Doctor rerum agriculturarum (Dr. rer. agr.)

der Humboldt-Universität zu Berlin

Präsident der Landwirtschaftlich-Gärtnerischen Fakultät:

Prof. Dr. J. Mlynek

Dekan der Landwirtschaftlich-Gärtnerischen Fakultät:

Prof. Dr. Dr. h. c. E. Lindemann

Dedicated to my wife Renate

Die Zukunft hat viele Namen.

Für die Schwachen ist sie das Unerreichbare.

Für die Furchtsamen ist sie das Unbekannte.

Für die Tapferen ist sie die Chance.

The future has many names.

For the weak it is the unattainable.

For the fearful it is the unknown.

For the brave it is the chance.

Victor Hugo (1802-1885)

1. Introduction

2. State of Art

2.1. Precision Farming in General

2.2. Site-Specific Plant Protection and Biomass Sensing

3. Objective

4. Methods and Materials

4.1. Measurement Principle

4.2. Mechanical Equations

4.3. Apparatus and Carrier

4.4. Software

4.5. Statistical Methods

4.6. Determination of Growth-Stages

4.7. Determination of Biomass as Fresh Mass and Dry Mass

4.8. Determination of Additional Plant Parameters

4.8.1. Determination of Plant Height

4.8.2. Determination of Crop Density

4.8.3. Determination of the Number of Hills

4.8.4. Determination of Fungi Infestation

4.9. Layout of Trials

4.9.1. Crops, Test Sites, Cultivation Methods and Climate

4.9.2. Optimisation Trials of the Biomass Sensor

4.9.3. Trials for Potentially Biasing Factors on the Biomass Measurement

4.9.4. Trials for Spatially Reduced Plant Protection

4.9.5. Spatial Crop Biomass Map

5. Results and Discussions

5.1. Biomass Sensor “Pendulum-Meter”

5.1.1. Measurement Principle

5.1.2. Mechanical Theory

5.1.3. Hardware

5.1.4. Pendulum Parameters to Optimise

5.1.5. Angle Force Relation

5.1.6. Discussion

5.2. Secondary Sensors, Data Recording and Processing

5.2.1. Slope Sensor

5.2.2. Speed Sensor

5.2.3. Plot Trigger

5.2.4. Recording and Processing of Measurement Data

5.2.5. Influence of Data-Reduction Method on Plot Angle of Deviation

5.2.6. Possible Biases in the Measurement Values

5.2.7. Discussion

5.3. Accuracy of Replicate of the Measurements

5.3.1. Winter Rye

5.3.2. Winter Wheat

5.3.3. Rice

5.3.4. Discussion

5.4. Dependency of Measured Angle on Biomass

5.4.1. Exemplified Growth-Stages

5.4.2. The Angle-Biomass-Relationship During Crop Growth

5.4.3. Earliest Possible Measurements with the Pendulum-Meter

5.4.4. Discussion

5.5. Dependency of the Measured Angle on Pendulum Parameters

5.5.1. The Combined Influence of the Parameters

5.5.2. The Influence of a Single Parameter on the Angle of Deviation

5.5.3. Discussion

5.6. Dependency of Measured Angle on Secondary Plant Parameters

5.6.1. Winter Rye

5.6.2. Winter Wheat

5.6.3. Rice

5.6.4. Discussion

5.7. Dependency of Measured Angle of Deviation on Carrier Velocity

5.8. The Influence of Potentially Biasing Factors on the Measurements

5.8.1. The Effect of Wind on the Measurements

5.8.2. The Effect of Water Height on Measurements in Irrigated Rice

5.8.3. The Influence of Weeds on Measured Angle

5.8.4. The Influence of Time and Variety on the Measurement

5.8.5. Discussion

5.9. Limits of Non-Destructiveness

5.10. Major Systemic and Random Errors

5.11. Obstacles

5.12. Trials for Spatially Reduced Plant Protection

5.12.1. Spatially Reduced Application of Plant Growth-Regulator in Winter Rye

5.12.2. Spatially Reduced Application of Fungicides in Winter Wheat

5.12.3. Tractor-Based On-Line Mapping of Cereal Crop Biomass

6. Conclusions

7. Summaries

7.1. Summary

7.2. Zusammenfassung

7.3. Summario Español

7.4. Resumo Português

- References
- List of Abbreviations
- Acknowledgements
Digital Appendix

Printed Appendix

Table 1: Publications on disk-meters and plate-meters.

Table 1 continued

Table 2: Cultivation of test sites for the optimisation trials in winter rye and winter wheat.

Table 3: Cultivation of test sites for the optimisation trials in irrigated rice.

Table 4: The self weight of the tested pendulum in rice and wheat / rye.

Table 5: Plot values for the three data reduction methods in 12 plots of the preliminary tests in winter rye, winter wheat and rice.

Table 6: The angle of deviation for each plot of 5 repeats, their standard deviation, coefficient of variation, and average of 12 plots.

