Interactions between soil organic matter dynamics and soil structure as affected by farm management

In the last century, agriculture has focussed primarily on attaining maximum yields of crop production with the use of large amounts of fertilizers, biocides and pesticides. Growing public awareness of the detrimental effects of modern agriculture on soil productivity, food quality and the environment, however, resulted in a demand for "sustainable land management". Soil processes play a crucial role in determining the sustainability of land management. In agricultural land, soil organic matter (SOM) content is a key indicator of soil quality and sustainability for its effects on physical, chemical and biological soil properties that regulate crop yields, nutrient emissions, leaching of agrochemicals and the susceptibility to soil erosion. More recently, the elevated atmospheric CO ₂ concentration and associated global warming receive much attention. Since SOM can act as a sink for atmospheric C, concern about the "greenhouse effect" provides a second reason to adopt management practices that increase SOM levels.

The beneficial effects of SOM are partly indirect via its influence on soil structure. SOM binds mineral particles into stable soil aggregates, which are important for aeration, water storage and rootability of the soil. In turn, stable aggregates can reduce SOM decomposition by physical protection of aggregate-occluded organic matter. To define management practices for sustainable land use and increased C sequestration, a better understanding of the interactions between SOM and soil structure, and how they are affected by management, is needed. Differences in organic matter inputs and soil disturbance in different management systems result in different processes of soil structure formation (biogenic versus physicogenic), which are reflected in different morphological characteristics. In general, however, SOM-soil structure interactions have been studied using physically fractionated soil separates, without regarding differences in structural morphology. This thesis was aimed to improve understanding of the mechanisms and consequences for SOM stabilization of biogenic aggregate formation as opposed to physicogenic aggregate formation in different management systems. Therefore, differences in soil structure morphology resulting from different aggregate formation processes were emphasized. To exclude interference of differences in soil type and texture, one soil series was selected. This soil series has developed in marine loam deposits and is located in the southwestern part of the Netherlands (Zeeland). (Chapter 1).

Chapter 2 describes a regional survey of land-use history and current SOM content in the considered soil series. SOM contents of 45 fields were determined and information about land use in the past 63 yr was provided by the farmers. Four land use types were distinghuished: conventional-arable, conventional-grass, organic-arable and organic-grass, resulting in a range of SOM contents from 17 to 88 g kg ^-1. Regression analysis showed that SOM content was increased by long-term grass, and to a lesser extent, by organic farming when compared to conventional arable farming.

For more detailed study of management effects on the interactions between SOM dynamics and soil structure, three long-term farming systems (>70 yr) were selected: a permanent pasture (PP), a conventional arable system (CA), and an organic arable system (OA). Due to the lack of soil tillage and arable cropping, the soil under permanent pasture represents optimal conditions for this particular soil type in terms of organic inputs, soil structure and biological activity. The differences in management resulted in a significant increase in (i) the impact of earthworms on soil structure, (ii) stable macroaggregation, and (iii) C and N mineralization in the order: CA < OA < PP (Chapter 3). However, as for conventional arable farming, organic arable farming had resulted in severe soil compaction, which negatively affects crop growth.

Chapter 4 describes an incubation experiment to determine C mineralization protection in undisturbed soil aggregates from the permanent pasture and the conventional arable system. The difference in respiration between undisturbed and crushed (<250mm) aggregates was used as an estimate of macroaggregate protected C. Under permanent pasture, C mineralization did not significantly increase due to crushing of the highly porous macroaggregates. In contrast, mineralization was considerably restricted in the macroaggregates from the conventional arable soil, which could be explained by a limited gas exchange due to the low porosity of the, dominantly physicogenic, macroaggregates. However, macroaggregates from the conventional arable soil were not very water-stable and are probably easily disrupted under field conditions, leading to rapid decomposition of the released SOM. The amount of C protected in microaggregates was estimated from the additional CO ₂ evolution due to grinding (< 50mm) of crushed soil material. The results suggested that microaggregate protected C was high compared to macroaggregate protected C, especially under permanent pasture.

To study management effects on SOM stabilization at the microaggregate scale, microaggregates were isolated from total soil samples and SOM distribution was determined using physical fractionation techniques (Chapter 5). Similar amounts of microaggregates (53-250mm) were isolated from permanent pasture and conventional and organic arable land. In all three systems, the main part of total soil organic C was present in these microaggregates. However, microaggregates under permanent pasture, especially in the top 10 cm, were more stable, contained a larger fraction of total soil C and more fine mineral particles, and had a higher C:N ratio than microaggregates in both arable soils. This indicated that microaggregates under permanent pasture more effectively protect SOM against decomposition than microaggregates in arable land. Differences in microaggregate characteristics between the two arable systems were mostly not significant. Microscopic analysis of undisturbed thin sections suggested that the high earthworm activity under permanent pasture plays an important role in the formation of stable microaggregates that are relatively enriched in C and fine mineral particles.

To study the impact of earthworms on the relation between soil structure and SOM dynamics in the three farming systems, aggregates were separated by hand into biogenic (worm casts) and physicogenic macroaggregates, based on morphological characteristics (Chapter 6). Small differences in total organic C content and C respiration were found between different aggregate types within each farming system. Quantification of particulate organic matter (POM) and microaggregates in thin sections, however, showed that biogenic aggregates of the permanent pasture contained distinctly larger amounts of POM and microaggregates, which were enriched in organic and fine mineral material, than physicogenic aggregates. In the arable soils, POM and microaggregates were rarely observed in physicogenic aggregates and hardly more abundant in biogenic aggregates, showing that the effects of earthworms on soil microstructure are dependent on management.

The morphological approach used in this study revealed that the formation of stable, highly SOM-enriched microaggregates can be initiated by the activities of earthworms. Current concepts of aggregate- and SOM dynamics do not account for differences in biogenic and physicogenic aggregate formation. This thesis provides strong indications that present conceptual models of aggregate formation and physical protection are incomplete. Therefore, I distinguished two different pathways of microaggregate formation: "Passive" microaggregate formation, which is the basis of current conceptual models of microaggregate formation, without interference of earthworms, and "active" microaggregate formation, that takes place in the earthworm gut and ageing casts after excretion (Chapter 7). Under permanent pasture biogenic aggregate formation by earthworms led to (i) larger amounts of microaggregates within macroaggregates, (ii) greater SOM storage in microaggregates, and (iii) stabilization of relatively little decomposed SOM. However, earthworms could not exert this active microaggregate formation effectively in arable land. The proposed extension of conceptual models of SOM-aggregate interactions needs further development, under a wider set of conditions, to further clarify the conditions and management practices that favour the beneficial effects of biogenic aggregate formation on SOM storage.