Testing simulation programs at farm and regional scale in Tasmania

Summary

The objectives defined for this project were:
1. Use the available models, GrassGro, Sustainable Grazing Systems (SGS) and Stockpol to model 2 -3 scenarios for typical Tasmanian sheep/beef areas (lower midlands, North East coast, Cressy area) 
2. Use technical and industry reference groups to develop the inputs and evaluate the outputs. 
3. With input from the model owners test the response of different animal genotypes. 
4. Extend the modelling to express the benefits and constraints of potential new meat production systems at both the farm and regional scale; 
5. Identify strengths and shortcomings of the methodology. 
6. Build Tasmania's farm system modelling capacity and promote the value of modelling to a core group of industry representatives. 
A Technical Reference Group consisting of Department of Primary Industry and Water and Tasmanian Institute of Agricultural Research extension and research specialists was established to assist in developing data inputs for the models and to review the outputs. An Industry Reference Group consisting of producers, processors and consultants was also established to provide input into the areas to model and review the model outputs. 
 The two major areas suggested as needing research by the project Industry Reference Group were the use of irrigation and nitrogen fertiliser to increase pasture growth. This has arisen as a result of the increase in the area under centre pivot irrigation in the State and the need to include a pasture rotation in these cropped areas A number of alternative production systems were modelled to investigate the potential return from use of irrigation water in the systems, and an attempt was made to estimate the costs and returns of some of the enterprises using the program Red Sky. 
 Each of the models used had limitations in the process attempted and there were a range of issues which arose when using an output from one model as input for another. These were primarily because of the different model backgrounds. GrassGro is a dryland model with no facilities for irrigation, unless rainfall files are modified manually, but good animal growth predictions. SGS has the ability to apply nitrogen and irrigation in a range of different ways but has limited animal growth facilities. In order to use these models to examine the effects of nitrogen and irrigation on animal growth and production, animal growth was predicted from GrassGro and plant production from SGS and these data were used in Stockpol, a model which has user defined animal and plant production rates to predict production from a property. 


The major problems were  

Difficulties in defining actual fertiliser application rates except for nitrogen in SGS, and comparing fertiliser practices in the two models.  

Problems in predicting animal growth rates for pastures with varying nitrogen and irrigation additions because of the differences in plant production predictions between the models. 

Determining the area used for irrigation, and therefore the cost of irrigation when it was used to grow animals. 



The State wide areas of each of the soil types used in the simulations was calculated from land capability data to give an indication of the area which could be influenced by any potential changes. Unfortunately there is no current method of determining what proportion of those soils could be irrigated, so it is not possible to put a regional financial value on any potential changes to production systems. Given the constraints of the project it was not possible to develop any different animal genotypes which could be modelled. It is would involve some model rewriting. Members of the Industry Reference Group indicated that their prime interest was in predicting the pasture which could be produced under irrigation and differential nitrogen addition. Because of their varying property enterprises they felt that they were individually able to determine the value of that production in their systems. This project was able to generate the information needed.