Previous Page Table of Contents Next Page


BRIDGING THE RICE YIELD GAP IN AUSTRALIA - J. Lacy1, W. Clampett2 and J. Nagy3

1 NSW Agriculture, P.O. Box 108, Finley, NSW 2713, Australia.

2 NSW Agriculture, P.O. Box 1087, Griffith, NSW 2680, Australia.

3 NSW Agriculture, P.O. Box 108, Finley, NSW 2713, Australia.

1. INTRODUCTION

The Rice Industry in Australia produces 1.3 million tonnes from an area of 150, 000 hectares, of which 85 percent is exported. One crop of japonica rice is grown per year under temperate climatic conditions. All rice is irrigated as the growing season rainfall of 180 to 220 mm is very low. The rice is milled and marketed by the Ricegrowers’ Cooperative Limited, which is a farmer owned cooperative.

In 1986 the New South Wales (NSW) Department of Agriculture developed the objective crop management and collaborative learning extension programme “Ricecheck” aimed at improving farmer yields. Together with the development of semi-dwarf varieties, Ricecheck has resulted in a significant increase in farmer yields. In the 20 years prior to Ricecheck there was little yield improvement.

Ricecheck has led to great changes in rice management. One of the great changes is that farmers monitor and check crops to see how their crops compare to the measures for high yields. Before Ricecheck farmers rarely ventured into rice crops. Ricecheck is an ongoing extension programme. Components of the programme are the Ricecheck Recommendations for improving yields and grain quality, the Crop Evaluation Database and Farmer Discussion Groups. Ricecheck has created a learning culture. Ricecheck has also led to closer links between farmers, researchers, extension and commercial agronomists and the rice industry.

2. STATUS OF RICE CULTIVATION

2.1 Area, Production and Yield Trends

Most of Australia’s rice is grown in southern New South Wales. The production is 1.3 million tonnes from an area of 150,000 hectares. Yields have been increasing. Average yield for the last 5 years (1995- 99) was 8.4t/ha compared to 6.8t/ha for the 1985-89 period. During this period the first semi-dwarf rice variety M7 was released in 1983, while the current main variety Amaroo was released in 1987.

2.2 Production Constraints

The biggest production constraint is the supply of irrigation water. Rainfall runoff collected in dams in the catchments is subject to large seasonal variation. Irrigation water is released from dams on the Murrumbidgee and Murray Rivers and fed by channel gravity systems to farms. Farming industries are increasingly competing for water with the needs for the environment and maintenance of clean river flows. Another constraint is cold temperature at the early pollen microspore stage, which is a limiting factor in 6 to 8 years out of 10. Minimum temperatures in January average 16° C. The rise of water tables to within 1 to 2 metres of the ground surface with potential for increased salting is another production constraint.

2.3 Yield Potential of Released Varieties

The yield potential of existing japonica varieties is 15t/ha although a few research plots yield above 14t/ha. Temperate climatic conditions allow only one crop per year which is sown in October (mid Spring) and harvested in March (mid Autumn).

2.4 Evidence of Yield Gaps

Farmer yields range from 8 to 12 t/ha with few if any farmers able to repeat 12t/ha yields every year. There is little difference between the best research plot yields and best farmer yields.

3. PROGRAMME FOR NARROWING THE YIELD GAP

3.1 Historical Perspective

In 1985 the Rice Research Committee convened a meeting to address the lack of progress in increasing yields. There had been little yield improvement in the previous 20 years (Figure 1). The cost-price squeeze was the trigger for the meeting with costs rising and incomes stagnant. Farmer yields tended to fluctuate widely. Most farmers could obtain high yields of 10t/ha in odd years but could not consistently reproduce 10t/ha yields. There was recognition that there were big gaps between research and farmer yields and between “top” and “bottom” farmers. There was a perception that the transfer of technology diffusion model was failing to deliver improved yields. There was a need for a new approach or model.

