@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix skos: . vivo:departmentOrSchool "Land and Food Systems, Faculty of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Roa-García, Clara"@en ; dcterms:issued "2018-08-25T00:00:00"@en, "2018"@en ; vivo:relatedDegree "Doctor of Philosophy - PhD"@en ; ns0:degreeGrantor "University of British Columbia"@en ; dcterms:description """Soil has a crucial role in the terrestrial component of the hydrologic cycle, regulating the availability of water for ecosystem services. Yet relationships between soil properties and land use for the major soil types in the Colombian Andes have not been extensively studied. This study evaluated soil water (SW) dynamics of two soils types, belonging to the most common soil orders in the Colombian Andes, Andisols and Inceptisols. The research was conducted in two watersheds at mid-elevation, and focused on the relationships between mineralogical, physical and chemical soil properties with soil water dynamics, including soil water retention (SWR) and field saturated hydraulic conductivity (Kfs). The Andisols and Inceptisols of this study have a large total porosity compared to typical clay soils, but Andisols, showed higher SWR at every soil tension relative to Inceptisols. Notwithstanding the high hygroscopic water (θPWP), both soils have a wide pore size distribution, with similar gravitational water and plant available water storage capacities. Despite differences in climate and soil parent material between watersheds, the presence of colloids with high specific surface area in both soils (allophane, imogolite, ferrihydrite and organo-metallic complexes in Andisols and ferrihydrite and Al/Fe oxides in Inceptisols) contribute to high SWR. Within each site, differences in SWR between land uses appear minimal, although soil organic carbon was lower under pasture in both soils. The limited differences in SWR between natural forest and pasture appear to reflect the effects of short-range order (SRO) minerals and organo-metallic compounds on SWR, which offset differences in SOC between natural forest and pasture. Quasi steady-state infiltration rates measured in the field did not correspond to expected values based on texture alone, highlighting the importance of field based measurements, particularly in soils with SRO minerals. Additionally, there was a pronounced seasonal difference in Kfs under pasture in both soil types, and a negative correlation of soil water content with Kfs in Inceptisols. Determination of physical, chemical and mineralogical properties was found to be crucial in understanding soil water dynamics in this study, and future work should include an assessment of SRO minerals in addition to SWR characteristics."""@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/66968?expand=metadata"@en ; skos:note " SOIL PROPERTIES AND LAND USE AFFECTING SOIL WATER DYNAMICS IN ANDISOLS AND INCEPTISOLS AT TWO MID-ELEVATION SITES IN THE COLOMBIAN ANDES by Clara Roa García B.Eng., Universidad del Valle, 1997 M.Sc., Universidad del Valle, 2006 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Soil Science) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) August 2018 © Clara Roa García, 2018 ii The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled: Soil properties and land use affecting soil water dynamics in Andisols and Inceptisols in the Colombian Andes submitted by Clara Roa-García in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Soil Science Examining Committee: Maja Krzic, Soil Science/Forestry Research Supervisor Sandra Brown, Soil Science Co-supervisor Leslie Lavkulich, Soil Science Supervisory Committee Member Cindy Prescott University Examiner Mark Johnson University Examiner Additional Supervisory Committee Members: Supervisory Committee Member Supervisory Committee Member iii ABSTRACT Soil has a crucial role in the terrestrial component of the hydrologic cycle, regulating the availability of water for ecosystem services. Yet relationships between soil properties and land use for the major soil types in the Colombian Andes have not been extensively studied. This study evaluated soil water (SW) dynamics of two soils types, belonging to the most common soil orders in the Colombian Andes, Andisols and Inceptisols. The research was conducted in two watersheds at mid-elevation, and focused on the relationships between mineralogical, physical and chemical soil properties with soil water dynamics, including soil water retention (SWR) and field saturated hydraulic conductivity (Kfs). The Andisols and Inceptisols of this study have a large total porosity compared to typical clay soils, but Andisols, showed higher SWR at every soil tension relative to Inceptisols. Notwithstanding the high hygroscopic water (θPWP), both soils have a wide pore size distribution, with similar gravitational water and plant available water storage capacities. Despite differences in climate and soil parent material between watersheds, the presence of colloids with high specific surface area in both soils (allophane, imogolite, ferrihydrite and organo-metallic complexes in Andisols and ferrihydrite and Al/Fe oxides in Inceptisols) contribute to high SWR. Within each site, differences in SWR between land uses appear minimal, although soil organic carbon was lower under pasture in both soils. The limited differences in SWR between natural forest and pasture appear to reflect the effects of short-range order (SRO) minerals and organo-metallic compounds on SWR, which offset differences in SOC between natural forest and pasture. Quasi steady-state infiltration rates measured in the field did not correspond to expected values based on texture alone, highlighting the importance of field based measurements, particularly in soils with SRO minerals. Additionally, there was a pronounced seasonal difference in Kfs under pasture in both soil types, and a negative correlation of soil water content with Kfs in Inceptisols. Determination of physical, chemical and mineralogical properties was found to be crucial in understanding soil water dynamics in this study, and future work should include an assessment of SRO minerals in addition to SWR characteristics. iv LAY SUMMARY Andisols and Inceptisols are the most predominant soils in the Colombian Andes, and these soils provide important ecosystem services related to food production and water supply. Despite their importance, little is known about the relationships between soil properties and water dynamics in this region. This study contributes to our understanding of the relationships between soil properties and water in two watersheds at mid-elevations in the Colombian Andes. It was found that in spite of differences in climate and geology, both soils have nano-size materials. These materials retain water inside their structure, and contribute to a wide range in soil pore sizes which store and release water. Differences in water retention between natural forest and pasture appear minimal, although forests had the highest soil carbon content. Infiltration under pasture was low, especially in Andisols, highlighting the importance of appropriate land management to minimize runoff. v PREFACE I was responsible for the development of the research questions and experimental design in consultation with my supervisory committee. Field sampling was conducted in collaboration with Dr. Sandra Brown. I performed the majority of the sample analyses, with the exceptions of particle size analyses and soil water retention which were run by Edinson Suarez and Miguel Angel Caicedo at the International Center for Tropical Agriculture (CIAT) in Palmira, Colombia. Additionally, Maureen Soon ran the inductively coupled plasma atomic emission spectrometer (ICP-AES) at Earth and Ocean Sciences, UBC, which was used in the determination of short range order minerals. I performed all data analysis. This dissertation is original, unpublished, independent work by the author, C.E. Roa-García. vi TABLE OF CONTENTS ABSTRACT .......................................................................................................................... iii LAY SUMMARY ................................................................................................................. iv PREFACE ............................................................................................................................... v TABLE OF CONTENTS ...................................................................................................... vi LIST OF TABLES ................................................................................................................. x LIST OF FIGURES ............................................................................................................. xiv LIST OF ABBREVIATIONS AND SYMBOLS .............................................................. xviii ACKNOWLEDGEMENTS ................................................................................................ xix DEDICATION ..................................................................................................................... xx 1. GENERAL INTRODUCTION ....................................................................................... 1 1.1 Background .............................................................................................................. 1 1.2 Thesis organization .................................................................................................. 4 2. DESCRIPTION OF THE STUDY AREA AND EXPERIMENTAL DESIGN ............. 7 2.1 Study area ..................................................................................................................... 7 2.1.1 Geology .................................................................................................................. 8 2.1.2 Climate ................................................................................................................. 10 2.1.3 Topography and watershed characteristics .......................................................... 11 2.1.4 Soils of the watersheds ......................................................................................... 12 2.1.4.1 Andisols of the Sonora watershed ................................................................. 12 2.1.4.2 Inceptisol of the El Chocho watershed .......................................................... 14 2.1.5 Land Use .............................................................................................................. 15 2.1.6 Vegetation characteristics .................................................................................... 18 2.2 Experimental design ................................................................................................... 19 2.3 Soil sampling .............................................................................................................. 22 3. MINERALOGY OF ANDISOLS AND INCEPTISOLS AT TWO MID-ELEVATION SITES IN THE COLOMBIAN ANDES .............................................................................. 25 3.1 Introduction ................................................................................................................. 25 3.2 Experimental conditions and laboratory analyses....................................................... 27 3.2.1 Mineralogy ........................................................................................................... 27 3.2.1.1 Crystalline minerals ....................................................................................... 27 vii 3.2.1.2 Short-range order (SRO) minerals ................................................................. 28 3.2.2 Soil chemical and physical properties .................................................................. 30 3.2.2.1 Soil pH ........................................................................................................... 30 3.2.2.2 Soil organic carbon (SOC) ............................................................................ 30 3.2.2.3 Particle size analysis ...................................................................................... 30 3.2.2.4 Soil bulk and particle density ........................................................................ 31 3.2.3 Comparative and statistical analyses .................................................................... 31 3.3 Results and discussion ................................................................................................ 33 3.3.1 Comparison of crystalline minerals...................................................................... 33 3.3.2 Short-range order (SRO) minerals and organo-metallic complexes .................... 38 3.3.2.1 Organo-metallic complexes ........................................................................... 40 3.3.2.2 Short-range order (SRO) minerals ................................................................. 41 3.3.2.3 Andic properties ............................................................................................ 43 3.3.2.4 Allophanic and non-allophanic soils ............................................................. 43 3.3.3 Relationships between short-range order (SRO) minerals and soil properties .... 46 3.3.3.1 Relationships between organo-metallic complexes, pH and soil organic carbon (SOC) ............................................................................................................. 48 3.3.3.2 Relationships between short-range order (SRO) minerals, pH and soil organic carbon (SOC) ................................................................................................ 50 3.3.3.3 Relationships between short-range order (SRO) minerals and organo-metallic complexes with bulk density (b) .............................................................................. 52 3.3.3.4 Relationships between short-range order (SRO) minerals and organo-metallic complexes with soil particle size ............................................................................... 54 3.4 Conclusions ................................................................................................................. 57 4. SOIL WATER RETENTION CHARACTERISTICS OF ANDISOLS AND INCEPTISOLS AT TWO MID-ELEVATION SITES IN THE COLOMBIAN ANDES ... 59 4.1 Introduction ................................................................................................................. 59 4.2 Experimental conditions and laboratory analyses....................................................... 61 4.2.1 Soil water retention curves ................................................................................... 61 4.2.2 Statistical analyses................................................................................................ 64 4.3 Results and discussion ................................................................................................ 66 4.3.1 Soil water retention characteristics ...................................................................... 66 viii 4.3.2 Relationships between soil water retention characteristics and soil properties ... 75 4.3.2.1 Relationships between soil water retention (SWR) characteristics and texture ................................................................................................................................... 77 4.3.2.2 Relationships between soil water retention characteristics (SWR), bulk density (ρb) and soil organic carbon (SOC) ............................................................... 77 4.3.2.3 Relationships between soil water retention (SWR) characteristics and short-range order (SRO) minerals ....................................................................................... 