Conservation tillage systems are increasingly adapted replacing conventional turnover moldboard plowing practices worldwide. This is part of a sustainable intensification of agriculture to meet future global food demand while at the same time sustaining environmental resources. The choice of tillage system affects soil structure and thereby also soil hydraulic properties (SHP) such as the water retention characteristic (WRC) and the hydraulic conductivity characteristic (HCC). Effects of agricultural management on SHP have been widely studied in the past decades. Thereby, temporal variations were identified as a major source of variability in the quantification of soil pore space and SHP. Such variability is introduced by tillage creating a loose soil matrix that eventually settles due to gravity, wetting-drying cycles and temperature fluctuations but also variable soil organic matter distributions in the soil and biological activity.
Past efforts to model soil water dynamics showed that consideration of time-variable SHP may significantly improve simulation results. This involves both the seasonal variability as well as long-term land-use changes from conventionally to untilled soil. A prerequisite for such an approach is the periodic quantification of the WRC and HCC in the field and laboratory. In addition to the direct provision of modeling parameters, the quantification of WRC and HCC over time yields information on soil structural changes in the shape of a soil pore size distribution (PSD). The evolution of derived PSDs can be modeled and with that, the evolution of SHP might be predicted. However, there is little data available and the processes happening over one cropping season or between land-use changes need to be better understood.
The aim of this dissertation was to shed light on soil pore space and associated hydraulic property changes on a long-term (23 years) tillage experiment in Eastern Germany. Three treatments with varying tillage intensity were investigated: conventional tillage with a turnover moldboard plow (CT), reduced mulch tillage with a cultivator (RT) and no tillage with direct sowing (NT). The soil was a Haplic Luvisol with silt loam texture. Objectives were twofold:
• Objective 1) was to quantify the temporal variability in PSD over one winter wheat cropping season by frequently measuring SHP. Soil physical quality of the three treatments was assessed using this data.
• Objective 2) was to characterize the soil structural differences between the treatments by relating hydraulic conductivity over a wide soil moisture range to other soil physical and chemical properties.
For Objective 1), undisturbed soil cores (250 cm3) were taken over one winter wheat cropping cycle on five occasions from December 2015 to after the harvest in August 2016. Those soil cores were used to determine the saturated hydraulic and the WRC as well as the HCC in the laboratory. The data was parametrized with the bimodal Kosugi and Mualem model. Soil physical quality was assessed by the relative field capacity and air capacity as suggested in recent literature.
Results showed that tilled soil, i.e. CT and RT, exhibited a distinct bimodal PSD with a structural and a textural mode. However, this structural mode was temporally instable and diminished after the winter and throughout the early growing season. Likely processes behind those changes were wetting-drying cycles, rainfall impact and freeze-thaw cycles. Shortly before and after the harvest some of the structural mode was restored which was probably induced by decomposing organic matter mixed into the topsoil from the previous winter wheat harvest during stubble breaking. Described changes were evident in decreases of transmission pores (⌀ 50 - 500 µm) during winter and increases during summer. Untilled soil, i.e. NT, tended towards a unimodal PSD with less transmission but more storage (⌀ 0.5 - 50 µm) pores. Temporally this soil was rather inert. This was attributed to natural compaction in absence of annual tillage for more than 20 years. Soil physical quality varied with the changes in PSD. Water availability was not an issue. Overall, the soil physical quality indicators for soil aeration were outside of an optimal range for indicators for most of the time.
For Objective 2), field infiltration measurements were conducted with a hood (tension) infiltrometer to obtain (near-) saturated hydraulic conductivity. Soil cores were taken to quantify unsaturated hydraulic conductivity. Other properties for correlation and multiple regression analysis were bulk density, the bubbling pressure, organic C, as well as macro- and mesoporosity. X-ray µCT imaging on undisturbed soil cores from CT and NT treatments gave additional information on soil pore metrics.
Results pointed towards a distinctly different soil structure between tilled and untilled soil. Near-saturated hydraulic conductivity of tilled soil was negatively correlated with bulk density as well as macro- and mesoporosity. None of the properties was meaningful for untilled soil. Imaging results confirmed the hypothesis, that (near-) saturated hydraulic conductivity on NT is governed by few well-connected large pores, while the soil matrix is comparably dense conducting only small amounts of infiltrating water. On tilled soil, the overall porosity is relevant for water transmission. Large continuous pore systems, however, get destroyed by annual tillage.
