Increasing Productivity and Water Use Efficiency in Australia’s Rice Industry through Nitrogen Management RIRDC Publication No. 09/173
RIRDC
Innovation for rural Australia
Increasing Productivity and Water Use Efficiency in Australia's Rice Industry through Nitrogen Management
by Ranjith Subasinghe and John Angus
November 2009 RIRDC Publication No 09/173 RIRDC Project N. DAN 207A
© 2009 Rural Industries Research and Development Corporation. All rights reserved. ISBN 1 74151 970 5 ISSN 1440-6845 Increasing Productivity and Water Use Efficiency in Australia's Rice Industry through Nitrogen Management Publication No. 09/173 Project No. DAN 207A The information contained in this publication is intended for general use to assist public knowledge and discussion and to help improve the development of sustainable regions. You must not rely on any information contained in this publication without taking specialist advice relevant to your particular circumstances. While reasonable care has been taken in preparing this publication to ensure that information is true and correct, the Commonwealth of Australia gives no assurance as to the accuracy of any information in this publication. The Commonwealth of Australia, the Rural Industries Research and Development Corporation (RIRDC), the authors or contributors expressly disclaim, to the maximum extent permitted by law, all responsibility and liability to any person, arising directly or indirectly from any act or omission, or for any consequences of any such act or omission, made in reliance on the contents of this publication, whether or not caused by any negligence on the part of the Commonwealth of Australia, RIRDC, the authors or contributors. The Commonwealth of Australia does not necessarily endorse the views in this publication. This publication is copyright. Apart from any use as permitted under the Copyright Act 1968, all other rights are reserved. However, wide dissemination is encouraged. Requests and inquiries concerning reproduction and rights should be addressed to the RIRDC Publications Manager on phone 02 6271 4165. Researchers Contact Details Ranjith Subasinghe Yanco Agricultural Institute, Department of Primary Industries, Yanco, NSW 2703 Phone: 02 69512679 Fax: 02 69557580 Email:
[email protected] In submitting this report, the researchers have agreed to RIRDC publishing this material in its edited form. RIRDC Contact Details Rural Industries Research and Development Corporation Level 2, 15 National Circuit BARTON ACT 2600 PO Box 4776 KINGSTON ACT 2604 Phone: Fax: Email: Web:
02 6271 4100 02 6271 4199
[email protected]. http://www.rirdc.gov.au
Electronically published by RIRDC in November 2009 Print-on-demand by Union Offset Printing, Canberra at www.rirdc.gov.au or phone 1300 634 313
ii
Foreword Nitrogen management is one of the critical issues for sustainable and profitable rice production in NSW. This is not just because nitrogen status of the crop is a key determinant of yield potential, but also because high nitrogen status is associated with extreme cold sensitivity. Rice farmers pay a large yield and profit penalty for under fertilising their crops, but also can increase their risk of cold damage by over fertilising. As new varieties are developed and released, the nitrogen and low temperature response of these varieties need to be understood and described. Furthermore, nitrogen requirement of a crop is site specific and time dependent. Constant upgrading of fertiliser recommendations is therefore necessary, especially for a crop like rice because the properties of rice growing soils change enormously with time due to the intensity of cultivation. To enable rice growers to manage the nitrogen fertiliser application and risk involved in it, a range of technologies are available. These include the nitrogen tissue test at panicle initiation (PI), soil testing and the computer based decision support system “maNage rice”. maNage rice, a user friendly computer based decision support system allows growers to experiment with a range of options for their crop and determine the best management regime for their particular crop. The primary aim of this project was to provide information to the NSW rice industry on how to manage nitrogen fertilisation based on field and glasshouse experiment results, so that the maximum benefit of new varieties can be quickly obtained on farm. This project has successfully improved the maNage rice and released several enhanced versions to rice growers. Findings of this project are the key to make new recommendations for nitrogen management under different water management systems. The recommended nitrogen management options published in the NSW DPI Ricecheck was considerably altered in the light of findings of this project. Many rice growers successfully adopted these recommendations during 2005-06 and 2006-07 rice seasons. This report provides the detail information of the findings of the project and will benefit the rice industry to further improve the productivity and water use efficiency through proper nitrogen management. This report was funded from RIRDC Core Funds provided by the Australian Government. This report, an addition to RIRDC’s diverse range of over 1900 research publications, forms part of our Rice Research and Development Program, which aims to improve the profitability and sustainability of the Australian rice industry through the organisation, funding and management of a research, development and extension program that is both market and stakeholder driven.
Most of RIRDC’s publications are available for viewing, downloading or purchasing online at www.rirdc.gov.au. Purchases can also be made by phoning 1300 634 313.
Peter O’Brien Managing Director Rural Industries Research and Development Corporation iii
Acknowledgments Authors thank the Rural Industries Research and Development Corporation, NSW Department of Primary Industries and CSIRO, Australia for financial and in-kind support. Authors also wish to express gratitude to Messrs Alexander Suladze, Dionne Wornes, Tina Dunn, Kathryn Bechaz and John Smith for their valuable contributions and support. Authors are grateful to rice growers and Rice Research Australia who provided their valuable resources to conduct the trials in their farms. Authors also wish to extend their gratitude to Dr. Laurie Lewin for his invaluable suggestions throughout this study.
Abbreviations N:
nitrogen
PI:
panicle initiation
Pre-flood:
before permanent water and sowing
PFN:
pre-flood nitrogen application
PIN:
nitrogen application at panicle initiation
iv
Contents Foreword ............................................................................................................................................... iii Acknowledgments................................................................................................................................. iv Abbreviations........................................................................................................................................ iv Executive Summary ............................................................................................................................ vii Introduction ........................................................................................................................................... 1 Objectives ............................................................................................................................................... 3 Methodology .......................................................................................................................................... 4 Environmental conditions at experimental locations .........................................................................4 Common cultural practices adopted...................................................................................................4 Trials 2002-03....................................................................................................................................4 Trials 2003-04....................................................................................................................................5 Trials 2004-05....................................................................................................................................5 Trials 2005-06....................................................................................................................................6 Trials 2006-07....................................................................................................................................7 Sample collection and analysis ..........................................................................................................7 Statistical analysis..............................................................................................................................8 Results and Discussions......................................................................................................................... 9 Continued development of maNage rice............................................................................................9 Trials 2002-03..................................................................................................................................11 Trials 2003-04..................................................................................................................................17 Trials 2004-05..................................................................................................................................25 Trials 2005-06..................................................................................................................................29 Trials 2006-07..................................................................................................................................34 Discussions ........................................................................................................................................... 36 Implications.......................................................................................................................................... 38 Recommendations ............................................................................................................................... 39 References ............................................................................................................................................ 40
v
Tables Table 1. Treatment structure of the field experiment at Wakool in 2004-05. ...................................................... 6 Table 2. Version of maNage rice, number of copies distributed and new features of each version .................. 10 Table 3. Average N uptake at PI (kg ha-1) of four varieties at different nitrogen treatments in Jerilderie, Leeton and Hay in 2002-03.................................................................................................................. 12 Table 4. Average grain yield (t ha-1) of four varieties at different nitrogen treatments in Jerilderie, Leeton and Hay in 2002-03.............................................................................................................................. 13 Table 5. Average N uptake at PI (kg ha-1) of four varieties at different nitrogen treatments with different sowing time in Bay 1 and 3 at Yanco in 2002-03. ............................................................................... 15 Table 6. Average grain yield (t ha-1) of four varieties at different nitrogen treatments with different sowing time in Bay 1 and 3 at Yanco in 2002-03. ............................................................................... 16 Table 7. Soil Chemical properties of 3 experimental sites in 2003-04. ............................................................. 19 Table 8. Average grain yield (t ha-1) of four varieties at different N treatments in Jerilderie, Leeton and Coleambally in 2003-04. ............................................................................................................... 21 Table 9
Incremental F-statistics for fitted fixed effects..................................................................................... 22
Table 10. Effect of midseason dry down on nutrient content of the flag leaf. ..................................................... 28 Table 11. Average grain yield at 180 N kg ha-1 with two N fertilisers at Griffith in 2005-06............................. 33
Figures Figure 1.
Air and water temperatures during January and February 2004 at 3 experiment sites...................... 18
Figure 2.
The relationships between N uptake at panicle initiation and grain yield of 3 cultivars at the 3 experimental sites in 2003-04 (Δ Leeton, □ Coleambally, ○ Jerilderie)......................................... 20
Figure 3.
Yield response of each variety to applied N at each 3 experimental sites in 2003-04 ...................... 23
Figure 4.
Relationship between sterility and applied N in Jerilderie in 2003-04.............................................. 24
Figure 5.
Yield responses to different timing of N application at Yanco in 2003/04 season ........................... 25
Figure 6.
Yield response to rate and timing of N application at Jerilderie in 2004-05..................................... 26
Figure 7.
Yield response to rate and timing of N application at Wakool in 2004-05. ...................................... 27
Figure 8.
Yield response to different nutrients treatments under two water regimes at Wakool in 2004-/05 .. 28
Figure 9.
Water and air temperatures during 2005-06 rice season at Griffith and Wakool.............................. 30
Figure 10. Yield response of Langi and YRL 125 to N treatments in 2005-06.................................................. 31 Figure 11. Yield responses of Amaroo in 2005-06 season to different N treatments at Griffith and Wakool (combined data for 2 sites), where midseason dry down was practiced............................................ 32 Figure 12. The relationship between N uptake at PI and grain yield in two systems of water management...... 33 Figure 13. Yield responses to different N treatment under two water management systems at Jerilderie in 2006-07. ............................................................................................................................................ 35 Figure 14. Relationship between grain yield and biomass accumulation at PI for Amaroo under continuous flooding (combined data for 2002- 2005). ........................................................................................ 36
vi
Executive Summary What the report is about? Research into nitrogen management in rice cultivation has been ongoing for many years. Constant upgrading of fertiliser recommendations is necessary because the properties of soils change enormously with time, release of new varieties and use of new management practices. Therefore, it is imperative to upgrade nitrogen management options to suit these changes. In addition, it is required to upgrade the maNage rice software, a user friendly computer based decision support system used by rice growers, to include new information. This project provided facilities to carry out investigations to further enhance the knowledge of nitrogen management in Australian rice paddocks. Since 2002, this project has investigated issues related to nitrogen management under different scenarios and developed nitrogen management options for newly released varieties and for different water management systems. The project has also gathered information on effects of other nutrients on rice productivity under different water management systems, and the influence of sowing date on nitrogen responses. 2006-07 rice season was the final year of the project. The key findings of each season’s work and the conclusions of the five seasons’ worth of work are presented in this report. Who is the report targeted at? This report is intended for Australian rice industry particularly growers, agronomists and consultants and also rice breeders, related scientific experts and range of government agencies responsible for land and water resource development. Background Nitrogen management is one of the critical issues for sustainable and profitable rice production in NSW. This is not just because N status of the crop is a key determinant of yield potential, but also because high N status is associated with extreme cold sensitivity. Nitrogen is the primary fertilizer used for the production of rice in Australia and N uptake by the plant is considered a key determinant for achieving high yields. The response of current commercial rice varieties to N has been shown to alter from year to year and site to site hence making accurate N recommendations for rice crops is a challenging task for researchers. Constant upgrading of fertiliser recommendations is therefore, necessary, especially for a crop like rice because the properties of rice growing soils change enormously with time due to the intensity of cultivation. As new varieties are developed and released the nitrogen and low temperature response of these varieties also need to be understood and described. Rice farmers pay a large yield and profit penalty for under fertilising their crops, but also can increase their risk of cold damage by over fertilising. To enable rice growers to manage the nitrogen fertiliser application and risk involved in it, they use a computer based decision support system “maNage rice”. maNage rice, a user friendly computer based decision support system allows growers to experiment with a range of options for their crop and determine the best management regime for their particular crop. Many rice growers use the tool for day to day management of their rice paddocks. This important software need to be upgraded to include information on new varieties and other practices.
