Factors Affecting The Distribution Of Bald Eagles And Effects Of Human Activity On Bald Eagles Wintering Along The Boise River

Robin Spahr

A Thesis
(submitted in partial fulfillment of the requirements for the degree of Master of Science of Raptor Biology)
Boise State University
March, 1990

The thesis presented by Robin Spahr entitled Factors affecting the distribution of bald eagles and effects of human activity on bald eagles wintering along the Boise River is hereby approved:

AUTOBIOGRAPHICAL SKETCH

I was born on 29 October 1963 in Indianapolis, Indiana to James and Bette Jeanne Spahr and graduated from Franklin Central High School in Indianapolis in May 1982. I attended the University of Tennessee in Knoxville from September 1982 to June 1986, and graduated with highest honors with a Bachelor of Science degree in Wildlife and Fisheries Science. In January 1987, I began studies at Boise State University working toward a Master of Science degree in Raptor Biology.

ACKNOWLEDGMENTS

I would like to thank my advisor, Karen Steenhof, for her guidance in all aspects of this project. I also wish to thank the other members of my committee, Dr. Marc Bechard for his advice and help in acquiring funding, and Dr. Kerry Reese for his suggestions and reviews of the thesis.

I am grateful to Karl Gebhardt for his efforts in initiating the study and aid in securing funding. I appreciate the help of Rich Howard of the U.S. Fish and Wildlife Service, who also helped secure funds.

I thank Lenny Young for his comments and suggestions regarding the study.

I wish to thank my field assistant, Greg Kaltenecker, for all his help.

I would like to thank Ted Swem for his help and suggestions.

I would also like to thank Keleigh Hague-Bechard for her assistance in preparing this report.

I would like to thank Steve Sweet and River Run Development Company, Bonneville Pacific Corporation, Golden Eagle Audubon Society, Idaho Foundation for Parks and Recreation, Rob Tiedemann, Chris Korte, and the Raptor Research Center at Boise State University for providing financial support for this study.

I also wish to thank the Idaho Department of Fish and Game, and the Boise District, Bureau of Land Management for the use of equipment.

LIST OF TABLES vii
LIST OF FIGURES x
LIST OF APPENDIXES xi
ABSTRACT xiii
CHAPTER 1 - OVERVIEW

Introduction
Study Area
Methods
Bald eagle census
General use patterns
Prey captures and foraging observations
Prey abundance
Data analysis
Results and Discussion
Bald eagle census
General use patterns
Prey captures and foraging observations
Prey abundance
CHAPTER 2 - FACTORS AFFECTING DISTRIBUTION OF WINTERING BALD EAGLES IN AN URBAN AREA 15

Introduction
Methods
Bald eagle census
River transects and segments
Prey captures
Data analysis
Results
Bald eagle census
General use patterns
River transects and segments
Regression analyses
Discussion
Management Implications
CHAPTER 3 - EFFECTS OF HUMAN ACTIVITY ON WINTERING BALD EAGLES IN AN URBAN AREA
Introduction
Methods
Observed human-eagle encounters
Experimental disturbances
Data analysis
Results
Observed human-eagle encounters
Experimental disturbances
Flushing frequencies
Discussion
Management Implications
CHAPTER 4 - RIVER FLOW TESTS 60
Introduction
Methods
Results
CHAPTER 5 - EFFECTS OF BARBER DAM CONSTRUCTION ACTIVITIES ON WINTERING BALD EAGLES
Introduction
Methods
Results
Conclusions
LITERATURE CITED
APPENDIXES

LIST OF TABLES

CHAPTER 1

Table 1.1. Mean number of perch trees and permanent structures per 250 m random segment of the Boise River, by zone, 1987-89 13

CHAPTER 2

Table 2.1. Means and standard deviations of river variables at bald eagle and random locations from November 1988 through February 1989 along the Boise River 30

Table 2.2. Mean numbers and standard deviations of perch trees, permanent buildings, m of road, and m of path within 250-m river segments at eagle and random locations in 1988 and 1989 along the Boise River 31

Table 2.3. Mean numbers and standard deviations of perch trees, permanent buildings, m of road, and m of path within 250-m river segments at eagle and random locations, by zone, in 1988 and 1989 along the Boise River 32

Table 2.4. Regression coefficients and r values for the logistic regression model predicting presence or absence of bald eagles along the Boise River, 1988-89 33

Table 2.5. Regression coefficients, t values, and probabilities for the multiple regression model predicting number of bald eagles along the Boise River, 1988-89 33

CHAPTER 3

Table 3.1. Numbers of encounters between bald eagles and 5 types of human activity and the percent flushing responses recorded along the Boise River during 2 winters 1987-89 48

Table 3.2. Percent of bald eagles flushing from all types of human activity recorded during individual eagle observations along the Boise River, winters 1987-89 49

Table 3.3. Means and ranges of bald eagle flushing distancesto 5 types of human activity recorded during individual eagle observations during the winters 1987-89 along the Boise River 50

Table 3.4. Means and ranges of bald eagle flushing distances from all types of human activity in 2 different zones along the Boise River recorded during individual eagle observations during the winters 1987-89 50

Table 3.5. Disturbance indexes developed for bald eagles wintering along the Boise River 51

Table 3.6. Means and ranges of bald eagle flushing distances to 4 types of human activity recorded during experimental disturbances along the Boise River, winter 1988-89. Values in parentheses exclude the outlier from calculations 52

Table 3.7. Comparison of mean flushing distances of bald eagles to human activity among 8 studies 53

CHAPTER 4

Table 4.1. R2 values for the variables measured at the test transects for flows at Diversion Dam and Glenwood St. along the Boise River during the winter 1988-89 62

CHAPTER 5

Table 5.1. Numbers of eagles and amount of time spent by perched eagles in the Barber Dam area of the Boise River prior to construction in 1988 68

Table 5.2. Numbers of eagles and amount of time spent by perched eagles in the Barber Dam area of the Boise river during construction in 1988 68

Table 5.3. Numbers of bald eagles observed in the Barber Dam area of the Boise River expressed as a proportion of the total counted in the preceding census in 1988 69

Table 5.4. Number of observations of bald eagles and time they spent at locations in the Barber Dam area of the Boise River before and during constructionin 1988 70

Table 5.5. Numbers of bald eagles using a communal roost near the Boise River prior to and during construction at Barber Dam for 12 observation periods in 1988 71

Table 5.6. Ranges of distances to 4 types of human activity to which eagles did not respond at the Barber Dam area of the Boise River before and during construction in 1988 72

LIST OF FIGURES

CHAPTER 1

Figure 1.1. Bald eagle wintering area along a 24.5-km stretch of the Boise River in southwestern Idaho in which bald eagles were studied during the winter 1987-89 14

CHAPTER 2

Figure 2.1. Diagram of bald eagle location transect and segment established along the Boise River during the winters 1987-89 34

Figure 2.2. Percentage of river habitat type at bald eagle and random locations along the Boise River during the winter 1988-89 35

CHAPTER 3

Figure 3.1. Percentage of wintering bald eagles that flushed at 5 distance categories along 2 stretches of the Boise River, 1987-89 55

Figure 3.2. Percentage of wintering bald eagles that flushed at 2 distance categories along 2 stretches of the Boise River, 1987-89 56

Figure 3.3. Cumulative probability of flushing curves predicting the percentage of wintering bald eagles that would be expected to flush from a disturbance type at a given distance along 2 different stretches of the Boise River 57

Figure 3.4. Percentage of eagles that flushed at 7 distance categories for 3 populations of wintering bald eagles, 1987-89 58

Figure 3.5. Cumulative probability of flushing curves predicting the percentage of eagles that would be expected to flush from a disturbance type at a given distance for 3 populations of wintering bald eagles 59

CHAPTER 5

Figure 5.1. Barber Dam observation area and eagle perch locations along the Boise River recorded in February and March 1988 73

LIST OF APPENDIXES

>Appendix 1. River flows in c.f.s. in the Boise River below Diversion Dam for November 1987 through March 1988 from the Bureau of Reclamation 81

Appendix 2. River flows in c.f.s. in the Boise River below Diversion Dam (DD) and the Glenwood Bridge (GB) for November 1988 through March 1989 from the Bureau of Reclamation 82

Appendix 3. Vehicle census route driven when counting bald eagles along the Boise River during the winters 1987-89 83

Appendix 4. Numbers of bald eagles counted along the Boise River during the winter 1987-88 84

Appendix 5. Distribution of bald eagles along the Boise River during winter censuses 1987-89 85

Appendix 6. Numbers of bald eagles counted along the Boise River during the winter 1988-89 86

Appendix 7. Amount of time bald eagles were observed in each zone during individual eagle observations along the Boise River, winter 1987-88 87

Appendix 8. Amount of time bald eagles and human activity were observed in each zone during individual eagle observations along the Boise River, winter 1988-89 87

Appendix 9. River characteristics of bald eagle prey captureand attempted prey capture points recorded along the Boise River during the winter 1987-88 88

Appendix 10. Observations of foraging by bald eagles along the Boise River during the winter 1987-88 89

Appendix 11. River characteristics of bald eagle prey captureand attempted prey capture points recorded along the Boise River during the winter 1988-89 90

Appendix 12. Observations of foraging by bald eagles along the Boise River during the winter 1988-89 93

Appendix 13. Number of observations of prey and their frequency of occurrence in bald eagle pellets collected in a communal roost along the Boise River during the winter 1988-89 92

Appendix 14. Number of waterfowl counted by zone on the Boise River between Diversion Dam and Glenwood St. during the winter 1988-89 93

Appendix 15. Number and species of ducks counted on the Boise River between Diversion Dam and Glenwood St. during the winter 1988-89 94

ABSTRACT

Bald eagle numbers, foraging, distribution, and response to human activity in an urban area were studied for 2 winters along the Boise River in Idaho. Eagles arrived on the study area in November and left by March during both winters. Peak counts of eagles (21 in 1987-88 and 25 in 1988-89) occurred during the first week of February both winters. Eagle numbers were negatively correlated with air temperature and with wind velocity. Eagles were observed foraging mainly on mule deer carcasses, small mammals, and fish.

Eagle locations were compared to random locations along the river, and regression models were developed to predict the presence or absence and abundance of eagles in an area. Eagles were found at pools more often than expected, and at areas with low human development containing a high number of perch trees. Variables that predicted eagle use of an area were the number of perch trees and the number of commercial buildings present, the river habitat type, and the river width.

Effects of human activity were studied by observing human/eagle encounters and conducting experimental disturbances. Based on disturbance indexes, developed from observed flushing frequencies and distances, eagles were least tolerant of walkers, then bicyclers and fishermen. Based on experimental disturbances, eagles were least tolerant of disturbers that approached slowly. Eagles were more tolerant of activity along the most developed stretch of the river. Eagles in this were, in general, more tolerant of activity than eagles studied in other areas. A probability of flushing curve was developed which predicts the proportion of eagles that will flush at a given distance.

CHAPTER 1 - OVERVIEW

INTRODUCTION

The Boise River supports a small concentration of bald eagles (Haliaeetus leucocephalus) during the winter months (November-March). A 24.5-km stretch between Lucky Peak Dam and the Glenwood Street Bridge flows through Boise, Idaho. This stretch is highly developed for residential and commercial uses, and provides recreation for the people of Boise. A path, used year round by walkers, joggers, and bicyclers, parallels the river through the city. The river is stocked with brown trout (Salmo trutta) and rainbow trout (Oncorhynchus mykiss) and is fished throughout the year. Despite its development and activity, eagles continue to use this stretch of river during the winter.

