Dynamic testing and imaging of living maize kernel moisture using Low-Field Nuclear Magnetic Resonance (LF-NMR)

Grain dehydration is a crucial aspect of the mechanized harvesting of maize. The high moisture of grains during harvest leads to a high percentage of damaged and mildew grains, and therefore, limiting the application of mechanized harvesting. Exploration of the moisture changes with the appropriate method is important for the development of inbred lines and hybrids with low grain moisture during harvesting. In this study, a popular planted hybrid Zhengdan958 (ZD958), and its parental inbred lines, Zheng58 (Z58) and Chang7-2 (C7-2), were used as models to test the live maize kernel moisture quantification and visualization using the Low-Field Nuclear Magnetic Resonance (LF-NMR). Self-pollination ears of ZD958, Z58, and C7-2 were harvested at 52 days after pollination, and then experienced three treatments, 1) naturally dehydration (D) under 26 ℃ environments for 5 days, samples were analyzed every 24 hours with LF-NMR; 2) hydration (H), grains from step1 were soaked in water at 26 ℃ for 9 hours, these samples were analyzed every 90 minutes with LF-NMR; 3) re-dehydration (T), repeat step 1 with grains from step 2, samples were analyzed every 24 hours with LF-NMR. For each analysis, 5 grains were used for moisture quantification with LF-NMR and 3 grains were used for grain moisture visualization with the MRI. Results showed that dehydration rate was negatively correlated with the grain moisture in both natural dehydration and re-dehydration among three materials. Under the same conditions, moisture reduction in T01-T02 was 25.13-27.69 percentage points, which was much higher than that of 20.93-21.94 percentage points, in D01-D02. Besides, significant differences were found in water loss among materials, water loss of C7-2 was significantly higher than that of Z58 and ZD958 in D01-D03 and T01-T03, while water loss of C7-2 was significantly lower than that of Z58 and ZD958 in D04-D06 and T04-T06. The visualization result showed a stronger moisture signal in the inner layer of endosperm than that in the outer layer of endosperm. In both D and T treatment, grain dehydration started from outside (endosperm) to inside (embryo). During H treatment, the moisture of the seed coat and navel increased rapidly, indicating that both seed coat and navel were important channels for water absorption. Because of the existence of grain oil which was mainly distributed in embryos and could be detected by MRI, embryos showed the strongest signals at all stages. Besides, the phase status of grain water was analyzed using transverse relaxation time (T2) of signal amplitude. Data showed that the T2 value of signal peaks decreased along with the water loss in either D treatment (D01-D06) or T treatment (T01-T06). By contrast, as water absorption went on in H treatment, the T2 value of signal peaks increased, demonstrating that phase status changes gradually along with both water absorption and dehydration treatment. Grains with low water content showed T2 value between 0.000 1-0.001 s, which meant tightly banding between water and other molecules, whole grains with high water content showed T2 value between 0.001-0.01 s (D01, T01), signifying loose banding or free from banding with other molecules. Overall, the results revealed that LF-NMR could be used to observe the continuous changes of maize grain moisture directly and accurately from three different perspectives, water content, visualization, and phase status. LF-NMR would have a high potential to be used as a powerful tool to evaluate water content and realize accurate single-kernel selection in maize breeding.