Open tools and standard for species interactions

Talk for the 4th GLOBIS-B workshop, Bari, Italia, Feb. 2018

Transcript

1. cb Tools for species interactions Open data, tools, and standards Timothée Poisot Université de Montréal February 26, 2018
2. But first... let’s talk about networks.
3. SPECIES INTERACTIONS VARY · β diversity of networks · Need to identify common species across networks Poisot et al. 2012 – Ecology Letters
4. SPECIES INTERACTIONS EVOLUTIONARY SIGNAL IS CONSERVED ACROSS SCALES · Co-phylogenetic signal in networks and in interactions · Need to link networks to phylogenies Poisot & Stouffer 2017 – Oikos
5. SPECIES AND INTERACTIONS DISTRIBUTE DIFFERENTLY · Impact of bioclimatic variables on species/interactions · Need to extract metadata from networks Poisot et al. 2017 – Global Ecology & Biogeography
6. SPECIES INTERACTION NETWORKS HAVE A BIOGEOGRAPHIC SIGNAL · Global distribution of network properties · Need to do all of the above Poisot et al. 2015 – Ecography
7. SPECIES INTERACTION NETWORKS HAVE A BIOGEOGRAPHIC SIGNAL · Global prediction of network properties · Need to do all of the above Gravel et al. 2016 – biorXiV
8. SPECIES INTERACTIONS NEED TRAITS TO BE PREDICTED · Predictive success with and without traits · Need to data on functional traits Desjardins-Proulx et al. 2017 – PeerJ
9. WWJHPS (WHAT WHOULD JULES HENRI POINCARÉ SAY)? The scientist must seek order. We build science with facts as we build houses with stones. But an accumulation of facts is no more a science than a heap of stones is a house.
10. BEYOND THE ACCUMULATION OF FACTS · Link network nodes to taxonomic backbones · Granular metadata (where, when, who, what) · Hierarchy of objects · Ability to join with external databases
11. DESIGN THE API FIRST · more data = less incentive to look at them · easier to link to other projects
12. CURRENT DATA FORMAT Poisot et al. 2013 – Ecography
13. MANGAL.IO
14. WEB OF LIFE
15. GLOBI
16. INTERACTION WEB DATABASE
17. COMPARISON Database Networks Interactions Metadata API mangal yes yes yes yes GLOBI no1 yes yes yes web of life yes no no 2 no IWDB yes no no no 1for some definitions of network… 2present, but not accessible programmatically
18. SO WHAT? · diversity is good (userbases, edge cases, …) · agree on a Minimum Data Specification for Ecological Interactions · use it to facilitate the mining of multiple databases