1978-2000 Surveys We began conducting bald eagle nest surveys in Oregon in 1978. Isaacs et al. (1983) described methods used to initiate the project and conduct the 1978-1982 surveys. Even though the project was renewed annually, there was never a long-term study plan because renewal was not guaranteed. There was consistency in the project because we were project leaders and collected much of the field data all years. We (the authors and others employed to work on this or related projects) made most of the observations at nest sites early in the study (e.g., 90%, n = 566 in 1980), but our share of observations declined (e.g., 44%, n = 1,300 in 2000) as the number of nest sites increased, and we relied more on others because of time constraints (e.g., >270 people contributed in 2000). The number of nest checks used to summarize annual surveys increased from ~201 in 1978 to ~1,300 in 2000. Nearly all nest checks (94%, n = 11,398 for 1971-1997) occurred from 1 March through 31 July (Figure 2). The average number of nest checks per site surveyed was 3.7 (n = 4,588) for 1978-2000. That number was larger than the minimum number of visits required to complete the survey, because some sites were checked many times during 1 or more breeding seasons for management purposes, or because of observer interest.
We chose to survey the entire population rather than a sample because we were often asked about the breeding status at nests by resource managers or landowners who needed the information to comply with Federal (U. S. Department of the Interior 1978) and State Endangered Species Acts (Oregon Department of Fish and Wildlife 1997), and Oregon Forest Practices Rules (Oregon Department of Forestry 1997). It was impossible to predict which nests would be involved in projects in a given year. Consequently, we wanted to know nest condition and nesting status at as many sites as possible, annually. As the nesting population increased, we modified methods in order to complete surveys and annual reports on time. Survey schedules, use of volunteers and biologists from other organizations, areas covered by air and ground surveys, record keeping systems, ways of collecting reports from others, and annual data summary methods were adapted to fulfill our objectives and take advantage of available technologies.
Standard site names based on nest tree names and nesting histories were used to designate breeding areas. Nest tree names were assigned based on geographic location; at some breeding areas, all nest trees had the same site name; at other sites, nest trees had different names. Standard site names were consistent within years; some changed between years as our knowledge of breeding area boundaries changed, or new nest trees were discovered within existing breeding areas. New breeding areas received new and unique standard site names upon discovery. From 1978-1992, each nest tree was assigned a six-digit number (e.g., 600-001). The first 3 digits designated region (e.g., 600 = Klamath Basin and Southern Oregon), and the second 3 digits designated nest tree number within the region, and were assigned sequentially as nests were discovered. In 1993, the region designation (first 3 digits) was dropped, and nest trees were reassigned numbers 1 through 600, because 600 nest trees had been cataloged by the end of that field season. After 1993, nest tree numbers were assigned sequentially, beginning with 601, as nests were discovered. The last nest tree discovered in 2000 was numbered 991. Nest tree numbers remained the same even if a nest fell completely apart and a new nest was built at a different location in the same tree several years later. That enabled us to track long-term history of use at individual nest trees.
We did not capture and mark eagles for this study. There were marked individuals in the population from other studies (Frenzel 1985, Garrett et al. 1993), but they were seldom observed during annual surveys. Band returns for bald eagles banded by Frenzel (1985) and Garrett (1993), and reports of birds marked elsewhere that were observed in Oregon were used to evaluate movements.
Terminology Our terminology and data summary methods were based on Postupalsky (1974, 1983), who recommended standard terminology for describing status of nests and breeding areas, and standard criteria for calculating reproductive success. The basic reproductive unit for bald eagles is a breeding pair. The area used by a breeding pair for nesting and raising young can be referred to in several ways: "territory", "breeding territory", "nesting territory", "breeding area", "nesting area", 'breeding site", "nest site", "site", etc. The term "territory" implies an exclusive defended space within a species home range (Lincoln and Boxshall 1990:377), and assumes occupancy. We did not determine defended spaces, or home ranges, and had unoccupied (therefore, not defended) breeding areas. Consequently, we used "breeding area", "nesting area", 'breeding site", "nest site", and "site", interchangeably to describe the area used by a breeding pair for nesting and raising young, and did not use terms containing the word "territory."