vii
Aims/Objectives This project was implemented to enhance the rice yields by improving the nitrogen management practices in rice paddocks and to upgrade/update the maNage rice software based on findings of the project. This project also aimed to provide information on nitrogen management option for new varieties to the NSW rice industry, so that the maximum benefit of new varieties can be quickly obtained on farm. Methods used During 2002-07, several investigations were carried out in Riverina region of New South Wales, Australia, centred at 35oS and 146oE.to evaluate different nitrogen management strategies in rice growing paddocks. Most of the investigations were carried out in farmers’ paddocks representing major rice growing soils and climatic zones for the area. Responses of key commercial rice varieties to rate and timing of nitrogen fertiliser application were evaluated in the investigations. Performances of new varieties released by rice breeders were also compared with their older counterparts. Investigations were carried out using local production standards under two water management systems currently practiced by farmers in the area. Many farmers keep the bays under continuous flooding whereas some practiced a midseason dry down where paddocks are kept without surface water for 12 – 14 days prior to panicle initiation. Results/Key findings •
Results indicated that there are no significant differences in nitrogen responses among Australian rice varieties and optimum nitrogen requirement lies around 170–180 kg N ha-1 depending on the inherent soil nitrogen supply
•
Five years worth of results show that the best yields are obtained when nitrogen application is split between pre-flood and panicle initiation applications in continuously flooded bays.
•
A minimum pre-flood application of 90 kg N ha-1 is necessary to ensure adequate nitrogen supply during the vegetative growth stage of the crop to produce sufficient biomass to sustain a good yield
•
A maintenance requirement of nitrogen at panicle initiation is necessary to ensure adequate nitrogen supply to the plant during the reproductive stages
•
The results suggest that N uptake at PI beyond 150 kg N ha-1 is detrimental when plants are exposed to low temperature spell during young microspore stage
•
Mid-tillering nitrogen applications are warranted if an inadequate amount of pre-flood nitrogen was applied, early nitrogen application was not managed correctly, or the soil is inherently low in fertility
•
Results indicated that it is better to carry out nitrogen topdressing just prior to re-irrigation in bay with midseason dry down.
Implications for relevant stakeholders Project contributed to develop, improve and distribute “maNage rice” model which contains up to date tools to best manage their rice crop. More than 750 “maNage rice” CDs were distributed annually among growers and advisors during last 4 years. Since January 2007, version 6.3 has been available for download on a grower-only part of the Sunrice website. viii
In addition, findings of investigations will enhance the farmers’ knowledge on nitrogen management of their rice paddocks resulting in increased yields. Recommendations The following recommendations could be made from the findings of this work. •
Split nitrogen application of 2/3 pre-flood and 1/3 panicle initiation could be recommended in continuously flooded bays to improve the grain yields and reduce the risk of cold damage and lodging.
•
Minimum application of 90 kg N ha-1 at pre-flood is required in fields with low nitrogen status, to sustain a good grain yield.
•
A maintenance requirement of nitrogen at panicle initiation is recommended to ensure adequate nitrogen supply to the plant during the reproductive stages.
•
It is important to target N uptake at panicle initiation below 150 kg N ha-1 to avoid a possible yield loss due low temperature spell during young microspore stage.
•
Mid-tillering nitrogen applications are warranted if an inadequate amount of pre-flood nitrogen was applied, early nitrogen application was not managed correctly, or the soil is inherently low in fertility
•
Mid-tillering nitrogen applications should be based on tissue tests that quantify the nitrogen requirement of the crop at that stage based on plant biomass and nitrogen uptake, rather than visual observations
•
It could be recommended to carry out nitrogen topdressing just prior to re-irrigation in bays with midseason dry down.
•
It would be wise to practice midseason dry down during the first half of December to avoid hot weather during the later half.
•
Nitrogen uptake at panicle initiation in mid-season dry down bays is different to that in conventionally managed bays therefore, new calibrations are necessary to assess the nitrogen status at panicle initiation in drained bays.
ix
Introduction Nitrogen (N) is the primary fertilizer used for the production of rice in Australia and nitrogen uptake by the plant is considered a key determinant for achieving high yields (Williams and Angus 1994). The response of current commercial rice varieties to nitrogen has been shown to alter from year to year and site to site hence making accurate nitrogen recommendations for rice crops is a challenging task for researchers. Constant upgrading of fertiliser recommendations is therefore, necessary, especially for a crop like rice because the properties of rice growing soils change enormously with time due to the intensity of cultivation. New rice varieties also need to be assessed for their response to nitrogen. Nitrogen is usually applied as a heavy single dose before seeding (pre-flood) and placed at 5 - 10 cm below the soil surface. However, depending on the N uptake at panicle initiation (PI) stage, a second dose of nitrogen may be applied as a top dress at PI. It is estimated that a 12 t ha-1 crop of rice would remove about 208 kg of N per hectare (NSW DPI, 2004). It is recommended to apply 120 – 180 kg of N per hectare at pre-flood to achieve a target nitrogen uptake of 130-150 kg N ha-1 at panicle initiation to obtain 12 t ha-1 rice yield (NSW DPI, 2004) depending on the history of the paddock. A rice crop after a 2 to 3 year pasture phase usually does not require any nitrogen fertiliser as pasture sufficiently builds up the soil nitrogen reserves. However, high nitrogen fertiliser rates at pre-flood can increase the risk of cold damage during reproductive stages due to low night temperatures during late January and early February (Farrell et al. 2001). Historical weather data illustrates that the average minimum temperature during this period is 17°C (Heenan 1984; Lewin and Heenan 1987). However, temperature variability, which can include extreme periods of low temperature, is a major problem that is shared by Australia and other temperate rice producing countries. High nitrogen content in the rice plant increases the risk of yield reductions due to low temperatures during reproductive development (Sasaki and Wada 1975; Yoshida 1981; Haque 1988) by reducing the number of engorged pollen grains per anther resulting increased spikelet sterility (Gunawardena et al. 2003). High nitrogen conditions can also delay panicle development (Gunawardena 2002), thereby increasing the risk of encountering low temperatures during the young microspore and flowering stages. The extra biomass produced due to heavy doses of nitrogen fertiliser applications at pre-flood could also increase the risk of exposing the rice plant to extreme temperatures during reproductive development. As such, timing and rate of nitrogen application is very critical in avoiding low temperature damage in Australian rice production. Research into the effect of the timing of nitrogen fertiliser application on yield has been ongoing for many years. Some findings have shown that total application of nitrogen fertiliser before flooding improves yield (Fageria and Baligar, 1999) as well as the nitrogen fertiliser efficiency (Patrick and Reddy, 1976; Humphreys et al., 1987; Russel et al., 2006). However, some findings revealed that midseason nitrogen top dressing is as efficient as pre-flood application (De Datta, 1987; Cassman et al., 1998). Other findings have demonstrated that heavy total nitrogen applications before flooding can increase the risk of damage to the crop in cold seasons, and induce lodging at harvest (Bacon and Heenan, 1984; Barmes, 1985; Fujisaka, 1993; Heenan and Lewin, 1982). Therefore, it is an imperative task for growers to manage the nitrogen requirement of their crop to achieve high yields and at the same time to avoid risk involved in it. A range of technologies are available to manage nitrogen in rice paddocks. These include the soil testing, tissue testing at various stages and use of computer based decision support systems. Currently, rice growers use a computer-based decision support system, “maNage rice”, developed jointly by CSIRO and NSW Department of Primary Industries, to evaluate the risk and rewards of 1
various management options. The maNage rice has shown to be a dependable tool for rice growers and farm advisers as it allows growers to experiment with a range of options for their crop and to determine the best management regime for their particular crop. The core of maNage rice is an integrated system first designed to estimates the yield response of rice to top-dressed nitrogen fertiliser. This system consists of a simple simulation model that has been calibrated on field experiments and tested against other field experiments. The strong feature of manage rice is that it considers all the available information in estimating the response to nitrogen fertiliser. This information includes sowing date, plant-N status, variety and risk of cold damage. In addition, it considers the grain price and nitrogen fertiliser cost in calculating the economic optimum rate of nitrogen fertiliser. Furthermore, most of the previous work on nitrogen management in Australian rice systems has been focused on nitrogen requirements under continuously flooded conditions. However, many rice farmers in Australia are currently practising a mid-season dry down to mitigate the occurrence of straighthead, a physiological condition which reduces the yield significantly. Moreover, very little information on the requirement of nutrients and their interactions in rice except nitrogen is available apart from their uptake at different nitrogen levels. If high target yields continue to be achieved, nutrients other than nitrogen should also be given consideration as they will be rapidly become depleted. In addition, time of sowing can alter the responses of rice plant to nitrogen fertiliser applications. When sowing is delayed, there is a high risk of yield loss due to cold damage during young microspore stage as it occurs after early February especially with high nitrogen rates. However, about 10-15% of rice farmers in Australia do not adhere to the recommended sowing windows due to variety of reasons. Therefore, nitrogen management in paddocks with delayed sowing is a key to achieve good yields. Our knowledge on the effect of delayed sowing on nitrogen responses is fragmentary. Since 2002, this project has investigated issues related to nitrogen management under different scenarios and developed nitrogen management options for newly released varieties and for different water management systems. This project also gathered information on effect of other nutrients on rice productivity under different water management systems. The influence of sawing date on nitrogen responses was also investigated. The information gathered during these investigations was used to further upgrade maNage rice for new varieties and to add new functions with each annual update.The key findings of each season’s work and the conclusions are presented here.
2
Objectives Improved farm yields through genetic and crop-management gains, based on knowledge of rice physiology by developing and improving maNage rice and other decision support systems for rice growers and advisers, with most emphasis on N nutrition and its interaction with low temperature.
