PREDATOR-PREY INTERACTIONS BETWEEN THE NATICIDS EUSPIRA-HEROS SAY ANDNEVERITA-DUPLICATA SAY AND THE ATLANTIC SURFCLAM SPISULA-SOLIDISSIMA DILLWYN FROM LONG-ISLAND TO DELAWARE

Interactions between the naticids Euspira heros (Say, 1822) and Neveri ta duplicata (Say, 1822) and their prey, the Atlantic surfclam Spisula solidissima (Dillwyn, 1817), were reconstructed from 1,300 bored shel ls collected at six localities from Long Island to Delaware. Both nati cids show comparable site selectivity on prey valves, with 90% of comp lete boreholes from any one locality situated in the umbonal region, s lightly posterior to the dorsoventral axis. Variation in complete bore hole location in biaxial scattergrams was least for the N. duplicata-a ffected locality in Great Egg Harbor. Incomplete boreholes show a simi lar distribution to complete boreholes, which suggests that failure to penetrate the shell was not a consequence of poor siting of the drill hole. Modal frequency of boreholes occurs in valves between 20 and 59 mm in anteroposterior width, with a size refuge from predation at widt hs greater than 120 mm, although unbored specimens reached a maximum w idth of 160 mm. Mean borehole diameter is greatest (5.7 mm) for the E. he,os-affected northernmost sample from Fire Island Inlet, NY, and le ast (3.3 mm) for the N. duplicata-affected Fenwick Island, DE, sample. In addition to possible taxonomic differences in radula size between naticid species, older and younger cohorts (age classes) may dominate the northern and southern naticid populations, respectively. Regressio ns of outer borehole diameter (OBD) on prey valve width indicate that naticids were prey size selective, although the degree of correlation varies significantly from E. heros-affected Barnegat Inlet (r = 0.81) to N. duplicata-affected Great Egg Harbor Inlet (r = 0.33) samples. Hi gh versus low correlations indicate few versus many predator-prey mism atches, i.e., oversized or undersized OBDs relative to surfclam valve width. Variation in prey size selectivity is attributed to age (size)- class dominance, or lack of among contiguous populations of surfclams and naticids. Significantly lower correlation coefficients,based on fi eld death assemblages relative to published statistics based on natici d predation in captivity are attributed to stochastically interrupted foraging of moonsnails on disturbance-prone exposed tidal flats versus disturbance-minimized aquarium experiments. Slopes of regression line s from E. heros-affected samples show a rate of increase in OBD with i ncreasing clam prey width that is three times that of the N. duplicata -affected Great Egg Harbor sample. Prey effectiveness, indexed by the ratio of valves with incomplete to total attempted boreholes, is 0.01 for Fenwick Island, a surfclam population dominated by young cohorts, to 0.22 for Barnegat Inlet, a population with many old-aged individual s.