Nest-Site Selection of American Oystercatchers (Haematopus palliatus) in Salt Marshes

Abstract

-We documented flexibility of American Oystercatcher (Haematopus palliatus) nest-site selection at marsh habitats in New York, New Jersey, and North Carolina. In New York we examined the relationship of nest-site selection to tidal flooding and found that birds chose nest sites nonrandomly with respect to marshy areas. Three types of habitat patch sites on marsh islands chosen were sand, wrack (tidally washed-up dead organic material), and Spartina patens grass. Nests on sandy sites were less susceptible to tidal flooding than nests on wrack or grass because they were higher in elevation. There was a positive correlation between the number of nests and the length of sand beach. Hatching success was higher for eggs placed on sand than for eggs on grass or wrack. The location of sandy nest sites used by birds changed little between years resulting in low turnover rate. Wrack and grass locations changed between years resulting in a high turnover rate. Birds used different habitats in different geographic areas. In North Carolina, all nests were on sand; none were on wrack or grass. In New York and New Jersey nests were placed on sand, wrack, and grass. Comparisons of nest characteristics showed that sand sites in North Carolina were larger and drier than sand sites in New York and New Jersey. Comparisons between years in New York and New Jersey were different because of the availability of wrack. Received 22 October 1987, accepted 16 September 1988. BIRDS select breeding habitats based on biotic and abiotic factors of the environment. This produces a nonrandom spatial distribution of nests. Breeding-habitat selection can be divided into three categories (Burger 1985): (1) General habitat use involves choice of a broad habitat type such as an oak forest, salt marsh, or prairie; (2) territory acquisition involves selection of an area vigorously defended by one or both members of a pair; and (3) nest-site selection involves the choice of the actual nest location including the substrate on which the nest is placed, cover around the nest, and food availability. Flexibility in nesting habitat is critical to reproductive success in birds because individuals within populations must adapt to differences in habitat physiognomy. This is particularly true for the American Oystercatcher (Haematopus palliatus) because habitat physiognomy changes over the breeding range (Nol 1984), which extends from Massachusetts to Southern Argentina (Hayman et al. 1986). Available descriptions of nesting habitat of the American Oystercatcher (Baird et al. 1884, Bent 1929, Stone 1967) provide no detailed quantitative or comparative examination of habitat selection. The traditional nesting habitat is described as sand dune on barrier island, although recently oystercatchers have nested in marsh habitat (Frohling 1965, Zaradusky 1985). Habitat studies have been conducted on other marsh-nesting species (Beer 1966; Bongiorno 1970; Burger 1974, 1977, 1980; Burger and Lesser 1978; Montevecchi 1978; Howe 1982), but we know of no comparisons of a salt-marsh nesting species in different geographical regions. Oystercatchers generally nest in open habitats that are sparsely vegetated (Webster 1941, Legg 1954, Harris 1967, Heppleston 1972, Hartwick 1974, Martinez et al. 1983, Summers and Cooper 1977). Most species nest near the shoreline, but European Oystercatchers (Haematopus ostralegus) and Magellanic Oystercatchers (Haematopus leucopodus) also nest in inland habitats (Baker 1973, Buxton 1961, Falla et al. 1966, Heppleston 1972, Miller and Baker 1980). Thus, some species are flexible in choice of nesting habitat. Another indication of flexibility in nesting-habitat selection is the recent range expansion of American Oystercatchers (Post 1960, Post and Rayner 1964, Zaradusky 1985), American Black 185 The Auk 106: 185-192. April 1989 This content downloaded from 152.1.106.107 on Thu, 18 Jul 2013 16:50:44 PM All use subject to JSTOR Terms and Conditions 186 LAURO AND BURGER [Auk, Vol. 106 Oystercatchers (Haematopus bachmani; Eley 1976), and European Oystercatchers (Buxton 1961, Dobbs 1970). STUDY AREA AND METHODS We observed 31 nests (1983) and 59 nests (1984) in New York at Great South Bay and South Oyster Bay (40?37'N, 73-24'W) between the Wantagh State Parkway and Oak Beach. Ninety-one marsh islands lay between the Jones Beach barrier island and