79 4.3.2.4 Relationships between soil water retention (SWR) characteristics and organo-metallic complexes ........................................................................................ 80 4.4 Conclusions ................................................................................................................. 82 5. LAND USE IMPACTS ON SOIL WATER RETENTION CHARACTERISTICS OF ANDISOLS AND INCEPTISOLS AT TWO MID-ELEVATION SITES IN THE COLOMBIAN ANDES ........................................................................................................ 84 5.1 Introduction ................................................................................................................. 84 5.2 Experimental conditions and laboratory analyses....................................................... 85 5.2.1 Statistical analyses................................................................................................ 85 5.3 Results and discussion ................................................................................................ 88 5.3.1 Soil water retention characteristics in Andisols and Inceptisols: differences between natural forest and pasture ................................................................................ 88 5.3.2 Soil water retention characteristics under natural forest and pasture as affected by soil properties ................................................................................................................ 92 5.3.2.1 Differences between land uses in Andisols ................................................... 92 5.3.2.2 Differences between land uses in Inceptisols ................................................ 99 5.4 Conclusions ............................................................................................................... 101 6. FIELD SATURATED HYDRAULIC CONDUCTIVITY DURING DRY AND WET SEASONS UNDER PASTURE ON ANDISOLS AND INCEPTISOLS AT TWO MID-ELEVATION SITES IN THE COLOMBIAN ANDES .................................................... 102 6.1 Introduction ............................................................................................................... 102 6.2 Experimental conditions and measurements ............................................................. 103 6.2.1 Measurement of field saturated hydraulic conductivity ..................................... 103 6.2.2 Estimation of overland flow ............................................................................... 104 6.2.3 Statistical analyses.............................................................................................. 105 6.3 Results and discussion .............................................................................................. 106 ix 6.3.1 Field saturated hydraulic conductivity (Kfs) in pasture: differences between Andisols and Inceptisols.............................................................................................. 107 6.3.2 Estimation of overland flow ............................................................................... 114 6.4 Conclusions ............................................................................................................... 119 7. GENERAL DISCUSSION AND CONCLUSIONS ...................................................... 120 7.1 Overview ................................................................................................................... 120 7.2 Summary ................................................................................................................... 120 7.3 Significance .............................................................................................................. 122 7.4 Implications for land use management ..................................................................... 123 7.5 Suggestions for future research ................................................................................. 123 REFERENCES ................................................................................................................... 126 APPENDICES .................................................................................................................... 145 Appendix A. General characteristics of Andisols and Inceptisols .................................. 145 Appendix B. Characteristics of soil profiles of Andisols (great group Hapludands) and Inceptisols (great group Dystrudepts) included in this study ......................................... 149 Appendix C. Family, species and Importance Value Index (IVI) of vegetation found in natural forest of Sonora (Andisol site) and El Chocho (Inceptisol site) watersheds ...... 151 Appendix D. Presence and intensities of crystalline minerals in Andisols and Inceptisols ........................................................................................................................................ 153 Appendix E. Correlations between short-range order minerals and soil water retention characteristics with soil physical and chemical properties with all data of Andisols and Inceptisols ....................................................................................................................... 155 Appendix F. Comparison of median values of soil water retention characteristics (SWR) between Andisols and Inceptisols ................................................................................... 157 Appendix G. Precipitation for a) Sonora and b) El Chocho watersheds during the overland flow assessment period (mm) .......................................................................... 158 Appendix H. Results of field saturated hydraulic conductivity (Kfs) and soil water content in Andisols and Inceptisols without high values (>50mm/hr) ........................... 159 x LIST OF TABLES Chapter 2 Table 2.1 Main characteristics of the two study watersheds .................................................. 8 Table 2.2 Land use in Sonora (Andisol site) and El Chocho (Inceptisol site) watersheds ... 15 Chapter 3 Table 3.1 Characteristics of selected soil colloids ................................................................ 26 Table 3.2 Peak intensities determined by the X-ray diffraction (XRD) analysis and the correspondent relative abundance of minerals ..................................................................... 28 Table 3.3 Mineral abundance in Andisols and Inceptisols ................................................... 37 Table 3.4 Median dissolution extraction values and indices in A, B and C horizons of Andisols and Inceptisols ....................................................................................................... 39 Table 3.5 Spearman´s Rho correlation coefficients (r) between short-range order (SRO) minerals and organo-metallic complexes (Alp and Fep) with soil physical and chemical soil properties in: a) Andisols all horizons and by horizon; b) Inceptisols all horizons and by horizon .................................................................................................................................. 47 Chapter 4 Table 4.1 Soil water component and associated abbreviations, symbols and pore size class .............................................................................................................................................. 63 Table 4.2 Median values for soil water retention (SWR) characteristics in A, B and C horizons for natural forest and pasture in Andisols and Inceptisols ..................................... 70 Table 4.3 First and third quartile of volumetric water content for Andisols and Inceptisols in A horizon and literature values for three pure textural classes ........................................ 72 Table 4.4 Water retention characteristics of Andisols of this study and other regional studies ................................................................................................................................... 74 Table 4.5 Water retention characteristics of Inceptisols of this study and other regional studies ................................................................................................................................... 75 xi Table 4.6 Spearman´s Rho correlation coefficients (r) between soil water retention (SWR) characteristics and soil physical and chemical soil properties in all horizons and in A and B horizons in: a) Andisols; and b) Inceptisols ......................................................................... 76 Chapter 5 Table 5.1 Median results of soil water retention (SWR) characteristics for forest and pasture in A and B horizons with all data for Andisols and Inceptisols ........................................... 90 Table 5.2 Median results of soil water retention (SWR) characteristics for forest and pasture in A and B horizons of Andisols and Inceptisols ................................................................. 91 Table 5.3 Spearman´s Rho correlation coefficients (r) between soil water retention (SWR) characteristics and soil physical and chemical soil properties for both soil orders in all horizons and in A and B horizons in: a) natural forest; and b) pasture ................................ 94 Table 5.4 Spearman´s Rho correlation coefficients (r) between soil water retention (SWR) characteristics and soil physical and chemical soil properties within Andisols in all horizons and in A and B horizons in: a) natural forest; and b) pasture ............................................... 95 Table 5.5 Spearman´s Rho correlation coefficients (r) between soil water retention (SWR) characteristics and soil physical and chemical soil properties within Inceptisols in all horizons and in A and B horizons in: a) natural forest; and b) pasture ................................ 96 Table 5.6 Median results of soil properties for natural forest and pasture in A and B horizons of Andisols and Inceptisols .................................................................................... 97 Table 5.7 Soil water retention (SWR) characteristics and soil properties of Andisols of the Sonora watershed and other regional studies ....................................................................... 98 Table 5.8 Soil water retention (SWR) characteristics and soil properties of Inceptisols of the El Chocho watershed and other regional studies ................................................................ 100 Chapter 6 Table 6.1 Dates when field saturated hydraulic conductivity (Kfs) were measured in the Sonora watershed (Andisol site) and in El Chocho watershed (Inceptisol site) ................. 103 Table 6.2 Rationale for overland flow (OF) estimation comparing rainfall intensity every 10 minutes (RI10) with field saturated hydraulic conductivity (Kfs) ....................................... 105 xii Table 6.3 Infiltration categories and hydraulic conductivity relative to texture class adapted from NCSS (1996) .............................................................................................................. 107 Table 6.4 Median field saturated hydraulic conductivity (Kfs) and soil water content (θ) in Andisols and Inceptisols ..................................................................................................... 108 Table 6.5 Spearman´s Rho correlation coefficients (r) between field saturated hydraulic conductivity (Kfs) and soil water content........................................................................... 110 Table 6.6 Spearman´s Rho correlation coefficients (r) between field saturated hydraulic conductivity (Kfs) and soil water content without high values (greater than 50mm/hr) .... 110 Table 6.7 Field saturated hydraulic conductivity (Kfs) values reported in other studies carried out in the tropics and in the Andes ......................................................................... 111 Table 6.8 Median field saturated hydraulic conductivity (Kfs) and soil water content (θ) in Andisols and Inceptisols ..................................................................................................... 112 Table 6.9 Median saturated hydraulic conductivity (Kfs), and soil water content (θ) by season in Andisols and Inceptisols ..................................................................................... 113 Table 6.10 Rainfall intensity classes for Sonora a) and El Chocho b) watersheds ............ 115 Table 6.11 Estimation of overland flow (OF) over total precipitation in: a) the Andisol site; and b) Inceptisol site ........................................................................................................... 118 Appendices Table B.1 General characteristics of soil profiles in Andisols watershed (great group Hapludands) ........................................................................................................................ 149 Table B.2 General characteristics of soil profiles in Inceptisols watershed (great group Dystrudepts) ....................................................................................................................... 150 Table C.1 Family, species and Importance Value Index (IVI) of vegetation found in the Andisols watershed ............................................................................................................. 151 Table C.2 Family, species and Importance Value Index (IVI) of vegetation found in the Inceptisols watershed .......................................................................................................... 152 Table D.1 Presence and intensities of crystalline minerals in Andisols ............................. 153 Table D.2 Presence and intensities of crystalline minerals in Inceptisols .......................... 154 xiii Table E.1 Correlations between short-range order (SRO) minerals and indices and soil physical and chemical properties ........................................................................................ 155 Table E.2 Correlations between soil water retention (SWR) characteristics and soil physical and chemical properties ...................................................................................................... 156 Table H.1 Field saturated hydraulic conductivity (Kfs) and soil water content comparison between soil orders ............................................................................................................. 159 Table H.2 Field saturated hydraulic conductivity (Kfs) and soil water content comparison between seasons .................................................................................................................. 159 xiv LIST OF FIGURES Chapter 2 Figure 2.1 Location of Sonora (Andisol site) and El Chocho (Inceptisol site) watersheds ... 7 Figure 2.2 Road cuts showing exposed parent materials of a) the Sonora watershed (Andisol site) and b) the El Chocho watershed (Inceptisol site) .......................................... 10 Figure 2.3 Mean annual precipitation and temperature from the nearest climate station for a) the Andisol site (Sonora watershed) and b) the Inceptisol site (El Chocho watershed)... 11 Figure 2.4 a) Hummocky topography in Sonora watershed (Andisol site); b) concave form of El Chocho watershed (Inceptisol site); c) layer of exposed C horizon (volcanic ash) near the soil surface in upper Sonora watershed (Andisol site); and d) land subsidence in the center of the El Chocho watershed (Inceptisol site) ............................................................. 12 Figure 2.5 a) Example of an Andisol profile from the Sonora watershed, and b) associated air-dried soil samples from the A horizon (left) and B horizon (right) ................................ 13 Figure 2.6 a) Biopores and b) termite burrows in Andisols of the Sonora watershed.......... 14 Figure 2.7 a) Example of an Inceptisol profile from the El Chocho watershed and b) associated air-dried soil samples from the A horizon (left) and B horizon (right) ............... 