4 October 1994 PROC. BIOL. SOC. WASH. 107(3), 1994, pp. 429-435 A GIANT PRESBYORNIS {AVES: ANSERIFORMES) AND OTHER BIRDS FROM THE PALEOCENE AQUIA FORMATION OF MARYLAND AND VIRGINIA

Storrs L. Olson

Abstract.•Presbyornis isoni, new species, is described from a humérus and an alar phalanx from marine sediments of the late Paleocene Aquia Formation in Maryland. About the size of the smallest living species of crane (Gruidae), it was much larger than any previously known member of the Presbyomithidae. Other fragmentary bird remains from the Aquia Formation are noted, several of which may be referable to the suborder Phaethontes of the Pelecaniformes. Additional Paleocene birds from eastern North America occur in the Hor- nerstown Formation of New Jersey, which is now considered to be Danian (early Paleocene) in age, rather than late Cretaceous.

The marine sediments of the Aquia For- Genus Presbyornis Wetmore, 1926 mation crop out in the Chesapeake Bay area Presbyornis isoni, new species of Maryland and Virginia. They are of late Figs. 1, 2 Paleocene (Landenian) age and overlie the Holotype. ~l^ñ humérus lacking proxi- Brightseat Formation (Danian), between mal third, collections of the Department of which there is a disconformity representing Paleobiology, National Museum of Natural about 3.6 million years, a period of time History, Smithsonian Institution, USNM approximately equivalent to that of the ex- 294116. Collected in March 1993 by Ron- posed Aquia Formation (Hazel 1969). The ald M. A. Ison. Aquia is divided into an upper Paspotansa Type locality. • Maryland, Charles Coun- Member and a lower Piscataway Member, ty, east bank of the Potomac River, from the latter having sometimes been designat- the area designated as Bluebanks, south of ed in earlier literature as the Piscataway Liverpool Point and north of Douglas Point Formation or Piscataway Indurated Marl (Widewater Quadrangle, USGS 7.5 minute Member of the "early Eocene." These sed- series). iments accumulated in a pelagic environ- Horizon and age.•Nea.T the base of the ment, well offshore, and are reasonably pro- Upper Paleocene (Landenian), Aquia For- ductive of vertebrate fossils, e.g., turtles mation, Piscataway Member. The specimen (Weems 1988), but so far have yielded very came from near the base of the blufFexposed few bird remains. A recently collected hu- at the type locality, which is probably upper mérus is the largest and most diagnostic avi- nannoplankton zone NP5, but possibly low- an fossil yet discovered in the Aquia For- er NP6 (Laurel M. Bybell, USGS, pers. mation, As it merits description, I have comm.). On the scale of Berggren et al. taken the opportunity to review the few oth- (1985), the age would be somewhere be- er avian fossils found so far in the same tween 61 and 62 million years. strata. Measurements ofholotype (mm).•Distal width, 23.3; depth through radial condyle, Class Aves 12.9; greatest diameter of brachial depres- Order Anseriformes sion, 8.8; width and depth of shaft at ap- Family Presbyomithidae Wetmore, 1926 proximate midpoint, 10.8 X 8.2. 430 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON

B C

Fig. I. Hiimeri in palmar view (A-C) and phalanx 1 of major alar digit (D-E) of Presbyomithidae: A, Presbyornis isoni new species, holotype (USNM 294116); B, Telmabates antiquus (AMNH 3170); C, Presbyornis pervetus (cast, USNM 483163); D, Presbyornis isoni new species, paratype (USNM 294117); E, Presbyornis pervetus (cast, USNM 483164). All figures natural size.

Paratype.•Left phalanx 1 of major alar size, there are no discernable diflerences be- digit, abraded along the posterior edge, tween P. isoni and P. pervetus, save that the USNM 294117. Collected in June 1984 by olecranal fossa appears less distinct in the Eugene Hartstein at the same locality as the former, which may be a size-related factor. holotype. Discussion. •Presbyornis is a remarkable Measurements of paratype (mm).• bird that exemplifies an early stage in the Length, 40.6; greatest diameter of proximal evolution of the Anseriformes, combining articulation, 9.5. a long-legged, shorebird-like body with the Etymology. •Named in honor of the col- head of a duck (Olson & Feduccia 1980). lector, Ronald M. A. Ison, an enthusiastic Presbyornis pervetus occurs abundantly in amateur paleontologist. lacustrine deposits of the early Eocene Green Diagnosis.•Much larger than Presbyor- River Formation of Wyoming, Colorado, nis pervetus Wetmore, 1926, or any other and Utah, and a slightly smaller form is known members of the family. Apart from found in the Paleocene of Utah and Mon- -^••'•..U:

VOLUME 107, NUMBER 3 431

Fig. 2. Left humeri oí Presbyornis: (A, C, E) P. isoni new species, holotype (USNM 294116); (B, D, F) P. pervetus (part of UCMP 136541). A-B, palmar view; C-D, anconal view; E-F, distal view. Specimens coated with ammonium chloride. 1.5 x. golia (Olson 1985a: 171). The contempora- Stockholm, pers. comm.). The type species, neous genus Telmabates, from Patagonia, is T. antiquus, was considerably larger than very similar to Presbyornis and was syn- Presbyornis pervetus. onymized with that genus by Feduccia & Presbyornis isoni was much larger than McGrew (1974). Further study, however, either. To put the relative differences in has shown that the limb proportions of Tel- terms of modem analogs for size, Presbyor- mabates differ sufficiently from Presbyornis nis pervetus would have been about the size to merit retention of the genus (Per Ericson, of the largest species of Burhinus, the Bush Swedish Museum of Natural History, Stone-curlew, B. magnirostris, of Australia. 432 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON

Telmabates antiquus was about the size of vergent of the Pelecaniformes, relegated to a Limpkin, Aramus guarauna, whereas a separate suborder Phaethontes. The first Presbyornis isoni would have been the size fossil assigned to the family was Prophae- of a Demoiselle Crane, Anthropoides virgo. thon shrubsolei Andrews (1899), known It is possible, given its much larger size, from a skull, mandible, and partial postcra- that the feeding adaptations oí P. isoni may nial skeleton from the early Eocene (Ypre- have been difierent from those of/*, perve- sian) London Clay of England. Harrison & tus, and that were the entire skeleton avail- Walker (1976) restudied the specimen and able it might be assigned to a different genus. created a new order and family for it, only As it is, however, there is no distinction the latter being justified, however (Olson whatever to be made between these taxa 1985a). The only other fossil member of the except on size. Phaethontes is Heliadornis ashbyi Olson The considerable assemblage of birds (1985b), known from three associated bones known from the Homerstown Formation of from the Middle Miocene (Langhian) Cal- New Jersey (Olson & Parris 1987) are now vert Formation of Maryland, which was re- believed to be Danian (early Paleocene) ferred to the Phaethontidae. rather than latest Cretaceous in age (see be- low), and are older than P. isoni. Most of Family Prophaethontidae these birds were referred to the "form fam- Harrison & Walker, 1976 ily" Graculavidae, which shares similarities Genus Prophaethon Andrews, 1899 with the Presbyomithidae. Presbyornis isoni Prophaethonl sp. was larger than any of these species except Material examined.•T>\%idl end of right Laornis edvardsianus, which was much larg- humérus, USNM 483158. Collected 1 m er than P. isoni. The species in the Hor- from top of chalky shell band of Aquia For- nerstown most similar to P. isoni is Ana- mation at Capital Beltway and Central Av- talavis rex, which was a much smaller bird enue, Prince Georges, County, Maryland in which the brachial depression is longer, (Lanham Quadrangle, USGS 7.5 minute se- narrower, and not as deep, and in which the ries), by Calvin F. Allison, Jr. in 1973. Dis- shaft appears to have been relatively shorter tal width, 7.6 mm. and more curved. Right coracoid lacking part of the acro- Although the alar phalanx referred to P. coracoid and the lateral process of the ster- isoni is not a particularly diagnostic ele- nal end, USNM 483159. Collected by T. B. ment, its size and overall similarity to the Ruhoffat a road cut on Indian Head High- same element in Presbyornis makes it highly way, about 200 m north of Piscataway Creek, unlikely that it would belong to some other Prince Georges County, Maryland (Piscata- family altogether. That two elements of this way Quadrangle, USGS 7,5 minute series). presumed land bird were discovered at the "About 10-12 feet above road." This is ev- same site, the sediments of which were de- idently the same as the type locality of the posited well offshore and where bird bones fossil turtle Catapleura ruhoßi (Weems are scarce, suggests that they may have come 1988). Length to internal distal angle 28.5 from the same individual, despite having mm. been collected years apart. Remarks.•The humérus is extremely similar to that in modem tropicbirds of the Order Pelecaniformes Sharpe genus Phaethon, practically the only notice- Suborder Phaethontes able difference being the indistinctness of The modem tropicbirds (Phaethontidae, the external tricipital groove, but this may Phaethon), are the most primitive and di- be an artifact as the specimen is consider- VOLUME 107, NUMBER 3 433 ably worn in this area. It comes from a bird Abraded distal end of left tibiotarsus lack- much smaller than any living tropicbird, ing internal condyle, USNM 294118. Col- about the size of a Common Tern, Sterna lected from "Top Gully" of Aquia For- hirundo, and thus much smaller than Pro- mation at the Hampton Mall Site, Central phaethon shrubsolei or Heliadornis ashbyi, Avenue near Capital Beltway, Prince although the distal end of the humérus is Georges County, Maryland (Lanham Quad- unknown in either of those species. rangle, USGS 7.5 minute series), in 1984 The coracoid from the Aquia was com- by Eugene Hartstein. Depth through exter- pared directly to that from the holotype of nal condyle, 3.7 mm. This is from a species P. shrubsolei and is very similar in overall small enough to be of comparable size to shape and proportions, but much smaller. Prophaethonl sp., but it bears no resem- At the sternal end, the specimens are lack- blance to the tibiotarsus in Phaethon. It is ing complementary portions and hence can similar, however, to that element in the ge- scarcely be compared. Both have a shallow nus Palaeotringa, from the Homerstown scapular facet. It is apparent that in the Formation (Olson & Parris 1987), though smaller species the procoracoid