Definition of the term "occupied", and criteria used to determine the number of young produced at successful sites are open to interpretations that can bias results of nest surveys. Postupalsky (1983) noted that the term "occupied" has been applied in different ways by different authors, and concluded that only nests with evidence of actual pairs be counted as occupied. Postupalsky (1983) and Steenhof (1987) pointed out the difficulties of being positive about the number of adults in a breeding territory, especially when nest checks are brief, and few in number. We chose to categorize sites as occupied on the basis of one adult and a nest in a breeding area because: (1) most nest checks were brief, often less than 10 seconds during air surveys; (2) only 2 vists per site were planned, although follow-up visits were often used when primary visits were inconclusive; (3) poor access and rugged terrain hampered ground searches for adults; (4) air searches for adults were time consuming, expensive, and sometimes dangerous, especially in rugged terrain; (5) evidence of recent nest use was difficult to assess, especially from the ground by inexperienced observers, and from the air when nests were within the canopy of a live tree, and (6) often, only 1 adult was observed, even at sites where young were raised. Consequently, our number of occupied sites may include single, unpaired adults at nest sites, but we could not distinguish those sites using our methods. This should be taken into account when comparing our results to studies that required 2 adults for occupancy. For instance, our number of occupied sites (minimum population) could be overestimated, because we counted sites occupied by single adults as part of the nesting population, and nesting success and productivity could be underestimated, because sites occupied by single adults became failed nesting attempts.
Ideally, nesting success is based on the number of young raised to fledging (Steenhof 1987). Postupalsky (1983) and Steenhof (1987) noted the difficulty of determining how many young fledge when surveys involve just 2 visits per pair for many breeding pairs distributed over large areas. Postupalsky (1983) recommended counting young when adults are no longer brooding consistently, and young are large and dark (partly featherd to feathered). Our goal was to determine the number of feathered nestlings, but occasionally downy nestlings were present during the last visit to a site. When this happened, we included the downy chicks in the number of young produced. Consequently, our number of young produced could be an overestimate if there was mortality between the downy and feathered nestling stages at those sites. On the other hand, it is difficult to accurately count the number of downy chicks because they are small and relatively inactive compared to feathered chicks, and often are hidden by brooding adults. Those conditions would result in the number of downy chicks being underestimated. The error caused by counting downy chicks was probably minor, because the above errors would tend to cancel one another, and because the number of outcomes that included downy chicks was small; i.e., 1.78% of the total number of outcomes determined (n = 4,044) from 1979-2000.
Ground Surveys The nest survey protocol we provided to cooperators summarizes our methods for checking nests from land or water: Nests should be observed from a distant location that does not disturb the eagles. We recommend observing with a spotting scope from a distance greater than 800 m. Our objective is to determine the outcome of nesting for every pair of bald eagles in Oregon and along the Washington side of the Columbia River. This requires at least two visits to nest sites during the breeding season. The first visit is to determine if the site is occupied, the second visit is to determine the outcome of nesting. The best timing for the first visit is late March or early April. If no eagles are observed during that visit, the site should be checked weekly until one or more adult bald eagles is observed, or 15 May, whichever comes first. The second visit should be as near to 1 June as possible. If feathered nestlings are observed, and you are sure how many there are, or you are sure that the nesting attempt failed, no further visits are required. If the number of nestlings is uncertain, downy nestlings are observed, or outcome is uncertain, additional visits are required. Return within a few days if nestlings are feathered and you need to verify the number of nestlings. Return weekly until outcome is determined if outcome is uncertain. Return in 4 weeks if nestlings are downy during the early June visit. The goal is to determine the number of feathered nestlings. Feel free to determine and report fledging dates. If you know that a pair are unusually early or late nesters, adjust the schedule accordingly. The earliest nesting pairs begin incubating in late February; late nesters begin incubating in late April. Initiation of incubation has never been documented after 1 May.
Cooperators were asked to complete a form we provided for each nest visit, and return it to us by U.S. mail; or to report their observations by phone, e-mail, copies of field notes, or agency data forms. We asked that cooperators provide details on the observation, rather than their interpretation of the data. That enabled us to review observations and keep interpretations consistent, and resulted in an archive of original observations that future researchers may find useful. Postupalsky (1983) emphasized the importance of keeping raw, uninterpreted data.
