2013 SOUTHEASTERN NATURALIST 12(2):399–412 Effects of Radio-transmitter Methods on Pileated Woodpeckers: An Improved Technique for Large Woodpeckers Brandon L. Noel1,2,*, James C. Bednarz1, Mark G. Ruder3,4, and M. Kevin Keel3,5 Abstract - We captured and radio-marked 64 Dryocopus pileatus (Pileated Woodpecker) in bottomland hardwood forests from February 2007 to June 2010. At least 12 (35.3%) of the first 34 birds radio-tagged died within 43 d of capture (x̅ = 8.2 d). Thus, we adjusted our radio-attachment techniques adaptively from a figure-eight harness to a tail-mount, and reduced handling times to minimize stress on woodpeckers. In 2009 and 2010, after the change in attachment type and modified handling protocol including a reduction of handling time (from ca. 1 h to 30 min), all 30 radio-marked birds (100%) survived the entire field season (≥3 mo). These data suggested that Pileated Woodpeckers, and perhaps other large woodpeckers, have an increased risk of death when tagged with figure eight harnesses, handled for longer periods and more obtrusively, and captured on days with relatively cold temperatures. We submit that future telemetry on this species or other large woodpeckers should not employ the figure-eight harnesses and should strive to minimize handling time and disturbance. We recommend that other ornithologists observing higher than expected mortalities possibly related to handling birds or transmitter attachment publish this information to minimize the adverse impacts on birds during future research. Introduction Birds are routinely marked with radio and global positioning system (GPS) transmitters (Kenward 2001), as many studies have indicated that transmitters have minimal to negligible effects (e.g., Anich et al. 2009, Neudorf and Pitcher 1997, Powell et al. 1998, Sykes et al. 1990, Vukovich and Kilgo 2009). However, there continues to be concern, that in certain situations, transmitters can alter behavior and increase stress (Bull 2001, Gilmer et al. 1974, Odom et al. 1982, Perry 1981, Whidden et al. 2007). Thus, further effort has been put into understanding the potential negative consequences of marking wildlife (Barron et al. 2010, Ponjoan et al. 2008). Central to this debate is whether transmitters increase mortality of subject animals. Unfortunately, few studies publish mortality rates of birds due to the negative effects of radio transmitters. The majority of these reports have been on 1Department of Biological Sciences, Graduate Program in Environmental Sciences, Arkansas State University, State University, AR 72467. 2Current address - Integrated Environmental Science, Bethune-Cookman University, Daytona Beach, FL 32114. 3Southeastern Cooperative Wildlife Disease Study, College of Veterinary Medicine, University of Georgia, Athens, GA 30602. 4Current address - USDA, Agricultural Research Service, Arthropod-Borne Animal Diseases Research Unit, Manhattan, KS 66502. 5Current address - UC Davis, School of Veterinary Medicine, Department of Pathology Microbiology and Immunology, Davis, CA 95616. *Corresponding author - bnoelmarinebio@hotmail.com. 400 Southeastern Naturalist Vol. 12, No. 2 game species such as waterfowl or gallinaceous birds (e.g., Conroy et al. 1989, Cox and Afton 1998, Höfle et al. 2004, Holt et al. 2009). Importantly, researchers may be reluctant to publish studies that report high mortality rates or negative impacts on species of conservation concern (Putman 1995) related to methodology. However, publishing the potential adverse effects of techniques would help future research avoid similar problems and reduce negative impacts on subject animals and their populations over the long-term (Ponjoan et al. 2008). Studies that have reported adverse effects of radio transmitters on marked individuals have emphasized the importance of handling and restraint time, age of the target individual, and trapping methods (Höfle et al. 2004, Nicholson et al. 2000, Ponjoan et al. 2008, Spraker et al. 1987). Dryocopus pileatus L. (Pileated Woodpecker) is a resident species in forested habitats in North America, and only one published paper has reported possible impacts on this species due to radio-transmitter attachments. Specifically, Bull (2001) found disparity between annual survivorship of color-banded (64% based on re-sighting data; Bull and Meslow 1988) and radio-marked Pileated Woodpeckers (47%, based on Kaplan-Meier models; Bull 2001), and she suggested that the backpack harness/transmitter might have contributed to the increased mortality. Bonar (2001) found approximately 60% annual survivorship (based on Kaplan-Meier models) for radio-marked Pileated Woodpeckers using a backpack- harness technique, and concluded these survivorship estimates were normal based on limited information in the literature. However, other woodpecker species have been shown to be negatively impacted by carrying transmitters (Hooge 1991, Odom et al. 1982, Rolstad and Rolstad 1995). In 2007, we began a 4-year project studying the ecology of the Pileated Woodpecker in bottomland hardwood habitats in southeastern Arkansas. As one of our original objectives was to examine the habitat and spatial use of Pileated Woodpeckers, we decided to use radio-telemetry. Based on experience and review of the literature, we elected not to use backpack harnesses because of potential