Figure 1. Average NSW rice yields from 1968 to 1999

The new approach was based on finding the answers for the high yields in high yielding farmer paddocks rather than from research plots. It is called the Check Approach (Lacy 1998). It was based on what was really happening in farmer paddocks rather than on what was thought to be happening. The new approach was already having success for increasing irrigated wheat yields (Lacy 1991). The Rice Research Committee provided funding for the use of the approach for rice with the commencement of Ricecheck in 1986.

3.2 Activities and Results of programmes to Narrow Gaps during the Last Two Decades

3.2.1 Ricecheck

Ricecheck, the crop management and collaborative learning system based on crop checking, has been used in the New South Wales rice industry to improve yields and grain quality. The checking and measuring of high yielding farmer crops identified 7 key recommendations or checks linked to high yields. Most farmers were not adopting these key checks hence yields were not increasing.

The most important feature of Ricecheck is to encourage farmers to monitor and check their crops to see how the crops compare to the 7 key checks. When Ricecheck commenced in 1986 the yield goal for the medium grain semi-dwarf varieties was 10t/ha. The number of key checks has recently increased to 9. The checks are described simply and objectively. This reduces information overload and aids communication and understanding of the total package. Another check targets improved quality. Farmers strive to adopt the key checks in order to increase yields. As part of the learning process farmers are encouraged to record their management of the checks by completing crop record sheets. The current key checks are:

· Develop a good layout with a landformed, even grade between banks and well-constructed banks of a minimum height of 40 cm.

· Sow on time during the ideal window for each variety.

· Achieve 200-300 plants/m2 established in standing water to ensure uniform establishment over 100 percent of the crop area.

· Apply only registered or approved pesticides to control weeds and insect pests to prevent economic yield loss.

· Apply sufficient nitrogen to achieve the target range nitrogen uptake at panicle initiation (P.I.) so that the P.I. topdressing requirement does not exceed 60 kg N/ha.

· Topdress nitrogen based on fresh weight and NIR analysis using the NIR Tissue Test.

· Apply 10-25 kg P/ha pre-flood where the Colwell phosphorus is below 20 ppm.

· Achieve P.I. before 10 January.

· Achieve a minimum water depth of 20 to 25 cm during the early pollen microspore stage.

The most important feature of Ricecheck is to encourage farmers to monitor and check their crops. This is achieved through a number of learning steps. These are:

Observing

¯
Measuring
¯
Recording
¯
Interpreting
¯
Acting

The aim is to educate farmers to improve their learning and performance at each step as well as moving from step to step over time. To assist paddock measuring and recording simple measuring aids and records are provided e.g. rice rings. The benefit of recording crop data is that crop growth and management can be related to the yield and grain quality check benchmarks.

3.2.2 Database analysis

Between 500 to 600 paddock records are received from farmers each year on 2 record sheets. The first sheet serves two purposes. It satisfies the needs for the objective NIR Nitrogen Topdressing Tissue Test and also satisfies the Ricecheck need for information on sowing and establishment up to panicle initiation. A second data sheet “Crop Evaluation P.I. to Harvest Data” records yield and other key check information. Records are entered into a Visual dBase for Windows software programme which uses the same template each year so that new data can be automatically added. The total number of crops on the Ricecheck Crop Evaluation Database (Lacy, Clampett and Nagy 1999) is currently 3822 crops for the 1994 to 1999 harvest years. A Report SmithTM software programme is used to produce 5 types of reports of farmer results.

The farmer data is analyzed to show comparisons of management practices between crops and indicate how high yields were achieved. The records allow farmer management practices to be compared to the Ricecheck key checks or recommendations and to determine the adoption of the recommendations. Farmers receive individual Ricecheck Crop Evaluation Reports, which compares their management to high yields and shows how they can improve. Another benefit of the records is that poorly adopted recommendations or checks can be quickly identified providing timely signals to extension, research and the rice industry as a whole on issues requiring further investigation.

