Dear all,
I would like to run a simulation on monoculture of maize, the soil was set as following:
(defhorizon "Ap F.S.L." FAO3
(description "0-30 cm")
(dry_bulk_density 1.32 [g/cm^3])
(clay 0.087)
(silt 0.278)
(sand 0.635)
(humus 0.032)
(SOM_fractions 0.30 0.15 0.55)
(C_per_N 11.0 [g C/g N]))
(defhorizon "B F.S.L." FAO3
(description "30-60 cm")
(dry_bulk_density 1.40 [g/cm^3])
(clay 0.135)
(silt 0.144)
(sand 0.721)
(humus 0.025)
(SOM_fractions 0.30 0.15 0.55)
(C_per_N 11.0 [g C/g N]))
(defhorizon "C F.S.L." FAO3
(description "60-90 cm")
(dry_bulk_density 1.40 [g/cm^3])
(clay 0.145)
(silt 0.156)
(sand 0.699)
(humus 0.024)
(SOM_fractions 0.30 0.15 0.55)
(C_per_N 11.0 [g C/g N]))
The target silage maize yield is 13 Mg/ha DW, 135 kg N/ha.
However, when I tried to run it, either with or without fertilizer at 105 kg N/ha, the soil mineral N after harvest was always much lower (around 5-10 kg) than my mearuements (100 kg N/ha). Even the N mineralization reached 200 kg N/ha and the leaching+denitrification was less than 30 kg N/ha at the that moment.
I am assuming that the N allocated to SMB was too high (250 kg N/ha). Does it make sense? How can I reduce the allocation of N to SMB?
I would appreciate any help or suggestions from you.
Thanks in advance!
kind regards,
Hongzhen