Phenology System

[Megachile]

This is an initial brainstorm for a set of features that are designed to facilitate collecting and rearing inducers. I think some of these features may be modular in the sense that we could build some and use them without doing them all at once.

Desired user-facing features (presented in AGDD or location-adjusted dates):

• A species-specific timeline on each gall page
• Calendar of upcoming gall events (a new display mode for the ID tool, or at least filterable using many of the same traits)
• List of adult cynipids known to be present at a given phenology

Admin-facing infrastructure needed:

• For each gall, a way to associate a set list of important events with AGDD values
• For each host, a way to associate a set list of important events with AGDD values

At some stage of the process, location+date needs to be converted into and out of AGDD. The most convenient for the admin and end user would be to have this conversion built into the site, but I think initially we could get away with skipping that and making both admins and end users do the conversions themselves, manually. That would allow us to begin collecting and presenting this data before solving those issues.

The exact details of the association between Gall and AGDD could be done several ways. We could just ask "what is the earliest and latest AGDD ever reported for 1) gall sightings 2) adult emergence" etc. This is likely the best we can do for galls with one or few observations, which would be a main target of the function anyway. But I'm toying with two other more complex/powerful approaches too.

One would involve storing multiple dates and doing some kind of statistical analysis to produce the graph. I find the idea of having to manually determine which AGDD is earliest and then erasing prior data suboptimal; it would be better to be able to just accumulate data in the site so we could use it to make the feature more informative somehow.

The other is using data about host plant phenology from NPN to infer possible date ranges for the gall. This is something that would be especially useful in cases where we've only seen an old gall and aren't sure when it first appeared, but it might still provide a reality check bound on estimates made with a dumb AGDD approach. It's also possible that correlation with host phenology is a better predictor than AGDD in general. These are two things I'd like to test as a proof of concept before we even think about implementing any of this, but I don't know how to pull iNat observations or NPN data in bulk to do it.