process much smaller than any of its known species. would not have projected as much anteri- orly, and the procoracoid foramen is more Size 2 stemally situated (i.e., farther from the scap- ular facet). Also the triossial groove in the Proximal half of phalanx 1 of major alar procoracoid is deeper in Prophaethon. digit, USNM 483161. Same collection data The Aquia coracoid is small and if not as USNM 483159. Greatest diameter of from a bird the same size as represented by proximal articulation, 5.0 mm. the humérus, would be from a slightly larger Proximal shaft of left humérus including individual or species. These two bones seem pectoral crest, USNM 483162. The label to be sufficiently diagnostic to be referred with this specimen, in Alexander Wet- to the Phaethontes, but they are assigned more's hand, says only "Piscataway For- only provisionally to the Prophaethontidae mation, Virginia, from Ted Ruhoff." Width on the basis of their age. and depth of shaft just distal to pectoral crest, 4.0 x 5.0 mm. Aves Incertae Sedis These two specimens come from a bird the size of modem Phaethon lepturus, and Size 1 such little morphology as remains on them is actually quite similar to that oí Phaethon, Distal third of right ulna, USNM 483160. so that they might well be from the same Same collection data as the coracoid of Pro- species and perhaps belong among the phaethon? sp, mentioned above. Depth of Phaethontes. distal end through external condyle, 4.4 mm. This specimen is from a small bird, possibly Size 3 smaller than Prophaethorü sp. Its mor- phology is quite unlike the distinctive con- Proximal end of right carpometacarpus dition in modem Phaethon, in which the New Jersey State Museum 13532. Collected carpal tubercle is developed as a strong hook, as float at Belvedere Beach, Potomac River, and is more similar to that in Procellariifor- King George County, Virginia (Passapatan- mes, but as we do not know what the ulna zy Quadrangle, USGS 7.5 minute series), looked like in ancient Phaethontes, it can- 27 Dec 1985 by Eugene Hartstein. Proximal not be ruled out that this fossil belongs in depth through alular metacarpal, 13.3 mm. that group. Although not found in situ, the specimen is 434 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON most probably Paleocene as the exposures vey will have none of this and place the at this locality are entirely a green glauco- outcropping portions of the basal Homers- nitic sandy marl of the Paspotansa Member town in nannoplankton zone NP 3 (Laurel of the Aquia Formation. The bone comes M. Bybell, USGS, pers. comm.), which is from a bird slightly larger than Phaethon Danian, Lower Paleocene, between about rubricauda, the largest modem tropicbird, 64 and 65 million years in age (Berggren et but differs considerably from that genus. It al. 1985). A Cretaceous assignment may has some resemblance to the carpometa- have arisen either through reworking of Me- carpus in the Sulidae but again is quite dif- sozoic macrofossils into the younger sedi- ferent, though it could still belong to some ments, or through the persistence of oth- member of the Pelecaniformes. erwise Mesozoic taxa into the earliest The age of the "Cretaceous" birds of New Tertiary. Jersey. •The record of Paleocene birds from Although it makes no great difference one marine deposits is quite scanty and the spec- way or another as far as avian evolutionary imens discussed above would probably be history is concerned, as birds with some of the only ones known from North America the same general morphological attributes were it not for the fact that the fairly diverse existed before and after the Homerstown avifauna derived from the marine sedi- Formation, the refinements brought about ments of the basal Homerstown Formation by modem microfossil analysis would ap- in New Jersey, long considered to be late pear to necessitate the New Jersey birds be- Cretaceous (Maastrichtian) in age, are prob- ing relinquished as part of Mesozoic history. ably Paleocene as well. These fossil birds have received attention since the early Acknowledgments 1870's, when many of the taxa were origi- nally described by O. C. Marsh. The fauna I thank Ron Ison and Eugene Hartstein, has been recently revised by Olson & Parris for collecting and donating the fossils of (1987). Presbyornis described here. Laurel M. By- The age of the basal Homerstown has bell, Thomas Gibson, and Robert Weems been, and still is, a highly contentious issue. of the U.S. Geological Survey provided The birds from here were originally regard- stratigraphie information, maps, and ref- ed as Cretaceous (Marsh 1870, 1872), then erences. For lending various fossils I am Paleocene (Wetmore 1930), then Creta- grateful to Charlotte Holton, American Mu- ceous (Baird 1967). Weems (1988) cites seum of Natural History, New York some of the more recent literature giving (AMNH); Howard Hutchison, Museum of evidence for the basal Homerstown being Paleontology, University of California, Paleocene, but still notes some dissention. Berkeley (UCMP); David C. Parris, New There seems to be consensus that the up- Jersey State Museum, Trenton (NJSM); and per Homerstown is definitely Paleocene, so Cyril Walker, British Museum (Natural that if the basal part were Cretaceous, then History). The photographs are by Victor E. the K/T boundary would have to lie within Krantz. I thank Robert Weems for com- the Homerstown, which, if true, should have ments on the manuscript and Mark Flor- attracted international attention. I get the ence for curatorial assistance. impression that local pride may be in- volved, with the New Jersey contingent Literature Cited holding out for a Cretaceous attribution, Andrews, C. W. 1899. On the remains of a new bird Mesozoic fossils presumably being more from the London Clay of Sheppey. •Proceed- glamorous than Cenozoic ones. Neverthe- ings of the Zoological Society of London 1899: less, personnel of the U.S. Geological Sur- 776-785. VOLUME 107, NUMBER 3 435

