Inferring the multi-host fitness landscape of endive necrotic mosaic virus from cross-inoculation experiments

A Bayesian framework infers an effective multi-host phenotypic fitness landscape for endive necrotic mosaic virus from cross-inoculation on five Asteraceae hosts, mapping host permissiveness heterogeneity.

• Why it matters: Fitness landscapes are hard to infer from sparse multi-environment data, limiting how adaptation across hosts can be summarized and interpreted mechanistically.
• What they did: The method links successful infection probabilities to Fisher’s geometrical model and decomposes establishment routes into direct establishment, rescue from standing variation, or de novo rescue in the target host.
• The result: Applied to the full cross-inoculation design, it estimates host-optimum distances and peak widths, yielding a phenotypic host-community map with geometry broadly concordant with host phylogeny.