Constructing Boxplots

For this exercise, you will be using the snapping turtle dataset, collected at the Cabin Creek wetlands in western Kentucky courtesy of hard-working biology students Emily Dustman and Amy Krzton-Presson. Dr. Chris Mecklin did not help in anything invovling handling snapping turtles.

require(mosaic)
snapping <- read.csv("http://campus.murraystate.edu/academic/faculty/cmecklin/snappingturtle.csv")
head(snapping)

##   Site  CL  CW  CH  PL
## 1  NPC 394 337 160 281
## 2  NPC 369 305 157 279
## 3  NPC 294 233 120 237
## 4  NPC 351 274 144 254
## 5  NPC 283 234 120 208
## 6  NPC 269 351 118 225

What are the measurements?

• CL=Carapace length (mm)
• CH=Carapace height (mm)
• CW=Carpace width (mm)
• PL=Plastron length (mm)

What are the 3 sites?

• EX=experimental site (i.e. where the herbicide was applied to kill the invasive species, a water reed known as Phragmites)
• PC=Phragmites control (i.e. no herbicide was applied)
• NPC=non-Phragmites control (i.e. a nearby stream that does not have the invasive species)

While eventually we might want to do a MANOVA, for today we'll just construct some wonderful boxplots, and maybe error-bar plots if that can be done in the mosaic package.

These are boxplots that I think are wonderful but some of my colleagues in biology and ecology do not like for reasons that escape me.

bwplot(CL ~ Site, data = snapping)

bwplot(CH ~ Site, data = snapping)

bwplot(CW ~ Site, data = snapping)

bwplot(PL ~ Site, data = snapping)

I don't think we can do error-bar plots with a function in mosaic.