Mining eBird Data – Checklists by County

Using eBird data from 2010 through 2012 — 1) which County in Oregon has submitted the most checklists?

2) Given the unequal distribution of the state’s population — which County in Oregon has submitted the most checklists per capita?

What variables are in play that would determine these metrics?  Can we assume that birder density per capita is relatively constant from county to county?  Can we assume the use of eBird will be a constant within the birding community once growth flattens out?

Possible set of variables that contribute to the number of checklists submitted in each County:

  1. Number of eBirders in the County
  2. eBirder interest in specific locations within a County
  3. Accessibility of the County
  4. Proximity of the County to large population centers
  5. Interest within a County by “hyper-active” eBirders
  6. others?

Well the answer to the first question is Benton County with 9,878 checklists submitted over the three years — averaging 63 checklists a week.  Benton county?  It has a population of 85 thousand.  The Portland Metro Tri-county Area by contrast has 735 thousand in Multnomah, 530 thousand in Washington and 376 thousand in Clackamas counties.  That’s 8.6, 6.2, and 4.4 times as many people respectively than Benton County.

OK, the William L. Finley National Wildlife Refuge is in Benton County – the crown jewel in the Willamette Valley’s system of NWR’s.  There’s almost certainly more going on here than Finley – variable number 5.  The single top contributor from each of the three years in Benton County submitted 50% of those 9,878 lists.  In this case it happens to be the same individual.   In contrast, the top five contributors combined only make up 37% of the lists submitted in Multnomah County.  Eight different individuals had one of the fifteen available top 5 slots over the three years.

The answer to the second question is, unsurprisingly, Harney with  338 lists submitted per 1000 residents – variable number 2.  Harney County has more than twice the lists per capita than any other county.  However, somewhat surprising was the second highest county with 166 lists per capita — Sherman.  Here another factor may be at work.  There are only 1700 residents in Sherman County.  It doesn’t take many checklists to bump this number up quite quickly.  Obviously the lists per capita is not a very meaningful metric.

What’s this all mean — i have no idea, other than we can plot the relative density of eBird usage by county.  Anyway —

I was loathe to subject anyone to another set of awful “heat maps” showing checklist data by county so i found a work around in ‘R’.  A free statistical software package; the ggplot2 library in particular.  Roughly one month later the following set of choropleths (yeah, not heat maps, who knew?) have been generated.  And just for reference, one with the county names layered over total specie count.  I’ll probably update the earlier choropleths with these cleaner ones.