UAV and field spectrometer based remote sensing for maize phenotyping, varietal discrimination and yield forecasting.
Year: 2020
Type: dissertation
Abstract: Maize is the major staple food crop in the majority of Sub-Saharan African (SSA) countries. \nHowever, production statistics (croplands and yields) are rarely measured, and where they are \nrecorded, accuracy is poor because the statistics are updated through the farm survey method, \nwhich is error-prone and is time-consuming, and expensive. There is an urgent need to use \naffordable, accurate, timely, and readily accessible data collection and spatial analysis tools, \nincluding robust data extraction and processing techniques for precise yield forecasting for \ndecision support and early warning systems. Meeting Africa’s rising food demand, which is \ndriven by population growth and low productivity requires doubling the current production of \nmajor grain crops like maize by 2050. This requires innovative approaches and mechanisms that \nsupport accurate yield forecasting for early warning systems coupled with accelerated crop \ngenetic improvement. \nRecent advances in remote sensing and geographical information system (GIS) have enabled \ndetailed cropland mapping, spatial analysis of land suitability, crop type, and varietal \ndiscrimination, and ultimately grain yield forecasting in the developed world. However, \nalthough remote sensing and spatial analysis afforded us unprecedented opportunities for \ndetailed data collection, their application in maize in Africa is still limited. In Africa, the challenge \nof crop yield forecasting using remote sensing is a daunting task because agriculture is highly \nfragmented, cropland is spatially heterogeneous, and cropping systems are highly diverse and \nmosaic. The dearth of data on the application of remote sensing and GIS in crop yield forecasting \nand land suitability analysis is not only worrying but catastrophic to food security monitoring \nand early warning systems in a continent burdened with chronic food shortages. Furthermore, \naccelerated crop genetic improvement to increase yield and achieve better adaptation to climate \nchange is an issue of increasing urgency in order to satisfy the ever-increasing food demand. \nRecently, crop improvement programs are exploring the use of remotely sensed data that can be \nused cost-effectively for varietal evaluation and analysis in crop phenotyping, which currently \nremains a major bottleneck in crop genetic improvement. Yet studies on evaluation of maize varietal response to abiotic and biotic stresses found in the target production environments are limited. \nTherefore, the aim of this study was to model spatial land suitability for maize production using \nGIS and explore the potential use of field spectrometer and unmanned aerial vehicles (UAV) \nbased remotely sensed data in maize varietal discrimination, high-throughput phenotyping, and \nyield prediction. Firstly, an overview of major remote-sensing platforms and their applicability \nto estimating maize grain yield in the African agricultural context, including research challenges \nwas provided. Secondly, maize land suitability analysis using GIS and analytical hierarchical \nprocess (AHP) was performed in Zimbabwe. Finally, the utility of proximal and UAV-based \nremotely sensed data for maize phenotyping, varietal discrimination, and yield forecasting were \nexplored. \nThe results showed that the use of remote sensing data in estimating maize yield in the African \nagricultural systems is still limited and obtaining accurate and reliable maize yield estimates \nusing remotely sensed data remains a challenge due to the highly fragmented and spatially \nheterogeneous nature of the cropping systems. Our results underscored the urgent need to use \nsensors with high spatial, temporal and spectral resolution, coupled with appropriate \nclassification techniques and accurate ground truth data in estimating maize yield and its spatiotemporal \ndynamics in heterogeneous African agricultural landscapes for designing appropriate \nfood security interventions. In addition, using modern spatial analysis tools is effective in \nassessing land suitability for targeting location-specific interventions and can serve as a decision \nsupport tool for policymakers and land-use planners regarding maize production and varietal \nplacement. \nDiscriminating maize varieties using remotely sensed data is crucial for crop monitoring, high throughput \nphenotyping, and yield forecasting. Using proximal sensing, our study showed that \nmaize varietal discrimination is possible at certain phenological growth stages at the field level, \nwhich is crucial for yield forecasting and varietal phenotyping in crop improvement. In addition, \nthe use of proximal remote sensing data with appropriate pre-processing algorithms such as auto scaling and generalized least squares weighting significantly improved the discrimination ability \nof partial least square discriminant analysis, and identify optimal spectral bands for maize \nvarietal discrimination. Using proximal sensing was not only able to discriminate maize varieties \nbut also identified the ideal phenological stage for varietal discrimination. Flowering and onset \nof senescence appeared to be the most ideal stages for accurate varietal discrimination using our \ndata. \nIn this study, we also demonstrated the potential use of UAV-based remotely sensed data in \nmaize varietal phenotyping in crop improvement. Using multi-temporal UAV-derived \nmultispectral data and Random Forest (RF) algorithm, our study identified not only the optimal \nbands and indices but also the ideal growth stage for accurate varietal phenotyping under maize \nstreak virus (MSV) infection. The RF classifier selected green normalized difference vegetation \nindex (GNDVI), green Chlorophyll Index (CIgreen), Red-edge Chlorophyll Index (CIred-edge), \nand the Red band as the most important variables for classification. The results demonstrated \nthat spectral bands and vegetation indices measured at the vegetative stage are the most \nimportant for the classification of maize varietal response to MSV. Further analysis to predict \nMSV disease and grain yield using UAV-derived multispectral imaging data using multiple \nmodels showed that Red and NIR bands were frequently selected in most of the models that gave \nthe highest prediction precision for grain yield. Combining the NIR band with Red band \nimproved the explanatory power of the prediction models. This was also true with the selected \nindices. Thus, not all indices or bands measure the same aspect of biophysical parameters or crop \nproductivity, and combining them increased the joint predictive power, consequently increased \ncomplementarity. \nOverall, the study has demonstrated the potential use of spatial analysis tools in land suitability \nanalysis for maize production and the utility of remotely sensed data in maize varietal \ndiscrimination, phenotyping, and yield prediction. These results are useful for targeting location-specific \ninterventions for varietal placement and integrating UAV-based high-throughput \nphenotyping systems in crop genetic improvement to address continental food security, \nespecially as climate change accelerates.
Author: Walter Chivasa
Language: English
Subfield: Ecology
Field: Environmental Science
Domain: Physical Sciences
SDG: Zero hunger
Open Access status: green
OpenAlex ID: W3185652160