.

This high-profile program is arguably the most successful example of an adaptive management

program in a large U.S. river (Ladson and Argent 2002). However, this recognition is primarily

the result of short-term experimentation with GCD operations designed to test policies for

sediment conservation (Collier et al. 1997).

Native fish conservation is also a key goal of the Glen Canyon Dam Adaptive

Management Program primarily because many of the species endemic to the Colorado River

Basin are protected under the US Endangered Species Act (ESA). This protected status

necessitates regular review of GCD operations to ensure that dam operations are not deleterious

to Grand Canyon native fish stocks. Current knowledge suggests that likely factors influencing

the population dynamics and ultimate recovery (as defined by the ESA mandated recovery

criteria) of native fish in Grand Canyon include: (1) non-native fish (Gorman et al. 2005; Olden

and Poff 2005), (2) water temperature (Robinson and Childs 2001), (3) flow regulation

(Osmundson et al. 2002), (4) juvenile rearing habitat (Stone and Gorman 2006), and (5) parasites

and disease (Choudury et al. 2004). Of these, previous modeling and data analyses have shown

that factors 1-3 are likely dominant drivers of native fish population dynamics in this system









(Walters et al. 2000), and suggests that improving rearing conditions for native fish in the

mainstem Colorado River will likely provide the most significant benefit to native fish.

Additionally, of the factors possibly influencing native fish population dynamics, controlled

manipulation of factors 1-3 in an experimental framework is most tenable and, in recent years,

has been the focus of efforts in adaptive management for native fish conservation.

Beginning in 2003, the first multi-year program of experimentation specifically designed

to test policies associated with native fish conservation was implemented in Grand Canyon. In

January 2003, an experiment was begun to experimentally manipulate GCD operations and the

abundance of non-native fishes in a 9.4 mile stretch of the Colorado River containing known

critical habitat for humpback chub Gila cypha and other native fish species. Although the

experimental design also called for manipulation of water temperature discharged from GCD in

subsequent years (Coggins et al. 2002), only the experimental fish manipulations were

implemented. The last of four years of manipulating the abundance of non-native fishes was

2006.

Fish Community Background

Over much of the last several decades, the fish community in the Grand Canyon stretch of

the Colorado River has been dominated by the non-native salmonids rainbow trout

Oncorhynchus mykiss and brown trout Salmo trutta (Gloss and Coggins 2005). Introductions of

non-native salmonids have been shown to adversely impact invertebrate (Parker et al. 2001),

amphibian (Knapp and Matthews 2000), and fish (McDowall 2003) communities. These two

species of fish have also been identified as among the top 100 worst invasive species (Lowe et al.

2000) principally because of the global scope of introductions rainbow trout have been

successfully established on every continent with the exception of Antarctica (Crawford and Muir

In Press). Although it is unclear how detrimental these fish are to native fish in the Colorado









River, interactions with various non-native fish have been widely implicated in the decline of

southwestern native fishes (Minckley 1991; Tyus and Saunders 2000). Non-native salmonids,

particularly brown trout, have been shown to be predators of native fishes (Valdez and Ryel

1995; Marsh and Douglas 1997) in Grand Canyon and rainbow trout predation on native fish has

also been documented in other southwestern U.S. systems (Blinn et al. 1993). Besides direct

mortality through predation, both rainbow trout and brown trout have demonstrated other

negative interactions with native fish in western U.S. river systems including interference

competition, habitat displacement, and agonistic behavior (Blinn et al. 1993; Taniguchi et al.

1998; Robinson et al. 2003; Olsen and Belk 2005). These lethal and sub-lethal effects of

interactions with native fishes have also been widely documented in New Zealand, Australia,

Patagonia, and South Africa (McDowall 2006).

Objective

While control of non-native species is widely considered as a management option, it is

rarely implemented and evaluated (Lessard et al. 2005; Pine et al. 2007), particularly for fish in

large river systems. Removal of non-native organisms to potentially benefit native species is

more frequently conducted in small streams (e.g., Meyer et al. 2006), in lakes and reservoirs

(e.g., Hoffman et al. 2004; Vrendenburg 2004; Lepak et al. 2006) and in terrestrial environments

(e.g., Erskine-Ogden and Rejmanek 2005; Donlan et al. 2007). However, recently much effort

has been expended to remove or reduce non-native fishes in the Colorado River (Tyus and

Saunders 2000). Unfortunately, little documentation is available to evaluate the efficacy of these

efforts (Mueller 2005). This study describes one such effort and evaluates the efficacy of a

program to reduce non-native fishes within humpback chub critical habitat. Given the ecological

and management interest in non-native species removals, this portion of the GCD adaptive

management program also represents an important first phase of active adaptive management to









benefit a focal biological resource, humpback chub. Specifically, the objectives of this study

were to: (1) evaluate the effectiveness of non-native control efforts in the mainstem Colorado

River, (2) investigate factors contributing to the effectiveness of control efforts, and (3)

characterize changes in the non-native and native fish communities.

Methods

Mechanical Removal Reach: Study Areas and Field Protocols

The Little Colorado River (LCR) inflow reach of the Colorado River extends from 56.3

river mile (RM) to 65.7 RM, as measured downstream from 0 RM at Lees Ferry, and is

recognized as having the highest abundance of adult and juvenile humpback chub in the

Colorado River (Valdez and Ryel 1995; Figure 2-1). This reach also has a relatively high

abundance of flannelmouth sucker Catostomus latipinus, bluehead sucker Catostomus

discobolus, and speckled dace Rhiniii, hijy % osculus owing to the availability of spawning and

rearing habitat in the LCR. Given the importance of this reach to native fishes, the LCR inflow

reach was selected as the area to test non-native mechanical removal efforts and was divided into

six river sections labeled A-F (Figure 2-1). Sections A and B are the right and left shore from

RM 56.3 to RM 61.8. Sections C and D are the right and left shore between RM 61.8 to RM

62.1 and include the LCR confluence and the mixing zone below the LCR. Sections E and F are

the right and left shore downstream of the LCR confluence from RM 62.1 to RM 65.7. I

stratified the study area into these 6 sections to control for the effect of the LCR discharge into

the mainstem Colorado River. Sections A and B are unaffected by the tributary and sections E

and F are believed to be of sufficient distance downstream of the mixing zone to be affected

uniformly throughout. Sections C and D include the LCR confluence and will be differentially

affected by LCR discharge throughout their lengths. Within river sections A-B and E-F, the









shoreline was divided into 500 m sites. The number of sites within each river section was:

A=19, B=19, E=13, and F=14. Sections C and D constitute single sites.

From January 2003 through August 2006, a total of 23 field trips were conducted to

remove non-native fish with serial depletion passes using boat-mounted electrofishing within the

mechanical removal reach. The majority of these trips removed fish during either 4 or 5

depletion passes; exceptions were in August 2003 (2 passes), September 2003 (3 passes), and

July 2004 (6 passes). All sites within sections A-B, C, and E-F were sampled during each pass.

Section D, encompassing the LCR confluence, was not sampled during any of the trips due to

concerns about equipment damage associated with high water conductivity issuing from the LCR

and possibly high native fish abundance near the confluence. All electrofishing occurred

following the onset of darkness and each depletion pass required 2 nights to complete.

Electrofishing crews consisted of a boat operator and a single netter. Two boat types (15-foot

rubber-hulled sport boat and 15-foot aluminum-hulled sport boat) and two types of electrofishing

control units (Coeffelt mark XXII and Smith-Root mark XXII) were used in this study. In an

attempt to standardize among boat and control unit type, current output was adjusted to produce

5000W of power during all electrofishing operations. Non-native fish were euthanized,

speciated, and total length (TL), and weight (g) recorded. Native fish were measured (TL) and

native fish larger than 150 mm TL were implanted with a passive integrated transponder (PIT)

tag.

To examine the effects of boat type, control unit type, location (either above or below the

LCR confluence), and boat operator on capture probability, I varied the deployment of boat and

control unit type to each river section in a systematic fashion during the first two years of the

study. The overall strategy was to ensure that a rubber- and an aluminum-hulled boat were









always deployed on opposite shorelines (e.g., sections A and B) and their positions reversed on

the subsequent trip. The two types of electrofishing control units were deployed on opposite

shorelines and reversed after each set of three trips. Four boat operators were randomly assigned

to a particular section and depletion pass within each trip. The same boat operators participated

in each trip with the exception of boat operator 2 (absent during July 2003 and September 2004)

and boat operator 3 (absent during January 2004). Experienced substitute boat operators (boat

operators 5 and 6) were employed in these instances. During the final 11 trips in the second two

years of the study, both electrofishing control unit type and boat operators were assigned to each

reach haphazardly.

Control Reach: Study Areas and Field Protocols

To determine if changes in the fish community in the mechanical removal reach were

related to environmental influences and not the mechanical removal, a control reach was

established upstream of the removal reach in an area of high rainbow trout density (44 RM-52.1

RM; Figure 2-2). This reach was stratified into 60, 500 m sites (30 on each shoreline). During

most trips, 24 sites were randomly chosen and sampled using identical capture methods as

outlined above in the mechanical removal reach. Exceptions occurred in January 2003 and

August 2003 when 25 and 11 sites were sampled. All captured non-native fish were speciated,

measured (TL), and fish > 200 mm TL were implanted with a uniquely numbered external tag

and their left pelvic fin removed prior to release all non-native fish were released alive.

Mechanical Removal Reach: Data Analysis

Following Dorazio et al. (2005), I used a hierarchical Bayesian modeling (HBM)

framework to estimate abundance and capture probability from data collected among the serial

removal passes. This framework assumes that the overall population is a collection of

subpopulations (defined below), each with different abundance and experiencing different









capture probability during removal efforts. Subpopulation abundance and capture probability are

sampled from common population level distributions conditional on unknown hyperparameters

(i.e., parameters that govern the population level distributions). This hierarchical structure

allows a model-based aggregation of data among subpopulations and can be thought of as an

intermediary between analyses that operate on data pooled over all subpopulations, and those

that operate on each subpopulation independently. The structure allows sharing of information

among subpopulations, particularly for subpopulations for which the data are relatively

uninformative or imply extreme parameter values. In these cases, the subpopulation parameter

values are more heavily influenced by the population distribution and are thus pulled, or shrunk

(Gelman et al. 2004), towards the population distribution means. The amount of shrinkage is a

function of both the difference between subpopulation and population distribution means and the

population distribution variance.

I defined closed subpopulations to correspond to fish within each mechanical removal site.

I assumed that the observed numbers of removals from site i (1,..., I) among removal pass

(1,..., J) were drawn from a multinomial distribution with number of trials equal to the site

abundance (N,) and cell probability vector 1, = [q,, I,...,r,]. If I first assume that capture

probability is constant among removal passes within each site i, then the cell probability in site i

in thejth depletion pass is given by

,r = 0, (1- 0)' (2-1)

where 0, is a constant capture probability in site i. The likelihood for the overall model is given

as


L(N,, )x,)= N' ( -0),) (N-- r,) (2-2)
j=i









where x, is the number of fish captured in site i and depletion pass, x, = x-' is the total


number offish captured in site i, and c, = I j= x,!.

Equation (2-2) is the familiar Zippin (1956) estimator and as above assumes that capture

probability 0, is constant within a site. To cast this model in a HBM framework, I assumed that

capture probability within a set of sites is sampled from a common distribution. The set of sites

could either be all sites within the removal reach, or a subset of sites belonging to a common

stratum. Because there is good reason to believe that electrofishing capture probability is

influenced by abiotic factors such as turbidity (Reynolds 1996) and because there is frequently

higher turbidity below the LCR confluence (Yard 2003), I chose to stratify the overall removal

reach into sites upstream (sections A and B) and downstream (sections C, E, and F) of the LCR

confluence and fit separate distributions to each strata. Similarly, fish abundance typically

differs upstream versus downstream (Gloss and Coggins 2005) of the LCR so separate

distributions of abundance were also used.

Following Dorazio et al. (2005), I assumed that the site specific capture probabilities were

sampled from beta distributions in each of the strata as 0k Bct,1 (. ,?k), where k is either 1

(upstream stratum) or 2 (downstream stratum) and ak and /k are the hyperparameters. The

mean ( k ) of the distribution is ak /(ck +,k ) and the variance is /Uk (1-/kA )/(k +l), where the

similarity parameter (rk) is ak + ,k. I assumed that the site specific abundances were sampled

from Poisson distributions with mean and variance Ak. For convenience, I estimated /, r, and

/ ( i/ = In A ) for each stratum. I chose diffuse prior distributions for each hyperparameter as:

/k -Uniform(0,1), k -Uniform(0,100), E(/k )-Normal(0,0.01), and SE(Vk )-Uniform(0,10).









To examine the effect of the covariates mentioned above on capture probability, I also re-

analyzed a subset of the data collected during 2003-2004 using a model that allowed capture

probability to vary among sites and passes as a function of covariate values. To accomplish this,

I assumed that capture probability was a logit function:

e1 (2-3)
J 1 + exp(- H (0 )+ 1(, 2X2( + ) + f/3x3(,,j) + 14x4(, ,) + -/5X5(,,j) + 6x6(,,j) + /X(,,) + A (,,))

where /, is the location coefficient (x, =1 for upstream sections A and B, x =0 for downstream

sections C, E, and F), f,2 is the boat hull type coefficient (x, =1 for rubber hull and x =0 for

aluminum hull), 3, is the electrofishing control unit coefficient (x, =1 for the Smith Root Mark

XXII and x =0 for the Coeffelt Mark XXII), 74 is the boat operator 2 coefficient (x4=1 for

operator 2 and x4=0 for not operator 2), 75 is the boat operator 3 coefficient (x =1 for operator

3 and x5=0 for not operator 3), /6 is the boat operator 4 coefficient (x6=1 for operator 4 and

x6=0 for not operator 4), f/ is the boat operator 5 coefficient (x,=l for operator 5 and x,=0 for

not operator 5), and p/ is the boat operator 6 coefficient (x, =1 for operator 6 and x, =0 for not

operator 6). Lastly, this coding scheme implies that /7 is the untransformed capture probability

for boat operator 1 in downstream sites in an aluminum-hulled boat outfitted with the Coeffelt

Mark XXII electrofishing control unit. I assumed that the site and removal pass specific values

of each of the coefficients (/ z) were sampled from normal distributions with hyperparameter

mean (k/u(z)) and standard deviation (op(z)), where Z=0, 1, ... 8. I specified diffuse priors for

each hyperparameters as: /u(z) -Normal(0,0.01), and "(z) -Uniform(0,10).

I implemented these analyses in programs R (R Development Core Team 2007) and

Winbugs (Lunn et al. 2000). For each trip analyzed, I characterized the distribution of each









parameter among 20,000 Markov Chain Monte Carlo samples with a thinning frequency of 10

and discarding the first 10,000 bum in samples. I examined convergence using Gelman and

Rubin's potential scale reduction factor (R Development Core Team 2007).

Control Reach: Data Analysis

I assessed the abundance of rainbow trout within the control reach using electrofishing

catch rate and mark-recapture-based open population abundance estimates. Because all rainbow

trout marked with external tags were also given a secondary fin clip, I attempted to incorporate

the rate of tag loss into the mark-recapture-based estimates of survival, capture probability, and

abundance. I estimated tag loss rate by comparing the observed and predicted proportion of

recaptured fish that retained tags each trip. This proportion is not influenced by survival or

capture probability under the assumption that survival and capture probability are independent of

tag retention.

To derive this estimator, I first assumed that tag loss rate during the first month after initial

tagging could be different from the rate experienced in subsequent months. This allows for the

possibility that tags may be lost at a higher rate initially (e.g., as a result of improper placement),

but that the rate of tag loss declines after this initial loss. I predict the number of tagged fish in

month t as

T, = S (,_, (1-/1,)+ F,_ (1-/) + R,_ (1 -/1)), (2-4)

where S is the monthly survival rate, T, is the number of tagged fish available for capture just

prior to sampling in month t, 12 is the monthly secondary tag loss rate, F, 1 is the number of

newly tagged fish in month t-1, 11 is the monthly initial tag loss rate (suffered in the month

following tagging), R, is the number of fish that had lost their tag prior to month t-1 and were









retagged in month t-1. Conversely, I predict the number of fish that have lost their tag in month t

as

iL, =S(L_ -R,_ + ,(12)+F, (11)+R, (1)), (2-5)

where L, is the number of fish that have lost their tag and are available for capture just prior to

sampling in month t. The predicted tag retention rate (,) of recaptured fish in the population in

month t is then

r, = Tl (2-6)
T1+L,

Note that equations (2-4) and (2-5) are linked by the R term such that when recaptured fish

without a tag are observed, they are fitted with a new tag and thus decremented from L and

added to T. To estimate 11 and 12, I minimized the sum of squares between observed and

predicted retention rate among the 22 sampling occasions following the first one. It is worth

noting that because the monthly survival rate (S) appeared in each term of equation (2-6), there

was no need to estimate it in order to estimate tag loss rates.

I estimated monthly survival rate (S,) and capture probability (A ) conditional on tag loss

rates generally following a single age recoveries only model (Brownie et al. 1985). However, for

computational simplicity, I assumed that observed recaptures followed a Poisson rather than a

multinomial distribution. Under this structure, the complete capture history is not used and the

predicted numbers of fish released in month e and recaptured with tags in a subsequent month t

is

ri = (F/ +Re)(1-l)(1-12)(-e-1) (2-7)









To reduce the number of parameters to be estimated, I set the monthly survival rate among

months not sampled equal to the survival rate of the next sampled month. Assuming that the

observed numbers of fish released in month e and recaptured with tags in a subsequent month t

(re,t) represent independent samples from Poisson distributions with means given by equation (2-

7), the log-likelihood function ignoring terms involving only the data is

22 23
InLr, S)= [- + n, (2-8)
e=l t>e

where p and S are the unknown capture probability and monthly survival rate vectors to be

estimated. The model was implemented in a Microsoft Excel spreadsheet using Solver (Ladson

and Allan 2002) as the non-linear search procedure. As a measure of uncertainty, I computed

95% likelihood profile confidence intervals on p and S using Poptools (Hood 2000). I

estimated the abundance of rainbow trout > 200 mm TL by dividing the numbers of fish captured

by the capture probability. Approximate 95% confidence intervals on these abundance estimates

were calculated using the confidence bounds on the capture probability estimates.

Results

Mechanical Removal Reach

Over 36,500 fish from 15 species were captured in the mechanical removal reach during

2003-2006 (Table 2-1). The majority of these fishes (23,266; 64%) were non-natives and were

comprised primarily by rainbow trout (19,020; 82%), fathead minnow Pimephalespromelas

(2,569; 11%), common carp Cyprinus carpio (802; 2%), and brown trout (479; 1%). Catches of

native fish amounted to 13,268 (36%) and were comprised of flannelmouth sucker (7,347; 55%),

humpback chub (2,606; 20%), bluehead sucker (2,243; 17%), and speckled dace (1,072; 8%).

The contribution of rainbow trout to the overall species catch composition fell steadily through

the course of the study from a high of approximately 90% in January 2003 to less than 10% in









August 2006 (Figure 2-3). Overall, non-native fish comprised more than 95% of the catch in

2003, but following July 2005 generally contributed less than 50%. Owing to particularly large

catches of flannelmouth sucker and humpback chub in September 2005, the non-native

contribution to the catch in that month was less than 20%. While the catch of non-native fish

generally fell through the course of the study, catches of non-native cyprinids (dominated by

fathead minnows) increased in 2006.

Using the HBM, the estimated abundance of rainbow trout in the entire removal reach

ranged from a high of 6,446 (95% credible interval (CI) 5,819-7,392) in January 2003 to a low of

617 (95% CI 371-1,034) in February 2006; a 90% reduction over this time period (Table 2-2).

Between February 2006 and the final removal effort in August 2006, the estimated abundance

increased by approximately 700 fish to 1,297 (95% CI 481-2,825). Though this increase was

more than double the February 2006 estimate, the August 2006 estimate was much less precise.

The estimated abundance in the downstream stratum of the mechanical removal reach was

typically approximately 30% of that in the upper stratum (Figure 2-4) and the density was also

typically lower (Table 2-2). The estimated capture probability in the upper stratum ranged from

4% to 34% (Figure 2-4). The estimated capture probability was generally lower in the lower

stratum and ranged between 2% and 19%.

Net immigration rate estimates indicate that fish were moving into both strata within the

removal reach at a higher rate during 2003-2004 than during 2005-2006 (Figure 2-5).

Additionally, it appears that net immigration may be lowest in the late fall through early winter,

and highest between January and March. During 2005-2006, there were only two time intervals

that suggest net immigration rate was different than zero in the downstream stratum and one in

the upstream stratum. However, since these estimates are the difference between two









distributions, each with their own error, the net immigration estimates were imprecise for many

of the time periods (Figure 2-5).

The results of the covariate analysis indicated that most of the factors had little influence

on capture probability (Figure 2-6). In fact, there was little indication that any factor had a

strong directional effect in either raising or lowering capture probability. The exceptions were

that on four of the trips, the boat operator 2 effects were significantly less than zero indicating

that this operator had a negative effect on capture probability compared to operator 1.

Additionally, operator 3 tended to have a slightly positive effect on capture probability.

Operators 5 and 6 participated in only 1 and 2 trips, respectively, but did not seem to affect

capture probability relative to operator 1. Neither boat type nor type of electrofishing control

unit had a strong directional effect on capture probability. The factor that had the largest effect

on capture probability was location in the overall reach. Though there was no indication that

capture probability was uniformly higher or lower in the upstream stratum versus the

downstream stratum, several of the trips indicated that capture probability was significantly

different among strata. This result supports the use of separate distributions to describe capture

probabilities in the overall HBM to estimate abundance, capture probability, and net immigration

rate.

Control Reach

A total of 11,221 fish representing 7 species were captured during control reach sampling

(Table 2-3). The majority of fish captured were rainbow trout (95%), followed by flannelmouth

sucker (3%), and brown trout (1%). A general pattern of decreasing rainbow trout abundance

was observed throughout the study, particularly following spring of 2005 (Figure 2-7). Initial

(11) and secondary (/) monthly tag loss rate estimates were 11% and <1%, respectively,









suggesting that most tag loss occurred shortly after tagging. Rainbow trout abundance within the

control reach was estimated at between 5,000 and 10,000 fish during 2003-2004 and between

2,000 and 5,000 during 2004-2005 (Table 2-4 and Figure 2-7). This analysis coupled with the

catch rate assessment (Figure 2-7) suggests that rainbow trout abundance likely declined by one

half or more between the first and last two years of the study. Capture probability ranged

between 3% and 13% with no strong temporal pattern (Figure 2-7). Estimated monthly survival

rate ranged from a low of approximately 0.72 to a high approaching 1 (Figure 2-8). The lowest

survival rates were observed during 2005.

Comparison of Mechanical Removal and Control Reaches

The abundance of rainbow trout declined through the study both in the mechanical removal

reach and in the control reach; however, the pattern of decline was dissimilar among reaches

(Figure 2-9). In the mechanical removal reach, the largest decline (62%) occurred between

January 2003 and September 2004. Rainbow trout abundance in this reach declined much less

rapidly from January 2005 to August 2006. In contrast, rainbow trout abundance in the control

reach was constant to slightly declining from March 2003 through September 2004, but

displayed a strong negative trend subsequently. These patterns suggest that removal efforts

likely affected abundance in the mechanical removal reach predominantly during 2003 and 2004.

Another difference between the mechanical removal and control reaches was the seasonal

patterns in rainbow trout abundance. In the removal reach, a pattern of declining abundance

during each three-month bout of removal efforts (e.g., January-March) was followed by an

increase in abundance at the beginning of the next series of removal efforts (e.g., July-

September), particularly during 2003-2004 (Figure 2-4). This pattern would be expected if the

removal rate was greater than the immigration rate only during each removal series. This pattern

was not evident in the control reach considering either the catch rate or abundance estimates









(Figure 2-7) suggesting that mechanical removal was influencing the abundance of rainbow trout

in the removal reach.

These seasonal patterns in rainbow trout abundance are mirrored by trends in the average

length (Figure 2-10). Average length among the two reaches tended to converge in July of each

year followed by a period of divergence. One possible explanation of this pattern is selective

removal of larger individuals followed by reinvasion, particularly following the winter/spring

removal efforts, of larger individuals from upstream sources. This pattern was not observed in

July 2005 and was preceded by a significant decline in average length in the control reach.

Taken together, these observations suggest a disruption in the net immigration pattern, possibly

from upstream sources, into the removal reach during 2005.

Discussion

Mechanical Removal: Effective Program?

Results suggest that the mechanical removal program was successful in reducing the

abundance of non-native fishes, primarily rainbow trout, in a large segment of the Colorado

River in Grand Canyon. However, maintenance of low rainbow trout abundance in the removal

reach was also facilitated by reduced immigration rates during 2005-2006 and a systemic decline

in abundance. Common features of this study and other successful non-native mechanical

removal efforts are significant and sustained removal effort. Bigelow (2003) demonstrated that

population level changes were not evident in removal efforts aimed at non-native lake trout

Salvelinus namaycush in Yellowstone Lake until the latter years of a four year study when

additional support for the project (e.g., funding and equipment) allowed increases in total

removal effort and efficiency. Similarly, removal objectives for non-native brook trout

Salvelinusfontinalis, golden trout Oncorhynchus mykiss aguabonita, and rainbow trout from

small high altitude lakes in the Sierra Nevada were achieved with year-round gillnet fishing









(Knapp et al. 2007). In combination with increased predation from stocked predators, Hein et al.

(2006; 2007) demonstrated effective control of non-native crayfish Orconectes rusticus using

mechanical removal, but only with sustained and significant removal effort. The necessity for

sustained "maintenance" control of non-natives is not uncommon (Pine et al. 2007) as many non-

native species demonstrate high resilience, and are well adapted to their introduced environment

as evidenced by their invasion success and warranted need for management action.

In contrast, Meyer et al. (2006) document a recent unsuccessful effort to remove brook

trout from a 7.8 km reach of a second-order stream in Idaho using electrofishing. Crews

removed fish over 3 days in August during each year for four years. While the capture technique

was likely appropriate, the overall effort was apparently insufficient to significantly reduce the

population. In the present study and considering the net immigration rates of rainbow trout into

the mechanical removal reach during 2003-2004, much smaller and possibly undetectable

reductions in overall abundance would have been realized had removal efforts been applied only

once per year.

Serendipitous Timing: What Led to the Decline of Rainbow Trout in the Control Reach?

The decline of rainbow trout abundance observed in the control reach was likely

precipitated by at least two factors. First, rainbow trout abundance in the Lees Ferry reach (-15

RM at GCD to RM 0) of the Colorado River increased during approximately 1992-2001 and

abundance in this reach steadily fell during 2002-2006 (Makinster et al. 2007). With the

exception of limited spawning activity in select tributaries of the Colorado River in Grand

Canyon, rainbow trout reproductive activity appears to be limited mainly to the Lees Ferry reach

(Korman et al. 2005). Examination of length frequency distributions of rainbow trout captured

using electrofishing from Glen Canyon Dam to RM 56 during 1991 through 2004 also supports

the idea that Lees Ferry is the primary spawning site, as the juvenile size class of rainbow trout is









largely absent from collections downstream of RM 10 (Figure 2-11). Thus, it is reasonable to

conclude that at least for the last 10-15 years, the natal source of most rainbow trout in this

system is the Lees Ferry reach. This is significant because it suggests that abundance of rainbow

trout in Grand Canyon is partially influenced by trends in rainbow trout abundance and

reproduction in the Lees Ferry reach.

Second, it has been widely demonstrated that the density of rainbow trout is not uniform in

the Colorado River below GCD and distribution patterns are likely influenced by food resources

and foraging efficiency (Gloss and Coggins 2005). Rainbow trout density generally declines

with downstream distance from GCD but exhibits punctuated declines below the confluences of

the Paria River and the LCR. The density of algae and invertebrates in the Colorado River also

decline along this gradient (Kennedy and Gloss 2005) suggesting a possible linkage between

distance from the dam and primary production. A major factor likely influencing these

distributional patterns is sediment delivery from tributaries and the subsequent effects of elevated

turbidity in the Colorado River in downstream sections. Yard (2003) demonstrated that these

tributary inputs of sediment contribute to high turbidity and limit aquatic primary production.

Trout are predominantly sight feeders thus, high turbidity is likely to adversely affect foraging

efficiency by decreasing encounter rate and reactive distance to prey items (Barrett et al. 1992).

From September 2004 through January 2005, the discharge and sediment load from the

Paria increased to the point that a threshold outlined in the Glen Canyon Dam Adaptive

Management Program related to rebuilding depleted Colorado River sandbars was reached

triggering an experimental high flow from GCD in November 2004. It is possible that the high

flow event and the associated period of elevated turbidity may have influenced rainbow trout

density downstream of the Paria River confluence, possibly through elevated mortality rates.









Estimated survival rates in the control reach generally support the notion that rainbow trout may

have experienced diminished survival rates during late 2004 and early 2005 (Figure 2-8).

Other Species

Beginning in September 2005, large increases in the catch of non-native fathead minnow

and black bullhead Ameiurus melas were observed compared to the previous 17 trips, suggesting

either increased immigration and/or survival of these fishes in the mechanical removal reach.

Since these fish are not captured with any regularity in the control reach nor in other sampling

upstream of RM 44 (USGS, unpublished data), it is reasonable to conclude that their source is

not upstream. Stone et al. (2007) documented the presence of these species and other warm

water non-natives in the LCR 132 km upstream from the confluence and suggested this

tributary as the likely source of fathead minnow, black bullhead, and 6 other non-native fish

frequently encountered in the lower LCR and the mechanical removal reach. Thus, one

possibility for the elevated catch of fathead minnow and black bullhead in the mechanical

removal reach during this latter timeframe is an elevated emigration rate of these fishes from the

LCR. Alternatively, increasing water temperature, particularly in 2005 (Figure 2-12), and the

concurrent reductions in rainbow trout biomass, may have influenced the survival and activity of

these fishes causing them to be both more abundant and more susceptible to capture.

Bias in Capture Probability and Abundance Estimates

Capture probability estimates from the upper stratum of the mechanical removal reach and

the control reach are surprisingly different. Neither of these reaches is differentially influenced

by large tributary inputs and they share similar overall channel morphology yet capture

probability estimates are generally nearly twice as high in the removal reach as in the control

reach. Several authors have demonstrated that capture probability is typically over-estimated

using electrofishing depletion-based methods (Peterson et al. 2004; Rosenberger and Dunham









2005). The typical mechanism for this finding is heterogeneity in capture probability among

individuals in the population. As a result, the fish remaining after each successive pass have

overall lower capture probability than those in preceding passes. This bias in capture probability

then leads to negative bias in abundance. Because of this bias, it is likely that the depletion-

based capture probabilities estimated in this study are higher than were actually realized.

In principle, the additional information available from mark-recapture should provide less

biased estimates of capture probability in the control reach. However, this may not be true due

to possible inadequate mixing of fish between each mark-recapture sampling event. Because

only 40% of the available sites were sampled during each mark-recapture event, if fish did not

mix completely between passes it is possible that capture probability may have been under-

estimated since not all fish had an equal probability of capture within each event. Theoretically

this assumption is met by the random selection of sampled sites each trip. In practice, however,

only 40% of fish (assuming uniform distribution) actually had an opportunity to be captured.

Additionally, it is possible that there was sampling induced heterogeneity in capture probability.

This is a common feature of mark-recapture experiments for small mammals (Otis et al. 1978)

and a recent paper by Askey et al. (2006) suggests that fish may develop anti-capture behavior as

well. Therefore, the realized capture probabilities were likely between those estimated in the

control reach and those estimated in the removal reach. If true, abundance and density estimates

are likely over-estimated in the control reach and under-estimated in the removal reach.

Recommendations for Future Mechanical Removal Operations

I recommend that further effort be spent better documenting the preferred habitat of target

non-native species. This information could then be used to more effectively distribute removal

effort among habitat types that contain the highest density of non-native species. Bigelow et al.

(2003) describe the use of hydroacoustic surveys to better target areas of high lake trout









abundance, increasing the efficiency of the control program. A possible technique to better

determine these high density areas in the mechanical removal reach would be to employ a finer

scale shoreline habitat-based delineation of removal sites, rather than the coarse 500 m sites used

in the present study. Serial depletion data could then be analyzed with the HBM to include a

habitat covariate for density. This approach has been successfully used to describe patterns in

the density of organisms as a function of habitat characteristics (Royle and Dorazio 2006).

Results from the present covariate analysis indicate that most variability in capture

probability is related to site location rather than methodological issues. In the context of

designing future removal efforts and the larger monitoring program for non-native salmonids in

Grand Canyon, this is a fortunate result as it suggests that current levels of standardization

among equipment will have a reasonably high likelihood of producing index data useful for

determining trends in salmonid abundance and distribution. However, the observed variability

among boat operators implies that additional training to reinforce consistent methodologies may

be useful to further minimize that source of variability. A more significant finding is the

heterogeneity in capture probability among upstream and downstream strata. If these differences

are primarily related to uncontrollable factors such as turbidity or shoreline and substrate type as

suggested by Speas et al. (2004), additional research should be conducted to better describe these

relationships. Unfortunately, I was not able to obtain robust measurements of turbidity with

sufficient regularity to investigate the effect of turbidity on capture probability. However,

assuming serial depletion methods are likely to be used in future fish control programs, these

efforts may provide the ideal setting to explore these associations.

Adaptive Management in Grand Canyon: The Future

As stated at the beginning of this chapter, this study documents the implementation of an

ecosystem-scale adaptive management action aimed at testing the efficacy of a particular









management policy (i.e., non-native control) in order to improve the status of native fish

resources in Grand Canyon. Though this study focuses on the efficacy of implementing the

policy, the more interesting, important, and difficult questions are related to evaluating whether

the policy will have the intended effect. I predict that if non-native salmonids are a significant

and uncompensated mortality source for native fish attempting to rear in the mainstem Colorado

River, then the survival rate and abundance of juvenile native fish in the mainstem should

increase during 2003-2006. I would further predict that humpback chub recruitment associated

with the 2003-2006 brood years should increase.

There are some indications that the abundance of native fish has increased in the removal

reach during 2003-2006 (Figure 2-3; Chapter 5) suggesting either increased survival rate,

increased production of juvenile fish, or both. However, these initial signals are not adequate to

infer the success of the policy for two important reasons. First, the unplanned increases in

release water temperature are nearly perfectly temporally correlated with the magnitude of the

non-native fish reduction (Figures 2-4 and 2-12). As water temperature is also a controlling

factor affecting quality of rearing habitat in the mainstem river (Gloss and Coggins 2005), this

confounding makes separation of these two effects impossible at this time. Second, since there is

not a monitoring program for estimating temporal trends in survival rate, likely the most reliable

available measure of a native fish response is humpback chub recruitment. Because the age-

structured mark-recapture model (Chapter 4) is not able to provide estimates of year class

strength until fish reach age-4, the best data to infer change in survival are not yet available.

One strategy to separate out the effects of non-native fish versus increased water

temperature would be to implement a future experiment with one of these factors changed from

the 2003-2006 condition. However, since temperature control is not available at this time and









since rainbow trout abundance appears diminished system-wide, there may be limited near-term

opportunities to manipulate the system further. Determining what future experiments to conduct

will be determined by the Glen Canyon Dam Adaptive Management Program using input from

studies such as this but it should be noted that the implementation and ultimate success of even

the best designed experiments will be dependent on unmanageable factors such as climate

(Seager et al. 2007) and unexpected biotic interactions.











Table 2-1. Electrofishing catch by species in the mechanical removal reach for each month, 2003-2006.

Species
Trip Date # Removal Passes BBH BHS BNT CCF CRP FHM FMS GSF HBC PKF RBT RSH SMB SPD STB SUC


Jan-03
Feb-03
Mar-03
Jul-03
Aug-03
Sep-03
Jan-04
Feb-04
Mar-04
Jul-04
Aug-04
Sep-04
Jan-05
Feb-05
Mar-05
Jul-05
Aug-05
Sep-05
Jan-06
Feb-06
Mar-06
Jul-06
Aug-06
Total


5 8 87 80 17 188
5 18 24 33 21 165
5 3 11 21 1 22 8 89
5 4 12 63 29 4 267
2 2 4 12 14 79
3 1 19 11 31 4 119
4 3 32 88 23 18 169
4 9 37 29 1 9 13 110
5 5 24 22 18 44 218
6 9 84 29 1 26 32 296
4 6 33 7 16 6 190
4 11 72 17 29 13 258
4 8 54 14 27 72 244
4 3 38 4 1 14 39 191
4 8 51 4 14 73 176
4 17 159 9 2 45 9 480
4 9 124 4 36 17 419
4 14 576 7 47 190 1,140
4 23 197 9 38 685 545
4 15 98 5 10 300 529
4 12 96 2 8 322 365
4 15 331 8 64 192 554
4 13 165 3 1 169 490 556
190 2,243 479 7 802 2,569 7,347


26 1 3,605
26 1,913
13 1 1,195
124 2,278
17 779
37 2 818
51 1,330
52 622
61 6 867
142 9 1,464
27 3 480
43 687
61 1 623
49 2 283
82 3 318
1 220 432
86 1 295
600 230
249 357
171 1 103
1 196 1 66
145 2 159
128 34 116
2 2,606 67 19,020


a BBH=black bullhead (Ameiurus melas), BHS=bluehead sucker (Catostomus discobolus), BNT=brown trout (Salmo trutta), CCF=channel catfish
(Ictalurus punctatus), CRP=common carp (Cyprinus carpio), FHM=fathead minnow (Pimephales promelas), FMS=flannelmouth sucker
(Catostomus latipinnis), GSF=green sunfish (Lepomis cyanellus), HBC=humpback chub (Gila cypha), PKF=plains killifish (Fundulus zebrinus),
RBT=rainbow trout (Oncorhynchus mykiss), RSH=red shiner (Cyprinella lutrensis), SMB=smallmouth bass (Micropterus dolomieu),
SPD=speckled dace (Rhinichthys osculus), STB= striped bass (Morone saxatilis), SUC=unidentified sucker.


18 4
53 3
34
92 3
47
7
19
52
39
51
38 2 1
24 24
187 4
115 1
70
84
56
1 63 9
1 1,072 2 59










Table 2-2. Estimated abundance and density of rainbow trout in the mechanical removal reach at the beginning of each month, 2003-
2006. Uncertainty estimates (95% CI) are 95% Bayesian credible intervals.


Total Reach Abundance


Trip Date
Jan-03
Feb-03
Mar-03
Jul-03
Aug-03
Sep-03
Jan-04
Feb-04
Mar-04
Jul-04
Aug-04
Sep-04
Jan-05
Feb-05
Mar-05
Jul-05
Aug-05
Sep-05
Jan-06
Feb-06
Mar-06
Jul-06
Aug-06


N
6,446
3,073
2,372
5,253
1,574
3,008
2,207
1,611
1,425
3,445
932
2,459
989
869
975
1,626
690
697
710
617
669
726
1,297


95% CI
5,819-7,392
2,802-3,492
1,939-3,014
4,249-7,616
1,253-2,199
1,964-4,197
1,953-2,635
1,098-2,809
1,227-1,710
2,533-5,284
734-1,536
1,647-3,752
819-1,275
519-1,785
636-1,548
742-5,837
498-1,080
460-1,291
514-1,121
371-1,034
280-1,460
376-2,210
481-2,825


Upper Stratum Abundance


N
4,977
2,437
2,023
3,614
1,237
2,399
1,684
845
1,075
1,718
677
1,980
722
386
782
736
415
411
502
479
367
538
767


95% CI
4,519-5,640
2,226-2,778
1,606-2,671
3,164-4,183
1,001-1,652
1,438-3,507
1,472-2,002
732-1,026
925-1,325
1,566-1,925
515-1,266
1,296-3,290
675-786
317-516
498-1,377
560-1,085
339-549
288-601
386-719
258-879
154-860
251-1,853
262-2,087


Lower Stratum Abundance


N
1,469
637
349
1,639
336
609
523
767
350
1,727
255
479
266
483
193
891
275
286
208
138
302
188
530


95% CI
1,168-1,996
489-879
289-485
902-3,776
178-845
345-1,187
385-851
293-2,009
269-516
856-3,627
183-455
199-1,060
115-539
142-1,388
80-427
128-5,056
115-638
108-893
100-580
61-290
69-992
89-410
136-2,090


Density (Fish/Km)
Upper Lower
Stratum Stratum
1,512 563
740 244
615 134
1,098 629
376 129
729 233
512 201
257 294
327 134
522 662
206 98
602 184
219 102
117 185
238 74
224 341
126 105
125 110
153 80
145 53
111 116
163 72
233 203










Table 2-3. Electrofishing catch by species in the control reach for each month, 2003-2006.
Species


Trip Date Control
Sites
Jan-03 25
Feb-03 24
Mar-03 24
Jul-03 24
Aug-03 11
Sep-03 24
Jan-04 24
Feb-04 24
Mar-04 24
Jul-04 24
Aug-04 24
Sep-04 24
Jan-05 24
Feb-05 24
Mar-05 24
Jul-05 24
Aug-05 24
Sep-05 24
Jan-06 24
Feb-06 24
Mar-06 24
Jul-06 24
Aug-06 24
Total


BBH BHS BNT
10
8
5
8
4
7
5
3
2 14
2
9
8
1
9
1 9
1 11
5
1 2
2 2
4
5
5 2
1 10 1
1 22 134


CRP FMS
1 1
1
1
1 2


1 1
3
9
6
1 4
1
4
5
34
21
1 72
1 31
2 53
23
5 47
1 52
7 378


HBC RBT
444
548
888
416
256
1,036
702
434
851
1 491
346
498
503
476
540
277
332
284
277
243
336
1 176
1 294
3 10,648


SPD SUC






1






2


15 3


a BBH=black bullhead (Ameiurus melas), BHS=bluehead sucker (Catostomus discobolus),
BNT=brown trout (Salmo trutta), CRP=common carp (Cyprinus carpio), FMS=flannelmouth
sucker (Catostomus latipinnis), HBC=humpback chub (Gila cypha), RBT=rainbow trout
(Oncorhynchus mykiss), SPD=speckled dace (Rhiiii /11/hiy osculus), SUC=unidentified sucker.









Table 2-4. Estimated abundance and density of rainbow trout in the control reach at the
beginning of each month, 2003-2006. Uncertainty estimates (95% CI) are 95% profile
likelihood confidence intervals.


Trip Date
Feb-03
Mar-03
Jul-03
Aug-03
Sep-03
Jan-04
Feb-04
Mar-04
Jul-04
Aug-04
Sep-04
Jan-05
Feb-05
Mar-05
Jul-05
Aug-05
Sep-05
Jan-06
Feb-06
Mar-06
Jul-06


Total
N
5,058
10,571
10,106
8,819
8,051
9,952
8,998
7,939
8,758
6,981
7,208
4,138
4,527
5,253
3,163
3,247
2,955
4,032
2,992
2,518
2,131


Abundance
95% CI
3,500-7,262
8,064-14,136
6,572-16,367
5,494-13,593
6,004-10,860
6,491-15,662
5,570-15,024
5,379-11,798
5,895-13,254
4,519-11,171
4,733-10,795
2,853-6,090
3,344-6,202
3,939-6,907
1,967-5,245
2,126-4,900
1,877-4,604
2,502-6,694
1,957-4,804
1,594-3,443
1,113-4,062


Density (Fish/Km)
1,018
2,128
2,034
1,775
1,621
2,003
1,811
1,598
1,763
1,405
1,451
833
911
1057
637
654
595
812
602
507
429




















































Figure 2-1. Map of the mechanical removal reach of the Colorado River within Grand Canyon,
Arizona. Depicted on the map are the reach sections (A-F) and the sites within each
reach (e.g., 1-19). The number of river miles downstream from Lees Ferry, Arizona
is indicated at demarcation lines.






45





















































Figure 2-2. Map of the control reach of the Colorado River within Grand Canyon, Arizona.
Depicted on the map are the sites within the reach. The number of river miles
downstream from Lees Ferry, Arizona is indicated at demarcation lines.







46













Humpback Chub
Flannelmouth Sucker
Bluehead Sucker
Speckled Dace
Cyprinids
Ictalurids
Centrarchids
Brown Trout
Rainbow Trout


A 100 -
90 -
80 -
70 -
- 60 -
-
2 50 -
-
S 40 -
30 -
20 -
10 -
0-
-
B 4500
4000 -
3500 -
3000
r 2500
-i

LL 2000
1500
1000
500 -
0-


C'o)o'CO n n n'3"- ':-I- ":T 971 T L LD U .) n LD U-))LD (.(DCD O (DCO (D
CO O OO O OO OOO OOL OOO OO-Q 0OO 6O
M (D 1( = ML-" 1D 00 L = (D. (D&J M
--D LL < -D -D

Figure 2-3. Percent composition (A) and number of fish (B) by species captured with
electrofishing in the mechanical removal reach among months, 2003-2006.


Humpback Chub
Flannelmouth Sucker
Bluehead Sucker
Speckled Dace
Cyprinids
Ictalurids
Centrarchids
Brown Trout
Rainbow Trout












2004








S--+.-r
.-0- -.


LO


Co

1%
tN


A








$-1


0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Trip Date


0 f 0 0 l 0
tfl c -0! 3 Ct
ID n i ^
M cn -; L


Figure 2-4. Estimated catch rate (A), abundance (B), and capture probability (C) for rainbow
trout in both the upstream and downstream strata in the mechanical removal reach
among months, 2003-2006. Error bars represent 95% Bayesian credible intervals.














48


2005 2006
Upstream
Downstream







_.- t- t~ 9- ., l


Upstream
Downstream






I I














200 2003 2004 2005 2006
150 -
100- V
50 -
0-
-50
S-100
-15 A
-200 Upstream Stratum
.o 200
o 150
E 100 -
50
z

-50 -
-100
-150 B
-200 'Downstream Stratum

_0 i >0 0) CL = -Q 0 0) C0 = -0 00 >, 0) 0 L = _0 >0 0)
(D CO (D Mo M c (D Mn (0 (D Mo M c c cD
LL < (f) < < O 7 <

CO CO COJ

Time Interval


Figure 2-5. Net immigration rate into the upstream (A) and downstream (B) strata of the
mechanical removal reach within time intervals between January 2003 and August
2006. Error bars represent 95% Bayesian credible intervals.













Upstream Effect





^ "-,J^I.


Rubber Hull Effect


GLc' -
C


Ax


-2 0 2 4 6 -4 -2 0 2 4 6


1J -







S-





C4



a -



---


LD -
t 4 -
c -



















0-
t" ----~




















s--


p 2
Boat Operator 2 Effect


-4 -2 0 2 4 6


Boat Operator 4 Effect


Electrofishing Control Unit Effect









-4 -2 0 2 4 6


Boat Operator 3 Effect





-I



-4 -2 0 2 4 6
pts
Boat Operator 5 Effect


-4 -2


4 6 -4 -2 0


Figure 2-6. Probability plots for coefficient values influencing capture probability among 10
trips (January 2003-July 2003; January 2004-September 2004). Within each plot,
each line is the estimated probability density for the coefficient in a trip.


-4 -2 0 2 4 6


Boat Operator 6 Effect


': -- W ,,- -._,,









2004


I' 1 11-


V Ip.;


I I I I I
"- 0..


Q C
I I


JJ1$-1--11-


I I I I I I I I I I I I I I I I

Trip Date


Figure 2-7. Estimated catch rate (A), abundance (B), and capture probability (C) for rainbow
trout in the control reach among months, 2003-2006. Error bars represent 95%
profile likelihood confidence intervals.







51


,t


2003


2005


2006


-I-- i
















1.0 -





0.8 -











o





0.2





0.0
0 .0 I 1 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I
000000000000000000000000000000000000000000
o OOOOwe L waOOw 0-C e OO o 0 oooTT OOx OOOOOOJ OwOOO
L L -' > < Z L- Z = < 0 Z Q .. L L < -

Month

Figure 2-8. Estimated monthly survival rate of rainbow trout in the control reach during 2003-
2006. Error bars represent 95% profile likelihood confidence intervals.

















o Removal Reach
- Removal Reach-Lowess Fit
--- Removal Reach-Linear Fit
* Control Reach
- Control Reach-Lowess Fit
- Control Reach-Linear Fit


7-




6-
CD
o
CD
0

5-
- -



r
(3
0-
<4-


ro


E
a)


Q-




1 -





0-


I I I
- o :- -O

< )O UO

Trip Date


I I I
LT L. CO
000

=4 D MB
a 0a


- 14




- 12




- 10 o
o



-4 -
-8



--




-4 0


Figure 2-9. Estimated rainbow trout abundance in both the mechanical removal and control
reaches at the beginning of each trip during 2003-2006. The solid lines represent the
locally weighted polynomial regressions (Lowess) fit to each time series. The dashed
lines represent linear regressions fit to either the 2003-2004 or 2005-2006 portions of
the time series.


* '
*


0 0
0D CD
I I)
C I .


I I
00

^^
















350 -


300 4 +



250 -



-c
200 -
7)
-J

O 150 -



100 -



50-
Control Reach
o Removal Reach

0 IIII I II I III I I I I I I I I I I I
f l 0 fl f f f "1 "= If1 Lf) I. L Lf Lf ) L CO CO C O C O
c i L M c, 0 c L i c ci C ifl -
LL < m -L L < m -L L < W -D LL-
Trip Date


Figure 2-10. Estimated total length and 95% confidence intervals of rainbow trout captured in
both the mechanical removal and control reaches during 2003-2006.













RM -15 to RM -10


S1 n= 17033


0 200 400 600

Total Length(mm)


RM 20 to RM 30

n n= 950


C-


1 I
0 200 400 600

Total Length(mm)


RM -10 to RM 0


F n= 31798


0 200 400 600

Total Length(mm)


RM 30 to RM 40


T f nn= 2365


0 200 400 600

Total Length(mm)


RM 0 to RM 10


















R0 40 to RM 50

0 200 400 6007

Total Length(mm)


RM 40 to RM 50

[ I [1n= 6477


0
Cn






0 200 400 600

Total Length(mm)


0


RM 10 to RM 20


n= 1272

















0 200 400 600

Total Length(mm)


RM 50 to RM 56

n= 2341

















0 200 400 600

Total Length(mm)


Figure 2-11. Length frequency distributions of rainbow trout captured using electrofishing in the
Colorado River from river mile -15 to river mile 56. Each panel represents captures
of fish within the identified river segment.

















55















1990-2002 Daily Mean
17 1990-2002 Lowess Fit o
2003 Daily Mean o
16- -- 2003 Lowess Fit 2
2004 Daily Mean
15 2004 Lowess Fit /, I -o
o 2005 Daily Mean
1 -- 2005 Lowess Fit A
+ 2006 Daily Mean .
1 -- 2006 Lowess Fit -.
13 -

| 12 4 T- t^ ^

E 11







7-

6-

,r-' ) Cj (D -- ") O CA CD D 0) --r-- U-) 0") NC (.0 0) r-- CO CI (0 CD
S- ) SO > > 00 W
D [- CLL3 ] <
Date

Figure 2-12. Daily mean water temperatures observed in the Colorado River at approximately
river mile 61, 1990-2006. Lines indicate locally weighted polynomial regressions
(Lowess) fits to the indicated data set.













56









CHAPTER 3
DEVELOPMENT OF A TEMPERATURE-DEPENDENT GROWTH MODEL FOR THE
ENDANGERED HUMPBACK CHUB USING MARK-RECAPTURE DATA

A primary interest of fisheries biologists is to understand how fishes grow and the

processes and factors that influence that growth. Such information is critical for research

addressing questions about basic ecological relationships such as the tradeoff between growth

and survival, and for management strategies associated with maximizing yield. In the latter case,

growth information is frequently used to populate assessment models with vital rates (Pauly

1980; Beverton 1992; Jensen 1996) and age-specific length, weight, fecundity, and vulnerability

to exploitation (Walters and Martell 2004). Additionally, information on growth may be used to

estimate the age of fish based on size (e.g., Kimura and Chikuni 1987). Given the importance of

understanding growth, much effort has been expended to understand factors that influence

growth, to develop models to describe observed growth patterns, and to estimate the parameters

of those models (Ricker 1975; DeVries and Frie 1996).

Though many models of fish growth have been proposed (Ricker 1975; Schnute 1981),

perhaps the most widely used is the von Bertalanffy model (Bertalanffy 1938). The parameters

of this model are typically estimated by comparing predicted with observed size-at-age or growth

increment data (Fabens 1965; Quinn and Deriso 1999). Obtaining growth increment data is

usually accomplished via mark-recapture studies where the sizes of individual fish are measured

before and after known times at liberty. A significant advantage of using growth increment data

to estimate growth model parameters is that ages of individual fish are not required growth is

simply measured over the times that individual fish have been at liberty. Since determination of

age frequently involves inspection of various calcareous structures (e.g., otoliths) that often

involves sacrificing the animal, use of increment data is preferable when working with

endangered or rare species.









The federally endangered cyprinid humpback chub Gila cypha is endemic to the Colorado

River drainage in the southwestern United States and is generally found in swift, canyon bound

river reaches (Minckley 1973). The Little Colorado River (LCR) population of humpback chub

within Grand Canyon is a focal resource of the Glen Canyon Dam Adaptive Management

Program (Gloss and Coggins 2005). Periodic stock assessments of this population serve as the

core monitoring tool and status metric for this resource. These assessments require accurate age

assignments of fish captured in a long-term sampling program (Coggins et al. 2006a) in order to

employ open population mark-recapture assessment methods that include age-dependent effects.

Due to endangered listing status, longevity, and difficulty determining the age of individual

humpback chub, little information is available on the relationship between size and age for this

species. At present, individual age assignments are based on size and rely on a growth curve

estimated from a limited set (z 60) age-length observations (USFWS 2002). This lack of growth

information promotes uncertainty and possibly bias in length-based age assignment, and this

potential bias has been identified as an area of concern by past external reviews of the humpback

chub assessment program used by the Glen Canyon Dam Adaptive Management Program

(Kitchell et al. 2003).

I used growth increment data to estimate the parameters of a generalized growth model for

the LCR population of humpback chub. This effort is undertaken to supplement the available

information on humpback chub growth and to inform length-based age assignments for stock

assessments. Because the older fish in this population exhibit a potadromous migration between

the seasonally warm LCR and the constant cold mainstem Colorado River within Grand Canyon

(Gorman and Stone 1999), I evaluated ontogenetic temperature-dependent effects in the growth

model. The results of this work should be useful to researchers studying humpback chub or









wishing to estimate temperature-dependent growth models using growth increment data for other

species that make major ontogenetic shifts in thermal habitat use.

Methods

An extensive monitoring program for the LCR population of humpback chub has been

ongoing since the late 1980s (Coggins et al. 2006a). As a result of routinely capturing and

implanting humpback chub with passive integrated transponder (PIT) tags, I was able to compile

over 19,000 growth increments with which to evaluate growth rate (or measurement error in the

case of recaptures made shortly after tagging). The basic technique for estimating growth model

parameters from growth increment data is to predict the amount of growth in the elapsed time

between capture and recapture. Assuming standard von Bertalanffy growth curve predictions of

length at time t and at time t+At, Fabens (1965) developed the most basic model where the

predicted growth increment is given as

AL = L(t + At) L(t) = (L -L(t))l e-kt), (3-1)

where t is time at initial capture, At is the elapsed time between initial capture and recapture,

and L. and k are the asymptotic length and the rate at which length approaches L. respectively

(Quinn and Deriso 1999). Parameter estimates are found by minimizing the difference between

predicted and observed growth increments.

Though this technique has been widely applied, numerous authors have pointed out

resulting parameter estimates will be biased if individual fish exhibit growth variability (e.g.,

Sainsbury 1980; Kirkwood and Somers 1984; Francis 1988). Using this technique, k will

typically be negatively biased and L. will be positively biased. Recognition of these problems

has led to the development of alternative models which attempt to minimize these biases (e.g.,

James 1991; Wang et al. 1995; Laslett et al. 2002). I attempted to estimate standard von









Bertalanffy growth parameters for humpback chub using two of these methods (Wang et al.

1995; Laslett et al. 2002) and generally obtained poor results, characterized by an inability of the

models to predict growth increments exhibited by small fish and large fish simultaneously.

Examination of growth rate as a function of size reveals that the basic problem with fitting a

standard von Bertalannfy model to these data is the lack of a simple linear relationship between

growth rate and length (Figure 3-1) as is implied by this model. It is apparent that the fish less

than approximately 250 mm TL have a larger von Bertalanffy k parameter value (i.e., more

negative slope of the growth rate vs. length plot) than do fish larger than 250 mm TL. These

results suggest a "kink" in the growth curve as would be found if fish grew along one curve

when small and then switched to another when larger.

Because water temperature is a major determinant of basal metabolic rate and hence the

von Bertalanffy k parameter among poikilotherms (Paloheimo and Dickie 1966; Essington et al.

2001), the "kink" hypothesis is consistent with fish that are demonstrating an ontogenetic shift

among habitats that have different water temperatures. For humpback chub, this would be a

transition from the warm LCR spawning and rearing habitat to the cooler mainstem Colorado

River adult habitat (Valdez and Ryel 1995; Gorman and Stone 1999). To account for this

apparent pattern of changing growth rate, I fit growth increment data to a general growth model

(Paloheimo and Dickie 1965) describing the rate of change in weight as

dW= HWd -mWn. (3-2)
dt

Here, the first term describes anabolism (i.e., mass acquisition) and is governed by a term

representing the mass normalized rate at which the animal acquires mass (H), the mass of the

animal (W), and a parameter (d) describing the scaling of anabolism with mass. The second term

represents catabolism (i.e., mass loss through basal metabolism or activity) where m is the mass









normalized rate at which the animal looses mass and n is the scaling factor of catabolism with

mass. Assuming a constant relationship between length and weight over time as

W= aLb, (3-3)

where L is length and a and b are constant, it is possible to derive an analogous relationship for

the rate of change in length as

dL
d = aL8 KL (3-4)
dt

Constants in this relationship are related to those in (3-2) and (3-3) as

ad- 1H
a=d (3-5)
b

a"n n
K=- (3-6)
b

= bd b +1, and (3-7)

7= bn b +1. (3-8)

Essington et al. (2001) review these relationships and describe the derivation of the standard von

Bertalanffy growth function as the integral of equation (3-4) when n=l, b=3, and d=2/3. This is

the situation where catabolism (plus mass loss to reproductive products) scales linearly with

mass, the length-weight relationship is isometric, and anabolism scales as the 2/3 power of mass

resulting in the standard von Bertalanffy growth model:

L(t) = L,(1-e -k(tto)), (3-9)

where to is the theoretical age where body length is equal to zero.

To estimate growth model parameters, I first assumed that measurement errors in the

length of fish are normal with variance oc and that all fish follow a standard von Bertalanffy









growth curve (equation 3-9) with shared k and individual L,. The predicted length offish at

time of recapture can be found by rearranging the Fabens (equation 3-1) as

L(t + At)= L(t)++ (L L(t)) -e e-"'). (3-10)

Assuming that individual L, is normally distributed with variance o- the variance of

eachL(t +At) is

t+At = ( (l+e 2 )+o(2_ (-e-k )2. (3-11)

Deviations between observed and predicted growth increment for individual fish i are given as

D, =L(t +At), L(t), -(L, L(t),)1 e-kAt ). (3-12)

It is then possible to estimate the parameter vector 0={ L, k k, oL, o- } by maximizing the log-

likelihood function:

InL( L(t),L(t + At))= 1 ) (3-13)
2 ,1 L(t+At), 2 ,

where s is the number of growth increments. This is essentially an inverse variance weighting

strategy where growth intervals that have high recapture length variance are down-weighted in

the fitting procedure.

Though this procedure is applicable assuming fish growth is described by equation (3-9), if

fish growth is described by equation (3-4) then there is no analytical solution for -L(t +t) as in

equation (3-11). However, by estimating 0, and in particular k, using equations (3-11 through 3-

13) and assuming that the individual variances computed using equation (3-11) are an adequate

approximation, deviations from the general model (equation 3-4) can be used in the log-

likelihood. These deviations are computed as









t=t,+Atz
Dr = L(t +At), fa(d U t .t (3-14)
t=t,

After specifying the parameters a and b for equation (3-3), estimation proceeds as above with the

parameter vector 0={H, d, m, n,&2L, &m }.

I implemented this procedure in both Microsoft Excel using Solver (Ladson and Allan

2002) and AD Model Builder (Foumier 2000) to obtain estimates of 0. I reduced the parameter

set by specifying a2m= 31.8 mm2 based on an analysis of the observed error between consecutive

measurements of identical fish within 10 days. I also specified the a and b parameters for

equation (3-3) as 0.01 and 3, respectively. To calculate the conditional variance of

eachL(t+ At), I specified k = 0.145 based on previous analyses. Additionally, I included penalty

terms in the log-likelihood equation (3-13) to constrain d and n so that they did not deviate too

far from the theoretical values assuming standard von Bertalanffy growth of 2/3 and 1,

respectively. I evaluated alternative weight values on these penalty terms to find an appropriate

tradeoff between minimum weights and decreased log-likelihood.

Because all the information contained in the mark-recapture data are for fish larger than

150 mm TL, extrapolating results to the growth rate of smaller fish could be problematic.

Fortunately, Robinson and Childs (2001) conducted monthly sampling of juvenile humpback

chub in the LCR during 1991-1994. They used these data to estimate (by modal progression

analysis) average monthly length from age-0 months to age-32 months. I utilized these data in

an additional log-likelihood term to constrain the predicted lengths from the general model to be

similar to those reported by Robinson and Childs (2001). Using these auxiliary data and

assuming normal deviations allowed me to incorporate information on the growth rate of fish









before they are large enough to be implanted with PIT tags. With these constraints in place, the

full log-likelihood is

InL(OL(t),,L(t+ At)) D 1 2
2 oL(t+At),


(d 2 1 (n-12 mos L(i)- (i) (3-15)
2A 3) 22 2 2 _, )

where A is the weighting value for the penalty terms, L(i) is the predicted length in month i

from the general model, and (i) is the predicted length over mos=32 months as reported by

Robinson and Childs (2001). I specified that the variance of the observed lengths -2 was unity.

The weighting term can be interpreted as the prior variance on the standard von Bertalanffy

parameters (d= 2/3, and n=l).

An important logical extension of the general model is to assume temperature dependence

in growth rate. Accounting for changes in growth rate as a function of temperature is likely to be

important for the analysis of this dataset for two reasons. The first is to account for the

differences in growth rate with occupancy in either the LCR or the mainstem Colorado River.

The second is to account for seasonal changes in water temperature within the LCR. The

importance of the second consideration is further magnified by the temporal distribution of

sampling within the LCR. Sampling in the LCR typically occurs in the spring and fall.

Therefore, much of the observed growth increment data corresponds to either summer growth

(i.e., observations of fish captured in spring and again in fall) or winter growth (i.e., observations

of fish captured in fall and the following spring). Because growth rate varies with temperature

(Paloheimo and Dickie 1966), I expect growth increments to be smaller during winter than

during summer. This general prediction is also consistent with both field (Robinson and Childs

2001) and laboratory (Clarkson and Childs 2000) observations of humpback chub.









To allow temperature dependence in equation (3-4), I defined temperature-dependent

multipliers of a and K as

(T 10)
fc(T) =Q 1, (3-16)

(T10)
fm(T)=Qm 10 (3-17)

where f (T) is the temperature-dependent multiplier of a and f,(T) is the temperature-

dependent multiplier of K. The consumption and metabolism coefficients (Q, and Q ) of a Q10

relationship allow these multipliers to increase or decrease with temperature (T). One can think

of these constants as the amount that anabolism or catabolism will change with an increase in

temperature from 100 C to 200 C. Inclusion of these temperature-dependent multipliers into

equation (3-4) yields


dt
= aL f, (T) L f.f (T) (3-18)

Equation (3-18) accounts for growth rate differences as a function of temperature, but does

not account for movement between the two thermal habitats. I used a logistic function to model

occupancy in either the LCR or the mainstem Colorado River. I assumed that the probability of

LCR occupancy is given as

0.8
PLCR =1 (L ,) (3-19)
1+e 20

where L is fish total length and L, is the fish total length where the probability of residing in the

LCR year round is 0.6. The behavior of this model is such that the probability of year-round

LCR residency approaches unity at lengths much less than L, and decreases to 0.2 at lengths

much larger than Lt. The number 20 in the denominator of the exponent governs the rate at









which the probability changes from near unity to near 0.2. Small denominator values cause a

more abrupt transition (i.e., complete the transition over a small length range) while larger values

imply a smoother transition. The asymptote at 0.2 requires at least some LCR residency for even

the largest fish and is consistent with the observation that adult humpback chub use the LCR for

spawning (Gorman and Stone 1999). This function can be considered analogous to the

proportion of the year that fish of given size occupy the LCR.

I then defined a weighted temperature function experienced by fish of a particular length as

T(t)= (PLCR)TL(t)+ (1- PLCR )T(t), (3-20)

where TL, (t) is the time-dependent water temperature in the LCR and TMs (t)is the time-

dependent water temperature in the mainstem Colorado River. This overall temperature

experienced by a fish of a given length is then used in equation (3-18) to predict growth rate

considering time-dependent changes in water temperature and size-dependent changes in LCR

versus mainstem Colorado River occupancy.

To model the time-dependent water temperature in the LCR, I utilized data reported by

Voichick and Wright (2007) to predict average monthly water temperature considering data

1988-2005. I fit these data with a sine curve as

TLCR ( Tve = +(max Tve )sin(2;(t + peak), (3-21)

where t is time in fraction of a year starting April 1, tpeak is a phase shift allowing predicted peak

temperature to align temporally with the observed peak temperature, T,e is the 1/2 amplitude

temperature and roughly corresponds to the average annual temperature, and Tmax is the

maximum annual temperature. I estimated tpeak, 'Tve, and Tmax by minimizing the squared

difference between observed and predicted monthly temperature.









Annual water temperature variation in the mainstem Colorado River near the confluence of

the LCR is much less variable (range 8-12C) than within the LCR (Voichick and Wright 2007).

Thus, I assumed constant water temperature in the mainstem Colorado River of 100 C. This

value corresponds roughly to the average water temperature within the LCR inflow reach of the

Colorado River during much of the time when the growth increments were observed (1989-

2006).

I estimated the parameter vector 0 =H, d, m, n, Qc, L } by maximizing the log-likelihood

equation (3-15). With this more complex model, predicted recapture lengths were found by

integrating the temperature-dependent growth model (equation 3-18) with respect to time. These

predictions were then used in the second term of equation (3-14) to compute the deviations

between observed and predicted growth. Following guidance from a meta-analysis by Clark and

Johnson (1999), I specified Qm as 2 to reduce the parameter set. I specified the weighting term

for the log-likelihood penalties equal to that used in the previous analysis. To further reduce the

parameter set, I specified C- = 2,000 to correspond with a coefficient of variation of about 10%

as is the minimum typically observed in fish populations (S. Martell and C. Walters, University

of British Columbia, personal communication). I compared model fit for the temperature-

independent growth model and the temperature-dependent growth model using AIC techniques

(Burnham and Andersen 2002).

Results

I fit both the temperature-independent (TIGM) and temperature-dependent (TDGM)

growth models to 14,971 observed growth increments extracted from the humpback chub mark-

recapture database. All fish were larger than 150 mm TL and the time interval between capture

and recapture exceeded 30 days. Although greater than 60% of the fish were at large for 1 year









or less, a small fraction of the observations were for much longer time intervals with the longest

interval about 15 years (5,538 days). I estimated the measurement error contained in the dataset

by computing the observed difference in measured lengths of fish captured and recaptured within

10 days. This resulted in a measurement error variance of 31.8 mm2 across all sizes of fish,

implying that most TL measurements were within 11 mm of the true TL. This amount of

measurement error is not unexpected considering the difficulty in measuring live fish. Although

this error rate is rarely reported, it is likely similar across a wide range of fish species and field

conditions when handling live animals.

I fit the TIGM with prior variance weighting terms on the d and n parameters A ={0.00001,

0.0001, 0.001, 0.01, 0.1, 0.5, 1, 10, 100, 1,000, and 10,000} to explore the effect of constraining

these parameters to values near standard von Bertalanffy values. The log-likelihood was nearly

identical for all values of A =0.01 and greater, but reducing A below 0.01 caused large changes

in the log-likelihood. Therefore, I specified A =0.01 as the weighting value for both the TIGM

and TDGM.

To estimate the parameters of the TDGM, I first had to fit the time-dependent LCR water

temperature model. Fortunately, the sine curve function with parameters tpeak =-0.011,

Te =17.9, and Tma =23.2 fit the observed average monthly temperatures very well (Figure 3-2).

The estimated parameters, log-likelihood, and AIC statistics for the TIGM and TDGM are

presented in Table 3-1. The parameter values for the TIGM suggest that anabolism scales as 0.5

mass and catabolism scales as 1.15 mass. These values are different than assumed by the

standard von Bertalanffy model and also result in an average L value that is smaller than would

be predicted from simple inspection of the data. In contrast, the estimated scaling parameters

(d=.61 and n=0.89) for the TDGM are not much different than what would be expected under the









standard von Bertalanffy model where the anabolic scaling parameter (d) should be close to 2/3

and the catabolic scaling parameter (n) should be close to unity. AIC results show strong support

for the TDGM over the TIGM (Table 3-1). However, the parameter correlation matrices for

each of these models show very high correlation, indicating that all of the parameters are not

separately estimable (Table 3-2 and Table 3-3). In situations such as this where the model is not

full rank, it has been shown that the AIC (Burnham and Andersen 2002) is undefined (Viallefont

et al. 1998; Bozdogan 2000) suggesting that the AIC criteria may not be appropriate for this

comparison.

An alternative way to arbitrate among these two models is to simply graphically examine

the model fit to the data. The measured growth rate as a function TL at the start of the interval is

extremely variable, particularly at smaller sizes (Figure 3-3). This variability is not surprising

considering that the rate is measured as a difference between two imprecise length measurements

and, for most measurements, expanded by dividing by a short time increment. It is also apparent

that all three lines differed from a strict linear relationship that would be implied using a standard

von Bertalanffy model, though the TDGM fits are reasonably linear through the portion of the

predicted curves populated with data. The temperature-independent model is somewhat of a

compromise between the temperature-dependent summer fit and the temperature-dependent

winter fit. I also extracted two subsets of seasonal data that had only either summer growth or

winter growth that suggest oscillating growth rate with temperature (Figure 3-4).

Each of the models was used to predict length as a function of age. In addition to the two

models fit above, I also predicted length-at-age using the growth function reported in the

USFWS recovery goals document (USFWS 2002) and length-at-age using the TDGM for a

constant temperature of 100 C (Figure 3-5). This last curve is equivalent to a fish experiencing a









constant 10 C temperature and is a prediction of length-at-age for a fish spending its entire life

in the mainstem Colorado River. Examination of these curves show that the USFWS growth

curve predicted somewhat smaller sizes at young ages and larger sizes at older ages than is

implied by the mark-recapture data. The TIGM and TDGM predict very similar length-at-age

with the exception of ages 10-25. Two features are apparent TDGM predictions: (1) a

temperature-dependent periodic change in growth rate at ages younger than about age-5, and (2)

an apparent "bend" in relationship at approximately age-4. This age corresponds to the length at

transition (Lt) where humpback chub are rapidly shifting from primarily LCR occupancy to

primarily mainstem Colorado River occupancy. A Lt length of 236 mm TL was most strongly

supported by the data and the TDGM (Table 3-1).

Finally, it is informative to utilize the TDGM to predict monthly growth increments as a

function of TL. These predictions based solely on field data can be compared to laboratory

observations of the same or similar species. I plotted growth rate predictions from both the LCR

population and a population that is experiencing constant 10 C temperatures (Figure 3-6). This

latter curve is presented as a prediction of monthly growth rates that would be observed in the

mainstem Colorado River.

Discussion

Growth model parameter estimation is typically accomplished using paired observations of

individual fish age and length (Quinn and Deriso 1999). Obtaining this information often

requires sacrificing the animal so that calcareous structures may be examined to determine age.

The TIGM and TDGM seek to obtain this information through non-lethal sampling using

information that is frequently collected in routine mark-recapture studies. Particularly for

endangered species such as the humpback chub, a non-lethal method to obtain information on

growth is mandatory.









Estimating growth model parameters using increment data is not a recent development.

Indeed, quite complicated frameworks have been developed that consider multiple growth model

formulations and model error in length measurements at both capture and recapture occasions

(Laslett et al. 2002). The model described herein takes a somewhat different approach by

starting with a very general growth model allowing many different functional forms to describe

the relationship between size and age. Additionally, the model parameterization allows intuitive

inclusion of the effect of temperature on anabolism and catabolism that have direct interpretation

in a bioenergetics framework (Essington et al. 2001).

Recent papers to estimate growth of marbled lungfish Protopterus aethiopicus (Dunbrack

et al. 2006) and wels catfish Silurus glanis (Britton et al. 2007) propose methods similar to those

described in this work. Interestingly, this methodology is clearly motivated by a similar problem

- the inability to obtain information on the age of individual fish. However, the methods

proposed in these studies do not explicitly allow growth to be influenced by temperature, even

though this is clearly an issue, particularly in the wels catfish case (Britton et al. 2007).

This study addresses the effect of temperature on humpback chub growth and attempts to

estimate the length at which fish transition from primarily LCR occupancy to primarily

mainstem occupancy. The general implication from my findings is that growth rate will increase

substantially with a temperature increase from 100 C to 200 C as indicated by the values of

Q =4.6 and Q =2.0. These coefficients suggest that anabolism will more than double relative to

catabolism across this temperature range. However, Petersen and Paukert (2005) constructed a

bioenergetics model for juvenile humpback chub and found Qc Qm, 2.4 suggesting much less

potential for increased growth with increased temperature. Though some of the difference in

estimated Qc between my analysis and that of Petersen and Paukert (2005) may be related to the









highly correlated parameters in the TDGM (i.e., may be able to obtain nearly as good a fit with

lower Qc and higher H, d, and m; see Table 3-3.), it is also likely that laboratory observations of

growth rates may not accurately represent field conditions (Rice and Cochran 1984). In

particular, the field estimate of Qc represents not only physiological (laboratory) constraints on

feeding, but also effects of any seasonal variations in food availability that are positively

correlated with temperature (e.g., insect emergence during spring and summer).

Clarkson and Childs (2000) conducted laboratory experiments to evaluate the growth rate

of juvenile humpback chub at 100 C, 140 C, and 200 C. They report monthly growth rates of 1

mm, 13 mm, and 17 mm per month for these temperatures, respectively. Considering the

estimated monthly growth rates from the TDGM in Figure 3-6, the TDGM tends to over-estimate

the growth rates reported by Clarkson and Childs (2000) at 100 C and under-estimate the growth

rate at 200. However, the TDGM results are in overall agreement with this laboratory study.

In their study of reproductive ecology of humpback chub in the LCR, Gorman and Stone

(1999) conclude that adult fish demonstrate a potadromous migration between the mainstem

Colorado River and the LCR to spawn. Based on catch rates of humpback chub within the LCR,

they suggest that fish larger than 300 mm TL remained in the LCR only long enough to complete

spawning activity. They also report that catch rate of fish between 200-300 mm TL declined by

only half following the spawning period. The implication is that fish between 200-300 mm TL

may occupy the LCR for longer periods of time than fish larger than 300 mm TL. The TDGM

estimate of L (236 mm TL) is in agreement with these observations suggesting that fish greater

than 236 mm TL should predominantly reside in the mainstem Colorado River.

This case history should be useful to those studying humpback chub throughout the

Colorado River Basin, and to researchers seeking to estimate the relationship between fish age









and size using non-lethal techniques. This technique shows considerable promise to extract

useful information on fish growth from field data, rather than laboratory studies where such

information is typically obtained.










Table 3-1. General growth model results.


Model H


d m n L 0 L Q


L, AIC


Parameters Rank


TIGM 163 0.52 0.0007 1.15 391 961 -- -- 133,658
TDGM 21.0 0.61 0.46 0.89 434 2000 4.59 236 95,165


AAIC
38,493
0










Table 3-2. Parameter correlation matrix for the temperature-independent growth model.
H d m n
H 1
d -0.99 1
m -0.66 0.73 1
n 0.62 -0.72 -0.99 1
2 0.14 -0.19 -0.38 0.38










Table 3-3. Parameter correlation matrix for the temperature-dependent growth model.
H d m n Q,


0.74
0.88
-0.86
-0.98


0.94
-0.93
-0.82


-0.99
-0.89


1
0.88


L, 0.55 0.16 0.35 -0.34


1
-0.46


















O







O
O
0q


200


300


* Observed growth rate offish < 250 mm TL
SObserved growth rate offish 250 mm TL
...... Predicted growth rate offish < 250 mm TL
Predicted growh rate offish 250 mm TL*









0m 0






o 9









"A" ,


400


500


Total Length at Start of Interval (mm)


Figure 3-1. Predicted and observed growth rate (dL/dt) as a function of total length (TL) at the
start of the time interval. Solid squares are observed growth rate of fish initially
captured with TL < 250 mm and open circles are observed growth rate of fish initially
captured with TL 250 mm. Predicted growth rates are simple linear regressions on
observed growth rate of fish initially captured with TL < 250 mm and of fish initially
captured with TL > 250 mm.


rt"

cl I'
0 c
o;
a8


C
C


C
o


I I I I I I






























C)




E
,- \

(I













0) M (ID 0 (
< E 0 LL

Month

Figure 3-2. Observed and predicted monthly Little Colorado River water temperature. The
points are the average observed monthly temperature and the line is the predicted
monthly temperature.


























Im v:*.-. c .* A






.*1 1 *















STemperature Indeperdent
S- Temperature Dependent Summer
-- Temperature Dependent Winter


200


300


400


500
500


Total Length at Start of Interval (mm)


Figure 3-3. Observed and predicted humpback chub growth rate (dL/dt) from the temperature-
independent growth model and the temperature-dependent growth model during
summer and winter.


C
CD
C\1





CD
CD


C
C





CD
C
o
0J










































0 100 200 300 400 500


Total Length at Start of Interval (mm)


Figure 3-4. Observed and predicted humpback chub growth rate (dL/dt) from the temperature-
dependent growth model during summer and winter.



















C
C





C
o

E
o


o)

0
0 -











STemperature Dependent LCR
C Temperature Dependent Mainstem Colorado River
I I I I I I
0 5 10 15 20 25 30

Age

Figure 3-5. Predicted humpback chub length-at-age from the U.S. Fish and Wildlife Service
(USFWS) growth curve, the temperature-independent growth model, the temperature-
dependent growth model for the Little Colorado River (LCR) humpback chub
population, and the temperature-dependent growth model for humpback chub living
in the mainstem Colorado River under a constant temperature of 10oC.


















Colorado River Growth
S-- LCR Growth








/ I







II
S',,




I r



4 9
I I
I I

II I I I I
i 0


I T I

++ o .


Figure 3-6. Predicted monthly growth rate from the temperature-dependent growth model for the
Little Colorado River (LCR) population of humpback chub and for humpback chub
living in the mainstem Colorado River under a constant temperature of 10C.


CD


(0 -


S-









CHAPTER 4
ABUNDANCE TRENDS AND STATUS OF THE LITTLE COLORADO RIVER
POPULATION OF HUMPBACK CHUB: AN UPDATE CONSIDERING DATA 1989-2006

Effective monitoring to evaluate endangered species status is a vital component of most

endangered species recovery plans (Campbell et al. 2002). Monitoring results are typically used

to evaluate species' status with regard to established goals towards changing its listing under the

U.S. Endangered Species Act. Resource monitoring is also a critical component of adaptive

management (Walters and Holling 1990). In the context of adaptive management, monitoring

not only evaluates the status of the resource, but combined with probing management

experiments it serves to inform managers about how the resource responds to various

management actions. This paradigm of "learning by doing" places a premium on an accurate

and reliable monitoring program (Parma et al. 1998).

Prompted by a National Research Council sponsored evaluation of the Glen Canyon Dam

Adaptive Management Program (GCDAMP; NRC 1999), the Grand Canyon Monitoring and

Research Center (GCMRC) has devoted significant resources to developing long-term

monitoring programs over the last 8 years. As an example, much effort has been expended to

synthesize existing data on fisheries resources in order to portray trends in these populations. Of

particular importance is the humpback chub (Gila cypha), a federally listed endangered cyprinid

endemic to the Colorado River Basin in the southwestern U.S. (Gloss and Coggins 2005).

