In-situ measuring and predicting dynamics of soil bulk density in a non-rigid soil as affected by tillage practices: Effects of soil subsidence and shrinkage

Non-rigid soils (e.g., Vertisols) present dynamics of bulk density (& rho;b) due to high shrinkage and swelling. However, the in-situ measurement and prediction of the dynamic of & rho;b in non-rigid soils are still great challenges. The objectives were to (1) evaluate the performance of the combined soil moisture and thermal property sensors in estimating in-situ & rho;b dynamics under different tillage practices, (2) and establish mathematic equations to predict the & rho;b dynamics associated with soil subsidence and shrink-swelling processes during wetting and drying cycles. The in-situ & rho;b monitoring and periodical intact soil core sampling were conducted in the 0-10 cm and 10-20 cm layers in a Vertisol under three tillage treatments, containing no-tillage (NT), rotary tillage (RT) and deep ploughing (DP). Our results showed that the dual-sensor combination provided accurate & rho;b estimates in the field over 2021-2022 year (R2 > 0.487, RMSE < 0.177 g cm-3), except for the early stage after deep tillage. The & rho;b dynamics in the 0-20 cm in the NT and the 10-20 cm layer in RT treatment were mainly caused by shrink-swelling. Whereas the & rho;b dynamics in the 0-10 cm and 10-20 cm in the DP and the 0-10 cm layer in RT treatment were predominantly determined by soil subsidence first and then shrink-swelling when the accumu-lative rainfall (Pt) reached 131.8 mm, 186.1 mm, and 79.3 mm, respectively. The & rho;b dynamics during soil subsidence were well-fitted by an exponential equation related to accumulative rainfall (R2 > 699, P < 0.01), while the & rho;b dynamics during shrink-swelling were well-fitted by a newly proposed SSC & rho;b equation derived from the Peng and Horn soil shrinkage model (R2 > 589, P < 0.05). Combined with the long-term monitored rainfall and soil moisture, The SSC & rho;b equation and the two-stage equation involving subsidence and SSC & rho;b exhibited good prediction of & rho;b dynamic from 2017 to 2022 (R2 > 0.453, RMSE < 0.070 g cm-3). The soil subsidence and shrink-swelling process accounted for 3.32%-12.5% and 2.84%-14.8% of the & rho;b variation in tilled non-rigid soils, respectively. Our results demonstrated that the dual-sensor combination can be applied for field & rho;b monitoring in non-rigid soil. The proposed two-stage equation has great potential for predicting the field dynamics of & rho;b.

Automated summary

This study evaluated the performance of combined soil moisture and thermal property sensors in estimating the dynamic of bulk density (ρb) in non-rigid soils under different tillage practices and established mathematic equations to predict ρb dynamics associated with soil subsidence and shrink-swelling processes. The results showed that the dual-sensor combination provided accurate ρb estimates and the proposed two-stage equation has great potential for predicting ρb dynamics.