Southeastern Naturalist 149 E. Zinsser, F.J. Sanders, P. Gerard, and P.G.R. Jodice 22001177 SOUTHEASTERN NATURALIST 1V6o(2l.) :1164,9 N–1o5. 62 Daily Survival Rate and Habitat Characteristics of Nests of Wilson’s Plover Elizabeth Zinsser1, Felicia J. Sanders2, Patrick Gerard3, and Patrick G.R. Jodice4,* Abstract - We assessed habitat characteristics and measured daily survival rate of 72 nests of Charadrius wilsonia (Wilson’s Plover) during 2012 and 2013 on South Island and Sand Island on the central coast of South Carolina. At both study areas, nest sites were located at slightly higher elevations (i.e., small platforms of sand) relative to randomly selected nearby unused sites, and nests at each study area also appeared to be situated to enhance crypsis and/or vigilance. Daily survival rate (DSR) of nests ranged from 0.969 to 0.988 among study sites and years, and the probability of nest survival ranged from 0.405 to 0.764. Flooding and predation were the most common causes of nest failure at both sites. At South Island, DSR was most strongly related to maximum tide height, which suggests that flooding and overwash may be common causes of nest loss for Wilson’s Plovers at these study sites. The difference in model results between the 2 nearby study sites may be partially due to more-frequent flooding at Sand Island because of some underlying yet unmeasured physiographic feature. Remaining data gaps for the species include regional assessments of nest and chick survival and habitat requirements during chick rearing. Introduction The beaches of coastal South Carolina support ~30 species of migratory shorebirds, but only 2 frequently nest in this habitat, Haematopus palliatus Temminck (American Oystercatcher) and Charadrius wilsonia Ord (Wilson’s Plover). The state of South Carolina has assigned both species the conservation status of “highest priority” within the state wildlife action plan and also lists the Wilson’s Plover as threatened (Sanders et al. 2013, South Carolina Department of Natural Resources 2015). The US Fish and Wildlife Service also considers Wilson’s Plover a species of high concern throughout its breeding range (Brown et al. 2001); there are ~8600 individuals in the southeastern US. The species is state-listed as endangered in Maryland and Virginia, threatened in Georgia, protected in Alabama, of special concern in North Carolina, and of greatest conservation need in Florida (Corbat and Bergstrom 2000, Florida Fish and Wildlife Conservation Commission 2012, Zdravkovic 2013). Wilson’s Plovers breed above the intertidal zone in areas with sparse vegetation (Corbat and Bergstrom 2000). Specific threats to nesting Wilson’s Plovers are similar to those faced by coastal shorebirds in general, and include anthropogenic 1Department of Forestry and Environmental Conservation and South Carolina Cooperative Fish and Wildlife Research Unit, Clemson University, Clemson, SC 29634. 2South Carolina Department of Natural Resources, 220 Santee Gun Club Road, McClellanville, SC 29458. 3Mathematical Sciences, Martin Hall O-114, Clemson University, Clemson, SC 29634. 4US Geological Survey, South Carolina Cooperative Fish and Wildlife Research, Clemson University, Clemson, SC, 29634. *Corresponding author - pjodice@clemson.edu. Manuscript Editor: Karl E. Miller Southeastern Naturalist E. Zinsser, F.J. Sanders, P. Gerard, and P.G.R. Jodice 2017 Vol. 16, No. 2 150 disturbance at nest sites, predation, barrier-beach stabilization, and flooding of nests (Brindock and Brown 2011, Brown et al. 2001, Corbat and Bergstrom 2000, Ray 2011). Recent research on the species in the southeastern US has expanded to focus on nest-site selection and reproductive success in North Carolina and South Carolina (DeRose-Wilson et al. 2013, Dikun 2008, Ray 2011). Here, we seek to extend that effort by examining nest-site selection and nest success along the central coast of South Carolina within the Tom Yawkey Heritage Preserve, where the beach is undeveloped and provides a relatively undisturbed system in which to study the species. Our objectives were to (1) identify microhabitat features associated with nest placement to better understand habitat preferences, and (2) assess the relationship between environmental features, habitat factors, and daily survival rate (DSR) of nests to identify factors that may be contributing to nest loss. Methods We conducted our study on South Island and Sand Island at The Tom Yawkey Wildlife Center and Heritage Preserve (33°15'N, 79°16'W; Fig. 1) in Georgetown, SC. Daily tidal ranges at the site can reach ~2 m. The 6.2 linear km of beachfront on South Island is adjacent to maritime forest. The 4.9 linear km of beachfront on Sand Island is