|
Colby, Kansas | Spike,
Do you have any yield data available, as others have mentioned I think that is a clearer starting point to look at before picking a soil sampling religion.
My personal preference is unaligned grid sampling, where the sampling point can be located randomly anywhere inside the grid cell. This provides a random distance between points and provides better data for use with interpolation and other statistical methods. That being said, I've also used some targeted sampling based on yield history, these points are often used as benchmarks to track soil fertility levels over time.
My problem with any type of zone management is that in order to buy into the zone philosophy you have to make the assumption that everything within a zone is homogeneous. We all know for a fact that is not the truth, sure I can make my yield within zones look the same by tweaking with the legend colors, but the fact is yield values, soil fertility, etc do not operate with very defined lines in most cases (as they are dealt with by zones) rather they are continuous in nature. Thats one thing that just amazes me about zone prescriptions, that somehow there is something magical about this zone boundary line where seeding rate or N or whatever jumps to another rate. Very few things change like that in the field in reality, they gradually go up and gradually go down.
If you believe your yield follows your soil type very closely then sampling by soil type may be an appropriate strategy for your operation. Its certainly better than taking no sample or just one sample per field. From what I've seen here and other places in Neb. and Kans. there is often something else going on with yield other than soil type, making soil type of marginal value. I only care about soil type if its affecting yield.
Just some more to help muddy the waters. I've enjoyed reading everyone's thoughts on this.
Lucas | |
|
grid sampleing
