## Info

Source: Adapted from Hayman and de Vries, 1995.

Source: Adapted from Hayman and de Vries, 1995.

Comparing this with an actual yield of 2.8 tonnes/ha:

2800 kg/ha * [(266 mm + 59 mm) - 110 mm] = 13 kg/ha/mm (10.2)

Calculate actual yield as a percentage of its potential:

(2.8 tonnes/ha -f- 3.2 tonnes/ha) x 100 = 88% (10.3)

Note: Calculations allow for a loss of 110 mm by direct evaporation and assume runoff and deep drainage over the growing period to be nil.

The previous comparison identifies that water use efficiency is 88 percent of the potential. This benchmarking exercise provides a general guide for farmers wanting to assess and improve their water use efficiency. Cornish and colleagues (1998) found that wheat farmers who had positive yield trends and showed higher productivity used rainfall more efficiently and more carefully managed soil fertility.

Making Tactical Adjustments—Farming to Season Type

Examining the climate record for many regions of Australia shows that there is no such thing as the typical or average season. The highly variable climate produces dramatic variations in crop yields from one season to the next. As a result, there can be large fluctuations in farm income between years. Egan and Hammer (1995) state that in some regions, the best three years in ten can generate up to 70 to 80 percent of income, while the poorest three years may result in a net loss of income.

The most critical decisions are made at sowing time, as growers commit farm resources (land, labor, machinery, finances, etc.) for the following season and beyond, with only limited opportunities for further modifications (Egan and Hammer, 1995). Farming to season type therefore requires a willingness to be flexible in production decisions, such as stocking rate, area to be cropped, and the level of inputs to be used (e.g., fertilizer), in order to reduce the risk to physical and financial resources and to maximize opportunities. These decisions should be based on climatic and soil conditions prior to sowing and on seasonal forecasts, which provide useful indicators of "season type" and yield prospects.

For example, Allen Lymn, a farmer in South Australia, has developed a risk management strategy to minimize the effects of climate on farming in his low-rainfall area. If his farm near Minnipa does not receive 40 mm of rain between April 1 and June 15, Lymn cuts back his cropping area and will even consider not sowing. In years when his farm receives between 40

and 100 mm, he sows an average area—about half the farm. In years with over 100 mm falling in this period, he increases the cropping area. By adjusting his cropping program depending on early-season rains, Lymn has increased the opportunity of gaining higher returns over a number of seasons.

Farm managers need to have strategies in place for both drier and wetter than "average" seasons. For a dryland cropping enterprise this may mean being prepared to alter the area sown, variety choice, time of sowing, and the amount of fertilizer used. The crop management options and strategies available to farmers depend on the region being farmed. Stored soil moisture is a critical factor in cropping decisions in northern grain-growing regions but is less significant for southern cropping areas, which depend more on growing-season rainfall and timing of the seasonal break (Egan and Hammer, 1995). For example, early research in northwest NSW into the relationship between wheat yields, time of seeding, and soil moisture demonstrated that as the depth of wet soil at planting increased, so did wheat yields, in an almost straight-line relationship (Figure 10.5). More recent research on the Liverpool Plains in NSW on the links between nitrogen fertilizer, climate forecasts, and stored soil water confirms this relationship (Hayman and Turpin, 1998). Although this relationship exists, there may be advantages in using, rather than storing, this soil moisture for long periods.

In the summer rainfall areas of Queensland and NSW, flexible cropping systems that adjust cropping in response to stored soil water give economic benefits. As rainfall increases, cropping frequency can also increase. Build-

FIGURE 10.5. A generalized pattern of wheat yields in northern parts of Australian grain belt as determined by time of sowing and stored soil moisture (Source: Fawcett, 1968.)

ing flexibility into the system gives more than direct economic benefits. At the same time, soil erosion and soil salinity will be reduced. Storing summer rainfall in the soil profile by means of a long (up to six months) fallow will give high yields in the following wheat crop, but if the seasonal outlook is favorable, double or opportunity cropping may give higher returns. Recommendations for the northern cropping zones now consistently advocate the use of farming systems based on opportunity cropping rather than fixed rotations. In light of the threat in many of these areas from dryland salinity caused by rising water tables, this recommendation not only makes good economic sense but also has important environmental implications.

Using Forecasts and the Southern Oscillation Index in Decision Making

A way of managing rainfall variability is to examine the rainfall probabilities of a given location. These show the chances of receiving a particular amount of rain at a given time. Probabilities are like odds. If the chance of something happening is one in four, scientists will express it as a percent-age—a 25 percent probability. Seasonal climate forecasts produced by the Bureau of Meteorology are usually given in terms of probabilities that can be linked to a property's rainfall history. Farmers will point out that within most decades about three years out of ten are poor years, four are average years, and three are good years. The probability of a poor rainfall season is 30 percent, an average season 40 percent, and a good season 30 percent. However, the Southern Oscillation Index shifts the probabilities. It works this way: Think of a wheel with equal segments (Figure 10.6, top left). The wheel is spun with an equal chance of landing on any segment. Now assume that the wheel has three segments, which represent a dry, average, or wetter than normal season. There is only one spin of the wheel per year. For eastern and northern Australia, in years when the SOI is very high, the probability shifts toward the wetter season category. This means there is a greater chance of landing on a good season when spinning the wheel. However, there is still a chance of landing on a poor season. For eastern Australia, when the SOI is very low, the probability shifts toward the dry category. When the wheel is spun this time there is a greater chance of landing on the "dry" segment. The odds never shift to give absolute certainty of a dry or wet year. We may not be able to obtain certainty, but we can obtain better chances, and "half a loaf is better than no bread." Managing climate risk is a process of assessing possibilities and turning them into probabilities.

