V. O. Snow

An overview of APSIM, a model designed for farming systems simulation

The Agricultural Production Systems Simulator (APSIM) is a modular modelling framework that has been developed by the Agricultural Production Systems Research Unit in Australia. APSIM was developed to simulate biophysical process in farming systems, in particular where there is interest in the economic and ecological outcomes of management practice in the face of climatic risk. The paper outlines APSIMs structure and provides details of the concepts behind the different plant, soil and management modules. These modules include a diverse range of crops, pastures and trees, soil processes including water balance, N and P transformations, soil pH, erosion and a full range of management controls. Reports of APSIM testing in a diverse range of systems and environments are summarised. An example of model performance in a long-term cropping systems trial is provided. APSIM has been used in a broad range of applications, including support for on-farm decision making, farming systems design for production or resource management objectives, assessment of the value of seasonal climate forecasting, analysis of supply chain issues in agribusiness activities, development of waste management guidelines, risk assessment for government policy making and as a guide to research and education activity. An extensive citation list for these model testing and application studies is provided.

DairyMod and EcoMod: biophysical pasture-simulation models for Australia and New Zealand

DairyMod and EcoMod, which are biophysical pasture-simulation models for Australian and New Zealand grazing systems, are described. Each model has a common underlying biophysical structure, with the main differences being in their available management options. The third model in this group is the SGS Pasture Model, which has been previously described, and these models are referred to collectively as ‘the model’. The model includes modules for pasture growth and utilisation by grazing animals, water and nutrient dynamics, animal physiology and production and a range of options for pasture management, irrigation and fertiliser application. Up to 100 independent paddocks can be defined to represent spatial variation within a notional farm. Paddocks can have different soil types, nutrient status, pasture species, fertiliser and irrigation management, but are subject to the same weather. Management options include commonly used rotational grazing management strategies and continuous grazing with fixed or variable stock numbers. A cutting regime simulates calculation of seasonal pasture growth rates. The focus of the present paper is on recent developments to the management routines and nutrient dynamics, including organic matter, inorganic nutrients, leaching and gaseous nitrogen losses, and greenhouse gases. Some model applications are presented and the role of the model in research projects is discussed.

Simulating pasture growth rates in Australian and New Zealand grazing systems

DairyMod, EcoMod, and the SGS Pasture Model are mechanistic biophysical models developed to explore scenarios in grazing systems. The aim of this manuscript was to test the ability of the models to simulate net herbage accumulation rates of ryegrass-based pastures across a range of environments and pasture management systems in Australia and New Zealand. Measured monthly net herbage accumulation rate and accumulated yield data were collated from ten grazing system experiments at eight sites ranging from cool temperate to subtropical environments. The local climate, soil, pasture species, and management (N fertiliser, irrigation, and grazing or cutting pattern) were described in the model for each site, and net herbage accumulation rates modelled. The model adequately simulated the monthly net herbage accumulation rates across the range of environments, based on the summary statistics and observed patterns of seasonal growth, particularly when the variability in measured herbage accumulation rates was taken into account. Agreement between modelled and observed growth rates was more accurate and precise in temperate than in subtropical environments, and in winter and summer than in autumn and spring. Similarly, agreement between predicted and observed accumulated yields was more accurate than monthly net herbage accumulation. Different temperature parameters were used to describe the growth of perennial ryegrass cultivars and annual ryegrass; these differences were in line with observed growth patterns and breeding objectives. Results are discussed in the context of the difficulties in measuring pasture growth rates and model limitations.

Modelling the seasonal and geographical pattern of pasture production in New Zealand

The pasture growth module AgPasture was integrated into the APSIM (Agricultural Production System Simulator) simulation model, allowing pasture-based systems to be modelled in combination with other land uses at farm scale or within land use change studies. The models predictions of pasture growth were evaluated against 32 pasture growth datasets from a diverse range of soil types and climatic zones across New Zealand. The pasture herbage accumulation simulated by the model closely matched actual measurements over varying intervals. Both predicted and measured pasture growth rate demonstrated the same seasonal pattern, including mean growth rate and inter-annual variation across measurement years. Predicted and measured annual average net herbage accumulation (NHA) on a dryland pasture was similar over 37 observation years (mean, 6.83 and 7.27 t DM/ha respectively; coefficient of variation, 29% and 27% respectively) and highly correlated (R 2 = 0.838, P < 0.0001; relative root mean squared deviation (RMSD) = 16%). The models prediction of annual average NHA of all simulated pastures, spanning a wide range of pasture environments, also matched the measurement data well (R 2 = 0.777, P < 0.0001; relative RMSD = 21%). However, discrepancies between simulated and observed values occurred in some seasons and at some sites. Analysis of these discrepancies identified areas where the model could be improved by incorporating more accurate descriptions of the effects of plant development and grazing, soil temperature and the interactive effects of high temperature and soil moisture dynamics.