Long Island. The islands ranged in size from <1 ha to 844 ha; the median was 40.58 ha. We defined a marsh island as an area surrounded by 30-cm-deep water at low tide and not connected to either Long Island or the barrier island. We chose 30 cm because young chicks were unable to walk to nearby islands at this depth. The dominant species of vegetation in the study area were Spartina alternaflora and S. patens. In 1984, we examined 19 nests in Barnegat Bay, New Jersey (39?45'N, 78?08'W). Barnegat Bay contains 259 salt-marsh islands between the barrier island and the mainland. The predominate species of vegetation were S. alternaflora and S. patens in the wetter areas with Iva frutescens and Baccharis halimifolia in drier areas (Burger and Lesser 1978). In 1984, we examined 18 nests at Battery Island, North Carolina (33?54'N, 78?01'W), a natural island with deposits of dredged material. It was dominated by a dense grass community of primarily S. patens and a maritime shrub thicket dominated by Xanthoxylum americanum, Ilex vomitoria, Quercus virginiana, Juniperus virgiana, Myrica cerifera, B. halimifolia and I. frutescens (Parnell and Soots 1979). In New York we collected data for each nest site to examine habitat-use flexibility and adaptations to marsh habitats. We defined a nest site as the nest scrape and the surrounding habitat patch. A habitat patch is classified by its substrate of sand, wrack, or S. patens grass (see Fig. 1). Sand patches are created when dredged spoil is deposited on the marsh. Wrack patches consist of dead organic material that washes up on the marsh during tidal flooding. Grass patches are on naturally high regions covered by S. patens, easily distinguished from S. alternaflora areas. Sandy patches are higher in elevation and less susceptible to flooding than wrack and grass patches. Low sites were those in which nests were placed on grass and wrack substrate; high sites were those in which nests were placed on sand substrate. In New York we used a measuring wheel to survey the dimensions of all high sites. The area of all sites was calculated from these measurements to determine if there was a difference in frequency of nest sites on the barrier island compared to marsh islands, and if there was a correlation between the length of sand beach and number of nests. We used length for the correlation calculation because most sandy sites were long but not very wide. We censused only the bay side of the barrier island. We calculated turnover rate (after Erwin et al. 1981) to examine the stability of nest sites from year to year. Nest sites from 1983 were surveyed in 1984 to note if sites had changed. We noted new sites in 1984. The turnover rate is T =1l + S2L 2 kN, N2' where S, is the number of sites occupied only on the first census, N, is the total number of sites occupied on the first census, S2 is the number of sites occupied only on the second census, and N2 is the total number of sites occupied on the second census. We collected the following data for nest sites and randomly chosen sites on islands: the percentage of vegetation, sand, and wrack within 1-m and 5-m radii around the nest, width and length of nest sites and distance to water (creek or bay). A random point for each nest was chosen from a grid system of each nesting island. Measures of these nest-site characters yielded information on the degree of protection provided from tidal flooding. We also compared years using the same nest characters. All comparisons were made using Kruskal-Wallis Chi-square tests rather than analysis of variance because the data were not distributed normally. We measured nest-site elevation on the sandy dredge-spoil area of North Line Island (40?38'N, 73?29'W) because we assumed competition for sites to be great due to the high concentration of breeding pairs (9 in 1983; 14 in 1984). We used a Nixon level accurate to 0.03 m to measure elevation (above sea level) with respect to survey markers. We collected data on hatching success in New York in 1983 and 1984. A nest was considered successful if at least one chick hatched. We did not determine fledgling success because the parents led the chicks away from the nest after hatching. We examined geographic variation in nest-site selection between New York, New Jersey, and North Carolina by comparing the percentage of vegetation, sand, and wrack within 1-m and 5-m radii around the nest, length and width of the nest site, and distance to water.