BULLETIN OF THE FLORIDA STATE MUSEUM BIOLOGICAL SCIENCES Volume 8 Number l THE TERNS OF THE DRY TORTUGAS William B. Robertson, Jr. UNIVERSITY OF FLORIDA Gainesville 1964 Numbers of the BULLETIN OF THE FLORIDA STATE MUSEUM are pub- lished at irregular intervals. Volumes contain about 800 pages and are not necessarily completed in any one calendar year. OLIVER L. AUSTIN, JR., Editor WALTER AUFFENBERG, Managing Editor Consultants for this issue: John W. Aldrich Joseph C. Moore Communications concerning purchase or exchange of the publication and all manuscripts should be addressed to the Managing Editor of the Bulletin, Florida State Museum, Seagle Building, Gainesville, Florida. Published 8 May 1964 Price for this issue $1.85 THE TERNS OF THE DRY TORTUGAS WTT T.T~M B. ROBERTSON, JR.1 4 SYNOPSIS: New information from unpublished sources and from published rec- or(is hitherto overlooked permit a re-evaluation of the history of the Dry Tortugas and of the terns that inhabit them. The geography and ecology of the 11 keys that have variously comprised the group since it was first mapped in the 1770's are described and their major changes traced. The recorded occurrences of the seven species of terns reported nesting on the keys are analyzed in detail. The Sooty Tern colony has fluctuated from a low of about 5,000 adults in 1908 to a reported peak of 190,000 in 1950;, for the past four years it has remained steady at about 100,000. The Brown Noddy population, which reached a peak of 85,000 in 1919, was rEduced by rats to about 400 adults in 1938; it is in the neighbor- hood of 2,000 today. A colony of 150 to 450 Roseate Terns has nested in most years from 1917 to the present. About 500 Least Terns nested regularly from 1916 to 1982, then unaccountably dwindled to a few pairs by 1937 and shortly afterward- disapp6ared. Royal and Salldwich Terns nested abundantly in the mid-19111 century, and a colony 6f Royals may have existed as late as 1890. Both species are believed to have been extirpated from the Tortugas by egging. No verifiable evidence exists for the nesting of the Common Tern, which has been reported several times. The Black Noddy, first reported for the continental United States at Dry Tortugas in 1960, has been found there each summer since. i The author is Park Biologist at Everglades National Park and Fort Jefferson National Monument, Homestead, Florida. Manuscript submitted 10 October 1963.-ED. Robertson, William B., Jr., 1964. The terns of the Dry Tortugas. Bull. Florida State Mus., vol. 8, no. 1, pp. 1-95. TABLE OF CONTENTS Introduction Acknowledgments 01 Co 14 ©Location and Phy5iography History and Names· of the Tortugan Keys Sooty Tern 17 Record of Nesting 17 Discussion 38 Brown Noddy -- 52 Record of Nesting - » 52 Discussion 58 Roseate Tern 64 Discussion _ 70 Least T6rn 78 Common Tern 75 Royal and Sandwich Terns 78 Black Noddy 7.9 Literatilre Cited 82 INTRODUCTION The tern colonies of the Dry Tortugas, in particular the great breed- ing aggregations of the Sooty Tern, Sterna fuscata Linnaeus, and the Brown Noddy, Anous stolidus (Linnaeus), have been of interest to ornithologists since Audubon visited them in 1832. Although the area is remote and difficult of access even today, few bird colonies in North America can boast so long a record of observations or so extensive a literature. During the early years of the Carnegie Institution of Washington's Tortugas Laboratory, John B. Watson and his co-workers made ex- tended observations on Sooty Terns and Brown Noddies (Watson, 1907, 1908, 1910; Watson and Lashley, 1915; Lashley, 1915). Their work provided nearly all of the detailed life history data available for these species until recently. It also included pioneer experimental studies. of behavior, homing, and orientation, as· well as an early in- stance of-the use of metal leg bands to mark'birds. Excepting the work of Watson and his associates, the literature consists- almost entirely.- of descriptions of the ternery as observed during brief visits. Many accounts since 1900 ·include estimates of the number of Sooty Terns, Brown Noddies, and other breeding spe- 1964 ROBERTSON: DRY TORTUGAS TEENS 3 cies. Lengthier studies by National Park Service personnel in 1987 and 1988 (Beard, 1989) were concerned particularly with predation upon Sooty Tern chicks by Magnifeent Frigate-birds, Fregata mag- nificens Mathews. Parts of the historical record were summarized by Bartsch (1919), Vinten (1948), Sprunt (1948b), and Moore and Dilley (1958). Modern banding at the Dry Tortugas began with the activities of Jack C. Russell in 1986 and was continued annually through 1941, principally on outings sponsored by the Florida Audubon Society. About 18,800 Sooty Terns and 246 Brown Noddies were banded. The bandings were reported separately by eight or more individuals and ho analysis of the data was undertaken until recently (Austin, 1962 MS.). In June 1959, the National Park Service, Florida State Museum, and Florida Audubon Society began a cooperative mass-banding study of the movements and demography of the Sooty Tern popula- tion. At the end of the 1963 season new bandings of Sooty Terns by project cooperators totalled approximately 82,300 adults and 41,900 juveniles. In the course of this work it became evident that a number of the widely scattered published reports and much unpublished in- formation had not been taken into account by previous compilers. Because of this, several apparent misinterpretations of the history of the colony had gained wide currency. The present summary resulted. The names of birds are those of the Check-List of North Ame,·ican Birds, American Orhithologists' Union, 1957, except for the changes resulting from the recent discovery of Anous tenuirostris (Temminck) at the Dry Tortugas (Robertson et al., 1961). ACKNOWLEDGMENTS It is possible to mention here only a few of the people who helped me to assemble the data this paper summarizes. The cooperation of those named and many others contributed greatly to this review. I am particularly indebted to Joseph C. Moore for permission to refer to the unpublished reports of tern censuses he made in 1958, 1954, and 1955. C. Russell Mason also made extensive field notes available to me. Others who contributed unpublished data or photo, graphs included Robert P. Allen, H. G. Deignan, John R, DeWeese, Willard E. Dilley, Theodore R. Greer, David O. Hill, James B. Meade, Dennis R. Paulson, Roger T. Peterson, Chandler S. Robbins, Alexander Sprunt IV, and Louis A. Stimson. 4 BULLETIN FLORIDA STATE MUSEUM Vol. 8 Albert Manucy provided a wealth of information from his re- search on the history of Dry Tortugas and useful advice on historical sources. Of those who helped me to obtain copies of rare publica- tions, maps, and material from archives, I must thank in particular Luis R. Arana, Charles M. Brookfield, C. Gordon Fredine, Lowell Sumner, and C. R. Vinten. Charles I. Park, Julius F. Stone, Jr., and C. C. Von Paulsen gave me their recollections of Dry Tortugas in the late 1920's and early 1980's, a period for which little written record exists. Recent work at Dry Tortugas has depended greatly upon the cooperation of National Park Service personnel in the area particu- larly.District Manager and Mrs. Wallace B. Elms, District Manager James A. Olson, and Park Rangers Roy Evenson, Carl S. Christensen, and James E. Markette. Finally, I am grateful to Oliver L. Austin, Jr. for assistance in locating references, friendly encouragement, and many helpful com- ments on the manuscript. LOCATION AND PHYSIOGRAPHY The Dry Tortugas, the westernmost outliers of the Florida Keys, are an area of shoals with several small, low islands located about 70 miles west by slightly north of Key West (figure 1). The shoals have the shape of a roughly elliptical atoll with its long axis north- east-southwest. They enclose a lagoon about 10 miles in greatest diameter, its center lying at approximately 24°40'N, 82°52'W. The 10-fathom line closely approaches the outer perimeter of the shoals. Depths within the lagoon are mostly 5 to 10 fathoms. According to Vaughan (1914) the shape and alignment of the shoals were deter- mined primarily by currents and antedate the present luxuriant growth of reef corals. The nearest land is the Marquesas Keys, about 50 miles east. The islands of the Dry Tortugas (Vaughan, 1914; Davis, 1942) are made up of coarse, unconsolidated calcareous sand and larger detrital fragments, chielly the remains of lime-secreting marine or- ganisms. Skeletons of corals predominate. Because of the strong currents and heavy wave action during storms, little fine sediment accumulates and the shorelines of the islands change frequently. Highest elevations on most of the present Tortugan islets do not exceed 8 or 4 feet above normal high tides. Except for Garden Key and Loggerhead Key, all are subject to some overflow by storm tides. Location Map t I ...*»rf ~ Jacksonville -4* S> 96 e " ' North ' Kby ·% .-' . Gulf of Mexico 0* = 1, * 4 Northeast' Key Miami ..'' 6 5.- 4/ i/.. 1Dry Tortugas ··r 43'* Fvt'"F -/ ' 4.1.:$ . 7 · L...15' Hospital Key s. Midae. 89., -f , 4-East-Key· 4 - ' , Bush Key /Long Key 5, DRY TORTUGAS %' /Garden Key 0 4 (1 Bird · Key ~ ''~ ...