Spatial summary: Aug. 2012–Jul. 2014

I have just completed two years of carefully tracking the specific locations at the Noble Research Center where I have found dead or trapped birds. To analyze those data statistically, I digitized the perimeter of the NRC in Google Earth and measured the apparent length of each segment (defining a segment as a straight line at least 2m in length). I identified 59 such segments for the NRC from a total apparent perimeter length of 847m. That’s right, the perimeter of that building is nearly 1 km long.

Next, I characterized segments according to whether or not they formed the terminus of one of the “alcoves”, which I defined as a recessed facade bordered by two segments of at least 4m oriented at right angles to the facade. So the question was, were more birds likely to be trapped and killed in alcoves than at random segments?

Predictably, that answer was “yes”: Birds were 49 times more likely to show up dead or trapped at alcove segments than at other segments along the 847m perimeter of the NRC.

The question now is “What’s the best way to represent these data graphically?” I like the idea of a map showing exactly where birds ended up – if you visit here regularly, you see those little red or blue dots overlain on a map of the building every time I find a bird there and report it here. Are those red and blue dots the best way to go for a two-year summary or does a numeric representation convey the information better? Please check out these images and use the comments feature to let me know which one you find to be more effective. Thanks!

Screen shot 2014-08-10 at 10.45.50 AM
Screen shot 2014-08-10 at 10.45.38 AM

3 thoughts on “Spatial summary: Aug. 2012–Jul. 2014

  1. I actually think I like the numbers better (which is weird because I usually prefer visuals). A few other questions, since I am not exactly sure what question your graphs are seeking to answer (beyond how many events at each microsite):

    – how is a bird “trapped” when not in an alcove?
    – have you thought about characterizing the relative sizes of the facades (in terms of linear feet)? Is that one top-facing stretch a bigger killer just because it is bigger?
    – does the compass orientation (north facing, etc) have an effect? Is north at top in the images?

  2. Good questions, TR. First yes, north is toward the top of the images. Seasonality appears to have little effect on where birds end up at this building. I get kills on south facades during fall migration and kills on north facades during spring migration.

    1) “Trapped” can include birds that are stunned from a collision or otherwise exhausted from flying back and forth trying to keep going in the direction that they cannot. On the big south side – which is not an alcove – there’s actually a very high area of metal rafters, and I’ve had some birds fly up into those rather than out and away from the building. I once watched a bird get confused up in there, come down, and turn right into the windows where it collided and died. There are a lot of things at play more complex that my simple dead or trapped dichotomy.

    2) No I haven’t compared mortality among alcoves yet. The surface area of the facade probably plays a role in making that north-facing facade so problematic, but so might its depth. That’s another problem with these long-term, real-world studies with an n = 1: Not only is the building unique, its various components are unique.

  3. I like the red and blue. The thing is; there seems to be a lot of overlap with some of the larger circles. The numbers make it easier to see and keep separate, but you don’t get the visualization of where they were found as well. So if you are trying to convey where they were, “you want to find a bird go here”, use the blue and red. If you want to convey how many you will find use the numbers.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s