Jensen (1981) recorded distribution and numbers of bald eagles along 30 km (18 miles) of the Boise River from Mores Creek Bridge to the Broadway Street Bridge on the east side of Boise. Reynolds et al. (1985) studied the distribution of eagles along 10 km of the river from Lucky Peak Dam to the Broadway Street Bridge and identified a roosting area on the river. Information on eagle use of the river from Lucky Peak Dam to the Glenwood Street Bridge has been recorded since 1982 (excluding 1985) during the Idaho midwinter bald eagle survey conducted in January. According to these data, eagles have been observed throughout the stretch.

Bald eagles wintering on the Boise River are thought to eat mountain whitefish (Prosopium williamsoni), black-tailed jackrabbits (Lepus californicus), California quail (Callipepla californica), Gray partridge (Perdix perdix), and waterfowl (Jensen 1981).

Information on the effects of human activity on use of the river by eagles is lacking and factors which determine their distribution are not known. Easy access to the river and the presence of several sites which provided excellent views of the river facilitated observation of eagles in this area. These features provided a unique opportunity to study the effects of development and human activity on use of the river by bald eagles in an urban setting.

The city of Boise is growing, and the development and accompanying human activity may threaten eagles wintering on the river. Therefore, it is important to document what influences eagle use, and what levels of human activity eagles can tolerate. This information is especially important along the Boise River because it has the potential to be altered by development.

I studied bald eagles along the Boise River for 2 winters 1987-89. The study had several objectives which are addressed in detail in the following chapters. This chapter reports the results of bald eagle censuses conducted during the 2 winters. Also contained in this chapter are results of several hours of observations of eagles. They include information on prey captures, foraging, and diet, and use of a communal roost. Chapters 2 and 3 present the results and discussion of the main objectives of this study. Those objectives were to determine factors influencing the distribution of bald eagles along the river during the winter, and how human activity them. Chapter 4 presents information on how different river variables were affected by the changes in flow rates. Chapter 5 presents results from a separate, but related, study on how construction at a dam within the study area affected bald eagles.

STUDY AREA

Bald eagles were studied along the Boise River, a tributary of the Snake River, in southwestern Idaho, Ada County, between Lucky Peak Dam and the Glenwood Street Bridge on the west side of Boise (Fig. 1.1) during the winters 1987-89. Land formations and use varied along this 24.5-km stretch of the river. Directly downstream from Lucky Peak Dam, the river narrowed into a 45-m deep canyon, then widened into a floodplain as it flowed toward Boise. Some pastureland was adjacent to the river between the canyon and the city, and residential and commercial development occupied floodplain sections of the river within the city.

A path, used by walkers, joggers, and bicyclers year round, paralleled the river through the city. The river was stocked with brown trout (Salmo trutta) and rainbow trout (Salmo gairdneri) and was fished throughout the study period.

Black cottonwood (Populus trichocarpa) dominated the riparian zone, but several willows (Salix spp.) and other native and introduced vegetation were also present.

River flow rates varied during the study period depending on precipitation, discharge from Lucky Peak Dam, and diversion of water for irrigation (Appendixes 1 and 2).

METHODS

Bald Eagle Census

To determine what areas of the river eagles were using, and to determine the number of eagles present in the study area throughout the winter, vehicle censuses were conducted twice a week throughout the winter (1 during midweek and 1 during weekend) between 7 November and 23 March during 1987-88 and between 29 October and 29 March during 1988-89. The census route extended from Lucky Peak Dam to the Glenwood Street Bridge (Appendix 3). The census was conducted in the morning during the week when human activity was low and in the afternoon during the weekend when human activity was high. Numbers, ages, and activities of all eagles seen were recorded. Activities included soaring, flying, perching, and foraging. Foraging was defined as consuming prey, capturing prey, and/or attempting to capture prey. Locations of all eagles were marked on a 7.5 minute topographic map.

An eagle location was defined as any site below Diversion Dam where an eagle was observed perching within 75 m of the river. To be considered a separate location, an eagle location had to be at least 50 m from an adjacent eagle location, or if an eagle location was closer than 50 m to another, but was at a different river habitat type (pool, riffle, or run), then it was considered a separate eagle location. If more than 1 eagle was seen at a location during a single census then that location was considered 1 eagle location only. If an eagle was observed at the same location on different census days then that location was considered 2 separate eagle locations. These locations were treated statistically independent. In 1988-89 all types of human activity and the number of people observed within 75 m of the river were recorded during the censuses. Air temperature and wind velocity were measured at the start and end of the census.

General Use Patterns

To compare the distribution of bald eagles at different locations along the river, the study area was divided into 2 zones. Zone 1 extended from Lucky Peak Dam to the Spring Meadow development area for 10.0 km and was upriver from the city limits of Boise. A communal roost was present at the upper end of Zone 1 approximately 350 m southwest of the river. Zone 2 extended from Spring Meadow to the Glenwood Bridge for 14.5 km and was within the city limits of Boise (Fig. 1.1).

The number of perch trees, number of residential buildings, number of commercial buildings, m of road, and m of path, within 75 m of the river banks, were tabulated in randomly selected 250-m segments of the river. Students t tests were conducted to determine whether or not there was a difference in the amount of development between these zones.

Prey Captures and Foraging Observations

To record prey captures and foraging I observed individual eagles 2 days per week from 5 December to 27 January during

1987-88 and 8 December to 20 March during 1988-89. A river km and direction (upriver or downriver) were randomly selected. Beginning 0.5 h before sunrise I went to the selected km, traveled in the selected direction, and followed, at a minimum distance of 200 m, the first eagle seen, recording all eagle activities. That eagle was followed until it could no longer be observed, at which time a new river km and direction were randomly selected, and the first eagle seen was followed for as long as possible. Eagles were followed and observed by vehicle, using binoculars and a spotting scope. This procedure was repeated until 0.5 h after sunset. Any prey captures or attempted prey captures that were observed were recorded, and the locations were marked on a 7.5 minute topographic map.

To identify other prey eagles were eating, I searched the communal roost for, and collected, pellets once a week. The search for pellets started after eagles were observed using the roost on 5 December. The pellets were analyzed by teasing them apart and identifying food items from bones, feathers, and hair. Prey was quantified by listing frequency of occurrence of food items. The night roost was searched for pellets 15 times during the winter between 9 December 1988 and 5 April 1989. Pellets were found the last 11 times the roost was searched.

Prey Abundance

To determine relative fish abundance in the 2 zones, information on fish abundance was obtained from the Idaho Department of Fish and Game. According to fish stocking information, catchable-sized rainbow trout (25-30 cm) were stocked during all months that eagles were present in the study area. Trout were stocked at 1 location in Zone 1 and at 4 locations in Zone 2 (Walt Rast, Idaho Dep. of Fish and Game, pers. commun.).

Waterfowl censuses were conducted 7 times during the winter along the 21.5-km stretch between 4 December 1988 and 19 February 1989.

Data Analysis

Correlation analyses were conducted to determine the relationship between eagle numbers and air temperature, and eagles numbers and wind velocity, at the start and end of the censuses.

RESULTS AND DISCUSSION

Bald Eagle Census

During the winter 1987-88 the first adult eagle was observed on 7 November in the Barber Pool area, and the first immature eagle was seen on 14 November (Appendix 4). The peak count occurred on 6 February when 13 adult and 8 immature eagles were counted. The last observation of an immature eagle occurred on 12 March and the last adult was seen on 19 March.

Eagles were observed 254 times during the 1987-88 censuses (Appendix 5). Perching was the most common activity observed at (79.9%, n=203). The next most frequent activity was soaring (9.4%, n=24), then flying (8.7%, n=22), and lastly, foraging (2.0%, n=5).

Adults accounted for 71% (n=180) of the observations and immatures for 29% (n= 74). Near the end of winter in March, the age composition changed to 64% adult and 36% immature.

A color marked eagle was observed on 26 January in Barber Park. The eagle was in a transition plumage (4 or 5 yrs.). Based on the color band and leg position the eagle was banded in the Greater Yellowstone Ecosystem Snake Unit-Wyoming.

The number of bald eagles counted on the 1987-88 census was negatively correlated with air temperature at the beginning of the census (r = -0.50, P = 0.01) and air temperature at the end of the census (r = -0.45, P = 0.02). The number of eagles and the wind velocity were also negatively correlated (r = -0.46, P = 0.03).

During the winter of 1988-89, the first bald eagle of the season was seen on 5 November. It was an adult observed flying at Barber Pool. The first immature was seen on 7 December just west of Barber Park (Appendix 6). The peak count occurred on 1 February when 25 bald eagles (16 adults and 9 immatures) were observed. The last census was conducted on 29 March. At that time 1 adult eagle was seen. An adult was also seen on 31 March. Eagles stayed in the area later during 1989 than 1988 probably because colder weather and increased snow and ice in 1989 caused starvation of deer in the area. Deer carcasses were readily available to eagles through March.

A total of 385 eagle observations was recorded during the 1988-89 censuses (Appendix 5). The most common activity observed during the censuses was perching (72.7% of observations, n=280). The second most common activity was soaring (18.2% of observations, n=70), then flying (8.8% of observations, n=34), and then foraging (0.3 % of observations, n=1). If the 11 unknown age observations are excluded, adults accounted for 60% (n=225) of the observations and immatures for 40% (n=149) of the observations. During March observations, the age composition changed to 35% adults and 65% immatures. Adults may have left earlier to find and establish breeding territories, and immatures may have stayed longer because food was still available.

The number of bald eagles counted on the 1988-89 census was negatively correlated with the air temperatures at the beginning of the census (r = -0.56, P = 0.0001) and air temperatures at the end of the census (r = -0.55, P = 0.0002). There was a significant negative correlation between the number of eagles and the wind velocity after the count (r = -0.44, P = 0.0038).

One communal roost was known in the study area. It was located approximately 350 m southwest of the river in the southwest end of the Barber Pool area within Zone 1. The roost was composed mainly of black cottonwood (Populus trichocarpa) trees both live and dead. The roost was bordered on the southwest by a large levy that rose above the roost floor approximately 6.5 m. Eagles were observed in the roost mornings, afternoons, and evenings, during heavy snow storms, and in foggy conditions. No regular roost counts were conducted, however eagles were observed in the roost on many of the censuses. The highest number of eagles observed in the roost at 1 time was 17 (12 adults and 5 immatures) at 0810 on 8 February 1989.

General Use Patterns

There were significantly more commercial buildings per segment (t = -7.40, P = 0.0001), and significantly more m of path per segment (t = -3.40, P = 0.001) in Zone 2 than Zone 1 (Table 1.1). Based on these results, Zone 2 was considered to be more developed than Zone 1.

Prey Captures and Foraging Observations

During the winter 1987-88, 20 days were spent following 22 eagles for a total of 2075 minutes (34 hours and 35 minutes). I observed 17 adults over 15 days for 1680 minutes (28 hours) and 5 immatures over 5 days for 395 minutes (6 hours and 35 minutes) (Appendix 7). Of the total observation time, 81% was spent watching adults and 19% was spent watching immatures.