% . ' Sif Under 5 'FathomsSou@eat Key* ss .5 ''4, 0 4 Existing islands r -r: Former Island 4 01 Scale (Nautical Miles) 1 1/2 0 1 FIGURE 1. Dry Tortugas, showing shoal areas and the location of existing and former islands. Based on C6ast and Geodetic Sur- vey 585 "Dry Tortugas." Location of former islands from U. S, Coast and Geodetic Survey 47la "Tortugas Harbor and Ap- proaches." 6 BULLETIN FLORIDA STATE MUSEUM : Vol. 8 4 HISTORY AND NAMES OF THE TORTUGAN KEYS According to Herrera's chronicle of the first Florida voyage (Davis, 1935:·21), Juan Ponce de Leon reached the Tortugas 21 June 1518. The islands had been sighted from the east as the expedition was rounding the tip of the Florida Keys some weeks earlier. Herrera speaks of an archipelago of "eleven rocky islets" named "Las Tortu- gas" because many sea turtles were captured there. The Tortugas offered a protected anchorage where sea birds, turtles, and seals (presumably the West Indian Seal, Monachus tropicalis, now ex- ceedingly rare if not extinct) could be taken to augment a ship's food supply. It is likely that the islands were visited frequently during the 250 years following their discovery, but little record of this period survives. TABLE 1. NAMES OF THE ToRTUGAN KEYS2 Tatnall Gednery Gauld Chart Chart Coast Survey Chart 47la Chart 585 1773-75 1829 1858-54 1868-75, 1896 1958 Booby Kay Bird Key Bird Key Bird Key Long Ke~ Long Key Long Key Bush Key East Kay East Key East Key East Key East Key Bush Kay Garden Key Garden Key Garden Key Garden Key Middle Kay Sand Key Sand Key Sand Key Hospital Key Logger Head Loggerhead Loggerhead Logg6rhead Loggerhead Turtle Kay Key Key Key Key Rocky Kay Bush Key Bush Key Bush Key Long Key Bird Kay Middle Key Middle Key Middle Key Middle Key North Kay North Key North Key Sandy Kay North East North East Key Key South West South West South West Kay Key Key * Blanks indicate that no island existed at the time of the survey. The frst modern chart, and the earliest I have seen that gives names to the individual keys, was based on a surv6y made by George Gauld for the British Admiralty in 1778-75 (Gauld, 1790). Gauld's chart applies the name "Dry Tortugas" to the group as a whole and shows 10 keys; the names it gives for 6 of these differ from those used later (table 1). The Dry Tortugas were next charted by Lieu- 1964 ROBERTSON: DRY TORTUGAS TERNS 7 tenants Josiah Tatnall and G. R. Gednery for the United States Navy Department in September 1829. A tracing of this chart is in the files of Castillo de San Marcos National Monument, St. Augustine, Florida (C. R. Vinten, in litt.). The 1829 chart has particular value because it gives areas and elevations for 6 of the 11 keys then emerged. Parties from the United. States Coast Survey worked at the Tortugas in 1853-54 ("Tortugas Island", Scale 1:81,680; and «Sec- tion No. VI", Scale 1:400,000, in Bache, 1858), and in 1868-75 ('T- 1410", Scale 1:10,000, in Coast Survey, 1878; Chart 47la, "Tortugas Harbor and Approaches", Scale 1:40,000, United States Coast and Geodetic Survey, 1896). The chart of the area presently in use is Coast and Geddetic Survey 585, "Dry Tortugas", Scale 1:80,000, first issued in 1922 and last revised in 1958. Table 1 shows the keys of the Dry Tortugas that existed at the time of each of the above sur- veys and the names applied to them on the various charts. Two general types of keys may be distinguislred in the Dry Tor- tugas, those little more than barren sandbanks slightly elevated above normal tides, and the larger, higher, and usually more permanent islands with considerable plant cover. The first group includes Hos- pital, Long, Middle, North, Northeast, and Southwest Keys; the second, Bird, Bush, East, Garden, and Loggerhead Keys. Bird Key was the principal nesting ground of Sooty Terns at the Dry Tortugas from at least 1882 (Audubon, 1885) and of Brown Noddies from at least 1857 (Wurdemann, 1861) until the island washed away in the early 19309. During periods of military activity at Fort Jeffer- son, Bird Key also served at times as a hospital site, quarantine sta- tion, and cemetery. The former hospital buildings later housed the Audubon and Biological Survey wardens guarding the tern colony. "The 1829 survey recorded the area of Bird Key as 4 acres 2 roods 20 poles", slightly more than 4% acres, and the elevation as «8 feet 8 inches" (Vinten, in litt.). Later comments on its areas dimensions, and 61evation vary Widely. The area in 1890 was stated as "about eight acres" (Scott, 1904: 278), in 1910-18 as "about 6,000 square yards" (Watson and Lashley, 1915: 85) and as "somewhat less than 5 acres (Lashley„ 1915: 61), in 1915 as "8 acres" (Pearson, 1915: 412), in 1918 as "about 6 acres" (Ashe and Lowe, 1918 MS.), and in about 1926 as "less than five acres" (England, 1928: 14). Dimensions giyen in various publications range from 500 x 250 feet in 1904 (Millspaugh, 1907: 283) to 400 x 300 yards in 1907 (Watson, 1908: 191), and the key is credited with various elevations up to "6 feet above mean tide level" (Watson and Lashley, 1915). A comparison of the representa- 8 BULLETIN FLORIDA STATE MUSEUM Vol. 8 tions of Bird Key on the charts of different periods suggests that much of the reported variation existed mainly in the eye of the obsefver, It is commonly stated that Bird Key was destroyed by a hurricane in 1935, the Labor Day hurricane that devastated the Florida Keys often being specified (Stevenson, 1988; Davis, 1942; Vinten, 1943; Sprunt, 1946b, 1948b). Other authors cite the "hurricane of 1988" (Robinson, 1940: 8; Peterson, 1950: 818) and "the big hurricane of 1988" (Peterson and Fisher, 1955: 142) as the storm re5ponsible. Many accounts suggest that the key was destroyed suddenly. Dilley (1950: 67) Wrote: "At times changes may be very sudden, as illustrated by the complete disappearance of Bird Key during the hurricane of 1985." Stevenson (1938) noted that Bird Key had be6n eroding grad- ually for some time before the 1985 storm, and Robinson (1940), Peterson (1950), and Peterson and Fisher (1955) state that it began to "sink" in 1928. The disappearance of Bird Key appears to have been an extended process following destruction of the vegetation, and without immedi- ate relation to any of the storms mentioned. In 1882 the key had a thick cover of bushes (Audubon, 1885), and in 1857 Wurdemann (1861: 426) described it as "covered with bay cedar ISuriana mari- timal bushes seven or eight feet in height interspersed here and there with the cactus." Later descriptions of the vegetation up to 1910 are almost identical to Wurdemann's. As early as 1904, however, some erosion had begun. Millspaugh (1907: 238) noted from Lan- sing's observations: "Wave action from the northwest appears to be rapidly eroding the western beach, the vegetation on the shore plain- ly showing the enchroachment." The severe hurricane of 15-17 October 1910 (Tannehill, 1950: 175-176) was the frst important eveht in the destruction of Bird Key. Of its effects Lashley (1915: 62-68) wrote: "The Key was formerly overgrown thickly with bay cedars, but the greater number of these were killed by the hurricane of 1910 and only a few living cedars remain." In 1915-16 the effects of the 1910 storm were still evident. Bird Key then had only scattered patches of bay cedar bushes (Bow- man, 1918: 124). On 10-11 September 1919, another severe hurri- cane passed directly over Dry T6rtugas (Tannehill, 1950: 186-187). In his assessment of the damage done on Bird Key, Warden. T. J. Ashe (1919 MS.) wrote: "All vegetation on island destroyed." Accounts of Visits to Bird Key after 1919 (Bartsch, 1928, 1981, 1982; England, 1928) trace the rapid erosion of the denuded island. The later stages are indicated in the following comments by Charles I. Park (in litt.): "When I went there in 1929, Bird Key had already 1964 ROBERTSON: DRY TORTUGAS TERNS 9 started to wash away. The house which the former warden had oc- cupied was considered unsafe so I lived on Garden Key and com- muted by boat to the other keys. ... Each year erosion on Bird Key progressed until in 1984 there was very little of the island above water level." As of June 1985, Longstreet (19860.37) stated: "the remains of Bird Key [are] now eroded to a negligible sandbar." /. .. .. . ·•4 ·· ../ &2 ..; . P.. A~- JIS -~:7 1-- '- - ··'-··•*-Se'lp'*~0~~~~ ... FIGURE 2. Aerial view looking west, Dry Tortugas, January 1945. Long Key in foreground, Bush Key and Garden Key next rear, and Loggerhead Key in background. The white spot on the shoal to the left of Garden Key and slightly~ above it is a sand bar at the former location of Bird Key. (Official photograph, U. S. Navy.) The 1985 Labor Day hurricane was a storm of extreme intensity but small diameter that struck the central Florida Keys (Tannehill, 1950). In reply to questions about this storm and the one of 4-6 November, which was the only other hurricane in the area in 1985. Gordon E. Dunn of the United States Weather Bureau, Miami, wrote me (in Ntt.): "Neither of these storms passed very close to Dry Tor- tugas or to Bird Key, and it is doubtful that either of these storms should have primary responsibility for the disappearance of Bird Key, I would expect that their effect on Bird Key would have been rela- tively minor." After storms in January 1940 (Felton, 1940 Ms.), a 40-foot sandbar elevated 2 feet above high water emerged at the former location of Bird Key. Other intermittent reappearances have occurred more recently (figure 2). Bush Ke!