From what I understand in your description, you have 40 treatments x 2 Blocks, with each treatment level being represented exactly once per block.

If so, I think it is amenable to the RCBD ANOVA. Alternatively, if your treatments can be expressed as a continuous variable (e.g., fertilizer level 1, 10, 100, etc.), you can fit a linear regression model to test the effect of treatment while controlling for the block.