Table 7: Standard deviation and coefficient of variation for the parameter optimisation trial in winter rye BBCH 39.

Table 8: Standard deviation and coefficient of variation for the parameter optimisation trial in winter wheat BBCH 39.

Table 9: Standard deviation and coefficient of variation for the parameter optimisation trial in rice BBCH 39, or DAT 42 respectively.

Table 10: Accuracy of biomass determination calculated for the parameter optimisation trial in winter rye BBCH 59.

Table 11: Accuracy of biomass determination calculated for the parameter optimisation trial in winter wheat BBCH 39.

Table 12: Accuracy of biomass determination calculated for the parameter optimisation trial in irrigated rice BBCH 39, DAT 42 respectively.

Table 13: Accuracy of biomass determination calculated for the parameter optimisation trial in winter rye BBCH 32.

Table 14: Accuracy of biomass determination calculated for the parameter optimisation trial in winter wheat BBCH 34.

Table 15: Accuracy of biomass determination calculated for the parameter optimisation trial in irrigated rice BBCH 25, DAT 28 respectively.

Table 16: Results of the multiple stepwise forward regression, calculated of 12 plots.

Table 17: Results of the multiple regression for each of the 12 plots in winter rye at BBCH 39, and the respective plot fresh mass FM.

Table 18: The dependency of the measured angle on the height of pivot point for each of the 12 plots, and the respective plot fresh mass FM.

Table 19: The dependency of the measured angle on the height of the cylindrical body for each of the 12 plots, and the respective plot fresh mass FM.

Table 20: The dependency of the measured angle on the mass of the pendulum for each of the 12 plots, and the respective plot fresh mass FM.

Table 21: Linear regression results in winter rye for two plant parameters versus the angle of deviation at different growth-stages.

Table 22: Linear regression results in winter wheat for the two plant parameters versus the angle of deviation at different growth-stages.

Table 23: Linear regression results in rice for various plant parameters versus the angle of deviation at different growth-stages.

Table 24: Results of the linear regression for the dependency of the angle of deviation on carrier speed in several plots of winter wheat BBCH 49 with different amounts of biomass.

Table 25: The measured angles in one way, the return way, and their respective difference.

Table 26: The angles of deviation measured in one way, the return way, and their respective difference for a stem inclination of 30°.

Table 27: Destruction limits for the pendulum-meter.

Table A 1: Cultivation of the winter rye field for the field trial 1999 for the site-specific application of growth-regulators and the test for the biasing effect of the year on the measurements.

Table A 2: Cultivation of the winter wheat field for the test of the biasing effect of the year and the cultivar on the measurements.

Table A 3: Cultivation of the rice field for the test of the biasing effect of the year, planting method, and the cultivar on the measurements.

Table A 4: Cultivation of the winter wheat field for the field mapping of the biomass with the prototype sensor.

Table A 5: Vibration time of the pendulum-meter for a full vibration, averaged from 10 replicates.

Table A 6: Results of the optimisation trials in winter rye at BBCH 32, measured with 75 Hz measurement frequency, 2.5 m s-1 velocity, the standard deviation and the coefficient of variation calculated from 5 identical replicates, the measured angle of deviation, and the accuracy of biomass determination in terms of linear (y = a + b·x) and square (y = a + b·x + c·x2) goodness of fit R2 and the standard error of estimate SE of the linear regression for the tested pendulum parameter.

Table A 7: Results of the optimisation trials in winter rye at BBCH 39.

Table A 8: Results of the optimisation trials in winter rye at BBCH 59.

Table A 9: Results of the optimisation trials in winter rye at BBCH 69.

Table A 10: Results of the optimisation trials in winter wheat at BBCH 34.

Table A 11: Results of the optimisation trials in winter wheat at BBCH 39.

Table A 12: Results of the optimisation trials in winter wheat at BBCH 59.

Table A 13: Results of the optimisation trials in winter wheat at BBCH 69.

Table A 14: Results of the optimisation trials in irrigated rice at BBCH 25, or DAT 28 respectively.

Table A 15: Results of the optimisation trials in irrigated rice at BBCH 39, respectively 42 DAT.

Table A 16: Results of the optimisation trials in irrigated rice at BBCH 49, respectively 62 DAT.

Table A 17: Results of the optimisation trials in irrigated rice at BBCH 65, respectively 80 DAT.

Table A 18: Linear regression equations for the change of the measured angle with increase of the height of pivot point hP by 1 m for 12 plots.

Table A 19: Linear regression equations for the change of the measured angle with increase of the height of cylindrical body hA0 by 1 m for 12 plots.

Table A 20: Linear regression equations for the change of the measured angle with increase of the mass of the pendulum m P by 1 kg for 12 plots.

Table A 21: Multiple regression equations for all 12 plots in winter wheat at BBCH 39 for the three independent variables: height of pivot point hP, height of cylindrical body hA0, and mass of pendulum m P.

