Bluefin tuna (Thunnus thynnus) distribution in relation to sea surface temperature fronts in the Gulf of Maine (1994–96)

Fishery‐linked aerial surveys for bluefin tuna (Thunnus thynnus) were conducted in the Gulf of Maine (GOM) from July through October, 1994–96. Each year, from 507 to 890 surface schools were detected and their locations examined in relation to oceanographic conditions. Correlations between bluefin tuna presence and environmental variables were explored for sea surface temperature (SST), distance to a SST front, frontal density (relative density of all SST fronts seen in a given 1 km area for 2 weeks prior to each tuna sighting), and bottom depth and slope. Mean SST associated with bluefin schools was 18.1°C (±2.8). Schools were located at a mean distance of 19.7 km (±19.6) from SST fronts, and in water masses with an average frontal density of 28.2 m km^?2 (±35.7). Mean bottom depth of detected schools was 139.0 m (±70.3), and mean bottom slope was 0.7% rise (±0.7). A binomial generalized linear model fit to these variables indicated that bluefin are seen closer to fronts than locations in which no tuna were seen. Using simple and partial Mantel tests, we investigated the spatial correlation between bluefin tuna presence and the environmental variables, controlling for spatial autocorrelation. For each day that schools were sighted, we performed 24 Mantel tests, on a combination of response and predictor variables. The spatial relationship between bluefin tuna and SST fronts was inconsistent. Our analysis identified significant spatial structure in the bluefin school locations that had no significant correlation with any of the measured environmental features, suggesting that other untested features, such as prey density, may be important predictors of bluefin distribution in the GOM.