Development of an automated system for extraction and quantification of soybean cyst nematode (SCN) eggs and cysts

Abstract

Soybean is the world’s foremost source of vegetable protein and oil. In order to meet the growing global soybean demand for feed and biodiesel, soybean production in the US will need to increase by 28 percent over the next decade. Addressing this demand necessitates increasing efficiency of soybean production, as increases in harvested area and yield per unit area alone are not viable solutions. The soybean cyst nematode (SCN) is the most devastating pathogen for US soybean, causing an estimated $1.5 billion in yield loss annually across the US. Diagnosis of SCN infestation in a field is currently performed with antiquated methods which yield very little actionable data. This thesis provides an alternative to these methods.

SCN infestation in a field tends to be highly localized. Most integrated pest management techniques require high resolution SCN infestation data to be effective at identifying or controlling SCN populations. The low throughput and high cost of existing quantification methods makes it virtually impossible to obtain such high resolution data. A novel design for an automated SCN extraction system is presented which (1) preserves spatial and temporal information associated with a soil sample, (2) is able to process multiple samples simultaneously using a novel apparatus, and (3) reduces the processing time and manual labor required for SCN egg extraction from soil. In addition, an automated image analysis technique is proposed to replace the current method of counting SCN eggs under a microscope. Using these improvements, a ten-fold increase in performance of SCN egg extraction and quantification is achieved.