Retention, Effect, and Utility of Tail-mounted Satellite-tracked Transmitters on Golden Eagles

Capture and marking studies deploying satellite-tracked transmitters (also known as Platform Transmitter Terminals [PTTs]) or Global System for Mobile Communications-GPS (GSM) packages on Golden Eagles have proliferated in recent years. U.S. Geological Survey (U.S.G.S.) Bird Banding Laboratory records indicate only 12 auxiliary marking authorizations for PTTs (or GSMs) on Golden Eagles were issued from 2006 to 2010 (D. Bystrak pers. comm.). However, from 2011 to 2015, auxiliary marking authorizations for PTTs (or GSMs) on Golden Eagles tripled for a total of 48 active authorizations by 2016 (D. Bystrak pers. comm.).

Backpack harnesses (e.g., Buehler et al. 1995) typically have been used to attach PTTs and other transmitter types to eagles. Although some researchers indicated lack of effect of backpack harnesses on eagles (Buehler et al. 1991, Hunt et al. 1996), others reported negative effects, particularly on Golden Eagles (Lockhart and Kochert 1980, Marzluff et al. 1997a, Gregory et al. 2003). More recently, no adult breeding Golden Eagles fitted with backpack PTTs in Sweden in 2012 (n = 12 eagles from seven pairs) produced young (Moss et al. 2014). The researchers attributed the total nesting failure of the tagged birds to “local food shortage” but did not address the possibility that failure may have been exacerbated by negative effects of the backpack-mounted telemetry system. Stahlecker et al. (2015) reported 55% of adult Golden Eagles (including three of seven territorial birds) died within 1 yr of being tagged with a backpack PTT and encouraged “closer scrutiny of possible negative effects of backpack PTTs on Golden Eagles.” In an ongoing analysis, Millsap may have detected a decrease in survival in a large sample of juvenile and immature Golden Eagles wearing backpack PTTs (B. Millsap pers. comm.). The evidence suggests backpack telemetry harnesses may negatively affect behavior, reproduction, and survival of Golden Eagles.

As a result, I tested retention, effect on reproduction, and utility for determining survival and range size of tail-mounted PTTs on a sample of Golden Eagles in southwestern Montana. I expected tail-mounted packages would have no or negligible effect on behavior, survival, or reproductive rate of tagged birds. Although Golden Eagles generally retain rectrices about 2 yr (Ellis and Kery 2004, pers. obs.), I anticipated a reduced retention (≤1 yr) of rectrices supporting tail-mounted PTT packages but accepted it as a trade-off to maximize the ethic of primum non nocere for eagles included in the study. However, I suspected that retention would at least be adequate to address the primary objective of the marking study, i.e., accurate estimates of survival (Harmata 2015a).

Golden Eagles were captured as free-flying birds with Coda net launchers (Coda Enterprises, Inc., Mesa, AZ U.S.A.) or padded leg-hold traps (Miner 1975). Eagles were tagged primarily in southwestern Montana roughly between 44.7°N (West Yellowstone) and 47.5°N (Great Falls), and between 110.8°W (Ringling) and 112.3°W (Twin Bridges) between February 2011 and January 2015. However, some tagged eagles ranged widely throughout the state and from southeastern Idaho and central Wyoming to northwestern Alaska.

Emphasis was placed on capturing adult Golden Eagles resident in Montana to study survival. Only one member of an adult, mated pair on territory was PTT-tagged. Younger age classes were captured opportunistically or in territories after target adults were tagged. Captured eagles were sexed by size (Harmata and Montopoli 2013) and aged by plumage (Humphrey and Parkes 1959; Table 1 in Bloom and Clark 2001, Ellis 2004). Age classes were juvenile, immature (Basic I and II plumage), subadult (Basic III plumage), or adult.

