A time-lapse imaging platform for quantification of soil crack development due to simulated root water uptake

Plants are major drivers of soil structure dynamics. Root growth creates new macropores and provides essential carbon to soil, while root water uptake may induce crack formation around roots. Cracks can facilitate root growth as they provide pathways of least resistance and improve water infiltration and soil aeration. Due to the lack of suitable quantification methods, knowledge on the effects of root water uptake on soil crack formation remains limited. In the current study, we developed a time-lapse imaging platform that allows i) simulating root water uptake through localized soil drying and ii) quantifying the development of two-dimensional crack networks. Customized soil boxes that were 50 mm wide, 55 mm high and 5 mm deep were designed. Artificial roots made of dialysis tubes were inserted into the soil boxes and polyethylene glycol solution was circulated through the tubes. This induced a gradient in osmotic potential at the contact area (150 mm²) between the soil and the dialysis tubes, resulting in controlled soil drying. Drying intensity was varied by using different polyethylene glycol concentrations. Experiments were conducted with three soils that were subjected to three drying intensities for 6.5 days. We developed a time-lapse imaging system to record soil crack formation at two-minute intervals in twelve samples simultaneously. Resulting crack networks were quantified with an automated image analysis pipeline. Across soils and drying intensities, crack network development slowed down after 24–48 h of soil drying. The extent and complexity of crack networks increased with drying intensity and crack networks were larger and more complex in the clay and clay loam soil than in the silt loam soil. Smaller and less complex crack networks were better connected than larger and more complex networks. These results demonstrate that the platform developed in this study is suitable to quantify crack network development in soil due to simulated root water uptake at high temporal resolution and high throughput. Thereby, it can provide information needed to improve our understanding on how plants modify soil structure.