3.2.3 Discussion groups

Farmer discussion groups have played a key part in the delivery of Ricecheck. About 45 discussion groups are run by 7 extension agronomists. The momentum for the success of Ricecheck and the discussion groups is from the focus on the key checks. At the group meetings farmers are encouraged to collaborate and learn from each other and give feedback on the check recommendations. This also allows them to influence changes to the Ricecheck management package and develop ownership of the programme. Between 3 to 5 rounds of meetings are held over the rice-growing season.

3.2.4 Yield evaluation results

Key checks and yield

Ricecheck is based on the principle that as the adoption of the 7 Key Checks increase, yields increase. Figure 2 shows the relation between the number of Key Checks achieved and yield for the main medium grain variety Amaroo from 1994 to 1999 based on 1,834 crop records. The results confirm that adopting more checks results in higher yield.

Figure 2. Yield response to checks adopted 1994 - 1999

Top and average yields

The database report “Statistics on high and low yielding crops” shows the check performance for high, average and low yields for any variety. This is important for the development of the Ricecheck recommendations. The check performance for high yields can be compared to the existing recommendations. This provides the opportunity to modify existing check recommendations to enable improvements to management and targeting of higher yields.

Check data comparing average Amaroo variety yields with the top 10 percent yields are shown in Table 1 for the 1994 to 1999 harvests. The results show that the top yielding crops have better adoption of each of the 7 Key Checks. The top crops were sown a little earlier, achieved better plant numbers with fewer weak areas, had higher bank heights, better weed control and higher nitrogen levels.

Table 1. Comparison of Average Amaroo yields with top 10% yields 1994 - 1999

Check

Yield

% Adoption

Average

12 t/ha

Average

12t/ha

Banks

42

44

69

78

Sow date

15 Oct

11 Oct

74

87

Plant numbers

174

192

74

88

Weak areas

5%

1%

89

99

Plants + weak area



68

86

Weeds 0 - 0.5t/ha



18

55

Fresh weight

3103

3280

54

54

NIR

1.61

1.62

44

57

N Uptake

114

123

32

37

P.I. N

39

53

83

75

Total N

112

131



% N pre-flood

62

54



P

6

7

29

39

EPM

20

20

56

71

Laser



81

91

Av. yield t/ha

8.8

12.6




Single check yield comparisons

The database option “Analysis of Interactions” has the ability to compare any one of 60 major factors with yield and produce graphs of the results. Figure 3 is an example of one of the graphs comparing yield with sowing date for the Murrumbidgee Irrigation Area (MIA). The use of graphs at farmer meetings is an excellent tool for promoting discussion and farmer learning and motivating farmers to improve practices.

Figure 3. Effect of Sowing Date on Yield for MIA/CIA 1994-1999

3.2.5 Adoption of the checks

One of the aims of Ricecheck is to improve the adoption of the Key Checks since the higher the adoption the higher the yields. Figure 4 shows the trends in adoption for the variety Amaroo over the period 1994 to 1999. Overall adoption of the checks is good with the exception of the nitrogen uptake at P.I. and early pollen microspore water depth checks. On an individual year basis the 1998 harvest year had the best adoption for five of the checks. The checks where adoption has generally improved over the past 5 years are bank height, establishment plant number, weed control and recommended nitrogen topdressing rate. Bank heights have increased as farmers realize the importance of deep water at the early pollen microspore stage. Sowing rates have significantly increased to help improve establishment. More farmers are using the recommendations from the NIR Tissue Test to improve decisions for nitrogen topdressing.

Figure 4. Adoption of key checks 1995-1999

Adoption of the nitrogen uptake at P.I. check though poor has slightly improved. Sowing date has remained static. Late announcements of water allocations often prevent farmers sowing in the recommended check window. Weed control has been improving which is surprising given the increasing problems with resistance to bensulfuron methyl. The adoption of deep water at the cold sensitive early pollen microspore (EPM) stage has tended to fall. This is the result of two factors. High temperatures at the P.I. to EPM stage prevented farmers raising water levels and uncertain water allocations, particularly in the Murray Valley in the last 2 years, has resulted in many farmers running much lower water levels than normal.

