Seasonal habitat preference and foraging behaviour of post-moult Weddell seals in the western Ross Sea

Weddell seals (Leptonychotes weddellii) are important predators in the Southern Ocean and are among the best-studied pinnipeds on Earth, yet much still needs to be learned about their year-round movements and foraging behaviour. Using biologgers, we tagged 62 post-moult Weddell seals in McMurdo Sound and vicinity between 2010 and 2012. Generalized additive mixed models were used to (i) explain and predict the probability of seal presence and foraging behaviour from eight environmental variables, and (ii) examine foraging behaviour in relation to dive metrics. Foraging probability was highest in winter and lowest in summer, and foraging occurred mostly in the water column or just above the bottom; across all seasons, seals preferentially exploited the shallow banks and deeper troughs of the Ross Sea, the latter providing a pathway for Circumpolar Deep Water to flow onto the shelf. In addition, the probability of Weddell seal occurrence and foraging increased with increasing bathymetric slope and where water depth was typically less than 600 m. Although the probability of occurrence was higher closer to the shelf break, foraging was higher in areas closer to shore and over banks. This study highlights the importance of overwinter foraging for recouping body mass lost during the previous summer.

Phytoplankton Blooms Expanding Further Than Previously Thought in the Ross Sea: A Remote Sensing Perspective

Accurate and robust measurements from ocean color satellites are critical to studying spatial and temporal changes of surface ocean properties. Satellite-derived Chlorophyll-a (Chl) is an important parameter to monitor phytoplankton blooms on synoptical scales, particularly in remote seas. However, the present NASA standard Chl algorithm tends to strongly underestimate the Chl in the Ross Sea. Based on a locally-tuned Chl algorithm in the Ross Sea and using the data record from MODIS between 2002 and 2020, here we investigated the spatial expansion of phytoplankton blooms in the Ross Sea. Our results show the geometric areas of the phytoplankton blooms could reach (7.20 ± 2.8) × 104 km2 on average, which was ~3.1 times that of those identified using the NASA default Chl algorithm. Spatially, blooms were frequently identified on the shelf of the Ross Sea polynya with a typical chance of ≥80%. In the context of climate change and global warming, the general decrease and interannual dynamics of sea ice cover tends to affect solar light penetration and surface seawater temperature, which were found to regulate the spatial expansion of the phytoplankton blooms over the years. Statistical analyses showed that the spatial coverages of the phytoplankton blooms were significantly correlated with sea surface temperature (Spearman correlation coefficient R = 0.55, at p < 0.05), sea surface wind speed (R = 0.42, at p < 0.05), and sea ice concentration (R = −0.84, at p < 0.05), yet without significant long-term (>10 years) trends over the study period. The stronger phytoplankton blooms than those previously observed may indicate larger carbon sequestration, which needs to be investigated in the future. More valid satellite observations under cloud covers will further constrain the estimates.

Data Reconstruction for Remotely Sensed Chlorophyll-a Concentration in the Ross Sea Using Ensemble-Based Machine Learning

Polar regions are too harsh to be continuously observed using ocean color (OC) sensors because of various limitations due to low solar elevations, ice effects, peculiar phytoplankton photosynthetic parameters, optical complexity of seawater and persistence of clouds and fog. Therefore, the OC data undergo a quality-control process, eventually accompanied by considerable data loss. We attempted to reconstruct these missing values for chlorophyll-a concentration (CHL) data using a machine-learning technique based on multiple datasets (satellite and reanalysis datasets) in the Ross Sea, Antarctica. This technique—based on an ensemble tree called random forest (RF)—was used for the reconstruction. The performance of the RF model was robust, and the reconstructed CHL data were consistent with satellite measurements. The reconstructed CHL data allowed a high intrinsic resolution of OC to be used without specific techniques (e.g., spatial average). Therefore, we believe that it is possible to study multiple characteristics of phytoplankton dynamics more quantitatively, such as bloom initiation/termination timings and peaks, as well as the variability in time scales of phytoplankton growth. In addition, because the reconstructed CHL showed relatively higher accuracy than satellite observations compared with the in situ data, our product may enable more accurate planktonic research.

Reconstruction of Ocean Color Data Using Machine Learning Techniques in Polar Regions: Focusing on Off Cape Hallett, Ross Sea

The most problematic issue in the ocean color application is the presence of heavy clouds, especially in polar regions. For that reason, the demand for the ocean color application in polar regions is increased. As a way to overcome such issues, we conducted the reconstruction of the chlorophyll-a concentration (CHL) data using the machine learning-based models to raise the usability of CHL data. This analysis was first conducted on a regional scale and focused on the biologically-valued Cape Hallett, Ross Sea, Antarctica. Environmental factors and geographical information associated with phytoplankton dynamics were considered as predictors for the CHL reconstruction, which were obtained from cloud-free microwave and reanalysis data. As the machine learning models used in the present study, the ensemble-based models such as Random forest (RF) and Extremely randomized tree (ET) were selected with 10-fold cross-validation. As a result, both CHL reconstructions from the two models showed significant agreement with the standard satellite-derived CHL data. In addition, the reconstructed CHLs were close to the actual CHL value even where it was not observed by the satellites. However, there is a slight difference between the CHL reconstruction results from the RF and the ET, which is likely caused by the difference in the contribution of each predictor. In addition, we examined the variable importance for the CHL reconstruction quantitatively. As such, the sea surface and atmospheric temperature, and the photosynthetically available radiation have high contributions to the model developments. Mostly, geographic information appears to have a lower contribution relative to environmental predictors. Lastly, we estimated the partial dependences for the predictors for further study on the variable contribution and investigated the contributions to the CHL reconstruction with changes in the predictors.