Question 1062667
The way I would do it is to make it a paired t-test, since he is using the two different fertilizers on the same parcel of land.  This removes a lot of variability. The test is appropriate here rather than adding up both groups, taking the mean and the sd and doing a 2-tailed t-test with df=28.  Here, the df=14, a lot less, but the test would be looking at the average difference between the two divided by the sd/ sqrt (n).  It is a far more powerful test.

Here, the d bar is 0.28, the sd is 0.22, and the t value is 4.84, with a p-value <0.0003.