Impact of the use of different temperature‐dependent larval development functions on estimates of potential large‐scale connectivity of American lobster

The way in which the effect of temperature on the development rate of crustacean larvae is simulated in larval dispersal models potentially impacts the inferences made about population recruitment and connectivity. In this study, we contrasted dispersal and connectivity predictions made by a large‐scale dispersal model of American lobster (Homarus americanus H. Milne Edwards, 1837) larvae using three temperature‐dependent larval development functions proposed in the literature: (1) “warm‐source lab”, (2) “warm‐source field”, and (3) “cold‐source lab”. Differences in predictions using each function were contrasted in the northern (colder) and southern (warmer) portions of the species' range. Using these different development functions resulted in significant and marked differences (61.3–162.4 km in the north and 30.9–81.9 km in the south) in the distances dispersed by larvae from hatch to settlement. In general, predicted self‐seeding, retention, and local connectivity were increased, and predicted connectivity among distant locations was decreased, when a function predicting faster development was used. The field‐derived function predicted much less connectivity and decreased dispersal overall than both lab‐derived functions. The cold‐source lab function predicted more retention in northern regions, but less in southern regions, than the warm‐source lab function. Our findings indicate the need for more studies to quantify the rate at which lobster larvae develop in nature, including how this may vary over space and time.