Strategic applications of nitrogen fertiliser to increase the yield and nitrogen use efficiency of wheat

Transcription

1 i Strategic applications of nitrogen fertiliser to increase the yield and nitrogen use efficiency of wheat A thesis submitted for the degree of Masters Of Agricultural Science School of Agriculture, Food and Wine The University of Adelaide Peter Hooper January 2010

2 ii TABLE OF CONTENTS ABSTRACT...v DECLARATION...vii ACKNOWLEDGMENTS...viii LIST OF FIGURES...ix LIST OF TABLES...xvii LIST OF ABBREVIATIONS... xix 1. General introduction Review of literature Introduction Nitrogen use in southern Australia The role of nitrogen fertiliser Nitrogen fertiliser use in southern Australia Grain yields in southern Australia Nitrogen use efficiency Definition of nitrogen use efficiency Current NUE values The physiological limit to NUE Factors limiting NUE High pre-anthesis water use Storing water soluble carbohydrates Methods for improving NUE Increasing nitrogen uptake Balancing pre-anthesis and post anthesis water-use Reducing soil evaporation Extending leaf area duration Delayed applications of nitrogen fertiliser Delayed nitrogen applications and NUE Benefits of delayed nitrogen applications Summary General materials and method Introduction Site descriptions Soils Climate Agronomy Sampling and measurement Nitrogen use efficiency Data analysis...44

3 iii 3.7 Experimental treatments Nitrogen application timing in relation to crop growth stage and rainfall Grain yield, grain quality and NUE Introduction Method Results Grain yield Grain protein Total grain nitrogen Screenings Harvest index Nitrogen harvest index Agronomic efficiency Apparent recovery Physiological Efficiency Discussion Conclusion Canopy development Introduction Method Results Dry matter Total plant nitrogen Total crop nitrogen uptake Green Area Index Light interception Discussion Conclusion Yield formation and components Introduction Method Results Shoot number at stem elongation (GS30) Head number and tiller mortality Grain number Individual grain weight Grains per spike Spikelets per head Grains per spikelet Discussion Conclusion Review of current nitrogen management research Introduction Nitrogen rate Time of application...147

4 iv 7.4 Plant density General discussion REFERENCES APPENDICES Appendix A: Grain yield, grain quality and NUE table of significance Appendix B: Canopy development: Dry matter, Shoot N% and Total shoot N table of significance Appendix C: Canopy development: Green area index and light interception table of significance Appendix D: Yield formation and components table of significance Appendix E: the Zadoks cereal growth scale (Zadoks et al. 1974)...171

5 v ABSTRACT In southern Australia a large proportion of crop nitrogen requirements are usually applied early in the season, either just before or at sowing, or during the first 6-8 weeks after sowing. However, the time of greatest demand for nitrogen by wheat crops is during the stem-elongation phase when the crop is growing fastest, while early demand for nitrogen is small The study utilised two field sites in South Australia (Hart and Mintaro) over two seasons ( ) to examine the effect of delayed applications of nitrogen fertiliser on nitrogen use efficiency (NUE) to dryland wheat. Treatments were applied to manipulate the crop canopy and to measure the balance between canopy size and structure and crop nitrogen uptake. In both seasons and sites significant (P<0.05) grain yield increases were obtained by later nitrogen application times. Grain protein also increases with rate of applied nitrogen and delayed time of application, with the 1 st node (GS31) treatment generally producing the highest values. Applying higher rates of nitrogen fertiliser reduced NUE, while later applications were able to increase NUE. Canopy manipulation was able to increase crop nitrogen uptake, coinciding with improved grain yield formation. Generally, the data suggest that the nitrogen taken up later in the season is used very efficiently by the crop.

6 vi The experiments outlined in this thesis showed that there is potential to improve NUE in dryland wheat grown in southern Australia, using tactical applications of nitrogen fertiliser.

7 vii DECLARATION This work contains no material which has been accepted for the award of any other degree or diploma in any university or other tertiary institution to Peter Hooper and, to the best of my knowledge and belief, contains no material previously published or written by another person, except where due reference has been made in the text. I give consent to this copy of my thesis, when deposited in the University Library, being made available for loan and photocopying, subject to the provisions of the Copyright Act I also give permission for the digital version of my thesis to be made available on the web, via the University s digital research repository, the Library catalogue, the Australasian Digital Theses Program (ADTP) and also through web search engines, unless permission has been granted by the University to restrict access for a period of time. Peter J Hooper

8 viii ACKNOWLEDGMENTS I would like to especially thank my supervisors; David Coventry and Glenn McDonald (University of Adelaide) and Allan Mayfield (Allan Mayfield Consulting) for initiating the study and providing continued contribution and support. An input of technical expertise, experimental guidance, academic advice, editorial direction and comments are some of the numerous roles they performed, all of which were greatly appreciated. I am grateful for the assistance I received from many researchers, technicians, agronomists and farmers from within Australia and overseas, the names of whom are too numerous to mention. A special thanks goes to my parents and friends for their continual support and encouragement. I also wish to thank the South Australian Grains Industry Trust for their financial support and the Hart Fieldsite Group for the provision of land and support. The use of data from Pivot fertiliser and various grower groups is also acknowledged.

