Previous PageTable Of ContentsNext Page

Using soil water content to predict pasture growth rates

Brendan R. Cullen1 and Ian R. Johnson2

1 Melbourne School of Land and Environment, University of Melbourne. Email bcullen@unimelb.edu.au
2
Melbourne School of Land and Environment, University of Melbourne. Email ian@imj.com.au

Abstract

Managing climate variability, particularly variation in rainfall, is an important challenge for livestock producers in southern Australia. In this project the role of stored soil water in predicting future monthly pasture growth rates was explored, using a phalaris (Phalaris aquatica)-based pasture at Hamilton in south-west Victoria. The influence of low, medium or high soil water content (SWC) on the first day of each month on pasture growth rates over the following 12 months was simulated with the Sustainable Grazing Systems (SGS) Pasture model using daily climate data from 1901-2010. At Hamilton, moderate or high SWC at the beginning of March or April indicated above median pasture growth rates for the following two months. In spring, low SWC in August, September or October indicated a high probability of low pasture production over the following 3 months. These effects of SWC on pasture growth rates were more pronounced on soils with higher plant-available water-holding capacity. The results indicate that knowledge of SWC can give valuable information about future pasture growth in autumn and spring when growth rate variability is high. Further research is needed to assess the value of these pasture growth forecast in farm systems decision-making, and to investigate ways to make this information available to producers.

Key Words

grazing systems, seasonal forecasting

Introduction

Climate variability has a large impact on agricultural production, rural livelihoods and sustainability of grazing systems (Ash et al. 2007). Farm managers need to make appropriate management decisions that take climate variability into account, in order to both reduce losses in the poor years and make the most of the good years. Agricultural systems models, such as the Sustainable Grazing Systems (SGS) Pasture model, Grassgro and Agricultural Production Systems sIMulator (APSIM), are tools that allow the investigation of the interactions between climate, soil types and agricultural production, and can offer insights into the relationships between climate and farm management decisions. A good example of this is the ‘Yield Prophet’ tool for evaluating seasonal climate risk in cereal cropping systems (Hochman et al. 2009). The ‘Yield Prophet’ tool utilises knowledge of soil moisture and nitrogen status to improve decision making for grain growers, for example in deciding how much nitrogen to apply. There has been no comparable evaluation and tool development for the grazing industries of southern Australia. The aim of this project was to examine the potential for soil water content to predict pasture growth over subsequent months. In this paper the effect of low, mid and high SWC on the first day of each month on average daily pasture growth rates over subsequent months was investigated for a phalaris based pasture on three different soil types at Hamilton in south-west Victoria.

Methods

The simulation modelling approach developed in this study used all available climate data to calculate the pasture growth rate over the following 12 months in relation to different initial soil water contents (SWCs) on the first day of each month of the year. The site modelled was at Hamilton in south-west Victoria (-37.83, 142.06) which has a temperate climate with a long-term average annual rainfall of 694 mm. The simulations were run using the SGS Pasture model (Johnson et al. 2003, 2008) with SILO climate data (Jeffrey et al. 2001) from 1901 to 2010. Three different soil types were used, with descriptions provided in Table 1.

Table 1. Physical characteristics of the light, medium and heavy soil types used in the simulations.

Parameter

Light

Medium

Heavy

Saturated hydraulic conductivity (cm/d)

50

10

5

Bulk density (g/cm3)

1.6

1.4

1.2

Saturated water content (% vol)

40

47

55

Field capacity (% vol)

25

40

50

Wilting point (% vol)

10

20

30

Air dry water content (% vol)

5

15

25

Monthly pasture growth rates were simulated by cutting the pasture on the last day of each month to a residual of 1 t DM/ha. The cut yield at the end of the month was divided by the number of days to determine the monthly average pasture growth rate (kg DM/ha/day). The simulations were run with soil nutrients being non-limiting to plant growth.

The essential feature of the analysis was to use the long-term climate data to assess the effect of initial SWC on subsequent pasture growth. The approach was as follows:

  • Run all simulations for the period 1901 to 2010 to get baseline simulation data;
  • Initialise the SWC to characteristic high (H), medium (M) or low (L) values on the first day of each month;
  • Using each of these H, M, L sets of starting conditions for the first of each month, run the simulation with all of the climate data, but with the SWC re-initialised at the start of the month. Thus, for example, with the L starting conditions for SWC in May, the simulation started on 1/5/1902 and ran until 30/4/2010 with the SWC re-set to the L conditions on 1 May each year.

Thus, apart from the initial simulation from 1901 to 2010, there were corresponding simulations run for each month with the SWC initialised to H, M, or L on the first of each month for every year. This allowed the analysis to utilise the modelled pasture growth rate based on SWC at the start of each month using the full climate data set.

The approach used to define H, M and L SWC starting conditions was to prescribe the surface SWC in relation to field capacity, ӨFC, and wilting point, ӨW. For the H, M, L these were ӨFC, 0.5 * (ӨFC + ӨW) and ӨW respectively. It was assumed that the SWC was at field capacity at depth. The profiles are shown in Figure 1. It should be noted that these profiles may not always be realistic at particular times of year.

(a)

(b)

(c)

Figure 1. Initial SWC distribution for the (a) High, (b) Mid and (c) Low categories for the ‘medium’ soil type. Also shown are the saturated water content (Sat), field capacity (FC), wilting point (WP) and air-dry water (AD) content, as indicated in the legend.