A review of literature on the land treatment of farm‐dairy effluent in New Zealand and its impact on water quality

Abstract Dairy farming is the largest agricultural industry in New Zealand, contributing 20% of export earnings but providing a challenge for the environmentally acceptable treatment of wastes from dairy farms. Nutrient‐rich farm‐dairy effluent (FDE), which consists of cattle excreta diluted with wash‐down water, is a by‐product of dairy cattle spending time in yards, feed‐pads, and the farm dairy. Traditionally, FDE has been treated in standard two‐pond systems and then discharged into a receiving fresh water stream. Changes brought about primarily due to the Resource Management Act 1991 have meant that most regional councils now prefer dairy farms to land treat their FDE. This allows the water and nutrients applied to land in FDE to be utilised by the soil‐plant system. Research on the effects of land‐treating FDE, and its affects on water quality, has shown that between 2 and 20% of the nitrogen (N) and phosphorus (P) applied in FDE is leached through the soil profile. In all studies, the measured concentration of N and P in drainage water was higher than the ecological limits considered likely to stimulate unwanted aquatic weed growth. Gaps in the current research have been identified with respect to the application of FDE to artificially drained soils, and the lack of research that has taken place with long term application of FDE to land and at appropriate farm scale with realistic rates of application. Whilst the land treatment of FDE represents a huge improvement on the loss of nutrients discharged to fresh water compared with standard two‐pond systems, there is room for improvement in the management of FDE land‐treatment systems. In particular, it is necessary to prevent the direct discharge of partially treated FDE by taking into account soil physical properties and soil moisture status. Scheduling effluent irrigations based on soil moisture deficits results in a considerable decrease in nutrient loss and may result in a zero loss of raw or partially treated effluent due to direct drainage.

Modelling pastoral farm agro‐ecosystems: A review

Abstract A pastoral farm is a complex agro‐ecosystem that produces agricultural outputs such as milk, meat and wool. Detailed consideration of the complexities of agro‐ecosystem components, and the interactions between these components, may be best approached with simulation modelling. We reviewed nine simulation models (APSIM, EcoMod, FASSET, GRAZPLAN, GPFARM, Hurley Pasture Model, IFSM, LINCFARM, and WFM) used primarily for research of grazing systems. GrassGro (one of the GRAZPLAN suite of decision support tools), LINCFARM, and WFM had particular strength in their simulation of animal production and farm management. The strengths of APSIM, EcoMod and FAS‐SET lie in their simulation of soil nutrient dynam ics, crop or pasture production and spatial variations in soil properties. The Hurley Pasture Model has particular capability to model nutrient cycling in the presence of climate change. The IFSM includes a comprehensive representation of machinery. GPFARM can be used to predict forage production of different plant functional groups. We suggest that more emphasis could be placed on including the effects of pests and diseases on pasture production and animal performance, more detailed representation of management practices, inclusion of more mechanistic voluntary feed intake and rumen processes and including the effect of specific genes and gene‐by‐environment interactions on plant quality and yields, nutrient use and animal performance. The Common Modelling Protocol, used in APSIM and the GRAZPLAN suite, may enable the development of a powerful and flexible pastoral agro‐ecosystem model which incorporates newly developed modules of the nature mentioned above.

The challenges - and some solutions - to process-based modelling of grazed agricultural systems

Pastoral systems are characterised by a number of features that are absent in arable cropping systems. These features include: (i) pastures are biologically diverse so interactions between plant species must be considered; (ii) economic return requires the inclusion of the animal as an additional trophic level; (iii) interaction between the grazing animal and the pasture is complex, influenced by the environment, plant species and animal behaviour and this creates feedbacks that can result in vicious cycles; (iv) animals spatially transfer substantial amounts of nutrients both randomly and systematically and this creates or exacerbates soil variability; and (v) whole farm management is both more complex and more important to system function in grazed compared to arable systems and it is harder to capture in simulation models. These challenges complicate the process-based modelling of pastoral systems and present significant obstacles to model developers and users.Here we discuss these challenges, describe the range of solutions used by different models and discuss the strengths and weaknesses of these solutions. We have placed particular emphasis on the analysis of a range of possible solutions with the point of view that diversity between and within models is important to provide the flexibility needed for future uses.We find that for most challenges there is a diversity of solutions incorporated into the models and that there is the potential to capture additional diversity, if needed, from other models. We note an apparent lack of development in the modelling of extreme events such as very high temperatures, systematic animal-mediated nutrient transfers, pests, weeds and gene-environment interactions in pastoral simulation models and suggest that these subject areas should receive more attention. We review the major challenges for simulating pastoral farming systems.Simple solutions can transfer the burden of conceptualisation from developer to user.The more complex solutions are probably not suitable for routine use.Modelling the effects of systematic nutrient transfers requires additional attention.Modelling of pests, diseases and gene-environment should receive more attention.

