2011 SOUTHEASTERN NATURALIST 10(2):345–356 Population Densities of Painted Buntings in the Southeastern United States J. Michael Meyers* Abstract - The eastern population trend of Passerina ciris (Painted Bunting) declined 3.5% annually during the first 30 yrs of the Breeding Bird Survey (BBS, 1966–1996). Recently, the US Fish and Wildlife Service listed Painted Buntings as a focal species. Surveys for this focal species for the next 10 yrs (BBS, 1997–2007), however, are too low (less than 1 bird per 50 stops) for determining trend estimates. Also, to monitor densities adequately, surveys should account for incomplete detections. I surveyed singing Painted Buntings from 13 May to 26 June 2003 at 582 point counts (50 randomly selected transects) within blocks (64 x 64 km) in coastal and river areas from Florida to North Carolina. I compared densities of Painted Buntings for major habitats. Painted Buntings were detected at 33.5% of points surveyed for 5 min. Densities varied from 9 singing males per km2 in young pine plantations to 42 per km2 in maritime shrub. Effective detection radii for habitats varied from 64 to 90 m and were slightly higher in developed than in undeveloped habitats. Distance sampling is recommended to determine densities of Painted Buntings; however, large sample sizes (70–100 detections/habitat type) are required to monitor Painted Bunting densities in most habitats in the Atlantic coastal region of the southeastern United States. Special attention should be given to maritime shrub habitats, which may be important to maintaining the Painted Bunting population in the southeastern US. Introduction Population trends of Passerina ciris (L.) (Painted Bunting) declined signifi- cantly at 3.5% annually in the southeastern United States during the first 30 yrs of the Breeding Bird Survey (BBS, 1966–1996, coastal flatwoods; Sauer et al. 2008). More recent trend estimates (1997–2007), however, are not conclusive and may indicate either a continual decline or that the population has stabilized at a lower level. The average survey of Painted Buntings in the coastal flatwoods declined from approximately 1.9 to 0.8 birds/BBS route from 1966 to 2007. Inconclusive trends of regional abundance from surveys with less than 1.0 bird/route (low abundance) indicate data with deficiencies (Sauer et al. 2008). In other words, Painted Buntings may have become too rare on the BBS for accurate trend estimates in the southeastern United States, and managers may need other methods to monitor the population. BBS data did warn us, however, of a rapid population decline, which led to additional more detailed monitoring by the Working Group of the Eastern Painted Bunting. Concern for the eastern and western populations of the Painted Bunting led the US Fish and Wildlife Service to declare it a focal species (USFWS 2005). At *US Geological Survey, Patuxent Wildlife Research Center, Warnell School of Forestry and Natural Resources, The University of Georgia, Athens, Georgia 30602-2152; jmeyers@warnell.uga.edu. 346 Southeastern Naturalist Vol. 10, No. 2 that time, the Working Group of the Eastern Painted Bunting also recommended creation and expansion of a monitoring program that could provide estimates of the total population from North Carolina to Florida. Prior to that decision, a preliminary estimate of Painted Bunting densities was completed in 2003 and is reported herein. Ninety-five percent of bird monitoring in the past has been done by index surveys, with no density estimate and with no accounting for incomplete detections; however, density estimates are preferred in monitoring, especially those that account for incomplete detections (Buckland et al. 1993, 2001; Farnsworth et al. 2002; Rosenstock et al. 2002). The goal of this research was to conduct a preliminary study to estimate and compare the densities of Painted Buntings (singing males) for the major breeding habitats in North Carolina, South Carolina, Georgia, and Florida. Based on field work conducted on Sapelo Island, GA during 1996 to 2000, I hypothesized that maritime shrub habitat would have the highest densities of Painted Bunting in the southeastern US. I designed a survey using distance sampling (program DISTANCE 5.0) to estimate densities of eastern Painted Buntings in their breeding region (see Lowther et al. 1999, Sykes and Holzman 2005) of the southeastern Atlantic coastline and three major river systems (up to 90 km inland). My objectives