J. Field Ornithol., 66(1):30-36 THE SHORT-TAILED NIGHTHAWK IS A TREE NESTER C, ILLES SEUTIN 1 SmithsonianTropical Research Institute Apartadopostal 2072 Balboa, Panampi MARK LETZER 24 Warren Manor Court Cockeysville,Maryland 21030, USA Abstract.--A nest of the Short-tailedNighthawk, Lurocalissemitorquatus noctivagus, was found in Panama.It waslocated on a 15-cmdiameter, horizontal branch in an Erythrinafuscatree, 6 m aboveground. The tree wasgrowing on the side of a pond in semi-openhabitat, close to old second-growth,lowland forest. The nestling period was at least 24 d. The nest was similar to that recently describedfor L. s. nattereri,confirming that unlike any other Capri- mulgidae,this nighthawkis an arborealnester. The validityof previousaccounts of ground nesting by the speciesis questioned,and the finding is discussedin the context of the uncertaintaxonomy of the Short-tailedNighthawk complex, and of the evolutionof ground nesting in Caprimulgiformes. LUROCALISSEMITORQUATUS ANIDA EN •tBOLES Sinopsis.--Un nido del Afiapero colicorto,Lurocalis semitorquatus noctivagus, fue encontrado en Panamir. Estaba ubicado sobre una rama horizontal de 15 cm de diSmetro, en un •trbol de Erythrinafusca, 6 m de altura. E1 •trbol estabacreciendo al lado de un estanqueen un firea semi abierta, cerca de un bosquesecundario de tierras bajas.E1 pich6n permaneceen el nido por lo menos24 dias.El nido fue similaral recientementedescrito para L. s. nattereri, confirmandoque a diferenciade otros Caprimulgidae,este tapacamino es arboreo. Nosotros ponemosen duda la validezde encuentrosprevios de anidamientoen el suelo para la es- pecie, y discutimosnuestro hallazgoen el contexto de la taxonomia incierta del complejo del Afiapero colicorto,y del anidamientoen el suelo entre los Caprimulgiformes. Although the Short-tailedNighthawk, Lurocalissemitorquatus, is prob- ably not rare through most of its range, this speciesis seldom seen and virtually nothing is known of its habits.Here, we provide the first descrip- tion of an arboreal nest of L. semitorquatusnoctivagus, which is similar to that of L. semitorquatusnattereri. We discussthis finding in the context of the uncertain taxonomy of the Short-tailed Nighthawk complex, and of the evolution of ground nestingin Caprimulgiformes. FIELD OBSERVATIONS The nest of a Short-tailedNighthawk of the subspeciesnoctivagus was found at Gamboa, Panam/t (9ø08'N, 79ø42'W). On 1 Feb. 1993, ML saw an adult sitting on a large horizontal branch of an Erythrinafusca tree. On 4,7,11 and 14 February,GS observeda bird roostingat exactlythe same place on that limb, suggestingthat it was nesting there. The 14 m high nesting tree was the penultimate tree of a row of 10-15 m tall Ery- • Mailing address:Smithsonian 7?opical Research Institute, Unit 0948, A.P.O. AA 34002-0948, USA. 30 66,•,,, 1 TreeNesting in a Nighthawk [31 thrina fusca, Spondiasmombin and Cercropiasp., which ended at a large human-made clearing and was bordered on one side by a paved road and on the other by a large pond. Overall, the habitat was relativelyopen, lowland second-growth,but was within 250 m of the limits of Parque Soberania,an extensivetract of dense, old, second-growthlowland forest. The 15 cm diameter horizontal portion of the branch on which the bird wasbrooding was6 m aboveground (Fig. 1A). The branch wasthe lowest one of the tree crown and protruded by about 8 m from the main trunk; the nest was approximately5 m from the trunk. The brooding bird was sittingclose to the point where the branch forked on the horizontal plane into two smaller branches;a lip-like ridge connecting the two arms cre- ated a wider area where the bird sat (Fig. lB). On 11 March, GS sawa small downy head emerging from the breast feathers of the attending adult. The nestling'seyes were open, its head and neck were uniformly medium gray, and its bill was black. Thereafter, a single nestling and a brooding adult were seen on severaloccasions until the end of March. On 3 April, GS photographed the nestlingas it stood alone on the nesting branch (Fig. lB). From a nearby tree it was possibleto see that the egg had been incubatedin a small ill-defined natural depressionon the upper surface of the branch, just before it split. There was no sign of nesting material in or around the depression.The next day K. Kaufman (pers. comm.) found the nesting branch unoccupied. Thus, the nestling period for this individual was at least 24 d (11 March-3 April), which is signifi- cantlylonger than the 16-20 d range for other caprimulgids(Perrins and Middleton 1985). DISCUSSION The nest of another subspeciesof the Short-tailed Nighthawk, L. sem- itorquatusnattereri, has been describedrecently (Stranecket al. 1987). The three neststhese authors had found were located on large