Résumé
Analysing the chemical composition of otoliths from fish is a commonly employed technique to search for a signal indicative of a particular habitat at some point in a fish's environmental history. In connectivity studies, this method frequently runs into a problem as the origin of most larval fish is unknown, reducing the ability of the technique to definitively identify a source location among locations with subtle water mass differences. Following an intensive self-recruitment study on an isolated island in Kimbe Bay, Papua New Guinea, genetic parentage analysis (genotyping) confirmed the origin of about 50% of 110 new recruits of the orange clownfish, Amphiprion percula, on Kimbe Island. Known-origin larvae were those confirmed to have been spawned by parents on Kimbe Island, while the remainder would have been spawned by parents on other reefs (with the nearest reef ~10km away). Using standard otolith composition analyses (laser ablation inductively coupled mass spectrometry, LA-ICPMS), we were able to compare the known-origin (self-recruiting) larvae against those known to have originated from reefs at least 10km away, testing for differences in elemental composition of their otoliths. We will discuss these findings and the implications for future studies attempting to use natural markers in studies of connectivity in coral reef fish populations