with wetland area than is the density of seabirds which feed on land (gulls) 

 and the pelagic zone in addition to marsh and bay areas. Nonetheless, water- 

 birds as a group show significant correlations between numerical abundance and 

 wetland area. Coastline length, used as another index of "coastal space", was 

 poorly correlated with waterbird abundance (Table 15). 



Correlations of nesting abundance with wetland area and coastline were 

 examined for six widely-distributed species of seabirds and wading birds 

 (Table 15). An a^ posteriori prediction was that the abundance of the marsh- 

 feeding species would show higher correlation with wetland area than the more 

 "general ist" feeders. The expectation was only partly supported. Of the 

 seabirds, only least terns showed a significant correlation between abundance 

 and wetland area. This species is strictly an inshore-feeder, fishing mostly 

 in inlets and marsh shallows (Erwin 1978). The other inshore-feeder, the 

 black skimmer, did not show a significant correlation, although only five 

 States were compared. Of the wading birds, snowy egret numbers were highly 

 correlated with wetland area, but black-crowned night heron abundance only 

 showed a correlation with coastline length. Snowy egrets seem to be more 

 dependent upon marsh-dwelling aquatic organisms than black-crowns which prey 

 upon birds, rodents, reptiles, etc. (Kushlan 1977). Similarly, Custer and 

 Osborn (1977) found that the abundance of the three species which feed nearly 

 exclusively in coastal salt marshes (great and snowy egrets, Louisiana herons) 

 was significantly correlated with wetland area. The species which were less 

 "specialized" or fed in fresh-water areas (black-crowns, cattle egrets, little 

 blue herons) showed no correlation between abundance and wetland area per 

 State. The herring gull showed the lowest correlation coefficient with wet- 

 land area. The herring gull's omnivorous and refuse-feeding habits have 

 probably emancipated it from a dependence upon food produced in coastal 

 wetlands. 



To make the above comparisons more meaningful, it would be useful to have 

 measures of nesting habitat availability and human disturbance. Although 

 attempts to determine breeding habitat availability have been made (Burger and 

 Lesser 1977), there is a problem in assessing "suitability" from a human as 

 opposed to a bird perspective. "Wetland area" is a composite of feeding, 

 nesting, and resting habitats and nesting abundance may be a complex function 

 of these interdependent variables. 



The locations of major waterbird colonies along the northeast coast are 

 shown on State coastal maps (Figures 2-10). These maps indicate waterbird 

 "hot spots" and may provide impetus and direction for further research and for 

 local and regional protection and management. 



Comparisons of mean colony sizes by States were made for several of the 

 more widely-distributed species (Table 16). The mean colony size varies 

 considerably for most species. Herring gulls nest in more widely-distributed, 

 larger colonies in southern Maine, Massachusetts, and Rhode Island than in 

 States from New Jersey south. However, this gull is still increasing south of 

 New York and, in time, may show similar colony size distribution throughout 

 its range. Common and least terns generally nest in many scattered colonies 

 but, roseate terns nest at only a few locations. Therefore, local disturbance 

 or environmental degradation could have a serious impact upon the roseate tern 

 population as a whole. A species with many breeding populations, or demes, 

 will be better buffered against extirpation than one which is confined to a 

 few locations (Buckley and Buckley 1977). 



19 



