843 A.L. Bryan, Jr., and R.R. Borkhataria 22001133 SOUTSoHuEthAeSaTsEteRrnN NNaAtTurUaRliAstLIST 1V2o(4l.) :1824,3 N–8o5. 04 Characterization of Tidally Influenced Wood Stork Foraging Habitats in Georgia A. Lawrence Bryan, Jr.1,* and Rena R. Borkhataria2 Abstract - To characterize tidally influenced Wood Stork foraging habitats, we documented the physical structure and potential prey populations of 17 known (based primarily on satellite telemetry locations) foraging and 20 “alternate” (similar habitat) sites, tidal creeks, in coastal Georgia. The majority of sites contained reaches partially impounded by three landscape features: oyster-shell dams, root/mud dams, or the junction of two or more creeks. Potential prey species, dominated by Fundulus heteroclitus (Mummichog) and shrimp, were highly variable among the tidal habitats but generally occurred in densities (average > 140 individuals/m2) far greater than those observed in an earlier inland Georgia study. There were no differences in potential prey densities between known foraging and alternate sites, confirming that the large salt marsh region of Georgia provides excellent foraging habitat for the regional Wood Stork population. Introduction The Wood Stork L.(Mycteria americana) is a large wading bird that breeds in the southeastern United States and feeds in shallow wetlands (Coulter et al. 1999, Kahl 1964). Wood Storks forage in a variety of aquatic habitats including forested wetlands, fresh- and saltwater marshes, ditches, and impoundments (Coulter and Bryan 1993, Coulter et al. 1999). A tactile foraging species, it forages most efficiently when its aquatic prey is concentrated in shrinking wetlands due to seasonal or tidal drawdowns (Coulter et al. 1999, Kahl 1964, Odum et al. 1995). Many of these wetlands are ephemeral, maintaining fish and other prey for a period of months prior to localized extinctions due to wetland drying. However, wetlands of varying hydroperiod are typically needed to provide sufficient food resources for parent storks and their young during the 85–105-day breeding season (Fleming et al. 1994). The Wood Stork is currently listed as a federal endangered species due to population declines resulting from foraging habitat loss (USFWS 1996). Common characteristics of habitats are shallow (10–30 cm) and relatively open water (Coulter and Bryan 1993, Coulter et al. 1999). Lacustrine and other deepwater habitats are typically unavailable for foraging storks (too deep), although suitable habitats occasionally occur in draining or drought-reduced reservoirs (Bryan et al. 2002). Fast-flowing freshwater systems are typically not employed, possibly due to incompatibility with their tactile foraging technique (Coulter and Bryan 1993), but storks forage in associated flood plain/backwater habitats of river and creek drainages affected by either tidal action or fluctuations due to annual rainfall. Tidal 1University of Georgia, Savannah River Ecology Laboratory, PO Drawer E, Aiken, SC 29802. 2University of Florida, Everglades Research and Education Center, Belle Glade, FL 33430. *Corresponding author - lbryan@srel.uga.edu. A.L. Bryan, Jr., and R.R. Borkhataria 2013 Southeastern Naturalist Vol. 12, No. 4 844 creeks are used extensively by foraging storks during low-tide stages (Bryan et al. 2001, 2005; Gaines et al. 1998, 2000). There are few ground-level characterizations of stork foraging sites and/or their prey densities, limited primarily to foraging habitats associated with the Birdsville Colony in east-central Georgia (Coulter and Bryan 1993) and the Everglades (Ogden et al. 1976). Generally, prey items were considerably more dense in the Everglades sites, and prey lengths were greater in the inland Georgia sites (Depkin et al. 1992). To date, there have been no characterizations of tidal foraging habitats/ prey densities. In this study, we characterized the structure and potential prey densities of known Wood Stork tidal creek foraging sites in coastal Georgia and compared them to other available tidal creek habitats. Study Area and Methods Site selection We utilized satellite telemetry locations of 15 storks associated with Harris Neck NWR to specify foraging sites and/or blocks of