were to randomly survey Painted Buntings in all major habitats within the four states and determine preliminary population densities and differences among the major habitats including developed areas. Methods Study design I surveyed habitat of Painted Buntings in all major breeding habitats including maritime shrub, maritime oak (adjacent to marsh), shrub-scrub agriculture, open pine forest (sawtimber) with less than 50% canopy cover, and young pine plantations (1–5 yr old) (Bellis 1995, Lowther et al. 1999). Habitats, except young pine plantations, also included points in undeveloped as well as developed lands (buildings, homes, farms, other human facilities, and other human habitats such as mowed grass). Habitat of shrub-scrub agriculture was classified as developed. Surveys began in Florida on 13 May and ended in North Carolina on 26 June 2003. Except for the portion in interior South Carolina, I sampled the entire southeastern Painted Bunting’s breeding range (Sykes and Holzman 2005), covering areas less than 90 km from the Atlantic coast of Florida, Georgia, North Carolina, and South Carolina. I stratified the Atlantic coastline and rivers into blocks of ≈64 x 64 km within the area, which resulted in the proportion of survey points as follows: ≈10% NC, ≈38% SC, ≈39% GA, and ≈13% FL. For a morning's survey, I randomly selected one transect (median = 14 points/transect) from all transects (center lines in areas of ca 2 km²) for each of the bunting’s habitats in the block. A total of 50 transects and 582 points was established. Rarer habitats, such as, maritime oak and open pine forest were sometimes limited to none or one possible transect in a block, but in most cases three to six transects were included for random selection for bunting habitats based on 2011 J.M. Meyers 347 examination of topographic maps and aerial photographs. I checked habitat of the randomly selected transect on the day before the survey. If a selected transect was determined unsuitable habitat for breeding Painted Buntings (i.e., 30-yr-old unthinned pine plantations, agriculture crop with no shrubs or grasslands, or oak woodlands with >100% cover and no understory), I used the next randomly selected transect for that habitat. I found suitable habitat in the first (≈90%) or second (≈10%) randomly selected transect. I used roadside surveys on most private lands in all habitats; however, roads were secondary and most were unpaved. Painted Buntings used the roads and roadsides, but roads with heavy traffic and primary roads were not considered bunting habitat because of wide, mowed rights of way or were unsuitable for aural sampling because of traffic noise. I surveyed private developed habitat from roadsides and rights of way. I randomly selected transect direction and routes (turns) on roads when more than one direction was available. I established points along transects for distance sampling at intervals no less than 200 m and averaging 355 m (13% were <250 m). Distances between points in young pine plantations varied considerably (0.4–2.0 km or more) because of the location of this habitat within the landscape and were not included in the average 355-m interval. Avian survey I conducted surveys from just before sunrise to 4.5 to 5.0 hrs after sunrise. I did not survey on rainy or windy days (>4 on the Beaufort Scale). An observer (n = 2) conducted each transect. Two observers each conducted approximately half of 582 points on 50 transects. Observers wore camouflage clothing and moved about the point (≈5–8 m area) slowly and recorded all Painted Buntings seen or heard. Movement about the point reduced the possibility of missing a bird nearby and thereby ensured that the probability of detection at the point is certain or g(0) = 1 (Buckland et al. 1993, 2001). Observations began about 1 to 2 minutes after arrival at the point (Buckland et al. 1993, 2001). Distances from the point to the first position of a singing bird were recorded to the nearest meter with a laser rangefinder (±1 m) and paced (at the end of the survey) when <15 m. I used binoculars (10 x 40 mm) to aid in determining a bunting’s position. I also mapped the first position (distance and bearing from point) of Painted Buntings and noted the birds’ movements during each five-minute survey. Distance analyses I used program DISTANCE 5.0, Release 2 for point-transects to determine density of Painted Buntings (Thomas et al. 2010). I ran half-normal, uniform, and hazard-rate