During the winter 1988-89, 29 days were spent following 43 eagles for a total of 8243 minutes (137 hours and 23 minutes). I observed 28 adults over 23 days for 6183 minutes (103 hours and 3 minutes) and 15 immatures over 14 days for 2060 minutes (34 hours and 20 minutes) (Appendix 8). Of the total time spent observing eagles, 75% was spent watching adult eagles and 25% was spent watching immature eagles. More time was spent following adults because there were more adults than immatures present. The immatures tended to leave the study area more often and were quickly lost from view.

During both winters adults and immatures were followed most often in Zone 1. The most minutes of human/eagle encounters also occurred in Zone 1.

During 1987-88 3 successful and 2 unsuccessful prey captures were observed (Appendix 9). Three of the attempts occurred in the same spot at Barber Park in Zone 1. All attempts were made on fish by adult eagles. The fish captured on 3 February was identified as a rainbow trout. Only 1 attempt was made in Zone 2, and it was successful. Several other instances of eagles foraging on fish were observed during the winter (Appendix 10).

During 1988-89, 2 successful prey captures and 2 unsuccessful prey captures were observed on the river at 3 different locations (Appendix 11). All attempts were by adults on fish. Two attempts, by the same eagle on the same day, occurred at a pool located in Zone 1. The first attempt was unsuccessful and the second was successful. The other 2 attempts occurred at riffles in Zone 1.

Eagles were observed foraging on fish and deer carcasses (Appendix 12). According to frequency of occurrence of food items in 186 pellets, they fed on deer carcasses more than any other food (Appendix 13). Ground squirrels and fish were the next most common prey found in pellets. Though waterfowl were common in the study area (Appendixes 14 and 15), no observations of eagles foraging on waterfowl were recorded, and only 3 pellets contained waterfowl remains.

I observed 2 instances of eagles with hatchery rainbow trout. One eagle dropped a fish while being chased by another eagle, and a fish was found below a perch tree with talon marks. Both were identified as hatchery rainbow trout by a former Idaho Dept. of Fish and Game fish hatchery employee.

Prey Abundance

According to fish stocking information from the Idaho Department of Fish and Game, catchable-sized rainbow trout (25-30 cm) were stocked during all months that eagles were present in the study area. Trout were stocked in 1 location in Zone 1 and 4 locations in Zone 2 with approximately 75% more hatchery trout/km stocked in Zone 2 than Zone 1. Mallards (Anas platyrhynchos) were the most common duck on the river. Most waterfowl were counted on areas of the river adjacent to city parks in Zone 2 (Appendixes 14 and 15).

CHAPTER 2 - FACTORS AFFECTING THE DISTRIBUTION OF WINTERING BALD EAGLES IN AN URBAN AREA

INTRODUCTION

Numbers of bald eagles in western North America have increased in the last few decades (Gerrard 1983, Swenson 1983). Although populations are increasing, wintering habitat in many areas is being altered or lost through development (Reed 1979, Stalmaster 1980, Schultz 1982, Anderson et al. 1988). Management and regulation of the development of remaining wintering areas requires knowledge of what factors are important about these areas to bald eagles.

Much is already known about the general habitat requirements of wintering bald eagles. An abundant food source is an important factor in determining bald eagle distribution during winter. In Washington spawned salmon (Oncorhynchus spp.) attract large concentrations of bald eagles (Stalmaster 1976). In Utah concentrations of eagles eat black-tailed jackrabbits (Lepus californicus) (Platt 1976). In other areas waterfowl are an important food source (Lish and Lewis 1975, Griffin et al. 1982). Other factors such as availability of perch trees and open water, and weather may influence bald eagle distribution during winter (Steenhof et al. 1980).

A few studies have reported on specific characteristics of eagle foraging sites. Nesting eagles in Arizona foraged in river areas containing deep pools bordered by riffles or sandbars (Haywood and Ohmart 1986). Shallow pools with no surface turbulence were identified as important foraging areas in northern California (Biosystems Analysis, Inc. 1985).

Human activity is another factor that may affect bald eagle distribution in winter (Steenhof 1976). Stalmaster and Newman (1978) found that human activity displaced eagles, causing them to seek areas with less human activity. Skagen (1980) reported a decrease in feeding when human activity was within 200 m of the feeding area. Stalmaster (1983) calculated energy budgets for wintering bald eagles and suggested that human activity could increase energy stress.

The purpose of this study was to determine specific factors that affect the distribution of wintering bald eagles. I studied a small concentration of bald eagles on the Boise River in Idaho. This concentration of eagles is found partially within the city of Boise which provided an opportunity to determine how human activity and development, along with habitat characteristics, affect distribution of wintering bald eagles.

METHODS

Bald Eagle Census

To determine what areas of the river eagles were using, vehicle censuses were conducted twice a week throughout the winter (1 during midweek and 1 during weekend) between 7 November and 23 March during 1987-88 and between 29 October and 29 March during 1988-89. See Chapter 1 Methods - Bald Eagle Census for details.

River Transects and Segments

At each eagle location a transect was established across the river during the same week eagles were observed at the locations. Along the transect, 3 points (0.25, 0.5, and 0.75 of the distance across the river) were sampled for river depth, velocity, temperature, and turbidity. River width was measured along the transect, and the location was classified as a pool, riffle, or run.

With the eagle location as the midpoint, a 250-m long segment of the river was established for sampling at each eagle location. The number of potential perch trees and the number and type of permanent structures within 75 m of the river were counted in each segment (Fig. 2.1). Potential perch trees were defined as any tree in which an eagle had been observed perching or any tree with a dbh of at least 60 cm and a height of at least 15 m. Permanent structures included buildings, walking/bicycling path, and roads. Buildings were categorized as residential (homes, apartment buildings, condominiums, dormitories), or commercial (office buildings, hotels, campus buildings, saw mills). Counts of perch trees and permanent structures were conducted in March after most eagles had left the area.

To compare river characteristics, perch trees, and permanent structures at eagle locations to the entire study area, the 21.5-km stretch between Diversion Dam and the Glenwood Street Bridge was divided into 86 250-m segments. Each time an eagle location was measured, one of the 86 segments was randomly selected. Within each segment selected, a transect was established n meters (n being a randomly selected number between 1 and 250) from the upriver segment boundary. The same measurements were conducted at these random locations as those at the eagle locations.

Prey Captures

To record prey captures I observed individual eagles 2 days per week from 5 December to 27 January during 1987-88 and 8 December to 20 March during 1988-89. See Chapter 1 Methods - Prey Captures and Foraging Observations for details. Transects were established across each prey capture location and measured the day of or the day after the observation using the same methods used for measuring the eagle locations. Perch trees and permanent structures were counted using the method described previously for eagle locations with the capture or attempted capture point as the midpoint of the 250-m segment.

Data Analysis

Chi-square goodness-of-fit analysis was used to test the hypothesis that eagles were distributed evenly along the 24.5-km study area, and to test the hypothesis that eagle observations were distributed at river habitat types similar to that expected based on the proportions from the random transects. Bonferroni-Z confidence intervals were used to determine which habitat types were used more or less than expected (Neu et al. 1974).

To test the hypothesis that the number of eagles counted on the river was independent of the number of people counted on the river, correlation analysis, comparing the number of eagles with the number of people, was conducted. Weekend observations were excluded to avoid any time of day bias that may have occurred because bald eagles are more likely to perch on the river in the morning than in the afternoon arch. The mean was 8, and the range was 1 to 25. The peak count was on 1 February (Appendix 6).

A total of 385 eagle observations was recorded during the censuses. If the 11 unknown age observations are excluded, adults accounted for 60% (n=225) of the observations and immatures for 40% (n=149) of the observations. However during March observations, the age composition changed to 35% adults and 65% immatures.

RESULTS

General Use Patterns

During the winter 1987-88, the linear density of perched eagle observations, expressed in eagle observations/km, was higher in Zone 1 (16 eagle observations/km) than in Zone 2 (3 eagle observations/km). In 1988-89, the linear density of perched eagle observations was again higher in Zone 1 (24/km) than in Zone 2 (3/km) (Appendix 5). The observed distributions were different from what would be expected if the eagles were distributed evenly along the study area (X2 = 10.90, d.f. = 1, P < 0.001 and X2 = 21.36, d.f. = 1, P < 0.001, respectively).

The number of bald eagles and people counted per segment during the morning censuses were not correlated (r = -0.11, P = 0.08). Eagles and people per segment within Zone 1 were not correlated (r = 0.05, P = 0.60). No people were counted in Zone 2 in the morning along the river. Although eagle numbers were not directly related to people numbers, the number of people was correlated with development within a segment. Along the entire study area there were slight positive correlations between the number of people in a segment and the number of commercial buildings in a segment (r = 0.28, P = 0.003) and the number of people and the m of path in a segment (r = 0.21, P = 0.03). In Zone 1 there was a positive correlation between people and commercial buildings in a segment (r = 0.35, P = 0.02). In Zone 2 people and commercial buildings were positively correlated (r = 0.25, P = 0.05), people and m of road were positively correlated (r = 0.27, P = 0.03), and people and m of path (r = 0.45, P = 0.0002) were positively correlated.

River Transects and Segments

Between 27 November and 29 March 1988-89, 90 eagle locations, 2 prey capture, and 2 attempted prey capture observations were recorded (Appendix 11). These 94 observations were recorded at 42 different locations. There were 94 random transects measured at 73 different locations. Twenty-three of the eagle locations were pools, 14 were riffles, and 5 were runs. This distribution was not different from that expected based on the random locations which contained 36 pools, 20 riffles, and 17 runs (X2 = 3.51, d.f. = 2, P > 0.10). However, when all 94 observations were considered, the distribution was different from random (X2 = 15.71, d.f. = 2, P < 0.001). Using the Bonferroni-Z confidence interval, eagles were observed at pools (n=57) more than expected, at riffles (n=31) the same as expected, and at runs (n=6) less than expected (Fig. 2.2).

I also compared the habitat type distributions by zone. When only the 42 individual locations were used, the distribution in Zone 1 was different from expected (X2 = 6.67, d.f. = 2, P < 0.05). Eagles were observed at pools and riffles the same as expected and at runs less than expected. When all 94 observations at the 42 locations were used in the analyses, there was a difference in the distribution in Zone 1 as well (X2 = 34.80, d.f. = 2, P < 0.001). Eagles were observed at pools (n=46) the same as expected, at riffles (n=18) more than expected and at runs (n=1) less than expected. Based on the random locations, Zone 1 contained 63% pools and Zone 2 contained 44% pools. This difference was not significant (X2 = 1.29, d.f. = 1, P > 0.25).

I compared river measurements at eagle and random locations by habitat type (Table 2.1). In riffles, widths were greater

(t = -2.37, P = 0.02), and temperatures were lower (t = 2.07, P = 0.02) at eagle locations. In runs, widths were greater at eagle locations (t = 2.07, P = 0.04).

In 1987-89, eagle locations had more perch trees/segment than random locations. Eagle locations also had fewer commercial buildings/segment, and fewer m of path/segment than random (Table 2.2).

The number of perch trees and structures between eagle and random segments were also compared by zones (Table 2.3). In Zone 1, there were more perch trees/segment and more commercial buildings/segment at eagle locations than at random locations. In Zone 2 eagle locations had more perch trees/segment and more residential buildings/segment than random locations. There were fewer commercial buildings/segment and fewer m of path/segment at eagle locations than at random locations.