adverse impacts reported with this attachment technique on woodpeckers (Bull 2001, Odom et al. 1982). Thus, we initially employed the figure-eight harness (Rappole and Tipton 1991) to attach transmitters on Pileated Woodpeckers because work by Ripper et al. (2007) indicated no adverse impacts on Picoides villosus L. (Hairy Woodpecker) and J. Bednarz (unpubl. data) found no negative impacts on two Picoides borealis Vieillot (Red-cockaded Woodpecker) radiotagged with this technique. However, when our radio-tagged birds started to exhibit higher than expected mortality, we abandoned the figure-eight harness and switched to a tail-mount attachment (Kenward 2001). Investigating radiotransmitter effects was not a primary objective in our study, but we experienced high bird mortality rates during the first 2 years of our study. Therefore, we used our data based on both these attachment methods here to examine the variables most likely influencing the survival of radio-marked Pileated Woodpeckers in bottomland hardwood forests. Importantly, we provide information that may help researchers reduce capture-related morbidity and mortality in Pileated Woodpeckers and other birds. 2013 B.L. Noel, J.C. Bednarz, M.G. Ruder, and M.K. Keel 401 Methods Project background This study was conducted in eastern Arkansas (ca. 34°04'N, 91°04'W) in the Mississippi Alluvial Valley. This habitat includes cypress-tupelo (Taxodium- Nyssa) swamps, sweetgum-oak (Liquidambar-Quercus) forests, and intermediate forest associations between these 2 forest cover types. Fieldwork was conducted in the Cache River National Wildlife Refuge, White River National Wildlife Refuge, Dagmar Wildlife Management Area, and Wattensaw Wildlife Management Area in eastern Arkansas. The purpose of our work was to investigate large woodpecker breeding ecology in bottomland hardwood habitats. Specifically, we set out to examine breeding success, habitat and spatial use, and nest-survival modeling for Pileated Woodpeckers in the cypress-tupelo swamps and sweetgum-oak forests of eastern Arkansas. We collected our data from February through June, 2007 to 2010. Capture and processing techniques Starting in February, we trapped woodpeckers using two stacked elevated mist nets (12 × 2.6 m, 60-mm mesh) and a taped call (York and Wilson 1998). We extracted captured birds from mist nets and placed them in cloth bags until we banded and fitted each with a radio-transmitter. Birds were difficult to capture; therefore, we set out to capture as many birds as possible. We recorded the time the bird hit the net and the time the bird was released. We recorded morphological measurements, and all birds were aged (Pyle 1997). All birds were marked with unique combinations of colored leg bands and an aluminum United States Geological Survey band. All animal-handling procedures were approved through Arkansas State University Institutional Animal Care and Use Committee (protocol no. JB01-22A). We captured and handled all woodpeckers under Banding Permit 21426 from the United States Geological Survey, and Scientific Collection Permit 21320076 from the Arkansas Game and Fish Commission. At the beginning of the study, in 2007 and part of 2008, we used a modified figure-eight harness (Fig. 1A; Rappole and Tipton 1991) on 22 Pileated Woodpeckers, including three fledglings. The modification included use of an elastic cord that would wear over time, causing the transmitter to eventually fall off. The cord wrapped around the sartorius (i.e., the most anterior muscle on the anterolateral surface of the thigh [Fig. 1A; Pettingill 1970]), and the transmitter (Holohil Systems, Inc., Carp, ON, Canada; 7.5 g) sat directly on the synsacrum (Fig. 1A). The transmitter-to-body-mass ratio was 2.7% (x̅ = 274.6 g, n = 43 adults) for males and 3.1% (x̅ = 241.5 g, n = 18 adults) for females. Ripper et al. (2007) used the modified figure-eight harness on Hairy Woodpeckers and observed no negative effects. However, 9 (40.9%) of the 22 Pileated Woodpeckers we initially radio-tagged were documented as mortalities (Table 1; through March 2008). Because this mortality rate was higher than reported in the literature for the Pileated Woodpecker (Bonar 2001, Bull 2001), we modified the transmitter attachment technique near the end of the 2008 field season to a tail-mount procedure (Fig. 1B; Kenward 2001). 402 Southeastern Naturalist Vol. 12, No. 2 The tail-mount procedure involved both 7.5- and 4.4-g transmitters (Holohil Systems, Inc.) attached with unwaxed dental floss tied on to the two central retrices (Fig. 1B; Kenward 2001). The transmitter-to-body-mass ratio for the lighter transmitters was 1.6% (x̅ = 274.6 g, n = 43 adults) for males and 1.8% (x̅ = 241.5 g, n = 18 adults) for females. The floss was tied tightly to one retrix, whereas the other was tied much looser to allow the spread of the retrices during perching and flight between trees and to permit independent molting of these two feathers (Fig. 1B). Near the end of our 2008 field season, six radio-transmitters (two 7.5-g, and four 4.4-g radio transmitters) were affixed to woodpeckers using the tail-mount procedure. All 4 individuals (100%) with lighter mass tail-mounted transmitters survived throughout the field season, whereas one of the two 7.5-g radio-marked birds survived. However, early in 2009, two of the first 6 birds