3 min read

Lists of plant growth forms of Wisconsin plants

Daijiang Li · 2014/03/17

When I analyze the vegetation data, there is a good chance that I want to split plants into different growth forms, e.g. trees, shrubs. However, I do not have a list of them to do the match in R. I searched online and did not find any list available for analysis. The only thing I found is from the University of Wisconsin - Steve point’ s herbarium website. They have an Identification guides. But when I click on, say, Trees link, they do not have a txt file ready to download, and a txt file table is all what I want!

As a result, I decide to create the txt table according to their website, with the help of Sublime text. Here is the procedure:

  1. Open the website, right click, choose view page source.
  2. Copy the page source into Sublime text, then we can see the species names are wrapped by SpCode= and </A>.
  3. In sublime text, Ctrl+I, turn on regular expression search (Alt+R), type SpCode=.+</A, then press Alt+Enter. All matches were selected! Cool! Then copy Ctrl+C all of them into a new file (Ctrl+N) in Sublime text!
  4. Not done yet. In the new file, Ctrl+I, then type SpCo.+> first, then press Alt+Enter, then press Delete. We are half way there, we search </A, Alt+Enter, then press Delete. Awesome! We got the list of that page!
  5. Repeat this for each page… Fortunately, they are not too many for trees and shrubs! Or copy and paste all page source into one file and then follow the step 2-4, for each growth habit.

I put the list of tree, shrub, vine, fern/fern allies, and graminoids online. Downlaod them if you need.

For each species, if you just want the genus and sp, not including the subsp/var info. You can read them in R and then using regular expression to extract them. Or you can do this in Sublime text. Here is my way doing in R:

library(stringr)
tree=read.table("data/tree.txt", sep="\n", stringsAsFactors=F, quote="")$V1
tree=unlist(str_extract_all(tree, "^[A-Za-z]+\\s{1,1}X?\\s?[a-z]+"))
tree=unique(tree)

In the regular expression part, ^ means at the begining of the string; [A-Za-z] means any letters; + means more than one times; \\s means any space, {1,1} means exactly one time; ? means one time or not present. If you do this in Sublime text, replace \\s with \s.

I hope this will be helpful if you are doing community ecological analysis.

You can do all of these in R easily. I did not write the R code to extract the information at the beginning since there are only several pages and I can do them by hand. If you look at the page source, you will find that the pages’ web address only differ with the number in the web address. As a result, you can get the web address for all of these pages, then read them in R using readLines(). Then extract the names you want by regular expressions.