Because of this species' unique ecological role as one of the few remaining endemic aquatic

species within Grand Canyon and their endangered listing status, the humpback chub is a focal

resource of the GCDAMP (Gloss et al. 2005).

The objective of this chapter is to provide updated information on the status and trend of

the Little Colorado River population of humpback chub in light of new information and refined

assessment methodology. Such information constitutes the cornerstone of the humpback chub









monitoring program within the adaptive management program and is also relevant to U.S. Fish

and Wildlife Service recovery goals for this species (USFWS 2002).

The unique life-history attributes of Little Colorado River population (LCR) of humpback

chub and the large variety of sampling and monitoring programs ongoing since the 1980s

(Coggins et al. 2006a) prompted the development of a new type of age-structured, open

population, mark-recapture model called the age-structured mark-recapture model (ASMR;

Coggins et al. 2006b). This model was subsequently used in combination with other mark-

recapture and index-based assessments to provide a comprehensive assessment of the LCR

humpback chub population (Coggins et al. 2006a). The ASMR approach has been subjected to a

series of independent peer evaluations both as part of the GCDAMP (e.g., Kitchell et al. 2003)

and through peer review as part of the publication process. Since publication of the last

assessment, I have continued development of the ASMR model to address concerns presented in

previous reviews (Kitchell et al. 2003). These improvements include incorporation of a formal

model comparison approach and consideration of the uncertainty in assigning age to individual

fish.

A central problem in conducting humpback chub stock assessment is the assignment of age

to individual fish. Though this problem is ubiquitous in fish assessment programs (Coggins and

Quinn 1998; Sampson and Yin 1998), it is particularly difficult when working with endangered

fish and when sacrificing the animal is necessary to determine age. In this situation, individual

fish ages must be assigned based on fish lengths and assuming some relationship between length

and age. In past humpback chub assessments (Coggins et al. 2006a), I assumed that this

relationship was adequately described from a growth curve based on a limited collection of

paired age and length observations (USFWS 2002). However, this age-length relationship is









based on an extremely small sample size and is therefore suspect. Additionally, when assigning

individual age based on this relationship, I assumed that fish could be aged without error. This is

clearly not a valid assumption and presumes much more certainty in the assignment of age than

is warranted. To alleviate these shortcomings, in Chapter 3, I presented a method to estimate the

relationship between fish age and length using mark-recapture data. In this chapter, I use that

relationship to translate uncertainty in the determination of age from length to uncertainty in

abundance and recruitment estimates from ASMR. These analyses offer insight into the

humpback chub assessment and other monitoring programs for aquatic and terrestrial species

where mark-recapture methodologies serve as the core of the assessment approach, but where

estimated trends in recruitment and mortality are influenced by uncertainty in age assignment.

The ongoing monitoring program for humpback chub in Grand Canyon has varied in

intensity over the years, but the primary sample locations, techniques, and personnel have

remained remarkably consistent (Coggins et al. 2006a). Conducting the annual stock assessment

and continuously evaluating the performance of the assessment through retrospective analyses,

independent peer evaluation, and testing of the model with simulated data all provide insight into

the performance of the model-based on the available data. This comprehensive examination may

prove useful to other adaptive management programs that seek to develop a robust monitoring

component, and in particular to provide insight into: (1) limitations of monitoring alone to assign

cause and effect associated with prescriptive management actions, (2) pathologies associated

with large changes in monitoring protocols, and (3) a realistic assessment of the considerable

uncertainty in results for a rare, elusive, long-lived animal even after many years of intensive

monitoring.









Methods

Monitoring efforts for the LCR population of humpback chub began in 1987 when a

standardized hoop net sampling program was implemented in the lower reaches of the LCR.

During the subsequent 19 years, 4 sampling periods can be generally defined corresponding to

different levels of sampling effort and protocol, particularly in the LCR (Coggins et al. 2006a).

The first period of sampling lasted until 1991 and consisted mainly of limited hoop netting in the

lower 1,200 m of the LCR. Sampling period 2 (1991-1995) involved an intensive sampling

effort in both the LCR and the mainstem Colorado River as part of an environmental impact

study on the operation of Glen Canyon Dam (USDOI 1995). The third sampling period began in

1996 with severely reduced intensities compared to period 2. Finally, beginning in fall 2000 a

period of higher sampling intensity relative to period 3 (but less intensive than period 2) began

and continued through 2006. During each of these sampling periods, humpback chub have been

collected using multiple gear types (by many of the same personnel) including hoop nets and

trammel nets in the LCR, and these same gears plus pulsed DC electrofishing in the mainstem

Colorado River (Valdez and Ryel 1995; Douglas and Marsh 1996; Gorman and Stone 1999;

Coggins et al. 2006a).

Index-Based Metrics

Although index-based metrics (e.g., catch rate) can be unreliable to track trends in

population size with great precision (MacKenzie et al. 2006), these indices are frequently

examined and are potentially useful for comparison to previous assessment efforts. With this

caveat in mind and following Coggins et al. (2006a), I updated two long-term catch rate time

series with data from 2003-2006: (1) hoop net catch rate of humpback chub in the lower 1,200 m

of the LCR, and (2) trammel net catch rate of humpback chub in the LCR inflow reach of the

Colorado River (defined as approximately 9 km upstream and 11 km downstream of the









confluence; Valdez and Ryel 1995). Details about these sampling programs are provided by

Coggins et al. (2006a).

Tagging-Based Metrics

The heart of the tagging-based assessment is the large number of uniquely tagged sub-adult

(150 mm-199 mm total length; TL) and adult (>200 mm TL) humpback chub that have been

captured, measured, and implanted with passive integrated transponder (PIT) tags. Since 1989,

over 19,000 humpback chub have been released with unique identifiers. These data are

maintained in a central database housed at the GCMRC.

Capture-recapture-based methods to assess population abundance and vital rates have been

widely used in fisheries and wildlife studies for well over 50 years, and numerous reviews have

been conducted highlighting the general approaches (e.g., Seber 1982; Williams et al. 2002).

Traditional methods (e.g., Jolly-Seber type methods) generally rely on recaptures of tagged or

marked individuals to estimate abundance, recruitment, and survival. The approach is to create a

known population of marked fish that are repeatedly sampled to obtain time series estimates of

mark rate (i.e., proportion of the overall population that is marked) and the number of marked

fish alive in the population. These estimates are subsequently used to estimate capture

probability, abundance, recruitment, and survival.

Here, I briefly describe the overall ASMR method and refer readers to Coggins et al.

(2006b) for full details. The ASMR model differs from the traditional approach because in

general it contains more structural assumptions through the specification of a population

accounting structure governing transition of both marked and unmarked animals through ages

and time. A standard fisheries virtual population analysis framework (Quinn and Deriso 1999) is

used to annually predict the numbers of marked and unmarked fish available for capture. The

total number of marked fish depends on numbers of fish recently marked as well as previously









marked fish decremented by mortality rate. Numbers of unmarked fish depend on the time series

of recruitment, the numbers of fish marked from those cohorts, and the mortality rate. These

annual predictions of the abundance of marked and unmarked fish are further segregated by age

such that age-specific survival and capture probability may be considered. Parameters are

estimated by comparing predicted and observed age- and time-specific captures of marked and

unmarked fish in a Poisson likelihood framework.

The ASMR model has three different parameterizations (ASMR 1-3) that vary in how the

terminal abundance is estimated and how age- and time-specific capture probability is modeled.

Both ASMR 1 and 2 assume that age- and time-specific capture probability can be modeled as

the product of an annual overall capture probability multiplied by age-specific vulnerability.

This is the common parameterization of fishing mortality in many assessment models under the

separabilityy assumption" (Megrey 1989) and diminishes the size of the parameter set since it is

not necessary to separately estimate each age- and time-specific capture probability. ASMR 1

and 2 further assume that vulnerability is asymptotic with age. As such, vulnerability is assumed

to be unity for fish age-6 and older and estimated only for the younger fish. Finally, annual age-

specific vulnerabilities are assumed to be equal among each sampling period as described above.

Implicit in this assumption is that within a sampling period, annual age-specific capture

probabilities differ only as a scalar value proportional to the annual overall capture probability.

The primary difference between ASMR 1 and ASMR 2 is how the terminal abundances are

calculated. ASMR 1 estimates an overall terminal year capture probability and calculates age-

specific terminal abundances (both marked and unmarked fish) as the ratio of age-specific catch

(both marked and unmarked fish) and age-specific capture probability (i.e., product of the

terminal year capture probability and sampling period 4 age-specific vulnerability). In contrast,









ASMR 2 treats age-specific terminal abundances up to age-13 as individual parameters.

Terminal abundances for subsequent ages are estimated by applying age-specific survivorship to

the age-13 abundance. This difference in formulations decreases the parameter count for ASMR

1 relative to ASMR 2 at the expense of assuming that the vulnerability schedule in the terminal

year is identical to the rest of period 4.

ASMR 3 is the most general model in that it makes no assumption as to the age- or time-

specific pattern in capture probability. The conditional maximum likelihood estimates of age-

and time-specific capture probability are used to predict the age- and time-specific catch of

marked and unmarked fish. Full details of each of the models are provided by Coggins et al.

(2006b).

In addition to the ASMR assessments, I also update the time series of the annual spring

abundance estimates in the LCR. Abundance of humpback chub in the LCR > 150 mm TL was

estimated during the early 1990s and 2001-2006 using closed population models. These models

included the program CAPTURE suite of models (Otis et al. 1978) and simple Chapman

modified Lincoln-Petersen length-stratified models (Seber 1982). The recent estimators use data

collected annually during two sampling occasions in the spring. Full details of the sampling and

estimation methods are provided by Douglas and Marsh (1996) and Coggins et al. (2006a).

Coggins et al. (2006b) recommended exploring the use of individual capture histories

within the ASMR framework to reduce confounding between capture probability and mortality.

Though the updated ASMR models presented in this chapter do not yet incorporate individual

capture histories, they do model recaptured fish by annual tagging cohort with the intent of

reducing parameter confounding by increasing the number of observations available for

parameter estimation. In the non-tag cohort or pooled version of ASMR described by Coggins et









al. (2006b) and above, age- and time-specific predictions of recaptured fish are not separated by

year of tagging. As an example, assume that ASMR 3 predicts that 50 marked age-6 humpback

chub should be captured in 2002. These 50 fish could be comprised offish tagged as: age-5 in

2001, age-4 in 2000, age-3 in 1999, or age-2 in 1998. However, as the model is currently

formulated, all age-6 fish recaptured in 2002 are pooled for a single observation. Assuming that

the age- and time-specific captures of marked and unmarked fish are Poisson distributed, the log-

likelihood ignoring terms involving only the data is computed as

A T A T
InL(On ,r)= I [-ma,t + ,t ln(l,^)]+ [- +,ln(,)], (4-1)
a=l t=l a=l t=2

where ma,t is the observed number of age a unmarked fish captured in year t, mi,, is the predicted

number of unmarked fish captured, ra,t is the observed number of marked fish captured (i.e.,

recaptures), r,, is the predicted number of marked fish captured, and 0is the parameter vector to

be estimated. Notice in the second term that the individual log-likelihood terms are summed

over age and time. However, it may be more informative to stratify the recapture data by tagging

cohort. The proposed log-likelihood is then

A T T TT-1
lnL(Om,r)= h [- ,,t +mtln(h,,^t)]+ [- atc+rtc In (4-2)
a=l t=l a=l t=2 c=l

where c is the tag cohort (i.e., all fish marked in year t). In principle, this stratified log-

likelihood should provide additional information on time-specific capture probability and may

improve parameter estimation.

Evaluating Model Fit

Following Baillargeon and Rivest (2007), I used standardized Pearson residuals of

observed and predicted age composition for both unmarked and marked fish to evaluate model fit









among the three different ASMR models. The standardized Pearson residual is the difference

between the observed and predicted values scaled by an estimate of the standard deviation as

Oa.,t Pa.,t (4-3)
r,= (4-3)
p (1- p,,)



where n, is the number of observations (e.g., the number of marked fish recaptured each year)

and o,, and p,,, are the proportions of fish in each year and age class observed and predicted,

respectively. I plotted the individual Pearson residuals for each combination of age and time to

look for consistent bias for individual brood year cohorts. Additionally, I used Quantile-Quantile

(Q-Q) plots to compare the distribution of the Pearson residuals to a theoretical normal

distribution. The slope of the theoretical curve is approximately the standard deviation of the

distribution of Pearson residuals where a small value of the slope indicates a narrow distribution

of the residuals. Deviations from the theoretical curve indicate a non-normal distribution of the

Pearson residuals and imply that the model error is not well distributed (e.g., tending to more

often either over- or under-predict age proportions) and possibly inducing bias in parameter

estimates.

In addition to examination of model fit using Pearson residuals, I chose to also rely on

information theory to aid in model evaluation. This approach is increasingly common in

ecological studies to arbitrate among competing models and is primarily concerned with

estimating the Kullback-Leibler (K-L) distance between the model and the "truth" as a measure

of model support (Burnham and Andersen 2002). The Akaike Information Criterion (AIC;

Akaike 1973) is the standard estimator for the relative K-L distance and is computed as a

function of model likelihood and number of model parameters. Following external review of the

ASMR method, Kitchell et al. (2003) pointed out that although ASMR uses a quasi-likelihood









structure of estimating equations and true likelihood, estimates of relative K-L distance using

AIC, though not strictly appropriate, would be valuable to consider for model selection.

Therefore, in addition to the evaluation based on Pearson residuals, I also conducted an AIC

evaluation.

Incorporation of Ageing Error in ASMR Assessments

As mentioned above, Coggins et al. (2006a) assigned age to individual fish strictly as an

inverse von Bertalanffy function. This procedure ignores variability in the age of fish of a

particular length and tacitly assumes that age assignments can be made much more precisely than

is true. To account for uncertainty in the assignment of age using length, I estimated the

probability of age for fish having length within a particular length interval P(a l). Following

methods reported by Taylor et al. (2005), I define this procedure by first specifying that the

probability of an age a fish having length within length bin / is


P(la) = 1 exp- dl, (4-4)
Ua 2 1 d 2"a2

where length bin / has mid-point length 1, minimum length 1-d, and maximum length I+d. These

probabilities can be thought of as a matrix with rows corresponding to length bins and columns

as ages. As is obvious from equation (4-4), entries within a particular column (age) can be

thought of as resulting from the integral over each length bin of a normal probability density with

mean la and variance Ua The mean length-at-age is computed from the temperature-

dependent growth model (Chapter 3) and the variance of length-at-age is 'a2 = 1cv1 assuming

that coefficient of variation in length (cv, ) is constant among ages.









With P([la) available, one option to compute P(a I) would be to normalize each matrix

cell by the sum of its row as:

/P(l1)
pLaI/= -^ (4-5)

1=1

However, Taylor et al. (2005) suggest that this procedure will induce bias if the population has

experienced size-dependent mortality such as size selective fishing or size-dependent changes in

natural mortality. This results because within a particular age class, fast growing individuals

(i.e., large L ) may experience either higher or lower mortality rate than their cohorts, and

therefore be either over- or under-represented in the population. This "sorting" by growth rate

can favor either slow growing individuals, as in the case of increasing vulnerability to

exploitation with size, or fast growing individuals, as in the case of reduced natural mortality

with size. Therefore, Taylor et al. (2005) suggest that an adjustment for mortality must be made

to accurately predict the proportion of individuals in each age and length bin. Accordingly, I

define the numbers of fish in each age and length bin as

N, N = NP(la), (4-6)

where N, is the abundance of fish at each age. If the age specific mortality rate (Ma) is

available and recruitment (R) is assumed constant, abundance-at-age is given by


N, =Re 1 (4-7)

With abundance at each age and length bin thus available, the proportion in each age and length

bin can then be calculated as


P = 0 ,(4-8)
N T









where N, = I N,, The probability of age given length is then calculated as
l a


P(a )=- (4-9)

1=1

Taylor et al. (2005) focus on age-specific mortality driven by vulnerability to exploitation.

For the unexploited humpback chub, age-specific mortality as a function of changes in natural

mortality was included. Lorenzen (2000) demonstrated that much variation in natural mortality

can be explained by size of fish. Thus, Lorenzen's allometric relationship between natural

mortality and length was used to calculate a declining mortality rate with age as


M = L (4-10)
'a

where M. is the mortality rate suffered by an adult fish of size L This mortality schedule was

calculated with M, specified as 0.148, as estimated by ASMR 3 considering tag-cohort specific

data (see Results).

I computed four seasonal P(a 1) matrices that I used to assign age to fish captured at

different times of year. Growth during the year could thus be accounted for by

recalculating P(la) such that length-at-age was computed as either 1(a), l(a+.25), l(a+0.50), or

l(a+. 75). The resulting seasonal P(a /) matrices were then used to assign age to a fish

depending on the quarter of the year in which it was captured.

To incorporate the uncertainty in assigning age based on length into the overall

assessment, I used a Monte Carlo procedure where age was stochastically assigned to each fish

based on the seasonal P(a /) matrices. To understand this procedure, it is first helpful to

recognize that given a fish with length in bin 1, the probability of belonging to each age is









multinomial with number of categories equal to the number of ages. I used the multinomial

random number generator within program R (R Development Core Team 2007) to randomly

assign age to marked fish. Recapture age was calculated as the sum of age-at-tagging and time-

at-large.

I then stochastically assigned age to each fish using this procedure. For each resulting

dataset of captures- and recaptures-at-age, I estimated adult (age-4+) abundance and recruitment

using ASMR 3. Additionally, I used AD Model Builder (Fournier 2000) to compute the 95%

profile confidence interval for adult abundance and recruitment. I repeated this procedure to

generate and analyze 1,000 datasets (i.e., Monte Carlo trials).

Results

Index-Based Assessments

Between 1987-1999 and 2002-present, the Arizona Game and Fish Department sampled

humpback chub using hoop nets in the lower 1,200 m section of the LCR. Examination of this

index suggests that the abundance of both sub-adult (150-199 mm TL) and adult (> 200 mm TL)

humpback chub declined during 1987-1992 and remained relatively constant through much of

the 1990s (Figure 4-1). Since 2003, there is a slight upward trend in the catch rates of sub-adult

fish. Note that several data points in this index are shifted slightly relative to those reported by

Coggins et al. (2006a). This adjustment is due to additional standardization of the data used to

construct this index (D. Ward, Arizona Game and Fish Department, Personal Communication).

The trammel net catch rate of adult abundance in the LCR inflow reach of the Colorado River

suggests a similar trend in adult fish (> 200 mm; Figure 4-1). In general, this index shows a

stable to declining trend through the 1990s with a slight indication of increased abundance in

most recent years. All monthly trammel net samples from the LCR inflow reach for 1990-2006

are presented in Figure 4-1. However, only samples from 1990-1993, 2001, and 2005-2006 (i.e.,









dark circles in Figure 4-1) represent robust sampling coverage throughout the entire reach.

Annual sample sizes in 1994-2000 and 2002-2003 (i.e., hollow circles in Figure 4-1) were

between 2% and 50% of the 1990-1993 average sample size, and in some years effort was

focused near the LCR confluence where humpback chub density may have been highest. Thus,

the 1990-1993, 2001, and 2005-2006 data are likely to best depict the overall trend of relative

abundance within this reach. Simple linear regression analyses provide estimated slopes that are

not significantly different from zero (p = 0.16 for all data, andp= 0.26 for the preferred data;

Figure 4-1).

Tagging-Based Assessments

As described above, the data required for the ASMR models are numbers of fish marked

and recaptured each year and for each age. For the results contained in this section, all ages are

assigned based on the standard von Bertalanffy growth curve as described in Coggins et al.

(2006b). With that in mind, examination of the age distribution of fish marked and recaptured

since 1989 provides insight into the trends in sampling effort, and also provides important

information related to humpback chub mortality (Figure 4-2). The top panel of figure 4-2 shows

the numbers of newly marked fish and is influenced by both trends in sampling effort and

numbers of un-marked fish alive. The most consistent period of sampling has been since about

2001 with about 1,100 fish marked annually (numbers of fish collected at the top of the bubble

columns). Because a large fraction of the population was marked in sampling periods 1 and 2,

the majority of fish marked in recent years are young fish and the number of new fish marked

each year declines with fish age. The bottom panel of figure 4-2 represents the numbers of fish

recaptured each year. Some of the same patterns related to sampling effort are evident, but there

are some very interesting patterns that result from the high sampling effort in the early to mid

1990s (Figure 4-2). For example in 1995, a total of 1,244 humpback chub were collected and









902 of these fish had been marked in previous years. This pattern is evident for several years of

data indicating that the high sampling effort in the early 1990s resulted in marking upwards of

70% of the humpback chub population. It is also apparent that because few age-3 to age-5 fish

were marked during period 3, there were few age-8 to age-10 fish recaptured in the early 2000s

five years later. This contributes to the "spoon" shapes to the lower panel of figure 4-2, where

there were relatively large numbers fish < age-10 recaptured in period 4, lower catches of age-10

to age-15, and relatively stable numbers of fish > age-15.

Another finding is the extreme longevity of these fish. This is evident by examining the

number of humpback chub of each age marked in each year and recaptured in subsequent years

(Figure 4-3a through Figure 4-3e). Figure 4-3a shows the number of humpback chub of each age

marked in 1989-1992 and recaptured in subsequent years. There is a remarkable number of old

fish (> age-15) first marked in the early 1990s that continue to be recaptured into 2006. This

slow decay pattern of marked fish demonstrates the low mortality rate suffered by older

humpback chub.

Closed Population Models

The time series of abundance estimates for humpback chub >150 mm TL in the LCR

during spring implies a decline in abundance from the early 1990s to present (Figure 4-4).

However, as is apparent in the data, these estimators are very imprecise with corresponding poor

ability to detect significant trends. Additionally, preliminary analyses of data collected in this

program suggest that the 2007 estimate may be up to twice as large as the 2006 estimate (R. Van

Haverbeke, U.S. Fish and Wildlife Service, Personal Communication).

ASMR Without Tag Cohort Specific Data

The three ASMR formulations generally agree that adult (age-4+) humpback chub

abundance has been gradually increasing since about 2001 (Figure 4-5). For the three models,









the 2006 adult abundance estimate is 6,690 (95% CI 6,403-6,994), 6,768 (95% CI 6,397-7,131),

and 6,648 (95% CI 6,222-7,102) for models ASMR 1, ASMR 2, and ASMR 3, respectively.

These results suggest that this population has increased from an estimated low of approximately

4,800-5,000 during 2000-2001. Estimated recruitment (age-2) among models is also in

agreement (Figure 4-6). Following low recruitments for brood years during the early 1990s, all

the models suggest that recruitment increased through the latter part of the 1990s. The biggest

discrepancy among the three models is that ASMR 1 suggests a decline in recruitment following

the 2001 brood year, while the other two models suggest stability. The structural assumptions of

model ASMR 3 (see Coggins et al. 2006b) do not permit a reliable recruitment (age-2) estimate

for brood year 2003. An additional difference in the models results are the estimates of

instantaneous adult mortality (M, ) where adult mortality ranges from 0.119 (ASMR 1) to 0.133

(ASMR 3).

Model Evaluation and Selection

With these results in hand, the question becomes which model is best? Stated another way,

which model produces results most consistent with or best supported by the data? The

discrepancies among model results related to adult abundance are not large, so from a

management or conservation perspective, selecting the "best" model is probably not critical.

However, the models do suggest rather different recruitment trends. Model ASMR 1 supports

the hypothesis that recruitment has declined following the 2000 brood year, while the other two

models suggest relative stability. Therefore, selecting which model is most consistent with the

data is desirable. The patterns in Pearson residuals for both ASMR 1 (Figure 4-7) and ASMR 2

(Figure 4-8) demonstrate systematic lack of fit for particular sets of cohorts. This is best seen in

the recapture residuals where it is apparent that there were more fish observed than predicted for









about eight pre-1990 cohorts, particularly for observations after 2000. Additionally, there were

fewer recaptures associated with the 1992 cohort than predicted. These systematic trends likely

impose bias in the model results for ASMR 1 and ASMR 2. In contrast, there is much less

systematic lack of fit in the residual patterns for ASMR 3 (Figure 4-9). Among the three models,

the Pearson residual standard deviation was smallest for ASMR 3.

The finding that ASMR 3 has the best fit among the three models is not surprising since it

has the largest parameter set. Although ASMR 3 only varies 13 parameters in the direct numeric

search, the conditional maximum likelihood estimates are used for each age- and time-specific

capture probability (Coggins et al. 2006b). Therefore, and assuming a liberal maximum

longevity of 50 years, ASMR 3 has 895 parameters. The question then becomes whether these

additional parameters are justified. To provide insight into this question, I estimated relative K-L

distance using AIC (Table 4-1). These results strongly indicate that model ASMR 3 is superior

to ASMR 1 and 2.

Since the fundamental difference between ASMR 1-2 and ASMR 3 is the amount of

flexibility in age- and time-specific capture probabilities, I examined the pattern in ASMR 3

estimated capture probabilities (Figure 4-10). The patterns in age-specific capture probabilities

during sampling period 2 (i.e., 1991-1995; heavy gray lines) and sampling period 4 (i.e., 2000-

2006; heavy black lines) differ markedly. These findings suggest that there was a major shift in

the gear selectivity; sampling since 2000 appears to be much less effective at capturing fish

between ages 9-20 than was sampling during the second period. Since structural assumptions in

ASMR 1 and ASMR 2 require that vulnerability is asymptotically related to age, it is not

surprising that these models are not able to account for this unexpected pattern, and thus display

poor model fit.









ASMR with Tag Cohort Specific Data

In addition to repeating the analyses by Coggins et al. (2006a) above, I also fit the ASMR

models to the tag cohort specific data using the log-likelihood in equation (4-2). The trends in

adult abundance and recruitment are similar to those found using the simpler log-likelihood

(Figures 4-11 and 4-12). In general, adult abundance estimates are slightly higher at the

beginning of the time series and slightly lower at the end. Adult abundance estimates for 2006

were 6,057 (95% CI 5,797-6,308), 6,138 (95% CI 5,842-6,458), and 5,893 (95% CI 5,554-6,242)

for the ASMR 1, ASMR 2, and ASMR 3 models, respectively. Adult mortality (M ) estimates

from the models fit to the stratified data indicate slightly higher adult mortality than when fit to

the pooled data and ranged from 0.128 (ASMR 1) to 0.148 (ASMR 3). This finding is consistent

with the more rapid decay observed in the time series estimates of adult abundance.

Model Evaluation and Selection

Examination of Pearson residuals for the tag cohort specific models suggests similar

patterns in model misspecification for ASMR 1 and ASMR 2 (Figures 4-13 and 4-14) relative to

ASMR 3 (Figure 4-15). As with the pooled tag cohort data, ASMR 3 displays better fit. Model

evaluation using AIC methods again suggests that ASMR 3 is preferable (Table 4-2) in general

agreement with the residual evaluation. Finally, the estimated capture probability from ASMR 3

suggests a similar mechanism to explain the poor performance of models ASMR1 and ASMR 2

(Figure 4-16) as was found for the without tag cohort specific analysis.

Incorporation of Ageing Error in ASMR Assessments

I used the temperature-dependent growth model (Chapter 3) and the procedures identified

above to construct seasonal P(a I ) matrices. I then plotted the resulting probability

distributions as surfaces to allow examination of the uncertainty in predicting age given length









(Figure 4-17). The most obvious feature of these probability surfaces is the increasing

uncertainty in age assignment with increasing length. For instance and considering the April-

June P(a | /) surface (Figure 4-17), one can see that a 150 mm TL fish is age-2 with highest

probability, but there is some chance that it is all ages between age-1 and age-4. In contrast, a

300 mm fish is approximately age-7 with highest probability but could be as young as age-4 or

as old as age-18. It is precisely this uncertainty that I sought to incorporate in the assessment.

I stochastically assigned age to each fish using the appropriate P(a 1/) matrix depending

on the time of year the fish was first captured. Using this procedure, I generated a total of 1,000

input datasets and fit the ASMR 3 model to each. For each model fit, I retained the estimated

annual adult abundance and 95% profile likelihood confidence bounds. I also retained the

estimated brood year recruitment and 95% confidence bounds. Note that because of the

uncertainty in assigning age to even the smallest fish, newly tagged fish had the possibility of

being assigned age-1. As a result, I expanded the age range of the model such that recruitment

estimates were for age-1 fish.

Estimated adult abundance (age-4+) from model ASMR 3 ranged from 9,322 (95% CI

8,867-9,799) in 1989 to 6,017 (95% CI 5,369-6,747) in 2006 (Figure 4-18). As expected, these

estimates have lower precision than those from ASMR 3 ignoring ageing error. The coefficient

of variation in adult abundance estimates considering ageing error ranges from approximately

1%-7% in contrast to 0.5%-3% if uncertainty in assignment of age is ignored. The recruitment

trend considering the new growth function and the incorporation of ageing error is much less

precise than when ageing error is ignored (Figures 4-12 and 4-19). Although the point estimates

from the two models are in agreement that recruitment has been increasing since about the mid









1990s, the uncertainty in the recruitment estimates from the latter assessment makes statements

about differences among years quite tenuous.

Results Summary

The adult portion of the LCR humpback chub population has increased in recent years as a

result of increased recruitment particularly associated with brood years 1999 and later. Model

evaluation procedures indicate that the results from model ASMR 3 are most consistent with the

data. Utilizing data stratified by tagging cohort appears to add little additional information to the

assessment as indicated by overall similarity in abundance and recruitment estimates and residual

analyses considering both pooled and stratified data. Inclusion of ageing error increases the

uncertainty about individual annual estimates, but gross trends remain the same.

Discussion

The overall result of the mark-recapture-based open population model assessment is that

the adult portion of the LCR humpback chub population appears to have increased in abundance

since 2001. The assessment model best supported by the data is ASMR 3 with a corresponding

2006 adult abundance estimate of approximately 5,900-6,000 fish. In addition, this model

suggests that there has been an increase in the adult abundance of approximately 20%-25% since

2001. This increase appears to be related to an increasing recruitment trend beginning perhaps as

early as 1996, but likely no later than 1999. Recruitment of juvenile humpback chub since 2000

appears stable, but the precision of these estimates is quite low when ageing error is included in

the assessment.

The LCR hoop net abundance index suggests a modest increase in the abundance of

juvenile fish and stability in the abundance of adult fish. In addition, the LCR inflow reach

trammel net abundance index indicates stability with a slight indication of increased abundance

in 2005 and 2006. Though there would be increased confidence in the mark-recapture-based









open population model results if the catch rate metrics indicated similar trends, it is not

surprising that these index measures are not able to detect a 25% increase in abundance. The

basic assumption of catch rate indexes is that capture probability must remain constant for the

index to be well correlated with abundance (MacKenzie and Kendall 2002). There is good

reason to suspect that this assumption is violated for the index data series presented in this update

due to the influence of turbidity on catchability (Arreguin-Sanchez 1996). Turbidity appears to

influence humpback chub catchability in the Little Colorado River (Dennis Stone, U.S. Fish and

Wildlife Service, Personal Communication) and turbidity varies greatly in the mainstem

Colorado and Little Colorado Rivers as a function of tributary freshets and dam operations.

A more significant concern is the lack of correlation between ASMR 3 results and the

mark-recapture closed population model estimates in the LCR. However, since the number of

fish in the LCR during sampling is influenced by migration magnitude and timing, this source of

variability may obfuscate expected correlations with the ASMR 3 results. It is also clear that the

low precision of these annual closed population model estimates may not permit detection of a

25% increase in adult abundance. Additionally, preliminary analyses of data collected during

2007 suggest that the abundance estimate for 2007 may be twice as large as the 2006 estimate

(R. Van Haverbeke, U.S. Fish and Wildlife Service, Personal Communication). Though this

result would provide support for the ASMR 3 results, it would also call into question the ability

of the LCR program to provide a consistent measure of overall population size. One would have

to reconcile whether that level of change was related to a very large age class entering the

sampled population, a larger than normal fraction of the population entering the LCR during the

sampling period, or some other factor.









Though the GCDAMP is fortunate to have such a large mark-recapture database for these

high-profile endangered animals, significant changes in sampling protocol over time continue to

cause ambiguity in these analyses. As identified by Melis et al. (2006), retrospective analyses of

the data suggest a continual updating of the adult mortality rate as additional information has

been collected since 2000. Following addition of the 2006 data, this updating is again apparent

(Figure 4-20). It appears that adult mortality rate may be stabilizing as more data are collected,

but it is difficult to be certain. I believe that the likely cause of this updating is the sampling

program essentially having to "catch-up" following the low sampling effort during period 3.

When focused analysis of this dataset began with open population models in 2000 (GCMRC

unpublished data), there had been so little sampling in the mid to late 1990s that the models

interpreted the lack of old fish captures as a relatively high adult mortality rate. As additional

data was collected through a more rigorous sampling program during 2000-2006, each time the

model "saw" a recent old fish recapture, mortality rate was adjusted downward. The hope is that

if the GCDAMP continues with a fairly uniform sampling program over time, adult mortality

rate will stabilize and only abundance estimates in the last few years of the dataset will be

subject to much updating.

An additional finding, identified by Martell (2006) and in this assessment, is the major

change in gear selectivity between periods 2 and 4. I am uncertain what is driving the trend to

lower the capture probability for the middle aged fish. However, it has been suggested that the

high capture probability for middle aged fish was due to extensive trammel netting effort in the

LCR inflow reach of the Colorado River during period 2. It is apparent from Figure 4-1 that

there is a large difference in the amount of trammel netting effort in period 2 versus period 4. To

investigate this possibility, I fit the ASMR 3 model to a subset of the database containing only









LCR data. The results indicated an almost identical pattern in age-specific capture probability as

is observed with the full dataset. This is not surprising since fish captured in the LCR inflow

reach of the Colorado River represent only about 10% of the entire humpback chub mark-

recapture database. It has also been suggested that reducing the use of large hoop nets in the

LCR during period 4 has reduced the catch rate of the largest fish. Though this is possible, the

net throat openings are the same size on all nets used during both sampling periods. It has also

been suggested that sampling in the LCR only 4 months of the year during period 4 as opposed

to 10 to 12 months of the year during period 2 may be the cause. This is also possible,

particularly if there is some differential migration timing for the middle aged fish relative the

oldest individuals.

Large changes in sampling protocol should be approached with caution in light of how

those changes may affect ability to infer population change. This is particularly true for

populations that are in low abundance and individuals difficult to capture. I suggest that careful

simulation of considered changes may help to expose potential problems or, at the very least,

help to clarify thinking related to proposed changes in sampling protocol. Finally, those

considering implementing a mark-recapture-based monitoring program should plan to expend

considerable sampling effort using similar protocols for the duration of the monitoring program.

I echo the recommendation of Williams et al. (2002) that the objectives of the monitoring

program with regard to issues such as precision of measured quantities should not only be clearly

identified, but that the measured quantities should be directly linked to the management

objectives.

A major criticism of the ASMR technique as previously applied is that it does not

explicitly account for uncertainty in the assignment of age to individual fish (Kitchell et al.









2003). As a result, abundance, recruitment, and mortality estimates may contain excessive bias.

Additionally, estimates of precision are likely overstated by not incorporating this important

source of uncertainty. This analysis attempts to address these concerns by incorporating

uncertainty from age assignments into estimates of abundance and recruitment. Coggins et al.

(2006b) conducted sensitivity analyses on the effect of random ageing error and found little

systematic bias in reconstructed recruitment trends. However, the current analysis is a more

rigorous treatment of the problem and has two major implications.

First, model results of estimated adult abundance are still very precise even when

uncertainty in the assignment of age is accounted for in the assessment. Following review by

Kitchell et al. (2003), this assessment lends additional credibility to results from ASMR

indicating that it provides a rigorous measure of the state of the adult portion of the Little

Colorado River humpback chub population. I recommend that this assessment be considered

"best available science" for use in contemplating management decisions both within the

GCDAMP and the U.S. Fish and Wildlife Service.

Second, this analysis points out the difficulty that open population models have generally

in the precise estimation of recruitment (Williams et al. 2002; Pine et al. 2003). Because many

of the most critical management questions for humpback chub center around how best to

improve humpback chub recruitment, particularly considering improved rearing conditions in the

mainstem Colorado River, it will be difficult for ASMR to detect statistically significant changes

in recruitment unless those changes are quite large. As a result, design of experimental adaptive

management actions intended to increase recruitment should consider first and foremost how to

achieve large changes in recruitment. Small scale experimental treatments of short time duration

or so called "mini-experiments" should be summarily discounted recognizing that the monitoring









program is unlikely to detect small recruitment change even if it occurs. Additionally, multi-year

experiments should be strongly favored in order to help offset not only unexpected and

uncontrollable effects, but the low precision in recruitment estimates.









Table 4-1. AIC model evaluation results among ASMR models fit to data pooled among tag
cohort.
Model AIC # Parameters Rank AAIC
ASMR1 -216274 18 3 2492
ASMR2 -217132 30 2 1634
ASMR3 -218766 895 1 0









Table 4-2. AIC model evaluation results among ASMR models fit to data stratified by tag
cohort.
Model AIC # Parameters Rank AAIC
ASMR1 -196278 18 3 2577
ASMR2 -197183 30 2 1672
ASMR3 -198856 895 1 0












0
a))-
0



," -
Sr
0

(D
Z 0
o 0
0
I

0
0 -
0





0


ro


~DCD
-c o



z

E 0
E
Fo


---- 150-199mm
- 200 mmTL











// I-


I I I I I I I I I I I I I I I I I I I I
r- O0 C CD C\1 CO LC c(D r--- O CD o cN 0o i Lc co
OD O3 O0 CD CD CD CG CM CD CD CD CM CM O O O 0 0 O O
CD CO ) C CD0 CD0 G CD C O C 0 0 CO O O 0 0 O O


B




00
o

0 0










(A C 'T 8D cD r- CO 0 0 n ci c' CD r-
)-- ", 0) ) .. D Gi---- 0 G 0 CD C CD CD D CD CD CD
CD 0) 0) C0 :) 0) 0) 0) CCD 0 0 0 0 0 0 0 0 0
0- 0 0 C0 0- 0- 0 0 0 0 0 0 0 0 0 0 0
CI CI ,:" c-I ",I ,"1 C' ,I


Year


Figure 4-1. Relative abundance indices of sub-adult (150-199 mm total length; TL) and adult
(>200 mm TL) humpback chub based on hoop net catch rate (fish/hour) in the lower
1,200 m section of the Little Colorado River (A) and trammel net catch rate
(fish/hour/100 m) of adult humpback chub in the Little Colorado River inflow reach
of the Colorado River (B). Error bars in panel (A) are 95% confidence intervals. In
panel (B), the solid line represents a regression model fit to the subset of data
representing robust sampling (solid circles) and the dashed line represents a
regression model fit to the entire dataset (all circles).








110


A




11


1W













30 -* :
30
2 0 :
< : ; : i '
II



1990 1995 2000 2005

Year


40 n o CO CO O L C O 0D C ; -

30 -

< 20

10 -


1990 1995 2000 2005


Year


Figure 4-2. Numbers of humpback chub marked (A) and recaptured (B) by age and year. The
annual sample size is indicated by the number at the top of each bubble column and
the distribution among ages indicated by relative size of bubbles within each column.














40 40 -


30 ..o 30 a *


60
100 .00 2


1 10.. o10 '



1990 1995 2000 2005 1990 1995 2000 2005

Year Year



B D

40 40 -


30 30- : c. a
0 0 Io;j 00 .
2 6o-': --,- j;


10 01 .1




1990 1995 2000 2005 1990 1995 2000 2005

Year Year

Figure 4-3a. Numbers of fish marked by age in years 1989 (A), 1990 (B), 1991 (C), and 1992
(D) indicated by dark circles and subsequently recaptured (light circles) by age and

years. The annual sample size is indicated by the number at the top of each bubble
column and the distribution among ages indicated by relative size of bubbles within
each column.
each column.




























1990 1995 2000 2005


Year


1990 1995 2000 2005

Year


ooo oo8,2 ,.
o o 8 "





1990 1995 2000 2005

Year



D















1990 1995 2000 2005

Year


Figure 4-3b. Numbers of fish marked by age in years 1993 (A), 1994 (B), 1995 (C), and 1996
(D) indicated by dark circles and subsequently recaptured (light circles) by age and
years. The annual sample size is indicated by the number at the top of each bubble
column and the distribution among ages indicated by relative size of bubbles within
each column.
















40 -


30 -


Year


i i i





1990 1995 2000 2005

Year



B









O O

N- o o 0-





1990 1995 2000 2005


Year


Figure 4-3c. Numbers offish marked by age in years 1997 (A), 1998 (B), 1999 (C), and 2000
(D) indicated by dark circles and subsequently recaptured (light circles) by age and
years. The annual sample size is indicated by the number at the top of each bubble
column and the distribution among ages indicated by relative size of bubbles within
each column.















114


8
8 o o
O



': C1 -




1990 1995 2000 2005

Year



D






0
0

0
0




0







1990 1995 2000 2005


30 -












A C

40 40 -


30 30 -



0* 10






1990 1995 2000 2005 1990 1995 2000 2005

Year Year



B D







4 20 20 :
i 0 o o .




















10 10
ii n o


























1990 1995 2000 2005 1990 1995 2000 2005

Year Year

Figure 4-3d. Numbers of fish marked by age in years 2001 (A), 2002 (B), 2003 (C), and 2004
(D) indicated by dark circles and subsequently recaptured (light circles) by age and
years. The annual sample size is indicated by the number at the top of each bubble
column and the distribution among ages indicated by relative size of bubbles within
each column.














40-


30 -


< 20 -


10









B

40 -


30 -

30 -


10-



1990 1995 2000 2005

Year

Figure 4-3e. Numbers offish marked by age in years 2005 (A) and 2006 (B) indicated by dark
circles and subsequently recaptured (light circles) by age and years. The annual
sample size is indicated by the number at the top of each bubble column and the
distribution among ages indicated by relative size of bubbles within each column.





















7



6-



S5



8 4

r0
m
,.-o




2 3





1 -






c CI 7 LD o oo 00 ) 0 D C-. O0 i'D (D
0) 0,) 0) 0) 0) O) 0) C CD C C C C C CO
0-) 0) 0') CD C) 0) 0) 0) 0 0 0 0 0 0 0
CN CN C". C\ CN CN CN
Year

Figure 4-4. Mark-recapture closed population model estimates of humpback chub abundance >
150 mm total length in the Little Colorado River. Error bars represent 95%
confidence intervals.

















117











10
ASMR 1


8-I
7 T



-- A


4
10 1
S I ASMR 2



CD
Co 9 E



-3:

C< I I
r< T T


S B
4
10 I ASMR 3


-5- T
















95% credible intervals from 200,000 Markov Chain Monte Carlo trials.
S118
I T I I




























118











5 -
ASMR 1
4
ST -- r T


3 -

2 -

1-, --- -- .
A

^ 5-
0

o ASMR 2
4- I I

33:


1 --



(D B


ASMR 3 i -








1 I I I
0 1 1 1



M CD CN f Lf LO -- COD M CD C f
S" n" m a) a) a a CD C: CZ c
Brood Year


Figure 4-6. Humpback chub recruit abundance (age-2) estimates from the ASMR 1 (A), ASMR
2 (B), and ASMR 3 (C) models using data pooled among tag cohorts. Error bars are
95% credible intervals from 200,000 Markov Chain Monte Carlo trials.


































-3 -2 -1 0 1 2 3


Theoretical Quantiles


. .



. 4 .... ... .

o .4 .... .... . ..
S : : : : : ;:


. . 4 0
0 .0 ..... 0 0



1990 1995 2000 2005


-3 -2 -1 0 1 2 3


Theoretical Quantiles


. + a .

. . o .. I f

*
44 40 0

. I
.


+, : f ... Q : ; :
'. c r 0


.o .* a o o [* 0 **
.. 0 1 o
~ o o *
o o i o <


1995


2000


1990


Year


2005


Year


Figure 4-7. Pearson residual plots for model ASMR 1 using data pooled among tag cohorts.
Individual plots are: Quantile-Quantile plots for marked (A) and recaptured (C) fish
and Pearson residuals-at-age and at-time for marked (B) and recaptured (D) fish.


















120


30 -

25 -






























-3 -2 -1 0 1 2 3


Theoretical Quantiles


Theoretical Quantiles


30 -

25 -


. . . . .
* I




i. .. .. . .
0 l f 0 . .





1990 1995 2000 2005


1990


1995


Year


2000


2005


Year


Figure 4-8. Pearson residual plots for model ASMR 2 using data pooled among tag cohorts.
Individual plots are: Quantile-Quantile plots for marked (A) and recaptured (C) fish
and Pearson residuals-at-age and at-time for marked (B) and recaptured (D) fish.
















121


; ; i 9 ti
: : : : : : : ; : i.I


* *
S . 0 .
. * .o ,


S 4 0 0 o o 0


-3 -2 -1 0 1 2 3


































-3 -2 -1 0 1 2 3


Theoretical Quantiles


Theoretical Quantiles


30 -

25 -


4 4 *
. . . ,
4, 4.. .4.. .
** ** .. 4 *
* .. . . l ,
'. : 0 4 : 4 : :'.



: . o o ,


1990


1995


Year


2000


2005


Year


Figure 4-9. Pearson residual plots for model ASMR 3 using data pooled among tag cohorts.
Individual plots are: Quantile-Quantile plots for marked (A) and recaptured (C) fish
and Pearson residuals-at-age and at-time for marked (B) and recaptured (D) fish.



















122


S. .. ..





+* Q .. . 0 .

I* : o t ?. : ?:. ,: ?7
n 'oC' .
0* O o
44 0> 044 0

* 9 1 _.o ..
*~ j *ii .


-3 -2 -1 0 1 2 3















1989
1.0-- 990
S1991
1992
S1993
1994
1995
-..... 1996
---- 1997
0.8 -- 1998
1999
--- 2000
2001
2002
2003
2004
2. 2005
0 6 2006

2 A
0


0.4 ,| ,,
I 0- II III AI
I ,* -i
0 ,.2'









0.0-


2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

Age

Figure 4-10. Capture probability by age and year estimated from model ASMR 3 using data
pooled among tag cohorts.

















123











11
ASMR 1
10

9
B--

7-


6 A
5 A A-

4
11
5o ASMR 2
rD in









4 -
r8

w7
iI


5- C6 I
SI I I I I I I I

11
I ASMR 3
10

9- 3:
F FCI

SYear
7
z mlII


SI I i i i
C









95% credible intervals from 200,000 Markov Chain Monte Carlo trials.














124
Year

Figure 4-11. Humpback chub adult abundance (age-4+) estimates from the ASMR 1 (A), ASMR
2 (B), and ASMR 3 (C) models using data stratified by tag cohort. Error bars are
95% credible intervals from 200,000 Markov Chain Monte Carlo trials.















124










5 -
ASMR 1
4 I -



2


A

^ 5-
0
o
o ASMR 2




I ^
CD




_2- T I I


5)
(D B


ASMR 3


3- T T
S1 iTT I



1
C
0 I I I I I I I II I I I I I


Brood Year

Figure 4-12. Humpback chub recruit abundance (age-2) estimates from the ASMR 1 (A), ASMR
2 (B), and ASMR 3 (C) models using data stratified by tag cohort. Error bars are
95% credible intervals from 200,000 Markov Chain Monte Carlo trials.














125































-3 -2 -1 0 1 2 3


Theoretical Quantiles


Theoretical Quantiles


30 -

25 -


.4... .............
* :




* .. .

.





1990 1995 2000 2005


1990


1995


Year


2000


2005


Year


Figure 4-13. Pearson residual plots for model ASMR 1 using data stratified by tag cohort.
Individual plots are: Quantile-Quantile plots for marked (A) and recaptured (C) fish
and Pearson residuals-at-age and at-time for marked (B) and recaptured (D) fish.

















126


S: '. : : : : : ; '.i


ar
*







0.00
** .o. o -
o 0 o o o,


-3 -2 -1 0 1 2 3

































-3 -2 -1 0 1 2 3


Theoretical Quantiles


Theoretical Quantiles


30 -

25 -


94 ..... .
S. . .




. ... .. . .
1990 199 2000 ....
o o . .
*, 9 o .9. 0 : 0




1990 1995 2000 2005


1990


1995


Year


2000


2005


Year


Figure 4-14. Pearson residual plots for model ASMR 2 using data stratified by tag cohort.
Individual plots are: Quantile-Quantile plots for marked (A) and recaptured (C) fish
and Pearson residuals-at-age and at-time for marked (B) and recaptured (D) fish.


















127


S

40 4 4 9 i
.* *




. o :
. 00. 00 0
.r ** ^ .


-3 -2 -1 0 1 2 3






































-3 -2 -1 0 1 2 3


Theoretical Quantiles


Theoretical Quantiles


30 -


25 -


-. . 9. .





...
0 .....


I: I I I2 0
1990 1995 2000 2005
: r ~ : : : : : :
o .


1990


1995


Year


2000


2005


Year


Figure 4-15. Pearson residual plots for model ASMR 3 using data stratified by tag cohort.
Individual plots are: Quantile-Quantile plots for marked (A) and recaptured (C) fish

and Pearson residuals-at-age and at-time for marked (B) and recaptured (D) fish.





















128


.4..
0 ... .



I 0 ....... .. 0 0 0
/. + 8 7 : ; : 0. .:b9
+ .. . .. => . i +.




I .. 41 9 .0 . 0 0 0
I . . ,. : 0 .
I' i 0 0 0 : . $ .
/O + o
I4 'l o 1:1 0 '' *,I,
* 0 O o 1 ^^
60 a
4 :* '* *:* *
4 4 4 > ** .
.;8 ^ 4. QO *.
i * ^ i 1: i i~i 4 i
*rr *0 *
rS o f : ** :.' ** *
+B* 0 ** A
~ ~ ??0 0-* ****


-3 -2 -1 0 1 2 3
















1989
1.0---1990
S1991
1992
S 1993
S1994
1995
-..... 1996
---- 1997
0.8 -- 1998
1999
--- 2000
2001
2002
2003
2004
2005
06 2006

o I



0.- \ I





0.2
j O A '












0.0-
I / II \ I' I








2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

Age

Figure 4-16. Capture probability by age and year estimated from model ASMR 3 using data
stratified by tag cohort.


















129





























150 200 250 300 350 400 450 500


Length Bin


150 200 250 300 350 400 450
150 200 250 300 350 400 450


150 200 250 300 350 400 450 500
160 200 260 300 360 400 450 600


Length Bin


150 200 250 300 350 400
150 200 250 300 350 400


Length Bin


450 500
450 500


Length Bin


Figure 4-17. Seasonal probability surfaces of age for a particular length bin. These surfaces sum
to unity in the vertical dimension (i.e., for each length bin) with the height of the
surface indicating the probability of a particular age given a particular length bin,
P(a 1). Individual plots are for April-June (A), July-September (B), October-
December (C), and January-March (D).












130


//// '


,
-~









10 -


III'''''


-


I ill I:


CD) 0 CI oO r L.. (D --- oo 00 ) OC .- C o (., D C
0D O0 0) 0) CD 0) 0), C 0) 0,) 0C) CDO O O O O O O
0) 0) M) 0) M) 0) 0) 0D 0) 0) O) 0 0 0 0D0 0 0
.- (cN C\ CIN C". C CNI CN
Year
Figure 4-18. Estimated adult abundance (age-4+) from ASMR 3 incorporating uncertainty in
assignment of age. Point estimates are mean values among 1,000 Monte Carlo trials
and error bars represent maximum and minimum 95% profile confidence intervals
among 1,000 Monte Carlo trials.




























I 6
o6
CD


u
CD
C---



r"













0-
I I I I I I I I I I I

00 0, 0) O) 0) C) 0:) C 0-) CD O) 0:) o 0 o
C) 0:) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0 C 0

Brood Year

Figure 4-19. Estimated recruit abundance (age-1) from ASMR 3 incorporating uncertainty in
assignment of age. Point estimates are mean values among 1,000 Monte Carlo trials
and error bars represent maximum and minimum 95% profile confidence intervals
among 1,000 Monte Carlo trials.













4 _
x-


-Data through 2006
- Data through 2005
- Data through 2004


S_ ---- Data through 2003
S--- Data through 2002
o ---- Data through 2001




C--
c
O0

<






o A






0
o-I






cP Data through 2005
S- -

S- Data through 2006
--- Data through 2005
D .". Data through 2004
Data through 2003
DData through 2002
S- B --- Data through 2001
o IlDI I I I I I I
D D C n T CD L- 0 MCO- n L 0 O OCD
0 CD0 0) 0) D 0) 0) 0) 0) 0) 0) D D D D D 0 0
-- '- -- c( o N o J ( o N o o
Year

Figure 4-20. Retrospective analysis of adult abundance (A) and mortality rate (B) considering
datasets beginning in 1989 and ending in the year indicated in the figure legend.









CHAPTER 5
LINKING TEMPORAL PATTERNS IN FISHERY RESOURCES WITH ADAPTIVE
MANAGEMENT: WHAT HAVE WE LEARNED AND ARE WE MANAGING
ADAPTIVELY?

With increased public awareness of the altered and degraded conditions prevalent in many

U.S. rivers over the last two decades, river restoration activities have increased exponentially

(Palmer et al. 2007). In the southwest, motivations for river restoration activities are varied and

include riparian zone and water quality management, in-stream habitat improvement, flow

modification, and concern over federally listed endangered species (Baron et al. 2002; Gloss et

al. 2005; Follstad-Shah et al. 2007). A recent review of U.S. river restoration activities

discovered that in many cases (>90%) little information was available from monitoring or other

activities to assess the success of these efforts (Bernhardt et al. 2005). Though this finding is

troubling considering the associated financial expenditures (>7.5 billion dollars between 1990

and 2003; Bernhardt et al. 2005), the science of river restoration and associated measures of

ecological "success" are yet in formative stages (Palmer et al. 2005). Nevertheless, this finding

shows that clear project goals and monitoring systems to evaluate progress towards those goals

are often lacking.

The Glen Canyon Dam Adaptive Management Program (GCDAMP) was formed as a

provision of the Record of Decision following the Final Environmental Impact Statement on the

operation of Glen Canyon Dam (USDOI 1995). Though the overarching goal of the GCDAMP

could be described as assisting the U.S. Secretary of Interior to comply with the body of law

governing the management of Colorado River water resources and Grand Canyon National Park

and Glen Canyon National Recreation area, this program has significant river restoration intent

as described in the Grand Canyon Protection Act of 1992 (Act). The geographic scope of the

GCDAMP is the Colorado River within Glen and Grand Canyons and associated riparian and









terrace zones influenced by dam operations (GCDAMP 2001). Additionally, the Act refers to

improving resources in Grand Canyon Nation Park and Glen Canyon National Recreation Area,

and the majority of the GCDAMP goals refer directly to aquatic or riparian resources (GCDAMP

2001).

The GCDAMP is generally viewed as a successful example of a large-scale adaptive

management and river restoration program (Ladson and Argent 2002; Poff et al. 2003), but

specific criteria defining resource goals are generally lacking in the foundational and working

documents of the program (GCDAMP 2001). Although the GCDAMP has attempted to develop

goals shared by all of the participating stakeholders, the result has been an unfocused set of

ambiguous resource goals that are neither well prioritized nor organized within an ecosystem or

"guiding image" perspective (NRC 1996; Palmer et al. 2005). According to the U.S. Department

of Interior technical guide on adaptive management (Williams et al. 2007), a requirement of

adaptive management is a statement of explicit and measurable goals. Therefore, and according

to guidance from the U.S. Department of Interior, this program is not operating within an

optimum adaptive management framework.

To meet agency guidelines from Williams et al. (2007), further program development is

clearly needed, but the GCDAMP has initiated several large-scale adaptive management

experiments since inception in 1996 and the program continues to attempt to learn as much as

possible from those efforts. This chapter provides an overview of past experimental actions,

uncontrolled factors, and attempts to evaluate the effects of these actions on Colorado River fish

populations. I describe trends in uncontrolled, but possibly important factors such as: (1)

Colorado River and major tributary hydrology, (2) release water temperature from Glen Canyon

Dam, and (3) variable production of native fish from the Little Colorado River (LCR).









Description of Adaptive Management Actions

Below I summarize the major management actions implemented by the GCDAMP since

inception. These actions are potentially biologically significant to Colorado River fish

populations (Chapter 2).

1996 Experimental High Flow

Prior to the construction of Glen Canyon Dam, annual spring flood discharges in the

Colorado River often exceeded 50,000 ft3/s (cfs) with infrequent events exceeding 210,000 cfs

(Topping et al. 2003). Following construction of Glen Canyon Dam, flows rarely exceeded

30,000 cfs. For seven days in April 1996, a widely publicized experimental high flow (hereafter

1996 EHF) was released from Glen Canyon Dam with a peak discharge of 45,000 cfs (Webb et

al. 1999). Though a discharge of this magnitude represents a minor pre-dam flood event with

recurrence interval < 1.25 years, it represented a post-dam event with recurrence interval of 5.1

years (Schmidt et al. 2001). The overarching restoration goal of the 1996 EHF was to

redistribute fine sediment within the channel to build shoreline sandbars, a primary program

objective. However, a broad set of physical and biological studies were undertaken in concert

with the 1996 EHF and are thoroughly documented in the literature (e.g., Webb et al. 1999;

Patten and Stevens 1999).

2000 Low Summer Steady Flow

A program of experimental flows (hereafter LSSF) from Glen Canyon Dam (Valdez et al.

2000) was initiated in 2000 to benefit native fish resources and to comply with the reasonable

and prudent alternatives of the U.S. Fish and Wildlife Service Biological Opinion on the

operation of Glen Canyon Dam (USFWS 1994). The hydrograph for this treatment is thoroughly

described elsewhere (Trammell et al. 2002), but consisted primarily of a period of constant 8,000

cfs discharges from June 1, 2000 to September 30, 2000 with the intent of promoting improved









native fish rearing conditions through increased temperature and hydraulically stable near-shore

environments. Preceding this prolonged constant 8,000 cfs discharge, there was a 52-day block

of high (17,000 to 30,000 cfs) and generally constant discharge to mimic spring flooding during

April and May. During the constant 8,000 cfs discharge in September, there were several spike

flows to 16,000 and 30,000 cfs for up to four days with the intent of disadvantaging non-native

fish.

Sampling during the LSSF indicated higher abundance of speckled dace RhiniL hI//y)

osculus, flannelmouth sucker Catostomus latipinnis, bluehead sucker Catostomus discobolus,

and fathead minnow Pimephalespromelas in backwater areas, particularly in the lower portions

of Grand Canyon (Trammel et al. 2002). These results suggested some benefit to warm water

fishes associated with LSSF. Unfortunately, few juvenile humpback chub were captured during

LSSF, perhaps related to low production of young of year fish in the LCR during 2000 (Dennis

Stone, U.S. Fish and Wildlife Service, Personal Communication).

2004 Experimental High Flow

Following large sediment inputs to the Colorado River from the Paria River in late summer

and fall of 2004, a second experimental high flow was implemented from Glen Canyon Dam

(hereafter 2004 EHF; Wright et al. 2005). In contrast to the 1996 EHF, the 2004 EHF occurred

during November and with slightly lower discharge (41,000 cfs) and shorter duration (60 hours;

Topping et al. 2006). As with the 1996 EHF, this management action was primarily focused on

restoring shoreline sandbars. Much of the data analysis and reporting for this effort are still

ongoing.

2003-2005 Non-native Fish Suppression Flows

Coupled with removal of non-native fish (see below), a program of increased flow

fluctuations (termed NNFSF, non-native fish suppression flows) from Glen Canyon Dam was









implemented during January-March, 2003-2005 to suppress rainbow trout Oncorhynchus mykiss

on a systemic basis by disrupting spawning and rearing. Korman et al. (2005) conducted

extensive studies on the spawning and rearing success of rainbow trout during the NNFSF.

These studies also investigated the extent of spawning in the Colorado River and tributaries

between River Mile 0 (RM 0) at Lees Ferry and RM 62 near the confluence of the LCR.

Korman et al. (2005) concluded that these flows were likely ineffective at substantially reducing

rainbow trout recruitment because of sub-optimal timing (i.e., too early-in the year) and possible

compensatory survival at older ages. Based on surveys of available spawning habitat and density

of young of year rainbow trout between RM 0 and RM 62, they further concluded that NNFSF

likely did not affect rainbow trout reproduction downstream of RM 0 since there was little to no

evidence of reproductive activity in this reach during 2004.

2003-2006 Mechanical Removal of Non-native Fish

This effort removed over 23,000 non-native fish between RM 56.3 and 65.7 between

January 2003 and August 2006. Concurrent with these removals, the fish community

composition within this reach shifted from one being numerically dominated by rainbow trout

(>90%) in 2003, to one dominated by native fishes and the non-native fathead minnow (>90%)

in 2006. Though trends in the abundance of rainbow trout in both the removal and the control

reaches imply a systemic decline in rainbow trout unrelated to removal efforts, mechanical

removal also contributed to the shift in community composition. Motivation, methodology, and

results of the non-native fish removal program are presented in Chapter 2.

Description of Uncontrolled Factors

Below I summarize some of the uncontrolled factors influencing both fish population

dynamics and the ability to detect potential fish population responses.









Paria, Little Colorado, and Colorado River Hydrology

Variability in the Paria, Little Colorado, and Colorado Rivers hydrology all potentially

affect fish populations in the Colorado River. Seasonal and episodic freshets in the Paria River

are of much smaller discharge than the Colorado River (Figure 5-1), yet in the post GCD system,

the Paria River is the largest source of fine sediment (Wright et al. 2005). As a result, flooding

in the Paria River is a dominant driver of turbidity in downstream portions of the Colorado

River. Because turbidity decreases primary and secondary production (Kennedy and Gloss

2005), the Paria River is a major factor structuring the downstream aquatic community both in

terms of food resources (Carothers and Brown 1991) and foraging efficiency for sight feeders

such as rainbow trout (Barrett et al. 1992). Examination of the Paria River hydrograph (Figure

5-1) demonstrates the "flashy" character of this system and reveals major fall flooding during

1997-2000, 2004, and 2006. Conversely, 1995-1997 and 2001-2003 were periods of relatively

less flooding and with corresponding lower turbidity downstream from the Paria River

confluence.

The LCR is also a major source of fine sediment to the system with a corresponding

influence on turbidity. In contrast to the Paria River, large floods in the LCR are frequently of

equal or greater discharge than the post-dam Colorado River (e.g., 1992 and 2002 events; Figure

5-1). These events not only increase turbidity, they also transport large numbers of native and

non-native fish to the Colorado River (Valdez and Ryel 1995; Stone et al. 2007). In turn, results

of fish sampling (e.g., species and length composition and catch rate) downstream of the

confluence can be highly influenced by recent LCR hydrology.

Annual release volume from Glen Canyon Dam is influenced by storage capacity in Lake

Powell, annual inflow volume to Lake Powell, and a set of laws and regulations aggregately

termed "The Law of the River". During periods of high storage capacity and low annual inflow,









relevant policies guarantee a minimum annual release volume of 8.23 million acre feet (maf) of

water. During periods of lower reservoir storage and larger annual inflows, annual release

volumes may exceed this amount. Since 1990, annual release volume exceeded 8.23 maf in

1995-2000 (Figure 5-1). Though the relationship between annual release volume and fish

population dynamics remains unclear, Paukert and Rogers (2004) found a positive relationship

between annual release volume and flannelmouth sucker condition factor. These authors

hypothesized that greater annual volumes provide increased euphotic volume and therefore

greater primary and secondary production.

Release Water Temperature from Glen Canyon Dam

Water temperature released from Glen Canyon Dam is largely dependent on the reservoir

elevation in relation to the dam penstock depth. By 1973, the last vestige of the pre-dam

thermograph disappeared as the reservoir filled to an elevation promoting annual hypolimnetic

releases of between 7 and 12 C (Vemieu et al. 2005). The reservoir water levels fell during a

drought in 2000-2005 prompting partial epilimnetic releases and the warmest release water

temperatures since the before the reservoir filled. These warmer water releases, coupled with

further downstream warming during the summer months, resulted in significantly increased

water temperature in the Colorado River near the LCR confluence during 2003-2006 as

compared to 1990-2002 (Figure 5-2). Though water temperatures during 2003-2006 were still

below pre-dam values, they are much closer to those required for successful spawning and

rearing by warm-water adapted native fishes (Hamman 1982; Valdez and Ryel 1995), and should

be conducive to increased growth of humpback chub Gila cypha (Chapter 3).

Juvenile Native Fish Production in the Little Colorado River

Native fish production in the LCR is influenced by a host of factors controlling egg, larval,

and juvenile survival. Dominant factors likely include: hydrology, temperature, food resources,









parasite infestation, and predation risk. As mentioned above, freshets in the LCR tend to

increase the abundance of young fish in the mainstem downstream of the confluence, particularly

associated with late summer rainstorms (Valdez and Ryel 1995). Valdez and Ryel (1995)

suggest that survival rates of these young native fish attempting to rear in the mainstem is

extremely low given post-dam conditions. It is, therefore, easy to conceive how LCR freshets

may affect year class strength if a significant portion of a cohort must attempt to rear in the

mainstem. Conversely, that portion of the cohort remaining in the LCR may experience a

compensatory increase in survival and ultimately contribute significantly to recruitment.

Valdez and Ryel (1995) hypothesized that flooding activity in the LCR tends to

disadvantage non-native fish and cleanse spawning gravels. This is the argument provided to

explain the very large number of young of year humpback chub observed from the 1993 brood

year (Valdez and Ryel 1995) following extensive flooding in late 1992-early 1993. Furthermore,

these authors hypothesized that the 1993 brood year consisted of such large numbers of young

humpback chub that LCR resources were insufficient for juvenile fish rearing leading to

reduced growth rates and condition, and prompting large numbers of fish to emigrate to the

mainstem during July 1993 base flow conditions. Clearly, it is difficult to predict how LCR

hydrology may affect native fish production and recruitment.

Another factor potentially influencing native fish production in the LCR is infection by the

non-native Asian Tapeworm Bothriocephalus acheilognathi and the copepod Lernaea

cyprinacea. These parasites are present in the LCR and both humpback chub and speckled dace

appear to be particularly susceptible to infection (Choudhury et al. 2004). Hoffnagle et al.

(2006) demonstrated diminished condition of humpback chub infested with these parasites and

suggested sub-lethal and lethal effects.









How Are Fish Populations Affected by Adaptive Management Actions and Uncontrolled
Factors?

The GCDAMP is fortunate to have long-term time series measures of the relative

abundance of native and non-native fish as indexed by electrofishing catch rate in the mechanical

removal reach (Figures 5-3 and 5-4), and by hoop net catch rate in the mainstem downstream

from the LCR confluence (RM 63.7-64.2; Figure 5-5). These measures, along with average size

(Figures 5-6 through 5-8), provide relatively good information to infer changes in population

demography over time.

Has Increased Turbidity Affected Fish Populations in Grand Canyon?

Turbidity is hypothesized to affect fish populations via a suite of direct and indirect

mechanisms. Decreased water clarity should generally result in reduced autochthony (Yard

2003) potentially limiting food resources for some fishes. However, increased tributary

discharge may also result in increased detritus available for consumption by simuliids a

dominant food item for fish species downstream of the Paria River and LCR (Kennedy and Gloss

2005). Turbidity may also affect foraging efficiency, particularly for rainbow trout (Barrett et al.

1992). Thus, turbidity may mediate negative interactions between non-native and native fish

(Gradall and Swenson 1982; Gregory and Levings 1998; Johnson and Hines 1999) and possibly

represent a mortality source for rainbow trout (Chapter 2).

With these hypotheses in mind and given minimal flooding activity from the Paria River

during 1995-1997 and 2001-2003, I would predict that the abundance of rainbow trout in the

mechanical removal reach should be highest during these time periods. Examination of the catch

rate data indicates that rainbow trout abundance was high during 2001-2003, but not during

1995-1997 (Figure 5-3). Similarly, if native fish are advantaged by high turbidity conditions,

their abundance should be lower during these time periods. Available data indicate no strong









trends in native fish abundance in these time periods as indexed by either electrofishing or hoop

net catch rate (Figure 4-5). Humpback chub recruitment estimates from the age-structured mark-

recapture model (Chapter 4) suggest an increasing recruitment trend beginning with the 1995

brood year. Though turbidity may affect rainbow trout abundance in downstream reaches, there

are clearly other controlling mechanisms such as immigration from the Lees Ferry reach

(Chapter 2; Korman et al. 2005). Additionally, there does not appear to be a simple relationship

between turbidity and native fish populations.

The preceding discussion illustrates the challenges involved in evaluating the impact of

adaptive management experiments and uncontrolled factors on fish populations in Grand

Canyon. For nearly every credible hypothesis suggesting a positive influence from turbidity, an

equally credible contradictory hypothesis can be generated. Additionally, correlative analyses to

assess various aposteriori hypotheses of cause and effect relationships provide weak inference

at best (Yoccoz et al. 2001; MacKenzie et al. 2007). This is particularly true with regard to the

highly variable and uncontrolled factors (e.g., tributary hydrology) and short duration adaptive

management experiments (i.e., 1996 and 2004 EHF and LSSF).

Has Reduced Non-native Fish Abundance and Increased Temperature Affected Native Fish
Populations in Grand Canyon?

The GCDAMP initiated the mechanical removal of non-native fish program (Chapter 2) in

response to the increases in the abundance of non-native rainbow and brown Salmo trutta trout

near the confluence of the LCR during 1996-2002 (Figure 5-3). Implicit in the action was the

hypothesis that native fish are negatively impacted by non-native fish. More specifically, that

juvenile native fish survival was negatively impacted by non-native fish via predatory and

competitory interactions. If this hypothesis is correct, the abundance of juvenile native fish in

the mechanical removal reach would increase after non-native removal. Additionally, the









removal effort was predicted to promote increased humpback chub recruitment associated with

the 2003-2006 brood years.

Examination of electrofishing catch rate in the mechanical removal reach suggest an order

of magnitude increase in the relative abundance of flannelmouth sucker and bluehead sucker

between 2003 and 2006 (Figure 5-4). Increases in humpback chub and speckled dace are also

apparent, though not as large. Hoop net catch rates of humpback chub also increased,

particularly in late 2004-2006 (Figure 5-5). Additionally, the relative abundance of non-native

fathead minnow increased markedly in 2005 and 2006 (Figure 5-3). Though these responses are

consistent with predictions, it is not clear that these responses were caused in whole or part by

diminished abundance of rainbow and brown trout.

The mechanical removal of non-native fish coincided with the warmest water temperatures

in the mainstem Colorado River near the LCR confluence observed since before 1990 (Figure 5-

2). Since there is near perfect temporal correlation between these factors both hypothesized to

control native fish survival, it is difficult to evaluate their effects separately. Indeed, these

factors acting in combination likely have a multiplicative effect on juvenile fish survival through

increased growth of native fish, increased food resources, and decreased predation risk (Paukert

and Petersen 2007).

Paukert and Petersen (2007) used bioenergetics models to evaluate the effects of increased

water temperature on the growth, food consumption, and rainbow trout predation on juvenile

humpback chub. Their results indicated that an 8g humpback chub could realize a 287% greater

annual increase in mass under the 2005 thermograph as compared to mean 1993-2002 water

temperatures. They further predicted that the 2005 thermograph would reduce the time juvenile

chub were vulnerable to predation by 2-3 months. Paukert and Petersen (2007) also predicted









that the rainbow trout population in the mechanical removal reach annually consumed more than

18,000 kg of invertebrates. If food availability is a limiting factor for native fishes in Grand

Canyon, liberation of these resources may have significantly increased their growth and survival.

The average length of native fish captured with electrofishing provides some evidence of

increased growth and survival consistent with the predictions of Paukert and Petersen (2007).

The declining trend in both flannelmouth sucker and humpback chub TL is a result of a larger

fraction of smaller fish in the electrofishing catch during 2005 and 2006 than previous years

(Figure 5-7). This is particularly evident for flannelmouth sucker as the 2006 length frequency

distribution displays multiple modes consistent with several successful year classes (Figure 5-9).

The increasing trend in speckled dace TL is consistent with the hypothesis of increased growth in

2005 and 2006 (Figure 5-7). The sudden appearance of adult bluehead sucker in 2002-2006

electrofishing samples is puzzling and is most consistent with immigration of adults from the

LCR (Figures 5-4 and 5-7). Sampling by U.S. Fish and Wildlife Service personnel in the LCR

has indicated large increases in bluehead suckers during 2006 (Van Haverbeke and Stone 2007)

in support of this hypothesis. Increased size of humpback chub captured in hoop nets in 2005-

2006 also supports the hypothesis of increased growth and survival, as this gear seldom captures

adult fish in the mainstem Colorado River (Figure 5-8). However, as described above,

interpreting patterns of relative abundance or length composition of juvenile fish is complicated

by LCR hydrology and associated fish migration.

Conclusions and Recommendations

Index data from electrofishing and hoop net sampling in the mechanical removal reach

indicate an increase in native fish abundance in 2005-2006, particularly for the sucker species.

These changes in abundance are consistent with the predicted response to removal of non-native

fish in this reach, but are also consistent with predictions of increased water temperature. Thus,









it is not possible to determine which factor is most responsible for the changes. Nevertheless,

from an adaptive management perspective, there is good evidence that a successful policy with

respect to native fish conservation may involve at least one of these factors.

A strategy that could be employed to determine the relative influence of these factors is to

manipulate the system so that one of the factors remained at the present state while the other was

varied. Perhaps the most tenable option to achieve this circumstance would be for the GCDAMP

to continue non-native fish control and wait for wetter hydrology to raise reservoir levels and

decrease release temperatures (though such hydrology may not soon occur according to Seager et

al. 2007). If under these conditions native fish abundance and subsequent recruitment remained

high, the conclusion would be that interactions with native fish were the dominant factor. In

contrast, if native fish resources suffered, the conclusion would be that water temperature was

the dominant factor. This latter conclusion would lend support to construction of a selective

withdrawal structure on Glen Canyon Dam in order to provide control of release water

temperature. If warmer release water conditions persist, there would be limited ability to

discriminate among these two factors, but additional evidence might accumulate supporting the

success of the combined factor policy. This would also have the benefit of achieving the primary

program objective of native fish conservation.

However, two GCDAMP review panels have concluded that the risk to native fish may be

compounded by warmer water releases by favoring non-native fishes (Mueller et al. 1999; Garret

et al. 2003). This advice might be interpreted to be wary of transitory behavior of native fish

populations following onset of warmer water conditions. Initial signs of native fish benefit could

be short lived as the non-native fish community shifted to favor warm-water species potentially

capable of even more detrimental interactions with native fishes than the present assemblage.









Additionally, such a shift in fish community composition might not be easily reversible in a short

period of time, particularly with limited control of release water temperature.

The ultimate success of a native fish conservation policy should be evaluated in light of

increased native fish recruitment. A policy that increased native fish survival for some life stage,

but failed to promote recruitment to adulthood, could not be deemed successful. This

observation highlights the importance of the humpback chub stock assessment program (Chapter

4). However, evaluating policy success solely on recruitment estimates may not allow critical

insights into policy performance since recruitment to adulthood is the integration of conditions

faced by a cohort for multiple years. Differential recruitment from both the LCR and the

mainstem Colorado River is a further complication. Consider a situation where a management

action is undertaken in the mainstem concurrent with uncontrolled factors benefiting rearing

conditions in the LCR. If recruitment for those brood years is determined largely by the LCR

and the management activity in the mainstem is ineffective, evaluation of the recruitment time

series might erroneously lead managers to believe that activities in the mainstem represented a

successful policy.

Because of these complications, I contend that a more rigorous mainstem monitoring

program to estimate growth and survival of native fish is needed in order to fully evaluate

adaptive management experiments. Such a program would need to monitor immigration of

juvenile native fish from the LCR into the mainstem and the abundance of native fish in the

mainstem near the LCR confluence. These data could then be used in a balance framework to

estimate time specific apparent survival rate. Korman et al. (2005) have successfully

implemented a similar framework to study the effect of dam operations on juvenile rainbow trout

in the Lees Ferry reach. Though the Lees Ferry program benefits from having high densities of









fish and is not confounded by tributary hydrology, it might be possible to modify their approach

and implement it to study native fish survival and growth dynamics in the mainstem Colorado

River. This information, coupled with the ongoing stock assessment program, would

significantly increase the ability of the GCDAMP to evaluate adaptive management experiments.

Williams et al. (2007) argue that critical elements of an adaptive management program

require specification of explicit and measurable program goals and the ability to predict policy

performance relative to those goals. The GCDAMP would benefit from additional specificity in

program goals so that it is clear if and when a policy could be judged successful. As an example,

the native fish index data suggests recent improved conditions for native fish, but without

specific goals the program cannot judge whether the policy should be pronounced successful.

The program could also benefit from additional predictive capability both to screen policy

options and to formalize and evaluate alternative hypotheses of system behavior. As discussed

by Anderson et al. (2006), predictive models of fish population dynamics in riverine systems

should recognize and model the dynamic feedbacks among various trophic components and the

forcing factors mediating such interactions. Apparently without recognizing it, these authors

have described many of the elements contained in the Ecopath/Ecosim modeling framework

(Christensen and Walters 2003). I recommend that the GCDAMP rely more heavily on these

types of quantitative approaches to predict native fish population dynamics under alternative

management policies.

With improved monitoring and predictive capabilities to evaluate explicit and measurable

goals, I believe the GCDAMP will be well placed to better conceive and implement future

adaptive management experiments. Such planning should recognize the uncontrolled variability

in the system and select experimental design options to most effectively alleviate potential









confounding factors, while attempting to implement experimental policies most likely to achieve

program goals. I contend that one of the biggest obstacles in this effort is the inability to

implement experiments of sufficient duration and magnitude to provide measurable results. The

mechanical removal effort combined with multi-year temperature modification has the potential

to elicit effective learning and improved resource condition. The ultimate success of this effort

will depend heavily on the commitment to subsequent well planned experimentation and

monitoring.














50000 -


C "40000
l

a)



20000
m











1500 -
L







-U
o 1000


0
2000





S1500oo -




L



> 1000 -
[F











0t1500
0




20000
C,



S15000 -



0
o 6






I (N m t L w m cD m I I t Ln (D -
C) 0 ) ) 0) 0)D CO CD CD CD CD CD CD C 0 0 0 0





Figure 5-1. Discharge in ft3/s (cfs) for the Colorado River at Lees Ferry (A), the Paria River at

Lees Ferry (B), and the Little Colorado River at Cameron, AZ (C), 1990-2006.


















1990-2002 Daily Mean
17 1990-2002 Lowess Fit
2003 Daily Mean
16 2003 Lowess Fit
2004 Daily Mean
15 2004 Lowess Fit
S2005 Daily Mean
2005 Lowess Fit
14 2006 Daily Mean
-- 2006 Lowess Fit


- LD O') C" (C

SA CLL
D -D -D LL LL


L, OM (, CO 0 o'
CO O C, 0 C, 0)

2 < < ^3


D--a
Date


CM" (Dc 0-) ) r--
-- CD ( CD0
= =S c CD T


Figure 5-2. Daily mean water temperatures observed in the Colorado River at approximately
river mile 61, 1990-2006. Lines indicate locally weighted polynomial regressions
(Lowess) fits to the indicated data set.


4*

a\


St


(D CD
0)0


























oM E M r^ n D Lt tr-- oM M = E M rLn -LO r--
O) a O O COa O ) C D C
O CJO4C1OO
CT)")"3T)C)0)0)0"0"0"CN "CNC 4 .O O


I ii


I r I i I I I I I r r r r I
= -C, r Lf CON-C a C ODc-= C Ifl O -
C F) aY) ay) l C) ay) ( C) C C CD CD C CD C
M ') Ca) M) CfCl) M) Cn M MCD C CD C COCD C
- - CA CN "" CN


oM E r Ml r LO r--C M M = EN t M r Lfl O r--
0) 0C 0)C)H3) H)0)0) CDOC CD CDCOC
ij~o~ncncjcncnccncCQQQ "Q "


S; CN Ml tM L Ol Cr--C a) C = EC t LM l MMr--
- ) 0 -o 0- C C CC C


Figure 5-3. Monthly electrofishing catch rate (fish/hour) in the Colorado River between river
mile (RM) 56.3 and RM 65.7 for rainbow trout (A), brown trout (B), common carp
(C), and fathead minnow (D). Error bars represent 95% confidence intervals.


j


































= F n M- r M tLO r-- Mc M = E C4 t M o M Mr--
M 7c MMMMMMMM= Cc==CC C C CC
- - - )0"CN .O .O. .