Solute transport in a layered field soil: Experiments and modelling using the convection-dispersion approach

Greater understanding of the processes affecting solute transport in field soils is required to meet the ever-increasing demand for improved management of field-applied chemicals. In this study, we sought to elucidate the effect of distinct interfaces between horizons of strongly-contrasting texture on solute transport. Field experiments of solute transport were performed on a soil consisting of three layers of different texture in which porous cup samplers had been installed at four depths in twenty sites. Similar experiments were done in a lysimeter of area 2 m2 and 1 m deep. Convection-dispersion modelling was first attempted using the lysimeter data. This was successful, provided that the surface 250 mm of soil were not used to calibrate the model coefficients. Layering within the profile appeared to have little effect on solute transport. The transport porosity was found to be just two-thirds of the water-filled porosity. However, convection-dispersion modelling of the field data was not particularly successful, probably due to the spatially-variable nature of solute transport coupled with variation in the water application pattern. Textural differences in the soil seemed to be overwhelmed by both small-scale heterogeneity of water application and local variation of solute movement through the soil, especially near to the soil surface. It appears that the hydraulic processes occurring in the surface soil require more attention by modellers of solute transport than they have been afforded in the past.

The single heterogeneous paddock approach to modelling the effects of urine patches on production and leaching in grazed pastures

Despite the fact that urine patches within grazed paddocks are the primary source of N leaching, virtually all pastoral simulation models assume a uniform spatial return of urinary-N to the soil. This simple spatial averaging might not be appropriate if the aim of the modelling is to explore leaching losses because of the non-linearity caused by the high N concentration in urine patches. Here we describe the single heterogeneous paddock (SHP) approach to modelling the dynamics of N in pastoral systems. We also examine the potential for manipulating rate parameters in a simpler uniform-return model (URM) to compensate for the lack of explicit description of urine patches. Comparison of simulation results from the URM and SHP showed some differences in the patterns of production and a substantial difference in leaching. Depending on soil and climate simulated, there was 5–30% higher pasture production in the URM because simulated leaching in the URM was 5–85% of that simulated by the SHP. Examination of the ratio of the outputs from the two models revealed that the differences in pasture production and N fixation in the URM could probably be corrected with a change in parameter values. This was not true of leaching where there was considerable variation and skew in the ratios, so at the very least, any correction factor would be highly soil and climate specific. We suggest that models of grazed grass–legume systems can probably adequately simulate production with a simple URM but that the simulation of leaching requires an explicit representation of the heterogeneous urine return. The SHP approach is one methodology for this but this has implications for model and software complexity and for model run-time duration.

Estimating nutrient loss to waterways - an overview of models of relevance to New Zealand pastoral farms.

Abstract Reliable estimation of nutrient losses from farmland is of increasing interest, driven by both economic and environmental concerns. Routine direct measurement of nutrient losses is currently impractical given the scale and variability of the problem. Simulation models are the best alternative and their use for assessing potential nutrient losses has been increasing worldwide. In New Zealand, there are a considerable number of models in use, or that are being developed, aiming to estimate N and P losses from pastoral fields. This range of alternative models reflects both the different level of detail and scale at which N and P losses can be estimated, and the diverse range of purposes assumed during the model development. Thus, it is important to understand the differences between models in order to select the one that will produce estimates appropriate to the intended use. This work presents an overview of the principal models for estimating nutrient loss being used or developed in New Zealand. It emphasises models that deal with N and P losses from pastoral farming systems, particularly via leaching, and that may allow the handling of different farm management procedures. Most of the models have gone through some testing and are supported by published works, although some are not fully operational yet and others need further evaluation. There is, in general, a lack of organised information about how several of these models work and what their main purposes are. We aim to supply some basic information about the available tools, sorting them into categories to highlight their primary differences and similarities. This is intended to assist discussions about model selection as well to highlight where information gaps about particular models need to be addressed.