For example, Stuart and Maxine Armitage (personal communication) farm an irrigated cropping property on the Darling Downs at Cecil Plains.

FIGURE 10.6. Probability of occurrence of average, good, and bad seasons (Source: Hayman and Pollock, 2000.)

They use climate probabilities and forecasts, taking into account the SOI, to determine for the season their enterprise mix, the area of crop to sow, and amounts of water needed for irrigation scheduling. Recently the Armitages had their irrigation dam half full and, faced with a potentially dry season, had some major decisions to make. The time was September and the Southern Oscillation Index was strongly negative in an El Niño year. Little rain had fallen during the winter, and there was no subsoil moisture. Using the computer software program Australian RAINMAN (Clewett et al., 1999), they found there was only a 20 percent chance of the 50 mm needed as planting moisture. The Armitages used the information to minimize their risk by making the following decisions. Rather than gamble on receiving planting rain, they used the water stored as prewatering to germinate and establish the cotton crop. They also reduced the crop area, because of the expected dry season and the reduced amount of water expected to be available for irri gation. By using the seasonal forecast and an assessment of water supply, the Armitages made a decision, which proved to be the right one for the season.

Seasonal forecasts can also help growers decide when to double crop. If, for example, the SOI is higher than 10 in May to June and some subsoil moisture is present in the profile, there is a good chance of a reasonable crop. If, on the other hand, the SOI is negative and subsoil moisture is low, the best decision is likely to be to fallow and not plant a crop (Hammer, Holzworth, and Stone, 1996; Greggery, 2000).

The following points should be considered when incorporating El Nino-Southern Oscillation events into seasonal decision making:

• Autumn is when to expect major changes in the movement and value of the SOI. The SOI usually becomes set as consistently negative, positive, or neutral by the end of May, and this phase can be used to indicate rainfall patterns over the next nine months, that is, until the start of the next autumn.

• A strong rise (to greater than +5) in the SOI in autumn indicates the probability of at least average rainfall for the following winter, spring, and summer.

• Positive (greater than +5) SOI readings in autumn months usually indicate the probability of above-average rainfall.

• Negative (lower than -5) SOI readings in autumn months usually indicate the probability of below-average rainfall.

• Look at the trend or phase of the SOI as well as individual numbers.

• Before using ENSO information in farm decision making, producers should be aware that (1) ENSO effects vary greatly between regions and with the time of year, (2) decisions are based on probabilities and not certainty, and (3) ENSO is only one of the factors controlling the climate.

The SOI is not a perfect forecasting tool. One of the limitations of forecasts is that they do not always provide enough lead time for farmers to prepare and adjust their winter cropping operations (Nicholls, 2000). However, this does not apply for summer cropping.

Climate forecasts and information only acquire value when decisions are modified in response to them (Hammer, 2000). However, overreaction to forecasts (e.g., selling large portions of a herd or not planting a crop) can be as detrimental to a farming system as ignoring or dismissing a forecast completely (Nicholls, 2000; Stafford Smith et al., 2000). For example, the Australian Farm Journal (February 1998, p. 10) reported about a farmer selling young sheep for \$14 a head because of a decline in the SOI, only to find a short time later that they were worth \$40 a head.

Planned and measured responses to seasonal forecasts should be applied with the understanding that forecasts are simply one more element to include in decision making. Examining the local climate record and using seasonal forecasts helps producers to bring realistic expectations into their decision making and increase their confidence when selecting management options (Clewett et al., 2000). The continued use of seasonal forecasts and the development of farming systems that respond to erratic rainfall and stored soil moisture have been shown to be superior to farming according to the calendar in Australia (Hammer, Nicholls, and Mitchell, 2000; Hayman, Cox, and Huda, 1996).

### Sources of Weather and Climate Information

Climate information is delivered through a range of media, which are mainly sourced from national and regional meteorological centers. In Australia, the official source of climate information is the Bureau of Meteorology. The media range includes the Internet, facsimile services, television, radio, telephone services, newspapers, journals, and computer software packages. The media provide only generalized information, for a mainly "suburban" population. In farming, weather and climate information need to be more specific and related to current enterprises. Bayley (2000) provides a comprehensive review of the national weather and climate information and services available.

The most useful information for decision making will depend on location and the enterprises being operated. Table 9.1 (Chapter 9) summarizes some of the major agricultural enterprises, the key decisions to be made, and the climatic information required in decision making.