20 -
D


C CLLC)O 14-CD CD CD CDOO CD CD
0)0) 0c 0)0)0)0)00)0) c CC Q CC C


)0M)0)C)0)0)C)0M) D0 CDCDOCDDCOCDOC
- - - OQOOQ


Figure 5-4. Monthly electrofishing catch rate (fish/hour) in the Colorado River between river
mile (RM) 56.3 and RM 65.7 for humpback chub (A), flannelmouth sucker (B),
bluehead sucker (C), and speckled dace (D). Error bars represent 95% confidence
intervals.


35 -
A


Cr)CrCncrCTI0 I I r F C OCDC O IDO

























0.20 0.00


-
0 15-

n-

8 0.10 -
r
-n

1 0.05 -
I


0.00 -


LI4+ 4


) 0C 0 0C
4 C


-c
.c
S0 004


oC)
0O

a 0.002
a)
_M


0.000


I I I I
S'n Ln LO
M M M
M M M


S CO CO CO CO CO CO CO CO


0 00 -


1 1 I I I I I 1 -
m C Y rC n CI in wl
" CO C oo o o .o
C CD CD CD C C CO
- ri rN i r, rCN rC r rN


-c
r
S0.004 -

0)



1-
'0







I 0.000 -


I'---
CO
C-


O D o C rC n M Li L r--
D M" CO CO CO CO CO CO CO CO
-r r- ri r^ rJ rNi rC
~f~i~t-~II

| | 1 I | | 1


Figure 5-5. Monthly hoop net catch rate (fish/hour) in the Colorado River between river mile
(RM) 63.7 and RM 64.2 for humpback chub (A), flannelmouth sucker (B), bluehead
sucker (C), and speckled dace (D). Error bars represent 95% confidence intervals.


0.008 -


-c

-c
0.006 -






0.002


--)
S0.004 -
0
E
c 0.002 -

u-

0.000 -


[
[
[


0.20 -


0.006
















{t


800 -
700 -
GOO -
600
500 -
400 -
300
200 -
100


oCD E r n t Lfl tD r-- oD CD ED- CMt Lfl! r--
) OCY) nCY) CT ) Y) M Y) Y) CD CC CDOCDCC
CY) a ')CL )0 ) l!a) )0 CD CDCD CD CCC
CN C4 CN CN C4 CN "


100 -


It.


*


CD E C fl t Lfl Nr--COD ) CD ED- C t Lf L r--
0)0) i ,L)C)0 )0 )0C)0D)0)CD CDCDOCDCC
ijcnnccjcnnccCN CN N C CNrN


* I


n -


lb


I I I I I I I I I I I I


I I I I I


= F CN flrt LflO r-- a) MC E CItA M l MMr--
O3 Y) aMMM) a) a) a) a) C C CDOCDCOCCCD
- ^ a) a) a) a)-C


4


44.


4s4


CD F Cfl r Ll Lo rN--OCC M)CDE CAt rl LO r--
- - - OQOOQ


Figure 5-6. Monthly average total length (TL; mm) observed in electrofishing sampling in the
Colorado River between river mile (RM) 56.3 and RM 65.7 for rainbow trout (A),
brown trout (B), common carp (C), and fathead minnow (D). Error bars represent
95% confidence intervals.


300 -
250 -
200 -
150 -
100 -


550 -
500 -
450 -
400 -
350 -
300 -
250 -
200 -
150 -
100 -
6 -
-


u












400 -
350 -
300 -
250 -
200 -
150 -
100 -
60-


oM E C4 fl t L LOr-- C OM C EC4Mr- LfL O r--
O- ^ fcLC O CFN0, CC14 R, CFN, RDC.OO
iOi30 O3T)C fl 0 O)"0 C0 COC CDOCCCC
^^^^^^^^ ^pjr Cjr J&rCr1r'r


11*~


120

100

80

60

40

20


CD E C flt Lfl CONr--C OD CY CDED- t Ll L r--
CYI ,L) O14 O CDCCD CCDC CC
0)0) 0c 0)0)0)0))0)0) c CC Q CCC


f*4


IIIIIIIIIIIIIIIIII
C -N fl t Lfl CONCO--COD C CD E- CI l O LO--
OM MMMMMMM = CDOC CD CDOCCC





D


- *i :l\' il







i I I I I I I I I I I I I I I I I
CD F CN fl r LflO r-- OD ) CD E CIlA r O LOr--
SC) CM CDC:COC DCDC
)- -)- -0 0-0 CN CN CN CN CN C\


Figure 5-7. Monthly average total length (TL; mm) observed in electrofishing sampling in the
Colorado River between river mile (RM) 56.3 and RM 65.7 for humpback chub (A),
flannelmouth sucker (B), bluehead sucker (C), and speckled dace (D). Error bars
represent 95% confidence intervals.


t


tb
















44


4{4fl


200 -
180 -
E 160-
E
. 140 -
F--
. 120 -
o 100 -
o 80-
-n
60
I- GO0-
E
S40
20 -
0-


oo M o ri n in t r-
C) m CC CD CD C CD C
CN ,- "- "- -


300 -
280 C
260 -
240 -
220 -
200 -
180 -
160
140 -
120 -
100 -
80 -
60
40-
20-
o I r n i t I --
a) fl O O O O O O O O


120 -

100 -

80 -

60 -

40 -

20 -


I I I I I I I I I I
O rC rCD rE r i r-
CY)CY)CD C:)CD D D C C
CY) M D D D C C C C C
- CN N N CN CN C


Figure 5-8. Monthly average total length (TL; mm) observed in hoop net sampling in the
Colorado River between river mile (RM) 63.7 and RM 64.2 for humpback chub (A),
flannelmouth sucker (B), bluehead sucker (C), and speckled dace (D). Error bars
represent 95% confidence intervals.


I I I I I I I I
o ED- Cri M in i t --
o C C CD C CD C
o C C C C C C C
CN .l .l Cl N "M "
0 0 0 00 0
0 0 0 00 0
( 1 1 1 (1 (1 ( ( (


Ji3"i


~~i c t*r



















2003 (n=900)
o 2006 (n=2548)





CD
C
C

0


CO
I


(-

CD
aC
o
o -
C
C





C
C
C




C
oC
o



0 100 200 300 400 500 600

Total Length (mm)


Figure 5-9. Smoothed kernel density plot of the total length of flannelmouth sucker captured
with electrofishing in the mechanical removal reach during 2003 and 2006. Numbers
in the legend represent total number (n) of fish captured.









LIST OF REFERENCES


Akaike, H. 1973. Information theory and an extension of the maximum likelihood principle.
Pages 267-281 in B. N. Petrov and F. Csaki, editors. Second International Symposium on
Information Theory. Akademiai Kiado, Budapest, Hungary.

Allan, C. and A. Curtis. 2005. Nipped in the bud: why regional scale adaptive management is not
blooming. Environmental Management 36:414-425.

Andersen, K. E., A. J. Paul, E. McCauley, L. J. Jackson, J. R. Post, and R. M. Nisbet. 2006.
Instream flow needs in streams and rivers: the importance of understanding ecological
dynamics. Frontiers in Ecology and the Environment 4:309-318.

Arreguin-Sanchez, F. 1996. Catchability: a key parameter for fish stock assessment. Reviews in
Fish Biology and Fisheries 6: 221-242.

Askey, P. J., S. A. Richards, J. R. Post, and E. A. Parkinson. 2006. Linking angling catch rates
and fish learning under catch-and-release regulations. North American Journal of Fisheries
Management 26:1020-1029.

Baillargeon, S. and L. P. Rivest. 2007. Rcapture: loglinear models for capture-recapture in R.
Journal of Statistical Software 19:1-31.

Baron, J. S. and 9 coauthors. 2002. Meeting ecological and societal needs for freshwater.
Ecological Applications 12:1247-1260.

Barrett, J. C., G. D. Grossman, and J. Rosenfeld. 1992. Turbidity-induced changes in reactive
distance of rainbow trout. Transactions of the American Fisheries Society 121:437-443.

Benke, A. E. and C. E. Cushing, editors. 2005. Rivers of North America. Elsevier Academic
Press, Burlington, MA.

Bernhardt, E. S. and 24 coauthors. 2005. Synthesizing U.S. River Restoration Efforts. Science
308:636-637

Bertalanffy, L. von. 1938. A quantitative theory of organic growth (Inquiries on growth laws II).
Human Biology 10:181-213.

Beverton, R. J. H. 1992. Patterns of reproductive strategy parameters in some marine teleost
fishes. Journal of Fish Biology 41:137-160.

Bigelow, P. E., T. M. Koel, D. Mahony, B. Ertel, B. Rowdon, and S. T. Olliff. 2003. Protection
of native Yellowstone cutthroat trout in Yellowstone Lake, Yellowstone National Park
Wyoming. National Park Service Technical Report NPS/NRWRD/NRTR-2003/314,
Yellowstone National Park, Wyoming.

Blinn, D. W., C. Runck, D. A. Clark, and J. N. Rinne. 1993. Effects of rainbow trout predation of
Little Colorado spinedace. Transactions of the American Fisheries Society 122:139-143.









Bozdogan, H. 2000. Akaike's information criterion and recent developments in information
complexity. Journal of Mathematical Psychology 44:62-91.

Britton, J. R., J. Pegg, R. Sedgwick, and R. Page. 2007. Investigating the catch returns and
growth rate of wells catfish, Silurus glanis, using mark-recapture. Fisheries Management
and Ecology 14:263-268.

Brownie, C., D. R. Andersen, K. P. Burnham, and D. R. Robson. 1985. Statistical inference from
band recovery data a handbook, 2nd Edition. U.S. Fish and Wildlife Service Resource
Publication 156.

Burnham, K. P. and D. R. Andersen. 2002. Model selection and multimodel inference. A
practical information-theoretic approach, second edition. Springer Science + Business
Media, New York.

Campbell, S. P. and 7 coauthors. 2002. An assessment of monitoring efforts in endangered
species recovery goals. Ecological Applications 12:674-681.

Carothers, S. W. and Brown, B. T. 1991. The Colorado River through Grand Canyon. University
of Arizona Press, Tucson, AZ.

Choudhury, A., T. L. Hoffnagle, and R. A. Cole. 2004. Parasites of native and nonnative fishes
of the Little Colorado River, Grand Canyon, Arizona. Journal of Parasitology 90:1042-
1053.

Christensen, N. L and 12 coauthors. 1996. The report of the Ecological Society of America
committee on the scientific basis for ecosystem management. Ecological Applications
6:665-691.

Christensen, V. and C. J. Walters. 2003. Ecopath with Ecosim: methods, capabilities, and
limitations. Ecological Modelling 172:109-139.

Clarke, A. and N. M. Johnston. 1999. Scaling of metabolic rate with body mass and temperature
in teleost fish. Journal of Animal Ecology 68:893-905.

Clarkson, R. W. and M. R. Childs. 2000. Temperature effects of hypolimnial-release dams on
early life stages of Colorado River big-river fishes. Copeia 2:402-412.

Coggins, L. G. Jr., M. D. Yard, and C. Paukert. 2002. Piscivory by non-native salmonids in the
Colorado River and an evaluation of the efficacy of mechanical removal on non-native
salmonids: an operational plan. Available: http://www.gcmrc.gov/library/reports/
biological/Fish_studies/GCMRC/Coggins2002.pdf. (January 2008).

Coggins, L. G., Jr. and T. J. Quinn II. 1998. A simulation study of the effects of aging error and
sample size on sustained yield estimates. Pages 955-975 in F. Funk, T. J. Quinn II, J.
Heifetz, J. N. Ianelli, J. E. Powers, J. F. Schweigert, P. J. Sullivan, and C. I. Zhang, editors.
Fishery Stock Assessment Models. Alaska Sea Grant College Program Report No. AK-
SG-98-01, University of Alaska Fairbanks.









Coggins, L. G., Jr., W. E. Pine, III, C. J. Walters, D. R. Van Haverbeke, D. Ward, and H. C.
Johnstone. 2006a. Abundance trends and status of the Little Colorado River population of
humpback chub. North American Journal of Fisheries Management 26:233-245.

Coggins, L. G., Jr., W. E. Pine, III, C. J. Walters, S. J. D. Martell. 2006b. Age-structured mark-
recapture analysis: a virtual-population-analysis-based model for analyzing age-structured
capture-recapture data. North American Journal of Fisheries Management 26:201-205.

Collier, M. P., R. H. Webb, and E. D. Andrews. 1997. Experimental flooding in Grand Canyon.
Scientific American 276:82-89.

Crawford, S. S. and A. M. Muir. In Press. Global introductions of salmon and trout in the genus
Oncorhynchus: 1870-2007. Reviews in Fish Biology and Fisheries. Available: www.
springerlink.com.lp.hscl.ufl.edu/content/m447581887406t4n/fulltext.pdf (January 2008)

DeVries, D. R., and R. V. Frie. 1996. Determination of age and growth. Pages 483-512 in B. R.
Murphy and D. Willis, editors. Fisheries techniques, 2nd edition. American Fisheries
Society, Bethesda, Maryland.

Donlan, C. J., K. Campbell, W. Cabrera, C. Lavoie, V. Carrion, and F. Cruz. 2007. Recovery of
the Galapagos rail (Laterallus spilonotus) following the removal of invasive mammals.
Biological Conservation 138:520-524.

Dorazio, R. M., H. L. Jelks, and F. Jordan. 2005. Improving removal-based estimates of
abundance by sampling a population of spatially distinct subpopulations. Biometrics
61:1093-1101.

Douglas, M. E., and P. C. Marsh. 1996. Population estimates/population movements of Gila
cypha, an endangered cyprinid fish in the Grand Canyon region of Arizona. Copeia
1996:15-28.

Dunbrack, R., J. M. Green, and C. M. Mlewa. 2006. Marbled lungfish growth rates in Lake
Baringo, Kenya, estimated by mark-recapture. Journal of Fish Biology 68:443-449.

Erskin Ogden, J. A. and M. Rejmanek. 2005. Recovery of native plant communities after control
of a dominant invasive plant species, Foeniculum vulgare: implications for management.
Biological Conservation 125:427-439.

Essington, T. E., J. F. Kitchell, and C. J. Walters. 2001. The von Bertalanffy growth function,
bioenergetics, and the consumption rates of fish. Canadian Journal of Fisheries and
Aquatic Sciences 58:2129-2138.

Fabens, A. J. 1965. Properties and fitting of the von Bertalanffy growth curve. Growth 29:265-
289.

Follstad-Shah, J. J., C. N. Dahm, S. P. Gloss, and E. S. Bernhardt. 2007. River and riparian
restoration in the southwest: results of the national river restoration science synthesis
project. Restoration Ecology 15:550-562.









Fournier, D. 2000. AD Model Builder. Otter Research, Sidney, British Columbia, Canada.
Available: http://otter-rsch.com/admodel.html. (January 2008).

Francis, R. I. C. C. 1988. Are growth parameters estimated from tagging and age-length data
comparable? Canadian Journal of Fisheries and Aquatic Sciences 45:936-942.

Garrett, D. and 12 coauthors. 2003. Evaluating a Glen Canyon Dam temperature control device
to enhance native fish habitat in the Colorado River: a risk assessment by Adaptive
Management Program Science Advisors. Prepared for the Glen Canyon Dam Adaptive
Management Program and the Grand Canyon Monitoring and Research Center, Flagstaff,
Arizona. Available: http://www.gcmrc.gov/library/reports/synthesis/Garrett2003.pdf (April
2008)

GCDAMP (Glen Canyon Dam Adaptive Management Program). 2001. Strategic Plan: Glen
Canyon Dam Adaptive Management Program, Final Draft. Bureau of Reclamation, Salt
Lake City, UT. Available: http://www.usbr.gov/uc/rm/amp/pdfs/sp_final.pdf (January
2008).

Gelman, A., J. B. Carlin, H. S. Stern, and D. B. Rubin. 2004. Bayesian data analysis, 2nd edition.
Chapman & Hall/CRC Press, New York.

Gloss, S. P. and L. G. Coggins, Jr. 2005. Fishes of Grand Canyon. Pages 33-56 in S. P. Gloss, J.
L. Lovich, and T. S. Melis, editors. The state of the Colorado River ecosystem in Grand
Canyon. U.S. Geological Survey Circular 1282.

Gloss, S. P., J. E. Lovich, and T. E. Melis, editors. 2005. The state of the Colorado River
ecosystem in Grand Canyon. U.S. Geological Survey Circular 1282.

Gorman, O. T., and D. M. Stone. 1999. Ecology of spawning humpback chub (Gila cypha) in the
Little Colorado River near Grand Canyon, Arizona. Environmental Biology of Fishes
55:115-133.

Gorman, O. T., R. G. Bramblett, R. M. Hervin, D. R. Van Haverbeke, and D. M. Stone. 2005.
Distribution and abundance of native and non-native fishes of the Colorado River
Ecosystem of Grand Canyon, Arizona. Pages 78-94 in M. J. Brouder, C. L. Springer, and
S. C. Leon, editors. The Lower Colorado River: Restoring natural function and native fish
within a modified riverine environment. Proceedings of two symposia and workshops, Las
Vegas, Nevada, July 8-9, 1998 and July 13-14, 1999. U.S. Fish and Wildlife Service,
Southwest Region, Albuquerque, New Mexico.

Gradall, K. S. and W. A. Swenson. 1982. Responses of brook trout and creek chubs to turbidity.
Transactions of the American Fisheries Society 111:392-395.

Gregory, R. S. and C. D. Levings. 1998. Turbidity reduces predation on migrating juvenile
pacific salmon. Transactions of the American Fisheries Society 127:275-285.









Gunderson, L. 1999. Resilience, flexibility and adaptive management antidotes for spurious
certitude? Conservation Ecology 3(1): 7. Available: http://www.consecol.org/vol3/
issl/art7/. (January 2008).

Hamman, R. L. 1982. Spawning and culture of humpback chub. Progressive Fish Culturist
44:213-216.

Hein, C. L., B. M. Roth, A. R. Ives, and M. J. Vander Zanden. 2006. Fish predation and trapping
for rusty crayfish (Orconectes rusticus) control: a whole-lake experiment. Canadian
Journal of Fisheries and Aquatic Science 63:383-393.

Hein, C. L., M. J. Vander Zanden, and J. J. Magnuson. 2007. Intensive trapping and increased
fish predation cause massive population decline of an invasive crayfish. Freshwater
Biology 52:1134-1146.

Hoffman, R. L., G. L. Larson, and B. Samora. 2004. Responses ofAmbystoma gracile to the
removal of introduced nonnative fish from a mountain lake. Journal of Herpetology
38:578-585.

Hoffnagle, T. L., A. Choudhury, and R. A. Cole. 2006. Parisitism and body condition in
humpback chub from the Colorado and Little Colorado Rivers, Grand Canyon, Arizona.
Journal of Aquatic Animal Health 18:184-193.

Holling, C. S. 1973. Resilience and stability of ecological systems. Annual Review of Ecology
and Systematics 4:1-23.

Holling, C. S., editor. 1978. Adaptive environmental assessment and management. John Wiley,
London.

Hood, G. M. 2000. Poptools Software for the Analysis of Ecological Models. Version 2.5.5.
Available: http://www.cse.csiro.au/poptools. (January 2008).

James, I. R. 1991. Estimation of von Bertalanffy growth curve parameters from recapture data.
Biometrics 47:1519-1530.

Jensen, A. L. 1996. Beverton and Holt life history invariants result from optimal trade-off of
reproduction and survival. Canadian Journal of Fisheries and Aquatic Science 53:820-822.

Johnson, J. E. and R. T. Hines. 1999. Effect of suspended sediment on vulnerability of young
razorback suckers to predation. Transactions of the American Fisheries Society 128:648-
655.

Kennedy, T. A. and S. P. Gloss. 2005. Aquatic ecology: the role of organic matter and
invertebrates. Pages 87-101 in S. P. Gloss, J. L. Lovich, and T. S. Melis, editors. The state
of the Colorado River ecosystem in Grand Canyon. U.S. Geological Survey Circular 1282.

Kimura, D. K. and S. Chikuni. 1987. Mixtures of Empirical Distributions: An iterative
application of the age-length key. Biometrics 43:23-35.









Kirkwood, G. P. and I. F. Somers. 1984. Growth of two species of Tiger Prawn, Penaeus
esculentus and P. semisulcatus, in the Western Gulf of Carpentaria. Australian Journal of
Marine and Freshwater Research 35:703-712.

Kitchell, J. F., C. Grimes, S. T. Lindley, D. Otis, and C. Schwarz. 2003. Report to the adaptive
management work group, Glen Canyon Dam Adaptive Management Program: An
independent review of ongoing and proposed scientific methods to assess the status and
trends of the Grand Canyon population of the humpback chub Gila cypha. Final report
submitted to the Grand Canyon Monitoring and Research Center, Flagstaff, Arizona.
Available: www.gcmrc.gov/library/reports/biological/Fish_studies/ Kitchell2003.pdf
(January 2008).

Knapp, R. A. and K. R. Matthews. 2000. Non-native fish introductions and the decline of the
mountain yellow-legged frog from within protected areas. Conservation Biology 14:428-
438.

Knapp, R. A., D. M. Boiano, and V. T. Vrendenburg. 2007. Removal of nonnative fish results in
population expansion of a declining amphibian (mountain yellow-legged frog, Rana
mucosa). Biological Conservation 135:11-20.

Korman, J. M. Kaplinski, J. E. Hazel, III, and T. S. Melis. 2005. Effects of the experimental
fluctuating flows from Glen Canyon Dam in 2003 and 2004 on the early life history stages
of rainbow trout in the Colorado River. Final Report, submitted to the Grand Canyon
Monitoring and Research Center, Flagstaff, AZ. Available: http://www.gcmrc.gov/library/
reports/biological/Fish_studies/Ecometric/Korman2005.pdf. (January 2008).

Ladson, A. R. and R. M. Argent. 2002. Adaptive management of environmental flows: lessons
for the Murray-Darling basin from three large North American rivers. Australian Journal of
Water Resources 5:89-102.

Ladson, L. and W. Allan. 2002. Solver. Frontline Systems, Incline Village, NV. Available:
http://www.solver.com. (January 2008).

Laslett, G. M., J. P. Eveson, and T. Polacheck. 2002. A flexible maximum likelihood approach
for fitting growth curves to tag-recapture data. Canadian Journal of Fisheries and Aquatic
Sciences 59:976-986.

Lee, K. N. 1999. Appraising adaptive management. Conservation Ecology 3(2): 3. Available:
http://www.consecol.org/vol3/iss2/art3/. (January 2008).

Lepak, J. M., C. E. Kraft, and B. C. Weidel. 2006. Rapid food web recovery in response to
removal of an introduced apex predator. Canadian Journal of Fisheries and Aquatic
Sciences 63:569-575.

Lessard, R. B., S. Martell, C. J. Walters, T. E. Essington, and J. F. Kitchell. 2005. Should
ecosystem management involve active control of species abundances? Ecology and
Society. 10(2):1. Available: http://www.ecology andsociety.org/voll0/iss2/artl/. (January
2008).









Lorenzen, K. 2000. Allometry of natural mortality as a basis for assessing optimal release size in
fish-stocking programmes. Canadian Journal of Fisheries and Aquatic Sciences 57:2374-
2381.

Lowe, S., M. Browne, S. Boudjelas, and M. De Poorter. 2000. 100 of the world's worst invasive
alien species: a selection from the global invasive species database. Invasive Species
Specialist Group, Species Survival Commission of the World Conservation Union.
Available: www.issg.org/booklet.pdf (January 2008).

Lunn, D. J., A. Thomas, N. Best, and D. Spiegelhalter. 2000. Winbugs -- a Bayesian modelling
framework: concepts, structure, and extensibility. Statistics and Computing 10:325-337.

MacKenzie, D. I. and W. L. Kendall. 2002. How should detection probability be incorporated
into estimates of relative abundance? Ecology 83:2387-2393.

MacKenzie, D. I., J. D. Nichols, J. A. Royle, K. H. Pollock, L. L. Bailey, and J. E. Hines. 2006.
Occupancy estimation and modeling: inferring patterns and dynamics of species
occurrence. Elsevier, Burlington, Massachusetts.

Makinster, A. S., R. S. Rogers, and W. R. Persons. 2007. Status of the Lee's Ferry trout fishery:
2006 Annual Report. 2006 Annual Report, submitted to the Grand Canyon Monitoring
and Research Center, Flagstaff, AZ. Arizona Game and Fish Department, Phoenix, AZ.

Mangel, M. and 41 coauthors. 1996. Principles for the conservation of wild living resources.
Ecological Applications 6:338-362.

Marsh, P. C. and M. E. Douglas. 1997. Predation by introduced fishes on endangered humpback
chub and other native species in the Little Colorado River, Arizona. Transactions of the
American Fisheries Society 126:343-346.

Martell, S. J. D. 2006. Review of the status and trends of humpback chub (Gila cypha) in the
Grand Canyon: 1989-2005. Report submitted to Grand Canyon Monitoring and Research
Center, Flagstaff, AZ.

McDowall, R. M. 2003. Impacts of introduced salmonids on native galaxiids in New Zealand
upland streams: a new look at an old problem. Transactions of the American Fisheries
Society 132:229-238.

McDowall, R. M. 2006. Crying wolf, crying foul, or crying shame: alien salmonids and a
biodiversity crisis in the southern cool-temperate galaxioid fishes? Reviews in Fish
Biology and Fisheries 16:233-422.

Megrey, B. A. 1989. Review and comparison of age structured stock assessment models from
theoretical and applied points of view. Pages 8-48 in E. F. Edwards and B. A. Megrey,
editors. Mathematical analysis offish stock dynamics. American Fisheries Society,
Symposium 6, Bethesda, Maryland.









Melis, T. S., S. J. D. Martell, L. G. Coggins, Jr., W. E. Pine, III, and M. E. Andersen. 2006.
Adaptive management of the Colorado River ecosystem below Glen Canyon Dam,
Arizona: using science and modeling to resolve uncertainty in river management. In
Proceedings of the American Water Resources Association, 2006. Available:
http://www.awra.org/ proceedings/cdrom.html#Missoula 2006. (January 2008).

Meyer, K. A., J. A. Lamansky, Jr., and D. J. Schill. 2006. Evaluation of an unsuccessful brook
trout electrofishing removal project in a small Rocky Mountain stream. North American
Journal of Fisheries Management 26:849-860.

Minckley, W. L. 1973. Fishes of Arizona. Arizona Game and Fish Department, Phoenix,
Arizona.

Minckley, W. L. 1991. Native fishes of the Grand Canyon region: An obituary? Pages 124-177
in Marzolf G. R., editor. Colorado River Ecology and Dam Management. National
Academy Press, Washington, DC.

Mueller, G. A. 2005. Predatory fish removal and native fish recovery in the Colorado River
mainstem: what have we learned? Fisheries 30:10-19.

Mueller, G. A., C. Walters, P. B. Holden, P. G. Walker, J. Landye, and B. Johnson. 1999.
Scientific review of the Glen Canyon Dam modifications to control downstream
temperatures plan and draft environmental assessment, January, 1999, and supporting
documentation. Prepared for the Bureau of Reclamation, Upper Colorado Region, Salt
Lake City, Utah. Available: http://www.gcmrc.gov/library/reports/Synthesis/
Muellerl999b.pdf (April 2008).

Nichols, J. D., F. A. Johnson, and B. K. Williams. 1995. Managing North American waterfowl in
the face of uncertainty. Annual Review of Ecology and Systematics 26:177-199.

Nilsson, C., C. A. Reidy, M. Dynesius, and C. Revenga. 2005. Fragmentation and flow
regulation of the world's largest river systems. Science 308:405-408.

NRC (National Research Council). 1987. River and dam management: a review of the Bureau of
Reclamation's Glen Canyon Environmental Studies. National Academy Press,
Washington, D.C.

NRC (National Research Council). 1996. River resource management in the Grand Canyon.
National Academy Press, Washington, D.C.

NRC (National Research Council). 1999. Downstream: adaptive management of Glen Canyon
Dam and the Colorado River Ecosystem. National Academy Press, Washington, D.C.

Olden, J. D. and N. L. Poff. 2005. Long-term trends of the native and non-native fish faunas in
the American Southwest. Animal Biodiversity and Conservation 28:75-89.









Olsen, D. G. and M. C. Belk. 2005. Relationship of diurnal habitat use of native stream fishes of
the eastern great basin to presence of introduced salmonids. Western North American
Naturalist 65:501-506.

Osmundson, D. B., R. J. Ryel, V. L. Lamarra, and J. Pitlick. 2002. Flow-sediment-biota
implications for river regulation effects on native fish abundance. Ecological Applications
12:1719-1739.

Otis, D. L., K. P. Burnham, D. R. Anderson, and G. C. White. 1978. Statistical inference from
capture data on closed populations. Wildlife Monographs 62:1-135.

Palmer, M. A. and 21 coauthors. 2005. Standards for ecologically successful river restoration.
Journal of Applied Ecology 42:208-217.

Palmer, M., J. D. Allan, J. Meyer, and E. S. Bernhardt. 2007. River restoration in the twenty-first
century: data and experiential knowledge to inform future efforts. Restoration Ecology
15:472-481.

Paloheimo, J. E. and L. M. Dickie. 1965. Food and growth of fishes. I. A growth curve derived
from experimental data. Journal of the Fisheries Research Board of Canada 22:521-542.

Paloheimo, J.E. and L.M. Dickie. 1966. Food and growth of fishes. II. Effects of food and
temperature on the relation between metabolism and body weight. Journal of the Fisheries
Research Board of Canada 23:869-908.

Parker, B. R., D. W. Schindler, D. B. Donald, and R. S. Anderson. 2001. The effects of stocking
and removal of a nonnative salmonid on the plankton of an alpine lake. Ecosystems 4:334-
345.

Parma, A. M. and 9 coauthors. 1998. What can adaptive management do for our fish, forests,
food, and biodiversity? Integrative Biology 1:16-26.

Pascual M. A. and R. Hilborn. 1995. Conservation of harvest populations in fluctuating
environments: the case of the Serengeti wildebeest. Journal of Applied Ecology 32:468-
480.

Patten, D. T. and L. E. Stevens. 1999. Restoration of the Colorado River ecosystem using
planned flooding. Ecological Applications 11:633-634.

Paukert, C. and R. S. Rogers. 2004. Factors affecting condition of flannelmouth suckers in the
Colorado River, Grand Canyon, Arizona. North American Journal of Fisheries
Management 24:648-653.

Paukert, C. P. and J. H. Petersen. 2007. Comparative growth and consumption potential of
rainbow trout and humpback chub in the Colorado River, Grand Canyon, Arizona, under
different temperature scenarios. Southwestern Naturalist 52:234-242.









Pauly, D. 1980. On the interrelationships between natural mortality, growth parameters, and
mean environmental temperature in 175 fish stocks. Journal du Conseil, Conseil
International pour L'Exploration de la Mer 39: 175-192.

Petersen, J. H. and C. P. Paukert. 2005. Development of a bioenergetics model for humpback
chub and evaluation of water temperature changes in the Grand Canyon, Colorado River.
Transactions of the American Fisheries Society 134:960-974.

Peterson, J. T., Thurow, R. F., and J. W. Guzevich. 2004. An evaluation of multipass
electrofishing for estimating the abundance of stream-dwelling salmonids. Transactions of
the American Fisheries Society 133:462-475.

Pine, W. E., III, K. H. Pollock, J. E. Hightower, T. J. Kwak, and J. A. Rice. 2003. A review of
tagging methods for estimating fish population size and components of mortality. Fisheries
28:10-23.

Pine, W. E., III, T. J. Kwak, and J. A. Rice. 2007. Modeling management scenarios and the
effects of an introduced apex predator on a coastal riverine fish community. Transactions
of the American Fisheries Society 136:105-120.

Poff, N. L. and 9 coauthors. 2003. River flows and water wars: emerging science for
environmental decision making. Frontiers in Ecology and the Environment 1:298-306.

Postel, S. L., G. C. Daily, and P. R. Ehrlich. 1996. Human appropriation of renewable fresh
water. Science 271:785-788.

Quinn, T. J., II and R. B. Deriso. 1999. Quantitative Fish Dynamics. Oxford University Press,
New York.

R Development Core Team. 2007. R: A language and environment for statistical computing. R
Foundation for Statistical Computing, Vienna, Austria. Available: http://www.R-project
.org. (January 2008)

Reisner, M. 1993. Cadillac desert. Penguin Books, New York.

Reynolds, J. B. 1996. Electrofishing. Pages 221-254 in B. R. Murphy and D. W. Willis, editors.
Fisheries techniques, 2nd edition. American Fisheries Society, Bethesda, Maryland.

Rice, J. A. and P. A. Cochran. 1984. Independent evaluation of a bioenergetics model for
largemouth bass. Ecology 65:732-739.

Ricker, W. E. 1975. Computation and interpretation of biological statistics of fish populations.
Fisheries Research Board of Canada Bulletin 191.

Robinson, A. T. and M. R. Childs. 2001. Juvenile growth of native fishes in the Little Colorado
River and in a thermally modified portion of the Colorado River. North American Journal
of Fisheries Management 21:809-815.









Robinson, A. T., S. D. Bryan, and M. G. Sweetser. 2003. Habitat use by non-native rainbow
trout, Oncorhynchus mykiss, and native Little Colorado spinedace, Lepidomeda vitta.
Environmental Biology of Fishes 68:205-214.

Rosenberger, A. E. and J. B. Dunham. 2005. Validation of abundance estimates from mark-
recapture and removal techniques for rainbow trout captured by electrofishing in small
streams. North American Journal of Fisheries Management 25:1395-1410.

Royle, J. A. and R. M. Dorazio. 2006. Hierarchical models of animal abundance and occurrence.
Journal of Agricultural, Biological, and Environmental Statistics 11:249-263.

Sainsbury, K. J. 1980. Effect of individual variability on the von Bertalanffy growth equation.
Canadian Journal of Fisheries and Aquatic Sciences 37:241-247.

Sainsbury, K. J. 1991. Application of an experimental approach to management of a tropical
multispecies fishery with highly uncertain dynamics. ICES Marine Science Symposium
193:301-320.

Sampson, D. B. and Y. Yin. 1998. A monte carlo evaluation of the stock synthesis program.
Pages 315-338 in F. Funk, T. J. Quinn II, J. Heifetz, J. N. Ianelli, J. E. Powers, J. F.
Schweigert, P. J. Sullivan, and C. I. Zhang, editors. Fishery Stock Assessment Models.
Alaska Sea Grant College Program Report No. AK-SG-98-01, University of Alaska
Fairbanks.

Schmidt, J. C., R. A. Parnell, P. E. Grams, J. E. Hazel, M. A. Kaplinski, L. E. Stevens, and T. L.
Hoffnagle. 2001. The 1996 controlled flood in Grand Canyon: flow, sediment transport,
and geomorphic change. Ecological Applications 11:657-671.

Schnute, J. 1981. A versatile growth model with statistically stable parameters. Canadian Journal
of Fisheries and Aquatic Science 38:1128-1140.

Schreiber, E. S. G., A. R. Bearlin, S. J. Nicol, and C. R. Todd. 2004. Adaptive management: a
synthesis of current understanding and effective application. Ecological Management and
Restoration 5:177-182.

Seager, R and 12 coauthors. 2007. Model projections of an imminent transition to a more arid
climate in southwestern North America. Science 316:1181-1184.

Seber, G. A. F. 1982. The estimation of animal abundance and related parameters, 2nd edition.
Griffin, London.

Sit, V. and B. Taylor, editors. 1998. Statistical methods for adaptive management studies.
Research Branch, British Columbia Ministry of Forests, Land Management Handbook No.
42, Victoria, British Columbia.

Speas, D. W., C. J. Walters, D. L. Ward, and R. S. Rogers. 2004. Effects of intraspecific density
and environmental variables on electrofishing catchability of brown and rainbow trout in
the Colorado River. North American Journal of Fisheries Management 24:586-596.









Stone, D. M. and O. T. Gorman. 2006. Ontogenesis of endangered humpback chub (Gila cypha)
in the Little Colorado River, Arizona. American Midland Naturalist 155:123-135.

Stone, D. M., D. R. Van Haverbeke, D. L. Ward, and T. A. Hunt. 2007. Dispersal of nonnative
fishes and parasites in the intermittent Little Colorado River, Arizona. Southwestern
Naturalist 52:130-137.

Taniguchi, Y., F. J. Rahel, D. C. Novinger, and K. G. Gerow. 1998. Temperature mediation of
competitive interactions among three fish species that replace each other along longitudinal
stream gradients. Canadian Journal of Fisheries and Aquatic Science 55:1894-1901.

Taylor, N. G., C. J. Walters, and S. J. D. Martell. 2005. A new likelihood for simultaneously
estimating von Bertalanffy growth parameters, gear selectivity, and natural and fishing
mortality. Canadian Journal of Fisheries and Aquatic Sciences 62:215-223.

Topping, D. J., D. M. Rubin, J. C. Schmidt, J. E. Hazel, Jr., T. S. Melis, S. A. Wright, M.
Kaplinski, A. E. Draut, and M. J. Breedlove. 2006. Comparison of sediment-transport and
bar-response results from the 1996 and 2004 controlled-flood experiments on the Colorado
River in Grand Canyon: Proceedings of the 8th Inter-Agency Sedimentation Conference,
April 2-6, 2006, Reno, Nevada.

Topping, D. J., J. C. Schmidt, and L. E. Vierra, Jr. 2003. Computation and analysis of the
instantaneous-discharge for the Colorado River at Lees Ferry, Arizona: May 8, 1921
through September 30, 2000. U.S. Geological Survey Professional Paper, Reston, Virginia,
1677. Available:http://pubs.usgs.gov/pp/pp1677/pdf/pp 677.pdf. (January 2008).

Trammell, M., R. Valdez, S. Carothers, and R. Ryel. 2002. Effects of a low steady summer flow
experiment on native fishes of the Colorado River in Grand Canyon, Arizona. Final Report
submitted to the Grand Canyon Monitoring and Research Center, Flagstaff, AZ.

Tyus, H. M. and J. F. Saunders, III. 2000. Nonnative fish control and endangered fish recovery:
lessons from the Colorado River. Fisheries. 25:17-24.

USDOI (U.S. Department of the Interior). 1995. Operation of Glen Canyon Dam Final
Environmental Impact Statement, Bureau of Reclamation, Salt Lake City, Utah.
Available: http://www.gcmrc.gov/library/reports/LawoftheRiver/ GCDEISFinal.pdf.
(January 2008).

USFWS (U.S. Fish and Wildlife Service). 1994. Final Biological Opinion, operation of Glen
Canyon Dam as the modified low fluctuating flow alternative of the Final Environmental
Impact Statement, operation of Glen Canyon Dam (2-21-93-F-167). Ecological Services,
Arizona State Office, U.S. Fish and Wildlife Service, Phoenix, AZ.

USFWS (U.S. Fish and Wildlife Service). 2002. Humpback chub (Gila cypha) Recovery Goals:
amendment and supplementation to the humpback chub Recovery Plan. U.S. Fish and
Wildlife Service, Mountain-Prairie Region (6). Denver, Colorado. Available:
http://www.gcmrc.gov/library/reports/biological/Fish_studies/USFWS/
Humpbackchub recoverygoals2002.pdf. (January 2008).









Valdez, R. A. and R. J. Ryel. 1995. Life history and ecology of the humpback chub (Gila cypha)
in the Colorado River, Arizona. Final Report. Contract No. 0-CS-40-09110. Salt Lake
City, UT. Available: http://www.gcmrc.gov/library/reports/biological/Fish_studies/
Biowest/Valdezl995f.pdf. (January 2008).

Valdez, R. A., S. W. Carothers, D. A. House, M. E. Douglas, M. Douglas, R. J. Ryel, K. R.
Bestgen, and D. L. Wegner. 2000. A program of experimental flows for the endangered
and native fishes of the Colorado River in Grand Canyon. Final Report, Grand Canyon
Monitoring and Research Center, Flagstaff, AZ. Available: http://www.gcmrc.gov/library/
reports/biological/Fish_studies/swca/99-FC-40-2260 LSSF/Valdez_2000_EXPFLOW
99FC402260.comp.PDF. (January 2008).

Van Haverbeke, D. R. and D. M. Stone. 2007. Stock assessment and fisheries monitoring
activities in the Little Colorado River within Grand Canyon during 2006. Final Report
submitted to the Grand Canyon Monitoring and Research Center. U.S. Fish and Wildlife
Service, Flagstaff, AZ.

Vernieu, W. S., S. J. Hueftle, and S. P. Gloss. 2005. Water quality in Lake Powell and the
Colorado River. Pages 69-85 in S. P. Gloss, J. L. Lovich, and T. S. Melis, editors. The
state of the Colorado River ecosystem in Grand Canyon. U.S. Geological Survey Circular
1282.

Viallefont, A., J. D. Lebreton, and A. M. Reboulet. 1998. Parameter identifiably and model
selection in capture-recapture models: a numerical approach. Biometrical Journal 40:313-
325.

Vitousek, P. M., H. A. Mooney, J. Lubchenco, and J. M. Melillo. 1997. Human domination of
the Earth's ecosystems. Science 277:494-499.

Voichick, N. and S. A. Wright. 2007. Water-temperature for the Colorado River and tributaries
between Glen Canyon Dam and Spencer Canyon, Northern Arizona 1988-2005. USGS,
Reston, Virginia. US Geological Data Series: 251.

Vrendenburg, V. T. 2004. Reversing introduced species effects: experimental removal of
introduced fish leads to rapid recovery of a declining frog. Proceedings of the National
Academy of Sciences 101:7646-7650.

Walters, C. J. 1986. Adaptive management of renewable resources. McMillan, New York.

Walters, C. J. 1997. Challenges in adaptive management of riparian and coastal ecosystems.
Conservation Ecology 1(2):1. Available: http://www.consecol.org/voll/iss2/artl/. (January
2008)

Walters, C. J. and C. S. Holling. 1990. Large-scale management experiments and learning by
doing. Ecology 71:2060-2068.

Walters, C. J. and Martell, S. J. D. 2004. Fisheries Ecology and Management. Princeton
University Press, Princeton, New Jersey.









Walters, C. J., L. H. Gunderson, and C. S. Holling. 1992. Experimental policies for water
management in the Everglades. Ecological Applications 2:189-202.

Walters, C., J. Korman, L. E. Stevens, and B. Gold. 2000. Ecosystem modeling for evaluation of
adaptive management policies in the Grand Canyon. Conservation Ecology 4(2): 1.
Available: http://www.consecol.org/vol4/iss2/artl/. (January 2008).

Wang, Y-G., M. R. Thomas, and I. F. Somers. 1995. A maximum likelihood approach for
estimating growth from tag-recapture data. Canadian Journal of Fisheries and Aquatic
Science 52:252-259.

Webb, R. H., J. S. Schmidt, G. R. Marzolf, and R. A. Valdez, editors. 1999. The controlled flood
in Grand Canyon. AGU Geophysical Monograph 110.

Williams, B. K., J. D. Nichols, and M. J. Conroy. 2002. Analysis and management of animal
populations. Academic Press, London, UK.

Williams, B. K., R. C. Szaro, and C. D. Shapiro. 2007. Adaptive Management: The U.S.
Department of Interior Technical Guide. Adaptive Management Working Group, U.S.
Department of Interior, Washington, D.C.

Wright, S. A., T. S. Melis, D. J. Topping, D. M. Rubin. 2005. Influence of Glen Canyon Dam
operations on downstream sand resources of the Colorado River in Grand Canyon. Pages
17-32 in S. P. Gloss, J. L. Lovich, and T. S. Melis, editors. The state of the Colorado River
ecosystem in Grand Canyon. U.S. Geological Survey Circular 1282.

Yard, M. D. 2003. Light availability and aquatic primary production: Colorado River, Glen and
Grand Canyons, AZ. Doctoral dissertation. Northern Arizona University, Flagstaff.

Yoccoz, N. G., J. D. Nichols, and T. Boulinier. 2001. Monitoring biological diversity in space
and time. Trends in Ecology and Evolution 16:446-453.

Zippin, C. 1956. An evaluation of the removal method of estimating animal populations.
Biometrics 12:163-189.









BIOGRAPHICAL SKETCH

I was born in Phoenix, AZ in 1967 to Lewis and Jan Coggins. After attending primary

school in both Phoenix and in Tuba City on the Navajo Reservation in Northern AZ, I graduated

from Coconino High School in Flagstaff, AZ in 1985. During these formative years, I was

fortunate to have a father who loved exploring Northern Arizona and particularly Marble and

Grand Canyons. Our explorations and his teachings of the natural world had a profound effect

on my career choices. After graduating from the University of Arizona in 1990 with a degree in

ecology and evolutionary biology, I moved to Alaska where I worked as a fisheries biologist for

the Alaska Department of Fish and Game in Bristol Bay and Kodiak. I married Jennifer (Gape)

Coggins in December 1993 and our first child, Elizabeth Tate, was born in Kodiak, AK in 1997.

While employed by the state, I was able to pursue and earn a master's degree in fisheries from

the University of Alaska, Fairbanks, in 1997, under the guidance of Dr. Terry Quinn. I returned

to Flagstaff in 1999 to study Grand Canyon fishes with the U.S. Geological Survey. In August

2002, our second child, Annie Rose, was born in Flagstaff, AZ.





PAGE 1

ACTIVE ADAPTIVE MANAGEMENT FO R NATIVE FISH CONSERVATION IN THE GRAND CANYON: IMPLEMENTATION AND EVALUATION By LEWIS GEORGE COGGINS, JR. A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008 1

PAGE 2

2008 Lewis G. Coggins, Jr. 2

PAGE 3

For Dad 3

PAGE 4

ACKNOWLEDGMENTS This research was funded by the U.S. Geological Survey through the Grand Canyon Monitoring and Research Center. I thank Dr Ted Melis, Dr Barbara Ralston, and Dr. Denny Fenn for supporting the agreement that allowed me to return to school. I also thank John Hamill and Matthew Andersen for their continued support. I thank my friend and colleague Dr. Mike Yard for his collaboration in the design and implem entation of the mechanical removal project. I thank Clay Nelson and his staff wi th the Arizona Game and Fish Department for data collection during the 2005-2006 mechanical rem oval efforts. I also thank Ca rol Fritzinger, Brian Dierker, Peter Weiss, Brent Berger, Steve Jones, Stew art Reeder, Danny Martinez, Yael Bernstein, Courtney Giaque, Emily Thompson, Dave Doring, Dave Baker, Melanie Caron, Scotty Davis, Ted Kennedy, Ally Martinez, Scott Perry, Lynn R hoder, Park Stefensen, John Taylor, Todd and Erica Tietjen, and Josh Winiecki for outstanding field and logistical support. I also would like to thank field biologists and tec hnicians from ASU, Bio West, USFWS, AGFD, and SWCA for their efforts to collect data on Grand Canyon fishes for the past two decades. I thank my advisor, Dr. William Pine, and the rest of my supervisory committee (Dr. Carl Walters, Dr. Micheal Allen, Dr. Tom Frazer, and Dr. Christina St audhammer) for their service and assistance with my research. I especially th ank Dr. Pine for encouraging me to return to school and for his assistance and guidance in maki ng it a rewarding and fruitful experience. I also owe a tremendous debt to my mentor and tr usted friend, Dr. Carl Walters, who has assisted and guided much of my research ac tivity for the last 8 years. I thank Dr. Steve Martell for his assistance in evaluating humpback chub stock assessment models. Finally, I thank my wife Jennifer and our daught ers Ellie and Annie for their love, support, and understanding of the long hours we were forced to endure apart. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS................................................................................................................. LIST OF TABLES................................................................................................................. ............ LIST OF FIGURES........................................................................................................................... ABSTRACT....................................................................................................................................... CHAPTER 1 GENERAL INTRODUCTION..............................................................................................14 2 NON-NATIVE FISH CONTROL IN THE COLORADO RIVER IN GRAND CANYON, ARIZONA: AN EFFECTIV E PROGRAM OR SERENDIPITOUS TIMING?................................................................................................................................16 Adaptive Management in Grand Canyon........................................................................18 Fish Community Background..........................................................................................19 Objective..........................................................................................................................20 Methods..................................................................................................................................21 Mechanical Removal Reach: Study Areas and Field Protocols......................................21 Control Reach: Study Areas and Field Protocols............................................................23 Mechanical Removal Reach: Data Analysis...................................................................23 Control Reach: Data Analysis.........................................................................................27 Results.....................................................................................................................................29 Mechanical Removal Reach............................................................................................29 Control Reach..................................................................................................................31 Comparison of Mechanical Removal and Control Reaches............................................32 Discussion...............................................................................................................................33 Mechanical Removal: Effective Program?......................................................................33 Serendipitous Timing: What Led to the D ecline of Rainbow Trout in the Control Reach?..........................................................................................................................34 Other Species...................................................................................................................36 Bias in Capture Probability and Abundance Estimates...................................................36 Recommendations for Future Me chanical Removal Operations.....................................37 Adaptive Management in Grand Canyon: The Future....................................................38 3 DEVELOPMENT OF A TEMPERATURE-DEPENDENT GROWTH MODEL FOR THE ENDANGERED HUMPBACK CHUB USING MARK-RECAPTURE DATA.........57 Methods..................................................................................................................................59 Results.....................................................................................................................................67 Discussion...............................................................................................................................70 5

PAGE 6

4 ABUNDANCE TRENDS AND STATUS OF THE LITTLE COLORADO RIVER POPULATION OF HUMPBACK CHUB: AN UPDATE CONSIDERING DATA 1989-2006...............................................................................................................................83 Methods..................................................................................................................................86 Index-Based Metrics........................................................................................................86 Tagging-Based Metrics...................................................................................................87 Evaluating Model Fit.......................................................................................................90 Incorporation of Ageing E rror in ASMR Assessments...................................................92 Results.....................................................................................................................................95 Index-Based Assessments...............................................................................................95 Tagging-Based Assessments...........................................................................................96 Closed Population Models...............................................................................................97 ASMR Without Tag Cohort Specific Data......................................................................97 Model Evaluation and Selection......................................................................................98 ASMR with Tag Cohort Specific Data..........................................................................100 Model Evaluation and Selection....................................................................................100 Incorporation of Ageing E rror in ASMR Assessments.................................................100 Results Summary...........................................................................................................102 Discussion.............................................................................................................................102 5 LINKING TEMPORAL PATTERNS IN FI SHERY RESOURCES WITH ADAPTIVE MANAGEMENT: WHAT HAVE WE LEARNED AND ARE WE MANAGING ADAPTIVELY?...................................................................................................................134 Description of Adaptive Management Actions....................................................................136 1996 Experimental High Flow......................................................................................136 2000 Low Summer Steady Flow...................................................................................136 2004 Experimental High Flow......................................................................................137 2003-2005 Non-native Fish Suppression Flows............................................................137 2003-2006 Mechanical Removal of Non-native Fish...................................................138 Description of Uncontrolled Factors....................................................................................138 Paria, Little Colorado, and Colorado River Hydrology................................................139 Release Water Temperature from Glen Canyon Dam...................................................140 Juvenile Native Fish Production in the Little Colorado River......................................140 How Are Fish Populations Affected by Adaptive Management Actions and Uncontrolled Factors?.......................................................................................................142 Has Increased Turbidity Affected Fish Populations in Grand Canyon?.......................142 Has Reduced Non-native Fish Abundance and Increased Temperature Affected Native Fish Populations in Grand Canyon?...............................................................143 Conclusions and Recommendations.....................................................................................145 LIST OF REFERENCES.............................................................................................................159 BIOGRAPHICAL SKETCH.......................................................................................................173 6

PAGE 7

LIST OF TABLES Table page 2-1 Electrofishing catch by species in the mechanical removal reach for each month, 2003-2006..........................................................................................................................41 2-2 Estimated abundance and density of rainbow trout in the mechanical removal reach at the beginning of each month, 2003-2006.......................................................................42 2-3 Electrofishing catch by species in th e control reach for each month, 2003-2006.............43 2-4 Estimated abundance and density of rai nbow trout in the control reach at the beginning of each month, 2003-2006................................................................................44 3-1 General growth model results............................................................................................74 3-2 Parameter correlation matrix for the temperature-independent growth model..................75 3-3 Parameter correlation matrix for the temperature-dependent growth model.....................76 4-1 AIC model evaluation results among AS MR models fit to data pooled among tag cohort......................................................................................................................... ......108 4-2 AIC model evaluation results among ASMR models fit to data stratified by tag cohort......................................................................................................................... ......109 7

PAGE 8

LIST OF FIGURES Figure page 2-1 Map of the mechanical removal reach of the Colorado River within Grand Canyon, Arizona...............................................................................................................................45 2-2 Map of the control reach of the Co lorado River within Grand Canyon, Arizona.............46 2-3 Percent composition (A) and number of fish (B) by species captured with electrofishing in the mechanical removal reach among months, 2003-2006.....................47 2-4 Estimated catch rate (A), abundance (B), and capture probability (C) for rainbow trout in both the upstream and downstream st rata in the mechanical removal reach among months, 2003-2006.................................................................................................48 2-5 Net immigration rate into the upstrea m (A) and downstream (B) strata of the mechanical removal reach within time intervals between January 2003 and August 2006....................................................................................................................................49 2-6 Probability plots for coefficient valu es influencing capture probability among 10 trips (January 2003-July 2003; January 2004-September 2004).......................................50 2-7 Estimated catch rate (A), abundance (B), and capture probability (C) for rainbow trout in the control reach among months, 2003-2006........................................................51 2-8 Estimated monthly survival rate of ra inbow trout in the control reach during 20032006....................................................................................................................................52 2-9 Estimated rainbow trout abundance in both the mechanical removal and control reaches at the beginning of each trip during 2003-2006....................................................53 2-10 Estimated total length and 95% confidence intervals of rainbow trout captured in both the mechanical removal and control reaches during 2003-2006...............................54 2-11 Length frequency distributions of rainbow trout captured using electrofishing in the Colorado River from river mile -15 to river mile 56.........................................................55 2-12 Daily mean water temperatures observe d in the Colorado River at approximately river mile 61, 1990-2006....................................................................................................56 3-1 Predicted and observed growth rate (dL/dt) as a function of tota l length (TL) at the start of the time interval..................................................................................................... 77 3-2 Observed and predicted monthly Litt le Colorado River water temperature......................78 8

PAGE 9

3-3 Observed and predicted humpback chub gr owth rate (dL/dt) from the temperatureindependent growth model and the temp erature-dependent growth model during summer and winter.............................................................................................................7 9 3-4 Observed and predicted humpback chub gr owth rate (dL/dt) from the temperaturedependent growth model dur ing summer and winter........................................................80 3-5 Predicted humpback chub length-at-age from the U.S. Fish and Wildlife Service (USFWS) growth curve, the temperat ure-independent growth model, the temperature-dependent growth model for th e Little Colorado River (LCR) humpback chub population, and the temperature-depe ndent growth model for humpback chub living in the mainstem Colorado River under a constant temp erature of 10C.................81 3-6 Predicted monthly growth rate from the temperature-dependent growth model for the Little Colorado River (LCR) population of humpback chub and for humpback chub living in the mainstem Colorado River under a constant temp erature of 10C.................82 4-1 Relative abundance indices of sub-adult (150-199 mm total le ngth; TL) and adult ( 200 mm TL) humpback chub based on hoop net catch rate (fish/hour) in the lower 1,200 m section of the Little Colorado Ri ver (A) and trammel net catch rate (fish/hour/100 m) of adult hum pback chub in the Little Colorado River inflow reach of the Colorado River (B)................................................................................................110 4-2 Numbers of humpback chub marked (A ) and recaptured (B) by age and year...............111 4-3a Numbers of fish marked by age in years 1989 (A), 1990 (B), 1991 (C), and 1992 (D) indicated by dark circles and subsequen tly recaptured (light circles) by age and years.................................................................................................................................112 4-3b Numbers of fish marked by age in years 1993 (A), 1994 (B), 1995 (C), and 1996 (D) indicated by dark circles and subsequen tly recaptured (light circles) by age and years.................................................................................................................................113 4-3c Numbers of fish marked by age in years 1997 (A), 1998 (B), 1999 (C), and 2000 (D) indicated by dark circles and subsequen tly recaptured (light circles) by age and years.................................................................................................................................114 4-3d Numbers of fish marked by age in years 2001 (A), 2002 (B), 2003 (C), and 2004 (D) indicated by dark circles and subsequen tly recaptured (light circles) by age and years.................................................................................................................................115 4-3e Numbers of fish marked by age in years 2005 (A) and 2006 (B) indicated by dark circles and subsequently recaptured (light circles) by age and years..............................116 4-4 Mark-recapture closed population mode l estimates of humpback chub abundance 150 mm total length in th e Little Colorado River............................................................117 9

PAGE 10

4-5 Humpback chub adult abundance (age-4+) estimates from the ASMR 1 (A), ASMR 2 (B), and ASMR 3 (C) models using data pooled among tag cohorts...........................118 4-6 Humpback chub recruit abundance (age-2) estimates from the ASMR 1 (A), ASMR 2 (B), and ASMR 3 (C) models using data pooled among tag cohorts...........................119 4-7 Pearson residual plots for model ASMR 1 using data pooled among tag cohorts...........120 4-8 Pearson residual plots for model ASMR 2 using data pooled among tag cohorts...........121 4-9 Pearson residual plots for model ASMR 3 using data pooled among tag cohorts...........122 4-10 Capture probability by age and year estimated from model ASMR 3 using data pooled among tag cohorts................................................................................................123 4-11 Humpback chub adult abundance (age-4+) estimates from the ASMR 1 (A), ASMR 2 (B), and ASMR 3 (C) models using data stratified by tag cohort.................................124 4-12 Humpback chub recruit abundance (age-2) estimates from the ASMR 1 (A), ASMR 2 (B), and ASMR 3 (C) models using data stratified by tag cohort.................................125 4-13 Pearson residual plots for model ASMR 1 using data stratified by tag cohort................126 4-14 Pearson residual plots for model ASMR 2 using data stratified by tag cohort................127 4-15 Pearson residual plots for model ASMR 3 using data stratified by tag cohort................128 4-16 Capture probability by age and year estimated from model ASMR 3 using data stratified by tag cohort.....................................................................................................12 9 4-17 Seasonal probability surfaces of age for a particular length bin......................................130 4-18 Estimated adult abundance (age-4+) from ASMR 3 incorporating uncertainty in assignment of age.............................................................................................................1 31 4-19 Estimated recruit abundance (age-1) from ASMR 3 incorporat ing uncertainty in assignment of age.............................................................................................................1 32 4-20 Retrospective analysis of adult abundance (A) and mort ality rate (B) considering datasets beginning in 1989 and ending in th e year indicated in the figure legend..........133 5-1 Discharge in ft3/s (cfs) for the Colorado River at Lees Ferry (A), the Paria River at Lees Ferry (B), and the Little Colo rado River at Cameron, AZ (C), 1990-2006............150 5-2 Daily mean water temperatures observe d in the Colorado River at approximately river mile 61, 1990-2006..................................................................................................151 10

PAGE 11

11 5-3 Monthly electrofishing ca tch rate (fish/hour) in the Colorado River between river mile (RM) 56.3 and RM 65.7 for rainbow trout (A), brown trout (B), common carp (C), and fathead minnow (D)...........................................................................................152 5-4 Monthly electrofishing ca tch rate (fish/hour) in the Colorado River between river mile (RM) 56.3 and RM 65.7 for humpback chub (A), flannelmouth sucker (B), bluehead sucker (C), and speckled dace (D)....................................................................153 5-5 Monthly hoop net catch rate (fish/hour) in the Colorado River between river mile (RM) 63.7 and RM 64.2 for humpback chub (A), flannelmouth sucker (B), bluehead sucker (C), and speckled dace (D)...................................................................................154 5-6 Monthly average total length (TL; mm) observed in electrofishing sampling in the Colorado River between river mile (RM) 56.3 and RM 65.7 for rainbow trout (A), brown trout (B), common carp (C), and fathead minnow (D).........................................155 5-7 Monthly average total length (TL; mm) observed in electrofishing sampling in the Colorado River between river mile (RM) 56.3 and RM 65.7 for humpback chub (A), flannelmouth sucker (B), bluehead sucker (C), and speckled dace (D)...........................156 5-8 Monthly average total length (TL; mm) observed in hoop net sampling in the Colorado River between river mile (RM) 63.7 and RM 64.2 for humpback chub (A), flannelmouth sucker (B), bluehead sucker (C), and speckled dace (D)...........................157 5-9 Smoothed kernel density plot of the to tal length of flannelmouth sucker captured with electrofishing in the mechani cal removal reach during 2003 and 2006..................158

PAGE 12

Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ACTIVE ADAPTIVE MANAGEMENT FO R NATIVE FISH CONSERVATION IN THE GRAND CANYON: IMPLEMENTATION AND EVALUATION By Lewis George Coggins, Jr. May 2008 Chair: William E. Pine, III Major: Fisheries a nd Aquatic Sciences My first objective was to evaluate the efficacy of a large scale non-native fish removal effort to benefit endemic fishes of the Colorado River within Grand Canyon. During 2003-2006, over 23,000 non-native fish, primarily rainbow trout Oncorhynchus mykiss were removed from a 9.4 mile reach of the Colorado River. These removals resulted in a rapid shift in fish community composition from one dominated by co ld water salmonids (>90%), to one dominated by native fishes and the non-native fathead minnow Pimephales promelas (>90%). Concurrent with the mechanical removal, data collected within a control reach of the river suggested a systemic decline in rainbow trout unrelated to the fish removal effort. Thus, the efficacy of the mechanical removal was aided by an external systemic decline, particularly in 2005-2006. My second objective was to improve cu rrent knowledge of humpback chub Gila cypha growth to aid in length-based age determination, and to provide a tool to evaluate temperaturedependent changes in growth rate. I estimated a temperature-dependent growth function for humpback chub by predicting more than 14,000 gr owth increments from a mark-recapture database. Results suggest that humpback chub gr owth is strongly dependent on temperature and that previous growth curves based on paired age-length data tend to over-estimate the age of small fish and under-estimate the age of large fish. 12

PAGE 13

13 My third objective was to update humpback chub stock assessment procedures following guidance from an external review panel. Th ese recommendations were primarily to develop model selection procedures and to evaluate the eff ect of error in length-b ased age determination. I used both Pearson residual analysis and Akaike Information Criterion to evaluate candidate models leading to the conclu sion that the most general asse ssment model was required to adequately model patterns in capture probability I used the temperature-dependent growth relationship to estimate probabi listic relationships between ag e and length. These age-length relationships were then used in Monte Carlo simu lations to capture the effect of ageing error on subsequent estimates of recruitment and adult abundance. The results indicate that the adult humpback chub population has likely in creased between 20-25% since 2001. My fourth objective was to evaluate whether there was any evidence of effect from past adaptive management actions or uncontrollable factors on Grand Canyon fish populations, and to make recommendations for further adaptive management program development. These results are largely inconclusive except that the combin ed policy of mechanical removal and increased water temperatures is temporally correlated with increased native fish abundance in the mainstem Colorado River near the confluence of the Little Colorado River, a reach deemed critical habitat for humpback chub. I recommend that the adaptive management program invest additional effort in developing more explicit a nd measurable resource goals, particularly for focal Colorado River resources. I further recommend that additional investment in monitoring of juvenile native fish survival and growth in the ma instem is needed to adequately evaluate future adaptive management experiments. Finally, addi tional predictive capabili ty is needed to both formalize a priori hypotheses about juvenile native fish su rvival and recruitment, and to screen future policy options.

PAGE 14

CHAPTER 1 GENERAL INTRODUCTION Modifications to river ecosystems to serve human interests are a ub iquitous feature of human occupied landscapes. Post el et al. (1996) estimated that 54% of global annual runoff was appropriated for human use in 1996, and forecast that this figure might approach 70% by the year 2025. A recent review by Nilsson et al. (2005) documented that 50% of the Earths large river systems are fragmented by dams. Thus, anthropogenic modifications are a major, and frequently detrimental, influence on ri verine ecosystems on a global basis. Alterations to riverine ecosystems in the U.S. have led to an increase in river restoration projects and an active dialogue be tween scientists and policy makers (Poff et al. 2003). This has led to increased research in bot h river restoration science and th e appropriate measures of river restoration success (Palmer et al. 2005). However, scientists frequently are unable to predict with great certainty the outcome of management actions designed to achieve restoration, and this uncertainty can lead to skepticism and mistrust on the part of policy makers. Given this uncertainty, adaptive management (Holling 1978, Walters 1986) has been widely advocated as a strategy to guide restoration programs (Poff et al. 2003). Adaptive management recognizes that predictions of system response to management actions are uncertain, a nd seeks to use thoughtful application of management actions to learn about system behavior and hence, how to achieve resource management goals. The 1,470 mile course of the Colorado River begins at high altitude in the Rocky Mountains and terminates at the northern extent of the Gulf of California. The Colorado River drainage area encompasses seven U.S. states (AZ, CA, CO, NM, NV, UT, WY) and a small portion of northwestern Mexico (Benke and Cu shing 2005). Described as the Life-blood of the southwestern U.S. (Reisner 1993) and one of the most highly regulated rivers in the world 14

PAGE 15

15 (Nilsson et al. 2005), this river has enormous social, economic, recreational, and political importance. Additionally, the Grand Canyon of the Colorado River is widely recognized as one of the 7 natural wonders of the world and a na tional treasure of the United States. Following recognition of degraded conditions in th e Grand Canyon reach of the Colorado River downstream of Glen Canyon Dam (NRC 1987), the Glen Canyon Dam Adaptive Management Program has attempted to use adaptive management for river restoration since its formation in 1996. A focal resource of the Glen Canyon Dam Ad aptive Management program is the native fishes endemic to this basin, particularly the federally listed e ndangered humpback chub Gila cypha This dissertation is focused on evaluating the efficacy of the implementation of a specific adaptive management experiment and developing improved monitoring capability for humpback chub. In Chapter 2, I describe and evaluate an adaptive management experiment to remove nonnative fish from a large section of the Colorado River heavily used and deemed critical for humpback chub and other native fishes. In Chapter 3, I used mark-recapture information to develop a temperature-depende nt humpback chub growth model in support of improved monitoring and stock assessment. In Chapter 4, I used the growth model along with all available monitoring data through 2006 to provide an updated evaluation of recent humpback chub population dynamics. In Chapter 5, I provide a s ynthesis of available fish monitoring data to evaluate the effect of past adaptive mana gement experiments, as well as to make recommendations for future adaptive management experimentation, mo nitoring, and research priorities.

PAGE 16

CHAPTER 2 NON-NATIVE FISH CONTROL IN THE COLORADO RIVER IN GRAND CANYON, ARIZONA: AN EFFECTIVE PROGRAM OR SERENDIPITOUS TIMING? Harvest, species introductions, and large-scale habitat alterati ons have resulted in dramatic changes in the structure and function of ecosyst ems on a global basis (Vitousek et al. 1997). Humans have modified all types of ecosystems through various interventions, yet despite more than a century of focused ecological researc h, there remains much uncertainty as to how ecosystems will respond to anthropogenic in terventions (Holling 1973; Walters and Holling 1990). Partly as a result of this uncertainty, efforts to manage human activities using prescriptive science-based policies to achieve basic goals as th ey relate to ecosystems (e.g., sustainability or species conservation) have been widely unsuccessful (Christens en et al. 1996; Mangel et al. 1996). In response to these fa ilures, adaptive environmental assessment and management or adaptive management (AM) has been proposed as a strategy to link scientific inquiry and natural resource management (Holling 1978; Walters 1986; Walters and Holling 1990). Adaptive management assumes that successful management of natural resources can occur only if objectives are clearly de fined and future management policy choices are informed and directed by past policy performance. Holling and Walters (1990) describe three classes of AM implementation: (1) trial and error, (2) passiv e adaptive, and (3) active adaptive. Trial and error is a structure where in itial policy choices are complete ly uninformed, and later policy choices are selected from the set of best perfor ming initial choices. A passive adaptive structure chooses polices based on historic data informi ng a single predictive model. A passive adaptive structure differs from trial and error by the use of a predictive model to screen policy choices. This model can be either concep tual or quantitative and is pres umed to be accurate until proven otherwise. Finally, active adaptive manageme nt recognizes that there are usually multiple predictive models that can explain the historic data equally well, and seeks to implement specific 16

PAGE 17

policies that can both optimize s hort-term system performance a nd provide insight into which model provides the best predictions The predictive models then become hypotheses of system behavior under different management policies. If it is possible to quantif y the likelihood of each hypothesis a priori then policy choices are further selected considering tradeoffs in the future value of increased understanding of system beha vior. Most basically, the concept of AM embraces three related ideas: (1) Predictive models of complex systems can never be fully trusted in their ability to structure management policies that unambiguous ly attain specific system objectives, (2) Detailed research into the processes that define system complexity (e.g., resilience, feedbacks, thresholds, or alternative stable states) can never fully resolve prediction ambiguity, and (3) Predictive models contain key uncertainties that may only be resolved (if ever) by observing the response of the system to particular interventions. Though AM has been widely adopted as a conceptual strategy for the management of natural resources (Williams et al. 2007) including : waterfowl (e.g., Nichols et al. 1995), forests (e.g., Sit and Taylor 1998), wildlife (e.g., Pasc ual and Hilborn 1995), fi sheries (e.g., Sainsbury 1991), large river systems (e.g., NRC 1999), wetlands (e.g., Walters et al. 1992), and others, critics argue that many management programs that supposedly operate within an adaptive framework have embraced this term as a buzz word, but fail to apply the strategy as originally proposed (Gunderson 1999; Lee 1999). A frequent failure in AM programs has been that following the rigorous knowledge as sessment and modeling that charact erizes the initial steps of the process, the subsequent implementation and monitoring of candi date policies is not completed (Walters 1997; Gunderson 1999; Ladson and Argent 2002; Schr eiber et al. 2004). I have extensively reviewed the primary literature and found few true empiri cal tests of AM as a strategy for management of ecosystems (e.g., Alla n and Curtis 2005). This study documents the 17

PAGE 18

implementation of an ecosystem-scale adaptive management experiment in the Colorado River within Grand Canyon, Arizona. Adaptive Management in Grand Canyon Following the Final Record of Decision from the Environmental Impact Statement on the operation of Glen Canyon Dam (USDOI 1995), the Glen Canyon Dam Adaptive Management Program was formed and charged with managi ng the Colorado River ecosystem (CRE) within Grand Canyon, Arizona. This program consists of a multi-stakeholder federal advisory committee that defines objectives for the CRE a nd makes recommendations to the U.S. Secretary of Interior regarding the operation of Glen Canyon Dam (GCD) and other management actions. This high-profile program is arguably the most successful example of an adaptive management program in a large U.S. river (Ladson and Argent 2002). However, this recognition is primarily the result of short-term experi mentation with GCD operations designed to test policies for sediment conservation (Collier et al. 1997). Native fish conservation is also a key goal of the Glen Canyon Dam Adaptive Management Program primarily because many of the species endemic to the Colorado River Basin are protected under the US Endangered Sp ecies Act (ESA). This protected status necessitates regular review of GC D operations to ensure that da m operations are not deleterious to Grand Canyon native fish stocks Current knowledge suggests that likely factors influencing the population dynamics and ultimate recovery (as defined by the ESA mandated recovery criteria) of native fish in Gr and Canyon include: (1) non-native fish (Gorman et al. 2005; Olden and Poff 2005), (2) water temperature (Robi nson and Childs 2001), (3) flow regulation (Osmundson et al. 2002), (4) juve nile rearing habitat (Stone an d Gorman 2006), and (5) parasites and disease (Choudury et al. 2004). Of these, previous modeling and data analyses have shown that factors 1-3 are likely domina nt drivers of native fish popul ation dynamics in this system 18

PAGE 19

(Walters et al. 2000), and suggests that improving rearing conditions for native fish in the mainstem Colorado River will likely provide th e most significant benefi t to native fish. Additionally, of the factors possibly influencing native fish population dynamics, controlled manipulation of factors 1-3 in an experimental fr amework is most tenable and, in recent years, has been the focus of efforts in adaptive management for native fish conservation. Beginning in 2003, the first multi-year program of experimentation specifically designed to test policies associated with native fish conservation was implemented in Grand Canyon. In January 2003, an experiment was begun to expe rimentally manipulate GCD operations and the abundance of non-native fishes in a 9.4 mile st retch of the Colorado River containing known critical habitat for humpback chub Gila cypha and other native fish species. Although the experimental design also called for manipulation of water temper ature discharged from GCD in subsequent years (Coggins et al. 2002), only the experimental fish manipulations were implemented. The last of four years of mani pulating the abundance of non-native fishes was 2006. Fish Community Background Over much of the last several decades, the fish community in the Grand Canyon stretch of the Colorado River has been dominated by the non-native salmonids rainbow trout Oncorhynchus mykiss and brown trout Salmo trutta (Gloss and Coggins 2005). Introductions of non-native salmonids have been shown to advers ely impact invertebrate (Parker et al. 2001), amphibian (Knapp and Matthews 2000), and fish (McDowall 2003) communities. These two species of fish have also been identified as among the top 100 wors t invasive species (Lowe et al. 2000) principally because of the global scope of introductions rainbow trout have been successfully established on every continent with the exception of Antarcti ca (Crawford and Muir In Press ). Although it is unclear how detrimental thes e fish are to native fish in the Colorado 19

PAGE 20

River, interactions with various non-native fish have been widely implicated in the decline of southwestern native fishes (Minckley 1991; Tyus and Saunders 2000). Non-native salmonids, particularly brown trout, have been shown to be predators of native fishes (Valdez and Ryel 1995; Marsh and Douglas 1997) in Grand Canyon and rainbow trout predation on native fish has also been documented in other southwestern U.S. systems (Blinn et al. 1993). Besides direct mortality through predation, both rainbow trout and brown trou t have demonstrated other negative interactions with native fish in western U.S. river systems including interference competition, habitat displacement, and agonistic behavior (Blinn et al. 1993; Taniguchi et al. 1998; Robinson et al. 2003; Olsen and Belk 2005). These lethal and sub-lethal effects of interactions with native fishes have also been widely documented in New Zealand, Australia, Patagonia, and South Africa (McDowall 2006). Objective While control of non-native species is widely considered as a management option, it is rarely implemented and evaluated (Lessard et al. 2005; Pine et al. 2007), particularly for fish in large river systems. Removal of non-native organisms to potentially benefit native species is more frequently conducted in small streams (e.g., Meyer et al. 2006), in lakes and reservoirs (e.g., Hoffman et al. 2004; Vrendenburg 2004; Lepak et al. 2006) and in te rrestrial environments (e.g., Erskine-Ogden and Rejmanek 2005; Donlan et al. 2007). However, recently much effort has been expended to remove or reduce non-nati ve fishes in the Colorado River (Tyus and Saunders 2000). Unfortunately, little documentation is available to evaluate the efficacy of these efforts (Mueller 2005). This study describes one su ch effort and evaluates the efficacy of a program to reduce non-native fishes within humpback chub critical habitat. Given the ecological and management interest in non-native species removals, this porti on of the GCD adaptive management program also represents an important first phase of active adaptive management to 20

PAGE 21

benefit a focal biological resource, humpback chub. Specifically, the obj ectives of this study were to: (1) evaluate the effec tiveness of non-native control e fforts in the mainstem Colorado River, (2) investigate factors contributing to the effectivenes s of control efforts, and (3) characterize changes in the non-nati ve and native fish communities. Methods Mechanical Removal Reach: Study Areas and Field Protocols The Little Colorado River (LCR) inflow reach of the Colorado River extends from 56.3 river mile (RM) to 65.7 RM, as measured dow nstream from 0 RM at Lees Ferry, and is recognized as having the highest abundance of adult and juvenile humpback chub in the Colorado River (Valdez and Ryel 1995; Figure 21). This reach also has a relatively high abundance of flannelmouth sucker Catostomus latipinus, bluehead sucker Catostomus discobolus and speckled dace Rhinichthys osculus owing to the availability of spawning and rearing habitat in the LCR. Given the importance of this reach to native fishes, the LCR inflow reach was selected as the area to test non-native mechanical rem oval efforts and was divided into six river sections labeled A-F (F igure 2-1). Sections A and B ar e the right and left shore from RM 56.3 to RM 61.8. Sections C and D are the right and left shore between RM 61.8 to RM 62.1 and include the LCR confluence and the mixing zone below the LCR. Sections E and F are the right and left shore downstream of th e LCR confluence from RM 62.1 to RM 65.7. I stratified the study area into these 6 sections to control for the e ffect of the LCR discharge into the mainstem Colorado River. Sections A and B are unaffected by the tributary and sections E and F are believed to be of suff icient distance downstream of th e mixing zone to be affected uniformly throughout. Sections C and D include the LCR confluence and will be differentially affected by LCR discharge throughout their lengths Within river sections A-B and E-F, the 21

PAGE 22

shoreline was divided into 500 m sites. The num ber of sites within ea ch river section was: A=19, B=19, E=13, and F=14. Sections C and D constitute single sites. From January 2003 through August 2006, a total of 23 field trips were conducted to remove non-native fish with serial depletion passes using boat-mounted electrofishing within the mechanical removal reach. The majority of these trips removed fish during either 4 or 5 depletion passes; exceptions were in August 2003 (2 passes), September 2003 (3 passes), and July 2004 (6 passes). All sites w ithin sections A-B, C, and E-F were sampled during each pass. Section D, encompassing the LCR confluence, was not sampled during any of the trips due to concerns about equipment damage associated with high water conductivity issuing from the LCR and possibly high native fish abundance near the confluence. All electrofishing occurred following the onset of darkness and each depl etion pass required 2 nights to complete. Electrofishing crews consisted of a boat operator and a single ne tter. Two boat types (15-foot rubber-hulled sport boat and 15-foot aluminum-hulle d sport boat) and two types of electrofishing control units (Coeffelt mark XXII and Smith-Root mark XXII) were us ed in this study. In an attempt to standardize among boat and control unit type, current output was adjusted to produce 5000W of power during all elec trofishing operations. Non-na tive fish were euthanized, speciated, and total length (TL), and weight (g) recorded. Native fish were measured (TL) and native fish larger than 150 mm TL were implanted with a passive integrated transponder (PIT) tag. To examine the effects of boat type, control unit type, location (either above or below the LCR confluence), and boat operator on capture pr obability, I varied the deployment of boat and control unit type to each river section in a syst ematic fashion during the first two years of the study. The overall strategy was to ensure that a rubberand an aluminum-hulled boat were 22

PAGE 23

always deployed on opposite shorelines (e.g., sections A and B) and their positions reversed on the subsequent trip. The two types of elect rofishing control units were deployed on opposite shorelines and reversed after each set of three trips. Four boa t operators were randomly assigned to a particular section and depletion pass within each trip. The same boat operators participated in each trip with the exception of boat opera tor 2 (absent during July 2003 and September 2004) and boat operator 3 (absent during January 2004). Experienced substitute boat operators (boat operators 5 and 6) were employed in these instan ces. During the final 11 trips in the second two years of the study, both electrofishing control unit type and boat operators were assigned to each reach haphazardly. Control Reach: Study Areas and Field Protocols To determine if changes in the fish community in the mechanical removal reach were related to environmental influences and not the mechanical removal, a control reach was established upstream of the removal reach in an area of high rainbow tr out density (44 RM-52.1 RM; Figure 2-2). This reach was stratified into 60, 500 m sites (30 on each shoreline). During most trips, 24 sites were randomly chosen a nd sampled using identical capture methods as outlined above in the mechanical removal reach. Exceptions occurred in January 2003 and August 2003 when 25 and 11 sites were sampled. All captured non-native fish were speciated, measured (TL), and fish 200 mm TL were implanted with a uniquely numbered external tag and their left pelvic fin remove d prior to release all non-native fish were released alive. Mechanical Removal Reach: Data Analysis Following Dorazio et al. (2005), I used a hierarchical Bayesi an modeling (HBM) framework to estimate abundance and capture proba bility from data collected among the serial removal passes. This framework assumes th at the overall populati on is a collection of subpopulations (defined below), each with di fferent abundance and e xperiencing different 23