horizontal branchesof forest trees, 10-18 m above ground. Their observationspro- vided the first evidence of arboreal nesting by a Caprimulgidae. As with all other caprimulgids,the single egg, as observedin the nest that could be inspectedclosely, was laid directly on the substratewith no lining or other constructionmaterial involved.Thus, the nestsof L. semitorquatus nattererifound by Stranecket al. (1987) closelyresemble that of L. semi- torquatusnoctivagus described above, although the latter waslocated low- er in a tree found in relativelymore open habitat. Apart from the report of Stranecket al. (1987), three descriptionsof L. semitorquatusnests are availablein the literature.Straneck et al. (1987) suggestedthat the nestsof L. semitorquatusnattereri described by Fiebrig (1921) and by Bertoni (1923) were misassignedto the Short-tailedNight- hawk becausethey did not correspondto the neststhey themselvesde- scribed for the same subspecies,and becauseBertoni (1926) himself ex- presseddoubts about his assignmentof the nest he found to the species. Herklots (1961) provided the third description, stating that L. semitor- quatussemitorquatus in Trinidad is a ground nester. This statementis 32] G. Seutinand M. Letzer j. FieldOrnitho]. Winter 1995 4,. % F•cu•,• 1. A. The tree (Erythrinafusca)in which a nestof Lurocalissemitorquatus noctiva•us waslocated. B. The nestingbranch on which the nestling(1 d before fledging)is sitting. On both panels,the arrow indicatesthe positionof the nest. Vol.66, No. ! TreeNesting in a Nighthawk [33 B F•6URI• 1. Continued. based on Herklots's observation of a bird incubating on the ground a single egg, much smaller in size (23.5 x 16 mm) than the one Straneck et al. (1987) measured(37 X 25 mm). The subspeciessemitorquatus and noctivagusare thought to be more closelyrelated to one another than either one is to nattereri(e.g., American Ornithologists'Union 1983, Gris- wold 1936, Meyer de Schauensee 1966). The fact that both noctivagus and nattereriare arboreal nesters, that the egg measured by Herklots (1961) wasdefinitely small for a caprimulgid (Fig. 2), and that the habitat in which the Trinidad nest was found (simply described as savannah) is quite atypical for the species,suggest that Herklots might have misas- signed the nest he found to L. semitorquatus.Until new evidence to the contrary becomesavailable, the Short-tailedNighthawk sensulargo should be considered as an arboreal nester. Unconfirmed differences in the breeding habits of some subspeciesshould not be consideredin the de- bate over the recognition of some forms as distinct species(see American Ornithologists'Union 1983, Fjeldsfiand Krabbe 1990, Meyer de Schauen- see 1966). One observationthat was not discussedby Stranecket al. (1987) in their report of L. semitorquatusnattereri breeding is the description of three eggsprovided by Sch6nwetter(1960-1967). Theseeggs were closer in size (31.2-32.3 X 22.0-24.0 mm) to that measured by Straneck et al. 34] G. Seutinand M. Letzer J.Field Ornithol. Winter 1995 15000 125OO 3 ß o 10O00 e© o2e 7500 ed• e eee 5000 ße%•l•ee © 1 2500 i i i i 0 40 80 120 160 200 BodyMass (g) FIGURE2. Relationshipbetween body massand egg volumein caprimulgidspecies. Each filled dot representsone species(as presentedin Dunning 1993). Open dots represent Lurocalissemitorquatus position in this allometric spacebased on egg measurements provided by: I = Herklots (1961); 2 = Sch6nwetter (1960-1967); and 3 = Straneck et al. (1987). Male bodymasses (from Dunning1993) were used whenever given separately from female massesbecause they are typicallybased on larger samplesize and are in- dependent of laying condition. Egg volumes (V) were estimatedusing Hoyt's (1979) formula: V = 0.51 X LB2;L (egg length) and B (maximum breath) were taken from Sch6nwetter(1960-1967), matchinggeographic origin of massand egg measurements wheneverpossible. than to the one describedby Herklots (1961), and have preciselythe volume expectedfor a nighthawkthe sizeof L. semitorquatusnattereri (Fig. 2). Still, we questionSch6nwetter's report for two reasons.Firstly, he de- scribedwhite unmarked eggs,whereas Straneck et al. observeda white egg speckledwith brown and gray.Secondly, although Sch6nwetterrarely provided information on nest location or substratein his Handbuch, we are confident that he would have recognizedthe significanceof tree nest- ing by a nighthawk, had it been known to him, and that he would have reported it. It is also possiblethat the eggsSch6nwetter reported upon were collected without nest data. Sibleyand Ahlquist (1990) haveproposed that the eared-nightjars(fam- ily Eurostopodidae)and the typicalnighthawks and nightjars (family Ca- primulgidae) are relativelydistantly related and, together,form a mono- phyletic group, the