models with adjustment (or no adjustment) functions of cosine, simply polynomial, and hermite polynomial for untruncated and right truncated (>100 m, ≈11% of observations truncated) distance data (global and by habitats). Preliminary data indicated similar mean detection distances of 60 to 62 m for singing males in five of seven habitats and 68 m for the remaining two habitats. Adjustment functions for the models were selected automatically by the program based 348 Southeastern Naturalist Vol. 10, No. 2 on AIC scores using all models and maximum adjustments set to five. I determined variances for n (number of observed singing males) by the empirical method from the sample. I tested model fit (P > 0.15) by chi-square goodness-of-fit (GOF). I removed models with multiple warnings for parameters highly correlated, with parameters constrained to obtain monotonicity, and with convergence failures. I entered distance data as transect code (in Region), point number code (in Point- Transect), and distances in m (in Observations) of program DISTANCE 5.0. I deleted one outlier distance of 203 m that was 48 m beyond the next maximum distance and 95 m more than the maximum distance in that habitat. High tide and sound transmission over water probably caused this outlier. I also compared EDR differences by observers (n = 2) for the global and maritime shrub models (truncated distance data >100 m) and found no difference: global (EDR1 = 57 m, 95% CI: 45 to 72 m, 141 df; EDR2 = 62 m, 95% CI: 52 to 72 m, 73 df) and maritime shrub (EDR1 = 59 m, 95% CI: 40 to 88 m, 38 df; EDR2 = 60 m, 95% CI: 47 to 76 m, 28 df). The observers’ probabilities of detecting a singing Painted Bunting within a 100-m radius were also similar for the global model (P1 = 0.32, 95% CI : 0.20–0.52, 110 df; P2 = 0.38, 95% CI: 0.28–0.52, 73 df) and for the maritime shrub model (P1 = 0.35, 95% CI : 0.16–0.76, 38 df; P2 = 0.36, 95% CI: 0.22–0.57, 28 df). Therefore, I pooled data for the two observers. Results Painted Bunting presence Painted Buntings were detected on 33.5% of the points (n = 582) during 5-min surveys in the coastal area of four southeastern states (Table 1). Point counts in open pine and maritime oak with development (homes, parks, and farm facilities) were occupied 4% to 13% less by Painted Buntings than in similar habitats without development. Buntings occupied maritime shrub habitat in developed areas 6% more than in similar undeveloped habitat. Detection rates of Painted Buntings varied by habitats with the lowest in young pine plantations (12% of points), highest in maritime shrub with development (48% of points), and ranging from 14% to 42% of points in the remaining six habitats (Table 1). Pine habitat overall had the lowest detection rate of Painted Buntings. Table 1. Percent of points occupied by singing male Painted Buntings and detections by habitats, southeastern Atlantic coastal states, 2003 (n = number of points by habitat, total = 582). Singing males detected to maximum of 155 m. Habitat n Occupied (%) Number of detections Maritime oak 117 30 43 Maritime oak developedA 101 26 28 Open pine (saw timber) 51 27 18 Open pine (saw timber) developedA 7 14 1 Maritime shrub 150 42 75 Maritime shrub developedA 29 48 17 Shrub-scrub agriculture 37 30 12 Young pine plantation 90 12 14 AHuman dwellings or buildings in the area of the survey. 2011 J.M. Meyers 349 Distance models Four global models and adjustments (half normal with no adjustments, halfnormal with simple polynomial, half-normal with hermite polynominal, and uniform with cosine adjustments) had ΔAICc < 0.50; however, the half-normal model without adjustment and with 1 parameter ran without any constraints and with similar densities and low CVs (0.10–0.11) compared to the others. The half-normal model without adjustment fit the data well for the global detection function (Fig. 1A). The model showed a good fit by Kolmogorov-Smirnov’s GOF Figure 1. The detection function and histogram of observed detection distances for the global model (all data), selected half-normal key with no adjustment (parameters = 1), and without truncations (A) and with truncation at 100 m (B) from program DISTANCE for the Painted Bunting in coastal areas of North Carolina, South Carolina, Georgia, and Florida (n = 208 detections from 50 point transects and 582 points), 2003. GOF = goodness-of-fit. 350 Southeastern Naturalist Vol. 10, No. 2 Table 2. DISTANCE models’ goodness-of-fit by habitats of