9 ix LIST OF FIGURES Figure 2.1. Seasonal pattern of nitrogen demand for wheat crops of high and low nitrogen status (Angus 2001)... 7 Figure 2.2. Total nitrogen consumption (kt) in Australia. Source (ABARE 2009) Figure 2.3. Five year moving average grain yield (t/ha) for Australia between 1969/70 and 2007/08. (ABARE 2009)...11 Figure 2.4. Trend in grain yield (t/ha) for the South East ( ), Mid-North and Yorke Peninsula ( ), Eyre Peninsula ( ) and state average ( ) for South Australia (ABARE 2009) Figure 2.5. Decreasing soil evaporation with higher ground cover at flowering (Angus and van Herwaarden 2001)...29 Figure 3.1. Nitrogen application timing shown as arrows in relation to rainfall events at Hart in Figure 3.2. Nitrogen application timing shown as arrows in relation to rainfall events at Hart in Figure 3.3. Nitrogen application timing shown as arrows in relation to rainfall events at Mintaro in Figure 3.4. Nitrogen application timing shown as arrows in relation to rainfall events at Mintaro in Figure 4.1. The grain yield response to nitrogen at Mintaro in a) 2003 and b) Values are the averages for Frame and H45 as the Variety x N treatment interaction was not significant. Symbols represent sowing ( ) midtillering ( ); 1 st node ( ); T1+T2+T3 (+); T2+T3 ( ); T2+T5 ( ); T3+T4 ( ); T3+T5 (Δ). (LSDs [0.05] for nitrogen rate x timing are 0.14 for 2003 and 0.32 for 2004 at Mintaro)...53 Figure 4.2. The grain yield response to nitrogen at Hart for a) Frame and b) H45 in Symbols represent sowing ( ) mid-tillering ( ); 1 st node ( ); T1+T2+T3 (+); T2+T3 ( ); T2+T5 ( ); T3+T4 ( ); T3+T5 (Δ). (LSD [0.05] for nitrogen rate x timing x variety is 0.38 in 2003 at Hart) Figure 4.3. The grain yield response to nitrogen at Hart in Values are the averages for Frame and H45 as the Variety x N treatment interaction was not significant. Symbols represent sowing ( ) mid-tillering ( ); 1 st node ( ); T1+T2+T3 (+); T2+T3 ( ); T2+T5 ( ); T3+T4 ( ); T3+T5 (Δ). (There were no significant differences (P<0.05) at Hart in 2004). 56

10 x Figure 4.4. The relationship between grain protein and nitrogen rate for both Frame and H45 at Mintaro and Hart in 2003 and Symbols represent sowing ( ) mid-tillering ( ); 1 st node ( ); T1+T2+T3 (+); T2+T3 ( ); T2+T5 ( ); T3+T4 ( ); T3+T5 (Δ). (LSDs [0.05] for nitrogen rate x timing are ns and 1.07 for Hart and Mintaro in 2003, and 0.74 and 0.91 for Hart and Mintaro for 2004)...58 Figure 4.5. The relationship between total grain nitrogen and nitrogen rate with application timing for both Frame and H45 at Mintaro in 2003 and Symbols represent sowing ( ) mid-tillering ( ); 1 st node ( ); T1+T2+T3 (+); T2+T3 ( ); T2+T5 ( ); T3+T4 ( ); T3+T5 (Δ). (LSDs [0.05] for nitrogen rate x timing were 1.07 in 2003, and 0.91 for 2004) Figure 4.6. The relationship between total grain nitrogen and nitrogen rate for both cultivars and all application timings at Mintaro in 2003 and 2004, and Hart Symbols represent Mintaro 2003 ( ) Mintaro 2004 ( ); Hart 2003 ( ). (LSDs [0.05] for nitrogen rate is 6.34 for Mintaro and ns for Hart in 2003, and 6.93 for Mintaro in 2004.) The dashed line represents 50% apparent recovery of applied fertiliser in the grain...61 Figure 4.7. The relationship between physiological efficiency and grain protein at Mintaro for 100 kg N/ha in 2003 ( ); 200 kg N/ha in 2003 ( ); 100 kg N/ha in 2004 ( ) or 200 kg N/ha in 2004 ( ) Figure 5.1 Dry matter production for nitrogen rate with days after sowing (DAS), between mid-tillering (GS22) and anthesis (GS65) (including maturity at Mintaro), averaged for the varieties Frame and H45 at a) Hart and b) Mintaro in 2003 and (LSDs [0.05] for nitrogen rate at Hart are 5.50, ns, and 51.7 in 2003, and ns in For Mintaro 5.50, 4.90, 81.5 and 51.5 in 2003, and 2.59, ns, 57.1 and in 2004)...81 Figure 5.2. Dry matter production with nitrogen rate for Frame and H45 at Mintaro in 2003 and 2004 measured at a) mid-tillering (GS22) b)1 st node (GS31). Symbols represent sowing ( ), mid-tillering ( ) and T1+T2+T3 ( ) in 2003, and sowing ( ), mid-tillering ( ) and T1+T2+T3 (Δ) in (LSDs [0.05] for nitrogen rate x timing in 2003 are 6.4 and 9.8 for GS22 and GS31 respectively, and for and 25.6 for GS22 and GS31 respectively)...82 Figure 5.3. Dry matter production with nitrogen rate and application timing for Frame and H45 at Mintaro in 2003 and 2004 measured at a) anthesis 2003 b) anthesis 2004 c) maturity 2003 d) maturity Symbols represent sowing ( ) mid-tillering ( ); 1 st node ( ); T1+T2+T3 (+); T2 + T3 ( ); T3 + T5 (Δ). (LSDs [0.05] for nitrogen rate x timing in 2003 are and 68.2 for anthesis and maturity respectively, and for and for anthesis and maturity respectively)...83