Three factors were considered when assessing the projections for pasture growth rates based on SWC:

  • Impact on mean monthly pasture growth rate: do the SWC categories show an increase or decrease in the simulated pasture growth rate compared to the long-term average?
  • Accuracy of the prediction: in what percentage of years is the effect observed? This was assessed as the percentage of years when the simulated pasture growth rate for each of the SWC categories was greater than the long-term median. A threshold level of >70% accuracy in predicting pasture growth rate above or below the median was used to determine an accurate forecast in this study. This threshold was used because it is commonly cited as the minimum accuracy required for farmers to have confidence in a forecast (Ash et al. 2007).
  • Persistence of the prediction: how many months during the following year does the effect of increased or decreased pasture growth rate occur for?

Results

Effects of soil water content on predicted monthly pasture growth rates

The simulation results showed that SWC at the beginning of the month had an influence on future pasture growth rates during the autumn and spring seasons, but not at other times of year. If SWC was M or H at the beginning of March or April, above median pasture growth rates resulted for the following 2 months (Figure 2a,b). The H and M SWC categories in March reliably (>70% accuracy) predicted above average pasture growth rates for the following 3 and 2 months respectively. In spring, low SWC in September or October indicated a high probability of low pasture production over the following 3 months (Figure 2 c,d).

(a)

(b)

(c)

(d)

Figure 2. Effect of High (blue), Mid (purple) and Low (orange) SWC at the beginning of (a) March, (b) April, (c) September and (d) October on simulated average monthly pasture growth rate (kg DM/ha/day) compared to the long-term median growth rate (green; top panel), and the percentage of years in each SWC category predicted above the long-term median (bottom panel) for a phalaris-based pasture at Hamilton. The ‘medium’ soil type is used.

Effect of soil type on predicted monthly pasture growth rates

The impact of soil type on the simulations for phalaris at Hamilton for the heavy and light soil types are illustrated in Figure 3. For the light soil the growth rate was generally lower, being around 20 kg DM/ha/day in autumn for the H simulations, whereas it reached 40 kg DM/ha/day for the heavy soil. The growth rate for the medium soil fell between the two (Figure 2). In addition, the probabilities that growth exceeded the median in 70% of years lasted for 4 months for the medium and heavy soils whereas the response, while quite noticeable, was less apparent for the light soil (Figure 3).

(a)

(b)

(c)

(d)

Figure 3. Effect of High (blue), Mid (purple) and Low (orange) SWC at the beginning of (a) March, (b) April, (c) September and (d) October on simulated average monthly pasture growth rate (kg DM/ha/day) compared to the long-term median growth rate on a light (top panel) and heavy soil type (bottom panel) for a phalaris-based pasture at Hamilton. Note different scales on figures.

These results indicate that at times of year when pasture production is most variable in southern Australia, that is during the autumn and spring (see for example, Chapman et al. 2009), SWC can be used to predict future pasture growth rates. This information may be valuable to producers in planning farm activities over the following 2-3 months. For example, it may inform tactical decisions about stocking rate, time of selling stock and supplementary feed requirements.

In the winter and summer months of the year the SWC categories did not impact on the future pasture growth rate. At these times of year pasture growth rates are usually more reliable, due to cool/moist conditions in winter and hot/dry conditions in summer. The soil type also influences the pasture growth predictions, so it is important that this is also taken into account.

Conclusion

In autumn at Hamilton in south-west Victoria, moderate or high SWC at the beginning of March or April indicated above median pasture growth rates for the following 2 months. In spring, low SWC in September or October indicated a high probability of low pasture production over the following 3 months. The effects of SWC on pasture growth were more pronounced on soils with higher plant available water holding capacity. These results highlight that knowledge of soil water content can give valuable information about future pasture growth in autumn and spring when growth rate variability is high. The modelling approach developed can be used to investigate the effects of stored soil water on pasture growth rates across a broad range of sites, soils and pasture types. There is potential to develop this approach to provide pasture growth forecasts to producers to aid in the management of seasonal climate variability.

References

Ash A, McIntosh P, Cullen B, Carberry P and Stafford Smith M (2007). Constraints and opportunities in applying seasonal climate forecasts in rural industries. Australian Journal of Agricultural Research 58, 952-965.

Chapman DF, Cullen BR, Johnson IR and Beca D (2009). Interannual variation in pasture growth rate in Australian and New Zealand dairy regions and its consequences for systems management. Animal Production Science 49, 1071-1079.

Hochman Z, van Rees H, Carberry PS, Hunt JR, McCown RL, Gartmann A, Holzworth D, van Rees S, Dalgliesh NP, Long W, Peake AS, Poulton PL and McClelland T (2009). Re-inventing model-based decision support with Australian dryland farmers. 4. Yield Prophet helps farmers monitor and manage crops in a variable climate. Crop and Pasture Science 60, 1057-1070.

Jeffrey SJ, Carter JO, Moodie KM and Beswick AR (2001). Using spatial interpolation to construct a comprehensive archive of Australian climate data. Environmental Modelling and Software 16, 309-330.

Johnson IR, Chapman DF, Snow VO, Eckard RJ, Parsons AJ, Lambert MG and Cullen BR (2008). DairyMod and EcoMod: biophysical pastoral simulation models for Australia and New Zealand. Australia Journal of Experimental Agriculture 48, 621-631.

Johnson IR, Lodge GM and White RE (2003). The Sustainable Grazing Systems Pasture Model: description, philosophy and application to the SGS National Experiment. Australian Journal of Experimental Agriculture 43, 711-728.

Previous PageTop Of PageNext Page