Large marine predator aerial survey data for Hauraki Gulf, New Zealand

Large marine predators, such as cetaceans and sharks, play a crucial role
in maintaining biodiversity patterns and ecosystem health. Despite the
recognised importance of these animals and their over-representation as
threatened species, distribution data at appropriate temporal and spatial
scales is often lacking or insufficient for effective
conservation.  Here, we present sightings of large marine
megafauna recorded from a replicate systematic aerial survey undertaken in
the Hauraki Gulf, Aotearoa New Zealand during a full year. Using flexible
machine learning models (Boosted Regression Tree models), we use these
sightings data to investigate relationships between large marine predator
occurrence (Bryde’s whales, common and bottlenose dolphins, bronze
whalers, pelagic and immature hammerhead sharks) and spatially explicit
environmental and biotic variables to predict species richness of large
marine predators and investigate their fine-scale spatiotemporal
distribution patterns. All models were considered informative (all, AUC
> 0.78), and temporally dynamic variables, such as the distribution
of prey, were important in predicting the occurrence of the study species
and species groups.  Our approach and data highlight the value of
multi-species surveys and the importance of considering temporally
variable abiotic and biotic drivers for understanding biodiversity
patterns when informing ecosystem-scale conservation planning and dynamic
ocean management. We provide data files of: Locations of species presence
/ pseudo absence location over time (in .csv format) and associated
environmental and biotic variables for Bryde’s whales, common and
bottlenose dolphins, bronze whaler, pelagic and immature hammerhead
sharks. Monthly estimates of 14 high-resolution spatially explicit
environmental and biotic variables (1 km grid resolution): Bathymetry;
Slope; Distance from shore; Distance to 40m depth; Sand; Mud; Gravel;
Seabed disturbance; Tidal current; Chlorophyll a; Sea surface temp;
Distance to plankton; Distance to prey (saved as .R data) Model objects, R
code, and model outputs of the Boosted Regression Tree modelling.
Predicted monthly distributions (and associated spatially explicit
uncertainty) of Bryde’s whales, common dolphins, bottlenose dolphins,
bronze whaler sharks, pelagic sharks, immature hammerhead sharks, and
richness of large marine predators in the Hauraki Gulf, New Zealand,
(January to December). Monthly richness estimates of large marine
predators.