11 xi Figure 5.4. The relationship between dry matter production measured at anthesis and time of nitrogen application for 100 kg N/ha ( ) and 200 kg N/ha (+) at Mintaro in 2003 and The treatments are the single applications of nitrogen at mid tillering (T2) and 1 st node (T3). (LSDs [0.05] for nitrogen rate x timing are as per Figure 5.3)...84 Figure 5.5. Shoot nitrogen concentration for nitrogen rate and days after sowing, across all varieties and application timings. Symbols represent Hart 0 kg N/ha ( ) Hart 40 kg N/ha ( ); Hart 80 kg N/ha ( ); Mintaro 0 kg N/ha ( ); Mintaro 100 kg N/ha ( ); Mintaro 200 kg N/ha (Δ) Figure 5.6. The effect of fertiliser rate on the shoot nitrogen content (%) at different growth stages at Mintaro in The data are the averages across all varieties and applications timings. Symbols represent midtillering (GS22) ( ) 1 st node (GS31) ( ); 1 st awn emergence (GS49) ( ); anthesis (GS65) ( ); maturity ( ). (LSDs [0.05] for nitrogen rate are 0.12, 0.27, 0.13, 0.17 and 0.08 for GS22, GS31, GS49, anthesis and maturity respectively)...87 Figure 5.7. The effect of nitrogen fertiliser rate on shoot nitrogen content (%) at 1 st node (GS31) averaged across both varieties at Hart in Symbols represent sowing ( ) sowing + mid-tillering + 1 st node split ( ); midtillering (GS22)( ). (LSDs [0.05] for nitrogen rate x timing is 0.31)...88 Figure 5.8. The effect of nitrogen fertiliser rate and time of application on the shoot nitrogen content (%) at awn emergence (GS49), averaged across both varieties at Mintaro in Symbols represent sowing ( ) sowing + mid-tillering + 1 st node split (+); 1 st node (GS31)( ); T2 + T3 ( ). (LSDs [0.05] for nitrogen rate x timing is 0.16 for Mintaro in 2004)...88 Figure 5.9. The effect of nitrogen rate and time of application on shoot nitrogen content (%) at anthesis (GS65), averaged across all varieties at a) Hart in 2003 and b) Mintaro in Symbols represent sowing ( ) sowing + mid-tillering + 1 st node split (+); mid-tillering (GS22)( ); 1 st node (GS31)( ); T1+T2 ( ); T2 + T3 ( ); T3 + T5 (Δ). (LSDs [0.05] for nitrogen rate x timing are 0.16 for Hart in 2003, and 0.23 for Mintaro in 2004)...89 Figure Total plant nitrogen content between mid-tillering (GS22) and anthesis (GS65) averaged for all varieties, timings and nitrogen rates at Mintaro 2003 ( ); Hart 2003 ( ); Mintaro 2004 ( ); Hart 2004 ( )...91

12 xii Figure The effect of fertiliser nitrogen rate on total plant nitrogen content averaged for all timings and varieties at a) Mintaro and b) Hart in 2003 and 2004 measured at different growth stages. For 2003 symbols represent mid-tillering (GS22) ( ); 1 st node (GS31)( ); anthesis ( ), and for 2004 mid-tillering ( ); 1 st node (Δ); anthesis ( ). (LSDs [0.05] for nitrogen rate x timing at Hart are ns, ns, and 14.9 for GS22, GS31 and anthesis respectively, in 2003, and 8.35 for anthesis in At Mintaro the LSDs are 3.1, 3.6 and 25.2 at GS22, GS31 and anthesis respectively for 2003 and 1.6, 8.3 and 10.9 at GS22, GS31 and anthesis respectively for 2004)...91 Figure Total plant nitrogen (kg/ha) and nitrogen applied (kg N/ha) for application timing at 1 st node (GS31), averaged for both Frame and H45 at Mintaro in a) 2003 and b) Symbols represent sowing ( ) midtillering ( ); T1+T2+T3 (+). (LSDs [0.05] for nitrogen rate x timing are 10.8 for 2003, and 4.5 for 2004) Figure Total plant nitrogen (kg/ha) and nitrogen applied (kg N/ha) for application timing at anthesis, averaged for both Frame and H45 at Mintaro in a) 2003 and b) Symbols represent sowing ( ) midtillering (GS22)( ); T1+T2+T3 (+);1 st node (GS31)( ); T1 + T2 ( ); T2 + T3 ( ); T3 + T5 (Δ). (LSDs [0.05] for nitrogen rate x timing are ns for 2003, and 14.7 for 2004)...93 Figure Change in GAI with days after sowing until anthesis (GS65) for both Frame and H45 at a) Mintaro and b) Hart for 2003 and Symbols represent Frame 2003 ( ) H ( ); Frame 2004 ( ); H ( ).The solid black line is GAI=3.5, which is approx 90% light interception Figure GAI and nitrogen applied (kg N/ha) at Hart, measured at a) 1 st node (GS31) or b) 1 st awn emergence (GS49) for Frame and H45 in Symbols represent sowing ( ) mid-tillering ( ); 1 st node (GS31)( ); T2 + T3 ( ); T1+T2+T3 (+); T2 + T5 ( ). Solid line represents Frame and dashed line H45. (LSDs [0.05] for nitrogen rate x timing are ns for GS31, and 0.92 for GS49) Figure GAI and nitrogen applied (kg N/ha) at Mintaro, measured at a) 1 st node (GS31) or b) 1 st awn emergence (GS49) in 2003 or tip of flag (GS37) in 2004, for Frame and H45. Symbols represent sowing ( ) mid-tillering ( ); 1 st node (GS31)( ); T2 + T3 ( ); T1+T2+T3 (+); T2 + T5 ( ). Solid line represents 2003 and dashed line (LSDs [0.05] for nitrogen rate x timing are 0.19 and 0.41 for GS31 in 2003 and 2004, and ns and 0.34 for GS49 and GS37 in 2003 and 2004) Figure GAI and nitrogen applied (kg N/ha) at Mintaro, measured at anthesis, for Frame and H45. Symbols represent sowing ( ) mid-tillering ( ); 1 st node (GS31)( ); T2 + T3 ( ); T1+T2+T3 (+); T2 + T5 ( ); T3 + T5 ( ). Solid line represents 2003 and dashed line (LSDs [0.05] for nitrogen rate x timing are 0.28 for 2003 and 0.19 for 2004)....98