/, where most of the Sooty Terns and Brown Noddies have nested in recent years, has an involved history complicated by con- fusion of names. The names Bush Key and Long Key have been 1() BULLETIN FLORIDA STATE MUSEUM Vol. 8 applied at various times to each of the two adjacent islands on the shoal east of Fort Jefferson (table 1, figures 2 and 8). The names in current use were established with the first edition of Chart 585 (1922), but the confusion persisted somewhat longer (viz., Coast Pilot, 1936: 78). FIGURE 3. Aerial view looking east. Garden Key and Fort Jefferson in fore- ground. Bush Key, the east spit, and the northernmost sand ridge of Long Key at upper right. Large trees in the parade of Fort Jefferson are mainly button- woods (Conocarpus erectus), possibly remnants of the original stand. Pilings at the north and south extremities of Garden Key formerly supported the coaling docks. The center of Bush Key is a thicket of bay cedar (Suriana maritima) enclosing several mangrove-fringed ponds. Brown Noddies nest at the edges of this area. The Sooty Tern colony occupies open areas between the bay cedars and the shore. (Official photograph, U. S. Navy, by U. S. Naval Air Station, Key West, 1959.) Now the second largest of the Tortugan islets, Bush Key has under- gone several cycles of building and erosion. Gauld's chart shows no land in the area. The 1829 survey reported (as "Long Key") an island with an area of «5 acres 8 roods 22 poles" and an elevation of "2 feet 4 inches" (Arana. in litt.). By 1882 this island (or possibly Long 1964 ROBERTSON: DRY TORTUGAS TERNS 11 Key) was thickly covered with bushes and low trees, and Audubon (1885) referred to it as "Noddy Key" because most of the population of Brown Noddies was nesting there. Maps of 1858-54 (Bache, 1858) show a sizable island at the present location of Bush Key. During the military occupation of Fort Jefferson in the 18609, the island served as a pasture and slaughter grounds for cattle and hogs brought in as food for the garrison (Holder, 1868: 262; Manucy, 1948: 321). Shortly after this time, Bush Key and Long Key are said to have been, "almost entirely obliterated by a hurricane" (Holder, 1892: 77). About 1889 Bush Key was a barren sandbank (Coast Pilot, 1889: 40) and Chart 47la of 1896 shows only a small area above high water. Scott's (1890) detailed account of Tortugan geography as of the spring of 1890 mentions no land at- this location. The history of Bush Key after 1900 is thoroughly bedeyilled by confusion of names. Significant observations on T6rtugan geography in this period were made by Lansing in 1904 (Millspaugh, 1907), Bow- man in 1915-16 (Bowman, 1918), and Bartsch in 1917 (Bartsch, 1919). Millspaugh does not mention Bush Key. Bowman (1918: 128-129) describes a large, irregular island that had shrubs about 12 years old. Bartsch (1919: 469, ·482) refers to Bush Key as «an elevated coral reef" with the statement "all the vegetation, in fact, most everything shiftable above the sea, has long since been swept away by the waves. These records appear to show that Bush Key did not exist in 1904, built up rapidly until 1915-16, and then was suddenly reduced and devegetated (presumably by storms) to produce the conditions Bartsch found in 1917. Dayis (1942: 187-189) and Sprunt (1948b: 5-6) adopt approximately this interpretation of its history. Davis ' also points out that Lansing may have overlooked a small island in 1904 because a considerable quantity of sand was removed from the area in 1901-05 for lise as fill during the construction of coaling sheds and piers on Garden Key (figure 8). The record is open to the alternative interpretation that Bush Key had a history of steady growth from before 1900. Close ex- amination of the accounts of Millspaugh (1907) and Bartsch (1919) strongly suggest that these authors, following the nomenclature of the charts then current, referred to the present Bush Key as Long Key and vice versa. Bartsch (1919: 469), for example, wrote of Long Key: "the northern end consists of a barren rim of coral boulders that curves eastward and southward, to join with the reef fringe of Bush Key. This is a fairly accurate description of present geogra- phy with the names of the keys revers6d. Bowman (1918) discussed 12 BULLETIN FLORIDA STATE MUSEUM Vol. 8 Bush and Long Keys together but appears to have followed present usage in his application of the names. "Long Key" is described in Millspaugh's (1907: 225) account as, «so low as to be awash during heavy weather" and "void of vegetation." Bartsch (1919) indicated that the southern part of "Long Key" supported a sparse vegetation of grass and bushes in 1917. The probable history of Bush Key may be summarized as fbllows. After having existed as a well-vegetated island for 40 or 50 years, it was destroyed by a hurricane around 1870. Sandbars soon re- appeared at the site, but as late as 1904 they were small and had no permanent vegetation. During the next decade some plants be- came established and a serjes of ridges and bars developed as shown in Bowman's (1918) sketch. By 1915 (Bowman, 1918) or 1919 (Davis, 1942) several of the sandbars had grown together to cut off ponds from the ocean. Most of the area between the coalesced bars grad- ually filled[ and a lohg sandspit built up from the east end to give the key approximately its present shape (Davis, 1942; figure 3). Bush Key continued to build up during the 1980's and 1940's and contained an estimated 110,000 square yards in 1946 (Sprunt, 1946b; 5). More recently some of the shores have eroded, but the island seems to be more or less stabilized at about 20 acres. East Key appears on all maps of the area and, unique among the present islands of the Dry Tortugas, it has borne the same name throughout its history. Although more stable than many Tortugan islets, East Key has undergone substantial changes in size and vegeta- tion. Gauld's chart shows it as the second largest island of the group. This is corroborated by the 1829 survey which recorded an eleyation of more than 4 feet and an area of about 12 acres, second in size only to Loggerhead Key. During the late 1800's and early 1900's, East Key may have suffered several periods 6f devegetation and erosion. About 1860 (Holder, 1892) it was covered with a dense stand of bay cedar bushes and numerous mangroves. In 1875 it was repofted to be "partly covered with a growth of cedar" (Coast Sur- vey, 1878) and later (Coast. Pilot, 1889) was said to have "a few bushes on it." At almost the same time, Scott (1890: 802) wrote of East Key: "It is a low, sandy, coral island, covered in parts with stunted bushes, and contains an area of perhaps eighteen acres." By 1904 little but herbaceous growth persisted and Millspaugh (1907: 224-225) described East Key as "little more than a mere sand bank 280 x 50 feet in area." He may, however, have been misinformed about its size. In 1915-16 (Bowman, 1918: 181-182) the island was said to be "almost entirely 1964 ROBERTSON: DRY TORTUGAS TERNS 18 covered with vegetation including "large, well-grown bushes," but no bay cedar. Its dimensions were given as about one-third mile long and less than one-sixth mile wide. Davis (1942: 191) found a thicket of bay cedar on the highest sand ridge and reported the island's dimensions to be about 1200 x 600 feet. He stated that East Key "has probably grown in size and become more stabilized in the past half-century." Sprunt (1948b: 17) wrote of East Key: "It com- prises about 85,000 square yards," indicating continued growth. At present sizable bushes of bay cedar, sea lavender (Tournefortia gnophalodes), and Scaevota plumieri are well distributed over East Key. It has often been said t:hat Sooty Terns and Brown Noddies were not known to have nested on East Key and several authors have re- marked upon the failure of the terns to use so suitable a nesting area, These comments overlook various records of the 19th century. Large breeding colonies of both species occupied East Key in the 1850's (Wurdemann, 1861; Bryant, 1859a). Sooties, at least, still nested there as late as 1890 (Scott, 1890). Continual persecution by eggers, mentioned by every early writer, may finally have driven the terns from East Key. Though a warden was in residence at Bird Key each nesting season from 1908 on, his surveillance is not likely to have ektended to the outlying islands. It is of interest that no terns have bred on East Key during the past 28 years of strict protection. Garden Key adjoins the best protected anchorage in Tortrgan waters and has long been the center of human activity in the area (Manucy, 1943). Most of the key is occupied by the immense ruin of Fort Jeffer- son (figures 2 'and 8). A lighthouse was built on Garden Key as early as 1825. Construction of the fbrt began in December 1846 and was discontinued about 20 years later with the work still far from com- plete. After use chiefly as a military prison, the post was abandoned in the 1870's. It was reoccupied during and after the Spanish-Ameri- can War and World War I, first as a coding