primarily bordered by marsh habitat. Human use on both beaches is minimal, with the exception of beach patrols to search for nesting sea turtles. We found nest sites by searching for Wilson’s Plovers that exhibited territorial or nesting behavior such as distraction behaviors, wing dragging, scraping, or mating behaviors (Bergstrom 1988). We recorded the lay date of the 1st egg whenever possible and calculated potential hatch date as 25 days from the date the 3rd egg was laid (Corbat and Bergstrom 2000). When a clutch was found after completion, we floated one of the eggs to estimate lay date (Mabee et al. 2006). We monitored nest survival by visually checking nests every 3 ± 1.3 (mean ± SD) days (range = 1–5 d). We considered a nest successful if ≥1 egg hatched. If we encountered an empty nest at the time of hatch and the parents failed to exhibit defensive behavior, we considered the nest to have failed. We classified nests as abandoned after 3 consecutive nest checks without a sign of parents, as flooded when eggs were absent from a nest immediately following a spring- or high-tide that also deposited fresh wrack or debris in the vicinity of the nest, and as depredated when eggs were absent from a nest at a date other than near the anticipated hatch date and if predator tracks or scat also were present. If we were unable to assign one of the above classifications due to a lack of visual cues, then we reported the nest-loss cause as unknown. We used only visual cues; thus, our ability to categorize nest loss was likely conservative. To minimize disturbance during the nesting season and because early-season measures are most likely to coincide with habitat conditions when pairs choose nest sites, we recorded habitat variables on the day the nest was detected. Variables included: distance from nest to the nearest dune > 1.5 m high; height of and distance from nest to the nearest plant >10 cm high with a spatial extent of at least 0.25 m2 (i.e., the area of ground covered by the extent of the vegetation); and the distance from nest to the high-tide line as defined by the highest dried-wrack line. Southeastern Naturalist 151 E. Zinsser, F.J. Sanders, P. Gerard, and P.G.R. Jodice 2017 Vol. 16, No. 2 We obtained a measure of maximum tide height during the observation interval from a tide station located within 2 km of both study sites (http://tidesandcurrents. noaa.gov/noaatidepredictions/NOAATidesFacade.jsp?Stationid=TEC2929). To assess the extent of natural beach debris (also referred to as “items”), we centered a 1-m2 quadrat on the nest and counted all shells >1.5 cm in diameter and sticks longer than 10 cm (pooled as a single category, items), and any live plant stems within the plot. We recorded the number of items and plants per m2. Items and plants within the quadrat appeared relatively constant throughout the nesting period; therefore, we only counted these at the time of nest detection to minimize disturbance. To assess the potential effect of microtopography on nest success, we Figure 1. Wilson’s Plovers nests were monitored on South Island and Sand Island, SC, USA, March–July, 2012 and 2013. The black border is the delineation for The Tom Yawkey Wildlife Center and Heritage Preserve. Southeastern Naturalist E. Zinsser, F.J. Sanders, P. Gerard, and P.G.R. Jodice 2017 Vol. 16, No. 2 152 measured the height of the nest relative to the adjacent ground 20 cm from center of nest. We did so in the direction of the greatest difference in elevation to assess maximum difference in platform height relative to the immediate surroundings of the nest. We chose the aforementioned variables primarily based on their biological relevance in recent studies of Wilson’s Plovers (DeRose-Wilson et al. 2013, Dikun 2008, Ray 2011). To compare microhabitat features at used and unused sites, we chose an unused site 5 m from each nest location in a randomly chosen direction (Compton et al. 2002, Fedy and Martin 2011). We measured the same microhabitat variables described above using the same approach, and collected measurements at the unused location immediately following measurements at the nest site. We used a MANOVA to assess the relationship between microhabitat variables that were measured on a continuous scale (i.e., item density, plant density, distance to nearest plant, plant height, and relative nest height) and the differences in those values between nest sites and unused sites. This approach accounts for the dependence of the location of the unused site on the location of the nest site. We calculated the difference between the values at nest sites and unused sites for each habitat variable and used this difference as the response variable. MANOVA indicated