13 xiii Figure The relationship between green area index and dry matter (grams per square metre) measured at 1 st node (GS31), averaged for variety, nitrogen rate and timing, at Hart 2003 ( ), Mintaro 2003 ( ) and Mintaro 2004 ( )...99 Figure The relationship between green area index at GS49 in 2003 or GS37 in 2004 and dry matter (grams per square metre) at anthesis, averaged for variety, nitrogen rate and timing, at Hart 2003 ( ), Mintaro 2003 ( ) and Mintaro 2004 ( ) Figure The relationship between green area index and total crop nitrogen (kg N/ha) measured at 1 st node (GS31), averaged for variety, nitrogen rate and timing, at Hart 2003 ( ), Mintaro 2003 ( ) and Mintaro 2004 ( ) Figure The relationship between green area index and total crop nitrogen (kg N/ha) measured at anthesis (GS65), averaged for variety, nitrogen rate and timing, at Hart 2003 ( ), Mintaro 2003 ( ) and Mintaro 2004 ( ) Figure Change in light interception with days after sowing until anthesis (GS65) for both Frame and H45, averaged for nitrogen rate and timing at a) Mintaro in 2003 and 2004 and b) Hart for Symbols represent Frame 2003 ( ) H ( ); Frame 2004 ( ); H ( ) Figure Light interception and nitrogen applied (kg N/ha) at Mintaro, measured at mid-tillering (GS22), for Frame and H45. Symbols represent sowing ( ) mid-tillering ( ); 1 st node (GS31)( ); T2 + T3 ( ); T1+T2+T3 (+); T2 + T5 ( ); T3 + T5 ( ). Solid line represents 2003 and dashed line (LSDs [0.05] for nitrogen rate x timing are ns for 2003 and ns for 2004) Figure Light interception and nitrogen applied (kg N/ha) at Mintaro, measured at GS31, for Frame and H45. Symbols represent sowing ( ) mid-tillering ( ); 1 st node (GS31)( ); T2 + T3 ( ); T1+T2+T3 (+); T2 + T5 ( ); T3 + T5 ( ). Solid line represents 2003 and dashed line (LSDs [0.05] for nitrogen rate x timing are 8.74 for 2003 and 7.78 for 2004) Figure Light interception and nitrogen applied (kg N/ha) at Mintaro, measured at GS49 in 2003 and GS37 in 2004, for Frame and H45. Symbols represent sowing ( ) mid-tillering ( ); 1 st node (GS31)( ); T2 + T3 ( ); T1+T2+T3 (+); T2 + T5 ( ); T3 + T5 ( ). Solid line represents 2003 and dashed line (LSDs [0.05] for nitrogen rate x timing are ns for 2003 and 2.84 for 2004)...104

14 xiv Figure Light interception and nitrogen applied (kg N/ha) at Mintaro, measured at anthesis, for Frame and H45. Symbols represent sowing ( ) mid-tillering ( ); 1 st node (GS31)( ); T2 + T3 ( ); T1+T2+T3 (+); T2 + T5 ( ); T3 + T5 ( ). Solid line represents 2003 and dashed line (LSDs [0.05] for nitrogen rate x timing are 2.89 for 2003 and 3.80 for 2004) Figure 6.1. The effect of nitrogen rate on shoot number at stem elongation (GS30) for Frame and H45 for all application timing treatments at Hart in 2003 (+); Hart in 2004 ( ); Mintaro in 2003 ( ) and Mintaro in 2004 ( ). (LSDs [0.05] for nitrogen rate are ns for Hart in 2003, ns for Hart in 2004, 67 for Mintaro 2003 and 97.5 for Mintaro 2004) Figure 6.2. The effect of nitrogen rate and application timing on shoot number (GS30) at Mintaro in a) 2003 and b) Symbols represent sowing ( ); mid-tillering ( ); 1 st node ( ); T1+T2+T3 (+); T2+T3 ( ); and T2+T5 (Δ). (LSDs [0.05] for nitrogen rate x timing are 86.5 for 2003 and for 2004) Figure 6.3. Relationship between shoot number at stem elongation (GS30) and total plant nitrogen uptake at 1 st node (GS31) for both varieties at Hart in 2003 ( ); Mintaro in 2003 ( ); and Mintaro in 2004 ( ) Figure 6.4. Relationship between shoot number at stem elongation (GS30) and GAI at 1 st node (GS31) for both varieties at Hart in 2003 ( ); Mintaro in 2003 ( ); and Mintaro in 2004 ( ) Figure 6.5. Relationship between grain yield response and shoot number at stem elongation (GS30) to either 40 or 100 kg N/ha for both varieties at Hart in 2003 ( ); Mintaro in 2003 ( ); Hart in 2004 ( ) and Mintaro in 2004 ( ) Figure 6.6. The effect of nitrogen rate and application timing on head number at Hart in a) 2003 and b) Symbols represent sowing ( ); midtillering ( ); 1 st node ( ); T1+T2+T3 (+); T2+T3 ( ); T2+T5 (Δ); T3 + T4 ( ); T3 + T5 ( ).(LSDs [0.05] for nitrogen rate x timing at Hart in 2003 and 2004 are ns) Figure 6.7. The effect of nitrogen rate and application timing on head number at Mintaro in a) 2003 and b) Symbols represent sowing ( ); midtillering ( ); 1 st node ( ); T1+T2+T3 (+); T2+T3 ( ); T2+T5 (Δ); T3 + T4 ( ); T3 + T5 ( ).(LSDs [0.05] for nitrogen rate x timing at Mintaro in 2003 are 47.8 and 64.0 in 2004) Figure 6.8. Relationship between shoot number and head number for both varieties at Hart in 2003 ( ); Mintaro in 2003 ( ); Hart in 2004 ( ) and Mintaro in 2004 ( )

15 xv Figure 6.9. Shoot mortality (%) for both Frame and H45 at Mintaro in a) 2003 and b) Symbols represent sowing ( ); mid-tillering ( ); 1 st node ( ); T1+T2+T3 (+); T2+T3 ( ); T2+T5 (Δ); T3 + T4 ( ); T3 + T5 ( ). (LSDs [0.05] for nitrogen rate x timing are 11.1 % for 2003, and 18.6 % for 2004) Figure Relationship between shoot number and shoot mortality for both varieties at Hart in 2003 ( ); Mintaro in 2003 ( ); Hart in 2004 ( ) and Mintaro in 2004 ( ) Figure The response of grain number (grains/m 2 ) to nitrogen rate and application timing for both Frame and H45 at Hart in a) 2003 and b) Symbols represent sowing ( ) mid-tillering ( ); 1 st node ( ); T1+T2+T3 (+); T2 + T3 ( ); T2 + T5 (Δ). (LSDs [0.05] for nitrogen rate x timing are ns for 2003, and 802 for 2004) Figure The response of grain number (grains per square metre) with nitrogen rate for both Frame and H45 at a) Hart and b) Mintaro. Symbols represent Hart 2003 ( ) Hart 2004 ( ); Mintaro 2003 ( ); Mintaro 2004 (+). (LSDs [0.05] for nitrogen rate x timing are 767 for Hart 2003, ns for Hart 2004, 505 for Mintaro 2003 and 1357 for Mintaro 2004). 125 Figure The response of grain number to nitrogen rate and application timing for a) Frame and b) H45 at Hart in Symbols represent sowing ( ) mid-tillering ( ); 1 st node ( ); T1+T2+T3 (+); T2 + T3 ( ); T2 + T5 (Δ). (LSDs [0.05] is 1434 for the variety x rate x timing interaction) Figure The response of grain number (grains per square metre) to nitrogen rate and application timing for both Frame and H45 at Mintaro in a) 2003 and b) Symbols represent sowing ( ) mid-tillering ( ); 1 st node ( ); T1+T2+T3 (+); T2 + T3 ( ); T2 + T5 (Δ). (LSDs [0.05] for nitrogen rate x timing are 668 for 2003, and ns for 2004) Figure Relationship between grains/m 2 and grain yield for variety, nitrogen rate and timing at Hart in 2003 ( ); Mintaro in 2003 ( ); Hart in 2004 ( ) and Mintaro in 2004 ( ) Figure Relationship between green area index at anthesis in 2003 and flag leaf emergence (GS37) in 2004 and grain number for variety, nitrogen rate and timing at Hart in 2003 ( ); Mintaro in 2003 ( ) and Mintaro in 2004 ( ) Figure Grain weight (mg) for Frame and H45 at a) Hart and b) Mintaro in 2003 and Symbols represent Frame in 2003 ( ) H45 in 2003 ( ); Frame in 2004 ( ) and H45 in 2004 (+). (LSDs [0.05] for nitrogen rate x timing at Hart are ns in 2003 and 2004 and at Mintaro are ns in 2003 and 1.8 in 2004)

