Big data opportunities and challenges for assessing multiple stressors across scales in aquatic ecosystems

Abstract

Abstract. Aquatic ecosystems are under threat from multiple stressors, which vary in distribution and intensity across temporal and spatial scales. Monitoring and assessment of these ecosystems have historically focussed on collection of physical and chemical information and increasingly include associated observations on biological condition. However, ecosystem assessment is often lacking because the scale and quality of biological observations frequently fail to match those available from physical and chemical measurements. The advent of high-performance computing, coupled with new earth observation platforms, has accelerated the adoption of molecular and remote sensing tools in ecosystem assessment. To assess how emerging science and tools can be applied to study multiple stressors on a large (ecosystem) scale and to facilitate greater integration of approaches among different scientific disciplines, a workshop was held on 10-12 September 2014 at the Sydney Institute of Marine Sciences, Australia. Here we introduce a conceptual framework for assessing multiple stressors across ecosystems using emerging sources of big data and critique a range of available big-data types that could support models for multiple stressors. We define big data as any set or series of data, which is either so large or complex, it becomes difficult to analyse using traditional data analysis methods.

Freshwater biomonitoring in the Information Age

“…The progressive adoption of metabarcoding for taxonom ical identification will substantially increase the volume of data produced by biomonitoring activities and modify the characteristics of these data (Dafforn et al 2016). It is often stated that characteristics of big data fulfill five "Vs": vol ume, velocity, variety, variability, and value (Fan and Bifet 2013).…”

New approaches to the ecological risk assessment of multiple stressors

“…Novel types of data and analyses are being generated Dafforn et al 2016) and a better screen of their quality (see Big data section below) will allow for more accurate and representative ecological models. This will, in turn, improve our ability to predict impacts caused by multiple stressors and their regulation.…”

“…Big data is an all-encompassing term for any collection of datasets so large and complex that it becomes difficult to process using traditional data-processing applications. The improved collection and analysis of big data can potentially contribute to ERA, by identifying and providing evidence of major ecosystem stressors, improve monitoring of the environment or provide information on the environmental values of interest (see Dafforn et al 2016). In the years to come, the volume and variety of big data that will be collected and analysed will be even greater (Hampton et al 2013;Ballard et al 2014).…”

“…So as to assess how emerging science and new tools can be applied to study multiple stressors at a large (ecosystem) scale and to facilitate greater integration of approaches among different scientific disciplines, a workshop was organised 10-12 September 2014 at the Sydney Institute of Marine Sciences, Sydney, Australia . This workshop aimed to explore the potential offered by these new approaches to characterise stressor regimes, to explore stressorresponse relationships among biota, to design better early warning systems and to develop smart tools to support sustainable management of human activities through better-informed risk assessment Dafforn et al 2016). A key focus of the workshop was to more closely integrate diverse approaches developed by aquatic scientists working on marine and freshwater systems, so as to improve the integrated assessment of risks of multiple stressors.…”

New diagnostics for multiply stressed marine and freshwater ecosystems: integrating models, ecoinformatics and big data

“…The first paper in the workshop series by Dafforn et al (2016) focuses on two critical issues in multiple-stressor ecology: (i) developing a conceptual framework for study, which takes into account issues of spatial and temporal scaling in relation to the DPSIR framework; and, (ii) the availability of suitable data, including 'big data' sources, which can provide information on Driver-Pressure-State relationships. Their paper includes a brief critique of existing multiple-stressor approaches, which are largely ad hoc and expert knowledge-driven, and thus suffer from a lack of formal rigour, and provide a poor model for a multiple-stressor paradigm, which should be extendable from ecosystem to planetary scale.…”