3
Methodology Environmental conditions at experimental locations The trials were located in Riverina region of NSW, centred at 35oS and 146oE. The altitude throughout the region is approximately 120m. The climate of the area is temperate and characterized by high temperatures and high total radiation during summers. The region has a large diurnal temperature variation of 12° to 15°C and low humidity. Due to the climate variability of the area, only one rice crop could be grown in a year. The rice growing season (October to March) is characterised by long days and high levels of solar radiation averaging 27MJ m-2 day-1 during December, with low temperatures at the beginning and end of the season. Although areas in the southern Riverina can be slightly cooler than in the northern Riverina, temperatures usually vary little within a given season across the whole rice-growing area. However, temperature variability, which can include extreme periods of low temperature, is a major problem of the area. Very low temperatures (<17oC) during late January and early February (during the reproductive stages) could cause very high spikelet sterility and result in severe yield losses. However, the low temperatures from late February through to March extend the duration of the grain filling process, leading to the production of high quality grain. Rice crops in the area are drained and harvested in March and April before the first frosts, when grain moisture is between 16% and 22%. Most of the experiments were located in paddocks having a history of at least 10 years of commercial rice cultivation with cracking, grey clay soils (50 -60% clay, Grey Vertosol).
Common cultural practices adopted Land preparation was carried out in all experimental sites to obtain a well aggregated 5 – 7 cm deep firm seedbed. Pre-flood nitrogen as urea (according to the treatment designs) was drilled in 5 - 7 cm below the soil surface 3 days before permanent flooding. Ridge rolling was done after urea application to create a smooth seedbed. Pre-germinated seeds were aerial sown at 120 - 150 kg ha-1 into shallow water (5 - 10 cm) depth. . The remaining nitrogen as urea was top dressed uniformly at various growth stages (mid-tillering, PI and booting) according to the treatment design. Water level was maintained at 5 - 10 cm until PI (except in midseason dry down bays). At PI, the water depth was increased up to 25 cm and maintained at that level until 35 days before harvesting. Other management practices followed the local production standards for rice (NSW DPI, 2004). In bays with midseason dry down, water level was maintained at 5 -10 cm until 3 wk before PI. Then the paddocks were subjected to a dry down for 12 - 14 days and re-irrigated to maintain a 10 cm water depth until PI.
Trials 2002-03 During the season, three field experiments were conducted at Leeton, Jerilderie and Hay to extend the information for expanding maNage rice for new varieties. Each experiment consisted of 3 replicates of six nitrogen treatments and 4 varieties in a 2 range by 24 row array. The experimental design was a randomised block design. Trials were planted in October 2002. The trials compared newer varieties Paragon and Quest with their older counterparts Amaroo and Millin. The nitrogen treatments consisted of three N rates (0, 90 and 240 kg N ha-1) applied at preflood with same three nitrogen rates applied at pre-flood with topdressing of 90kg N ha-1 at PI.
4
Two field experiments were also conducted in summer 2002-03 in adjoining bays at Yanco to evaluate the effect of planting time on nitrogen response of 4 varieties. Experiment 1 was sown on 5th October 2002 while Experiment 2 was sown on 31st October 2002. Each experiment consisted of 3 replicates of six nitrogen treatments and 4 medium grain cultivars Amaroo, Millin, Paragon and Quest. Amarro and Paragon are long duration while Quest and Millin are short duration verities. The first four nitrogen treatments consisted of 4 N levels (0, 90, 240 and 360 kg N ha-1) applied at pre-flood. Treatments 5 and 6 received 90 and 240 kg N ha-1 at pre-flood with 90 kg N ha-1 at PI as a top dress. The experimental design was a randomized complete block with 3 replicates. The trials were harvested in April-May 2003.
Trials 2003-04 Four nitrogen trials were conducted at Yanco, Leeton, Jerilderie and Coleambally in the 2003-04 rice season. The objectives were to quantify the nitrogen requirement of new varieties and to assess the response of Australian rice cultivars to split nitrogen application. Responses of four varieties (Amaroo, Millin, Quest and Reiziq) to four nitrogen levels (0, 90, 180 and 270 kg N ha-1) and two nitrogen timing treatments (a. total N pre-flood, b. 2/3 pre-flood and 1/3 PI) were investigated in the Leeton, Jerilderie and Coleambally trials. Each experiment consisted of 3 replicates in a split plot design with nitrogen treatments randomised to main plots and varieties randomised to subplots within main plots. At Yanco, the response of Amaroo to four nitrogen levels (0, 90, 180 and 270 kg N ha-1) and three nitrogen timing treatments (a. total N pre-flood, b. 2/3 pre-flood and 1/3 PI and c. 1/3 pre-flood and 2/3 PI) were examined. The experiment was consisted of 3 replicates in a randomised block design. Leeton, Jerilderie and Coleambally trials were planted in late October - early November 2003 while Yanco trial was planted in early October 2003. Trials were harvested in April-May 2004.
Trials 2004-05 During the 2004-05 rice season, 3 trials were conducted at Jerilderie and Wakool to further investigate some of these issues relating to N fertiliser use and straighthead occurrence. The objectives of these trials were to compare the effects of timing of nitrogen application on rice yield and the effect of N, S, Zn, Cu and midseason draining on straighthead occurrence. Responses of Amaroo to four nitrogen levels (0, 90, 180 and 270 kg N ha-1) and four timings of nitrogen application were investigated at Jerilderie. The four timings of nitrogen application were; a) total N pre-flood, b) 2/3 pre-flood and 1/3 PI, c) 2/3 pre-flood and 1/3 at mid-tillering and d) 3 equal doses at pre-flood, mid-tillering and PI. Experiment consisted of 3 replicates in a randomized block design. At Wakool, responses of Opus to four nitrogen levels (0, 90, 180 and 270 kg N ha-1) and four timings of nitrogen application were investigated. Four nitrogen application times were; a) total N pre-flood, b) 2/3 pre-flood and 1/3 just before re-irrigation after the midseason dry down, c) 1/3 pre-flood and 2/3 just before re-irrigation after the midseason dry down and d) 2/3 at pre-flood and 1/3 PI. Experiment consisted of 3 replicates in a randomised block design. In addition, an experiment consisted of 11 treatments of different levels and combinations of N, S, Zn and Cu as shown in Table 1 with two irrigation treatments (continuous flooding and midseason dry down) was conducted at Wakool to investigate the effect of theses treatments on straighthead occurrence. The set of nutrients treatments are shown in Table 1. Total N and S were drilled in 5 - 7
5
cm below the soil surface while Zn and Cu fertilisers were broadcasted before pre-flood. Foliar Zn and Cu treatments were carried out during mid-tillering. Experiment consisted of 3 replicates in a split plot design with irrigation treatments randomised to main plots and nutrient treatments randomised to subplots within main plots. Trials were sown in late October 2004 and harvested in April-May 2005. Table 1. Treatment structure of the field experiment at Wakool in 2004-05.
Treatments
N
S
Zn
Cu
Zn
Cu
No.
(kg ha-1)
(kg ha-1)
(kg ha-1)
(kg ha-1)
(Foliar)
(Foliar)
1
-
-
-
-
-
-
2
-
100
-
-
-
-
3
200
100
-
-
-
-
4
-
-
15
-
-
5
-
-
-
5
-
-
6
-
-
-
-
0.5%
-
7
-
-
-
-
-
0.5%
8
-
100
15
-
-
-
9
-
100
-
5
-
-
10
-
100
-
-
0.5%
-
11
-
100
-
-
-
0.5%
Trials 2005-06 During the 2005–06 rice season, 3 trials were carried out at Leeton, Jerilderie and Wakool to quantify the nitrogen requirement of new long grain variety, YRL 125, in conjunction with its older counterpart Langi. The trials further assessed the response of current rice cultivars to split nitrogen application. Four nitrogen levels (0, 90, 180 and 270 kg N ha-1) and two nitrogen timing treatments (a. total N preflood and 2. 2/3 pre-flood and 1/3 PI) were evaluated. Each experiment consisted of 3 replicates in a split plot design with nitrogen treatments randomised to main plots and varieties randomised to subplots within main plots. Another two trials were conducted in 2005-06 season at Griffith and Wakool to investigate the effect of timing of nitrogen application on the performance of rice in bays with midseason dry down. Response of Amaroo to four nitrogen levels (0, 90, 180 and 270 kg N ha-1) and four timings of nitrogen application were evaluated. The four timing of nitrogen application were; a) total N preflood, b) 2/3 pre-flood and 1/3 just before re-irrigating after the midseason dry down c) equal split at pre-flood, just before re-irrigating and booting and d) 2/3rd pre-flood and 1/3 PI. Each experiment consisted of 3 replicates in a randomised block design.
6
In addition, a trial was carried out in Griffith to compare the performance of new slow releasing N fertilizer Entect with that of urea. Entect is coated with nitrification inhibitor (3,4-dimethyle pyrazole phosphate). Response of Amaroo to two nitrogen fertilisers at 180 kg N ha-1 applied at pre-flood was evaluated. Application rate of 180 kg N ha-1 was chosen here because it was found to be the optimum rate for Amaroo in previous investigations. The experimental design was randomised block design with 3 replicates. Trials were planted in late September-October 2005 and harvested in April-May 2006.
Trials 2006-07 In the 2006–07 season, a trial was conducted at Jerilderie to evaluate the responses of Reiziq to different rates and timings of nitrogen application in continuously flooded and midseason dry down bays. Five nitrogen rates (0, 60, 120, 180 and 240 kg N ha-1) and three timings of nitrogen application (a. total N pre-flood, b. 2/3 pre-flood and 1/3 just before re-irrigating after the midseason dry down, and c. 2/3 pre-flood and 1/3 panicle initiation) were used in the experiments. Experiment consisted of 3 replicates in a split plot design with irrigation treatments randomised to main plots and nitrogen treatments randomised to subplots within main plots. The trial was planted in mid October 2006 and harvested in April 2007.
Sample collection and analysis Soil sampling and analysis Soil sampling was carried out before sowing and fertiliser application. Ten soil cores of 50 mm diameter were taken from each replicate randomly at each experimental site to a depth of 100 mm. After thorough mixing, a composite sample of each replicate was taken for analysis. Soils samples were then stored in a cold room at – 4oC until analysis. The samples were air dried and ground to pass 2 mm sieve before the chemical analysis. Soil pH, organic carbon, exchangeable cations, available P and mineral N were determined by methods described in Rayment and Higginson (1992). Soil pH was determined from a 1:5 soil/0.01M CaCl2. The Walkley and Black method was used to determine organic carbon while phosphorus was measured by the bicarbonate-extractable phosphorus method (Colwell extraction). Exchangeable cations (K+, Ca2+, Mg2+) were extracted using ammonium chloride extraction and subsequently measured using Inductively Coupled Plasma Spectrometry (ICP). Total N was determined by LECO combustion (Wang and Anderson, 1998). Chemical analyses of the samples were performed by Diagnostic and Analytical Services Environmental Laboratories of NSW Department of Primary Industries at Wollongbar and Wagga Wagga, NSW, Australia.
Plant sampling and analysis Plant samples were taken randomly at mid tillering, just before dry down in bays with midseason dry down, PI, flowering and physiological maturity within the first half of each plot leaving a 0.5m margin from the boundary.