16 xvi Figure Grain weight (mg) for Frame and H45 at Mintaro in a) 2003 and b) Symbols represent sowing ( ) mid-tillering ( ); 1 st node ( ); T1+T2+T3 (+); T2 + T3 ( ); T2 + T5 (Δ). (LSDs [0.05] for nitrogen rate x timing are 1.67 for 2003, and 2.33 for 2004) Figure The relationship between grain weight and grain number for both varieties at Hart in 2003 ( ); Mintaro in 2003 ( ) and Mintaro in 2004 ( ) Figure The relationship between light interception, at anthesis, and grain weight for both varieties at Hart in 2003 ( ); Mintaro in 2003 ( ); Hart in 2004 ( ); and Mintaro in 2004 ( ) Figure The response of grains per spike with nitrogen rate and application timing for Frame and H45 at Hart in a) 2003 and b) Symbols represent sowing ( ) mid-tillering ( ); 1 st node ( ); T1+T2+T3 (+); T2 + T3 ( ); T2 + T5 (Δ). (LSDs [0.05] for nitrogen rate x timing are ns for 2003, and 4.15 for 2004) Figure The relationship between head number and grains per spike at Hart in 2003 for H45 ( ) and Frame ( ), r = Figure The relationship between head number and grains per spike at Mintaro for Frame in 2003 ( ); H45 in 2003 ( ); Frame in 2004 ( ) and H45 in 2004 ( ) Figure The relationship between grain number and grains per spike for both varieties at Hart in 2003 ( ); Mintaro in 2003 ( ); Hart in 2004 ( ) and Mintaro in 2004 ( ) Figure Relationship between grains per spike and light interception measured at anthesis for both varieties at Hart in 2003 ( ); Mintaro in 2003 ( ); Hart in 2004 ( ) and Mintaro in 2004 ( ) Figure The relationship between grain number and spikelets per head for both varieties at Hart in 2003 ( ); Mintaro in 2003 ( ) and Mintaro in 2004 ( ) Figure The relationship between grain number and grains per spike for both varieties at Hart in 2003 ( ); Mintaro in 2003 ( ); Hart in 2004 ( ) and Mintaro in 2004 ( ) Figure The relationship between grains per spikelet and head number for both varieties at Hart in 2003 ( ); Mintaro in 2003 ( ) and Mintaro in 2004 ( )

17 xvii LIST OF TABLES Table 2.1a. A selection of rain-fed trials on wheat conducted in South Australia between 1981 and Table 2.1b. A selection of rain-fed trials on wheat conducted in South Australia between 1981 and Table 2.2a. A selection of rainfed trials on wheat conducted outside South Australia between 1981 and Table 2.2b. A selection of rainfed trials on wheat conducted outside South Australia between 1981 and Table 2.3. Comparison of the effects of 1 mm of evapotranspiration (ET) before or after anthesis on biomass production of low (no applied N) and high (80 kg N ha -1 ) wheat crops (Angus and van Herwaarden 2001)...27 Table 3.1. Soil descriptions for three sample horizons for each trial site...38 Table 3.3. Annual rainfall for 2003 and 2004 and the long term average rainfall (mm) for Hart and Mintaro sites, including the growing season (April- October) sub-total and long-term averages Table 3.2. Pre-sowing soil mineral nitrogen (kg/ha) (nitrate + ammonium) and total soil moisture (mm) for each site Table 3.4. Average monthly maximum and minimum temperatures ( C) for 2003, 2004 and the long term average at Hart and Mintaro trial sites...40 Table 3.5. Rain and temperature conditions compared to long term averages for key stages of the growing season at Hart and Mintaro during 2003 and Table 3.6 The nitrogen treatments used at Hart The values are the percentage of the nitrogen applied at the specified growth stage and either 40 kg N/ha or 80 kg N/ha were applied at each treatment...46 Table 3.7. The nitrogen treatments used at Mintaro The values are the percentage of the nitrogen applied at the specified growth stage and either 100 kg N/ha or 200 kg N/ha were applied at each treatment...46 Table 3.8. Additional nitrogen treatments included at Hart The values are the percentage of the nitrogen applied at the specified growth stage and either 40 kg N/ha or 80 kg N/ha were applied at each treatment...47 Table 3.9. The additional nitrogen treatments included at Mintaro The values are the percentage of the nitrogen applied at the specified growth stage and either 100 kg N/ha or 200 kg N/ha were applied at each treatment....47

