Data_Sheet_1_Drivers of Spatial Distributions of Basking Shark (Cetorhinus maximus) in the Southwest Pacific.pdf

• Main Authors: Brittany Finucci (9663671), Clinton A. J. Duffy (10698414), Tom Brough (6725690), Malcolm P. Francis (7360979), Marco Milardi (4678516), Matthew H. Pinkerton (10528142), Grady Petersen (10698417), Fabrice Stephenson (7360973)
• Format: Dataset
• Language: unknown
• Published: 2021
• Subjects: Oceanography; Marine Biology; Marine Geoscience; Biological Oceanography; Chemical Oceanography; Physical Oceanography; Marine Engineering; New Zealand; species distribution models; boosted regression tree models; threatened species; elasmobranch; Pacific; Auckland Island; Cetorhinus maximus
• Online Access: https://doi.org/10.3389/fmars.2021.665337.s001

Basking sharks (Cetorhinus maximus) were widely reported throughout New Zealand waters. Once commonly observed, and sometimes in large numbers, basking sharks are now infrequently reported. Basking shark observations are known to be highly variable across years, and their distribution and occurrence have been shown to be influenced by environmental predictors such as thermal fronts, chl-a concentration, and the abundance of prey (zooplankton). Little is known of basking sharks in the South Pacific and more information on distribution, habitat use, and migratory patterns is required to better understand the species’ regional ecology. Here, we used bootstrapped Habitat Suitability Models [HSM, ensembled from Boosted Regression Tree (BRT) and Random Forest (RF) models] to determine the drivers of basking shark distribution, predict habitat suitability and estimated uncertainty in the South Pacific for the first time. High−resolution environmental (1 km 2 grid resolution) and biotic data, including inferred prey species, and all available basking shark records across New Zealand’s Exclusive Economic Zone (EEZ) were included in the ensemble HSMs. The most influential driver of modeled basking shark distribution was vertical flux of particulate organic matter at the seabed, which may indicate higher levels of primary production in the surface ocean and higher prey density in the mesopelagic zone and at the seafloor. The BRT and RF models had good predictive power (AUC and TSS > 0.7) and both models performed similarly with low variability in the model fit metrics. Areas of high basking shark habitat suitability included the east and west coasts of the South Island, Puysegur Ridge, and Auckland Island slope. The outputs produced here could be incorporated into future management framework for assessing threat and conservation needs (e.g., spatially explicit risk assessment) for this regionally protected species, as well as providing guidance for future research efforts (e.g., areas of interest for sampling).