Golden Eagles were tagged with one of two Argos-tracked PTT configurations (Fig.1). One was a 32-g, 63.5 × 22 × 18 mm, battery-powered Doppler unit (NorthStar Science and Technology, LLC, King George, VA U.S.A.) attached ventrally to central rectrices. The second was a 22-g, 63 × 20 × 17 mm, solar-powered GPS PTT (Model 22GPS, Northstar Science and Technology, LLC, King George, VA U.S.A.) attached to central rectrices dorsally to expose solar panels to sunlight. Duty cycles were 6 hr on, 2 d off for both units. Solar units provided up to six hourly GPS locations daily between 0800 and 2200 H MST with adequate light. GPS units also provided Doppler data. Antennas on both units were 195 mm long and sutured adjacent to the ventral rachis, and both units were attached with mechanisms to distribute package weight between both rectrices while allowing feathers to spread and contract (Fig. 2). All eagles tagged with tail-mounted PTTs were also marked with a U.S.G.S. pop-rivet leg band, and blood and feather samples were collected from some. I suspected that ventrally mounted systems would be retained longer than dorsally mounted systems, mostly due to visual prominence and proximity of the dorsal package, especially when the eagle accessed the uropygial gland during preening. Hence, I invested in fewer solar GPS-capable PTTs.

Behavior and movements of tagged eagles permitted confirmation of residency and mated status. Eagles present within Montana between 15 April and 15 October were considered residents. Eagles located outside Montana at any time during tracking were considered migrants. Resident adults observed engaged in breeding behavior or commonly associated with nest sites or another adult eagle of distinctly dissimilar size were considered territorial and mated. Resident adult eagles not commonly observed associated with nest sites or another adult eagle of distinctly dissimilar size and ranged widely were considered floaters (Brown 1969, Hunt 1998) and were assumed to be unmated. Nonadult eagles were classified as either resident or migrant based on above criteria.

Lack of movement of a PTT during one duty cycle as determined by Argos activity sensor data indicated either a dead or debilitated eagle, or a shed (i.e., molted or forcibly removed by eagle) PTT and stimulated recovery effort by vehicle or on foot. At the most recent telemetry location, visual scanning and a Communications Specialists, Inc., AL-1 ARGOS PTT LOCATOR™ receiver programmed to receive 401.650 mHz frequency with an omnidirectional or two-element, handheld directional antenna were employed during the “on” portion of the PTT's duty cycle. Patience was required when locating PTTs, as signal pulse rate was only 1/min for Doppler units and 1/hr for solar units.

Fate (alive/dead) of eagles whose PTTs were recovered was obvious but fate of eagles with PTTs that stopped transmitting data or were unrecovered was not. However, situational criteria associated with cessation of movement or PTT failure (e.g., subsequent observations of eagles of appropriate age/sex at habitual perches and missing central tail feathers) allowed for assessment of fate. Unrecovered PTTs that stopped moving later than early July, especially those in Alaska, were assumed to have been molted by an eagle that continued to survive because mode of Golden Eagle molt occurs in early summer (Jollie 1947, Ellis and Kery 2004, A. Harmata unpubl. data).

Range size as a function of time monitored (days tracked) was analyzed in an attempt to determine if retention of tail-mounted PTTs was adequate to provide representative annual range-size descriptions for territorial eagles. Only GPS-tracked eagles were analyzed because GPS locations (termed “fixes”) are inherently more accurate than Doppler fixes. The cumulative proportions of total range size (using minimum convex polygons [MCP; Mohr 1947] in km²) per number of days tracked were used to compare area use among four territorial males and one territorial female Golden Eagles tracked by satellite.

Latitude and longitude of each PTT-generated fix were tabulated sequentially in PTT Tracker (GeoTrak 2011) for eagles with ≥25 GPS fixes. Sequential fix locations were exported as .kml files and imported to Google Earth™ ver. 7.1.2.2041 and displayed. Minimum convex polygons were constructed around peripheral fixes of each plot using the polygon utility in Google Earth. To quantify an eagle's expanding range recorded over time during tracking, polygon files were exported to Earth Point (http://www.earthpoint.us/Shapes.aspx) and MCP area was calculated sequentially at 5-d intervals using the Polygon Area utility.

I compared differences in group retention times (d) by parametric procedures. I accepted P-values ≤ 0.11 as indicating statistical difference. This minimized potential type II error due to small sample size. Geometric means (GM) also are presented to minimize effect of outliers.

