Methods to create and assess deep and open-ocean ecosystem timeseries

In Section 1, we discuss the value of collating ecosystem time series in the deep and open ocean. In the context of planning for blue growth in ocean-based economies, our time series now provide the critical backbone to support decision-making in exclusive economic zones and in the High Seas by enabling trends and tipping points to be identified in these otherwise remote hard to access places.

In Section 2, we introduce the focal ecosystems and living compartments in our 12 Study Regions, and briefly explain the multidisciplinary approaches we took to collect these data. We then feature six methods used to create new time series from data originally gathered for other purposes. These include: (i) photogrammetry at a hydrothermal vent; (ii) capture-recapture modelling of humpback whales based on fluke identification photos; (iii) zooplankton biomass estimates from acoustic Doppler conductivity profilers; (iv) reconstruction of historic and ancient deep and open ocean ecosystems from geochemical samples; (v) palaeoreconstructions of ecosystems from ancient DNA in sediment cores; and (vi) Lagrangian particle tracking models based on capelin life history information and ocean physics.

Section 3 provides the full list and more in-depth descriptions of each time series organised across four categories: (i) marine ocean observatories (MOBs), (ii) monitoring programmes and long-term surveys, (iii) fisheries dependent surveys, and (iv) geochemistry.

Heterogeneity of our time series datasets was a critical issue for iAtlantic to consider. Section 4 describes three main sources of heterogeneity that needed consideration, including: spatial, temporal and simply the huge range of different types of data in each Study Region.

Section 5 describes our approach to overcome this heterogeneity by harmonising subsequent analyses to be reported in Deliverable 3.2, “Ecosystem drivers of change and tipping points”. First, we established three common research questions for each Study Region; second, we coupled each time series to a common set of VIKING20X and INALT hindcasts of temperature, salinity and currents; third, we created a common statistical workflow for univariate and multivariate analyses to standardise our mathematical approaches to quantify drivers and tipping points.

International multidisciplinary collaborations and networking opportunities called for by the Belém and Galway Statements were critical for extending our capacity to locate or create time series in the deep- sea and open Atlantic. The Atlantic community must work towards standardisation and align ocean observations. However, Deliverable 3.1 shows that a flexible place-based approach still allows existing time series to be integrated into ecosystem assessments as part of any basin-scale observation programmes, provided this is underpinned by a common set of over-arching questions answered with common design principles. We propose methodological approaches that can be applied in the future to many additional existing samples and time-series in the Atlantic, thus setting the methodological ground for future large scale ecosystem assessment.