Systematics

Editor’s Note: Study of the mitochondrial DNA of terns, along with their plumage characteristics, have suggested that the heretofore broadly defined genus Sterna is paraphyletic. Reclassification of this genus now places Royal Tern in the genus Thalasseus. See the 47th Supplement to the AOU Check-list of North American Birds for details. Future revisions of this account will account for this change.

One of so-called “crested terns” formerly all placed in separate genus, Thalasseus . Recent authors merge crested terns in genus Sterna based on molecular evidence (Hackett 1989, Sibley and Ahlquist 1990). Originally described as Sterna maxima by Boddaert in 1783 from “Cayenne.”

Two subspecies recognized following Cramp 1985 and Gochfeld and Burger 1996 . Austral (se. South American) S. m. maxima and w. African S. m. albididorsalis differ in being smaller, slenderer, and longer-winged (especially S. m. albididorsalis) than North American S. m. maxima populations, and in se. South American populations and w. African birds having longer and thinner bills that are usually paler orange, tending toward yellow. In these respects, se. South American and w. African populations more similar to each other than to boreal S. m. maxima . Further study needed to establish relationship of se. South American populations to rest of S. m. maxima, with which they are currently placed; possibility exists that austral populations may be subspecifically or even specifically distinct. Indeed, each of 3 major and allopatric populations (w. African, se. South American, and North American [including both Pacific coast and Atlantic-Caribbean]) could be separate biological species.

Sterna maxima maxima Boddaert, 1783. Boreal population: Breeds in North America on Atlantic Coast from Maryland south along Gulf of Mexico shore into Yucatán, in Sea of Cortez (north sporadically to California) and in limited numbers in West Indies south to Venezuela, Netherlands Antilles, French Guiana, and Tobago. Winters along Atlantic Coast from N. Carolina south along Gulf Coast, throughout Caribbean to n. South America, and along Pacific coast from California to Peru. Austral population: Breeds se. Brazil, Uruguay (presumably), and Argentina, although few sites known as most of area unsurveyed. Apparently disperses unknown distance northward postbreeding, but many may winter in general breeding area. Populations and movements barely known.

Sterna maxima albididorsalis Hartert, 1921. Breeds locally in w. Africa, largely Mauritania but also Senegal and Gambia. Winters north to Morocco and w. Mediterranean, south to Gulf of Guinea.

Population interactions complex and not fully understood. For example, although Yucatán and w. Mexican S. m. maxima apparently mix with other boreal populations of S. m. maxima in areas where courtship begins (on wintering grounds), some have intimated their morphological and possibly taxonomic distinctness. Relationship of austral, summer-breeding populations from Brazil to Argentina described by Escalante (1968, 1985, 1991) and currently ascribed to S. m. maxima, yet no adequate samples to our knowledge examined critically by taxonomists; even its biometrics scant. No banding recoveries of boreal breeders southeast of Suriname, nor of austral breeders north of se. Brazil, despite occasional statements that boreal breeders occur in winter as far south as Argentina (evidence suggests austral breeders). As Escalante (1985: 938) notes, “Virtually nothing known about [breeding] status of Royal Tern along ± 4000 km of northern and eastern coast [of Brazil] from boundary with French Guiana (5°N) to Rio de Janeiro (23°S).” Thus, increasingly likely that austral and boreal breeders wholly allopatric in addition to breeding at opposite times of year.

Curiously, austral-breeding Cayenne Tern (S. [sandvicensis ?] eurygnatha) exactly mirrors this situation, even down to occasional individuals with all-red bills (Escalante 1970). Disjunct populations of Cayenne Tern also not examined exhaustively by taxonomists, although Voous (1968) tantalizingly suggests that Cayenne Tern populations in s. Caribbean and austral-breeding populations different enough to warrant at least subspecific designation. And, as in Royal Tern, apparently large gap between 2 breeding populations in Brazil. Parallels between boreal S. m. maxima and S. [s. ?] eurygnatha versus austral S. m. maxima and S. [s. ?] eurygnatha on one hand, and boreal Common Tern (S. hirundo) and Arctic Tern (S. paradisaea) versus austral South American Tern (S. hirundiunacea) and Antarctic Tern (S. vittata) on other hand most intriguing.

Based on our studies, S. m. maxima and S. m. albididorsalis differ in several features (contra C. S. Roselaar’s comments in Cramp 1985; cf. Malling Olsen and Larsson 1995). Structurally, S. m. albididorsalis ’s bill more similar to Elegant Tern’s and to Great Crested Tern’s in lacking strong gonydeal angle, being relatively slender and gently curving (= less robust) and in tending toward yellow-green rather than red-orange; dorsally, S. m. albididorsalis much lighter than S. m. maxima, and overall smaller and slenderer than S. m. maxima . Gochfeld and Burger (1996), commenting on S. m. albididorsalis, suggest its affinities might lie with austral S. m. maxima, while Roselaar (in Cramp 1985: 35) claims that differences between Royal and Great Crested terns “largely bridged by nominate S. b. bergii from South Africa”—with which our own observations of S. b. bergii in Namibia and South Africa differ sharply.

Species limits in crested terns not worked out, and almost all currently recognized subspecies allopatric. One of few examples of true sympatry occurs between Sandwich Terns (S. s. acuflavida and boreal S. s. eurygnatha) in Netherlands Antilles (Ansingh et al. 1960), where mixed pairs and intermediates apparently prevalent. Still, no one has, to our knowledge, yet looked at mating patterns or selective fate of “hybrids” there. Adding complexity, S. s. eurygnatha seems to be in process of advancing as breeder north through West Indies and beyond (Mitra and Buckley 2000), with first phenotypically clear individual appearing in North America in mid-1980s (Buckley and Buckley 1984b). Still, incidence of hybridization might only reflect genetic similarity and chromosomal compatibility typifying larids in general.

We know of no published molecular studies of any of world’s largely allopatric populations of crested terns. Besides Atlantic/Gulf, Yucatán, Caribbean, Pacific Coast, Argentine, and w. African Royal Terns; Elegant Tern; North American and European Sandwich Terns; recently rediscovered Chinese Crested Tern (S. bernsteini); and West Indian and Argentine Cayenne Terns, there are some 6 currently recognized races of Great Crested Tern and 3 or 4 races of Lesser Crested Tern (S. bengalensis; Howard and Moore 1991, Clements 2000). Even with classic taxonomic techniques, racial limits confused in 2 most widespread tropical species (Great Crested and Lesser Crested), and species limits and population genetic distances never determined molecularly for group. Not inconceivable that each allopatric crested-tern taxon is a valid biological species, and also possible (but less likely) that currently recognized species may be reduced to 2–4. While crested terns surely monophyletic, until objective molecular support for reality of superspecies concept advanced, we make no predictions about Royal Tern’s closest relative(s).