PAGE 24

capture probability during removal efforts. Subpopulation abundance and capture probability are sampled from common population level distributions conditi onal on unknown hyperparameters (i.e., parameters that govern th e population level distributions). This hierarchical structure allows a model-based aggregation of data am ong subpopulations and can be thought of as an intermediary between analyses that operate on data pooled over all subpopulations, and those that operate on each subpopulation independently. The structure allows sharing of information among subpopulations, particular ly for subpopulations for which the data are relatively uninformative or imply extreme parameter values In these cases, th e subpopulation parameter values are more heavily influenced by the popul ation distribution and ar e thus pulled, or shrunk (Gelman et al. 2004), towards the population distribution means. The amount of shrinkage is a function of both the difference between subpopula tion and population distribution means and the population distribution variance. I defined closed subpopulations to correspond to fi sh within each mechanical removal site. I assumed that the observed num bers of removals from site i (1,, I ) among removal pass j (1,, J ) were drawn from a multinomial distribution w ith number of trials equal to the site abundance ( Ni) and cell probability vector Jiiii ,2,1,,...,,. If I first assume that capture probability is constant among re moval passes within each site i then the cell probability in site i in the j th depletion pass is given by 11j iiij, (2-1) where i is a constant capture probability in site i The likelihood for the overall model is given as J j x ji xNJ i iii i iiiij iiixNc N xNL1 ,.1 (2-2) 24

PAGE 25

where xij is the number of fish captured in site i and depletion pass j is the total number of fish captured in site i and iJ j ij ixx1 iJ j ij ixc1!. Equation (2-2) is the familiar Zippin (1956) estimator and as above assumes that capture probability i is constant within a site. To cast this model in a HBM framework, I assumed that capture probability within a set of sites is samp led from a common distribution. The set of sites could either be all sites within the removal reach, or a subset of sites belonging to a common stratum. Because there is good reason to beli eve that electrofishing capture probability is influenced by abiotic factors such as turbidity (Reynolds 1996) and because there is frequently higher turbidity below the LCR confluence (Yard 2003), I chose to stratif y the overall removal reach into sites upstream (sections A and B) and downstream (sections C, E, and F) of the LCR confluence and fit separate distributions to each strata. Similarly, fish abundance typically differs upstream versus downstream (Gloss a nd Coggins 2005) of the LCR so separate distributions of abundance were also used. Following Dorazio et al. (2005), I assumed that the site specific capture probabilities were sampled from beta distributions in each of the strata as kk kiBeta ,~,, where k is either 1 (upstream stratum) or 2 (downstream stratum) and k and k are the hyperparameters. The mean (k ) of the distribution is k /(k +k ) and the variance is k (1-k )/(k +1), where the similarity parameter (k ) is k +k I assumed that the site specific abundances were sampled from Poisson distributions with mean and variancek For convenience, I estimated and ( ln ) for each stratum. I chose diffuse prio r distributions for each hyperparameter as: k ~Uniform(0,1), k ~Uniform(0,100), E(k )~Normal(0,0.01), and SE(k )~Uniform(0,10). 25

PAGE 26

To examine the effect of the covariates men tioned above on capture probability, I also reanalyzed a subset of the data collected during 2003-2004 using a model that allowed capture probability to vary among sites and passes as a f unction of covariate values. To accomplish this, I assumed that capture probabi lity was a logit function: ji ji ji ji ji ji ji jixxxxxxxx,88,77,66,55,44,33,22,110 ji,exp1 1 (2-3) where1 is the location coefficient ( =1 for upstream sections A and B, =0 for downstream sections C, E, and F), 1x1x2 is the boat hull type coefficient ( =1 for rubber hull and =0 for aluminum hull), 2x2x3 is the electrofishing control unit coefficient ( =1 for the Smith Root Mark XXII and =0 for the Coeffelt Mark XXII), 3x3x4 is the boat operator 2 coefficient ( =1 for operator 2 and =0 for not operator 2), 4x4x5 is the boat operator 3 coefficient ( =1 for operator 3 and =0 for not operator 3), 5x5x6 is the boat operator 4 coefficient ( =1 for operator 4 and =0 for not operator 4), 6x6x7 is the boat operator 5 coefficient ( =1 for operator 5 and =0 for not operator 5), and 7x7x8 is the boat operator 6 coefficient ( =1 for operator 6 and =0 for not operator 6). Lastly, this coding scheme implies that 8x8x0 is the untransformed capture probability for boat operator 1 in downstream sites in an aluminum-hulled boat outfitted with the Coeffelt Mark XXII electrofishing control unit. I assumed that the si te and removal pass specific values of each of the coefficients (Z ) were sampled from normal dist ributions with hyperparameter mean (Z ) and standard deviation (Z ), where Z=0, 1, 8. I specified diffuse priors for each hyperparameters as: Z ~Normal(0,0.01), and Z ~Uniform(0,10). I implemented these analyses in programs R (R Development Core Team 2007) and Winbugs (Lunn et al. 2000). For each trip analyz ed, I characterized the distribution of each 26

PAGE 27

parameter among 20,000 Markov Chain Monte Carl o samples with a thinning frequency of 10 and discarding the first 10,000 burn in samples. I examined convergence using Gelman and Rubins potential scale reduction factor (R Development Core Team 2007). Control Reach: Data Analysis I assessed the abundance of rainbow trout wi thin the control reach using electrofishing catch rate and mark-recapture-based open populati on abundance estimates. Because all rainbow trout marked with external tags were also give n a secondary fin clip, I attempted to incorporate the rate of tag loss into the mark-recapture-base d estimates of survival, capture probability, and abundance. I estimated tag loss rate by compar ing the observed and predicted proportion of recaptured fish that retained tags each trip. This proportion is not influenced by survival or capture probability under the assumption that surviv al and capture probabil ity are independent of tag retention. To derive this estimator, I first assumed that tag loss rate during the first month after initial tagging could be different from the rate experience d in subsequent months. This allows for the possibility that tags may be lost at a higher rate initially (e.g., as a result of improper placement), but that the rate of tag loss declines after this in itial loss. I predict the number of tagged fish in month t as )1()1()1( 11 11 21lRlFlTSTt t t t ( where S is the mont 2-4) hly survival rate, is the number of tagged fish available for capture just ere tT prior to sampling in month t, 2l is the monthly secondary tag loss rate, 1tF is the number of newly tagged fish in month t-1, 1l is the monthly initial tag loss rate (sued in the month following tagging), 1tR is the number of fish that ha d lost their tag prior to month t-1 and w ffer 27

PAGE 28

retagged in month t-1. Conversely, I predict the number of fish that have lost their tag in month t as )()()( 11 11 2111lRlFlTRLSLt t ttt t (2-5) where is the number of fish that have lost their tag and are available for capture just prior to sampling in month t. The predicted tag retention rate (tL t ) of recaptured fish in the population in month t is then tt t tLT T (2-6) Note that equations (2-4) and (2-5) are linked by the R term such that when recaptured fish without a tag are observed, they are fitted with a new tag and thus decremented from and added toL T To estimate and I minimized the sum of squares between observed and predicted retention rate among the 22 sampling occas ions following the first one. It is worth noting that because the monthly survival rate ( ) appeared in each term of equation (2-6), there was no need to estimate it in order to estimate tag loss rates. 1l2l SI estimated monthly survival rate ( ) and capture probability ( ) conditional on tag loss rates generally following a single age recoveries only model (Brownie et al. 1985). However, for computational simplicity, I assumed that observed recaptures followed a Poisson rather than a multinomial distribution. Under th is structure, the complete capture history is not used and the predicted numbers of fi sh released in month e and recaptured with tags in a subsequent month t is tS tpt t ei i et eetepS llRFr 11 1 1 21 (2-7) 28

PAGE 29

To reduce the number of parameters to be esti mated, I set the monthly survival rate among months not sampled equal to the survival rate of the next sampled month. Assuming that the observed numbers of fish released in month e and recaptured with tags in a subsequent month t ) represent independent samples 7), the log-likelihood func tion ignoring terms involving only the data is (ter,from Poisson distributions with mean s given by equation (222 1 23 ,,, ee t tetetewhere ln ,ln rrr SprL (2-8) p and S are the unknown capture probability and m onthly survival rate vectors to be estimated. The model was implemented in a Mr osoftxcel spreadsheet using Solver (L and Allan 2002) as the non-linear search procedure. As a meas ure of uncertainty, I computed 95% likelihood profile c onfidence intervals on ic E adson p and S using Poptools (Hood 2000). I estimated the abundance of rainbow trout 200 mm TL by dividing the nu mbers of fish captured by the capture probability. Approximate 95% nce intervals on th ese abundance estimates dence bounds on the capture probability estimates. f c onfide were calculated using the confiResults Mechanical Removal Reach Over 36,500 fish from 15 species were captu red in the mechanical removal reach during 2003-2006 (Table 2-1). The majority of these fishes (23,266; 64%) were non-natives and were comprised primarily by rainbow tr out (19,020; 82%), fathead minnow Pimephales promelas (2,569; 11%), common carp Cyprinus carpio (802; 2%), and brown trout (479; 1%). Catches o native fish amounted to 13,268 (36%) and were comprised of flannelmouth sucker (7,347; 55%) humpback chub (2,606; 20%), bluehead sucker (2,243; 17%), and speckled dace (1,072; 8%). The contribution of rainbow trout to the overall species catch compositi on fell steadily through the course of the study from a high of approxima tely 90% in January 200 3 to less than 10% in 29

PAGE 30

August 2006 (Figure 2-3). Overall, non-native fish comprised more than 95% of the cat 2003, but following July 2005 generally contributed less than 50%. Owing to particularly larg catches of flannelmouth sucker and humpb ack chub in September 2005, the non-native contribution to the catch in that month ch in e was le ss than 20%. While the catch of non-native fish gener of s captu re probability in the upper stratum ranged from 4% to intervals since thes e estimates are the difference between two ally fell through the course of the study, catches of non-native c yprinids (dominated by fathead minnows) increased in 2006. Using the HBM, the estimated abundance of ra inbow trout in the entire removal reach ranged from a high of 6,446 (95% cr edible interval (CI) 5,819-7,392) in January 2003 to a low 617 (95% CI 371-1,034) in February 2006; a 90% re duction over this time period (Table 2-2). Between February 2006 and the final removal ef fort in August 2006, the estimated abundance increased by approximately 700 fish to 1,297 (95% CI 481-2,825). Though this increase wa more than double the February 2006 estimate, the August 2006 estimate was much less precise. The estimated abundance in the downstream stratum of the mechanical removal reach was typically approximately 30% of th at in the upper stratum (Figure 2-4) and the density was also typically lower (Table 2-2). The estimated 34% (Figure 2-4). The estimated capture probability was generally lower in the lower stratum and ranged between 2% and 19%. Net immigration rate estimates indicate that fi sh were moving into both strata within the removal reach at a higher rate during 2003-2004 than during 2005-2006 (Figure 2-5). Additionally, it appear s that net immigration may be lowest in the late fall through early winter, and highest between January and March. Duri ng 2005-2006, there were only two time that suggest net immigration rate was different than zero in th e downstream stratum and one in the upstream stratum. However, 30

PAGE 31

distrib s were ting est effect ent among strata. This resu lt supports the use of separate di stributions to describe capture e overall HBM to estimate abunda nce, capture probability, and net immigration rate. h nce utions, each with their own error, the ne t immigration estimates were imprecise for many of the time periods (Figure 2-5). The results of the covariate analysis indicated that most of the factors had little influence on capture probability (Figure 2-6). In fact, th ere was little indication that any factor had a strong directional effect in either raising or lowering capture probability. The exception that on four of the trips, the boat operator 2 effects were signif icantly less than zero indica that this operator had a negative effect on cap ture probability compared to operator 1. Additionally, operator 3 tended to have a sli ghtly positive effect on capture probability. Operators 5 and 6 participated in only 1 and 2 tr ips, respectively, but did not seem to affect capture probability relative to operator 1. Neith er boat type nor type of electrofishing control unit had a strong directional effect on capture probability. The fact or that had the larg on capture probability was location in the overall reach. Though there was no indication that capture probability was uniformly higher or lower in the upstream stratum versus the downstream stratum, several of the trips indicated that capture probability was significantly differ probabilities in thControl Reach A total of 11,221 fish representing 7 species were captured during control reach sampling (Table 2-3). The majority of fish captured we re rainbow trout (95%), followed by flannelmout sucker (3%), and brown trout (1 %). A general pattern of decr easing rainbow trout abunda was observed throughout the study, particularly following spring of 2005 (Figure 2-7). Initial (1l) and secondary (2l) monthly tag loss rate estimates were 11% and <1%, respectively, 31

PAGE 32

suggesting that most tag loss occu rred shortly after tagging. Ra inbow trout abundance within th control reach was estimated at between 5, 000 and 10,000 fish during 2003-2004 and between 2,000 and 5,000 during 2004-2005 (Table 2-4 and Figur e 2-7). This analysis coupled with catch rate assessment (Figure 27) suggests that rai nbow trout abundance likely declined by one half or more between the first and last two years of the study. Capture probability ranged between 3% and 13% with no strong tempo e the ral patt ern (Figure 2-7). Estimated monthly survival hing 1 (Figure 2-8). The lowest surviv oval h less ol displa 04. ance e control reach considering either the ca tch rate or abundance estimates rate ranged from a low of approximately 0.72 to a high approac al rates were observed during 2005. Comparison of Mechanical Removal and Control Reaches The abundance of rainbow trout declined thr ough the study both in th e mechanical rem reach and in the control reach; however, the pa ttern of decline was dissimilar among reaches (Figure 2-9). In the mechanical removal reach the largest decline ( 62%) occurred between January 2003 and September 2004. Rainbow trout abundance in this reac h declined muc rapidly from January 2005 to August 2006. In co ntrast, rainbow trout abundance in the contr reach was constant to slightly declini ng from March 2003 through September 2004, but yed a strong negative tre nd subsequently. These patterns suggest that removal efforts likely affected abundance in the mechanical re moval reach predominantly during 2003 and 20 Another difference between the mechanical removal and control reaches was the seasonal patterns in rainbow trout abundance. In the re moval reach, a pattern of declining abund during each three-month bout of removal effo rts (e.g., January-March) was followed by an increase in abundance at the be ginning of the next series of removal efforts (e.g., JulySeptember), particularly during 2003-2004 (Figure 2-4). This patte rn would be expected if th removal rate was greater than th e immigration rate only during each removal series. This pattern was not evident in the 32

PAGE 33

(Figu t h g e length in the control reach. Taken together, these observations suggest in the net immi gration pattern, possibly during 2005. oval that t re 2-7) suggesting that mech anical removal was influencing the abundance of rainbow trou in the removal reach. These seasonal patterns in ra inbow trout abundance are mirror ed by trends in the average length (Figure 2-10). Average length among the tw o reaches tended to converge in July of eac year followed by a period of divergence. One po ssible explanation of this pattern is selective removal of larger individuals followed by reinva sion, particularly following the winter/sprin removal efforts, of larger individuals from upstream sources. This pattern was not observed in July 2005 and was preceded by a significant declin e in averag a di sruption from upstream sources, into the removal reachDiscussion Mechanical Removal: Effective Program? Results suggest that the mechanical rem oval program was successful in reducing the abundance of non-native fishes, primarily rainbo w trout, in a large segment of the Colorado River in Grand Canyon. However, maintenance of low rainbow trout abundance in the rem reach was also facilitated by reduced immigra tion rates during 2005-2006 and a systemic decline in abundance. Common features of this study and other successful non-native mechanical removal efforts are significant a nd sustained removal effort. Bigelow (2003) demonstrated population level changes were not evident in re moval efforts aimed at non-native lake trouSalvelinus namaycush in Yellowstone Lake until the latte r years of a four year study when additional support for the project (e.g., funding and equipment) allowed increases in total removal effort and efficiency. Similarly, removal objectives for non-native brook trout Salvelinus fontinalis, golden trout Oncorhynchus mykiss aguabonita, and rainbow trout from small high altitude lakes in the Sierra Nevada were achieved with year-round gillnet fishing 33

PAGE 34

(Knapp et al. 2007). In combination with increased predation from stocked pr edators, Hein et al (2006; 2007) demonstrated effective control of non-native crayfish Orconectes rusticus using mechanical removal, but only with sustained and significant removal effort. The necessity for sustained maintenance control of non-natives is not uncommon (Pine et al. 2007) as many nonnative the removal reach during 2003-2004, much smaller and possibly undetectable once (-15 d out is species demonstrate high resilience, and ar e well adapted to their introduced environment as evidenced by their invasi on success and warranted need for management action. In contrast, Meyer et al. (2006) document a recent unsuccessful effort to remove brook trout from a 7.8 km reach of a second-order st ream in Idaho using electrofishing. Crews removed fish over 3 days in August during each year for four years. While the capture technique was likely appropriate, the overall effort was appare ntly insufficient to significantly reduce population. In the present study and considering th e net immigration rates of rainbow trout into the mechanical reductions in overall abundance would have been r ealized had removal efforts been applied only per year. Serendipitous Timing: What Led to the Declin e of Rainbow Trout in the Control Reach? The decline of rainbow tr out abundance observed in the control reach was likely precipitated by at least tw o factors. First, rainbow trout a bundance in the Lees Ferry reach RM at GCD to RM 0) of the Colorado River increased during approximately 1992-2001 an abundance in this reach steadily fell during 2002 -2006 (Makinster et al. 2007). With the exception of limited spawning activity in select tributaries of the Colorado River in Grand Canyon, rainbow trout reproductive activity appears to be limited ma inly to the Lees Ferry reach (Korman et al. 2005). Examination of length frequency distributions of rainbow trout captured using electrofishing from Glen Canyon Dam to RM 56 during 1991 through 2004 also supports the idea that Lees Ferry is the primary spawning s ite, as the juvenile size cl ass of rainbow tr 34

PAGE 35

largely absent from collections downstream of RM 10 (Figure 2-11). Thus, it is reasonable to conclude that at least for the last 10-15 years, the natal source of most rainbow trout in th system is the Lees Ferry reach. This i is s signif icant because it suggests that abundance of rainbow trout in of er also ed Trout g 1992). e s. in Grand Canyon is partially influenced by trends in rainbow trout abundance and reproduction in the Lees Ferry reach. Second, it has been widely demonstrated that th e density of rainbow trout is not uniform the Colorado River below GCD and distribution patterns are likely influenced by food resources and foraging efficiency (Gloss and Coggins 2005). Rainbow trout density generally declines with downstream distance from GCD but exhibits punctuated declines below the confluences the Paria River and the LCR. The density of alg ae and invertebrates in th e Colorado Riv decline along this gradient (K ennedy and Gloss 2005) suggesting a possible linkage between distance from the dam and primary production. A major factor likel y influencing these distributional patterns is sediment delivery from tributaries an d the subsequent effects of elevat turbidity in the Colorado River in downstream sections. Yard (2003) demonstrated that these tributary inputs of sediment c ontribute to high turbidity and li mit aquatic primary production. are predominantly sight feeders thus, high turbidity is likely to adversely affect foragin efficiency by decreasing encount er rate and reactive distance to prey items (Barrett et al. From September 2004 through January 2005, the discharge and sediment load from th Paria increased to the point that a threshold outlined in the Glen Canyon Dam Adaptive Management Program related to rebuilding de pleted Colorado Rive r sandbars was reached triggering an experimental high fl ow from GCD in November 2004. It is possible that the high flow event and the associated period of elevated turbidity may have influenced rainbow trout density downstream of the Paria River confluence possibly through elevated mortality rate 35

PAGE 36

Estimated survival rates in the control reach ge nerally support the notion that rainbow trout may have g g e is h and the nfluenced the survival and activity of these h and d experienced diminished survival rate s during late 2004 and ear ly 2005 (Figure 2-8). Other Species Beginning in September 2005, large increases in the catch of non-native fathead minnow and black bullhead Ameiurus melas were observed compared to th e previous 17 trips, suggestin either increased immigration and/or survival of these fishes in the mechanical removal reach. Since these fish are not captured with any regul arity in the control reach nor in other samplin upstream of RM 44 (USGS, unpublished data), it is reasonable to conclude that their sourc not upstream. Stone et al. (2007) documented the presence of these species and other warm water non-natives in the LCR 132 km upstream from the conf luence and suggested this tributary as the likely source of fathead minnow black bullhead, and 6 other non-native fis frequently encountered in the lower LCR and the mechanical removal reach. Thus, one possibility for the elevated catch of fathead minnow and black bullhead in the mechanical removal reach during this latter timeframe is an el evated emigration rate of these fishes from the LCR. Alternatively, increasing water temperatur e, particularly in 2005 (Figure 2-12), concurrent reductions in rainbow trout biomass, ma y have i fishes causing them to be both more abundant and more susceptible to capture. Bias in Capture Probability and Abundance Estimates Capture probability estimates from the upper st ratum of the mechanical removal reac the control reach are surprisingly different. Neit her of these reaches is differentially influence by large tributary inputs and th ey share similar overall ch annel morphology yet capture probability estimates are generally nearly twice as high in the removal reach as in the control reach. Several authors have demonstrated that capture probability is typically over-estimated using electrofishing depletionbased methods (Peterson et al 2004; Rosenberger and Dunham 36

PAGE 37

2005). The typical mechanism for this finding is heterogeneity in cap ture probability among individuals in the population. As a result, the fish remaining after each successive pass have overall lower capture probability than those in preceding passes. This bias in capture prob ability then l ss e y the dance and density estimates are lik moval al. eads to negative bias in abundance. Because of this bias, it is likely that the depletionbased capture probabilities estimated in this study are higher than were actually realized. In principle, the additional information avai lable from mark-recapture should provide le biased estimates of capture probability in the co ntrol reach. However, this may not be true due to possible inadequate mixing of fish between each mark-recapture sampling event. Becaus only 40% of the available sites were sampled dur ing each mark-recapture event, if fish did not mix completely between passes it is possible th at capture probability may have been underestimated since not all fish had an equal probability of capture within each event. Theoreticall this assumption is met by the random selection of sampled sites each trip. In practice, however, only 40% of fish (assuming uniform distribution) actually had an opportunity to be captured. Additionally, it is possible that there was sampling induced heterogeneity in capture probability. This is a common feature of ma rk-recapture experiments for small mammals (Otis et al. 1978) and a recent paper by Askey et al (2006) suggests that fish may develop anti-capture behavior as well. Therefore, the realized capture probabilities were likely between those estimated in control reach and those estimated in the removal reach. If true, abun ely over-estimated in the control re ach and under-estimated in the removal reach. Recommendations for Future Mechanical Removal Operations I recommend that further effort be spent better documenting the preferred habitat of target non-native species. This information could then be used to more effectively distribute re effort among habitat types that contain the highest density of non-native sp ecies. Bigelow et (2003) describe the use of hydroacoustic surveys to better target areas of high lake trout 37

PAGE 38

abundance, increasing the efficiency of the control program. A possibl e technique to better determine these high density areas in the mechan ical removal reach would be to employ a fine scale shoreline habitat-based deline ation of removal sites, rather than the coarse 500 m sites use in the present study. Serial depletion data coul d then be analyzed with the HBM to include r d a habita s in in ility as these probability. However, futu re fish control programs, these effort of an t covariate for density. This approach has been successfully used to describe pattern the density of organisms as a function of habitat characteristics (Royle and Dorazio 2006) Results from the present covariate analysis indicate that most variability in capture probability is related to site lo cation rather than methodological issues. In the context of designing future removal efforts and the larger monitoring program for non-native salmonids Grand Canyon, this is a fortunate result as it suggests that current leve ls of standardization among equipment will have a reasonably high lik elihood of producing index data useful for determining trends in salmonid abundance and dist ribution. However, the observed variab among boat operators implies that additional trai ning to reinforce consistent methodologies may be useful to further minimize that source of variability. A more significant finding is the heterogeneity in capture probability among upstream and downstream strata. If these differences are primarily related to uncontrollabl e factors such as turbidity or shoreline and substrate type suggested by Speas et al. (2004), additional resear ch should be conducted to better describe relationships. Unfortunately, I was not able to obtain robust measuremen ts of turbidity with sufficient regularity to investig ate the effect of turbidity on capture assuming serial depletion methods are likely to be used in s may provide the ideal setti ng to explore these associations. Adaptive Management in Grand Canyon: The Future As stated at the beginning of this chapter, this study documents the implementation ecosystem-scale adaptive management action aime d at testing the efficacy of a particular 38

PAGE 39

management policy (i.e., non-native control) in order to improve the status of native fish resources in Grand Canyon. Though this study focuses on the efficacy of implementing the policy, the more interesting, important, and diffi cult questions are related to evaluating whether the policy will have the intended effect. I predic t that if non-native salmonids are a signific and uncompensated mortality source for native fish attempting to rear in the mainstem Colorado River, then the survival rate and abundance of juv ant enile native fish in the mainstem should increa re is liable struct ss se during 2003-2006. I would further pred ict that humpback chub recruitment associated with the 2003-2006 brood y ears should increase. There are some indications that the abundance of native fish has increased in the removal reach during 2003-2006 (Figure 2-3; Chapter 5) suggesting either increased survival rate, increased production of juvenile fish, or both. Ho wever, these initial signa ls are not adequate to infer the success of the policy for two important reasons. First, the unplanned increases in release water temperature are nearly perfectly temporally correlated with the magnitude of the non-native fish reduction (Figures 2-4 and 2-12). As water temperature is also a controlling factor affecting quality of rear ing habitat in the mainstem river (Gloss and Coggins 2005), this confounding makes separation of these two effects impossible at this time. Second, since the not a monitoring program for estimating temporal trends in survival rate, likely the most re available measure of a native fish response is humpback chub recruitment. Because the ageured mark-recapture model (Chapter 4) is not able to provide estimates of year cla strength until fish reach age-4, the best data to infer change in survival are not yet available. One strategy to separate out the effects of non-native fish versus increased water temperature would be to implement a future expe riment with one of these factors changed from the 2003-2006 condition. However, since temperatur e control is not availa ble at this time and 39

PAGE 40

40 rom ementa tion and ultimate success of even the best designed experiments will be dependent n unmanageable factors such as climate (Seager et al. 2007) and unexp ected biotic interactions. since rainbow trout abundance appears diminished system-wide, there may be limited near-term opportunities to manipulate the syst em further. Determining what future experiments to conduct will be determined by the Glen Canyon Dam Adap tive Management Program using input f studies such as this but it should be noted that the impl o

PAGE 41

Table 2-1. Electrofishing catch by species in the mechanical rem oval reach for each month, 2003-2006. Speciesa Trip Date # Removal Passes BBH BHS BNT CCF CRP FHM FMS GSF HBC PKF RBT RSH SMB SPD STB SUC Jan-03 5 8 87 80 17 188 26 1 3,605 7 2 Feb-03 5 18 24 33 21 165 26 1,913 1 2 Mar-03 5 3 11 21 1 22 8 89 13 1 1,195 1 8 Jul-03 5 4 12 63 29 4 267 124 2,278 1 6 3 Aug-03 2 2 4 12 14 79 17 779 5 Sep-03 3 1 19 11 31 4 119 37 2 818 1 18 4 Jan-04 4 3 32 88 23 18 169 51 1,330 53 3 Feb-04 4 9 37 29 1 9 13 110 52 622 34 Mar-04 5 5 24 22 18 44 218 61 6 867 92 3 Jul-04 6 9 84 29 1 26 32 296 142 9 1,464 3 47 Aug-04 4 6 33 7 16 6 190 27 3 480 2 7 Sep-04 4 11 72 17 29 13 258 43 687 5 19 Jan-05 4 8 54 14 27 72 244 61 1 623 9 52 Feb-05 4 3 38 4 1 14 39 191 49 2 283 2 39 Mar-05 4 8 51 4 14 73 176 82 3 318 4 51 Jul-05 4 17 159 9 2 45 9 480 1 220 432 2 38 2 1 Aug-05 4 9 124 4 36 17 419 86 1 295 4 24 24 Sep-05 4 14 576 7 47 190 1,140 600 230 15 187 4 Jan-06 4 23 197 9 38 685 545 249 357 13 115 1 Feb-06 4 15 98 5 10 300 529 171 1 103 70 Mar-06 4 12 96 2 8 322 365 1 196 1 66 2 84 Jul-06 4 15 331 8 64 192 554 145 2 159 2 56 Aug-06 4 13 165 3 1 169 490 556 128 34 116 1 1 63 9 Total 190 2,243 479 7 802 2,569 7,347 2 2,606 67 19,020 68 1 1,072 2 59 41 a BBH=black bullhead ( Ameiurus melas ), BHS=bluehead sucker ( Catostomus discobolus ), BNT=brown trout ( Salmo trutta ), CCF=channel catfish (Ictalurus punctatus ), CRP=common carp ( Cyprinus carpio), FHM=fathead minnow ( Pimephales promelas ), FMS=flannelmouth sucker (Catostomus latipinnis ), GSF=green sunfish ( Lepomis cyanellus ), HBC=humpback chub ( Gila cypha ), PKF=plains killifish ( Fundulus zebrinus ), RBT=rainbow trout ( Oncorhynchus mykiss ), RSH=red shiner ( Cyprinella lutrensis ), SMB=smallmouth bass ( Micropterus dolomieu ), SPD=speckled dace ( Rhinichthys osculus ), STB= striped bass ( Morone saxatilis), SUC=unidentified sucker.

PAGE 42

42Table 2-2. Estimated abundance and density of rainbow trout in the mechanical removal reach at the beginning of each month, 20 032006. Uncertainty estimates (95% CI) ar e 95% Bayesian credible intervals. Total Reach Abundance Upper Stratum Abundan ce Lower Stratum Abundance Density (Fish/Km) Trip Date N 95% CI N 95% CI N 95% CI Upper Stratum Lower Stratum Jan-03 6,446 5,819-7,392 4,977 4,519-5,640 1,469 1,168-1,996 1,512 563 Feb-03 3,073 2,802-3,492 2,437 2,226-2,778 637 489-879 740 244 Mar-03 2,372 1,939-3,014 2,023 1,606-2,671 349 289-485 615 134 Jul-03 5,253 4,249-7,616 3,614 3,164-4,183 1,639 902-3,776 1,098 629 Aug-03 1,574 1,253-2,199 1,237 1,001-1,652 336 178-845 376 129 Sep-03 3,008 1,964-4,197 2,399 1,438-3,507 609 345-1,187 729 233 Jan-04 2,207 1,953-2,635 1,684 1,472-2,002 523 385-851 512 201 Feb-04 1,611 1,098-2,809 845 732-1,026 767 293-2,009 257 294 Mar-04 1,425 1,227-1,710 1,075 925-1,325 350 269-516 327 134 Jul-04 3,445 2,533-5,284 1,718 1,566-1,925 1,727 856-3,627 522 662 Aug-04 932 734-1,536 677 515-1,266 255 183-455 206 98 Sep-04 2,459 1,647-3,752 1,980 1,296-3,290 479 199-1,060 602 184 Jan-05 989 819-1,275 722 675-786 266 115-539 219 102 Feb-05 869 519-1,785 386 317-516 483 142-1,388 117 185 Mar-05 975 636-1,548 782 498-1,377 193 80-427 238 74 Jul-05 1,626 742-5,837 736 560-1,085 891 128-5,056 224 341 Aug-05 690 498-1,080 415 339-549 275 115-638 126 105 Sep-05 697 460-1,291 411 288-601 286 108-893 125 110 Jan-06 710 514-1,121 502 386-719 208 100-580 153 80 Feb-06 617 371-1,034 479 258-879 138 61-290 145 53 Mar-06 669 280-1,460 367 154-860 302 69-992 111 116 Jul-06 726 376-2,210 538 251-1,853 188 89-410 163 72 Aug-06 1,297 481-2,825 767 262-2,087 530 136-2,090 233 203

PAGE 43

Table 2-3. Electrofishing catch by species in the control reach for each month, 2003-2006. Speciesa Trip Date Control Sites BBH BHS BNT CRP FMS HBC RBT SPD SUC Jan-03 25 10 1 1 444 Feb-03 24 8 1 548 Mar-03 24 5 1 888 Jul-03 24 8 1 2 416 Aug-03 11 4 1 256 Sep-03 24 7 2 7 1,036 1 Jan-04 24 5 702 Feb-04 24 3 1 1 434 Mar-04 24 2 14 3 851 Jul-04 24 2 9 1 491 Aug-04 24 9 6 346 Sep-04 24 8 1 4 498 2 Jan-05 24 1 1 503 Feb-05 24 9 4 476 1 Mar-05 24 1 9 5 540 Jul-05 24 1 11 34 277 Aug-05 24 5 21 332 Sep-05 24 1 2 1 72 284 1 Jan-06 24 2 2 1 31 277 1 Feb-06 24 4 2 53 243 Mar-06 24 5 23 336 Jul-06 24 5 2 5 47 1 176 12 Aug-06 24 1 10 1 1 52 1 294 Total 1 22 134 17 378 3 10,648 15 3 a BBH=black bullhead (Ameiurus melas), BHS=bluehead sucker (Catostomus discobolus), BNT=brown trout (Salmo trutta), CRP=common carp (Cyprinus carpio), FMS=flannelmouth sucker (Catostomus latipinnis), HBC=humpback chub (Gila cypha), RBT=rainbow trout (Oncorhynchus mykiss), SPD=speckled dace (Rhinichthys osculus), SUC=unidentified sucker. 43

PAGE 44

Table 2-4. Estimated abundance and density of rainbow trout in the control reach at the beginning of each month, 2003-2006. Uncertain ty estimates (95% CI) are 95% profile likelihood confidence intervals. Total Abundance Trip Date N 95% CI Density (Fish/Km) Feb-03 5,058 3,500-7,262 1,018 Mar-03 10,571 8,064-14,136 2,128 Jul-03 10,106 6,572-16,367 2,034 Aug-03 8,819 5,494-13,593 1,775 Sep-03 8,051 6,004-10,860 1,621 Jan-04 9,952 6,491-15,662 2,003 Feb-04 8,998 5,570-15,024 1,811 Mar-04 7,939 5,379-11,798 1,598 Jul-04 8,758 5,895-13,254 1,763 Aug-04 6,981 4,519-11,171 1,405 Sep-04 7,208 4,733-10,795 1,451 Jan-05 4,138 2,853-6,090 833 Feb-05 4,527 3,344-6,202 911 Mar-05 5,253 3,939-6,907 1057 Jul-05 3,163 1,967-5,245 637 Aug-05 3,247 2,126-4,900 654 Sep-05 2,955 1,877-4,604 595 Jan-06 4,032 2,502-6,694 812 Feb-06 2,992 1,957-4,804 602 Mar-06 2,518 1,594-3,443 507 Jul-06 2,131 1,113-4,062 429 44

PAGE 45

Figure 2-1. Map of the mechanical removal reach of the Colorado River within Grand Canyon, Arizona. Depicted on the map are the reach sections (A-F) and th e sites within each reach (e.g., 1-19). The number of river miles downstream from Lees Ferry, Arizona is indicated at demarcation lines. 45

PAGE 46

Figure 2-2. Map of the contro l reach of the Colorado River within Grand Canyon, Arizona. Depicted on the map are the sites within the reach. The num ber of river miles downstream from Lees Ferry, Arizona is indicated at demarcation lines. 46

PAGE 47

A B Figure 2-3. Percent composition (A) and numbe r of fish (B) by species captured with electrofishing in the mechanical removal reach among months, 2003-2006. 47

PAGE 48

A B C Figure 2-4. Estimated catch rate (A), abundan ce (B), and capture prob ability (C) for rainbow trout in both the upstream and downstream st rata in the mechanical removal reach among months, 2003-2006. Error bars repres ent 95% Bayesian credible intervals. 48

PAGE 49

A B Figure 2-5. Net immigration rate into the upstream (A) and downstream (B) strata of the mechanical removal reach within time intervals between January 2003 and August 2006. Error bars represent 95% Bayesian credible intervals. 49

PAGE 50

Figure 2-6. Probability plots for coefficient values influe ncing capture probability among 10 trips (January 2003-July 2003; January 2004-September 2004). Within each plot, each line is the estimated probability de nsity for the coefficient in a trip. 50

PAGE 51

A B C Figure 2-7. Estimated catch rate (A), abundan ce (B), and capture prob ability (C) for rainbow trout in the control reach among months 2003-2006. Error bars represent 95% profile likelihood confidence intervals. 51

PAGE 52

Figure 2-8. Estimated monthly survival rate of rainbow trout in th e control reach during 20032006. Error bars represent 95% prof ile likelihood confidence intervals. 52

PAGE 53

Figure 2-9. Estimated rainbow trout abundance in both the m echanical removal and control reaches at the beginning of each trip duri ng 2003-2006. The solid lines represent the locally weighted polynomial regressions (Low ess) fit to each time series. The dashed lines represent linear regre ssions fit to either the 2003-2004 or 2005-2006 portions of the time series. 53

PAGE 54

Figure 2-10. Estimated total length and 95% conf idence intervals of rainbow trout captured in both the mechanical removal and control reaches during 2003-2006. 54

PAGE 55

Figure 2-11. Length frequency dist ributions of rainbow trout captured using electrofishing in the Colorado River from river mile -15 to river mile 56. Each panel represents captures of fish within the identified river segment. 55

PAGE 56

56 Figure 2-12. Daily mean water temperatures ob served in the Colorado River at approximately river mile 61, 1990-2006. Lines indicate lo cally weighted polynomial regressions (Lowess) fits to the indicated data set.

PAGE 57

CHAPTER 3 DEVELOPMENT OF A TEMPERATURE-DEPENDENT GROWTH MODEL FOR THE ENDANGERED HUMPBACK CHUB USING MARK-RECAPTURE DATA A primary interest of fisheries biologists is to understand how fishes grow and the processes and factors that influe nce that growth. Such informa tion is critical for research addressing questions about basic ecological relati onships such as the tradeoff between growth and survival, and for management strategies associat ed with maximizing yield. In the latter case, growth information is frequently used to popul ate assessment models w ith vital rates (Pauly 1980; Beverton 1992; Jensen 1996) and age-specific length, weight, fecundity, and vulnerability to exploitation (Walters and Martell 2004). Additionally, informa tion on growth may be used to estimate the age of fish based on size (e.g., Ki mura and Chikuni 1987). Given the importance of understanding growth, much effort has been ex pended to understand factors that influence growth, to develop models to describe observed growth patterns, and to estimate the parameters of those models (Ricker 1975; DeVries and Frie 1996). Though many models of fish growth have been proposed (Ricker 1975; Schnute 1981), perhaps the most widely used is the von Bertalanffy model (Ber talanffy 1938). The parameters of this model are typically estimated by compari ng predicted with observed size-at-age or growth increment data (Fabens 1965; Quinn and Deriso 1999). Obtaining growth increment data is usually accomplished via mark-recapture studies where the sizes of individual fish are measured before and after known times at liberty. A significant advantage of using growth increment data to estimate growth model parameters is that ages of individual fish are not required growth is simply measured over the times that individual fish have been at liberty. Since determination of age frequently involves inspection of various calc areous structures (e.g., otoliths) that often involves sacrificing the animal, use of increm ent data is preferable when working with endangered or rare species. 57

PAGE 58

The federally endangered cyprinid humpback chub Gila cypha is endemic to the Colorado River drainage in the southwestern United St ates and is generally found in swift, canyon bound river reaches (Minckley 1973). The Little Colorado River (LCR) population of humpback chub within Grand Canyon is a focal resource of the Glen Canyon Dam Adaptive Management Program (Gloss and Coggins 2005). Periodic stock assessments of this population serve as the core monitoring tool and status metric for this resource. These assessments require accurate age assignments of fish captured in a long-term samp ling program (Coggins et al. 2006a) in order to employ open population mark-recapture assessment methods that include age-dependent effects. Due to endangered listing status, longevity, a nd difficulty determining the age of individual humpback chub, little information is available on the relationship between size and age for this species. At present, individual age assignments are based on size and rely on a growth curve estimated from a limited set ( 60) age-length observations (USF WS 2002). This lack of growth information promotes uncertainty and possibly bi as in length-based age assignment, and this potential bias has been identified as an area of concern by past external reviews of the humpback chub assessment program used by the Glen Canyon Dam Adaptive Management Program (Kitchell et al. 2003). I used growth increment data to estimate the parameters of a genera lized growth model for the LCR population of humpback chub. This effort is undertaken to supplement the available information on humpback chub growth and to inform length-based age assignments for stock assessments. Because the older fish in this population exhibit a potadromous migration between the seasonally warm LCR and the constant cold mainstem Colorado River within Grand Canyon (Gorman and Stone 1999), I evalua ted ontogenetic temperature-depe ndent effects in the growth model. The results of this work should be useful to researchers studying humpback chub or 58

PAGE 59

wishing to estimate temperature-dependent growth models using growth increment data for other species that make major ontogenetic shifts in thermal habitat use. Methods An extensive monitoring program for the LCR population of humpback chub has been ongoing since the late 1980s (Coggi ns et al. 2006a). As a re sult of routinely capturing and implanting humpback chub with passive integrated transponder (PIT) tags, I was able to compile over 19,000 growth increments with which to evalua te growth rate (or m easurement error in the case of recaptures made shortly after tagging). The basic technique for estimating growth model parameters from growth increment data is to predict the amount of growth in the elapsed time between capture and recapture. Assuming standard von Bertalanffy growth curve predictions of length at time t and at time t+t, Fabens (1965) developed the most basic model where the predicted growth increment is given as tketLLtLttLL 1 )() ( (3-1) where t is time at initial capture, is the elapsed time between initial capture and recapture, and and k are the asymptotic length and the rate at which length approaches respectively (Quinn and Deriso 1999). Parameter estimate s are found by minimizing the difference between predicted and observed growth increments. t LLThough this technique has been widely app lied, numerous authors have pointed out resulting parameter estimates will be biased if individual fish exhibit growth variability (e.g., Sainsbury 1980; Kirkwood and Somers 1984; Francis 1988). Using this technique, k will typically be negatively biased and will be positively biased. Recognition of these problems has led to the development of alternative models which attempt to minimize these biases (e.g., James 1991; Wang et al. 1995; Laslett et al. 2002) I attempted to estimate standard von L 59

PAGE 60

Bertalanffy growth parameters for humpback chub using two of these methods (Wang et al. 1995; Laslett et al. 2002) and gene rally obtained poor results, charac terized by an inability of the models to predict growth increments exhibited by small fish and large fish simultaneously. Examination of growth rate as a function of si ze reveals that the basic problem with fitting a standard von Bertalannfy model to these data is the lack of a simple linear relationship between growth rate and length (Fi gure 3-1) as is implied by this model. It is apparent that the fish less than approximately 250 mm TL have a larger von Bertalanffy k parameter value (i.e., more negative slope of the growth rate vs. length plot ) than do fish larger than 250 mm TL. These results suggest a kink in the growth curve as would be found if fish grew along one curve when small and then switched to another when larger. Because water temperature is a major determin ant of basal metabolic rate and hence the von Bertalanffy k parameter among poikilotherms (Palohe imo and Dickie 1966; Essington et al. 2001), the kink hypothesis is consistent with fi sh that are demonstra ting an ontogenetic shift among habitats that have different water temper atures. For humpback chub, this would be a transition from the warm LCR spawning and rear ing habitat to the cooler mainstem Colorado River adult habitat (Valdez and Ryel 1995; Go rman and Stone 1999). To account for this apparent pattern of changing grow th rate, I fit growth increment data to a general growth model (Paloheimo and Dickie 1965) describing the rate of change in weight as n dmWHW dt dW (3-2) Here, the first term describes anabolism (i.e ., mass acquisition) and is governed by a term representing the mass normalized rate at which the animal acquires mass (H), the mass of the animal (W), and a parameter (d) describing the scaling of anabolism with mass. The second term represents catabolism (i.e., mass loss th rough basal metabolism or activity) where m is the mass 60

PAGE 61

normalized rate at which the animal looses mass and n is the scaling factor of catabolism with mass. Assuming a constant relationship between length and weight over time as baLW, (3-3) where L is length and a and b are constant, it is possible to de rive an analogous relationship for the rate of change in length as LL dt dL (3-4) Constants in this relationship are rela ted to those in (3-2) and (3-3) as b Had1, (3-5) b man1, (3-6) 1 bbd and (3-7) 1bbn (3-8) Essington et al. (2001) review thes e relationships and describe th e derivation of the standard von Bertalanffy growth function as the integral of equation (3-4) when n =1, b =3, and d =2/3. This is the situation where catabolism (plus mass loss to reproductive pr oducts) scales linearly with mass, the length-weight relationship is isometric, and anabolism scales as the 2/3 power of mass resulting in the standard von Bertalanffy growth model: )1()()(0ttkeLtL (3-9) where t0 is the theoretical age where body length is equal to zero. To estimate growth model parameters, I firs t assumed that measurement errors in the length of fish are normal with variance and that all fish follow a standard von Bertalanffy 2 m 61

PAGE 62

growth curve (equation 3-9) with shared k and individual The predicted length of fish at time of recapture can be found by rearranging the Fabens (equation 3-1) as L tketLLtLttL 1 (3-10) Assuming that individual is normally distributed with variance the variance of each is L2LttL2 2 2 2 21 1i itk L tk m i ttLe e (3-11) Deviations between observed and predicted growth increment for individual fish i are given as itk i ii ietLLtLttLD 1 (3-12) It is then possible to estimate the parameter vector ={ k, } by maximizing the loglikelihood function: L2L2 m s i ttL s i ittL iDs ttLtLL2 22 1 2 ,ln i, (3-13) where s is the number of growth increments. This is essentially an inve rse variance weighting strategy where growth intervals that have high recapture length variance are down-weighted in the fitting procedure. Though this procedure is applicab le assuming fish growth is de scribed by equation (3-9), if fish growth is described by equation (3-4) then there is no analytical solution for as in equation (3-11). However, by estimating and in particular k, using equations (3-11 through 313) and assuming that the individual variances co mputed using equation (3-11) are an adequate approximation, deviations from the general mo del (equation 3-4) can be used in the loglikelihood. These deviations are computed as 2ttL 62

PAGE 63

dtLLttLDii ittt tt i i (3-14) After specifying the parameters a and b for equation (3-3), estimation proceeds as above with the parameter vector ={ H d m n ,2 L, 2 m }. I implemented this procedure in both Micr osoft Excel using Solver (Ladson and Allan 2002) and AD Model Builder (Fournier 2000) to obtain estimates of I reduced the parameter set by specifying 2 m= 31.8 mm2 based on an analysis of the observed error between consecutive measurements of identical fish with in 10 days. I also specified the a and b parameters for equation (3-3) as 0.01 and 3, respectively. To calculate the conditional variance of each I specified k = 0.145 based on previous analyses Additionally, I included penalty terms in the log-likelihood e quation (3-13) to constrain d and n so that they did not deviate too far from the theoretical values assuming sta ndard von Bertalanffy growth of 2/3 and 1, respectively. I evaluated alternative weight valu es on these penalty terms to find an appropriate tradeoff between minimum weight s and decreased log-likelihood. ttLBecause all the information contained in the ma rk-recapture data are for fish larger than 150 mm TL, extrapolating results to the growth rate of smaller fish could be problematic. Fortunately, Robinson and Childs (2001) conducte d monthly sampling of juvenile humpback chub in the LCR during 1991-1994. They used th ese data to estimate (by modal progression analysis) average monthly length from age-0 months to age-32 mont hs. I utilized these data in an additional log-likelihood term to constrain the predicted lengths from the general model to be similar to those reported by Robinson and Childs (2001). Using these auxiliary data and assuming normal deviations allowed me to incor porate information on the growth rate of fish 63

PAGE 64

before they are large enough to be implanted with PIT tags. With these constraints in place, the full log-likelihood is s i ittL s i ittL i i iD s ttLtLL2 22 1 2 ,ln mos iiiL mos n d1 2 2 2ln 2 1 2 1 3 2 2 1 (3-15) where is the weighting value for the penalty terms, iL is the predicted length in month i from the general model, and is the predicted length over mos =32 months as reported by Robinson and Childs (2001). I specified th at the variance of the observed lengths was unity. The weighting term can be interpreted as the prior variance on the standard von Bertalanffy parameters ( d = 2/3, and n =1). i2 An important logical extension of the genera l model is to assume temperature dependence in growth rate. Accounting for ch anges in growth rate as a function of temperature is likely to be important for the analysis of this dataset for two reasons. The first is to account for the differences in growth rate with occupancy in either the LCR or the mainstem Colorado River. The second is to account for seasonal changes in water temperature within the LCR. The importance of the second consideration is furt her magnified by the temporal distribution of sampling within the LCR. Sampling in the LC R typically occurs in the spring and fall. Therefore, much of the observed growth increm ent data corresponds to either summer growth (i.e., observations of fish captured in spring and ag ain in fall) or winter growth (i.e., observations of fish captured in fall and the following spring). Because growth rate varies with temperature (Paloheimo and Dickie 1966), I e xpect growth increments to be smaller during winter than during summer. This general prediction is also consistent with both field (Robinson and Childs 2001) and laboratory (Clarkson and Childs 2000) observations of humpback chub. 64

PAGE 65

To allow temperature dependence in equati on (3-4), I defined temperature-dependent multipliers of and as 10)10( Tc cQTf, (3-16) 10)10( Tm mQTf, (3-17) where is the temperature-dependent multiplier of and Tfc Tfm is the temperaturedependent multiplier of The consumption and metabolism coefficients ( and ) of a Q10 relationship allow these multipliers to incr ease or decrease with temperature ( T ). One can think of these constants as the amount that anabolism or catabolism will change with an increase in temperature from 10 C to 20 C. Inclusion of these temperature-dependent multipliers into equation (3-4) yields cQmQ TfLTfL dt dLm c (3-18) Equation (3-18) accounts for grow th rate differences as a function of temperature, but does not account for movement between th e two thermal habitats. I used a logistic function to model occupancy in either the LCR or the mainstem Colo rado River. I assumed that the probability of LCR occupancy is given as 201 8.0 1tLLe PLCR (3-19) where L is fish total length and Lt is the fish total length where the probability of residing in the LCR year round is 0.6. The behavior of this model is such that the probability of year-round LCR residency approaches unity at lengths much less than Lt and decreases to 0.2 at lengths much larger than Lt. The number 20 in the denominator of the exponent governs the rate at 65

PAGE 66

which the probability changes from near unity to near 0.2. Sm all denominator values cause a more abrupt transition (i.e., complete the transitio n over a small length range) while larger values imply a smoother transition. The asymptote at 0. 2 requires at least some LCR residency for even the largest fish and is consiste nt with the observation that a dult humpback chub use the LCR for spawning (Gorman and Stone 1999). This func tion can be considered analogous to the proportion of the year that fish of given size occupy the LCR. I then defined a weighted temperature function ex perienced by fish of a particular length as tTPLCR tTPLCRtTMS LCR 1, (3-20) where is the time-dependent water temperature in the LCR and is the timedependent water temperature in the mainstem Colorado River. This overall temperature experienced by a fish of a given le ngth is then used in equation (3-18) to predict growth rate considering time-dependent changes in water te mperature and size-dependent changes in LCR versus mainstem Colorado River occupancy. tTLCRtTMSTo model the time-dependent water temperature in the LCR, I utilized data reported by Voichick and Wright (2007) to predict average monthly water temperature considering data 1988-2005. I fit these data with a sine curve as peak ave ave LCRtt TTTtT 2sinmax, (3-21) where t is time in fraction of a year starting April 1, is a phase shift allowing predicted peak temperature to align temporally w ith the observed peak temperature, is the amplitude temperature and roughly corresponds to the average annual temperature, and is the maximum annual temperature. I estimated and by minimizing the squared difference between observed and predicted monthly temperature. peaktaveTmaxTpeaktaveTmaxT 66

PAGE 67

Annual water temperature varia tion in the mainstem Colorado River near the confluence of the LCR is much less variable (range 8-12C) th an within the LCR (Voich ick and Wright 2007). Thus, I assumed constant water temperature in the mainstem Colorado River of 10 C. This value corresponds roughly to the average water te mperature within the LCR inflow reach of the Colorado River during much of the time when the growth increments were observed (19892006). I estimated the parameter vector ={H, d, m, n, Lt } by maximizing the log-likelihood equation (3-15). With this more complex m odel, predicted recapture lengths were found by integrating the temperature-dependent growth mode l (equation 3-18) with respect to time. These predictions were then used in the second term of equation (314) to compute the deviations between observed and predicted growth. Followi ng guidance from a meta-analysis by Clark and Johnson (1999), I specified as 2 to reduce the parameter set. I specified the weighting term for the log-likelihood penalties equal to that used in the previous analysis To further reduce the parameter set, I specified = 2,000 to correspond with a coefficient of variation of about 10% as is the minimum typically observed in fish popul ations (S. Martell and C. Walters, University of British Columbia, personal communication). I compared model fit for the temperatureindependent growth model and the temperaturedependent growth mode l using AIC techniques (Burnham and Andersen 2002). cQmQ2 LResults I fit both the temperature-independent (T IGM) and temperature-dependent (TDGM) growth models to 14,971 observed growth increm ents extracted from the humpback chub markrecapture database. All fish were larger than 150 mm TL and the time interval between capture and recapture exceeded 30 days. Although greater than 60% of the fish were at large for 1 year 67

PAGE 68

or less, a small fraction of the observations were for much longer time intervals with the longest interval about 15 years (5,538 days). I estimate d the measurement error contained in the dataset by computing the observed difference in measured lengths of fish captured and recaptured within 10 days. This resulted in a m easurement error variance of 31.8 mm2 across all sizes of fish, implying that most TL measurements were with in 11 mm of the true TL. This amount of measurement error is not unexpected considering the difficulty in measuring live fish. Although this error rate is rarely reported, it is likely similar across a wide range of fish species and field conditions when handling live animals. I fit the TIGM with prior variance weighting terms on the d and n parameters ={0.00001, 0. early 0001, 0.001, 0.01, 0.1, 0.5, 1, 10, 100, 1,000, and 10,000} to explore the effect of constraining these parameters to values near standard von Bertalanffy values. The log-likelihood was n identical for all values of =0.01 and greater, but reducing below 0.01 caused large changes in the log-likelihood. Therefore, I specified =0.01 as the weighting value for both the TIGM and TDGM. To estimate the parameters of the TDGM, I first had to fit the time-dependent LCR water temperature model. Fortunately, the sine curve function with parameters =-0.011, =17.9, and =23.2 fit the observed average monthly temperatures very well (Figure 3-2). peaktaveTmaxTThe estimated parameters, log-likelihood, and AIC statistics for the TIGM and TDGM are presented in Table 3-1. The parameter values fo r the TIGM suggest that anabolism scales as 0.5 mass and catabolism scales as 1.15 mass. These values are different than assumed by the standard von Bertalanffy model and also result in an average value that is smaller than would be predicted from simple inspection of the data. In contrast, the estimated scaling parameters (d=.61 and n=0.89) for the TDGM are not much different than what would be expected under the L 68

PAGE 69

standard von Bertalanffy model where the anabolic scaling parameter (d) should be close to 2/3 and the catabolic scaling parameter (n) should be close to unity. AIC results show strong support for the TDGM over the TIGM (Table 3-1). However, the parameter correlation matrices for each of these models show very high correlation indicating that all of the parameters are not separately estimable (Table 3-2 a nd Table 3-3). In situations such as this where the model is not full rank, it has been shown that the AIC (Bur nham and Andersen 2002) is undefined (Viallefont et al. 1998; Bozdogan 2000) suggest ing that the AIC criteria may not be appropriate for this comparison. An alternative way to arbitrate among these tw o models is to simply graphically examine the model fit to the data. The measured growth rate as a function TL at the start of the interval is extremely variable, particularly at smaller sizes (Figure 3-3). This variability is not surprising considering that the rate is m easured as a difference between two imprecise length measurements and, for most measurements, expanded by dividing by a short time increment. It is also apparent that all three lines differed from a strict linear relationship that would be implied using a standard von Bertalanffy model, though the TDGM fits are reasonably linea r through the portion of the predicted curves populated with data. The temp erature-independent model is somewhat of a compromise between the temperature-dependent summer fit and the temperature-dependent winter fit. I also extracted tw o subsets of seasonal data that had only either summer growth or winter growth that suggest oscillating gr owth rate with temperature (Figure 3-4). Each of the models was used to predict length as a function of age. In addition to the two models fit above, I also predic ted length-at-age using the gr owth function reported in the USFWS recovery goals document (USFWS 2002) and length-at-age using the TDGM for a constant temperature of 10 C (Fig ure 3-5). This last curve is equivalent to a fish experiencing a 69

PAGE 70

constant 10 C temperature and is a prediction of length-at-age for a fish spending its entire life in the mainstem Colorado River. Examination of these curves show that the USFWS growth curve predicted somewhat smaller sizes at young ag es and larger sizes at older ages than is implied by the mark-recapture data. The TIGM and TDGM predict very similar length-at-age with the exception of ages 10-25. Two feat ures are apparent TDGM predictions: (1) a temperature-dependent periodic change in growth rate at ages younger than about age-5, and (2) an apparent bend in relationshi p at approximately age-4. This age corresponds to the length at transition (Lt) where humpback chub are rapidly shifti ng from primarily LCR occupancy to primarily mainstem Colorado River occupancy. A Lt length of 236 mm TL was most strongly supported by the data and the TDGM (Table 3-1). Finally, it is informative to utilize the TDGM to predict monthly growth increments as a function of TL. These predictions based solely on field data can be compared to laboratory observations of the same or similar species. I pl otted growth rate predictions from both the LCR population and a population that is experiencing constant 10 C temp eratures (Figure 3-6). This latter curve is presented as a prediction of mont hly growth rates that w ould be observed in the mainstem Colorado River. Discussion Growth model parameter estimation is typically accomplished using paired observations of individual fish age and length (Quinn and Deriso 1999). Obtaining this information often requires sacrificing the animal so that calcareous structures may be examined to determine age. The TIGM and TDGM seek to obtain this information through non-lethal sampling using information that is frequently collected in ro utine mark-recapture studies. Particularly for endangered species such as the humpback chub, a non-lethal method to obtain information on growth is mandatory. 70

PAGE 71

Estimating growth model parameters using incr ement data is not a recent development. Indeed, quite complicated frameworks have been developed that consider multiple growth model formulations and model error in length measur ements at both capture and recapture occasions (Laslett et al. 2002). The model described here in takes a somewhat different approach by starting with a very general gr owth model allowing many different functional forms to describe the relationship between size and ag e. Additionally, the model parameterization allows intuitive inclusion of the effect of temperature on anabolis m and catabolism that have direct interpretation in a bioenergetics framework (Essington et al. 2001). Recent papers to estimate gr owth of marbled lungfish Protopterus aethiopicus (Dunbrack et al. 2006) and wels catfish Silurus glanis (Britton et al. 2007) propos e methods similar to those described in this work. Interestingly, this methodology is clearly motivated by a similar problem the inability to obtain information on the ag e of individual fish. However, the methods proposed in these studies do not explicitly allow gr owth to be influenced by temperature, even though this is clearly an issue, particularly in the wels catfish case (Britton et al. 2007). This study addresses the effect of temperature on humpback chub growth and attempts to estimate the length at which fish transition from primarily LCR occupancy to primarily mainstem occupancy. The general implication from my findings is that growth rate will increase substantially with a temperature increase from 10 C to 20 C as indicated by the values of =4.6 and =2.0. These coefficients su ggest that anabolis m will more than double relative to catabolism across this temperature range. However, Petersen a nd Paukert (2005) constructed a bioenergetics model for juvenile humpback chub and found 2.4 suggesting much less potential for increased growth with increased temperature. Though some of the difference in estimated Q between my analysis and that of Petersen and Paukert (2005) may be related to the cQmQc cQmQ 71

PAGE 72

highly correlated parameters in th e TDGM (i.e., may be able to obtain nearly as good a fit with lower and higher H, d, and m; see Table 3-3.), it is also like ly that laboratory observations of growth rates may not accur ately represent field cond itions (Rice and Cochran 1984). In particular, the field estimate of Qc represents not only physiologi cal (laboratory) constraints on feeding, but also effects of a ny seasonal variations in food availability that are positively correlated with temperature (e.g., ins ect emergence during spring and summer). cQClarkson and Childs (2000) conducted laboratory experiments to evaluate the growth rate of juvenile humpback chub at 10 C, 14 C, and 20 C. They report monthly growth rates of 1 mm, 13 mm, and 17 mm per month for these temp eratures, respectively. Considering the estimated monthly growth rates from the TDGM in Figure 3-6, the TDGM tends to over-estimate the growth rates reported by Clarkson and Childs (2000) at 10 C and under-estimate the growth rate at 20. However, the TDGM results are in overall agreement with this laboratory study. In their study of reproductive ecology of hu mpback chub in the LCR, Gorman and Stone (1999) conclude that adult fish demonstrate a potadromous migration between the mainstem Colorado River and the LCR to spawn. Based on catch rates of humpback chub within the LCR, they suggest that fish larger than 300 mm TL remained in the LCR only long enough to complete spawning activity. They also re port that catch rate of fish between 200-300 mm TL declined by only half following the spawning period. The im plication is that fish between 200-300 mm TL may occupy the LCR for longer periods of time th an fish larger than 300 mm TL. The TDGM estimate of Lt (236 mm TL) is in agreement with these observations suggesting that fish greater than 236 mm TL should predominantly reside in the mainstem Colorado River. This case history should be useful to those studying humpback chub throughout the Colorado River Basin, and to researchers seeking to estimate the relationship between fish age 72

PAGE 73

73 and size using non-lethal techniques. This tech nique shows considerable promise to extract useful information on fish growth from field da ta, rather than laborat ory studies where such information is typically obtained.

PAGE 74

Table 3-1. General growth model results. Model H d m n L 2 L cQ Lt AIC Parameters Rank AIC TIGM 163 0.52 0.0007 1.15 391 961 --133,658 6 2 38,493 TDGM 21.0 0.61 0.46 0.89 434 2000 4.59 236 95,165 8 1 0 74

PAGE 75

Table 3-2. Parameter correlation matrix for the temperature-independent growth model. H d m n H 1 d -0.99 1 m -0.66 0.73 1 n 0.62 -0.72 -0.99 1 2 L 0.14 -0.19 -0.38 0.38 75

PAGE 76

Table 3-3. Parameter correlation matrix fo r the temperature-dependent growth model. H d m n cQ H 1 d 0.74 1 m 0.88 0.94 1 n -0.86 -0.93 -0.99 1 cQ -0.98 -0.82 -0.89 0.88 1 Lt 0.55 0.16 0.35 -0.34 -0.46 76

PAGE 77

Figure 3-1. Predicted and observe d growth rate (dL/dt) as a func tion of total length (TL) at the start of the time interval. Solid squares are observed growth rate of fish initially captured with TL < 250 mm and open circles are observed growth rate of fish initially captured with TL 250 mm. Predicted growth rates ar e simple linear regressions on observed growth rate of fish initially cap tured with TL < 250 mm and of fish initially captured with TL 250 mm. 77

PAGE 78

Figure 3-2. Observed and pred icted monthly Little Colorado Ri ver water temperature. The points are the average observed monthly te mperature and the line is the predicted monthly temperature. 78

PAGE 79

Figure 3-3. Observed and predicted humpback chub growth rate (dL/dt) from the temperatureindependent growth model and the temp erature-dependent gr owth model during summer and winter. 79

PAGE 80

Figure 3-4. Observed and predicted humpback chub growth rate (dL/dt) from the temperaturedependent growth model during summer and winter. 80

PAGE 81

Figure 3-5. Predicted humpback chub length-at-age from the U. S. Fish and Wildlife Service (USFWS) growth curve, the temperature-in dependent growth model, the temperaturedependent growth model for the Little Colorado River (LCR) humpback chub population, and the temperat ure-dependent growth model for humpback chub living in the mainstem Colorado River under a constant temperature of 10C. 81

PAGE 82

82 Figure 3-6. Predicted monthly gr owth rate from the temperaturedependent growth model for the Little Colorado River (LCR) population of humpback chub and for humpback chub living in the mainstem Colorado River under a constant temperature of 10C.

PAGE 83

CHAPTER 4 ABUNDANCE TRENDS AND STATUS OF THE LITTLE COLORADO RIVER POPULATION OF HUMPBACK CHUB: AN UPDATE CONSIDERING DATA 1989-2006 Effective monitoring to evaluate endangered sp ecies status is a vital component of most endangered species recovery plans (Campbell et al 2002). Monitoring results are typically used to evaluate species status with regard to esta blished goals towards chan ging its listing under the U.S. Endangered Species Act. Resource monitoring is also a critical component of adaptive management (Walters and Holling 1990). In th e context of adaptive management, monitoring not only evaluates the status of the resour ce, but combined with probing management experiments it serves to inform managers about how the resource responds to various management actions. This paradigm of l earning by doing places a premium on an accurate and reliable monitoring pr ogram (Parma et al. 1998). Prompted by a National Research Council s ponsored evaluation of the Glen Canyon Dam Adaptive Management Program (GCDAMP; NRC 1999), the Grand Canyon Monitoring and Research Center (GCMRC) has devoted significant resources to developing long-term monitoring programs over the last 8 years. As an example, much effort has been expended to synthesize existing data on fisherie s resources in order to portray trends in these populations. Of particular importance is the humpback chub (Gila cypha), a federally listed endangered cyprinid endemic to the Colorado River Basin in the southwestern U.S. (Gloss and Coggins 2005). Because of this species unique ecological role as one of th e few remaining endemic aquatic species within Grand Canyon and their endangere d listing status, the humpback chub is a focal resource of the GCDAMP (Gloss et al. 2005). The objective of this chapter is to provide updated information on the status and trend of the Little Colorado River populati on of humpback chub in light of new information and refined assessment methodology. Such information constitutes the cornerstone of the humpback chub 83

PAGE 84

monitoring program within the adap tive management program and is also relevant to U.S. Fish and Wildlife Service recovery goa ls for this species (USFWS 2002). The unique life-history attributes of Little Colorado River population (LCR) of humpback chub and the large variety of sampling and monitoring programs ongoing since the 1980s (Coggins et al. 2006a) prompted the developmen t of a new type of age-structured, open population, mark-recapture model called the ag e-structured mark-recapture model (ASMR; Coggins et al. 2006b). This model was subseque ntly used in combination with other markrecapture and index-based assessments to provide a comprehensive assessment of the LCR humpback chub population (Coggins et al. 2006a). The ASMR approach has been subjected to a series of independent peer eval uations both as part of the GC DAMP (e.g., Kitchell et al. 2003) and through peer review as part of the publication process. Since publication of the last assessment, I have continued development of the ASMR model to address concerns presented in previous reviews (Kitchell et al. 2003). These improvements include incorporation of a formal model comparison approach and consideration of the uncertainty in assigning age to individual fish. A central problem in conducting humpback chub stock assessment is the assignment of age to individual fish. Though this problem is ubiqui tous in fish assessment programs (Coggins and Quinn 1998; Sampson and Yin 1998), it is particular ly difficult when working with endangered fish and when sacrificing the animal is necessary to determine age. In this situation, individual fish ages must be assigned based on fish length s and assuming some relationship between length and age. In past humpback chub assessments (Coggins et al. 2006a), I assumed that this relationship was adequately described from a growth curve based on a limited collection of paired age and length observati ons (USFWS 2002). However, th is age-length relationship is 84

PAGE 85

based on an extremely small sample size and is therefore suspect. Additionally, when assigning individual age based on this relations hip, I assumed that fish could be aged without error. This is clearly not a valid assumption and presumes much more certainty in the assignment of age than is warranted. To alleviate these shortcomings, in Chapter 3, I presented a method to estimate the relationship between fish age and length using mark -recapture data. In this chapter, I use that relationship to translate uncertainty in the dete rmination of age from length to uncertainty in abundance and recruitment estimates from ASMR These analyses offer insight into the humpback chub assessment and other monitoring programs for aquatic and terrestrial species where mark-recapture methodologies serve as the core of the assessment approach, but where estimated trends in recruitment and mortality ar e influenced by uncertainty in age assignment. The ongoing monitoring program for humpb ack chub in Grand Canyon has varied in intensity over the years, but the primary sample locations, techniques, and personnel have remained remarkably consistent (Coggins et al. 2006a). Conducting the annual stock assessment and continuously evaluating the performance of the assessment through retrospective analyses, independent peer evaluation, and testing of the m odel with simulated data all provide insight into the performance of the model-based on the availa ble data. This comprehensive examination may prove useful to other adaptive management programs that seek to develop a robust monitoring component, and in particular to provide insight into: (1 ) limitations of monitoring alone to assign cause and effect associated with prescriptive management actions, (2) pathologies associated with large changes in monitoring protocols, and (3) a realistic assessment of the considerable uncertainty in results for a rare, elusive, long-liv ed animal even after many years of intensive monitoring. 85

PAGE 86

Methods Monitoring efforts for the LCR population of humpback chub began in 1987 when a standardized hoop net sampling program was implem ented in the lower reaches of the LCR. During the subsequent 19 years, 4 sampling peri ods can be generally defined corresponding to different levels of sampling effo rt and protocol, partic ularly in the LCR (Coggins et al. 2006a). The first period of sampling lasted until 1991 and consisted mainly of limited hoop netting in the lower 1,200 m of the LCR. Sampling period 2 (1991-1995) involved an intensive sampling effort in both the LCR and the mainstem Colorado River as part of an environmental impact study on the operation of Glen Canyon Dam (USDOI 1995). The third sampling period began in 1996 with severely reduced intens ities compared to period 2. Finally, beginning in fall 2000 a period of higher sampling intensity relative to period 3 (but less intensive than period 2) began and continued through 2006. During each of these sampling periods, humpback chub have been collected using multiple gear types (by many of the same personnel) including hoop nets and trammel nets in the LCR, and these same gears plus pulsed DC electrofishing in the mainstem Colorado River (Valdez and Ry el 1995; Douglas and Marsh 1996; Gorman and Stone 1999; Coggins et al. 2006a). Index-Based Metrics Although index-based metrics (e.g., catch rate ) can be unreliable to track trends in population size with great precis ion (MacKenzie et al. 2006), th ese indices are frequently examined and are potentially useful for comparis on to previous assessment efforts. With this caveat in mind and following Coggins et al. (20 06a), I updated two long-term catch rate time series with data from 2003-2006: (1) hoop net ca tch rate of humpback chub in the lower 1,200 m of the LCR, and (2) trammel net catch rate of humpback chub in the LCR inflow reach of the Colorado River (defined as approximately 9 km upstream and 11 km downstream of the 86

PAGE 87

confluence; Valdez and Ryel 1995). Details ab out these sampling programs are provided by Coggins et al. (2006a). Tagging-Based Metrics The heart of the tagging-based assessment is the large number of uniquely tagged sub-adult (150 mm-199 mm total length; TL) and adult (>200 mm TL) humpback chub that have been captured, measured, and implanted with passive integrated transponder (PIT) tags. Since 1989, over 19,000 humpback chub have been released with unique identifier s. These data are maintained in a central database housed at the GCMRC. Capture-recapture-based methods to assess population abundance and vital rates have been widely used in fisheries and wildlife studies fo r well over 50 years, and numerous reviews have been conducted highlighting the general approaches (e.g., Seber 1982; Williams et al. 2002). Traditional methods (e.g., Jolly-Seber type met hods) generally rely on recaptures of tagged or marked individuals to estimate abunda nce, recruitment, and survival. The approach is to create a known population of marked fish th at are repeatedly sampled to obtain time series estimates of mark rate (i.e., proportion of the overall population that is marked) and the number of marked fish alive in the population. These estimates are subsequently used to estimate capture probability, abundance, recruitment, and survival. Here, I briefly describe the overall ASMR me thod and refer readers to Coggins et al. (2006b) for full details. The ASMR model differs from the trad itional approach because in general it contains more structural assump tions through the specification of a population accounting structure governing tran sition of both marked and unmarked animals through ages and time. A standard fisheries virtual populati on analysis framework (Quinn and Deriso 1999) is used to annually predict the num bers of marked and unmarked fi sh available for capture. The total number of marked fish depends on numbers of fish recently marked as well as previously 87

PAGE 88

marked fish decremented by mortality rate. Nu mbers of unmarked fish depend on the time series of recruitment, the numbers of fish marked fr om those cohorts, and the mortality rate. These annual predictions of the abundance of marked a nd unmarked fish are further segregated by age such that age-specific survival and capture pr obability may be considered. Parameters are estimated by comparing predicted and observed ageand time-specific captures of marked and unmarked fish in a Poisson likelihood framework. The ASMR model has three diffe rent parameterizations (ASMR 1-3) that vary in how the terminal abundance is estimated and how agean d time-specific capture probability is modeled. Both ASMR 1 and 2 assume that ageand time-s pecific capture probability can be modeled as the product of an annual overall capture probabil ity multiplied by age-specific vulnerability. This is the common parameterization of fishi ng mortality in many assessment models under the separability assumption (Megrey 1989) and dimini shes the size of the parameter set since it is not necessary to separately estimate each ageand time-specific capture probability. ASMR 1 and 2 further assume that vulnerability is asympto tic with age. As such, vulnerability is assumed to be unity for fish age-6 and older and estim ated only for the younger fish. Finally, annual agespecific vulnerabilities are assumed to be equa l among each sampling period as described above. Implicit in this assumption is that within a sampling period, annual age-specific capture probabilities differ only as a scalar value propo rtional to the annual ove rall capture probability. The primary difference between ASMR 1 and AS MR 2 is how the terminal abundances are calculated. ASMR 1 estimates an overall termin al year capture probability and calculates agespecific terminal abundances (both marked and unmarked fish) as the ratio of ag e-specific catch (both marked and unmarked fish) and age-specif ic capture probability (i.e., product of the terminal year capture probability and sampling period 4 age-specific vulnerability). In contrast, 88

PAGE 89

ASMR 2 treats age-specific terminal abundances up to age-13 as individual parameters. Terminal abundances for subsequent ages are es timated by applying age-sp ecific survivorship to the age-13 abundance. This difference in formul ations decreases the parameter count for ASMR 1 relative to ASMR 2 at the expe nse of assuming that the vulnerabi lity schedule in the terminal year is identical to the rest of period 4. ASMR 3 is the most general model in that it makes no assumption as to the ageor timespecific pattern in capture proba bility. The conditional maximum likelihood estimates of ageand time-specific capture probability are used to predict the ageand time-specific catch of marked and unmarked fish. Full details of each of the models are provided by Coggins et al. (2006b). In addition to the ASMR assessments, I also update the time series of the annual spring abundance estimates in the LCR. Abundance of humpback chub in the LCR 150 mm TL was estimated during the early 1990s and 2001-2006 usi ng closed population models. These models included the program CAPTURE suite of mode ls (Otis et al. 1978) and simple Chapman modified Lincoln-Petersen length-stratified models (Seber 1982). The recent estimators use data collected annually during two sampling occasions in the spring. Full details of the sampling and estimation methods are provided by Douglas and Marsh (1996) a nd Coggins et al. (2006a). Coggins et al. (2006b) recomme nded exploring the use of individual capture histories within the ASMR framework to reduce confounding between capture probability and mortality. Though the updated ASMR models presented in th is chapter do not yet incorporate individual capture histories, they do model recaptured fish by annual tagging cohort with the intent of reducing parameter confounding by increasing th e number of observations available for parameter estimation. In the non-tag cohort or po oled version of ASMR described by Coggins et 89

PAGE 90

al. (2006b) and above, ageand ti me-specific predictions of recaptured fish are not separated by year of tagging. As an example, assume that ASMR 3 predicts that 50 marked age-6 humpback chub should be captured in 2002. These 50 fish c ould be comprised of fish tagged as: age-5 in 2001, age-4 in 2000, age-3 in 1999, or age-2 in 1998. However, as the model is currently formulated, all age-6 fish recaptured in 2002 are pooled for a single observation. Assuming that the ageand time-specific captures of marked and unmarked fish are Poisson distributed, the loglikelihood ignoring terms involving only the data is computed as A a T t tatata A a T t tatatarrr mmm rmL12 ,,, 11 ,, ln ln ,ln, (4-1) where ma,t is the observed number of age a unmarked fish captured in year t is the predicted number of unmarked fish captured, ra,t is the observed number of marked fish captured (i.e., recaptures), is the predicted number of marked fish captured, and is the parameter vector to be estimated. Notice in the second term that the individual log-likelihood terms are summed over age and time. However, it may be more info rmative to stratify the recapture data by tagging cohort. The proposed lo g-likelihood is then tam,tar, A a T t T c ctactacta A a T t tatatarrr mmm rmL12 1 1 ,,,,,, 11 ,,, ln ln ,ln, (4-2) where c is the tag cohort (i.e., all fish marked in year t ). In principle, this stratified loglikelihood should provide additional information on time-specific capture probability and may improve parameter estimation. Evaluating Model Fit Following Baillargeon and Rivest (2007), I us ed standardized P earson residuals of observed and predicted age composition for both unma rked and marked fish to evaluate model fit 90

PAGE 91

among the three different ASMR models. The st andardized Pearson residual is the difference between the observed and predicte d values scaled by an estimate of the standard deviation as t tata tata tan pp po r )1(, ,, (4-3) where is the number of observations (e.g., the number of marked fish recaptured each year) and o and are the proportions of fish in each year and age class obser ved and predicted, respectively. I plotted the indi vidual Pearson residuals for each combination of age and time to look for consistent bias for individual brood year cohorts. Additionally, I used Quantile-Quantile (Q-Q) plots to compare the dist ribution of the Pearson residu als to a theoretical normal distribution. The slope of the theoretical curve is approximately the sta ndard deviation of the distribution of Pearson residuals where a small value of the slope indicates a narrow distribution of the residuals. Deviations from the theoretic al curve indicate a non-no rmal distribution of the Pearson residuals and imply that the model error is not well di stributed (e.g., tending to more often either overor under-predi ct age proportions) and possibly inducing bias in parameter estimates. tnta tap,In addition to examination of model fit using Pearson residuals, I chos e to also rely on information theory to aid in model evaluation. This approach is increasingly common in ecological studies to arbitrate among competing models and is primarily concerned with estimating the Kullback-Leibler (K-L) distance be tween the model and the truth as a measure of model support (Burnham and Andersen 2002). The Akaike Information Criterion (AIC; Akaike 1973) is the standard estimator for th e relative K-L distance and is computed as a function of model likelihood and num ber of model parameters. Follo wing external review of the ASMR method, Kitchell et al. (2003) pointed out that although ASMR uses a quasi-likelihood 91

PAGE 92

structure of estimating equations and true likelihood, estimates of relative K-L distance using AIC, though not strictly appropriate, would be valuable to consider for model selection. Therefore, in addition to the evaluation base d on Pearson residuals, I also conducted an AIC evaluation. Incorporation of Ageing Error in ASMR Assessments As mentioned above, Coggins et al. (2006a) assigned age to indi vidual fish strictly as an inverse von Bertalanffy function. This procedure ignores variability in the age of fish of a particular length and tacitly assumes that age assi gnments can be made much more precisely than is true. To account for uncertainty in the assignment of age using length, I estimated the probability of age for fish having lengt h within a particular length interval laP. Following methods reported by Taylor et al (2005), I define this proced ure by first specifying that the probability of an age a fish having length within length bin l is dl dl a a adl ll alP2 22 exp 2 1 (4-4) where length bin l has mid-point length l minimum length l-d and maximum length l+d These probabilities can be thought of as a matrix w ith rows corresponding to length bins and columns as ages. As is obvious from equation (4-4), entr ies within a particular column (age) can be thought of as resulting from the integral over eac h length bin of a normal probability density with mean and variance The mean length-at-age is computed from the temperaturedependent growth model (Chapter 3) and the variance of length-at-age is assuming that coefficient of variation in length ( ) is constant among ages. al2alaacvl2lcv 92

PAGE 93

alP laP available, one option to com pute would be to normalize each matrix With cell by the sum of its row as: iilP alP lap1|. (4-5) However, Taylor et al. (2005) s uggest that this procedure will induce bias if the population ha experienced size-dependent mortality such as size selective fishing or size-dependent chang natural mortality. This results because within a particular age class, fast growing individuals (i.e., largeL) may experience either higher or lower mortality rate than their cohorts, a therefore be either overor unde r-represented in the population. This sorting by growth rate can favor either slow growing individuals, as in the case of increasing vulnerability to exploitation with size, or fast growing individual s, as in the case of reduced natural mortality with size. Therefore, Taylor et al. (2005) suggest that anAs es in nd adjust ment for mortality must be made to accurately predict the proportion of individuals in each age and length bin. Accordingly, I define the numbers of fish in each age and length bin as alPNNaal,, (4-6) where is the abundance of fish at each age. (4-7) With abundance a bin can then be calculated as aN If the age specific mortality rate (aM) is available and recruitment ( R ) is assumed constant, abundance-at-age is given by 1 i iM aeRN. 1 at each age and length bin thus available, the proportion in each age and length T al alN N P, (4-8) 93

