In:
Agronomy Journal, Wiley, Vol. 89, No. 5 ( 1997-09), p. 723-729
Abstract:
Further improvement of residue decomposition models may help to optimize the use of N released from cover crop residues. CERES models are widely used to simulate crop‐soil systems and have a common subroutine that describes N dynamics (CERES‐N). In this work, we tested CERES‐N with results from a litterbag study over 120 d in which crimson clover ( Trifolium incarnatum ( L.), rye ( Secale cereale( L.), wheat ( Triticum aestivum( L.), and oat ( Avena sativa L.) residues were allowed to decompose on the soil surface. The field data were used to (i) evaluate the original CERES‐N and a modification of CERES‐N that allows users to enter actual sizes for the carbohydrates, cellulose, and lignin pools and (ii) determine the best‐fitting rate constants for decomposition of the carbohydrates and cellulose pools in the original CERES‐N and in CERES‐N with user‐supplied pool sizes. CERES‐N with the original rate constants or with rate constants and parameters proposed by other workers either underestimated (root mean square error RMSE = 1.23 g N m −2 ) or overestimated (RMSE = 2.18 g N m −2 ) N remaining in the crop residues after 120 d of decomposition. CERES‐N with user‐supplied pool sizes improved the simulation of N remaining after 120 d (RMSE = 0.69 g N m −2 ), but the simulated values generally tended to overestimate the field data. The best‐fitting rate constants for carbohydrates and cellulose pools were 0.67 and 0.021 d −1 , respectively, for the original CERES‐N (RMSE = 0.77 g N m −2 ), and 0.17 and 0.011 d −1 for CERES‐N with user‐supplied pool sizes (RMSE = 0.55 g N m −2 ). Future work should further evaluate these fitted rate constants under field conditions.
Type of Medium:
Online Resource
ISSN:
0002-1962
,
1435-0645
DOI:
10.2134/agronj1997.00021962008900050003x
Language:
English
Publisher:
Wiley
Publication Date:
1997
detail.hit.zdb_id:
1471598-3