I need to start thinking about how we define reproducibility for a) ecological models b) ecological models in the context of decision support.
Reviewing Fiona’s article on metaresearch on ecology would be a good place to start thinking about reproducibility in ecology.
-> Reproducibility in ecology… why important? “the inherent complexity of the inference chain needed to advance ecology as well as the importance of ecological results to challenges important to society” https://eco.confex.com/eco/2015/webprogram/Paper50853.html(https://eco.confex.com/eco/2015/webprogram/Paper50853.html)
surveying reproducibility attempts in ecology
There seems to be a lot of scattered, grass-roots efforts to develop reproducibility standards in ecology, especially at the level of the individual researcher or the research group level. These are often in the form of blog posts. There also seem to be quite a few at the ‘application’ level, for example, looking at niche models, and often have some software package distributed with them that tend towards the modular and the scalable, e.g. model-R and Wallace R packages (Kass et al 2018).
Increasingly there is more work being done at the institutional level (e.g. the British Ecological Society has published guidelines for implementing reproducible ecological analyses, Ecography also has a special edition.). ESA hosted a special session / plenary on cutting edge developments in data analyses and management in ecology https://eco.confex.com/eco/2015/webprogram/Session10539.html. There is also emerging work over recent years examining the extent of the ‘reproducibility’ crisis in ecology (fidler’s work).
Workflows and Data management
Reproducibility emphasised from a data management perspective, coming at the problem from a bioinformatics discipline / approach: focus on process of locating, understanding and accessing data, as well as data-derived products such as visualisations andmodelling that form the basis for scientific conflucions. Yet methods sections in journals are inadequate, current data repositories support minimal descriptions. MB Jones et al designed DataONE repository based on this.
Focuses on data storage, description, as well as describing the complete workflow used in scientific analyses. Aim to postulate standards and tools.