Painted Buntings for Kolmogorov-Smirnov’s (K-S) and Cramér-von Mises’s (W 2 and C 2) goodness- of-fit statistics, southeastern Atlantic coastal states, 2003. All models had good fit (P > 0.15), but most had excellent fit (P > 0.60). Habitat Model (parameters) K-S P W 2 P C 2 P Maritime oak Uniform cosine (1) 0.092 0.86 0.046 0.9 < P < 1.0 0.033 0.8 < P < 0.9 Maritime oak developed Uniform cosine (2) 0.102 0.93 0.039 0.9 < P < 1.0 0.023 0.9 < P < 1.0 Open pine (saw timber) Uniform cosine (1) 0.181 0.60 0.077 0.7 < P < 0.8 0.058 0.6 < P < 0.7 Maritime shrub Half-normal simple polynomial (2) 0.059 0.96 0.051 0.8 < P < 0.9 0.034 0.8 < P < 0.9 Shrub-scrub developed Uniform simple polynomial (2) 0.183 0.62 0.139 0.4 < P < 0.5 0.097 0.3 < P < 0.4 Shrub-scrub agriculture Uniform simple polynomial (2) 0.289 0.27 0.217 0.2 < P < 0.3 0.121 0.3 < P < 0.4 Young pine plantation Uniform cosine (1) 0.124 0.98 0.043 0.9 < P < 1.0 0.022 0.9 < P < 1.0 2011 J.M. Meyers 351 statistic (Dn = 0.070, P = 0.27) and Cramér-von Mises’s GOF statistic (W ² = 0.112, 0.5 < P ≤ 0.6; C ² = 0.0.072, 0.5 < P ≤ 0.6). Truncation (>100 m) of distance data produced an acceptable model (Fig. 1B, note GOF was not as good as in the untruncated model); however, CI’s for density estimates were larger in the truncated model (loss of precision) for global data and also in models run by habitats. Therefore, distance truncation was not used in the models to determine densities. All DISTANCE models for habitats had Kolmogorov-Smirnov’s GOF statistics and Cramér-von Mises’s GOF statistic that indicated fair to excellent fits (Table 2). Hazard-rate models did not perform as well as other models (both in global and by habitats) in determining Painted Bunting densities. Distance estimates EDR in DISTANCE 5.0 analysis by habitats indicated that EDR expanded in developed areas, although only developed and undeveloped habitats (combined all) were different (Table 3). Probability of observing a singing Painted Bunting was lower in maritime shrub than in maritime oak, open pine, shrub-scrub agriculture, and developed shrub-scrub. The probability of observing singing buntings in combined developed habitat was more than twice that in combined undeveloped habitat (Table 3). Only the global model and combined undeveloped habitats had detection sample sizes >100 and CVs (0.10–0.11) for density estimates considered adequate (CV ≤0.15), although maritime oak, maritime oak developed, and maritime shrub (75 detections) were reasonably adequate (CV = 0.19–0.25). Densities of Painted Buntings for distance sampling ranked highest in maritime shrub (41.8 singing males per km2, CV = 0.25, Fig. 2). Singing buntings in maritime shrub occurred at much higher densities than those in young pine plantations and developed maritime oak, which recorded only 9.2 to 12.7 singing Table 3. Comparisons by habitat of similarity in effective detection radius (EDR) and probability (P) of observing a singing Painted Bunting within the survey area. Detection data truncated at >100 m for comparisons among habitats. Models: Ahalf-normal with no adjustment, Buniform with cosine or simple polynomial adjustment, and Chalf-normal with simple polynomial adjustment in DISTANCE program, southeastern Atlantic coastal states, 2003. Habitat (maximum detection, m) EDR (m) (95% CI) CV P (95% CI) CV df Maritime oak developed (155)B 80 (72–89) 0.05 0.64 (0.51–0.80) 0.11 24 Maritime oak (110)B 60 (48–74) 0.10 0.36 (0.24–0.55) 0.21 37 Open pine (120)B 65 (50–86) 0.13 0.43 (0.25–0.73) 0.25 13 Maritime shrub (137)A 61 (49–76) 0.11 0.37 (0.24–0.58) 0.23 67 Shrub-scrub developed (118)B 91 (71–118) 0.12 0.83 (0.50–1.00) 0.24 15 Shrub-scrub agriculture (115)B 82 (70–95) 0.07 0.67 (0.49–0.90) 0.14 10 Young pine plantation (115)B 77 (67–88) 0.06 0.59 (0.44–0.78) 0.13 11 Developed combined (155)B,D 86 (80–92) 0.03 0.74 (0.65–0.84) 0.07 41 Undeveloped combined (137)C,E 57 (51–64) 0.06 0.33 (0.26–0.40) 0.11 142 Global, all (155)A 70 (65–75) 0.05 0.48 (0.40–0.58) 0.10 185 DCombined maritime oak developed and shrub-scrub developed habitats. Open pine developed habitat did not have sufficient sample. ECombined maritime oak, maritime shrub, and open pine habitats. 