3.2.5 Evaluation of farmer crops

Ricecheck is a collaborative learning system. An important aspect of adult learning is participation and feedback. As part of the learning process farmers like to compare themselves with others. The Ricecheck Crop Evaluation Report provides feedback to each participating farmer as to how their crop compared to the Ricecheck Key Checks and to other farmers. The Reports are produced by the computer programme for each growing district for each variety and for each farmer crop. The Reports are sent to farmers after harvest and after entry of all data sheet records and yield analysis. Appendix 1 shows an example of a Ricecheck Crop Evaluation Report 1999 for one crop from the Eastern Murray Valley for the medium grain variety Amaroo. In the Report the crop data is compared to the achievement of the Key Checks, to the average yields and the highest 10 percent yields. The Report shows the crop achieved 7 of the 7 Key Checks.

4. ISSUES AND CHALLENGES IN NARROWING THE RICE YIELD GAP

Figure 2 showed that the more checks adopted, the higher the yield. The Crop Evaluation Database results show that there is an average of 60 to 70 percent adoption of the key checks. This is creditable but the challenge is to improve this level of adoption to close the yield gap. The relative importance of the checks changes each year so it is important for farmers to try and adopt all checks. Often farmers may adopt most of the checks and only miss out on one or two. However, if the one or two non-adopted checks in any year are crucial to yields there can be a significant yield penalty. Invariably the top farmers get more of the checks right more of the time than the average farmers. Any small lift in the performance of the average yielding farmers could make a big difference to the yield gap. Extension programmes need to highlight this and to motivate farmers to improve.

There is poor adoption of the nitrogen uptake at P.I. check and lower adoption of the early pollen microspore (EPM) water depth check. These are barriers to increasing yields. A recent survey of EPM water depths supports the Ricecheck results. Years with significant cold at EPM result in significant yield differences between the top and average farmers. There is a need to investigate the barriers to farmers adopting both checks.

5. CONCLUSIONS AND RECOMMENDATIONS

5.1 Conclusions

Ricecheck has provided the framework for collaboration between farmers, researchers and extension officers. As a method of on-farm research it recognises that farmer learning and knowledge is just as important as research and extension knowledge.

Before Ricecheck, farmers used to pinpoint single factors as the main determinant of yield. Extension programmes focussed on single technologies. Ricecheck as a systems approach has demonstrated that there are many factors to get right for yields to increase.

Importantly Ricecheck has changed the culture and management of ricegrowing from managing from a distance, to walking in and checking the crop. Farmers learn by critically observing and measuring their crops. The Ricecheck Database reports and the individual Ricecheck Crop Evaluation Reports allow farmers to compare their management with high yields, and also assists learning and the bridging of the yield gap. The discussion groups have provided an ideal learning environment for extension delivery.

An independent evaluation of Ricecheck in 1997 based on random interviews with 124 farmers found that 83 percent of the farmers said Ricecheck was useful in producing higher yields. With the rating of the extension components discussion groups rated highly at 76 percent, the Crop Evaluation Reports next highest at 63 percent and Ricecheck Recommendation booklet at 54 percent.

Ricecheck will continue to develop and change to meet the demands of farmers and the rice industry.

5.2 Recommendations

· The close collaboration and teamwork between farmers, researchers, extension and commercial agronomists and industry needs to continue.

· The Ricecheck Recommendations need to be revised and improved each year through the incorporation of any new research results or new results from the Ricecheck Crop Evaluation Database.

· The discussion groups should be maintained using facilitation methods to keep meetings dynamic and to motivate both the top and average farmers.