18 xviii Table Nitrogen application timing, days after sowing (DAS) and corresponding crop growth stages at Hart...48 Table Nitrogen application timing, days after sowing (DAS) and corresponding crop growth stages at Mintaro Table 4.1. Grain yield (t/ha) and nitrogen rate (kg N/ha) for both Frame and H45 at Mintaro and Hart in 2003 and Values for the control treatments are 2.24 and 2.68 for 2003 and 2004 at Mintaro, and 2.38 and 0.73 for 2003 and 2004 at Hart Table 4.2. Grain protein and nitrogen rate for both Frame and H45 at Mintaro and Hart in 2003 and Values for the control treatments are 8.5 and 7.0% for 2003 and 2004 at Mintaro, and 9.46 and 11.8% for 2003 and 2004 at Hart Table 4.3. Screenings (%) for variety (H45 and Frame) and nitrogen rate at Hart and Mintaro in 2003 and Table 4.4. Harvest index (%) for variety and nitrogen rate at Hart and Mintaro in 2003 and Table 4.5. Nitrogen harvest index (%) for nitrogen rate and timing at Mintaro in 2003 and Values are an average of Frame and H45. Values for the control treatments were 78.2 in 2003 and 71.5 in 2004 at Mintaro...65 Table 4.6. Agronomic efficiency (kg/kg) for nitrogen timing and rate (kg N/ha) at Hart and Mintaro in 2003 and The values are an average of Frame and H Table 4.7. The agronomic efficiency (kg/kg) of Frame and H45 wheat at 40 kg N/ha with nitrogen timing, at Hart in Table 4.8. Apparent recovery (%) and nitrogen fertiliser application timing, averaged for variety and nitrogen rate at Hart and Mintaro in 2003 and Table 4.9. Physiological efficiency (kg/kg) for nitrogen timing and rate at Hart and Mintaro in 2003 and The values are an average of Frame and H Table 7a. Trial results for nitrogen management trials 1 conducted across Southern Australia between 2002 and 2008 in wheat Table 7b. Trial results for nitrogen management trials 1 conducted across Southern Australia between 2002 and 2008 in wheat...152

19 xix LIST OF ABBREVIATIONS AE ANOVA AR DAS ET GAI HI LAD LAI LI LSD NHI NUE PE WSC WUE Agronomic efficiency Analysis of variance Apparent recovery Days after sowing Evapotranspiration Green area index Harvest index Leaf area duration Leaf area index Light interception Least significant difference Nitrogen harvest index Nitrogen use efficiency Physiological efficiency Water soluble carbohydrate Water use efficiency

20 1 CHAPTER 1 1. General introduction The cropping regions of southern Australia experience a typically Mediterranean type climate, with cool, wet winters and hot dry summers (Aschmann 1973). Average annual rainfall in these regions is generally between mm and the growing season is constrained by the onset of autumn rains, frost at flowering and terminal drought in late spring. The amount and timing of rainfall during the growing season varies considerably such that droughts are likely to occur prior to anthesis and the season will always end in drought. Until the 1990s ley farming was the predominant farming system in which high density legume based pastures provided most of the nitrogen inputs for the following cereal crops. Inputs of nitrogen fertiliser were very low (McDonald 1989). However, over the past 15 years the decline of the ley farming system and the intensification of cropping in southern Australia have resulted in increased use of nitrogen fertiliser (Angus 2001). A large proportion of a cereal crop s nitrogen is usually applied early in the season, either just before or at sowing, or during the first 6-8 weeks after sowing (McDonald 1989). This practice was aided by the development of tillage equipment with narrow sowing tynes and deep banding of fertiliser, which allowed the fertiliser to be applied below the seed during sowing. The purpose of this nitrogen input is to improve early crop vigour, increase weed competition, and ultimately to improve grain yield and grain protein content. However, the time of greatest crop nitrogen demand for wheat is normally during the stem-elongation

21 2 phase when the crop is accumulating dry matter most rapidly, while demand for nitrogen prior to this is small (Ortiz-Monasterio; Angus 2001; Lopez-Bellido et al. 2005). In the dryland farming systems of southern Australia, there is a strong interaction between available moisture and response to nitrogen that is associated with the trade-off between the need to produce sufficient biomass to establish a high yield potential and the need to have adequate reserves of soil moisture at critical periods of growth later in the growing season (Fischer 1979; Sadras 2002). Haying off is a term used to describe cereal crops that senesce rapidly in conditions of high soil nitrogen and post-anthesis drought (van Herwaarden et al. 1998b). The challenge is to identify the optimum rate of nitrogen to be applied at planting and later at stem elongation that will result in the optimal combination of light interception (driven by the leaf area index, LAI) and biomass production (Ortiz- Monasterio 1999). A major determinant of the response to nitrogen fertiliser will be the weather conditions during spring, which will influence the occurrence and severity of the crop water deficit. Uncontrolled vegetative growth associated with high levels of soil nitrogen can lead to high levels of water stress in spring if rainfall is low, leading to low or negative responses to nitrogen. Controlling biomass production by crop management can help reduce this risk. This seasonal uncertainty and its associated economic risk has been a major factor affecting nitrogen fertiliser practices in Australia (McDonald 1989).

22 3 Experience overseas has shown that nitrogen can be used successfully to manipulate the wheat crop canopy. Studies in the high rainfall areas ( mm) of southern Australia have indicated that nitrogen fertiliser may be used more effectively if more nitrogen is applied post emergent and less at or before sowing. There are also some environmental advantages connected with this improved use of nitrogen, reducing the potential for losses by less leaching of nitrogen. In high input cropping systems, such as wheat production on the South Island of New Zealand or irrigated wheat production in Mexico, very little or no nitrogen is applied at sowing. In such situations crops experience an initial period of nitrogen deficiency until the nitrogen is applied at tillering. Compared to applying most of the nitrogen at sowing, these crops produce similar or higher grain yields, better grain protein concentrations and have increased nitrogen use efficiencies (FAR 2002). This tactical approach to nitrogen application is often described as canopy management. That is, managing the green surface area of the crop in order to optimise nitrogen use efficiency and profitability (HGCA 1998). While it is clear that nitrogen timing and rate is a key component of successful management, it is essential that it is considered in conjunction with the availability of soil moisture and soil nitrogen, as well as planting date, planting rate and cultivar. In southern Australia, deferred or split applications of nitrogen fertiliser have been trialed for many years, with many positive results. Deferred and split applications have proved to be efficient strategies, particularly as they have allowed for