captured and outfitted with tail-mounted transmitters were confirmed mortalities (Table 1). Thus, handling procedures were adjusted further to decrease potential stress on the birds. While our adjustments of procedures were not limited to just time, we used the actual time as a variable for modeling survivorship, but recognize any of the following modifications could have contributed to an increase in survival: decreased handling time from ca. 1 hr to 30 min, attached all bands and the transmitter with the bird in the bag, discontinued the collection of some morphological measurements, discontinued age determination with bird in hand, and minimized talking and other human noises while handling. After these adjustments were made to the procedures, all remaining birds captured in 2009 (n = 12) survived Figure 1. Attachment of radio transmitters on Dryocopus pileatus (Pileated Woodpecker) in bottomland hardwood forests of eastern Arkansas. A) The figure-eight harness (Rappole and Tipton 1991) cord wrapped around the sartorius with the transmitter sitting on the synsacrum, and B) tail-mount (Kenward 2001) attached tightly to one retrix, but loosely attached to an adjacent retrix to allow mobility of the tail. 2013 B.L. Noel, J.C. Bednarz, M.G. Ruder, and M.K. Keel 403 the entire field season (≥3 months). In addition, all birds captured in 2010 (n = 18) survived the entire field season. Because woodpeckers were radio-tagged, all individuals were followed throughout the field season (February–June). All “confirmed mortalities” included feather piles or an intact carcass with transmitter. Some radio transmitters were found without feather piles, but with markings (scrapes, scratches, apparent bite marks) on the transmitter. Because of the possibility that these scrapes could have been caused by vegetation or the birds themselves, and also the possibility that these individuals may have been killed by predators, these were classified as “probable mortalities”, which would increase overall mortality estimates. Thus, we modeled survivorship for woodpeckers both using “confirmed mortalities”, and “total mortalities” that included “confirmed” and “probable” mortalities. Statistical analyses Distributions of estimated days of survival for individuals identified as mortalities, both “confirmed” and “probable” mortalities, were non-normal. Therefore, we performed a Mann-Whitney U-test to compare differences between estimated days of survival after release for “confirmed mortalities” and “probable mortalities”. In addition, the proportion of birds that did not survive Table 1. “Confirmed” and “probable” mortalities of Dryocopus pileatus (Pileated Woodpecker) in bottomland forests of eastern Arkansas from 2007 to 2010. Age of woodpeckers is defined as HY (hatch-year), SY (second-year), or TY (third-year or older) (Pyle 1997). Transmitter attachment refers to figure-eight harness (F8) or tail-mount (TM), and mass of the transmitter in grams. Handling time is given in minutes and refers to the time the bird hit the net until the time the bird was released. Transmitter Minimum Estimated Confirmed Bird attachment Handling days days or probable ID Capture date Age (mass [g]) Time survived survived mortalityA A 10 Mar 2007 SY F8 (7.5) 61 12 12 Confirmed B 13 Apr 2007 TY F8 (7.5) 33 50 54 Probable C 17 Apr 2007 SY F8 (7.5) 60 40 43 Confirmed D 21 Apr 2007 SY F8 (7.5) 38 2 4 Confirmed E 11 May 2007 HY F8 (7.5) 22 3 7 Confirmed F 22 May 2007 HY F8 (7.5) 40 4 6 Confirmed G 10 Feb 2008 TY F8 (7.5) 52 46 46 Probable H 22 Feb 2008 TY F8 (7.5) 45 59 59 Probable I 1 Mar 2008 SY F8 (7.5) 51 5 5 Confirmed J 14 Mar 2008 SY F8 (7.5) 47 6 7 Confirmed KB 20 Mar 2008 TY F8 (7.5) 30 1 2 Confirmed LB 22 Mar 2008 TY F8 (7.5) 41 6 7 Confirmed M 16 May 2008 TY TM (7.5) 104 1 2 Confirmed NB 26 Feb 2009 TY TM (4.4) 58 1 1 Confirmed O 2 Mar 2009 TY TM (4.4) 60 40 43 Probable P 3 Mar 2009 TY TM (4.4) 65 2 2 Confirmed A“Confirmed mortalities” included feather piles, whereas “probable mortalities” included the transmitter on the ground with scratches on the epoxy . BBirds found intact and carcasses were retrieved and submitted to the Southeastern Cooperative Wildlife Disease Study (SCWDS; Athens, GA) for postmortem analysis (clinical case 299-09; Ruder et al. 2012). 404 Southeastern Naturalist Vol. 12, No. 2 (both “confirmed” and “total” mortalities) a field season were compared to those that survived the entirety of a field season using a model I chi-square contingency table. Specifically, the proportion of mortalities before final adjustments was compared to the proportion of mortalities after final adjustments. Similarly, we conducted separate analyses using only “confirmed mortalities”, in the event “probable mortalities” were false assumptions. Because multiple adjustments (e.g., change in transmitter attachment type, change in transmitter mass, reduced handling time) were made throughout the study, we considered multiple predictors in the modeling analysis (Table 2). Specifically, we used logistic regression to estimate survivorship and Akaike’s information criterion corrected for small sample sizes (AICc) to examine alternative models to predict the likelihood of survival (Anderson and Burnham 2002). Birds that survived the entirety of a field season in which they were captured (from capture date until the end of