A Suas Based High Throughput Data Acquisition and Processing Pipeline for Breeding Programs.

A sUAS-based high throughput data acquisition and processing pipeline for plant breeding programs

*Daljit Singh¹, Atena Haghighattalab², Kyle McGahee³, Mark Lucas⁴, Xu Wang⁴, Nan An⁵, Stephen Welch⁵, Dale Schinstock³ and Jesse Poland⁴

¹Interdepartmental Genetics, Kansas State University, Manhattan, KS, USA

²Department of Geography, Kansas State University, Manhattan, KS, USA ³Department of Nuclear and Mechanical Engineering, Kansas State University, Manhattan, KS, USA

⁴Department of Plant Pathology, Kansas State University, Manhattan, KS, USA ⁵Department of Agronomy, Kansas State University, Manhattan, KS, USA

Efficient and cost-effective phenotyping and genotyping tools are required to achieve desired genetic gains in breeding programs. Recent advancements in sequencing and informatics technologies have made it easier to genotype and analyze large number of individuals efficiently and cost-effectively. Although relatively slow, high throughput phenotyping (HTP) technologies are also catching momentum. Ease of deployment and cost-effectiveness are key considerations for an ideal HTP platform. Small Unmanned Aerial System (sUAS) is a promising tool to address the existing limitations in current HTP tools. A sUAS fitted with lightweight spectral sensors facilitate analysis of a large number of test plots in a high throughput and non-destructive fashion. A sUAS-based data acquisition protocol and a prototype data management pipeline were developed to enable HTP analysis of large breeding nurseries. The pipeline was tested on preliminary datasets collected from multiple sUAS flights over wheat nurseries in India, Mexico and Kansas, USA in 2014 and 2015. Approximately 100 datasets containing information from visible and near-infrared light spectrum were processed. Vegetation indices extracted from over 50,000 breeding plots are being evaluated to assess its relationship with crop agronomic and physiological characteristics. The current work aims to develop a robust, scalable and easily deployable sUAS workflow to enable efficient phenotyping analysis in an accelerated breeding context.