352 Southeastern Naturalist Vol. 10, No. 2 males per km2 , respectively (CVs = 0.38 and 0.19). Densities in open pine, shrub-scrub agriculture, developed shrub-scrub, and maritime oak were similar (15.9–23.2 singing males per km2; CVs = 0.37, 0.34, 0.35, and 0.19, respectively) and 45% less than in maritime shrub (Fig. 2). I found no strong indications of differences between developed and undeveloped areas within habitats or for combined developed and undeveloped habitats (Fig. 2). Discussion The half-normal model for distance sampling illustrated detection distances declining gradually with a good model fit (Fig. 1). The assumption is most likely valid that all singing Painted Buntings where detected at 0 distances, (see 0 to 8 m zone, Fig. 1). Actually, 4 singing males were observed in this zone, when only 2.3 would be expected for the model. One was detected at 0 m and above the observer, which would be a rare occurrence. A relatively high detection rate of ca. 0.90 was evident to 43 m, after which detection dropped rapidly to <0.50 (Fig. 1). A second assumption that buntings were detected prior to any evasive movement was most likely met. For about half of the point counts, I observed locations Figure 2. Density estimates (95% CI) of Painted Buntings for distance sampling by habitats. SSM = maritime shrub, SSdev = shrub-scrub developed, MO = maritime oak, MOdev = maritime oak developed, SSA = shrub-scrub agriculture, OP = open pine (sawtimber), YPP = young pine plantations (≤5 yrs old). Undev = combined maritime oak and maritime shrub. Dev = combined maritime oak developed and shrub-scrub developed habitats. Dev and Undev habitats are not comparable to other habitats. Only seven points and one detection occurred in open pine developed habitat. 2011 J.M. Meyers 353 of buntings as I entered the area, then stopped at the center, and waited 1 to 2 minutes to begin the survey. Painted Buntings singing while I entered the survey area were not observed moving away from the observer; therefore, distance sampling in a short 5-min period should provide an accurate “instantaneous picture” of birds around the point (Rosenstock et al. 2002, Scott and Ramsey 1981). A third assumption that distances were measured accurately was met by using a laser rangefinder with ±1 m accuracy and also by walking towards measured singing locations after the survey period to be sure that birds were singing from that location. Distances <15 m were also measured by pacing at the completion of the survey. Slowly moving around the point, while observing and listening to singing buntings also provided better accuracy of location data by listening and watching from different angles. Four global DISTANCE models provided similar density estimates (20–23 singing males per km2) overall for Painted Buntings, with ΔAICcs ≤ 0.50. Within habitats, uniform and half-normal models with cosine or simple polynomial adjustments provided the best density estimates according to ΔAICc (≤ 0.50), CVs, and model goodness-to-fit statistics (Table 2). Program DISTANCE requires 60 to 100 detections in each habitat to obtain a coefficient of variation (CV ≤ 0.15) that provides a precise density estimate (recommended by Buckland et al. 1993, 2001; Rosenstock et al. 2002, Thomas et al. 2010). This was met for three density estimates in this study (developed habitats: n = 46 detections, CV = 0.15; undeveloped habitats combined: n = 137 detections, CV = 0.12; global – all habitats; n = 207 detections, CV = 0.11); however, maritime oak and maritime oak developed habitats had reasonable CVs (0.19) with detections of 43 and 28, respectively (Fig 2A). CVs of density estimates in all other habitats ranged from 0.27 to 0.38 with detections below (n ≤ 18) what is recommended for estimates. Maritime shrub was an exception to this pattern (n = 75 detections) with a CV = 0.27, which may be related to variance associated with extremely high density or, in this case, a lack of data near the inflection point of detection curve (50–60 m). CIs were less for some habitats with <60 detections, such as maritime oak when compared to CIs from maritime shrub with 75 detections (Fig. 2), which means that increasing sampling for more detections may not reduce CIs in maritime shrub habitat. Obtaining 80 distance samples of Painted Buntings for young pine plantations would require >500 points surveyed just for that density estimate alone. For this study, some preliminary density estimates of Painted Buntings may lack statistical power needed to predict changes in the population using program DISTANCE. In future surveys, density estimates of Painted Buntings from program DISTANCE may require ≥200 points or more per habitat and considerably more for young