Twenty-seven Golden Eagles were tagged with tail-mounted PTTs (Table 1). Adults made up 63% of tagged eagles, subadults approximately 15%, immatures 7%, and juveniles approximately 15%; 52% were females and 48% were males. Territorial eagles made up 76% of adults and 18% of adults were floaters. Migrants composed 22% of tagged eagles; three summered in Alaska, two in British Columbia and one (identified as “Bull Sitz” [Table 1], after the location where it was captured) likely wintered in the continental United States at latitudes lower than 43°N (Table 1). One territorial adult female was previously banded as a 1-yr-old, 10 yr prior and 6 km NNW of the 2011 capture site (“Ringling Gal,” Table 1). An adult male was previously banded as a 1-yr-old, 22 yr prior and 37 km NE of the capture site (Harmata and Restani 2015). One territorial adult male was developing both tail feathers adjacent to a central rectrix when captured in December and could not be tagged with a tail-mounted PTT.

Of 27 PTTs attached to rectrices of Golden Eagles, date of shedding was known for 16, of which 13 (81%) were recovered (Table 1). All but one recovered PTT were retrieved within one duty cycle of searching. All recovered solar units were located without telemetry by visually searching the coordinates where movement ceased. Four (15%) PTTs failed or were suspected of failing on the eagle and were not recovered, and two PTTs that ceased movement failed before they could be recovered. Observations of four eagles (“Bear Crk,” “Camp,” “Ruby Dude,” and “N Cottonwood,” Table 1) without central rectrices, each within their respective home ranges, soon after PTT movement ceased, confirmed that their PTTs were shed.

The mean length of time tail-mounted PTTs indicated movement was 132.8 d (n = 27, SD = 84.5 d, range 7–381 d, GM = 101.0 d). Mean length of time eagles retained tail-mounted PTTs where date of molt or removal was confirmed was 143.7 d (n = 16, SD = 96.9 d, range 15–381 d, GM = 108.6 d). Females tended to retain tail-mounts (GM = 138 d) longer than males (GM = 94 d; t₁₄ = 1.7, P = 0.11, Fig. 3) and dorsally mounted PTTs tended to be retained longer (K = 2.5, GM = 181 d) than those mounted ventrally (GM = 86 d; t₁₄ = 1.76, P = 0.09, Fig. 3). Attachment techniques evolved and improved over the course of study and later attachments (Fig. 2) tended to be retained longer (Table 1).

Of 13 recovered tail-mounts, six (46%) were forcibly removed by eagles: five by males and one by a female. All packages that were forcibly removed were 32-g ventral mounts. Two males broke off the mounting tabs attached to rectrices where they were soldered to the PTT case, and four others broke the supporting feather shafts proximal to the PTT mounting point. Seven (54%) of the 13 recovered tail-mounts were found attached to at least one molted rectrix, and six of these seven were found attached to both rectrices. Females tended to molt PTTs in early to mid-summer (June, July) but males were more erratic, possibly related to age, e.g., an adult male molted the PTT in May and immature male molted the PTT in October. The dorsal tail-mounted PTT of “Ruby Dude” was molted in July 2015 but not recovered until 7 mo later (Table 1). Initial intensive search at the coordinates of the last GPS locations within 3 d after movement ceased failed to locate it. The eagle survived, as no carcass was located at the site and an adult was subsequently observed in the home range without central rectrices. I surmised that the molted PTT was not detected on the initial search because a corvid or fossorial mammal cached it soon after it was dropped. It later emerged or was excavated because it was easily found at the same GPS coordinates on a subsequent incidental search several months later. Of eight unrecovered PTTs (four from males and four from females) where situational criteria indicated molt from a live bird, seven occurred in July. One territorial female was still wearing her ventrally mounted PTT when it failed in late September (“Yellow Dog,” Table 1).

No Golden Eagle wearing a tail-mounted PTT was confirmed to have died during the study. However, one mortality was highly suspected (“Swingley,” Table 1). Activity characteristics recorded prior to the abrupt cessation of movement, location, and landowner refusal to allow access to recover the PTT all strongly suggested a dead eagle, likely not as a result of PTT effects or natural causes. All other unrecovered PTTs (Table 1) that ceased movement were likely molted or shed as suggested by location, prior activity, or time of year. Adult annual survival rate estimate obtained during the study from which retention data used here were drawn was 86% (CI = 60–100%) (Harmata 2015a).

