Moore Notes 1 11 11

Group Call

• Niche mapping
  • Tom: Environmental layer associations plots
    • MICROBIS data only: remote data vs. measurement at sampling site by MICROBIS researchers
    • Don't know units for MICROBIS (hopefully scales linearly with remote scale)
    • Can't be sure that type/mode of measurement are identical between the two (e.g., soluble vs. non-soluble iron)
    • Need to keep in mind that time of year of measurements could differ
    • Silicate only has 3 values - can't really compute correlation
    • All other variables except phosphate have significant positive correlations
    • Josh and Tom will check if nitrate and/or the nitrate/phosphate ratio have good correlations - may substitute nitrate and/or add N/P ratio
  • Josh: GLM results
    • Richness vs. latitude plot
      • Sample area lower at poles and extreme latitudes
      • Richness is higher at mid latitudes
      • If we did correct for area, it would accentuate this pattern
      • People usually don't do this, but some people do use diversity per unit area (log of proportion of area sampled to total area sampled)
      • Alternatively could do a simulation by resampling grid cells with probability based on their relative area to account for correlations
  • Alex: individual taxon analyses
    • Generated genus level data
    • Will run GLM predictions separately for each phylum with at least 20 genera
  • James can run species-area analyses once we have final predictions
  • We probably will not include beta-diversity in this paper
    • Map doesn't show strong patterns - could be interesting result
    • But there are better ways to generate beta diversity maps so not confident enough to report it
  • Methods draft
  • Author list
    • Josh, ..., Jessica, Katie?
    • Every one will write an author statement and we'll discuss
• Next week: protein family database (Guillaume/Tom)