pine plantations (ca 1800 total points for this study). As an alternative, points could be surveyed more than once to collect additional distance samples (Buckland et al. 1993, 2001); however, this approach would also require more time and funds. EDR was higher for habitat of developed maritime oak with more open understory and relatively open canopy (<50%) than for maritime oak. EDR was higher, as expected, in more open developed habitat compared to undeveloped habitats 354 Southeastern Naturalist Vol. 10, No. 2 (combined). More similarities of EDRs existed than differences, however, indicating that different habitats for the most part had similar effects on the ability of the observer to detect singing male Painted Buntings in undeveloped habitats. Because of this similarity, it may be reasonable, with more testing, to use a standard 70- or 75-m fixed radius for surveys that are at least within 95% CIs for EDRs for all habitats or the average of all (70 m; Table 3). Different habitat data could be then compared; however, density data from fixed-radius plots would be underestimated. Other factors, such as use of highly trained observers (Kepler and Scott 1981) and camouflage clothing (Gutztwiller and Marcum 1997) may have also contributed positively to detectability issues. Multiple surveys of the same area in most habitats would be needed to detect buntings that may not have been present during one five-minute survey. Painted Buntings in maritime shrub habitats do not leave territories to forage at distant sites, so this may be a source of bias for density estimates between this habitat and others where buntings may be feeding far from territories (see Springborn and Meyers 2005). Multiple surveys at the same point have been incorporated in range-wide surveys of eastern Painted Buntings currently underway (Rua Mordecai, Eastern Painted Bunting Working Group, 2008 pers. comm.). Few data exist for comparing density estimates of Painted Buntings obtained in this study. There are, however, home-range data from Sapelo Island, GA, that indicate densities may be fairly accurate for two habitats herein (Springborn and Meyers 2005). In a two-year study of radio-tracked male Painted Buntings (n = 23) in open pine and maritime shrub, kernel home ranges averaged 7.0 ha (95% CI: 4.9–9.1) and 3.1 ha (95 % CI: 2.2–3.9), which converts to approximately 14.3 singing males/km2 (95 % CI: 11.0–20.4, open pine) and 32.3 singing males/km2 (95 % CI: 25.6–45.5, maritime shrub). Both of these densities and CIs overlap those in distance sampling for open pine (15.9 singing males/km2) and maritime shrub (41.8 singing males/km2) (Fig. 2). More sampling of each habitat may produce smaller CIs and CVs for density estimates of Painted Buntings using distance sampling. At high densities, e.g., in maritime shrub, sampling may not improve CVs. I recommend that managers of Painted Buntings use distance sampling with large sample sizes (60–100 detections per habitat) for long-term studies in the region and on management areas to determine densities and CIs of this focal species. Use of roads was the only method that some areas could be surveyed in this study because of time and funding. One should be aware that surveys next to roads may not provide a random sample of the breeding habitat for Painted Buntings because of a road effect. The use of road surveys in a comparison study without roads to determine any adjustment to density estimates may be helpful; however, buntings are species of habitat edges and use edges created within the habitat (clumps of shrubs) and on linear edges (salt marsh and maritime oak). Recently, the Working Group for the Eastern Painted Bunting learned quickly that conducting random surveys without using secondary or unpaved roads was not practical because of time and permission required to collect the data. A comparison of off road and roadside surveys from the Working Groups’ three-year study may 2011 J.M. Meyers 355 provide density estimates of Painted Bunting that will clarify how roadside surveys affect estimates. Managers should consider maritime shrub the most important habitat for Painted Buntings in the southeastern US. Maritime shrub has the highest bunting densities and also high adult survival (Springborn and Meyers 2005). This coastal habitat is also prime real estate and it’s unlikely that much of it which is in private ownership will be maintained without development. Development may reduce bunting