Differential distribution patterns of Balaenoptera spp. from tropical areas

• ¹ Oceanographic Center of Málaga, Spanish Institute of Oceanography (IEO), Spain
• ² University of A Coruña, Spain
• ³ Centro Oceanográfico de Canarias, Instituto Español de Oceanografía (IEO), Spain

Tropical tunas (i.e. Katsuwonus pelamis, Thunnus albacares, and Thunnus obesus) tend to aggregate around all types of floating objects (both natural and artificial), including whales. Cetaceans, especially spotted dolphins in the eastern Pacific Ocean, are also found in association with tuna in tropical areas. Some authors suggest that records from the fishing industry is the only data source for reporting whales occurrence in theTropical open ocean waters. Currently, 6 species of the genus Balaenoptera are known in the Atlantic Ocean and only three of them have been recorded in tropical regions: B. physalus (fin whale), B. edeni (Eden’s whale) and B. musculus (blue whale). The industrial Spanish purse seine fleet targeting tropical tunas perform an important activity in the tropical Atlantic region. The main aim of present study was to understand the differential distribution patterns of Balaenoptera in relation to other cetaceans in the tropical Atlantic region. The Spanish Institute of Oceanography (IEO) observers’ program on board commercial purse seine vessels follows a scientific programme developed under the EU Data Collection Framework. The main aim of this observer program is to obtain direct information on bycatch species, as well as on discards of target species (e.g. catch and bycatch species identification, number of individuals, size, and other biological data). In the present study, we used data recorded by IEO from 2003 to 2016 from the above-mentioned programme. These data included 189 cetaceans sightings observed during the fishing operation: 19 Balaenoptera spp., 39 dolphins, and 131 other cetaceans (including Ziphiidae, whales no Balaenoptera) (Figure 1). Balenoptera ocurrence was modelled as a function of longitude, latitude and the following environmental variables: sea surface temperature (SST), sea surface height (SSH), meridional surface velocity (v), zonal surface velocity (u), mixed layer depth (MLD), dissolved oxygen concentration (DO), and sea surface salinity (SAL). Moreover, we included the spatial variables: longitude and latitude. Model coefficients were assessed by means of an omnibus test and the goodness-of-fit between expected and observed proportions of by-catch events was evaluated using the Hosmer and Lemeshow test (which also follows a Chi-square distribution; low p-0.05 would indicate lack of fit of the model). On the one hand, the Omnibus test examines whether there are significant differences between the -2LL (less than twice the natural logarithm of the likelihood) of the initial step, and the -2LL of the model, using a Chi-squared test with one degree of freedom. On the other hand, the Hosmer and Lemeshow test compares the observed and expected frequencies of each value of the binomial variable according to their probability. In this case we expected that there are no significant differences for a good model fit. In addition, the discrimination capacity of the model (trade-off between sensitivity and specificity) was evaluated with the receiver operating characteristic (ROC) curve. Furthermore, the area under the ROC curve (AUC) provides a scalar value representing the expected discrimination capacity of the model. The relative importance of each variable within the model was assessed using the Wald test. The most parsimonious model obtained included SST and u as explanatory variables. According to the Omnibus test, the adjusted model is significant (χ² = 39.678, 2 degrees of freedom, P <0.001). In addition, no significant differences were observed between the observed values of presences and absences, and those expected by the model, by means of the Hosmer & Lemeshow test (χ² = 2.802, 8 degrees of freedom P = 0.946). The likelihood (-2LL) was 83.642 and R2 of Nagelkerke = 0.395. The logit connection function is expressed as: Y = 25.189 – 1.147 * SST – 9.44 * v According to the Wald test (where α = 0.05) these parameters of the model were considered to be significant. Finally, the AUC obtained was 0.890, which reflects a good adjustment. Our results suggested that the habitat preference of the genus Balaenoptera from the tropical zone is related to lower SSTs and the surface current component. On the other hand, the sightings distribution evidences that the majority of the cetaceans observed, and especially of the genus Balaenoptera, are concentrated in areas close to the coast (figure 1). This could be due to the upwelling occurring next to the coast, which result in lower SSTs and increased productivity.