Well...I like math.

That's a lot of treatments. How many replications each? This is where the raw data would be handy -- to get a sense of variance, for starters.

After seeing past data posted here, what would interest me most would be multiple reps of 0# N versus 0# N + GL. That would be the clearest way to demonstrate that GL does, in fact, provide N to the crop -- since no one has apparently figured out a way to test GL water for N.