Loading...
HomeMy WebLinkAboutAPA2271BEHAVIORAL RESPONSES OF WINTERING BALD EAGLES TO HUMAN ACTIVITY MARK V. STALMASTER, Huxley College of Environmental Studies, Western Washington University, Bellingham 982251 JAMES R. NEWMAN, Huxley College of Environmental Studies, Western Washington University, Bellingham 982252 Abstract: The effects of human activity on wintering bald eagles (Haliaeetus leucocephalus) were studied in Washington for 2 yr. Tolerance to disturbance was determined by analyzing eagle distribution in relation to human activity and by measuring flight distances of eagles from simulated human disturbances. Human activity was found to affect eagle distribution and behavior adversely. Distribution patterns were signif- icantly (P < 0.001) changed, resulting in displacement of eagles to areas of lower human activity. Older birds were more sensitive (P < 0.001) to disturbances. Flight distances were highest for simulated dis- turbances in water and on gravel bars, intermediate on land, and shortest under a vegetation canopy. Habituation to normal activities seemed to occur. Reduced human interferences, creation of vegetation buffer zones, and establishment of activity restriction zones are recommended for wintering grounds. J. WILDL. MANAGE. 42(3):506-613 Declines in bald eagle populations have been attributed, in part, to direct human disturbance (Sprunt 1969). Urban sprawl, recreational developments, and increasing outdoor activity in prime ea- gle habitat have directly and adversely affected the bald eagle (Sprunt and Ligas 1966). Evaluations of the effects of dis- turbance on nesting eagles are difficult to generalize. Mathisen (1968), Grier (1969), Grier and Fyfe (in press), and Newman et al. (1977) found no evidence of human activity disrupting reproductive success. Juenemann (1973) and Grubb (1976a), however, have shown an indirect rela- tionship between human activities and nesting success. Concern over the sensi- tivity of wintering populations to human interferences has also been expressed (Shea 1973, Servheen 1975, Steenhof 1976, Spencer 1976, Stalmaster 1976). Quantitative information on the avoid- ance responses of eagles to disturbance is lacking. The purpose of this study was to measure the effects of human disturb- ance on the behavior of wintering bald eagles in northwest Washington. Support for this study was provided by Huxley College of Environmental Stud- ies and the Bureau for Faculty Research at Western Washington University. The U.S. Forest Service provided living quar- ters for the senior author. STUDY AREA Bald eagles winter along the Nooksack River in northwest Washington feeding on spawned-out salmon (Oncorhynchus spp.). The Nooksack River drains the northwestern slopes of the North Cas- cade Range, an area of approximately 1,671 km2. The primary study area was a 40-km segment of the river. The river there was braided with numerous gravel bars and anastomosing side-channels, and provided optimal spawning habitat for salmon, thereby supporting large numbers of eagles. Riparian vegetation in the study area consisted primarily of red alder (Alnus rubra), black cottonwood (Populus tri- chocarpa), western red cedar (Thuja pli- cata), Sitka spruce (Picea sitchensis), and bigleaf maple (Acer macrophyllum). Sur- ' Present address: Department of Biology, Utah State University, Logan 84322. 2 Present address: Environmental Science and Engineering, Inc., P. O. Box 13454, University Sta- tion, Gainesville, Florida 32604. 506 J. Wildl. Manage. 42(3):1978 BEHAVIORAL RESPONSES OF WINTERING BALD EAGLES TO HUMAN ACTIVITY MARK V.STALMASTER,Huxley College of Environmental Studies,Western Washington University,Bellingham 98225 1 JAMES R.NEWMAN,Huxley College of Environmental Studies,Western Washington University,Bellingham 98225% Abstract:The effects of human activity on wintering bald eagles (Haliaeetus leucocephalus)were studied in Washington for 2 yr.Tolerance to disturbance was determined by analyzing eagle distribution in relation to human activity and by measuring flight distances of eagles from simulated human disturbances.Human activity was found to affect eagle distribution and behavior adversely.Distribution patterns were signif- icantly (P <0.001)changed,resulting in displacement of eagles to areas of lower human activity.Older birds were more sensitive (P <0.001)to disturbances.Flight distances were highest for simulated dis- turbances in water and on gravel bars,intermediate on land,and shortest under a vegetation canopy. Habituation to normal activities seemed to occur.Reduced human interferences,creation of vegetation buffer zones,and establishment of activity restriction zones are recommended for wintering grounds. J.WILDL.MANAGE.42(3):506-613 Declines in bald eagle populations have been attributed,in part,to direct human disturbance (Sprunt 1969).Urban sprawl,recreational developments,and increasing outdoor activity in prime ea- gle habitat have directly and adversely affected the bald eagle (Sprunt and Ligas 1966).Evaluations of the effects of dis- turbance on nesting eagles are difficult to generalize.Mathisen (1968),Grier (1969), Grier and Fyfe (in press),and Newman et al.(1977)found no evidence of human activity disrupting reproductive success. Juenemann (1973)and Grubb (19700), however,have shown an indirect rela- tionship between human activities and nesting success.Concern over the sensi- tivity of wintering populations to human interferences has also been expressed (Shea 1973,Servheen 1975,Steenhof 1976,Spencer 1976,Stalmaster 1976). Quantitative information on the avoid- ance responses of eagles to disturbance is lacking.The purpose of this study was to measure the effects of human disturb- 1 Present address:Department of Biology,Utah State University,Logan 84322. 2 Present address:Environmental Science and Engineering,Inc.,P.O.Box 13454,University Sta- tion,Gainesville,Florida 32604. 