significant differences for at least 1 habitat parameter; thus, we subsequently conducted a series of paired t-tests on the values for item density, plant density, distance to nearest plant, and plant height. We used a sign test to compare relative nest height at nest sites and unused sites because those data did not display a normal distribution. Five of the 8 habitat variables we measured differed between sites (P < 0.10, ANOVA); therefore we conducted all subsequent analyses separately by site. We modeled daily survival rate (DSR) of Wilson’s Plovers using a logistic exposure model (Schaffer 2004). The data were underdispersed for both study sites (deviance/ DF = 0.54 at Sand Island; deviance/DF = 0.36 at South Island), so we used a liberal alpha value of 0.15 during the backwards-elimination process. We assessed multicollinearity of continuous independent variables with a correlation analysis and avoided pairing any strongly correlated terms (r > 0.60) in the same model. Independent variables for the DSR model included density of items (defined above) within the quadrat, density of live plants within the quadrat, distance to the nearest plant, distance to the nearest dune, distance to the high-tide line, height of the nearest plant, initiation date of nest, midpoint date of nest ([date of last observation - initiation date]/2), midpoint2, maximum tide height during the observation interval as obtained from tide records, and exposure days (the number of days a nest was monitored). We included both initiation date and midpoint to capture potential date-effects on DSR based on both nest initiation and the core phase of the incubation cycle. We considered P-values ≤ 0.10 as marginally significant given the exploratory nature of the study, and P-values ≤ 0.05 as significant. We conducted all analyses in SAS 9.3 (SAS Institute, Cary, NC). Southeastern Naturalist 153 E. Zinsser, F.J. Sanders, P. Gerard, and P.G.R. Jodice 2017 Vol. 16, No. 2 Results We documented 38 nests on South Island and 34 nests on Sand Island. Several microhabitat features differed significantly between nest sites and paired unused sites at each site (MANOVA: P < 0.01). At South Island, density of natural beach debris (i.e., nest items; paired t37 = 2.7, P = 0.01), distance to nearest plant (paired t37= 2.9, P < 0.01), and relative nest height (t37 = 3.0, P ≤ 0.01) were greater at nest sites compared to unused sites (Table 1). At Sand Island, plant density (t33 = 1.9, P ≤ 0.07) was slightly higher, relative nest height was greater (t33 = 4.8, P ≤ 0.01), and plant height was lower (t33 = -2.6, P = 0.01; Table 1) at nest sites compared to unused sites. Flooding and predation were the 2 primary causes of nest failure at both sites (65% at South Island and 70% at Sand Island). In addition, 4 nests failed because of abandonment, 1 nest failed because of windblown sand, and 1 nest was buried by a nesting Caretta caretta L. (Loggerhead Turtle). The remainder failed due to unknown causes. The DSR of nests on South Island was negatively related to maximum tide height (Table 2). We also detected a moderate effect of year (the odds of a nest Table 2. Significant coefficients (SE) from logistic exposure models of DSR of nests of Wilson’s Plovers nesting at South and Sand islands, SC, March–July 2012 and 2013. Variable Estimate Pr > ChiSq South Island Year -1.27 (0.67) 0.06 Maximum tide height (m) -3.15 (1.86) 0.01 Distance to dune (m) -0.08 (0.03) 0.09 Item density (items/m2) 0.04 (0.02) 0.08 Sand Island No significant variables Table 1. Mean (SE) values for habitat variables at nest sites of Wilson’s Plovers and at nearby unused sites on South and Sand islands, SC, March–July 2012 and 2013. Parameter Nest site Paired unused site P-value South Island Item density (items/m2) 17.8 (2.5) 8.76 (2.6) 0.01 Plant density (plants/m2) 9.0 (3.4) 4.11 (1.4) 0.93 Distance to vegetation (cm) 335.9 (37.2) 95.4 (32.1) less than 0.01 Plant height (cm) 68.4 (5.63) 59.6 (8.4) 0.32 Distance to high tide line (cm) 4409 (551.8) 4323 (492.0) 0.81 Distance to dune (cm) 640.8 (141.7) 576.8 (125.3) 0.33 Relative nest height (cm) 2.3 (0.4) - - Sand Island Item density (items/m2) 4.9 (2.4) 4.12 (2.6) 0.51 Plant density (plants/m2) 5.3 (1.9) 1.0 (1.00) 0.07 Distance to vegetation (cm) 607.9 (96.3) 737.9 (104.1) 0.17 Plant height (cm) 48.0 (5.5) 60.6 (6.2) 0.01 Distance to high tide line (cm) 4819 (520.7) 4576 (478.2) 0.15 Distance to dune (cm) 2108 (308.6) 2113 (301.6) 0.96 Relative nest height (cm) 2.6 (0.5) - - Southeastern Naturalist E. Zinsser, F.J. Sanders, P. Gerard, and P.G.R. Jodice 2017 Vol. 16, No. 2 154 surviving increased by 3.5 times during 2013 compared to 2012), a moderate and positive effect of item