Modelling responses of perennial ryegrass pastures to future climate scenarios in Tasmania

Evidence for the warming of the climate system is considered unequivocal. Observations from all continents and most oceans show many natural systems are being disturbed by regional climate change (IPCC 2007). Agricultural production is strongly influenced by climate and so climate change will have profound effects on food production and the natural resources on which agriculture depends. In Tasmania, dairy is the largest agriculture industry contributing nearly 30% of the gross value of agricultural production. Production and consumption of high quality temperate pastures and the availability of irrigation underpin the competitiveness of the Tasmanian dairy industry. As a result, the Tasmanian dairy industry will need to adapt to any changes in climate that affect the productivity of temperate pastures or the availability of irrigation water. Therefore, there is an urgent need to evaluate climate change impacts on agricultural production in Tasmania. Although relatively small in area, Tasmania has a diverse range of regional climates that are the result of different synoptic influences. Consequently, climate change might be expected to vary in different regions of the state. The aim of this thesis is to quantify the regional impacts of projected climate change on the productivity of pasture based dairy systems for six regions within Tasmania. Six sites were selected to represent the dairying regions of the north west (Woolnorth and Flowerdale), central north (Merseylea), the northern Midlands (Cressy), the north east (Ringarooma) and the south (Ouse). Observed and interpolated historical climate data from four different climate data sources are available. A meteorological and biophysical comparison between the four sources of observed and interpolated gridded daily climate data for the period 1971 to 2007 was undertaken at each site. The data obtained were SILO Patched Point data (observed), SILO Data Drill (0.05°), Australian Water Availability Project (AWAP 0.05°) and Australian Water Availability Project (AWAP 0.1°). Differences were observed between the four daily climate data sources for each region for the climate variables mean daily minimum temperature, rainfall and potential evaporation. In addition, outputs from a biophysical model DairyMod (version 4.9.2) compared simulated monthly and annual pasture production. There were few significant differences in annual pasture production but there were significant differences for the months of October and November. Climate Futures for Tasmania (CFT) used CSIROs Conformal Cubic Atmospheric Model (CCAM) to dynamically downscale five General Circulation Models (GCMs) (ECHAM5/MPI-OM, GFDL-CM2.0, GFDL-CM2.1, MIROC3.2 (medres) and UKMO-HADCM3) reported in the IPCC fourth assessment report and a sixth GCM, CSIRO-Mk3.5. The CFT project provided projections under the A2 emissions scenario at a scale of 0.1° (≈10km) grid across Tasmania for the period 1961 to 2100. Climate models have variable skill and inherent biases. To manage these biases, a bias-adjustment method was undertaken to scale the climate modelling outputs to the historical interpolated data of AWAP 0.1. The aim was to preserve the change in the frequency distributions from the six GCMs projections. The suitability of the downscaled bias-adjusted GCM data for use in a biophysical model was evaluated by comparing simulated pasture yields from DairyMod from the downscaled GCMs with that from the AWAP 0.1° data source for the period 1990 to 2007. The mean annual simulated pasture yields from the downscaled bias-adjusted GCMs and from the AWAP 0.1° data source were generally comparable, giving confidence that the bias-adjusted GCM simulations were suitable for projections of pasture production. The projected climate for the A2 emissions scenario at each site was described over the period of 1971 to 2100. The mean of daily maximum and daily minimum temperatures at each site are projected to increase from a baseline period (1971 to 2000) to 2085 (2071 to 2100) ranging from 2.4°C at Woolnorth to 2.7°C at Ringarooma. Mean annual rainfall is projected to slightly increase at each site except Flowerdale, from the baseline period to 2085. Seasonal rainfall is projected to increase slightly at each site except Woolnorth (summer) and Flowerdale (summer and spring). Inter-annual rainfall variability is projected to increase for each region. Mean annual potential evapotranspiration rates as calculated by Morton‟s wet method (Bennett et al. 2010) are projected to increase by 3% to 4% for each region from the baseline period to 2085. No changes in solar radiation are projected for any of the regions. The Tasmanian dairy industry is a vital component of the state‟s agricultural sector. The industry is comprised of approximately 450 farms and produces approximately 7% of the national milk production. Regional variation in the projected climate is likely to result in regional variation in pasture production across Tasmania. The biophysical model DairyMod was used to simulate the growth of perennial ryegrass (Lolium perenne L.) under rainfed and irrigated conditions at each site. The simulations were nutrient non-limited, pasture was harvested monthly to a residual of 1400 kg /DM/ha, and identical soil physical and chemical parameters were used for each site. Throughout the 21st century mean annual pasture yields under rainfed conditions at each site are projected to increase above the baseline varying from a 29% increase at Woolnorth to 150% increase at Cressy by 2085. Mean annual irrigated pasture yields are projected to increase until mid-century, then decrease to 2085. The reduction of irrigated pasture yields in the latter half of this century is the result of an increase in the number of warm days with maximum temperatures exceeding 28°C. The impact of higher daily maximum temperatures are more evident under the irrigated simulations, because the irrigated simulations were not limited by soil moisture. In contrast, the major limitation to growth under the rainfed simulations during the summer months is inadequate soil moisture. Yield increases under both rainfed and irrigated simulations are projected to occur during late winter and spring. The increase in yield is a result of increases in both daily maximum and minimum temperatures and the progressive increase of atmospheric CO\(_2\) concentrations under the A2 scenario. Inter-annual yield variability for each region is projected to decrease throughout the 21\(^{st}\) century under both the rainfed and irrigated simulations, except at Cressy under irrigation. The decrease in inter-annual yield variability is being driven by a marked reduction in yield variability during autumn and spring. Irrigation demand progressively increases throughout the 21\(^{st}\) century for each region, except at Cressy and Ouse. Although irrigation requirements are projected to increase, there is a corresponding increase in the water use efficiency (WUE) of the pastures. This increase in WUE is driven by increasing atmospheric CO\(_2\) concentrations resulting in an increase in the net influx of CO\(_2\), increased stomatal resistance and a decline in transpiration per unit of CO\(_2\) fixed (Clark et al. 1995). Surface runoff was quantified at each site, using the downscaled bias-adjusted GCMs simulations within the hydrological model SIMHYD (Chiew et al. 2002) for the period 1971 to 2100 (Bennett et al. 2010). At each site, except Flowerdale, mean annual surface runoff at 2085 was projected to increase as a result of increased runoff during the winter months. A case study farm at Ringarooma was selected to quantify the projected surface runoff and river flow impacts on an irrigated perennial ryegrass based dairy system. Currently the farm is 166 ha in size with 75 ha of the farm supported by irrigation infrastructure. Irrigation water is accessed from both an on-farm catchment (dam) and access to River flows (Ringarooma River). Under the projected climate, hydrological modelling indicates that irrigation demand by 2085 will be met by the on-farm storage dam 50% of the time in comparison to the current baseline of 23%. In contrast, water accessed from the Ringarooma River during summer is projected to continually fail to meet irrigation demand from the baseline to 2085. Modelling the climate change impacts indicate a progressive increase in pasture growth within Tasmania throughout the 21\(^{st}\) century (particularly under rainfed conditions). The low cost of milk production associated with pasture based dairy systems underpins the national and international competitive advantage of the Tasmania dairy industry. This study has found the current forage base of Tasmanian dairy regions is resilient to future climate scenarios and that adaptations are likely to be within system adaptations, with the industry continuing to focus on milk production per hectare and pasture consumption per hectare as key determinants of business success. This will allow the Tasmanian dairy industry to retain its comparative advantage