Abstract
Mesopelagic fish, being in the middle of the trophic web, are important key species for the marine environment; yet limited knowledge exists about their biology and abundance. This is particularly true in the Mediterranean Sea where no regional assessment is currently undertaken regarding their biomass and/or distribution. This study evaluates spatial and temporal patterns of mesopelagic fish biomass in the 1994–2011 period. We do that for the whole Mediterranean Sea using two well-established statistical models, the Generalized Additive Model (GAM) and Random Forest (RF). Results indicate that the bathymetry played an important role in the estimation of mesopelagic fish biomass and in its temporal and spatial distribution. The average biomass over the whole time period reached 1.08 and 0.10 t/km2, depending on the model considered. The Western Mediterranean and Ionian Seas were the sub-regions with the highest biomass, while the Adriatic was the area with the lowest. Temporal trends showed different trajectories with steep decrease and a fluctuation, using respectively RF and GAM. This study constitutes the first attempt to estimate the biomass and the spatial temporal patterns of mesopelagic fish using environmental variables as predictors. Given the growing interest in mesopelagic fish, our study sets a baseline to further develop mesopelagic fish biomass assessments in the region. Our results stress the need to improve data collection and quality in the region while identifying appropriate tools to better understand and assess the processes behind mesopelagic fish dynamics in the basin.
Highlights
Mesopelagic fish species are considered the marine vertebrates with the highest biomass in the world ocean (Mann, 1984)
The most statistically significant results were obtained with Random Forest (RF)
Bathymetry was the most important predictor for both models followed by salinity, temperature, and primary production (PPR) for RF and by dissolved oxygen (DO), PPR, and salinity for Generalized Additive Model (GAM) (Table 1)
Summary
Mesopelagic fish species are considered the marine vertebrates with the highest biomass in the world ocean (Mann, 1984). The net-based method, on the other hand, is likely to underestimate the mesopelagic biomass through escapement (e.g., large trawls and large meshes) or avoidance (e.g., small nets and fine meshes) phenomena (Davison et al, 2015; Escobar-Flores et al, 2020). This method is still the most used and informative (e.g., providing coordinates and exact number of specimens) and provides a ground-truth of size and species composition of the sampled fish (Cotter, 2009; Davison et al, 2015)
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call