Portal de Periódicos da CAPES

A novel use of passive integrated transponder (PIT) tags as nest markers

2007; Association of Field Ornithologists; Volume: 78; Issue: 1 Linguagem: Inglês

10.1111/j.1557-9263.2006.00088.x

1557-9263

Travis L. Booms, Brian J. McCaffery,

Avian ecology and behavior

ABSTRACT Double-observer methodology requires independent collection of data to accurately estimate population parameters. Use of visual nest markers to facilitate matching, relocating, and monitoring nests as part of a double-observer study violates this assumption, but few reliable alternatives exist, especially when working with cryptic nests and high nest densities in homogeneous habitat. We used passive integrated transponder (PIT) tags to nonvisually mark the nests of ground-nesting birds at the Yukon Delta National Wildlife Refuge in western Alaska in a double-observer study of nest density. We marked 70 nests with PIT tags and naïve observers subsequently detected tags at 44 of 50 re-scanned nests (88% correct identification). Failed detections were likely due to either suboptimal tag orientation or tags falling through nest material, and such failures may be an inherent, but uncommon, feature of this nest-marking technique. PIT tags facilitated nest monitoring among independent observers, uniquely and reliably marked nests, provided a minimum of cues to potential nest predators, and allowed us to estimate densities in a double-observer framework while not violating assumptions. These tags should be useful in other studies of nesting birds where nonvisual, reliable nest markers are needed, and they provide a new tool for double-observer studies. El método de doble observador require el tomar datos de forma independiente para estimar con precisión parámetros poblacionales. El uso de marcadores visuales en nidos para facilitar el pareo de observaciones, relocalizar y monitorear nidos, como parte de un estudio de doble observador, viola la premisa básica. Sin embargo, hay pocas alternativas confiables, particularmente cuando se trabaja con nidos crípticos, alta densidad de nidos y habitat homogéneos. Utilizamos bitácoras electrónicas pasivas integradas (SPI) para marcar de forma no-visual nidos de aves que anidan en los suelos en el Refugio de Vida Silvestre de Yukon Delta. Esto como parte de un estudio de doble observador con el objetivo de determinar la densidad de nidos. Marcamos 70 nidos con las bitácoras de los cuales 44 o el 88% (de 50 nidos) fueron localizados por observadores sin adiestramiento en la identificación de los nidos.Los seis nidos que no fueron encontrados fueron posteriormente identificados una vez se discutieron asuntos particulares entre los observadores independientes. Las SPI facilitaron el monitoreo de los nidos por observadores sin adiestramiento, marcaron los nidos de forma confiable, y ofrecieron muy pocas pistas a depredadores potenciales. No menos importante permitieron estimar la densidad de los nidos utilizando la técnica de doble observador sin violar la premisa básica. Los SPI deben ser de utilidad en otros estudios de anidamiento de aves, en donde son necesarios marcadores que no sean visuales y además proveen de una nueva herramienta para estudios de doble-observador. Double observer/sampling research designs have long been used with numerous taxa to estimate population parameters (Anderson 2001). The method requires two or more independent observers or sampling occasions where detection of the item of interest is not influenced by its detection history by other independent observers or previous sampling occasions. However, when surveying the nests of ground-nesting birds in homogeneous habitat to estimate density, nests are typically marked to facilitate relocation and monitoring of nests and to permit comparison of results among sampling occasions or independent observers. Unfortunately, marking nests typically violates the assumption of independence in a double-observer research design. However, not marking nests may limit comparisons between observers and accurate estimates of population parameters. Therefore, an inconspicuous nest marker is needed that meets the assumptions of double-observer designs and permits accurate geographic and numerical comparisons of independently collected survey data. A modified Lincoln–Peterson estimator (Anthony et al. 1999, Thompson 2002) can be used to estimate the total number of nests or other items of interest, but detections must be classified into one of three categories: detected only by observer A, detected only by observer B, or detected by both observers. To correctly classify detections for analysis, nests must be marked. Assumptions of a double-observer study, however, preclude marking nests with conspicuous visual cues. These assumptions include: (1) the detection of the item of interest by one observer must be independent of detection by the other observer, and (2) detection probabilities