23 4 assessments of early seasonal conditions, crop vigour, colour and nitrogen status before nitrogen fertiliser use was contemplated (Elliott et al. 1985). However, alternative nitrogen application strategies such as post emergent nitrogen have not been widely adopted because of the ease and affordability of applying more nitrogen than is needed at or before sowing (Raun and Johnson 1999). Despite considerable research conducted on nitrogen, and the subsequent development of guidelines and monitoring tools, much potential still exists in Mediterranean environments to improve the effective use of nitrogen. Improvements in nitrogen management has the ability to increase potential yields, particularly in intensive production systems, where water use efficiency is low, or it can provide an opportunity to increase grain yields in systems where water use efficiency is already high. In the medium to high rainfall areas of southern Australia ( mm) the use of high nitrogen rates at seeding may be less efficient than delayed or split applications and could be detrimental to crop yield and quality by reducing water use efficiencies and/or causing haying off. New ideas for nitrogen management suggest that nitrogen applied at seeding does not allow for the full use of nitrogen and its capacity to control the development and size of a crop. There is potential to increase wheat yields and profitability through improvements in nitrogen use efficiency in the medium and high rainfall regions in the state of South Australia. Thus the overarching hypothesis for the study reported here is that manipulating wheat canopy development can improve nitrogen use efficiency in

24 5 the medium-high rainfall areas of southern Australia. To address this hypothesis field experiments were conducted in the two successive seasons, 2003 and 2004 at two sites in South Australia with different annual rainfall (medium rainfall site at Hart and high rainfall site at Mintaro), to examine the effect of post sowing applications of nitrogen fertiliser applications to wheat. In particular the study recorded here has the following specific objectives. 1. to use field experiments at medium and high rainfall sites to establish if improved nitrogen use efficiency can be achieved 2. to examine the effects of different methods of nitrogen application, timing and rates will manipulate the canopy of wheat to improve: i. the balance between canopy size and light interception ii. the balance between canopy size and crop evapotranspiration iii. total plant nitrogen uptake

25 6 CHAPTER 2 2. Review of literature 2.1 Introduction Improvements in the efficiency of nitrogen fertiliser usage through agronomic or genetic means could have a substantial beneficial impact on the profitability of wheat in southern Australia. Targeting nitrogen fertiliser applications in response to crop growth and seasonal conditions might contribute to increased nitrogen use efficiency (NUE, defined in Section 2.3.1). Improvements in NUE will be discussed in relation to environmental constraints and agronomic variables. Throughout this thesis it is proposed that NUE can be enhanced by increasing crop nitrogen uptake, maintaining and extending leaf area duration at key yielddetermining stages of development, and by improvements in the capture and efficient use of water. An aim of crop management in rainfed systems is to optimise the balance between water and nitrogen use efficiency, which largely centres on managing the crop canopy. This can be achieved by manipulating the amount and the timing of applications of nitrogen fertiliser. The purpose of this review of literature is to develop the concept that NUE can be improved by manipulating the canopy development of dryland wheat.

26 7 2.2 Nitrogen use in southern Australia The role of nitrogen fertiliser Nitrogen is the key element in achieving consistently high yields in cereals as it affects numerous processes in the plant/soil system and within the plant (Delogue et al. 1998). Nitrogen is essential for plant growth and plays a central role in plant biochemistry as a vital constituent of cell walls, cytoplasmic proteins, nucleic acids, and chlorophyll (Hay and Walker 1989). The main sources of plant nitrogen are nitrate and ammonium, principally taken up from the soil solution. The time of greatest demand for nitrogen by winter cereals is during the stemelongation phase when the crop is growing fastest, while early demand for nitrogen is small (Angus 2001; Lopez-Bellido et al. 2005) (Figure 2.1). Nitrogen has a dominant role in dry matter formation and accumulation (Delogue et al. 1998): the consequences of insufficient nitrogen in seedlings may be reduced tillering and loss of soil water from soil evaporation, while the consequences of excessive seedling nitrogen may be lodging, foliar diseases and high early water use leading to haying off (Angus 2001). Figure 2.1. Seasonal pattern of nitrogen demand for wheat crops of high and low nitrogen status (Angus 2001).

27 8 The grain yield of a wheat crop is strongly associated with crop growth rate in the immediate pre-anthesis period (Fischer 1979). Maintaining a high rate of photosynthesis during this period is therefore important to achieving high yields, water and nitrogen use efficiency. When water availability is not the overriding limitation to growth, the crop growth rate depends on the amount of light intercepted by the crop and the efficiency with which it utilises the intercepted radiation. Nitrogen in a wheat crop will mainly affect: (i) leaf expansion, leaf area development and light interception; and (ii) leaf nitrogen concentration, which affects radiation use efficiency (Angus 1995; Ortiz-Monasterio 1999). Hence, it is important for soil nitrogen supply to be high at tillering, stem elongation, booting, heading and grain filling, as the plant requires a greater amount for the development and growth of its reproductive organs and for an enhanced and high accumulation of proteins in the kernel (Delogue et al. 1998) Nitrogen fertiliser use in southern Australia In Australia, the consumption of nitrogen fertiliser in agriculture has increased steadily through the 1990 s (Figure 2.2), and fertiliser has steadily replaced biological fixation by pasture legumes as a source of nitrogen (Angus 2001; Evans et al. 2003). An increase in the frequency of cereal cropping, retention of crop residues, harvesting of pasture and grain legumes and decrease in area of legume based pastures have contributed to this increase (Ladd and Amato 1986). Drought, reduced grain yields and tight financial margins have caused nitrogen consumption to stabilise at about 1 million t, with values being even lower than this figure during the and growing seasons. Changes in nitrogen fertiliser

28 9 strategies which have led to more efficient use of nitrogen fertiliser have also contributed to this reduction. Historically in southern Australia nitrogen fertiliser was usually applied early in the season, either just before or at sowing, or during the first 6-8 weeks after sowing (McDonald 1989). This practice was aided by the development of tillage equipment with narrow sowing tynes and deep banding of fertiliser, which allowed the fertiliser to be applied below the seed, during sowing. The increased use of diammonium phosphate or mono-ammonium phosphate also contributed to this practice. Topdressing of nitrogen was not a common practice and where it was employed low-biuret granular urea was the major form of nitrogen used. The experimental program for this study was established on knowledge up to 2004, with the full realisation that fertiliser practices have already evolved. Further developments in fertiliser practice mean that many growers are now applying less fertiliser at sowing and applying post-emergent urea based on the early paddock growth and likely predicted grain yield. The predominance of applications of nitrogen while the crop demand is low (Figure 2.1) is risky, as significant losses of nitrogen can occur. Also, large amounts of nitrogen applied early can promote vegetative growth at the expense of yield due to haying-off (van Herwaarden et al. 1998b; Angus and van Herwaarden 2001). Practices can be tailored to minimise this risk: for example in the UK 90% of the fields are top-dressed twice and 50% are top-dressed 3 times or more (Sylvester-Bradley 1993).