15 Figure 2.8 Land uses in a) Sonora (Andisol site) and b) El Chocho (Inceptisol site) watersheds ............................................................................................................................ 17 Figure 2.9 Experimental design ............................................................................................ 20 Figure 2.10 Locations of soil pits in a) Sonora (Andisol site) and b) El Chocho (Inceptisol site) watersheds that were sampled during this study........................................................... 21 Figure 2.11 Overview of samples collected and field measurements: a) initial number of samples collected, b) number of samples after slope position was amalgameted, and c) number of samples for specific analyses and field measurements ....................................... 24 Chapter 3 Figure 3.1 Overview of samples collected to compare: a) crystalline minerals of Andisol and Inceptisols; b) short-range order (SRO) minerals of Andisol and Inceptisol; and c) relationships between SRO minerals and soil properties ..................................................... 32 xv Figure 3.2 X-ray diffractograms of A, B1 and C horizons of the P4 pasture profile located in the Sonora watershed (Andisol site) ................................................................................. 34 Figure 3.3 X-ray diffractograms of A, B and C horizons of the P2 pasture profile located in El Chocho watershed (Inceptisol site) .................................................................................. 35 Figure 3.4 Median and quartiles of pyrophosphate extractable aluminum (Alp) in the A, B, and C horizons of: a) Andisols and b) Inceptisols ................................................................ 40 Figure 3.5 Median and quartiles of pyrophosphate extractable iron (Fep) in the A, B, and C horizons of: a) Andisols and b) Inceptisols .......................................................................... 41 Figure 3.6 Median and quartiles of allophane and imogolite in the A, B, and C horizons of: a) Andisols and b) Inceptisols .............................................................................................. 42 Figure 3.7 Median and quartiles of ferrihydrite in the A, B, and C horizons of: a) Andisols and b) Inceptisols .................................................................................................................. 42 Figure 3.8 Relationship between pyrophosphate and oxalate extractable aluminum (Alp/Alo) and soil organic carbon (SOC) in A, B and C horizons of Andisols of the Sonora watershed .............................................................................................................................................. 44 Figure 3.9 Relationships between: a) pyrophosphate extractable aluminum (Alp) and pHH2O; b) pyrophosphate extractable iron (Fep) and pHH2O; c) Alp and SOC; and d) Fep and SOC in A, B and C horizons of Andisols .......................................................................................... 48 Figure 3.10 Relationships between: a) pyrophosphate extractable aluminum (Alp) and pHH2O; b) pyrophosphate extractable iron (Fep) and pHH2O; c) Alp and SOC; and d) Fep and SOC in A, B and C horizons of Inceptisols .......................................................................... 50 Figure 3.11 Relationships between: a) allophane and imogolite and pHCaCl2; b) ferrihydrite and pHH2O; c) allophane and imogolite and soil organic carbon (SOC); and d) ferrihydrite and SOC in A, B and C horizons of Andisols ...................................................................... 51 Figure 3.12 Relationship between ferrihydrite and soil organic carbon (SOC) in A, B and and C horizons of Inceptisols ............................................................................................... 52 Figure 3.13 Relationships between: a) allophane and imogolite with bulk density (b); b) ferrihydrite with b; c) pyrophosphate extractable aluminum (Alp) with b; d) pyrophosphate extractable iron (Fep) with b in A, B and C horizons of Andisols .............. 53 xvi Figure 3.14 Relationships between: a) allophane and imogolite and sand particle size (sand); b) ferrihydrite and sand; c) pyrophosphate extractable Al (Alp) and sand; and d) pyrophosphate extractable Fe (Fep) with sand in A, B and C horizons of Andisols ............ 55 Figure 3.15 Relationships between: a) allophane and imogolite and sand particle size (sand); and b) ferrihydrite and sand in A and B horizons of Inceptisols .............................. 56 Figure 3.16 Soil texture triangles showing textural classes for the A, B, and C horizons of: a) Andisols and b) Inceptisols .............................................................................................. 56 Chapter 4 Figure 4.1 Overview of samples collected to a) compare soil water retention (SWR) characteristics between Andisols and Inceptisols and b) determine relationships between SWR characteristics and soil properties ............................................................................... 65 Figure 4.2 Median results by soil horizon for soil water retention (SWR) curves of Andisols and Inceptisols in: a) A horizon, b) B horizon and c) C horizon .......................................... 67 Figure 4.3 Median results and quartiles for soil water retention (SWR) curves in Andisols and Inceptisols for forest in: a) A horizon and b) B horizon ................................................ 68 Figure 4.4 Median results and quartiles for soil water retention (SWR) curves in Andisols and Inceptisols for pasture in: a) A horizon and b) B horizon ............................................. 69 Figure 4.5 Relationship between proportion of size fractions of clay size (clay) and sand size (sand) with soil water retention at field capacity (θFC) in A horizon of Andisols, n= 17 .............................................................................................................................................. 77 Figure 4.6 Relationships between soil water content at saturation (θSat) and bulk density (ρb) in A, B and C horizons of: a) Andisols; and b) Inceptisols .................................................. 78 Figure 4.7 Relationships between soil water content at saturation (θSat) and soil organic carbon (SOC) in A, B and C horizons of: a) Andisols; and b) Inceptisols ........................... 79 Figure 4.8 Relationships between short range order (SRO) minerals in A and B horizons of Andisols or Inceptisols: a) ferrihydrite and soil water content at saturation (θSat) in Andisols; b) allophane and imogolite and soil water content at 100kPa (θ100kPa) in Andisols; and c) allophane and imogolite and soil water content at PWP (θPWP) in Inceptisols .......... 80 xvii Figure 4.9 Relationships in Inceptisols between: a) pyrophosphate extractable Fe (Fep) and gravitational water (GW) in A horizon; and b) pyrophosphate extractable Al (Alp) and soil water content at saturation (θSat) in B horizon ...................................................................... 81 Chapter 5 Figure 5.1 Overview samples collected to a) compare soil water retention (SWR) characteristics between natural forest and pasture in Andisols and Inceptisols; b) determine relationships between SWR characteristics and soil properties; and c) compare overall differences in soil properties between natural forest and pasture ......................................... 87 Figure 5.2 Median results and quartiles for soil water retention (SWR) curves in A horizon of natural forest and pasture for a) Andisols, and b) Inceptisols .......................................... 88 Figure 5.3 Median results and quartiles for soil water retention (SWR) curves in B horizon of natural forest and pasture for a) Andisols, and b) Inceptisols .......................................... 89 Figure 5.4 Relationship between soil organic carbon (SOC) and sand size fraction in A horizon of natural forest and pasture of Andisols................................................................. 93 Chapter 6 Figure 6.1 Overview of field measurements to a) compare field saturated hydraulic conductivity (Kfs) in pasture between Andisols and Inceptisols and b) determine relationships between Kfs and soil water content .............................................................. 106 Figure 6.2 Gravimetric soil water content (θgrav) vs field saturated hydraulic conductivity (Kfs) in the wet and dry seasons in: a) Andisols; and b) Inceptisols ................................. 110 Figure 6.3 Number of days and total precipitation (TP) for each rain intensity (RI) class in: a) Andisol site; and b) Inceptisol site ................................................................................. 116 xviii LIST OF ABBREVIATIONS AND SYMBOLS Alo, Feo, Sio: oxalate-extractable aluminum (Al), iron (Fe) and silicon (Si) Alp, Fep, Sip: pyrophosphate-extractable aluminum (Al), iron (Fe) and silicon (Si) Alsro, Fesro, Sisro: aluminum (Al), iron (Fe) and silicon (Si) associated with short-range order inorganic material CVC: Corporación Autónoma del Valle del Cauca, local environmental authority from the Valle del Cauca department (province) ETc: crop evapotranspiration ETo: potential evapotranspiration f: total porosity FC: field capacity GW: gravitational water or hydrological buffering capacity Kfs: field saturated hydraulic conductivity or field based quasi steady-state infiltration rate Ksat: saturated hydraulic conductivity or steady-state infiltration rate OF: overland flow PAWS: plant available water storage PWP: permanent wilting point RI: rainfall intensity Sat: saturation SOC: soil organic carbon SRO: short-range order SWR: soil water retention TP: total precipitation θFC: soil water content at field capacity θgrav: gravimetric soil water content θvol: volumetric soil water content θPWP: soil water content at permanent wilting point θSat: maximum retention capacity or soil water content at saturation ρb and ρs: soil bulk density and soil particle density xix ACKNOWLEDGEMENTS I wish to acknowledge the University of Bristish Columbia, UBC, that gave me the flexibility to undertake my research in Colombia, and to my financial supporters for field work and laboratory analyses: the International Development Research Center, IDRC and the International Foundation for Science, IFS. I want to thank my sister Dr. Cecilia Roa, with who I share many interests, and gave me the space to work on this research within the project she directed “Climate change in the Colombian Andes: the role of the water governance” financed by the International Development Research Center, IDRC. Special thanks to the community water organizations Golondrinas and Tribunas Corcega, and their directors Oliverio Suarez, Daciely Gomez and Oscar Gomez for their support and interest in this project. A special gratitude to the field assistants Edinson Suarez, Willmar Herrera and Natalia Gonzalez and the teams for infiltration measurements: Jaidilvia Suárez, Alcides Gonzales, Dairo Salazar, Jairo de Jesus Jimenez and William Henao at La Sonora watershed and Hernando Giraldo Varón, David Perez and Humberto Marin Castillo at El Chocho watershed. My deepest gratitude to my supervisors Dr. Sandra Brown, Dr. Maja Krzic and Dr. Les Lavkulich for their support and their recommendations during these years, especially during the writing process. Thanks to Ines Restrepo for her support since my Master studies, and to the CINARA Institute for the working space and the opportunity to share with other researchers. To my parents Maria Eugenia y Gilberto, from who I received the joy of appreciating and respecting nature. All my gratitude to my husband Alexander Diaz, for his trust and support, to Juan Antonio and Alejandra for their patience and love, and to all my family members, especially Teresa Bedoya, Daniela Diaz and Celina López, without whom I wouldn´t have been able to face this project. xx DEDICATION To Alex 1 1. GENERAL INTRODUCTION 1.1 Background Soil has a crucial role in the terrestrial component of the hydrologic cycle, vital to dry season water supplies, food production, and the resilience of ecosystems.Communities living in the Andean region, are largely dependent on the capacity of soils to regulate water availability (Roa et al., 2011). The Andean region lacks storage infrastructure and relies mainly on soils, glaciers and vegetation to regulate the flows of water, storing it during the wet seasons and releasing it during the dry seasons. Land use and climate change are altering the functions of these storage components of the water cycle, increasing the vulnerability of communities to water and food insecurity. Thus, understanding the ability of soils to hold and release water is critical to rural livelihoods and reducing the vulnerability of these communities to variable climate. From 1939 to 2006, mean annual air temperatures in the Andes have increased 0.7°C (Vuille et al., 2008); with higher elevations (>2,000 m) anticipated to experience a further increase of 3°C by 2090 (Bradley et al., 2006; Urrutia and Vuille, 2009; Vuille et al., 2015). Precipitation is expected to increase in the wet season and decrease in the dry season (Vuille et al., 2008), with fluctuations from 10 to 40% (IDEAM, 2015). In addition, climatic variability related with the El Niño/Southern Oscillation (ENSO) phenomenon further increases the vulnerability of Andean communities to water scarcity and extreme events (IPCC, 2007; IDEAM, 2010). The Andes, with its moderate and diverse climate, is the most density populated region of Colombia. The Andean area (287,720 km2; 500 m and above) represents 25% of the total country (Armenteras and Rodríguez, 2007) but is home to 77% of the urban population and 71% of the rural population (DANE, 2005). Despite being home to the majority of the population of Colombia, the Andean region contributes only 13% of the national water supply (IDEAM, 2010) and “over 80% of the municipalities are supplied by small sources (streams, creeks, brooks) with low regulation capacity and high vulnerability\" (IDEAM, 2 2000, p. 38). Some rivers from the Andes have shown stream flow decreases of 40% in the El Niño phase compared to the long-term average (IDEAM, 2010). Poverty is concentrated in the rural area, where 65% of people are considered to be poor (DNP, 2010) and the majority of the population are reliant on small scale food production for home consumption. It is recognized that the poorest and most vulnerable groups will disproportionately experience the negative effects of climate change. Land use in the Andes has changed drastically since the introduction of cattle by the Spaniards in the 16th century. Natural land cover such as wetlands, páramo1 and forest have been converted to pasture for cattle grazing and to agricultural crops, and by the year 2000 less than 40% of the natural vegetation in the Andes still remained (Rodríguez et al., 2006). In 1984, three national institutes determined that the appropriate land cover for 68% of the Colombian Andes is forest; however, just 26% of this area was still covered by forest vegetation (INDERENA et al., 1984) and deforestation continued in the Andes from 1985 to 2005 as montane forest decreased a further 13% (Armenteras et al., 2011). Due to the dependence of Andean population on ecosystems to provide water for crops and drinking water, it is important to gain a better understanding of the factors (including soil) influencing the regulation and storage of water within ecosystems. This thesis documents the soil properties in two watersheds located on Andisols and Inceptisols, the most common soil orders in the Colombian Andes, and the effects of natural forest and pastures on soil water dynamics. Andisol is a soil order in the US Department of Agriculture soil taxonomy (Soil Survey Staff, 2014) that refers to soils of dominantly volcanic origin. Their main features are a high proportion of short-range order (SRO) minerals and the presence of a dark surface horizon rich in organic matter. They are very common in the Andean páramos and are characterized by large water holding capacity (Buytaert et al., 2006). On the other hand, Inceptisols lack a well-developed soil profile and are characterized by indistinct horizons. 