June) were classified as survived (1) and birds that did not, were classified as dead (0). All of the proposed predictors (Table 2) were tested for multicollinearity prior to model building (Graham 2003). One predictor excluded from the list (Table 2) was proportion of radio transmitter mass to body mass. Because the change in attachment method coincided with change in transmitter size, the transmitter-to-body-mass ratio and attachment type were confounded. When the relationship between this ratio and survival was investigated separately for each attachment type, there was no relationship (indicated by confidence intervals that overlapped zero), so we dropped the proportion of radio-transmitter-mass-to-body-mass from subsequent analyses. All analyses were performed using SAS (2003). Results Mortalities We captured and radio-marked 64 Pileated Woodpeckers, of which at least 12 died (confirmed mortalities; 18.8%). Of the birds handled before final adjustments in the procedures, 12 of 34 (35.3%) Pileated Woodpeckers died. After final adjustments in radio-transmitter attachment and a reduction of handling time, all 30 birds captured and handled (100%) survived the entirety of the field seasons in Table 2. Predictors considered in our logistic regression models for estimating the likelihood of Dryocopus pileatus (Pileated Woodpecker) survival after being captured, restrained, and processed. Predictor Description Handling timeA Total minutes bird was restrained, which includes the moment bird hit the net until bird was released. Attachment Whether or not the figure-eight harness or tail-mount was used on the individual. Capture Date Day of the year the bird was captured. Year Year of the study (i.e., 2007, 2008, 2009, or 2010). Age Hatch-year, second-year, and third-year or older were the classifications used (Pyle 1997). Sex Male or female. AWhile limited to handling time, this variable also accounts for handling procedures that were modified throughout 4 years of capturing and tagging of Pileated Woodpeckers. 2013 B.L. Noel, J.C. Bednarz, M.G. Ruder, and M.K. Keel 405 which they were radio-marked (≥3 months). Nine (75.0%) of the 12 individuals classified as “confirmed mortalities” had the figure-eight harness (Table 1, Fig. 1A). The remaining 3 individuals had tail-mounted transmitters (Fig. 1B), but one individual (male; Bird M) had a heavier (7.5-g) transmitter , potentially contributing to that bird’s death. All “confirmed mortalities” were documented after a shorter period of time after transmitter attachment (x̅ = 8.2 days; range 1–43 days) than those individuals that were “probable mortalities” (x̅ = 50.5 days; range 43–59 days; U = 47.5, df = 1, P = 0.01). Four individuals were identified as “probable mortalities,” due to the lack of a feather pile. If these individuals are included in the mortality totals, the “total mortality” becomes 16 of 64 birds (25.0%). Importantly, 3 of these 4 “probable mortality” transmitters were found under water or in areas where feather piles and other evidence may have drifted away. Three of these 4 individuals had the modified figure-eight harness, and one had the tail-mounted transmitter (Table 1, Fig. 1). Further, in 2008, we successfully re-sighted 7.7% of the birds colorbanded in 2007 (1/13). In 2009, 26.7% of the individuals marked in 2008 (4/15) were re-sighted. Finally, in 2010, 55.6% of the birds marked from 2009 (10/18) were re-sighted. After all adjustments (e.g., tail-mount and reduced handling time) were made, survival (100%) through one field season was significantly higher in Pileated Woodpeckers compared to the period prior to adjustments for “confirmed mortalities” (64.7%; χ2 = 13.87, df = 1, P = 0.0002) and “total mortalities” (52.9%; χ2 = 20.04, df = 1, P < 0.0001). Logistical regression models Candidate AICc models of “total mortalities” suggest that, of the variables examined, the attachment type, handling time, and capture date explained the most Table 3. Results of the top 5 logistic regression models predicting Dryocopus pileatus (Pileated Woodpecker) survival in Arkansas for “total mortalities” and “confirmed mortalities.” Attachment = figure-eight harness or tail-mount, handling time = moment bird hits the net until it is released, capture date = capture day of the year, and year = study year (i.e., 2007, 2008, 2009, or 2010). Model KA AICc B ΔAICc C wi D Total Mortalities Attachment + handling time 4 56.68 0.00 0.47 Attachment + handling time + capture date 5 57.42 0.74 0.33 Attachment + handling time + year 8 61.09 4.41 0.05 Attachment 3 61.13 4.45 0.05 Attachment + capture date 4 61.54 4.85 0.04 Confirmed Mortalities Attachment + handling time 4 52.26 0.00 0.57 Attachment + handling time + capture date 5 54.61 2.35 0.18 Attachment 3 55.78 3.52 0.10 Year 5 58.01 5.75 0.03 Attachment + capture date 4 58.03 5.77 0.03 A Number of parameters. BAICc = -2 log L + 2K + 2K(K + 1) / (n - K - 1). CΔAICc = AICci - minAICc· Dwi = exp [-{ΔAICci / 2}] / Σ exp [-{ΔAIC ci / 2}]. 