PAGE 94

la ta TN N,. The probability of age given length is then calculated as where A ilP, i alP laP1 ,|. (4-9) Taylor et al. (2005) focus on ag e-specific mortality driven by vul nerability to exploitation. For the unexploited humpback chub, age-specific mortality as a function of changes in natural mortality was included. Lorenzen (2000) demonstrated that much variation in natural mortality can be explained by size of fish. Thus, Loren zens allometric relationship between natural mortality and length was used to calculate a declining mortality rate with age as a al LM M, (4-10) where is the mortality rate suffere d by an adult fish of size This mortality schedule was calcuith specified as 0.148, as estimated by ASMR 3 considering tag-cohort specific data (see Results). I computed four seasonal matrices that I used to a ssign age to fish captured at different times of year. Growth during the year could thus be accounted for by Mlated wLMrecalculatinglaP| alP such that length-at-ag was computed as either l(a) l(a+.25) l(a+0.50) e or l(a+0.75) The resulting seasonal l aP| matrices were then used to assign age to a fish depending on the quarter of the year in which it was captured. To incorporate the uncertainty in assigni ng age based on length into the overall assessment, I used a Monte Carlo procedure wher e age was stochastically assigned to each fish based on the seasonal matrices. To understand this pr ocedure, it is first helpful to recognize that given a fish with length in bin l the probability of belonging to each age is laP| 94

PAGE 95

multinomial with number of categories equal to the number of ages. I used the multinomial random number generator within program R (R Development Core Team 2007) to randomly assign enier 2000) to compute the 95% profile confidence interval for a dult abundance and recruitment. I repeated this procedure to atase ts (i.e., Monte Carlo trials). ). 006 (i.e., age to marked fish. Recapture age was ca lculated as the sum of age-at-tagging and tim at-large. I then stochastically assigned age to each fish using this pr ocedure. For each resulting dataset of capturesand recaptu res-at-age, I estimated adult (a ge-4+) abundance and recruitment using ASMR 3. Additionally, I used AD Model Builder (Four generate and analyze 1,000 dResults Index-Based Assessments Between 1987-1999 and 2002-present, the Arizona Game and Fish Department sampled humpback chub using hoop nets in the lower 1,200 m section of the LCR. Examination of this index suggests that the abundance of both sub-adult (150-199 mm TL) and adult ( 200 mm TL) humpback chub declined during 1987-1992 and rema ined relatively constant through much of the 1990s (Figure 4-1). Since 2003, there is a slight upward trend in the catch rates of sub-adult fish. Note that several data points in this index are shifted slightly relative to those reported by Coggins et al. (2006a). This adjustment is due to additional standardizati on of the data used to construct this index (D. Ward, Arizona Game and Fish Department, Personal Communication The trammel net catch rate of adult abundance in the LCR inflow reach of the Colorado River suggests a similar trend in adult fish ( 200 mm; Figure 4-1). In general, this index shows a stable to declining trend through the 1990s with a slight indica tion of increased abundance in most recent years. All monthly trammel ne t samples from the LCR inflow reach for 1990 are presented in Figure 4-1. However, onl y samples from 1990-1993, 2001, and 2005-2006 95

PAGE 96

dark circles in Figure 4-1) re present robust sampling coverage throughout the entire reach. Annual sample sizes in 1994 and 2002 (i.e., hollow circles in Figure 4-1) were between 2% and 50% of the 1990 average samp le size, and in some years effort was focused near the LCR confluence where humpback chub density may have been highest. Thus, the 1990, 2001, and 2005-2006 data are likely to best depict the overa ll trend of relative abundance w ithin this reach. Simple linear regre ssion analyses provide estimated slopes that are zero (p = 0.16 for all data, and p= 0.26 for the preferred data; Figur ed re not significantly different from e 4-1). Tagging-Based Assessments As described above, the data required for the ASMR models are numbers of fish mark and recaptured each year and for each age. For the results contained in this section, all ages are assigned based on the standard von Bertalanffy gr owth curve as described in Coggins et al. (2006b). With that in mind, examination of the age distribution of fish marked and recaptured since 1989 provides insight into the trends in sampling effort, and also provides important information related to humpback chub mortality (F igure 4-2). The top pa nel of figure 4-2 shows the numbers of newly marked fish and is infl uenced by both trends in sampling effort and numbers of un-marked fish alive. The most cons istent period of sampling has been since about 2001 with about 1,100 fish marked annually (numbers of fish collected at the top of the bubble columns). Because a large fract ion of the population was marked in sampling periods 1 and 2, the majority of fish marked in recent years are young fish and the number of new fish marked each year declines with fish age. The bottom pa nel of figure 4-2 represents the numbers of fish recaptured each year. Some of the same patterns related to sampling effort are evident, but the are some very interesting patterns that result from the high sampling effort in the early to mid 1990s (Figure 4-2). For example in 1995, a tota l of 1,244 humpback chub were collected and 96

PAGE 97

902 of these fish had been marked in previous year s. This pattern is evid ent for several years of data indicating that the high sampling effort in the early 1990s resulted in marking upwards of 70% of the humpback chub population. It is also apparent that because few age-3 to age-5 fish were marked during period 3, there were few age8 to age-10 fish recaptured in the early 2000s five years later. This contribu tes to the spoon shapes to the lower panel of figure 4-2, where there 0 ge of old st marked in the early 1990s that continue to be recaptured into 2006. This fish demonstr ates the low mortality rate suffered by older hump f data collected in this e up to twice as large as the 2006 estimate (R. Van Have were relatively large numbers fish < age-10 recaptured in period 4, lower catches of age-1 to age-15, and relatively stable numbers of fish > age-15. Another finding is the extreme longevity of thes e fish. This is evident by examining the number of humpback chub of each age marked in each year and recaptured in subsequent years (Figure 4-3a through Figure 4-3e). Figure 4-3a shows the number of humpback chub of each a marked in 1989-1992 and recaptured in subsequent years. There is a remarkable number fish (> age-15) fir slow decay pattern of marked back chub. Closed Population Models The time series of abundance estimates for humpback chub 150 mm TL in the LCR during spring implies a decline in abundance fr om the early 1990s to present (Figure 4-4). However, as is apparent in the data, these esti mators are very imprecise with corresponding poor ability to detect significant trends. Additionally, preliminary analyses o program suggest that the 2007 estimate may b rbeke, U.S. Fish and Wildlif e Service, Personal Communication). ASMR Without Tag Cohort Specific Data The three ASMR formulations generally agree that adult (a ge-4+) humpback chub abundance has been gradually increasing since ab out 2001 (Figure 4-5). For the three models 97

PAGE 98

the 2006 adult abundance estimate is 6,690 (95% CI 6,403-6,994), 6,768 (95% CI 6,397-7,131), and 6,648 (95% CI 6,222-7,102) for models ASMR 1, ASMR 2, and ASMR 3, respectiv These results suggest that this population has increased from an estimated low of approximately 4,800-5,000 during 2000-2001. Estimated recrui tment (age-2) among models is also in agreement (Figure 4-6). Following low recruitm ents for brood years dur ing the early 1990s, all the models suggest that recruitm ent increased through the latter part of the 1990s. The biggest discrepancy among the three models is that ASMR 1 suggests a decline in recruitment following the 2001 brood year, while the other two models sugge st stability. The structural assump model ASMR 3 (see Coggins e 2006b) do not permit a reliable recruitment (age-2) estimate for brood ye ely. tions of t al ar 2003. An additional difference in the models results are the estimates of ortality ranges from 0.119 (ASMR 1) to 0.133 (ASM other way, instantaneous adult mortality (M) where adult m R 3). Model Evaluation and Selection With these results in hand, the question become s which model is best? Stated an which model produces results most consistent with or best supported by the data? The discrepancies among model result s related to adult abundance are not large, so from a management or conservation perspective, selectin g the best model is probably not critical. However, the models do suggest rather different recruitment trends. Model ASMR 1 supports the hypothesis that recruitment has declined fo llowing the 2000 brood year while the other two models suggest relative stability. Therefore, selecting which model is most consistent with the data is desirable. The patterns in Pearson re siduals for both ASMR 1 (F igure 4-7) and ASMR 2 (Figure 4-8) demonstrate systematic lack of fit for pa rticular sets of cohorts. This is best seen in the recapture residuals where it is apparent that there were more fish observed than predicted for 98

PAGE 99

about eight pre-1990 cohorts, partic ularly for observations after 2000. Additionally, there w fewer recaptures associated with the 1992 cohort than predicted. These systematic trends likely impose bias in the model results for ASMR 1 an d ASMR 2. In con ere trast, there is much less system umeric rs are justified. To provide in sight into this question, I estimated relative K-L distan rior 0h MR 2 require that vulnerability is asymptotically related to age, it is not o account for this unexpected pattern, and thus display poor atic lack of fit in the residual patterns for ASMR 3 (Figur e 4-9). Among the three models, the Pearson residual standard devi ation was smallest for ASMR 3. The finding that ASMR 3 has the best fit among the three models is not surprising since it has the largest parameter set. Although ASMR 3 only varies 13 parameters in the direct n search, the conditional maximum likelihood estimates are used for each ageand time-specific capture probability (Coggins et al. 2006b). Therefore, a nd assuming a liberal maximum longevity of 50 years, ASMR 3 has 895 paramete rs. The question then becomes whether these additional paramete ce using AIC (Table 4-1). These results st rongly indicate that model ASMR 3 is supe to ASMR 1 and 2. Since the fundamental difference between ASMR 1-2 and ASMR 3 is the amount of flexibility in ageand time-specific capture pr obabilities, I examined the pattern in ASMR 3 estimated capture probabilities (Figure 4-10). The patterns in age-specific capture probabilities during sampling period 2 (i.e., 1991-1995; heavy gr ay lines) and sampling period 4 (i.e., 200 2006; heavy black lines) differ markedly. These fi ndings suggest that there was a major shift in the gear selectivity; sampling si nce 2000 appears to be much le ss effective at capturing fis between ages 9-20 than was sampling during the s econd period. Since structural assumptions in ASMR 1 and AS surprising that these models are not able t model fit. 99

PAGE 100

ASMR with Tag Cohort Specific Data In addition to repeating the analyses by Coggi ns et al. (2006a) above, I also fit the A models to the tag cohort specific data using the log-likelihood in equation (4 -2). The trends in adult abundance and recruitment are similar to those found using the simpler log-likelihood (Figures 4-11 and 4-12). In general, adult abundance estimates are slightly higher at the beginning of the time series and slightly lowe r at the end. Adult abundance estimates for 2006 were 6,057 (95% CI 5,797-6,308), 6,138 (95% CI 5,842-6,458), and 5,893 (95% CI 5,554-6,242) for the ASMR 1, ASMR 2, and ASMR 3 m odels, respectively. Adult mortality (MSMR ) estimates data indicate slightly higher adult mortality than when fit to the po istent (Table 4-2) in general m ASMR 3 sugge entified utions h from the models fit to the stratified oled data and ranged from 0.128 (ASMR 1) to 0.148 (ASMR 3). This finding is cons with the more rapid decay observed in the time series estimates of adult abundance. Model Evaluation and Selection Examination of Pearson residuals for the ta g cohort specific m odels suggests similar patterns in model misspecification for ASMR 1 and ASMR 2 (Figures 4-13 and 4-14) relative to ASMR 3 (Figure 4-15). As with the pooled tag cohort data, ASMR 3 displays better fit. Model evaluation using AIC methods again suggests that ASMR 3 is prefer able agreement with the residual evaluation. Finally, the estimated capture probability fro sts a similar mechanism to explain the poo r performance of mode ls ASMR1 and ASMR 2 (Figure 4-16) as was found for the w ithout tag cohort sp ecific analysis. Incorporation of Ageing Error in ASMR Assessments I used the temperature-dependent growth model (Chapter 3) and the pr ocedures id above to construct seasonal laP| matrices. I then plotte d the resulting probability distrib as surfaces to allow examination of the uncertainty in predicting age given lengt 100

PAGE 101

(Figure 4-17). The most obvious feature of these probability surf aces is the increasing uncertainty in age assignment with increasing length. For instance and considering the AprilJune laP| surface (Figure 4-17), one can see that a 150 mm TL fish is age-2 with highest probability, but there is some chan ce that it is all ages between ag e-1 and age-4. In contrast, a 300 mm fish is approximately ag e-7 with highest probability but could be as young as age-4 or as old as age-18. It is precise ly this uncertainty that I sought to incorporate in the assessment. I stochastically assigned age to each fish using the appropriate laP| matrix depend on the time of year the fish was first captured. Using this procedure, I generated a total o input datasets and fit the ASMR 3 model to each. For each model fit, I retained the estimated annual adult abundance and 95% profile likelih ood confidence bounds. I also retained the estimated brood year recruitment and ing f 1,000 95% c onfidence bounds. Note that because of the unce t ates fromodels are in agreement that recr uitment has been increasing since about the mid rt the two m ainty in assigning age to even the smallest fish, newly tagged fish had the possibility of being assigned age-1. As a re sult, I expanded the age range of the model such that recruitment estimates were for age-1 fish. Estimated adult abundance (age-4+) from model ASMR 3 ranged from 9,322 (95% CI 8,867-9,799) in 1989 to 6,017 (95% CI 5,369-6,747) in 2006 (Figure 4-18). As expected, these estimates have lower precision than those from ASMR 3 ignoring ageing error. The coefficien of variation in adult abundance estimates considering ageing error ranges from approximately 1%-7% in contrast to 0.5%-3% if uncertainty in assignment of age is ignored. The recruitment trend considering the new growth function and the incorporation of ageing error is much less precise than when ageing error is ignored (Fig ures 4-12 and 4-19). Although the point estim 101

PAGE 102

1990s, the uncertainty in the recr uitment estimates from the latte r assessment makes statements about the tes and residual analyses considering both pooled and strat clusion of ageing error increases the uncer ance t likely no late r than 1999. Recruitment of juvenile humpback chub since 2000 appea d in ndance differences among years quite tenuous. Results Summary The adult portion of the LCR humpback chub po pulation has increased in recent years as a result of increased recruitment particularly asso ciated with brood years 1999 and later. Model evaluation procedures indicate that the results from model ASMR 3 are most consistent with the data. Utilizing data stratified by tagging cohort appears to add little additiona l information to assessment as indicated by overall similarity in abundance and recruitmen t estima ified data. In tainty about individual annual estimat es, but gross trends remain the same. Discussion The overall result of the mark-recapture-bas ed open population model assessment is that the adult portion of the LCR humpback chub populat ion appears to have increased in abund since 2001. The assessment model best supporte d by the data is ASMR 3 with a corresponding 2006 adult abundance estimate of approximately 5,900-6,000 fish. In addition, this model suggests that there has been an increase in the adult abundance of appr oximately 20%-25% since 2001. This increase appears to be related to an increasing recruitment trend beginning perhaps as early as 1996, bu rs stable, but the precision of these estimate s is quite low when ageing error is include the assessment. The LCR hoop net abundance index suggests a modest increase in the abundance of juvenile fish and stability in the abundance of adult fish. In addition, the LCR inflow reach trammel net abundance index indicates stability w ith a slight indication of increased abu in 2005 and 2006. Though there would be increased confidence in the mark-recapture-based 102

PAGE 103

open population model results if the catch rate metrics indicated sim ilar trends, it is not surprising that these index measur es are not able to detect a 25% increase in abundance. Th basic assumption of catch rate indexes is that capture probability must remain constant for the index to be well correlated with abundan ce (MacKenzie and Kendall 2002). There is good reason to suspect that this assump tion is violated for the index data series presented in this update due to the influence of turbid ity on catchability (Arreguin-Sanc hez 1996). Turbidity appe influence humpback chub catchabil ity in the Little Colorado Rive r (Dennis Stone, U.S. Fish e ars to and Wildl f e ility e was related to a very large age class entering the samp ife Service, Personal Communication) and turbidity varies greatly in the mainstem Colorado and Little Colorado Rive rs as a function of tributar y freshets and dam operations. A more significant concern is the lack of correlation between ASMR 3 results and the mark-recapture closed population model estimates in the LCR. However, since the number of fish in the LCR during sampling is influenced by migration magnitude and timing, this source o variability may obfuscate expected correlations with the ASMR 3 result s. It is also clear that th low precision of these annual closed population model estimates may not permit detection of a 25% increase in adult abundance. Additionally, pr eliminary analyses of data collected during 2007 suggest that the abundance estimate for 200 7 may be twice as la rge as the 2006 estimate (R. Van Haverbeke, U.S. Fish and Wildlife Service, Personal Communi cation). Though this result would provide support for the ASMR 3 resu lts, it would also call in to question the ab of the LCR program to provide a consistent meas ure of overall population size. One would have to reconcile whether that level of chang led population, a larger than normal fracti on of the population entering the LCR during the sampling period, or some other factor. 103

PAGE 104

Though the GCDAMP is fortunate to have such a large mark-recapture database for thes high-profile endangered animals, significant cha nges in sampling protocol over time continue to cause ambiguity in these analyses. As identified by Melis et al. (2006), re trospective analyses of the data suggest a continual updating of the adu lt mortality rate as additional information has been collected since 2000. Following addition of the 2006 data, this updating is again apparen (Figure 4-20). It appears that adult mortality ra te may be stabilizing as more data are collected but it is difficult to be certain. I believe that the likely cause of this updating is the sampling program essentially having to catch-up following the low sampling effort during period 3. When focused analysis of this dataset began with open population models in 2000 (GCMRC unpublished data), there had been so little sampli ng in the mid to late 1990s that the models interpreted the lack of old fish captures as a relatively high adult mortality rate. As additional data was collected through a more rigorous sampling program during 2000-2006, each time the model saw a recent old fi e t sh recapture, mortality rate was adjusted downward. The hope is that if the To nly GCDAMP continues with a fairly unif orm sampling program over time, adult mortality rate will stabilize and only abundance estimates in the last few years of the dataset will be subject to much updating. An additional finding, identified by Martell (200 6) and in this assessment, is the major change in gear selectivity between periods 2 and 4. I am uncertain what is driving the trend to lower the capture probabili ty for the middle aged fish. Howeve r, it has been suggested that the high capture probability for middle aged fish was due to extensive trammel netting effort in the LCR inflow reach of the Colorado River during peri od 2. It is apparent from Figure 4-1 that there is a large difference in the amount of trammel netting effort in period 2 versus period 4. investigate this possibility, I fit the ASMR 3 mo del to a subset of the database containing o 104

PAGE 105

LCR data. The results indicated an almost identical pattern in age-specific capture probability is observed with the full dataset. This is no t surprising since fish captured in the LCR inflow reach of the Colorado River represent only about 10% of the entire humpback chub markrecapture database. It has also been suggested that reducing th e use of large hoop nets in the LCR during period 4 has reduced the catch rate of the largest fish Though this is possib net throat openings are the same size on all nets used during both sampling periods. It has also been suggested that as le, the sampling in the LCR only 4 months of the year during period 4 as opposed to 10 the l ogram. commendation of Williams et al. ( 2002) that the objectives of the monitoring progr e clearly l. to 12 months of the year during period 2 may be the cause. This is also possible, particularly if there is some differential migr ation timing for the middle aged fish relative oldest individuals. Large changes in sampling protocol should be approached with cau tion in light of how those changes may affect ability to infer populat ion change. This is particularly true for populations that are in low abundan ce and individuals difficult to cap ture. I suggest that carefu simulation of considered changes may help to ex pose potential problems or at the very least, help to clarify thinking relate d to proposed changes in samp ling protocol. Finally, those considering implementing a mark-recapture-base d monitoring program should plan to expend considerable sampling effort using similar protoc ols for the duration of the monitoring pr I echo the re am with regard to issues such as precision of measured quantities sh ould not only b identified, but that the measured quantities should be directly linked to the management objectives. A major criticism of the ASMR technique as previously applied is that it does not explicitly account for uncertainty in the assignm ent of age to individual fish (Kitchell et a 105

PAGE 106

2003). As a result, abundance, recruitment, and mortality estimates may c ontain excessive bias Additionally, estimates of precisi on are likely overstated by not incorporating this important source of uncertainty. This analysis attempts to address these con cerns by incorporating uncertainty from age assignments into estimates of abundance and r ecruitment. Coggins et al. (2006 ttle state of the adult portion of the Little Color duration niz ing that the monitoring b) conducted sensitivity analyses on the effect of random ageing error and found li systematic bias in reconstructed recruitment tren ds. However, the current analysis is a more rigorous treatment of the problem and has two major implications. First, model results of estimated adult a bundance are still very precise even when uncertainty in the assignment of age is accounted for in the assessment. Following review by Kitchell et al. (2003), this assessment lends additional credibility to results from ASMR indicating that it provides a rigo rous measure of the ado River humpback chub population. I reco mmend that this assessment be considered best available science for use in contempla ting management decisions both within the GCDAMP and the U.S. Fish and Wildlife Service. Second, this analysis points out the difficulty that open population models have generally in the precise estimation of recruitment (Williams et al. 2002; Pine et al. 2003). Because many of the most critical management questions for humpback chub cente r around how best to improve humpback chub recruitment, particularly considering improved rearing conditions in the mainstem Colorado River, it will be difficult for AS MR to detect statistically significant changes in recruitment unless those changes are quite large. As a result, design of experimental adaptive management actions intended to increase recruitment should consider first and foremost how to achieve large changes in recruitment. Small scal e experimental treatments of short time or so called mini-experiments should be summ arily discounted recog 106

PAGE 107

107 ear ents should be strongly favored in order to help offset not only unexpected and uble effectsw pren in recruent program is unlikely to detect small recruitment ch ange even if it occurs. Additionally, multi-y xperime ncontrolla but the lo cisio itm estimates.

PAGE 108

Table 4-1. AIC model evaluation results among ASMR models fit to data pooled among ta cohort. g # Par n A Model AIC ameters Ra k IC ASMR1 -216274 1832492 ASMR2 -217132 302 163 8951 4 ASMR3 -218766 0 108

PAGE 109

Table 4-2. AIC model evaluati on results among ASMR models f it to data stratified by tag cohort. Model AIC # Parameters Rank AIC ASMR1 -196278 1832577 ASMR2 -197183 302 1672 ASMR3 -198856 8951 0 109

PAGE 110

A Figure 4-1. 1,200 m section of the Little Colorado River (A) and tramme l net catch rate (fish/hour/100 m) of adult humpback chub in the Little Colorado River inflow reach of the Colorado River (B). Error bars in panel (A) are 95% confidence intervals. In panel (B), the solid line represents a regr ession model fit to the subset of data representing robust sampling (solid circles) and the da shed line represents a regression model fit to the entire dataset (all circles). B Relative abundance indices of subadult (150-199 mm tota l length; TL) and adult ( 200 mm TL) humpback chub based on hoop net catch rate (fish/hour) in the lower 110

PAGE 111

A B Figure 4-2. Numbers of humpback chub marked (A) and recap tured (B) by age and year. The annual sample size is indicated by the numb er at the top of each bubble column and the distribution among ages indicated by rela tive size of bubbles within each column. 111

PAGE 112

A C B D Figure 4-3a. Numbers of fish marked by ag e in years 1989 (A), 1990 (B), 1991 (C), and 1992 (D) indicated by dark circles and subsequent ly recaptured (light circles) by age and years. The annual sample size is indicat ed by the number at the top of each bubble column and the distribution among ages i ndicated by relative size of bubbles within each column. 112

PAGE 113

A C B D Figure 4-3b. Numbers of fish marked by ag e in years 1993 (A), 1994 (B), 1995 (C), and 1996 (D) indicated by dark circles and subsequent ly recaptured (light circles) by age and years. The annual sample size is indicat ed by the number at the top of each bubble column and the distribution among ages i ndicated by relative size of bubbles within each column. 113

PAGE 114

A C B D Figure 4-3c. Numbers of fish marked by ag e in years 1997 (A), 1998 (B), 1999 (C), and 2000 (D) indicated by dark circles and subsequent ly recaptured (light circles) by age and years. The annual sample size is indicat ed by the number at the top of each bubble column and the distribution among ages i ndicated by relative size of bubbles within each column. 114

PAGE 115

Figure 4-3d. Numbers of fish marked by ag e in years 2001 (A), 2002 (B), 2003 (C), and 2004 (D) indicated by dark circles and subsequent ly recaptured (light circles) by age an years. The annual sample size is indicat ed by the number at the top of each bubb column and the distribution among ages i ndicated by relative size of bubbles withi d le n each column. D C A B 115

PAGE 116

Figure 4-3e. Numbers of fish marked by age in years 2005 (A) and 2006 (B) indicated by dark circles and subsequently recaptured (light circles) by age and years. The ann sample size is indicate ual d by the number at the top of each bubble column and the distribution among ages indica ted by relative size of bubbles within each column. B A 116

PAGE 117

Figure 4-4. Mark-recapture closed populati on model estimates of humpback chub abundance 150 mm total length in the Little Colorado River. Error bars represent 95% confidence intervals. 117

PAGE 118

A B C Figure 4-5. Humpback chub adult abundance (age -4+) estimates from the ASMR 1 (A), ASMR 2 (B), and ASMR 3 (C) models using data pooled among tag cohorts. Error bars are 95% credible intervals from 200,000 Markov Chain Monte Carlo trials. 118

PAGE 119

Humpback chub recruit abundance (a ge-2) estimates from the ASMR 1 (A), ASMR 2 (B), and ASMR 3 (C) models using data pooled among tag cohorts. Error bars a Figure 4-6. re 95% credible intervals from 200,000 Markov Chain Monte Carlo trials. C A B 119

PAGE 120

A B C D Figure 4-7. Pearson residual plots for model ASMR 1 using data pooled among tag cohorts. Individual plots are: Quantile-Quantile plots for marked (A) and recaptured (C) fish and Pearson residuals-at-age and at-time for marked (B) and recaptured (D) fish. 120

PAGE 121

A C B D Figure 4-8. Pearson residual plots for model ASMR 2 using data pooled among tag cohorts. Individual plots are: Quantile-Quantile plots for marked (A) and recaptured (C) fish and Pearson residuals-at-age and at-time for marked (B) and recaptured (D) fish. 121

PAGE 122

Figure 4-9. Pearson residual plots for model ASMR 3 using data pooled among tag cohorts. Individual plots are: Quantile-Quantile plots for marked (A) and recaptured (C) fish and Pearson residuals-at-age and at-time for marked (B) and recaptured (D) fish. B D C A 122

PAGE 123

Figure 4-10. Capture probability by age and ye ar estimated from model ASMR 3 using data pooled among tag cohorts. 123

PAGE 124

A B C Figure 4-11. Humpback chub adu lt abundance (age-4+) estimates from the ASMR 1 (A), ASMR 2 (B), and ASMR 3 (C) models using data stratified by tag cohort. Error bars are 95% credible intervals from 200,000 Markov Chain Monte Carlo trials. 124

PAGE 125

. Humpback chub recr uit abundance (age-2) estimates from the ASMR 1 (A), ASMR 2 (B), and ASMR 3 (C) models using data stratified by tag cohort. Error bars are Figure 4-12 95% credible intervals from 200,000 Markov Chain Monte Carlo trials. B C A 125

PAGE 126

A C B D Figure 4-13. Pearson residual plots for model ASMR 1 using data stratified by tag cohort. Individual plots are: Quantile-Quantile plots for marked (A) and recaptured (C) fish and Pearson residuals-at-age and at-time for marked (B) and recaptured (D) fish. 126

PAGE 127

A C B D Figure 4-14. Pearson residual plots for model ASMR 2 using data stratified by tag cohort. Individual plots are: Quantile-Quantile plots for marked (A) and recaptured (C) fish and Pearson residuals-at-age and at-time for marked (B) and recaptured (D) fish. 127

PAGE 128

Figure 4-15. Pearson residual plots for model ASMR 3 using data stratified by tag cohort. Individual plots are: Quantile-Quantile plots for marked (A) and recaptured (C) fish and Pearson residuals-at-age and at-time for marked (B) and recaptured (D) fish. C D B A 128

PAGE 129

Figure 4-16. Capture probability by age and ye ar estimated from model ASMR 3 using data stratified by tag cohort. 129

PAGE 130

Figure 4-17. Seasonal probability surfaces of age fo r a particular length bin. These surfaces su to unity in the vertical dimension (i.e., fo r each length bin) with the height of the surface indicating the probability of a particular age give n a particular length bin, laP|. Individual plots are for A m pril-June (A), July-September (B), OctoberDecember (C), and January-March (D). B C D A 130

PAGE 131

Figure 4-18. Estimated adult abundance (age-4+) from ASMR 3 incorporating uncertainty in assignment of age. Point estimates are mean values among 1,000 Monte Carlo trials and error bars represent maximum and mi nimum 95% profile confidence intervals among 1,000 Monte Carlo trials. 131

PAGE 132

Figure 4-19. Estimated recruit abundance (age-1 ) from ASMR 3 incorporating uncertainty in assignment of age. Point estimates are mean values among 1,000 Monte Carlo trials and error bars represent maximum and mi nimum 95% profile confidence intervals among 1,000 Monte Carlo trials. 132

PAGE 133

133 Figure 4-20. Retrospective analysis of adult ab undance (A) and mortalit y rate (B) considering A B datasets beginning in 1989 and ending in th e year indicated in the figure legend.

PAGE 134

CHAPTER 5 MANAGEMENT: WHAT HAVE WE LEARNED AND ARE WE MANAGING LINKING TEMPORAL PATTERNS IN FI SHERY RESOURCES WITH ADAPTIVE ADAPTIVELY? With increased public awareness of the altere d and degraded conditions prevalent in many U.S. rivers over the last two decades, river re storation activities have increased exponentially (Palmer et al. 2007). In the southwest, motivations for river restoration ac tivities are varied and include riparian zone and water quality mana gement, in-stream hab itat improvement, flow modification, and concern over federally listed en dangered species (Baron et al. 2002; Gloss et al. 2005; Follstad-Shah et al. 2007). A recent review of U.S. river restoration activities discovered that in many cases (>90%) little inform ation was available from monitoring or other activities to assess the success of these efforts (Bernhardt et al. 2005). Though this finding is troubling considering the associ ated financial expenditures (> 7.5 billion dollars between 1990 and 2003; Bernhardt et al. 2005), th e science of river re storation and associated measures of ecological success are yet in formative stages (Palmer et al. 2005). Nevertheless, this finding shows that clear project goals and monitoring sy stems to evaluate progress towards those goals are often lacking. The Glen Canyon Dam Adaptive Management Program (GCDAMP) was formed as a provision of the Record of Decision following th e Final Environmental Impact Statement on the operation of Glen Canyon Dam (USDOI 1995). Though the overarching goal of the GCDAMP could be described as assisting the U.S. Secretary of Interior to comply with the body of law governing the management of Colorado River wa ter resources and Grand Canyon National Park and Glen Canyon National Recreation area, this pr ogram has significant river restoration intent as described in the Grand Canyon Protection Act of 1992 (Act). The geographic scope of the GCDAMP is the Colorado River within Glen and Grand Canyons and associated riparian and 134

PAGE 135

terrace zones influenced ly, the Act refers to impro CDAMP viewed as a succ essful example of a large-scale adaptive mana elop nt ccording s AMP has initiated several larg e-scale adaptive management 6 and the program continues to attempt to learn as much as possib h by dam operations (GCD AMP 2001). Additional ving resources in Grand Canyon Nation Park and Glen Canyon National Recreation Area, and the majority of the GCDAMP goals refer dir ectly to aquatic or riparian resources (G 2001). The GCDAMP is generally gement and river restoration program (L adson and Argent 2002; Poff et al. 2003), but specific criteria defining resour ce goals are generally lacking in the foundational and working documents of the program (GCDAMP 2001). Although the GCDAMP has attempted to dev goals shared by all of the participating stakeholders, the result has been an unfocused set of ambiguous resource goals that are ne ither well prioritized nor orga nized within an ecosystem or guiding image perspective (NRC 1996; Palmer et al. 2005). According to the U.S. Departme of Interior technical guide on adaptive manage ment (Williams et al. 2007), a requirement of adaptive management is a statement of explicit and measurable goals. Therefore, and a to guidance from the U.S. Department of Inte rior, this program is not operating within an optimum adaptive management framework. To meet agency guidelines from Williams et al. (2007), further program development i clearly needed, but the GCD experiments since inception in 199 le from those efforts. Th is chapter provides an overview of past experimental actions, uncontrolled factors, and attempts to evaluate the effects of these actions on Colorado River fis populations. I describe trends in uncontrolled, but possibly im portant factors such as: (1) Colorado River and major tributary hydrology, (2 ) release water temperature from Glen Canyon Dam, and (3) variable production of native fish from the Little Colorado River (LCR). 135

PAGE 136

Description of Adaptive Management Actions Below I summarize the major management actions implemented by the GCDAMP since inception. These actions are pot entially biologically significant to Colorado River fish populations (Chapter 2). 1996 Experimental High Flow Prior to the construction of Glen Canyon Da m, annual spring flood discharges in the Color cfs e 1996 EHF was to the channel to build shore line sandbars, a primary program objec e nt alternatives of the U.S. Fish and Wildlife Service Biological Opinion on the he hydrograph for this treatment is thoroughly descr t 8,000 ado River often exceeded 50,000 ft3/s (cfs) with infrequent events exceeding 210,000 (Topping et al. 2003). Following construction of Glen Canyon Dam, flows rarely exceeded 30,000 cfs. For seven days in April 1996, a wide ly publicized experimental high flow (hereafter 1996 EHF) was released from Glen Canyon Dam w ith a peak discharge of 45,000 cfs (Webb et al. 1999). Though a discharge of this magnitude represents a minor pre-dam flood event with recurrence interval < 1.25 years, it represented a post-dam event with recurrence interval of 5.1 years (Schmidt et al. 2001). The overarc hing restoration goal of th redistribute fine sediment with in tive. However, a broad set of physical an d biological studies were undertaken in concert with the 1996 EHF and are thoroughly documente d in the literature (e.g., Webb et al. 1999; Patten and Stevens 1999). 2000 Low Summer Steady Flow A program of experimental flows (hereafter LSSF) from Glen Canyon Dam (Valdez et al. 2000) was initiated in 2000 to be nefit native fish resources and to comply with the reasonabl and prude operation of Glen Canyon Dam (USFWS 1994). T ibed elsewhere (Trammell et al. 2002), but consisted primarily of a period of constan cfs discharges from June 1, 2000 to September 30, 2000 with the intent of promoting improved 136

PAGE 137

native fish rearing conditions th rough increased temperature and hydraulically stable near-shore environments. Preceding this prolonged constant 8,000 cfs discharge, there was a 52-day of high (17,000 to 30,000 cfs) and generally cons tant discharge to mimic spring flooding dur April and May. During the constant 8,000 cfs discharge in September, th ere were several spi flows to 16,000 and 30,000 cfs for up to four days with the intent of disadvantaging non-n fish. Sampling during the LSSF indicated hi gher abundance of speckled dace Rhinichthys osculus, flannelmouth sucker Catostomus latipinnis, bluehead sucker Catostomus discobolus and fathead minnow Pimephales promelas in backwater areas, particularly in the lower portions of Grand Canyon (Trammel et al. 2002). These re sults suggested some benefit to warm water fishes associated with LSSF. Unfortunately, few juvenile hu block ing ke ative mpback chub were captured during ar fish in the LCR during 2000 (Dennis Stone er urs; is effort are still ongoing. 2003-of increased flow fluctuations (termed NNFSF, non-native fish suppression flows) from Glen Canyon Dam was LSSF, perhaps related to low production of young of ye U.S. Fish and Wildlife Se rvice, Personal Communication). 2004 Experimental High Flow Following large sediment inputs to the Colorado River from the Paria River in late summ and fall of 2004, a second experimental high flow was implemented from Glen Canyon Dam (hereafter 2004 EHF; Wri ght et al. 2005). In contrast to the 1996 EHF, the 2004 EHF occurred during November and with sligh tly lower discharge (41,000 cfs) and shorter duration (60 ho Topping et al. 2006). As with the 1996 EHF, th is management action was primarily focused on restoring shoreline sandbars. Mu ch of the data analysis and re porting for th2005 Non-native Fish Suppression Flows Coupled with removal of non-native fish (s ee below), a program 137

PAGE 138

implemented during January-March, 2003-2005 to suppress rainbow trout Oncorhynchus mykiss on a s cing ible ty en and August 2006. Concurrent w ith these removals, the fish community comp ut ors ystemic basis by disrupting spawning and rearing. Korman et al. (2005) conducted extensive studies on the spawni ng and rearing success of rainbow trout during the NNFSF. These studies also investigated the extent of spawning in the Colorado River and tributaries between River Mile 0 (RM 0) at Lees Ferry and RM 62 near the confluence of the LCR. Korman et al. (2005) concluded that these flows were likely ineffective at substantially redu rainbow trout recruitment because of sub-optimal timing (i.e., too early-in the year) and poss compensatory survival at older ages. Based on surveys of available spawning habitat and densi of young of year rainbow trout between RM 0 and RM 62, they further concluded that NNFSF likely did not affect rainbow trout reproduction downstream of RM 0 since there was little to no evidence of reproductive activity in this reach during 2004. 2003-2006 Mechanical Removal of Non-native Fish This effort removed over 23,000 non-native fish between RM 56.3 and 65.7 betwe January 2003 osition within this reach shifted from one being numerically dominated by rainbow tro (>90%) in 2003, to one dominated by native fi shes and the non-native fathead minnow (>90%) in 2006. Though trends in the abundance of rain bow trout in both the re moval and the control reaches imply a systemic decline in rainbow trout unrelated to removal efforts, mechanical removal also contributed to the shift in community composition. Motivation, methodology, and results of the non-native fish removal program are presented in Chapter 2. Description of Uncontrolled FactBelow I summarize some of the uncontrolled factors influencing both fish population dynamics and the ability to detect potential fish population responses. 138

PAGE 139

Paria, Little Colorado, and Colorado River Hydrology Variability in the Paria, Little Colorado, and Colorado Rivers hydrology all potentiall affect fish populations in the Colorado River. Seasonal and episodic freshets in the Paria Riv are of much smaller discharge than the Colorado River (Figure 5-1), yet in the post GCD sy the Paria River is the largest source of fine sedi ment (Wright et al. 2005). As a result, flooding in the Paria River is a domina nt driver of turbidity in dow nstream portions of the Colorado River. Because turbidity decreases primary and secondary production (Kennedy and Gloss 2005), the Paria River is a major factor struc y er stem, t uring the downstream aquatic community both in foraging efficien cy for sight feeders such ing f ; Figure sults d by storage capacity in Lake Powe terms of food resources (Carothers and Brown 1991) and as rainbow trout (Barrett et al. 1992). Examination of the Paria River hydrograph (Figure 5-1) demonstrates the flashy character of th is system and reveals major fall flooding dur 1997-2000, 2004, and 2006. Conversely, 1995-1997 and 2001-2003 were periods of relatively less flooding and with corresponding lower tu rbidity downstream from the Paria River confluence. The LCR is also a major source of fine sediment to the system with a corresponding influence on turbidity. In contrast to the Paria River, large floods in the LCR are frequently o equal or greater discharge than the post-dam Colorado River (e.g., 1992 and 2002 events 5-1). These events not only increase turbidity, they also transport large numbers of native and non-native fish to the Colorado River (Valdez and Ryel 1995; Stone et al. 2007). In turn, re of fish sampling (e.g., species and length co mposition and catch rate ) downstream of the confluence can be highly influenced by recent LCR hydrology. Annual release volume from Glen Canyon Dam is influence ll, annual inflow volume to Lake Powell, and a set of laws and regulations aggregately termed The Law of the River. During periods of high storage capacity and low annual inflow, 139

PAGE 140

relevant policies guarantee a minimum annual rel ease volume of 8.23 million acre feet (m water. During periods of lower reservoir stor age and larger annual inflows, annual release volumes may exceed this amount. Since 1990, annual release volume exceeded 8.23 maf in 1995-2000 (Figure 5-1). Though the relationship between annual release volume and fish population dynamics remains unclear, Paukert a nd Rogers (2004) found a positive relationship between annual release volume and flannelmout h sucker condition factor. These authors hypothesized that greater annual volumes provide increased e uphotic volume and therefore greater primary and se condary production. af) of Relear comp nd is influenced by a host of factors controlling egg, larval, and juvenile survival. Dominant factors likel y include: hydrology, temp erature, food resources, se Water Temperature from Glen Canyon Dam Water temperature released from Glen Canyon Dam is largely dependent on the reservoi elevation in relation to the dam penstock depth. By 1973, the last vestige of the pre-dam thermograph disappeared as the reservoir filled to an elevation promoting annual hypolimnetic releases of between 7 and 12 C (Vernieu et al. 2005). The re servoir water levels fell during a drought in 2000-2005 prompting par tial epilimnetic releases a nd the warmest release water temperatures since the before the reservoir fille d. These warmer water releases, coupled with further downstream warming during the summer months, resulted in significantly increased water temperature in the Colorado River near the LCR conflu ence during 2003-2006 as ared to 1990-2002 (Figure 5-2). Though wa ter temperatures during 2003-2006 were still below pre-dam values, they are much closer to those required for successful spawning a rearing by warm-water adapted native fishes (Hamman 1982; Valdez and Ryel 1995), and should be conducive to increased growth of humpback chub Gila cypha (Chapter 3). Juvenile Native Fish Production in the Little Colorado River Native fish production in the LCR 140

PAGE 141

p increase the abundance of young fish in the mainst em downstream of the confluence, particularly associated with late summer rainstorms (Valdez and Ryel 1995). Valdez and Ryel (1995) suggest that survival rates of these young native fish attempting to rear in the mainstem is extremely low given post-dam conditions. It is, therefore, easy to conceive how LCR freshets may affect year class strength if a significant po rtion of a cohort must attempt to rear in the mainstem. Conversely, that portion of the cohort remaining in the LCR may experience a compensatory increase in survival and ultimately contribute significantly to recruitment. explain the very large number of young of year humpback chub observed from the 1993 brood ore, mainstem during July 1993 base flow conditions. Clearly, it is difficult to predict how LCR Another factor potentiall y influencing native fish production in the LCR is infection by the appear to be particularly su sceptible to infection (Choudhury et al. 2004). Hoffnagle et al. suggested sub-lethal and lethal effects. arasite infestation, and pred ation risk. As mentioned above freshets in the LCR tend to Valdez and Ryel (1995) hypothesized that flooding activity in the LCR tends to disadvantage non-native fish and cleanse spawning gravels. This is the argument provided to year (Valdez and Ryel 1995) following extensive flooding in late 1992-early 1993. Furtherm these authors hypothesized that the 1993 brood year consisted of such large numbers of young humpback chub that LCR resources were insuffici ent for juvenile fish rearing leading to reduced growth rates and condition, and prompti ng large numbers of fish to emigrate to the hydrology may affect native fish production and recruitment. non-native Asian Tapeworm Bothriocephalus acheilognathi and the copepod Lernaea cyprinacea These parasites are presen t in the LCR and both humpback chub and speckled dace (2006) demonstrated diminished condition of hum pback chub infested with these parasites and 141

PAGE 142

How Are Fish Populations Affected by Adap tive Management Actions and Uncontrolled Factors? The GCDAMP is fortunate to have long-term time series measur es of the relative abundance of native and non-native fish as indexed by electrofishing catch ra te in the mechanical removal reach (Figures 5-3 and 5-4), and by hoop net catch rate in the mainstem downstream from the LCR confluence (RM 63.7-64.2; Figure 5-5). These measures, along with average size (Figures 5-6 through 5-8), provide relatively good information to infer changes in population demography over time. Has Increased Turbidity Affected Fi sh Populations in Grand Canyon? 2003) potentially limiting food resources for some fishes. However, increased tributary dominant food item for fish species downstream of the Paria River and LCR (Kennedy and Gloss repres ch ng Turbidity is hypothesized to affect fish populat ions via a suite of direct and indirect mechanisms. Decreased water clarity should generally result in reduced autochthony (Yard discharge may also result in increased detritus available for consumption by simuliids a 2005). Turbidity may also affect foraging efficiency particularly for rainbow trout (Barrett et al. 1992). Thus, turbidity may mediate negative inte ractions between non-native and native fish (Gradall and Swenson 1982; Gregory and Levi ngs 1998; Johnson and Hines 1999) and possibly ent a mortality source fo r rainbow trout (Chapter 2). With these hypotheses in mind and given mini mal flooding activity from the Paria River during 1995-1997 and 2001-2003, I would predict th at the abundance of ra inbow trout in the mechanical removal reach should be highest duri ng these time periods. Examination of the cat rate data indicates that ra inbow trout abundance was high during 2001-2003, but not during 1995-1997 (Figure 5-3). Similarly, if native fish are advantaged by high turbidity conditions, their abundance should be lower during these time periods. Ava ilable data indicate no stro 142

PAGE 143

trends in native fish abundance in these time periods as indexed by either electrofishing or hoop net catch rate (Figure 4-5). Humpback chub recr uitment estimates from the age-structured markrecap re ationship sting a positive influence from turbidity, an equal ive Populations in Grand Canyon? The GCDAMP initiated the mechanical removal of non-native fish program (Chapter 2) in response to the increases in the a bundance of non-native rainbow and brown Salmo trutta trout near the confluence of the LCR during 1996-2002 (F igure 5-3). Implicit in the action was the hypothesis that native fish are negatively impact ed by non-native fish. More specifically, that juvenile native fish survival was negatively impacted by non-native fish via predatory and competitory interactions. If this hypothesis is correct, the abundance of juvenile native fish in the mechanical removal reach would increase after non-native removal. Additionally, the ture model (Chapter 4) suggest an incr easing recruitment trend beginning with the 1995 brood year. Though turbidity may affect rainbow trout abundance in downstream reaches, the are clearly other controlling mechanisms such as immigration from the Lees Ferry reach (Chapter 2; Korman et al. 2005). Additionally, th ere does not appear to be a simple rel between turbidity and native fish populations. The preceding discussion illustrates the challe nges involved in evaluating the impact of adaptive management experiments and uncontrolled factors on fish populations in Grand Canyon. For nearly every credible hypothesis sugge ly credible contradictory hypothesis can be ge nerated. Additionally, co rrelative analyses to assess various a posteriori hypotheses of cause and effect rela tionships provide weak inference at best (Yoccoz et al. 2001; MacKenzie et al. 2007). This is particularly true with regard to the highly variable and uncontrolled factors (e.g., tributary hydrology) and short duration adapt management experiments (i.e., 1996 and 2004 EHF and LSSF). Has Reduced Non-native Fish Abundance and In creased Temperature Affected Native Fish 143

PAGE 144

removal effort was predicted to promote increased humpback chub recruitment associated with the 2003-2006 brood years. Examination of electrofishing catch rate in th e mechanical removal r each suggest an order of ma e res 5h n, and rainbow trout predation on juvenile humpback chub. Their result ould realize a 287% greater annua gnitude increase in the re lative abundance of flannelmouth sucker and bluehead sucker between 2003 and 2006 (Figure 5-4). Increases in humpback chub and speckled dace are also apparent, though not as large. Hoop net ca tch rates of humpback chub also increased, particularly in late 2004-2006 (Figure 5-5). A dditionally, the re lative abundance of non-nativ fathead minnow increased markedly in 2005 and 2006 (Figure 5-3). Though these responses are consistent with predictions, it is not clear that these responses were cause d in whole or part by diminished abundance of rainbow and brown trout. The mechanical removal of non-native fish coincided with the warmest water temperatu in the mainstem Colorado River near the LCR confluence observed since before 1990 (Figure 2). Since there is near perfect temporal co rrelation between these f actors both hypothesized to control native fish survival, it is difficult to evaluate their effects separately. Indeed, these factors acting in combination likel y have a multiplicative effect on juvenile fish survival throug increased growth of native fish, increased food resources, and decreased pr edation risk (Paukert and Petersen 2007). Paukert and Petersen (2007) used bioenergetics models to evaluate th e effects of increased water temperature on the growth, food consumptio s indicated that an 8g humpback chub c l increase in mass under the 2005 thermograph as compared to mean 1993-2002 water temperatures. They further predicted that the 2005 thermograph would reduce the time juvenile chub were vulnerable to predation by 2-3 months Paukert and Petersen (2007) also predicted 144

PAGE 145

that the rainbow trout population in the mechani cal removal reach annually consumed more tha 18,000 kg of invertebrates. If food availability is a limiting factor for na tive fishes in Grand Canyon, liberation of these resources may have si gnificantly increased their g n rowth and survival. h in e LCR 007) re 5-8). However, as described above, interp e fish in this reach, but are also consistent with predictions of increased water temperature. Thus, The average length of native fish captured wi th electrofishing provides some evidence of increased growth and survival consistent with the predictions of Paukert and Petersen (2007). The declining trend in bo th flannelmouth sucker and humpback chub TL is a result of a larger fraction of smaller fish in the electrofishi ng catch during 2005 and 2006 than previous years (Figure 5-7). This is particularly evident fo r flannelmouth sucker as the 2006 length frequency distribution displays multiple modes consistent with several successful year classes (Figure 5-9). The increasing trend in speckled dace TL is consis tent with the hypothesis of increased growt 2005 and 2006 (Figure 5-7). The sudden appear ance of adult bluehead sucker in 2002-2006 electrofishing samples is puzzling and is most consistent with immigration of adults from the LCR (Figures 5-4 and 5-7). Sampling by U.S. Fish and Wildlife Servi ce personnel in th has indicated large increases in bluehead su ckers during 2006 (Van Haverbeke and Stone 2 in support of this hypothesis. Increased size of humpback chub captured in hoop nets in 20052006 also supports the hypothesis of increased growth and survival, as this gear seldom captures adult fish in the mainstem Colorado River (F igu reting patterns of relative abundance or length composition of juvenile fish is complicated by LCR hydrology and associ ated fish migration. Conclusions and Recommendations Index data from electrofishing and hoop net sampling in the mechanical removal reach indicate an increase in native fi sh abundance in 2005-2006, particul arly for the sucker species. These changes in abundance are consistent with th e predicted response to removal of non-nativ 145

PAGE 146

it is not possible to determine which factor is mo st responsible for the changes. Nevertheless, from an adaptive management perspective, there is good evidence that a successful policy with respe other was t ained these two factors, but addi tional evidence might accumulate supporting the succe mary arret nefit could ct to native fish conservation may i nvolve at least one of these factors. A strategy that could be employed to determine th e relative influence of these factors is to manipulate the system so that one of the factors remained at the present state while the varied. Perhaps the most tenable option to ach ieve this circumstance would be for the GCDAMP to continue non-native fish control and wait fo r wetter hydrology to rais e reservoir levels and decrease release temperatures (though such hydrology may not soon occur according to Seager e al. 2007). If under these conditions native fish abundance and subsequent recruitment rem high, the conclusion would be that interactions w ith native fish were the dominant factor. In contrast, if native fish resources suffered, the conclusion would be that water temperature was the dominant factor. This latter conclusion w ould lend support to constr uction of a selective withdrawal structure on Glen Canyon Dam in order to provide control of release water temperature. If warmer release water conditions persist, there would be limited ability to discriminate among ss of the combined factor policy. This woul d also have the benefit of achieving the pri program objective of native fish conservation. However, two GCDAMP review pa nels have concluded that the risk to native fish may be compounded by warmer water releases by favoring nonnative fishes (Mueller et al. 1999; G et al. 2003). This advice might be interpreted to be wary of transitory behavior of native fish populations following onset of warmer water conditions Initial signs of native fish be be short lived as the non-native fish community shifted to favor warm-wat er species potentially capable of even more detrimental interactions wi th native fishes than the present assemblage. 146

PAGE 147

Additionally, such a shift in fish community composition might not be easily reversible in a short period of time, particularly with limite d control of release water temperature. The ultimate success of a native fish conservation policy should be evaluated in increased native fish recruitment. A policy that in creased native fish survival for some life stage, light of but fa r m the LCR into the ma instem and the abundance of native fish in the mains s of iled to promote recruitment to adult hood, could not be deemed successful. This observation highlights the importance of the humpb ack chub stock assessment program (Chapte 4). However, evaluating policy success solely on recruitment estimates may not allow critical insights into policy performance since recruitment to adulthood is the integration of conditions faced by a cohort for multiple years. Differential recruitment from both the LCR and the mainstem Colorado River is a further complicati on. Consider a situation where a management action is undertaken in the mainstem concurre nt with uncontrolled fact ors benefiting rearing conditions in the LCR. If recruitment for t hose brood years is determined largely by the LCR and the management activity in the mainstem is ineffective, evaluation of the recruitment time series might erroneously lead managers to believe that activities in the mainstem represented a successful policy. Because of these complications, I contend that a more rigorous mainstem monitoring program to estimate growth and survival of native fish is needed in order to fully evaluate adaptive management experiments. Such a pr ogram would need to m onitor immigration of juvenile native fish fro tem near the LCR confluence. These data could then be used in a balance framework to estimate time specific apparent survival rate Korman et al. (2005) have successfully implemented a similar framework to study the effect of dam operations on juvenile rainbow trout in the Lees Ferry reach. Though the Lees Ferry program benefits from having high densitie 147

PAGE 148

fish and is not confounded by tributary hydrology, it might be possible to modify their approach and implement it to study native fish survival and growth dynamics in the mainstem Color River. This information, coupled with the ongoing stock assessment program, would significantly increase the ability of the GCDAMP to evaluate adaptive management experiments. Williams et al. (2007) argue that critical elements of an adaptive management program require specification of explicit and measurable program goals a nd the ability to predict po performance ado licy relative to those goa ls. The GCDAMP would benefit fr om additional specificity in program goals so that it is clear if and when a policy could be judged successful. As an example, the native fish index data suggests recent im proved conditions for native fish, but without specific goals the program cannot judge whether the policy should be pronounced successful. The program could also benefit from additional predictive capability both to screen policy options and to formalize and eval uate alternative hypothe ses of system behavior. As discussed by Anderson et al. (2006), predic tive models of fish population dynamics in riverine systems should recognize and model the dynamic feedbacks among various trophic components and the forcing factors mediating such interactions. Ap parently without recognizing it, these authors have described many of the elements contained in the Ecopath/Ecosim modeling framework (Christensen and Walters 2003). I recommend that the GCDAMP rely more heavily on these types of quantitative approaches to predict native fish popul ation dynamics under alternative management policies. With improved monitoring and predictive capabilities to evaluate explicit and measurable goals, I believe the GCDAMP will be well placed to better conceive and implement future adaptive management experiments. Such pla nning should recognize the uncontrolled variability in the system and select experimental design options to most effectively alleviate potential 148

PAGE 149

149 confounding factors, while attempting to implement experimental policies most likely to achieve program goals. I contend that one of the biggest obstacles in this effort is the inability to implement experiments of sufficient duration and magnitude to provide measurable results. The mechanical removal effort combined with multi-year temperature modification has the potential to elicit effective learning and improved resource condition. The ultimate success of this effort will depend heavily on the commitment to s ubsequent well planned experimentation and monitoring.

PAGE 150

Discharge in ft3/s (cfs) for the Colorado River at Lees Ferry (A), the Paria River at Lees Ferry (B), and the Little Colora do Figure 5-1. River at Cameron, AZ (C), 1990-2006. B C A 150

PAGE 151

Figure 5-2. Daily mean water temperatures obs erved in the Colorado Ri ver at approximately river mile 61, 1990-2006. Lines indicate lo cally weighted polynomial regressions (Lowess) fits to the indicated data set. 151

PAGE 152

Figure 5-3. Monthly electrofis hing catch rate (fish/hour) in th e Colorado River between river mile (RM) 56.3 and RM 65.7 for rainbow trout (A), brown trout (B), common ca (C), and fathead minnow (D). Error ba rs represent 95% confidence intervals. rp 152

PAGE 153

Figure 5-4. Monthly electrofis hing catch rate (fish/hour) in th e Colorado River between river mile (RM) 56.3 and RM 65.7 for humpback chub (A), flannelmouth sucker (B), bluehead sucker (C), and speckled dace (D ). Error bars represent 95% confidence intervals. 153

PAGE 154

Figure 5-5. Monthly hoop net ca tch rate (fish/hour) in the Colorado River between river mile (RM) 63.7 and RM 64.2 for humpback chub (A ), flannelmouth sucker (B), bluehea sucker (C), and speckled dace (D). Error bars represen t 95% confidence intervals. d 154

PAGE 155

Figure 5-6. Monthly average to tal length (TL; mm) observed in electrofishing sampling in the Colorado River between river mile (RM) 56.3 and RM 65.7 for rainbow trout (A), brown trout (B), common carp (C), and fa thead minnow (D). Error bars represent 95% confidence intervals. 155

PAGE 156

Figure 5-7. Monthly average to tal length (TL; mm) observed in electrofishing sampling in Colorado River between river mile (RM) 56.3 and RM 65.7 for humpback chub (A), flannelmouth sucker (B), bluehead sucker (C), and speckled dace (D). Error bars represent 95% confidence intervals. the 156

PAGE 157

Figure 5-8. Monthly average total length (TL; mm) observed in hoop net sampling in the Colorado River between river mile (RM) 63.7 and RM 64.2 for humpback chub (A), flannelmouth sucker (B), bluehead sucker (C), and speckled dace (D). Error bars represent 95% confidence intervals. 157

PAGE 158

158 Figure 5-9. Smoothed kernel density plot of th e total length of flannelmouth sucker captured with electrofishing in the mechanical re moval reach during 2003 and 2006. Numbers in the legend represent total number (n) of fish captured.

PAGE 159

LIST OF REFERENCES Akaike, H. 1973. Information theory and an ex tension of the maximum likelihood principle. on Allan, C. and A. Curtis. 2005. Nipped in the bud: why regional scale adaptive management is not Andersen, K. E., A. J. Paul, E. McCauley, L. J. Jackson, J. R. Post, and R. M. Nisbet. 2006. Instream flow needs in streams and rivers : the importance of unde rstanding ecological Arreguin-Sanchez, F. 1996. Catchability: a key parameter for fish stock assessment. Reviews in Fish Biology and Fisheries 6: 221-242. Askey inking angling catch rates and fish learning under catch-and-release regulat ions. North American Journal of Fisheries Baillargeon, S. and L. P. Rivest. 2007. Rcapture: loglinear models for capture-recapture in R. Baron, J. S. and 9 coauthors. 2002. Meeting ecological and societal needs for freshwater. Ecological Applicat ions 12:1247-1260. Barre ainbow trout. Tr ansactions of the American Fisheries Society 121:437-443. ic ). Bigel iff. 2003. Protection e National Park Blinn e American Fisheries Society 122:139-143. Pages 267 in B. N. Petrov and F. Csaki, editors Second International Symposium Information Theory. Akademiai Kiado, Budapest, Hungary. blooming. Environmental Management 36:414-425. dynamics. Frontiers in Ecology and the Environment 4:309-318. P. J., S. A. Richards, J. R. Post, and E. A. Parkinson. 2006. L Management 26:1020-1029. Journal of Statistical Software 19:1-31. tt, J. C., G. D. Grossman, and J. Rosenf eld. 1992. Turbidity-induced changes in reactive distance of r Benke, A. E. and C. E. Cushing, editors. 2005. Rivers of North America. Elsevier Academ Press, Burlington, MA. Bernhardt, E. S. and 24 coauthors. 2005. Synthesizing U.S. River Restoration Efforts. Science 308:636-637 Bertalanffy, L. von. 1938. A quantitative theory of organic growth (Inquiries on growth laws II Human Biology 10:181-213. Beverton, R. J. H. 1992. Patterns of reproductive strategy parameters in some marine teleost fishes. Journal of Fish Biology 41:137-160. ow, P. E., T. M. Koel, D. Mahony, B. Erte l, B. Rowdon, and S. T. Oll of native Yellowstone cutthroat trout in Ye llowstone Lake, Yellowston Wyoming. National Park Service Tec hnical Report NPS/NRWRD/NRTR-2003/314, Yellowstone National Park, Wyoming. D. W., C. Runck, D. A. Clark, and J. N. Rinne. 1993. Effect s of rainbow trout predation of Little Colorado spinedace. Transactions of th 159

PAGE 160

Bozdogan, H. 2000. Akaikes information criteri on and recent developments in information complexity. Journal of Math ematical Psychology 44:62-91. Britton, J. R., J. Pegg, R. Sedgwick, and R. Page. 2007. Investigating the catch returns and Brownie, C., D. R. Andersen, K. P. Burnham, an d D. R. Robson. 1985. Statistical inference from Carothers, S. W. and Brown, B. T. 1991. Th e Colorado River through Grand Canyon. University Choudhury, A., T. L. Hoffnagle, and R. A. Cole 2004. Parasites of native and nonnative fishes of the Little Colorado River, Grand Ca nyon, Arizona. Journal of Parasitology 90:1042Christensen, N. L and 12 coauthors. 1996. The report of the Ecological Society of America committee on the scientific basis for ecosyst em management. Ecological Applications Christensen, V. and C. J. Walters. 2003. Ec opath with Ecosim: methods, capabilities, and limitations. Ecological Modelling 172:109-139. Clark rature in teleost fish. Journal of Animal Ecology 68:893-905. Clark iver fishes. Copeia 2:402-412. ve vailable: http://www.gcmrc.gov/library/reports/ biological/Fish_studies /GCMRC/Coggins2002.pdf (January 2008). Coggins, L. G., Jr. and T. J. Quinn II. 1998. A si mulation study of the effects of aging error and sample size on sustained yi eld estimates. Pages 955-975 in F. Funk, T. J. Quinn II, J. itors. growth rate of wells catfish, Silurus glanis, using mark-recapture. Fisheries Management and Ecology 14:263-268. band recovery data a handbook, 2nd Edition. U.S. Fish and Wildlife Service Resource Publication 156. Burnham, K. P. and D. R. Andersen. 2002. Model selection and multimodel inference. A practical information-theoretic approach, s econd edition. Springer Science + Business Media, New York. Campbell, S. P. and 7 coauthors. 2002. An assessment of monitoring efforts in endangered species recovery goals. Ecol ogical Applicat ions 12:674-681. of Arizona Press, Tucson, AZ. 1053. 6:665-691. e, A. and N. M. Johnston. 1999. Scaling of metabolic rate with body mass and tempe son, R. W. and M. R. Childs. 2000. Temp erature effects of hypo limnial-release dams on early life stages of Colorado River big-r Coggins, L. G. Jr., M. D. Yard, and C. Paukert 2002. Piscivory by non-native salmonids in the Colorado River and an evaluation of the e fficacy of mechanical removal on non-nati salmonids: an operational plan. A Heifetz, J. N. Ianelli, J. E. Powers, J. F. Schweigert, P. J. Sullivan, and C. I. Zhang, ed Fishery Stock Assessment Models. Alaska Sea Grant College Program Report No. AK SG-98-01, University of Alaska Fairbanks 160

PAGE 161

Coggins, L. G., Jr., W. E. Pine, III, C. J. Walte rs, D. R. Van Haverbeke, D. Ward, and H. C Johnstone. 2006a. Abundance trends and status of the Little Colo rado River population of humpback chub. North American Journa l of Fisheries Management 26:233-245. Cogg red markrecapture analysis: a virtual-population-analysis -based model for analyzing age-structured Collie n. Crawford, S. S. and A. M. Muir. In Press. Global introductions of salm on and trout in the genus Oncorhynchus : 1870-2007. Reviews in Fish Biology and Fisheries. Available: www. ) DeVr R. and D. Willis, editors. Fisheries techniques, 2nd edition. American Fisheries Society, Bethesda, Maryland. Donla ie, V. Carrion, and F. Cruz. 2007. Recovery of the Galapagos rail (Laterallus spilonotus ) following the removal of invasive mammals. Dorazio, R. M., H. L. Jelks, and F. Jordan. 2005. Improving removal-based estimates of abundance by sampling a population of spatia lly distinct subpopulations. Biometrics Douglas, M. E., and P. C. Marsh. 1996. Popul ation estimates/populati on movements of Gila Dunbrack, R., J. M. Green, and C. M. Mlewa. 2006. Marbled lungfish gr owth rates in Lake Erski l Essin The von Bertalanffy growth function, Fabens, A. J. 1965. Properties and fitting of 29:265Follstad-Shah, J. J., C. N. Dahm, S. P. Gloss, and E. S. Bernhardt. 2007. River and riparian restoration in the southwest: results of the national river restoration science synthesis project. Restoration Ecology 15:550-562. ins, L. G., Jr., W. E. Pine, III, C. J. Walte rs, S. J. D. Martell. 2006b. Age-structu capture-recapture data. North American Jour nal of Fisheries Management 26:201-205. r, M. P., R. H. Webb, and E. D. Andr ews. 1997. Experimental flooding in Grand Canyo Scientific American 276:82-89. springerlink.com.lp.hscl.ufl.edu/cont ent/m447581887406t4n/fulltext.pdf (January 2008 ies, D. R., and R. V. Frie. 1996. De termination of age and growth. Pages 483 in B. Murphy n, C. J., K. Campbell, W. Cabrera, C. La vo Biological Conservation 138:520-524. 61:1093-1101. cypha, an endangered cyprinid fish in the Grand Canyon region of Arizona. Copeia 1996:15. Baringo, Kenya, estimated by mark-recapture. Journal of Fish Biology 68:443-449. n Ogden, J. A. and M. Rejmanek. 2005. Rec overy of native plant communities after contro of a dominant invasive plant species, Foeniculum vulgare : implications for management. Biological Conservation 125:427-439. gton, T. E., J. F. Kitchell, and C. J. Walters. 2001 bioenergetics, and the consumption rates of fish. Canadian Journal of Fisheries and Aquatic Sciences 58:2129-2138. the von Bertalanffy growth curve. Growth 289. 161

PAGE 162

Fournier, D. 2000. AD Model Builder. Otter Re search, Sidney, British Columbia, Canada. Available: http://otter-rsch.com/admodel.html (January 2008). Francis, R. I. C. C. 1988. Are growth parame ters estimated from tagging and age-length data Garrett, D. and 12 coauthors. 2003. Evaluating a Glen Canyon Dam temperature control device e ng and Research Center, Flagstaff, Arizona. Available: http://www.gcmrc.gov/library/re ports/synthesis/Garrett2003.pdf (April GCD Adaptive Management Program). 2001. Strategic Plan: Glen Canyon Dam Adaptive Management Program, Final Draft. Bureau of Reclamation, Salt Gelman, A., J. B. Carlin, H. S. Stern, and D. B. Rubin. 2004. Bayesian data analysis, 2nd edition. Gloss nd Canyon. Pages 33-56 in S. P. Gloss, J. L. Lovich, and T. S. Melis, editors The state of the Colorado River ecosystem in Grand Gloss, S. P., J. E. Lovich, and T. E. Melis, editors. 2005. The state of the Colorado River Gorman, O. T., and D. M. Stone. 1999. Ecology of spawning humpback chub ( Gila cypha ) in the of native and non-native fishes of the Colorado River Gradall, K. S. and W. A. Swenson. 1982. Respons es of brook trout and cr eek chubs to turbidity. Grego comparable? Canadian Journal of Fish eries and Aquatic Sciences 45:936-942. to enhance native fish habitat in the Co lorado River: a risk assessment by Adaptive Management Program Science Advisors. Prepared for the Glen Canyon Dam Adaptiv Management Program and the Grand Canyon Mon itori 2008) AMP (Glen Canyon Dam Lake City, UT. Available: http://www.usbr .gov/uc/rm/amp/pdfs/sp_final.pdf. (January 2008). Chapman & Hall/CRC Press, New York S. P. and L. G. Coggins, Jr. 2005. Fishes of Gra Canyon. U.S. Geological Survey Circular 1282. ecosystem in Grand Canyon. U.S. Geological Survey Circular 1282. Little Colorado River near Grand Canyon, Ar izona. Environmental Biology of Fishes 55:115. Gorman, O. T., R. G. Bramblett, R. M. Hervin D. R. Van Haverbeke, and D. M. Stone. 2005. Distribution and abundance Ecosystem of Grand Canyon, Arizona. Pages 78-94 in M. J. Brouder, C. L. Springer, and S. C. Leon, editors. The Lower Colorado River: Restoring natural function and native fish within a modified riverine environment. Proceedings of two symposia and workshops, Las Vegas, Nevada, July 8-9, 1998 and July 13-14, 1999. U.S. Fish and Wildlife Service, Southwest Region, Albuquerque, New Mexico. Transactions of the American Fisheries Society 111:392-395. ry, R. S. and C. D. Levings. 1998. Turb idity reduces predation on migrating juvenile pacific salmon. Transactions of the American Fisheries Society 127:275-285. 162

PAGE 163

Gunderson, L. 1999. Resilience, flexibility and adaptive management antidotes for spuri certitude? Conservation Ecol ogy 3(1): 7. Available: http://www.consecol.org/vol3/ iss1/art7/ (January 2008). ous Hein, C. L., B. M. Roth, A. R. Ives, and M. J. Vander Zanden. 2006. Fish predation and trapping Hein, C. L., M. J. Vander Zanden, and J. J. Magnuson. 2007. Intensive trapping and increased Hoffm L., G. L. Larson, and B. Samora. 2004. Responses of Ambystoma gracile to the Hoffnagle, T. L., A. Choudhury, and R. A. Cole. 2006. Parisitism and body condition in Hollin Hood ware for the Analysis of Ecological Models. Version 2.5.5. James, I. R. 1991. Estimation of von Bertalanffy growth curve parameters from recapture data. Jense val. Canadian Journal of Fisheries and Aquatic Science 53:820-822. ican Fisheries Society 128:648Kenn -101 in S. P. Gloss, J. L. Lovich, and T. S. Melis, editors The state 1282. Kimu Hamman, R. L. 1982. Spawning and culture of humpback chub. Progressive Fish Culturist 44:213. for rusty crayfish (Orconectes rusticus) control: a whole-lake experiment. Canadian Journal of Fisheries and Aquatic Science 63:383-393. fish predation cause massive population declin e of an invasive crayfish. Freshwater Biology 52:1134-1146. an, R. removal of introduced nonnative fish from a mountain lake. J ournal of Herpetology 38:578-585. humpback chub from the Colorado and Little Colorado Rivers, Grand Canyon, Arizona Journal of Aquatic Animal Health 18:184-193. g, C. S. 1973. Resilience and stability of ecological systems. A nnual Review of Ecology and Systematics 4:1-23. Holling, C. S., editor. 1978. Adaptive environm ental assessment and management. John Wiley, London. G. M. 2000. Poptools Soft Available: http://www.cse.csiro.au/poptools (January 2008). Biometrics 47:1519. n, A. L. 1996. Beverton and Holt life histor y invariants result from optimal trade-off of reproduction and survi Johnson, J. E. and R. T. Hines. 1999. Effect of suspended sediment on vulnerability of young razorback suckers to predati on. Transactions of the Amer 655. edy, T. A. and S. P. Gloss. 2005. Aqua tic ecology: the role of organic matter and invertebrates. Pages 87 of the Colorado River ecosystem in Grand Canyon. U.S. Geological Survey Circular ra, D. K. and S. Chikuni. 1987. Mixtures of Empirical Distributions: An iterative application of the age-length key. Biometrics 43:23-35. 163

PAGE 164

Kirkwood, G. P. and I. F. Somers. 1984. Growth of two species of Tiger Prawn, Penaeus esculentus and P. semisulcatus, in the Western Gulf of Carpentaria. Australian Journal of Marine and Freshwater Research 35:703-712. nd k chub Gila cypha. Final report submitted to the Grand Canyon Monitoring and Research Center, Flagstaff, Arizona. servation Biology 14:428Knap results in ountain yellow-legged frog, Rana mucosa ). Biological Conservation 135:11-20. Korm s / etric/Korman2005.pdf (January 2008). of c Lee, K. N. 1999. Appraising adaptive manageme nt. Conservation Ecology 3(2): 3. Available: Lepak l 2006. Rapid food web recovery in response to Lessa e c ontrol of species ab undances? Ecology and Society. 10(2):1. Available: http://www.ecology andsociety.o rg/vol10/iss2/art1/. (January 2008). Kitchell, J. F., C. Grimes, S. T. Lindley, D. Otis, and C. Schwarz. 2003. Report to the adaptive management work group, Glen Canyon Dam Adaptive Management Program: An independent review of ongoing and proposed scie ntific methods to assess the status a trends of the Grand Canyon population of the humpbac Available: www.gcmrc.gov/library/reports /biological/F ish_studies/ Kitchell2003.pdf. (January 2008). Knapp, R. A. and K. R. Matthews. 2000. Non-native fish introductions an d the decline of the mountain yellow-legged frog from within pr otected areas. Con 438. p, R. A., D. M. Boiano, and V. T. Vrende nburg. 2007. Removal of nonnative fish population expansion of a declining am phibian (m an, J. M. Kaplinski, J. E. Hazel, III, and T. S. Melis. 2005. Effects of the experimental fluctuating flows from Glen Canyon Dam in 2003 and 2004 on the early life history stage of rainbow trout in the Colorado River. Final Report, submitted to the Grand Canyon Monitoring and Research Center Flagstaff, AZ. Available: http://www.gcmrc.gov/library reports/biological/Fish_st udies/Ecom Ladson, A. R. and R. M. Argent. 2002. Adaptiv e management of environmental flows: lessons for the Murray-Darling basin from three large North American rivers. Australian Journal Water Resources 5:89-102. Ladson, L. and W. Allan. 2002. Solver. Frontlin e Systems, Incline Village, NV. Available: http://www.solver.com (January 2008). Laslett, G. M., J. P. Eveson, and T. Polach eck. 2002. A flexible maximum likelihood approach for fitting growth curves to tag-recapture data. Canadian Journal of Fisheries and Aquati Sciences 59:976-986. http://www.consecol.org/vol3/iss2/art3/ (January 2008). J. M., C. E. Kraft, and B. C. Weide removal of an introduced apex predator. Ca nadian Journal of Fisheries and Aquatic Sciences 63:569-575. rd, R. B., S. Martell, C. J. Walters, T. E. Essington, and J. F. Kitchell. 2005. Should ecosystem management involve activ 164

PAGE 165

Lorenzen, K. 2000. Allometry of natural mortality as a basis for assessing optimal release siz fish-stocking programmes. Canadian Journal of Fisheries and Aquatic Sciences 57:2 2381. S., M. Browne, S. Boudjelas, and M. De Poorter. 2000. 100 of the worlds worst e in 374 Lowe invasive alien species: a selection from the global invasive species database. Invasive Species Lunn, D. J., A. Thomas, N. Best, and D. Sp iegelhalter. 2000. Winbugs -a Bayesian modelling 37. MacKenzie, D. I. and W. L. Kendall. 2002. How should detection probability be incorporated MacK yle, K. H. Pollock, L. L. Bailey, and J. E. Hines. 2006. Occupancy estimation and modeling: inferr ing patterns and dynamics of species Makinster, A. S., R. S. Rogers, and W. R. Persons. 2007. Status of the Lees Ferry trout fishery: ng x, AZ. Mang Marsh, P. C. and M. E. Douglas. 1997. Predation by introduced fishes on endangered humpback the oring and Research Center, Flagstaff, AZ. McDo ealand ew look at an old problem Transactions of the American Fisheries Society 132:229-238. McDo r crying shame: alien salmonids and a biodiversity crisis in the s outhern cool-temperate galaxioid fishes? Reviews in Fish Megrey, B. A. 1989. Review and comparison of age structured stock assessment models from ociety, Symposium 6, Bethesda, Maryland. Specialist Group, Species Survival Commissi on of the World Conservation Union. Available: www.issg.org/booklet.pdf (January 2008). framework: concepts, structure, and extens ibility. Statistics a nd Computing 10:325-3 into estimates of relative abundance? Ecology 83:2387-2393. enzie, D. I., J. D. Nichols, J. A. Ro occurrence. Elsevier, Burlington, Massachusetts. 2006 Annual Report. 2006 Annual Report, submitted to the Grand Canyon Monitori and Research Center, Flagsta ff, AZ. Arizona Game and Fi sh Department, Phoeni el, M. and 41 coauthors. 1996. Principles for the conservation of wild living resources. Ecological Applications 6:338-362. chub and other native species in the Little Colorado River, Arizona. Transactions of the American Fisheries Society 126:343-346. Martell, S. J. D. 2006. Review of the status and trends of humpback chub ( Gila cypha ) in Grand Canyon: 1989-2005. Report submitted to Grand Canyon Monit wall, R. M. 2003. Impacts of introduced salmonids on native galaxiids in New Z upland streams: a n wall, R. M. 2006. Crying wolf, crying foul o Biology and Fisheries 16:233-422. theoretical and applied points of view. Pages 8 in E. F. Edwards and B. A. Megrey, editors. Mathematical analysis of fish stock dynamics. Am erican Fisheries S 165

PAGE 166

Melis, T. S., S. J. D. Martell, L. G. Coggins Jr., W. E. Pine, III, and M. E. Andersen. 2006. Adaptive management of the Colorado River ecosystem below Glen Canyon Dam, Arizona: using science and modeling to reso lve uncertainty in river management. In Proceedings of the American Water Res ources Association, 2006. Available: Meyer, K. A., J. A. Lamansky, Jr., and D. J. Schill. 2006. Evaluation of an unsuccessful brook trout electrofishing removal pr oject in a small Rocky Mount ain stream. North American Minckley, W. L. 1973. Fishes of Arizona. Ar izona Game and Fish Department, Phoenix, Minckley, W. L. 1991. Native fishes of the Grand Canyon region: An obituary? Pages 124-177 Mueller, G. A. 2005. Predatory fish removal a nd native fish recovery in the Colorado River Mueller, G. A., C. Walters, P. B. Holden, P. G. Walker, J. Landye, and B. Johnson. 1999. of Reclamation, Upper Colorado Region, Salt Nichols, J. D., F. A. Johnson, and B. K. W illiams. 1995. Managing North American waterfowl in Nilsso 5. Fragmentation and flow NRC of mations Glen Canyon Environmen tal Studies. National Academy Press, NRC gement in the Grand Canyon. National Academy Press, Washington, D.C. NRC Canyon er Ecosystem. Na tional Academy Press, Washington, D.C. in http://www.awra.org/ proceedi ngs/cdrom.html#Missoula_2006 (January 2008). Journal of Fisheries Management 26:849-860. Arizona. in Marzolf G. R., editor. Colorado Rive r Ecology and Dam Management. National Academy Press, Washington, DC. mainstem: what have we learned? Fisheries 30:10-19. Scientific review of the Glen Canyon Da m modifications to control downstream temperatures plan and draft environmental assessment, January, 1999, and supporting documentation. Prepared for the Bureau Lake City, Utah. Available: http://www.gcmrc.gov/library/reports/Synthesis/ Mueller1999b.pdf (April 2008). the face of uncertainty. Annual Review of Ecology and Systematics 26:177-199. n, C., C. A. Reidy, M. Dynesius, and C. Revenga. 200 regulation of the worlds largest river systems. Science 308:405-408. (National Research Council). 1987. River and dam management: a review of the Bureau Recla Washington, D.C. (National Research Council). 1996. River resource mana (National Research Council). 1999. Downst ream: adaptive management of Glen Dam and the Colorado Riv Olden, J. D. and N. L. Poff. 2005. Long-term tr ends of the native and non-native fish faunas the American Southwest. Animal Bi odiversity and Cons ervation 28:75-89. 166

PAGE 167

Olsen, D. G. and M. C. Belk. 2005. Relationship of diurnal habitat use of native stream fishes the eastern great basin to presence of in troduced salmonids. Western North Amer Naturalist 65:501-506. of ican ce from Palmer, M. A. and 21 coauthors. 2005. Standard s for ecologically successful river restoration. Palmer, M., J. D. Allan, J. Meyer, and E. S. Be rnhardt. 2007. River restora tion in the twenty-first century: data and experiential knowledge to in form future efforts. Restoration Ecology Paloh th of fishes. I. A growth curve derived from experimental data. Journal of the Fi sheries Research Boar d of Canada 22:521. Paloh temperature on the relation between metabolis m and body weight. Journal of the Fisheries Parker, B. R., D. W. Schindler, D. B. Donald, a nd R. S. Anderson. 2001. The effects of stocking Pascual M. A. and R. Hilbor n. 1995. Conservation of harv est populations in fluctuating al of Applied Ecology 32:468480. Patten planned flooding. Ecological Applications 11:633-634. Pauke fecting condition of flannelmouth suckers in the Colorado River, Grand Canyon, Arizona. North American Journal of Fisheries Paukert, C. P. and J. H. Petersen. 2007. Comp arative growth and cons umption potential of Osmundson, D. B., R. J. Ryel, V. L. Lamarra, and J. Pitlick. 2002. Flow-sediment-biota implications for river regulation effects on na tive fish abundance. Ec ological Applications 12:1719-1739. Otis, D. L., K. P. Burnham, D. R. Anderson, and G. C. White. 1978. Statistical inferen capture data on closed populations Wildlife Monographs 62:1. Journal of Applied Ecology 42:208-217. 15:472-481. eimo, J. E. and L. M. Dickie. 1965. Food a nd grow eimo, J.E. and L.M. Dickie. 1966. Food a nd growth of fishes. II. Effects of food and Research Board of Canada 23:869-908. and removal of a nonnative salmonid on the plan kton of an alpine lake. Ecosystems 4:334 345. Parma, A. M. and 9 coauthors. 1998. What can adaptive management do for our fish, forests, food, and biodiversity? In tegrative Biology 1:16. environments: the case of the Serengeti wi ldebeest. Journ D. T. and L. E. Stevens. 1999. Rest oration of the Colorado River ecosystem using rt, C. and R. S. Rogers. 2004. Factors af Management 24:648-653. rainbow trout and humpback chub in the Colorado River, Grand Canyon, Arizona, under different temperature scenarios. S outhwestern Naturalist 52:234-242. 167