station, later as a sea- plane base and wireless station. Gauld's chart shows Garden Key with an irregular shoreline and the 1829 survey reported its area as about 7% acres. An interesting map in the files of the U.S. Corps of Engineers (Bache, 1845 MS.) is a detailed topographic survey of Garden Key as it was immediately before the construction of Fort Jefferson began. The shape is rough- ly elliptical, highest land elevations are just over 5 feet above mean low water, and the center of the island is shown as low and evidently swampy. The exact scale of the map is uncertain. Calculations (by 14 BULLETIN FLORIDA STATE MUSEUM Vol. 8 William M. Alexander, Assistant Park Engineer, Everglades National Park) based upon the scale taken from a superimposed outline draw- ing of the Fort, laid out on the original map presumably by Major Bache, give a land area of 8.8 acres above high tide line. The size of Garden Key was increased by filling when the Fort was being built and again about 1900 when the coaling structures were built. Davis (1942: 185) gave the area as 16 acres, of which 5 acres lay outside the walls of the Fort. No terns are known to have nested on Garden Key until relatively recent years. Detail shown on the 1845 map suggests that the in- terior of the island may originally have been too heavily vegetated to attract nesting Sooties, although much of it was apparently suit- able for Brown Noddies. Any that may have nested there undoubt- edly were displaced soon after 1845. A few pairs of Brown Noddies have nested on pilings and in the ruins of the north coaling dock in a number of years since at least 1932 (Bartsch, 1982). In 1987 (Long- street, 1987), 1938 (Beard, 1988), and 1947 (Sprunt, 1947a) large num- bers of Sooty Terns nested along the east side of Garden Key. A substantial part of the Brown Noddy population also nested there in 1987 and 1938, but not in 1947 (Sprunt, 1948a). Hospital Key, although always a small, shifting sandbar with little yegetation, has existed since the earliest surveys of the Dry Tor- tugas. The present name, which was used as early as 1875 (Coast Survey, 1878) stems from the isolation hospital for yellow fever patients built there in the 1860's. Sand Key, an earlier name, re- mained in common use until the 1940's. Various plants have been recorded from Hospital Key, but the island is so often awash in rough weather that no permanent plant cover ha5 become estab- lished. Least Terns nested on Hospital Key in 1907 (Watson, 1907) and 1987 (C. R. Mason, in litt.) and a colony of Roseate Terns has occu- pied the key in a number of recent years since 1987 (Mason, in liM.). Sprunt (1948b: 17) suggested that Sooty Terns might End Hospital Key a suitable nesting area, a prediction fulfilled when a few Sooties nested there in 1957 and 1959. Loggerhead Key is the largest, highest, and most heavily vegetated of the Tortugas and the site of the 150-foot Loggerhead Light (fig- ure 4) built in 1856-60. The size and shape of the key have been remarkably constant: It had an area of about 80 acres in 1829 and is, approximately the same size at present; erosion of the west shore 1964 ROBERTSON: DRY TORTUGAS TERNS 15 having been balanced by the growth of sandspits at the northeast and southwest ends. Loggerhead Key has been credited with an elevation of 9 feet above mean tide (Millspaugh, 1907: 235; Davis, 1942: 179) but it seems likely that this estimate is excessive. The 1829 survey gave the elevation as -4 feet 4 inches." Least Terns nested on the Loggerhead Key sandspits intermit- tently from before 1900 to 19,86 (Russell, 1988 MS.: 4). No other tern is known to have nested on the island. 4 FIGURE 4. Loggerhead Key about 1945, looking southwest from the north tip. In the foreground is the former site of the Tortugas Laboratory, Carnegie Insti- tution of Washington. In the center of the island, Loggerhead Light, (Omcial photograph, U. S. Navy, by U. S. Naval Air Station, Key West.) Long Key is a bar or shoal of reef debris with several dune-like ele- vations of broken coral (figure 2). Davis (1942: 189) estimated that more than one-third of the key was flooded by normal high tides and that the sparse vegetation of herbaceous halophytes and scattered small mangroves covered less than one-third of the area above high tide. If allowance is made for apparent confusion of names in the past (see Bush Key), it appears that Long Key has never been greatly different. Gauld's chart of 1778-75 which shows a small island at the north end of the bar and below it the notation "Ridge of rocks almost dry and very steep, closely approximates present conditions. 16 BULLETIN FLORIDA STATE MUSEUM V61. 8 A few Least Terns occasionally nested on the higher sandbanks at the north end of Long Key as late as 1948 (Sprunt, 1948c). Roseate Terns have nested there from time to time, most recently in 1962 (Robertson, 1962). Some Sooties and Brown Noddies probably nested there in 1982 and 1983 (Bartsch, 1982, 1983). In 1948 (Budlong, 1944 Ms.), 1952 (Moore and Dilley, 1958), 1956 (Robertson, 1956 MS,), and perhaps in other years, many Sooties have tried to nest in rocky spots between the dunes and farther south on Long Key, but because even moderate storm waves wash over this section, the attempts are be- lieved to have been largely unproductive. Middle Ke!/ is shown on Gauld's chart as a fair-sized island, and the map symbols indicate that it supported some vegetation at that time. In 1875 (Coast Survey, 1878) Middle Key was still considerably larger than it is now but without established vegetation. More recently the key has existed only intermittently as a low strip of bare sand with few or no plants. Several pairs of Least Terns may have nested on Middle Key in 1947 (Sprunt, 1948a), and a small colony of Roseates nested there in 1958 (DeWeese, 1953 MS.), and possibly also in 1960. Gauld's name for the island, "Bird Kay," suggests that it was once a more important nesting locality. North Key, Northeast Key, and Southwest Key all were barren sand islands that had washed away by 1875 (Coast Survey, 1878). They have shown no tendency to reappear, but the former location of Southwest Key is marked on present charts as bare at low water. No plants are recorded from any of these keys and no terns are known to have nested on Southwest Key, which may never have been much more than a high place in the reef. Northeast Key harbored a large colony of Royal and Sandwich Terns in the late 1850's (Bryant, 1859a). The ohly definite reference to nesting on North Key seems to be Holder's (1892: 155) mention of "a solitary gull's egg" (from the context possibly a Sooty Tern egg) found on the bare summit of a sand ridge. In addition, Bartsch (1919: 492-498) believed that the island-about 8 miles northeast of Tortugas Lighthouse-"a small sand-bar a few acres in extent, called Booby Island"-where Audu- bon found large numbers of some species of Booby, was probably North Key. 1964 ROBERTSON: DRY TORTUGAS TERNS 17 SOOTY TERN Appearance and behavior combine to make the Sooty Tern a con- spicuous bird, and it has usually been the most abundant species in the Tortugas terneries. Little wonder, then, that the crowded and noisy breeding colonies of Sooties have claimed most of the attention of observers who visited the Dry Tortugas. Perhaps inevitably, much of the comment on the Sooty Tern at the Dry Tortugas has centered on the question, how many? Early ornithologists contented themselves with word pictures that suggest merely large numbers of birds, but few 20th century authors have failed to attempt a numerical reckoning of the size of the colony. Their 8gures range in quality from guesses made after brief observa- tion to estimates calculated from measurements of colony area and density of nests. Table 2 shows what I consider the soundest figures available for numbers of adult Sooties in each year of record from 1903 to 1956. Population figures for several of the years have had an eventful history in the hands of compilers, and quantitative data were found for a number of years previously thought to be gaps in the record. With these corrections and additions the broad outline of the history of the colony seems clear, though many details remain obscure. The Dry Tortugas ternery has been called 'The Oldest Bird Col- ony (Peterson, 1950) on the assumption that its known history reach- es back to the discovery of the area in 1518. It is reasonable to sup- pose that the "other birds" of Herrrera's statement (Davis, 1985), "... there were killed many pelicans and other birds that amounted to five thousand...", included Sooty Terns. The accounts of other early visitors, such as John Hawkins (Longstreet, 1986a), and much later ones, such as George Gauld (1796), contain similar imprecise allusions to the abundance of sea.fowl at the Dry Tortugas. No cer- tain record of any tern is known for the area, however, prior to Audu- bon's visit in May 1882. RECORD OF NESTING 1832. Audubon (1835: 268-269) reported Sooties breeding in great numbers on Bird Key and Noddies breeding on Bush Key. His account shows that both colonies were then being heavily exploited as a source of food. Besides several references to the killing of adult birds and the gathering of eggs it includes the