Relationship Between Field-Measured Agronomic Variables on Winter Wheat: LAI, CC, Biomass and NDVI.

Destructive measurements such as leaf area index (LAI) and biomass can be time and resource consuming; and faster, cheaper, and non-destructive estimates of these variables are warranted. Our objectives were to predict LAI and biomass using normalized difference vegetative index (NDVI) and percent green canopy cover (CC) for winter wheat (Triticum aestivum L.). We collected NDVI, LAI, CC, and biomass at bi-weekly intervals from two different studies encompassing 17 site-years in the central region of Oklahoma, through the planting years 2012 to 2014, for a total of 646 biomass and 218 LAI measurements. We developed regression models for measurements taken before or after the start of senescence, using 37% CC or the milk stage of grain development as senescence threshold. Inverse exponential models explained the relationship between biomass or LAI and NDVI or CC with a high level of predictability. A strong relationship between LAI and CC existed regardless of CC range (r²= 0.92^*** for measurements above and r² = 0.87^** for below 37% CC). However, the relationship between biomass and CC depended on CC and was strong for pre-senescence measurements (r²= 0.68^***) and non-existing for values below 37% CC (r²= 0.008^ns). Similarly, the relationship between biomass and NDVI was strong pre-senescence (r²= 0.76^***) and weak post-senescence (r²= 0.17^ns). LAI and NDVI showed similar trend, with a strong relationship before the milk stage (r²= 0.83^***) and a poor relationship afterwards (r²= 0.5^***). The poor relationships between biomass and NDVI or CC after senescence are likely resultant from the increase in biomass with the concomitant decrease in green leaf area during that period. Our results suggest that for wheat, CC can predict LAI throughout the growing season and NDVI can predict LAI before the grain milk stage. Additionally, CC and NDVI can predict wheat biomass during pre-senescence measurements.