406 Southeastern Naturalist Vol. 12, No. 2 variation in survival of Pileated Woodpeckers after release (Table 3). Age, sex, and year of study did not explain enough variation to be considered factors influencing the likelihood of survivorship. Summation of Akaike’s weights for all models considered indicated that attachment type (97.8%) and handling time (89.0%) best predicted survival of Pileated Woodpeckers (Fig. 2), but capture date (39.0%) also contributed (Fig. 3). Similar to our previous analysis, candidate AICc models of “confirmed mortalities” suggested that the attachment type, handling time, and capture date explained the most variation in survival of Pileated Woodpeckers after release (Table 3). Somewhat different to the previous analysis on “total mortalities”, summation of Akaike’s weights for all combined models predicting survival indicated that capture date accounted for 22.9% of the weight. However, similar to previous analysis on “total mortalities”, summation of Akaike’s weights indicated that attachment method (91.9%) and handling time (81.2%) were the best predictors of survival. Although ranked second in both analyses, the 3-variable models containing attachment type, handling time, and capture date competed with the best-ranked model from each analysis (i.e., attachment type and handling time; Fig. 2). The second competing AICc model predicts that survivorship increased for those indi- Figure 2. Predicted values of survival with 95% confidence intervals for Dryocopus pileatus (Pileated Woodpecker) in bottomland hardwood forests of eastern Arkansas, based on the top candidate model from “total mortalities” including attachment type (i.e., figure-eight harness and tail-mount) and handling time (min) as predictors. Total mortalities includes, “confirmed mortalities” and “probable mortalities”. “Confirmed mortalities” included feather piles or an intact carcass, whereas “probable mortalities” included a recovered transmitter with markings, but no feather pile or carcass. 2013 B.L. Noel, J.C. Bednarz, M.G. Ruder, and M.K. Keel 407 viduals captured later in the year and outfitted with a radio transmitter using the figure-eight harness, holding handling time constant (Fig. 3). However, there is no obvious trend on the likelihood of survivorship for individuals outfitted with a radio transmitter using the tail-mount procedure (Fig. 3). Discussion Our analyses indicated that attachment type and handling time were 2 variables that likely affected the survival of captured and radio-marked Pileated Woodpeckers in eastern Arkansas (Fig. 2). In addition, capture date might have had some likelihood of affecting the survival of woodpeckers in Arkansas, especially with the figure-eight harness (Fig. 3). Multiple studies have concluded that behavioral and survivorship data collected from color-banded or radio-marked individuals might be biased and should be interpreted cautiously (Hooge 1991, Massey et al. 1988, Schulz et al. 2001, White and Garrott 1990). Based on our analyses, the figureeight harness (Rappole and Tipton 1991) decreased survivorship of Pileated Woodpeckers when compared to tail-mounts. Pileated Woodpeckers hitch up and down trees and the elastic cord wrapped around the base of their legs may have added resistance to these movements and increased the expenditure of energy by these Figure 3. Predicted values of survival with 95% confidence intervals for Dryocopus pileatus (Pileated Woodpecker) in bottomland hardwood forests of eastern Arkansas, based on the second candidate model from “total mortalities” including attachment type (i.e., figure-eight harness and tail-mount), and capture date as predictors, holding handling time (min) constant (i.e., x̅ = 44 and 41 min, respectively). Total mortalities includes, “confirmed mortalities” and “probable mortalities”. “Confirmed mortalities” included a recovered transmitter with markings, but no feather pile or car cass. 408 Southeastern Naturalist Vol. 12, No. 2 woodpeckers (Fig. 1). Therefore, we discourage the use of the figure-eight harness (Rappole and Tipton 1991) on the Pileated Woodpecker, or any species, such as other large woodpeckers, that use their legs as a primary means of locomotion. Consistent with our results, Barron et al. (2010) indicated that radio-tagged birds experience higher mortality rates when using harness attachments or collars, as opposed to glue or tail-mounts. However, Barron et al. (2010) only included three studies of birds with tail-mounted radios, as opposed to 27 studies using harness techniques in their meta-analysis. We found that woodpeckers captured earlier in the year (i.e., late winter) in Arkansas were less likely to survive (Fig. 3), suggesting that researchers should also be conscientious of trapping and radio-tagging large woodpeckers in colder, winter conditions. During these conditions, cold stress could be a factor and food resources may be less available, thereby making the Pileated Woodpecker more vulnerable to heightened stress when captured and processed. Therefore, capture date is cautiously considered as another variable that could impact survivorship of Pileated Woodpeckers and possibly other birds. The tag-mass-to-body-mass ratio recommended for radio transmitters is less than 5% for vertebrates species (Murray and Fuller 2000). Aldridge and Brigham (1988) attempted to test the “5%” rule using bats and suggested maneuverability decreased with increased mass of transmitters. Recently, most avian researchers have adopted the “rule” that transmitters and harnesses should be ≤3% of body mass to minimize adverse effects (Barron et al. 2010). In the first 2 years of radio-tagging Pileated Woodpeckers, we used