habitat, including only ground-level locations. Other foraging sites were found by opportunistic observations (e.g., direct sightings of storks foraging in the marsh). Seventeen known Wood Stork foraging sites were sampled (Fig. 1). In order to assess Wood Stork foraging-site quality, 20 additional sites (hereafter “alternate”) of similar habitat type were sampled within the same locale as the known foraging sites (Fig. 1). These sites were selected based on aerial overviews (GoogleEarth) and on-site observations and were sampled with the same protocol (see below) as the known foraging sites. All sampled sites occurred within an approximately 16,150-ha area of coastal habitat in Georgia’s McIntosh and Liberty counties, east of the Harris Neck NWR stork colony (31.6314°, -81.2760°). This region of salt marsh was bounded by Medway River/St. Catherine’s Sound on the north and by Sapelo Sound on the south. Sampling Protocol We documented physical characteristics and potential prey densities of all sampled tidal creeks from July–September 2012. Physical characteristics included occurrence of pools/impounded areas, structures causing the pools, water depths, and the surrounding vegetation during low-tide levels. Based on earlier findings relative to the timing of tidal foraging-habitat use (Bryan et al. 2001, Gaines et al. 1998), all sites were visited/sampled within a 4-hour time block centered on the predicted low-tide level for that area. Habitats were sampled for potential prey with dipnets 10–15 times per site. Given the size of the dipnet and average length of the dipnet “haul” (approximately 1 m), each haul resulted in estimated potential prey per 0.32 m2, which we converted to prey per 1.0 m2. We randomly selected 8–10 individual prey items of each species and measured their total lengths. One foraging site was sampled twice at different times during the low-tide period to examine the potential change in prey density relative t o tide level. Statistical comparison of potential prey densities between foraging and alternate 845 A.L. Bryan, Jr., and R.R. Borkhataria 2013 Southeastern Naturalist Vol. 12, No. 4 sites, and among pool-impounding features (e.g., oyster dams, root dams, creek junctions, and no pools) occurred with either Wilcoxon two-sample or Kruskal- Wallis tests. Prey data were presented as the mean ± 1 standard error to allow for comparison with other data sets. Results Physical structure All characterized foraging (n = 17) and alternate (n = 20) sites varied considerably relative to physical structure. Most sites (n = 31, 84%) had pools/impounded areas that were formed by features such as oyster-shell dams (n = 14) or root/mud dams (n = 13), or were formed at conjunctions of two or more tributaries (n = 4). Six (all alternate) sites were un-impounded tidal creek reaches with no pooling. We collected water depth (cm) and width (m) data for both the tidal creek reach and tidal pool components of the foraging and alternate sites. The average depths Figure 1. Locations of sampled Wood Stork foraging and aternate sites in the salt marsh near Harris Neck NWR, Mc- Intosh County, GA. A.L. Bryan, Jr., and R.R. Borkhataria 2013 Southeastern Naturalist Vol. 12, No. 4 846 and widths of the pools were significantly greater (Wilcoxon: P < 0.0003) than those of the creek reaches (Table 1). Pool length was highly variable and ranged from 2 to 40 m. We often encountered a series of smaller (2–3 m in length) consecutive pools. All tidal creeks sampled in this study were surrounded by homogenous stands of Spartina alterniflora Loisel. (Smooth Cordgrass) marsh. There were no other plants or shrubs within these areas. Potential prey The most prevalent species discovered in the tidal creeks were Fundulus heteroclitus (L.) (Mummichog), found in every site, and shrimp (Peneaus spp.), found in all but one of the sampled sites. Mugil curema Valenciennes (Mullet) were less common but found at several (n = 5) sites, whereas the following species were occasionally/rarely sampled: Menidia menidia (L.) (Atlantic Silversides), Paralichthys lethostigma D.S. Jordan & C.H. Gilber (Southern Flounder), and Strongylura marina (Walbaum) (Atlantic Needlefish). Adult Callinectes sapidus