Proximal remote sensing of cotton growth habits in water use efficiency trials



Journal Title

Journal ISSN

Volume Title



Development of elite cotton germplasm is primarily a numbers game. Many lines are evaluated in order to identify elite candidates. This is time consuming and expensive. Remote sensing can be a tool in cutting resources and labor requirements for many of the assessments and could aid breeders in selections and advancements. The objectives of this research were to 1) evaluate the use of a Normalized Difference Red Edge (NDRE)-based growth inflection point (GIP) to identify when vegetative growth ceases during the season and identify the relationship between GIP and manual maturity assessments nodes above white flower (NAWF), percent open boll (POB) and upper and lower boll fraction, 2) identify the relationship between lint yield and individual vegetation indices at single assessment dates and to identify the relationship between final lint yield and temporal analysis of NDRE integrated over several dates, 3) identify a relationship between maturity and drought tolerance/avoidance as well as the influence of maturity and irrigation on boll distribution. Field trials took place in 2015 through 2017 in Lubbock, TX. Trials consisted of nine commercial cultivars, categorized into three maturity classifications, and evaluated under three irrigation treatments (20% evapotranspiration (ET), 40% ET, and 60% ET). Weekly reflectance measurements were taken from pin-head square to boll opening. Maturity assessment NAWF was taken once during peak bloom followed by POB taken at the end of the growing season, prior to application of harvest aids. Plant mapping was conducted prior to harvest on one square meter in each plot. Growth inflection point (GIP) was generated based on a quadratic fit of the NDRE growth curves. Growth Inflection Point was the number of heat units associated with the inflection point of the NDRE values during the season. Each year had environmental conditions that changed growth habits and maturity of the cultivars. Cultivar and irrigation affected maturity in all 3 years. The GIP had significant correlations with each maturity assessment, with the highest correlations found with NAWF (r2 from 0.38 to 0.88) by irrigation. In most cases, the relationship between NAWF and GIP was not cultivar specific, suggesting that GIP may be used across multiple cotton genotypes within multiple growing environments. The GIP method provides a method to more rapidly and objectively evaluate maturity characteristics of cotton cultivars, as well as the effects of management on these characteristics. The second objective evaluated remote sensing parameters including NDRE, NDVI, ∆T (difference between canopy temperature and air temperature), and plant height. Of these measurements, NDRE was most consistent in terms of coefficient of determination, r2, slope, and normality in predicting yield. From these findings, a temporal analysis was calculated as NDRE integrated over the season, namely NDRE-days. Statistically significant r2 values were identified for all the individual remote sensing measurements with the highest r2 values occurring around peak bloom. In 2015, a r2 value of 0.81 was identified between ∆T and lint yield, while in 2017, the highest r2 value with lint yield was with NDRE (r2=0.71). The temporal analysis of NDRE showed a significant relationship in both years of this study (p<0.0001; r2= 0.55 in 2015 and r2=0.71 in 2017). The relationship between NDRE-days and lint yield was not cultivar specific, indicating that this method could be used across multiple genotypes. This study presents a suitable method that breeders could use to efficiently evaluate plant growth and estimate yield for variety selections while cutting resource requirements. In 2015, early maturing cultivars tended to yield higher in the 20% ET than the later maturing cultivars due to their ability to set majority of their fruit on lower nodes and avoid late season stress. This observation was the inverse in the 60% ET where ample heat units and higher irrigation supported boll development on upper nodes. In 2016 the hotter and wetter conditions during boll production aided later maturing cultivars, with higher fruit retention and greater distal position fruit. The 2017 season was cooler and had more rainfall than the other two years, which equated to more advantageous conditions for the early maturing cultivars with lower set fruit. Understanding how cultivars respond to different environments can aid producers and breeders in cultivar selection based on irrigation capabilities and geographies.



Physiology, Remote sensing