Regression Analyses

Before entering variables into the regression equations, the independent variables were tested for collinearity by conducting correlation analyses. Some variables were correlated. Width was negatively correlated with depth and velocity, and velocity was negatively correlated with depth. Also the number of commercial buildings/segment was positively correlated with the m of path/segment. Because of these correlations, width and number of commercial buildings were used in the regression equation and depth, velocity, and m of path were not included. River temperature and turbidity changed through the season and were negatively correlated with the number of eagles in the study area throughout the winter. Because of these correlations, both variables were also excluded from the regression equations.

Variables that were entered into the regression equations were river habitat type, river width, number of trees/segment, number of residential buildings/segment, number of commercial buildings/segment, and m of road/segment.

The significant variables in the logistic regression model, which predicted the presence or absence of an eagle in a segment, were the number of trees in a segment, the number of commercial buildings in a segment, the river habitat type at the transect location, and river width (Table 2.4). The total R2 was 0.2294.

In the multiple regression analysis, the number of eagles counted in the segments was the dependent variable. The same independent variables as those in the logistic regression model were used.

The significant variables in this model were the same as those for the logistic model (Table 2.5). The R2 value was 0.1924 for this model.

DISCUSSION

Eagles were not distributed evenly along the study area. They were observed most often in Zone 1 where development was lower, prey abundance was lower, and the distance to the roost was closer. Several researchers have reported that eagles avoid areas of high human activity (Steenhof 1976, Stalmaster and Newman 1978, Russell 1980, Skagen 1980, Conrad and Stillwell 1984, McGarigal 1988). Although prey abundance was generally higher in Zone 2 than Zone 1, prey may have been more available to eagles in Zone 1 because there was less human activity to disturb foraging eagles. Steenhof (1976) observed that eagles were more tolerant of human activity at preferred feeding sites than at less preferred sites. However, Stalmaster and Newman (1978) found that human activity disrupted feeding and in fact feeding birds were the most sensitive to human activity. Skagen (1980) also reported that eagles were most sensitive to human activity while feeding. Eagles were observed capturing prey in both zones, however, eagles did not stay in Zone 2 for long periods. They may only have used Zone 2 while actively foraging. I was not able to measure prey abundance or availability at specific eagle and random locations. Differences in prey abundance and availability at these specific locations may have been a determining factor in eagle distribution.

The presence of a communal roost in Zone 1 may have accounted partially for the higher distribution of eagles there. However, bald eagles are known to travel several km from roosts to diurnal perches and feeding areas. Distances reported range from 0.25 to 24 km (Edwards 1969, Hansen et al. 1980, Steenhof et al. 1980, and Keister and Anthony 1983). During 1 eagle observation, I followed an eagle from the roost to 18.5 km below the roost on the river.

Data on river habitat use were consistent with those of BioSystems Analysis, Inc (1985) who observed that bald eagles selected pools above riffles or runs in riverine conditions. Haywood and Ohmart (1986) reported that benthic-feeding fish were most vulnerable to breeding bald eagles at deep pools bounded by riffles or sandbars.

Eagles along the Boise River were observed capturing and eating trout and other unidentified fish species (Appendixes 9-12). Allen (1969) reported that in temperate climates stream-dwelling salmonids move to pools to seek cover during winter. Because pools in the Boise River during winter are long and shallow and may not have provided adequate cover for trout (Asbridge 1988), eagles may have foraged at these areas to increase their hunting success.

Because eagles seemed to be selecting perch locations at pools one might expect that pool and eagle numbers would be similar between zones. The proportion of pools was approximately equal in both zones, while the density of eagles was higher in Zone 1.

Along several areas of the river development was only present along 1 bank. Eagles may perch at wider areas along the river to provide themselves with a wider buffer from activity on the opposite bank. Stalmaster (1980) reported that disturbances to eagles are reduced on wide channels because of the longer protective distances. However, this hypothesis was not tested in my study. The difference in temperature at riffles between eagle and random locations was small, although statistically significant. However, this difference may not have biological significance.

Prey capture and attempted prey capture points had slower velocities than either the eagle location or random locations in riffles. Eagles may have foraged at riffles with slow velocities (Appendixes 9 and 11) because fish were more likely to be present in slower water. Cunjak and Power (1987) reported that during the winter trout were found most often in riffles with reduced water velocities because they were better able to maintain their position in the water with less energy expenditure. Rimmer et al. (1985) documented that hatchery trout were less able to hold their positions in the water column, especially in winter water conditions (temperatures less than 8 0C), than wild trout, suggesting that they are more likely to be found in slower water velocities. However, the number of prey capture observations was small, and no statistical tests were conducted.

Eagles were associated with areas that had a high number of perch sites near the river. Several researchers have observed that eagles are more tolerant of human activity if they are buffered from it by vegetation (Krauss 1977, Stalmaster and Newman 1978, Fraser et al. 1979, Reed 1979). Eagles in this study may have been attracted to areas with a high number of perch trees because they were buffered from human activity.

The number of commercial buildings and m of walking/biking path in an area also influenced eagle distribution. Eagles may have avoided developed areas because they were associated with higher levels of activity than undeveloped areas.

Both regression models concluded that the number of perch trees, the number of commercial buildings, the river habitat type, and the river width were significant variables in predicting the presence or absence of eagles in a segment and numbers of eagles in a segment. These results support the eagle distribution findings and the eagle versus random location comparisons. The total amount of variance explained in these models was low. Because prey availability was not measured at specific locations, it was not entered into the model. It is probably an important factor determining eagle distribution.

In this study the most important factors that determined eagle distribution along the river were the number of perch trees in an area, the amount of development in an area, the river habitat type, and the river width. Eagles were most likely to be found along wide stretches of the river, or at pools, with a high number of perch trees, and low amount of development.

MANAGEMENT IMPLICATIONS

Eagles wintering near urban areas are probably more tolerant of human activity than eagles wintering in more isolated areas.

Several researchers have observed that eagles may habituate to human activity in area where human activity is high (Stalmaster and Newman 1978, Russell 1980, Skagen 1980, Harmata 1984, Knight and Knight 1984). Individual eagles vary however, in their responses to human activity based on their past experiences. Eagles wintering on the Boise River were relatively tolerant of human activity, otherwise they would winter in more isolated areas. However, these eagles still avoided areas of relatively high development within the study area.

Based on my findings, development of remaining land along river corridors should be limited or at least regulated to limit impacts on eagle habitat. Set backs from the river and vegetative buffer zones might help limit these impacts (See Chapter 3 - Management Implications). Perch trees should be left along the river, and new perch sites should be established through regeneration of forested riparian areas.

Table 2.1. Means and standard deviations of river variables at bald eagle and random locations from November 1988 through February 1989 along the Boise River.

Pool Riffle Run

Measure Type (n=30) (n=17) (n=3)

Width (m) Eagle 35.5 + 10.5 34.0 + 11.0* 33.9 + 5.4*

Random 33.9 + 10.9 28.0 + 14.1 27.4 + 7.2

Temp. (0C) Eagle 1.6 + 1.6 1.5 + 1.2* 1.7 + 1.1

Random 1.7 + 1.5 2.0 + 2.0 1.5 + 1.5

Depth (cm) Eagle 66.0 + 24.1 32.6 + 20.3 40.0 + 12.4

Random 63.3 + 31.1 37.4 + 18.3 45.7 + 11.2

Velocity (cm/s) Eagle 21.5 + 11.6 70.0 + 46.4 51.3 + 14.6

Random 24.0 + 15.0 82.5 + 46.1 54.7 + 20.9

Turbidity (FTU) Eagle 32.0 + 40.5 24.6 + 37.7 38.9 + 40.8

Random 32.8 + 41.4 22.1 + 32.2 45.9 + 50.2

*denotes a significant difference between the eagle and random

locations (P < 0.05)

Table 2.2. Mean numbers and standard deviations of perch trees, permanent buildings, road, and paths within 250-m river segments at eagle and random locations in 1988 and 1989 along the Boise River.

Eagle Random

Type of structure (n=61) (n=76) t P

Perch trees* 33.5 + 21.6 20.2 + 16.4 3.97 0.0001

Residential buildings 1.0 + 1.9 0.8 + 1.4 0.51 0.6077

Commercial buildings* 0.3 + 0.8 1.1 + 1.4 -4.21 0.0001

Roads (m) 36.3 + 84.6 66.4 + 139.9 -1.56 0.1220

Path (m)* 166.6 + 148.8 255.6 + 201.0 -2.98 0.0035

*denotes a significant difference between the eagle and random , 1988-89.

Regression

Variable coefficient t P

Number of perch trees 0.012 6.035 0.0001

Number of commercial buildings -0.206 -6.021 0.0001

River habitat type -0.170 -3.687 0.0003

River width 0.007 2.230 0.0263

Figure 2.1. Diagram of bald eagle location transect and segment established along the Boise River during the winters 1987-89.

Figure 2.2. Percentage of river habitat type at bald eagle and random locations along the Boise River during the winter 1988-89.

CHAPTER 3 - EFFECTS OF HUMAN ACTIVITY ON WINTERING BALD EAGLES IN AN URBAN AREA

INTRODUCTION

Several studies (Steenhof 1976, Krauss 1977, Stalmaster and Newman 1978, Russell 1980, Schonholtz 1980, Skagen 1980, Walter and Garrett 1981, Knight and Knight 1984) have documented responses of wintering bald eagles to human activity, but few have investigated wintering eagles in an urban area. In general, researchers have reported negative responses of eagles to human activity. Stalmaster and Newman (1978) found that human activity displaced eagles causing them to seek areas with less activity. Skagen (1980) reported a decrease in feeding when human activity was within 200 m of the feeding area. Stalmaster (1983) calculated energy budgets for wintering bald eagles and suggested that human activity could increase energy stress.

Several researchers, however, have reported that eagles may be more tolerant of people in some situations, such as prime foraging areas (Steenhof 1976) or areas with high levels of human activity (Stalmaster and Newman 1978, Russell 1980, Skagen 1980, Harmata 1984). Eagles may become tolerant of people as encounters with people increase over time (Conrad and Stillwell 1984, Knight and Knight 1984). If bald eagles habituate to human activity, one would predict that eagles wintering in an urban area would be more tolerant of human activity than eagles wintering in more isolated areas.

As development of bald eagle wintering habitat increases, bald eagles may be forced to winter in close contact with people. It is important to document what levels of human activity eagles can tolerate so that activity can be regulated to protect wintering habitat. I studied a small concentration of bald eagles wintering along the Boise River, partially within the city of Boise, Idaho for 2 winters from 1987-89.

The goals of this study were to determine types of human activity to which eagles wintering near an urban area were most sensitive, to determine bald eagle flushing frequencies and flushing distances from different types of activity, to develop a probability of flushing curve for this population of eagles, and to compare flushing distances and frequencies of this population to other populations of eagles in North America.

METHODS

Observed Human-eagle Encounters

To record human-eagle interactions I observed individual eagles 2 days per week from 5 December to 27 January during 1987-88 and 8 December to 20 March during 1988-89. See Chapter 1 Methods - Prey Captures and Foraging Observations for details. All responses to human activity were quantified by recording flushing distances to walkers/skiers, joggers, fishermen, bicyclers, and vehicles with a rangefinder. If no flushing response was observed, the distance of the eagle to the closest activity within 300 m was measured using a rangefinder. The number of walkers/skiers, joggers, fishermen, bicyclers, and vehicles that came within 350 m of the eagle were also recorded.

