403 Status of Tern Colonies in the Honduras Bay Islands DAVID T. SHOCH 1 AND DAVID L. ANDERSON 2 1 Winrock International, 1621 North Kent Street, Suite 1200, Arlington, VA 22209 2 Museum of Natural Science, Department of Biological Sciences, 119 Foster Hall, Louisiana State University, Baton Rouge, LA 70803 Abstract.—In the first comprehensive survey of tern colonies in the Honduras Bay Islands, 46 cays were surveyed in 2005 and 2006 and three species of terns were observed nesting in colonies on ten individual cays. Apparent col- ony turnover was high between years for Least (Sternula antillarum) and Roseate (Sterna dougallii) Terns, from 71% to 100%, respectively. The first nest records of Bridled Tern (Onychoprion anaethetus) for Honduras are document- ed. Received 10 October 2006, accepted 18 December 2006. Key words.—Caribbean, Honduras Bay Islands, nesting, colonies, terns. Waterbirds 30(3): 403-411, 2007 The Honduras Bay Islands (henceforth Bay Islands) are a discrete group of islands in the Caribbean Sea off the coast of mainland Honduras. Despite wide recognition of the marine ecosystems of the Bay Islands (Har- borne et al. 2001; Almada-Villela et al. 2002) and their conservation designation as part of the multinational Mesoamerican Reef, com- paratively little is known about marine birds in the region. To date, information on sea- birds has been derived from a handful of published notes (Udvardy et al. 1973; Seutin et al. 1997; Thorn and Medina 2005), and knowledge of these bird populations is in- complete and outdated. Terns in particular are a significant com- ponent of the marine avifauna breeding in the Bay Islands, representing six of seven seabird species known or suspected to nest there (Shoch and Canfield 2006). Several of these species are of conservation importance, such as the regionally threatened Roseate Tern (Sterna dougallii)(USFWS 1993), and the Royal (Thalasseus maxima), Sandwich (T. sandvicen- sis) and Least (Sternula antillarum) terns, all at the southern limit of their known breeding distribution in Mesoamerica. To fill this gap in knowledge and to provide a base against which future observations can be evaluated, we conducted the first comprehensive survey of tern breeding colonies in the Bay Islands. METHODS Study Area This study was conducted in the Bay Islands (15°94’ to 16°54’N, 87°00’ to 85°81’W), consisting of the main islands and associated cays of Utila, Roatán, and Guana- ja, and the Cayos Cochinos. Cays off Roatán are concen- trated around the smaller island of Barbareta and are hereafter referred to as the Barbareta cays. These is- lands are on the continental shelf from 19 to 56 km off the coast of mainland Honduras (Fig. 1). The nearest neighboring islands, cays off Belize and Moskitia, and the Swan Islands, are approximately 110, 270 and 220 km distant, respectively, and thus the Bay Islands are a natural geographic unit. We identified 46 cays from maps and satellite imag- ery, and surveyed these in 2005 and 2006, to produce a complete survey of cays in the Bay Islands (Table 1). Field Observations Field observations were carried out in 2005 from 12 to 28 August, between 06.00 and 15.00 h, and in 2006 from 11 July to 1 August, between 07.30 and 13.00 h. Weather conditions were uniform with clear skies and low to moderate winds. At each cay surveyed, complete counts were made from the ground or by boat and verified independently by two or more observers. For a given cay, only birds ob- served in specific association with that cay (e.g., roost- ing, incubating, feeding dependent young, in transit to or from) were tallied. Birds were identified to species, and age classes were simplified to distinguish juveniles and adults (i.e., first and second summer individuals were not distinguished from adults). The juvenile age class was used to refer to fledged juveniles only. UTM co- ordinates were recorded at each cay surveyed using a global positioning system (GPS) unit. Because all potential colony sites within each of the four principal island groups were surveyed in a single day, and the entire study area surveyed within a maxi- mum period of three weeks (16 and 21 days for 2005 and 2006, respectively), surveys were designed to mini- mize risk of double counting. Sites were considered to represent breeding colo- nies where one or more of the following was observed: territorial behavior (e.g., agonistic behavior toward the observers), courtship, incubating posture, eggs, or flightless young. The minimum number of adults at each colony was based on either a direct count or dou- ble the number of observed nests, whichever was great- er. Annual colony turnover rates were calculated as per Erwin et al. (1981).