506 ance on the behavior of wintering bald eagles in northwest Washington. Support for this study was provided by Huxley College of Environmental Stud- ies and the Bureau for Faculty Research at Western Washington University.The U.S.Forest Service provided living quar- ters for the senior author. STUDY AREA Bald eagles winter along the Nooksack River in northwest Washington feeding on spawned-out salmon (Oncorhynchus spp.).The Nooksack River drains the northwestern slopes of the North Cas- cade Range,an area of approximately 1,671 km2 •The primary study area was a 40-km segment of the river.The river there was braided with numerous gravel bars and anastomosing side-channels, and provided optimal spawning habitat for salmon,thereby supporting large numbers of eagles. Riparian vegetation in the study area consisted primarily of red alder (Alnus rubra),black cottonwood (Populus tri- chocarpa),western red cedar (Thuja pli- cata),Sitka spruce (Picea sitchensis),and bigleaf maple (Acermacrophyllum).Sur- J.Wildl.Manage.42(3):1978 BALD EAGLE AVOIDANCE BEHAVIOR *Stalmaster and Newman 507 Table 1. Criteria used for classifying high, moderate, and low human activity types in the Nooksack River study area. Activity types High Moderate Low Human activity frequently Human activity occasionally Human activity unlikely within sight of eagles. within sight of eagles. within sight of eagles. High recreational use; sport Moderate recreational use; Low recreational use; sport fishing common. sport fishing occasional. fishing rare. Heavily traversed roads Roads paralleling river with Roads not paralleling river paralleling river; bridges; easy some vegetation buffer zone; and/or extensive vegetation access to river, good access to river. buffer zone; poor access to river. Habitat altered; numerous Habitat slightly altered; Habitat largely undisturbed; human developments, occasional human human developments absent. developments. Activity almost always Activity occasionally Activity unlikely to encroach encroaches on flight distance. encroaches on flight distance. on flight distance. rounding forests were composed of Douglas fir (Pseudotsuga menziesii), western hemlock (Tsuga heterophylla), and western red cedar. Human activities along the Nooksack River have increased in intensity and fre- quency in recent years. The major influ- ences are logging, housing, mining, and recreation. Primary and secondary roads parallel most of the river. Recreational use is common and numerous recreation- al land developments are present. Sport steelhead (Salmo gairdneri) fishing is popular in the study area. METHODS Data on eagle avoidance behavior were collected during the autumns and win- ters of 1974-75 and 1975-76. Analyses of these responses were made by correlat- ing the distribution of eagles in relation to existing human activities and by mea- suring the flight or flushing distances of birds from simulated disturbances. The 40-km study area was divided into 100 units of either high, moderate, or low human activity types. The criteria for classifying the extent of activity (Table 1) was similar to that of Mathisen (1968). Each unit consisted of 0.8 km of bank on 1 side of the river. Eagle distribution re- corded during censuses was correlated with each activity type. Forty-four semiweekly censuses were made during 1975-76 at predetermined dates and at approximately the same time each census day. Counts were conducted by driving or walking to 30 observation points along the river. We assumed that a large percentage of birds were migra- tory transients, thus reducing the fre- quency of repeated observations of the same birds; however, some repetition was unavoidable. Simulated human disturbances consist- ed of approach toward a single eagle on foot by a single investigator and obser- vation of the avoidance behavior. Three types were used: (1) Vegetation zone- approaching an eagle under a heavy veg- etation canopy where the observer was partially and intermittently visible to the bird; (2) Riverbank-approaching an ea- gle in open meadows adjacent to the riv- er's edge; and (3) River channel-ap- proaching an eagle along gravel bars or in the river (wading). Both Riverbank J. Wildl. Manage. 42(3):1978 BALD EAGLE AVOIDANCE BEHAVIOR·Stalmaster and Newman 507 Table 1.Criteria used for classifying high,moderate,and low human activity types in the Nooksack River study area. Activity types High Human activity frequently within sight of eagles. High recreational use;sport fishing common. Heavily traversed roads paralleling river;bridges;easy access to river. Habitat altered;numerous human developments. Activity almost always encroaches on flight distance. Moderate Human activity occasionally within sight of eagles. Moderate recreational use; sport fishing occasional. Roads paralleling river with some vegetation buffer zone; good access to river. Habitat slightly altered; occasional human developments. Activity occasionally encroaches on flight distance. Low Human activity unlikely within sight of eagles. Low recreational use;sport fi shing rare. Roads not paralleling river and/or extensive vegetation buffer zone;poor access to river. Habitat largely undisturbed; human developments absent. Activity unlikely to encroach on flight distance. rounding forests were composed of Douglas fir (Pseudotsuga rnenziesii), western hemlock (Tsuga heterophylla), and western red cedar. Human activities along the Nooksack River have increased in intensity and fre- quency in recent years.The major influ- ences are logging,housing,mining,and recreation.Primary and secondary roads parallel most of the river.Recreational use is common and numerous recreation- al land developments are present.Sport steelhead (Salrno gairdneri)fishing is popular in the study area. METHODS Data on eagle avoidance behavior were collected during the autumns and win- ters of 1974-75 and 1975-76.Analyses of these responses were made by correlat- ing the distribution of eagles in relation to existing human activities and by mea- suring the flight or flushing distances of birds from simulated disturbances. The 40-km study area was divided into 100 units of either high,moderate,or low human activity types.The criteria for classifying the extent of activity (Table 1) J.Wildl.Manage.42(3):1978 