1 Ecosystems in the regions above the continuous forest line, yet below the permanent snowline, characterized by vegetation composed largely of tussock grasses, cushion plants and frailejónes (Espeletia). 3 They include soils developed on recent alluvium and lacustrine materials that are quite high in clay. A detailed description of these two soil orders are provided in Appendix A. In regions of the Colombian Andes with humid climate, the most abundant great groups are Hapludands (within the Udands suborder) and Dystrudepts (within the Udepts suborder), respectively. Hapludands are characterized by the dark color of the surface horizon and their high porosity, while Dystrudepts are characterized by removal of soluble compounds, the addition of organic carbon favored by humid climate, and by its vulnerability to soil erosion (Malagón, 2003). Since these two soil great groups occupy 66% of the Colombian Andes (Malagón, 2003) they were selected as the focus of this study. Within this document, the terms “Andisol” and “Inceptisol” will be used as short forms to distinguish between the soils at the two study sites, and “soil orders” used as a descripter when referring to both soils. Most studies carried out in the Andes have focused on the effects of different land management practices on crop yields, soil erosion or soil nutrient dynamics; however, only a limited number of studies have evaluated the effects of land uses on the soil water characteristics of the two soil orders in relation to water security. The major land uses in Colombia are natural forest and pasture, covering 53.2% and 30.6% of the country, respectively; while crops cover an area of 4.7% (IGAC, 2012). Most studies on soil water characteristics carried out in the Andes were located in páramo ecosystems at elevations >3,500 m, and have shown that the conversion from natural vegetation (natural forest or páramos) to crops and pasture reduce the proportion of macro and mesopore volume (Diaz and Paz, 2002; Daza et al., 2014; Buytaert et al., 2002; Buytaert et al., 2005b; Podwojewski et al., 2002). However, results from high elevation (> 3,000 m) studies may not be directly applicable to mid-elevation conditions, due to colder and wetter conditions and generally higher soil organic carbon (SOC). In addition, most of these studies do not include an evaluation of soil mineralogical, chemical and physical properties, and their relationships to soil water retention (SWR) and field saturated hydraulic conductivity (Kfs). Overall, Kfs has been poorly studied in tropical ecosystems, despite being related to events such as erosion, floods, landslides and water scarcity (Bonell, 1993; Ilstedt et al., 2007). 4 The goal of the study was to assess soil properties as they affect soil water dynamics in a watershed context at mid-elevation (1,700-2,300 m) under the two most common soil orders in the Colombian Andes: Andisols located in the Central mountain range and Inceptisols located in the Western mountain range. The assessed soil subgroups of these soil orders were Acrudoxic Hapludands (Andisols) and Typic Dystrudepts (Inceptisols). The specific objectives of the research were to: • Determine the mineralogy of the soils and relate these properties to water retention characteristics • Determine soil water retention (SWR) characteristics of A and B horizons • Compare the effects of natural forest and pasture land uses on SWR characteristics, and • Determine field saturated hydraulic conductivity (Kfs) during dry and wet seasons. The contribution of a reconnaissance study of this type, in which two soils from the same region (i.e., mid-elevation region of Colombian Andes) are compared, highlights the influence of climate and geology, on soil properties and soil water dynamics. Understanding the soil processes, and the differences in water availability in two soils (Andisols and Inceptisols), is important for adaptation strategies needed to reduce vulnerability to water scarcity in the Andean region. 1.2 Thesis organization The thesis is organized in the following seven chapters: 1) General introduction 2) Description of the study area and experimental design 3) Mineralogy of Andisols and Inceptisols at two mid-elevation sites in the Colombian Andes 4) Soil water retention characteristics of Andisols and Inceptisols at two mid-elevation sites in the Colombian Andes 5) Land use impacts on soil water retention characteristics of Andisols and Inceptisols at two mid-elevation sites in the Colombian Andes 5 6) Field saturated hydraulic conductivity during dry and wet seasons under pasture of Andisols and Inceptisols at two mid-elevation sites in the Colombian Andes 7) General conclusions Description of the two study watersheds, representative of the two most common soils in the Andean region of Colombia (Andisols and Inceptisols) including climate, ecology, land use, geology and soil characteristics are presented in Chapter 2. This chapter also contains a description of the experimental design and the soil sampling design. Soil water retention depends on soil chemical and physical properties such as SOC, texture and bulk density (ρb) (Rawls et al., 2003) as well as soil mineralogy. Chapter 3 focuses on the mineralogy of Andisols and Inceptisols. Crystalline minerals and concentrations of SRO minerals are compared between Andisols and Inceptisols, and correlations are assessed between mineralogical characteristics and physical and chemical soil properties. Crystalline minerals are estimated by X-ray diffraction, and the concentrations of SRO minerals are quantified using dissolution methods. Andisols and Inceptisols are common in the Andean mountain region where the majority of the population is located; consequently, they are of particular importance for the regulation of water for local food production and domestic water supply. Some research has been conducted on the SWR characteristics of Andisols; however, they are limited to the páramo and coffee regions. On soils developed on volcanic ash, SRO minerals enhance SWR (Nanzyo et al., 1993), but clay-dominated soils such as the Inceptisols of this study, may have variable SWR characteristics depending on their mineralogy (Hodnett and Tomasella, 2002). In Chapter 4, SWR characteristics are compared between Andisols and Inceptisols from mid-elevation sites of the Colombian Andes, SWR characteristics are correlated with soil properties, and similarities and differences are discussed. With the introduction of cattle, land use in the Andes changed dramatically. Natural forests on sloping terrain were transformed into extensive grazing lands with negative effects such as landscape homogenization and erosion (Etter and Van Wyngaarden, 2000). Although some studies (Diaz and Paz, 2002; Daza et al., 2014; Buytaert et al., 2002; Buytaert et al., 2005b; Podwojewski et al., 2002) have assessed the change in SWR with the conversion of 6 natural land cover to other land uses on páramo ecosystems (located at elevations >3,500 m) the findings of those studies are not directly applicable to lower elevation sites in the Andes. To better understand the impacts of land use type (natural forest and pasture) on SWR of Andisols and Inceptisols at mid-elevation sites in the Colombian Andes, SWR characteristics were compared between pasture and natural forest in both soils. Results were correlated with soil properties, and are presented in Chapter 5. Although water infiltration may be directly related to events such as erosion, floods, landslides and water scarcity, Kfs in tropical ecosystems is poorly understood (Bonell, 1993; Ilstedt et al., 2007). The Colombian National Institute of Meteorology and Environmental Studies (IDEAM, 2000) estimated that by 2025, 55% of total population in Colombia could be affected by water scarcity, 70% of which live in the Andean region. In addition, within Colombia, landslides and floods rank first and third, respectively among catastrophic events leading to deaths (World Bank, 2012). These events may be related with low infiltration rates, potentially associated with the conversion of natural forest to pasture or crops (Bruijnzeel, 1989, 2004; Chaves et al., 2008). However, limited studies have been carried out in the Andes to evaluate quasi steady-state infiltration rates in the field (Kfs) on different soil types. In this study, double ring infiltrometers were used to measure Kfs in the wet or rainy season and the dry season. Differences between soils and seasons were assessed. Results are presented in Chapter 6. Overview, summary, significance, implications for land use management, and suggestions for future research are given in Chapter 7. 7 2. DESCRIPTION OF THE STUDY AREA AND EXPERIMENTAL DESIGN 2.1 Study area This study was conducted in two watersheds in the Colombian Andes, the Sonora watershed dominated by Andisols (great group Hapludands) located in the Central branch, and the El Chocho watershed dominated by Inceptisols (great group Dystrudepts) located in the Western branch (Fig. 2.1). These two watersheds, located 230 kilometers apart, were selected as they are representative of the most common soil types (Andisols and Inceptisols) and land uses (natural forest and pasture) in the Colombian Andes, and are headwater ecosystems on which rural populations depend for their water supply. Both watersheds are located at similar elevation (>1,700 m), and are of similar size. Figure 2.1 Location of Sonora (Andisol site) and El Chocho (Inceptisol site) watersheds 8 The main characteristics of the two watersheds in terms of their location, geology, climate, and soil type are summarized in Table 2.1. Table 2.1 Main characteristics of the two study watersheds Sonora watershed El Chocho watershed Mountain range or branch of the Colombian Andes Central branch Western branch Department (provinces) Risaralda Valle del Cauca Drainage basin Barbas Aguacatal Latitude, Longitude 4.41N, 35.74W 3.30N, 76.34W Elevation range (m) 2,088-2,331 1,768-2,111 Area (ha) 204 149 Mean annual air temperature (°C) 16.6 16.7 Average annual rainfall (mm) 2,955 1,393 Geology Fluvio-volcanic sediments above a volcanic and methamorphic basement1 Volcanic formation, diabase, basalts and lateritic rocks of volcanic formation2 Soil order Andisol Inceptisol Soil suborder Udands Udepts Soil great group Hapludands Dystrudepts Soil subgroup Acrudoxic Hapludands Typic Dystrudepts 1 Guarín, et al. (2004); 2 SGC (1984b) 2.1.1 Geology The Central Branch is the oldest of the three mountain ranges that form the Colombian Andes. It is dominated by poly-metamorphic and igneous rocks represented by plutons, batholiths and large volumes of volcanic rocks of different ages (Malagón, 2003). The Sonora watershed is located on the Quindío-Risaralda fan, a mass flow of fluvio-volcanic sediments deposited during the last million years above a Cretaceous volcanic and metamorphic basement (Guarín et al., 2004). The geology is classified as volcanic lahar 9 (Qfs)2 (SGC, 1984a). The thickness of the lahar deposits varies from more than 200 m in the proximal regions of the volcanoes to less than 20 m in distal regions, and is overlain by a volcanic ash layer that varies from 5 to 20 m thick (Guarín et al., 2004) (Fig. 2.2a). The ice-capped active volcanoes, which are under permanent monitoring in the region, are Huila, Del Ruiz, Tolima, and Santa Isabel (Duque-Escobar, 2007). The Western branch of the Colombian Andes was formed by igneous, plutonic, and volcanic rocks, partially covered by clastic sedimentary rocks of limestone, which in turn are covered by thick and extensive Quaternary deposits of volcanic, fluvial-volcanic, fluvial and colluvial origin (Malagón, 2003). The El Chocho watershed is part of the Farallones system formed by a cluster of mountains in which stream flows are controlled by multi-directional rock fractures (PNNC, 2005). The geology in the region of the El Chocho watershed consists of: (1) quartz conglomerate, siltstones, dirty sandstones, shales and coal (Tog), (2) volcanic formations, diabase, basalts (Kv), and (3) lateritic rocks of volcanic formation (Ql/Kv) (SGC, 1984b). The dominant rock type in the watershed is basalt, with a composition of approximately equal parts of pyroxenes and plagioclase with traces of olivine and magnetite (Nivia, 2014) (Fig. 2.2b). 2 Landslide of wet volcanic debris on the side of a volcano 10 a) b) Figure 2.2 Road cuts showing exposed parent materials of a) the Sonora watershed (Andisol site) and b) the El Chocho watershed (Inceptisol site) 2.1.2 Climate Average monthly precipitation and temperature data for the study sites are given in Figure 2.3. Both sites experience a bimodal annual precipitation cycle, with two wet seasons (April-May and October-November) and two dry seasons (December-February and June-August). The annual precipitation in the region of the Sonora watershed (Andisols) averages 2,955 mm (CRQ, 2015), which is more than double the 1,393 mm average annual precipitation in the region of the El Chocho watershed (CVC, 2014). Given their similar 11 elevation, the two regions have similar average annual air temperatures (around 16.6°C) (CRQ, 2015; CVC, 2014). a) Andisol site (Sonora watershed) b) Inceptisol site (El Chocho watershed) Figure 2.3 Mean annual precipitation and temperature from the nearest climate station for a) the Andisol site (Sonora watershed) and b) the Inceptisol site (El Chocho watershed) 2.1.3 Topography and watershed characteristics The Sonora watershed has a hummocky topography (Fig. 2.4a) with dominantly east-west trending slopes (Guarín, et al., 2004). Slopes in riparian areas around tributary streams and in the upper watershed are up to 88% gradient, while flatter areas in the central and lower watershed are < 20%. In some parts of the upper Sonora watershed the volcanic ash layer is visible near the soil surface (Fig. 2.4c), while in the lower depositional areas of the watershed, the ash layer is located at depths of two to five meters. In the El Chocho watershed (Inceptisols), the stream flow direction is north-south, and the watershed form is concave with steep side slopes (Fig. 2.4b). Slopes around streams and in the upper watershed exceed 70%, while flatter regions located in the mid and lower watershed are <20% slope. A dirt road traverses the western slope, with sidecast material deposited downslope. In the center of the watershed land subsidence is visible (Fig. 2.4d), and may be related to subsurface fractures (PNNC, 2005). 12 a) b) c) d) Figure 2.4 a) Hummocky topography in Sonora watershed (Andisol site); b) concave form of El Chocho watershed (Inceptisol site); c) layer of exposed C horizon (volcanic ash) near the soil surface in upper Sonora watershed (Andisol site); and d) land subsidence in the center of the El Chocho watershed (Inceptisol site) 2.1.4 Soils of the watersheds 2.1.4.1 Andisols of the Sonora watershed Soils in the Sonora watershed are classified as Andisol (soil order), Udand (suborder for Andisols of humid climates) and are Acrudoxic Hapludands (soil subgroup) (IGAC, 1996). 