for the observers are the same (Nichols et al. 2000, Thompson 2002). Both assumptions could be violated if conspicuous nest markers were used to identify nests. Nests marked conspicuously (e.g., with a pin-flag) upon initial discovery by one observer would probably be more likely to be discovered by the second observer. Thus, detections would not be independent. In addition, nest markers may serve as cues to predators. An increased probability of predation after marking the nest at initial discovery violates the second assumption because the second observer would have little chance of finding depredated nests (i.e., the detection probabilities of the two observers would differ). Because we needed to mark nests without violating the assumptions of the analysis, nest markers had to: (1) provide no visual cues to surveyors, (2) identify individual nests by assigning unique nest numbers, (3) be detectable by naïve surveyors (i.e., those with no previous experience with a particular nest) only after they found the nest, (4) facilitate coordinating nest monitoring among surveyors, and (5) provide no apparent cue(s) that predators could use to locate nests. Here we describe the use of passive integrated transponder (PIT) tags to inconspicuously mark nests and evaluate the performance of these tags in a double-observer sampling framework based on the above criteria. Our study was conducted at the Kanaryarmiut Field Station on the Yukon Delta National Wildlife Refuge in western Alaska (61°22′N, 165°07′W) from approximately 15 May–10 June 2003. The study site was a mosaic of dry and moist tundra interspersed with numerous ponds and lakes. Dominant vegetation included mosses, lichens, grasses, sedges, and low-growing willow (Salix spp.) and birch (Betula spp.). Shorebird nesting density exceeded 200 pairs/km2 (McCaffery et al. 2002, McCaffery and Ruthrauff 2004). On each of four randomly selected, 10-ha plots, three independent crews (a primary observer, a secondary observer, and a two-person rope-dragging team) conducted repeated nest searches during the period from clutch initiation through early incubation. Primary and secondary observers searched for nests on alternate days on specific plots, and each searched for 6–8 h/d for 10–11 d/plot. The rope-dragging crew visited each plot four times over a 3-week period, and searched for nests when primary and secondary observers were not present. No information about nest searching on specific plots was shared among crews; these blind searches permitted independent model-based estimates of nest density (Nichols et al. 2000). Once located, we marked nests using PIT tags (American Veterinary Identification Devices Company, Norco, California). PIT tags are microchips encoded with a unique identification number and originally designed to permanently mark (by subcutaneous injection) domestic or wild animals. The unique numerical code of each PIT tag is read by passing a hand-held scanner over the tag (or injection site). PIT tags and scanners cost $4.25 and $450 (US), respectively. The cylindrical PIT tags (12 × 2 mm) are coated with an acrylic exterior finish and prepackaged in small, rotating dispensers. PIT tags can be re-used in subsequent field seasons and, according to the manufacturer, can be accurately read for decades. The effective scanning distance varies with tag orientation. Tags oriented horizontally can be detected up to 6 cm away, and those standing vertically up to 3 cm away. We field-tested effective scanning distances and overall detection rates by randomly placing PIT tags in five of ten nest cups used by Western Sandpipers (Calidris mauri) in previous years. Four naive observers independently scanned each of the 10 nest cups and recorded the codes when PIT tags were detected. All observers detected all five microchips and identified each correctly. To mark a nest with a PIT tag, the observer rotated the dispenser to dislodge a tag, slid the tag into the lid of the dispenser, then scanned and recorded the unique identification number. The tag was then dropped into the center of the nest and the nest was scanned by passing the scanner just above the eggs to reconfirm the tag number. Observers relocated nests using detailed field notes including compass bearings from plot boundaries or conspicuous habitat features, written descriptions of the nest area, GPS coordinates, and sketches of the nest location. When observers found a nest they had not located previously, they scanned the nest to determine if it had been marked with a PIT tag. If a tag was detected, the observer recorded the tag number, but did not collected nest data (clutch size, presence of adult, and egg float angle/age) because those data had already been collected, thereby preventing unnecessary, duplicate nest visits by more than one observer. If no tag was detected, nest data were collected and the nest marked with a PIT tag. We marked 70 nests with