Trying to include spatially explicit interactions in mutualistic networks

[guifandos]

Interactions between neighbouring plants can influence mutualistic bipartite networks.
Spatial approaches to modelling neighbour interactions are necessary, as the distance between neighbouring plants can influence the strength of the interaction.
Recent Bayesian advancements offer the flexibility to fit models of spatially explicit neighbour interactions.
The suggestion is to try to see if it is possible to adapt those spatially models when you are inferring plant−animal interaction accounting for uncertainty derived from sampling.

• Review literature to understand approaches that have been used to leverage spatial relationships to inform mutualistic networks
• Start with a building block of a neighbourhood model, an interaction kernel. Following Canham and Uriarte (2006), we assume that the kernel alters the expected value of an interaction rate measured at a target plant.
• Using ragged matrices in a neighbour interaction model. See methods section here