a 7.5-g transmitter (ca. 2.7 and 3.1% of mean male and female mass, respectively), whereas after adjustments, we switched to a 4.4-g transmitter (ca. 1.6 and 1.8% of mean male and female mass, respectively). Due to multicollinearity, we did not include tag-mass-to-body-mass ratio and used only attachment type (Fig. 1), not proportion of radio-transmitter mass to body mass, as a predictor influencing survivorship. Our study was not designed to separate effects of radio-tag mass from effects of the attachment type. We are aware that tag-mass-to-body-mass ratio could also be as important of a predictor for survivorship of Pileated Woodpeckers. We adaptively adjusted the techniques, but not in an experimental way and could not tease the effects of these 2 factors apart. Thus, it is possible that if handling time is reduced (Bollinger et al. 1989, Ponjoan et al. 2008) and tail mounts are used, woodpeckers with the heavier transmitters could exhibit similar survival rates as those woodpeckers with lighter-mass tail-mounted transmitters. However, heavier-mass transmitters attached to tail feathers could cause premature molting or loss of the tail feathers. Specifically, 4 Pileated Woodpeckers (9.5%; 3 males and 1 female, respectively) dropped their tail-mounted transmitters (4.4 g) before the end of the radio-tracking study period, but we were still able to collect sufficient data prior to tail-loss for spatial-use analysis. In addition, all these individuals were re-sighted in subsequent years indicating that individuals could survive pre-mature tail feather loss. Moreover, one male Pileated Woodpecker had a larger mass tail-mounted transmitter (7.5 g), successfully bred and fledged young, carried the heavier transmitter throughout the field season, and was resighted in subsequent years. Many wildlife species exhibit some change in behavior in response to radiotags, such as increased preening and shaking, decreased flying ability, and altered 2013 B.L. Noel, J.C. Bednarz, M.G. Ruder, and M.K. Keel 409 time-energy budgets, which may make these individuals more susceptible to predation (Paton et al. 1991, Thirgood et al. 1995, White and Garrott 1990). Factors predisposing radio-tagged individuals to predation, such as altered behavior or capture myopathy, can be difficult to identify and document conclusively. Ponjoan et al. (2008) observed changes in mobility of Tetrax tetrax L. (Little Bustards) after capture and handling, and capture myopathy was histologically diagnosed in all 4 carcasses necropsied (Marco et al. 2006). Our findings are consistent with those of Ponjoan et al. (2008) in that most carcasses found were consumed by predators; thus, they were not available for pathological analysis. However, all 3 of the Pileated Woodpecker carcasses that we retrieved intact were submitted to the Southeastern Cooperative Wildlife Disease Study and exhibited clinical signs of capture myopathy (Ruder et al. 2012). Pathology can be conclusive pertaining to diagnosis of capture myopathy, but alterations in behavior can be indicative that organisms may have capture myopathy (Ponjoan et al. 2008). Collectively, these findings support our conclusion that longer handling time was a factor that influenced the survival of Pileated Woodpeckers in Arkansas. However, we adaptively made adjustments to our handling procedures (e.g., banding birds while still in bird bags, reduced human noises during handling) while reducing handling time, and emphasize that these adjustments could have contributed to the observed increase in survivorship as well. Two of 12 individuals documented as confirmed mortalities were handled in less than 35 min. (16.7%; Table 1), but all 30 individuals that survived after final adjustments were handled in less than 35 min. (x̅ = 29.6 min. ± 1.0 SE) suggesting that reduced handling time contributed to improved survivorship of Pileated Woodpeckers. Further, capture myopathy is an important issue to consider when trapping and marking the Pileated Woodpecker, and probably other large woodpeckers (Ruder et al. 2012). Thus, we strongly recommend that handling time be minimized and that other handling procedures be executed to minimize stress (e.g., minimize human noises) when capturing and marking lar ge woodpeckers. Vukovich and Kilgo (2009) found no difference in feeding rates, foraging behavior, vigilance, and preening behavior between radio-harnessed Melanerpes erythrocephalus L. (Red-headed Woodpecker) and non-radio-tagged individuals. Moreover, Ripper et al. (2007) found no negative effects using the figure-eight harness (Rappole and Tipton 1991) on Hairy Woodpeckers. Specifically, all 23 radio-marked birds survived and 88% of the Hairy Woodpeckers successfully fledged young (Ripper 2002). Importantly, Hairy Woodpeckers typically fly from location to location in a tree while foraging and use hitching locomotion less commonly than large woodpeckers (J.C. Bednarz, pers. observ.), which may make the figure-eight harness more suitable for this species and other smaller woodpeckers. In addition, Melanerpes formicivorus L. (Acorn Woodpecker) with 3.0-g harnessed transmitters (ca. 3.6% of body mass) did not spend more time preening, flying, or foraging than those individuals without transmitters (Hooge 1991). However, Acorn Woodpeckers with harnessed transmitters spent more time