7
Above ground plant samples were taken by cutting plants at the base within a quadrat (area of the quadrat used varies from 0.16 to 1 m2). Sampling at PI was carried out just before N topdressing at PI. A 100 g sub-sample was taken from the bulk sample and dried in a microwave oven for 5 – 6 minutes. Each sub-sample was ground and analysed for N by near-infrared spectroscopy (NIRS) using a locally developed calibration (Batten et al. 1991). The remainder of the bulk sample was dried at 70oC till constant weight is attained. The dry weights of the microwave sample and bulk sample were combined to determine the biomass accumulation at PI. At physiological maturity, above ground samples taken within a 1 m2 quadrat were dried at 70oC for 3 days to determine the total biomass at the stage. Samples were then threshed to determine the total dry weight of unhusked grain (paddy). The total dry weight of grain in the sample is used to estimate the final yield at 14% moisture. Harvest index (HI) is calculated as the ratio of grain yield to the total cumulative biomass at harvest. In addition, a sample of 30 tillers collected from next to the quadrats were dried at 70oC and separated into straw, full grains and empty grains to determine number of floret per panicle, percentage empty grain (% sterility) and the grain weight. Grain yield (14 % moisture) was determined by mechanically harvesting a 1.8 – 2.0m wide strip (depending on the harvester) from the remaining half of each plot with a small plot harvester.
Statistical analysis Data was analysed using analysis of variance to determine significant differences at their 5% and 1% level of probability using Genstat 8.1 (Lawes Agricultural Trust 2003).
8
Results and Discussions Continued development of maNage rice maNage rice is a computer-based decision support tool which has been developed for use by Riverina ricegrowers and advisers by CSIRO and NSW Department of Primary Industries. Development of maNage rice was supported by RIRDC from 1994 until 2001 on previous projects, and since 2002 as part of this project (DAN207A). The core of maNage rice is an integrated system first designed to estimates the yield response of rice to top-dressed N fertiliser. This system consists of a simple simulation model that has been calibrated on field experiments and tested against other field experiments. It remains the core part of maNage rice and during the course of the project it has been upgraded as new information became available from the experimental parts of the project. The strong feature of manage rice is that it considers all the available information in estimating the response to N fertiliser. This information includes sowing date, plant-N status, variety and risk of cold damage. In addition, it considers the grain price and Nfertiliser cost in calculating the economic optimum rate of N fertiliser. The original simulation model has been retained, apart from small changes to reflect reported increases in the efficiency of topdressed N fertiliser. The original model was calibrated with a Root Mean Square Error (RMSE) of 0.75 t/ha and was tested on independent data with a RMSE of 1.0 t/ha. These are as good as simulation models for upland crops. During 2003 the impact of research on rice nutrition, including maNage rice and NIR-related projects, was evaluated by the Centre for International Economics. This analysis showed a net benefit:cost =11 and an internal rate of return of 180%. These values indicate remarkably high returns to the rice industry for its investment in this research area. Details of the versions and their new components are shown in Table 2. Each new component mostly consisted of a screen providing information that could not readily be conveyed on hard copy. One reason for continuous updates was to keep the core parts of the system related to N management before users and advisers as they explore the new features. The original aim in developing the system was to provide information that is not obtainable as hard copy. This aim was met with the development slide and screens reporting N management, water use, harvest time, and sowing time. The feature most widely used by growers was the Water Use Calculator. It enabled growers to budget their allocated water use based on current and recent weather automatically downloaded from the internet and projected future water. It has been useful in helping to resolve misunderstandings about water use. The least widely used feature was the Harvest Date calculator. At the 2006 pre-season meetings we explained its use to intending users and anticipate great use in future. In addition to the new features shown in the Table, the existing parts of the program were improved, bugs were fixed and the programs updated annually to refer to new varieties and to give access to new experimental results and updated weather. Since version 6.1, the CD has included pdfs of documents from DPI. In version 6.3 the documents have expanded to include Sunrice and Ricegrowers’ Association documents and weblinks. These have been consolidated into a Ricegrowers’ Toolkit. The aim is to be more inclusive and present technical information from the range of organisations servicing the industry, while fully recognising the origin and ownership of the data. With these additions the package is drifting from the original aim of providing only information that could not be provided as hard copy. They were added at the request of ricegrowers and advisers, so maNage rice is evolving as a one-stop shop for technical information about Riverina Rice. It is evolving in a similar way to the CottonLogic package.
9
Table 2. Version of maNage rice, number of copies distributed and new features of each version
Version
Year
Distribution
New features
5.3
2002
~600 registered users
Weather download
Water use calculator
Harvest date Grain quality
6.0
2003
~650 registered users
Zone Management
Developmental slide
6.1
2004
~700 registered users
Rice diagnostics
Ricecheck pdf
6.2
2005
All ricegrowers
pdfs of DPI guides
Manual pdf
(~1700) 6.3
2006
All ricegrowers (~1700)
Sowing date & yield
Ricegrowers’ toolkit
Since version 5.3, maNage rice has been distributed on CDs. Distribution of versions 5.3-6.1 was restricted to users who signed a 1-page Terms and Conditions. Since release of version 6.2 CDs have been posted to all ricegrowers, based on a list updated annual by Ricegrowers Pty. Ltd. The main reason for the change was that it was becoming difficult to keep an up-to-date list of ricegrowers and we were beginning to post copies to some who had left the industry. In addition, the marginal cost of additional CDs was relative small. The signed Terms and Conditions have been replaced by a click/accept system when the user installs the program on a computer. The conditions are displayed in Help and referred to in hard-copy output. The change was made in consultation between legal staff at CSIRO and NSW Department of Primary Industries. The agreement between CSIRO and NSW Department of Primary Industries was revised to allow for the new distribution system. In January 2007, version 6.3 has been available for download on a grower-only part of the Sunrice website. The file is accessible only with a password supplied to ricegrowers registered with Sunrice. It is intended that future versions of maNage rice will be generally be provided by direct download. This change is possible because of improved download speeds and the prospect for future improvements.
10
Trials 2002-03 Yield responses at Jerilderie, Leeton and Hay Results of the trials conducted in Jerilderie, Leeton and Hay were very variable with no clear cut guidance on the newer cultivars. There has been a suggestion in past trials that Paragon became more sterile with higher N rates. This was not reflected in yields in this season although Paragon showed significantly higher N uptake at PI compared to other 3 varieties in most instances (Table 3). Quest showed consistently higher yields in most situations compared with other 3 cultivars (Table 4). Millin, Paragon and Quest yields were optimized around 180 kg N ha-1 and split application of 90 kg N ha-1 at pre-flood and 90 kg N ha-1 at PI showed comparatively higher yields showing the benefits of N top dressing at PI. All the varieties showed reasonably higher yields at higher nitrogen rates, even at the 330 kg N ha-1 rate. Higher yields at high N rates were expected as there was no extended spells of low temperatures (<15oC) during the young microspore development which would have occurred during the last 3 weeks of January 2003 for varieties and locations. Overall, the results demonstrated the importance of applying nitrogen fertiliser at pre-flood to produce sufficient biomass especially in paddocks with low inherent N. For example, grain yield in treatment 2 which only had 90 kg N ha-1 applied at PI was significantly lower compared to that of the treatment 3 which had 90 kg N ha-1 applied at pre-flood even though both treatments had the same total N. This difference was more in Hay trial where inherent N was lower compared to other two sites as can be deduced from grain yields of the control treatment (zero N treatment). However, results also showed that when adequate N is provided at pre-flood, topdressing of N at PI is better than if the total amount of N is applied at pre-flood. For example, treatment 4 which has split application (90+90) gave the best yield in most occasions. Results of the trials conducted in 2002-03 indicate that the optimum N requirement of four cultivars in the trials lies around 180 kg N ha-1 depending on the inherent N supply.
11
Table 3. Average N uptake at PI (kg ha-1) of four varieties at different nitrogen treatments in Jerilderie, Leeton and Hay in 2002-03.
Location
N treatment
Jerilderie
Leeton
Hay
PFN
PIN
Amaroo
(kg N ha-1)
Millin
Paragon
Quest
N uptake (kg ha-1)
1
0
0
47.3
62.8
58.1
48.2
2
0
90
55.1
54.1
67.3
45.4
3
90
0
101.7
90.1
100.0
83.7
4
90
90
81.0
97.8
97.3
92.7
5
240
0
171.0
170.8
182.2
180.9
6
240
90
179.7
188.9
200.8
174.3
1
0
0
47.1
54.6
55.9
51.4
2
0
90
54.7
63.0
52.3
40.4
3
90
0
74.0
106.9
102.8
85.7
4
90
90
96.3
83.1
100.4
102.9
5
240
0
183.5
111.3
182.0
179.5
6
240
90
154.9
145.4
212.1
114.2
1
0
0
30.8
31.9
38.2
31.3
2
0
90
39.1
31.0
41.1
30.4
3
90
0
89.4
70.5
82.9
77.6
4
90
90
91.9
74.9
93.0
72.8
5
240
0
110.2
99.5
127.8
120.6
6
240
90
117.7
88.7
121.0
86.0
9.02 to compare the response of a variety to different N treatments within a location. LSD (5%)
7.37 to compare the response between varieties to the same N treatment within a location. 36.82 to compare any two data points irrespective of variety, N treatment and the location.
Note: PFN = Pre-flood N application, PIN = N application at panicle initiation, Total N applied = PFN+PIN.
12
Table 4. Average grain yield (t ha-1) of four varieties at different nitrogen treatments in Jerilderie, Leeton and Hay in 2002-03
Location
N treatment
Jerilderie
Leeton
Hay
PFN
PIN
Amaroo
(kg N ha-1)
Millin
Paragon
Quest
Grain yield (t ha-1)
1
0
0
7.58
10.15
9.16
10.31
2
0
90
10.64
11.60
10.98
10.53
3
90
0
11.64
13.83
11.34
12.29
4
90
90
13.21
11.64
11.24
12.84
5
240
0
10.84
10.13
11.30
9.01
6
240
90
7.73
11.67
9.62
10.13
1
0
0
9.81
7.42
8.77
7.43
2
0
90
10.73
9.82
10.07
10.57
3
90
0
11.07
11.56
11.11
12.58
4
90
90
11.85
10.42
12.52
12.60
5
240
0
13.52
11.36
9.75
12.32
6
240
90
12.17
11.39
12.00
10.77
1
0
0
6.00
5.24
6.33
5.06
2
0
90
9.02
6.53
6.17
5.03
3
90
0
9.70
8.70
9.86
7.95
4
90
90
11.36
10.65
9.98
8.19
5
240
0
12.12
10.19
10.51
10.51
6
240
90
11.60
9.83
11.10
10.11
0.63 to compare the response of a variety to different N treatments within a location. LSD (5%)
0.52 to compare the response between varieties to the same N treatment within a location. 2.35 to compare any two data points irrespective of variety, N treatment and the location.
Note: PFN = Pre-flood N application, PIN = N application at panicle initiation, Total N applied = PFN+PIN.
