Elucidating ecological complexity: Unsupervised learning determines global marine eco-provinces
Abstract
Gaining ecological insight from ocean colour observations could address global and regional sustainability challenges with applications from carbon budgets to fisheries. However, gaining insight with conventional methods is difficult as colour data to date remains overwhelmingly complex, and links to ecology obscure. A key step towards colour data utilization is discovering emergent structure in ecology with unique colour signatures, where data-science methods could offer key advantages. An unsupervised learning method is presented for determining global marine ecological provinces (eco-provinces) from plankton community structure and nutrient flux data. The systematic aggregated eco-province (SAGE, figure below) method identifies eco-provinces within a highly nonlinear global ocean ecosystem model, which incorporates colour data. To accommodate the non-Gaussian covariance of the data, SAGE uses t-stochastic neighbor embedding (t-SNE) to reduce dimensionality. Over a hundred eco-provinces are identified with the density-based spatial clustering of applications with noise (DBSCAN) algorithm. Using a connectivity graph with ecological dissimilarity as the distance metric, robust aggregated eco-provinces (AEPs) are objectively defined by nesting the eco-provinces. Using the AEPs, the control of nutrient supply rates on community structure is explored. By allowing the unique identification of ecological assemblages on a range of spatial scales, planning for sustainable resource use, as well as vulnerability assessments, could be facilitated. The AEPs are already in use to inform the New Zealand marine protected area legislation. Eco-provinces and AEPs are unique and could serve as the basis for remote observation of ecosystem dynamics, a capacity that is increasingly needed under global climate change. Figure caption: The SAGE method workflow.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2021
- Bibcode:
- 2021AGUFMGC41A..08S