Development of a lucerne model in APSIM next generation: 1 phenology and morphology of genotypes with different fall dormancies

Abstract

Prediction of lucerne phenological and morphological development is important for optimising the defoliation schedule and time of other management events. A challenge for any lucerne phenology module is to capture the seasonality of development processes in response to environment, management and genotype. To date, lucerne phenological modules have not been evaluated under different defoliation regimes or with genotypes of different fall dormancy (FD) classes. This research integrated data of lucerne phenological development into the Agricultural Production Systems sIMulator (APSIM) next generation (APSIM NextGen) model framework to develop and verify a phenology module. Relationships derived from the FD5 genotype, grown under a 42 day (LL) defoliation treatment were used for model development. These were further tested for two genotypes with contrasting FD (FD2 and FD10) under frequent (28 day: S) or long (84 day: H) defoliation regimes, all under irrigated conditions. Development was parameterized based on thermal time targets to reach specific phenological stages and modified by photoperiod responses. Development stage and node appearance were shown to be independent of defoliation treatment and FD class. Simulation results showed good agreement for prediction of development stages (NSE of 0.77 for days to buds visible and 0.67 for days to flowering stage) and number of main stem nodes (NSE values were ranged from 0.53 to 0.84). However, both defoliation management treatment and FD classes affected stem height. For FD5, there was good agreement for the 84 day treatment (NSE of 0.83) and the 42 day treatment (NSE of 0.66), but it was poor for the 28 day treatment (NSE of -0.08). This was probably due to reduced stem extension rates, limited by low C and N reserves in perennial organs under the frequent (28 day) defoliation regime. For FD2 and FD10, two separate sets of parameters were used to improve model prediction of height to account for their contrasting seasonal C partitioning patterns. These results show that the APSIM NextGen lucerne phenology module was able to simulate crops grown under unconstrained growing conditions. However, the reason for under estimation of stem height for the 28 day treatment needs further investigation.