13
Yield responses at Yanco N uptake at PI and grain yield of two trials conducted at Bay 1 and 3 in Yanco are shown in Table 5 and 6 respectively. The trials in Bay 1 and 3 were planted on 5th and 31st October 2002 respectively. The treatment 4 which has N split application (90+90) in Bay 1 showed the highest yields in all 4 varieties confirming the benefits of N topdressing at PI. There is about 46 – 55% yield decrease in Bay 3 compared to Bay 1. This could not be explained by cold damage as there was no extended spells of low temperatures (<15oC) during the young microspore development and flowering. Sterility data (not shown) for the two trials confirm this. There is no significant difference in sterility % between two planting times. If there is a cold damage in Bay 3, sterility would have been significantly higher in Bay 3 compared to that of Bay1. In addition there is a reduction in N uptake at panicle initiation in Bay 3 compared to that of Bay 1. This was due to less biomass development at PI in Bay 3 compared to that of Bay 1. All these indicate that a factor other than nitrogen and sowing time may have contributed to lower yield and N uptake at PI in Bay 3 compared to Bay 1. Unfortunately soil analysis data is not available to access the contribution of soil factors for this yield and N uptake differences between two bays.
14
Table 5. Average N uptake at PI (kg ha-1) of four varieties at different nitrogen treatments with different sowing time in Bay 1 and 3 at Yanco in 2002-03.
Sow Time
N treatment
Bay 1 (5th Oct)
Bay 3 (31st Oct)
PFN
PIN
Amaroo
(kg N ha-1)
Millin
Paragon
Quest
N uptake (kg ha-1)
1
0
0
62.4
40.6
82.8
46.1
2
0
90
61.6
45.3
56.3
48.9
3
90
0
111.5
129.9
125.1
114.4
4
90
90
174.0
119.2
189.1
84.7
5
240
0
231.7
161.6
220.1
291.9
6
240
90
230.1
192.7
287.5
190.9
7
360
0
292.1
258.8
246.3
243.9
1
0
0
34.1
34.6
33.8
28.2
2
0
90
60.0
48.5
30.2
30.1
3
90
0
84.1
59.9
126.5
73.3
4
90
90
61.1
59.1
90.8
66.9
5
240
0
163.2
126.5
160.8
164.2
6
240
90
121.2
117.9
164.0
114.9
7
360
0
233.4
135.5
303.1
138.0
23.2 to compare the response of each variety to different N treatments within a sow time. LSD (5%)
17.6 to compare the response between varieties to the same N treatment within a sow time. 65.82 to compare any two data points irrespective of variety, N treatment and the sow time.
Note: PFN = Pre-flood N application, PIN = N application at panicle initiation, Total N applied = PFN+PIN.
15
Table 6. Average grain yield (t ha-1) of four varieties at different nitrogen treatments with different sowing time in Bay 1 and 3 at Yanco in 2002-03.
Sow Time
Bay 1 (5th Oct)
Bay 3
(31st Oct)
LSD (5%)
N treatment
PFN
PIN
Amaroo
(kg N ha-1)
Millin
Paragon
Quest
Grain yield (t ha-1)
1
0
0
9.55
7.78
8.63
6.84
2
0
90
10.10
9.80
10.59
11.30
3
90
0
12.33
10.35
12.92
12.36
4
90
90
13.89
14.53
14.75
13.32
5
240
0
12.03
11.81
13.04
9.84
6
240
90
10.52
10.31
12.09
9.39
7
360
0
10.22
10.29
12.16
7.05
1
0
0
5.74
4.90
5.30
5.65
2
0
90
7.30
6.71
3.97
6.02
3
90
0
5.71
7.75
8.88
7.76
4
90
90
7.69
9.31
7.24
9.45
5
240
0
5.89
6.28
7.34
7.48
6
240
90
11.98
8.10
7.48
7.76
7
360
0
7.52
4.47
6.95
6.67
0.93 to compare the response of each variety to different N treatments within a sow time. 0.70 to compare the response between varieties to the same N treatment within a sow time. 2.63 to compare any two data points irrespective of variety, N treatment and the sow time.
Note: PFN = Pre-flood N application, PIN = N application at panicle initiation, Total N applied = PFN+PIN.
16
Trials 2003-04 Temperatures during critical reproductive phase Air and water temperature data taken at every 30 minute interval during critical reproductive phase (16 January – 15 February 2005) for 3 sites is shown in Figure 1. All 3 sites showed a similar temperature patterns and experienced spells of minimum air temperatures below 17o C during the periods 18-20 January and 22 January to 7 February. Some nights even experienced temperatures below 10oC. Of the 3 sites, Jerilderie experienced the lowest minimum air temperatures. However, water temperatures remained above 170C at 3 sites indicating the buffer capacity of water to keep bays warmer during low temperature spells.
17
Air Water
40
30
20
17
10
Leeton
Temperature (oC)
40
30
20 17
10
Coleambally
40
30
20 17
10
Jerilderie 16 -J an -0 4
21 -J a
n04
26 -J a
n04
31 -J an -0 4
05 -F eb -0 4
10 -F eb -0 4
15 -F eb -0 4
Figure 1. Air and water temperatures during January and February 2004 at 3 experiment sites.
18
Soil analysis Results of soil analysis shown in Table 7 indicate that soils at Jerilderie had significantly lower total inherent soil N compared to soils at other two sites. Whereas, soils at Leeton site has significantly higher extractable phosphorus (Colwell P) content compared to other two sites. However, the differences in pH, organic carbon and exchangeable cation contents among three sites are not significant. Only single composite sample for the whole trial was analysed at Yanco and therefore the results were not considered for the statistical analysis. However, it shows very high content of extractable soil phosphorus compared to other 3 sites. Table 7. Soil Chemical properties of 3 experimental sites in 2003-04.
Location
pH (CaCl2)
Organic carbon -1
Total N
Colwell P
(g kg-1)
(mg kg-1)
K+
Ca2+
Mg2+
Exchangeable cations
(g kg )
(cmol+ kg-1)
Leeton
4.4
14.7
0.93
49.6
0.90
6.5
4.0
Coleambally
4.6
14.3
0.79
26.3
0.88
4.5
5.0
Jerilderie
4.4
13.7
0.49
26.7
0.87
6.5
4.3
LSD (5%)
0.4
1.8
0.22
10.1
0.19
1.6
1.5
Yanco
5.3
16.0
0.61
160.0
1.00
6.5
2.5
Yield responses at Leeton, Coleambally and Jerilderie Yield in relation to N uptake at PI Results showed that N uptake at PI was significantly varied between cultivars, sites and N treatments (Table 9). Figure 2 shows the relationships between N uptake at PI and the yield at three sites. Only data for treatments with total N applied at pre-flood was considered in this relationship as the yields of split N treatments could be influenced by N application at PI. Results demonstrated that the optimum N uptake (to produce maximum grain yield) at PI was between 145 and 170 kg N ha-1 for the four varieties at Coleambally and Jerilderie. The optimum N uptakes for 4 varieties at Leeton were higher compared with other two sites and can not be determined at Leeton as the yield continued to increase with increasing PI N uptake. Jerilderie had the lowest optimum N uptake at PI and the relationship between yield and N uptake at PI was weak for Quest and Reiziq at Jerilderie. The differences in optimum N uptake at PI between sites may be attributed to differences in inherent soil N and the extent of exposure to low temperature spell at young microspore stage. Across the three sites, the results suggest that N uptake at PI beyond 150 kg N ha-1 is detrimental when plants are exposed to low temperature spell during young microspore stage, and this was particularly evident at Jerilderie. Therefore, it is important to target N uptake at PI below 150 kg N ha-1 to avoid a possible yield loss due low temperature spell during young microspore stage.
19
0 12 10
50
100
150
Leeton Coliambally Jerilderie
8
200
250 2 r = 0.97**
2 r = 0.95** 2 r = 0.90**
2 r = 0.62*
2 r = 0.75**
2 r = 0.91**
6
Grain yield (t ha-1)
4 2
Amaroo
Millin
0 2 r = 0.96**
12 2 r = 0.97**
2 r = 0.76**
2 r = 0.97**
10 8
2 r = 0.19 2 r = 0.23
6 4 2
Quest
Reiziq 0
50
100
150
200
0 250
N uptake at PI (kg ha-1)
Figure 2. The relationships between N uptake at panicle initiation and grain yield of 3 cultivars at the 3 experimental sites in 2003-04 (Δ Leeton, □ Coleambally, ○ Jerilderie).
Yield responses to applied N Yield responses of cultivars to nitrogen application at three experimental sites are shown in Table 8 and Figure 3. Significance of each of these data point to the corresponding point of the response curve of total pre-flood N application is indicated by stars (*). Varieties showed significant variation in grain yield (Table 9) with Millin showing the highest yield at each N level compared with the other three varieties, and Reiziq giving the lowest yield (Fig. 3). Moreover, there is significant variation in grain yield between N treatment and between sites (Table 9). The interaction between N treatment and cultivar was also significant (Table 9). The interaction between site, variety and N treatment was also highly significant (Table 9). The highest yields for all four varieties were attained at Leeton while Jerilderie had the lowest among 3 experimental sites. At Leeton, all cultivars showed a continuous upward trend in yield with increasing N application and may not have reached the maximum yield despite the application of relatively high rate of total nitrogen fertiliser (270 kg N ha-1). At Coleambally and Jerilderie, the optimum nitrogen requirement of all the four varieties was somewhere around 180 kg N ha-1. At Jerilderie, all varieties showed a significant yield decline above 180 kg N ha-1 rate (Fig. 3). At Coleambally, Millin and Quest showed a similar yield decreasing trend above 180 kg N ha-1 rate, while Amaroo and Reiziq showing a plateau beyond 180 kg N ha-1 rate. These differences in yield responses to N application at three locations may be due to varying levels of cold exposure during the young microspore stage. The Leeton trial was sown on the 24 October, whereas the Coleambally and Jerilderie trials were sown on 5 and 7 November respectively. At Leeton panicle initiation occurred during the first week of January, whereas at Coleambally and Jerilderie, it occurred during the 3rd and 4th week of January. Therefore, young microspore stages of plants (15 to 16 days from panicle initiation) at Coleambally and Jerilderie may have been exposed to
20
low temperature spell occurred during the latter part of January and early part of February (Fig. 1). Yield decrease with increasing N application at Coleambally and Jerilderie therefore, could be attributed to cold damage during young microspore stage. Whereas, the young microspore stage of plants at Leeton may have occurred before the low temperature spell. However, the young microspore in Leeton trial may have coincided with the low temperature spell during 18 to 20 January but the impact would have been minimal as the spell was shorter. Sterility percentages also confirm the variation of levels of cold damage at three locations. Sterility percentage was significantly and positively correlated with the applied N at Jerilderie. Figure 4 shows the relationship between sterility % and applied N for 4 varieties at Jerilderie. All four varieties showed significant and strong positive relationship between total applied N and grain sterility %. Reiziq had the highest sterility among 4 varieties at each N level. Both split and no-split treatments showed a similar relationship between total applied N and grain sterility %. However, split N application treatment showed significantly lower sterility than the no-split N application treatment especially at higher N application rates. However, the relationship between sterility % and applied N was weaker at Leeton. At Coleambally, the relationships were significant but weaker than that of Jerilderie (data not shown). This suggests that the relationship between applied N and sterility is determined by the level of exposure of plants to low temperature spells during their young microspore stage. Table 8. Average grain yield (t ha-1) of four varieties at different N treatments in Jerilderie, Leeton and Coleambally in 2003-04.
