|
" Modeling Yield Loss due to Soybean Sudden Death Syndrome at Different Spatial Scales "
Raza, Muhammad Mohsin
Leandro, Leonor F. S.
| Document Type
|
:
|
Latin Dissertation
|
| Language of Document
|
:
|
English
|
| Record Number
|
:
|
1106055
|
| Doc. No
|
:
|
TLpq2363990746
|
| Main Entry
|
:
|
Leandro, Leonor F. S.
|
|
|
:
|
Raza, Muhammad Mohsin
|
| Title & Author
|
:
|
Modeling Yield Loss due to Soybean Sudden Death Syndrome at Different Spatial Scales\ Raza, Muhammad MohsinLeandro, Leonor F. S.
|
| College
|
:
|
Iowa State University
|
| Date
|
:
|
2019
|
| student score
|
:
|
2019
|
| Degree
|
:
|
Ph.D.
|
| Page No
|
:
|
163
|
| Abstract
|
:
|
Sudden death syndrome (SDS), caused by Fusarium virguliforme (Fv), is one of the major yield-limiting diseases of soybean in the Midwest. In the U.S., annual yield losses due to SDS have ranged between 0.6 to 1.9 million metric tons during the years from 2006 to 2014, and monetary losses of up to 6.75 billion dollars have been attributed to SDS for the period of 1996 to 2016. Our understanding of the relationships between varying disease levels, pathogen density, and yield is still limited. The first objective in this study was to quantify the relationship between SDS disease intensity (both foliar and root rot), pathogen density (in roots and soil) and soybean yield (yield components and grain yield) at different spatial scales. Individual soybean plants (2018) and quadrats (2016–2018) were surveyed in commercial fields and experimental plots located in Iowa to monitor SDS and yield. In individual soybean plants, SDS foliar severity and root rot had a positive relationship with each other (0.1 < R2 < 0.53), and a negative relationship with soybean yield components (0.10 < R2 < 0.50). The relationship of Fv population in roots with foliar severity, root rot and yield components were inconsistent among locations. No significant relationship was observed between Fv populations in soil and any of the disease or yield measures in individual soybean plants. In soybean quadrats, however, SDS foliar intensity (severity and incidence) had a negative relationship with soybean yield (0.05 < R2 < 0.46). Unlike individual plants, Fv population in soil in soybean quadrats had a positive relationship with foliar SDS intensity in soybean quadrats (0.20 < R2 < 0.42). Prior research suggests that time of SDS foliar symptom onset (DOY) influences disease progress and soybean yield. However, no quantitative information is available to date about these relationships. The second objective in this study was to assess and quantify the impact of DOY of SDS foliar symptom onset on final SDS intensity, soybean yield, and yield components. DOY of SDS foliar symptom onset and progress of foliar SDS were recorded weekly on individual soybean plants (severity) and quadrats (incidence and severity). Beta regression analysis showed that DOY of SDS onset has a consistent and stable effect on final disease intensity, both at individual plant and quadrat levels. The slope of the relationship between time of SDS onset and final SDS severity was common across all field sites and years, except at one site. Weighted linear regression analysis revealed that SDS onset time explained 60 to 96% of the variation in number of pods, number of seeds and total seed weight in individual plants, and 88 to 97% of the variation in seed yield (kg/ha) in quadrats. Soybean yield damage functions (slopes) were 9.6–31.3 kg/ha per day, indicating that for each day SDS onset was delayed, soybean yield increased by 9.6–31.3 kg/ha in soybean quadrats. Effective management for SDS requires early and accurate detection in soybean fields. However, current scouting methods for SDS are time-consuming, labor-intensive, and often destructive. The development of alternative methods to monitor SDS in soybean fields is necessary. The third objective of this study was to detect SDS using high-resolution (3 m) satellite imagery in large soybean plots. Quadrats were marked in a soybean field experiment with different rotation treatments, located in Boone, Iowa. Canopy reflectance in the red, blue, green, and near-infrared (NIR) spectral bands, a calculated normalized difference vegetation index (NDVI), and crop rotation information were used to detect healthy and SDS-infected quadrats. Satellite images collected during the 2016, 2017, and 2018 soybean growing seasons were analyzed using the Random Forest classification algorithm. Results indicate that spectral features, when combined with ground-based information, can predict SDS risk in soybean plots with high accuracy, even before foliar symptoms develop. Accuracy of classification of healthy and diseased soybean quadrats was > 74%, and area under the receiver operating characteristic curve (AUROC) was > 70% in the three growing seasons. Results from this study improve our understanding of the relationship between SDS, pathogen population and soybean yield, and how the time of SDS foliar symptom onset impact on final disease intensity and soybean yield. This information can be used when developing SDS risk and yield loss predictive models. Finally, our findings highlight that high-resolution satellite imagery may be useful for detecting SDS in soybean fields. Future research is needed to determine if this technology may facilitate large-scale monitoring of SDS, and possibly other economically important soybean diseases, to guide recommendations for site-specific management in current and future seasons.
|
| Subject
|
:
|
Plant pathology
|
|
|
:
|
Remote sensing
|
| |