What we do Don is to analyze several years of corn yield maps to find the areas that consistantly yield higher. Map Calc is the program we use although in the beginning just looked at the maps and drew lines manually. We use a base population rate that we consider the "proper" rate for our area and yield levels - this happens to be a little higher that most farmers will plant in this area but not all. From there we will vary the population upwards for the consistantly higher yielding area. If you have areas that consistantly yield lower you could lower the population in that area but in our case, depending on the year, these areas that yield lower some years do very well if we recieve timely rains so for us the seed savings in the tough years will be much more than lost in the better years from reduced yields - that's a management decision and will vary from area to area, field to field.

We have a long term average of about 130 bpa. We have a lot of fields that have a 4 corn crop year yield average that in spots exceed 175 bpa (we have a yield layer in Map Calc that is a map of the "average" within a 50' grid with the gridws covering the whole field). Before VR seed we were managing inputs on all acres for 120-140 bpa yield levels. It just made sense to me to manage acres that consistantly yield above 175 differently. I think Becky is putting in the 5th corn year (usually corn/bean fotation) in for some fields this year so we have a pretty good handle on where the "better" part of the fields are.

Before we had enough corn data we tried "normalizing" wheat and bean data to use in the equasion and that might work for some areas but here beans are in a totally seperate "season" than corn and corn and bean yields did not always correlate well. Wheat didn't correlate to anything at all. We gave up on using anything other than corn data for corn decisions.

We have been doing it long enough now that I will make a variable rate map for population on a single years corn data "if" the data was collected in a year that was conducive to showing where the higher water holding areas are (for us read "topsoil depth" in our clay pan soils). This last year was a very good example of the kind of stress year that I would use and in fact it was on a lot of newly rented ground without yield histories and I will use this one year of data to make variable rate maps for the next corn planted there. Most self respecting consultants or experts would never recomend that - that's my management decision. In fact I will use last years double crop bean yields with only one year data on the land to direct corn VR for next year even though I just told you bean and corn yields don't correlate very well for us "some of the time". My management decision is that the timing of last years moisture stress on the dc beans (in some fields) appears to correlate well with the moisture holding capacity of the field. On a single years experience I will not vary things a lot - just tweek the population up and the N up a bit where there is better moisture holding capacity.

With fields with several years of data we make up a formula and let the computer through Map Calc spit out the populations. It could be argued that increment populations 2000 at a time is too fine of an increment to be of much value and too fine of distinction to determine. I would not disagree with that but I see no harm in it and it is just an easier method to let the computer do the work - that may not be too clear. Ask if you have questions.

John