was similar to that of Mathisen (1968). Each unit consisted of 0.8 km of bank on 1 side of the river.Eagle distribution re- corded during censuses was correlated with each activity type. Forty-four semiweekly censuses were made during 1975-76 at predetermined dates and at approximately the same time each census day.Counts were conducted by driving or walking to 30 observation points along the river.We assumed that a large percentage of birds were migra- tory transients,thus reducing the fre- quency of repeated observations of the same birds;however,some repetition was unavoidable. Simulated human disturbances consist- ed of approach toward a single eagle on foot by a single investigator and obser- vation of the avoidance behavior.Three types were used:(1)Vegetation zone- approaching an eagle under a heavy veg- etation canopy where the observer was partially and intermittently visible to the bird;(2)Riverbank-approaching an ea- gle in open meadows adjacent to the riv- er's edge;and (3)River channel-ap- proaching an eagle along gravel bars or in the river (wading).Both Riverbank 508 BALD EAGLE AVOIDANCE BEHAVIOR *Stalmaster and Newman Table 2. Distribution of bald eagles in the 3 human activity types in the Nooksack River study area. High Moderate Low activity activity activity No. % No. % No. % Human activity units 23 23 41 41 36 36 Juveniles and subadultsb 42 9 269 60 135 30 Adultsb 60 7 447 55 311 38 Totala 102 8 716 57 446 35 a P < 0.001. b Age-class difference significant (P < 0.025). and River channel types were in open areas where the disturbance was initially visible to the bird between 400 and 600 m. Subjects of the Vegetation zone dis- turbances were initially aware of the in- vestigator between 75 and 100 m. All 360 simulated disturbances were conducted during full daylight on birds perched on trees along the river. Disturbance exper- iments were made intermittently throughout the 1974-75 and 1975-76 wintering seasons and on different stretches of the river to allow sampling of different birds. We suspect repetition to be minor by this process. Some birds were subjected to repeated disturbances, but only responses to the first disturb- ance were used in the analyses. In addition, approaches were made to groups of eagles where 2 or more differ- ent age-classes were present. Flight distances of eagles were record- ed during all simulated disturbances and used as a measure of the disturbance re- action (Hediger 1968:40-41). Distances over 500 m were not recorded due to the likelihood of an unknown factor eliciting the escape response. A Rangematic, Mark V, optical range finder was used to mea- sure distances. Eagles were aged by plumage color- ation as initially described by Southern (1964, 1967) and further verified by Servheen (1975). Aging is most reliably done by combining as many plumage characteristics as possible; however, some birds are difficult or impossible to age (Servheen 1975). Only birds positive- ly aged by this method were used in the disturbance experiments. Some individ- ual variations were evident, but we feel that our classification errors by this meth- od were minor. We have presented data grouping ea- gles into 3 age-classes and also into 6 age- classes; however, results of the statistical analyses were similar for both methods. Birds in their 1st year (1-year-olds) are referred to as juveniles; whereas, 2- through 5-year-olds are grouped as sub- adults. Six-year-olds (Southern's plum- age F) and older birds were recorded as adults. Statistical analyses were performed ac- cording to Zar (1974). Differences in the distribution patterns for totals and age- class dissimilarities were tested with chi- square goodness of fit tests and contin- gency tables, respectively. Analyses of covariance were used to compare regres- sion line elevations and slopes among the 3 disturbance types. Where birds were grouped as juveniles, subadults, and adults, a 3 x 3 factorial design was used to evaluate differences among the 3 dis- turbance types and the 3 age-classes. The Newman-Keuls multiple range test de- termined if differences were significant between 2 specific types. Mean life ex- pectancy, longevity, and hence mean J. Wildl. Manage. 42(3):1978 508 BALD EAGLE AVOIDANCE BEHAVIOR •Stalmaster and Newman Table 2.Distribution of bald eagles in the 3 human activity types in the Nooksack River study area. No.% High activity Human activity units Juveniles and subadults b Adults b Totala a P <0.001. b Age-class difference significant (P <0.025). 23 42 60 102 23 9 7 8 Moderate Low activity activity No.%No.% 41 41 36 36 269 60 135 30 447 55 311 38 716 57 446 35 and River channel types were in open areas where the disturbance was initially visible to the bird between 400 and.600 m.Subjects of the Vegetation zone dis- turbances were initially aware of the in- vestigator between 75 and 100 m.All 360 simulated disturbances were conducted during full daylight on birds perched on trees along the river.Disturbance exper- iments were made intermittently throughout the 1974-75 and 1975-76 wintering seasons and on different stretches of the river to allow sampling of different birds.We suspect repetition to be minor by this process.Some birds were subjected to repeated disturbances, but only responses to the first disturb- ance were used in the analyses. In addition,approaches were made to groups of eagles where 2 or more differ- ent age-classes were present. Flight distances of eagles were record- ed during all simulated disturbances and used as a measure of the disturbance re- action (Hediger 1968:40-41).Distances over 500 m were not recorded due to the likelihood of an unknown factor eliciting the escape response.A Rangematic,Mark V,optical range finder was used to mea- sure distances. Eagles were aged by plumage color- ation as initially described by Southern (1964,1967)and further verified by Servheen (1975).Aging is most reliably done by combining as many plumage characteristics as possible;however, some birds are difficult or impossible to age (Servheen 1975).Only birds positive- ly aged by this method were