Application of a robust design occupancy model for assessing fish recruitment

Abstract

Fish recruitment is complex, regulated by environmental factors that induce high mortality early in life. Additionally, age-0 fish can be difficult to sample, making recruitment difficult to detect. We used a robust design occupancy model to evaluate the effects of biotic (age-0 and adult common carp (Cyprinus carpio), bluegill (Lepomis macrochirus), walleye (Sander vitreus), and northern pike (Esox lucius) relative abundance, prey availability, age-0 carp length) and abiotic (water level, temperature) factors on age-0 carp occupancy, detection, and extinction in 13 lakes in South Dakota, USA, for July–April 2008–2010. Age-0 carp occupancy decreased with increasing adult carp abundance and increased with increasing water levels. Age-0 carp detection probability was high during summer (>0.75) but decreased in fall and spring (0.34). Most lakes were occupied in July but overwinter extinction probability was high (59%), resulting in 51% occupancy probability by April. Other environmental factors were not supported, suggesting they had little effect on reproduction and survival. Our results indicate reproduction was universally successful but difficult to detect and that overwinter mortality often resulted in recruitment failure.