29 10 Nitrogen recommendations for South Australia, are based on a locally developed nitrogen balance (Ladd and Payne 1994) which takes into account residual soil nitrogen, crop yield potential and the target grain protein concentration. Recommendations on applications of nitrogen focus on nitrogen applied before or at sowing to target an expected yield at 11% to 12% protein. Where yield potentials are high, or when to the chance of leaching is high, split applications between seeding and tillering are recommended Grain yields in southern Australia Since 1970, Australian grain yields have steadily increased at 1.9% per year. This was associated with an intensification of cropping (Angus 2001), including an increased use of nitrogen fertiliser (Figure 2.2). However, since 2002 yields have been declining (Figure 2.3) (ABARE 2009). This is evident in South Australia (Figure 2.4) with different regions showing similar trends to the seasonal conditions. To cope with this sequence of poor seasons and a decline in grain Figure 2.2. Total nitrogen consumption (kt) in Australia. Source (ABARE 2009).

30 11 yields Australian growers are now focussing on cost risk management and are striving for more efficient use of resources and inputs, including nitrogen fertiliser and water. Figure 2.3. Five year moving average grain yield (t/ha) for Australia between 1969/70 and 2007/08. (ABARE 2009). Figure 2.4. Trend in grain yield (t/ha) for the South East ( ), Mid- North and Yorke Peninsula ( ), Eyre Peninsula ( ) and state average ( ) for South Australia (ABARE 2009).

31 Nitrogen use efficiency Definition of nitrogen use efficiency Nitrogen use efficiency (NUE) is a universal description used to describe how well a crop has utilised applied nitrogen fertiliser. There are a number of different parameters used to describe NUE which may be useful to identify the pathways where nitrogen can be used more efficiently: Agronomic efficiency (AE) is an indicator of the ability of the plant to increase grain yield in response to applied nitrogen (Delogue et al. 1998), and reflects the overall efficiency with which applied nitrogen is used (Craswell and Godwin 1984). Poor AE might be due to the inability of the plant to recover nitrogen or the inability of the plant to utilise that nitrogen for growth and yield production (Craswell and Godwin 1984; Anderson 1985b). Apparent recovery (AR) is used to consider how well the crop has taken up the applied nitrogen. The efficiency of nitrogen uptake is related to soil fertility and nitrogen fertiliser use, and also depends on the ability of the crop to recover nitrogen from the soil (Anderson 1985a; Whitfield and Smith 1992). Physiological efficiency (PE) can be viewed as the efficiency with which crops utilise nitrogen already absorbed in the plant, for the synthesis of grain yield and protein (Craswell and Godwin 1984; Ortiz-Monasterio 1999), and therefore is affected by environmental stresses and the plant genotype (McDonald 1989).

32 13 Both PE and AR of fertiliser are derived on the assumption that unfertilised and fertilised crops take up similar amounts of soil nitrogen (McDonald 1989; Wuest and Cassman 1992) hence the term apparent recovery (Craswell and Godwin 1984). Although it can be argued that applications of nitrogen fertiliser frequently lead to a larger and more vigorous root system that can recover more non-fertiliser nitrogen from the soil than can the root system of an unfertilised crop (McDonald 1989). This research review will focus only on wheat, spring applications of fertiliser and rain fed crops Current NUE values Worldwide, nitrogen fertiliser recovery tends to be low, being about 33% for wheat production (Raun and Johnson 1999). In the UK recovery of fertiliser nitrogen by most crops averages at only 60% and varies from 40 to 90% (Sylvester-Bradley et al. 2001). Results like this are common and are consistent with worldwide NUE information (Raun and Johnson 1999) (Tables 2.1 and 2.2). The NUE of fertiliser applications in Australia is not consistently different compared to overseas values. For dryland wheat crops in Australia nitrogen recovery is about 40-50%, and analysis of efficiency show that fertiliser recoveries are often generally less than 50% (McDonald 1989; Fillery and McInnes 1992; van Herwaarden et al. 1998a) (Tables 2.1 and 2.2). Worldwide, AE for wheat averages at about kg/kg, while in Australia 33 kg/kg is the highest AE, with the average for South Australia being kg/kg

33 14 (Tables 2.1 and 2.2). Russell (1967) measured an average AE of 7.2 kg grain yield/kg nitrogen fertiliser applied at 25 kg N/ha and 5.1 kg/kg N at 50 kg N/ha over numerous trials in South Australia (Table 2.1). While in 29 responsive wheat experiments, maximum AE values ranged from 7.6 to 42.8 kg grain/kg applied N, and averaged 23 ± 10 kg/kg between 1990 and 1992 (Elliott et al. 1993). PE is not measured as often, but in Australia and around the world it varies between 30 and 60 kg/kg. For dryland wheat crops in southern Australia PE was measured at kg/kg (Angus and Fischer 1991; McDonald 1992) (Tables 2.1 and 2.2). Throughout the world there is a large variation in NUE values. In southern Australia based on experimental data over an extended period of time, average values can be considered to be 15 kg/kg for AE, 50% for AR and 40 kg/kg for PE The physiological limit to NUE Plants have a physiological limit for growth and resource use efficiency. Understanding these limits is important for benchmarking current levels of productivity and to measure the potential for further and realistic improvements. For instance, the harvest index of modern semi-dwarf wheat varieties is approaching 50% and the theoretical limit is estimated to be 60% (Passioura 1976; Fischer 1979).