Of 13 territorial adults tagged with tail-mounted PTTs, eight (62%; six males and two females) were known to attempt breeding the year they were tracked via telemetry. Two males (15%) failed in their breeding attempt and six birds (46%; four males and two females) produced young. Two females did not attempt to breed, two PTTs were shed (one female) or stopped transmitting (one male) too early in the season to determine breeding activity, and breeding activity was undetermined for one male.

Territorial Golden Eagles tracked with GPS-capable PTTs all had 50% of their total MCP area recorded by 71 d (Fig. 4). Logistic curves of three of the four GPS-tracked eagles indicated that 99% of the MCP area was recorded by 140 d (Fig. 4). However, the slope of the logistic curve of “Hiline Dad” suggests MCP area was still expanding at cessation of tracking (79 d; PTT shed).

Despite a low overall duration of tail-mounted PTT retention (where date of molt or removal was confirmed) of just over 3 mo (but up to 10 mo), results indicated that tail-mounts were adequate for estimating survival. Adult annual survival rate estimate during this study was consistent with a recent annual survival rate estimate by U.S. Fish and Wildlife Service (U.S.F.W.S.) for after-third-year (i.e., adult) Golden Eagles of 87%, (CI = 84–86%; U.S.F.W.S. 2016). Both analyses were conducted in Program MARK (White and Burnham 1999) and U.S.F.W.S. estimates were based on dead-bird band recoveries during 1968–2014.

In addition to a lack of major effect of tail-mounted PTTs on survival of adult Golden Eagles, lack of effect on reproduction also was demonstrated by the similarity of tagged adult breeding activity with Golden Eagles included in recent population surveys in southwestern Montana (Harmata 2015b). Specifically, the proportion of territorial adults with PTTs (each representing a pair) attempting to breed and breeding successfully (61% and 46%) was not different from the proportion of breeding attempts and breeding success of the Golden Eagle population surveyed (231 occupied territories) in southwestern Montana from 2010 to 2015 (69% and 36%; Z < −0.60, P > 0.23; Harmata 2015b).

Some researchers attributed varied effects of backpack-mounted telemetry systems on birds primarily to differences in transmitter mass, shape, and size and their effects on energetics and aerodynamics (Obrecht et al. 1988, Gessaman et al. 1991a, 1991b, Pennycuick et al. 2012). However, Michael et al. (2013) and Peniche et al. (2011) found significant skeletal and soft tissue damage to birds wearing backpack harnesses, inducing Michael et al. (2013) to state “These findings have implications for all avian species deployed with backpack harnesses.” I suspect that tolerance of backpack-mounted telemetry systems by Golden Eagles is likewise more directly related to discomfort and restriction of movement imparted by the encumbering harness rather than the PTT hardware configuration. Tolerance is further influenced by biologist talent (or lack thereof) in correctly fitting the harness.

Despite the absence of effects on survival and reproduction and the lack of observed injury or abrasion imparted by tail-mounted PTTs used here, effects were not totally benign. Nearly half of recovered tail-mounts were forcibly removed by eagles and males tolerated packages less well than females. Two males actually broke the PTT casing off the soldered mounting tabs, and the other PTTs forcibly removed were by the eagle breaking the supporting feather shafts proximal to the PTT mounting point. Clearly, they were an annoyance to some eagles, particularly males. I expected that ventral tail-mounts on Golden Eagles were less likely to affect behavior than any mounted dorsally, but, curiously, no dorsal tail-mounted PTT was forcibly removed. Ventral mounts removed by force were possibly due to the weight (32 g) exceeding a threshold of tolerance on critical aerodynamic structures (although <1% of body mass), physical size of the package (note dimensions, above), or improper attachment creating discomfort or an impediment to preening. Despite the annoyance factor, the effects were not debilitating, injurious, an impediment to reproduction, or fatal.