In summary, the study showed distinct differences in soil structure and inherently also SHP between conservation and conventional tillage treatments. Differences in SHP, both in (near-) saturated hydraulic conductivity as well as WRC and HCC were large between some occasions. Therefore, this study confirmed the notion that on arable soils one-off measurements of SHP are not enough for their proper quantification. This was especially true for tilled soil. Modeling tasks over one cropping period, i.e. for example for irrigation schedules, will make periodic measurements necessary, i.e. unless an accurate modeling of the PSD becomes feasible. Current restraints are that most PSD models only consider a short-term post-tillage loss of porosity while a restored macropore system is not accounted for. In contrast to CT and RT, NT soil was temporally stable. While water retention was improved, (near ) saturated hydraulic conductivity was overall lower than on tilled soil. Correlation and regression analysis in combination with X-ray µCT explained some of the differences observed by tension infiltration measurements. Results highlighted that for arable soil, tillage treatments and probably other agricultural management practices, need to be considered when developing pedotransfer functions for an accurate estimation of SHP.:Table of Contents
Declaration of conformity I
Acknowledgements II
Table of Contents IV
List of Figures VII
List of Tables XI
Nomenclature XIII
Abstract XV
Zusammenfassung XVIII
1 Introduction 1
1.1 The sustainable development agenda and conservation tillage 1
1.2 Soil structure and soil hydraulic properties 3
1.3 Effects of conservation tillage on soil hydraulic properties 5
1.4 Temporal variability of soil hydraulic properties 8
1.5 Objectives and hypotheses 10
1.6 Structure of the dissertation 12
2 Materials and methods 15
2.1 Study area 15
2.1.1 Tillage experiment Lüttewitz (‘Schlag Gasthof’) 15
2.1.2 Treatments and agricultural management 16
2.2 Sample design 20
2.3 Field measurements 22
2.3.1 Hood infiltrometer measurements 22
2.3.2 Analysis of hood infiltrometer measurements 24
2.3.3 Macropore stability indicator 24
2.3.4 Undisturbed and disturbed soil sampling 25
2.4 Laboratory measurements 26
2.4.1 Saturated hydraulic conductivity 26
2.4.2 Water retention and hydraulic conductivity characteristic 26
2.4.3 Other soil properties 27
2.5 Model fitting procedure 28
2.5.1 Bimodal models for the water retention and hydraulic conductivity characteristic 28
2.5.2 Parametrization to quantify changes in the pore size distributions and pore volume fractions 29
2.5.3 Parametrization to infer unsaturated hydraulic conductivity for variability analysis 31
2.6 Capacitive soil physical quality indicators 32
2.7 Relationship between imaged pore metrics and field near-saturated hydraulic conductivity 32
2.8 Statistical analysis 33
3 Results 35
3.1 Rainfall patterns 35
3.2 Field (near-) saturated hydraulic conductivity 36
3.3 Threshold pore radius 37
3.4 Laboratory saturated hydraulic conductivity 38
3.5 Unsaturated hydraulic conductivity 39
3.6 Soil pore size distributions and pore volume fractions over one cropping season 40
3.7 Capacitive soil physical quality indicators 45
3.8 Correlation and linear regression of hydraulic conductivity with other soil properties 46
3.9 Other soil properties 49
3.9.1 Bulk density 49
3.9.2 Soil organic carbon and nitrogen 50
3.10 Imaged soil structure and hydraulic conductivity 52
3.10.1 Comparison of hydraulic conductivity obtained through three methods in Spring 2018 52
3.10.2 Soil pore metrics 52
3.10.3 Correlation between hydraulic conductivity and pore metrics 53
4 Discussion 55
4.1 Soil pore size distributions over one cropping cycle 55
4.1.1 Soil pore size distribution is bimodal on tilled soil and varies with time 55
4.1.2 Summary Objective 1) Hypotheses A and B 58
4.2 The effects of a changing pore system on soil physical quality 59
4.2.1 Suboptimal soil physical quality indicators change with time 60
4.2.2 Summary Objective 1) Hypothesis C 62
4.3 Tillage effects on variability of hydraulic conductivity 62
4.3.1 (Near ) saturated hydraulic conductivity 62
4.3.2 Unsaturated hydraulic conductivity 64
4.3.3 Summary: Objective 2) Hypothesis D and E 65
4.4 Factors influencing water transmission and its temporal variation 65
4.4.1 Soil properties partly explain variability in hydraulic conductivity on CT 65
4.4.2 Imaged pore metrics explain differences in field hydraulic conductivity 67
4.4.3 Summary Objective 2) Hypothesis F 68
5 Summary and outlook 69
References 73
Appendix 93
Identifer | oai:union.ndltd.org:DRESDEN/oai:qucosa:de:qucosa:72958 |
Date | 01 December 2020 |
Creators | Kreiselmeier, Janis Leonhard |
Contributors | Feger, Karl-Heinz, Schwärzel, Kai, Wessolek, Gerd, Technische Universität Dresden, United Nations University |
Source Sets | Hochschulschriftenserver (HSSS) der SLUB Dresden |
Language | English |
Detected Language | English |
Type | info:eu-repo/semantics/publishedVersion, doc-type:doctoralThesis, info:eu-repo/semantics/doctoralThesis, doc-type:Text |
Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.004 seconds