Location
N
Coleambally
1 2 3 4 5 6 7
PFN PIN (kg N ha-1) 0 0 90 0 180 0 270 0 60 30 120 60 180 90
Leeton
1 2 3 4 5 6 7
0 90 180 270 60 120 180
0 0 0 0 30 60 90
4.51 7.55 9.27 10.70 8.04 11.71 12.52
4.48 7.98 9.45 11.27 7.98 11.00 11.75
4.53 7.39 8.93 11.30 6.83 9.31 10.80
4.42 6.91 9.13 10.37 6.53 9.83 11.16
Jerilderie
1 2 3 4 5 6 7
0 90 180 270 60 120 180
0 0 0 0 30 60 90
4.95 7.22 8.48 5.92 7.54 10.15 8.87
5.67 8.34 8.76 7.93 9.20 10.77 9.77
6.63 7.61 8.28 6.91 7.85 10.29 8.47
5.21 6.18 6.92 4.33 6.70 7.52 6.30
LSD (5%)
Amaroo 5.56 8.42 9.62 9.73 9.49 10.17 11.60
Millin Quest Grain yield (t ha-1) 5.42 3.53 10.61 8.98 11.41 11.42 9.08 9.13 9.47 9.48 11.84 10.79 10.72 10.72
Reiziq 4.59 8.87 10.50 10.58 8.79 10.10 11.49
0.52 to compare the response of a variety to different N treatments within a location. 0.25 to compare the response between varieties to the same N treatment within a location. 1.28 to compare any two data points irrespective of variety, N treatment and the location.
Note: PFN = Pre-flood N application, PIN = N application at panicle initiation, Total N applied = PFN+PIN.
21
Table 9: Incremental F-statistics for fitted fixed effects (** denotes P<0.01, * denotes P<0.05).
Term
df
N uptake at PI
Grain yield
Location
2
24.78**
773.56**
Variety
3
9.37**
34.79**
N Treatment (N Trt)
6
420.38**
101.53**
Location x Variety
6
6.33**
12.47**
Location x N Trt
12
1.89
12.72**
Variety x N Trt
18
1.77*
1.93*
Location x Variety x N Trt
36
1.28
2.6**
Split application treatments at or above 180 kg N ha-1 resulted in higher grain yield compared with corresponding no-split treatments at all three sites. In most instances, the yield differences were significant (Fig.3). These differences were most prominent at Jerilderie. Some split nitrogen application treatments resulted in nearly 2.0 t ha-1 yield advantage over the corresponding no-split (total pre-flood) application. Amaroo showed the best responses to split application over no-split application at all three sites. Millin and Quest also had significant responses to split N application at several N rates at 3 sites. i.e. with 180 and 270 kg N ha-1 at Jerilderie and with 270 kg N ha-1 at Coleambally. Millin also showed a significant higher yield at Leeton with split application of 180 kg N ha-1 over corresponding no-split N application. Reiziq only showed significant responses to split application over no-split application at the 270 kg N ha-1 rate at Coleambally and 180 and 270 kg N ha-1 rates at Jerilderie. However, split N application did not result in significantly lower grain yield compared to no-split application except with 90 kg N ha-1 rate at Coleambally. Lodging was minimal at all locations, even at 270 kg N ha-1. This may be due to zero rainfall during the harvesting season. This shows that the rice crop can stand without lodging at high nitrogen levels if a rainfall event does not occur during the harvest season.
22
Leeton Amaroo
0
12
**
**
90
180
270
Millin
0
90
Quest
Reiziq
Quest
Reiziq
180
270
** 10 8 6
No split Split
4
Coleambally Amaroo
Millin **
*
Yield (t/ha)
*
12
**
**
10
*
8 6 4
23
Jerilderie 12
Amaroo
Millin
Quest
Reiziq
** **
10
**
*
**
**
**
8
* **
6 4 0 90 * significant at 5% ** significant at 1%
180
270
0
90
180
270
Total N applied (kg/ha)
Figure 3. Yield response of each variety to applied N at each 3 experimental sites in 2003-04. Curves indicate N response for total pre-flood application.
0
90
180
270 80
No-split Split
70 60 r2 = 0.90**
50 r2 = 0.86**
r2 = 0.83**
40 r2 = 0.78**
30
Sterility %
20 10
Millin
Amaroo
0 80 70 60 r2 = 0.94**
50
r2 = 0.86**
r2 = 0.82**
40 r2 = 0.78**
30 20 10
Quest
Reiziq
0 0
90
180
270
Total N applied (kg ha-1)
Figure 4. Relationship between sterility and applied N in Jerilderie in 2003-04
Yield responses at Yanco Results indicated a significant yield increase due to N fertiliser application (Fig. 5). However, the yield increase above 90 kg N ha-1 rate was not significant. The trend of increasing yield with increasing N application rate seen at Leeton trial in the same season was not evident in Yanco trial, even though the temperature regimes during young microscope stage were the same at two locations. At the 180 kg N ha-1 rate, the split nitrogen application of 2/3 pre-flood and 1/3 PI showed significant yield advantage over the other two timing treatments. There were no differences in the timing treatments at 90 and 270 kg N ha-1. Total pre-flood application resulted in lower yield at all three application rates compared to that of split nitrogen application treatment of 2/3 pre-flood and 1/3 PI. The performance of the other split application treatment (1/3 pre-flood and 2/3 PI) was not consistent. At the 90 kg N ha-1 rate it resulted in the lowest yield among 3 treatments. Whereas, at the 270 kg N ha-1 rate it had the highest grain yield. This may be attributed to less biomass accumulation at 90 kg N ha-1 rate compared to other two timing treatments. The 1/3 pre-flood and 2/3 PI treatment had the lowest pre-flood applied N (30 kg N ha-1) among all the treatments. Whereas, no-split treatment (total N at pre-flood) and split treatment (2/3 pre-flood and 1/3 PI) received 90 and 60 kg N ha-1 respectively at pre-flood at 90 kg N ha-1 rate. The results suggest that the split nitrogen application is beneficial over no-split application if sufficient nitrogen is available to plants during their vegetative phase.
24
14
12
No split rd rd 1/3 PF & 2/3 PI rd rd 2/3 PF& 1/3 PI Control (0 N)
-1 Grain yield (t ha )
10 LSD = 2.8 8
6
4
2
0 0
90
180
270
-1 Total N applied (kg ha )
Figure 5. Yield responses to different timing of N application at Yanco in 2003/04 season
Trials 2004-05 Yield response at Jerilderie and Wakool There was marked variation in response to N timing and rate among the treatments. Even though there was an extremely low temperature spell during the first week of February, all treatments with split N application showed continuous yield increase with increase in N fertiliser rate at Jerilderie (Fig. 6). However, the no-split treatment (total N pre-flood) showed a significant yield decline after 90 kg N ha1 and resulted in lowest yield among the four timing treatments in higher N rates. The results lead to the conclusion that exposure of plants with high biomass to low temperatures during the first week of February may have contributed to low yield in no-split treatment above the 90 kg N ha-1 rate. In addition, high availability of nitrogen during the vegetative phase may have delayed the young microscope stage and thus the critical stage would have coincided with the low temperature spell. The results suggest that application of the total amount of nitrogen at pre-flood is detrimental in seasons with low temperature spells during the young microspore stage especially at high nitrogen rates. Split treatment with 2/3 pre-flood and 1/3 PI out yielded the other two split treatments except at 270 kg ha-1. The treatment with 3 equal N splits gave the highest yield at 270 kg ha-1. The grain yield of split treatment with top dressing at mid tillering was significantly lower than the other two split treatments. The results suggest that a second application at panicle initiation is better than one at midtillering. Earlier investigations also showed the same result. The low efficiency of nitrogen top dressing at mid-tillering could be attributed to high losses of applied nitrogen through denitrification, as the plants are not quick enough to absorb applied nitrogen due to their small root mass at midtillering.
25
10
Equal split PI split Midseason split No-split
9
8 -1
Grain yield (t ha )
LSD at 5% = 0.57 7
6
5
4
3 0
90
180
270
-1
Total N applied (kg ha )
Figure 6. Yield response to rate and timing of N application at Jerilderie in 2004-05.
However, at Wakool, all the treatments showed continuous decline with increase in N application (Fig. 7). This may be attributed to coincidence of young microspore stage of plant with extremely low temperature spell during the first week of February. The no-split treatment yet again showed the lowest yield at each N level compared to the other 3 timing treatments. Two split treatments with top dressing just before re-irrigation resulted in higher yields compared to PI - split treatment. This suggests that application of N fertiliser just before re-flooding of a dry down bay is more beneficial than applying at PI. In addition, application of N just before re-flooding could minimise the losses as N may go into cracks created during the dry down. Therefore, these results speculate that efficiency of N application is high if applied just before re-flooding than that of PI application in places where midseason dry down is practiced. These results again demonstrated that split N application is beneficial and improves the rice yield. However, the results may be attributed to the unusually cold weather during the first week of February in 2005, which could result in temperature N interactions.
26
7
-1
Grain yield (t ha )
6
5
4 LSD at 5% = 0.57
1/3 pre-flood and 2/3 re-irrigation PI split 2/3 pre-flood and 1/3 re-irrigation No-split
3
2 0
90
180
270
Total N applied (kg ha-1)
Figure 7. Yield response to rate and timing of N application at Wakool in 2004-05.
Yield response to midseason draining and other nutrients at Wakool There were no significant yield differences between 10 nutrient treatments involving nutrients other than N (Fig. 8). However, the treatment with 200 kg N ha-1 gave the lowest yield. This may be attributed to the N and low temperature interactions due to extremely low temperature spell during the first week of February. The results also showed that there is a small yield increase in bays with midseason dry down compared to continuously flooded bays. This yield advantage in midseason dry down could be due to increased availability of certain nutrients resulting from oxidation of rhizosphere during the dry down. Results of flag leaf analysis (Table 10) at anthesis support this hypothesis as midseason drainage showed increased leaf K, Mg, Mn and Cu contents compared to that of continuous flooding. Due to the mild weather during the season, straighthead symptoms were not present and therefore, it is difficult to evaluate the effect of midseason dry down on straighthead occurrence.
27
12
Continuous Midseason dry down
LSD 5% = 1.01
-1 Grain yield (t ha )
10
8
6
4
2
0 1
2
3
4
5
6
7
8
9
10
11
Nutrient treatments
Figure 8. Yield response to different nutrients treatments under two water regimes at Wakool in 2004-/05. The detail treatment structure is given in methodology. Table 10. Effect of midseason dry down on nutrient content of the flag leaf.