13 A typical Andisol profile in the Sonora watershed and its associated air-dried soil samples is shown in Figure 2.5. Based on field observations, these Andisols have a sandy texture and lack coarse fragments. After drying, Andisol samples had a light brown color. Munsell colors of the dry samples from this profile were 10YR 5/3 for the A horizon and 10YR 6/4 for the B horizon. The color of the ash was 10YR 8/1. Typically, the A horizon is about 0.20 m thick. The B horizon ranges from 0.20 to 1.0 m depth in the upper watershed, and is thicker in lower sections (0.20 to 5.0 m depth), likely due to erosion in the upper watershed and deposition in lower sections. In one of the analyzed profiles, a placic layer was found at 2.8 m and a perched water table was found at 2 m. Detailed characteristics of Andisol profiles are provided in Appendix B. a) b) Figure 2.5 a) Example of an Andisol profile from the Sonora watershed, and b) associated air-dried soil samples from the A horizon (left) and B horizon (right) Compared to the El Chocho watershed (Inceptisol site), there were fewer earthworms observed in the Sonora watershed, but termites were found in three pasture sites when the infiltration measurements were taken (Fig. 2.6); no termites were present in any of the sampled soil profiles. 14 a) b) Figure 2.6 a) Biopores and b) termite burrows in Andisols of the Sonora watershed 2.1.4.2 Inceptisol of the El Chocho watershed Soils in the El Chocho watershed are classified as Inceptisol (soil order), Udepts (suborder for humid climates) and Typic Dystrudepts (soil subgroup) (IGAC and CVC, 2004). The Typic Dystrudepts from the El Chocho watershed are moderately deep soils, with fine texture, moderate drainage, strong to highly acidic pH, high aluminum saturation (>60%), and low fertility (IGAC and CVC, 2004). A representative Inceptisol profile and its associated air-dried soil samples are shown in Figure 2.7. Based on field observations, the texture of Inceptisols in the study watershed is described as clay, with cobbles and stones present throughout the soil profile (as opposed to the Sonora watershed where no cobbles or stones were encountered). Small yellow and reddish concretions can be seen as inclusions in the A horizon of this soil profile. Munsell colors of the air-dry samples from the profile were: 10YR 4/3 for the A horizon and 10YR 6/6 for the B horizon. The red colors in subsurface horizons indicate the presence of iron oxides and hydroxides. The thickness of the A horizon in El Chocho watershed averaged 0.30 m, B horizons had a relatively constant thickness (from 0.30 to 2.0 m depth) throughout the watershed. The C horizon which found below 2 m, transitions to saprolite, overlying bedrock which based on field observations is > 30 m thick. Detailed descriptions of the Inceptisol profiles are provided in Appendix B. 15 a) b) Figure 2.7 a) Example of an Inceptisol profile from the El Chocho watershed and b) associated air-dried soil samples from the A horizon (left) and B horizon (right) 2.1.5 Land Use The two most common land use types in the country, natural forest and pasture, occupy more than 80% of the area of each watershed (Table 2.2 and Fig. 2.8). Table 2.2 Land use in Sonora (Andisol site) and El Chocho (Inceptisol site) watersheds Soil order Watershed Area (ha) % Natural forest % Pasture Other land uses % Other land uses Andisol Sonora 204 48 33 Plantation forest 19 Inceptisol El Chocho 149 50 42 Culinary herbs 8 Logging of the primary forest in both watersheds likely occurred with the expansion of cattle grazing in the early 1800s, and was coupled with exponential population growth and the settling of mid-elevation regions in the country over the last 50 – 60 years (Etter and Van Wyangaarden, 2000). Natural forests in the watersheds today are secondary forests established through natural regeneration, and are recognized by the small diameter at breast 16 height (<10 cm) of the majority of the vegetation (Cardona, 2015a and 2015b). Although there is no tracked history of land use change in either watershed, the naturally regenerated secondary forest was likely established from vegetation not removed during logging of the primary forest. Today, natural forests are located only on slopes > 20% gradient within both watersheds, commonly in stream canyons and at higher elevations. These natural forests are mainly affected by human activity along trails to farms and water intakes in lower parts of the watershed. Pasture species in both watersheds appeared after logging of the primary forest and the introduction of cattle (Sanchez, 2017) in the early 1800s. Pasture areas in both watersheds are used for cattle grazing. During the study period, around 25 cattle were found in the Sonora watershed (0.4 animals/ha), while around 20 cattle were located in El Chocho watershed (0.3 animals/ha). Animals are grazed throughout the year, utilizing a pasture rotation system with areas separated by fencing. Pasture in the watersheds are located on both flat (<20%) and sloping lands. In general, pasture in both watersheds provides a complete soil cover. Exceptions were small areas (about 40 m2) near drinking troughs or streams where vegetation was heavily grazed, trampled and soil was exposed, by cattle concentration in these areas. Machinery and fertilizers are not used in either natural forest or pasture. In addition to the two main land uses, there are secondary land uses which occupy less than 20% of the area of each watershed: plantation forest in the Sonora watershed (Andisol site) and crops in the El Chocho watershed (Inceptisol site). The plantation forest in the Sonora watershed, corresponds to planted eucalyptus and pine trees managed by the Smurfit Kappa Company for the manufacturing of paper and packing products. The secondary land use in the El Chocho watershed corresponds to culinary herbs. As these land uses are small in extent and not common between the watersheds, soils under these land uses were not evaluated in this study. 17 a) Sonora watershed- Andisol b) El Chocho watershed- Inceptisol Figure 2.8 Land uses in a) Sonora (Andisol site) and b) El Chocho (Inceptisol site) watersheds 18 2.1.6 Vegetation characteristics Natural forest from the studied watersheds, despite being secondary forests, are rich in diversity. Cardona (2015a, 2015b) identified 63 species in the Sonora watershed (Andisol site) and 81 species in El Chocho watershed (Inceptisol site). The Shannon-Wiener index (Shannon, 1948), was greater than 3 in both watersheds, indicating a high species diversity (Cardona, 2015a and 2015b). Tree species that stand out by their flowers and leaf color in Sonora watershed were Heliocarpus popayanensis (Malvaceae family) and Cecropia telealba (Urticaceae family) (Cardona, 2015a). Based on the IVI Index (Gentry, 1982), the natural forests of the Sonora watershed are dominated by species of the Cyatheaceae (tree ferns) and Arecaceae family (palm trees) interspersed with flowering shrub and trees from the Melastomataceae, Rubiaceae and Solanaceae families (Cardona, 2015a). Wettinia kalbreyeri, a palm tree from the Arecaceae family, an endemic species locally known as palma bolillos, is also found in the watershed. Tree species that stand out by their height and leaf color in the El Chocho watershed were Cedrela Montana (Meliaceae family) and Cecropia telealba (Urticaceae family) (Cardona, 2015b). The most dominant plant families, based on the IVI Index (Gentry, 1982) in the natural forest of the El Chocho watershed, were Heliconiaceae (flowering plants) and Arecaecea (palm trees) (Cardona, 2015b). Detailed lists of plant species found in the natural forests of these two watersheds are provided in Appendix C. Pasture species that were present in both of the study watersheds and that are widespread in the South and Central America (Cardona Mejía, 2012, INATEC, 2016), are Cynodon plectostachius (Estrella) and Pennisetum Clandestinum (Kikuyo) species. These species occupy approximately 80% of the pasture area of both watersheds. Additional pasture species throughout the Sonora watershed (Andisol site) include Paspalum fasciculatum (Gramalote) and Axonopus micay (Micay) (Sanchez, 2017). In the upper watershed there were patches of Rhynchospora nervosa. Furthermore, a shrub layer of up to 1.5 m, covered about 5% of pastures on both sloping and flat positions. 19 Additional pasture species throughout the El Chocho watershed (Inceptisol site) include Hyparrhenia rufa (Yaragua). Some patches of Saccharum sinense (King grass) were found in the upper watershed and Stipa charruana (Espartillo) in the mid watershed (Sanchez, 2017). The second tier of vegetation in the El Choco pasture land is Pteridium aquilinim, a native fern species which is known as a pioneer species. About 40% of the pasture on slopes and 20% in flat areas was covered by this fern layer. 2.2 Experimental design A stratified random experimental design was used in this study. Two watersheds with the most common soil orders and land uses in the Colombian Andes were selected: 1) Sonora watershed – Andisols, and 2) El Chocho watershed – Inceptisols. Within each watershed, the two dominant land uses (natural forest and pasture) were delineated. Each land use category was further subdivided by slope into: flat positions (<20% slope) and sloping lands (>20% slope). Flat position was defined as areas feasible for cultivation with low gradients (i.e., <20%) that do not require technologies to protect against erosion (Pasolac, 1999). Note that forests are only located on slopes in both watersheds. Pasture areas with exposed soil and forest areas on slopes >70% and with remote access were excluded. In each watershed three types of land units were identified: natural forests on slopes, pasture on slopes, and pasture on flat positions. Each land unit was subdivided into three geographical areas (e.g., north east, north west, and south). The geographical areas were also subdivided in 10 blocks, from which two blocks were randomly selected and soil pits were excavated. In this manner two sites were located in each geographical area for a total of 6 sites per land unit (Fig. 2.9). 20 Figure 2.9 Experimental design All soil pits were georeferenced (Fig. 2.10) and soil samples were taken by horizon. Details on soil sampling are provided in Section 2.3. Field measurements of Kfs were also taken at locations near the excavated pits. 21 a) Sonora watershed- Andisol b) El Chocho watershed- Inceptisol Note: P sites referred to pasture sites, while B sites referred to natural forest sites Figure 2.10 Locations of soil pits in a) Sonora (Andisol site) and b) El Chocho (Inceptisol site) watersheds that were sampled during this study 22 2.3 Soil sampling Soil pits were excavated to a depth of 1.20 m; with the exception of one pit in Andisols which was excavated to a depth of 3.1 m. Samples were taken from A and B, and the C horizon when encountered. Horizons were delineated, horizon thickness measured, and the coarse fragment content was estimated. To obtain additional information about the variability of the soil parent material, three additional C horizons in Andisols were sampled from exposed C horizons at locations near sampling sites. All soil samples, in both watersheds were collected in May 2012. Composite soil samples, comprised of 3 individual samples vertically distributed within each horizon (i.e. upper, mid and lower depths), were taken for chemical, physical, and mineralogical analyses. The chemical analyses conducted were pHH2O, pHCaCl2 and soil organic carbon (SOC); the physical analysis was particle size (% sand, % silt and % clay) and the mineralogical analyses were short-range order (SRO) mineral and crystalline mineral composition. In addition to the composite sampling, undisturbed soil core samples were taken from the mid-depth of each horizon (average depth of 12 cm). For SWR analysis, these undisturbed soil samples were collected in 45 cm3 steel cores (diameter 4.8 cm and height 2.5 cm) and were also used for soil bulk and particle density analyses. The number of samples for each soil analysis or measurement are summarized in Figure 2.11. Typically, in each land unit, six samples per soil horizon were collected. Exceptions were: 1) a soil pit of Andisols and two soil pits of Inceptisols, where two B horizons (B1 and B2) were delineated, leading to a total of seven samples, and 2) one Inceptisol pit, which did not have a B horizon. The number of C horizons sampled in Andisols were eight, including three additional C horizon samples from locations near sampling sites on pastures with steep slopes. Five C horizons were sampled in Inceptisols (Fig. 2.11a) Pasture samples from flat and sloping positions were pooled, leading to six samples per horizon for forest and 12 samples per horizon for pasture (with the exceptions explained above). The reason to unify this pool, was that slope showed no effect on SWR characteristics under pastures based on statistical analysis (Mann Whitney U test) (Fig. 2.11b). 23 Additionally, one sample from the A horizon of forest in Andisols had a very high SOC (26.5%) and was considered anomalous since all other samples in this study and other studies from this region (Roa-García, 2009) had SOC at around 13%. High values of SOC (32-38%) (Diaz and Paz, 2002; Buytaert et al., 2006), are common in páramo ecosystems at high elevations (>3,500 m) but not at mid-elevations. Consequently, this sample was removed from the statistical analyses of physical and chemical soil properties (Fig. 2.11b). Data normality was assessed using skewness and kurtosis values. As soil parameters were not normally distributed, non-parametric tests were used, specifically Mann Whitney U test to compare soil characteristics and Spearman correlations to measure the strengths of association. For X-ray diffraction (XRD) analysis, one natural forest profile and one pasture profile for each soil order were analyzed (Figs. 2.11c and 2.11d). All samples were analyzed for physical and chemical properties. Results for crystalline minerals and SRO minerals are shown in Chapter 3, while results of physical and chemical analyses are used in Chapters 3, 4 and 5. Field saturated hydraulic conductivity (Kfs) was determined using a double ring infiltrometer, which provides an index of quasi steady-state infiltration rate as measured in the field (Bagarello et al., 2014; Nimmo et al., 2009). As slope limits the use of the double ring infiltrometer, the six blocks of natural forest, all on slopes >20%, and the six blocks of pasture on slopes >20%, were not evaluated. The measurements of quasi steady-state infiltration rate on pasture were assessed in the six randomly selected blocks with flat slope positions (Fig. 2.11e). In each of the blocks, measurements were taken in duplicate using two sets of double ring infiltrometers. Infiltration measurements are reported in Chapter 6. 