Oceanic whitetip shark (Carcharhinus longimanus) is an important top predator in pelagic ecosystems currently classified as globally Critically Endangered by the International Union for the Conservation of Nature. This species is incidentally caught by fisheries targeting highly migratory tunas and billfishes throughout the Indian Ocean. Understanding the temporal, spatial and environmental factors influencing the capture of this species is essential to reduce incidental catches. In this study, we used generalized additive models to analyze the spatio-temporal distributions of the juvenile oceanic whitetip shark catches and the environmental conditions in the western Indian Ocean using observer data from 2010 to 2020 of the European Union and associated flags purse seine fishery. We found sea surface temperature and nitrate concentration to be the most important environmental variables predicting the probability of catching an oceanic whitetip shark. A higher probability of capture was predicted in areas where sea surface temperature was below 24°C and with low nitrate concentrations close to zero and intermediate values (1.5-2.5 mmol.m-3). We also found a higher probability of capture in sets on fish aggregating devices than in sets on free schools of tuna. The Kenya and Somalia basin was identified to have higher probabilities of capture during the summer monsoon (June to September) when upwelling of deep cold waters occurs. We provide the first prediction maps of capture probabilities and insights into the environmental preferences of oceanic whitetip shark in the western Indian Ocean. However, the causal mechanisms behind these insights should be explored in future studies before they can be used to design spatial management and conservation strategies, such as time-area closures, for bycatch avoidance.

Background: Harmful algal blooms (HABs) of the toxic dinoflagellate Karenia brevis impact the overall ecosystem health. Methods: K. brevis cell counts were extracted from Harmful Algal BloomS Observing System (HABSOS) in situ data and matched with 0.25º resolution environmental information from the Copernicus database to generate spatio-temporal maps of HABs in the Gulf of Mexico (GoM) between 2010 and 2020. The data was used to analyze the relationship between spatial and temporal variability in the presence/absence of K. brevis blooms (≥100,000 cells/L) and biotic and abiotic variables using Generalized Additive Models (GAM). Results: The variability of blooms was strongly linked to geographic location (latitude and salinity), and temporal variables (month and year). A higher probability of K. brevis blooms presence was predicted in areas with negative sea surface height (SSH) values, silicate concentration (0, 30-35 mmol. m-3), sea surface temperature of 22-28 oC, and water currents moving south-westward (225º). The smooth effect of each environmental variable shows a bimodal pattern common in semi-enclosed basins such as GoM. The spatial predictions from the model identified an important permanent area in (1) Southwest Florida (25.8-27.4o latitude), and four seasonally important areas, (2) North Central Florida (3) Central West Florida, (4) Alabama on Gulf Shores and (5) Mississippi with higher bloom probabilities during the fall to winter season (November-January). Results also suggest that HABs can extend until ≥ 300 km offshore; starting to form in March and reaching a peak in September, and were swept to the coastal area during fall and winter. This suggests the role of upwelling and water circulation in GoM for the accumulation of cells and HABs. Information on the spatio-temporal dynamics of K. brevis blooms and understanding the environmental drivers are crucial to support more holistic spatial management to decrease K. brevis blooms incidence in bodies of water.