PAGE 168

Pauly, D. 1980. On the interrelationships betw een natural mortality, growth parameters, and mean environmental temperature in 175 fi sh stocks. Journal du Conseil, Conseil International pour LExploration de la Mer 39: 175-192. nsactions of the American Fisheries Society 133:462-475. Pine, ice. 2003. A review of tagging methods for estimating fish population size and components of mortality. Fisheries Pine, Kwak, and J. A. Rice. 2007. Modeling management scenarios and the effects of an introduced apex predator on a co astal riverine fish community. Transactions Poff, Postel, S. L., G. C. Daily, and P. R. Ehrlic h. 1996. Human appropriation of renewable fresh Quinn Oxford University Press, New York. R De nvironment for statistical computing. R Foundation for Statistical Compu ting, Vienna, Austria. Avai lable: http://www.R-project Reisner, M. 1993. Cadillac desert. Penguin Books, New York. hesda, Maryland. Ecology 65:732-739. ns. Robinson, A. T. and M. R. Childs. 2001. Juvenile growth of native fishes in the Little Colorado Petersen, J. H. and C. P. Paukert. 2005. Deve lopment of a bioenergetics model for humpback chub and evaluation of water temperature ch anges in the Grand Canyon, Colorado River Transactions of the American Fisheries Society 134:960-974. Peterson, J. T., Thurow, R. F., and J. W. Guzevich. 2004. An evaluation of multipass electrofishing for estimating the abundance of stream-dwelling salmonids. Tra W. E., III, K. H. Pollock, J. E. Hightower, T. J. Kwak, and J. A. R 28:10. W. E., III, T. J. of the American Fisheries Society 136:105-120. N. L. and 9 coauthors. 2003. River fl ows and water wars: emerging science for environmental decision making. Frontiers in Ecology and the Environment 1:298-306 water. Science 271:785-788. T. J., II and R. B. Deriso. 1999. Quantitative Fish Dynamics. velopment Core Team. 2007. R: A language a nd e .org. (January 2008) Reynolds, J. B. 1996. Electrofishing. Pages 221 in B. R. Murphy and D. W. Willis, editors Fisheries techniques, 2nd edition. American Fisheries Society, Bet Rice, J. A. and P. A. Cochran. 1984. Indepe ndent evaluation of a bioenergetics model for largemouth bass. Ricker, W. E. 1975. Computation a nd interpretation of biological statistics of fish populatio Fisheries Research Board of Canada Bulletin 191. River and in a thermally modified portion of the Colorado River. North American Journal of Fisheries Management 21:809-815. 168

PAGE 169

Robinson, A. T., S. D. Bryan, and M. G. Sw eetser. 2003. Habitat use by non-native rainbow trout, Oncorhynchus mykiss, and native Little Colorado spinedace, Lepidomeda vitta Environmental Biology of Fishes 68:205-214. Rosen echniques for rainbow trout captured by electrofishing in small streams. North American Journal of Fisheries Management 25:1395-1410. Royle Sainsbury, K. J. 1991. Application of an experimental approach to management of a tropical um Samp uinn II, J. Heifetz, J. N. Ianelli, J. E. Powers, J. F. Schweigert, P. J. Sullivan, and C. I. Zhang, editors. Fishery Stock Assessment Models. Schm d T. L. Hoffnagle. 2001. The 1996 controlled flood in Grand Canyon: flow, sediment transport, Schnu n Journal of Fisheries and Aqua tic Science 38:1128-1140. Schre 2004. Adaptive management: a synthesis of current understanding and effective application. Ecological Management and Seage estern North America. Science 316:1181. Sit, V Speas ific density ow trout in berger, A. E. and J. B. Dunham. 2005. Validation of abundance estimates from markrecapture and removal t J. A. and R. M. Dorazio. 2006. Hierarch ical models of animal abundance and occurrence. Journal of Agricultural, Biological, and Environmental Statistics 11:249-263. Sainsbury, K. J. 1980. Effect of individual va riability on the von Bertalanffy growth equation Canadian Journal of Fisheries and Aquatic Sciences 37:241-247. multispecies fishery with highly uncertain dynamics. ICES Marine Science Symposi 193:301-320. son, D. B. and Y. Yin. 1998. A monte car lo evaluation of the st ock synthesis program. Pages 315-338 in F. Funk, T. J. Q Alaska Sea Grant College Program Report No. AK-SG-98-01, University of Alaska Fairbanks. idt, J. C., R. A. Parnell, P. E. Grams, J. E. Hazel, M. A. Kaplinski, L. E. Stevens, an and geomorphic change. Ecologi cal Applicati ons 11:657-671. te, J. 1981. A versatile grow th model with statistically stab le parameters. Canadia iber, E. S. G., A. R. Bearlin, S. J. Nicol, and C. R. Todd Restoration 5:177-182. r, R and 12 coauthors. 2007. Model projections of an imminent transition to a more arid climate in southw Seber, G. A. F. 1982. The estimation of animal abundance and related parameters, 2nd edition Griffin, London. and B. Taylor, editors. 1998. Statisti cal methods for adaptive management studies. Research Branch, British Columbia Ministry of Forests, Land Management Handbook No. 42, Victoria, British Columbia. D. W., C. J. Walters, D. L. Ward, and R. S. Rogers. 2004. Effects of intraspec and environmental variables on electrofishing ca tchability of brown and rainb the Colorado River. North American Journal of Fisheries Management 24:586-596. 169

PAGE 170

Stone, D. M. and O. T. Gorman. 2006. Ontogenesis of endangered humpback chub ( Gila cypha ) in the Little Colorado River, Arizona. American Midland Naturalist 155:123-135. D. M., D. R. Van Haverbeke, D. L. Wa rd, a Stone nd T. A. Hunt. 2007. Dispersal of nonnative fishes and parasites in the intermittent Li ttle Colorado River, Arizona. Southwestern Tanig Taylor, N. G., C. J. Walters, and S. J. D. Martell. 2005. A new likelihood for simultaneously ng Topping, D. J., D. M. Rubin, J. C. Schm 2006. Comparison of sediment-transport and orado Topping, D. J., J. C. Schmidt, and L. E. Vi erra, Jr. 2003. Computation and analysis of the Available:http://pubs.usgs.go v/pp/pp1677/pdf/pp1677.pdf. (January 2008). er in Grand Canyon, Arizona. Final Report submitted to the Grand Canyon Monitoring a nd Research Center, Flagstaff, AZ. Tyus, : 4. y/reports /LawoftheRiver/ GCDEIS_Final.pdf (January 2008). USFWS (U.S. Fish and Wildlife Service). 1994. Fi nal Biological Opinion, operation of Glen al a State Office, U.S. Fish and Wildlife Service, Phoenix, AZ. s: n to the humpback chub Recovery Plan. U.S. Fish and Wildlife Service, Mountain-Prairie Regi on (6). Denver, Colorado. Available: January 2008). Naturalist 52:130-137. uchi, Y., F. J. Rahel, D. C. Novinger, a nd K. G. Gerow. 1998. Temperature mediation of competitive interactions among three fish speci es that replace each other along longitudinal stream gradients. Canadian Journal of Fisheries and Aquatic Science 55:1894-1901. estimating von Bertalanffy growth parameters gear selectivity, and natural and fishi mortality. Canadian Journal of Fish eries and Aquatic Sciences 62:215-223. idt, J. E. Hazel, Jr., T. S. Melis, S. A. Wright, M. Kaplinski, A. E. Draut, and M. J. Breedlove bar-response results from the 1996 and 2004 c ontrolled-flood experiments on the Col River in Grand Canyon: Proceedings of the 8th Inter-Agency Sedimentation Conference April 2-6, 2006, Reno, Nevada. instantaneous-discharge for the Colorado River at Lees Ferry, Arizona: May 8, 1921 through September 30, 2000. U.S. Geological Surv ey Professional Paper, Reston, Virginia 1677. Trammell, M., R. Valdez, S. Carothers, and R. Ryel. 2002. Effects of a low steady summer flow experiment on native fishes of the Colorado Riv H. M. and J. F. Saunders, III. 2000. Nonna tive fish control and en dangered fish recovery lessons from the Colorado River. Fisheries. 25:17-2 USDOI (U.S. Department of the Interior). 1995. Operation of Glen Canyon Dam Final Environmental Impact Statement, Bureau of Reclamation, Salt Lake City, Utah. Available: http://www.gcmrc.gov/librar Canyon Dam as the modified low fluctuating fl ow alternative of th e Final Environment Impact Statement, operation of Glen Ca nyon Dam (2-21-93-F-167). Ecological Services, Arizon USFWS (U.S. Fish and Wildlife Service). 2002. Humpback chub ( Gila cypha ) Recovery Goal amendment and supplementatio http://www.gcmrc.gov/library/report s/biological/Fish_studies/USFWS/ Humpbackchub_recoverygoals2002.pdf. ( 170

PAGE 171

Valdez, R. A. and R. J. Ryel. 1995. Life history and ecology of the humpback chub ( Gila c in the Colorado River, Arizona. Final Repor t. Contract No. 0-CS-40 ypha) -09110. Salt Lake City, UT. Available: http://www.gcmrc.gov/library/repor ts/biological/Fish_studies/ Valde R. J. Ryel, K. R. Bestgen, and D. L. Wegner. 2000. A program of experimental flows for the endangered n ry/ _s tudies/swca/99-FC-40-2 260 LSSF/Valdez_2000_EXPFLOW Van H nt and fisheries monitoring r and Canyon. U.S. Geological Survey Circular Vialle rn Arizona 1988-2005. USGS, Vrendenburg, V. T. 2004. Reversing introduced species effects: expe rimental removal of introduced fish leads to rapid recovery of a declining frog. Proceedings of the National Academy of Sciences 101:7646-7650. Walters, C. J. 1986. Adaptive management of renewable resources. McMillan, New York. Walters, C. J. 1997. Challenges in adaptive mana gement of riparian and coastal ecosystems. Conservation Ecology 1(2):1. Available: http://www.consecol.org/vol1/iss2/art1/ (January 2008) Walters, C. J. and C. S. Holling. 1990. Larg e-scale management expe riments and learning by doing. Ecology 71:2060-2068. Walters, C. J. and Martell, S. J. D. 2004. Fisheries Ecology and Management. Princeton University Press, Princeton, New Jersey. Biowest/Valdez1995f.pdf (January 2008). z, R. A., S. W. Carothers, D. A. House, M. E. Douglas, M. Douglas, and native fishes of the Colorado River in Grand Canyon. Final Report, Grand Canyo Monitoring and Research Center Flagstaff, AZ. Available: http://www.gcmrc.gov/libra reports/biological/Fish _99FC402260.comp.PDF (January 2008). averbeke, D. R. and D. M. Stone. 2007. Stock assessme activities in the Little Co lorado River within Grand Canyon during 2006. Final Report submitted to the Grand Canyon Monitoring and Re search Center. U.S. Fish and Wildlife Service, Flagstaff, AZ. Vernieu, W. S., S. J. Hueftle, and S. P. Gl oss. 2005. Water quality in Lake Powell and the Colorado River. Pages 69-85 in S. P. Gloss, J. L. Lovich, and T. S. Melis, editors The state of the Colorado River ecosystem in G 1282. font, A., J. D. Lebreton, and A. M. Reboulet. 1998. Parameter identifiably and model selection in capture-recapture models: a numerical approach Biometrical Journal 40:313325. Vitousek, P. M., H. A. Mooney, J. Lubchenco, and J. M. Melillo. 1997. Human domination of the Earths ecosystems. Science 277:494-499. Voichick, N. and S. A. Wright. 2007. Water-tempe rature for the Colorado River and tributaries between Glen Canyon Dam and Spencer Canyon, Northe Reston, Virginia. US Geological Data Series: 251. 171

PAGE 172

172 Walters, C. J., L. H. Gunderson, a mental policies for water management in the Everglades. Ecological Applications 2:189-202. Walters, C., J. Korman, L. E. Stevens, and B. Gold. 2000. Ecosystem modeling for evaluation of Available: http://www.consecol.org/vol4/iss2/art1/. (January 2008). Wang, Y-G., M. R. Thomas, and I. F. So mers. 1995. A maximum likelihood approach for Science 52:252-259. Webb, R. H., J. S. Schmidt, G. R. Marzolf, and R. A. Valdez, editors. 1999. The controlled flood n: Colorado River, Glen and na University, Flagstaff. Yoccoz, N. G., J. D. Nichols, and T. Boulinie r. 2001. Monitoring biological diversity in space and time. Trends in Ecology and Evolution 16:446-453. Zippin, C. 1956. An evaluation of the remova l method of estimating animal populations. Biometrics 12:163-189. nd C. S. Holling. 1992. Experi adaptive management policies in the Gra nd Canyon. Conservation Ecology 4(2): 1. estimating growth from tagrecapture data. Ca nadian Journal of Fisheries and Aquatic in Grand Canyon. AGU Geophysical Monograph 110. Williams, B. K., J. D. Nichols, and M. J. C onroy. 2002. Analysis and management of animal populations. Academic Press, London, UK. Williams, B. K., R. C. Szaro, and C. D. Shapiro. 2007. Adaptive Management: The U.S. Department of Interior Technical Guide. Adaptive Management Working Group, U.S. Department of Interior, Washington, D.C. Wright, S. A., T. S. Melis, D. J. Topping, D. M. Rubin. 2005. Influence of Glen Canyon Dam operations on downstream sand resources of the Colorado River in Grand Canyon. Pages 17-32 in S. P. Gloss, J. L. Lovich, and T. S. Melis, editors The state of the Colorado River ecosystem in Grand Canyon. U.S. Geological Survey Circular 1282. Yard, M. D. 2003. Light availabi lity and aquatic primary productio Grand Canyons, AZ. Doctoral dissertation. Northern Arizo

PAGE 173

173 BIOGRAPHICAL SKETCH I was born in Phoenix, AZ in 1967 to Lewi s and Jan Coggins. After attending primary school in both Phoenix and in Tuba City on the Navajo Reservation in Northern AZ, I graduated from Coconino High School in Flagstaff, AZ in 1985. During these formative years, I was fortunate to have a father who loved exploring Northern Arizona and particularly Marble and Grand Canyons. Our explorations and his teachi ngs of the natural world had a profound effect on my career choices. After gra duating from the University of Arizona in 1990 with a degree in ecology and evolutionary biology, I moved to Alaska where I worked as a fisheries biologist for the Alaska Department of Fish and Game in Bris tol Bay and Kodiak. I ma rried Jennifer (Gape) Coggins in December 1993 and our first child, E lizabeth Tate, was born in Kodiak, AK in 1997. While employed by the state, I was able to pursue and earn a master’s degree in fisheries from the University of Alaska, Fairbanks, in 1997, under the guidance of Dr. Terry Quinn. I returned to Flagstaff in 1999 to study Grand Canyon fishes with the U.S. Geological Survey. In August 2002, our second child, Annie Rose, was born in Flagstaff, AZ.