preening than those who had their transmitter attached with glue, suggesting the harness was an annoyance (Hooge 1991). Bull (2001) stated that adult Pileated Woodpeckers outfitted with backpack radio-tags were observed pecking 410 Southeastern Naturalist Vol. 12, No. 2 at the harness material after a week of attachment. Although several Pileated Woodpeckers died during previous radio telemetry studies, researchers suggested that backpack harnesses were not the cause (Bull 2001). Three of 4 (75%) harnessed radio-marked Red-cockaded Woodpeckers were consumed by predators within one week after translocation, whereas 5 of 12 (42%) color-banded birds could still be accounted for 8 months later (Odom et al. 1982). In addition, Acorn Woodpeckers spent greater amounts of time preening and trying to remove 4.5-g radio packages (ca. 5.5% of body mass) attached by harness techniques than those individuals without transmitters, and were less mobile overall (Hooge 1991). These observations along with our data suggest that harnesses could adversely impact many woodpecker species. If one considers a bird closer in size to the Pileated Woodpecker, 26 of 80 (32.5%) harnessed radio-marked Strix occidentalis Xantus de Vesey (Spotted Owl) died during a 2-year study in California, but only 5 of these 26 deaths were confirmed due to predation (Paton et al. 1991). In addition, Foster et al. (1992) found that radio-marked Spotted Owls produced fewer young, and they recommended researchers use lighter-weight tail-mounted transmitters. For a relatively long-lived species such as a Pileated Woodpecker, we suggest mortality rates ≥10% should be considered as the “threshold” for high mortality during a 3-month field season. Therefore, researchers observing relatively high rates of mortality (≥10%) should carefully review and adjust their capture, handling, and radio-tagging protocols in an attempt to minimize adverse impacts to their study organism. Acknowledgments This work would have not been possible without the logistical and financial support from US Fish and Wildlife Service (USFWS), The Bobolink Foundation, and the Georgia Ornithological Society. Other funding was generously provided by the Arkansas Audubon Society Trust. We thank T.J. Benson for his assistance with statistical analyses, and J. Gehring and past PIWO researchers for providing support and critical advice. We thank the Bednarz lab, C. Ray Chandler, and anonymous reviewers of this manuscript that have improved earlier versions of this manuscript. D. Noel, C. Johnston, N. Thompson, B. Furfey, B. Wilson, C. Eliason, Z. Rowe, R. Kessler, A. Brunetti, D. Topolewski, and other assistants contributed in the field. M. Lammertink, M. Piorkowski, R. Rohrbaugh, D. Luneau, M. Powers, J. Fitzpatrick, A. Mueller, R. Crossett, R. Hines, S. Reagan, K. Weaver, USFWS staff (Cache and White River National Wildlife Refuges), Arkansas Wildlife Management Area staff (Dagmar and Wattensaw Wildlife Management Areas), and Nature Conservancy staff provided important help with logistics and supplies. Final and special thanks to E. Bull, R. Renken, J. Cummings, P. Newell, K. Aubry, C. Raley, and J. Marzluff for their communication throughout regarding our capture and handling techniques during the study. Literature Cited Aldridge, H., and R.M. Brigham. 1988. Load carrying and maneuverability in an insectivorous bat: A test of the 5% “rule” of radio telemetry. Journal of Mammalogy 69:379–382. Anderson, D.R., and K.P. Burnham. 2002. Avoiding pitfalls when using informationtheoretic methods. Journal of Wildlife Management 66:912–918. 2013 B.L. Noel, J.C. Bednarz, M.G. Ruder, and M.K. Keel 411 Anich, N.M., T.J. Benson, and J.C. Bednarz. 2009. Effect of radio transmitters on return rates of Swainson’s Warblers. Journal of Field Ornithology 80:206–211. Barron, D.G., J.D. Brown, and P.J. Weatherhead. 2010. Meta-analysis of transmitter effects on avian behaviour and ecology. Methods in Ecology and Evolution 1:1–8. Bollinger, T.G., G. Wobeser, R.G. Clark, D.J. Nieman, and J.R. Smith. 1989. Concentration of creatine kinase and aspartate aminotransferase in the blood of wild Mallards following capture by three methods for banding. Journal of Wildlife Diseases 25:225–231. Bonar, R.L. 2001. Pileated Woodpecker habitat ecology in the Alberta Foothills. Ph.D. Dissertation. University of Alberta, Edmonton, AB, Canada. 75 pp. Bull, E.L. 2001. Survivorship of Pileated Woodpeckers in northeastern Oregon. Journal of Field Ornithology 72:131–135. Bull, E.L., and E.C. Meslow. 1988. Breeding biology of the Pileated Woodpecker: Management implications. US Department of Agriculture, Forest Service, Pacific Northwest Research Station, Research Note 474. Portland, OR. Conroy, M.J., G.R. Costanzo, and D.B. Stotts. 1989. Winter survival of female American Black Ducks on the Atlantic Coast. Journal of Wildlife Management 53:99–109. Cox, R.R., Jr., and A.D. Afton. 1998. Effects of capture and handling on survival of female Northern Pintails. Journal of Field Ornithology 69:276–287 . Foster, C.C., E.D. Forsman, E.C. Meslow, G.S. Miller, J.A. Reid, F.F. Wagner, A.B. Casey, and J.B. Lint. 1992. Survival and reproduction of radio-marked adult Spotted Owls. Journal of Wildlife Management 56:91–95. Gilmer, D.S., I.J. Ball, L.M. Cowardin, and J.H. Riechmenn. 