Quantitative analysis of foraging behaviour in a field population of the South American sun-star Heliaster helianthus

Foraging behaviour in the South American sunstar Heliaster helianthus (Lamarck) was investigated quantitatively on a subtropical rocky shore in central Peru. H. helianthus feeds mainly on two mussel species, Semimytilus algosus (Gould) and Perumytilus purpuratus (Lamarck). A sequence of foraging behaviour was described and observations were made of the timing of attacks on mussel beds by sun-stars. Although H. helianthus is capable of foraging out of water, an unusual trait for asteroids, the timing of such foraging appears to be well adjusted to avoid the risk of prolonged heat and desiccation. Foraging activity began 4 h 52 min before high tide and 3 h 19 min after high tide, with a mode between 2 h 30 min and 2 h before high tide. Fifty-two percent of all foraging activity began between 3 h 30 min and 2 h before high tide, while only 12.4% began after high tide. This suggests that H. helianthus mainly relies upon changes in the rate of tidal increase as a cue to begin foraging. Foraging activity ceased between 3 h 40 min before and 4 h 29 min after high tide, with 48% ending between 2 and 0.5 h before high tide. The duration of foraging ranged from 19 to 190 min, with values between 30 and 80 min accounting for 67.4% of all observations. The median duration was 62 min. No significant correlation was detected between the time when foraging activity commenced and its duration. The intensity of foraging activity varied on consecutive days, with a general pattern of decreasing intensity after a day of relatively high activity. Foraging location in relation to a mussel bed was analysed on a marked, 8 m stretch of rocky shore. The numbers of foraging H. helianthus observed on different sections of the shore were related neither to the width of the mussel zone nor to the vertical position of the lower edge of the mussel zone, indicating that sun-stars do not rely upon these factors to assess prey availability and that ideal free distribution with regard to prey abundance does not occur on the spatial scale examined.