Maize yield potential: critical processes and simulation modeling in a high-yielding environment

Abstract

Understanding processes of maize ( Zea mays L.) growth and production of grain in high-yielding, irrigated conditions offers hope to understand yield potential in many other environments. In this study we investigated such processes at the plant level, and attempted to simulate maize yields at the field level and county level in the high yielding region of the High Plains of Texas. In addition, we used the normalized difference vegetation index (NDVI) from satellite data of year 2000 to update leaf area index for yield simulation in three counties. In the field study, we measured maize leaf area index (LAI), the fraction of photosynthetically active radiation intercepted (FIPAR), and the harvest index (HI) in irrigated plots near Dumas, Texas. The light extinction coefficient ( k) for Beer's law was calculated with the FIPAR and the LAI. The radiation use efficiency (RUE) was determined with sequential measurements of the fraction of photosynthetically active radiation (PAR) intercepted and biomass harvests. The RUE was 3.98 g of above-ground biomass per MJ of intercepted PAR in 1999 and 3.41 in 2000 for three sampling dates prior to silking. These values are 106 and 93% of the expected RUE values at the measured vapor pressure deficits, using a previously published response function. The mean k value was −0.46 in 1999 and −0.47 in 2000, similar to the expected value of −0.43 reported in the literature for this row spacing. The mean HI measured in 2000 was 0.52, similar to values of 0.53 and 0.54 in the literature. Application of these parameters to maize simulation with the Agriculture Land Management Alternatives with Numerical Assessment Criteria (ALMANAC) model for 13 center pivot irrigated fields near Dumas in 1999 provided simulations within 1.0 Mg ha −1 with a mean error of 0.03 Mg ha −1 and a mean square error of 0.10. For five years of grain yields reported for each of four counties in this region of Texas, ALMANAC simulations were within 5% of the mean measured yields. Introduction of PAR interception, based on the satellite-derived normalized difference vegetation index (NDVI), into ALMANAC resulted in slight increases in accuracy of yield prediction for two counties and a slight decrease in accuracy in one county for year 2000. Consistency in values of RUE, k, and HI in this study as compared with values reported in the literature will aid modelers simulating maize growth and grain yields in similar high-yielding, irrigated conditions.