Blog Post 8: Visualizing Data

After I collected my first transect experimented with how best to organize it my data. I knew that I’d be analyzing it with a free program I found called JASP but needed to put it into a spreadsheet first, which wound up teaching me a lot about how those work and how scripts can automate things.

Part of my study was using tree type as a proxy for summer moisture. I looked up biogeoclimatic zones, found out how common each tree type was in them and how much precipitation they got in their driest month, weighting each zone by abundance and calculating a “tree value”. I could then have the spreadsheet import those values into where the trees were. My method was clumsy, but likely helped me avoid errors I may have made.

I wanted to look at my data and see if plant distribution demonstrated zonation. JASP gave me things that looked like this:

The light area around the line is .05 confidence. That’s useful to some extent, but hard to use if I want to see if want to compare plants for observing zonation, since if there’s an effect it’s going to be somewhat subtle. I was getting the distinct impression that I hadn’t found any significant effects.

I began experimenting with overlaying and applying colour to the graphs using a Paint.net (basically free Photoshop). Different colours stand in for different types of plants. The green is sword ferns and the blue is ivy, for example.

Aside from making it easier to share my results, I can also choose “layers” in Paint.net to make some plants visible and hide others. I thought I wouldn’t find any statistically significant effect  but I was wrong. In this graph I can tell that ivy becomes less likely to occur when the “Tree Value” (and driest month precipitation) was highest. This suggests that either ivy is drought tolerant, or it’s growing in places that other plants can’t and will die back.