following: "At Bird Key we found a party of Spanish Eggers from Havannah. They had already laid in a cargo of about eight tons of the eggs of 18 BULLETIN FLORIDA STATE MUSEUM Vol. 8 TABLE 2. BREEDING POPULATIONS OF SOOTY TERNS AT THE DRY TORTUGAS Number Year of Method Reference Adults 1908 5000 Estimate Burton (in Dutcher, 1904) 1907 18;858 Area x Density Watson (1908) 1909 40,000 Estimate Peacon (1909 ms.) 1911 48,000 Estimate Peacon (1911 ms.) 1912 48,000 Estimate Peacon (1912 ms.) 1918 80,000 Estimate Peacon (1918 ms.) 1914 97,500 Estimate Peacon (1914. ms.) 1915 102,000 Estimate Ashe and Bethel (1915 ms.) 1916 66,000 Estimate Bethel (1916 ms.) 1917 80,000 Estimate Ashe (in Pearson, 1917) 1918 100,000 Estimate Ashe (in Pearson, 1918) 1919 110,000 Estimate Ashe (1919 ms.) 1929 80,000 Estimate Park (ms. notes) 1935 30,000 Estimate Mason (I986) 1986 40,000 Estimate Doe and Russell (1986) 1937 72,000 Area x Density Longstreet (1987) 1938 64,057 Area x Density Beard (1988) Direct Count 1939 70,000 Estimate Robinson (1989) 1940 100,000 Estimate Robinson (1940) 1941 100,000 Estimate Peterson (1950) 1942 65,000 Estimate Budlong (in Vinten, 1943) 1943 100,000 Estimate Budlong (1943 ms:) 1944 180,000 Estimate Vinten (in titt:) 1945 109,000 Area x Density Sprunt (1946a) 1946 97,200 Area x Density Sprunt (19466) 1947 64,276 Area x Density Sprunt (1947a) Direct Count 1948 104,000 Area x Density Sprunt (19480) 1949 120,220 Area x Density Dilley (1950) 1950 190,876 Area x Density Moore and Dilley (1958) 1951 167,770 Area x Density Moore and Dilley (1958) 1952 76826 Area x Density Moore and Dilley (1958) 1953 84,569 Area x Density Moore (1954 ms.) 1954 88,776 Area x Density Moore (1954 ms.) 1955 71,102 Area x Density Moore (1955 ms.) 1956 90,452 Area x Density Robertson (1956 ms.) Direct Count 1964 ROBERTSON: DRY TORTUGAS TERNS 19 this Tem and the Noddy. On asking them how many they supposed they had, they answered that they never counted them, even while selling them, but disposed of them at seventy-five cents per gallon; and that one turn to market sometimes produced upwards of two hundred dollars, while it took only a week to sail backwards and forwards and collect their cargo. Some eggers, who now and then come from Key West, sell their eggs at twelve and a half cents the dozen; but wherever these eggs are carried, they must soon be dis- posed of and eaten, for they become putrid in a few weeks." Sprunt (1948b:8) points out that Audubon's account contains nothing definite about the number of terns. Despite this, later writers almost without exception have supposed that Audubon found Sooty Terns in far greater numbers than were ever seen again at the Dry Tortugas. The statement, "both species were on their respective breeding-grounds by millions," has been cited both as evidence of former abundance and as typical of Audubon's bent for extravagant language. Although attributed to Audubon, in fact it is only re- ported by him as the remark of an omcer of the Marion, made as the ship approached Dry Tortugas and before Audubon had seen the tem colony. Peterson (1950: 818) used one of the statistics of the Cuban egg trade cited above to obtain an estimate of the number of terns in the «colony in 1882. He wrote: A sooty's egg weighs about thirty grams, or about fifteen eggs to the pound. Eight tons would come to about 240,000 eggs. As sooties and noddies normally lay but one egg this shows irrefutably that the concentration. was far larger than it is now." A repetition of the exercise (Peterson and Fisher, 1955: 142) arrived at an estimate of about 250,000. Had Audubon mentioned no other statistics, this ingenious reason- ing might indeed be difficult to dispute. The eggers who spoke of an eight-ton cargo, however, also told Audubon that they sometime5 realized "upwards of two hundred dollars" per trip to market. If this is interpreted to have been as much as $250, the 250,000-egg cargo was sold at ten for a penny. This seems too good a bargain in eggs even for 1882, especially as the price in Key West is given as "twelve and a half cents the dozen." The Sooty Tem population can also be estimated on the basis of a return of $250 per successful trip and the stated Havana price of "seventy-five cents per gallon, if the latter is taken to mean fluid egg contents. Worth (1940: 56) calculated the volume of a Sooty Tern egg as 1.95 cubic inches or about I18 eggs to the standard gallon. 20 BULLETIN FLORIDA STATE MUSEUM VoI. 8 At 75 cents per gallon a $250 cargo would amount to about 884 gal- lons, therefore equalling about 89,412 eggs. Some or all of Audubon's statistical. information about egging evi- dently is inexact. Attempts to derive a pepulation estimate from any of the details he gives seem unwarranted. The question of the size of the colony in 1882 can be approached by considering the number of nests Bird Key could accommodate. Although Sooty Terns nest in dense aggregations, a limit of colony compressibility exists. Watson (1908: 200) wrote of Sooties nesting on Bird Key in 1907: "Each pair ... defended a circular territory roughly 14 inches to two feet in diameter." If the smaller figure is taken to represent maximum density of nesting observed by Watson, then the minimum area of the territory of a nesting Sooty was 154 square inches and the maximum density of nesting about 8.4 nests per square yard. Detailed observations of the density of nesting of Sooty Terns on Bush Key were made in 1958-56 on 20 to 80 plots each of 8 square yards distributed throughout the parts of the island judged suitable for nesting. The largest numbef of eggs laid on a plot was 56 (7.00 square yard) on one plot in 1954. Field maps show four instances in which 10 eggs occurred in areas of one square yard within the larger plots. In each case, however, some were located at the edges, and no one square yard area appears to contain 10 entire territories. The average number of eggs per square· yard for all occupied plots and the number of plots that contained one or more eggs were: 1953-8.00 per square yard (14 plots), 1954-8.12 (21), 1955-2.1 (26) (Moore I954 ms., 1955 ms.); and 1956-2.58 (20) (Robertson, 1956 ms.). Measured nesting densities reported for other Sooty Tern colohies are mostly similar to or lower than those found on Bush Key. Data for two breeding seasons on Ascension Island (Ashmole, 1963a: 809), for example, show maximum densities (on plots of 25 square yards area) of 5.28 and 5.00 eggs per square yard; average densities for all plots occupied of 1.95 and 2.00 eggs per square yard. The 1958-56 data from Bush Key suggest that 10 nests per square yard is about the maximum density breeding Sooties will tolerate. Few colonies are this crowded, except locally, because vegetation or terrain limit the number of acceptable nest sites. Nesting Sooties ordinarily avoid areas with dense shrubbery or heavy herbaceous ground cover. Ashmole (1968a) found that nests also were fewer on featureless bare ground deficient in the local clues that enable 1964 ' ROBERTSON: DRY TORTUGAS TERNS 21 a bird to return to the proper egg. The Tortugas ternery, however, lacks extensive bare areas. From the Tatnall-Gednery survey Bird Key is known to have had an area of about 4% acres (21,780 square yards) in 1829. Assuming for the moment that the Sooty Tern colony occupied its entire sur- face, an average density of 11.5 nests per square yard Would be nec- essary to accommodate 250,000 nests. Parts of Bird Key, however, were thickly covered with bay cedar bushes in the 19th century. Photographs taken much later, after the hurricane of 1910 had great- ly reduced the amount of plant cover (e.g., Bartsch, 1919: Plate 18), show large areas still not available t6 nesting Sooties because of the dense bush growth. Therefore, I think it unlikely that the maximum breeding population of Sooty Terns on Bird Key much exceeded 50,000 pairs. Audubon's manner of reference to his visit to Bird Key suggests that he saw tremendous numbers of Sooty Terns. A large subjective element, however, seemingly must be allowed in Verbal descriptions of 8rst visits to Sooty Tern colonies. When Herbert K. Job saw the Bird Key ternery at its lowest ebb in 1908 he wrote (1905: 87) of the Sooties: «There are such clouds of them that accurately to estimate their numbers was impossible ..." This language also could be taken to indicate great abundance were it not for the rest of the Rev- erend Job's sentence which reads: "'. . . but my guess of six or eight thousand I think cannot be far out of the way." It seems characteristic of moderns to suppose that Audubon saw all bird concentrations in their pristine glory. However true this may have been of many places he visited, it does not apply in the case of the Dry Tortugas. Bird Key was adjacent to a fine anchorage, itself adjacent to a major shipping lane that had been used for more than three centuries. That there were no accurate charts before Gauld's survey of the 1770's can scarcely have deterred mariners from using Tortugas harbof. Audubon was told that the terns had frequented Tortugas "since the oldest wrecker on that coast can recollect." It is altogether likely that the ternery was first disturbed on the day of its discovery, and as often thereafter as ships put in to Tortugas in appropriate season. The most that can be assumed is that exploitation up to about 1882 had been infrequent enough to permit the Sooties to rear young in most years. 