density, and a moderate and negative relationship to distance to the nearest dune (Table 2). The DSR for nests at South Island was 0.977 ± 0.02 in 2012 and 0.988 ± 0.02 in 2013. The probability of a nest surviving from laying to hatching was 0.56 ± 0.2 in 2012 and 0.76 ± 0.2 in 2013 at South Island. The DSR of nests from Sand Island was not significantly related to any of the variables we measured (Table 2). The DSR and probability of success at Sand Island for both years combined were 0.969 ± 0.01 and 0.40 ± 0.2, respectively. Discussion During our study, the probability of a nest succeeding ranged from 40% to 76% among sites and years. Nest success of Wilson’s Plovers reported from other southeastern states was: 0–31% in 2 study years in Georgia (apparent nest success; Corbat 1990), ~35% during 2 study years in North Carolina (logistic exposure method; DeRose-Wilson et al. 2013); ~45% in North Carolina in 2 study years (Mayfield method; Ray 2011), 58% in Louisiana in 1 study year (logistic regression method; Zdravkovic 2010), and 25–66 % in 2 studies in Texas (apparent nest success; Bergstrom 1988, Zdravkovic 2005). Our rates of nest success appear moderate to high relative to these other reports, although caution is warranted when comparing nest success calculated using different methods. On South Island, DSR was negatively related to maximum tide height, positively related to density of beach debris in nest quadrats, and negatively related to distance to nearest dune. We suggest that these relationships reflect and are partly driven by flooding and predation. Predation and flooding of nests during extreme high tides and even daily high tides are common causes of nest failure among other beachnesting birds along the central coast of South Carolina (Brooks et al. 2013, 2014; Jodice et al. 2014), and flooding was a common cause of nest failure for Wilson’s Plovers in Georgia (Corbat 1990). Our data also suggest that nest placement may be chosen to minimize flooding and predation. For example, nests at South Island were on small (≤10 cm), slightly elevated platforms of accumulated sand, vegetation, and beach debris, compared to nearby unused sites without these features. Elevated platforms appeared to reduce flooding during high tides and storm events and were absent from all of the unused sites we assessed for comparison. Nest sites also had a higher density of natural beach debris within the nest quadrant compared to unused sites. The higher density of shells or plants near nest sites, and their locations with respect to dunes, may have enhanced crypsis of eggs (Smith et al. 2012). Sternula albifrons (Pallas) (Little Tern), which nest in habitat similar to Wilson’s Plovers, also had higher nest-success when they nested closer to dunes, perhaps because these structures obstruct a predator’s view, and, hence, lower predation rates (Medeiros et al. 2012). In contrast to South Island, DSR at Sand Island was not related to any of the variables we measured during the course of this study. Nest height at Sand Island was higher than the height of nearby unused sites, which might suggest nest sites were located to reduce the risk of flooding and predation. However, such relationships Southeastern Naturalist 155 E. Zinsser, F.J. Sanders, P. Gerard, and P.G.R. Jodice 2017 Vol. 16, No. 2 were not apparent in the results of the logistic exposure models. We posit 2 possible explanations for the lack of a significant relationship. First, some underlying yet unmeasured feature of Sand Island, perhaps the lack of mature dunes, may have reduced the strength of any of the associated variables in our models. Second, flooding may have occurred not just when tides peaked in height (i.e., maximum tide height), but also during intermediate or lower tides that still reached nest sites (i.e., failure was associated with a wider range of tide heights). Our data demonstrate that nests of Wilson’s Plovers were more likely to occur on elevated platforms and include natural beach debris that may enhance crypsis of eggs. Despite frequent flooding and high levels of predation, Wilson’s Plovers had moderately high rates of nest success on undisturbed islands on the South Carolina coast, particularly compared to other data from the southeastern US. Remaining gaps in data concerning this species in the southeastern US include rates of nest and chick survival, and habitat requirements during chick-rearing. Acknowledgments This research was supported and funded by the US Fish and Wildlife Service Threatened and Endangered Species Grant Fund and the South Carolina Department of Natural Resources. The US Geological Survey South Carolina Cooperative Fish and Wildlife Research Unit, in particular Carolyn Wakefield, provided administrative and logistical support. We are grateful to the staff of Tom Yawkey Wildlife Center and Heritage Preserve for their logistical support, particularly Jamie Dozier. Steve Coker and Mark Spinks, South Carolina Department of Natural Resources, also provided support during field research. The study was approved by the Clemson University Institutional Animal Use and Care Committee. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the US Government. Literature Cited Bergstrom, P.W. 1988. Breeding biology of Wilson’s Plovers (Charadrius wilsonia). Wilson Bulletin 100:25–35. Brindock, K., A.C. Brown. 2011. Breeding success and nest-site selection by a Caribbean population of Wilson’s Plovers. The Wilson Journal of Ornithology 123:814–819. Brooks, G.L., F.J. Sanders, P.D. Gerard, and P.G.R. Jodice. 2013. Daily survival rate for nests and chicks of Least Terns (Sternula antillarum) at natural nest sites in South Carolina. Waterbirds 36:1–10. Brooks, G.L., F.J. Sanders, P.D. Gerard, and P.G.R. Jodice. 2014. Daily survival rates for nests of Black Skimmers from a core breeding area of the Southeastern USA. Wilson Journal of Ornithology 126:443–450. Brown, S., C. Hickey, B. Harrington, and R. Gill (Eds.). 2001. United States Shorebird Conservation Plan, 2nd Edition. Manomet Center for Conservation Sciences. Manomet, MA. 70 pp. Compton, B.W., J.M. Rhymer, and M. McCollough. 2002. Habitat selection by Wood Turtles (Clemmys insculpta): An application of paired logistic regression. Ecology 83:833–843. Corbat, C. 1990. Nesting ecology of selected beach-nesting birds in Georgia. Ph.D. Dissertation, University of Georgia, Athens, GA. Southeastern Naturalist E. Zinsser, F.J. Sanders, P. Gerard, and P.G.R. Jodice 2017 Vol. 16, No. 2 156 Corbat, C.A., and P.W. Bergstrom. 2000. Wilson’s Plover (Charadrius wilsonia), No. 516, In A. Poole (Ed.). The Birds of North America OnLine. Cornell Lab of Ornithology, Ithaca, NY. Available online at http://bna.birds.cornell.edu./bna/species/516/. Accessed 24 March 2017. DeRose-Wilson, A., J.D. Fraser, S.M. Karpanty, and D.H. Catlin. 2013. Nest-site selection and demography of Wilson’s Plovers on a North Carolina barrier island. Journal of Field Ornithology 84:329–344. Dikun, K.A. 2008. Nest-site selection of Wilson’s Plovers (Charadrius wilsonia) in South Carolina. M.Sc. Thesis. Coastal Carolina University, Conway, SC. Fedy, B., and K. Martin. 2011. The influence of fine-scale habitat features on regional varaition in population performance of alpine White-tailed Ptarmigan. Condor 113:306–315. Florida Fish and Wildlife Conservation Commission. 2012. Florida’s wildlife legacy initiative: Florida’s state wildlife action plan. Tallahassee, FL. Jodice, P.G.R., J.M. Thibault, S.A. Collins, M. Spinks, F.J. Sanders. 2014. Reproductive ecology of American Oystercatchers nesting on shell rakes. Condor 116:588–598. Mabee, T.J., A.M. Wildman, and C.B. Johnson. 2006. Using egg flotation and eggshell evidence to determine age and fate of Arctic shorebird nests. Journal of Field Ornithology 77:163–172. Medieros, R., J.A. Ramos, P. Pedro, and R.J. Thomas. 2012. Reproductive consequences of nest-site selection by Little Terns breeding on sandy beaches. Waterbirds 35:512–524. Ray, K.L. 2011. Factors affecting Wilson’s Plover (Charadrius wilsonia) demography and habitat use at Onslow Beach, Marine Corps base Camp Lejeune, North Carolina. M.Sc. Thesis. Virginia Polytechnic Institute and State University, Blacksburg, VA. Sanders, F.J., M. Martin, M.D. Spinks, and N. J. Wallover. 2013. Abundance and breeding distribution of Wilson’s Plovers during the breeding season in South Carolina. The Chat 76:117–124. Schaffer, T.L. 2004. A unified approach to analyzing nest success. Auk 121:526–540. Smith, P.A., I. Tulp, H. Schekkerman, H.G. Gilchrist, and M.R. Forbes. 2012. Shorebird incubation behavior and its influence on the risk of nest predation. Animal Behavior 84:835–842. South Carolina Department of Natural Resources. 2015. South Carolina’s state wildlife action plan, Columbia, SC. Zdravkovic, M.G. 2005. 2004 Coastal Texas breeding Snowy and Wilson’s Plover census and report. Coastal Bird Conservation Program, National Audubon Society, Science Department, New York, NY. Zdravkovic, M.G. 2010. Wilson’s Plover (Charadrius wilsonia) breeding biology study at select sites in coastal Louisiana, 2009 Breeding summary report, Coastal Bird Conservation Program Conservian, Big Pine Key, Florida. Submitted to Barataria-Terrebonne National Estuary Program, Thibodaux LA. Zdravkovic, M.G. 2013. Conservation plan for the Wilson’s Plover (Charadrius wilsonia) Version 1.0. Manomet Center for Conservation Sciences, Manomet, MA.