Rathbun (Blue Crab) were observed at many sites. Fiddler crabs (Uca spp.) and Panopeus herbstii H. Milne-Edwards (Mud Crab) were also observed on the mud banks at most of the sampled sites, although these species have not been confirmed as stork prey and were not sampled/quantified. Prey densities Average densities of potential prey (fish and shrimp) varied considerably among the 17 foraging sites, with an average density of slightly >140 individuals/m2 and ranging from 104–377 individuals/m2 (Table 2). Of the potential prey, on average approximately 117 and 24 individuals/m2 were shrimp and fish, respectively. Concurrently, densities of potential prey averaged approximately 186 individuals/m2 for the alternative sites, with a greater proportion of shrimp than observed at the actual foraging sites (Table 2). However, comparisons of densities of shrimp, fish, and combined prey between foraging and alternate sites were not significantly different (Wilcoxon: P > 0.1100). Therefore, all data were merged for subsequent analyses of prey relative to site landscape features. Examination of prey densities (foraging and alternate sites) relative to tidal pool impounding features (e.g., oyster dam, root/mud dam, creek junction, and no Table 1. Water depth and width of sampled tidal creek foraging and alternate sites in Georgia salt marshes.A n Average Std. error Range Depth (cm) Creek 26 25.6 1.6 20–50 Pool(s) 32 34.2 1.4 20–50 Width (m) Creek 26 0.8 0.1 0.2–2.0 Pool(s) 32 1.7 0.2 0.8–6.0 ADepths and widths of sampled tidal creek pools were significantly greater than creek reaches (Wilcoxon: P ≤ 0.0003). 847 A.L. Bryan, Jr., and R.R. Borkhataria 2013 Southeastern Naturalist Vol. 12, No. 4 pool) appeared to indicate an increasing gradient of prey from un-impounded creek reaches to creek-junction pools (Table 3). However, statistical comparison did not indicate significant differences (Kruskal-Wallis: P = 0.2033) among these features, possibly due to low sample sizes for the habitats on either end of the gradient and/ or the overall variation in densities among samples. Average fish densities were relatively consistent among the habitats (range = 15–30 fish/m2), whereas average shrimp densities ranged considerably from 96 to 235 shrimp/m 2 (Table 3). Densities of potential prey relative to tidal stage were examined at one site. An approximately 70–80-minute difference in tidal stage resulted in a 15-cm change in water depth and a substantial difference in prey densities. Sampling during the lower tidal stage resulted in 3.5 times more fish (31.2 fish/m2) and 10 times more shrimp (180.2 shrimp/m2) than the higher tidal stage (9.4 fish/m2, 17.7 shrimp/m2). Prey lengths Average total lengths of the predominant potential prey species among sites ranged from 27.7 ± 0.9 (SE) to 59.8 ± 1.8 (SE) mm for Mummichogs and 28.6 + 0.9 (SE) to 61.7 ± 5.1 (SE) mm for shrimp. Mullet were sampled at five sites, and lengths ranged from 72–200 mm. The largest individual prey items collected via dipnets among the sites were 71, 82, and 200 mm for Mummichogs, shrimp, and Mullet, respectively. Discussion The structures of both the foraging and alternate sites sampled in this study appeared indicative of the tidal creek habitats Wood Storks would encounter in this region. Our alternate sites may well have been utilized as foraging habitats by Table 2. Potential Wood Stork prey densitiesA at salt marsh foraging sites and alternative sitesB associated with the Harris Neck NWR colony. Density per m2 SitesB n Prey type Average SE Minimum Maximum Foraging 17 Shrimp 117.5 25.9 13.1 371.1 Fish 23.7 4.0 3.1 51.0 All 141.2 25.3 104.4 377.3 Alternate 20 Shrimp 161.2 19.8 19.2 369.8 Fish 24.2 3.7 3.1 80.0 All 185.6 21.9 25.4 401.0 ADensities of shrimp, fish, and combined prey at foraging and alternate sites were not significantly different (Wilcoxon: P > 0.1100). BPresumed foraging sites were determined by either direct observation (ground level and/or airplane) or by ground locations of storks carrying satellite transmitters (see Methods). Alternate sites were selected after assessing aerial photography and/or ground observations as being similar structurally to the foraging sites. A.L. Bryan, Jr., and