1840-1902. Although much of this period was. marked by intensive human activity at the Dry Tortugas, the record of the tern colonies in the 19th century after Audubon's visit is limited to observations 22 BULLETIN FLORIDA STATE MUSEUM Vol. 8 by Bryant (1859a), Wurdemann (1861), Maynard (1881), and Scott (1890). Comments by the first three of these authors are brief. Scott discusses the Dry Tortugas in greater detail, but most of his in- formation about terns is hearsay, because the colonies were not active at the dates of his visit, 19 March to 10 April 1890. Data accompanying bird specimens from the Dry Tortugas in several collections show that other ornithologists may have visited the tern colonies during this period, but left little or no published record of their observations. One such visit was by A. L. Heermann and John Krider, probably in May 1848. Howell (1932: 13) mentions this expedition but does not include the Dry Tortugas among the places visited. Heermann (1858: 34), however, lists eggs of the Sooty Tern and Brown Noddy from "Tortugas Islands" presented by him to the oological collection of the Academy of Natural Sciences of Philadel- phia. At least a part of this material (ANSP Catalogue Nos. 32055, 32060, and 32061) is still in the collection (Henry M. Stevenson, in litt.) The specimens bear no date, but H6ermann is known to have visited F16rida only once. In the book in which John Krider sum- marized his career, including "only those species of birds of the United States that I have myself collected and mounted", he refers to the Sooty Tern as follows (1879: 81) : "Common on the Keys of Florida and the Tortugas, where it breeds in large numbers. I have two specimens in my collection. The two main items to be gleaned from the later 19th century papers are: Sooties and Noddies then nested on East Key as well as Bird Key, and the colonies were under increasing pressure from eggers. Dr. Henry Bryant traveled and collected extensively in Florida in the decade 1850-1860, but his obituary in the Annual Report of the Boston Society of Natural History for 1867 gives no details of his work in Florida, and little seems to be recorded elsewhere. All that ha-s been known of his visit to Dry Tortugas is that he was there 6n 8 May. Data on bird specimens he collected now in the Museum of Comparative Zoology (David O. Hill, in litt.) suggest that the year was 1850. Two Sboty Terns (MCZ Catalogue Nos. 42097 and 42099) carry the dates 10 May and 11 May, respectively, with no year; a Great White Heron (MCZ No. 42584) he collected at Sand Key off Key West, however, places Bryant near the Tortugas on 16 April 1850. His visit there can have occurred no later than the 1858 nest- ing season, because on 19 April 1854 he donated his collection of birds' eggs from Florida, including eggs of the Sooty and Noddy to the Boston Society of Natural History. His own account of his visit 1964 ROBERTSON: DRY TORTUGAS TERNS 28 (1859a: 19-21) states merely that he found Sooties and Noddies nest- ing principally on East Key and "in as great numbers as at the time of Audubon," and that Royal and Sandwich Terns were breeding "in great numbers on Northeast Key. In a paper on Bahaman birds (1859b: 184) he remarks that Sooties and Noddies occur there "in immense numbers, as at the Tortugas. Gustavus Wurdemann (1861: 426) described his visit to the Tor- tugas the last week of June 1857 in a letter that accompanied a ship- ment of bird specimens to the Smithsonian, published two years after his death in 1859. *At the Tortugas are two keys or islands, East Key and Bird Key, which serve as places of resort to the noddies and laying gulls to deposit their eggs and raise their young. They are watched closely at East Key by boatmen, who gather the eggs to carry them to Key West for sale . But at Bird Key the birds are under special protection of Captain D. P. Woodbury, the ofRcer in charge of construction of the fortifications.... The keys are covered with Bay cedar bushes seven or eight feet in height, inter- spersed here and there with the cactus, among which some young laying gulls sought refuge. Their eggs are laid on the sand, whilst the noddies lay in nests built from two to six feet from the ground of dried sticks or twigs. Only one egg was found in each noddie's nest, and about two in the laying gull's. Their eggs are said to have been taken some time previous to our visit, and that they lay usually two or three. I picked up several female laying gulls with my hands, and might have caught noddies if I had not been encumbered with the gun, birds, and eggs. No young noddies were seen at this time, which was the last week of June. ... Other specimens in the National Museum (Deignan, in litt.) and a Sooty Tern in the Museum of Comparative Zoology he took there 10 June 1858 (Hill, in litt.) show Wurdemann also visited the Tortugas the next year. C. J. Maynard never visited the Dry Tortugas in person. As Howell (1982: 16) notes "In 1874 he worked at Cedar Keys from January 26 to March 1. From there, in a small yacht he went down the coast as far as Clearwater, but from that point he was obliged to return home on account of illness, leaving his assistants to com- plete the trip, which took them as far as the Tortugas." This may partly explain the several geographical and historical inaccuracies in his account (1881: 480): -The Sooty Terns are now only found in any numbers on the small islands which lie to the southward [sic] of Key West and which are known as the Dry Tortugas. Here they breed on Bird Key which is about four miles [sic] from Fort.Jeffer- son, depositing their eggs early in May. The birds are extremely 24 BULLETIN FLORIDA STATE MUSEUM Vol. 8 tame when nesting, insomuch so, that they may be killed with sticks or even caught with the hand, and they deposit the eggs on the naked sand. There were thousands of these birds on this little key in 1874, but as the soldiers of Fort Jefferson had been in the habit of taking the eggs regularly every other day, but few or no young were raised. The officer who had command of the fort, prohibited shooting the birds On the island, but the continual robbing of the eggs must ulti- mately drive the Sooty Terns from this breeding ground. The actual date of Maynard's assistants' visit is indicated by a Sooty Tern in the Museum of Comparative Zoology (No. 204310, Hill, in litt.) taken there 25 May 1874, though no collector is named. The account contains no first-hand comment on the Noddy, nor does it mention East Key, which lay approximately 4 miles from Fort Jefferson. Regarding the comments on egging by soldiers in the area, the Fort Jefferson ,garrison had been withdrawn 11 January 1874, leaving only a small detail to guard ordnance stores and a much reduced construction crew engaged mainly in closing down the operation (Manucy, 1961 MS.). In addition, Captain Woodbury, the only commanding officer known to have shown an interest in protecting the tern colony, had left the Dry Tortugas in 1860 and died in Key West of yellow fever 15 August 1864 (Cullum, 1891: 496-497; Manucy, 1961 MS.). It seems likely that the report Maynard received from his assistants in 1874 blended considerable hearsay with their actual observations. Most of W. E. D. Scott's information about nesting terns was sent to him after his return from the Dry Tortugas by Dr. F. S. Goodman, who was stationed at the Quarantine Station on Garden Key. Scott reports (1890: 807) Sooties nesting on East Key and Bird Key, and Noddies "mainly confined" to Bird Key, but his comments on egging are of greatest interest: "All of the Gulls and Terns that breed at the Dry Tortugas have been much diminished in numbers in the past ten years. It has always been the custom for some of the boats engaged in fishing and sponging about Key West to resort to these islands during the breeding season, and lately their depreda- tions have really made a very appreciable difference in the birds that resort to this breeding ground. I am told that the eggs have a commercial value as an article of food in the markets of Key West, where barrels of birds' eggs from the Tortugas are brought every season of late years. Vinten (1943: 54) suggests that search of the records of govern- ment agencies that had maintained operations at the Dry Tortugas might reveal additional data about the tern population during this 1964 ROBERTSON: DRY TORTUGAS TERNS 25 period. Though most of the search of the voluminous Army archives of Fort Jefferson remains to be accomplished, the studies of historians show these archives do indeed contain information pertinent to the history of the ternery. Albert Manucy (in litt.) advises me that among records he examined he recalls having seen correspondence relating to the visit of Louis Agassiz to the Dry Tortugas in I858, and that the Fort Jefferson Letter Books include such items as a letter from Mordecai and Co. to Woodbury on 8 May 1859 concerning ship- ment of Woodbury's bird specimens to the Smithsonian Institution. The historical records Manucy (1961 MS.) studied suggest Bird Key