Environmental variables and machine learning models to predict cetacean abundance in the Central-eastern Mediterranean Sea

Abstract

Although the Mediterranean Sea is a crucial hotspot in marine biodiversity, it has been threatened by numerous anthropogenic pressures. As flagship species, Cetaceans are exposed to those anthropogenic impacts and global changes. Assessing their conservation status becomes strategic to set effective management plans. The aim of this paper is to understand the habitat requirements of cetaceans, exploiting the advantages of a machine-learning framework. To this end, 28 physical and biogeochemical variables were identified as environmental predictors related to the abundance of three odontocete species in the Northern Ionian Sea (Central-eastern Mediterranean Sea). In fact, habitat models were built using sighting data collected for striped dolphins Stenella coeruleoalba, common bottlenose dolphins Tursiops truncatus, and Risso’s dolphins Grampus griseus between July 2009 and October 2021. Random Forest was a suitable machine learning algorithm for the cetacean abundance estimation. Nitrate, phytoplankton carbon biomass, temperature, and salinity were the most common influential predictors, followed by latitude, 3D-chlorophyll and density. The habitat models proposed here were validated using sighting data acquired during 2022 in the study area, confirming the good performance of the strategy. This study provides valuable information to support management decisions and conservation measures in the EU marine spatial planning context.

Endocrine-Disrupting Chemical Exposure Induces Adverse Effects on the Population Dynamics of the Indo-Pacific Humpback Dolphin

“…The impacts of defined threats on population abundance were explored using a random forest (RF) modeling approach, a widely utilized method in cetacean distribution and abundance studies. , Model cross-validation was executed by using 75% of the data set for training and the remaining portion for validation in each resampling iteration. Model training utilized 1000 trees (ntree value), and 2 to 15 predictors were sampled at each node for splitting (mtry value), with the selection aimed at minimizing the root mean square error.…”

Environmental variables influencing occurrence and distribution ofDelphinus delphisin the eastern Aegean Sea (eastern Mediterranean Sea)

“…Sixteen different physio‐chemical variables were used to investigate the presence–absence and distribution of dolphins in the study area (Table 1). Some of these variables are among the main physio‐chemical variables used in other studies on cetaceans, such as geographic coordinates, depth, distance from the coast, sea bottom temperature (BottomT), water column temperature (WCT), salinity, net primary production, Chl‐ a and phytoplankton and pH (Gaskin, 1968; Bush, 2006; Forney, 2006; Parra, Schick & Corkeron, 2006; Cañadas & Hammond, 2008; Di Tullio, Fruet & Secchi, 2015; Hornsby et al, 2017; Zanardo et al, 2017; Passadore et al, 2018; Chavez‐Rosales et al, 2019; Giralt Paradell, Díaz López & Methion, 2019; Correia et al, 2021; Milani et al, 2021; Torreblanca et al, 2022; Maglietta et al, 2023), and phosphorus and nitrogen (Muckenhirn, Bas & Richard, 2021). Other variables, such as dissolved oxygen, dissolved carbon, and dissolved ammonium (NH 4 ) were tested as proxies of local productivity.…”

“…In this framework, to achieve holistic and effective protection of Obviously, further analysis should be performed by testing other modelling approaches and/or different oceanographic variables (e.g., the euphotic depth, sea-level anomaly, or sea-surface current speed) and spatio-temporal variable aggregations (Cañadas & Hammond, 2008;Moura et al, 2012;Cañadas & Vázquez, 2017;Giannoulaki et al, 2017;Giménez et al, 2017;Giménez et al, 2018;Karamitros et al, 2020;Bonizzoni, Furey & Bearzi, 2021;Gannier, 2021;Maglietta et al, 2023).…”

“…Currently, studies on the topic have demonstrated the importance of different environmental, ecological and human pressures factors in influencing the distribution and occurrence of cetaceans (e.g., Carlucci et al, 2016; Carlucci et al, 2018a; Carlucci et al, 2018b; Derville et al, 2018; Passadore et al, 2018). The most common variables used to model and/or predict occurrence/distribution of species include characteristics of the habitat, such as bathymetry, distance to coast, slope, salinity, and sea‐surface temperature, as well as prey distribution (Giannoulaki et al, 2017) and variables used as proxies of prey availability or of oceanographic processes that enhance local productivity, such as chlorophyll a (Chl‐ a ), primary productivity, and phytoplankton carbon biomass (Gaskin, 1968; Bush, 2006; Forney, 2006; Parra, Schick & Corkeron, 2006; Cañadas & Hammond, 2008; Di Tullio, Fruet & Secchi, 2015; Hornsby et al, 2017; Zanardo et al, 2017; Passadore et al, 2018; Chavez‐Rosales et al, 2019; Giralt Paradell, Díaz López & Methion, 2019; Correia et al, 2021; Milani et al, 2021; Torreblanca et al, 2022; Maglietta et al, 2023) or phosphorus and nitrogen (Muckenhirn, Bas & Richard, 2021). Environmental information can be collected locally during field surveys (uncommon) or retrieved from online databases that provide long‐term time series of satellite‐derived images, allowing a better assessment of mesoscale, seasonal, and long‐term variability of the marine ecosystem (Skliris et al, 2010).…”