densities in maritime shrub by ca 50% (this study) and may also reduce survival because of increased predation pressures. Similarly, undeveloped maritime oak habitat, although 38% less dense than maritime shrub for Painted Buntings, still supports more buntings than other habitats. Maritime oak is also under pressure for development in the coastal southeastern US. Development also reduces Painted Bunting densities by ca 50% in maritime oak. All other habitats should be considered important for Painted Buntings, although lower densities (10–15 singing males/km²) make them less likely to be potential source habitats (Pulliam 1988). Acknowledgments United States Fish and Wildlife Service, Region 4, provided partial funding for this project. K. Bettinger assisted with field work and managed data. J. Peterson, J. Kubel, and two anonymous reviewers provided helpful comments and suggestions for the manuscript that improved the final draft. J. Nichols inspired me to conduct this monitoring research. I thank my colleagues of the Eastern Painted Bunting Working Group for discussions, which also helped me develop ideas for this monitoring research project. K. Chapman of the USFWS assisted with administration of the project. Many persons facilitated permits for surveys at parks, wildlife management areas, and refuges, and their help was important to completing this project during a period of special care for my terminally ill mother, Jane Rojahn Meyers. Literature Cited Bellis, V.J. 1995. Ecology of maritime forests of the southern Atlantic coast: A community profile. Biological Report 30, US Department of Interior, Washington, DC. 95 pp. Buckland, S.T., D.R. Anderson, K.P. Burnham, and J.L. Laake. 1993. Distance Sampling: Estimating Abundance of Biological Populations. Chapman and Hall, London, UK. 446 pp. Buckland, S.T., D.R. Anderson, K.P. Burnham, J.L. Laake, D.L. Borchers, and L. Thomas. 2001. Introduction to Distance Sampling: Estimating Abundance of Biological Populations. Oxford University Press, Oxford, UK. 432 pp. Farnsworth, G.L., K.H. Pollack, J.D. Nichols, T.R. Simons, J.E. Hines, and J.R. Sauer. 2002. A removal model for estimating detection probabilities from point-count surveys. Auk 119:414–425. Gutzwiller, K.J., and H.A. Marcum. 1997. Bird reactions to observer clothing color: Implications for distance-sampling techniques. Journal of Wildlife Management 61:935–947. Kepler, C.B., and J.M. Scott. 1981. Reducing bird count variability by training observers. Studies in Avian Biology 6:366–371. 356 Southeastern Naturalist Vol. 10, No. 2 Lowther, P.E., S.M. Lanyon, and C.W. Thompson. 1999. Painted Bunting (Passerina ciris). In A. Poole (Ed.). The Birds of North America Online. Cornell Lab of Ornithology, Ithaca, NY. Online at http://bna.birds.cornell.edu/bna/species/398doi:10.2173/ bna.398. Accessed 2 June 2010. Pulliam, H.R. 1988. Sources, sinks, and population regulation. American Naturalist 132:652–661. Rosenstock, S.S., D.R. Anderson, K.M. Giesen, T. Leukering, and M.F. Carter. 2002. Landbird counting techniques: Current practices and an alternative. Auk 119:46–53. Sauer, J.R., J.E. Hines, and J. Fallon. 2008. The North American Breeding Bird Survey, Results and analysis 1966–2007. Version 5.15.2008. USGS, Patuxent Wildlife Research Center, Laurel, MD. Available online at http://www.mbr-pwrc.usgs.gov/bbs/ bbs.html. Accessed 11 February 2009. Scott, J.M., and F.L. Ramsey. 1981. Length of count period as possible source of bias in estimating bird densities. Studies in Avian Biology 6:409–413. Springborn, E.G., and J.M. Meyers. 2005. Home range and survival of breeding Painted Buntings on Sapelo Island, Georgia. Wildlife Society Bulletin 33:1432–1439. Sykes, P.W., Jr., and S. Holzman. 2005. Current range of the eastern population of Painted Bunting (Passerina ciris) Part I: Breeding. North American Birds 59:4–17. Thomas, L., S.T. Buckland, E.A. Rexstad, J.L. Laake, S. Strindberg, S.L. Hedley, J.R.B. Bishop, T.A. Marques, and K.P. Burnham. 2010. Distance software: Design and analysis of distance sampling surveys for estimating population size. Journal of Applied Ecology 47:5–14. United States Fish and Wildlife Service. 2005. The US Fish and Wildlife Service’s focal species strategy for migratory birds. Division of Migratory Bird Management, Arlington, VA. Available online at http://www.fws.gov/migratorybirds/FocalSpecies/ The%20Focal%20Species%20Fact%20Sheet%20and%20Table.pdf. Accessed 11 February 2009.