PIT tags, including nests of Western Sandpipers (N= 30), Red-necked Phalaropes (Phalaropus lobatus; N= 16), Dunlin (C. alpina; N= 11), Willow Ptarmigan (Lagopus lagopus; N= 8), Rock Sandpipers (C. ptilocnemis; N= 3), and Black-bellied Plovers (Pluvialis squatarola; N= 2). Tags were not visible to observers because of their small size and resemblance to the fragments of lichen, moss, or twigs typically found in bird nests. Fifty of seventy marked nests were scanned by at least one subsequent naive observer, and tags were not detected at 6 of those 50 nests (12% failure rate). We subsequently were able to identify these six “missed” nests using descriptions of nest locations and GPS coordinates. Our results suggest that PIT tags could be useful in other studies of densely nesting birds where visual markers may affect nest success or violate analytical assumptions. Using the Mayfield method (Manolis et al. 2000) and considering a nest successful if ≥1 egg hatched, shorebird nest success on our four study plots was 35% (95% CI = 24%–52%). For Western Sandpipers, our most abundant species, nest success was 40% (95% CI = 24%–64%). Similarly, mean nest success for Western Sandpipers at the Kanaryarmiut Field Station over a 7-year period was 30.6 ± 3.3 (SE)% (Johnson et. al 2005). Because of the small size of PIT tags, the relative ease and speed of deployment and scanning, a protocol that eliminates direct human handling of tags, and the absence of any apparent effect on nest success, we believe that PIT tags provide few, if any, cues to potential nest predators. Additional study is, of course, needed to confirm this hypothesis. We failed to detect PIT tags at six nests. At five of these nests, tags may have fallen deep into the nest material, been oriented vertically and beyond the effective distance of the scanner, or both. Adults may have removed PIT tags from nests, but this seems unlikely because of their small size and resemblance to nest materials. In the sixth nest, two tags apparently fell from the dispenser simultaneously because two were found in the nest cup after the nest failed, and both had been assigned to the observer who marked the nest. In later lab experiments, two tags placed near each other sometimes cancelled each other's signal so neither was detected, possibly explaining why no tag was detected in the nest with two tags. Failed detections caused by unknowingly placing two PIT tags in a nest can be avoided by strictly adhering to the placement protocol described above. Failed detections due to suboptimal tag orientation or tags falling through nest material may be an uncommon, irresolvable feature of this nest-marking technique. After the field season, written descriptions of nest locations plus GPS coordinates allowed us to identify the six nests where PIT tags were not detected. Identification was possible because these nests were not close to other nests. In cases where nests are near each other, correct identification of nests with undetected PIT tags may not always be possible. In summary, PIT tags met our five study design criteria and were a simple and effective nest marker in our study. They provided independent observers and, perhaps, potential nest predators with no apparent cues and served as unique nest markers that allowed us to coordinate nest monitoring among surveyors and estimate nest densities. Alternative nest markers that may not violate double-observer assumptions include detailed descriptions of nest sites (Martin et al. 1997), GPS coordinates, and inconspicuous markers such as tongue depressors inserted into the ground at a predetermined distance and direction from the nest so that only the extreme tip is visible (S. Earnst, pers. comm.). Unfortunately, in areas where cryptic nests may be just a few meters apart, detailed descriptions and GPS coordinates would not eliminate all uncertainty about whether a nest had been found previously by the same or a different observer. This alternative would also make it difficult to facilitate nest monitoring among observers while not violating the independence assumption. Tongue depressor nest markers may leave more foreign odors near nests than the small, acrylic-coated PIT tags and detecting tongue depressors likely requires more time than scanning for PIT tags. Funding and logistic support for this project were provided by the U.S. Fish and Wildlife Service, Yukon Delta National Wildlife Refuge and its employees, and the University of Alaska Fairbanks. We thank C. Fitzpatrick, R. Hill, M. Swaim, H. Swenson, and H. Woodward for their nest searching efforts and participation. J. Bart, V. Johnston, and R. Lanctot of Arctic PRISM (Program for Regional and International Shorebird Monitoring) assisted in developing the original sampling design. G. Ritchison, an anonymous reviewer, and S. Earnst, in particular, provided helpful comments. Our use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.