· More farmers need to be encouraged to record crops because of the improved knowledge gained from the Crop Evaluation Reports.

· The Ricecheck Crop Evaluation Database includes a number of very high yielding 12t/ha and over crops which need to be analyzed to identify new key checks which can be incorporated into the Ricecheck recommendations and hence lift yield potential.

· There is a need to add crop grain quality to the database which will allow feedback and crop comparisons on grain quality and provide linkage to crop management. Quality assurance is becoming a significant issue relating to food safety and market access.

REFERENCES

Lacy, J. 1991. Finley Irrigated Wheat Five Tonne Club 1991. Technical Publication.

Lacy, J.1994. Ricecheck-A Collaborative Learning Approach for Increasing Productivity. In: Proceedings of the Temperate Rice Conference: 21-24 Feb. 1994: Leeton, New South Wales, Australia. p 247-254.

Lacy, J. 1998. Learning from Farmers - the Check Approach. Proceedings 9th Australian Agronomy Conference, Wagga Wagga, NSW, Australia p 58-65

Lacy, J., Clampett, W. and Nagy, J. 1999. Development and Use of a Crop Management Database to Evaluate Rice Crop Performance in NSW, Australia. In: Proceedings 2nd Temperate Rice Conference 13-17 June 1999: Sacramento, California, USA.

Ricecheck CROP EVALUATION REPORT 1999-AMAROO MVE
Report from the R&D Project “Performance Evaluation of Commercial Rice Crops”

GROWER NAME:

Address:

Farm Number:

Variety:

Field Name:

RCL Sample Number:

NB The results in this Report are based on information provided by Ricegrowers. The accuracy is that of the data provided.

Ricecheck

YOUR RESULTS

Average Results

Results for Highest 10% yield

% of all growers achieving Key Checks

MANAGEMENT AREA

CROP DATA

Achievement
of each
Key Check

YIELD TARGET
- minimum of
10 tonne/hectare

Yield
11.6 t/ha


Yes


10.0 t/ha


12 t/ha


52%

FIELD LAYOUT
Key Check 1 - “well constructed banks of a minimum height of 40cms”

Bank Height
42 cm


Yes


44 cm


43 cm


72%

SOWING TIME
Key check 2 - “aim to sow between

Sowing Date
14/10/1998


Yes


15 Oct


11 Oct


57%

CROP ESTABLISHMENT

Key Check 3 - “aim to establish 150 - 300 plants/m2

Seedling Number
225 m2


Yes


188


183


82%

with less than 5% weak areas”

Weak area
0 %


Yes


1%


0%


98%

CROP PROTECTION

Key check 4 - “prevent economic yield loss from weeds

Good Weed Control
Nil


Yes


98%


100%


98%

CROP NUTRITION

Key Check 5 - Pre-flood Nitrogen “apply sufficient nitrogen to achieve 2500-3700 g/m2 fresh weight and

Fresh weight
2956 g/m2
NIR


Yes


3677 g/m2


3220 g/m2


43%


1.56 %N

Yes




1.4 - 2.0 N%
to give

N Uptake

Yes

1.67 %N

1.92 %N

67%

90-130 kg N/ha Uptake at P.I.”

114 Kg N/ha


136 Kg N/ha

142
Kg N/ha

27%

Key Check 6 - P.I. Nitrogen “at P.I. topdress with nitrogen according to the NIR Tissue Test Recommendation

Deviation from NIR Test Recommendation
-2 Kg N/ha

Yes

7 Kg N/ha

-1 Kg N/ha

57%

WATER MANAGEMENT Key check 7 - “achieve a water depth of 20 - 25 cm during Early Pollen Microspore

E.P.M. Water Depth
20 cm


Yes


19 cm


20 cm


55%


You achieved 7 of the 7 recorded key checks (out of a total of 7) and the RIRDC Rice Research & Development Committee

NSW Agriculture


Previous Page Top of Page Next Page