Simulating grazing beef and sheep systems

Abstract

Context: Ruminant livestock make an important contribution to global food security by converting feed that is unsuitable for human consumption into high value food protein, demand for which is currently increasing at an unprecedented rate because of increasing global population and income levels. Factors affecting production efficiency, product quality, and consumer acceptability, such as animal fertility, health and welfare, will ultimately define the sustainability of ruminant production systems. These more complex systems can be developed and analysed by using models that can predict system responses to environment and management.

Objective: We present a framework that dynamically models, using a process-based and mechanistic approach, animal and grass growth, nutrient cycling and water redistribution in a soil profile taking into account the effects of animal genotype, climate, feed quality and quantity on livestock production, greenhouse gas emissions, water use and quality, and nutrient cycling in a grazing system.

Methods: A component to estimate ruminant animal growth was developed and integrated with the existing components of the SPACSYS model. Intake of herbage and/or concentrates and partitioning of the energy and protein contained in consumed herbage and/or concentrates were simulated in the component. Simulated animal growth was validated using liveweight data from over 200 finishing beef cattle and 900 lambs collected from the North Wyke Farm Platform (NWFP) in southwest England, UK, between 2011 and 2018. Annual nitrous oxide (N₂O), ammonia, methane and carbon dioxide emissions from individual fields were simulated based on previous validated parameters.

Results and conclusions: A series of statistical indicators demonstrated that the model could simulate liveweight gain of beef cattle and lamb. Simulated nitrogen (N) cycling estimated N input of 190 to 260 kg ha^-1, of which 37-61% was removed from the fields either as silage or animal intake, 15-26% was lost through surface runoff or lateral drainage and 1.14% was emitted to the atmosphere as N₂O. About 13% of the manure N applied to the NWFP and excreta N deposited at grazing was lost via ammonia volatilisation.

Significance: The extended model has the potential to investigate the responses of the system on and consequences of a range of agronomic management and grazing strategies. However, modelling of multi-species swards needs to be validated including the dynamics of individual species in the swards, preferential selection by grazing animals and the impact on animal growth and nutrient flows.

Keywords: Grazing; Liveweight; Modelling; North Wyke Farm Platform; SPACSYS.

Graphical abstract

Fig. 1

Extension of the SPACSYS model, component linkages, inputs and outputs. Solid lines show the components and linkages included in the latest version of the model. Dashed lines indicate components and flows for future inclusion.

Fig. 2

A schematic representation of the factors limiting intake and the metabolisable energy and protein system. MEI: metabolisable energy intake (MJ head⁻¹ day⁻¹); FME: fermentable metabolisable energy (MJ head⁻¹ day⁻¹); CP: crude protein (g head⁻¹ day⁻¹); QDP: quickly degradable protein content (g head⁻¹ day⁻¹); SDP: slowly degradable protein content (g head⁻¹ day⁻¹); UDP: undegradable dietary protein content (g head⁻¹ day⁻¹); ERDP: effective rumen degradable protein content (g head⁻¹ day⁻¹); MCP: microbial crude protein supply (g head⁻¹ day⁻¹); MTP: microbial true protein (g head⁻¹ day⁻¹); and MTP: true protein content of MCP (g head⁻¹ day⁻¹).

Fig. 3

Map of the North Wyke Farm Platform (NWFP), showing the permanent pasture (PP) farmlet, sub-catchments, fields, soil class, topography and the locations of flume outlets where water and nutrient fluxes are measured. The soil moisture and rain gauge in Top Borrows is situated within the North Wyke Met station.

Fig. 4

Comparison of measured and simulated animal liveweight by age for given breeds (BRBX, CHX, HEX, LIMX, SMX, ST and STX for cattle; CHA, LLE and SUFMU for lambs), where n is number of observations. Solid circles are measured average liveweight and open circles are simulated average liveweight. Solid and dotted lines are fitted polynomial functions for measured and simulated average liveweight against age, respectively.

Fig. 4

Comparison of measured and simulated animal liveweight by age for given breeds (BRBX, CHX, HEX, LIMX, SMX, ST and STX for cattle; CHA, LLE and SUFMU for lambs), where n is number of observations. Solid circles are measured average liveweight and open circles are simulated average liveweight. Solid and dotted lines are fitted polynomial functions for measured and simulated average liveweight against age, respectively.

Fig. 4

Comparison of measured and simulated animal liveweight by age for given breeds (BRBX, CHX, HEX, LIMX, SMX, ST and STX for cattle; CHA, LLE and SUFMU for lambs), where n is number of observations. Solid circles are measured average liveweight and open circles are simulated average liveweight. Solid and dotted lines are fitted polynomial functions for measured and simulated average liveweight against age, respectively.

Fig. 4

Comparison of measured and simulated animal liveweight by age for given breeds (BRBX, CHX, HEX, LIMX, SMX, ST and STX for cattle; CHA, LLE and SUFMU for lambs), where n is number of observations. Solid circles are measured average liveweight and open circles are simulated average liveweight. Solid and dotted lines are fitted polynomial functions for measured and simulated average liveweight against age, respectively.