Scanning electron micrograph of Arctic diatoms, including several chains of Thalassiosira. Credit: Beth Caissie.
Diatoms are one of the most important primary producers in the world’s oceans. They secrete siliceous frustules that are often well preserved in sediments, and due to their diversity and selective environmental preferences, they are ideal proxies for paleoceanographic studies. In the Arctic, marine sedimentary diatoms are widely used to reconstruct sea-surface conditions, including temperature (SST), salinity, and sea-ice cover.
Reconstructions range from qualitative (e.g., presence of sea ice) to quantitative (e.g., seasonal SST values), but their robustness as paleoenvironmental indicators fundamentally depends on our understanding of the modern autecology of individual species. Although this can be studied through plankton surveys, sediment traps, and lab-culturing, the most common approach is by the study of surface sediment diatom assemblages and linking these to surface ocean conditions (so-called calibration data sets).
Over the last decades, hundreds of Arctic surface sediments have been analyzed for this purpose. This work has however been carried out by several independent research groups around the world, often using slightly different methodologies and diatom taxonomies. To enable their full potential for quantitative diatom-based surface ocean reconstructions, MARDI aims to align and integrate the various surface datasets resulting in an open-access Pan-Arctic diatom dataset.
The goal of the MARDI working group is to improve the precision and reliability of quantitative diatom-based palaeoceanographic reconstructions in the Arctic and subarctic. This will give diatom workers tools to create a wider spatial and temporal set of environmental reconstructions that can be used by paleoclimate modelers to calibrate models and improve data-model comparisons.
- MARDI Launch and start of collection of the various diatom surface sample datasets.
- Workshop 1 (January 2023): Establishing protocols
This initial workshop will:
1) identify existing Arctic and subarctic diatom-based calibration data sets including surface sediment assemblages and associated environmental surface ocean data and agree upon common criteria for integrating the databases. Additional new sites will adhere to these criteria; here special emphasis will be given to best practices for sampling surface sediment data.
2) Identify the challenges ahead for aligning the taxonomy of the various datasets and outline a plan towards Workshop 2.
- Workshop 2 (June 2023): Taxonomy
This workshop will be dedicated to integrating and agreeing upon a common taxonomy for all datasets. A key outcome of this workshop will be the outline of the publication of a taxonomic guide that includes descriptions of species ecology, images, and geographical distribution. Workshop 2 will be followed by several online taxonomic workshops on specific taxa during the following months.
- Taxonomy archiving (June 2023 – June 2024)
Gradual collection and archiving of diatom species information to online databases, based on the plans laid out in Workshop 2.
- Workshop 3 (July 2024): Statistical relationships
This workshop will focus on relating diatom assemblages to environmental data in the Arctic. Key questions to be answered will include:
1) how to integrate samples collected as far back as the 1960s with samples collected more recently,
2) how to determine whether surface samples are modern,
3) how much time is integrated in various surface samples,
4) how and where to access accurate and time appropriate environmental data, and
5) modern best practices in statistical analyses, transfer functions, and data mining. This workshop will not only include diatomists, but also data scientists and experts in quantitative paleoceanography.
- Synthesis Production (August 2024-September 2025)
Final integration of the datasets and association with environmental metadata as agreed in Workshop 3. Construction and publication of the final diatom-based transfer functions.
- Workshop 4: Training (October 2025)
In this concluding MARDI workshop, the final transfer functions will be presented in the form of a short course. The instructors will outline the principles and the limitations of the new approach