Experimental Disturbances

Beginning the first week in January 1989 I conducted experiments that simulated disturbances to eagles. These experiments were run 2 days per week throughout January and February, 1 day/week in Zone 1 for a total of 105 disturbances and 1 day/week in Zone 2 for a total of only 7 disturbances. During several weeks no disturbances were conducted in Zone 2 because no eagles were observed in the zone.

A total of 112 disturbances of 4 different types were simulated. These disturbance types included single walkers (n=47), groups of walkers (n=27), pointers (n=27), and fishermen (n=11). Single walkers consisted of 1 person approaching an eagle. Groups of walkers consisted of 2-3 people approaching an eagle. Pointers consisted of 2-3 people approaching an eagle, stopping every 10 m, and looking and pointing at the eagle for 30 s. Fisherman which consisted of 1 person approaching an eagle from the water.

Experimental disturbances started approximately 350 m from perched eagles, and disturbers moved toward the bird until it flushed. I randomly selected the disturbance type and starting location. Flushing distances were recorded for all 4 disturbance types along with age of the eagle (adult or immature), location (Zone 1 or Zone 2), and time of day (sunrise-1000, 1000-1400, or 1400-sunset). Few disturbances were conducted between 1400 and sunset because eagles were off the river or at a roost at the upper end of the study area at this time. No disturbances were conducted at the roost.

Data Analysis

Flushing data recorded during individual eagle observations were analyzed using an analysis of variance to test the hypothesis that mean flushing distance did not differ among activity types or location of the activity. For experimental flushing data, an analysis of variance also was used to test the hypothesis that mean flushing distance did not differ among disturbance types, age of the eagle, location of the disturbance, and time of day of the disturbance. Chi-square goodness-of-fit analyses were used to compare flushing percentages among types and between zones for both individual eagle observations and experimental disturbances. The 0.05 level of significance was used in all tests.

I calculated an index of disturbance, based on both flushing distance and frequency, by multiplying the mean flushing distance by the flushing frequency for each activity type.

To develop a probability of flushing curve, the flushing data from the individual eagle observations and the experimental flushing data were combined (n=178). The percent of eagles that flushed in 7 distance categories, 0-50 m, 51-100 m, 101-150 m, 151-200 m, 201-250 m, 251-300 m, and >300 m were determined. These categories were used to correspond with categories used in 2 other studies analyzing flush distances (Stalmaster and Newman 1978, Knight and Knight 1984). The cumulative flushing percentage, proceeding from the farthest to the closest distance category, was then calculated. I also developed flushing curves from data reported in Stalmaster and Newman (1978) and Knight and Knight (1984). I used Kolomogorov-Smirnov 2 sample tests to determine if the curve developed from my study was different from the curves developed from data in Stalmaster and Newman (1978) and Knight and Knight (1984).

The flushing frequencies in 5 distance categories, 0-50 m, 51-100 m, 101-150 m, 151-200 m, and >200 m, were also compared between zones using Chi-square analysis to determine if eagles were more tolerant in the developed or undeveloped zone. Probability of flushing curves were developed for both zones and tested as described above.

RESULTS

Observed Human-eagle Encounters

During the winters 1987-89, 234 eagle-human encounters were recorded during individual eagle observations. These observations were divided into 5 categories (Table 3.1). Of these 66 (28%) encounters resulted in the eagle flushing. Walkers caused the highest percentage of flushing responses. Vehicles, then joggers caused the lowest percentage to flush. These relative frequencies were different (X2 = 37.64, d.f. = 5, P < 0.0009).

A higher percentage of eagles flushed in Zone 1 than in Zone 2 (X2 = 9.95, d.f. = 1, P = 0.002, Table 3.2).

Mean flushing distances did not differ among the 5 types of encounters (F = 0.89, P = 0.47, Table 3.3). Bicyclers and vehicles caused eagles to flush at the highest mean distances and fishermen and joggers the lowest. Because some categories had a small number of observations, I combined the walkers and joggers, and the bicyclers and vehicles categories. Mean flushing distances still did not differ among the 3 categories (F = 1.42, P = 0.25). Bicyclers and vehicles caused eagles to flush at the highest mean distance, followed by walkers and joggers, then fishermen. Mean flushing distances differed by zone, however, (t = 2.33, P = 0.02) with eagles flushing at a higher mean distance in Zone 1 than Zone 2 (Table 3.4).

Walkers had the highest disturbance index, followed by bicyclers, fishermen, then joggers and vehicles (Table 3.5).

Experimental Disturbances

Mean flushing distances did not differ among the 4 types of experimental disturbances (F = 2.18, P = 0.10, Table 3.6). Pointers caused the highest mean flushing distance, followed by fishermen, then walkers. However 1 observation for the pointer category had an extremely low flushing distance. An eagle did not flush until the disturbance was 5 m away. The next lowest value for this category was 32 m. If the extreme value was excluded from the analysis then there was a difference among the means (F = 2.81, P = 0.04). Fisher’s least significant difference test for multiple comparisons indicated that pointers and fishermen caused eagles to flush at significantly greater distances than walkers-single and walkers-group (t = 1.98, P = 0.05). The overall mean flushing distance was 86 m including the extreme value and 87 m excluding the extreme value (Table 3.6).

There were no differences among experimental disturbance flushing distance means due to the age of the eagle (adult n=95 and immature n=17), the zone in which the disturbance took place (Zone 1 n = 105 and Zone 2 n = 7), or the time of day of the disturbance (sunrise-1000 n=39, 1000-1400 n=62, and 1400-sunset n=11) with or without the extreme value (F = 1.56, P = 0.16 and F = 1.94, P = 0.07, respectively).

Flushing Frequencies

There was no difference in the relative flushing frequencies between zones among the 5 distance categories (Fig. 3.1). When distances were combined into 2 categories of >100 and 0-100, there was still no difference in flushing frequencies (Fig. 3.2). Comparison of the probability of flushing curves for each zone show that a lower proportion of eagles would be expected to flush in Zone 2 than Zone 1 (Fig. 3.3), however these curves were not statistically different (P = 0.40).

Relatively more flushes occurred at closer distances (<100 m) in Boise than in Northwest Washington (Fig. 3.4). The probability of flushing curve shows the percentage of eagles that would be expected to flush at a particular distance for this population of eagles (Fig. 3.5). A lower proportion of eagles wintering in Boise would be expected to flush at a given distance than the eagles studued in northwest Washington (Fig. 3.5). The probability of flushing curve from this study was significantly different than both curves developed from data in Washington

(P < 0.0001 for both comparisons).

DISCUSSION

Bald eagles were more likely to flush when the disturber approached slowly, and at farther distances if the disturbance focused on the eagles. Walkers, especially those stopping to look at an eagle, were the most disturbing to eagles in terms of flushing frequencies. Bicyclers and vehicles which approached and passed by quickly, and at a constant speed, were less likely to cause eagles to flush. This finding has been documented by other researchers. Eagles are more likely to tolerate activity if it is not directed toward them (Stalmaster 1976). Jacobsen (1977) reported that "people who noticed eagles and made direct eye contact made the birds more nervous then if they were ignored." McGarigal (1988) reported that eagles flushed more than expected from boats that approached slowly than from boats that approached rapidly. Fraser (1981) also noted that slow pedestrian approaches to a nest were more disturbing because they gave the eagle more time to "decide" to fly.

The disturbance indexes, which reflect both flushing distance and frequency, indicated that walkers were the most disturbing to eagles. Bicyclers, followed closely by fishermen, were the next most disturbing.

Eagles flushed from human activity at a higher frequency and at a higher mean flushing distance in Zone 1 than in Zone 2. These results indicate that eagles may have habituated to human activity in the more developed stretch of the river. Several studies have shown that eagles habituate to people both spatially and temporally. Stalmaster and Newman (1978) reported that tolerance to human activity was related to the location of the disturbance. Eagles wintering on a river with little human activity were more sensitive to the human activities than eagles wintering on a river with higher levels of human activity (Russell 1980). Knight and Knight (1984) reported that eagles were less likely to flush from activity later in the winter suggesting that they became used to people as time passed.

The results of this study suggest that eagles along the Boise River are spatially habituated to human activity. There was no difference in flushing percentages among different distances or between the flushing curves, however, both mean flushing distance and flushing frequency were higher in Zone 1 than Zone 2. No eagles were marked so individuals could not be identified, however at least a few eagles used both zones because eagles were observed going from zone to zone during 5 different individual eagle observation days. This suggests that eagles were more tolerant of people where development was higher within the study area. However, eagles did not habituate to types of disturbance. Walkers caused the highest percentage of eagles to flush, yet it was the most common activity type that eagles encountered.

Eagles may be more tolerant of human activity at preferred foraging areas (Steenhof 1976). The apparent increased tolerance of humans in Zone 2 may have been related to more abundant prey in Zone 2.

The overall mean flushing distance for the individual eagle observations and for the experimental disturbances was lower than has been reported in other studies measuring flushing distances of bald eagles to human activity (Table 3.7). However, because individual eagles were not marked, it is not possible to determine how long a particular eagle stayed in the area. Some eagles may have been driven away by the high level of human activity.

Individual bald eagles vary widely in their response to human activity depending on the type, frequency, and duration of the activity, and the birds past experiences. However, the Boise eagle population in general flushed at lower proportions within a given distance then eagles studied in northwest Washington (Fig. 3.5).

MANAGEMENT IMPLICATIONS

In comparing the results of this study to other studies of bald eagle response to human activity, it appears that the Boise eagle population is more tolerant of human activity. While this is an indication that bald eagles may be able to adapt, to a certain degree, to humans, protection for this population is still necessary. Bald eagles use an energy conserving strategy for survival during winter (Stalmaster and Gessaman 1984). Increasing their activity through avoidance flights from humans could increase energy stress (Stalmaster 1983). Minimizing disturbance may be important to survival of some eagles, especially energy stressed eagles.

Stalmaster and Newman (1978) documented that eagles were more tolerant of human disturbance when that disturbance was obscured, at least partially, from their view. Others have reported this behavior as well (Krauss 1977, Fraser et al. 1979, Reed 1979). Vegetative buffer zones consisting of thick understory vegetation established below perch trees and an increased number of perch trees may further decrease flush distances and rates.

Because eagles were less tolerant of activities that approached slowly, people coming in contact with wintering eagles should be encouraged to approach an eagle at a steady, relatively fast pace. Bicyclers caused eagles to flush at greater distances and walkers caused eagles to flush at higher frequencies. Both of these disturbance types should be buffered from eagles either by distance or vegetation.

The flushing curve developed from this population should be considered specific to this population or a similar population of eagles wintering in urban, riverine conditions. This curve could be used by city planners and developers when deciding on distance buffer zones that are desirable to protect eagles from human activity. Stalmaster (1980) suggested buffer zones wide enough to protect 96-99% of an eagle population in Northwest Washington. A buffer zone of 200 m would protect 96% of the eagles along the Boise River. Planners should decide what proportion of a population of eagles they are willing to disturb, and plan buffers accordingly.

Table 3.1. Numbers of encounters between bald eagles and 5 types of human activity and the percent flushing responses recorded along the Boise River during 2 winters 1987-89.

Activity n Percent flushing (n)

Walker 99 46 (46)

Fisherman 35 34 (12)

Bicycler 13 15 (2)

Jogger 8 13 (1)

Vehicle 79 6 (5)

234 28 (66)

Table 3.2. Percent of bald eagles flushing from all types of human activity recorded during individual eagle observations along the Boise River, by zone, winters 1987-89.

