Construction was initiated following the 2000 tern breeding season for Phase 1 of a planned two-phase ?Shoreline Protection and Erosion Control Project? at the Falkner Island Unit of the USFWS Stewart B. McKinney National Wildlife Refuge located in Long Island Sound off the coast of Guilford, CT. When the Common Tern (Sterna hirundo) and federally endangered Roseate Tern (S. dougallii) arrived in spring 2001, they encountered several major habitat changes from what had existed in previous years. These changes included: a rock revetment covering most of the former nesting habitat on the beach from the northwestern section around the northern tip and covering about 60% of the eastern side; an elevated 60- ? 4-m shelf covering the beach and lower bank of the southwestern section; and about 2,000 sq m of devegetated areas on top of the island on the northeast side above the revetment, and about one-third of the southern half of the island. The southwest shelf was created by bulldozing and compacting extra construction fill and in situ materials. This shelf differed in internal structure from the main revetment on the north and eastern sections of the island because it lacked the deep internal crevices of the revetment. The deep internal crevices were created from the large stones and boulders (up to 2 tons) used in the construction of the main revetment. Small rock and gravel was used to fill the crevices to within 3 feet (0.9 m) of the surface of the revetment. Because half-buried tires and nest boxes for the six Roseate Tern (Sterna dougallii) sub-colony areas were deployed in similar patterns on the remaining beach, and nest boxes were placed on the newly elevated shelf areas several meters above previous locations on the now-covered beach areas, the distribution of Roseate Tern nests did not change much from 2000 to 2001. However, the movements of Roseate Tern chicks ? in many cases led by their parents towards traditional hiding places ? into the labyrinth of subterranean channels, especially in the main revetment area, made it difficult to measure chick growth and productivity as had been done for more than 12 years prior to construction. Also, observations of color-banded adults that were unable to locate and feed their young inside the main revetment, and of adults returning to courtship behavior and renesting after having hatched chicks from their initial clutches, indicated that a minimum of 20% of the chicks (mostly first hatched A-chicks, which usually have high rates of survival to fledging) that entered the main revetment died after doing so in 2001. The mortality rate of Roseate Tern chicks that entered the secondary revetment on the southwest shelf, however, was not unusually high in 2001. In an attempt to reduce the likelihood of nesting and chick losses in these sub-colonies, a research team led by the USGS in 2002-2003 did not put nest boxes on the northeast and east shelf areas where previous losses had been high. However, losses of tern eggs and young chicks to predatory Black-crowned Night Herons (Nycticorax nycticorax) were so great in 2002?2003 that few Roseate Tern chicks survived long enough to move into the main revetment area and little comparative survival data were collected. The USFWS continues to remove predatory night herons, to monitor the location and success of nesting Roseate Terns and to fill in some of problem areas near the tern nests in the main revetment. The extensive chick-searching fieldwork and observational procedures used by the USGS-led research team to determine the growth and survival of the Roseate Tern chicks at Falkner Island were not used in 2004?2005. The number of chicks lost in the revetment during these years is not known. Without additional fill, loss of chicks of this endangered species in the main revetment may rise again even though the night heron predation problem has been reduced.