Krill Motion in the Southern Ocean: Quantifying in Situ Krill Movement Behaviors and Energetics

Abstract

Krill are a major link in the transfer of carbon between primary production and upper trophic level organisms, like whales and penguins, in the Southern Ocean. However, not much is known about in situ individual krill behaviors, and there have been no seasonal comparisons of individual krill motility in the Southern Ocean. To address how individual krill movement behaviors and energetic requirements shift between seasons, we used a novel stereo-camera and environmental sensor system to observe krill in three bays in the Western Antarctic Peninsula in May-June 2013 (i.e., the late austral autumn) and December 2014 (i.e., the late austral spring). Krill abundances and movement behaviors were determined from in situ image sequences collected for up to 10 minutes at specific depths throughout the water column, up to 625 m deep; using a semi-automated tracking method, 3345 individual krill tracks were collected. We found that seasonal changes in individual krill behaviors coincided with seasonal shifts in krill vertical distributions. During late spring, net upward swimming direction (0.9 ± 2.1° from horizontal) and vertical velocity (0.3 ± 0.2 Body Lengths (BL) s-1) resulted in shallower maximum abundances of krill within the water column proximate to near-surface phytoplankton distributions. During late autumn, krill swimming patterns tended downward, including swimming direction (-5.2 ± 0.8° from horizontal) and vertical velocity (-0.1 ± 0.0 BL s-1), leading to deeper distributions proximate to the benthos. Individual krill motility was greater in spring than autumn, as evidenced by a significant increase in swimming speeds (5.4 ± 0.2 BL s-1 vs. 2.8 ± 0.0 BL s-1) and turning rates (120 ± 5° s-1 vs. 106 ± 2° s-1); however, krill were capable of swimming just as quickly in late autumn as late spring. Furthermore, we found that krill consumed up to 18% of their carbon weight (CW) in late spring and up to 11% CW in late autumn, larger than the estimates of carbon requirements currently assumed in most krill models. Despite krill consuming more carbon in late spring than late autumn, krill required 15 - 48% higher concentrations of carbon in the water column in late autumn than late spring to cover energetic costs, likely due to the decrease in prey encounter rates with decreasing swimming speeds. Moreover, modeled diel shifts in krill motility demonstrated how shifts in krill swimming speeds can result in different energetic costs and prey concentration requirements. The most ideal motility pattern for