Modelling feedlots using the MEDLI model framework

Abstract

Model for Effluent Disposal using Land Irrigation (MEDLI) is a biophysically-based daily time-step model released in 1996 to facilitate designing effluent irrigation schemes. The model simulates a waste stream generator producing effluent that is treated in a pond system with a wet weather storage pond from which the effluent is irrigated as required to an area of land growing vegetation (Gardner et al. 1996). To complement the existing waste stream generator options, MEDLI is undergoing further development to include rainfall-dependent waste streams, including that generated by rainfall wash-off from feedlot production pens. This will facilitate MEDLI's use for designing effluent irrigation schemes associated with feedlots. The feedlot pen model attempts to model the complex dynamic processes within feedlot production pens that impact on the quantity and quality of runoff using a daily time-step mass balance approach. An early description of the feedlot model for MEDLI, focusing on runoff quantity, was provided by Atzeni et al. (2001). Since then, the hydrology component has been substantially improved to generate daily surface and sub-surface pad moisture output for use in predicting odour emissions (Atzeni et al. 2015), as well as runoff quantity and quality. In this paper, we present the modelling approach and model algorithms used to simulate the waste stream from the feedlot production pens. Supporting references are detailed in Atzeni et al. (2015). The MEDLI feedlot pen model is designed to simulate a modern feedlot yard with equal-sized production pens having adequate slope, and operating within the recommended Australian guidelines. Cattle can be designated to up to four markets, with market-specific entry and exit weights, daily weight gain, proportion of total herd designated, and proportion of pens occupied. Daily calculations are performed on a pen by pen basis, to model the key processes of herd dynamics, manure (faeces+urine) production, assimilation of the fresh manure into the pad, pen hydrology and pen cleaning. Herd dynamics include modelling animal mortality and pen stocking. When animals in a pen reach the exit weight for their market type, the model flags that the pen is vacant and drafts another mob (of the same market type) into another vacant pen if possible, or else the same pen. Manure production relies on BEEFBAL (QPIF 2004) or similar model to provide the market-specific annual manure production (total solids, volatile solids, total nitrogen, total phosphorus, salts and water) of each animal which is then used to determine the solids, nutrient, salt and water loading onto the manure pad. Assimilation of the fresh manure into the pad uses a two-layer model for the manure pad, assuming no loss of water or solids below the lower layer of the pad. The two layers capture the dynamics of pad hydrology and composition, including the impacts of rainfall, evaporation, animal stocking, manure accumulation, volatile solids decay, pen cleaning, runoff and manure erosion during runoff. Pens are cleaned at intervals to remove the excess manure, and involve considering the specified minimum number of days since a pen is cleaned, the pen's pad moisture content, pad depth, and the number of pens being cleaned each day. By modelling these processes, the fate of the nutrients, salts and solids from the manure pads is simulated as shown in Figure 1.Validation of the feedlot pen model hydrology was undertaken using four field-collected data sets from three South East Queensland feedlots. The prediction of runoff quantity appears closely correlated with measured data. However, the runoff quality predictions require calibration of the total nitrogen, total phosphorus, and salt runoff concentrations with actual or expected holding pond chemistry. Data collection is in progress to allow further testing and validation of the feedlot pen module. Copyright © 2019 The Modelling and Simulation Society of Australia and New Zealand Inc. All rights reserved.