24 1 Two B horizons in one pit; 2 No B horizon found in one pit; 3 C horizon was sampled when found at ≤1.2m depth; 4 C horizon samples including three exposed C horizons from nearby locations; 5 One A horizon sample excluded due to unusually high SOC content Figure 2.11 Overview of samples collected and field measurements: a) initial number of samples collected, b) number of samples after slope position was amalgameted, and c) number of samples for specific analyses and field measurements 25 3. MINERALOGY OF ANDISOLS AND INCEPTISOLS AT TWO MID-ELEVATION SITES IN THE COLOMBIAN ANDES 3.1 Introduction Soil water characteristics are governed to a large extent by the type and amount of soil colloids (Hodnnett and Tomasella, 2002). Yet in the Andes limited research has been conducted on soil mineralogy, how mineralogy relates to other soil properties, and the implications for soil water retention (SWR). Short range order (SRO) minerals are of particular interest due to their high water holding capacity and their stabilizing effect on soil organic matter (Buytaert et al., 2002; Krammer et al., 2012). In Andisols and Inceptisols, the soil orders predominant in the Colombian Andes, allophane, imogolite and Fe/Al oxides and hydroxides are SRO minerals of interest as they may play a key role in SWR. SRO minerals are formed from the weathering of framework silicates and ferromagnesian minerals, and are characterized by high specific surface area (SSA) (Table 3.1) and a large number of hydroxyl groups (Krammer et al., 2012; Eusterhues et al., 2005). Allophane and imogolite are SRO alumino-silicates commonly associated with volcanic ash soils (Parfitt, 2008; Nanzyo, 2002). Their nanoparticle size (1-100 nm) and hollow structure result in microscopic pores that can store water within the mineral framework (Wada, 1985). The presence of (pH dependent) positive and negative charges allow SRO minerals to bind to organic compounds and participate in aggregate formation (García et al., 2018). These characteristics also promote low soil bulk density (ρb) and a wide pore size distribution (Nanzyo, 2002). Iron and aluminum oxides also play a significant role in SWR and aggregate formation as they also have pH-dependent charges. This especially applies to the non-crystalline ferrihydrite, with its nanoparticle size and high SSA (Table 3.1), which despite its small contribution to soil mass (commonly <1%) (Duiker et al., 2003; Regelink et al., 2015), binds silt and sand particles to soil organic matter leading to creation of stable aggregates (Arias et al., 1996; Sei et al., 2002). 26 Table 3.1 Characteristics of selected soil colloids Soil colloid /Formula Specific surface area (m2/g) References for specific surface area Total charge at pH 7 (cmolc+/kg)1 pH dependent charge (%) Alumino-silicate SRO minerals Allophane Al2O3 (SiO2)1.3-2 (2.5-3) H2O and imogolite Al2SiO3(OH4), 700-1500 Parfitt, 2008 302 90 Fe oxides Magnetite Fe3O4 90 Dixit and Hering, 2003 0 - Hematite α- Fe2O3 14.4 Singh et al., 1996 0 - Goethite α- Fe3+O(OH) 134-139 Matis et al., 1997 42 100 Ferrihydrite Fe5HO8.4H2O 220-560 Shoji et al., 1984 -0.53 100 Al oxides Boehmite γ-AlOOH 108 Singh and Yadava, 2003 5 to 104 100 Gibbsite α- Al(OH) 3 100-220 Kämpf et al., 2012 42 100 Humus 800-900 Handreck and Black, 2005 2002 100 1Centimoles of charge per kilogram of colloid (cmolc+/kg); 2Dixon and Weed (1989); 3Bompoti et al. (2017); 4Goldberg et al. (1996)27 Within Colombia, most soil mineralogical studies have been conducted in Andisols. These soils are largely allophanic3 with a high content of SRO minerals; although some non-allophanic soils (high in organo-metallic complexes) are found in southern Colombia (Malagón, 1995). Hapludands, the great group of the study area, are allophanic and occupy 11% of the region (Malagón, 1995). In contrast, few studies consider Inceptisols, even though Dystrudepts (the dominant great group) occupy 55% of the region (Malagón, 1995). Given the great diversity of Inceptisols, generalization of their soil mineralogy is difficult. This chapter provides a comparison of the mineralogical properties of Andisols and Inceptisols located at two mid-elevation sites in the Colombian Andes. This study will help bridge the knowledge gap on the soil mineralogy of the two main soil types in this region and will highlight soil properties that may impact SWR characteristics. 3.2 Experimental conditions and laboratory analyses Soils were sampled as described in Section 2.3. From the 18 sampled soil pits, one pit from a natural forest and one pit from a pasture from each soil order were randomly selected for mineralogical analysis by X-ray diffraction (XRD). All composite samples of the 18 pits of each soil order, were analyzed for SRO minerals by dissolution extractions and for the following soil chemical and physical properties: pH H2O, pHCaCl2, soil organic carbon (SOC), particle size, bulk density (ρb) and particle density (ρs). Soil samples taken in 45 cm3 steel cores were used to determine ρb and ρs. Laboratory analyses of composite soil samples were performed on air dried samples that were passed through a 2-mm sieve, except for the XRD analysis, that was performed on the clay size fraction (<2 µm). 3.2.1 Mineralogy 3.2.1.1 Crystalline minerals X-ray diffraction (XRD) was performed using Bruker D8 Focus and D8 Advance diffractometers producing Co-K radiation (Thorez, 1976). The search-match software by 3 Allophanic soils are young, dark-colored soils derived mainly from volcanic ash (Juo and Franzluebbers, 2003). These soils typically have a low bulk density (< 900 kg/m3), a high water retention capacity (100% by weight at field capacity), and contain predominantly allophanes, imogolite, halloysite, and amorphous Al silicates in the clay size fraction. 28 Bruker (DIFFRACplus EVA 16) was used to identify mineral peaks with d-spacing and their relative intensities. As every mineral has a set of unique d-spacings, the minerals present were identified. Based upon their basal peak areas (Thorez, 1976), the relative abundance of the minerals was classified as dominant, significant, present, or trace as shown in Table 3.2. Table 3.2 Peak intensities determined by the X-ray diffraction (XRD) analysis and the correspondent relative abundance of minerals Peak relative intensities Relative abundance of minerals >40% to ≤100% Dominant >20% to ≤40% Significant >10% to ≤20% Present ≤10% Trace 3.2.1.2 Short-range order (SRO) minerals The standard XRD technique for identification of the mineral components in soils is very useful in the determination of crystalline minerals but given the amorphous nature of SRO minerals, XRD is very limited in their determination. Therefore, chemical dissolution extractions were used to assess the SRO minerals in the sampled soils. Pyrophosphate and oxalate extractions were done following the methods of Mizota and Van Reeuwijk (1989). Pyrophosphate extracts aluminum and iron (Alp and Fep) that are associated with organic matter (Dahlgren, 1994). Alp and Fep were determined by extraction after shaking 1 g of soil with 50 mL of 0.1 M pyrophosphate solution adjusted to pH 10, for 10-14 hours and filtering through a 2.5 m cellulose filter paper (Whatman #42) (Mizota and Van Reeuwijk, 1989). Oxalate solutions extract aluminum, iron, and silicon (Alo, Feo, and Sio) from organic complexes, and SRO minerals such as Fe-hydroxides (ferrihydrite) and aluminosilicates (imogolite and allophane) (Dahlgren, 1994). Oxalate extractable Alo, Feo, and Sio were determined by extraction after shaking 0.5 g of soil with 25 mL of 0.03 M ammonium 29 oxalate for 4 hours at pH 3 in the dark, filtering through cellulose filter paper (Whatman # 42) and diluting to 50 mL with 10% HNO3. Extractable Al, Fe, and Si were analyzed using an inductively coupled plasma atomic emission spectrometer (ICP-AES). The results of the pyrophosphate and oxalate extractions are used to determine the allophanic or non-allophanic character of Andisols and to quantify the amounts of SRO mineraloids. The following indices were determined: • Alp/Alo ratio is used to distinguish between soils rich in SRO minerals (Alp/Alo <0.5) from soils rich in organo-metallic complexes (Alp/Alo >0.5) (Shoji et al., 1993). Andisols rich in SRO are referred to as allophanic, while Andisols rich in humus complexes are non-allophanic. • Alo + ½ Feo (%), referred to as oxalfe, is used to identify andic properties, a key attribute of Andisols. If oxalfe is greater than 2% then this requirement is met (Van Wambeke, 1992). The quantity of imogolite and allophane was estimated according to the equation developed by Mizota and Van Reeuwijk (1989): % Allophane and imogolite=100 % Sio23.4-5.1 x Eqtn 2.1 Where: x = (Alo-Alp)/Sio represents the molar ratio of Alsro/Sio and Alsro = Alo – Alp, gives a measure of Al associated with SRO inorganic material (Parfitt and Childs, 1988; Dahlgren, 1994). Ferrihydrite occurs in soils undergoing rapid weathering, and in soils containing soluble silicate or organic anions which inhibit the formation of more crystalline iron oxides (Childs, 2007). Ferrihydrite concentration was estimated from Fesro, according to Childs et al. (1991): % Ferrihydrite = 1.7 x % Fesro Eqtn. 2.2 30 Where: Fesro = Feo – Fep, gives a measure of Fe associated with SRO inorganic material (Parfitt and Childs, 1988; Dahlgren, 1994). 3.2.2 Soil chemical and physical properties 3.2.2.1 Soil pH Soil pH was determined in both distilled water and 0.01 M CaCl2 on a 1:2 (v/v) ratio (Hendershot and Lalande, 1993). For samples high in organic matter (>12% by weight), the liquid volume was doubled. The pH results for the samples analyzed with double liquid of volume are indicated in Appendix B. This appendix also provides data on soil depth, slope, air dried color, SOC, ρb, and texture. 3.2.2.2 Soil organic carbon (SOC) Soil organic carbon (SOC) was determined by loss-on-ignition (LOI); 5 g of air-dried soil was dried at 105°C for 24 hours and then heated to 300°C for 8 hours in a muffle furnace. Weights were recorded before and after combustion. Soil organic carbon was calculated using the equation formulated by Rahman et al. (2011): SOC = (0.5663LOI) - 0.7589 Eqtn. 2.3 LOI = (W105°C - W300°C) / W105°C Eqtn. 2.4 Where: LOI is the % of weight associated to organic matter, W105°C: Initial soil weight after drying at 105°C for 24 hours, and W300°C: Final soil weight after drying at 300°C for 8 hours. 3.2.2.3 Particle size analysis Determination of texture in Andisols by mechanical (or particle) analysis is challenging due to incomplete dispersion of the mineral particles even with vigorous shaking and the addition of sodium hexametaphosphate as a dispersion agent (Shoji et al., 1993). The incomplete dispersion of mineral particles in Andisols is due to the presence of amorphous 31 minerals and the associated high stability of soil aggregates. Ultrasonic dispersion with a pH adjustment to 4 or 10 using HCl or NaOH may be utilized; however, there is no standardized procedure (Shoji et al., 1993). For this study, the hydrometer method was used for both Andisols and Inceptisols for comparison purposes. Interpretation of particle size analysis for the Andisol samples should consider the potential impact of incomplete dispersion. Particle size analysis was done with a hydrometer according to the Bouyoucos method (Bouyucous, 1962). All samples were oxidized using H2O2 to remove organic matter and dispersed with sodium hexametaphosphate. For 10 samples, insufficient sample volume was available for the Bouyoucos method and particle size was determined following the sieving / sedimentation method of Kettler et al. (2001). Pre-treatment procedures with H2O2 and sodium hexametaphosphate were identical for all samples. The samples analyzed by the Kettler method are identified in Appendix B. 3.2.2.4 Soil bulk and particle density Bulk density (ρb) was determined gravimetrically. Soil samples from the 45 cm3 steel cores were dried at 105°C for 24 h and then weighed. Bulk density was assessed by dividing the dry soil weight over the core volume (Blake, 1965). Soil particle density (ρs) was determined using the pycnometer method (Blake, 2008). 3.2.3 Comparative and statistical analyses Crystalline minerals were compared by soil horizon between Andisols and Inceptisols (Fig. 3.1a). Chemical dissolution results (i.e., Alp, Fep, allophane and imogolite, ferrihydrite, and the indicators: Alp/Alo and Alo + ½ Feo,), were compared by horizons between Andisols and Inceptisols to evaluate significant differences using the Mann Whitney U test and probability values (p) of 0.01, 0.05, and 0.1 (Fig. 3.1b). In addition, relationships were assessed between SRO, and soil physical and chemical properties utilizing the non-parametric Spearman’s rank-order correlation. Spearman’s Rho correlation coefficients (r) with values > 0.4 and probability values (p) of 0.01 and 0.05 were used to indicate notable relationships. 32 Correlations were conducted for: • Andisol and Inceptisol all horizons ▪ Andisol and Inceptisol by horizon (A, B, and C horizons) o Andisol only, all horizons ▪ Andisol by horizon (A and B horizons) o Inceptisol only, all horizons ▪ Inceptisol by horizon (A and B horizons) Correlations for the C horizon in individual soil types were not assessed since there were only eight samples for Andisols and five for Inceptisols. Sample numbers for each analysis are provided in Figure 3.1. Figure 3.1 Overview of samples collected to compare: a) crystalline minerals of Andisol and Inceptisols; b) short-range order (SRO) minerals of Andisol and Inceptisol; and c) relationships between SRO minerals and soil properties33 3.3 Results and discussion 3.3.1 Comparison of crystalline minerals X-ray diffraction (XRD) results presented in this section will be used to assess differences in primary and secondary minerals between Andisols and Inceptisols, the degree of weathering, and the mineralogical characteristics of the soil parent material. X-ray diffractograms and primary and secondary minerals for the pasture and forest samples within each soil order were similar (Appendix D). Thus, results for XRD from the randomly selected pasture soils are presented here (Figs. 3.2 and 3.3); A, B and C horizons were analyzed for both soil orders. XRD results suggest the presence of SRO minerals in both Andisols and Inceptisols. In Andisols (Fig. 3.2) the initial gently decreasing slope, notably in the B horizons of the diffractograms, suggests short-range order (SRO) minerals. In Inceptisols, although the slope was less steep than that recorded for Andisols, it still suggests the presence of SRO minerals (Fig. 3.3). 34 1Mineral names in gray are of secondary peaks Figure 3.2 X-ray diffractograms of A, B1 and C horizons of the P4 pasture profile located in the Sonora watershed (Andisol site) 05000100001500020000250000 10 20 30 40 50 60Counts2AB1QuartzCIllite or micasAntigorite and crysitileCristobaliteMagnetiteChlorite or vermiculiteAntigorite and crysotileHydrobiotiteAmphibolesMetahalloysite, kaoliniteQuartzCa and Mg carbonatesHematiteNa plagioclaseK feldsparsHydrobiotite1HydrobiotiteAmphibolesAmphibolesNa plagioclase35 1Mineral names in gray are of secondary peaks Figure 3.3 X-ray diffractograms of A, B and C horizons of the P2 pasture profile located in El Chocho watershed (Inceptisol site) 020004000600080001000012000140001600018000200000 10 20 30 40 50 60Counts2CMeta halloysite,kaoliniiteQuartzMagnetiteGoethiteNa feldsparsQuartzCaand Mg carbonatesQuartzHematitePartially dehydrated halloysiteMeta halloysite, kaoliniite1GoethiteNa feldsparsCa and Mg carbonatesCa and Mg carbonatesBoehmiteBA36 Primary and secondary minerals found in the Andisol and Inceptisol samples and their relative abundances (Table 3.3) indicate the state of weathering of both soils and the predominantly mafic nature of the parent material. Results show the young pedogenic age of Andisols. In Andisols, halloysite may be formed by the weathering of imogolite (Cortes and Franzmeier, 1972). Halloysite was found only in the C horizon and there were no crystalline silicates in the A or B horizons. This limited presence of halloysite and other crystalline silicates indicates the limited soil weathering processes in Andisols. The predominance of primary minerals with a mafic nature, such as cristobalite, micas, feldspars, and amphiboles, suggest the basic nature of the volcanic ash. Cristobalite was the most abundant mineral in the A, B and C horizon samples of Andisols. Cristobalite normally occurs in metamorphosed sandstones and sandstone xenoliths in basic rocks (Kämpf et al., 2012). This, combined with the presence of illite or micas, Na feldspars and amphiboles in the C horizon, suggests the predominantly mafic nature of the volcanic ash in the Sonora watershed. When sufficient bases, notably calcium (Ca) and magnesium (Mg) are present, they neutralize the carboxyl groups of organic acids and relatively high pH values prevail suppressing the formation of Al- and Fe-humus complexes (Van Breemen and Wielemaker, 1974). In C horizon samples, quartz was also found but was less abundant than in the A horizon, which may indicate that quartz (a primary mineral), may have been transported by alluvial processes (Guarín et al., 2004) and is therefore predominant in surface horizons. In contrast to the limited weathering in Andisols, Inceptisols show more advanced weathering. There were abundant secondary minerals such as kaolinite and Ca and Mg carbonates. Metahalloysite and partially dehydrated halloysite were found in all horizons in the Inceptisols. The presence of kaolinite and halloysite suggests an intense weathering of the sedimentary, diabase and basalt parent rocks, the typical parent material in this region (Section 2.1.1). These parent rocks, diabase and basalt, are also mafic rocks. Magnetite and goethite (Fe oxides), and boehmite (Al oxide) present in the C horizon; and hematite (Fe oxide) present in the A and B horizons, may have also been formed as weathering products of the parent material. 