It’s turtles all the way down!

I created this map as part of the #SlowMapChallenge proposed by Daniel Wood — an alternative to the #30DayMapChallenge where instead of making a map a day, you tackle one map idea you’ve been putting off over the course of 12 weeks.

I decided to work on this novelty map of habitat ranges for map turtles, which is a genus known for the topo-reminiscent patterns on their shells. Each section of the turtle’s shell shows the range of one of the 14 species of map turtle found in North America.

The map is, essentially, a giant visual pun—but it’s also meant to raise awareness for a fragile genus of turtles that face a lot of pressure from habitat degradation and the exotic pet trade. The layout is designed to be eye-catching, and sacrifices a bit of practicality to do so, but in prioritizing the theme I hope it makes the map more memorable and keeps readers around long enough to glean some of the deeper information it holds. 

The “small multiple” maps were made using ArcGIS Pro, but all the graphics, text and layout were done using Inkscape.

This is the most I had used Inkscape for any one map, up to this point. Previously, I’d been using ArcGIS Pro’s layout tools for most tasks, occasionally importing an svg graphic from Inkscape, but for this one I’d estimate more than half the work was done in Inkscape itself—mostly just to see what the workflow would be like. The giant turtle took a long time, getting the surrounding shell and appendages to line up with the small multiples and to look like they were part of the rest of the shell without being too distracting.

Ironically, I think the maps themselves turned out to be the weakest aspect of this, even though that part of the design was what kept bringing me back to the idea. The intent of the hex bins for habitat ranges was to match up with the hexagonal shell pattern, for a sort of repeating, ‘turtles-all-the-way-down’ effect. But in hindsight, I think they’re too tiny (even when printed at the intended poster size). When I experimented with larger bins, the habitat range data became coarse and misleading. I also tried using multiple scales for the small multiples, depending on how far the species was dispersed, but it ruined the overall pattern effect and made it difficult to tell what region was being shown. All that aside, I’m pretty content with how the rest of it came out. The background especially, which is a Turing pattern that was generated via a series of blurs and filters, before getting vectorized. It’s meant to mimic the skin patterns many map turtles have, such as Barbour’s map turtle.

Although now that I write that, I can’t unsee a bizarre Ed Gein-esque turtle nightmare.

The map won 1st place in the Most Innovative category at the 2022 ESRI User Conference, and also has a few hidden turtledoms scattered for those who look closely.