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20101111_AAAACK INGEST_TIME 2010-11-11T22:22:46Z PACKAGE UFE0021996_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 8937 DFID F20101111_AABUAS ORIGIN DEPOSITOR PATH coggins_l_Page_168thm.jpg GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
8d8ed6066fdfd92d12161a11b68a89e1
SHA-1
3e8e3dfddffbf4f47117b7585fee2181295f4fab
35409 F20101111_AABTWB coggins_l_Page_067.QC.jpg
801b1e17b542b2d98e3da3916b1bc4d4
c98d6cf23b728f35dcbb0628d131796696c6c598
18135 F20101111_AABTVM coggins_l_Page_050.QC.jpg
d2fa5fc64c9c9398abc58dc033f7d162
1a7c166f8f2b5887579e57c25fc1ea7372dfa1bf
39411 F20101111_AABTUY coggins_l_Page_037.QC.jpg
3e7416f18376774073264d7a74214996
135ba3d9b708e3b64e9fe15a3c68fe8c324ee2f9
1051915 F20101111_AABSSK coggins_l_Page_035.jp2
159199dd0c87d98179f97972fc6a4c9f
76c9d459bcb3396e21dbd0518a40bdbb604d7315
1051960 F20101111_AABSRW coggins_l_Page_020.jp2
894efc726abf61d20170012da67e5986
8ae413cd6d0c3b74e710777b58d1672509f50d46
39077 F20101111_AABUAT coggins_l_Page_169.QC.jpg
ad5ff11669c573ed99d233e261a76bd1
0d14fb267a9d6d58c7c58ba854467e7784e5fa1e
8673 F20101111_AABTWC coggins_l_Page_067thm.jpg
8c136f08058178703367fd3dea5aa920
286720e823b66f85f9e7978dafb4c13571d1914e
5378 F20101111_AABTVN coggins_l_Page_050thm.jpg
d5f3ba3475a69e107971a0d0cf500d9a
370ea98ac3b4a8e84bab35113564e3854b0764f6
9240 F20101111_AABTUZ coggins_l_Page_037thm.jpg
3b5b12e78f0c94bb16dab18a79ad67b1
c0e4ff35ec54eb6becf3da83258d0c2d2ef70a29
463030 F20101111_AABSTA coggins_l_Page_052.jp2
436db402faf1f1ea7de461257a0cde11
4c94e0855653aa3f8870b50f4473dd1bac114b16
1051982 F20101111_AABSSL coggins_l_Page_036.jp2
7082892f43a060d546d31af1cc5a52df
547d89a7ced4d3960c60b68cfa980f09128c44a4
1051980 F20101111_AABSRX coggins_l_Page_022.jp2
d4265104ec193b5e2071bf09010e39f5
62206f4270fbe5da1c20d04cbbbb6f33d8ba363a
43024 F20101111_AABUAU coggins_l_Page_170.QC.jpg
49b126fd0c96abc1e7591f1920ee008a
f9e541067536ff5e21c779676cdfea2031c3cf65
8781 F20101111_AABTWD coggins_l_Page_068thm.jpg
a6af9f61d9561f1be919dbc8edb86390
745058b06ab78cb2eb70e790912ba47ab7c0d403
19451 F20101111_AABTVO coggins_l_Page_051.QC.jpg
f79d26b62d20737db606d6eef90ac324
ad897a2c9fd889ba358d51a7c3d0e1108b8431c0
718686 F20101111_AABSTB coggins_l_Page_053.jp2
f9bf527e8629a9ced98b19b6317abd7a
c81d4b167696e41a3dcafc167a0efcfdd5ca41cc
1051983 F20101111_AABSSM coggins_l_Page_037.jp2
233309b53266d8c9f0ce7d7255132719
18120bcc8438272f4fbbdb910fb93c1ebdc4b2e3
1051935 F20101111_AABSRY coggins_l_Page_023.jp2
6f64be9e01c15fc893daa2ee1aacb5bc
52ed40306a05d103d4ff1e0ac97b833ee6ad36e4
9603 F20101111_AABUAV coggins_l_Page_170thm.jpg
9f201a8d23eafeba0d1f593a39170c2c
b1a4f6d278017e5930c486fce49bd3d29fe577ac
8926 F20101111_AABTWE coggins_l_Page_069thm.jpg
0561b66bc131c4bfa2f120f365ff0a40
b800a4474a0ac5b441dc51221a96e252d42d8684
13127 F20101111_AABTVP coggins_l_Page_052.QC.jpg
a3392ee852e517acd75f2e4d0acc11a3
9fd34653090534e5d1718ef8317ebcc99980394b
362908 F20101111_AABSTC coggins_l_Page_054.jp2
3e5c5166a5552132ee28472a48732667
90a5f7bbc47b4b1f266f4482ed8e8d0bab96de95
1051968 F20101111_AABSSN coggins_l_Page_038.jp2
24d47d5cc9d013ff161f511b74974501
d9bf037159730fa5ea0295da405683f531427a60
1034933 F20101111_AABSRZ coggins_l_Page_024.jp2
b2b31ed60ed9787fea3b65ce20ea37e5
2519639abbae0b70d75326d33dd334efb77c3175
22416 F20101111_AABUAW coggins_l_Page_173.QC.jpg
0bead526f5593eb7580aaedb9b76a1eb
f90ce007282fae116efa471aa87d85697e42dfd3
35194 F20101111_AABTWF coggins_l_Page_070.QC.jpg
41614473f848e7841e366ddce380b17d
3f8937507ff1c96729fb9554bb4dabf2403cda31
3353 F20101111_AABTVQ coggins_l_Page_052thm.jpg
b20fef9507df3c2e228585b35d134ed5
c64ba2e201025505544c14887a80002b103a2bb9
773086 F20101111_AABSTD coggins_l_Page_055.jp2
39de62dd89465aa6c00c9856195f790a
2648fa34e2b346d470ee4c2cf32d517f79441b71
1051912 F20101111_AABSSO coggins_l_Page_039.jp2
aee5c607b40f22a68e6f05a2477bc705
bbe8709b60ac2b3bb579bfcd635df500d6656365
5586 F20101111_AABUAX coggins_l_Page_173thm.jpg
ad804f2ec85ac44547b40fde145dedb6
814e192405814cb3fa52416110ef0439124c214a
8927 F20101111_AABTWG coggins_l_Page_070thm.jpg
96ce1704a86241991c451b6af05f9f3c
997ae3bb804cb69e1dd9cfd8820e012c2eaa2bc4
10628 F20101111_AABTVR coggins_l_Page_054.QC.jpg
b915ff91020a118ecdb0e6be29bb30a4
bb0e4ecf18124f9e6c9213cbb546aa4d32279c2c
970443 F20101111_AABSTE coggins_l_Page_056.jp2
79ae7aa3ee412ec51348c687f823a6d7
9683554eef5f02debf658dbce402c460187cb1fd
323129 F20101111_AABSSP coggins_l_Page_040.jp2
636370ef0e93dc3020d9639dc8f840e4
5c54237537ae19a1270770e4def1cfbf91c7274c
38362 F20101111_AABTWH coggins_l_Page_071.QC.jpg
a2d2ce149d32cdcc63bb97863f37dcd7
4e2f710898a56ca124ca5bee3f486d2242240477
3319 F20101111_AABTVS coggins_l_Page_054thm.jpg
1e81650a0f6504654f45841c430434ce
caa5017804a609315d76a37e2980f6592382eaf5
997889 F20101111_AABSSQ coggins_l_Page_041.jp2
e45aa723bbf305bca100d450b38a3810
0c600aed60676108e06730ca79186be573308b88
1051981 F20101111_AABSTF coggins_l_Page_057.jp2
47d5fb240e3abbd5f960efab8304fe7c
639f42844136fd2b5708fa25a49b2d992f90a8a1
20318 F20101111_AABTVT coggins_l_Page_056.QC.jpg
1bd8dd6ccb201e5b6438fc63c371fd80
d33c72e5026a47f6a060d2b7ef2881cea24d2ae7
996169 F20101111_AABSSR coggins_l_Page_042.jp2
6effc4d69553595c5386e055967293fd
06c9f28265c04cc3cc3ae57e3a57e88991a1b0bf
9056 F20101111_AABTWI coggins_l_Page_071thm.jpg
becf7bcffe3ebfe28ab83574796cbd97
ef6958ff59ab9a6ca7bdd2dc607e2789f8bbec7b
37545 F20101111_AABTVU coggins_l_Page_057.QC.jpg
e6e1345c14f33fdc398cca8baf081010
607dbb907dc4b06c33dec536e0b85ea415c8f228
644308 F20101111_AABSSS coggins_l_Page_043.jp2
3bc9dd353172d969c8167d0dca309c1c
2ce9e73ba7835c4b2e920fd57a2432a2bd821c68
1051934 F20101111_AABSTG coggins_l_Page_058.jp2
2bb288087682b9737b15824b2e30b50d
3e1913998eb7a46d1027bc82690d3c1da011b237
9236 F20101111_AABTWJ coggins_l_Page_072thm.jpg
3453883cf20710d54ff9ba9e5d0e3e97
01ab55605c762370db130df372eb8bbf35866482
9054 F20101111_AABTVV coggins_l_Page_057thm.jpg
bb126375b760105327ea3b60a9c9b3a2
1316580c19a15d09c33e41cf0461ba23647c7c1a
562637 F20101111_AABSST coggins_l_Page_044.jp2
1c84a50b5d4199e8d280b93b5daf9891
de15c35920b06c9126d0e9a4c00605131a6ae1ea
1051979 F20101111_AABSTH coggins_l_Page_059.jp2
e6c938a3fc95479530bcfc7aae79c5cd
109f7a178d4b9cb3aaf47af3d5b1691cb623f19c
1479 F20101111_AABTWK coggins_l_Page_073thm.jpg
30105012ec9383b5f7b26d418ec9819c
e1f3761a2d3ddaa9bad5c3f53bf45436f461245e
33878 F20101111_AABTVW coggins_l_Page_059.QC.jpg
76d1071f9b22a3cd74d0c77586056496
74968bf604164011cc22f40fbf16067eebaeca0e
446255 F20101111_AABSSU coggins_l_Page_045.jp2
533b58c129fa25ef9bc6730112e5c329
db31dee26612b466afebf957fd2b4cdeadbbf9a1
1051977 F20101111_AABSTI coggins_l_Page_060.jp2
18ec1eee4b34bae52766e18a6ff141d6
aa5187d77e4bfc10f8c38598ca40d3308a21e785
36343 F20101111_AABTXA coggins_l_Page_086.QC.jpg
f726d7ee65e7b3b47a900ddf6dcf9097
1d1aa5c4f48bc5180bd922dde07884f47d1eef58
5178 F20101111_AABTWL coggins_l_Page_075.QC.jpg
b587aca6f8692ad17333fa2757ea7c6e
458b6e160ceabc8413972032709f73a50190126e
8883 F20101111_AABTVX coggins_l_Page_060thm.jpg
35c0413797a12fe16d29e7f841b773da
462655f44722cbf5f3a640faa9b23b48bcc4d1c5
371240 F20101111_AABSSV coggins_l_Page_046.jp2
438d6a47fa89204e129242b9c928d91a
3fb39e0e6247a3ce0e2ddfc952ad1e8418cf0bb8
719953 F20101111_AABSTJ coggins_l_Page_061.jp2
b92d991d7ab779f5749c8f0a7eec9e41
8904b14e56acaa0f67db6cc0e527752182c91e6e
8554 F20101111_AABTXB coggins_l_Page_086thm.jpg
d542477a3311d009787b3c2d41a8ec38
3454a39e2f53ecf3c289ea10f4492258056429c3
1357 F20101111_AABTWM coggins_l_Page_075thm.jpg
f0f0842bbcff4cb9f9d5b471b53f21cd
f319f63de654fa55d8261e05fc8097eff5db6774
21842 F20101111_AABTVY coggins_l_Page_061.QC.jpg
bc2dd17d88e54dba9aa859f734f905ad
478fd5e6ab3747c01a68084d84931e3e23e36606
836947 F20101111_AABSSW coggins_l_Page_047.jp2
58304f53b66689e54ff0b1e4c6297e67
cb70aa3dacc21562e241ca46fb013beb958511ee
845954 F20101111_AABSTK coggins_l_Page_062.jp2
85ff1c3e75b17802a0ec5a8d62b5bf22
af966b50fa6b4ace18da9ce31f8be998a6a4cf01
8745 F20101111_AABTXC coggins_l_Page_087thm.jpg
5324fdcd733ead5ebe282c5dd915a91a
fd7633e3faf4d3af72215f3d84eb56fb3f20de50
5845 F20101111_AABTWN coggins_l_Page_076.QC.jpg
79a2cf754fae60f1bcd1d1c0316b3e3a
1817df950b057d91ca2d9b06dc84ba91c95c8ee8
34822 F20101111_AABTVZ coggins_l_Page_064.QC.jpg
1562935284d5d7b21312791b2a80f843
1beacd0619c2427521197df1188bfc5db069ada4
654638 F20101111_AABSSX coggins_l_Page_049.jp2
28448ec6cb0bdcae6ce1ebcd950fdcb2
8fd5bc723db60d9c5983cf1032e58718b68f10d4
532204 F20101111_AABSUA coggins_l_Page_080.jp2
89a27c8f83d15386bf30f9157050d7f4
21d64b2ea8d0475473c762517537cb644934ba16
F20101111_AABSTL coggins_l_Page_063.jp2
b9686748b5df2a32667cbbec6065aa84
8bb43b304361923392b3bd4b5fce0633e9a68ecb
36832 F20101111_AABTXD coggins_l_Page_088.QC.jpg
b2e2fe6b188f9bd08fe773b93c9f1f7a
c69f35902e86c331349bc1968cedd6334def1ee6
1487 F20101111_AABTWO coggins_l_Page_076thm.jpg
cd5635e325b050a7beb3cd06312816e9
710995bc47d0cfb45571a66ec951ec507fccbe5b
634154 F20101111_AABSSY coggins_l_Page_050.jp2
8ebff7625e47643dd4cd5aa38e013ca7
71abe943257cf8a19bbbb6d11b58c0e5ee924a4e
553720 F20101111_AABSUB coggins_l_Page_081.jp2
692fbbb9cb0d8313f0f5de1e81dd6038
0a4b3ec473b6f0a5881108891c119e1345090ad1
1051971 F20101111_AABSTM coggins_l_Page_064.jp2
e5c9cb3f37686d45c961b11367907799
bb714e3b0a2496ad660f46ee4be0ac71ee06fb98
32382 F20101111_AABTXE coggins_l_Page_090.QC.jpg
035a44f82e9b26079d0bf39aa9c322a4
a61f41387049248840decc8219c9bf909e98a284
3311 F20101111_AABTWP coggins_l_Page_078thm.jpg
45d341491d07cec9328cc57fe9fd2c54
fd04d31469f8b7476c0491c790953820f21b89ec
632684 F20101111_AABSSZ coggins_l_Page_051.jp2
bc9337976520ba6b052c58204e4b2696
53c3f7995fff22277f03457403dc4b55929ce1dc
505182 F20101111_AABSUC coggins_l_Page_082.jp2
282013fdd6efda7e3707c1d8ad212d4b
0b37a633b38e90de4d59d3d185d3b892409dc886
834281 F20101111_AABSTN coggins_l_Page_065.jp2
1fd6ac7d474fa4e1abb457cebcada29b
e0fed50bbbc7491f4680de837a91407f134448b2
8263 F20101111_AABTXF coggins_l_Page_090thm.jpg
448ecec583732232efbb6350689fa1d3
bb2186c9b25cda01b2cdd090f77462348c7ab59d
4553 F20101111_AABTWQ coggins_l_Page_079thm.jpg
6c2c426edd49da4df9fb926ab205bc7e
4cf5d7fd6cd0def78a5367f93b9c7f8c07d419a2
1051938 F20101111_AABSUD coggins_l_Page_083.jp2
5e29e8f200f2c4af67fc2e4789468962
75d07879cda8d2c576c8d3d45187fb9f17a751db
1051986 F20101111_AABSTO coggins_l_Page_066.jp2
b562c930956326ef0cb81ee8a305852a
c125dccef42eed15a4f952d4185b585291ed41ac
34193 F20101111_AABTXG coggins_l_Page_091.QC.jpg
ef4b6828f2b0fbd90aeccb9a3c42d9db
71766bfbf32c728a7dbfd6b5d3fad9738332f2f5
4187 F20101111_AABTWR coggins_l_Page_080thm.jpg
4630fdcf45246de01902c29fae4995a5
0d6cf2def4569db6a9cfde49e2e853def4d6a9f2
1051933 F20101111_AABSUE coggins_l_Page_084.jp2
04da70244a5b359c1d102fb67c0fc6a2
1d6f0dd8445c201072d32cf90aad9395aefae3fc
F20101111_AABSTP coggins_l_Page_069.jp2
fc4f2d3198712c3c9af05f06f4d643bd
2ab9c0f7f1f50ccff9ccce17c73e81a554ffccc7
30877 F20101111_AABTXH coggins_l_Page_092.QC.jpg
ac6f41b9e0c89b711d65ee0863f29050
90013717408dfda14e9845197c52651d7619be6f
16081 F20101111_AABTWS coggins_l_Page_081.QC.jpg
4800c0e93c9c2b85b425648788a7079d
159833a9b648cd4fad107641dc4ad902f11a8296
1051975 F20101111_AABSUF coggins_l_Page_085.jp2
3ef4821fce27fe4f3b766123e3925c5c
6f20f2bbb1a624eed99b41dac812ad36474c42ad
1051961 F20101111_AABSTQ coggins_l_Page_070.jp2
8fdf1ebb0bc833720b8756d77f553a36
f51537ce9b2d12743f68860f8cbd8694a44f4d73
7814 F20101111_AABTXI coggins_l_Page_092thm.jpg
1d3053a8bcdd80d16beb21c50accc8fa
f5b4ab957aeca721a9e10b371fd8c1506e4e867b
4485 F20101111_AABTWT coggins_l_Page_081thm.jpg
ac6c8a0b2a9711e833f4632f5af3b6e0
d5f7c0773dbe723873de6943079252d8a445991c
25271604 F20101111_AABTAA coggins_l_Page_080.tif
c618c62b6f68ea86caab207fdcfab18f
15522b5d0e065c66d4a3cbf9d4cb3c2a5ede731e
1051969 F20101111_AABSUG coggins_l_Page_086.jp2
bb08f626c995549b4fbfa075e1772ef1
628644896242758d4e90defa1c406c013005b132
1051973 F20101111_AABSTR coggins_l_Page_071.jp2
12de4c8a68fd1b499673c505d80b310a
7fe2e8608aab5d641bdd04f16de781ae5444d9c1
15043 F20101111_AABTWU coggins_l_Page_082.QC.jpg
39813a0cdf8d975ed8b3d45e38482254
b0cd7354acc609aa6c32c86de2083b78eeb0f970
F20101111_AABTAB coggins_l_Page_081.tif
22cf36aa8238e48d4394b90f4589bd56
4a23dcb9ed5e163ddb653613b725ab9d14e0e396
1051890 F20101111_AABSTS coggins_l_Page_072.jp2
bdb21a952e18a10cb1532cd73a0c6c8b
0b9222dfcd7f95644f937c0c805cd32ce89d0689
25402 F20101111_AABTXJ coggins_l_Page_093.QC.jpg
266ce4d578d624929e45a88ab7eafb67
8fce32002f5b10edfe02a68a4203008941e8adbb
4332 F20101111_AABTWV coggins_l_Page_082thm.jpg
46ec9d41b1a318278407a2c3c601fd57
d950f7ce899115e40aedd561ff1fa2a1134c6662
F20101111_AABSUH coggins_l_Page_087.jp2
4937b4bdac5268526feeffebb900fb35
4f35f5b0c03f389838c392e0ddd850897acf9c42
139576 F20101111_AABSTT coggins_l_Page_073.jp2
3ef88a7ef92ed8340c6ba2b062fef341
a1c1ebf17cf78181f3342e2dd0b0b39924ec885e
F20101111_AABTAC coggins_l_Page_082.tif
a1a226b17930640251659a2541df89f0
21eddd834f50bedc43408aef154e5dd618344544
6935 F20101111_AABTXK coggins_l_Page_093thm.jpg
11c7ddd8a6c9f63311e69cd48a693610
8fa8905dbc3eaad9650c360b9412054bb52e9bf2
37614 F20101111_AABTWW coggins_l_Page_083.QC.jpg
5935406eb60c0f80f93ce95c3a90676c
56a69e039442b05672ade87d1e6126a2e4462a57
1051967 F20101111_AABSUI coggins_l_Page_088.jp2
2ba64d82f61a520a58b511d4cac9a336
f72b9d60920d0f852e2595369fab1b75fadb5e9f
146385 F20101111_AABSTU coggins_l_Page_074.jp2
a1ee88018c655de86c5ea5a306416027
6c3bebe38e4f04c7dc33978ed93ffe7efcbf918e
F20101111_AABTAD coggins_l_Page_083.tif
2a0ca72343130e3d917369c36a8111d8
f6957c09c47582f9347b72a0c5f380b41d8b58bf
5900 F20101111_AABTYA coggins_l_Page_109.QC.jpg
5d378d0f88b00a08b0c081bcbfdb2b81
6a65baabc24e0d9fa24b9a7633e878ea8e276c46
29059 F20101111_AABTXL coggins_l_Page_094.QC.jpg
12b5729d33a1ef58199a2850383399e8
d93363176f384c9b6e05f662c246e026511c2eae
8989 F20101111_AABTWX coggins_l_Page_083thm.jpg
534369f65c149a18770a60422aa59811
2ba5ee1fc04124b6696620a3e1af3f572eaa5741
F20101111_AABSUJ coggins_l_Page_089.jp2
820db3d5c4bfafdeb6a6e3c07681c7c3
9542144302b3c7afec237cdf780c7366e2499713
118190 F20101111_AABSTV coggins_l_Page_075.jp2
58369b7d4ceddf07d34e515393a83fe8
9826b5c4ccf9a49f410d117fbd236ca6d901a584
F20101111_AABTAE coggins_l_Page_084.tif
f426c710f1abe70fb5f3c0c20b5eb83b
7c4b00bce36342dea08a3c04e121429faf46b5b5
18067 F20101111_AABTYB coggins_l_Page_113.QC.jpg
8acc1a5c4761b4c65c580696dbadb587
4cd472bd68db8fa26c74bc4248e6669f21b3bc10
8836 F20101111_AABTXM coggins_l_Page_095thm.jpg
0e3236eee69e1f258bf48e169b0cc4a5
5b31645b4569c085c17d0e486993f79d3a6cf327
8846 F20101111_AABTWY coggins_l_Page_084thm.jpg
95976dcb6f249ae007c6e9a29a5b4127
4ecdcdb1200bcb5c5be819e72cfeaf8408f8d6ae
1043106 F20101111_AABSUK coggins_l_Page_090.jp2
51f92c103762c0fc3562ebd48423f3eb
acc3fb92e796dc3a63b43394947bbcef91b051a5
140262 F20101111_AABSTW coggins_l_Page_076.jp2
0a2d3267f84ca58b8d1245c9f8032432
f25e80e9956e2862da0a76cdeaf74acbeff1230d
F20101111_AABTAF coggins_l_Page_085.tif
b45cad4babb4cbd0d26c4b883445f38b
d0e73387e4fecd0872a04cb2be692593d39942fc
5695 F20101111_AABTYC coggins_l_Page_113thm.jpg
55735f39e0137e0037311a4683d0426f
a2d88ddaabae0ea635a5d9f05c714d13013312d4
F20101111_AABTXN coggins_l_Page_096thm.jpg
6ae737bd88d6432f027b96eebb453d86
257575d6de247a6d211fe7b330f15b87a8a27a3c
8304 F20101111_AABTWZ coggins_l_Page_085thm.jpg
d7c7dff9e5b8d52ccda9f3e86257e17b
e18ff69f0ff06b662652c3ae8a5c0d4863a42a8a
160970 F20101111_AABSVA coggins_l_Page_107.jp2
0e13c72da46f873b79c1ecf1f7ad9ab2
a1f787c791149300059e9f58b56ad610e9f2f4fd
1051943 F20101111_AABSUL coggins_l_Page_091.jp2
d52a19f52c8c3db338547c2d95822b89
5cbe8a7d0dcd401fc884c06fd82801193b4c8962
818137 F20101111_AABSTX coggins_l_Page_077.jp2
7385eacd8160b71ea45c301c97bdfef2
1829c52f7b6f2eb33e2e2a0c9f0a1b85e7beb71a
F20101111_AABTAG coggins_l_Page_086.tif
7f86b8e0027329d35d9648af4dcf22bc
2ca58b7f63ec645c9d35de3fa69e8bf931395a4e
16781 F20101111_AABTYD coggins_l_Page_114.QC.jpg
c7e9ea714c1ce8cb15af0f38ff7c0571
16c992b49f05d21274037ecfc5c9322247e4fbd8
35944 F20101111_AABTXO coggins_l_Page_098.QC.jpg
afeb963de5debca6cdf4d1e0c0ce1646
38ab1b23b6ce424f486c6320bfae8b64b14ef8e5
149204 F20101111_AABSVB coggins_l_Page_109.jp2
ea8083035bdb24008da13172c5aa7f24
71b4eda5141875d9cbabafd078118542a7061816
982440 F20101111_AABSUM coggins_l_Page_092.jp2
6d06b4bbf68936b70bec31aac9633154
28c28b79db56643ec9f0a646c57b49733dde3155
299570 F20101111_AABSTY coggins_l_Page_078.jp2
e0f87fbbb411797e3ecefd2845f32afd
bcf48751a0db6f4a7cda6886796fed82524f1e3f
F20101111_AABTAH coggins_l_Page_087.tif
a997c6cff583a22eb147e370f6503e59
5e6abf52d7a8ff7dbfd0398951a40f4a74900922
5391 F20101111_AABTYE coggins_l_Page_114thm.jpg
adaff6130b07b189af6d4f10d16d5231
afda425758ed8c64ef2ecb1671df9c20e3cdffa1
35240 F20101111_AABTXP coggins_l_Page_099.QC.jpg
7d6488a4993b155e858dd7c107ce8d31
03952afda0095965835cb0aa99c5c7e3b7ff4dbd
767978 F20101111_AABSVC coggins_l_Page_111.jp2
8f79913e814c5fc2eedfe1298d784c8a
536ee81477c6a306c20a429344747946ce205eae
939327 F20101111_AABSUN coggins_l_Page_094.jp2
f96ddf18ff39d7e4a670e71d6a532b97
748881986d36e3a9100d557bc14984ab040e6c88
657797 F20101111_AABSTZ coggins_l_Page_079.jp2
8ee160ccf008c057dc7a6d7902c668ac
2429d48e60b4ed67184a4b1101199ffd2bcb2c55
F20101111_AABTAI coggins_l_Page_089.tif
e08435004da66a97bd39b4a113391714
9598339a0991f898c3b08203f628d1e1744d516e
15562 F20101111_AABTYF coggins_l_Page_115.QC.jpg
e998f1dc5ecd0209bfcc638c793c752f
61b095e0b28ce0a04568e45c95bc052e259de8a4
34413 F20101111_AABTXQ coggins_l_Page_100.QC.jpg
945cb5f84a3c25a1c87891c7b392e056
4e0a337c31f41e3360077927d9c20b5b3cfc9021
931127 F20101111_AABSVD coggins_l_Page_112.jp2
f7acdd6722e5da1bbef8e428cdc1c4c6
53c3d25a405c219a6343c97c79edb9df25a1b9af
1051951 F20101111_AABSUO coggins_l_Page_095.jp2
c675a5e7af00e4f7042a3ef21d1879c8
2598d130194b57bc9006bb1566537b56a7964df3
F20101111_AABTAJ coggins_l_Page_090.tif
dc19635781d074e5716ca6a8609ae8c5
a0b015867ffea9ea4d4eea9e4594600ee3fb1e00
5001 F20101111_AABTYG coggins_l_Page_115thm.jpg
a9cb296e3c28bdcafdbc76365357c0a4
78a007262de93946e13a8f174d33d2a88eda0c44
35990 F20101111_AABTXR coggins_l_Page_101.QC.jpg
9a68e60d547eacbf745c79ab1aa52853
9ba2838cc3530c92919e77a884750b77f36be389
711236 F20101111_AABSVE coggins_l_Page_113.jp2
594c9c9bf97bfc8512d9e929e56d429a
150abfc02316bfb0dfb6d2198a91ba39d22a59d8
1051964 F20101111_AABSUP coggins_l_Page_096.jp2
f8a1dc8b474975292b65029662a84c38
92efdf5d1c1992b272117905e8eb435a3f0667b5
F20101111_AABTAK coggins_l_Page_091.tif
ca88a302c84e4fc0770a390237a06a59
7545717f411c2673d4b762062c51f3db482298c0
10005 F20101111_AABTYH coggins_l_Page_116.QC.jpg
b8788e7cd571acdbb639984f7f2deeb5
362f84fa7f046872eb9aad7900202e69194c33be
8542 F20101111_AABTXS coggins_l_Page_101thm.jpg
1bcb64eb374d1567b77949f6446e6c80
896f72347b857a5da5d9a993dada4dabf857e5a7
631970 F20101111_AABSVF coggins_l_Page_114.jp2
abf445e32ec9cb5dce55926e08d76722
71698e9a862dc150b9941e592eb4a4e1971f7a71
1051984 F20101111_AABSUQ coggins_l_Page_097.jp2
cbaa9313bc2e5bdce586f3b371409dbc
50f5b7ab7bf5d4b577892e7749b9e2157497d99c
F20101111_AABTAL coggins_l_Page_092.tif
c9c2b7091f742b9281d45c169130298b
e8029983b73cb63c99122fb6a6ebbb2a67b833eb
10886 F20101111_AABTYI coggins_l_Page_117.QC.jpg
b2178a7bf802ae213ce209747b787d17
a59eaecd4c6dbd3aa6d3207c054c371001a42fa3
35585 F20101111_AABTXT coggins_l_Page_102.QC.jpg
74dc8e4fc6d149a10db16ee08cf933d1
9743c1f30cdaefd704ae9600ac0a42cce428e848
601912 F20101111_AABSVG coggins_l_Page_115.jp2
46b2b25e27772efb2953d9a2f9f31b84
2a21d59b381a905eb5736021a71ef8e412c3f4d8
F20101111_AABSUR coggins_l_Page_098.jp2
3ee51c0405c238097430bf75d9a480e0
bcbf1e27fe79cdde12fc989ebf50a536cd593bc1
F20101111_AABTBA coggins_l_Page_112.tif
95fb8d73a45c21d8fe14fd8069555c6b
dd18d90625a29a0cb6a41eb40f4eb203d35b6dee
F20101111_AABTAM coggins_l_Page_093.tif
754d285114eb6df374d90bd979aa161a
6719bd4f1d75ebc44fd286272faac10608d23c4d
3285 F20101111_AABTYJ coggins_l_Page_117thm.jpg
7ce3e3f2a7f1e9ce8a26d86d245a5ec1
7836965407a8c5d6b4bf1445176cf193a675eeaa
35389 F20101111_AABTXU coggins_l_Page_103.QC.jpg
4a0a11bb95c02df8471f1408ae14720f
b99da4ef852a43cf79067a13bd76097c488d04e0
364993 F20101111_AABSVH coggins_l_Page_116.jp2
0f7a005f2c37d9902317c986dd7c944c
a17a12fa557aca4c41c7fb82a5999e6f896bb302
1051937 F20101111_AABSUS coggins_l_Page_099.jp2
1879023585ce22f42c126eeb6984259b
9c522ed77a16207bb0cf134d626ac4a331c92db0
F20101111_AABTBB coggins_l_Page_113.tif
cb7abf245d1640c8ae71b09dada3d1d7
64c7bbeb7211c312018564a9d90f9c9a41ce3c1a
F20101111_AABTAN coggins_l_Page_097.tif
fe0b095b981f45efb728b754e817bc93
75b7de62b34229c052aba78c4875af4afd986995
8544 F20101111_AABTXV coggins_l_Page_103thm.jpg
2d7fa571609e6cecd7fe8c58d4c59bc2
6abdc36f6d250033772f2998fc61662075c33425
1051940 F20101111_AABSUT coggins_l_Page_100.jp2
62c6a718dd45d39697cbdebef4671822
1460c7aebec8499edc7bb65ec9ac606bfe44a2e9
F20101111_AABTBC coggins_l_Page_114.tif
36d9d53efbeaf18cf3340e708f1372c2
68356f55bf63ac37a5231564126313d805be1710
F20101111_AABTAO coggins_l_Page_098.tif
7505bad43eb16026feaafb8b4c690828
a6dc2e574254de0ddf1aca2760d22ef78e1ea340
15793 F20101111_AABTYK coggins_l_Page_118.QC.jpg
90477402c05e5e24dc3cee2822764132
e30e65e214d96974d9878f2de8b86d4bd87efd22
36839 F20101111_AABTXW coggins_l_Page_104.QC.jpg
e92e18bdf4497ebd68b72b89bf0361de
01a882a04151149d2d8a4eb032babf5b98ba4ac5
314620 F20101111_AABSVI coggins_l_Page_117.jp2
fb4a754a9de401b63d410ddd75eb6628
a642605568bc4005f722340ca27e4e5e5e79d15f
1051965 F20101111_AABSUU coggins_l_Page_101.jp2
ba98f9ba684275fb641faf0c258a5cd9
eafe2bf48bf1641202df09e0f2828de086a5aa0f
F20101111_AABTBD coggins_l_Page_115.tif
7fc736dba572546b55ab6b882b1f7696
6edaa2c6ef369785d47d6668c7528644c512f5b4
F20101111_AABTAP coggins_l_Page_099.tif
88c85540bf5e71e868efda02f0d39da2
8fe45ab39ddcd782b16b6aa7d902c81476172bf5
6134 F20101111_AABTZA coggins_l_Page_130thm.jpg
a72bc0d0e465746b591ae5a1022ed4d7
46e67caf3f9437aa39004a2ccd9c0c1da94835b3
4898 F20101111_AABTYL coggins_l_Page_118thm.jpg
f73f526b0bb773ac156c227981ab017c
2c0af56f6b5107c2648dff6d896e04d73d8b3696
6021 F20101111_AABTXX coggins_l_Page_107.QC.jpg
faba9ecf1be18c58689b58fb719e2a1f
c9c3a7345e1955dbfda7939b36de1938ebdf5873
462720 F20101111_AABSVJ coggins_l_Page_118.jp2
00c51751031c8af8941593142d94391d
a7590f5ad20f7090fe6c010951328746d00e607f
1051946 F20101111_AABSUV coggins_l_Page_102.jp2
6ce1e7b00c73fa498fc1ff6616198561
839377764a9793d104b56cfcdd618c2644fff4fd
F20101111_AABTBE coggins_l_Page_116.tif
bbd464b49bf3bdc029f955dcadbf193b
e298563b9561870339f0794333202ae360659f15
F20101111_AABTAQ coggins_l_Page_100.tif
5441f77c0c1824f7fa0da48cd8fa2d3a
e343fe9af051aaa67922cdfb8ec2d170831d1ee9
13112 F20101111_AABTZB coggins_l_Page_131.QC.jpg
347731cd738ab9cc9e8b866428be2af9
babfc134dee7ffa92fe06abd0f7ffe7effe0ecc2
14821 F20101111_AABTYM coggins_l_Page_119.QC.jpg
08b6d2001691d8d4f3f5b2cb32f8459e
6dd6d1b99e0089aba39e5f10ec69cb7c55fbd947
5847 F20101111_AABTXY coggins_l_Page_108.QC.jpg
384838978ad676d4c76c15045eed0ad7
894372fd5921c6e1fff6e45cb0b3423c6ac15c0a
421388 F20101111_AABSVK coggins_l_Page_119.jp2
d75a9ced57018b16bb145574b4c80a46
5bc025aa1dc41122154639d524da379bb798a707
F20101111_AABSUW coggins_l_Page_103.jp2
052ba3b6eb556b336183c0ce4b2f0e33
f0f4e8f851d707df2f694683433468950c60be8a
F20101111_AABTBF coggins_l_Page_117.tif
e08b0f7c67266b8947dad85d9f11397a
92b5d1ecda0d4e8c8095ee1171b0b690409223c2
F20101111_AABTAR coggins_l_Page_101.tif
14819aaad341186ab26dbefb8ee9434e
f18e5b7a7c184816e4a70927e5133c98aaa099eb
12593 F20101111_AABTZC coggins_l_Page_132.QC.jpg
77ffe364c7b27469d951638b75d5e09d
3d3ba1b369f9d788b7597496689e2a39018f2d58
4773 F20101111_AABTYN coggins_l_Page_119thm.jpg
aa6520fafb7cd45fac07889ff333a603
b8f48c20ac0a0935cd219b36e621184f891f30e2
1532 F20101111_AABTXZ coggins_l_Page_108thm.jpg
19c670aca43ab924812500508d028bcb
a99a89bff7df73ce7066395af30b8ec5bba6fdec
826109 F20101111_AABSVL coggins_l_Page_120.jp2
d6e0565744e742a96fceb2e20b57dba3
33fd4c52c779c00c3ae91bfa897eca9bb85ea8d5
1051945 F20101111_AABSUX coggins_l_Page_104.jp2
597299fb05639ac3a07f1d1a59111ad3
55107b9180e60356598e30f663cd15c1bc10830d
F20101111_AABTBG coggins_l_Page_118.tif
b127ea9155710e520ad144c5c5cb9b00
1485763e638787e8eacb551101439491444de230
1051970 F20101111_AABSWA coggins_l_Page_137.jp2
d2c59bec93633ce6aab83001127d483d
c8caf2ae20cc5ba5831cd3172c882601543f806b
3795 F20101111_AABTZD coggins_l_Page_132thm.jpg
64a13a7bd86ef6c5cb0ae1fb9b048c50
5e335264fba62b92d9bdca80b45aabd83285537c
20489 F20101111_AABTYO coggins_l_Page_121.QC.jpg
2aae76df3ef8c9da0a57fe1d47f00d98
8cf07435cf6036cb9b2aca577170ec5d43cf5e3a
821032 F20101111_AABSVM coggins_l_Page_121.jp2
518e2cf312c069a9ecb4c0a3632207b9
27d121a306cca5545b23c2de3a527b466f215912
F20101111_AABSUY coggins_l_Page_105.jp2
23f4af572378de972e0973290e9e9e6d
a5378d38e9b395e8b9d5e8cfbc5761e6ec2ddaba
F20101111_AABTBH coggins_l_Page_119.tif
aa03bab48b29e3e353e32883b1ffeb09
6341c353abc1f11429efe90605ec40b523203221
1051976 F20101111_AABSWB coggins_l_Page_138.jp2
92bb19418d06f11294dffe05c24c7390
45b4490b188b281001e64ec8357acd0e8d406527
F20101111_AABTAS coggins_l_Page_102.tif
a1d634557863c8cea037d36210bdd459
b52edf262c957c25616d812d937f1f29f519b70f
16414 F20101111_AABTZE coggins_l_Page_133.QC.jpg
d45d1d8a6b71ba5ec331c004290e9ced
b42c7f56c504d9f7934cfaf415d92b163b2656f5
5528 F20101111_AABTYP coggins_l_Page_121thm.jpg
8f8108f4bdded3e0bb1f47ed518a3c04
ab0aa6168b97aacad6b4f09ac65cd6d42a532623
811726 F20101111_AABSVN coggins_l_Page_122.jp2
c65921261c7b4a3a8350b31889b81826
c0342f36bdd309925a2c2eae1c9564d732bc5c23
F20101111_AABSUZ coggins_l_Page_106.jp2
1d4ade993c292b5f878dc45836e0b90c
3d66cf5299c3320d2b7f491cbbf92e4f5db8f303
F20101111_AABTBI coggins_l_Page_120.tif
8ef7ca66445d00e2ddaf0dc5ae97ed16
8a9e25296ad65e987edece7092cc7f29f83c6f3e
F20101111_AABSWC coggins_l_Page_139.jp2
394e7edaab318cb1c949aaa29381af7a
69c39b3495d589cb3fc3c803c27a0e592379170f
F20101111_AABTAT coggins_l_Page_103.tif
23501ee766b2a080cf8ac01d8a7d0338
5d1ba3f8979594bb67a9fa4d8a32a2933bdb0ca5
4512 F20101111_AABTZF coggins_l_Page_133thm.jpg
7afc39bd5b2442b6d08b3081e2259fc7
61532293ef2f948d0ed2a454a1be02998812d93f
4605 F20101111_AABTYQ coggins_l_Page_123thm.jpg
35a2719116add3fc0b2393ba6182d429
5cdff8192ad941aa9fb271830174b38a3aa75703
700477 F20101111_AABSVO coggins_l_Page_123.jp2
3f6db323b18da3c683b3771dca8e211c
5cca59d0166981d108388e7d22f6812f3990d63c
F20101111_AABTBJ coggins_l_Page_121.tif
10b73b6c3dac34950c2d12c424a1658e
a1e31ea1bd1f195290d26a062f25b38e8ac048c4
1051913 F20101111_AABSWD coggins_l_Page_140.jp2
26444669582a95711848d1c03b701ee0
0fe802fe6a52b9cbf9521aff9c893510577c1732
F20101111_AABTAU coggins_l_Page_104.tif
aee7ce10a6ebca79cee97a16f3837e12
7f5542891bf405471b49e755227f33a335d0da06
35517 F20101111_AABTZG coggins_l_Page_135.QC.jpg
7f911118f1bcdf39da15a68f2fac48ec
165ba39b17f5e565c7e5fd080cb1ad81eef2fcc7
15630 F20101111_AABTYR coggins_l_Page_124.QC.jpg
a46577a576a8cdc6963d8e888664224b
10959b045547534c60d58d4e6418b3fb56870434
453848 F20101111_AABSVP coggins_l_Page_124.jp2
8c5ce3a0e87bc3bddabb1b72bac91b11
128de6a015c85c20a5b88bef3c6e8ebab61f9123
F20101111_AABTBK coggins_l_Page_122.tif
524d4fef85780760d420fdb55565092a
15c09bb7d14f160cd0efd9bc41af1e490f9be928
F20101111_AABSWE coggins_l_Page_141.jp2
7cd30681e4a042edde766a393869ba60
47f92fff443d4714c4c973d28ad450e4047310f2
F20101111_AABTAV coggins_l_Page_105.tif
12d7446837953aaa978ec71f24428d76
aeae7736572d01c82a8239242ea804b27c42e559
33792 F20101111_AABTZH coggins_l_Page_136.QC.jpg
6cd5c3b0ff1aec11c380c46bc711db65
312d91845dd12f77d08bb074ce806b91d041e3c9
4865 F20101111_AABTYS coggins_l_Page_124thm.jpg
a17519d60609bc343636097e7fbfefff
6e71cb0ba6fa0c0e8b7da51538323b9f07affaef
823241 F20101111_AABSVQ coggins_l_Page_126.jp2
b028730573b23a2a9e9b04a6e1d880eb
8b9ea3377d11fb8eb80142e5a455a9219d6b77b4
F20101111_AABTBL coggins_l_Page_124.tif
24e28fc1e5d208a5ab6e3ef681965392
eb8498ed3f89728641b66c57b8010523b044fe3a
1051956 F20101111_AABSWF coggins_l_Page_143.jp2
5033680ea218931720d8a7d14ee38af3
7cabd5c83030d0c7427f89cc6097082ba9079646
F20101111_AABTAW coggins_l_Page_106.tif
f780999fbd35ad8938775cedc503018e
fbd6d19a86241f8a4a47a695b51ab4ab6ffff954
8193 F20101111_AABTZI coggins_l_Page_137thm.jpg
bb91d7f5443ffe02bd7ff09822fc2f5a
afb04b02ca7f353546c52ebae390feae7529c044
15100 F20101111_AABTYT coggins_l_Page_125.QC.jpg
ce1d8060d30bd5f2a62a760cc056d22b
84ae9b356910e4eb752be5e7707e35836e6d0613
819970 F20101111_AABSVR coggins_l_Page_127.jp2
9d0132751952d88383ffe676ebcbbeb6
f7048c573f2777482d172d620a54ce47ed78e917
F20101111_AABTCA coggins_l_Page_141.tif
00905a975f246de9d706f947897e2726
553d2c1d275e8535761aa62e51b7b9f38add059a
F20101111_AABTBM coggins_l_Page_125.tif
f180491b17e0145da72126ef92707fe1
0ae33c4006ed9eb5fde909ccef28f5392d35ad47
F20101111_AABSWG coggins_l_Page_144.jp2
562feb85359159daa6098652865f44a1
165055fd4e5371754534596a0bbe5d47d5b859ef
F20101111_AABTAX coggins_l_Page_108.tif
5b79ce0f05f665a4c7ff673505f2ea7e
9147d315d8a4bf1af221e2413ca8097c9ad457d7
35471 F20101111_AABTZJ coggins_l_Page_138.QC.jpg
f8f9b6acb80c1b2ecf31c52f1def938f
d01965a4636594bf4c1b3923e341264913a38f0b
5545 F20101111_AABTYU coggins_l_Page_126thm.jpg
5b979e8761a14a7548e35d417c761b56
36accd9ce4e8deb9246aefef0ae5e4ee4b621425
736517 F20101111_AABSVS coggins_l_Page_129.jp2
71633f9e8702bff0954bad80815fc683
6199ac3612bcad41a7f925bacc97ecf6b1add179
F20101111_AABTCB coggins_l_Page_142.tif
87a1c581b1619aa2cd2fa932de3a3b43
735a7291f8b1e5bd038aa3cdda7d0725113161be
F20101111_AABTBN coggins_l_Page_126.tif
46e1612b44d64027a6cd89cdba41912e
844fa0df42ca581190baaf7e0b7b3c4a028da8eb
F20101111_AABSWH coggins_l_Page_145.jp2
fb7a35f50026a83cab764f881f2496f7
7f8b09f39283e4994c20287610f4701f9fc5a39e
F20101111_AABTAY coggins_l_Page_109.tif
e2f3a95fbef734cbe3e4030e9ae1e545
665f082065e98afd8cd8995bfa42b9be9ac0b02e
37000 F20101111_AABTZK coggins_l_Page_139.QC.jpg
96dd80801cb8268194bbdd4d3d96a9a5
9385bcb5394c966649b59c09ff9ddefd2d93a35d
20506 F20101111_AABTYV coggins_l_Page_127.QC.jpg
f9cd8e6a2af12c3e6db8b8cb084404fe
684dd2dc6e7e72d75c1854d9bd35a31802c67841
958509 F20101111_AABSVT coggins_l_Page_130.jp2
a77fb1b71bb7cfcccf70eecd860d73ec
a150d1f59ad9ae978c20d8ceb7dbab25cebcde79
F20101111_AABTCC coggins_l_Page_143.tif
fd183de63100aecd3fcf5087d19d1ecb
3c0e8b4ed36e2ddaa3de61b79a4611918963ace1
F20101111_AABTBO coggins_l_Page_127.tif
1f9ed66aa78f8e297fd761da55d716db
028479f7612d00ff6d335eab5ef59b5cd774eab3
F20101111_AABSWI coggins_l_Page_146.jp2
1e05ea0c7be16083e3a84e2f084bf289
b236472177fa7e4eb304e0f076e29da60e9d3529
F20101111_AABTAZ coggins_l_Page_111.tif
43f0c49614253e29c072ca50e884805e
9c665fd69b6e5e55e134902fc139896e5830113a
5515 F20101111_AABTYW coggins_l_Page_127thm.jpg
c46bc838af40a8be52a8585c62f3f1e7
20db0230ac43f369fd5c8c2af7992dd133158797
398490 F20101111_AABSVU coggins_l_Page_131.jp2
8d3bc611c6058bfb0f88d76032b92217
cbdef73ffffc9a9189a4a718c5f8458e667589f0
F20101111_AABTCD coggins_l_Page_144.tif
93f7b45f788e7c6a72829096b0f0af39
b8f1984d9c6ac43964564e1e7531071557cff3fe
F20101111_AABTBP coggins_l_Page_129.tif
7c2252f11a8e4e8707ca525b0e96bd23
997e714c1e47fc1d86393e2ea0b485b7fe8930ac
8810 F20101111_AABTZL coggins_l_Page_139thm.jpg
4b18148131bb96a8eda71e359e1e78d1
43242bfc8167cabac1fca73d48d0de1cecc1b46f
17964 F20101111_AABTYX coggins_l_Page_129.QC.jpg
cb33aa94b917235839db09d2edda1324
4972b1f65fba8feda49cc1c64e8ccfc873987a7c
379716 F20101111_AABSVV coggins_l_Page_132.jp2
c50f029c281e2be14ee7edd1141d5c39
c40330f17db139f26661564fa70e1a25e67574ef
F20101111_AABTCE coggins_l_Page_145.tif
b259efec41e7a6b7103c12a7decf963d
3784f2189c5581e0eb27eef14b4599ba3b6633b1
F20101111_AABTBQ coggins_l_Page_130.tif
f7241403ac6313b52a125ef604d9dcb0
3a5ccf3bac396102d0b71c974ae041327c0892b5
1051972 F20101111_AABSWJ coggins_l_Page_147.jp2
1f3fefc3dc1a72de231496cb500f5cde
5f6d59385eb8e3bf5b801f11f2ba3d5b053ef9f6
8969 F20101111_AABTZM coggins_l_Page_140thm.jpg
70026f16b01ba5cd08b132a80dcc212f
51e6e41029a4bc99e76655b4b4bb207b1b9783a4
4673 F20101111_AABTYY coggins_l_Page_129thm.jpg
ad17ea56f4156aaa056ac5c034a10460
7926317489bfa59634f64f34146d67c0db82529d
586622 F20101111_AABSVW coggins_l_Page_133.jp2
439443f515aef08a0804f8cd8306b6a8
91e2057be80dbe751aa2fcd9db5720d940fa7d53
F20101111_AABTCF coggins_l_Page_146.tif
f97b89bbab7a4420b3cc0a3efb618393
99ae96735891c4fd018449acc970124eb0f697bd
F20101111_AABTBR coggins_l_Page_131.tif
6b05b29f17c446df1c28d6941d702e77
78a1cdc68355be6e4f036219caa019cfeb7f35c9
F20101111_AABSWK coggins_l_Page_148.jp2
702734e97761161acd5787cd752d2bf3
675262f4cae0d89d8d7e50ccfaa9a1cd9a69dad0
35915 F20101111_AABTZN coggins_l_Page_142.QC.jpg
11a8db94b776473af7f29fb344ad4ad3
576923dfdd1665825bd1e7e299c45b2818ffc46e
22197 F20101111_AABTYZ coggins_l_Page_130.QC.jpg
e5af95d786632fcc93e0c286b6b066b6
af08350164224c147ffd32fde1427714d3385ca3
1051936 F20101111_AABSVX coggins_l_Page_134.jp2
b8045a97d34738116daa0bf032b100e6
6dfe4a356f0448600e362928dd4de8233bda2e3f
F20101111_AABTCG coggins_l_Page_147.tif
16d0928a29cdd4edd82bd8ec92f3eaf0
4c316222cbd7807aa3f006b7473c690ef23a07f3
F20101111_AABSXA coggins_l_Page_164.jp2
b665419d1da3bf7be715c3b588f5f887
4c0639705ca42fe5390f167a29e5075d00bfaeff
F20101111_AABTBS coggins_l_Page_132.tif
dc73cb2e8e6d246d8a37f23a86fab8ec
79ce14831ab596fb375311853b4f84ac693b698d
350728 F20101111_AABSWL coggins_l_Page_149.jp2
f2d7f67a0121c6d77e41e11b856f453f
2d2adab77e9df83210d033177fc403e24c48610e
F20101111_AABTZO coggins_l_Page_142thm.jpg
c7a0b634b1b50fd8f1707a42043b9240
ef067f10d71340dc7d9e92ecdc7da0e6f7f320f8
F20101111_AABSVY coggins_l_Page_135.jp2
2ee9d50d7db2e76aa719501bd9613968
05e50b3f51396f46af65df4881aa099bc5a537ed
F20101111_AABTCH coggins_l_Page_148.tif
7f45784521ca7fc407b0291be1819472
d018284433e7b364dcb3001c096390e8a555ff22
F20101111_AABSXB coggins_l_Page_165.jp2
d17f80c0751b442afccd2666b704e371
59480d04f421213342224953bd9ba4cc9fa40c06
544511 F20101111_AABSWM coggins_l_Page_150.jp2
30866ceb8ea315cf32fe3356bbac5472
01ac243114a546e724b842ae62ff2144ffd17166
9091 F20101111_AABTZP coggins_l_Page_143thm.jpg
e0c6a37492176ea2611bba9b2ca99a41
118d1104956098b8d43d28c33fb673e0098ce1bb
1051944 F20101111_AABSVZ coggins_l_Page_136.jp2
2959167614445bc1a7271ce24a52c1bc
187fe349cd1dfbcd76c3ddf84698c65f4f21614f
F20101111_AABTCI coggins_l_Page_149.tif
0730d343e8075679c08870a50833756f
49f3b5e5cb04d0d4c17e6ed2b048254668e93540
F20101111_AABSXC coggins_l_Page_166.jp2
47d1c127ead7e76b22007f8f2d566024
0dfcabd4489b5c908f2d29171f3108a64edd9bc3
F20101111_AABTBT coggins_l_Page_133.tif
9536b2956b22e51f9d1c7c788cae5c89
e587e8e8ab6e563130e6d5e41fe3b759a6241f3a
969854 F20101111_AABSWN coggins_l_Page_151.jp2
eb7c7742cb256a9ee61c33a63d44648f
5a28eab620d1e9b0641529e17d11039997f12697
38382 F20101111_AABTZQ coggins_l_Page_145.QC.jpg
f551bc925c3f3f54f713d6eab86931c0
2d505c702d15a2fae3abc68c203822dd687ec102
F20101111_AABTCJ coggins_l_Page_150.tif
048063a59ceccc90a2b1df648342a605
fd8e14d1e54b2d9deebc2db71904deb093bbfef5
F20101111_AABSXD coggins_l_Page_167.jp2
4fb10e206f19ffd2d25af334849c9478
8fee05f2fbdc4a746233e07c5fc60e66af572ba0
F20101111_AABTBU coggins_l_Page_134.tif
ea711bcd6d0e6b7bbd1f3fb9f918a420
0c2a6633eb37dfb7d4ae36354662dff3e9537615
689045 F20101111_AABSWO coggins_l_Page_152.jp2
67d791f30f90fc04830d08e5c16184b9
ea20f379ff676eb57567a9ab79b7710997bb7b91
8984 F20101111_AABTZR coggins_l_Page_145thm.jpg
c9c48f7df7d9d8192972a2b2c62c80ea
d73527a7de95193601319f266d854aaff8cd8c09
F20101111_AABTCK coggins_l_Page_151.tif
6f03b7d587371ae2e1a31254db888373
2c691112a8e727b59eebf2089632cb2f19d7f087
F20101111_AABSXE coggins_l_Page_168.jp2
55d9b65b9c2f87df522ec351fbdaf945
4e87953e596a165bf879391d61df06f1a864520a
F20101111_AABTBV coggins_l_Page_135.tif
f9f60abc217d83c23257578159d6a3ab
aaa37da9b04f5dd89e3d2f077998664240b10dc0
682712 F20101111_AABSWP coggins_l_Page_153.jp2
bbfb1ce66451ff417022d59d474495a5
f88d4b9346d873fe982e1ec18214f879a9dec5a2
36551 F20101111_AABTZS coggins_l_Page_146.QC.jpg
2ee0f251e38b6a82a8d4930d39ead1aa
f44ee87b6eaa19b070d7c3a9c7323fee44d574a0
F20101111_AABTCL coggins_l_Page_152.tif
c32129a05f58966d982e0c6b5a2174f4
136619b8178d2a204ed550afa42ea48010e3d9c9
F20101111_AABSXF coggins_l_Page_169.jp2
852a9f25214593ebee7b26f7a5f4b047
d8bb5ee482643378621889544daab5c69143bcde
F20101111_AABTBW coggins_l_Page_136.tif
bb826ac22d39698fdba9417cbcfc7f6b
a77b6c08f0a2617a6b63380f59b2cc438fddafdd
656401 F20101111_AABSWQ coggins_l_Page_154.jp2
57de11646946694c7f3958f7f2be71bf
4373633285ac78f99bcc08398a91ce47249da3ff
35139 F20101111_AABTZT coggins_l_Page_147.QC.jpg
86f8205f809e63099c3d995e23de0e58
d1bc3ca24f91237d6ca705f28d23374cfa9451fb
F20101111_AABTCM coggins_l_Page_154.tif
551ff79f57c769e28d923ddb6b8186df
e1d317e867cb8dcc39d2411114d429b18603e71e
1051919 F20101111_AABSXG coggins_l_Page_170.jp2
e5fad25e0e67246f99be71b61a2f682f
8d0531471662146b5d81bcc431eff02cfdae93d9
F20101111_AABTBX coggins_l_Page_138.tif
19b55f7c32c374a2dc6f60958c04be17
b5bcac4efaf8f33af70c4c206fd7801b459d24e8
798274 F20101111_AABSWR coggins_l_Page_155.jp2
c4f08cb289b337d55a82513b1744757d
1d50c83e5586428d0148d2c3ad111c43436c4607
F20101111_AABTDA coggins_l_Page_169.tif
9d1f2f98501e076a9f965de819bdfa12
b092684cf1e61f4163f89ab524e994c63075b5e7
35618 F20101111_AABTZU coggins_l_Page_148.QC.jpg
1c8014e5c5cda158aa6e2ddddb03c20a
efe4da0e7098e7ebc54b8fce9129313d72fb04c1
F20101111_AABTCN coggins_l_Page_155.tif
ce3eeb554295bc1c9addf1125936fb68
0fa7975e41985f324e721647cc013225f0974aae
1051942 F20101111_AABSXH coggins_l_Page_171.jp2
972c97e617921e9e9c3a3d7d05d1cf81
c75df57a840d4fa41ca8e025845d28170f9f5219
F20101111_AABTBY coggins_l_Page_139.tif
49dfca9c466e18d0b001b41b5748f3c0
5e7afe36cd425697aba25bcb89318be4cf2b2b8c
833131 F20101111_AABSWS coggins_l_Page_156.jp2
b11190ec43b21a5f7cb53cb35c413143
1675343edbbe6a2c6912d71993e490d2dd63d0d9
F20101111_AABTDB coggins_l_Page_170.tif
2476c19e13de8dd5c64564fdcacee51c
02d0e507a6c17db0a549a52a17f53903b3689984
8775 F20101111_AABTZV coggins_l_Page_148thm.jpg
6adbcd0100060cc72ede97cd89579764
5fa48f2ff54f0c7af205f8c7dd528d2f7697d945
F20101111_AABTCO coggins_l_Page_156.tif
d3987c99b7d76735a03d7f7e9504b618
a4dd1782190bc3b66685106d5d1ac2bb8ebd651d
F20101111_AABSXI coggins_l_Page_172.jp2
57d081c8da11d3bf082792370a1e46c7
f494c030beda93215ee5df122476a434185056fa
F20101111_AABTBZ coggins_l_Page_140.tif
dcb192e442858586e0d5c99456bd2dc0
e24bdf72d1eed32694e00738d6ba607d0c2c9e69
668287 F20101111_AABSWT coggins_l_Page_157.jp2
7ad82ab887cbc6446fc04dbbbc22cdef
e65b3ddde1cb15848d834d8754ace374ec78d8f7
F20101111_AABTDC coggins_l_Page_172.tif
91676fb397f629f0cd2d6c851fc42d83
94bc5c5871b0eebcb84c39af03dfb0a95810b339
17583 F20101111_AABTZW coggins_l_Page_150.QC.jpg
d06832dc8f2af62fec93eadd5402cff4
ee1797aad39597ae4446fbe182f21d62fe91d693
F20101111_AABTCP coggins_l_Page_157.tif
2aa3f50f36a1b467f46effb5a7e66b03
0ade1f8efacda5f71d006e72274a11a6a1ef6ff8
71559 F20101111_AABSXJ coggins_l_Page_173.jp2
0eb85bfa965887bd17320f3030392ab3
d34b1909e19b7dce6a0bfc860fdc4817b0e22b8f
385974 F20101111_AABSWU coggins_l_Page_158.jp2
ca78ac246e53399b9842f4e1b42d8d0a
f93a70771023d7b98ad16aab8a12d54647d91c74
9206 F20101111_AABTDD coggins_l_Page_001.pro
11e1b06e65aac563cf97fec4606fbf8c
e812fec1d18557962ec329d75e65b324a097f23c
20272 F20101111_AABTZX coggins_l_Page_151.QC.jpg
153c2b798e853b36e35ed066e374abd3
69786ad6bfa6ef79b226cbbcbd33401b5398d831
F20101111_AABTCQ coggins_l_Page_158.tif
c23c257648c73dc0d977d4964ba223d9
658077e1ac3bfed38d08df304f5895f2236826e9
F20101111_AABSWV coggins_l_Page_159.jp2
c8a1a16a05cc4ff367d72b6326c266cf
7a01ecced91941a10db645fefb271c167f215be8
470 F20101111_AABTDE coggins_l_Page_003.pro
dd177c08c0add64a4e5522fc62de02ae
900349d20c8f39203dfab7076f4afd7dfe039605
5272 F20101111_AABTZY coggins_l_Page_151thm.jpg
f01053d1830fecea2c6ac7f9f06a88cb
21eca34cd46895e25bfb435480361e3367333146
F20101111_AABTCR coggins_l_Page_159.tif
8bf028bda6bf6ab759076b0de30491bb
336a527f3e48090a4b51606c1da5facf617f9817
F20101111_AABSXK coggins_l_Page_002.tif
dad6eaab2e099ed566ee8e4d366e1f88
f400366f78556f1ee6be009514976273f4beab28
F20101111_AABSWW coggins_l_Page_160.jp2
10bedb2ec9353d67db1c3aa30651bceb
65ca635b7fae93f12ebe6df47ed666ddb5df20e2
52545 F20101111_AABTDF coggins_l_Page_004.pro
ded4cd38055a0057a4cfcbfc2ec6c944
da9dec80d1a3c0a2943f350a81d8226e9a0307b2
18154 F20101111_AABTZZ coggins_l_Page_152.QC.jpg
f9534fc3224a2e6585ee2a42b05cc4f0
8c00ea587a570f40baf4ff647a6cf5c0a44ef136
F20101111_AABSYA coggins_l_Page_021.tif
f4f78f44e7ed7ea742498e53651613bd
15e01b57d74adea35917c3fd944fd15ec93cd100
F20101111_AABTCS coggins_l_Page_160.tif
80663aef766599d273bd64e0916a4d03
d814c0fb4fe47021b638efc2c2486e3ffbc1dc41
F20101111_AABSXL coggins_l_Page_003.tif
a93bc31b25aca7ace8b5e8fe72402bf1
a8ee632f97bf5be24e21744273a14c1dee221cc5
F20101111_AABSWX coggins_l_Page_161.jp2
896a1522bb932c9d82f8b9fc7c2e065b
c7221312b76fcc74db1690d74681ea06a5a4213c
84392 F20101111_AABTDG coggins_l_Page_005.pro
b1ec2ac881e0646838e35ad9dac473f0
33b79b5971521a22d18a093186e952a97bc57503
F20101111_AABSYB coggins_l_Page_022.tif
c75d3f75e9f43ae007a1028bb193d274
841ba8d4420f90cdaf6fb52f032d903fbda60afe
F20101111_AABTCT coggins_l_Page_162.tif
5973e3ec24486cf675c89069ae860a73
4154cabb390de4e3ce97aea9811b57490b7729f9
F20101111_AABSXM coggins_l_Page_005.tif
be1b9f71a17bfa25043a007951e37dc0
a1e431752c100ada366297f4b170c8f56bf2c7c2
F20101111_AABSWY coggins_l_Page_162.jp2
8499e603b212d32068fceddf4297a09c
0b18517f99e2932171b30f984a7fb087ce2db8dc
108277 F20101111_AABTDH coggins_l_Page_006.pro
7cb80008c8baea98e8aa8f270236a353
8d59f338c1f2626eaba0292992cc88163deb7f85
F20101111_AABSXN coggins_l_Page_006.tif
5ddee3204bc8bda835bf0ee1ad6a2088
60f4afde473efc83760b89847c828b01197991c7
1051930 F20101111_AABSWZ coggins_l_Page_163.jp2
280fdaa388da7ec9b3a51cc158fd926d
6b478da8faa2481751620a86644f17fb7f424aa8
73211 F20101111_AABTDI coggins_l_Page_008.pro
20fcf66c80d96e9af0e37820422eb077
e2bf7be032e940389349aa9efc3095e3b3f87578
F20101111_AABSYC coggins_l_Page_023.tif
59fe684f7bebb5c1795b185754e4c826
e8acd153b7b1ac0c40061d0d29ad70fb025cb6fb
F20101111_AABTCU coggins_l_Page_163.tif
9a0520868b98711328dc74def20a087a
53ccdcf1e534671939463299ab3919b6de4cc13a
F20101111_AABSXO coggins_l_Page_007.tif
6f9fe069834e63a151bc84f2c10c5cc3
d2a180f246daad6bb8c775a554d521c8b87058f3
83810 F20101111_AABTDJ coggins_l_Page_009.pro
fee563e7cce4123bc5641a4ff42eb884
a6429fa7f3291494aad3c5e1470f8d110f252469
F20101111_AABSYD coggins_l_Page_024.tif
b53f4b1994ee7ca8181089c16d2b261b
e62de327ed7d83f8342eff8864887ea166c252e5
F20101111_AABTCV coggins_l_Page_164.tif
1fa3b9c6d428bf48e8ce6d376be6a26a
961cdaf76d85ee60d887bdc21d7003d6da093b52
F20101111_AABSXP coggins_l_Page_008.tif
5db0fa6d9dd9a817d565dc700c4db2be
0cea095c0de296090fd7469039dcf58e2924bb11
72508 F20101111_AABTDK coggins_l_Page_010.pro
935da47b7aa724297c6250c368843c34
f5b1a6ece3b32c3fba2873f795e22200b115232e
F20101111_AABSYE coggins_l_Page_025.tif
cfc2347c5d039ba93917884ca093c48b
b4478082444bf3cfad0d0755a7c5c6d72510d0e0
F20101111_AABTCW coggins_l_Page_165.tif
8e724a11be1dc387e3d59f8e1aebaf69
4d2f0568c1822b9744ce72ecadc0f81cb49b0acf
F20101111_AABSXQ coggins_l_Page_009.tif
ad6426ac9ecf4194c3326977f8f5bdee
1ab60b1e268b11c5699458805bc2d01c291b8872
47911 F20101111_AABTDL coggins_l_Page_012.pro
c853cf9a060bb0d2112b0ddb5c0c1364
7355677303ca6f9eeddd0ecde2b16d441c61289f
F20101111_AABSYF coggins_l_Page_027.tif
42be75b73cd99cf6029c39e6198e0f61
4cf4dc178e7ab331261f9e98d72994743b9c3197
F20101111_AABTCX coggins_l_Page_166.tif
1f1a780d5ad16655941eae30f9fab524
8688c297ea433cf4d8eaa20c99d2df8f01bb4f44
F20101111_AABSXR coggins_l_Page_010.tif
5d961d859b68c888a1efcc3fc354ac8c
d88a08cdbf539edf0f16f0a5155d55709efe63a4
53074 F20101111_AABTEA coggins_l_Page_030.pro
b2a0b218ffb2b4ae9b049ce51d26f01a
8f6cbcf1a31ebaf8b6ec61d104ee24e30572c38e
54018 F20101111_AABTDM coggins_l_Page_013.pro
9891043c0042c70f32e5ee6daba0a120
7b7dece549fd54ce771a3d6ed98b1e9e38997eb8
F20101111_AABSYG coggins_l_Page_028.tif
f266d01e476172110c0ec17fc054ed8a
c10bcb899bcba0eaa4d77817897afa6259deb8a8
F20101111_AABTCY coggins_l_Page_167.tif
93c71e6a59db4aff1e97f902986e67f1
3a2f736e3a21bd84a3d39a0df171ca2c16f5283a
F20101111_AABSXS coggins_l_Page_011.tif
6fa1360c80af6025812d468d7508eddb
cb96619a16bc8a0883ae8af55eaf7aa1099d2812
51564 F20101111_AABTEB coggins_l_Page_031.pro
a0a0421fa91f512e35210f0373a0f3d3
217ce33cb52e7d6ca9ccb5da5c4e3e593beb7044
58309 F20101111_AABTDN coggins_l_Page_016.pro
6d5940fd064e22f19f35a4cd30c72ca5
a84b4cc6217f1dfb7eea94ab6a619314b044aa4a
F20101111_AABSYH coggins_l_Page_029.tif
b914247bcd5c4c5ab905d2a447ec7bac
570aaa92e533d2544ecb9a9f87773a85ce8caa6f
F20101111_AABTCZ coggins_l_Page_168.tif
ebbfc11a489341486f78c6fce594c35f
4d9969e3dac5dba15e0411427e537130cc179ff6
F20101111_AABSXT coggins_l_Page_012.tif
b56ae520d6ee8311949801c8ce59fce7
c767f019dc457410d2525259444505ba42ab345b
56537 F20101111_AABTEC coggins_l_Page_032.pro
23c1a11b80dc380720c7088e715ec54d
97201f4fb04be6cfb9498daab2ac3eea4b2ea52e
56472 F20101111_AABTDO coggins_l_Page_017.pro
5c84e144aa72071ea24f3454a6281faa
fb94801edda55457b76b0e3141a9a01b676e7276
F20101111_AABSYI coggins_l_Page_030.tif
35a99c934cecf6f9120481db0267ee9b
08f90536756c9923fab659dd061adc3f021dd3a9
F20101111_AABSXU coggins_l_Page_013.tif
63172c7efbb89066b72a601f12c163db
0ab57e95934d61ae90013b169ff9ad56ee87830a
52672 F20101111_AABTED coggins_l_Page_033.pro
ee06f676e89d27a0668b135458b2373c
1f0d96b0ec5118bf0a047512cfc74dce5b679b91
53475 F20101111_AABTDP coggins_l_Page_018.pro
0352cf84894f3e14916b58e21387fa94
72fbaaad26b7d445d5196a7ac43eb938672c59a1
F20101111_AABSYJ coggins_l_Page_032.tif
6319e154fd2a18e951f7621ec3b68441
c94c938af4f8f1b1348b23dce0d6b0830ccd5585
F20101111_AABSXV coggins_l_Page_014.tif
cb8a4ba0c3055ea802605fd6f44135d7
29555b9aca73c053b093d1413ab3a05257440021
56981 F20101111_AABTEE coggins_l_Page_034.pro
8566091b98de065d03dc5948b82ea0dc
770d31a18340c03ffd5aaf42914fc47d1eb74e78
54252 F20101111_AABTDQ coggins_l_Page_019.pro
cd52022d2fbb9dc8df963e647302f774
c32b153decf37845c4e997c8ce23f0765cdc2fc4
F20101111_AABSYK coggins_l_Page_033.tif
56d25cc7fb121c678384ee1500f049bb
30fd3e3cf4f2ab61e1095362f73bcc30bb289a7f
F20101111_AABSXW coggins_l_Page_015.tif
94a135557cea3a95e96b9dedc7766ba0
a95564f1b619fb21d85040a95d4bbf5f15fc02ce
55303 F20101111_AABTEF coggins_l_Page_035.pro
552c18de479e4a0eb34853cca8272058
e4492f5fff5998efd60630366cad8ce86007be48
55922 F20101111_AABTDR coggins_l_Page_020.pro
343a180aea06f461739aeb443c34d9d1
2033fb2833f7d854f4783b432d991c6e84698ea9
F20101111_AABSXX coggins_l_Page_016.tif
6dd3df010a4620da2b18ee76ac34d538
ea8c3a661081ac7df4f35ed17ab877b48ae09745
55924 F20101111_AABTEG coggins_l_Page_036.pro
4d9da4ccbe03173c1a5410de7d13f0ba
dc2c497d8014f674c02c97b7ce90dfc16662456d
F20101111_AABSZA coggins_l_Page_051.tif
84db296e814d74de2eaa4e8766fef8d7
79108d9d1813a641d370001e06151417f8302e0c
52282 F20101111_AABTDS coggins_l_Page_021.pro
d27c2f8cae8e16b5aeefd4d4318613e4
70c8bd418afe80e2e4bfb18baa8d3e7ff8c53385
F20101111_AABSYL coggins_l_Page_034.tif
80a3366367f82b75b4e1dc31a7471b3b
aea1da89515a4d7f70194623c0c476cf7cc52ee3
F20101111_AABSXY coggins_l_Page_017.tif
44e0199c9cd794bd13afa9abef6d8530
424289cd18b7cadc93b03663b63e319c84d7c916
59414 F20101111_AABTEH coggins_l_Page_037.pro
62570eedf4ae0c7cdf37c0bac922d750
aaac378c985e8a0cc5c557af8060ff8c7bb82699
F20101111_AABSZB coggins_l_Page_052.tif
a04e33480062449895ed92a093633729
46762a8d660ce1a3139fe9dfc9650aa90ea4c2ed
53532 F20101111_AABTDT coggins_l_Page_022.pro
a341e3a71d65a2d0c060096fd3adec99
7db2b802d7ad965d8cf9c6afad1ac8d36aae99c5
F20101111_AABSYM coggins_l_Page_035.tif
28d40ee213f0547f09d0f81a3aa78722
840dcad752655f886030b638261ebfc4eb47c1da
F20101111_AABSXZ coggins_l_Page_020.tif
d58a2f83924e1694200b4795c698ffb1
cc13109a65a30706a3e8fd4e209fb6246d0cd107
58072 F20101111_AABTEI coggins_l_Page_038.pro
5bd3ecb9c85f87fb5d7f46d60fe5b54d
f1c272dfac0c3b9b0e428f7b42eab288b072c74d
F20101111_AABSZC coggins_l_Page_053.tif
dc96c0bc222d96b183579a7b3e4fa59c
02804ff20fb12bc4b6fdb677c21573735cb5802a
54330 F20101111_AABTDU coggins_l_Page_023.pro
a010ef698f8c557bdf26c314d4ece783
0e855699af1785f1c23a01600dd546aba73a8413
F20101111_AABSYN coggins_l_Page_036.tif
a796ed83e1f14a5bf62a255f496d6260
69b4fe9b8098a8b0ea222ffe13e106a04e65ff69
56138 F20101111_AABTEJ coggins_l_Page_039.pro
50bb17ed2765f58a575dd3849762b045
044ca208956746cd62f5bee07397c744cd8511bb
F20101111_AABSZD coggins_l_Page_054.tif
2b2004f51b7c1e92c1e485c975b73310
b198db2c7331b66260b6ebcc83ed06631f1ccf12
F20101111_AABSYO coggins_l_Page_037.tif
91e076c21175bc141ec7f602f2cef649
1f28c227d03985578ac2ab7d0181206f84890af1
13831 F20101111_AABTEK coggins_l_Page_040.pro
8f47e6a8e335ff23018c6de4131c2086
91ce96088e5d7257f63c701f57b73980049a234a
F20101111_AABSZE coggins_l_Page_055.tif
81b43cce35bb1a3cd231edafb0cc62cd
3199acd0ee43890d91df6b70fac9f3c2ba49eb1f
47358 F20101111_AABTDV coggins_l_Page_024.pro
1863c14b298d1dc4c77e07387bb55e5d
d110f031f928f8b1bc5c9c3b29edc7bc38492035
F20101111_AABSYP coggins_l_Page_038.tif
252000886492af6d008db8e566973454
be95dca429f0520399c854e18c5829665d346a02
60466 F20101111_AABTEL coggins_l_Page_041.pro
65f2e708683c89a2c1a1f817138626d8
b5965b0140a2c1c2b4d1f553196f534b3db3f9d2
F20101111_AABSZF coggins_l_Page_056.tif
2a134d39fc56d8d893c4a56225bfd356
3ebac487e5fad9b24baa058f97c910db778565c2
52739 F20101111_AABTDW coggins_l_Page_026.pro
372cdd0baa9b56e468f85029f36b833b
8013202516a23be3d6e0b09957e1dd5ac21b77e7
F20101111_AABSYQ coggins_l_Page_039.tif
643b5d6ce7a43b49c1e4a1d8a6a99f34
6d49bc41a9d7b1accd40ce5b905245c4451db4de
50880 F20101111_AABTEM coggins_l_Page_042.pro
49e06b332b11d0f220dfa3d745acae89
602891ced3408c792246f7d6be2d8b69882a103f
F20101111_AABSZG coggins_l_Page_057.tif
f222d45df4ab0ac47129441221b2c3e4
94fdd27cbeaf41f3dfc1bb2185e4ba58925036a7
48244 F20101111_AABTDX coggins_l_Page_027.pro
31b884f6505a0ba23902a6d57e1954c6
1fae0f8926af5a2efe58a652f03cb43e0af92378
F20101111_AABSYR coggins_l_Page_040.tif
cac3f3f1caf81eba20b0c3dc1bbb537d
817c4f85b677794e25fa86b728afc8552ec61b3a
54453 F20101111_AABTFA coggins_l_Page_058.pro
37021114b557e08a89839f99e5102eda
1c39f9fa8ecfccec6a24c59669abfb21678a916b
31770 F20101111_AABTEN coggins_l_Page_043.pro
fc9a7939f3cd797f3d7d234056d4f348
0b9686f9ea925e4ef069e3e5557e8cd88e413032
F20101111_AABSZH coggins_l_Page_058.tif
7ac578fd0f5a15575be1122bf3093e32
2ba1a24ea3545429df1ce02f43598ceee97fdb68
39511 F20101111_AABTDY coggins_l_Page_028.pro
b8c69e7368d38cd329e282e571e0c7ae
43c1ffc87990756d1e13666fbe5129a84dfb8dc6
25265604 F20101111_AABSYS coggins_l_Page_042.tif
a22093284cb8a4b40f42ea317e6df3b1
84250144a6ee68580c2d47c5070cad3e0a772d21
50114 F20101111_AABTFB coggins_l_Page_059.pro
8865e170f0e6cafe6bdb9c7c6b3a85d7
5d03ed284ca8fb745f3d34a7105a5a02e10def2b
25324 F20101111_AABTEO coggins_l_Page_044.pro
6d0b43a743ca09c5f2f91e19132c2198
43ca3a35d7233c037a4dd9637b1f1bc772cd1e2a
F20101111_AABSZI coggins_l_Page_059.tif
e1c73417f2b9cf9fd3b4247aab140b1b
63537a900637b96070d838a270b2fa677fd8c433
50902 F20101111_AABTDZ coggins_l_Page_029.pro
1fe20da22c8241395c28570c0033ff6a
67c6fff10ea7423d2c911bd76420fe1882c1af34
F20101111_AABSYT coggins_l_Page_043.tif
7dfc74510f1ab278d8fba37be1e99b9c
18e249ea1f15e0bf3c3c42dd0141ce7329d249be
54381 F20101111_AABTFC coggins_l_Page_060.pro
f10ab1057c423cf982e881a16b001d65
5382eaecbccbea8ec2ddad9f82f7d996184113a6
8584 F20101111_AABTEP coggins_l_Page_045.pro
c65cb9f313e40eec670f3e5f4b016721
b669599e34c9cbef33ab768f0a42c748f1e93094
F20101111_AABSZJ coggins_l_Page_060.tif
1bed6ab5a9a0a310b3ce21b898fea5c4
4143793ff347fd1f2ab49caa93031b4b12482e14
F20101111_AABSYU coggins_l_Page_044.tif
67b655a03758451cc6e351271195a6f9
14a34498c356af2d23d5a812326077c2c8c46986
38646 F20101111_AABTFD coggins_l_Page_061.pro
77779c43230e6d33a0c135b714671bc0
b93e15e54c57adb283d6f7717409cc576f859304
20657 F20101111_AABTEQ coggins_l_Page_047.pro
6064f0caf78558e7c8a56f3c6ecb36da
567da50e275b18614d1972b80149eaedfd998fa4
F20101111_AABSZK coggins_l_Page_061.tif
25d66a2241790a4d0b70e9a854f67e1a
f441fc705234b5184101c6b8429bfad49366960a
F20101111_AABSYV coggins_l_Page_045.tif
bebfd26fc91f9741399c30c19340be7f
9e651f17a91b0a390060314cf1b6d6a613a75a13
39464 F20101111_AABTFE coggins_l_Page_062.pro
a27046981243f43f2d34fcb5863ac2b6
ec440d516a158b36eaf0fbcc5dfac788af79354b
20357 F20101111_AABTER coggins_l_Page_049.pro
539511adc49ba5285a49a8f25a9c56b7
9afdfcd5ed8bf796516f07c27df037e5ecb35cc4
F20101111_AABSZL coggins_l_Page_062.tif
19b6cb9ce35211344d552d4a839adbce
d9ba80085ba64fbea34d5e429a08b63b214415fc
F20101111_AABSYW coggins_l_Page_047.tif
355308913105629f885f09b5a850d66c
1bf4b63ab68181f83eb14bb36faa7f3b1f3063ca
48860 F20101111_AABTFF coggins_l_Page_063.pro
83dc220983634fc19166f56d1ba94f8c
5c6b756720a9cd626ecd4b941c2150358bc66d88
21881 F20101111_AABTES coggins_l_Page_050.pro
c0c808d3c3a4d57ef894c2917034ccf3
1cd1bf4996d74442fecf275652f676910c517619
F20101111_AABSYX coggins_l_Page_048.tif
ffe63e1ada98019f2d4b51a205cca437
48adb30484fc9d45740ec786230100f47b9c6128
49257 F20101111_AABTFG coggins_l_Page_064.pro
88d82e5954e8a602610856c9a9468cc2
8662b1b14c0159bb81fd9a7482cc0d2e0d3f4d26
13577 F20101111_AABTET coggins_l_Page_051.pro
4c218885bf32570f325f2ff1ac133bce
c5499622195a6864cbea1375c36756245d6d61bc
F20101111_AABSZM coggins_l_Page_063.tif
0f818b0d938e7bbd19afc83f1905ebf7
f7d90ec45e11ac384c7f8098a0d03c5db98d9558
F20101111_AABSYY coggins_l_Page_049.tif
f3479156f888f533dfaab727f0b54167
1004f69904b21ee428618c6da509ca4f21aab4e5
40640 F20101111_AABTFH coggins_l_Page_065.pro
933194d53ff888221741697076e20489
05071c7a152c2420e16609fc6f08ceb4aeb8a28c
13063 F20101111_AABTEU coggins_l_Page_052.pro
8b47d1e5603877b5029ca81a3be17a18
121e635fd59eee0886ef027c3f34e7e61d900728
F20101111_AABSZN coggins_l_Page_065.tif
40effbe9b3e472b7891abdc47cb4f4b6
acff28df94ceb91195d498ced8c779a95ec1db00
F20101111_AABSYZ coggins_l_Page_050.tif
97590cd1695ce59e0ab7c7611f391db2
2e8f89344de695d9426c4c992ed223c99187ed61
48128 F20101111_AABTFI coggins_l_Page_066.pro
34c14634ec94cf6a84069944be61901f
0165034e5dc519e3388248b969c706c4200cdb69
20417 F20101111_AABTEV coggins_l_Page_053.pro
98d70472eb0892dcdf41d7549d71e011
a9e35b669ac45513fc341ef0b09bd9fe3b14d116
F20101111_AABSZO coggins_l_Page_066.tif
ce8bd6908955114d4551787b4c89fc08
63561b85c4470ab9a6904e95142a4eaf6b05acd7
50073 F20101111_AABTFJ coggins_l_Page_067.pro
d0d241538035a451c4cbee32de738f66
aab6768957b1a29ab257f8cf0f2b93f5258809ce
F20101111_AABSZP coggins_l_Page_067.tif
3b99e2cb2cc1316a98858927d197898e
915e4fc0577cd3d81c7167b5bb691bcc1204fe54
53751 F20101111_AABTFK coggins_l_Page_068.pro
71c3f35733cc4f65e35ada9607b24a45
8f92ba91cf8fe7d991115789059f2d1eba5c66a5
13524 F20101111_AABTEW coggins_l_Page_054.pro
07cb55b137ee2b9006f8a358ec23597d
941dcd37bd0ef0acb52f79a1c5c391c29ea79b04
F20101111_AABSZQ coggins_l_Page_069.tif
910e68f1626ce1215585087a973275f8
0732c8765c44df458d45390e3b44d9fa2f850114
55010 F20101111_AABTFL coggins_l_Page_069.pro
ff1e5f4668e4f8b603ac698a9cc08192
6435d63dc1a9505d14888ce84001d982b989dc9c
21148 F20101111_AABTEX coggins_l_Page_055.pro
46c5e22225b07e778901d802d4779800
f86387bc4ccb0c6777fb66e3b6bfa1b23a6c3fd6
F20101111_AABSZR coggins_l_Page_070.tif
4981a9167f1876f52dd6b6a59a06c8e8
6c73ca49e06b56ec3e9469188e9fd96c3f195967
52415 F20101111_AABTGA coggins_l_Page_086.pro
2b2e613f0d52db88860c4fe3536698d2
2254d441f1653b9b2f329ef08d46d3197a382f7a
55746 F20101111_AABTFM coggins_l_Page_071.pro
2854de0d90852047531b5e935d1b5333
8f7e6b785a3f17922ee648af3f005c6a62bbdb6d
21080 F20101111_AABTEY coggins_l_Page_056.pro
640477a227168525b47b28f2c584ef6e
ea32b3b959788ca26fc1642bf6976b2c8227f5ad
F20101111_AABSZS coggins_l_Page_071.tif
c67d145893e0be7cd2adae3a9698b447
7bd50ce08dae4992b00c32271e3f37a1ea44c9df
52984 F20101111_AABTGB coggins_l_Page_087.pro
0bc08cf6d0528d232425da453c0f0472
87fb28562a30f31ad76c53fe2dd69dd8da85bff7
55340 F20101111_AABTFN coggins_l_Page_072.pro
c5876b9f200c3c959fcf1b576fd91c88
d8823168a29783e5ea0b76a61f7aeab31470e86d
55023 F20101111_AABTEZ coggins_l_Page_057.pro
44d05280c64df8fb77e3afbd728ce26c
ec1185997a47b50dd64701e833d30b24bf4e9897
F20101111_AABSZT coggins_l_Page_072.tif
db98bb2b21ad2ad7ff6a2bd1a0f9af63
fc06afb9aebd9f1f346700af75b2e0e952aa54f1
56108 F20101111_AABTGC coggins_l_Page_088.pro
b919726f83471416a24b10bac5ba64cc
7f6793b410cfc06e1aa3fd107730ffdae0a36417
5986 F20101111_AABTFO coggins_l_Page_073.pro
4b09919d274966b4ca9e0ec7662ddcb1
bfc989c1bbf0a1ce15451c5297c92dd1108bfc9e
F20101111_AABSZU coggins_l_Page_073.tif
53b13d3106e1d2f9690d29eaf94df551
b3dbab2bf100c61481896422bd07b553670d0594
52253 F20101111_AABTGD coggins_l_Page_089.pro
673c2b7ebdb59a90503f21e6a593b832
eb89dca804ec3ce8125884126cb64cf589b58d19
6798 F20101111_AABTFP coggins_l_Page_074.pro
7c011594573bfcac079a1e1294963276
2ef9321d394699fc17a0331a61670851fd222b22
F20101111_AABSZV coggins_l_Page_074.tif
6ea62d5d5d23099b8b9d6e7d55ac1d79
4cf49561f35dcb0b1ce681c8f9f82b5c97668104
47779 F20101111_AABTGE coggins_l_Page_090.pro
57d60712b883528c3dae428ff17dc324
2fd6f9fe39e7371ed5ea0cabadefce52da0f81fd
5767 F20101111_AABTFQ coggins_l_Page_075.pro
b49e07d93f925d794d169e72ee736bff
0aca5474fe66400bfc04e247d7e639b60e0d3b37
F20101111_AABSZW coggins_l_Page_075.tif
1bc3ea8fb1daa7ba4ad99d1ec04f4526
7efdf5b64dd8c6a17de7efaa583af707121c5581
52933 F20101111_AABTGF coggins_l_Page_091.pro
669dc89cc8b1f358e10e3d3cc2a101d0
73bada458f384bda0d5bb82c00cd73826aaeea28
6207 F20101111_AABTFR coggins_l_Page_076.pro
629bdd334ecda823d25db229ddc6ea2e
d26fed8cce29e166becde96ae6d11d1a7b5aaadc
F20101111_AABSZX coggins_l_Page_076.tif
2cc4b95291b41b066661b798dbcc78c8
c015d9b180f01a139d6defd9dd5f1fa2c18fa90f
44271 F20101111_AABTGG coggins_l_Page_092.pro
96353fa857dc4a9904500709687b7d55
f3c5a2754563b03391cad1cd6c5c1cce3e197ea8
21928 F20101111_AABTFS coggins_l_Page_077.pro
0bea3489929ba53ce7e0728b1dda1bba
5731e60fe06f05cd9d86db36504999bb96755bbe
F20101111_AABSZY coggins_l_Page_077.tif
401567deff55a994f85b48194e5b91ae
b49ffca57a09ae33c58f03d0f0c66e5e0f67670d
40282 F20101111_AABTGH coggins_l_Page_093.pro
dcfb5fdd0b12a5082272a77e666825d7
1a413b93aeb2be2cbf7aff5af1bbab60933dd8ba
8445 F20101111_AABTFT coggins_l_Page_078.pro
70a7603137adc27a8ae80ea24a644c02
2e6a5180c6b0ed0e9957d2eafab84a6477b94aca
F20101111_AABSZZ coggins_l_Page_078.tif
d2330cefa9372ade64f64cb995e2a610
fc0869cbbf128a976c827a957c75bb270c9c4979
52098 F20101111_AABTGI coggins_l_Page_095.pro
e82ee29a93bbc33ddc7752e25e7a9bde
a4b86e40d6a07aa00ef0a98b47a938b151c3f33c
12310 F20101111_AABTFU coggins_l_Page_079.pro
79a2a1c9a08a349ae195aacc5ce83828
8a347b7259996b732f7c09add1bf39def679e105
55761 F20101111_AABTGJ coggins_l_Page_096.pro
e7e85917c362c29a37857f91cb7c5502
03d140c1ef46926caa82972ea9cc7e13cc1fe355
6281 F20101111_AABTFV coggins_l_Page_080.pro
af46dd3ba87b58cfaa407402f3ad4e41
1ec6fd8a861168a239689e1bfb308c3b331ee179
52385 F20101111_AABTGK coggins_l_Page_097.pro
623179c2ca5ee58d9a42ddea09b7995a
b220ff32ec816fe1df30eea0c84256ea8863315d
16104 F20101111_AABTFW coggins_l_Page_081.pro
fe6a493a7d41eb258a66216a5e10b4ba
127f5b3d3102cbe3b8fd3858780f9b25f872d3ef
52996 F20101111_AABTGL coggins_l_Page_098.pro
0caa7854e5835b66557be02d575ed183
0a3aa8db09aeefeb0b11be171f713577b826b365
16579 F20101111_AABTHA coggins_l_Page_114.pro
e00a81e0965318c3e7620aded36da8ca
2cc4ebae715c6fb775775d30286e5163cb72ec57
52285 F20101111_AABTGM coggins_l_Page_099.pro
b2d1bdad32c925decb87bdf2f6dafcdf
afbdff9bd72f2bc162ec52e453eafd29903031ea
54210 F20101111_AABTFX coggins_l_Page_083.pro
4821243e2b604c58748a394486d3cc97
8c17ada092d06d6d96cc5606e4b00d09abb567c0
22522 F20101111_AABTHB coggins_l_Page_115.pro
d891607d5954c4f3a07b42b2238aae3b
62cb7cffc4f8986ef01a1ea73bb2853a0b247b16
51391 F20101111_AABTGN coggins_l_Page_100.pro
f63fbe18f8e885112879444efa0573d8
91a2fdbcc3e0ffca80ce1c8c8b72e4d3fe50b94d
52825 F20101111_AABTFY coggins_l_Page_084.pro
de3610bf66ebde12f6c42c1add9a6a30
7f022fabef74424e2290862886382534342241e0
11004 F20101111_AABTHC coggins_l_Page_116.pro
ea84fb87f868f6940bbc7c5f75fab6b3
134f943633f19c862d7fbce0e3b86148efa078a8
52651 F20101111_AABTGO coggins_l_Page_101.pro
c98583c36eb70a5fbdadf91018bc6c22
adc057d3ab6a4c04df42aca53453d6e56d8af083
52442 F20101111_AABTFZ coggins_l_Page_085.pro
91b40934529d302ab6bf1e31e3609911
47a506dfd244e13358b01c87306bb70d7aba5925
12034 F20101111_AABTHD coggins_l_Page_117.pro
0af1e4832ce8c0a378cb7b784aaf2275
35905e1b770e2d0866878929883b0515ea21df7c
51915 F20101111_AABTGP coggins_l_Page_102.pro
646e159a7b34cce62a7f7cc49da262a8
d60e05072d815af287deb783ba287214330f2c64
10321 F20101111_AABTHE coggins_l_Page_118.pro
85e104f882b820d65a1a171b87fec53c
d1799a39c9458cd1cb4a5363830caafc00ff2b07
52965 F20101111_AABTGQ coggins_l_Page_103.pro
6e9fb403ad90e3f93ef32b3141de3755
86dfef70611c5df31fa3d29670f5c9d56f689a60
15869 F20101111_AABTHF coggins_l_Page_119.pro
b2f11638f02608d46ed241a767a34156
118a9654d04c229054e00fd26a86ad49e20bc251
56171 F20101111_AABTGR coggins_l_Page_104.pro
a87e5a1599388b1e328444ff32796ddc
e1f1c0a5394edb3f4baf8ad85f4a14fd49ec456f
23970 F20101111_AABTHG coggins_l_Page_120.pro
a98478b33c3245cc49cf567c7de11698
f5cbee7b502cf2ed895c60306622c97d23c3f56a
52657 F20101111_AABTGS coggins_l_Page_105.pro
4bd754e67d0e75a78b14cff88590c82a
4f10ab54517fa3e3395fe5aa0d11bee0befa42ff
19338 F20101111_AABTHH coggins_l_Page_121.pro
6f3bc5ee04441b61a4e08eb96c59a37c
2d5885f2df58b000f93d3be2d3a0e4e00ae3cc98
54878 F20101111_AABTGT coggins_l_Page_106.pro
a76e29ff1522193f47cb20748d2e8f74
aba3bb980086ed740ef0300590493d66a98225f6
22517 F20101111_AABTHI coggins_l_Page_122.pro
e830c63815345b169c930eecf57683bc
2cdb3fed0220d1f6e9cf31aaba42b29efdba36d9
7000 F20101111_AABTGU coggins_l_Page_107.pro
e4492f63fafbb0253b1693b6733b6431
3b9869b440cadf197f8c414acd6396b4427e9d2a
18474 F20101111_AABTHJ coggins_l_Page_123.pro
8ef9d0a16321173c551e0e28705c973a
4574d0b218cf44d634821543acfe4917168218cd
6997 F20101111_AABTGV coggins_l_Page_108.pro
2fd06ccd9e33efb18f871f551570d90d
4766b79bf17467d4ba382af90ed23baa96bf0903
18273 F20101111_AABTHK coggins_l_Page_124.pro
2ccc7f3283caedbc0ff42b998ab7faed
a081a780cb9e54a6f4f9a63aa7fc61c64b8fddec
38127 F20101111_AABTGW coggins_l_Page_110.pro
d1e2592ea190917c4087ec628dd8e526
313ff03e2d9e4ba1199f41914873ace643e08304
15032 F20101111_AABTHL coggins_l_Page_125.pro
ee9fb6f5f0bd5cc1ddb08ddd8bd74662
e3bd35fbabaaac09d482ab38b24a09d841918a77
15061 F20101111_AABTGX coggins_l_Page_111.pro
517f7b8f62292995f7884fb84cdd16e6
a7c56a21fa98722428799aea341ea6c027dd8dde
21119 F20101111_AABTHM coggins_l_Page_127.pro
eacd8d41e3314f0b207d33c67744822b
4a3f949c1f2375bd19fcaa4e1bf4b77c3fbd9c02
53331 F20101111_AABTIA coggins_l_Page_141.pro
decf979019654998ae15a47a5d2b50e1
c3638717342878eb6e44d969decec0c242f4879d
32098 F20101111_AABTHN coggins_l_Page_128.pro
67c814a23d447104793f80b709d33b7c
450f80b68fac08a7fb11e17614d1c1d72c5b1b94
25613 F20101111_AABTGY coggins_l_Page_112.pro
1a93d72b2a9c9881e0176e27547059ce
144f948950e7859bb17221cefbc0ceac95aff71a
52785 F20101111_AABTIB coggins_l_Page_142.pro
8207159a0f73f48ae75ddadb5fe2e4b1
aedda105224307e9d05558516d36912fbab76b68
17749 F20101111_AABTHO coggins_l_Page_129.pro
750860cb7369d7c923b927da7c81c793
45c94e6f7483c6c7de35065f508d63d87d691436
14378 F20101111_AABTGZ coggins_l_Page_113.pro
9f50b3b78670dc3ff0694c5c1bdba405
f4dc0e089e97402681880b778bbef32f084ba199
55617 F20101111_AABTIC coggins_l_Page_143.pro
7910a5abca6458cf658743a91779798f
d2d89396b840e25dc3f38f80bc2b09a911cd267a
18589 F20101111_AABTHP coggins_l_Page_130.pro
1429d03cee0188331617a741b493da2c
66e02f7439c80217a2de68df75cd4e2e4eb82c18
57265 F20101111_AABTID coggins_l_Page_145.pro
5bae42d5f53e1c802a9bedf919679766
9213a1e2e704d05a0c15937786b70866768c2943
15018 F20101111_AABTHQ coggins_l_Page_131.pro
40d164bed9b0e6facedf58567fdf4729
881d44a3d1b1de164baa33994bf7b0ae2e8ef153
54160 F20101111_AABTIE coggins_l_Page_147.pro
e4b0178b71e23d8eba8bab1b000236bd
28f53a09641e8df2b057df8b7edc18b807bfb3b2
12603 F20101111_AABTHR coggins_l_Page_132.pro
6288537519b5c8c7adbab09a0d2ad94d
bb8675f65d3c8ff39c49b62039e582db94d3469a
54255 F20101111_AABTIF coggins_l_Page_148.pro
846b0e7ae68becfeba4eecc941dad43e
f37cba775d9fc45272269c67a47e57b301657d98
24247 F20101111_AABTHS coggins_l_Page_133.pro
2aeacefcf7c665ad68974db528616d2f
aa0fd512c67b284a8cfe0475e6d4bf80c7f34282
15250 F20101111_AABTIG coggins_l_Page_149.pro
8eb24816fe09a72834de03c2e42a4c02
940825f3371809f6f21c7fe84b7adc00707fb3a5
53378 F20101111_AABTHT coggins_l_Page_134.pro
1c612c4b4eb0972b44a779e2aa8aa04c
9dbbd504ebb58ae8edab82be82eac43977865e66
16747 F20101111_AABTIH coggins_l_Page_151.pro
f8725731a31f693745668e28e71ff436
80c9d5d55c82fc549dce7c03875a3093d2243153
52193 F20101111_AABTHU coggins_l_Page_135.pro
e2d6afd74d3953d0d9ccdb2622127fc5
81c3acb40a429cd5fc56566c2ff51fcd212eb588
21410 F20101111_AABTII coggins_l_Page_152.pro
98fd405f14e7343922d6197409940356
6b60935680687204ecbecefcf7ee7c3d21a40c07
50145 F20101111_AABTHV coggins_l_Page_136.pro
984beca0ced5518551b0b363ab8258b9
10ac1809fa1d48b949ffa3b0b5b40fdcaec889ab
15997 F20101111_AABTIJ coggins_l_Page_153.pro
367a16ddfbfe93471de9a6380633bcbd
376073a00b97284bccc63bcfff0ee7e26a2754bf
50401 F20101111_AABTHW coggins_l_Page_137.pro
6dcd6d9c79421c6048259c5059111b39
8e601ccad30c9412a8847a3216f14fbb14137af9
27774 F20101111_AABTIK coggins_l_Page_155.pro
c7753f1a6bddd93b27b85a060cf6d2cf
414ab9a2739cba2d399deafd6cdb60c4fee8be7d
51015 F20101111_AABTHX coggins_l_Page_138.pro
354ededb70870e332ba2a16087fb5f6c
825958f543c74fd0971b80f5c4786962f2bf500f
25564 F20101111_AABTIL coggins_l_Page_156.pro
21200f09b031cb6f985d292fa7ad7b5d
f7a2b2598507d9cddbd58a364048e2258e042b9f
31528 F20101111_AABTJA coggins_l_Page_173.pro
a7a95c293dc83cad30b0d8bc2c7eac4f
12371c1e98d4856f26310f50b7cf21c35f7fe94b
27746 F20101111_AABTIM coggins_l_Page_157.pro
520d96b5b824d22217857d11b89dc77f
46597666552dfd7a731dbd1d305bf66fbcf2d9cd
55184 F20101111_AABTHY coggins_l_Page_139.pro
9fc9e37141b4b11d7ba387f9863a36b7
2d0807b8430a9c2ca6d7ff2a4aebba89cd3ad71d
498 F20101111_AABTJB coggins_l_Page_001.txt
476f1f418904eec32c64b4e33c334823
8e54c936f6b9893a37af6c13fd3981b6612c44af
12215 F20101111_AABTIN coggins_l_Page_158.pro
73db2448d68b58dc810c81b85ded9f98
e8868f95715e3c050f420d0a6efc8626880cfb0a
99 F20101111_AABTJC coggins_l_Page_002.txt
a793ca5435adcb8ffa466bb2e17db557
cea83766e381b2556cc0d8dc9489f95d9bd5c4dc
63705 F20101111_AABTIO coggins_l_Page_159.pro
6fcd3497887230edbd75dcfea366ed27
760bbdd4c2781bece0fcbe661aa42dd30a0da7b3
54337 F20101111_AABTHZ coggins_l_Page_140.pro
afa673f3e5a3f32ab9e8bc08e5178955
cba64c13729e228bc48d390a2aa8a21c0326732a
78 F20101111_AABTJD coggins_l_Page_003.txt
a2a7038b799ed23cd95c4fec7b947f03
3ed5e44467e838a45db11e86294990ff7325c812
68537 F20101111_AABTIP coggins_l_Page_160.pro
45dc922c6b7e06ffe629aa0b93b8e37d
5826baebe320db8ea0bc1e60a3b431691817e7d4
2104 F20101111_AABTJE coggins_l_Page_004.txt
0d58e4a97fc6b43a86b3fbb5c56e5d33
f98dfd6b957083f3189dea280261d6163b65b197
69916 F20101111_AABTIQ coggins_l_Page_161.pro
64e9b6b26bfb302ed30c0354a4e5dac8
213b4c55dce0e4756ae70ceee804ccba3bb0123c
3506 F20101111_AABTJF coggins_l_Page_005.txt
df527f14573212f2428a55fdabe4840f
0be6f2d9d13ea64117038a63cf744e630b47504b
68860 F20101111_AABTIR coggins_l_Page_162.pro
64b336873edde41b84c61ebf3554c6aa
d3a77d38256858123558f0aa8adf170e9cd3ba70
31846 F20101111_AABSGE coggins_l_Page_012.QC.jpg
5cf45071620f931b4a38087491c6f335
8b6c048acbc85ae872e829595fc7d69c3ddff676
4370 F20101111_AABTJG coggins_l_Page_006.txt
703d7a516e4a06bce7aaa74b3ce8ace7
ecb6a89f07b8f2200ce9060fb504f9be7053bced
64067 F20101111_AABTIS coggins_l_Page_163.pro
9a649bae1f68a354aafaf936a2f85d91
071adc6233f7904768e7b25f18e1ceac5cadd748
F20101111_AABSGF coggins_l_Page_153.tif
ebb176810096a50aefe2a0006897dde9
dd3d38553cc09eed6a3c190948fa6b1580a21e68
1577 F20101111_AABTJH coggins_l_Page_007.txt
d17c02052352ea0c1de02db85e45cc6c
c5e2643173feff02c173f3b5a610032fa9b861be
72463 F20101111_AABTIT coggins_l_Page_164.pro
bfdf9a39d5a7d3cd0cc218d9b93ec46f
c8be2d28d51fcbd35f0db0083aa2ea7acf893f6f
7191 F20101111_AABSGG coggins_l_Page_109.pro
91ce02ecbc44cc2d03db044a954c46bc
89bca124ad27a30ed7c57bb330afabe6150ef141
2993 F20101111_AABTJI coggins_l_Page_008.txt
23c505813e93d0e9fff7d2b1f8e38471
4605d45a0aecef9c21bc360bf47ad3742647064c
63375 F20101111_AABTIU coggins_l_Page_167.pro
b4b7c9560bb20f8f0b85ea60c8414585
321de1ef88e2286ef2666424833dd1389c79c402
149670 F20101111_AABSGH coggins_l_Page_108.jp2
885b34377a47b146d8c11f3ad0ebd129
7e16b9892f2f258f496bd2ee90e31b8eb6f7e6e7
3408 F20101111_AABTJJ coggins_l_Page_009.txt
bbfd50cae01e9faae6d8719e3ec49a48
bf49ea81ab7f21a430efd9cc608d52e25154ffde
64607 F20101111_AABTIV coggins_l_Page_168.pro
4a0a26d04c93dbcfa2a45b14a046fd9e
670a7fc2b7df080c92fdcf0653d3c70263df77d3
1559 F20101111_AABSGI coggins_l_Page_109thm.jpg
a9402e2d2612d58c8914a29de160119d
f010b5f9dc35884a3d4a970df1d6222901474213
2922 F20101111_AABTJK coggins_l_Page_010.txt
c9b305d43e5cfa067a7849bf1a796c21
8ef980c81b46722b129ef4ce0ec88ccffb06c2dd
71150 F20101111_AABTIW coggins_l_Page_169.pro
77bba4f815ad159ee48f57b92761c79c
2087e86b427c909d5464b466564160580df9f4ac
638 F20101111_AABSGJ coggins_l_Page_002thm.jpg
f6d6a6b511f7bf3bff6de04f981828aa
6e7ebbb319d84f561b650cc21edb102fccb28ca9
1942 F20101111_AABTJL coggins_l_Page_011.txt
0dc1e3c250727ec915bdfafceed861f2
f1926051f497594fc04d0a818618fc9c671fc3b2
79421 F20101111_AABTIX coggins_l_Page_170.pro
dc0c266efcb5ef926d7f86b104e22d71
40604f5fef86c576f77cd0fef915133642194ffb
95982 F20101111_AABSGK coggins_l_Page_015.jpg
7bf9db6c14f8b528b395c7bf946eba45
f875d4fb536a1a28426651d198b94872c6250321
2067 F20101111_AABTJM coggins_l_Page_012.txt
88fbf36fc587ee88d677398eec3b6f87
6c2368f075bd4ac37ee11da9ad0493e9ee0e7afd
68882 F20101111_AABTIY coggins_l_Page_171.pro
a4351186ba0c4ea4bf475b946eab4f8e
a4d3eb1cb704c50bff0b382ccec318a633eeda6a
1931 F20101111_AABTKA coggins_l_Page_027.txt
6b724832254e234cced440f364ae3d97
f998ac1fba0a565dc35f573c9637802c1e0381fe
F20101111_AABSGL coggins_l_Page_142.jp2
0a5e5583523bb8a489a825a913a9b732
e1fda6bf42248023e07320c370e89f7cb8dbe762
2126 F20101111_AABTJN coggins_l_Page_013.txt
69a687fa6881d78cb8989179a2d57c81
9f215196e79e1617ae348c64f2920b60b98de156
48437 F20101111_AABTIZ coggins_l_Page_172.pro
315653541197a80e84bbd18fff263105
ee81591e66df472e1475a5a975f4d8ab34f5fcaa
1686 F20101111_AABTKB coggins_l_Page_028.txt
e7d91ec09c0b53041bf2aaeba31dbb8b
066d31b2c5db70ecfe18039d0799b20175025196
114104 F20101111_AABSGM coggins_l_Page_104.jpg
c1af1d00d2a26f720203a20b3883a02b
ab46c177779173e9bd0e52772435ae4635488374
2148 F20101111_AABTJO coggins_l_Page_014.txt
6fb6c2bce98ba20b78aaad28aaf08e47
b9e6a3796da10aec468051def762815ece5baf0e
2083 F20101111_AABSHA coggins_l_Page_103.txt
41a281b5412e237691d74ec90b4f818d
d2d8817149ba4ac3dc9f58a0483d2980fdde61cf
2084 F20101111_AABTKC coggins_l_Page_030.txt
4ca9245f404ac97beb585331cd18e75a
84ed6097786ac3cd6d417ae98af56464d7eb779f
1199 F20101111_AABSGN coggins_l_Page_074thm.jpg
82f383a6ab43b034c95cb3539cb04e6e
71f3ec357291cd7b618921cf3230f4d0eaa5ac13
1806 F20101111_AABTJP coggins_l_Page_015.txt
e0ad5262654be66734127a2ce8e1c3f7
d638f5319eece990e475a030008782e13b279504
2787 F20101111_AABSHB coggins_l_Page_171.txt
c6cdb5c5462477b9bbe7d40ff46f33e6
635b1019c685ece7079e7a90d7d60d95fcc78054
2028 F20101111_AABTKD coggins_l_Page_031.txt
df917dca075170b88b96a0a250ec1d14
cd8c308a742088ad31a0f08165ad1b2f7049fd96
38107 F20101111_AABSGO coggins_l_Page_007.pro
eea363381f5dc86c9ec2386efeb97ffe
7d5749dd54baddf73b4bdc3ff1df23b74f236a4e
2328 F20101111_AABTJQ coggins_l_Page_016.txt
c91486610cf493629bb3a723b2442d74
53402bc5154a93a989bf0685480129be87f9a3df
108626 F20101111_AABSHC coggins_l_Page_135.jpg
347a179207d20b1e2e7334bac3b39f15
89148b42b888caf1a3b94f5c6a32c9b6c7a613a9
2214 F20101111_AABTKE coggins_l_Page_032.txt
5c5f1b738aefdbcd6d3b0433af9f2f97
ae11e8e709599a6aaaa42224d12e4b6be965bb8d
9114 F20101111_AABSGP coggins_l_Page_088thm.jpg
2707e664429e9b3feeb2ce8c6de597c6
8ca76cb8fb989d306a4cb22db400c96f65c05cc4
2263 F20101111_AABTJR coggins_l_Page_017.txt
b9f1e77d389918c13a684ec2fc8b9037
5bb4dc461fe1e25b4c6945c8278d340dd43efca4
1053954 F20101111_AABSHD coggins_l_Page_173.tif
b1a151422c8ba9393cb4658c22734398
14508d400a5cef0c2273002b1f9a88955e1955aa
2108 F20101111_AABTKF coggins_l_Page_033.txt
e22539d4a5dc0c6dd83de71f7719fe43
55cb0faf9e4bfc7bc51a2b4565b7d1b5cd475ac0
35839 F20101111_AABSGQ coggins_l_Page_141.QC.jpg
a85c7849351738158962cfdd9a826f90
18c146f240bccd01e6aaab4a17117b61ed5d208d
2094 F20101111_AABTJS coggins_l_Page_018.txt
61994ae668bd7a21bc5decb157ac8d30
5e76d24347dd1e9e5e5ebcbacab3d27d70bfdfaa
116625 F20101111_AABSHE coggins_l_Page_032.jpg
cbcd16b2e4cf6a66dd705a0719013415
16a2762c3857324990e0c7aeea1ccd5838d872ee
2230 F20101111_AABTKG coggins_l_Page_034.txt
09bb4be14b1bfedf847e9dab7349085b
f111c3662a5fed3a7c785f3d94e122f30d0ffe6e
5893 F20101111_AABSGR coggins_l_Page_110thm.jpg
57e4e2e085d03dee468b5ad93407ddc6
e5aa6a307f8728eddbab1e077b05cf7fb9ad317f
F20101111_AABTJT coggins_l_Page_019.txt
f68c57fe8df51f213cd70786e48a0436
fe266646302f6b051c5e366652b02e9a60e6f060
56012 F20101111_AABSHF coggins_l_Page_146.pro
b44c8129e3cd644d5a2fe2138298dba3
9b314dc6ce33b1f87a24ed51bb826ce1bbb50582
2171 F20101111_AABTKH coggins_l_Page_035.txt
42e1cd3735aaceb2fd90e3510b3bd22a
accf21ba15076fed4504b1593b2511668ac38769
