Rütt, Marius Andreas: Decision analysis and hyperspectral imaging to support farmers in ornamental heather production. - Bonn, 2022. - Dissertation, Rheinische Friedrich-Wilhelms-Universität Bonn.
Online-Ausgabe in bonndoc: https://nbn-resolving.org/urn:nbn:de:hbz:5-64973
@phdthesis{handle:20.500.11811/9591,
urn: https://nbn-resolving.org/urn:nbn:de:hbz:5-64973,
author = {{Marius Andreas Rütt}},
title = {Decision analysis and hyperspectral imaging to support farmers in ornamental heather production},
school = {Rheinische Friedrich-Wilhelms-Universität Bonn},
year = 2022,
month = feb,

note = {Ornamental Heather (Calluna vulgaris) is an important crop for the farmers of Germany’s Lower Rhine region. These farmers produce more than 78% of the ornamental heather in Germany. Successful production of this crop requires management of a variety of fungal pathogens, including Botrytis cinerea, Glomerella cingulata, and Phytophthora spp., which threaten the economic success of the farms. Fungal pathogens can lead to sudden mass disease outbreak and reduce the attractive and healthy look of ornamental plants, which is critical at the market. It is nearly impossible to make a profit with heather plants that have disease symptoms. To prevent the occurrence of disease symptoms, farmers often apply frequent and intensive prophylactic fungicide applications. However, the future of heather cultivation will likely require new cultivation techniques without prophylactic spraying. This is partly because intensive pesticide applications can favor the development of pathogen resistance and because some of the existing plant protection product approvals are set to expire. Moreover, consumers favor products with environmentally friendly cultivation strategies and farmers do too. Therefore, farmers are looking for more sustainable and less pesticide-intensive crop management options. In this thesis I outline a successful approach to working together with farmers, stakeholders, and experts to understand and make forecasts about changes to these cultivation systems. Following holistic research techniques, I gathered the many uncertainties and risks and made scientifically supported recommendations for more sustainable production techniques. The chapters of this thesis will outline my process of understanding the complexity of ornamental heather production, generating probabilistic impact pathway models in direct collaboration with experts and farmers, developing methods to analyze the vitality of heathers using hyperspectral sensors, and generating forecasts to support decision-making and assessment of farmers’ individual risk preferences:
1. In chapter 2, I report the results of a model-based simulation of management options in heather production. A general reduction in prophylactic fungicide applications does not currently appear to be beneficial to farmers. In contrast, implementing a monitoring plan to monitor disease symptoms is likely to result in a positive net benefit. We conclude that more intensive visual monitoring of disease symptoms has the potential to optimize crop management in heather production.
2. In chapter 3, I present a method for hyperspectral analysis of ornamental plants and the potential of sensor-based monitoring of heather plants. We applied a trained Partial Least Squares Regression model on the spectral reflectance data collected from measured heather plants. The model classified plants into healthy and stressed with an accuracy of 98.1% and identified the most important wavelengths for the classification process. The method is promising for high-resolution measurements of ornamental plants and particularly well suited for small plant samples.
3. In chapter 4, I report the projected impact of different monitoring approaches on the profitability and on the expected utility of heather farmers. The results show that heather production is inherently risky. Financial benefits appear to be better with the intensive visual monitoring strategy, which is more preferred by risk-taking farmers who want to maximize profits and optimize their system. Risk-averse farmers, on the other hand, would rather stay with currently applied management. Sensor-based monitoring incurs a risk of financial losses that currently seems to be too high for application in the heather production system.
The collaborative research approaches outlined in the thesis could be widely applied for research into risks and uncertainties of decision-making in agricultural production systems. These processes could also be used by decision-makers and policy-makers working in the agricultural sector. The specific results of the model building processes and the resulting forecasts generated in this work have helped farmers and producers of ornamental plants who seek to implement changes to optimize their horticultural crop management, to assess applicability of new technologies, and to improve disease control strategies while considering individual risk preferences.},

url = {https://hdl.handle.net/20.500.11811/9591}
}

The following license files are associated with this item:

InCopyright