Element
Effect of dry down
S
Reduced 3%
Zn
No effect
Cu
Increased 23%
Ca
Reduced 20%
Cl
Increased 10%
Fe
Reduced 8%
K
Increased 7%
Mg
Increased 12%
Mn
Increased 17%
Na
Increased 15%
P
No effect
Si
No effect
28
Trials 2005-06 Temperatures during the season Air and water temperature data collected at 30 minute intervals at Jerilderie, Wakool, Leeton and Griffith indicated that the season was quite warm, without distinctive low temperature (<15oC) spells compared with previous rice seasons. However, some nights experienced temperature below 15oC during the reproductive stage. Nevertheless, deep water levels maintained after PI were able to keep bay temperatures above the critical minimum (170C). However, temperature within bays exceeded 40°C during the dry down period at Griffith and Wakool mainly due to very high ambient temperatures during the period, and the complete dryness of the soil surface in the bay (Fig. 9). Had the hot weather continued for another couple of days, the plants would have suffered severe stress and recovery would have been a difficult task. This suggests that it is advisable to carry out the midseason dry down during the first half of December to avoid the periods of highest risk of hot weather, which are towards the end of December.
29
60
Griffith
Air Water
50
40
30
20
o Temperature ( C)
10
Dry down 0 60
Wakool
50
40
30
20
10
Dry down Nov
Dec
0 Jan
Feb
Mar
Date
Figure 9. Water and air temperatures during 2005-06 rice season at Griffith and Wakool. Water temperatures shown during dry down period are actually air temperatures taken 2 cm above the soil surface as the bays were dry.
Yield responses of Langi and YRL 125 at Jerilderie, Wakool and Leeton There were significant differences in yield among the treatments. However, two cultivars showed similar N responses in all 3 sites. Unlike previous rice seasons, all sites showed a trend of continuous yield increase with increasing nitrogen rates, except with the total pre-flood application (no-split) at Jerilderie and Wakool (Fig. 10). This may be due to favourable weather conditions during the season and no damaging low temperatures at the young microspore stage. However, the no-split treatment showed a significant yield decline after 90 kg N ha-1 rate at Jerilderie and Wakool. At all three sites, the split nitrogen application of 2/3 pre-flood and 1/3 panicle initiation resulted in significantly higher grain yield compared with the no-split treatment, at rates beyond 90 kg N ha-1. Even though the season did not experience low temperature spells during critical growing stages, these 30
findings are parallel to the findings of previous rice seasons which experienced low temperature weather spells. The season’s results suggest that split application of nitrogen is beneficial over total pre-flood application even in years with no low temperature spells during critical reproductive phase. Jerilderie Langi
YRL125
**
12
**
**
**
**
** 10
8
6
Wakool ** **
Langi Grain yield (t/ha)
12
**
**
YRL 125
** **
10
8
6
Leeton Langi
YRL125
12
**
**
** ** 10 ** 8
No-split PI split
6
0
90
180
270
0
90
180
270
Total N applied (kg/ha)
Figure 10. Yield response of Langi and YRL 125 to N treatments in 2005-06.
Yield responses to N in mid-seasonally dry down bays The yield responses to different rates and timings of nitrogen for the two sites are shown in Figure 11. It shows a continuous upward trend in yield with an increase in nitrogen application rate as expected in a normal season without low temperature events. Results also indicated that, the yield responses to nitrogen treatments were highly significant (< p=0.001). The treatment with total N application at pre-flood (no-split) showed the lowest yield for all N rates beyond 90 kg N ha-1 in both experimental sites. The treatment with split N application of 2/3 pre-flood and 1/3 just before re-irrigation after the midseason dry down showed the highest yields at N rates beyond 90 kg N ha-1, followed by the split application of 2/3 pre-flood and 1/3 at PI. These two split N application methods gave about 1–2 t ha-1 yield advantage over the no-split treatment at higher N rates.
31
The results are similar to findings in 2004-05 rice season and suggest that recommendation could be made to encourage farmers to carry out their top dressing just prior to re-irrigating their dry down bays. This will give an option for farmers to apply their N top dressing using tractors as the bays are fairly dry after a midseason dry down. In addition, application of N just before re-irrigation could minimise the N losses as N may go into cracks created during the dry down.
14
Total pre-flood 2/3rd pre-flood, 1/3rd re-irrigation 2/3rd pre-flood, 1/3rd PI Equal split at pre-flood, re-irrigation and booting
2 r =0.91** 2 r =0.88**
Grain yield (t/ha)
12 2 r =0.87** 2 r =0.80** 10
8
6 0
90
180
270
N applied (kg/ha)
Figure 11. Yield responses of Amaroo in 2005-06 season to different N treatments at Griffith and Wakool (combined data for 2 sites), where midseason dry down was practiced.
Results show that N uptake at PI in bays with midseason dry down are significantly different to that of conventional bays (Fig. 12). For example, N uptake at PI in mid-seasonally dry down bays is around 80 kg N ha-1 at 10 t ha-1 yield target, whereas it is around 150 kg N ha-1 in conventional bays. This may be attributed to slow recovery process of plants after the stress during the dry down. This could result in low biomass production by PI in mid-seasonally dry down bays and hence showing low N uptake at PI as the N uptake figure is calculated on the basis of biomass. Yet, they produce similar yield to that of conventional bays. This suggests that biomass levels and N uptake figures at PI in bays with midseason dry down are misleading and does not reveal the actual N status at PI. However, farmers may have wrong impression that their fields are low in N and tend to apply top dressing as recommended in Rice Check for conventional bays. This is not advisable as the fields may be over fertilised. N assessment of mid-seasonally drained bays at PI may, therefore, not be a good practice as plants will not have sufficient time to recover by PI. Therefore, it is necessary to carry out investigations to find out the appropriate age of plant sampling for N assessment and a set of N calibrations for bays with midseason dry down.
32
14
Midseason dry down Conventional 2 r =0.67
12
2 r =0.88
-1
Yield (t ha )
10
8
6
4
25
50
75
100
125
150
175
200
-1
Nitrogen uptake at panicle initiation (kg ha )
Figure 12. The relationship between N uptake at PI and grain yield in two systems of water management.
Yield response to Urea and Entect Results indicated that there are no significant yield differences due to application of two nitrogen fertilisers at 180 kg N ha-1 (Table 11). However, slow N releasing fertilisers like Entect are expected to give better performances as they will lessen the opportunity for plants to access to high N doses during the vegetative phase resulting less biomass development and thus reducing the risk of cold damage and lodging. Therefore, one would expect to have similar performances as with split N applications. However, the results obtained here is contrary to this expectation. This may be due to inability of rice plants to take up pre-flood N fertiliser placed below the soil surface after a certain age therefore, reducing the required N supply during later stages of development. These speculations are worth while to investigate. Table 11. Average grain yield at 180 N kg ha-1 with two N fertilisers at Griffith in 2005-06.
Treatment
Yield (t ha-1)
Urea
10.81
Entect
10.83
Lsd at 5%
1.02
33
Trials 2006-07 Yield response to N in dry down and continuous bays at Jerilderie Results indicate that split N topdressing at PI (PI split) was the best nitrogen timing treatment under continuously flooded conditions, whereas topdressing just before re-irrigation gave the best yield in bays with midseason dry down (Fig. 13). The treatment with topdressing N at PI gave the lowest grain yield among the treatments in bays with midseason dry down. This suggests that there is a greater loss of N when applied at PI in bays with midseason dry down. This may be attributed to lower rate of N recovery by plants which were stressed during the dry down just before PI. The treatment with N topdressing just before re-irrigation showed the best performances in bays with midseason dry down confirming the previous two seasons’ findings. This may be due to the reduction of loss of N applied at this stage because of aerobic soil conditions developed during the dry down. Under both conditions, total pre-flood nitrogen application resulted in yield decline with increasing nitrogen application after the 90 kg N ha-1 rate. This may be attributed to exposure of plots with excessive biomass to a low temperature spell during the first week of February which would have coincided with the critical young microspore stage. The high initial nitrogen rates of the no-split treatment would have encouraged the development of biomass during the vegetative phase.
34
8
Total N pre-flood Topdressing at re-irrigation Topdressing at PI 7
Grain Yield (t ha-1)
6
Conventional Bays 5 8
7
6
Dry down Bays 5 0
60
120
180
240
Applied N (kg ha-1) LSD at 5% = 0.87
Figure 13. Yield responses to different N treatment under two water management systems at Jerilderie in 2006-07.
35
Discussions Overall, the five seasons’ results indicate that split application of nitrogen with optimal proportions between splits is better than total pre-flood application as a nitrogen management strategy for Australian rice crops. The split strategy lowers the risk of cold damage due to uncertain low temperature events during the critical young microspore stage. It reduces lodging that can occur at high pre-flood nitrogen application rates and it increases grain yield – even in seasons without low temperature spells during the critical reproductive phase. Results also indicated that adequate quantities of nitrogen should be applied at pre-flood so as to produce sufficient biomass during the vegetative growth of the rice crop to sustain a good grain yield. This explains why the split treatments did not show the yield advantage over the no-split treatments at nitrogen fertiliser rates below 90 kg ha-1. Figure 14 shows that it is necessary to have a biomass accumulation of more than 6.5 t ha-1 at panicle initiation to achieve a target yield of 10 t ha-1 in an Amaroo crop under continuous flooding. Therefore, it is advisable to apply minimum of 90 kg N ha-1 if rice is to be grown on continuously cultivated bays. A maintenance requirement of nitrogen during the reproductive phase could be applied at panicle initiation. Fujisaka (1993) also confirmed that rice yield in Philippine do not increase with basal nitrogen applications, and that nitrogen is optimally applied at mid-tillering and panicle initiation. However, the findings of this work are contrary to earlier findings where the practice of applying total nitrogen requirement of the crop pre-flood showed a yield advantage over the split nitrogen application – in years without low temperature spells during the critical reproductive phase.
16
-1 Grain yield (t ha )
14
12 r2 = 0.22 10
8
6
4
2 0
2
4
6
8
10
12
-1 Biomass at PI (t ha )
Figure 14. Relationship between grain yield and biomass accumulation at PI for Amaroo under continuous flooding (combined data for 2002- 2005).