suffered even more disturbance than the authors of ornithologi- cal works on the Dry Tortugas have appreciated. Shortly after war began in 1861, for instance, concern for the safety of Fort Jefferson, still unBnished and weakly armed, led to the appropriation early in 1862 of $200,060 to fortify Bird Key. The preliminary survey, in- cluding extensive borings to determine subsoil structure, was de- layed by personnel changes and slow delivery of materials, and was not C0mpleted until the spring of 1864. The project then seems to have lapsed, but it can hardly have failed to disrupt the terns attempt- ing to nest during the survey. Manucy also cites a letter of 18 July 1865 to the Post Commandant from Edward Frost, Assistant Engineer in Charge, complaining of the removal of a number of hogs "from their ranging ground on Long Key to Bird Key" which contained «the scattered graves of many Union Soldiers who have died at this Post during the war." Whether or not the hogs were returned to Long Key seems to be unrecorded. Most probably the Sooty Terns failed to rear young at the Dry Tor- tugas in most of the years from 1860 to the early 1870's when Fort Jefferson was heayily garrisoned. This loss of annual recruitment plus an undoubtedly heavy mortality of adults must have reduced the population rapidly. Little definite information about the ternery exists for the years 1890-1908. It may be presumed that the colony was raided regularly by eggers, and that some time in this period Sooty Terns nested for the last time on East Key. J. W. Atkins, a well-known resident col- lector of Key West, collected specimens now in the Museum of Com- parative Zoology at the Dry Tortugas in May 1896, but no other record of his trip is known. A. G. Mayer visited Bird Key in 1898, but the Only datum published (in Dutcher, 1906) is his impression that Sooties were then about one-third as numerous as at his next visit in 1906. 26 BULLETIN FLORIDA STATE MUSEUM Vol. 8 With the outbreak of the Spanish-American War, Fort Jefferson was garrisoned once more from 1898 until ab6ut 1906. In 1900 the Dry Tortugas were transferred to the Navy Department and con- struction of a coaling depot at Garden Key began. Sprunt (1948b: 9) suggests that the renewed military activity at Fort Jefferson probably put additional pressure on the tern colony. The Navy at the command level was aware of the need to protect Bird Key, for a letter from Captain T. C. Treadwell quoted by Dutch- er (1908:120,121) states Treadwell ordered egging stopped soon after he assumed command of the U. S. Naval Station, Key West, in June 1901. Unfortunately orders from Key West were not altogether ef- fective at the Dry Tortugas, for according to Thompson (1903: 77-78) "the terns suffered very seriously from eggers in 1902. Thompson adds, presumably with reference to the recent past and to both Sooties and Noddies: "There have been years when not a single individual was raised, every egg having been taken shortly after it was laid." Thanks to William Dutcher's untiring efforts sterner measures to protect Bird Key followed in 1908. The Secretary of the Navy issued an order on 24 April prohibiting the taking of eggs or disturbing of terns at Dry Tortugas, and in May W. R. Burton was detailed there as a special warden representing the American Ornithologists' Union with the permission and logistical support of the Navy (Dutchet, 1904). Burton arrived at Bird Key accompanied by H. K. Job 19 May 1903. The modern history of the ternery can fairly be said to begin on that date. 1903. Four estimates of the Sooty Tern population in 1908 are avail- able from the published comments of the original observers. They are: «8600" by Job and Burton inade before Job returned to Key West on 22 May; "at least 5000" by Burton in a letter to Dutcher dated 15 July 1908, the increase accounted for by birds that began nesting after Job's departure; "five to six thousand- by Job ih a letter to Dutcher (all three Bgures published in Dutcher, 1904); and, "six or eight thousand" (Job, 1905: 87). The context of the accounts sug- gests that the figures refer to number of adult Sooties rather than number of nests, but nowhere is this clearly stated. Compilers have given the 1908 populatien as 8600 (Longstreet, 1936a; Vinten, 1948; Sprunt, 1947b; Peterson 1950), 6-8000 (Sprunt, 1948b), ·and "about 7000 nests" (Fisher and Lockley, 1954: 60; Peterson and Fisher, 1955: 142). The figures, where identified, are in all cases credited to Job. I consider the warden's Rgure of 5,000, based upon observation of the 1964 ROBERTSON: DRY TORTUGAS TERNS 27 colony through the entire nesting season, to be the soundest estimate available. I have found no record of the condition of the colony in 1904 and no estimates of the population for the seasons of 1904 through 1906. Charles Russell, the warden in 1905, reported "a very success- ful season" (Dutcher, 1905). After visiting the colony in 1906 A. G. Mayer informed Dutcher (Dutcher, 1906) that the Sooties appeared to be three times as numerous as they were in 1898. 1907 . John B. Watson began his studies of the tern colony in 1907 and also served as the warden of the National Association of Audubon Societies f6r that season. In addition to his other work Watson made a careful estiniate of the nesting population of Sooties. He divided the colony into 10 sections presumably distinguished by conspicuous features of vegetation or terrain. By determining the area and sam- pling the density of distribution of nests within each section, he ar- rived at an estimated 9,429 nests or 18,858 breeding adults (Watson, 1908; 198). 1908. Most summaries of the changes in size of the Tortugas Sooty Tern population include an estimate of 20,000 (or 10,000 nests) as the population in 1908. All authors who cite an auth6rity credit this figure to Watson who, according to the Carnegie annual reports, was not at the Dry Tortugas in 1908 or 1909. The earliest reference I Bnd to it is Lashley's (1915: 61) statement that Sooty Tern nests to- talled "more than 10,000 in 1908," with no mention of the source of his information. I have omitted the figure from table 2 because I can Rnd no authority for it. 1909-1916. On 6 April 1908 Executive Order No. 779 of President Theodore Roosevelt established the. Tortugas Keys Reservation for protection of birds nesting in the area. The order specified that use as a bird reservation was not to interfere with military uses (under President Polk's Executive Order of 17 September 1845 establishing the Dry Tortugas Military Reservation) except that military use of Bird Key was prohibited. Protection of the Tortuga5 Keys Reserva- tion became the responsibility of the Bureau of Biological Survey. After 1908 warden protection at the Dry Tortugas was supported jointly by the Bioldgical Survey and the National Association of Au- dubon Societies. T. J. Ashe of Key West, who was in general charge of bird protection activities in the Florida Keys during most of the ensuing decade, hired and supervised the men stationed at Bird Key. 28 BULLETIN FLORIDA STATE MUSEUM Vol. 8 These were John Peacon (1909-1914), Ludwig Bethel (1915-1916), and William E, Lowe (1917-1919). Warden's reports on the condition and size of the tern colony were made annually to both supporting organizations. From the annual reports to the Biological Survey I have seen only the data entered in the bird distribution file now at Patuxent Wildlife Research Center of the Bureau of Sport Fisheries and Wildlife. The annual reports to the National Association of Audubon So- cieties for this period are still in the files of that organization. Some were prepared by Ashe and submitted in his name; others seem to have been prepared by his wardens at Bird Key. They include esti- mates of the population of Sooties in all the years 1909 through 1916 except 1910. Watson apparently prepared the 1910 rep6rt for Bird- Lore, but it was not published and has beeh lost. This is unfortunate because comments in a later report (Watson, 1912 MS.) indicate he made the 1910 count of Sooties by the same method he used in 1907. Previous summaries of the colony include no mention of Sooty Terns in these years, but skip directly from the questionable 1908 8gure to 1917. In addition to the annual warden's reports, several published comments for this period have been generally overlooked. Of the population in 1913 Watson and Lashley (1915: 88) wrote: "There are probably more than 18,000 (possibly 80,000) sooties on Bird Key." On 28 May 1915 Herbert K. Job and H. R. Mills visited Bird Key to take motion pictures for the National Association of Au- dubon Societies. A brief excerpt published from Job's report (Pear- son, 1915) gives the number of Sooties as -possibly 75,000." Pearson also prepared a longer article about this trip (1915 Ms.), evidently copy for Bird-Lore that wasn't used, which quotes more extensively from Job. It reveals that the 75,000 population figure was based on area-density calculations by Mills. Because these calculations con- tain obvious inaccuracies impossible to resolve today, I have used the 1915 population estimate from the warden's report in table 2. 