37 Table 3.3 Mineral abundance in Andisols and Inceptisols Minerals abundance Andisols Inceptisols Primary minerals Secondary minerals Primary minerals Secondary minerals A and B horizons Dominant Cristobalite and quartz - Quartz Kaolinite/ Metahalloysite Significant Hydrobiotite and plagioclase feldspars - Plagioclase feldspar Hematite Low amounts Amphiboles, chrysotile and antigorite Chlorite or vermiculite - Partially dehydrated halloysite, Ca and Mg carbonates, and magnetite Traces K feldspars Ca and Mg carbonates, magnetite and hematite - Boehmite C horizon Dominant Cristobalite Kaolinite/ metahalloysite Quartz Kaolinite/ metahalloysite Significant - - - Boehmite Low amounts Quartz, crysotile and antigorite, and Na feldspars - - Partially dehydrated halloysite Traces Hydrobiotite, amphiboles, illite or micas and K feldspars Chlorite or vermiculite, Ca and Mg carbonates, magnetite and hematite Plagioclase feldspars Hematite, magnetite, goethite, Ca and Mg carbonates The mineralogical characteristics of soils in the Sonora watershed are in agreement with studies conducted by Malagón et al. (1995) in the central part of Colombia, where they 38 concluded that: first, Andisols are young soils (decades to centuries) suggested by the limited presence of halloysite and other crystalline silicates; and second, Andisols are developed on mafic ash, with cation rich primary minerals such as amphiboles and plagioclase feldspars. In contrast to the young age of these soils, the Andisols studied by Buytaert et al. (2005a) in páramo ecosystems of Ecuador, were pedologically older as suggested by the abundance of kaolinite and gibbsite. In addition, the Andisol at the Ecuador site, contained more felsic minerals such as K-micas. In a study of 14 soil profiles from the Western and Central branches of the Andes, Mejia el al. (1968) found a profile of the Western branch showing kaolinite > quartz, cristobalite, and gibbsite. There are two similarities between the Inceptisols of this study and the profile studied by Mejía et al. (1968): first, kaolinite is the predominant secondary mineral in soils in both studies, showing the intense weathering; and second, diabase was the parent material at both sites. 3.3.2 Short-range order (SRO) minerals and organo-metallic complexes Chemical dissolution was used for comparative purposes between Andisols and Inceptisols, although extractions with pyrophosphate and ammonium oxalate are largely used for Andisols and Spodosols, which are commonly associated with SRO minerals or organo-metallic complexes (Ugolini and Dahlgren, 1991; Algoe et al., 2012). The results of pyrophosphate and oxalate extractions are used to determine parameters for soils containing SRO minerals or organo-metallic complexes. In this study these parameters include: andic properties, the type of Andisols (allophanic or non-allophanic) and the quantity of SRO minerals and organo-metallic complexes. Table 3.4 gives median dissolution extraction values for SRO minerals and indices, and identifies statistically significant differences between Andisols and Inceptisols. All parameters were significantly different in at least one horizon, and most were different in both A and B horizons. 39 Table 3.4 Median dissolution extraction values and indices in A, B and C horizons of Andisols and Inceptisols 1Pyrophosphate-extractable aluminum (Al) and iron (Fe) (Alp and Fep); 2Values in parenthesis are first and third quartile; 3Indicator for determining andic properties in soils, called “oxalfe” (Alo +1/2 Feo); 4Indicator for determining allophanic or non-allophanic soils (Alp/Alo); Number of samples (n) for Andisols were 17, 19 and 8 and for Inceptisols 18, 19 and 5 for A, B and C horizons, respectively; *, ** significant differences between Andisols and Inceptisols with Mann Whitney U test at p<0.05 and p<0.01, respectively Horizon Andisols InceptisolsA 1.4 (1.1-1.7)2 0.06 (0.05-0.13)**B 0.4 (0.3-0.6) 0.08 (0.03-0.14)**C 0.1 (0.1-0.2) 0.05 (0.02-0.14)A 0.3 (0.2-0.5) 0.02 (0.01-0.05)**B 0.009 (0.006-0.030) 0.012 (0.004-0.041)C 9x10-4 (2x10-4-0.01) 0.007 (0.001-0.02)A 6.2 (4.5-8.0) 1.4 (1.2-2.1)**B 14.8 (9.9-16.4) 1.2 (1.1-1.6)**C 4.0 (1.5-6.7) -A 0.5 (0.4-0.5) 1.2 (0.9-1.4)**B 0.7 (0.3-0.8) 0.6 (0.4-1.1)C 0.2 (0.1-0.3) 0.6 (0.5-0.8)*A 1.7 (1.4-1.8) 0.7 (0.6-0.8)**B 3.7 (2.3-4.2) 0.5 (0.4-0.7)**C 1.0 (0.4-1.5) 0.6 (0.4-0.7)A 0.10 (0.06-0.13) 0.02 (0.01-0.03)**B 0.011 (0.07-0.016) 0.02 (0.01-0.04)*C 0.014 (0.005-0.045) 0.01 (0.01-0.03)Alp/Alo4Alo + 1/2 Feo3 (%)Allophane and imogolite (%)Ferrihydrite (%)Alp1 (g/kg)Fep1 (g/kg)40 3.3.2.1 Organo-metallic complexes Organo-metallic complexes were not high in either Andisols or Inceptisols. However, in Andisols formation of organo-metallic complexes was observed; Alp and Fep were higher in the surface horizon relative to deeper horizons (Figs. 3.4 and 3.5), and this increasing trend suggests the formation of Al and Fe-humus complexes in the epidedon of Andisols. Furthermore, Fep is lower than Alp in both soils and was not different between soil orders in the B horizon, indicative of less formation of Fe-humus complexes relative to Al-humus complexes. The predominance of organo-metallic complexes is common in weathered Andisols. Therefore, the low amounts of organo-metallic complexes in the Andisols of this study is in accordance with its young pedogenic stage. a) Andisols b) Inceptisols Figure 3.4 Median and quartiles of pyrophosphate extractable aluminum (Alp) in the A, B, and C horizons of: a) Andisols and b) Inceptisols 0,0 0,1 1,0 10,0ABCLog Alp (g/kg)Horizon0.0 0.1 1.0 10.0 0,0 0,1 1,0 10,0ABCLog Alp (g/kg)Horizon0.0 0.1 1.0 10.0n = 5 n = 17 n = 19 n = 8 n = 18 n = 19 41 a) Andisols b) Inceptisols Figure 3.5 Median and quartiles of pyrophosphate extractable iron (Fep) in the A, B, and C horizons of: a) Andisols and b) Inceptisols 3.3.2.2 Short-range order (SRO) minerals Oxalate extraction results show substantial amounts of allophane and imogolite (> 8%) in the B horizon of Andisols, while ferrihydrite was highest (<2%) in the A horizon of Inceptisols. Allophane and imogolite in Andisols were significantly higher in A and B horizons in comparison to Inceptisols (Table 3.4 and Fig. 3.6). Unlike allophane and imogolite, which were higher in Andisols, ferrihydrite was more abundant in Inceptisols and in greater concentration in comparison to Andisols (Table 3.4 and Fig. 3.7). The higher active Feo in Inceptisols may be related to the basaltic origin of these soils, which in turn could release compositional Fe to form Fe-hydroxides. Even in humus rich horizons, Fe-humus complexes are very low due to the greater stability of Fe-hydroxides compared to Fe-humus complexes (Shoji et al., 1993). 0,0001 0,0010 0,0100 0,1000 1,0000ABCLog Fep (g/kg)Horizon0. 1 . 01 0.01 0.1 1. 0,0001 0,0010 0,0100 0,1000 1,0000ABCLog Fep (g/kg)Horizon0.0 . 0.01 0.1 1.n = 5 n = 17 n = 19 n = 8 n = 18 n = 19 42 a) Andisols b) Inceptisols Figure 3.6 Median and quartiles of allophane and imogolite in the A, B, and C horizons of: a) Andisols and b) Inceptisols a) Andisols b) Inceptisols Figure 3.7 Median and quartiles of ferrihydrite in the A, B, and C horizons of: a) Andisols and b) Inceptisols The grouping criteria formulated by Mizota and Van Reeuwijk (1989) was used to determine if the amounts of SRO minerals found in the allophanic Andisols of the Sonora 0,0 0,5 1,0 1,5 2,0ABCFerryhidrite (%)Horizon0.0 0.5 1.0 1.5 2.0 0,0 0,5 1,0 1,5 2,0ABCFerryhidrite (%)Horizon0.0 0.5 1.0 1.5 2.0n = 17 n = 19 n = 8 n = 18 n = 19 n = 5 n = 17 n = 19 n = 8 n = 18 n = 19 43 watershed were substantial (i.e., >8% of SRO minerals in at least one horizon). Based on this criteria, the Andisols of the Sonora watershed have substantial amounts of allophane and imogolite, as 17 of 19 samples of B horizons registered values >8% allophane (Fig. 3.6a). 3.3.2.3 Andic properties Oxalfe or Alo + ½ Feo (%), is one of the requirements for identifying andic properties, a key attribute of Andisols. Andic properties reflect the presence of volcanic ejecta such as ash, pumice, lava or they indicate the presence of SRO minerals. If oxalfe is greater than 2% then the andic requirement is met (Van Wambeke, 1992). Oxalfe was significantly higher in A and B horizons of Andisols in comparison to Inceptisols. Oxalfe in the Sonora watershed was greater than 2% in 18 of 19 samples from the B horizon, with a median value of 3.7% (Table 3.4), while in the A horizon only 3 samples met the index. In Inceptisols, due to the low quantities of Alo, oxalfe was below 2% and therefore Inceptisols from El Chocho watershed did not meet this criterion for andic properties, which is in accordance with the parent material of this region: igneous and sedimentary rocks (SGC, 1984b). 3.3.2.4 Allophanic and non-allophanic soils Alp/Alo ratio is an indicator used to distinguish between soils rich in SRO minerals (Alp/Alo <0.5) from soils rich in metal humus complexes or organo-metallic complexes (Alp/Alo >0.5) (Shoji et al., 1993). Andisols rich in SRO are classified as allophanic while Andisols rich in humus complexes are classified as non-allophanic. Median Alp/Alo ratios in the Sonora watershed were all <0.5, indicating the predominance of SRO and that Andisols of the Sonora watershed are allophanic in A and B horizons. Alp/Alo was significantly higher in the A and B horizons of Andisols in comparison to Inceptisols, showing the low amounts of organo-metallic Al in Inceptisols (Table 3.4). The ratio of Alp/Alo versus soil organic carbon (SOC) for Andisols from the Sonora watershed are presented in Figure 3.8, and the low Alp/Alo ratios are evident. Mizota and 44 Van Reeuwijk (1989) defined an evolutionary criteria using Alp/Alo: a weathered Andisol will have two opposite Alp/Alo values; one close to 1 in the epidedon (highly humic and acid soil with lower pH) and the other close to 0 in the deeper horizons (lower organic matter and higher pH). Andisols of the Sonora watershed showed a very low value in the B horizon, suggesting a weathered or old Andisol but also a low value in the A horizon, which suggests a young Andisol. Even though these results appear contradictory, there are two conditions that may explain these results. First, the presence of minerals such as amphiboles and plagioclase feldspars, that due to their basic nature, promote the formation of allophane and imogolite rather than Fe- and Al-humus complexes (Shoji and Fujiwara, 1984; Parfitt and Saigusa, 1985) and second, there may be an ongoing input of volcanic ash from active volcanos in the vicinity of the Sonora watershed, that inhibits the formation of metal-humus complexes in the A horizon. Figure 3.8 Relationship between pyrophosphate and oxalate extractable aluminum (Alp/Alo) and soil organic carbon (SOC) in A, B and C horizons of Andisols of the Sonora watershed The mineralogical results for the Andisol site of this study are in accordance with the definition of the evolutionary tendency of Andisols in Colombia by Malagón et al. (1995). They state that the evolution of Andisols in Colombia is commonly the formation of 05101520250,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0SOC (%)Alp/AloA horizon, n= 17B horizon, n= 19C horizon, n= 80.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0Allophanic soils Non-allophanic soils45 humus, and allophane and imogolite (with pH between 5.2 and 5.7) rather than the formation of organo-metallic complexes. However, Malagón et al. (1995) noted two exceptions: Andisols with a tendency to form organo-metallic complexes (pH between 4.3 and 5.3) found in Nariño and Cauca departments meaning non-allophanic soils; and an allophanic soil in Tolima department but with less moisture and higher pH (pH between 5.8 and 6.0). In Ecuador, Buytaert et al. (2005a) obtained results indicating a non-allophanic soil where Alp/Alo >0.58 in the epipedon with a pH between 4.3 and 5.1. Although the mineralogy of the sites in Colombia studied by Malagón et al. (1995), were similar to that found in this study, where the mafic minerals plagioclase and amphiboles are the prevailing ones, the ecosystem conditions in Nariño and Cauca departments are different and may explain the differences in the soil; a higher elevation (>3,100 m), with a wetter and colder climate, may favor the accumulation of humus, which in turn decreases the pH, favoring the formation of metal-humus complexes. In the case of the non-allophanic Andisols of Ecuador, they have developed over older volcanic ash deposits with more felsic minerals and lower pH, which also favor the formation of metal-humus complexes. Another interesting observation of the Andisols in the Sonora watershed was the SOC content found in the A and B horizons (Fig. 3.8). It has been suggested by Shoji et al. (1993) that a SOC value of 6% separates the A and B horizons, since the majority of non-allophanic soils (Alp/Alo >0.5) have SOC >6%. The Andisol of Sonora watershed has SOC >6% in the A horizon even though it is an allophanic soil. A high Alp/Alo ratio (dominance of Al-humus complexes or non-allophanic soil) is associated with a high organic matter content, but the opposite is not necessarily true (i.e., an Andisol with high organic matter content is not necessary associated with a non-allophanic soil) (Mizota and Van Reeuwijk, 1989). These results show that the Andisols of this study are allophanic, due to the predominantly mafic nature of the volcanic ash (on which these soils formed) and the ongoing input of ash. This is in agreement with the pedogenic young age of these Andisols. Interestingly, the allophanic Andisols of this study have considerable amounts of SOC (>6%), even though 46 these levels of SOC are more common in non-allophanic soils, in which organo-metallic complexes are dominant. 3.3.3 Relationships between short-range order (SRO) minerals and soil properties Correlations among all parameters for the two soil orders are given in Appendix E, Table E1. There were few significant correlations using this combined data set, therefore correlations between SRO minerals, and physical and chemical soil properties were assessed for each soil independently. Spearman´s Rho correlation coefficients (r) for Andisols and Inceptisols are presented in Table 3.5 for all horizons and by A and B horizons, individually. Statistically significant relationships are found in Andisols between SRO minerals and organo-metallic complexes and pH, texture, SOC and bulk density. In contrast, within Inceptisols only a limited number of correlations were found. 