| Summary: | This study shows the characterization of krill (Euphausia superba) aggregations identified in the Bransfield Strait and around of Elephant Island. Data were collected using a multifrequency SIMRAD EK80 echosounder during three austral summers: 2018, 2019 and 2020. For detection of krill aggregations, two frequencies (38 and 120 kHz) and an automated Echoview version 9 algorithm with the EchoviewR package in R were used. A total of 22,221 aggregations were detected. Acoustic descriptors were analyzed with Pearson’s correlation. For the characterization of krill aggregations, principal component analysis (PCA) was applied, followed by hierarchical clustering. To determine temporal differences of clusters, an ANOVA was applied. In addition, krill aggregations were assigned to surface environmental variables to apply a generalized additive model (GAM). Three clusters were identified using the first three dimensions of the PCA (which explained 81% of the total variability). The first cluster was characterized by krill aggregations having lower height (2 m), backscattering acoustic energy (7 m2 mn-2), and being located at a greater depth (81 m). The second cluster had the shallowest swarms (34 m), shortest length (75 m) and compactness (202). Finally, the third cluster had the largest swarms in length (849 m), volume (207,412 m3) and height (11 m); in addition of having greater acoustic energy (637 m2 mn-2), obliquity (6), compactness (2,436) and coefficient of variation (213). Spatially, cluster I was located with a greater presence around Elephant Island during 2018 and 2019, while for the same period, clusters I and II were located scattered throughout the study area but focused on the Bransfield Strait. By 2020, thermal anomalies of approximately + 2 °C were presented and a dispersion of the three clusters was noted throughout the study area, where cluster I was located with a greater presence in the Bransfield Strait. Significant differences (p < 0.05) were found among the clusters per year. However, such ...
|
|---|