Preliminary results of parameterisation of DairyMod pasture model for tropical pasture; Brachiaria

Abstract

Exploring improved tropical forages is considered to be an important approach in delivering quality and consistent feed options in tropical and even subtropical regions under changing climate scenarios. Pasture modelling has been an effective tool in simulating pasture growth and obviating expensive field research under a range of soil, climate, and management strategies. Many models lack parameters for tropical pasture species, hindering their use in tropical regions. This study aims to adapt the generic parameters in the DairyMod pasture model to parameterise and evaluate the model for the tropical pasture species Brachiaria ruziziensis x B. decumbens x B. brizantha ‘Brachiaria Mulato II’ (BM). Data were collected from plots of BM established at the Gatton Research Dairy (27° 54´S, 152°33`E), Queensland, Australia from 19 November 2020 to 06 May 2021 to parameterise the model. Model evaluation was performed through various statistical indices for accuracy and precision. Canopy structure and carbon partitioning, photosynthesis and respiration, senescence and leaf appearance parameters were mainly modified from the generic C4 grass parameters in the model. Results showed that, species specific parameters setup for BM in the model simulated the total above ground yield (R2=0.92), leaf (R2=0.97), and LAI (R2=0.93) at a reasonable accuracy. Stem production also ranged under acceptable level except for the second defoliation due to decreased cutting height. Despite these reasonable simulated results, the model tended to underestimate stem production. Reasons could be higher variation of residual weight across the seasons and model failure to explicitly capture the plant physiological changes like anthesis, accelerated growth rate and increased stem production associated with tropical pasture phenological developments. Results suggest that the developed BM parameters in DairyMod need further testing under range of locations and seasons to improve the model.