Spatial and temporal distribution of horseshoe crab (Limulus polyphemus) spawning in Delaware Bay: Implications for monitoring

Abstract

Concern for the status of horseshoe crab (Limulus polyphemus) has increased as harvest for conch and eel bait has increased and spawning habitat has decreased. In early 1999 a workshop was held at the behest of the Atlantic States Marine Fisheries Commission to design a statistically valid survey of horseshoe crab spawning in Delaware Bay. The survey that resulted was a redesign of a volunteer-based spawning survey that began in 1990, and its network of volunteers was relied on to implement the three-stage sampling design in 1999. During May and June of 1999, 163 participants surveyed during the highest of the daily high tides on 16 beaches (8 on each site of Delaware Bay). During the first half of the spawning season, spawning was associated with lunar phases, but moderated by wave height. Disproportionately more spawning occurred within 3 d of the first new and full moons, and spawning activity (measured by an index of female density) was correlated inversely to the percent of beaches with waves ≥0.3 m. Spawning was heaviest on the Delaware shore around the full moon in May in spite of low waves in New Jersey during the new and full moons in May. Number of beaches sampled was the most important factor in determining the precision of the spawning index and power to detect a decline. Explicit consideration of statistical power has been absent from the current debate on horseshoe crab status and harvest. Those who argue against harvest restrictions because of a lack of statistically significant declines take on a burden to show that the surveys they cite have high statistical power. We show the Delaware Bay spawning survey will achieve high statistical power with sufficient sampling intensity and duration. We recommend that future Delaware Bay spawning surveys sample on 3 d around each new and full moon in May and June and increase the number of beaches to ensure high statistical power to detect trends in baywide spawning activity.