Zone n Percent flushing (n)

1 107 38 (41)

2 127 20 (25)

234 28 (66)

Table 3.3. Means and ranges of bald eagle flushing distances to 5 types of human activity recorded during individual eagle observations during the winters 1987-89 along the Boise River.

Type of activity n Mean (m) Range (m)

Bicycler 2 148 96 - 200

Vehicle 5 107 5 - 250

Walker 46 87 17 - 300

Fisherman 12 64 20 - 90

Jogger 1 50 50 - 50

Total 66 86 5 - 300

Zone n Mean (m) Range (m)

1 41 99* 20 - 300

2 25 64 5 - 200

66 86 5 - 300

*significantly higher than Zone 2 at P = 0.02

Table 3.5 Disturbance indexes developed for bald eagles wintering along the Boise River.

Flushing Mean flushing Index

Activity Frequency distance (m) (Frequency*Mean)

Walker 0.46 87 40.0

Bicycler 0.15 148 22.2

Fisherman 0.34 64 21.8

Jogger 0.13 50 6.5

Vehicle 0.06 107 6.4

Type of activity n Mean (m) Range (m)

Pointer 27 106 5 - 350

(26) (110)* (32 - 350)

Fisherman 11 92* 30 - 170

Walker - single 47 78 10 - 160

Walker - group 27 77 25 - 215

Total 122 86 5 - 350

(121) (87) (10 - 350)

*significantly higher at P = 0.04

Table 3.7. Comparison of mean flushing distances of bald eagles to human activity among 8 studies.

Mean

Study Flushing flushing

location Season rate (%) distance (m) Source

Bear Valley, Winter not available ~ 183 Krauss (1977)

Oregon

Nooksack River, Winter 0- 50 m = 17 131 Stalmaster and

Washington 51-100 m = 30 Newman (1978)

101-150 m = 27

151-200 m = 19

201-250 m = 6

251-300 m = 1

>300 m = 0

Shasta-Trinity Breeding not available ~ 227 Detrich (1980)

National Forest,

California

Eldorado National Winter not available ~ 130 Schonholtz (1980)

Forest, California

Skagit and Winter 0- 50 m = 11 150 and Knight and

Nooksack Rivers, 51-100 m = 31 168 Knight (1984)

Washington 101-150 m = 21 151-200 m = 17

201-250 m = 13

251-300 m = 5

>300 m = 2

Table 3.7 continued.

Mean

Study Flushing flushing

location Season rate (%) distance (m) Source

Chippewa National Breeding <200 m = 91 476 Fraser et al.

Forest, Minnesota >200 m = 9 (1985)

Lower Columbia Breeding 0-100 m = 21 197 McGarigal (1988)

River, Washington 101-200 m = 12

and Oregon >200 m = 9

Boise River, Winter 0- 50 m = 30 86 This study

Idaho 51-100 m = 43

101-150 m = 17

151-200 m = 6

201-250 m = 2

251-300 m = 1

>300 m = 1

Figure 3.1. Percentage of wintering bald eagles that flushed at 5 distance categories along 2 stretches of the Boise River, 1987-89.

Figure 3.2. Percentage of wintering bald eagles that flushed at 2 distance categories along 2 stretches of the Boise River, 1987-89.

Figure 3.4. Percentage of eagles that flushed at 7 distance categories for 3 populations of wintering bald eagles, 1987-89.

CHAPTER 4 - RIVER FLOW TESTS

INTRODUCTION

During the first field season, eagle location river measurements and random location river measurements were not always measured during the same flow rates. To test whether or not the difference in flow rates during the winter caused a difference in river variables, I measured the variables at different flows throughout the second field season.

METHODS

On 4 November 1988, after the irrigation season was over and the river was at low flow, I established sample transects at 9 locations in the river, 3 at riffles, 3 at runs, and 3 at pools. These transects were randomly selected and measured as the flows changed throughout the winter. Information on river flows was available weekly from the Bureau of Reclamation (See Appendix 2). The transects were measured weekly at different flow rates. This information was used to determine how flow affected the river measurements.

RESULTS

The 9 test transects were measured a total of 17 times between 4 November 1988 and 15 March 1989. Regression analyses were conducted 2 separate times using the flows at the Diversion Dam and the flows at Glenwood St.

The flows at Diversion Dam fluctuated from 139 to 157 c.f.s. The width of the river at pools varied significantly with the Diversion Dam flow, and the turbidity of all habitat types varied significantly with flow. The flows at Glenwood St. fluctuated from 160 to 246 c.f.s. The velocities of pools, and the temperature and turbidity of all habitat types varied significantly with the Glenwood St. flow (Table 4.1). The temperature was probably not affected by flow but by fluctuating air temperatures throughout the winter. The results of this analysis showed that water turbidity in all habitat types, and river widths and water velocities in pools were affected by the changing river flows.

Because flow rates affected the river variables, river measurement data from the first field season were not used in the analyses.

Table 4.1. R2 values for the variables measured at the test transects for flows at Diversion Dam and Glenwood St. along the Boise River during the winter 1988-89.

Location

Habitat of flow

type reading Width Temp. Depth Vel. Turb.

Pool Diversion 0.0700* 0.0004 0.0011 0.0174 0.0462*

Glenwood 0.0027 0.2777* 0.0024 0.0383* 0.4431*

Riffle Diversion 0.0033 0.0170 0.0000 0.0006 0.0570*

Glenwood 0.0005 0.2095* 0.0229 0.0001 0.4672*

Run Diversion 0.0182 0.0203 0.0002 0.0174 0.0647*

Glenwood 0.0022 0.2344* 0.0053 0.0019 0.4438*

*denotes a P value less than 0.05 for the hypothesis that the regression coefficient for the variable is equal to zero

CHAPTER 5 - EFFECTS OF BARBER DAM CONSTRUCTION ACTIVITIES ON WINTERING BALD EAGLES

INTRODUCTION

Bald eagles use the area around Barber Dam because of its proximity (1.4 km) to a night roost and to a foraging area (1.2 km) in Barber Park. This study was undertaken to determine if construction at the dam, beginning 1 March 1988, would affect eagle use of the area and the roost.

METHODS

The Barber Dam area was intensively observed for bald eagles prior to and during construction. The observation area covered approximately 325 hectares and included Barber Dam, the Barber pool area, and the east end of Barber Park (Fig. 5.1). Observations began 18 February and ended 14 March and were conducted 2 days during the week (Mondays and Thursdays) and 1 day on the weekend (Sundays). Observations were made from a point on the south side of the river directly across from the dam.

To record daily activity patterns of bald eagles within this area, it was observed from sunrise to sunset. Due to the proximity of the construction site to a bald eagle night roost, the roost was observed during the same days as the dam area. The roost was observed prior to the dam area from a different observation point (Fig. 5.1), which was approximately 850 m from the roost, 0.5 h before sunrise until sunrise, and again after completion of observations of the dam area, from sunset to 0.5 h after sunset. During this time the number of adult and immature eagles in the roost were counted.

All eagles observed within the dam area were recorded. Age, location, distance of the location to the construction site, time spent in the area, and activity of the eagles were recorded. Human activity was quantified by recording the numbers of walkers, fishermen, and construction workers during the day, and the time each spent in the area. The construction activity that occurred in the area was described. All eagle responses to human activity were quantified by recording flushing distances, by estimation, to the activities that came within 300 m of the eagle. If no response was observed, the closest distance of the eagle to the activity was estimated up to 300 m. Temperature and wind velocity were measured at the beginning of the day and at 4 hour intervals during the day.

During this same time period, 18 February - 13 March, vehicle censuses were conducted twice a week along 24.5 km of the river from Lucky Peak Dam to the Glenwood Street Bridge and included the Barber Dam observation area (Appendix 4). These censuses were run every Wednesday and Saturday between 0730 and 1130 hours. A Spearman’s rank correlation analysis was run to see if the numbers of eagles observed in the Barber Dam study area correlated with the numbers of eagles observed in the rest of the 24.5-km study area.

Numbers of eagles and the time they spent in the area before and during the construction period were compared using a nonparametric Mann-Whitney test. The numbers of observations of perched eagles and the time they spent perching at all locations within 500 m of the construction site were compared before and during construction using a Mann-Whitney test. Because construction only occurred on weekdays, eagle use of the area during the weekdays was compared to use during the weekend with the Mann-Whitney test. Numbers of eagles using the roost before and during construction were compared using a Mann-Whitney test. All significance levels were set at 0.05.

RESULTS

I observed the dam area for 6 days during the preconstruction period and 6 days during construction. A total of 135.5 hours was spent observing on 12 days, 67.5 hours before construction and 68 hours during construction. The amount of time spent by eagles in the area was not different between the 2 periods (U = 27, P > 0.05). The number of eagles observed in the area was higher during the preconstruction period (U = 31.5, P < 0.05) (Tables 5.1 and 5.2). However, the total number of eagles along the entire 24.5-km study area declined from preconstruction to the construction period. The number of eagles observed in the Barber Dam study area (Fig. 5.1) was positively correlated (r = 0.73) with the number of eagles observed in the rest of the 24.5-km study area and with the total number of eagles along the entire 24.5-km census route (r = 0.89). Therefore, if the numbers of eagles in the Barber Dam area are adjusted as a proportion of the census total for the census just prior to the observation day, then the average number of eagles seen is not significantly different before construction than during construction (U = 21.5, P > 0.05) (Table 5.3).

The number of observations of eagles and the time they spent at locations within 500 m of the construction site were not different before and during construction (U = 14, P > 0.05 and

U = 14, P > 0.05, respectively) (Fig. 5.1 and Table 5.4). Weekend and weekday use of the Barber Dam area by eagles was not different either in numbers of eagles observed, unadjusted, or in amount of time they spent in the area (U = 19, P > 0.05 and U = 17, P > 0.05, respectively).

The maximum number of eagles using the roost before and during construction was 2. Numbers of eagles using the roost before and during construction were not different (U = 81, P > 0.05) (Table 5.5).

All encounters with human activity caused no response

(Table 5.6). The closest distance from an eagle to an activity ranged from 50 to 300 m.

On 3 March a forklift was operated for 90 min. from 1330 - 1500 unloading concrete blocks from 2 flatbed trailers. On 7 March the same activity occurred for 80 min. from 1450 - 1610. On 10 March, a person operated a backhoe for 305 min. He was digging near the dam off and on between 0840 and 1745. A bulldozer was moving dirt for 375 min. between 1110 and 1820. A frontloader was moving concrete blocks for 95 min. from

0840 - 1015. On 14 March a backhoe was digging for 165 min. off and on between 0755 and 1730. A bulldozer was operated for 120 min. from 1230 - 1430.

CONCLUSIONS

Construction at Barber Dam had no noticeable effect on bald eagle use of the area. When eagle numbers were adjusted for the overall census decline there was no difference in eagle numbers before and during construction in the Barber Dam area. The amount of time eagles spent in the area did not differ nor did the eagles use of perch areas within 500 m of the construction site. There was no difference in eagle use of the area between the weekdays and weekends. Roost use by eagles did not differ before and during construction.

No encounters with human activity resulted in eagles flushing. I think this is partly because most of the activity did not occur close to eagles. The closest encounter was 50 m and that only happened once. Eagles did not perch close to the dam often, even before construction began.

Table 5.1. Numbers of eagles and amount of time spent by perched eagles in the Barber Dam area of the Boise River prior to construction recorded 1988.