R.R. Borkhataria 2013 Southeastern Naturalist Vol. 12, No. 4 848 storks that were not monitored or tracked. It is possible that the tidal pools along these creek reaches would be more attractive as foraging habitat to Wood Storks due to their affinity for still/slow flowing water (Coulter and Bryan 1993). These pools were also thought to support larger prey, although this was not specifically examined in this study. The dominant potential prey species we found at the foraging and alternate sites were Mummichogs and shrimp. This finding may have been influenced by the timing of our sampling (late summer) as well as our methodology (dipnets), although Mummichogs are considered resident species of the Georgia tidal marshes (Dahlberg 1972). Use of other sampling methods (e.g., seine, throw traps) was found to be logistically impractical due to site conditions (e.g., deep mud, un-even/ irregular creek bottoms, etc.). We believe our findings indicate the more prevalent potential prey species available to storks during this season. Mummichogs, the most common fish collected in our samples, were the most common prey item associated with coastal Wood Stork nestlings, and shrimp, Mullet, and Blue Crabs have all been documented as prey fed to nestlings (Bryan and Gariboldi 1 998). Densities of potential prey in the Georgia tidal creeks were much higher than those of inland (east-central) Georgia foraging sites and more similar to those reported for the Everglades. In east-central Georgia, the average density of all fish, crayfish, and amphibians for multi-year sampling of stork foraging sites (inland/ freshwater habitat types) was 13.4 individuals per m2, with a range of 0.2–298 Table 3. Potential Wood Stork prey densitiesA at salt marsh sites relative to tidal pool occurrence/ structureB. Pool-forming Density per m2 structureB n Prey type Average SE Minimum Maximum Oyster Shell 14 Shrimp 117.4 22.8 13.1 294.3 Fish 25.6 3.9 3.1 51.0 All 143.0 23.9 25.4 313.0 Root Mass 13 Shrimp 158.5 24.7 15.2 371.0 Fish 24.8 3.2 6.2 50.0 All 183.3 23.1 52.6 377.3 Creek Junction 4 Shrimp 234.8 76.5 24.8 369.8 Fish 30.2 17.8 3.1 80.0 All 265.0 81.8 27.9 401.0 No Pools 6 Shrimp 96.4 31.8 17.7 193.3 Fish 15.0 4.8 3.1 31.2 All 111.4 36.4 27.1 219.3 ADensities were not significantly different among impounding features (Kruskal-Wallis: P = 0.2033). BTidal pools along marsh creeks were typically impounded by three landscape features: oyster-shell dams, root/mud dams, and locations where two creeks joined together. We also sampled 6 tidal creek reaches without pools. 849 A.L. Bryan, Jr., and R.R. Borkhataria 2013 Southeastern Naturalist Vol. 12, No. 4 individuals/m2 (Coulter and Bryan 1993). Narrowing this to likely potential prey, those species known to be consumed by storks and ≥24 mm in length (Depkin et al. 1992), the resulting average density was 7.8 individuals/m2 (range = 0.7–250/ m2). The average prey density for Everglades foraging sites was estimated at approximately 141 individuals/m2 (Kushlan et al. 1975, Ogden et al. 1976). Sizes of potential prey from Georgia tidal habitats were also more similar to those of Everglades prey than inland Georgia prey, based on previous nestling regurgitation studies. Depkin et al. (1992) reported that average prey sizes in inland Georgia were almost two times greater than those from the Everglades. Also, Bryan and Gariboldi (1998) reported estuarine prey fed to stork nestlings were shorter in length than prey fed inland nestlings, based on nestling regur gitation. The findings of this study support the concept that the Georgia tidal salt marshes provide excellent foraging habitat for Wood Storks, although the availability of the prey is limited to 1–2 hours around low tide (Bryan et al. 2001, Gaines et al. 1998). The large tidal amplitude (8–10 feet) found in the tidal marshes in the Georgia– South Carolina Bight results in higher productivity (including prey densities) than the marshes to the north and south where the amplitude is less (Odum et al. 1995). Microhabitat differences among tidal creeks and variation in tides in different regions of the marsh system cause them to be suitable as foraging habitat at different stages of tidal draw down