1974. Effects of radio packages on wild ducks. Journal of Wildlife Management 38:243–252. Graham, M.H. 2003. Confronting multicollinearity in ecological multiple regression. Ecology 84:2809–2815. Höfle, U., J. Millan, C. Gortazar, F.J. Buenes-Tado, I. Marco, and R. Villafuerte. 2004. Self injury and capture myopathy in net-captured juvenile Red-legged Partridge with necklace radiotags. Wildlife Society Bulletin 32:344–350. Holt, R.D., L.W. Burger, Jr., S.J. Dinsmore, M.D. Smith, S.J. Szukaitis, and K.D. Goodwin. 2009. Estimating duration of short-term acute effects of capture and handling and radiomarking. Journal of Wildlife Management 73:989–995. Hooge, P.N. 1991. The effects of radio weight and harnesses on time budgets and movements of Acorn Woodpeckers. Journal of Field Ornithology 62:230–238. Kenward, R.E. 2001. A Manual for Wildlife Radio Tagging. Academic Press (Harcourt Science and Technology), London, UK. 311 pp. Marco, I., G. Mentaberre, A. Ponjoan, G. Bota, S. Mañosa, and S. Lavín. 2006. Capture myopathy in Little Bustards after trapping and marking. Journal of Wildlife Diseases 42:889–891. Massey, B.W., K. Keane, and C. Boardman. 1988. Adverse effects of radio transmitters on the behavior of nesting Least Terns. Condor 90:945–947. Murray, D.L., and M.R. Fuller. 2000. A critical review of the effects of marking on the biology of vertebrates. Pp. 15–64, In F.T. Boitani (Ed.). Research Techniques in Animal Ecology. Columbia University Press, New York, NY. 464 pp. Nicholson, S.D., R.L. Lochmiller, M.D. Stewart, R.E. Masters, and D.M. Leslie. 2000. Risk factors associated with capture-related death in Eastern Wild Turkey hens. Journal of Wildlife Diseases 36:308–315. Neudorf, D.L., and T.E. Pitcher. 1997. Radio transmitters do not affect nestling feeding rates by female Hooded Warblers. Journal of Field Ornithology 68:64–68. Odom, R.R., J. Rappole, J. Evans, D. Charbonneau, and D. Palmer. 1982. Red-cockaded Woodpecker relocation experiment in coastal Georgia. Wildlife Society Bulletin 10:197–203. 412 Southeastern Naturalist Vol. 12, No. 2 Paton, P.W.C., C.J. Zabel, D.L. Neal, G.N. Steger, N.G. Tilghman, and B.R. Noon. 1991. Effects of radio tags on Spotted Owls. Journal of Wildlife Management 55:617–622. Perry, M.C. 1981. Abnormal behavior of Canvasbacks equipped with radio transmitters. Journal of Wildlife Management 45:786–789. Pettingill, O.S., Jr. 1970. Ornithology in Laboratory and Field. Burgess Publishing Company, Minneapolis, MN. 403 pp. Ponjoan, A., G. Bota, E.L. Garcia de la Morena, M.B. Morales, A. Wolff, I. Marco, and S. Mañosa. 2008. Adverse effects of capture and handling Little Bustard. Journal of Wildlife Management 72:315–319. Powell, L.A., D.G. Krementz, J.D. Lang, and M.J. Conroy. 1998. Effects of radio transmitters on migrating Wood Thrushes. Journal of Field Ornithology 69:306–315. Putman, R.J. 1995. Ethical considerations and animal welfare in ecological field studies. Biodiversity and Conservation 4:903–915. Pyle, P. 1997. The Information Guide to North American Birds: Part 1. Slate Creek Press, Bolinas, CA. 732 pp. Rappole, J.H., and A.R. Tipton. 1991. New harness for attachment of radio transmitters to small passerines. Journal of Field Ornithology 62:335–337. Ripper, D. 2002. Habitat use by the Hairy Woodpecker (Picoides villosus) in the managed forests of the Olympic Peninsula, Washington. M.Sc. Thesis. Arkansas State University, Jonesboro, AR. Ripper, D., J.C. Bednarz, and D.E. Varland. 2007. Landscape use of habitat by Hairy Woodpeckers in managed forests of Northwestern Washington. Journal of Wildlife Management 71:2612–2623. Rolstad, J., and E. Rolstad. 1995. A note on the use of backpack radio-tags on mediumsized woodpeckers. Ornis Fennica 72:177–179. Ruder, M.G., B.L. Noel, J.C. Bednarz, M.K. Keel. 2012. Exertional myopathy in Pileated Woodpeckers (Dryocopus pileatus) subsequent to capture. Journal of Wildlife Diseases 48:514–516. SAS. 2003. SAS for Windows. Version 9.1. SAS Institute, Inc., Cary, NC. Schulz, J.H., A.J. Bermudez, J.L. Tomlinson, J.D. Firman, and Z. He. 2001. Comparison of radiotransmitter attachment techniques using captive Mourning Doves. Wildlife Society Bulletin 29:771–782. Spraker, T.R., W.J. Adrian, and W.R. Lance. 1987. Capture myopathy in Wild Turkeys (Meleagridis gallopavo) following trapping, handling, and transportation in Colorado. Journal of Wildlife Diseases 23:447–453. Sykes, P.W., Jr., J.W. Carpenter, S. Holzman, and P.H. Geissler. 1990. Evaluation of three miniature radio transmitter attachment methods for small passerines. Wildlife Society Bulletin 18:41–48. Thirgood, S.J., S.M. Redpath, P.J. Hudson, M.M. Hurely, and N.J. Aebischer. 1995. Effects of necklace radio transmitters on survival and breeding success of Red Grouse, Lagopus lagopus scoticus. Wildlife Biology 1:121–126. Vukovich, M., and J.C. Kilgo. 2009. Effects of radio transmitters on the behavior of Redheaded Woodpeckers. Journal of Field Ornithology 80:308–313. Whidden, S.E., C.T. Williams, A.R. Breton, and C.L. Buck. 2007. Effects of transmitters on the reproductive success of Tufted Puffins. Journal of Field Ornithology 78:206–212. White, G.C., and R.A. Garrott. 1990. Analysis of Wildlife Radio-tracking Data. Academic Press, New York, NY. 383 pp. York, D.L., and E.A. Wilson. 1998. Pileated Woodpecker capture using a mist net and taped call. North American Bird Bander 23:81–82.