Brian:
 
I'm a bit confused with your question.  The total N fixation amounts that you
give are quite normal for a good soybean crop (40-50 bu/ac) on a soil that is
fairly low in N-supplying capacity.  I think the "old book" maximum of 200
kg/ha is low, conditioned on lower, historical yields.  The only other
option, is that the modeled N uptake is low.  Actually, this is something I
have complained about for a year or two.  The problem is using the
Ritchie-CERES maize approach for N uptake as a function of root length
density.  That is fine for maize which has 3 to 4 fold greater root length,
but gives an underestimate of soil N uptake for soybean's lower root length.
We need soil N uptake to be driven partly by mass flow of water to the root.
This would reduce the simulated leaching a bit.
 
Any way, you are discussing two problems that you have not separated well.
Is the problem excess simulated N leaching from insufficient root N uptake?
Or is it excess simulated N leaching a year later because of N coming from
the soybean crop (that you think fixes too much N).  For a Millhopper fine
sand with good water supply, Bragg soybean accumulated a total of 320 kg N/ha
(210 in harvested seed, 35 kg N in standing stem and podwall, and 75 kg in
abscised leaf and petiole).  The total N accumulation came from 292 kg
N-fixed and 28 kg N uptake.  The model says 110 kg of N is left in the field
in the residue.  Some of this would decompose rapidly, providing an N credit
of about 50 kg N.  Is that the problem, too much left-over N for the next
crop?
 
As I see it, we do not want to make the model fix less N.  That is not the
problem (high yielding soybean crops require a lot of total N).  We could
have too much leaching: 1) during the season if root N uptake is too low, 2)
after harvest if the residue left is too high in N conc. (which is a possible
problem with the modeled N mobilization) and thus decomposes too fast.  I
need more information on what the problem is?
 
Regards, Ken Boote