Soil structure formation along an agricultural chronosequence

Highlights

• X-ray μCT disentangles effect of tillage and plant roots on soil structure over time.

• A new segmentation method allowed for quantification of biopore-network

• A maximum biopore density (18 cm cm⁻³) was already reached after 6 years 18 cm cm⁻³.

• Tillage led to total different macropore characteristics.

Abstract

During soil formation, the interaction of different biota (plants, soil fauna, microbes) with weathered mineral material shapes unique structures depending on the parental material and the site specific climatic conditions. While many of these interactions are known, the relative importance of the different biota is difficult to unravel and therefore difficult to quantify. Biological soil structure formation is often superimposed by soil management and swell-shrink dynamics, making it even more difficult to derive mechanistic understanding.

We here explore soil structure formation within a “space-for-time” chronosequence in the Rhenish lignite mining area. Loess material from a depth of 4–10 m has been used for reclamation in a standardized procedure for 24 years.

Changes in soil pore system are characterized by properties such as connectivity (Euler number) and pore size distribution using undisturbed soil columns with a diameter of 10 cm. They were taken from two different depths (0–20 cm and 40–60 cm) at different sites ranging in age from 0 to 24 years. X-ray CT is used for scanning the original columns as well as undisturbed subsamples of 3 and 0.7 cm diameter. This hierarchical sampling scheme was developed to overcome the trade-off between sample size and resolution.

For the first time also information on the development of biopores could be measured by separating them from other structural pores based on their unique shape. The data were complemented by destructive sampling and determination of root length with WinRHIZO to give an estimate of how many biopores are filled with roots. Furthermore HYPROP measurements of water retention curves were conducted and showed a general agreement with the image-derived pore size distribution merged across three scales.

An increase in biopore density throughout year zero to year 12, in particular in 40–60 cm soil depth, was observed. The biopore length densities of approximately 17 cm/cm³ obtained in year 12 was similar to the one measured in year 24, suggesting that equilibrium was reached. Only about 10% of these biopores were filled with roots. In the topsoil (0–20 cm) the equilibrium value in biopore density is already reached after six years due to a higher root length density. Ploughing lead to higher mean pore size and to lower connectivity compared to the well-connected, very stable pore network in 40–60 cm depth.

This study shows how fast plant roots create a stable and connected biopore system and how this is disrupted by soil tillage, which produces completely contrasting pore characteristics.

Introduction

Soil structure is a prerequisite for the functioning of soil and thus its ability to support life of plants and animals. It controls various important soil properties and processes such as soil water conductivity and retention, gaseous exchanges and erosion. In addition, soil organic matter and nutrient dynamics, root penetration and crop yield are also strongly influenced by soil structure (Bronick and Lal, 2005; Rabot et al., 2018).

Soil structure changes constantly during soil formation, i.e. it is shaped by the interaction of different biota (plants, soil fauna, microbes) with weathered mineral material, and is depending on the parental material and the site specific climatic conditions. The relative contribution of the different biota to soil structure formation is difficult to disentangle, in particular in managed systems. Growing roots or burrowing earthworms reorganize the spatial arrangements of soil particles as they align individual minerals of various types and sizes and organic substances and may compact the soil along the biopores they form (Bruand et al., 1996; Kautz, 2015). Biopores can extend along the whole soil profile, are cylindrical in shape, show a low tortuosity and high vertical continuity. Thus, they significantly affect infiltration and preferential flow phenomena (Koestel and Larsbo, 2014; Luo et al., 2010; Naveed et al., 2013; Rasse et al., 2000; Wuest, 2001). It has been suggested, that in dense soils and with increasing depth, root growth depends more on an existing pore network (Gao et al., 2016), which results in an intimate relation between old biopores, new root growth and water extraction (Stirzaker et al., 1996).

While vegetation and earthworms have the ability to directly change the pore system and deliver carbon sources for soil organisms, microbes are responsible for the vast majority of turnover processes and, therefore, have been described as “soil architects” (Ramirez et al., 2014). By secretion of extracellular polymeric substances (EPS), microbes modulate the pore wall surface, which leads to the formation of habitat patches at the micro scale (Colica et al., 2014). Together with plant polysaccharides, EPS forms the “glue” for the stabilisation of mineral particles in soil and therefore stabilises soil structure (Totsche et al., 2018; Watteau et al., 2006).

In addition to biota, agricultural management, more specifically soil cultivation, may have a strong impact on soil structure. Increasing macroporosity with little change in micro- and mesoporosity has been reported for conventionally managed soils (Ambert-Sanchez et al., 2016; Kay and VandenBygaart, 2002; Kravchenko et al., 2011; Pires et al., 2017; Rasmussen, 1999). Soils cultivated without tillage show lower air capacity and higher bulk densities due to lower macroporosity. However, in no-till systems biotic factors dominate, e.g. higher amounts of earthworms occur (Jarvis et al., 2017; Rasmussen, 1999; Schlüter et al., 2018b).

Soil structure is typically considered as the spatial arrangement of solids and voids across various scales. Therefore, soil structure can be described both from the solid phase perspective and from the pore space perspective, as these are complementary aspects (Rabot et al., 2018). From the solid phase perspective bulk density or aggregate stability and aggregate size distribution are often used as an indicator for soil structure. However, describing aggregates does not seem to be the most suitable way to link soil structure with soil functions and processes (Rabot et al., 2018). For example water flow and gas diffusion are both directly affected by the architecture of pores (Koestel et al., 2018; Naveed et al., 2013). Nowadays different imaging techniques allow for visualizing and describing the soil pore space of undisturbed samples directly. X-ray computed tomography (X-ray CT) as a non-invasive imaging method, has been increasingly used in recent years to describe pore systems and their dynamics using parameters such as connectivity, macroporosity and pore size distribution (Kravchenko et al., 2011; Kuka et al., 2013; Naveed et al., 2013; Pires et al., 2017; Schlüter et al., 2011; Schlüter et al., 2018b).

In X-ray CT-analysis the image resolution is limited by the sample size, with a fixed factor of 1000–2000 between the size of a voxel and the size of the sample depending on the X-ray detector hardware (Rabot et al., 2018). Therefore, small samples in the size range of a few centimetres must be used to describe changes in mesopore structure (here defined as pores <50 μm) (Kravchenko et al., 2011; Schlüter and Vogel, 2016). However, such small samples may not include the structure of soil in a representative way. To overcome this trade-off, we have extended the nested strategy from Schlüter et al. (2018b) to three different sample diameters, which allowed us to representatively describe changes in soil structure down to pore sizes of 5 μm.

In order to describe soil structure from the very beginning, we have chosen a “space-for-time” chronosequence approach. We extracted undisturbed samples from reclamation sites of ages up to 24 years within an open-cast lignite mining area. Samples were taken both in 0–20 cm (directly affected by tillage) and 40–60 cm (no direct tillage effect) depth in order to separate the influence of soil management that periodically disrupts the biopores formed during a year, from the uninterrupted, biopore formation accumulated over decades beneath it.

Our objectives are 1) to characterize the influence of plants on soil structure (e.g. the temporal development of biopores towards an equilibrium biopore density) and 2) to characterize how this process is interrupted by soil cultivation.