If the subjects of your test remains the same it is pairing.
For example if you apply the pest. to the crop and measured "before" and "after" for each test subject (crop) then it is a paired test.
So, you should have exactly 2 data for each crop: before and after.
But my feeling is that yours should not be paired t-test. You got to give us more info about the mechanics of your test and data collection to rule that out.