Baird, D. 1967. Age of fossil birds from the green- . 1985b. A new genus of tropicbird (Pelecan- sands of New Jersey.•Auk 84:260-262. iformes: Phaethontidae) from the Middle Mio- Berggren, W. A., D. V. Kent, J. J. Flynn, & J. A. Van cene Calven Formation of Maryland.•Pro- Couvering. 1985. Cenozoic geochronology.• ceedings of the Biological Society of Washington Geological Society of America Bulletin 96; 1407- 98:851-855. 1418. , & A. Feduccia. 1980. Presbyornis and the Feduccia, A.,&P. O. McGrew. 1974. A flamingolike origin of the Anseriformes (Aves: Charadrio- wader from the Eocene of Wyoming.• Contri- morphae). • Smithsonian Contributions to Zo- butions to Geology University of Wyoming 13: ology 323:1-24. 49-61. , & D. C. Parris. 1987. The Cretaceous birds Harrison, C. J. O., and C. A. Walker. 1976. A re- of New Jersey. • Smithsonian Contributions to appraisal of Prophaethon shrubsolei Andrews Paleobiology 63:1-22. (Aves). •Bulletin of the British Museum (Nat- Weems, R. E. 1988. Paleocene turtles from the Aquia ural History), Geology 27:1-30. and Brighlseat Formations, with a discussion of Hazel, J. E. 1969. Faunal evidence for an unconfor- their bearing on sea turtle evolution and phy- mity between the Paleocene Brighlseat and Aquia togeny.•Proceedings of the Biological Society Formations (Maryland and Virginia), •U.S. of Washington 101:109-145. Geological Survey Professional Paper 650-C: Wetmore,A. 1926. Fossil birds from the Green River C58-C65. deposits of eastern Utah,•Annals of the Car- Marsh, O. C. 1870. Notice of some fossil birds from negie Museum 16:391^02. the Cretaceous and Tertiary formations of the . 1930. The age of the supposed Cretaceous United States.•American Journal of Science ser birds from New Jersey. Auk 47:186-188. 2., 49:205-217. . 1872. Preliminary description of Hesperornis Department of Vertebrate Zoology, Na- regalis, with notices of four other new species of Cretaceous birds. • American Journal of Sci- tional Museum of Natural History, Smith- ence ser 3., 3:360-365. sonian Institution, Washington, D.C. 20560, Olson, S. L. 1985a. The fossil record of birds. Pp. U.S.A. 79-238 in D. Earner, J. King, and K. Parkes, eds. Avian Biology, volume 8. Academic Press, New York.