Over half of tail-mounts recovered had been molted. Females tended to molt PTT supporting feathers in early summer (June, July) but one female wore her PTT from January until at least late September (“Yellow Dog,” Table 1). Males were more erratic, possibly related to age and sex-specific intolerance of the package. Adult males shed PTT-supporting feathers in May and July and immature males in April, July, and October. Of unrecovered PTTs where situational criteria indicated shedding from a live bird, nearly all occurred in July. These observations were consistent with those of rectrice molt recorded for captive (Jollie 1947, Servheen 1976, Ellis and Kery 2004) and wild (Bloom and Clark 2001) Golden Eagles. However, Ellis and Kery (2004, Table 1) noted that a captive Golden Eagle molted both rectrices in the same year only twice over 15 yr, but six of the seven PTTs recovered with feathers attached had both rectrices, suggesting that PTT mounts used here may have induced abnormal, simultaneous expulsion of both rectrices.

Logistic regression analysis of cumulative MCP area by monitoring duration showed four territorial eagles with tail-mounted GPS PTTs had 99% of their total range recorded within 140 d of tracking (Fig. 4). This was approximately the mean length of time eagles retained tail-mounted PTTs where date of molt or removal was confirmed (144 d). MCP ranges of territorial Golden Eagles in Montana (x̄ = 1032 km², Harmata 2015a) were not dissimilar in size to MCP ranges of territorial Golden Eagles in Idaho during the nonbreeding season (x̄ = 317 km²; Marzluff et al. 1997b) and monthly ranges of Golden Eagles in the Mojave Desert (309 km²; Braham et al. 2015). Idaho habitats were primarily shrub-steppe grassland ≤1000 m in elevation, with moderate topographical relief, and abundant patches of shrubs suitable for lagomorphs (Marzluff et al. 1997a). MCP ranges of territorial Golden Eagles in Montana were also comparable in size to multiyear home ranges of territorial Golden Eagles in Washington and Oregon (Watson et al. 2014), where most nests were <1000 m in elevation and eagle prey was dominated by mid- to large-size mammals and upland game birds. However, range sizes were calculated by vastly different methods (c.f., Methods in this study and Watson et al. 2014). Regardless, analysis suggests dorsal tail-mounted PTTs may be adequate for describing annual ranges of resident, territorial Golden Eagles if attached at dates and with techniques recommended here.

Future deployments of tail-mounted PTTs on Golden Eagles should be attached as soon after 1 December as possible (although one adult male had a central rectrix still growing in early December) to promote maximum retention time. However, molt phenology likely varies with latitude and attachment timing may need to be adjusted accordingly. Capture and handling during late autumn–early winter at higher continental latitudes (ca. 45°N) or autumn at lower latitudes (ca. 36°N) may be least likely to negatively affect breeding individuals, or to be inhibited by absent/developing rectrices. To minimize any effect of added mass on premature induction of molt of the supporting rectrix, especially on males, PTT mounts should be ≤22 g and include a support brace (pin) from the central mounting rectrix to an adjacent rectrix (not necessarily central) that also permits spreading of feathers (Fig. 2). PTTs should be attached dorsally (Fig. 1) and if technology permits, be of smaller physical size to minimize the “annoyance” factor. Antennas should be attached ventrally, adjacent to the feather shaft on both dorsal and ventral mounted systems to restrict access by the preening eagle.

This study could not have been conducted without the enlightened cooperation and permission of a vast number of private landowners in southwestern Montana. Beth Madden, Eric Johnston, and John “JT” Thornburg, in addition to assisting in capture and tagging, demonstrated a truly extraordinary ability to locate and recover shed PTTs. Jeremiah Smith, Marco Restani, and Kevin Frey assisted in capture and tagging of eagles. J. Rotella and John Marzluff conducted and interpreted survival analysis and results. Valuable comments on early drafts of this report were provided by Robert “Murf” Murphy and two anonymous reviewers. Thanks to Danny Bystrak and the U.S.G.S. Bird Banding Laboratory for providing auxiliary marking statistics presented here. Funding for the primary study from which these data were drawn was provided by PacifiCorp, as administered through the National Fish and Wildlife Foundation. Research was conducted under Federal Bird Banding Permit #20357, Montana Dept. of Fish, Wildlife, and Parks Scientific Collectors Permit 2015-26 and Montana State Univ. IACUC Authority 2013-46.