INVESTIGATORS: R. MICHAEL ERWIN, JOHN G. HAIG, DANIEL B. STOTTS, U.S. National Biological Service, Patuxent Environmental Science Center, Laurel MD 20708-4015; BARRY TRUITT, Virginia Coast Reserve, Brownsville, Nassawadox VA 23419; and C. RANDOLPH CARLSON, LTER Project, University of Virginia, Oyster VA 23412.
In 1994, we initiated a project on the nesting Gull-billed Tern, Sterna nilotica, a cosmopolitan species whose nesting distribution along the Atlantic extends from Florida to western Long Island, New York. The study began because of rising concerns over the nesting population declines since the early 1970s. Numbers of adults have diminished from more than 2000 during the 1975-76 period down to about 1000 in the early 1980s, now down to 400-500 in the last three years (TNC-VCR data files). Similar declines have been found in Florida (H. Smith, pers. comm.). The species has been listed on the U.S. Fish & Wildlife Service's Migratory Bird Species of Management Concern (Office of Migratory Bird Management 1986) because of unstable numbers throughout the country.
Primary concerns for the survival of this species are based on several threats: sea level rise, expanding populations of Herring Gulls, Larus argentatus, and the dependence of the species on invertebrates, some of which are taken in agricultural fields where pesticides are used. This study began to address some basic ecological questions concerning this virtually unknown species.
No research on nesting ecology has been published on this species in the U.S., primarily because of the difficulty in monitoring the survival of young; they are reported to often leave the nest vicinity within a few days of hatching, although there is conflicting information about this. Thus, a first step was to attempt to follow the fate of young after hathing.
We began the field study in May 1994, attempting to locate as many colonies along the Virginia barrier island complex as possible. By early June, we had located 10 colonies in the barrier island region from Fisherman Island north to the Maryland line in Chincoteague Bay. For several reasons, we selected six colonies to study intensively, ranging from 8 to 55 nests. The largest colony, about 150 pairs on Wreck Island, was not chosen because of high disturbance potential to several species of terns and Black Skimmers, Rynchops niger. We also did not study 2 small colonies near Congers Sloop Channel (W. of Hog Island) or one at Coard's Marsh in Chincoteague Bay. We selected a small colony at Wire Narrows (Chincoteague), two at Middle Metomkin Island, Cedar Sandbar, Chimney Pole Marsh, and Brant Hill.
After most clutches were complete, we marked nests with numbered stakes. About two weeks later, we enclosed a sample of from 8 to 22 nests in each colony to try to determine fledging success. Where possible, an approximately equal number of "control" nests was marked outside of the enclosures to allow a comparison of enclosure effect on hatching success. Enclosures consisted of 1/2" mesh nylon screening stapled to firring strips that were then driven ca. 15 cm into the sand or shell substrate. Stakes were positioned about 3-4 m apart and enclosed from 2-7 nests in areas from 10 to 25 sq.m. Nests were monitored from 2-3 times per week during the incubation and hatching periods until young were large enough to fly. The fate of each nest was recorded on nest cards. Hatchlings were marked with fingernail polish using different colors (n=4) on the two wings to distinguish brood orders. They were banded when at least 5 days of age with USFWS inkalloy bands. On each visit, all young were weighed with an Ohaus scale, and culmen and wing chord lengths were recorded (nearest mm) with calipers. We recorded any prey items found near the nest or regurgitated by young.
To document elevations of nests and vulnerability to flooding, we used a laser theodolite and Global Positioning System with reference markers to measure x,y, and z coordinates to the nearest cm. Because of equipment malfunction, we only obtained data at one site (Chimney Pole).
We also recorded the presence of other species nesting near the Gull-billed Terns at each site and recorded evidence of predation.
A summary of the nesting results is shown in Table 1. Success in general was fairly low, with the average hatching success about 55% and fledging rate of 0.45 young per nest. In most cases, we could not determine the cause of nest failure or chick loss. Storm flooding at our colonies was minor, with 3 of 15 nests being lost in June at Brant Hill. Flooding also washed out a small number of nests at the two Congers Sloop Channel colonies that we did not study. At all other 5 study colonies, no nests were flooded during the nesting period (May to July); however September storms washing over several of these sites (B. Truitt, pers. observ.). Herring Gull predation was witnessed on young Black Skimmers and Gull-billed Terns at the Chimney Pole site, but this may have been caused by our intrusion.
We examined growth rates of mass, culmen, and wing chord to evaluate whether food limitation may be affecting differential growth among brood members (a,b, vs. c chicks). Linear regression results during the linear phase of growth of the chicks (5 to 20 days) revealed significant differences between a and b chicks for mass and wing chord, but not culmen growth. This suggested food limitation in the colonies in 1994. A comparison of standardized mass growth of Gull-bills with other terns species showed a much reduced rate comparing this species with others (Table 2). This is apparently not a function of the invertebrate (vs. fish) diet, since the only other invertebrate- feeding species, the Black Tern, had the highest growth rate.
A small amount of information collected on diet of young revealed a variety of prey items were fed to the chicks, including fish, shrimp, and a variety of insects. Most noteworthy were piles of Japanese beetles found next to nests in several locations at the Metomkin colony sites.
The effect of enclosures was evaluated with respect to clutch size and hatching rate (Table 3). No significant differences were found using t tests.
Finally, elevations were taken of nest heights for 21 Gull- billed Tern nests at Chimney Pole before the GPS malfunctioned. We found the lowest nest was only 23 cm above the high water mark recorded in June. The highest nest was 37 cm above, therefore the range of elevations was only 14 cm.
Gull-billed Terns have declined in Virginia since the mid 1970s rather dramatically. Elsewhere in the U.S. some declines have also occurred. We found low productivity and hatching success at six colonies in the barrier island region in 1994. Growth rates of young also seemed to be among the lowest reported among a variety of other studies with other species. Differential growth of a vs. b chicks was found for mass and wing chord suggesting food limitation. Enclosing nests seems to be an effective and nonbiased method for evaluating nesting success for this species.
Parameters N Mean + 1 S.D. Colony Size 6 21 (range 8-55) Study nests 115 Clutch Size 2.38 + 0.89 No. chicks hatched/nest 1.37 + 1.08 No. chicks fledged/nest 0.45 + 0.67
Adult Peak Standardized Species Mass (g) Rate (g/d) Peak Ratea Diet Leasta 45 3.2 7.1 (CA) Fish Blacka 63 5.7 9.0 (Canada) Insect Roseatea 112 6.1 5.5 (N.E.) Fish Forster'sa 123 8.1 6.6 (WI) Fish Commona 124 7.9 6.5 (England) Fish Gull-Billed 178 6.7 3.8 (VA) Invert. Sandwicha 237 10.7 4.5 (England) Fish Caspiana 672 26.0 4.0 (CA) FishaFrom Schew 1990 (MS thesis, unpubl.); Standardized peak rate = peak rate / adult mass x 10-2.
No. hatching Group Clutch Size per nest Enclosed (136) 2.43 + 0.9 1.44 + 1.0 Unenclosed (94) 2.32 + 0.9 1.26 + 1.2