used in the disturbance experiments.Some individ- ual variations were evident,but we feel that our classification errors by this meth- od were minor. We have presented data grouping ea- gles into 3 age-classes and also into 6 age- classes;however,results of the statistical analyses were similar for both methods. Birds in their 1st year (I-year-olds)are referred to as juveniles;whereas,2- through 5-year-olds are grouped as sub- adults.Six-year-olds (Southern's plum- age F)and older birds were recorded as adults. Statistical analyses were performed ac- cording to Zar (1974).Differences in the distribution patterns for totals and age- class dissimilarities were tested with chi- square goodness of fit tests and contin- gency tables,respectively.Analyses of covariance were used to compare regres- sion line elevations and slopes among the 3 disturbance types.Where birds were grouped as juveniles,subadults,and adults,a 3 x 3 factorial design was used to evaluate differences among the 3 dis- turbance types and the 3 age-classes.The Newman-Keuls multiple range test de- termined if differences were significant between 2 specific types.Mean life ex- pectancy,longevity,and hence mean J.Wildl.Manage.42(3):1978 BALD EAGLE AVOIDANCE BEHAVIOR Stalmaster and Newman 509 Table 3. Comparison of the distribution of bald eagles along riversides where the human activity types differed for adjacent sides of the river. Juveniles and subadultsb Adultsb Totala Distribution No. % No. % No. % On high activity side 53 31 87 22 140 25 On low activity side 118 69 300 78 418 75 a P < 0.001. b Age-class difference significant (P < 0.05). adult age are unknown for bald eagles (Brown and Cade 1972). To facilitate sta- tistical analyses, adults were grouped as 10-year-olds. RESULTS AND DISCUSSION Distribution in Relation to Human Activity The distribution of bald eagles on the Nooksack River reflected the effect of hu- man activity. There was a lower (P < 0.001) number of eagles in units of high activity compared to the units of moder- ate activity (Table 2). Distribution in the low activity units was near the expected level. Where the activity classification differed for adjacent sides of the river, a greater (P < 0.001) number of birds were observed on the riverside with lower hu- man disturbance (Table 3). Disturbances in high activity units were beyond the tolerance limits of most wintering eagles. Moderate human activ- ity seemed to be tolerated; however, the shift in distribution resulted in a greater number of birds displaced to marginal habitat and confined the population to a smaller area. Since the major feeding grounds were disproportionately located in high and moderate human activity units, the number of birds in low activity units was lower than expected. On the Nooksack River, the shift in distribution was evident only in high and moderate units and prevented effective use of all feeding sites. Many unsuccessful feeding attempts were observed to occur in these units due to the existing human activi- ties. Human activity had a significant influ- ence on the feeding behavior of eagles. Mathisen (1968) stated that human activ- ity could interfere with food gathering and possibly cause general unrest among an eagle population. However, Steenhof (1976) suggested that tolerance levels were high on preferred feeding sites. Feeding behavior on the Nooksack was disrupted by the mere presence of hu- mans. Disturbed birds did not return to the same feeding area until several hours after the disturbance occurred and only when the disturbance no longer persist- ed. Compared to other eagle activities, feeding birds were most sensitive to hu- man interferences. A comparison of juvenile and subadult to adult distribution indicated a signifi- cant difference in tolerance to existing human activities (Tables 2, 3). Adults were more sensitive to disturbance than younger birds and preferred units of low- er activity. Proportions of adults were lower in areas where activity was partic- ularly prevalent. Reaction to Simulated Disturbances General Effects.-Variations in the flight distances of the entire population were high, but distances were generally J. Wildl. Manage. 42(3):1978 BALD EAGLE A VOIDANCE BEHAVIOR·Stalmaster and Newman 509 Table 3.Comparison of the distribution of bald eagles along riversides where the human activity types differed for adjacent sides of the river. Distribution On high activity side On low activity side a p <0.001. b Age-class difference significant (P <0.05). Juveniles and subadultsb No. 53 118 % 31 69 No. 87 300 Adults b % 22 78 No. 140 418 % 25 75 adult age are unknown for bald eagles (Brown and Cade 1972).To facilitate sta- tistical analyses,adults were grouped as 10-year-olds. RESULTS AND DISCUSSION Distribution in Relation to Human Activity The distribution of bald eagles on the Nooksack River reflected the effect ofhu- man activity.There was a lower (P < 0.001)number of eagles in units of high activity compared to the units of moder- ate activity (Table 2).Distribution in the low activity units was near the expected level.Where the activity classification differed for adjacent sides of the river,a greater (P <0.001)number of birds were observed on the riverside with lower hu- man disturbance (Table 3). Disturbances in high activity units were beyond the tolerance limits of most wintering eagles.Moderate human activ- ity seemed to be tolerated;however,the shift in distribution resulted in a greater number of birds displaced to marginal habitat and confined the population to a smaller area.Since the major feeding grounds were disproportionately located in high and moderate human activity units,the number of birds in low activity units was lower than expected.On the Nooksack River,the shift in distribution was evident only in high and moderate units and prevented effective use of all J.Wildl.Manage.42(3):1978 feeding sites.Many unsuccessful feeding attempts were observed to occur in these units due to the existing human activi- ties. Human activity had a significant influ- ence on the feeding behavior of eagles. Mathisen (1968)stated that human activ- ity could interfere with food gathering and possibly cause general unrest among an eagle population.However,Steenhof (1976)suggested that tolerance levels were high on preferred feeding sites. Feeding behavior on the Nooksack was disrupted by the mere presence of hu- mans.Disturbed birds did not return to the same feeding area until several hours after the disturbance occurred and only when the disturbance no longer persist- ed.Compared to other eagle activities, feeding birds were most sensitive to hu- man interferences. A comparison of juvenile and subadult to adult distribution indicated a signifi- cant difference in tolerance to existing human activities (Tables 2,3).Adults were more sensitive to disturbance than younger birds and preferred units of low- er activity.Proportions of adults were lower in areas where activity was partic- ularly prevalent. Reaction to Simulated Disturbances General Effects.-Variations in the flight distances of the entire population were high,but distances were generally 510 BALD EAGLE AVOIDANCE BEHAVIOR* Stalmaster and Newman 350 500 - ] VEGETATION ZONE 300 I RIVER-BANK M RIVER-CHANNEL c,, W 250 w w 1 200 z o 150 I-- (50) 50- i(50) (20) ( ) (20) (20) (80) (80) 8) (32) JUVENLE SUBADULT ADULT AGE CLASS Fig. 1. Comparison of the responses (flight distances) of juvenile, subadult, and adult bald eagles to Vegetation zone, Riverbank, and River channel simulated disturbance types. Mean ? standard deviation, range, and sample sizes (in parentheses) are presented. between 25 and 300 m (Table 4). The mean flight distance in open areas (Riv- erbank and River channel combined) was 196 m for adults and 99 m for juveniles and subadults. We suggest using flight co r 150 - VEGETATION ZONE HA w Y = 24.793 + 4.201X 7 r = 0.7204 1'O0 - n = 60 w z (1) 50 5 (8) (8) (20) ( )(8) r 0 (8) 1 2 3 4 5 ADULT(IO) AGE CLASS Fig. 2. Relationship of the responses (flight distances) to the 6 age-classes of bald eagles for the Vegetation zone simulated disturbance type. Mean _ standard deviation, range, and sample sizes (in parentheses) are presented. distances to determine sensitivity to dis- turbance and to establish boundaries for activity restriction zones. When disturbed, eagles flew avoidance flights between 50 and 500 m and rarely left the river system. Flight was usually directly or quartering away from the dis- turbance. Only 1 simulated disturbance elicited apparent aggressive behavior. An adult made 6 diving flights between 10 and 20 m at the observer and vocalized extensively. Nesting eagles commonly exhibit aggressive behavior (Grier 1969, Table 4. Percentages of bald eagles responding at designated flight distances for the 3 simulated disturbance types. Flight distances (m) Age-class N 15-50 51-100 101-150 151-200 201-250 251-300 >300 Vegetation zone Juveniles 8 88 12 Subadults 32 84 16 Adults 20 20 80 Total 60 63 37 Riverbank Juveniles 20 65 30 5 Subadults 80 15 39 31 12 3 Adults 50 2 14 28 38 16 2 Total 150 17 30 27 19 6 1 River channel Juveniles 20 45 35 10 10 Subadults 80 8 34 36 18 5 Adults 50 8 8 20 34 18 12 Total 150 10 26 23 17 14 6 4 J. Wildl. Manage. 42(3):1978 510 BALD EAGLE AVOIDANCE BEHAVIOR •Stalmaster and Newman Fig.1.Comparison of the responses (flight distances)of juvenile,subadult,and adult bald eagles to Vegetation zone,Riverbank,and River channel simulated disturbance types.Mean ±standard deviation,range,and sample sizes (in parentheses)are presented. between 25 and 300 m (Table 4).The mean flight distance in open areas (Riv- erbank and River channel combined)was 196 m for adults and 99 m for juveniles and subadults.We suggest using flight o VEGETATION ZONE _RIVER-BANK _RIVER-CHANNEL VEGETATION ZON E y=24.793 +4.201x r =0.7204 n =60 ~150 I.LJ to- I.LJ ~ -100 I.LJ U Z ~en 50o to- I 5:2....J 0 L......L._...L.-----'-_...L----&-.L.--_ u..3 4 5 ADULT(fO) AGE CLASS Fig.2.Relationship of the responses (flight distances)to the 6 age-classes of bald eagles for the Vegetation zone simulated disturbance type.Mean ±standard deviation, range,and sample sizes (in parentheses)are presented. distances to determine sensitivity to dis- turbance and to establish boundaries for activity restriction zones. When disturbed,eagles flew avoidance flights between 50 and 500 m and rarely left the river system.Flight was usually directly or quartering away from the dis- turbance.Only 1 simulated disturbance elicited apparent aggressive behavior.An adult made 6 diving flights between 10 and 20 m at the observer and vocalized extensively.Nesting eagles commonly exhibit aggressive behavior (Grier 1969, ADULT 500 (20) at (50)r (50) SUBADULT AGE CLASS II JIt(80)(80) JUVENLE ill Jt (20) 8 50 350 300 ena:: ~250 L£J ~ ~200z ~en 25 150 l-x: (!) ::::i 100 LL Table 4.Percentages of bald eagles responding at designated flight distances for the 3 simulated disturbance types. Flight distances (m) Age-class N 15--50 51-100 101-150 151-200 201-250 251-300 >300 Vegetation zone Juveniles 8 88 12 Subadults 32 84 16 Adults 20 20 80 Total 60 63 37 Riverbank Juveniles 20 65 30 5 Subadults 80 15 39 31 12 3 Adults 50 2 14 28 38 16 2 Total 150 17 30 27 19 6 1 River channel Juveniles 20 45 35 10 10 Subadults 80 8 34 36 18 5 Adults 50 8 8 20 34 18 12 Total 150 10 26 23 17 14 6 4 J.Wildl.Manage.42(3):1978 BALD EAGLE AVOIDANCE BEHAVIOR * Stalmaster and Newman 511 RIVER-BANK 250 - = 56.353+10.986x r = 0.6365 n=150 200 w 7- w 150- z oI I-- - -r (D9 - 50- SL (50) (20) (20) (20) 0 ' 1 I1 1 11 1 2 3 4 5 ADU LT(10) AGE CLASS Fig. 3. Relationship of the responses (flight distances) to the 6 age-classes of bald eagles for the Riverbank simu- lated disturbance type. Mean + standard deviation, range, and sample sizes (in parentheses) are presented. Grubb 1976b); however, no documenta- tion of this type exists for wintering birds. Repeated disturbances were made to the same individual in several instances. Responses were similar in both the flight distance and the distance of the avoid- ance flight. Normally occurring auditory disturb- ances were not unduly disruptive to ea- gle behavior. Gunshots were the only noises that elicited overt escape behav- ior. Edwards (1969) found that gunshots could be used effectively in flushing birds from their roosts. Eagles were es- pecially tolerant of auditory stimuli when the sources were partially or totally con- cealed from their view. Age-Class Variation.