Afternoon Breakout Roundup

Condensed list of ideas that occurred in multiple groups or inspired lively conversation:

  • Speciation needs to be more explicitly modelled in both the local and regional pools, perhaps with some kind of ‘knob’ or ‘dial’.
    • Not clear whether this knob should go only up to or perhaps beyond 10.
    • Also, need to allow for endemics and endemic radiations, and adaptive radiations.
  • Investigate GDM-style varying carrying capacity through time in the local community.
  • Add trait evolution to the model (even with a simple quantitative genetics model)?
    • How do species traits evolve in these communities and how it relates to abundances?
    • Do island species explore their niches faster?

Unvarnished thoughts from all the groups following the monday afternoon breakout session.

Group 1

  • Having interactions in the local community feed back on the phylogeny is important.
  • Rare things often aren’t rare because they are bad at being a species, very often rare and persist through time as rare species, perhaps specializing on scarce resources. Are there more specialists on rare resources in the tropics. or in hyperdiverse systems in general.
    • The evolutionary ecology of rarity. What do we need to do to the model as it stands to account for this.
    • Evolution of traits w/in the local environment feeding back on speciation dynamics.
  • Full spatially explicit
    • More realistic coexistence mechanisms than we currently do.
    • Space will allow frequency dependence to emerge more readily.

  • Question: Is introducing all this complexity really increasing explanatory power?

  • “Can I say something about unpublished data that I’m not involved in?”

  • Devoloping good summary statistics
    • What’s diagnostic of competitive coexistence, for example?
    • SGDC could be another good place to look.
  • Rapid temporal dynamics? Antique data? Comparing to contemporary samples.

Group 3

  • Put a couple of key papers on the github
    • Rosindell harmon 2013
  • Volcanic and continental islands can be incorporated. Don’t want to limit to oceanic islands.
  • The tree in the regional pool is confusing. Why is it static w/ respect to the local community can we keep the regional community dynamic.
  • Corollary: Introduce speciation on the island.
  • Clustered immigration? Already exists.
  • Very early stage of colonization may be unrealistic if early colonizers expand to consume all available resources too quickly.
  • Have K be variable either as a parameter to estimate or as a GDM style variable island size through time.

Group 1

  • Types of questions empiricists want to ask from their data.
    • It wouldn’t be that interesting to classify on the neutral/compexclu/envfiltering that we’re offering right now.
    • We need some mechanism of speciation in our model.
    • Need a way for endemics to form, and also endemic radiations to form
  • Islands have a lifespan. Should be a high priority to capture this in the model (GDM-style).
  • Can we find the dial that spans lake victoria cichlids to a fully invasive species model?
  • Add trait evolution to the model (even with a simple quantitative genetics model)?
    • How do species traits evolve in these communities and how it relates to abundances?
    • Do island species explore their niches faster?
  • Non-equilibrium models? But now we’re satisfied.
    • Sense of equilibrium and sense of time are based on initial conditions which are arbitrarily set.

Group 5

  • They want a knob instead of a dial for the “speciation”, where evolution is happening, either all the way one way so evolution is happening at the regional scale or dialed the other way so all the action is happening in the local community.
    • Some of the species in the local community aren’t in the regional community at all, in some systems.
    • Speciation happening on the island is an important addition.
    • Potentially add another model to capture adaptive radiations (if the knob/dial doesn’t work).
  • Trait evolution and both anagenesis and cladogenesis happening on the islands.
  • Data organization: Source pool and one island in the simulations. For most of the groups we are going to have information from multiple islands, also islands that have been colonized multiple times.
    • Define what really is going to be useful in terms of data. Also for the empiricists
  • Difference between species abundance and effective population size. Specifically for species like plants that are clonal. Can we have a more nuanced approach to scaling abundance to Ne?
  • Broad goals: Comparing different systems: Birds, plants, gastropods. Comparing across timescales: similar things on young islands and old islands.

Group 2

  • Simplicity vs complexity in models. How to have so much but have it be tractable and elegant.
  • How to facilitate the hive-mind of everyone in the room:
    • How to make this operation worth-while?
    • 1a. List of potential papers and outputs
    • 1b. List of specific questions
      1. List of processes (Importance and universality)
        • Dispersal, establishment, competition, facilitation
      1. List of systems
      1. List of data types
      1. Link data and system to processes
        • Id priority and importance of processes
    • Come up with other models, evaluate current and perhaps propose new.
  • The overall island structure is the sum of local communities at sites within the island. Different islands may have the same regional structure, whereas the structure of sites might be quite different.
    • Perhaps more niche stuff at local scale and more neutral stuff at regional scale.
  • Data types specify how detailed the model needs to be.
  • What’s our question? Is it variation in diversity? Or variation in diversification?
  • From an empirical point of view what is the minimum we need?
  • Perhaps start the morning with the empirical people leading the groups to walk through steps 2-5 above?