Estimating nitrogen release from Brassicacatch crop residues: Comparison of different approaches within the APSIM model

TY - JOUR

T1 - Estimating nitrogen release from Brassicacatch crop residues

T2 - Comparison of different approaches within the APSIM model

AU - Vogeler, Iris

AU - Cichota, Rogerio

AU - Thomsen, Ingrid K.

AU - Bruun, Sander

AU - Jensen, Lars Stoumann

AU - Pullens, Johannes W.M.

PY - 2019/12

Y1 - 2019/12

N2 - Determining the nitrogen (N) release from catch crop residues is important for N management in arable farming systems. The use of catch crops and subsequent incorporation of its residues represents a nutrient provision for the following crop and has the potential of reducing N leaching. To investigate whether the Agricultural Production Simulator (APSIM) can be used to predict the N release from Brassica catch crop residues with different carbon to nitrogen ratios (C:N) under different temperatures the model was set-up using three different approaches: (i) the generic approach, in which the residue is distributed to the various carbon pools with default APSIM parameters (Gen), (ii) the generic approach but with partitioning of the various pools based on the biochemical composition (Gen-BC), and (iii) the inclusion of an additional carbon pool in which mineralisation is based on first order decay and the C:N of the crop residue, termed simple mineralisation model (SMM). The SMM was parameterized based on results from various previous incubation studies. APSIM was then used to predict N release through mineralisation from these experiments. The simulations indicate that the generic approach in APSIM greatly underestimates the amount of N released from the cover crop residues, especially at low temperatures; the model efficiency (based on the Nash Sutcliffe efficiency score) was negative for the incubation study done at 2 °C. APSIM linked with the SMM shows a substantially higher prediction accuracy. The Gen-BC, which could only be done for one of the incubation experiments, had a higher prediction accuracy than the generic approach, but was not as good as the SMM. This indicates that, for accurate predictions of N mineralisation from Brassica catch crop residues, the generic approach with default partitioning needs to be modified, either by assigning various carbon pools based on the biochemical composition or by including an additional carbon pool.

AB - Determining the nitrogen (N) release from catch crop residues is important for N management in arable farming systems. The use of catch crops and subsequent incorporation of its residues represents a nutrient provision for the following crop and has the potential of reducing N leaching. To investigate whether the Agricultural Production Simulator (APSIM) can be used to predict the N release from Brassica catch crop residues with different carbon to nitrogen ratios (C:N) under different temperatures the model was set-up using three different approaches: (i) the generic approach, in which the residue is distributed to the various carbon pools with default APSIM parameters (Gen), (ii) the generic approach but with partitioning of the various pools based on the biochemical composition (Gen-BC), and (iii) the inclusion of an additional carbon pool in which mineralisation is based on first order decay and the C:N of the crop residue, termed simple mineralisation model (SMM). The SMM was parameterized based on results from various previous incubation studies. APSIM was then used to predict N release through mineralisation from these experiments. The simulations indicate that the generic approach in APSIM greatly underestimates the amount of N released from the cover crop residues, especially at low temperatures; the model efficiency (based on the Nash Sutcliffe efficiency score) was negative for the incubation study done at 2 °C. APSIM linked with the SMM shows a substantially higher prediction accuracy. The Gen-BC, which could only be done for one of the incubation experiments, had a higher prediction accuracy than the generic approach, but was not as good as the SMM. This indicates that, for accurate predictions of N mineralisation from Brassica catch crop residues, the generic approach with default partitioning needs to be modified, either by assigning various carbon pools based on the biochemical composition or by including an additional carbon pool.

U2 - 10.1016/j.still.2019.104358

DO - 10.1016/j.still.2019.104358

M3 - Journal article

SN - 0167-1987

VL - 195

SP - 1

EP - 10

JO - Soil & Tillage Research

JF - Soil & Tillage Research

M1 - 104358

ER -