-Sensitivity to human disturbance increases with age. Flight distances of older birds were greater (P < 0.001) than younger birds for all 3 simulated disturbance types (Fig. 1). A gradual increase in distance was ap- parent among the subadults, but a steep rise occurred with adults (P < 0.001 for all 3 disturbance types) (Figs. 2, 3, 4). 350 RIVER-CHANNEL 500 S= 57.872 + 16.828 x r = 0.6971 300 n = 150 c/ w 250 I.- w w 200 0 z 3 150 I-- J100 (50) (20) 50- (20) (20) (20) )(20 ) (20) 1 2 3 4 5 ADULT(10) AGE CLASS Fig. 4. Relationship of the responses (flight distances) to the 6 age-classes of bald eagles for the River channel sim- ulated disturbance type. Mean + standard deviation, range, and sample sizes (in parentheses) are presented. This differential sensitivity was support- ed by distribution patterns (Tables 2, 3) and by the upper limit in the range of flight distances exhibited by juveniles and subadults (Table 4). The tolerance of juveniles (1-year-olds) was relatively high, whereas, adults were highly sensi- tive to human approach. The accuracy of our ground censuses may have been affected by this differen- tial response behavior. The more easily displaced adults may have been dupli- cated in our counts, whereas juveniles and subadults were more easily missed due to their sedentary behavior and less conspicuous plumages. When approaches were made to groups of eagles, the oldest bird flushed first 81 percent of the time (P < 0.001). This fur- ther confirmed the contention that older birds were more sensitive to disturbance. Young birds seemed to react to the flight J. Wildl. Manage. 42(3):1978 BALD EAGLE AVOIDANCE BEHAVIOR·Stalmaster and Newman 511 RIVER-CHANNEL 500 1\ Y =57.872 +16.828 x r =0.6971 n =150300 350 en ffi 250 I-w :E ~200 z ~enCi 150 l- I (!) :J 100 L&.. o '--.L.-2-'----3L.--.-I.4--'-5------A-DU-L~T-(-,0) AGE CLASS 50 (50) ADULT(10) RIVER-BANK y=56.353+10.986X r =0.6365 n=150 100 250 enffi 200 ~ IJJ ~ lLI 150 <.Jz ~eno ~ I (!):J 50 L&.. 2 345 AGE CLASS Fig.3.Relationship of the responses (flight distances)to the 6 age-classes of bald eagles for the Riverbank simu- lated disturbance type.Mean ±standard deviation,range, and sample sizes (in parentheses)are presented. Grubb 1976b);however,no documenta- tion of this type exists for wintering birds. Repeated disturbances were made to the same individual in several instances. Responses were similar in both the flight distance and the distance of the avoid- ance flight. Normally occurring auditory disturb- ances were not unduly disruptive to ea- gle behavior.Gunshots were the only noises that elicited overt escape behav- ior.Edwards (1969)found that gunshots could be used effectively in flushing birds from their roosts.Eagles were es- pecially tolerant of auditory stimuli when the sources were partially or totally con- cealed from their view. Age-Class Variation.-Sensitivity to human disturbance increases with age. Flight distances of older birds were greater (P <0.001)than younger birds for al13 simulated disturbance types (Fig.1). A gradual increase in distance was ap- parent among the subadults,but a steep rise occurred with adults (P <0.001 for all 3 disturbance types)(Figs.2,3,4). Fig.4.Relationship of the responses (flight distances)to the 6 age-classes of bald eagles for the River channel sim- ulated disturbance type.Mean ±standard deviation,range, and sample sizes (in parentheses)are presented. This differential sensitivity was support- ed by distribution patterns (Tables 2,3) and by the upper limit in the range of flight distances exhibited by juveniles and subadults (Table 4).The tolerance of juveniles (I-year-olds)was relatively high,whereas,adults were highly sensi- tive to human approach. The accuracy of our ground censuses may have been affected by this differen- tial response behavior.The more easily displaced adults may have been dupli- cated in our counts,whereas juveniles and subadults were more easily missed due to their sedentary behavior and less conspicuous plumages. When approaches were made to groups of eagles,the oldest bird flushed first 81 percent of the time (P <0.001).This fur- ther confirmed the contention that older birds were more sensitive to disturbance. Young birds seemed to react to the flight J.Wildl.Manage.42(3):1978 512 BALD EAGLE AVOIDANCE BEHAVIOR* Stalmaster and Newman behavior of their older counterparts when disturbed and not to the observer, thus elevating their flight distances. Pos- sibly, young eagles were sensitized to human activity by the behavior of older birds on the wintering grounds. Vegetation Zone Effect.-Wintering bald eagles were more tolerant when hu- man disturbances were partially ob- scured from their line of sight by buffers of vegetation. When birds were ap- proached through heavy riparian vegeta- tion (Vegetation zone), flight distances were significantly shorter than distances recorded in open areas (compared to Riverbank, Fig. 1, q = 12.51, P < 0.001 or Figs. 2, 3, q = 15.51, P < 0.001). In- creasing sensitivity with age was also ev- ident for Vegetation Zone (Fig. 2, r = 0.72, P < 0.001). We expected juveniles and subadults to respond at the same flight distance as adults since the observer was initially visible to the birds at between 75 and 100 m. At this distance in open areas, flight had usually occurred, but this was not observed in Vegetation zone. This diminished response pattern in veg- etation zones was probably related to the eagles' reliance on visual identification of the disturbance before flight. Vegetation buffer zones on wintering grounds can be effective in reducing the disturbance caused by human activities. Buffer zones are used to protect nesting territories on lands administered by the U.S. Forest Service (Juenemann 1973). Strips of vegetation, which efficiently re- duce line-of-sight contact, will allow a closer presence of human activity and also provide perching and roosting trees. Based on flight distance information (Ta- ble 4 and Fig. 2), we recommend zones 75 to 100 m in width to protect critical wintering grounds where disturbances are common. Habituation to Activity.