F20101111_AABSGS coggins_l_Page_031.tif
7653cc13148a9723799338e714e73866
5834566d4e71e2edf66f8d2c8f02b31b464ddc0b
2184 F20101111_AABTJU coggins_l_Page_020.txt
1c236180f752c03b30cfbfb5d6926599
449741da28431fc3865f48510d65171c10ed8bfa
F20101111_AABSHG coggins_l_Page_107.tif
4c03c6c14e06d889890594715dca74a6
b7c35fb7edc7756c44c8c3cd3bd4a957254c9e70
2326 F20101111_AABTKI coggins_l_Page_037.txt
8f32d0d2a94f125d02b421098ffe36e8
074224a1a55ae46ddd51d64c963a7e014aef2e3f
1247 F20101111_AABSGT coggins_l_Page_157.txt
82d5f6c6f675d25c603886fcb4d2371b
cdb1b6a956b4492ea029b26e3e0947545e5a2992
2092 F20101111_AABTJV coggins_l_Page_021.txt
800467dc1ae5436323fd24149d6ade48
047a47bb1b10efe42f8ddcf92e70877ace3bcd88
F20101111_AABSHH coggins_l_Page_064.tif
ab377123aa50baab3e156799908d476e
abf0bb28176c7fd842338eb6136c79b3860a360d
2278 F20101111_AABTKJ coggins_l_Page_038.txt
112b0d7302aa0301cea019cd2bcdff25
93d356dda19604eb88bade5f17977fa5294088dc
2203 F20101111_AABTKK coggins_l_Page_039.txt
4423cb4c9764ce33325f255a208799ef
9b05b2663567c3108400a839b9a3b2bff9b39258
846710 F20101111_AABSGU coggins_l_Page_093.jp2
07a398a9498608016f1d8296128b1c90
5584e68222ceb5f28725a4bd6b3f2a8a66b76bb8
2134 F20101111_AABTJW coggins_l_Page_023.txt
fe1279272aaf50f6757b044165814173
b2e076b67e212f083f106ecb7b514a52e6d7c2c0
14941 F20101111_AABSHI coggins_l_Page_044.QC.jpg
6132cb98bf95f66469956f176ce959de
f562352a79d3a6e541eb68a5286e79469e8c2f50
550 F20101111_AABTKL coggins_l_Page_040.txt
7134db383dddba34a540dcad43ad2f18
301609f43d00d20f75cfda05dfdf02bc7eec6438
F20101111_AABSGV coggins_l_Page_019.tif
cd22ff88f2974d32e13b728bdb192172
b8533aa445a24c41be8dcbe8272ae87f22d29438
1984 F20101111_AABTJX coggins_l_Page_024.txt
1f57d51d10e37ee1cd6a7aeadfccb1e7
9490ba3bd673e1675fae48cc523abb69e55aae9d
2784 F20101111_AABSHJ coggins_l_Page_162.txt
7c8d3cfdb73988514a454c8a60652815
98da1cc15b60c03e7087f70911e516992c96dce6
1203 F20101111_AABTLA coggins_l_Page_056.txt
b5b0874dc281a7d6e82eea5034a2f7f0
994505e66977a9277e0206cb2dc9d3fe1fbb321c
3166 F20101111_AABTKM coggins_l_Page_041.txt
5e8cbff89678ca162dd24d48b6dda123
a676aef9202141b9cd3366af0e142cf8d9c96501
11811 F20101111_AABSGW coggins_l_Page_150.pro
25af9e6d654d4c27662b42d6454b8e19
06143e843346b525d137d9ca08a097d468ab50f3
1960 F20101111_AABTJY coggins_l_Page_025.txt
2943dc32d2dfce3d6ef7b709b6f0e3c3
547f0aa79574fa2a59882e7d85800cc57c7a925e
36529 F20101111_AABSHK coggins_l_Page_058.QC.jpg
3bfde1db207f26706233439093ad4ce4
4430ed4fdaebefb69756c58dd7752ff4b6f58e8d
2208 F20101111_AABTLB coggins_l_Page_057.txt
bc1b8d61c93fb91cf5bcebaba3cd9089
a65a14910bfca5a84edec3afab7eb3547a8282a4
2507 F20101111_AABTKN coggins_l_Page_042.txt
c6fdf90519121f38d03ae7a2601628c6
714d7aa25a504516e31b71560e5211f58c6e0ba6
9085 F20101111_AABSGX coggins_l_Page_166thm.jpg
3fe3eee97daa975b49893e49ae70fd7d
a518c77232c7f59df91c9586a716485b8c103172
2128 F20101111_AABTJZ coggins_l_Page_026.txt
b61fa3d9f7da6cbdd16a2b6d7aae98de
179d6fb34e8fd2ee179f1f4138c3ec2ed114a1a9
11828 F20101111_AABSHL coggins_l_Page_082.pro
4cd5cc75e1b90eca82b17e820e7a9971
fc0adb314d0bb185503167adc5011e40e9a71da1
2138 F20101111_AABTLC coggins_l_Page_058.txt
d3e36c11752fe283096615bcfee9ef0c
e124bba3fba0bf71223323765f00b0cb44ab1961
2077 F20101111_AABTKO coggins_l_Page_043.txt
35bfdceadd14b3f42de716bea6f76da7
07cf2e6edaaff7f3835df99cc5378d785cbdbd0b
784603 F20101111_AABSGY coggins_l_Page_048.jp2
9ab1a7e5ef263d7f5ab72efc537b6196
379a2e7d9ef02c27af806c37968a3cae76f573e0
12054 F20101111_AABSIA coggins_l_Page_046.QC.jpg
92c44dea47f6845c8c4ef3e5183c1c5e
f1158f8485577fa8883cd9dd41da7f54c0f3e7d5
798981 F20101111_AABSHM coggins_l_Page_128.jp2
4e5aeca33f2a020e6f590d2deaec0e46
753e36c09eb817fdb13f4848aa0643fb6b5e1252
2037 F20101111_AABTLD coggins_l_Page_059.txt
86af06e715e215e95d01928e9ed75932
da0ab04988d355a5d85818c04e9100704c75e404
1262 F20101111_AABTKP coggins_l_Page_044.txt
7c39f6535afb69eb04562a776f29ad2e
ade0ec0013ee2cf04848ab969ecba74bc3286f5c
65617 F20101111_AABSIB coggins_l_Page_166.pro
e29c3422d0e8e445845401cacf72bf03
15defa82614eae7e62b73e5ff4f070deec20f097
113223 F20101111_AABSHN coggins_l_Page_096.jpg
5a26a707488c6ba093dcbf97f9780d1c
c6075eedca97365e96686077e453fc1af64d65c7
2173 F20101111_AABTLE coggins_l_Page_060.txt
5ae5ee6358c76d8e92c4dff3a8979236
ee952198de6d245ef684f3f440d30214d789ba0e
423 F20101111_AABTKQ coggins_l_Page_046.txt
6ba993b2a7473b5f7b2ebb8ebdacb364
9c8b4b461eb58dbc08c58223aafcfc508c4c5315
F20101111_AABSGZ coggins_l_Page_095.tif
18c99f4d9c5f70bd334752405cf7523b
94474ff9e3256a826828037d40c669a79e7cf587
8452 F20101111_AABSIC coggins_l_Page_066thm.jpg
61c886939cae4a91a278318bb7a80b77
a6bbb9d966fe8a0c61916b1601edb203c90a23f8
2192 F20101111_AABSHO coggins_l_Page_036.txt
7905e1ad2dffbe220a8e2e4262434c65
12ef5db1833532f4fef70501488f9bfb0603c22e
1871 F20101111_AABTLF coggins_l_Page_061.txt
c6f1ad8d1b99235ac8b6f0d4589ea23a
f510db7ab99fc5a1c766e7c965b1ea83b337781e
924 F20101111_AABTKR coggins_l_Page_047.txt
d8c695ff7193c48e73d741c8c824821f
323b1147f14392f0d7c8a590f6f5a640bf44948d
9291 F20101111_AABSID coggins_l_Page_104thm.jpg
c81923bcb368513bf8a994d7259713c9
15651fe3b17e40c1b2744ba6a2caf78373429d80
3193 F20101111_AABSHP coggins_l_Page_170.txt
ff5ec16854d30bec4e3e1f0a337d72e4
34ccbb4c9596e4b657fc676a40a50c16096b6f94
1705 F20101111_AABTLG coggins_l_Page_062.txt
720009fbf6b2f4647468d28f9db0ddbf
05f9ac33beea101f13f6596dd3b40f7992d69ec8
912 F20101111_AABTKS coggins_l_Page_048.txt
4de425b7126b955bb8ab1c34c8ba4875
9330fb622395cf1089d2ddfb97a15fe7eef8fa7d
F20101111_AABSIE coggins_l_Page_041.tif
b1a060eb42fd0c50643805fb91b2750a
d906ec13ef9a5f3ca015e99e5514af095b4a0e54
8701 F20101111_AABSHQ coggins_l_Page_018thm.jpg
cddac82e5ea6190ccfda1c458ed81e42
314e6acf9fa1af19e6ae3ce1c8d222407cc1eb79
1998 F20101111_AABTLH coggins_l_Page_063.txt
6fe8cb337ac4ff19fd51e0663d40d138
6ee6d0eddabb8e108f273ac4258b1e90add7f213
1213 F20101111_AABTKT coggins_l_Page_049.txt
9dc335ffed32af38b3a4b8c1e6b24205
686994aac98be232bc633b1288122ad96f477612
12161 F20101111_AABSIF coggins_l_Page_158.QC.jpg
ed3ae818460a373c44ef5e52a3959c6e
75d3106e3b9e03272d526f29eaceaaec6128122a
36473 F20101111_AABSHR coggins_l_Page_087.QC.jpg
4ca258c6346efeb4b09e7168f6938449
1708689ee632cf4f4ace278721beb260d3b7f61a
2047 F20101111_AABTLI coggins_l_Page_064.txt
d805d4ac8856f0dc66e378890e275aad
cbf958d67af4b30d35adb3aafe866f748c517e2c
1227 F20101111_AABTKU coggins_l_Page_050.txt
342577bec2060d9d4402d84d254bd97d
3a14a73b1235d71879f5f500f9e8947cb8480788
106730 F20101111_AABSIG coggins_l_Page_097.jpg
473f0cc032065da3b2f9b0c2a9b3fedc
2067f3f61505a40c6f4fa39def1d0254676a767b
43753 F20101111_AABSHS coggins_l_Page_094.pro
63244148c53f900d372108828a507d6c
75411383a40812460b9c8929c9e07f7513ded118
1955 F20101111_AABTLJ coggins_l_Page_066.txt
78bcff5c36e5ea453283f257b62c94f8
8decb231a70ae8fca88eb33becfa531b6b0e1d3a
594 F20101111_AABTKV coggins_l_Page_051.txt
0fbe81cd1f0416cc192049b1727faa59
46b79556a14dba00157993fe58a12f711a04fe24
2060 F20101111_AABSIH coggins_l_Page_099.txt
c7444ec8023bb893e8e5cc5eb75d819f
54b7f774688fd9578340ae169369e5af0fcf142e
F20101111_AABSHT coggins_l_Page_171.tif
194123102b981c071d5082568b64b21c
2b1ce32b0e0969fed929e86960fc73f4e500c044
2018 F20101111_AABTLK coggins_l_Page_067.txt
52b3212447d17fa1c31aa24405c14e73
6eef41554131a37ecc499b83c8c244a186681786
647 F20101111_AABTKW coggins_l_Page_052.txt
ec766279c353f6987d7b3d11155c6fd2
0b7c90e66b2eeec0a8e940de683586c971c6472c
8669 F20101111_AABSII coggins_l_Page_001.QC.jpg
7dda5b3f4863f47f7a6e4b581e233923
fee782b29567e642ba39ca10a4817c9238ab3720
36288 F20101111_AABSHU coggins_l_Page_019.QC.jpg
bb4ef18a68c401b3502537c2f4b4eb7b
1212541079ad177ce01631996a90e5cff480ea66
2116 F20101111_AABTLL coggins_l_Page_068.txt
b6346dc0d775c2a140472118a3b05233
c7e7c283d9f285673edd7f5a74adcec3e73737d3
1008 F20101111_AABTKX coggins_l_Page_053.txt
05e3d76335d59c27fc0eb4c6dffda0a6
ce27fd49b5587e6edd9d818d782ed4bf9610b0d9
F20101111_AABSIJ coggins_l_Page_079.tif
3795f0cf97dfac2f5cf24c382082eb59
3df502ceb41c58258cfbfed79e6a56235a202de9
F20101111_AABSHV coggins_l_Page_161.tif
a5ff749639f439ebb9c855a2ab8ac484
3e58eaf72b362b4590f44ae20e74bf7b83cfeb14
2097 F20101111_AABTMA coggins_l_Page_086.txt
10f3d14725a7e7ad0ac44d2316af9263
b4e4380485212fb397eff0be417af06d84c13dfd
2078 F20101111_AABTLM coggins_l_Page_070.txt
fb8a00d804111db2f16fcc389947f34d
7665237efc8b5401c72e335c3d30ac1eea0af3c6
739 F20101111_AABTKY coggins_l_Page_054.txt
2a9b7a6cd4b5ea5dac0fc6d4c04561b3
9cad7d751609a1b20122679bfbfbf41991a651ca
851925 F20101111_AABSIK coggins_l_Page_110.jp2
b29e49e88f50da6d6155a77e26b2df59
42fbe4b669b84c69127c975e637446f757b60a57
F20101111_AABSHW coggins_l_Page_123.tif
48978ceb35a3ee1edbee282b1c716b89
a8e1dbe2001a3303f5841e761235ecca90684fb6
2082 F20101111_AABTMB coggins_l_Page_087.txt
bce4d5174d8af68d6ef513642872ffaf
4d826972aefd4b12f3110cac7e58f608ce0a8a22
2193 F20101111_AABTLN coggins_l_Page_071.txt
cefecff13762bb4261ad5ed434c1e4d0
045a813487f84655430926531e62992ca49e2005
1185 F20101111_AABTKZ coggins_l_Page_055.txt
3bc6c66e08454309597955fecbf9ab05
45c5a3a45b67b886f8b0648878e67a47d79b6d23
654 F20101111_AABSIL coggins_l_Page_079.txt
91ac34dcd9ace71c741e7df5b2522247
f51b550bc0e611f984f4cba8811246b03c231243
36762 F20101111_AABSHX coggins_l_Page_166.QC.jpg
2f14496e571086e4fc48a4599d1e2735
f4f5eab1e7e422de2d0dbc1963f1779e15eda47f
2201 F20101111_AABTMC coggins_l_Page_088.txt
0d1eb7038f3d223636fea8960a952d3e
76c663e115da7701580966c567899024fb5a31e4
2179 F20101111_AABTLO coggins_l_Page_072.txt
57af97bdf06f0d8be7a0fc0a4794e049
2a7807fc220f9f4554c894d9c193bbb4ad4814e6
1285 F20101111_AABSJA coggins_l_Page_173.txt
22239a22854adb5882cae8f65d288364
e2e41a4b82f2eacecd30a0782a9eb72fed8576b7
F20101111_AABSIM coggins_l_Page_004.tif
3af4b0ac968068cf317e51001c4748e5
56c3926dbdff0ef068bcd3124c4b11c0418a74b4
50811 F20101111_AABSHY coggins_l_Page_081.jpg
1b3a1aeefb4b004ceca22162f2d84cf2
883e89da3eb77817852fb189a44214ba65b65767
F20101111_AABTMD coggins_l_Page_089.txt
3376c59fa3abbc2ec7a55d851906ccd0
0d7902f674f4315a3ab0b82b2c7190854fcfc1ad
241 F20101111_AABTLP coggins_l_Page_073.txt
6d3d5c28720cd91ebef4074d5b6efad9
c7967b72363e7e2bebd1eac2c62009df3307e33f
4181 F20101111_AABSJB coggins_l_Page_002.jpg
7127c2b921a923e8baaf0c715760cda5
62e316dfda3bf4ca4c06f45fcfebcca4ea541daf
52124 F20101111_AABSIN coggins_l_Page_150.jpg
083f5ad082fb30bff4e0ad88cc5e249f
60ee5040acbf7ef4dbeadaffb063ac583ddf2f11
108058 F20101111_AABSHZ coggins_l_Page_014.jpg
541310561b86ce6c48f0ef0c60fa2340
230716f8dfb3e2e036046a6e0359ca007eb310b0
2049 F20101111_AABTME coggins_l_Page_090.txt
14412be5314c13102372671657b1d792
07ffa35f83ef1e1782a345d1dda44fc5b419945c
326 F20101111_AABTLQ coggins_l_Page_074.txt
89df6815db707660b25e4d7ebece7233
87613a9423c7d1942a9e7b931263adb44e4b072c
51993 F20101111_AABSJC coggins_l_Page_144.pro
ce5ae1d15aa64e97f3532a049f2ada71
c4795c40407e8cb61bc7824e92d99e5d86291692
1274 F20101111_AABSIO coggins_l_Page_120.txt
4799219af4f0911ffbe446add48653ba
78e90c49b8f2f1c8488dc02d8db3712ab45f174e
2166 F20101111_AABTMF coggins_l_Page_091.txt
e16891bfa012b1dd36f045ba11c5f732
b0001e6e1aaba8f8a55fc463d74d53c2b5eb3178
289 F20101111_AABTLR coggins_l_Page_075.txt
3e992fc9be37eff86e8ed6e7cd09aed5
1b53bf9bee5888b94266036664ab2edc97ac90af
20451 F20101111_AABSJD coggins_l_Page_122.QC.jpg
4fb07db084b9c121f475d032728a6920
9658056d4e2971e6ad9559b93a282dd47a79892e
110149 F20101111_AABSIP coggins_l_Page_023.jpg
d6dc8f524404badd2194cc4465e6d098
646324ea293709e203df2ea2037cad957dab9f66
1805 F20101111_AABTMG coggins_l_Page_092.txt
81db9deb17a095aa1624baf35b454e6e
7adb43ad0936690d1f42fafb561bf816d95a88fe
295 F20101111_AABTLS coggins_l_Page_076.txt
8910c560cb41048d4369e7d4b2a6d75b
d14a02a089aaef709a211627f5e61b6840b02da9
884 F20101111_AABSJE coggins_l_Page_111.txt
b54188093e66f65ecc290617b17e0c32
3ef57a59db89159a45a8d1964cbbe7bf64642915
17880 F20101111_AABSIQ coggins_l_Page_126.pro
910a259ba734ff543b2e5f62a45e55f2
1873213554269718c902d29b7c4d18404f970f37
1768 F20101111_AABTMH coggins_l_Page_093.txt
d07425ffa228f469780c455776d75865
f54dd441dd222c7a3dda3566117e1cadbfc9b682
1156 F20101111_AABTLT coggins_l_Page_077.txt
4cfff5801c49e5c6f99bf787bdae52c0
d057a74304e29d7e0f86a2eb883026f47a001d0c
F20101111_AABSJF coggins_l_Page_068.jp2
bf17250f8b09cdaa1eeb650160abc775
a1a93fa7501b27e46e6c27bfe96e10169a1924ce
68069 F20101111_AABSIR coggins_l_Page_165.pro
24a4f594159285a70bdd5323a7e62edb
1f778adde7100a3e69aa9f988c7012d78befcb6b
1862 F20101111_AABTMI coggins_l_Page_094.txt
f2673a118983cb478a2ddf93b2d318cb
f9065372196347ec12161a3f2c7af693cf3855d3
588 F20101111_AABTLU coggins_l_Page_078.txt
e703540679d5bd93fe8e1d8f873168a5
77962b5466601975165d230947e3a77ba60d55ef
5384 F20101111_AABSJG coggins_l_Page_128thm.jpg
dbcbc7852fb977e825187c4c73cd249e
6fe3f92b2809730e6ca869bc80fb76e57426aaef
8976 F20101111_AABSIS coggins_l_Page_135thm.jpg
277ae6872044f7c3fb658459f3aceb25
8db4600078c20a943eed34d53f42ec298c78f24f
2085 F20101111_AABTMJ coggins_l_Page_095.txt
2ef5ff782399f3096afb84ea1a1afe59
c43b23fa9e7e02956e02251383dfa7aec5c4c3eb
933 F20101111_AABTLV coggins_l_Page_081.txt
65faac6d1c66d35379b6579e615eff1f
f4f478f1d66e2c7751ba8e6b608ec8fa0ef87fd3
F20101111_AABSJH coggins_l_Page_046.tif
81177c8bd664b2d5240719f39d9ae21d
6da9c3c403988fe92ed159cde0cb4aea99062ba6
6533 F20101111_AABSIT coggins_l_Page_061thm.jpg
059a2a828bb3ff82a38dff18b23e19c8
4a12984868bd969d5fd2f7ff6f3d308ddca5697f
2177 F20101111_AABTMK coggins_l_Page_096.txt
8c0f93425ceecb67c4d074c809185e0f
74e73eb71d17337474fc99cf22f8ef1502227406
771 F20101111_AABTLW coggins_l_Page_082.txt
1eeb92049c5bfd28ad56040bbcd0b2e3
3e8969d2c19723db494ae700bd52715eb06cdc2b
38472 F20101111_AABSJI coggins_l_Page_171.QC.jpg
954da928a47e63b25e308c0686842bc0
b8b05c4047d974a5bec98c925b080a5ec1355a79
111051 F20101111_AABSIU coggins_l_Page_069.jpg
74b2b70d826cf6d6ba52e6a3221ff54e
2c1f9672771304aea89bc31a4593e0ef40e551d6
F20101111_AABTML coggins_l_Page_097.txt
f0322ee28e70c6d9c1cc9b8910b48dda
e591bf598b8c49a1f8e3fb4bfc3947ea302a2504
2175 F20101111_AABTLX coggins_l_Page_083.txt
f8288598119806cad784fca4c29d8e9b
5ae88f47ef2d94af1c37a0de38427d8a73023b83
14059 F20101111_AABSJJ coggins_l_Page_048.pro
8dc89c1ddffe2835d55596cb337b83cc
f268001efec0d4cf02aaf092ffe5bd08b8fe0a6e
2102 F20101111_AABSIV coggins_l_Page_022.txt
818c3f0efc72f561087bc212c27bd6e1
f1bf90bc1be32c8c44f8f5ae20d241e05a2f2adc
1436 F20101111_AABTNA coggins_l_Page_115.txt
a4bd38b37003d5e007484922ff0f66d6
007436f27a3abe887906d52fb024eb16121c32bc
2074 F20101111_AABTMM coggins_l_Page_098.txt
7be327475afd77a66aa4fed0b30e68a4
d096192993a94b7bba0e11688b41ea358aabbe4e
2076 F20101111_AABTLY coggins_l_Page_084.txt
78de920ccdc6c33a0d0919cbabf9883c
12976e3ab725c931d897115ec1ee67d4db6b6de1
421662 F20101111_AABSIW coggins_l_Page_125.jp2
1b83e3a31aa3a971761995200d79e3b7
137661f80c22fe89d443ab61d3b3544d9759aa7f
F20101111_AABSJK coggins_l_Page_088.tif
34f8ee6374f401d11026f181893c90f2
990e760ce05c6f20b83281485b10690fc5903b45
608 F20101111_AABTNB coggins_l_Page_116.txt
a855ee2ae493803aeaee58548570a4ea
cb58874ee8d90192839b3653df95386169750554
2030 F20101111_AABTMN coggins_l_Page_100.txt
c49278c267b946b5c1ef179ee62fb688
33cddc2359d1f0ef4dd0fd78371a17e138938cea
2065 F20101111_AABTLZ coggins_l_Page_085.txt
a104a39371347d76f4ecacee0363a1e4
c501d0589c0b22727eb01973fc582573fbf14734
6761 F20101111_AABSIX coggins_l_Page_055thm.jpg
d6401d6c62cc39fec86c58826430402f
94138307f18bd00d844e44b690060c4c1cfdca9d
F20101111_AABSJL coggins_l_Page_026.tif
5ede3c9ac69532790e144e790834fe21
682b421d0fb29709708f8324b4e2871156e6f16f
839 F20101111_AABTNC coggins_l_Page_117.txt
f6f632070fc893eaeadc488e00ca3a60
4ece7ce9272e50bffc030f61f05d858320790552
2072 F20101111_AABTMO coggins_l_Page_101.txt
31ab83829791cfce2f59f6126b736276
b58b936c600affac05effdb6a70721aa9a24a9f0
8240 F20101111_AABSIY coggins_l_Page_063thm.jpg
0c2a1e1da3442648ce7f60d8d47dd6ce
44641c1ee171d499a5e09060e8f3d9e4f6213031
8830 F20101111_AABSKA coggins_l_Page_026thm.jpg
6cb84cc6d329661df2357ce1020c2eac
d81607f8a0523f2c5e24f032fa242f7cc5763140
24942 F20101111_AABSJM coggins_l_Page_154.pro
4fe613229051288197df77622eb978bd
4e92cb8c5bad55245e19080f5f4d7091fc190bd5
749 F20101111_AABTND coggins_l_Page_118.txt
d0fd827cd52e20daeccd93948dd4288b
e02f2abe14b4bb001ca191561cf18aeb0ab1895e
2088 F20101111_AABTMP coggins_l_Page_102.txt
f73370dbd994d7859e7dfa55aa06aa73
438fa1c226146d28398224fa789deadb12238f7a
F20101111_AABSIZ coggins_l_Page_001.tif
365ae5d9025b66b6e94fa4705f536a82
d70f07a791c33884a324c498b351627b980f624b
1834 F20101111_AABSKB coggins_l_Page_065.txt
ec123a6764e62f9ccf353d0852aa2d69
f6eb3c93547e0360312983211ed634ac39397de6
F20101111_AABSJN coggins_l_Page_128.tif
c1548dc8efa1c10a00ee8640cfcb4817
d51ce9690cb0cf8caacfae8ccd6fe4f9a8e18908
1016 F20101111_AABTNE coggins_l_Page_119.txt
9503663d9e04a85a2ac102d7132ea64b
f4cc2b9cec290cb428a2e359f3d180aad739c66a
F20101111_AABTMQ coggins_l_Page_104.txt
09ed7bc21bb8920f08dc4241cf228897
674b68c83494c56600aeb8600b74dad30877ccf4
351 F20101111_AABSKC coggins_l_Page_080.txt
397d91bdd860cc72cbe6fdb03947dea3
f6f5ae8580c515a1407dafe4a252382011c55da7
19081 F20101111_AABSJO coggins_l_Page_154.QC.jpg
9cc9e5230930e8d6010d559eb54072a8
ac7f64b189d73b38b0c358d45db839325f10f5dc
1001 F20101111_AABTNF coggins_l_Page_121.txt
b1bde7f795a27a0487386c1ef67a33d2
db8bb9b8767728decc70db5ed830137393364b28
F20101111_AABTMR coggins_l_Page_105.txt
10ec1ce5823859dbb2fdba3bcf5d0084
007d9c977512b378faaa98d7ef293788fd533ebc
72886 F20101111_AABSKD coggins_l_Page_112.jpg
a363e29d32086762442b6457804e280b
eca202a94df5c47e4b2c9377da42a4690c463c54
7057 F20101111_AABSJP coggins_l_Page_046.pro
74f271146b886737082c02e1343ebd7c
4f58be47ac83cd075894d3ef97472351961b915e
1187 F20101111_AABTNG coggins_l_Page_122.txt
ff0b07d3700d2b4778a94b9a4db950fb
a4b5f2b430a8fed7df727957babb397f75536a7e
2153 F20101111_AABTMS coggins_l_Page_106.txt
b3ee360e922e8068d3ee8e3674317f4e
79d1d57faff7a63152c5543f6bace7fe897ef516
1018 F20101111_AABSKE coggins_l_Page_002.pro
2c70f22108d2d0cac567f0562ac4eeee
f9ff5aa271eaffe55ec01c4d154822cb4541b709
1106 F20101111_AABSJQ coggins_l_Page_127.txt
a1117c902927415aecbea0a4ccbdf4ff
b1cccf49874812ffe005cac4e9194ad820e2201e
1838 F20101111_AABTNH coggins_l_Page_123.txt
398b83f9cb7231ac0cbc5dd96cf8be97
6cc90a9397a653fc12fb29b43d32e7cca425cf73
280 F20101111_AABTMT coggins_l_Page_107.txt
419ffb517588b3a6621de6f8d4781c4a
e10418272afb66c4e863e7fde717270609b9293c
F20101111_AABSKF coggins_l_Page_094.tif
d485fba4fda0ce5c23dcc4ac5e9f3b12
0f620ededd08946398ce79d85e40b96c98684e27
51707 F20101111_AABSJR coggins_l_Page_070.pro
554e58697ab28aa40afb954f315f358c
8d6b4d3cc8cf1387b0435827f34b0bf124aea490
1060 F20101111_AABTNI coggins_l_Page_124.txt
bbe65b59e31670b44ad3184f5f8712a0
ba9ca3f121dfc47326fe2dcaf6f0b5938168a8c2
332 F20101111_AABTMU coggins_l_Page_108.txt
d51de952a80e382865d039620abf9334
777f10fcce3aa0641378df05307cdc1a89a6e175
35883 F20101111_AABSKG coggins_l_Page_021.QC.jpg
46725b26394e04324d9e576d7c5eb9a0
c78e038ad812a2c650c875d88b273a0dac99da38
F20101111_AABSJS coggins_l_Page_110.tif
4210a257f3f50d6e8ba9d432fea6ecfc
9f25111077d1ee3cf279d77d80488da1a081f1b4
923 F20101111_AABTNJ coggins_l_Page_125.txt
4e68d285ede51c196e7802a0ac4c116e
af8f7ee0294675921c6e6384b32857f581f73d03
353 F20101111_AABTMV coggins_l_Page_109.txt
4936421f74e49da7758b3252dc923adf
5c10d1b5e1c24f248f39edbbda2e9e96607d1ade
7859 F20101111_AABSKH coggins_l_Page_006thm.jpg
417a78627437035b10caddc2e8e86fce
28fc1c73e427433adb8774c2e2c77de133bd46cb
108200 F20101111_AABSJT coggins_l_Page_103.jpg
49351a4ddd103cd360d14820d70925a5
66e5dd15e4a57eefa82dbdf86befe574fc635ec5
1013 F20101111_AABTNK coggins_l_Page_126.txt
51780ecfac64af2ee0fa122ee208205d
4f55fe2adf56869146bd1cfc554c07faec4e6a58
1969 F20101111_AABTMW coggins_l_Page_110.txt
618e5941dc4a29216bf7585f7784c3b3
c930b2b721822ae2ec6131ebdd7ab68da0fd376f
484 F20101111_AABSKI coggins_l_Page_045.txt
8c9e5669045d2231f01c1aa6bf6ab588
71fdb463d43bfc7e604f1cb05ee2e5926ed6072e
60650 F20101111_AABSJU coggins_l_Page_113.jpg
f1d7afb6a5a92a0bce1d54b73df2ae9a
717f016eeea292de99bf090fa5e922f5de456872
1582 F20101111_AABTNL coggins_l_Page_128.txt
68254773bb7dfa67afdd863fe406a00a
919f6efe8c432a2828db32dfd25fb96b00c2b454
1326 F20101111_AABTMX coggins_l_Page_112.txt
dd83d103bee3c08ee8a6346d00bb3c6f
91ecb3b08a819c067dca32549a24a3119d9c75fe
F20101111_AABSKJ coggins_l_Page_096.tif
108322eb52a8a8800028b0320d4ee156
cb55bfea3c4581a4de0a3c4d9a4854cc06ee495a
F20101111_AABSJV coggins_l_Page_137.tif
04fe7dac641f4e6dcd610d731b183e94
5b937c0b3647e9eb07417f34fa1c90db031c82cc
2185 F20101111_AABTOA coggins_l_Page_143.txt
e8cf2c1ba1e414d017d2bbea503b385d
2afb115a686d2aa5ab039456723ca7b670e7d322
1909 F20101111_AABTNM coggins_l_Page_129.txt
5021d76faad8a2acdb7a0e7db681cf86
bbf412f63bed65d8bc827110ba1a7eaff29429b6
786 F20101111_AABTMY coggins_l_Page_113.txt
3d9b786db00604c1b55fadb6ed560d59
1061615ec4f5c22d36e236eacaf87a0351576e1f
47132 F20101111_AABSKK coggins_l_Page_011.pro
402f008bd2a13e6939a0571aa554af62
4b971fbb3ea8ab500d7014980dbe092d8a4f9dc8
9238 F20101111_AABSJW coggins_l_Page_058thm.jpg
50a8a1c90b64dfe4bbb12bf980b8d2df
71cf2d08edd87e96247d58d235e6bfc0e976d45e
2050 F20101111_AABTOB coggins_l_Page_144.txt
8a721d5e21efc17a7db4f56fab7771bd
9c1bfa7be4390adbbe356295f7efa854a2e60124
900 F20101111_AABTNN coggins_l_Page_130.txt
80eb41237b60c5ef044ae63d2cc0706e
40faf9fb079dfb683ace9d4474dcfc4b7a48ddad
1223 F20101111_AABTMZ coggins_l_Page_114.txt
c7959483a8577750cca0f0c49d9bc948
8dd05821783dd42f82ee67934d146187856e13f9
11579 F20101111_AABSKL coggins_l_Page_149.QC.jpg
feef35d355aba885477ea71ab824bbb5
e6de0208aa1eb15dad31f56be7a675594885ce6b
2100 F20101111_AABSJX coggins_l_Page_029.txt
33b31754b509b552f27fcd01bdf64751
96af25444f88b598da629950d719108518ef0bc4
2273 F20101111_AABTOC coggins_l_Page_145.txt
5fc861bb8ee6775f176de28640bcf596
a26ba455a514ad12e456624a61d35110e076fd65
746 F20101111_AABTNO coggins_l_Page_131.txt
9a19fa0f365c8fe466ca37e06bc8c218
226a1232385345a954aaeef2240b25d3a0469cab
200433 F20101111_AABSLA UFE0021996_00001.mets FULL
c16a20bba4304d45e4de7e572f535178
013419d61cfafb01aedfe1e45a001953cc873a22
8820 F20101111_AABSKM coggins_l_Page_023thm.jpg
60f77ef58eecd24278773f228f3c587e
8fb9df69d6367bd8d4470545e9b2f187dbb7d362
F20101111_AABSJY coggins_l_Page_021.jp2
bda1912347b42e86c0d2d5bfe16eb34d
60b85c4e06cec7fb8f2388af8ccbe1eb4cd186a2
2197 F20101111_AABTOD coggins_l_Page_146.txt
18390a0ac89feb7b029c24eb89454eb3
5c20aba877993007ad57b5fa378b22a79a08e1ba
734 F20101111_AABTNP coggins_l_Page_132.txt
8f34fe2b6af4406ce01564986b1e332f
ad6945be1834d6783195c4523d27ff04bc8fa023
49793 F20101111_AABSKN coggins_l_Page_025.pro
27f0a165526573ed5fb7b6ba58190b52
2aa64287ba85376058e135eb797842a61d4a0c7b
22300 F20101111_AABSJZ coggins_l_Page_156.QC.jpg
1194cbc01b6eeffad7ddf91cfa5a47cb
168a396b0e592d470ecd7ff388b9dc443ce948c1
2127 F20101111_AABTOE coggins_l_Page_147.txt
9a3cc15cf361e97dbe9bb7affd665527
6e47efc1a36ab6931a8bc09df647992ffa1367b7
1494 F20101111_AABTNQ coggins_l_Page_133.txt
e4411a45bd37032912308a0abf74d08d
2cbb0079eb9e6f5950b95f2d52df6cce0b718756
F20101111_AABSKO coggins_l_Page_068.tif
a0566c5a71444fab7c88eb48193afa1f
f34a747538339439b45fed45c4292b9eb6dc0fdf
2131 F20101111_AABTOF coggins_l_Page_148.txt
566ae8600d666fbe733da7366b564aac
1b7d399ad2d12d94894a371848f0869692b1a2b9
2178 F20101111_AABTNR coggins_l_Page_134.txt
f0ee4b599ed931763734e59f9e91bdbe
d492bd3e0c12c22bfa4208a27344ed6bd6a93b3c
29233 F20101111_AABSLD coggins_l_Page_001.jpg
02bae2b2eef468499e02149b8ee92ab7
71b9d06613419d341a834956a158aa4a6fc681a8
113306 F20101111_AABSKP coggins_l_Page_143.jpg
d5118222c79c09f4d83d3641ccd0433e
c1de6cd7c73c91c4f718aaebf5d564d100a8531b
607 F20101111_AABTOG coggins_l_Page_149.txt
3467a59bcbe3833cebee129303412abf
08cba1b1fd67b2a8e7e3026dda56eba00fe1ed36
2051 F20101111_AABTNS coggins_l_Page_135.txt
b4bf7ca827a40c068d983223ffec1fa8
bf970e0c7f6ddd781f1e93ea4493474679e121fd
2880 F20101111_AABSLE coggins_l_Page_003.jpg
b69d10fd8cc0305f3875d071f6f0d27f
97e11d78f66fa2bf7f4786d6b977a44b05574013
52763 F20101111_AABSKQ coggins_l_Page_014.pro
0a3069cd336332b4d7098c9612ffc852
02f322fda7d20799bc518cdf4dc1b2cedb316145
F20101111_AABTOH coggins_l_Page_150.txt
43296bfc6f3d62c1251fae5696a6b7b8
9d35c3e90e15672abc3a45ad181d15759a796db2
1995 F20101111_AABTNT coggins_l_Page_136.txt
4f2a3bb1611257003cb56ac9f1f262f8
2800b49dd878971018b9e6985895577eddb77054
110155 F20101111_AABSLF coggins_l_Page_004.jpg
5cb7663510204225aa69dbb3f9ed6ca0
14b24995a060178d0d54dd7e00aef57e99eab994
36846 F20101111_AABSKR coggins_l_Page_134.QC.jpg
e909d4905c5e8bdcae540f00595a49a4
a599d256e01545abdad188bc88bceaf9f3c6c0a6
774 F20101111_AABTOI coggins_l_Page_151.txt
27851126554cc9cd45bfbf4109862d8b
d024ffcb601ff942a704f80cc471f34dafd862ae
1993 F20101111_AABTNU coggins_l_Page_137.txt
eb9b305b9922e9697e3de440cbf4a6a9
a88e25d7dcd2567fd6433ac6e9db9336d6be44a8
130102 F20101111_AABSLG coggins_l_Page_005.jpg
e61155c09e27a1cb3ecdd84fcba0bf8e
abfb1c5ed4794e4c339f13f08a1e4e966db2ebe4
F20101111_AABSKS coggins_l_Page_018.tif
a08545a80906b49b0f7650d91a528196
d8f5ca580898e19b9862485942ff6965348cd5fa
F20101111_AABTOJ coggins_l_Page_152.txt
40382d5d5be1a9e238349046e9fb5e2b
c63cf500cb7b2c2b82a6c2cd7e4281718762d196
2040 F20101111_AABTNV coggins_l_Page_138.txt
4398b37fe5bb823b4e4eeb7bc0c33bc2
c3e0f509865b18c352e29c197f9331ec50c06c5d
159134 F20101111_AABSLH coggins_l_Page_006.jpg
9db9f4da99f0666385fd6a0cf3469937
a0b5e6abbdfec082e9c1a77612b4c17c39fdf97d
35875 F20101111_AABSKT coggins_l_Page_095.QC.jpg
6a062fe62fbe8a1bf9b2712406911146
2b8d65422aed8ff343dd4c41c2574bafe20692d6
712 F20101111_AABTOK coggins_l_Page_153.txt
f772cd33bfe6752f11aab58b933b555f
a864409d385a932a2b534b5c1c23db161282dc3b
F20101111_AABTNW coggins_l_Page_139.txt
03ff41788dd9ed1393d23de9444eb19a
56df57aa3700958ef14394c60958437056a9752c
70553 F20101111_AABSLI coggins_l_Page_007.jpg
fc0fb7a9a1775f3afe04a9b32e2c1d41
49bfe8d2c6a8764a0b98498267d1de749d0df39e
99913 F20101111_AABSKU coggins_l_Page_059.jpg
978664dce3e3e7a677975a6e14947358
00c58dbe3a970f889c8bde61638d0bb0ae795926
F20101111_AABTPA coggins_l_Page_001thm.jpg
978bf7173d3504d2bf560dbff492cfbe
fa4b029d169323df4408b8ace7bcd14d4f285176
1208 F20101111_AABTOL coggins_l_Page_154.txt
41da89a3aec5dcacb219f510b48840e1
801a97fc20f9c6bddde06d4f57cf5927b61a5f54
2136 F20101111_AABTNX coggins_l_Page_140.txt
98b1d8f61b555875b5dad8dcb01754da
4e888f3d4987ac72d97245006dede27096193095
131664 F20101111_AABSLJ coggins_l_Page_008.jpg
b014e76c8db6fc706f483539a39c01f4
58a7b76ba26d85f2c6578458f531810907bf4577
38409 F20101111_AABSKV coggins_l_Page_038.QC.jpg
a8005b891bf54e0466d30284e546454d
1c62cbfce91fd4338c2869a195904d2943a728b8
F20101111_AABTOM coggins_l_Page_155.txt
83f9dad3b204d5322a9d4e1da9387d26
e1e4acc04d5e31dfa220f03637975390d3c26b66
2099 F20101111_AABTNY coggins_l_Page_141.txt
8acfff5bc1beb22056d78b8358d6644d
aa456a15f0e65ade1bf7fc5bbeaf46eee64868d5
170286 F20101111_AABSLK coggins_l_Page_009.jpg
9c310166dd54e64a42e1ff43cfffff82
0e7609f5b4e73a91175332459fc204787225049b
45577 F20101111_AABSKW coggins_l_Page_015.pro
22e30cdbb90c84e442041567dedc18ca
6a18dbea457ae4093dc342961dbd95a6fa1bfd04
3384150 F20101111_AABTPB coggins_l.pdf
c79570d1ec99281723c43b9d828d97e9
54663b6bb620cf262e6637fc907c7fb63cc626da
1091 F20101111_AABTON coggins_l_Page_156.txt
c3cca005de4ff81b38f96c0eb6f068e9
a272ce40bf8f5353d71a00ca1e9587953a26af62
2122 F20101111_AABTNZ coggins_l_Page_142.txt
38e55b3cda5be807fba37af96c39162a
3cf0c5fdaa121e30c2028e5869a2354ce728226d
150313 F20101111_AABSLL coggins_l_Page_010.jpg
ab5779412956de7e681ddc1e3c78e0bc
92745e980500c89e692b297e66d6611e0e27f174
9136 F20101111_AABSKX coggins_l_Page_162thm.jpg
0a85cd8238f4d2826547f1577b733bcb
77b564fe093c05ed45ca71799926be524ce3381d
23023 F20101111_AABTPC coggins_l_Page_055.QC.jpg
3ecd0f9fdaa1d9426febbf48d9450f8b
2af8496983a88a0667bdbe5fdd7a2aa820723a8e
809 F20101111_AABTOO coggins_l_Page_158.txt
b6d13e30b5332018155f18f5171d2183
91cd7ba9b3953c62be6639b61e38a36c249e8b78
101613 F20101111_AABSLM coggins_l_Page_011.jpg
524b3f6d1c9bd683687be8d3eeb517a4
746750ab5cc1b6f295c5f7536dae7ececc343bd8
2160 F20101111_AABSKY coggins_l_Page_069.txt
005d5073fa375b8a0cbab6db3b005d3f
2090fbc34a79c2848a5a72adefae3f43a3db9fe1
79711 F20101111_AABSMA coggins_l_Page_028.jpg
717260057e97e9fe661e926868a2cf8b
2b3cf58a01cad21c6962062b7a101a54608f1d85
32551 F20101111_AABTPD coggins_l_Page_063.QC.jpg
363b173ea8d2d1053c5fe3d6b039121b
2c92ea106d0801d9b23946625bba2c3e964c79f5
2556 F20101111_AABTOP coggins_l_Page_159.txt
7edeb987961708b1bf078c56a0a49f30
73ce72ccfe0862f599d3636ac4427d9ddc509f27
102626 F20101111_AABSLN coggins_l_Page_012.jpg
ee7cbece35f94d9f1225447c87200e7e
3257c55aa8296bb3fb405f957941a987ea53978e
F20101111_AABSKZ coggins_l_Page_067.jp2
34135c65b8412a3159e3612d362f3ca2
02ed2aaa22e335954d87996a5c4b96b79c8f03e4
104580 F20101111_AABSMB coggins_l_Page_029.jpg
cef1e503801b829e40f59512c6c6a96e
b1e9fc8516358d11f17e4339cd6b4011ec42b9ee
5262 F20101111_AABTPE coggins_l_Page_051thm.jpg
b6094926cbfc3dee8d30f8f0222e2f4a
922f33182a3e92ff80e180d4adb4cda047c20d82
2751 F20101111_AABTOQ coggins_l_Page_160.txt
004bf57c5c95d0aaf411787de2c8669c
ee9787da9c1d19e6e529cb671484781b1981bbca
110514 F20101111_AABSLO coggins_l_Page_013.jpg
f8682d4994f66c992c56e1d2f4b825f9
e68a53ddb774b4b1b97d18dc5aa86ac15ce35ad9
108548 F20101111_AABSMC coggins_l_Page_030.jpg
c9dfbbc937023b58374e3f421ec6fb91
51de4e49bf634daa920f8db699890e1feeee83fa
8763 F20101111_AABTPF coggins_l_Page_147thm.jpg
415df7953548b1d386e92459c67bd868
675e1a98e95cc3b44882b45918a94e5e79e7503e
2823 F20101111_AABTOR coggins_l_Page_161.txt
740c553f330306751d43f497210df9fa
bb9b0a4f7aaf1c955350c64525ab79a1af34ac33
120343 F20101111_AABSLP coggins_l_Page_016.jpg
025461dfde934726da463173eea16773
9de65b32cc2ba43ee922352c351e5472bd65a7be
101367 F20101111_AABSMD coggins_l_Page_031.jpg
a962d1c513f9a56cab4fe416de70557f
ae5fc05fa2b630eefc0f3d0c406e94d3d358014c
38674 F20101111_AABTPG coggins_l_Page_016.QC.jpg
7b931428e094c375a1824b3ed2bf5af7
118eba4c41c7b89440b037740b5e27e938b9a95d
2588 F20101111_AABTOS coggins_l_Page_163.txt
1d35e417a8e939df6a317ee8890639c8
8512900d6ddedb2bd52ec2dd901db4a6c950d73a
110389 F20101111_AABSLQ coggins_l_Page_017.jpg
ba525bd85caa9f113c64abafdad363d3
d597ef59a620baf22f098e05ef2b46f2708896ef
105855 F20101111_AABSME coggins_l_Page_033.jpg
03f0794eda799a6362dc6de88a967fe2
832c1185eb14055ed3921be522cb2deeb252c4e6
17343 F20101111_AABTPH coggins_l_Page_123.QC.jpg
513f6f69088381cb1fd9a646b880bc6b
3e0f55cd623b33fb77de7e8948b1eefa85651b9e
2930 F20101111_AABTOT coggins_l_Page_164.txt
093919cfc96f4bb90d9bfab33a178131
78de8198c1bcd92676de5228b17e4aa85e31b600
112414 F20101111_AABSLR coggins_l_Page_018.jpg
cb79911fe80b14208edc2abf9b39612e
337dfab9e2c7015a52f5cd7a1978985e55c58d81
115758 F20101111_AABSMF coggins_l_Page_034.jpg
92e74c2b328393eafd7e7522dcb46ed3
ad34f5f5cb7e00b06ce6cfdb95f47577be1f42eb
38240 F20101111_AABTPI coggins_l_Page_096.QC.jpg
238c3fd49ad355b57b68fbd5e1759f41
8292253da1195871bb647dc536d3197cb3692377
2759 F20101111_AABTOU coggins_l_Page_165.txt
21ab2fe9d9308c48abe717c0a8c93a33
e916806e6fbbba013507a2a29e5f8d0c66bd6c77
111685 F20101111_AABSLS coggins_l_Page_019.jpg
3cc50d5e0aa2ec18bf6ca2a61f93349f
e9234efb4f40a51f8cc1b5d36d85f566309f009b
113257 F20101111_AABSMG coggins_l_Page_035.jpg
c3f633006b104690bf8d83e161af26fc
c9a3842258a0fe7a690c786ed8ac76d3f64fa931
9300 F20101111_AABTPJ coggins_l_Page_169thm.jpg
f482b3f1ec8adf43a11a0847589408de
dedfdcbe85d603f05d728731a70f726d806c78bd
2650 F20101111_AABTOV coggins_l_Page_166.txt
a86e3ef2677f8d558a50baa8547add57
6f97c1605947860aaa4125adb0d781db5d4b3931
114282 F20101111_AABSLT coggins_l_Page_020.jpg
5ccc7c86b364bc113d2c266ef3a38386
5877c7be009b51856b87540bb08137678dabb18a
113087 F20101111_AABSMH coggins_l_Page_036.jpg
45763040e86940ecb4fa41b2f5587c72
cd35e3d739f31f9639314ce0b482ae0fda3872b5
26740 F20101111_AABTPK coggins_l_Page_028.QC.jpg
328b84b448891640bd0397ed0114f433
08a464aa66bf9dbad99a7804c2069a7f2665ce64
F20101111_AABTOW coggins_l_Page_167.txt
f1dcb5e4ad305eb12acf31ec99fdca48
be7c2243bb7420553644364081388ccea0dfe7ab
109320 F20101111_AABSLU coggins_l_Page_021.jpg
78251a7e64538adb07ac9fcd73637793
e77f5c63fd37420f19dfc39ac507e88ae38ceb50
119637 F20101111_AABSMI coggins_l_Page_037.jpg
d98f0808d6215ff91ec9170cce4272c2
50ba716c4944921f78686a20147643ac75f0b0dd
7302 F20101111_AABTQA coggins_l_Page_005thm.jpg
3ab9bf24125398fcb612cf685059dd84
338040f0e3ba133f6d27f7b3749e0e780afce344
8665 F20101111_AABTPL coggins_l_Page_029thm.jpg
586b15fea2f24707ffb743f3d6939a26
890d4fcf89d42c480044dcae348146facfd906bf
2601 F20101111_AABTOX coggins_l_Page_168.txt
bcb1f79e989e9d309fe1eaa2909050de
30bef7f9ca54e5087bbce11f738029e7a0c24e7a
108793 F20101111_AABSLV coggins_l_Page_022.jpg
77b3f017f38731508ec11e0810c1e631
d95b5eb6028a0daa878994bb5e041a99e5c477af
116464 F20101111_AABSMJ coggins_l_Page_038.jpg
b664b16b33cf8b7ee6a58cbd3b458efe
a4355f24521f600495bbbaf2f939038184b9a42a
8713 F20101111_AABTQB coggins_l_Page_097thm.jpg
bda75a8c8ff3a77256444b1034a80970
066f10506cb8a3ef2d0115c4252bc42d29561d4e
20297 F20101111_AABTPM coggins_l_Page_077.QC.jpg
62e995461ef33f133e08a6e5c87f48e5
28d5dbbee8703ba96d16eefa3bca3cdb3d8f6ee9
2861 F20101111_AABTOY coggins_l_Page_169.txt
9610f0400ab14bfd86bbec6e1f9006c2
b37a6d35766c5d21d66c4ad6b69f76c0d7d0d3f9
94294 F20101111_AABSLW coggins_l_Page_024.jpg
f3e59341e7a823c4e99e4d7f57bd6754
ebe4fb05c1eea0abd91a494ad352fd3eca16e736
110963 F20101111_AABSMK coggins_l_Page_039.jpg
9788f7a704a1adabd681c42d1ed795e2
d0ca0d6a69ec58a7ebf0a859ca99b86c2201682c
8636 F20101111_AABTPN coggins_l_Page_019thm.jpg
a52f74f2b0c35e3aff6d947ac3e39c76
2b92fac6049f9c20fdfcc22b66edb9b409ee9bee
1972 F20101111_AABTOZ coggins_l_Page_172.txt
a333a5d7c19be20a18c8a7fdd12eb3f2
9ff2bb922f2948dd439151b681fde4044c30f116
101602 F20101111_AABSLX coggins_l_Page_025.jpg
6d53b3d181465d0f37517c92287b95d6
56196e7a8c2d5897e1f0acc12a9c04ab60d955ef
31138 F20101111_AABSML coggins_l_Page_040.jpg
4a95d9692866b0e47617cfcd7e67a225
5cf6fd97bb02a4fb288ba8e9e2577f9a03f1f671
35287 F20101111_AABTQC coggins_l_Page_097.QC.jpg
9267045a3a9ba2dc5b614537683c73e3
e7dea01672bda8f4ff502b5378efcca872c3bb4e
9901 F20101111_AABTPO coggins_l_Page_009thm.jpg
1461cecd28a36aece71f1ce6cf3ca68b
6aeb9b59e7dac8463e825bbf06fe35d936dc550c
102195 F20101111_AABSLY coggins_l_Page_026.jpg
46be1cae33f6f1fee913f972293ebf9b
aaca1b083036c5ef5a003665c12e204937d06b56
69909 F20101111_AABSNA coggins_l_Page_055.jpg
0174201381ba9c547f9e8d55198237d6
47449e881c61e30546b0e998e3028a12540716aa
64293 F20101111_AABSMM coggins_l_Page_041.jpg
8c7a33c939d309381f1e4985c23c814b
74165e657888736e00e41055c7344d21583d1254
21216 F20101111_AABTQD coggins_l_Page_111.QC.jpg
42eb19918aa1a306d333f3ec50e6071b
79616fa29078d916190a218443313b1c4f269ce8
5568 F20101111_AABTPP coggins_l_Page_120thm.jpg
4f0adbf502d5949e4674e2265789337b
e1ae21ef1d486d5721ad6b0eff42882eeedd4742
96680 F20101111_AABSLZ coggins_l_Page_027.jpg
73fb7e3a9d060ada7701ff38a1b398ae
88cdf1e32d91730d784b3243fcaa851284bd31b1
70632 F20101111_AABSNB coggins_l_Page_056.jpg
740bf0f61e34ce5b7338b1aaf375be98
6f6dcd797fa71545d2f9869d461f0be26de9454a
58449 F20101111_AABSMN coggins_l_Page_042.jpg
72bb9e5db68b4e70cb44c9ab6eb1fcf6
21704746cdd7c9b0ebc6b0be11bf081906b9a142
21298 F20101111_AABTQE coggins_l_Page_155.QC.jpg
e402261e12cab629b775a94615fd9d7a
0ff7615360837fd6b9eed1def56a9a5b24613a3d
34849 F20101111_AABTPQ coggins_l_Page_084.QC.jpg
7790b71f83dae2c31472506246570e32
0d227455453b7239d52d2983ee2aceae45ce8476
115285 F20101111_AABSNC coggins_l_Page_057.jpg
9d474e7639fac51a9628c3dfac618d11
b981ec5fbf22c13b8c409fdc6efbf671b623ea2e
61439 F20101111_AABSMO coggins_l_Page_043.jpg
19f5d376fc075744bdbb23ffe29aca6f
162dc194ff7e39a418251f35cd3f32e9a4ab6838
37761 F20101111_AABTQF coggins_l_Page_143.QC.jpg
fca77e672cc3304aede73279ef146ceb
f1df2efc2d95f50d868700b102089f14f5de6a01
37968 F20101111_AABTPR coggins_l_Page_162.QC.jpg
d48d8cbf6bd6ecabf5a4baa5cb9a2832
1e7b7056e46d4908d59f78534af31311fc3761d4
113919 F20101111_AABSND coggins_l_Page_058.jpg
c09b3ba426bbe9fdb898d43cbf90c352
87ded5737e00d3bfef1f58c48c834527aaa8dc88
54310 F20101111_AABSMP coggins_l_Page_044.jpg
7e11824fe0795f5d64e6fa3b5574fc9a
27c1d5651b4499b40dbbf3bdd166757c4504a14d
5914 F20101111_AABTQG coggins_l_Page_111thm.jpg
07c5f03cba312083b599cb4521d64b3d
050cf31bc1c447fcfc5db5a3a84655df29446488
37173 F20101111_AABTPS coggins_l_Page_018.QC.jpg
9c17d5b4b4b57861c26cad828dccb56e
96ae75da50ff9eb34cbf8432be4688a0494eab86
107789 F20101111_AABSNE coggins_l_Page_060.jpg
431b1212a21d196193c852f181e0ee4d
6727e443822ed3dff403cb9ab757b14e5cf716a0
46008 F20101111_AABSMQ coggins_l_Page_045.jpg
3ca597c53c32b2704e989c72561f308f
8cc627a9028df5d0398f430624c54710e9218ba5
1234 F20101111_AABTQH coggins_l_Page_002.QC.jpg
e6ea370a31f4c61d19e44ca7e48516be
8b8360b374a5b7c96813292d805ee7ec2c0d890d
4816 F20101111_AABTPT coggins_l_Page_045thm.jpg
d2ef09018440b2feb2b6ea222254cf03
6dbac8f0f1edb15fe988f3e1429b5eadbed9a7f1
67093 F20101111_AABSNF coggins_l_Page_061.jpg
8e0d2f37aff26d8e46f5858abedf094b
06c951c1480258728a1ab4af5f3740995b0acafb
38052 F20101111_AABSMR coggins_l_Page_046.jpg
1f61636ec87a792317f248e1469fe586
741d757c75818dcbfc51ce12f8d6ed2d047e40b5
5479 F20101111_AABTQI coggins_l_Page_122thm.jpg
cce22ad2c4f0be42cc1cbe2d48d53ef4
f13727f70dd7166d9633dbf862b44aacede36298
34921 F20101111_AABTPU coggins_l_Page_089.QC.jpg
0505a086ac29f2796e6edd6cafe48f85
17744f3a259645709c38e61d0721444260891bf2
78152 F20101111_AABSNG coggins_l_Page_062.jpg
0ca99746dcc3becf09a90892ab97dc2f
53d65fad12463da1cf65b8db077fbce829ce5326
70256 F20101111_AABSMS coggins_l_Page_047.jpg
fa6e3cc264cddec6a400dc9f31ffa77b
b56a7ef5a4b8030138b3ea6c8beacd4c79cb63c8
5315 F20101111_AABTQJ coggins_l_Page_056thm.jpg
b01ab95f4eaa8ace82a77d76e9728ed5
b5fc2afc7637ae79bc34fae317e0fa5cff917f67
34190 F20101111_AABTPV coggins_l_Page_029.QC.jpg
81b4ddca2c26bbf079f1837bbb0fa795
c2ef1b5d68575c08eb8115ee13d9d98c0eb36888
99208 F20101111_AABSNH coggins_l_Page_063.jpg
71237745e208fb5867eb4dc25e6824d1
f2a5311c2150d6cd764574b145d29049c90f6d9f
71024 F20101111_AABSMT coggins_l_Page_048.jpg
db2ed4c9d11918f1a9a8a2e0f05742c2
fe5d2cdf78152a1f199c94c25c26a7e076791651
5474 F20101111_AABTQK coggins_l_Page_043thm.jpg
66ecebcd14d044c6ae93c6400fd05fec
c883f1a3687141e60e11e11e537f0d9e49a1dc4f
37463 F20101111_AABTPW coggins_l_Page_140.QC.jpg
48edf590eb5db6eed38e0a9685cef528
30f009b69a2136e46f8b2700137e65cff6a79a50
100678 F20101111_AABSNI coggins_l_Page_064.jpg
e4af9b396a950d1b510a27102b543635
561897f48aff4f4bd6f15c051e15c6f88110a84d
60745 F20101111_AABSMU coggins_l_Page_049.jpg
af8fabbaeda4f671bc52260b3fa0096f
b11d967f20a2333f828758e61560ec4ceb43b06f
7473 F20101111_AABTRA coggins_l_Page_015thm.jpg
a9ce4833a4e7501155fc1e1cada26800
21080eff432faa74c717445668ab2ab92cb4657e
30215 F20101111_AABTQL coggins_l_Page_066.QC.jpg
12c79438af3eb4d246813a1f3312c0ac
0911084213821dec19d50ad7929dc4e0b0070227
20608 F20101111_AABTPX coggins_l_Page_120.QC.jpg
f49f963dc8034198642d1d149e07ddf3
5d628514a866b80fcae0734708d8c3fc9f61dd19
77331 F20101111_AABSNJ coggins_l_Page_065.jpg
afb9223433729373e373e773fbbf7fa9
b53ae2c55395058eed842a6ac5f25240330768f2
53545 F20101111_AABSMV coggins_l_Page_050.jpg
76fc21c1ffc54a85be3ee83e2890bf0d
1ec7332069f10036753fde83c14fcf88c3911308
30764 F20101111_AABTRB coggins_l_Page_027.QC.jpg
9828c213b418795a8f74a7b3deef03b8
30a906914030f7253339ca5e36886306df1d61e6
8676 F20101111_AABTQM coggins_l_Page_022thm.jpg
847ff14df327e6ecedd31f88e4039b70
cd1d84d13e172fd128f7b8f40b156eddd799d055
8933 F20101111_AABTPY coggins_l_Page_039thm.jpg
35d6070a9dc2b86f921d878836a5c142
578e65a751f70e5c2aa911aea635cde449e4192e
96858 F20101111_AABSNK coggins_l_Page_066.jpg
91611a0ef4924d9743a61ad848f84e3f
a5324e794dc1420792b83d88aa47cece1d43ae5b
57383 F20101111_AABSMW coggins_l_Page_051.jpg
74ff144ac0210047b8d4d91a7f3d9f71
6a5853bf87035f2ff80f53f577c2f01217859d17
9133 F20101111_AABTRC coggins_l_Page_159thm.jpg
1304569847db280ef653990df6772d4c
dccd584e46b73c9038715f3ef92e3f6ed91357cc
8801 F20101111_AABTQN coggins_l_Page_106thm.jpg
b2cd2fb5c49018ae1f092e6bdd47f6ff
f093728f7ed55b8547db32e6626adae4bc1d04ad
6889 F20101111_AABTPZ coggins_l_Page_172thm.jpg
5a33411dd1f129e82f4a8c79ce8cda08
db588642b25a1ec698a91707a73ed0c60301362a
109355 F20101111_AABSOA coggins_l_Page_084.jpg
674c42ff8ae5a29d9dd9edbea025040b
9bfca6ff176add29ab92f52ecac50003a50bbac9
104109 F20101111_AABSNL coggins_l_Page_067.jpg
543893e81cb982dda502a0d8d03b1f2b
3f54ab5be549b387c17c80734e43b4bf4f358f9a
41709 F20101111_AABSMX coggins_l_Page_052.jpg
c016ac458a98c33edae1ec7486a48355
3ad9793ee01ae4967b13a6e17f4f69106ef13c61
9480 F20101111_AABTQO coggins_l_Page_164thm.jpg
4a3962f12bf30872ef1159311ae2ac2b
6071fff46848edf7bda92448ba8b627f181964f1
107251 F20101111_AABSNM coggins_l_Page_068.jpg
8015c32b3a74e8e95b1fa99da828e41b
bc91c3204c3750d90b71a9bb652369700138e173
62124 F20101111_AABSMY coggins_l_Page_053.jpg
e2386e81148ec1404912584d159ed9b5
62f04400e9eb0d6be4cffdb7fcda016b66f649bf
10088 F20101111_AABTRD coggins_l_Page_078.QC.jpg
1a21caa1499e293cceab014b96db588e
2353f94014f5ac1fdb9965a30f31b5ed40d27956
4827 F20101111_AABTQP coggins_l_Page_153thm.jpg
7b4a038a8886c94027f4284e28bfe933
f986d758632a6adc450e31d0f615281ab5a17f0c
106214 F20101111_AABSOB coggins_l_Page_085.jpg
1b24eb05df4ae8c95666c3854cce9164
80372da30539baecc17eca8d98b6d807484e4b9b
104935 F20101111_AABSNN coggins_l_Page_070.jpg
8ae8bc6ec710f0d0017fe0be1b71097b
0270291bbb76543ce2a2b0c41248e56cf4a10dd0
32823 F20101111_AABSMZ coggins_l_Page_054.jpg
281314af000d31cbcdfee3f8938a4108
28b59a131895a9ded9dfb9d573e56dc2c82de3cd
14674 F20101111_AABTRE coggins_l_Page_080.QC.jpg
9fdc3da163a40d5acedc4f9f18a2720a
2e287a5ec1dff50f16ec21e988c06c3501ce3567
35459 F20101111_AABTQQ coggins_l_Page_006.QC.jpg
0a696cc5d20976b1aea03c83f61c3a6b
ea65f9fefcb0ba3606f575cfc791fcf3faebf557
110007 F20101111_AABSOC coggins_l_Page_086.jpg
1f995eb06f6429514e8ca84ef82f26f4
0a464a5be284781ba444ce8e8857be36e9c78150
112625 F20101111_AABSNO coggins_l_Page_071.jpg
a40c8411ccf5fa67dfdeaf777c082af2
d9cac7d0ab032596dd4c9e3a7c2fcd1c12b63ae9
2811 F20101111_AABTRF coggins_l_Page_149thm.jpg
2143869886cda22672ced2576a5a3d17
0ed6988736222a5e959dc933f9829786c1cea58e
23244 F20101111_AABTQR coggins_l_Page_110.QC.jpg
e195e4b622326ed7d38a8edb1ceffdb7
38ca65736d93654659dfe97eae572da1b7ae972c
109015 F20101111_AABSOD coggins_l_Page_087.jpg
97995e7f231d05a43dadb54103519272
ee8ae7bc5afb53eb2283a7fb97d077eb229e70fa
115444 F20101111_AABSNP coggins_l_Page_072.jpg
936f790db3de5b528df44d92923bec64
afef782ff039edea69b6d67919457828d645fa36
3802 F20101111_AABTRG coggins_l_Page_131thm.jpg
196eda15bd90fd99917a5f47a4c89e0e
fb571241ef44fc05a3d93eb4f49d45d21c57dfe7
8671 F20101111_AABTQS coggins_l_Page_099thm.jpg
d472bc2a8e973dd15a084f0e77d89e8a
2a3b0324a30cf9115aa4ae27cc4cd1832b2686c4
113729 F20101111_AABSOE coggins_l_Page_088.jpg
25a4338739184d0705d776d2cd31af8f
ea124f77976741eaec0d6078e87843ac452599cd
15201 F20101111_AABSNQ coggins_l_Page_073.jpg
342b849a46f71f97ca8fd4de5e9d9ee3
83834190303fce6e8600494db7722b3c3e44c834
914 F20101111_AABTRH coggins_l_Page_003.QC.jpg
31555a51872a0897530e07f3a0ae5227
22dd7184a0e1f711cdc955f822313e5e52285f8c
6152 F20101111_AABTQT coggins_l_Page_048thm.jpg
6f37b0923879e9914c720dbaa91b7aff
cb68d8b248d9135b64b7e3771cc9106c86107932
108321 F20101111_AABSOF coggins_l_Page_089.jpg
4ff34c8608e130a783dae3eee3bba677
61114c711d408189dd8768bfe8a0759d63f677b6
11658 F20101111_AABSNR coggins_l_Page_074.jpg
5d30c432bab122d8480cf05ef4e5f192
d494ca9e2e41d99c5b6793e104a7504e3c1de0e4
5542 F20101111_AABTRI coggins_l_Page_073.QC.jpg
da957521afe6e7ff326e801c54504199
a8c68bc6075b0d88a00fa1d0e00e02029a48e153
10588 F20101111_AABTQU coggins_l_Page_040.QC.jpg
7b1d566bf7b9ebcb8fb3e8ba5fc6aa8e
da0bc90ec6ce97770d81c80110d54b4fa6e8f142
95885 F20101111_AABSOG coggins_l_Page_090.jpg
3e468eb0c9e26f46166654b5bd36f4aa
14f24aecdf3596b19141ba9140796a9c2c76ed5a
14686 F20101111_AABSNS coggins_l_Page_075.jpg
208c0b7856439ea442e9580f0de605d0
b987925efff89734afe80dc4263a997e288fbfaf
8777 F20101111_AABTRJ coggins_l_Page_100thm.jpg
6f28ad57d601d5b1c0992452a9ead97c
b9791555d64fb3c8c607bd03d11c83da84b5549c
3855 F20101111_AABTQV coggins_l_Page_074.QC.jpg
b15796d4e7da7ffe7d90578a81aba24c
8a7cc667f8c7bd343075480ec6f943c07bae82ba
16986 F20101111_AABSNT coggins_l_Page_076.jpg
d2c4f80f4c808fe4fe16c5cae4b41466
b19ced1baea559cd6358071184b94f65d9feedf5
103597 F20101111_AABSOH coggins_l_Page_091.jpg
204c0d73ecb63aa99e6fece2cb03a944
948f91e3517d820ba09b1353b2352a750b35a952
3240 F20101111_AABTRK coggins_l_Page_116thm.jpg
5e436033b2e7d30524a4bc06e5cda229
0480ebd6e1cced757641b21ae0d90e460d1c1444
28944 F20101111_AABTQW coggins_l_Page_172.QC.jpg
66adc6e18ea70a41b7a0f9d54874b357
da486fa63736767151f9fed78379c843333a6a2f
68590 F20101111_AABSNU coggins_l_Page_077.jpg
3931c74848299486d7af2f1b7d9f627d
7fc80e57bb0c120fb415a73de03f1f93b0424ecc
89614 F20101111_AABSOI coggins_l_Page_092.jpg
3128ae990ecf488ff5ca8d4a658ff844
672530a60e9beb85043eca38c7319cae19bb49b9
34854 F20101111_AABTSA coggins_l_Page_033.QC.jpg
be82266acc144185230356cfb5b02d86
42b1ca05d2c12cc47cc63298a380750d2ad749be
8886 F20101111_AABTRL coggins_l_Page_020thm.jpg
84b67c2502588571fdafd0f325c01fe8
a4af4e012b0421b2cbf8d4354638c5bf77bf8b18
36518 F20101111_AABTQX coggins_l_Page_159.QC.jpg
05929bca36bd08f42b3edbd57c66348e
e093b665eec2717a34bc03712747d11216cb0cc4
28615 F20101111_AABSNV coggins_l_Page_078.jpg
90ffe0e231264d4921054e9082e2cdf9
4a6ce150190d73b47d80a37cd2fa2dbd563bad2b
76268 F20101111_AABSOJ coggins_l_Page_093.jpg
98c65736515bf3b009ee8ae056d09df2
985ac47497a4baad191f25a87e0640d6b892bd2a
5359 F20101111_AABTSB coggins_l_Page_077thm.jpg
0ef5566d6f0da54a9ac01e0ef6c5327e
e8315e3dda77578cb461d8f51a142545a6544b68
8838 F20101111_AABTRM coggins_l_Page_141thm.jpg
efef65c094004a657016fd493305dc59
3f1d5186a95fd7bc23d91d8bf2d4f4755dd7d989
6180 F20101111_AABTQY coggins_l_Page_112thm.jpg
874bc7c7c4c29dd51f83b691b8230ee1
45da9e6398f9f853ce4782ca020e2ac42df4f116
56219 F20101111_AABSNW coggins_l_Page_079.jpg
21b4fd579304044bea1f8d577b9d131b
405ea2a44de045278470a62103fc9c4d4e5120a5
85579 F20101111_AABSOK coggins_l_Page_094.jpg
98e7ff3ccca29a414e3c7fc86f585f9d
c4da40dc2e1a328437ba18f15a1be2af0b0138cf
37851 F20101111_AABTSC coggins_l_Page_020.QC.jpg
42495c309c8809904282d8c649af4521
4f1ac3c72d95776beea8f095b44a98e2932235ea
8408 F20101111_AABTRN coggins_l_Page_136thm.jpg
b217893c540dcc4fb3ca034fe0ae28c1
58871116581e8bb5fcdce8c02c7c6278e5254517
20226 F20101111_AABTQZ coggins_l_Page_047.QC.jpg
c63ef5816cc9f2f7846ba4b43c025183
0f59493a9d94be8876d530cff4fd0839659af0f9
43011 F20101111_AABSNX coggins_l_Page_080.jpg
56cf4030ea95ce0a3113286d342a74c5
43df7a1bb8b288e83740d3670ddcc7fedc1a6cae
34355 F20101111_AABSPA coggins_l_Page_116.jpg
d7d5bb370c887c33a00474897d924a44
5fb283c3867e69064e6b717be44e033439e2790a
107175 F20101111_AABSOL coggins_l_Page_095.jpg
facb129a01d168aecdee1bdee17e6a10
5d8e31689717f541156e9805f011b4e816fb07c3
8066 F20101111_AABTSD coggins_l_Page_024thm.jpg
9f088d0c6be5c19298de5e7026f6517f
6556db71fc12904ea96d40f72a8c7d59a030a1ce
9371 F20101111_AABTRO coggins_l_Page_171thm.jpg
ffe50f2ca420d43f57e0ab221942b634
2d4fcac6560e41725ba7a2c11cc85ba78e6435df
46135 F20101111_AABSNY coggins_l_Page_082.jpg
366269c1a1b610b0055393cd7ce62f6e
f23caf41c10abfcea154dffc928bb17305bda755
30888 F20101111_AABSPB coggins_l_Page_117.jpg
98b44ad348d374f4b4e4917acd8b4930
8fbaad0f38525288df991c8b8369f4a86459ca44
109030 F20101111_AABSOM coggins_l_Page_098.jpg
66e3b24fb0c01e5adbc87f11a9400667
38fc780d5d3d291d765c44337421759ef5ab0ab7
8887 F20101111_AABTRP coggins_l_Page_098thm.jpg
e3126013d93cc6c69cd589769dd80285
5c809b6d330c66c6562a8826a858bfb993a56df5
116574 F20101111_AABSNZ coggins_l_Page_083.jpg
f2d464736b85e1eb9424214d875723d8
8bd1266fbd5827c72f825b6d514da9b29251f609
107484 F20101111_AABSON coggins_l_Page_099.jpg
f3a3c1fe60eb60f3fad7429d07fd062e
b9c3fe95ac2ebc3bfe8a10179cbbf9f4d51df213
37581 F20101111_AABTSE coggins_l_Page_072.QC.jpg
6faf90e3c26d06e0ff8a80f767d4f5f9
18a48eaa4152cd7f37f89ace26af8e29b8470f5f
18988 F20101111_AABTRQ coggins_l_Page_053.QC.jpg
9a83707d9e4801fb2d19fc3b02718ba6
1a461a470edb7591c2cf0a34a0b69c085b597bb1
45837 F20101111_AABSPC coggins_l_Page_118.jpg
6a2a4a0e2f7e8d2fd285097c69a9b7eb
93c2f20bf9f5764c475cb0a9fb4ba7e6065e7032
108678 F20101111_AABSOO coggins_l_Page_100.jpg
ba24c7abb90c629f437f8c1877afde83
0788b96e61c86d78f327881f56673f3de7f017d6
8564 F20101111_AABTSF coggins_l_Page_017thm.jpg
5bd456fac201c94479d29ba1c7e10c62
5bb1c3d93a99d6335e51f87e7c7a9bd382523f43
8765 F20101111_AABTRR coggins_l_Page_021thm.jpg
f9ce70734555258df614b23cc229020e
2878803d21ed4dc06edf2713107af06d991c4fe0
42674 F20101111_AABSPD coggins_l_Page_119.jpg
244decac4c219c990901af2599568c35
34db233c83e489e8c334c9b352ffb797ab13a0a8
106899 F20101111_AABSOP coggins_l_Page_101.jpg
0166f088fc0763cb669d1c03ce02107a
289a063353f93c5904d69d351fa8af8deec18ea0
5195 F20101111_AABTSG coggins_l_Page_150thm.jpg
c46e305c96ab9b859547edf0862be4e0
f108d360a5fdaf32edf3475084f083d8a4861daf
8752 F20101111_AABTRS coggins_l_Page_105thm.jpg
1a631fc4b706086c3786ebeefaf1e13b
42ee1c1ae322570f1bb93e76a996d1e88071eaea
71887 F20101111_AABSPE coggins_l_Page_120.jpg
738e4cdce68867cf6649bfec9876b0fc
c3ebfecff5191f09e00da213158eaa5fbfe694d2
106022 F20101111_AABSOQ coggins_l_Page_102.jpg
5e2e3859a037e84f50bf0084b2847ff8
bb7c79a129262895ab2dbdd0504a087855cad3e9
17106 F20101111_AABTSH coggins_l_Page_079.QC.jpg
e68133198abf2782f889f9cded819d85
dbec177961f1e8b9bb75ac77bf2b0ee1ad3687a5
4698 F20101111_AABTRT coggins_l_Page_125thm.jpg
9ba1f8dd0dbe054d781d4e1b153d9207
b0231fa77779131326ffd867df5b1cdbadfdd75b
71637 F20101111_AABSPF coggins_l_Page_121.jpg
3f3706f92a9f05ea92a9ef2f077c4709
8f1e64eacd835c4d364086930d8d6563ec537189
106546 F20101111_AABSOR coggins_l_Page_105.jpg
7b61e55ed14e290bac21cdb6400e5b9c
ee68b0b9aacf97217680a7072aeb3e7c89251de9
F20101111_AABTSI coggins_l_Page_091thm.jpg
c30e37a576003118e1b3982239ab0d8d
d3cb3d2c33a18aedfd3cb364c2ef4f13f960eed0
8843 F20101111_AABTRU coggins_l_Page_144thm.jpg
72a5feb9fa924335ba54fa3f6acaa8a0
ad5d57a5974e111ae44276f70dac9e310e99574c
71374 F20101111_AABSPG coggins_l_Page_122.jpg
8f4de28116a9cc2e34202aa7ec322618
2f017fae2a468e95e44eff35ee6e31d118b7abb7
111006 F20101111_AABSOS coggins_l_Page_106.jpg
db3384967657d6104967eb185e6cba70
f3ba33057a40f34a50d40df1d58450168e3dc34e
19282 F20101111_AABTSJ coggins_l_Page_007.QC.jpg
4acf60ba83d8afb469ef5d9e336a6f22
35ba8781c35a73c4a773da68911253e6a51384ae
35118 F20101111_AABTRV coggins_l_Page_137.QC.jpg
18c277a7fa0817ac8aa0c727d7974759
d9211a31ddb6005edf587de79a263bb5b668f97e
54866 F20101111_AABSPH coggins_l_Page_123.jpg
52f9ef1f0cd1254324dcf17cdbfd2d26
1a4edadff5b166f92c02a2aa72d3033084327f2c
16825 F20101111_AABSOT coggins_l_Page_107.jpg
7c28b96fe922b57ab02329b8cf8c9701
249fb0531e9c72d89489113517a5da2591266453
36234 F20101111_AABTSK coggins_l_Page_068.QC.jpg
9f7a82a51e1b8031955387a05e25d125
d14af6b6a3d6ccc68e6e1a0006fcb24ac896baea
F20101111_AABTRW coggins_l_Page_134thm.jpg
fa9a13dfec65181b9c773f85eed6e3f6
8885a888b58cdfc081623e19e8e4410451256d10
45572 F20101111_AABSPI coggins_l_Page_124.jpg
f664fd4816432077464f85992709ae66
e56587734136f470fb38cf87428bc897da0e9584
16505 F20101111_AABSOU coggins_l_Page_108.jpg
0a70ad6fe4c06519fc84afae70da2c36
67759a05f55f0a812df9fe32d763cb486eae9465
8476 F20101111_AABTTA coggins_l_Page_138thm.jpg
b0243b8a966623a2c1e00b52b2042122
bbbf430ad6c77025b8c00e225009f7633599446c
1617 F20101111_AABTSL coggins_l_Page_107thm.jpg
e6f033b0c424a4fc5afe0d6482c4c8c4
29504af12237cd57701bd8225974b8c33b7b5c36
8784 F20101111_AABTRX coggins_l_Page_004thm.jpg
5f6a8495734855d50b6489b2f1ac18f2
2c050395c6e12d182d8b1087b67cd9afc08520cf
42780 F20101111_AABSPJ coggins_l_Page_125.jpg
2686dbd9f22b2d817aaf3008d2a45385
2f88e02cb92d5ef5a98c84539d499a5e2a7f4fd4
16672 F20101111_AABSOV coggins_l_Page_109.jpg
282e8e2b29c0ece23b72a914816c5d8b
4f7c43b072babb2202f56c9570390c26c2cade19
20248 F20101111_AABTTB coggins_l_Page_128.QC.jpg
0b9b3fb8a2e560b0c52a380dac9f06ad
97aeabd70d2a5abfede0fe01508fe9bd554192df
34645 F20101111_AABTSM coggins_l_Page_105.QC.jpg
f276d02aa60b78462e63d4c236fd8e16
3a2d01d44c74bf102388964b11d4e4619561dd2c
7669 F20101111_AABTRY coggins_l_Page_094thm.jpg
9b3ace74bf61f42282ad415f18c18ccb
698a23139eb17eae4ffe8e2dfa3a7f20c8b975bc
71698 F20101111_AABSPK coggins_l_Page_126.jpg
c3a5122bd22ddee857162e33ceaff6cc
c14a3be73fde02e5e5e12256fff63a0d4ca793eb
76987 F20101111_AABSOW coggins_l_Page_110.jpg
3eb99429078807c5b31dc5131837dd6e
3d287974862e3ca0c29d3a9066c6d8735409b231
35796 F20101111_AABTTC coggins_l_Page_106.QC.jpg
62023adeb393e7ab3a7b1c0585102602
bc430ab7f0987a40020ce4a54f65a78cb43179e4
8912 F20101111_AABTSN coggins_l_Page_165thm.jpg
60461b5841e4f2ce370471da4c3a6eaa
80596c390995ad4c0ecfb0bd4f7a563c1f8a198b
33462 F20101111_AABTRZ coggins_l_Page_031.QC.jpg
5b4bf6d73e61d1714bf44a84da5db3d5
ebd826522b4f3bd987b0ae50301b1024c71ba014
105693 F20101111_AABSQA coggins_l_Page_144.jpg
c142deba48d3d7f9492fe6779396908e
d0e4568bdf1c1c0f9c4a8859ef355d0696168ac5
71728 F20101111_AABSPL coggins_l_Page_127.jpg
7921b897a8093975afe7ee486d8ca1c1
b5ebd0f85154199394c5072226327bdbe18a261e
64863 F20101111_AABSOX coggins_l_Page_111.jpg
49b226a91b208794793f4844650f7556
8c8ecdd929a716efde50cddf2e169c996129ee75
8605 F20101111_AABTTD coggins_l_Page_102thm.jpg
33824490ffe7bbdab1a57c7fe54ea765
e5a52140e08d694df08fa2661215a79df2101ba3
5774 F20101111_AABTSO coggins_l_Page_156thm.jpg
6b0867299abe484ba0b7678a768e03cb
8478901aefe78bb7ff34e9175f3a682439055aa4
116684 F20101111_AABSQB coggins_l_Page_145.jpg
0654f895f07b0f132c16e9487aaadb48
39fb03a58ad67a3dbe5030c0a9b255bab4596e03
71180 F20101111_AABSPM coggins_l_Page_128.jpg
96c31bf8bc480c20c46a47747100499c
0c9fa206fe4399b8d730a496e1bcef2705d61d23
55369 F20101111_AABSOY coggins_l_Page_114.jpg
7628f26d32d6674aeba0300c2f579673
9ec2c3535fccb60d9ec600b4ea4a32aede857e60
8583 F20101111_AABTTE coggins_l_Page_059thm.jpg
744bf4f1c86fd486af40775ded9270d5
75a193424b5487afcc0aef1689a0a0abee551e03
31623 F20101111_AABTSP coggins_l_Page_015.QC.jpg
75c22eaf4d24eaeee1847c63445a3aeb
71e88d9638fe998f51af69396c53b1d235e3dc0d
111227 F20101111_AABSQC coggins_l_Page_146.jpg
8e0a79ba4990562f311bde59974fdd09
3beb866c2a05b185fd17b326cbcba3868befd492
57275 F20101111_AABSPN coggins_l_Page_129.jpg
1d9be5d31959e06313c9bbadf85996e4
a75162b8dd0582d17277cc1a7d6260f55fdbca4f
52941 F20101111_AABSOZ coggins_l_Page_115.jpg
1a3679a080bb315d19cc4bc8e1cd23b6
ae95261e2c0539b5e0eebc9e9114f8bd380d332b
8425 F20101111_AABTSQ coggins_l_Page_025thm.jpg
67b342010701500827232682cc540a2f
b4f4d1882b5630571d012247f0d81070e8595b6a
77216 F20101111_AABSPO coggins_l_Page_130.jpg
2165cb58e127cb3d354595e0a1a25a9c
e996a52f8dea624382840ca4af683885767e5b50
9150 F20101111_AABTTF coggins_l_Page_016thm.jpg
d616741d4cd4957eaea4bb1557f53b50
60d5755050fea2bcf7761f80d5753f9d3ede5122
34654 F20101111_AABTSR coggins_l_Page_085.QC.jpg
2bd49b2046da1fece10396ac813a0aff
643f37c4919caad5c52171ecaa2cca4f73392e6a
108945 F20101111_AABSQD coggins_l_Page_147.jpg
d1d3e854d653f4391fb10f9edcb3e719
8010dc4e144051323a23dd0211dbd290a3aa5f58
38237 F20101111_AABSPP coggins_l_Page_131.jpg
bd0f3ee2350c8aab538736fb5c60c30d
71df99c1b8fb3efe24dc5b4567b095bba6a2b6ac
36996 F20101111_AABTTG coggins_l_Page_069.QC.jpg
6443ec61ef79ff3be7bc767b2954accb
486cb64b12e9de782fad908ae3bb46104412edd9
F20101111_AABTSS coggins_l_Page_065thm.jpg
73bb5ed96194ff64c782c50333864c13
2fcf97673240a25676817bba4aadd168c5cbbb96
111116 F20101111_AABSQE coggins_l_Page_148.jpg
42ad2c514ac9e266561fc8fa9774ae83
e8d6aa76b3cb93a1c4160231651d034b5bc2ec91
37048 F20101111_AABSPQ coggins_l_Page_132.jpg
8a143de2d9eb1def18cf48a1bef1f9ee
a3b0225b7aeb1f2d472f731f246fc11b1206e4ac
7065 F20101111_AABTTH coggins_l_Page_062thm.jpg
e54fa134e152b4787eff35b38863370f
02e9e2d83f72797862eed577d1c3950c72b78203
9089 F20101111_AABTST coggins_l_Page_089thm.jpg
30dfb0686f377ed723525c88b97b8421
160797196f06a2ed80947f83f6b3f491979020cf
33790 F20101111_AABSQF coggins_l_Page_149.jpg
ba726cda09b19b45d31e43835f21391b
cfeeb317eee092a595c947a81a13bdc886146ab6
50614 F20101111_AABSPR coggins_l_Page_133.jpg
926cfae8066feb1129be2bd610119ae7
50ce7b01338538544af5533cc0146caea2413409
25187 F20101111_AABTTI coggins_l_Page_062.QC.jpg
ee626093a1338b3d7aec2880064a384f
0d7b1653b5748103aff3f68546ba0b308b0e10c7
34946 F20101111_AABTSU coggins_l_Page_060.QC.jpg
0a2a2a162f00a82106ccac65d2077fa4
cbc38791b3b6dc4557d5f7dc7a7413f5b7650ecb
70539 F20101111_AABSQG coggins_l_Page_151.jpg
f1a2f09035eb5b3a2b48d4c82b533f3b
7e5667ebd0f14a09df502008112de63a1ecff1eb
114196 F20101111_AABSPS coggins_l_Page_134.jpg
85db62d4876d35645f66803f7d3a3c2a
fd78e145534d5f5b5c285c0d3ee4cb467e2fdd2f
19347 F20101111_AABTTJ coggins_l_Page_043.QC.jpg
a7b351c9a1214d73f0e1a64dcdd7865c
cb758a218bbedc8c92a7ec4f3df499718091a176
20686 F20101111_AABTSV coggins_l_Page_112.QC.jpg
15bb961a20adfd91eae1efe24b15bbf2
5cfc80fc0e89d5ce19d4007c9b398838bc278e6d
58494 F20101111_AABSQH coggins_l_Page_152.jpg
55a0a6cb3c81ac65087796df8fb05ef6
b1493cf7a29cc8205adc2f3074e37bc3469fa7c2
104685 F20101111_AABSPT coggins_l_Page_136.jpg
4459623fe70a57f4675f7a4af4435cfa
42749393ec528d1b0dd16affb04ad1e5f55d1aec
8223 F20101111_AABTTK coggins_l_Page_064thm.jpg
18d4e29225835e4f5c1fc01b925c4e0d
72eac29506b24d8bca2f95b444ec719fc190ed42
9106 F20101111_AABTSW coggins_l_Page_146thm.jpg
4988182c600ff0dfeb191b177e8ac3e3
09ccf244c0e21e38582937a107aba5ad5f18e3c5
57525 F20101111_AABSQI coggins_l_Page_153.jpg
781995b7f8e3389136fe318c1c7c54de
faff880c100a098d234ce3c15bcb3de27477665d
105632 F20101111_AABSPU coggins_l_Page_137.jpg
ee7a62b6bd7e0aaeb4e060b88eca585e
12a520781f705173d8dca9f5284cf761b90dd742
6265 F20101111_AABTUA coggins_l_Page_011thm.jpg
9866421dcd78c4b686e7ed28aa20cc66
c99f86c74b548a0cbb0060949771ebba1a43461e
20568 F20101111_AABTTL coggins_l_Page_126.QC.jpg
ed199cbe9dfa7aa580a5291dcb18b081
dae0a3760a042ea0fdb914ccd04c48047ec9261a
9134 F20101111_AABTSX coggins_l_Page_038thm.jpg
fde50b3320ed49fe0f30f908dea91937
5b9bc18a3eaed78f0758a1fdb1f6f200da28d351
57837 F20101111_AABSQJ coggins_l_Page_154.jpg
4fe22e4da8d70cb88e788b6159d93a0e
3531ee8bddeeb73340b7140639bb80820592d0c8
106066 F20101111_AABSPV coggins_l_Page_138.jpg
76bc35ec3cc6ec3fae6e708afc3806b8
61d4064d2d0b60478c58bdfb345bf1aad0100409
7854 F20101111_AABTUB coggins_l_Page_012thm.jpg
63212337a744c7839c2d10987955293e
aa1fc985c7a881aad905447f17315893d29b004e
34918 F20101111_AABTTM coggins_l_Page_014.QC.jpg
62b08cf9455e2c2997551b0c15501866
fc2ce5bd32945a7793524d82543b51000c28e53c
5043 F20101111_AABTSY coggins_l_Page_049thm.jpg
276c8616356ff57735279eaf706c70eb
aac17f077960575070583fe48d0b1dfb7ae1ea76
67084 F20101111_AABSQK coggins_l_Page_155.jpg
e33708227af264a8e0d52f5d6529a6e1
cbbfbdec7e8fa5b732f76736c037cbcac11dbf94
113031 F20101111_AABSPW coggins_l_Page_139.jpg
2957efc3b1244b66f533e7afea5b2492
ca185f788fc63abba272cfb5d237e93bbec4d327
35568 F20101111_AABTUC coggins_l_Page_013.QC.jpg
f7d59693de21bd0ffc750a048578794c
b436116c8f2047cc4f10543362786935d4e36e26
5356 F20101111_AABTTN coggins_l_Page_053thm.jpg
40392e3a9c3e89b57cff12f66b7fc61b
728a7bd4e11e9d6b0a7f18d37287b3f227cd4da7
36903 F20101111_AABTSZ coggins_l_Page_023.QC.jpg
95c7cd788c65f6233fbcef628150893d
9ed9ca52227663a02345d99817bdeebcd3d0763a
70675 F20101111_AABSQL coggins_l_Page_156.jpg
ac8919918ea9ebf4b2e1478fb3d67fdc
72de55be4e17649eea35b327c96323e8e19ff64e
112358 F20101111_AABSPX coggins_l_Page_140.jpg
4a5b0dbc27144b2207cb8cbc44fcc723
efec5c83e738683f6bd4924847fc0a3c678aa32f
137930 F20101111_AABSRA coggins_l_Page_171.jpg
5f2333aa9088f1b153f5f47fa45a5b68
6659ce1c31d0788b63c015cf6e97adf7e50160db
8726 F20101111_AABTUD coggins_l_Page_013thm.jpg
d11e8f9230e88bf6a8260f986e6e98cb
05f676569753a04aaea70126a14d4ecadcf1a8f0
34329 F20101111_AABTTO coggins_l_Page_144.QC.jpg
50ba2fbf287c0c74be64348772c02cf0
948dc356b19ee252d2be5d40755d9851406277ba
60130 F20101111_AABSQM coggins_l_Page_157.jpg
f760b24ab7e7dd8452d480d7ae1ed0ce
79eb42957e0943e040530d2596e36c5b3d19adb6
110050 F20101111_AABSPY coggins_l_Page_141.jpg
17f37869f7544f0aa4661851a6038b5b
e92d08837932795458a297282208427957a2c2f2
103407 F20101111_AABSRB coggins_l_Page_172.jpg
eede28289fb75cfcb7fd266b0cf2ff99
350d482581cd3e0f2a1b80b009b50ee5dca996e7
8723 F20101111_AABTUE coggins_l_Page_014thm.jpg
e77c6b70278ad7bc822628ca7b41c46e
54f9d062b9323693db05bc0c741b146ac694e7e5
259599 F20101111_AABTTP UFE0021996_00001.xml
ac44af6c753b92d699d6b2a1b2649306
d73c7fb30de8566308011bc5d7f8208bc0779ed1
36192 F20101111_AABSQN coggins_l_Page_158.jpg
edcd0b181fd7673f2eb5ba689be06128
7af9ad2f29cdf8294912ccf4bd30d28c13846aae
109467 F20101111_AABSPZ coggins_l_Page_142.jpg
698a3190715e98df8fb56d02a2c70397
b609bfba33508e6c4ebedbe9ea840d0080b48b65
68512 F20101111_AABSRC coggins_l_Page_173.jpg
b6da97c39f8ec270c02f4a41f19687e7
69ad79fe5de5791affb3ac147a51701fe3a41812
36100 F20101111_AABTUF coggins_l_Page_017.QC.jpg
a922ebaf5a6bfa64367b23507392b258
591ee285100dcf2508e7ba29ff897b06f1e3a883
466 F20101111_AABTTQ coggins_l_Page_003thm.jpg
2bf59aaf92d415450b16b33604d343ee
4b992a609bd80ff7fc22e5b208bd07f9c4c6505f
120752 F20101111_AABSQO coggins_l_Page_159.jpg
2c7f2476af430c40b09dd070a3e60511
6827d449d9d2e917b9340420c6763d842ae47603
280010 F20101111_AABSRD coggins_l_Page_001.jp2
c151a20f1f59069bdb56cf8bb467ddde
b61106d922a527c9359e92028f04d740c6ad42f7
4967 F20101111_AABUAA coggins_l_Page_152thm.jpg
ae0c49aaacc721abbcf61118d6d88b9b
1e53244ed03a89a3e2f9eef9a923fc72280a1b56
35553 F20101111_AABTTR coggins_l_Page_004.QC.jpg
c31d4ea213ecb694580ac3c248e49e0a
3ddaa22783702684d4069c19e0c782abd066ff5b
139994 F20101111_AABSQP coggins_l_Page_160.jpg
a068c76beed26f4a74aecbc312831954
6648bb746ae5ab32c653d3c46683b6129e0bb4fa
18165 F20101111_AABUAB coggins_l_Page_153.QC.jpg
ea2a75d43d0e2e9928c4a0e47a7a3599
66d9b11ade482edaee198c18c1db00ffaae24711
F20101111_AABTUG coggins_l_Page_022.QC.jpg
6ab0e55ee267f6863e1e19ab1ea91fc4
e979c2d9fad0d9d8966920fb41e80fb33e73214c
29204 F20101111_AABTTS coggins_l_Page_005.QC.jpg
eec50ef8955004a0971c462d27ca595b
6b352b204ac433b672f9527e1b8d61eda6100af6
138362 F20101111_AABSQQ coggins_l_Page_161.jpg
d927e2f5104d40b7b86fac0de5c6134e
40fdf8dba2c90144aba50819d2dc53b6272ad837
28654 F20101111_AABSRE coggins_l_Page_002.jp2
ebb8a61eb7c1eae8feee6f5c295d0546
f4bcfbc413fb5f1f3aff925d1905b78d0019fa6b
5334 F20101111_AABUAC coggins_l_Page_154thm.jpg
afb279c72f44d9dc8d16f2d61f410c7e
4e9556dce5c1f7a80874d6eff3adffb534f75747
31703 F20101111_AABTUH coggins_l_Page_024.QC.jpg
a4cdd2ef2a0ee9a2a08f6309fda516fe
c2371762c6ad9e69849eab95d37bdc040f22c2d6
5110 F20101111_AABTTT coggins_l_Page_007thm.jpg
915a463620d179ab3ecea8ab5b389e69
c7d3cbe66067a57600fc1d60e4cbac510e7b9124
137846 F20101111_AABSQR coggins_l_Page_162.jpg
dc0ca9cd323d32c4b6c7c678b7ff231e
936381da87c855925729c22bef8c363e340dfc27
14828 F20101111_AABSRF coggins_l_Page_003.jp2
cd9e2024ee0626f9aebba229ba554500
d85ee499207bca204c7efaef58cfd402d61e1d7f
5701 F20101111_AABUAD coggins_l_Page_155thm.jpg
403bbd70d61e6f33f817430de8a1a8d1
e8f5da42b97191e0a86cc823348d4d3392b050c5
35292 F20101111_AABTUI coggins_l_Page_025.QC.jpg
38173639ea18f0238181053486325fbb
f3f7380f1c21b8f0b94672858886725ec4fea580
36084 F20101111_AABTTU coggins_l_Page_008.QC.jpg
730070b542460fc177337c634e833f2b
0a4375ea30912d957473b76cb952bf3359f7b78c
131945 F20101111_AABSQS coggins_l_Page_163.jpg
b0f220061e6c0da4132540ee6a58faed
62c261d9214d234400122ee8c9aabff726cb697a
F20101111_AABSRG coggins_l_Page_004.jp2
bf2c094e93c44b49e0ec0da4d65d7d67
6295e4f03f039995983e7f1fac3ae8f13f38ad18
19744 F20101111_AABUAE coggins_l_Page_157.QC.jpg
f1865afb764c892670d8724ae3687153
5aeb506d3a6d1fc8c99a497700a69bc86db06792
38903 F20101111_AABTUJ coggins_l_Page_026.QC.jpg
daa706811e40fe4477aec66aa1ca2c67
ce936b0677f0b541a42c5e75b2ebf45c224a2ce4
9154 F20101111_AABTTV coggins_l_Page_008thm.jpg
5643a27740d7c574dafd91ea4bfc0a50
da670cfc20fb00505736a8c7ea8f2411f68aba02
148075 F20101111_AABSQT coggins_l_Page_164.jpg
fff840a285f0854f3e61466a8048fa67
eff381e61136b9c5852c02c4b0f1a256ffe6b4e6
F20101111_AABSRH coggins_l_Page_005.jp2
55698fa0996e00f531ce2c79a399ad3b
8df9e34a4aa33422ad88f283e4ed72ff17ae3cd7
5520 F20101111_AABUAF coggins_l_Page_157thm.jpg
93d37e3d6d2e6e76ec6b0edad8e3738a
0f48e28e7bf6fe9f4362a5212405880ec77f2e75
8225 F20101111_AABTUK coggins_l_Page_027thm.jpg
32708489d994a436f3b3d9e6f95eb60b
468c065125b7b15ebc8bff1cd77482d96684110e
43377 F20101111_AABTTW coggins_l_Page_009.QC.jpg
82407dff90078d197625dcccae78e1cf
87b86bb9f8daf42c80e8913d096338b0e53292cc
132148 F20101111_AABSQU coggins_l_Page_165.jpg
95ac5d505157bae02193a4d3d911f087
9b6d5cef3a0dd42b031c163b6e196d2182e0d7ee
F20101111_AABSRI coggins_l_Page_006.jp2
9ccc0edc5a47e556e69b09b93cd8a2ac
fbe5d9863da71a6586253f632921feae15c47f8c
3807 F20101111_AABUAG coggins_l_Page_158thm.jpg
40245a8162205b4352b85c66c0f437f8
4d5a675db04b5ba973e8faa7cf02920bcac7f156
36391 F20101111_AABTVA coggins_l_Page_039.QC.jpg
9bdcc662fbe155f5dd8d5f4116c3571e
54744b8316d19e8456d4ba3b842b48859ac6d39b
6873 F20101111_AABTUL coggins_l_Page_028thm.jpg
b550231acb2cdf072cd650e9e35cc5e5
ec4aee1e70481f8cae9aa89e6c52e0b16cdb55e9
42339 F20101111_AABTTX coggins_l_Page_010.QC.jpg
239fa04d71bf2283d3bd6b4a47a89475
69ef9a4dd60b80791b811bcb54f940bd65ad9349
131280 F20101111_AABSQV coggins_l_Page_166.jpg
e6ca2b9f36e483ad179d12b41ad564e3
235f694720f93e9c3a2e7f684d92b177a34ad236
1051985 F20101111_AABSRJ coggins_l_Page_007.jp2
4ebeff8831566a6467488e9b33dc0ada
5b4536eea27f9d8d2409f963c05ae45a98be6cc1
39037 F20101111_AABUAH coggins_l_Page_160.QC.jpg
750c56256621e3aa84afc00698678f15
18566704981f1829a69b8cb25e6cb05d29dfeacb
2598 F20101111_AABTVB coggins_l_Page_040thm.jpg
56e0fbb6e75b5733e87f27b8c03965fa
d8de3f766c2be3ed266eb01aa7d4bbafeb4d8a5d
35316 F20101111_AABTUM coggins_l_Page_030.QC.jpg
09b04155635a92b7d07f47cc90fd5629
060f95ac528ae26e72efff26339c4da1af9599fa
F20101111_AABTTY coggins_l_Page_010thm.jpg
7b7e0de885a8282ba32892e54d7dafae
5bfeba8947c229221d31e64a1ef7888b73a2f7db
131862 F20101111_AABSQW coggins_l_Page_167.jpg
97138f3702987f4286e4b1be2a3237c7
dc7a735210f310f68f17386ff5c7063e0badd0eb
F20101111_AABSRK coggins_l_Page_008.jp2
adbd250d015d554fdef6873396bb0a72
568ab4c6618017b5828e763c96259e371aeee517
9528 F20101111_AABUAI coggins_l_Page_160thm.jpg
1b0e68c9d160bc226c3b3ae692188e66
c3542d3ac58cb346e765acb6a8259ff347d4f3c9
19110 F20101111_AABTVC coggins_l_Page_041.QC.jpg
2352bb9ef0eeeed5c94a54edb9bc3ca3
9bebf90dadd33c1259ae19fb2f2b75738a78e88a
8874 F20101111_AABTUN coggins_l_Page_030thm.jpg
b73f7aac5a53f71d242be6659a36d1d3
3e3f4786278118cdc2c29655f66abc7b7c2ce9d6
26577 F20101111_AABTTZ coggins_l_Page_011.QC.jpg
f4be3acc17b02ed72812b1401ee4b1fc
925a56a5c1aceb03ddebfb781859bad428d695ac
129973 F20101111_AABSQX coggins_l_Page_168.jpg
1b379a6c7bd7a957ec011cb6805e8ece
49e4a71f3daf893d6e0bca9441c2b0e86746008d
F20101111_AABSSA coggins_l_Page_025.jp2
a7119be29871d3244e207b49c67d1bd6
f5d9909d4ecd9b0a178392404b71dc5e6720c5f1
F20101111_AABSRL coggins_l_Page_009.jp2
8e527dc4b0ba0c54d6006a929356bdac
44cee769430fdd69172838304dfc8712e0a99946
38608 F20101111_AABUAJ coggins_l_Page_161.QC.jpg
1fb650c5a0ba7ddca4d17b7f9140956b
8093044744eb555fb6aa0ffb403589d8574e7fa3
4614 F20101111_AABTVD coggins_l_Page_041thm.jpg
851689ac355ac19a01a9bf0d5c925470
e46437332adcfc290d5295a81562175002cc7246
8517 F20101111_AABTUO coggins_l_Page_031thm.jpg
b453df90e3fad574c8380022ddaa37d6
05abb9d6f80efeb8d858c1addca64e1240418b11
141674 F20101111_AABSQY coggins_l_Page_169.jpg
dcdad35453957305142d194ff2623882
758f009ea31b29345076cd56ef90a6f382f12039
1051950 F20101111_AABSSB coggins_l_Page_026.jp2
210ea1a15e73a46253d30c483e4cc45e
a4e6d6f8b4f7db0e387330b1fce3b896a1bd49b3
F20101111_AABSRM coggins_l_Page_010.jp2
f583e86145848d183bdd73721fb40a21
ac2be9857ff94da3dfb13ed51c5893911892060c
9441 F20101111_AABUAK coggins_l_Page_161thm.jpg
a33c42332d95489c42f17f2feddbb302
d99760fd9a7729a1bab90ab8d64f53a07e423a42
15649 F20101111_AABTVE coggins_l_Page_042.QC.jpg
ab2ae4772f095828f7da1d4e4ad360e1
48185b45aaed2e45e960f0839dfe09fb1e59130a
38478 F20101111_AABTUP coggins_l_Page_032.QC.jpg
941a350962c719db3dc2b88861be3c35
dcd0beea51c9e150aa0f07df3fec22c1ea377e13
158961 F20101111_AABSQZ coggins_l_Page_170.jpg
9b981f69c2d7c1cb63c2ca820dd97a0c
f9bc03359a2bd6316a069bd8310e54de629dd739
1049274 F20101111_AABSSC coggins_l_Page_027.jp2
49522b18409492e5f6430a1dd57736ae
2bbc425a66a9bcb98d47a8f9ca3fce97674d905a
F20101111_AABSRN coggins_l_Page_011.jp2
1cd6d1af9d0f0cac36ab42e704c4d09f
4974708dd524ba72b989f87ea0d4c8251c6b8ad9
37223 F20101111_AABUAL coggins_l_Page_163.QC.jpg
2796982d397fcb04b18f311145694819
860397f21e95aa8220057430e371e0e1a011b0ee
4263 F20101111_AABTVF coggins_l_Page_042thm.jpg
7dae2e49184a2f6adef683ebc439ff3c
dc37da61a22720d8e0953e8b0326fb31221357fa
9173 F20101111_AABTUQ coggins_l_Page_032thm.jpg
ba7037d02bd6bb8097c72e38725e9bea
beeb2dc44c40877e5043ebcee5fe52847f728d2f
835364 F20101111_AABSSD coggins_l_Page_028.jp2
254db1c6b5c53544a7de4c3e356c60a0
34920c1f1935b197678e3a093dad921948bfed7f
1051947 F20101111_AABSRO coggins_l_Page_012.jp2
62d28573ec3e5961b8eef1bf4b3dbbb1
fd5063f0c4f029d8d10b01b439453397847c18dd
9163 F20101111_AABUAM coggins_l_Page_163thm.jpg
1b7a16219668f62f134836b074d5584d
7633d3a187e9ac38a777792f92117bd5d011ec41
3954 F20101111_AABTVG coggins_l_Page_044thm.jpg
71eed0c6dbc825dea6c5da704b5c9c74
5a9a4e3c0a0814aa9339b776e2420d986604768f
8520 F20101111_AABTUR coggins_l_Page_033thm.jpg
15cd11cf0d0cf1b58ab521f1d5df6dc7
3a7b84f49daf95eec6687a3d0c5b660dd177ad72
F20101111_AABSSE coggins_l_Page_029.jp2
7694acb9f284b75a6922204395e8e884
ce4376ca67824ec8e0245a0fd06e40c9f3463b11
1051910 F20101111_AABSRP coggins_l_Page_013.jp2
8d305626267d9b3bb8a6f655fbe8f494
e681e84c3ff50d15ad1d830c6355e8808f4ab712
39883 F20101111_AABUAN coggins_l_Page_164.QC.jpg
354b1990f8e19705b21e9e1909a16a78
266e1153993f2682f4ec526e1db2d8157f3a4ad5
38138 F20101111_AABTUS coggins_l_Page_034.QC.jpg
2fd332ad1ba53d2fb51b55fced3c8781
70354211d85cd295c49216f943919d36dbd9acc6
F20101111_AABSRQ coggins_l_Page_014.jp2
eaa6cfa6f4e547300af6bc2ac6f3682d
71cf39519ec96d3bf8e0192d88e23e315bb79b13
37779 F20101111_AABUAO coggins_l_Page_165.QC.jpg
cd762418a082f781708393caab1a2636
e59d4c9c889f9565970d0ebe41a1a9bd580087b1
14935 F20101111_AABTVH coggins_l_Page_045.QC.jpg
0fd8f07434c0c0aa19330acb5e815b96
49dc00eada576c071a7b8223f503945a6fc4f1a8
8987 F20101111_AABTUT coggins_l_Page_034thm.jpg
89cd44846b4dac1f9a95b9833fb8e0a5
52971f748d869573b8f2a7d4842b47840a7b92a7
1051916 F20101111_AABSSF coggins_l_Page_030.jp2
d09ec1fe4ad6eaed260c3340a0a4cde4
2ea46d2229c93e30084a406734c721719946fbaf
1049323 F20101111_AABSRR coggins_l_Page_015.jp2
a366653baf5e3b178dbe4087e45f4c65
cbb8e83a6a20b3e2ec3f212b41d1d2ba006b9747
36649 F20101111_AABUAP coggins_l_Page_167.QC.jpg
6d6170c4e3cd16c4d2cea6f110658825
6a8258c2a7ffbe3d06ea3803c93148e84a55802e
4006 F20101111_AABTVI coggins_l_Page_046thm.jpg
3b0bfe4f569db6633957e6daa3f83c70
bd168f1b9c308aedaa37e9ed2367415b283edb3c
36751 F20101111_AABTUU coggins_l_Page_035.QC.jpg
76496e7b8bb4264b0301ea8eb40efdfa
4f56fe1850a0f281569ef76466f141b491eda38b
1051926 F20101111_AABSSG coggins_l_Page_031.jp2
cc9d0e7e31046ffc47b7019f34cfc3df
cc675bf5f53b7b0e49f4084bc6581bd0ca65766e
F20101111_AABSRS coggins_l_Page_016.jp2
967b1f1b8af42f122b0594c5f145f8e3
fe41bc729d365eab685c1aaa11558937c68027a3
8853 F20101111_AABUAQ coggins_l_Page_167thm.jpg
fa28f0f21389fe233b971ff05f6168e1
89bdc92e3d1b8ed01b5b04fb0c9f8128db999ae8
5574 F20101111_AABTVJ coggins_l_Page_047thm.jpg
4d78c31b4532eb52ab8d96559ffbb999
cca4b077299b6bc0810bfdca2d8e898f01699c00
9045 F20101111_AABTUV coggins_l_Page_035thm.jpg
4892e1fcf2390c7a3420109b31ad1af5
5a2724c1d6ed556dddf9966edc1563aaf8c9ab63
F20101111_AABSSH coggins_l_Page_032.jp2
0331bcd3f2acbde72f2559528228b8bf
36f05b2eecfc2baf34ca425bde22362d65f85b5f
F20101111_AABSRT coggins_l_Page_017.jp2
2b8eac57c8fb98ebaf3974f428d9925b
cff7b75493eee5628d0e893822f3a9aa2a2d299f
23224 F20101111_AABTVK coggins_l_Page_048.QC.jpg
22221d63dbef22f1eb1c059a5c21bcff
e4ba8955beb16f0cd35683b467a6004ba6744023
37416 F20101111_AABTUW coggins_l_Page_036.QC.jpg
1b3ad4c5447691280c01cac1ed34bb95
4ebcec5bde4bb26cda97c8ad5b7ce4f1cb6f9507
F20101111_AABSSI coggins_l_Page_033.jp2
0c07bf1d8f8380084566c3c854ab6b78
8ab51f1e79d7b106ecb5eedddc157e42a1caf3b6
1051978 F20101111_AABSRU coggins_l_Page_018.jp2
5d4d001c3b2c9a3070bfa19bb085747c
29f4d5c58015a6c5cd9a884ec5d269bce1a85fb9
36676 F20101111_AABUAR coggins_l_Page_168.QC.jpg
16ce6b9eed06a2ffcc43280fbed9310c
e094e3fca14978b876f5ed5f1d9231262f60f4df
27882 F20101111_AABTWA coggins_l_Page_065.QC.jpg
d90ffdf8db5fdc1fe4159670e117fd89
6098d291e9b648b76393297bf4a6c846d12720cc
19105 F20101111_AABTVL coggins_l_Page_049.QC.jpg
fc660f009678b7e421574231f913211c
7cb7e3f82dfd3c3b57505988164ac6478b311d17
9076 F20101111_AABTUX coggins_l_Page_036thm.jpg
90aa0e47ed6bc291e7341e80a66593b0
a7276a703ee19c801bfa333bbc4a9805dea5f460
F20101111_AABSSJ coggins_l_Page_034.jp2
1b32a9240f2e14fa6c66d27648845be2
9617b88af3842d8cc2bb292d2d67a4f1aff1a008
F20101111_AABSRV coggins_l_Page_019.jp2
cbbc4634e94e4ce51941f51a082bd246
190d03d7c6a90cf005d9aeb78678e7ade1d18357