36
These results lead to speculation that rice roots are unable to take up nitrogen placed below the soil surface during its reproductive growth stages, especially after the formation of a superficial root mat – even though most of the pre-flood applied nitrogen is available in lower layers of the soil. As a result, the nitrogen requirement of plants during the reproductive phase is not fully met when the entire nitrogen application is placed in the soil before sowing. Moreover, the nitrogen applied at pre-flood may be immobilised due to soil fixation or leaching down below the root layers thus decreasing the nitrogen availability at panicle initiation which is just before maximum plant need. These speculations are worth investigating in future research work. Australian rice growing soils are capable of retaining ammonium based fertilizers for long periods of time because of their fine textures with high cation exchange capacities. Therefore, it is possible to apply the total amount of fertiliser containing ammonium nitrogen at once, without appreciable loss by leaching. The presence of nitrogen in the cationic form, however, does not ensure its loss against leaching and denitrification in submerged soils, since there is a constant ionic movement between clay surfaces and the surrounding soil solution. Those ions which move out of the clay surfaces are vulnerable to leaching and denitrification processes. High nitrogen rates can enhance these processes due to the saturation of exchangeable sites with ammonium ions. Therefore, at high nitrogen rates in submerged soils, some loss of ammonium is possible through leaching and denitrification even though the soils in the Australian rice growing regions are capable of retaining ammonium based fertilisers for long periods of time. Results also suggest that application of nitrogen fertilisers during mid-tillering is not as effective as application at panicle initiation. This may be due to lesser root mass available at that stage to utilise topdressed nitrogen rapidly, to impede losses through denitrification. However, if an inadequate amount of pre-flood nitrogen is applied, the early nitrogen is not managed correctly, or if the soil is inherently low in fertility, an application of nitrogen at mid-tillering may be necessary. While visual observations have been used in the past to decide about a mid-tillering nitrogen application, investigations should be carried out to quantify the nitrogen requirement at that stage based on plant biomass and N uptake, as practiced at panicle initiation. The research showed that just before re-irrigation is the best time to apply nitrogen topdressings in paddocks subjected to midseason dry down. This may be due to minimisation of nitrogen losses through denitrification compared with topdressing being carried out at other stages. In addition, some of the nitrogen fertiliser applied just prior to re-irrigation could enter cracks in the soil that developed during dry down, thus further reducing the potential for denitrification. Further research is necessary to quantify the reduction of nitrogen losses through this practice. This option also gives an opportunity for farmers to apply the fertiliser using tractors and ground spreaders, as the bays are fairly dry, rather than using aerial contractors.
37
Implications Project contributed to develop, improve and distribute “maNage rice” model which contains up to date tools to best manage their rice crop. More than 750 “maNage rice” CDs were distributed annually among growers and advisors during last 4 years. The feature most widely used by growers was the Water Use Calculator. It enabled growers to budget their allocated water use based on current and recent weather automatically downloaded from the internet and projected future water. It has been useful in helping to resolve misunderstandings about water use. Since January 2007, version 6.3 has been available for download on a grower-only part of the Sunrice website. During 2003 the impact of research on rice nutrition, including maNage rice and NIR-related projects, was evaluated by the Centre for International Economics. This analysis showed a net benefit:cost =11 and an internal rate of return of 180%. These values indicate remarkably high returns to the rice industry for its investment in this research area. In addition, findings of investigations will enhance the farmers’ knowledge on nitrogen management of their rice paddocks resulting in increase in yields.
38
Recommendations The following recommendations could be made from the findings of this work. •
Split nitrogen application of 2/3 pre-flood and 1/3 panicle initiation could be recommended in continuously flooded bays to improve the grain yields and reduce the risk of cold damage and lodging.
•
Minimum application of 90 kg N ha-1 at pre-flood is required in fields with low nitrogen status, to sustain a good grain yield.
•
A maintenance requirement of nitrogen at panicle initiation is recommended to ensure adequate nitrogen supply to the plant during the reproductive stages.
•
It is important to target nitrogen uptake at panicle initiation below 150 kg N ha-1 to avoid a possible yield loss due low temperature spell during young microspore stage.
•
Mid-tillering nitrogen applications are warranted if an inadequate amount of pre-flood nitrogen was applied, early nitrogen application was not managed correctly, or the soil is inherently low in fertility.
•
Mid-tillering nitrogen applications should be based on tissue tests that quantify the nitrogen requirement of the crop at that stage based on plant biomass and nitrogen uptake, rather than visual observations.
•
Results indicated that there are no significant differences in nitrogen responses among Australian rice varieties and optimum nitrogen requirement lies around 170–180 kg N ha-1 depending on the inherent soil nitrogen supply.
•
It could be recommended to carry out nitrogen topdressing just prior to re-irrigation in bays with midseason dry down.
•
It would be wise to practice midseason dry down during the first half of December to avoid hot weather during the later half.
•
Nitrogen uptake at panicle initiation in mid-season dry down bays is different to that in conventionally managed bays, thus new calibrations are necessary to assess the nitrogen status at panicle initiation in drained bays.
39
References Bacon PE and Heenan DP (1984) Response of Inga rice to application of nitrogen fertilizer at varying growth stages. Australian Journal of Experimental Agriculture and Animal Husbandry 24, 250 – 254. Barmes JE (1985) The response of rice to time and rate of application of nitrogen fertiliser in the Burdekin Valley. Queensland Journal of Agricultural and Animal Services 42, 71-77. Batten GD, Blakeney AB, Glennie-Holmes M, Henry RJ, McCaffery A, Bacon PE and Heenan DP (1991) Rapid determination of shoot nitrogen status in rice using near infrared reflectance spectroscopy. Journal of the Science of Food and Agriculture 54: 191-7. Cassman KG, Peng S, Olk DC, Ladha JK, Reicharrdt W, Dobbermann A and Singh U ( 1998). Opportunities for increasing nitrogen-use efficiency from improved resource management in irrigated rice systems. Field Crop Research 56, 7-39. De Datta SK (1987). Advances in soil fertility research and nitrogen management for low land rice. In: Efficiency of nitrogen fertilizers for rice, International Rice Research Institute, Los Banos, Philippines, pp. 27-41. Fageria NK and Baligar VC (1999) Yield and yield component of lowland rice as influenced by timing of nitrogen fertilization. Journal of Plant Nutrition 22, 23-32. Farrell TC, Williams RL, Fukai S (2001) The cost of low temperature to the NSW rice industry. In ‘Proceeding of the 10th Australian Agronomy Conference’. Australian Society of Agronomy www.regional.org.au/au/asa/2001/1/d/farrell.htm. Fujisaka S (1993) Were farmers wrong in rejecting a recommendation? The case of nitrogen at transplanting for irrigated rice. Agricultural Systems 43, 271-286. Gunawardena TA (2002) Spikelet sterility in rice (Oryza sativa L.) induced by low temperature and nitrogen fertilisation. PhD Thesis. The University of Queensland, Australia. Gunawardena TA, Fukai S, Blamey FPC (2003) Low temperature induced spikelet sterility in rice. I. Nitrogen fertilisation and sensitive reproductive period. Australian Journal of Agricultural Research 54, 937-946. Haque MZ (1988) Effect of nitrogen, phosphorus and potassium on spikelet sterility induced by low temperature at the reproductive stage of rice. Plant and Soil 109, 31-36. Heenan DP (1984) Low temperature induced floret sterility in the rice cultivars Calrose and Inga as influenced by nitrogen supply. Australian Journal of Experimental Agriculture and Animal Husbandry 24, 255-259. Heenan D P and Lewin L G (1982) Response of Inga rice to nitrogen fertilizer rate and timing in New South Wales. Australian Journal of Experimental Agriculture and Animal Husbandry 22, 62-66. Humphreys E, Murihead WA, Melhuish FM and White RJG (1987) Effect of time of urea application on combine-sown Calrose rice in south-east Australia. I. Crop response and N uptake. Australian Journal of Agricultural Research 38, 101 -112. Lawes Agricultural Trust. (2003) GENSTAT Statistical Package. Rothamsted Experimental Station: Harpenden, UK.
40
Lewin LG and Heenan DP (1987) The agronomy of rice production in the Riverina region of southeastern Australia. Efficiency of nitrogen fertilizers for rice: In ‘Proceedings of the meeting of the International Network on Soil Fertility and Fertilizer Evaluation for Rice’. Griffith, Australia, 10-16 April 1985. 69-80. NSW DPI (2004) Ricecheck recommendations: A guide to objective rice crop management for improving yields, grain quality and profits, and for economic and environmental sustainability. NSW Department of Primary Industries and the RIRDC Rice Research Development Committee, Australia. pp20. Patrick WH Jr and Reddy KR (1976) Fate of fertilizer nitrogen in a flooded rice soil. Soil Science Society of America Journal 40, 678-681. Rayment GE and Higginson FR (1992) Australian laboratory handbook of soil and water chemical methods. Inkata Press, Melbourne. Russel CA, Dunn BW, Batten GD, Williams RL and Angus JF (2006) Soil test to predict optimum fertiliser nitrogen rate for rice. Field Crop Research 97, 286-301. Sasaki K and Wada S (1975) Effects of nitrogen, phosphoric acid and potash on the percentage of sterile grains in rice plants. Proceedings of the Crop Science Society of Japan 44, 250-254. Van Dijk DC (1961) Soils of the southern portion of the Murrumbidgee Irrigation Areas. CSIRO Division of Soils Technical Bulletin No. 40. Wang D and Anderson DW (1998) Direct measurement of organic carbon content in soils by the Leco CR-12 Carbon Analyser. Communication in Soil Science and Plant Analysis 29, 15-21. Williams RL and Angus JF (1994) Deep flood water protects high nitrogen rice crop from low temperature damage. Australian Journal of Experimental Agriculture 34, 927-932. Yoshida S (1981) Fundamentals of rice crop science. International Rice Research Institute, Los Banos, Philippines, p269.
41
Increasing Productivity and Water Use Efficiency in Australia’s Rice Industry through Nitrogen Management RIRDC Publication No. 09/173 By Ranjith Subasinghe and John Angus
Nitrogen management is one of the critical issues for sustainable and profitable rice production in NSW. This is not just because nitrogen status of the crop is a key determinant of yield potential, but also because high nitrogen status is associated with extreme cold sensitivity. Rice farmers pay a large yield and profit penalty for under fertilising their crops and can also increase their risk of cold damage by over fertilising. As new varieties are developed and released, the nitrogen and low temperature response of these varieties need to be understood and described. Furthermore, nitrogen requirement of a crop is site specific and time dependent. Constant upgrading of fertiliser recommendations is therefore necessary, especially for a crop like rice because the properties of rice growing soils change enormously with time due to the intensity of cultivation.
Most RIRDC books can be freely downloaded or purchased from www.rirdc.gov.au or by phoning 1300 634 313 (local call charge applies).
www.rirdc.gov.au
This research aims to provide information to the NSW rice industry on how to manage nitrogen fertilisation based on field and glasshouse experiment results, so that the maximum benefit of new varieties can be quickly obtained on farm. The Rural Industries Research and Development Corporation (RIRDC) is a partnership between government and industry to invest in R&D for more productive and sustainable rural industries. We invest in new and emerging rural industries, a suite of established rural industries and national rural issues. IRDC books can be purchased by phoning 1300 634 313 or online R at: www.rirdc.gov.au.
Contact RIRDC: Level 2 15 National Circuit Barton ACT 2600 PO Box 4776 Kingston ACT 2604 Ph: 02 6271 4100 Fax: 02 6271 4199 Email:
[email protected] web: www.rirdc.gov.au