-Wintering eagles can become habituated to routine human activities. Birds on the Nooksack were more easily disturbed by ap- proaches that occurred on the river chan- nel than from adjacent farm meadows where activity was common. Flight dis- tances were longer for the River channel simulated disturbances than for the Riv- erbank type (Fig. 1, q = 7.99, P < 0.001 or Figs. 3 and 4, q = 7.36, P < 0.001). This aspect was particularly prevalent with adults as evidenced by comparison of the regression line slopes (q = 4.60, P < 0.05) for these 2 types. Habituation is the temporary or per- manent waning of responsiveness by an- imals to repetitious stimuli (Marler and Hamilton 1966:642-644). In bald eagles, habituation to human activity has been suggested by Grier (1969) and Edwards (1969). Grier contended that birds accus- tomed to activity were less disturbed by nest climbing. Edwards found that ap- proaches could be made closer at roosts which were closer to usual activity. Ea- gles on the Nooksack were also easier to approach when the approach was made from areas regularly high in activity (Riv- erbank). Tolerance to human activity is related to the location of the disturbance. Our data indicated that activities involved di- rectly on the channel of the river, such as boating and fishing, were most disturbing if activities did not regularly occur there. Normal daily activity patterns were changed by human presence. Activity restriction zones are suggested for bald eagle wintering grounds. In open regions where activities are common (Riverbank), boundaries of 250 m would be sufficient to protect 99 percent of the population (Table 4). Similarly, bound- aries on river channels of 250 m would protect 90 percent of the birds. A com- bination of both activity restriction zones J. Wildl. Manage. 42(3):1978 512 BALD EAGLE AVOIDANCE BEHAVIOR·Stalmaster and Newman behavior of their older counterparts when disturbed and not to the observer, thus elevating their flight distances.Pos- sibly,young eagles were sensitized to human activity by the behavior of older birds on the wintering grounds. Vegetation Zone Effect.-Wintering bald eagles were more tolerant when hu- man disturbances were partially ob- scured from their line of sight by buffers of vegetation.When birds were ap- proached through heavy riparian vegeta- tion (Vegetation zone),flight distances were significantly shorter than distances recorded in open areas (compared to Riverbank,Fig."I,q =12.51,P <0.001 or Figs.2,3,q =15.51,P <0.001).In- creasing sensitivity with age was also ev- ident for Vegetation Zone (Fig.2,r =0.72, P <0.001).We expected juveniles and subadults to respond at the same flight distance as adults since the observer was initially visible to the birds at between 75 and 100 m.At this distance in open areas,flight had usually occurred,but this was not observed in Vegetation zone. This diminished response pattern in veg- etation zones was probably related to the eagles'reliance on visual identification of the disturbance before flight. Vegetation buffer zones on wintering grounds can be effective in reducing the disturbance caused by human activities. Buffer zones are used to protect nesting territories·on lands administered by the u.s.Forest Service (Juenemann 1973). Strips of vegetation,which efficiently re!"" duce line-of-sight contact,will allow a closer presence of human activity and also provide perching and roosting trees. Based on flight distance information (Ta- ble 4 and Fig.2),we recommend zones 75 to 100 m in width to protect critical wintering grounds where disturbances are common. Habituation to Activity.-Wintering eagles can become habituated to routine human activities.Birds on the Nooksack were more easily disturbed by ap- proaches that occurred on the river chan- nel than from adjacent farm meadows where activity was common.Flight dis- tances were longer for the River channel simulated disturbances than for the Riv- erbank type (Fig.1,q =7.99,P <0.001 or Figs.3 and 4,q =7.36,P <0.001). This aspect was particularly prevalent with adults as evidenced by comparison of the regression line slopes (q =4.60,P <0.05)for these 2 types. Habituation is the temporary or per- manent waning of responsiveness by an- imals to repetitious stimuli (Marler and Hamilton 1966:642-644).In bald eagles, habituation to human activity has been suggested by Grier (1969)and Edwards (1969).Grier contended that birds accus- tomed to activity were less disturbed by nest climbing.Edwards found that ap- proaches could be made closer at roosts which were closer to usual activity.Ea- gles on the Nooksack were also easier to approach when the approach was made from areas regularly high in activity (Riv- erbank). Tolerance to human activity is related to the location of the disturbance.Our data indicated that activities involved di- rectly on the channel of the river,such as b.oating and fishing,were most disturbing if activities did not regularly occur there. Normal daily activity patterns were changed by human presence. Activity restriction zones are suggested for bald eagle wintering grounds.In open regions where activities are common (Riverbank),boundaries of 250 m would be sufficient to protect 99 percent of the population (Table 4).Similarly,bound- aries on river channels of 250 m would protect 90 percent of the birds.A com- bination of both activity restriction zones J.Wildl.Manage.42(3):1978 BALD EAGLE AVOIDANCE BEHAVIOR *Stalmaster and Newman 513 and vegetation buffer zones is desirable. Establishment of sanctuaries with these elements is encouraged. Variations of tol- erance between populations is likely; therefore, we suggest monitoring of adult eagle behavior in developing manage- ment recommendations since they are most sensitive to human presence. LITERATURE CITED BROWN, L. H., AND T. J. CADE. 1972. Age classes and population dynamics of the bateleur and African fish eagle. Ostrich 43(1):1-16. EDWARDS, C. C. 1969. Winter behavior and pop- ulation dynamics of American eagles in west- ern Utah. Ph.D. Thesis. Brigham Young Univ., Provo. 