Any reporting method was acceptable as long as the following information was included: site name, survey date, observer name and phone number, survey method (ground, boat, fixed-wing aircraft, helicopter), nest tree number if there was evidence of eggs, eagle use of nest (e.g., new material present, eagle(s) nestbuilding, eagle in incubating posture, etc.), number of adults observed, number of nestlings observed, and status of nestlings (downy, 0-2 weeks old; partly feathered, 2-7.5 weeks old; feathered, 7.5+ weeks old; fledged). In addition, we asked for details on nest and nest tree condition, age of other eagles observed, behavior of all eagles observed, and any other observations pertinent to bald eagle ecology. Nest site names and nest tree numbers came from our annual reports (most recent = F.B. Isaacs and R.G. Anthony. 2000. Bald eagle nest locations and history of use in Oregon and the Washington portion of the Columbia River Recovery Zone, 1971 through 2000. Annual report. Oregon Cooperative Fish and Wildlife Research Unit, Oregon State University, Corvallis, Oregon, USA). We requested that observers send detailed maps of nest locations not cataloged previously (new nests). We recorded similar information in field notes when checking nests from the ground. We always searched for and mapped new nests during ground and air surveys.
Air Surveys Air surveys were conducted by fixed-wing aircraft (Cessna 172, 180, or 185) from 1978 until 1989, when a plane crash convinced us that helicopters were safer. Helicopters (Bell 206B and Bell 206L) were used from 1990-2000. Nest visits by air were as brief as possible, and we stayed as far as possible from nests and eagles, while striving to achieve the goals of the survey (i.e., find incubating adults during the first visit; count nestlings during the second visit). Watson (1993) reported on bald eagle behavior during helicopter surveys of nests in western Washington. He recommended moving approachs rather than hovering, staying >60 m from nests and eagles, and keeping time at nests to <10 seconds. Generally, we followed those recommendations, and we planned approaches and departures to avoid passing between adult eagles and the nest in use. We also changed our flight path to avoid close encounters with flying eagles. We never recorded an obvious attack on survey aircraft by eagles. That contrasts with Watson's (1993) report that 11% of flushed eagles (n = 97) approached the helicopter to attack. Our caution when approaching nests or encountering flying eagles may have reduced the likelyhood of attack, or there may have been behavioral differences between the populations.
Air surveys were conducted by a pilot and 1-5 observers; the number of observers varied with aircraft size, and personnel available. The pilot concentrated on flight safety, positioning the aircraft for the best view of the nest, and avoiding flying and perched eagles. Passengers observed nests by eye or with binoculars, looked for and alerted the pilot about perched and flying eagles, wrote field notes, and read maps. Pilot and passengers searched for and communicated the locations of potential hazards to aircraft (e.g., power lines, other aircraft, towers, etc.), and searched for new nests. Experienced pilots were valuable because their knowledge of nest locations, and eagle habits improved flight efficiency (e.g., they were often the first to spot nests or eagles). Early in the project, all nests were checked by air, then follow-up visits were made on the ground. In 1980, 30% of nest checks (n = 547) were by air. As the nesting population increased, we used air surveys to check areas with large concentrations of nests, or difficult surface access; ground surveys were used to visit nests not checked by air, and to follow-up on inconclusive air surveys. In 2000, 35% of nest checks (n = 1,405) were by air.
Air survey timing initially was based on nest surveys conducted in Washington (Grubb et al. 1975), then were modified to fit local nesting phenology (Isaacs et al. 1983). Air survey dates were scheduled as close to 1 April as possible for the first nest checks, because that was within the period when most pairs were incubating (Figure3a), and as close to 1 June as possible for the second visit, because many nestlings were feathered, and that was just before fledging began at sites with early nesters. Flights were sometimes rescheduled due to bad weather. In 2000, the Columbia River downstream of Portland, the Willamette Valley, and the Klamath Basin were flown twice (late March or early April, and early June), and the Umpqua River downstream of Roseburg, and the Coos Bay area were flown once (early June). Air survey notes included: site name, nest number (if needed), number and age of eagles observed, nest status, status of nestlings, locations of "new" nests, and other pertinent information.