Number Time (min.)

Date Adults Immatures Adults Immatures

18 Feb. 4 0 755 0

21 Feb. (weekend) 3 2 591 242

22 Feb. 4 0 182 0

25 Feb. 2 2 565 438

28 Feb. (weekend) 5 1 1660 80

29 Feb. 2 0 375 0

Totals 20 5 4128 760

Table 5.2. Numbers of eagles and amount of time spent by perched eagles in the Barber Dam area of the Boise River during construction in 1988.

Number Time (min.) Date Adults Immatures Adults Immatures

3 March 1 1 450 10

6 March (weekend) 1 0 280 0

7 March 2 1 200 127

10 March 1 0 30 0

13 March (weekend) 2 0 60 0

14 March 3 1 1337 10

Totals 10 3 2357 147

Table 5.3. Numbers of bald eagles observed in the Barber Dam area of the Boise River expressed as a proportion of the total counted in the preceding census in 1988.

Biweekly

Barber Dam Census Proportion of

Observation Date Census Date Total Census Total

Preconstruction

18 Feb. 17 Feb. 8 0.50

21 Feb. 20 Feb. 6 0.83

22 Feb. 20 Feb. 6 0.67

25 Feb. 24 Feb. 6 0.67

28 Feb. 27 Feb. 2 3.00

29 Feb. 27 Feb. 2 1.00

Mean + SD 1.11 + 0.9

During construction

3 March 2 March 1 2.00

6 March 5 March 4 0.25

7 March 5 March 4 0.75

10 March 9 March 1 1.00

13 March 12 March 3 0.67

14 March 12 March 3 1.33

Mean + SD 1.00 + 0.6

Table 5.4. Number of observations of bald eagles and time they spent at locations in the Barber Dam area of the Boise River before and during construction in 1988.

Location Distance (m) No. of observations Time (min.) (see Fig. 5.1) from dam before during before during

1 1400 9 7 20 105

2 1100 5 6 524 1642

3 840 13 3 1347 130

4 600 3 0 230 0

5 360 1 0 25 0

6 360 21 7 1940 335

7 140 2 0 265 0

8 0 0 1 0 12

9 360 0 1 0 60

10 840 3 0 132 0

11 880 1 0 75 0

12 1200 9 8 320 220

Totals 67 33 4878 2504

Table 5.5. Numbers of bald eagles using a communal roost near the Boise River prior to and during construction at Barber Dam for 12 observation periods in 1988.

Morning Evening

Date adult immature adult immature

Before construction

18 Feb. 0 0 0 0

21 Feb. 2 0 0 0

22 Feb. 0 0 1 0

25 Feb. 0 1 1 1

28 Feb. 1 0 0 0

29 Feb. 0 0 2 0

Totals 3 1 4 1

During construction

3 Mar. 0 0 0 0

6 Mar. 1 0 0 0

7 Mar. 0 0 0 1

10 Mar. 0 0 0 0

13 Mar. 0 0 2 0

14 Mar. 2 0 0 0

Totals 3 0 2 1

Table 5.6. Ranges of distances to 4 types of human activity to which eagles did not respond at the Barber Dam area of the Boise River before and during construction in 1988.

Activity n Range (m)

Before construction

walker

adult eagle 11 100 - 200

vehicle

adult eagle 283 100 - 200

fisherman

adult eagle 1 100

cycler

adult eagle 1 100

Total 296

During construction

vehicle

adult eagle 158 100 - 300

imm. eagle 2 50 - 100

Total 160

Figure 5.1. Barber Dam observation area and eagle perch locations along the Boise River recorded in February and March 1988.

LITERATURE CITED

Allen, K. R. 1969. Limitations on production in salmonid populations in streams. Pages 3-18 in T. G. Northgate, ed. Symp. on salmon and trout in streams. Univ. of British Columbia, Vancouver, Can.

Anderson, B., J. Frost, K. McAllister, D. Pinea, and P. Crocker-Davis. 1988. Bald eagles in Washington. Unpubl. Rep. Washington Dep. of Game, Olympia. 8pp.

Asbridge, G. M. 1988. The effect of discharge on type and physical characteristics of habitat in the Boise River, Idaho, and an evaluation of the suitability of habitat in the Boise River for salmonids. M.S. Thesis, Univ. of Idaho, Moscow. 70pp.

Biosystems Analysis, Inc. 1985. Pit 3, 4, and 5 project bald eagle and fish study. Pacific and Electric Company. San Ramon, Calif.

Conrad, E. T., and J. E. Stillwell. 1984. Wintering bald eagles at Millerton Lake, 1983-84. Unpubl. Rep. Millerton State Recreation Area, Calif. Dept. of Parks and Recreation. 76pp.

Cunjak, R. A., and G. Power. 1987. Cover use by stream-resident trout in winter: a field experiment. North Am. J. of Fisheries Manage. 7:539-544.

Detrich, P. J. 1980. Pit 3, 4, and 5 bald eagle study. Unpubl. Rep. U. S. For. Serv., Shasta-Trinity Natl. For., Redding, Calif. 22pp.

Edwards, C. C. 1969. Winter behavior and population dynamics of American eagles in western Utah. Ph.D Thesis, Brigham Young Univ., Provo, Utah. 150pp.

Fraser, J. D. 1981. Breeding biology and status of the bald eagle on the Chippewa National Forest. Ph.D Thesis, Univ. of Minnesota, St. Paul 236pp.

, L. D. Frenzel, J. E. Mathisen, and N. S. Fraser. 1979. Experimental disturbance of nesting bald eagles. Presentation at Raptor Res. Found. Annu. Meet. Davis, Calif. 10pp.

, , and . 1985. The impact of human activities on breeding bald eagles in North-central Minnesota. J. Wildl. Manage. 49:585-592.

Gerrard, J. M. 1983. A review of the current status of bald eagles in North America. Pages 5-21 in D. M. Bird, chief ed. Biology and Management of Bald Eagles and Ospreys. Harpell Press, Ste. Anne de Bellevue, Que.

Griffin, C. R., T. S. Baskett, and R. D. Sparrowe. 1982. Ecology of bald eagles wintering near a waterfowl concentration. U. S. Fish and Wildl. Serv. Spec. Sci. Rep. Wildl. 247 12pp.

Hansen, A. J., M. V. Stalmaster, and J. R. Newman. 1980. Habitat characteristics, function, and destruction of bald eagle communal roosts in western Washington. Pages 221-229 in R. L. Knight, G. T. Allen, M. V. Stalmaster, and C. W. Servheen, eds. Proc. of the Wash. Bald Eagle Symp. The Nat. Conserv., Seattle, Wash.

Harmata, A. R. 1984. Bald eagles of the San Luis Valley, Colorado: their winter ecology and spring migration. Ph.D Thesis, Montana State Univ., Bozeman. 222pp.

Haywood, D. D., and R. D. Ohmart. 1986. Utilization of benthic-feeding fish by inland breeding bald eagles. Condor 88:35-42.

Holubetz, T. 1988. Fish population assessment on Boise River (Boise to Star). Idaho Dep. of Fish and Game contract report, Boise, Idaho.

Jacobsen, S. 1977. McCumber Lake bald eagle study. Unpubl. Rep. Calif. Dep. of Fish and Game, Redding. 52pp.

Jensen, S. W. 1981. Wintering bald eagles of the lower Boise River. Unpubl. manuscript. U. S. D. I., Fish and Wildl. Serv., Division of Endangered Species. Boise, Id.

Keister, G. P., and R. G. Anthony. 1983. Characteristics of bald eagle communal roosts in the Klamath Basin, Oregon and California. J. Wildl. Manage. 47:1072-1079.

Knight, R. L., and S. K. Knight. 1984. Responses of wintering bald eagles to boating activity. J. Wildl. Manage. 48:999-1004.

Krauss, G. D. 1977. Report on the 1976-77 Klamath Basin bald eagle winter use area investigation. Unpubl. Rep. Klamath Natl. For., Goosenest Ranger Dist., Mt. Hebron, Calif. 68pp.

Lish, J. W., and J. C. Lewis. 1975. Status and ecology of bald eagles wintering in Oklahoma. Proc. Southeast Assoc. Game and Fish Comm. 29:415-423.

McGarigal, K. 1988. Human-eagle interactions on the lower Columbia River. M.S. Thesis, Oregon State Univ., Corvallis. 108pp.

Neu, C. W., C. R. Byers, and J. M. Peek. 1974. A technique for analysis of utilization-availability data. J. Wildl. Manage. 38:541-545.

Platt, J. B. 1976. Bald eagles wintering in a Utah desert. Am. Birds 30:783-788.

Reed, S. B. 1979. Bald eagle winter management plan for the Lake Tahoe Basin Management Unit. U. S. For. Serv., South Lake Tahoe, Calif. 24pp.

Reynolds, J. D., N. D. Ertter, D. K. Broemeling, and R. P. Howard. 1985. Winter distribution of bald eagles along a segment of the Boise River. Northwest Sci. 59:93-96.

Rimmer, D. M., R. L. Saunders, and U. Paim. 1985. Effects of temperature and season on the position holding performance of juvenile Atlantic salmon. Can. J. Zool. 63:82-96.

Russell, D. 1980. Occurrence and human disturbance sensitivity of wintering bald eagles in the Sauk and Suiattle Rivers, Washington. Pages 165-174 in R. L. Knight, G. T. Allen, M. V. Stalmaster, and C. W. Servheen, eds. Proc. of the Wash. Bald Eagle Symp. The Nat. Conserv., Seattle, Wash.

Schonholtz, R. 1980. Bald eagles on the Eldorado National Forest. Unpubl. Rep. U. S. For. Serv., Placerville, Calif. 36pp.

Schultz, L. B. 1982. Wintering bald eagles in the Big Bear Valley annual report 1981-82. Unpubl. Rep. U. S. For. Serv. San Bernardino Natl. For., Big Bear Ranger Dist. Fawnskin, Calif. 25pp.

Skagen, S. K. 1980. Behavioral responses of wintering bald eagles to human activity on the Skagit River, Washington. Pages 231-242 in R. L. Knight, G. T. Allen, M. V. Stalmaster, and C. W. Servheen, eds. Proc. Wash. Bald Eagle Symp. The Nat. Conserv., Seattle, Wash.

Stalmaster, M. V. 1976. Winter ecology and effects of human activity on bald eagles in the Nooksack River Valley, Washington. M.S. Thesis, Western Washington Univ., Bellingham. 100pp.

. 1980. Management strategies for wintering bald eagles in the Pacific Northwest. Pages 49-67 in R. L. Knight, G. T. Allen, M. V. Stalmaster, and C. W. Servheen, eds. Proc. Wash. Bald Eagle Symp. The Nat. Conserv., Seattle, Wash.

. 1983. An energetics simulation model for managing wintering bald eagles. J. Wildl. Manage. 47:349-359.

, and J. A. Gessaman. 1984. Ecological energetics and foraging behavior of overwintering bald eagles. Ecol. Monographs 54:407-428.

, and J. R. Newman. 1978. Behavioral responses of wintering bald eagles to human activity. J. Wildl. Manage. 42:506-513.

Steenhof, K. 1976. Ecology of wintering bald eagles in southwestern South Dakota. M.S. Thesis. Univ. of Missouri, Columbia. 146pp.

. 1978. Management of wintering bald eagles. U. S. D. I. Fish and Wildl. Serv., Washington D. C. FWS/OBS - 78/79. 59pp.

