Open ocean ecosystems represent the largest habitat on Earth and are highly dynamic in time and space. Mesoscale eddies are a primary driver of this variability and serve a key structural role in ocean ecosystems. Eddies modulate marine biodiversity beyond their impacts on plankton, influencing many ecologically and commercially important predators that may preferentially occupy anticyclonic eddies. However, how animal-eddy interactions scale across predator species and the mechanistic drivers of these relationships remain an area of active research. We integrated satellite tracking data for sharks with observations of mesoscale eddies to determine how four shark species interact with eddies in the Gulf Stream region. Based on over 24,000 tracking days, we found that blue, white, and shortfin mako sharks selected for the cores of anticyclones while use of eddies by tiger sharks was less conspicuous. Some particularly large and long-lived anticyclones were occupied by tagged sharks for multiple weeks suggesting that these eddies may serve as hotspots for pelagic predators.

Aim: Species distribution models (SDMs) are an important tool for marine conservation and management, yet guidance on leveraging diverse data to build robust models is limited. While various approaches can be used to integrate different datasets, studies comparing their performance, particularly for highly migratory and mobile species, are scarce. Here, we assess whether a model-based integrative framework improves performance over traditional data pooling or ensemble approaches when synthesizing multiple data types. Location: North Atlantic Ocean Time Period: 1993 - 2019 Major Taxa Studied: Blue shark (Prionace glauca) Methods: We trained traditional, correlative SDMs and integrated SDMs (iSDMs) with three distinct data types: fishery-dependent marker tags, fishery observer records, and fishery-independent electronic tag data. We evaluated data pooling and ensemble approaches in a correlative SDM framework and compared performance to an iSDM approach designed to explicitly account for data-specific biases while retaining the strengths of each dataset. Results: While each integration approach yielded robust models, model performance varied among data types, with all models predicting fishery-dependent data more accurately than fishery-independent data. Differences in performance were primarily attributed to each model’s ability to explain the spatiotemporal dynamics of the training data. iSDMs that explicitly accounted for seasonal variability yielded the most accurate and ecologically realistic estimates. However, such approaches are computationally intensive and warrant identifying model purpose as an important step in the data-integration process. Main Conclusions:  Our findings reveal important trade-offs among the current techniques for integrating data in SDMs, including variability in accurately estimating species distributions, generating ecologically realistic predictions, and practical feasibility. With increasing access to growing and diverse data sources, maximizing our ability to leverage available data with robust analytical approaches will be instrumental in enhancing conservation and management efforts and for understanding current and future species distributions in a dynamic ocean.

Species distribution models (SDMs) are becoming an important tool for marine conservation and management. Yet while there is an increasing diversity and volume of marine biodiversity data for training SDMs, little practical guidance is available on how to leverage distinct data types to build robust models. We explored the effect of different data types on the fit, performance and predictive ability of SDMs by comparing models trained with four data types for a heavily exploited pelagic fish, the blue shark (Prionace glauca), in the Northwest Atlantic: two fishery-dependent (conventional mark-recapture tags, fisheries observer records) and two fishery-independent (satellite-linked electronic tags, pop-up archival tags). We found that all four data types can result in robust models, but differences among spatial predictions highlighted the need to consider ecological realism in model selection and interpretation regardless of data type. Differences among models were primarily attributed to biases in how each data type, and the associated representation of absences, sampled the environment and summarized the resulting species distributions. Outputs from model ensembles and a model trained on all pooled data both proved effective for combining inferences across data types and provided more ecologically realistic predictions than individual models. Our results provide valuable guidance for practitioners developing SDMs. With increasing access to diverse data sources, future work should further develop truly integrative modeling approaches that can explicitly leverage strengths of individual data types while statistically accounting for limitations, such as sampling biases.