157pp. GRIER, J. W. 1969. Bald eagle behavior and pro- ductivity responses to climbing to nests. J. Wildl. Manage. 33(4):961-966. ,AND R. FYFE. Assessing the impact of re- search activities: Birds of prey. Can. Wildl. Serv. Publ. In press. GRUBB, T. G. 1976a. A survey and analysis of bald eagle nesting in western Washington. M.S. Thesis. Univ. of Washington, Seattle. 87pp. . 1976b. Nesting bald eagles attack research- er. Auk 93(4):842-843. HEDIGER, H. 1968. The psychology and behavior of animals in zoos and circuses. Dover Publ., Inc., New York. 166pp. JUENEMANN, B. G. 1973. Habitat evaluations of se- lected bald eagle nest sites on the Chippewa National Forest. M.S. Thesis. Univ. of Minne- sota, Minneapolis. 170pp. MARLER, P. R., AND W. J. HAMILTON, III. 1966. Mechanisms of animal behavior. John Wiley and Sons, Inc., New York. 771pp. MATHISEN, J. E. 1968. Effects of human disturb- ance on nesting of bald eagles. J. Wildl. Man- age. 32(1):1-6. NEWMAN, J. R., W. H. BRENNAN, AND L. M. SMITH. 1977. Twelve-year changes in nesting patterns of bald eagles (Haliaeetus leucoce- phalus) on San Juan Island, Washington. Mur- relet 58(2):37-39. SERVHEEN, C. W. 1975. Ecology of the wintering bald eagles on the Skagit River, Washington. M.S. Thesis. Univ. of Washington, Seattle. 96pp. SHEA, D. S. 1973. A management-oriented study of bald eagle concentrations in Glacier National Park. M.S. Thesis. Univ. of Montana, Missoula. 78pp. SOUTHERN, W. E. 1964. Additional observations on winter bald eagle populations: Including re- marks on biotelemetry techniques and imma- ture plumages. Wilson Bull. 76(2):121-137. . 1967. Further comments on subadult bald eagle plumages. Jack-Pine Warbler 45(3):70- 80. SPENCER, D. A., ED. 1976. Wintering of the mi- grant bald eagle in the lower 48 states. Nat. Agr. Chem. Asso. Publ., Washington D.C. 170pp. SPRUNT, A., IV. 1969. Population trends of the bald eagle in North America. Pages 347-351 in J. J. Hickey, ed. Peregrine falcon populations: Their biology and decline. Univ. of Wisconsin Press, Madison. 596pp. - , AND F. J. LIGAS. 1966. Audubon bald eagle studies, 1960-1966. Proc. 62nd Ann. Cony. Nat. Aud. Soc., Sacramento, California. 6pp. STALMASTER, M. V. 1976. Winter ecology and ef- fects of human activity on bald eagles in the Nooksack River valley, Washington. M.S. The- sis. Western Washington Univ., Bellingham. 100pp. STEENHOF, K. 1976. The ecology of wintering bald eagles in southeastern South Dakota. M.S. The- sis. Univ. of Missouri, Columbia. 148pp. ZAR, J. H. 1974. Biostatistical analysis. Prentice- Hall, Inc., Englewood Cliffs. 620pp. Received 13 May 1977. Accepted 24 February 1978. J. Wildl. Manage. 42(3):1978 BALD EAGLE A VOIDANCE BEHAVIOR·Stalmaster and Newman 513 and vegetation buffer zones is desirable. Establishment of sanctuaries with these elements is encouraged.Variations of tol- erance between populations is likely; therefore,we suggest monitoring of adult eagle behavior in developing manage- ment recommendations since they are most sensitive to human presence. LITERATURE CITED BROWN,L.H.,AND T.J.CADE.1972.Age classes and population dynamics of the bateleur and African fish eagle.Ostrich 43(1):1-16. EDWARDS,C. C.1969.Winter behavior and pop- ulation dynamics of American eagles in west- ern Utah.Ph.D.Thesis.Brigham Young Univ., Provo.157pp. GRIER,J.W.1969.Bald eagle behavior and pro- ductivity responses to climbing to nests.J. Wildl.Manage.33(4):961-966. --,AND R.FYFE.Assessing the impact of re- search activities:Birds of prey.Can.Wildl. Servo Publ.In press. GRUBB,T.G.19700.A survey and analysis of bald eagle nesting in western Washington.M.S. Thesis.Univ.of Washington,Seattle.87pp. --.1976b.Nesting bald eagles attack research- er.Auk 93(4):842-843. HEDIGER,H.1968.The psychology and behavior of animals in zoos and circuses.Dover Publ., Inc.,New York.166pp. JUENEMANN,B.G.1973.Habitat evaluations of se- lected bald eagle nest sites on the Chippewa National Forest.M.S.Thesis.Univ.of Minne- sota,Minneapolis.170pp. MARLER,P.R.,AND W.J.HAMILTON,III.1966. Mechanisms of animal behavior.John Wiley and Sons,Inc.,New York.771pp. MATHISEN,J.E.1968.Effects of human disturb- ance on nesting of bald eagles.J.Wildl.Man- age.32(1):1-6. J.WildI.Manage.42(3):1978 NEWMAN,J.R.,W.H.BRENNAN,AND L.M. SMITH.1977.Twelve-year changes in nesting patterns of bald eagles (Haliaeetus leucoce- phalus)on San Juan Island,Washington.Mur- relet 58(2):37-39. SERVHEEN,C.W.1975.Ecology of the wintering bald eagles on the Skagit River,Washington. M.S.Thesis.Univ.of Washington,Seattle. 96pp. SHEA,D.S.1973.A management-oriented study of bald eagle concentrations in Glacier National Park.M.S.Thesis.Univ.of Montana,Missoula. 78pp. SOUTHERN,W.E.1964.Additional observations on winter bald eagle populations:Including re- marks on biotelemetry techniques and imma- ture plumages.Wilson Bull.76(2):121-137. --.1967.Further comments on subadult bald eagle plumages.Jack-Pine Warbler 45(3):70- 80. SPENCER,D.A.,ED.1976.Wintering of the mi- grant bald eagle in the lower 48 states.Nat.Agr. Chern.Asso.Pub!.,Washington D.C.170pp. SPRUNT,A.,IV.1969.Population trends of the bald eagle in North America.Pages 347-351 in J.J. Hickey,ed.Peregrine falcon populations: Their biology and decline.Univ.of Wisconsin Press,Madison.596pp. --,AND F.J.LIGAS.1966.Audubon bald eagle studies,1960-1966.Proc.62nd Ann.Conv.Nat. Aud.Soc.,Sacramento,California.6pp. STALMASTER,M.V.1976.Winter ecology and ef- fects of human activity on bald eagles in the Nooksack River valley,Washington.M.S.The- sis.Western Washington Univ.,Bellingham. l00pp. STEENHOF,K.1976.The ecology of wintering bald eagles in southeastern South Dakota.M.S.The- sis.Univ.of Missouri,Columbia.148pp. ZAR,J.H.1974.Biostatistical analysis.Prentice- Hall,Inc.,Englewood Cliffs.620pp. Received 13 May 1977. Accepted 24 February 1978.