Table A 22: Multiple regression equations for all 12 plots in rice at BBCH 39 for the three independent variables: height of pivot point hP, height of cylindrical body hA0, and mass of pendulum m P.

Figure 1: Research carrier.

Figure 2: Locations of test sites in Germany and the Philippines.

Figure 3: Precipitation and average daily temperature in 1998 at ATB, Germany.

Figure 4: Precipitation and average daily temperature in 1999 at ATB, Germany.

Figure 5: Precipitation and average daily temperature in 1998 / 1999 at IRRI, Philippines.

Figure 6: Layout of the parameter optimisation trials in winter rye with differences in crop height (red line) and indicated plots (black dotted lines).

Figure 7: Research carrier with fans for wind trials.

Figure 8: Advanced measurement principle of the pendulum-meter.

Figure 9: Force diagram of the pendulum-meter.

Figure 10: Bending types of cereals during measurements.

Figure 11: Construction of the pendulum-meter and its operating in rice.

Figure 12: Length cut of the potentiometer.

Figure 13: Angle versus resultant force relationship.

Figure 14: Slope sensor inside the electronic cabinet (red arrow).

Figure 15: Velocity sensor of the carrier for the optimisation trials.

Figure 16: Trigger used to separate the plots in the field.

Figure 17: Research carrier and signal transmission.

Figure 18: Screen shot of the laptop during biomass measurement.

Figure 19: Original angle of deviation versus average angle per plot.

Figure 20: Typical residual plot of the linear and square regressions in wheat and rice.

Figure 21: Residual plot of the linear and square regressions in rye.

Figure 22: Range of the standard error of estimate for the linear and square goodness of fit at different growth-stages of winter rye.

Figure 23: Range of the linear and square goodness’ of fit R2 in various growth-stages in winter rye.

Figure 24: Range of the linear and square goodness’ of fit R2 in various growth-stages in winter wheat.

Figure 25: Range of the linear and square goodness’ of fit R2 in various growth-stages in irrigated rice.

Figure 26: The influence of carrier velocity on the measured angle of deviation.

Figure 27: Linear regressions in 9 plots for the change of the measured angle of deviation at different carrier velocities.

Figure 28: Effect of wind speed of head wind on the measurements.

Figure 29: Effect of water height in irrigated rice on the biomass measurements.

Figure 30: Non-reduced measurement data of the pendulum-meter in winter wheat including and excluding a single thistle in the crop.

Figure 31: Measurements of the pendulum-meter during various times of two consecutive days in two plots with different growth pattern.

Figure 32: Square regressions for two growth-stages in winter rye 1998.

Figure 33: Pooled square regression in rice for dry season, transplanted IR 64, and wet season, direct-seeded IR 72.

Figure 34: Angle versus biomass relation for 12 plots of natural biomass and of one plot randomly thinned out in increments of 25 stems.

Figure 35: The pendulum-meter measurements of the site-specific and the uniform tramlines in the trial for a reduced application of plant growth-regulators in winter rye, the respective change between sites in the site-specific tramline, and additional results.

Figure 36: Test tramlines for the site-specific trial of plant growth-regulator and lodging area in the control.

Figure 37: Some of the processes determining lodging in cereals (source: GRACE 1977).

Figure 38: The pendulum-meter measurements of the site-specific and the uniform tramlines in the trial for a reduced application of fungicides in winter wheat, the respective change between sites in the site-specific tramline, and additional results.

Figure 39: The angle of deviation in the uniformily sprayed tramline at BBCH 41 and the percentage of powdery mildew infected leaf area at BBCH 75.

Figure 40: On-line recorded measurements with the tractor-based pendulum-meter.

Figure 41: Interpolated map of the measured angle of deviation.

Figure A 1: Listing of days after transplanting DAT, the names for the growth-stages, and Zadoks code and BBCH code for the growth-stages in cereals.

Figure A 2: Biomass differences in winter wheat.

Figure A 3: Biomass differences in irrigated rice.

Figure A 4: Circuitry between sensors and electronic cabinet.

Figure A 5: Circuitry between electronic cabinet and computer.

Figure A 6: Relationship between the angle of deviation and the horizontal force component of the resultant force.

Figure A 7: Relationship between the angle of deviation and the vertical force component of the resultant force

Figure A 8: Linear relationship between the angle of deviation and the carrier speed for various growth-stages in rice, winter wheat, and winter rye.

Figure A 9: Linear regressions for four growth-stages in winter wheat 1999, measured with 0.8 m hP, 0.3 m hA0, 1 kg m P, 75 Hz frequency, and 2.5 m s-1.

Figure A 10: Linear regressions for four growth-stages in irrigated rice, measured with 0.4 m hP, 0.1 m hA0, 0.497 kg m P, 75 Hz frequency, and 2.5 m s-1.