Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Ben Raymond - Visualising the Atlas.pdf

Ben Raymond - Visualising the Atlas.pdf

Atlas of Living Australia

August 05, 2013
Tweet

More Decks by Atlas of Living Australia

Other Decks in Science

Transcript

  1. [email protected] What is this? some ideas on using ALA data

    in different ways minimal examples … not polished or complete … not necessarily even useful!
  2. [email protected] Get the data … • divide Australia into 2x1

    degree grid cells • extract Plantae species list for each cell: http://biocache.ala.org.au/ws/webportal/species.csv? q=plantae&wkt=POLYGON((120:-26,122:-26,122:- 25,120:-25,120:-26)) • species-level records only • calculate dissimilarities between pairs of grid cells – (assumptions about absences!) • calculate geographic distance between each grid cell pair
  3. [email protected] Process … • adjust the geographic positions of the

    grid cells so that their distances better match their dissimilarities • (like MDS, but with additional constraint of planar warping of positions)
  4. [email protected] Do the distortion … • nudge each grid point

    to make the geographic distances a better match to the compositional dissimilarities
  5. [email protected] Get the data … • extracted Charadriiformes (gulls, waders,

    shorebirds, etc) from spatial portal – 1.1 million obs; 118 species • for each species, compared winter and summer distributions looking for indications of migration (differences in latitudinal distributions)
  6. [email protected] Shortlist Actitis hypoleucos (Common sandpiper) Charadrius (Eupoda) veredus (Oriental

    plover) Gelochelidon nilotica (Gull-billed tern) Onychoprion fuscata (Sooty tern) Rostratula australis (Australian painted snipe) Stiltia isabella (Australian pratincole)
  7. [email protected] Stiltia isabella simulated migration • randomly sampled 200 points

    from each distribution using the Maxent habitat suitability as sampling weights • constructed plausible* tracks between summer and winter point-pairs • created an animated “artist's impression” of migration using Matlab and Processing
  8. [email protected] A bit of fun … • extract observations by

    "interest group" and compare distance to road – birds (Aves), beetles (Coleoptera), fish (Pisces), moths & butterflies (Lepidoptera), plants (Plantae), reptiles (Reptilia)
  9. [email protected] A bit of fun … Percentage of records within

    10km of a road Pisces: 53% Coleoptera: 72% Reptilia: 72% Plantae: 75% Aves: 78% Lepidoptera: 88%
  10. [email protected] A bit of fun … Conclusions? (Recognizing that this

    is a small and possibly unrepresentative data set!) → Fish do not swim on roads. → Either butterflies like roads, or lepidopterists are lazy.
  11. [email protected] The idea … • each taxon has an information

    page • many have a representative photograph • can we link occurrence records with photos to reconstruct spatial variations in colour?
  12. [email protected] The data • 2° longitude by 1° latitude grid

    cells • extract Grevillea species list in each via web service • get each species information page via web service • if it has a photo, get it
  13. [email protected] Assemble • for each grid cell, take sample of

    colours in from the corresponding species – only those that have photos (obviously) – colours in proportion to species prevalence, and prevalence of colours in species palettes – only for grid cells with at least 5 species