, S. S. Berlinger, and L. H. Frederickson. 1980. Habitat use by wintering bald eagles in South Dakota. J. Wildl. Manage. 44:798-805.

Swenson, J. E. 1983. Is the northern interior bald eagle population in North America increasing? Pages 23-34 in D. M. Bird chief ed. Biology and Management of Bald Eagles and Ospreys. Harpell Press, Ste. Anne de Bellevue, Que.

Walter, H., and K.L. Garrett. 1981. Effects of human activity on wintering bald eagles in the Big Bear Valley, California. Unpubl. Rep. Univ. of California, Los Angeles. 84pp.

APPENDIXES

Appendix 1. River flows in c.f.s. in the Boise River below Diversion Dam for November 1987 through March 1988 from the Bureau of Reclamation.

Day November December January February March

1 131 144 129 144 125

2 133 159 129 124 128

3 132 159 131 137 136

4 132 156 141 124 147

5 132 158 147 125 146

6 132 159 149 124 149

7 132 159 147 124 152

8 132 148 147 124 155

9 135 147 148 124 151

10 135 142 148 124 151

11 135 139 150 124 147

12 135 139 150 124 147

13 136 139 150 123 148

14 135 138 149 122 147

15 135 138 146 124 147

16 134 138 146 125 148

17 126 139 147 125 148

18 120 139 142 132 146

19 127 138 137 131 136

20 132 137 132 130 136

21 132 139 165 129 147

22 132 138 160 129 159

23 130 139 155 130 159

24 133 138 159 132 160

25 136 139 178 132 160

26 135 139 166 131 160

27 135 139 156 130 160

28 132 139 159 130 165

29 132 139 160 129 167

30 130 133 161 161

31 128 161 152

Appendix 2. River flows in c.f.s. in the Boise River below Diversion Dam (DD) and the Glenwood Bridge (GB) for November 1988 through March 1989 from the Bureau of Reclamation.

Day November December January February March

DD GB DD GB DD GB DD GB DD GB

1 139 165 149 179 144 163 152 166 157 225 2 140 182 149 173 147 168 151 165 157 286

3 140 190 149 172 149 170 151 165 146 222

4 139 173 150 170 151 172 239 160 152 217

5 139 170 151 179 151 174 149 176 161 227

6 137 176 151 173 151 175 176 183 169 259

7 138 168 151 171 152 169 151 194 157 251

8 138 178 151 172 160 173 151 228 150 246

9 136 169 151 170 148 181 144 215 152 276

10 137 179 151 170 147 209 149 167 154 287

11 136 171 152 168 146 189 149 164 164 290

12 137 181 154 170 146 182 151 166 169 288

13 136 180 155 172 147 181 148 166 161 288

14 136 178 148 172 149 180 146 164 164 267

15 136 171 152 160 149 179 146 164 152 243

16 136 171 162 179 149 182 146 166 152 221

17 139 196 159 180 149 183 148 218 158 223

18 146 187 156 175 159 184 149 197 165 227

19 153 183 152 174 154 184 149 238 169 278

20 160 180 147 163 152 173 149 210 164 240

21 159 187 149 173 153 173 154 201 168 259 22 159 212 149 170 154 178 151 223 164 254

23 159 202 149 168 154 173 152 257 160 249

24 157 194 150 166 153 166 146 250 160 247

25 157 190 151 164 154 165 151 232 162 309

26 159 189 151 163 157 167 153 239 160 285

27 158 184 149 160 154 170 160 227 161 282

28 159 208 147 160 154 169 157 242 150 292

29 158 189 189 153 155 169 145 280

30 153 185 145 176 157 174 148 268

31 144 165 155 172 152 322

Appendix 3. Vehicle census route driven when counting bald eagles along the Boise River during the winters 1987-89.

Appendix 4. Numbers of bald eagles counted along the Boise River during the winter 1987-88.

Appendix 5. Distribution of bald eagles along the Boise River during winter censuses 1987-89.

No. of total Total No. of perch Perch No. of perch

observations obs./km observations obs./km locations

Zone Km 87-88 88-89 87-88 88-89 87-88 88-89 87-88 88-89 87-88 88-89

1 10.0 204 337 20 34 160 244 16 24 25 43

2 14.5 50 48 3 3 43 37 3 3 18 17

24.5 254 385 203 281 43 60

Appendix 6. Numbers of bald eagles counted along the Boise River during the winter 1988-89.

Appendix 7. Amount of time bald eagles were observed in each zone during individual eagle observations along the Boise River, winter 1987-88.

Number of Number of

Days eagles minutes

Location Ad. Imm. Ad. Imm. Ad. Imm.

Zone 1 10 2 11 2 1115 190

Zone 2 5 3 6 3 565 205

15 5 17 5 1680 395

Appendix 8. Amount of time bald eagles and human activity were observed in each zone during individual eagle observations along the Boise River, winter 1988-89.

Number of Number of Number of

Days eagles minutes human minutes

Location Ad. Imm. Ad. Imm. Ad. Imm. Ad. Imm.

Zone 1 22 19 23 20 4441 1895 300 34

Zone 2 14 1 15 1 1742 165 40 1

Totals 36 20 38 21 6183 2060 340 35

Appendix 9. River characteristics of bald eagle prey capture and attempted prey capture points recorded along the Boise River during the winter 1987-88.

River

Habitat

Date Time Location Success Type Temperature Depth Velocity Turbidity

7 Nov. 1005 Barber Pool yes pool NA NA NA NA

26 Dec. 1005 47th St. yes pool NA NA NA NA

28 Jan. 0940 Barber Park no riffle 3.6 0C 27 cm 10 cm/s 30 FTU

3 Feb. 1515 Barber Park yes riffle 3.6 0C 24 cm 10 cm/s 30 FTU

10 Feb. 0945 Barber Park no riffle 3.6 0C 24 cm 10 cm/s 25 FTU

NA - not available

Appendix 10. Observations of foraging by bald eagles along the Boise River during the winter 1987-88.

Date Age Location Time Prey type

7 Nov. 1 adult Barber Pool 1005 Unidentified fish*

26 Dec. 1 adult 47th St. 1005 Unidentified fish*

13 Jan. 2 adults 1800 m above

1 immature Diversion Dam 0900 Unidentified fish

28 Jan. 1 adult Barber Park 1000 Rainbow trout

28 Jan. 1 adult Barber Park 1015 Hatchery rainbow trout

30 Jan. 1 immature Diversion Dam 0900 Unidentified fish

3 Feb. 1 adult Barber Park 1105 Unidentified fish*

3 Feb. 1 immature Barber Pool 1625 Unidentified small mammal

prey captures were observed

Appendix 11. River characteristics of bald eagle prey capture and attempted prey capture points recorded along the Boise River during the winter 1988-89.

River

habitat

Date Time Location Success type Temperature Depth Velocity Turbidity

12 Dec. 1027 35th St. no pool 5 0C 40 cm 35 cm/s 95 FTU

12 Dec. 1258 35th St. yes pool 5 0C 40 cm 35 cm/s 95 FTU

16 Jan. 0920 Barber Park no riffle 2 0C 20 cm 30 cm/s 5 FTU

17 Feb. 1310 just west of yes riffle 2 0C 30 cm 65 cm/s 6 FTU

Barber Park

Appendix 12. Observations of foraging by bald eagles along the Boise River during the winter 1988-89.

Date Age Location Time Prey Type

12 Dec. 1 adult 35th St. 1258-1259 Trout*

20 Jan. 1 adult foothills north of 5 immatures Barber Pool 1110-1144 Deer carcass

23 Jan. 5 immatures canyon rim southeast of Barber Pool 1045-1210 Deer carcass

1 Feb. 1 adult Barber Park 0845-0900 Rainbow trout

13 Feb. 1 immature foothills north of Barber Pool 1505-1530 Deer carcass

17 Feb. 1 adult west of Barber

Park diversion 1310-1330 Trout*

10 Mar. 1 adult below Lucky Peak Dam 0720-0735 Unidentified

fish

*prey captures were observed

Appendix 13. Number of observations of prey and their frequency of occurrence in bald eagle pellets collected in a communal roost along the Boise River during the winter 1988-89. Numbers of pellets collected are in parentheses.

Frequency of

Jan. Feb. Mar. Apr. Total occurrence

Prey type (12) (110) (57) (7) (186) (Total/186)

Mammals:

Mule deer

(Odocoileus hemionus) 10 89 49 6 154 82.8

Townsend’s ground squirrel

(Spermophilus townsendii) 2 20 6 1 29 15.6

Nuttall’s cottontail rabbit

(Sylvilagus nuttallii) 0 3 2 0 5 2.7

Unidentified rabbit 0 4 2 0 6 3.2

Unidentified rodent 0 2 0 0 2 1.1

Birds:

Mallard

(Anas platyrhynchos) 0 1 0 0 1 0.5

Unidentified duck 0 1 1 0 2 1.1

Unidentified bird 0 0 1 0 1 0.5

Fish:

Unidentified fish 0 2 7 1 10 5.4

Appendix 14. Number of waterfowl counted by zone on the Boise River between Diversion Dam and Glenwood St. during the winter 1988-89.

Zone 1 Zone 2 Total

Date n n/km n n/km n n/km

4 Dec. 68 6.8 578 39.9 646 26.4

11 Dec. 32 3.2 413 28.5 445 18.2

18 Dec. 30 3.0 456 31.4 486 19.8

8 Jan. 8 0.8 404 27.9 412 16.8

22 Jan. 29 2.9 364 25.1 394 16.1

5 Feb. 26 2.6 365 25.2 391 15.9

19 Feb. 57 5.7 366 25.2 423 17.3

Total 250 25.0 2946 203.2 3197 130.5

Appendix 15. Number and species of ducks counted on the Boise River between Diversion Dam and Glenwood St. during the winter 1988-89.

Green- Pied-

Common American Common Wood winged billed

Date Mallards merganser Widgeon goldeneye duck teal Gadwall Coot grebe

12 Dec. 438 21 125 55 0 0 0 6 1

12 Dec. 307 47 27 24 10 0 10 7 0

12 Dec. 387 41 25 25 0 0 0 8 0

8 Jan. 298 62 9 27 5 1 0 0 0

22 Jan. 308 33 20 18 0 6 8 0 0

5 Feb. 307 37 27 15 0 5 0 0 0

19 Feb. 317 43 30 27 6 0 0 0 0

Help | Site Map | Copyright

Factors Affecting The Distribution Of Bald Eagles And Effects Of Human Activity On Bald Eagles Wintering Along The Boise River

AUTOBIOGRAPHICAL SKETCH

ACKNOWLEDGMENTS

TABLE OF CONTENTS

LIST OF TABLES

LIST OF FIGURES

LIST OF APPENDIXES

ABSTRACT

CHAPTER 1 - OVERVIEW

CHAPTER 2 - FACTORS AFFECTING THE DISTRIBUTION OF WINTERING BALD EAGLES IN AN URBAN AREA

CHAPTER 3 - EFFECTS OF HUMAN ACTIVITY ON WINTERING BALD EAGLES IN AN URBAN AREA

CHAPTER 4 - RIVER FLOW TESTS

CHAPTER 5 - EFFECTS OF BARBER DAM CONSTRUCTION ACTIVITIES ON WINTERING BALD EAGLES

LITERATURE CITED

APPENDIXES