1917 . According to Warden T. J. Ashe's annual report (Pearson, 1917: 398) "... there were probably 80,000 of these birds [Sooty Terns] nesting on the island." This figure has been overlooked by compilers, who instead have misquoted the 1917 population of Sooties from Bartsch as -18,000" (Longstreet, 1936a; Fisher and Lockley, 1954) or "25,000" (Vinten, 1943; Sprunt, 1947b, 1948b). Bartsch's list of the birds seen at the Dry Tortugas 19-31 July 1917 (1919: 471) in- cludes under Sooty Tem "adult 1IB,000 young 27,200." The figures are keyed to footnotes that read: -1Based upon Doctor Watson's cen- 1964 ROBERTSON: DRY TORTUGAS TERNS 29 sus of 1908.", and "2An estimate admitting two-fifths as many off- spring as we had parents." Bartsch also (1919: 473) wrote of the 4,Sooties: . . . probably more than 25,000 are present on Bird Key at the close of the breeding season." Apparently Bartsch made no independent estimate of the Sooty Tern population in 1917; the- Bgure 18,000 is an approximation of Watson's total for 1907 (not 1908) and the "more than 25,000" is merely 18,000 adults plus Bartsch's arbitrary 0gure for young of the year. 1918-1934. None of the earlier compilations mentions these years. I have seen warden's reports only for 1918 and 1919. A Federal law effective 1 July 1919 ended the National Association of Audubon Societies' participation in the protection of Bird Key (Pearson, 1919). The Biological Survey continued to employ a warden at the Dry Tor- tugas during the summer at least through 1930, but no wardens' re- ports can be located in the files now stored at Patuxent Wildlife Research Centef (Robbins, in litt.), and the distribution files contain only the warden's estimate of the Sooty Tern population in 1929 (Park, Ms. notes). Several popular articles published in the 1920's refer in passing to the number of terns at the Dry Tortugas. England (1928: 86) men- tions a population of 50,000, and a photograph in an article by Long- ley (1927: 66) is captioned: "The west shore of Bird Key showing some of the 83,000 birds that breed here annually. These 8gures are not considered bona #de population estimates. Neither can be associated with a definite year, and the 88,000 is suspiciously near Bartsch's (1919: 471) total of 82,810 for all the birds (19 species) he identified at the Dry Tortugas in July 1917. Bartsch visited Bird Key several times during the 1920's and in August of 1931, 1982, and 1938. Existing records of his trips contain no reference to the t6tal numbers of Sooty Terns. The brief published accounts of the later visits (Bartsch, 1981, 1982, 1983) have great in- terest because Bird Key was then eroding rapidly. In 1982 Bartsch noted that a few Sooties were nesting on Bush [Long] Key. The following year he reported (1988: 267) that more than half the popu- lation had left Bird Key and "It is beginning to look as if the major portion would eventually establish itself on Long [Bush] Key." C. C. Von Paulsen of Homestead, Florida, then an officer of the U. S. Coast Guard, visited the Dry Tortugas frequently ,in the years 1982- 1984. As he remembers it (personal communication) a substantial part of the Sooties nested on Bird Key in 1983 and smaller numbers remained there in 1984. 80 BULLETIN FLORIDA STATE MUSEUM Vol. 8 From the scant information available it appears that the tern colony mayhave been without warden protection in the early 1980's. Charles I. Park, the last Bird Key warden and now a resident of Key West, wrote me in a letter 14 December 1959: «As well as I can remember, I served as warden in the Tortugas area from 1929 through the summer of 1984, a total of six years." As G. A. England (1928: 14) refers to Charles Park as the Bird Key warden during his visit there the 5ummer of 1926 or 1927, apparently Mr. Park began his six years of service two or three years earlier than he recollects. Others who knew the Dry Tortugas in the early 19309 do not recall a warden in the area during those years (Julius F. Stone, Jr,, Charles M. Brookfield, and C. C. Von Paulsen, personal' communications). Absence of warden protection would explain the apparently well- founded rumors that the early depression years saw a vigorous re- newal of egging at the Dry Tortugas. It seems likely that protection of the ternery at least was less vigilant in the nesting seasons of 1931 through 1934, although Sooties are known to have succeeded in rear- ing many young in some of these years (Bartsch, 1932). The National Park Service assumed administrative responsibility for the Dry Tortugas early in 1935. Mason (1936: 18) mentions that the Custodian of Fort Jefferson turned away many boat parties from Key West that came to gather eggs on Bush Key in the spring of 1985. Correspondence in National Park Service files suggests that the colony was raided late in the 1935 season and a number of young birds taken. Protection of the colony by the National Park Service probably was not fully effective until the nesting season of 1986. From 1935 through 1941 one or tw6 groups of observefs visited the Dry Tortugas each June on trips sponsored by the Florida Audubon Society; The visits were brief, each group spending from two to five days at the Tortugas. Adult and young terns were banded in 1937 through 1941, and the published accounts of all the trips, except that of 1941, include estimates of the number of Sooty Terns. In 1987 and 1938 many Sooties nested along the east side of Garden Key (figures 5 and 6) as well as on Bush Key. 1935. The population figure in table 2 is an average of estimates by members of the party (Mason, 1986). Some thought as many as 50,000 Sooties were present. 1936. Doe and Russell (1986: 6-7) state of the published population estimate: "It was the general opinion of those who had been on the 1964 ROBERTSON: DRY TORTUGAS TERNS 81 " Masontrip in 1985 that the tern colony had increased one-third. (Ms. notes) entered an estimate of 48,000 in his field notes with the .comment that he considered it very conservative because the nest- ing colony of Sooties covered a much larger area on Bush Key than it had in 1985. 4,«ir=® 1.r= 4942 -,tl--4 .... M / 1/. *·V-. *f : '- 9- &-· « 1 . 1V':-' :i k - Ut :t 1/7-: I FIGURE 5. Portion of the Sooty Tern colony on the east side of Garden Key in 1987: (top) June; and, (bottom) August, showing many well-grown juveniles. (National Park Service photographs by Philip C. Puderer.) 82 BULLETIN FLORIDA STATE MUSEUM Vol. 8 1937. The colony was said to occupy an arer of 8000 square yards on Garden Key and 4000 square yards on Bush Key. From this area "and a nesting density of about six sooties to the square yard, de- .'termined from one sample plot of 9 square yards in a typical sec- tion" of the Garden Key colony, Longstreet (1987: 8) calculated a total of -72,000 [adult Sooties] actually present at one time." Though 72,000 birds present at one time would represent a total of 144,000 breeding adults by the usual methods of reckoning, Longstreet (1987: 8) continued: "It would seem not far wrong to calculate the number of adult sooty terns at the Tortugas in June 1987, as approximating 100,000. This would be a tremendous increase over any previous estimates, and for that reason may be seriously in error. But, at any rate, it is an estimate based on actual count of birds in a given area, multiplied by the number of times that area is found in the total area occupied by the birds." All summaries of the history of the Tortugas Sooty Tern population have cited the 1937 population as 100,000 from this source. Russell (1988 MS.) also "estimated the number of Sooties to exceed 100,000." Other observers appear to have considered this estimate too high. Young and Dickinson (1937) believed that Bush Key had no more than 20,000 Sooties, and Mason (MS. notes) recorded an estimate of 75,000 for the total adult population. Longstreet (1987: 7) includes a photograph, taken from the terreplein of Fort Jefferson, of the sample plot on which the figure for density of nesting was based. The picture shows most of the Sooties are either incubating or brood- ing small young, and hence distributed one adult per territory. Be- cause of the angle it is not possible to tell exactly how many Sooties are on nests within the 9 square yard plot, but the number is 20 to 80, certainly not 54. Thus, Longstreet's figure of 6 birds per square yard is apparently based on a nesting density 6f about 8 nests per square yard with allowance for the absent member of each pair. Accordingly, 72,000 is considered the soundest estimate of the breed- ing population of adult Sooties in 1937. 1938. This year Sooty Terns again nested on both Garden (figure 6) and Bush Keys, but the colony divided more equally between the two. Considerable effort was devoted to careful measurements of areas occupied and nest densities on both keys, and the resulting es- timate (Beard, 1988) is undoubtedly one of the more accurate of the population Bgures for adult Sooties in the Tortugas ternery. Di- rect counts of nests on the coaling docks (figure 6a) and in small, irregular patche5 of dense vegetation on Garden Key totalled 8950. 1964 ROBERTSON: DRY TORTUGAS TERNS 88 - r , /.-Ii/#/.05,6.- - '*~pr - *#*-*64*.4./4*Mal/*r#* I - 1.. ' . «f .7... - - t; 0¥3111,-4-7,11"'ll :iet_~*bater//4.*·. f.s... ,JAA<12/401"iryLITJF,#Itik#.i'jillib:ki ..Lip...flimi - <:9 .Mw