The response of marine predators to global climate change and shifting ocean conditions is tightly linked with their environment and prey. Environmental data are frequently used as proxies for prey availability in marine predator distribution models, as the ephemeral nature of prey makes sampling difficult. For this reason, the functional, ecological links between environment, prey, and predator are rarely described or explicitly tested. We used 3 years of vessel-based whale survey data paired with oceanographic sampling and hydroacoustic backscatter to model trophic relationships between water column structure, krill availability, and blue whale Balaenoptera musculus brevicauda distribution in New Zealand’s South Taranaki Bight region under typical (2014 and 2017) and warm (2016) austral summer oceanographic regimes. The warm regime was characterized by a shallower mixed layer, and a stronger, thicker, and warmer thermocline. Boosted regression tree models showed that krill metrics predicted blue whale distribution (typical regime = 36% versus warm regime = 64% cross-validated deviance explained) better than oceanography (typical regime = 19% versus warm regime = 31% cross-validated deviance explained). However, oceanographic features that predicted more krill aggregations (typical regime) and higher krill density (warm regime) aligned closely with the features that predicted higher probability of blue whale presence in each regime. Therefore, this study confirms that environmental drivers of prey availability can serve as suitable proxies for blue whale distribution. Considering changing ocean conditions that may influence the distribution of marine predators, these findings emphasize the need for models based on functional relationships, and calibrated across a broad range of conditions, to inform effective conservation management.

No abstract

Micronekton are a key component of the pelagic food web of the Chatham Rise east of New Zealand. The Chatham Rise is an important fishing area for hoki (Macruronus novaezelandiae), New Zealand's largest finfish fishery, and a predator on mesopelagic fish. Four fisheries oceanographic voyages provided multi-frequency acoustic data (18, 38, 70, 120, and 200 kHz) and midwater trawls, which were used to define a classification tree to separate micronektonic organisms. We carried out validation and sensitivity analyses that showed that we were able to classify pearlside (Maurolicus australis) and euphausiids. Other mesopelagic targets (mainly myctophids) were classified together based on their acoustic frequency response. Using scripting in the open-source software ESP3, we applied our classification tree to an independent time series of acoustic data from trawl surveys on the Chatham Rise between 2009 and 2018, that was not used for model development or validation. Our methodology allowed us to study temporal and spatial patterns of M. australis, euphausiids, and total backscatter in the water column. Total backscatter associated with micronekton has varied over the last 10 years, with no clear trend. The abundance of euphausiids showed a significant decreasing trend over the last 10 years. Abundance of M. australis also decreased since 2012, though this was not significant. This work contributes to ongoing efforts to monitor and detect changes in the pelagic ecosystems.

Acoustic transects by vessel Fig. S1. Acoustic transects (n = 6) collected between 2010 and 2014 along the transit between New Zealand and the Southern Ocean by fishing vessel Janas.