资源环境科技发展态势分析平台

For the first time, we combine depth-specific soil information obtained from the quantitative inversion of ground-based multicoil electromagnetic induction data with the airborne hyperspectral vegetation mapping of 1 x 1-m pixels including Sun-induced fluorescence (F) to understand how subsurface structures drive above-surface plant performance. Hyperspectral data were processed to quantitative F and selected biophysical canopy maps, which are proxies for actual photosynthetic rates. These maps showed within-field spatial patterns, which were attributed to paleo-river channels buried at around 1-m depth. The soil structures at specific depths were identified by quantitative electromagnetic induction data inversions and confirmed by soil samples. Whereas the upper plowing layer showed minor correlation to the plant data, the deeper subsoil carrying vital plant resources correlated substantially. Linking depth-specific soil information with plant performance data may greatly improve our understanding and the modeling of soil-vegetation-atmosphere exchange processes.

Plain Language Summary Plants interact with soil. This is intuitive, although we know little about the subsurface structure because we cannot see it. At first glance, all soil may look the same, yet healthy plants can survive beside withered ones. We investigate the soil-plant interaction in an agricultural field situated in an area characterized by ancient (paleo-) river channels. These channels formed in sandy-gravelly material due to melting water after last glaciation, were then filled up with fine aeolien sediments, overlaid with soils up to 1 m thick, and are no longer visible at the surface. However, crops grow in meandering/braiding patterns that can be seen on satellite images, for example. To explain this, the subsurface structural geometry must be known. We combine ground-based electromagnetic induction data inversion results with airborne hyperspectral measurements to reveal the soil depths driving plant performance (photosynthetic activity and growth). Contrary to expectations, the deeper subsoil and not the plowing layer controls plant performance at the investigated site. Plants above the buried paleoriver channels find nutrients and water, whereas the surrounding plants in gravelly soil suffer, especially during drought. These results improve our understanding of soil-plant interaction, which may improve soil-vegetation-atmosphere exchange process modeling and harvest predictability.