and for varying durations (Gaines et al. 1998). Foraging Wood Storks have been observed “hopping” among neighboring tidal creeks as foraging conditions/prey availability presumably changed. Thus, the tidal marshes during lower tide levels provide very dynamic feeding areas resulting in ephemeral prey-rich patches. Finally, Bryan and Gariboldi (1998) suggested that the presumed dense prey concentrations in draining tidal creeks might result in reduced foraging effort for storks, thus making their consumption of smaller estuarine prey more energetically efficient. Our study confirms that prey densities are high in draining tidal creeks. Depkin et al. (2005) reported foraging success rates by storks as 30 times greater in a draining tidal creek (3 prey items/min) as compared to foraging success rates at the Kathwood impoundments (0.1 prey item per min), a freshwater system created and stocked to provide foraging habitat for storks. These findings support the concept of greater consumption rates of smaller, but highly dense, prey and the value of tidal creeks and marshes as Wood Stork foraging habitat. Acknowledgments We received funding for these studies from the US Fish and Wildlife Service. Our study goals were aided by discussions with Billy Brooks (USFWS). The studies were partially supported by DOE Award Number DE-FC09-07SR22506 to the University of Georgia Research Foundation. We also thank Chris Depkin for occasional logistical support. Literature Cited Bryan, A.L., Jr., and J.C. Gariboldi. 1998. Food of nestling Wood Storks in coastal Georgia. Colonial Waterbirds 21:152–158. A.L. Bryan, Jr., and R.R. Borkhataria 2013 Southeastern Naturalist Vol. 12, No. 4 850 Bryan, A.L., Jr., J.W. Snodgrass, J.R. Robinette, J.C. Daly, and I.L. Brisbin, Jr. 2001. Nocturnal activities of post-breeding Wood Storks. Auk 118:508–513. Bryan, A.L., Jr., K.F. Gaines, and C.S. Eldridge. 2002. Coastal habitat use by Wood Storks during the non-breeding season. Waterbirds 25:429–435. Bryan, A.L., Jr., J.W. Snodgrass, J.R. Robinette, and L.B. Hopkins. 2005. Parental activities of nesting Wood Storks relative to time of day, tide level, and breeding range. Waterbirds 28:139–145. Coulter, M.C., and A.L. Bryan, Jr. 1993. Foraging ecology of Wood Storks (Mycteria americana) in east-central Georgia. I. Characteristics of foraging sites. Colonial Waterbirds 16:59–70. Coulter, M.C., J.A. Rodgers, J.C. Ogden, and F.C. Depkin. 1999. Wood Stork (Mycteria americana). No. 409, In A. Poole and F. Gill (Eds.). The Birds of North America. The Birds of North America, Inc., Philadelphia, PA. Dalhberg, M.D. 1972. An ecological study of Georgia coastal fishes. Fishery Bulletin 70:323–353. Depkin, F.C., M.C. Coulter, and A.L. Bryan, Jr. 1992. Food of nestling Wood Storks in eastcentral Georgia. Colonial Waterbirds 15:219–225. Depkin, F.C., L.K. Estep, A.L. Bryan, Jr., C.S. Eldridge, and I.L. Brisbin, Jr. 2005. Comparison of Wood Stork foraging success in selected tidal and non-tidal habitats. Wilson Bulletin 117:386–389. Fleming, D.M., W.F. Wolff, and D.L. DeAngelis. 1994. Importance of landscape heterogeneity to Wood Storks in Florida Everglades. Environmental Management 18:743–757. Gaines, K.F., A.L. Bryan, Jr., P.M. Dixon, and M.J. Harris. 1998. Foraging habitat use by Wood Storks in the coastal zone of Geor gia, USA. Colonial Waterbirds 21:43–52. Gaines, K.F., A.L. Bryan, Jr., and P.M. Dixon. 2000. The effects of drought on foraging habitat selection of breeding Wood Storks in coastal Georgia. Waterbirds 23:64–73. Kahl, M.P., Jr. 1964. Food ecology of the Wood Stork (Mycteria americana) in Florida. Ecological Monographs 34:97–117. Kushlan, J.A., J.C. Ogden, and A.L. Higer. 1975. Relation of water levels and fish availability to Wood Stork production in the southern Everglades, Florida. US Departement of the Interior Geological Survey, open file report 75434, Tallahassee, FL. Odum, W.E., E.P. Odum, and H.T. Odum. 1995. Nature’s pulsing paradigm. Estuaries 18:547–555. Ogden, J.C., J.A. Kushlan, and J.T. Tilmant. 1976. Prey selectivity by the Wood Stork. Condor 78:324–330. US Fish and Wildlife Service (USFWS). 1996. Revised recovery plan for the US breeding population of the wood stork. Atlanta, GA.