47 Table 3.5 Spearman´s Rho correlation coefficients (r) between short-range order (SRO) minerals and organo-metallic complexes (Alp and Fep) with soil physical and chemical soil properties in: a) Andisols all horizons and by horizon; b) Inceptisols all horizons and by horizon 1Pyrophosphate-extractable aluminum (Al) and iron (Fe) (Alp and Fep); 2SOC: soil organic carbon; 3ρs: particle density; ρb: bulk density; correlations shown are the ones r > 0.4; Gray cells show correlations with p < 0.01 and white cells show correlations with p < 0.05 Alp1 (g/kg)Fep1 (g/kg)Allophane (%)Ferrihydrite (%)Alp1 (g/kg)Fep1 (g/kg)Allophane (%)Ferrihydrite (%)Alp1 (g/kg)Fep1 (g/kg)Allophane (%)Ferrihydrite (%)a) Andisols All horizons (n= 44) A horizon (n= 17) B horizon (n= 19)pH H2O -0.46 -0.47 0.56 0.50pH CaCl2 0.69 0.51Sand (%) 0.52 -0.65 -0.55 0.68Silt (%) 0.41 0.46 0.59 0.57 -0.66Clay (%) -0.59 0.62SOC2 (%) 0.82 0.82 -0.55 0.71ρs3 (kg/m3)ρb3 (kg/m3) -0.45 -0.43 -0.51 -0.76b) Inceptisols All horizons (n= 42) A horizon (n= 18) B horizon (n= 19)pH H2O -0.53 -0.60pH CaCl2 -0.57 -0.67Sand (%) 0.55 0.59Silt (%) -0.52Clay (%)SOC2 (%) 0.51 0.59ρs3 (kg/m3)ρb3 (kg/m3)48 3.3.3.1 Relationships between organo-metallic complexes, pH and soil organic carbon (SOC) Within Andisols, strong relationships were found among organo-metallic complexes, pH and SOC. Organo-metallic complexes are predominately formed in the A horizon with higher SOC (Figs. 3.9c and 3.9d). Higher quantities of SOC and organometallic complexes in turn are related to lower pH in the A horizon (Figs. 3.9a and 3.9b). Although SRO minerals were predominant in the Andisols of this study, organo-metallic complexes were formed in the A horizon where there was higher SOC and lower pH in comparison to the B horizon. a) b) c) d) Figure 3.9 Relationships between: a) pyrophosphate extractable aluminum (Alp) and pHH2O; b) pyrophosphate extractable iron (Fep) and pHH2O; c) Alp and SOC; and d) Fep and SOC in A, B and C horizons of Andisols n = 44 r = -0.46 p = <0.01 n = 44 r = -0.47 p = <0.01 n = 19 n = 17 n = 19 n = 17 n = 44 r = 0.82 p = <0.01 n = 19 n = 17 n = 17 n = 19 n = 44 r = 0.82 p = <0.01 49 In contrast to the relationships found in Andisols, there was no significant correlation between pyrophosphate extractable iron (Fep) and SOC in Inceptisols (Table 3.5b and Fig. 3.10); however, there was a relation with pH. The lack of relationships between Fep and Alp with SOC, suggests that these metals were not bound strongly to organic matter in Inceptisols. As shown by Kaiser and Zech (1996), pyrophosphate may extract Al attached to hydroxides such as boehmite and gibbsite. It may also extract Fe attached to Fe-oxides such as ferrihydrite and goethite (Parfitt and Childs, 1988). The presence of these minerals in Inceptisols was confirmed by XRD. Boehmite was found in significant amounts in the C horizon and trace amounts in both A and B horizons (Table 3.3). Also, goethite, hematite and magnetite were present in most horizons (Table 3.3). Ferrihydrite was also assessed via oxalate extraction and it was found in higher concentrations in the A horizon of Inceptisols compared to any horizon in Andisols (Table 3.3). There was a correlation between Alp and pHH2O in the B horizon of Inceptisols (Fig. 3.10a), indicating a lower pH with higher concentrations of aluminum oxides (i.e., boehmite). The lack of a relationship between organo-metallic complexes and SOC in Inceptisols, indicates that pyrophosphate extractable Al and Fe are related to iron and aluminum oxides. These results are in accordance with XRD results (Section 3.3.1). 50 a) b) c) d) Figure 3.10 Relationships between: a) pyrophosphate extractable aluminum (Alp) and pHH2O; b) pyrophosphate extractable iron (Fep) and pHH2O; c) Alp and SOC; and d) Fep and SOC in A, B and C horizons of Inceptisols 3.3.3.2 Relationships between short-range order (SRO) minerals, pH and soil organic carbon (SOC) As was the case with the organo-metallic complexes in Andisols, SRO minerals were also correlated with pH and SOC. Andisols of the Sonora watershed, although being allophanic, show a clear differentiation in pH between the two horizons: a pHCaCl2 value of 4.5 separates the A and B horizons in Andisols (Fig. 3.11a). This difference in pH, favors allophane and imogolite formation in the B horizon, while inhibiting SRO mineral formation in the A horizon. It is known that pH is an important factor in the formation of SRO minerals; at lower pH, higher amounts of organic matter and felsic minerals favor the formation of metal-humus complexes, while at higher pH, mafic minerals favor the 05101520250,0 0,1 0,2 0,3 0,4SOC (%)Alp (g/kg)0.0 . 0.2 . 0.4 05101520250,00 0,05 0,10 0,15 0,20SOC (%)Fep (g/kg)0.00 0.05 0.10 . 0.20 n = 42 r = -0.53 p <0.01 n = 19 r = -0.60 p <0.01 n = 18 n = 42 n = 19 n = 18 n = 42 n = 19 n = 18 n = 42 n = 19 n = 18 51 formation of SRO minerals. This can be seen by the correlation between pHCaCl2 and allophane and imogolite in Andisols (Fig. 3.11a). In the B horizon of Andisols, there was a positive correlation between ferrihydrite and SOC (Fig. 3.11d), which suggests a stabilizing effect of SRO on SOC (Broadbent et al., 1964; Dahlgren et al., 2004; Egli et al., 2008; Parfitt, 2008). This is in accordance with Kaiser et al. (2011) and Regelink et al. (2013) who conclude that ferrihydrite dominates the surface area available for sorption of SOC, stabilizing it and forming organo-mineral assemblages, despite their small contribution to soil mass (<1%) (Regelink et al., 2013). a) b) c) d) Figure 3.11 Relationships between: a) allophane and imogolite and pHCaCl2; b) ferrihydrite and pHH2O; c) allophane and imogolite and soil organic carbon (SOC); and d) ferrihydrite and SOC in A, B and C horizons of Andisols 048121620240,0 0,2 0,4 0,6 0,8 1,0 1,2SOC (%)Ferrihydrite (%)0.0 0.2 0.4 0.6 0.8 1.0 1.2 n = 19 r = 0.71 p < 0.01 n = 44 r = 0.69 p < 0.01 n= 17 r = -0.55 p = 0.02 n = 44 n = 19 n = 17 r = 0.50 p = 0.04 n = 44 n = 44 n = 19 n = 17 n = 19 r = 0.51 p = 0.02 n = 17 52 There was also a positive correlation between ferrihydrite and SOC in the B horizon of Inceptisols (Fig. 3.12) as occurred with Andisols. This may be interpreted as ferrihydrite having a stabilizing effect on SOC in Inceptisols as well. Figure 3.12 Relationship between ferrihydrite and soil organic carbon (SOC) in A, B and and C horizons of Inceptisols The stabilizing effect of ferrihydrite on SOC is important since assemblages may be less susceptible to decomposition (Broadbent et al., 1964; Dahlgren et al., 2004; Egli et al., 2008; Parfitt, 2008). Both the rate and stability of aggregation generally increases with SOC and surface area (Bronick and Lal, 2005), both of which are relatively high in Andisols and Inceptisols. Thus, both Andisols and Inceptisols at mid-elevations in the Colombian Andes have mineralogical properties that benefit carbon stabilization, in comparison to soils without SRO minerals. 3.3.3.3 Relationships between short-range order (SRO) minerals and organo-metallic complexes with bulk density (b) Bulk density (b) was correlated with SRO minerals and organo-metallic complexes in Andisols, but not in Inceptisols. Low bulk density (b) is characteristic of Andisols with values typically ranging from 400 to 800 kg/m3 (Shoji et al., 1993). Median b in the Sonora watershed was 566 kg/m3 in the A horizon and increased with depth to 667 kg/m3 0481216200,0 0,5 1,0 1,5 2,0SOC (%)Ferrihydrite (%)A horizon B horizon All including C horizon. . 1.0 . . n = 19 r = 0.59 p < 0.01 n = 18 n = 42 r = 0.51 p < 0.01 53 and 766 kg/m3in the B and C horizons, respectively. ρb decreased with the increase of SRO minerals in the B horizon (Figs. 3.13a and 3.13b), and with the increase of organo-metallic Fe, which was higher in the A horizon (Fig. 3.13d). a) b) c) d) Figure 3.13 Relationships between: a) allophane and imogolite with bulk density (b); b) ferrihydrite with b; c) pyrophosphate extractable aluminum (Alp) with b; d) pyrophosphate extractable iron (Fep) with b in A, B and C horizons of Andisols Median b in Inceptisols of the El Chocho watershed was 852 kg/m3 in the A horizon and increased with depth to 990 kg/m3 and 972 kg/m3 in the B and C horizons, respectively. 20040060080010001200140016000,0 0,2 0,4 0,6 0,8 1,0 1,2ρb(kg/m3)Ferrihydrite (%)A horizon B horizon All including C horizon0.0 . 0.4 . . . . 2004006008001000120014000,0 1,0 2,0 3,0ρb(kg(m3)Alp (g/kg)0.0 0.5 1.0 1.5 . 2.5 3.02004006008001000120014000,0 0,2 0,4 0,6ρb(kg/m3)Fep (g/kg)0.0 . . 0.6 n = 19 r = -0.51 p = 0.03 n = 19 r = -0.76 p < 0.01 n = 17 n = 44 n = 17 n = 44 r = -0.45 p < 0.01 n = 44 r = -0.43 p < 0.01 n = 19 n = 19 n = 17 n = 44 n = 17 54 3.3.3.4 Relationships between short-range order (SRO) minerals and organo-metallic complexes with soil particle size The high specific surface area and the pH dependent charge (Eusterhues et al., 2005), allow SRO and organo-metallic complexes to combine silt and clay particles with organic matter into secondary structures or aggregates (Arias et al., 1996; Sei et al., 2002; Kaiser et al., 2011; Regelink et al., 2013). In both Andisols and Inceptisols, the measured sand size fraction increased with an increase in SRO minerals. In the B horizon of Andisols, the sand fraction increased with higher content of allophane and imogolite (Fig. 3.14a) and with less Al and Fe organo-metallic complexes (Figs. 3.14c and 3.14d). In Inceptisols, the sand size fraction increased in the A horizon with the increase of allophane and imogolite (Fig. 3.15a) and in the B horizon with the increase of ferrihydrite (Fig. 3.15b). Allophane, imogolite, and ferrihydrite can cement and coat aggregates (Goldberg et al., 2012; Shoji et al., 1993) and contribute to an over estimation of the sand size fraction. In addition, the polyphenolic groups of organic matter through their hydrophobic properties, may further stabilize aggregates (Bronick and Lal, 2005; Lal, 2007). 55 a) b) c) d) Figure 3.14 Relationships between: a) allophane and imogolite and sand particle size (sand); b) ferrihydrite and sand; c) pyrophosphate extractable Al (Alp) and sand; and d) pyrophosphate extractable Fe (Fep) with sand in A, B and C horizons of Andisols 0204060801000,0 0,2 0,4 0,6 0,8 1,0 1,2Sand (%)Ferrihydrite (%)A horizon B horizon All including C horizon0.0 0.2 0.4 . 0.8 1.0 .2 0204060801000,0 0,5 1,0 1,5 2,0 2,5 3,0Sand (%)Alp (g/kg). 0.5 . . . . 3.00204060801000,0 0,2 0,4 0,6Sand (%)Fep (g/kg)0.0 . . . n = 19 r = 0.68 p < 0.01 n = 19 r = -0.65 p < 0.01 n = 19 r = -0.55 p = 0.01 n = 44 r = 0.52 p < 0.01 n = 17 n = 19 n = 44 n = 17 n = 44 n = 17 n = 44 n = 17 56 a) b) Figure 3.15 Relationships between: a) allophane and imogolite and sand particle size (sand); and b) ferrihydrite and sand in A and B horizons of Inceptisols The majority of the Andisol samples had < 40% clay size fraction, with lower variability in the B horizon, thus classifying the soils as sandy loam, loam or sandy clay loam. C horizon samples ranged from sandy loam to clay (Fig. 3.16a). In contrast, most samples of Inceptisols had > 40% clay size fraction (Fig. 3.16b). The majority of A and B horizon samples had clay texture followed by silty clay loam. There are only a few samples, largely in the C horizon, in the sandy clay loam texture class. a) Andisols b) Inceptisols Figure 3.16 Soil texture triangles showing textural classes for the A, B, and C horizons of: a) Andisols and b) Inceptisols 0204060801000,0 0,5 1,0 1,5 2,0Sand (%)Ferrihydrite (%)A horizon B horizon0.0 . . . . n= 18 r = 0.55 p = 0.02 n = 19 r = 0.59 p < 0.01 n= 19 n = 18 57 Although texture in Andisols is not a differentiating property for separating Andisols from other orders, findings of this study were similar to studies in Colombia by Diaz and Paz (2002) and Hincapié (2011) that classified Andisols as sandy soils, with >40% sand size particles. Diaz and Paz (2002) used the Bouyucous method and Hincapié (2011) used the pipette method after oxidizing the organic matter with H2O2. Previous studies of the Typic Dystrudepts (Inceptisols) in the El Chocho watershed reported that the soils have a clay texture (IGAC and CVC, 2004). Those results are in agreement with the texture classes found in this study. 3.4 Conclusions The Andisols and Inceptisols of this study are both formed on mafic parent materials. In Andisols, primary minerals of mafic nature such as amphiboles and micas were predominant, while in Inceptisols, secondary minerals which are products of the weathering of mafic rocks (diabase and basalt) were predominant. The Andisols, despite being located in the oldest branch of the Andean mountains, are young soils. This may be due to the sporadic deposition of ash and its mafic nature. These mafic minerals increase pH and favor the formation of SRO minerals. Yet these soils are dominated by SOC with limited development of organo-metal complexes in the A horizon, indicating they are young soils. In contrast, the Inceptisols, despite being located in a geologically younger branch of the Andes, contain more weathered secondary minerals such as kaolinite, halloysite, boehmite and hematite, indicating a more advanced stage of weathering. Andisols and Inceptisols of my study, despite of being developed on different parent materials and climatic conditions, both contain SRO minerals: allophanes, imogolite and ferrihydrite in Andisols, and Al/Fe oxides, especially ferrihydrite in Inceptisols. Although the kind and amounts of these high specific surface SRO minerals vary between both soil types, they all may contribute to the water retention characteristics. SRO minerals appear important in stabilizing SOC in both Andisols and Inceptisols. Ferrihydrite was in low concentrations in both soils, with the highest concentrations (<2%), 58 found in the A horizon of Inceptisols. However, even at low concentrations, ferrihydrite may stabilize SOC. Correlations between SOC and ferrihydrite were stronger than between SOC and allophane and imogolite, particularly in the B horizons of both soils. Allophane and imogolite may also stabilize SOC in the B horizon of Andisols, where concentrations were the highest. In addition to stabilizing SOC, SRO minerals appear to increase the apparent proportion of the sand size fraction in both soil orders, by aggregation processes. Significant correlations were found between SRO minerals and the sand size fraction, predominantly in A and B horizons. SRO minerals in the Andisols and Inceptisols of this study are important for carbon stabilization and aggregate formation, which also may contribute to SWR. As these are the predominant soils in the Colombian Andes, occupying 66% of the region, knowing their mineralogical characteristics is particularly important for understanding their SWR characteristics. 59 4. SOIL WATER RETENTION CHARACTERISTICS OF ANDISOLS AND INCEPTISOLS AT TWO MID-ELEVATION SITES IN THE COLOMBIAN ANDES 4.1 Introduction Soil water processes have not received much research attention in Colombia, nor the Andes (Buytaert et al., 2005b; Quintero et al., 2009), or the tropics in general (Hodnett and Tomasella, 2002), despite their relevance for water supplies, food production and the resilience of ecosystems. Very little is known about SWR characteristics of Inceptisols in Colombia despite the fact they occupy 55% of the Colombian Andean area, in contrast to Andisols which occupy about 11%. The few studies carried out in Andisols and the even fewer carried out in Inceptisols in Colombia have focused on natural forest and páramo ecosystems at high elevation zones (i.e., >2,700 m) and in coffee plantations at lower elevations (i.e., 1,200-1,800 m) (Daza et al., 2014; Diaz and Paz, 2002; Hincapié, 2011). The interest in soils of the páramo ecosystems is based on the importance of these ecosystems for water supply, especially since large cities such as Quito and Bogotá rely almost entirely on surface water from the páramo (Buytaert et al., 2007). The interest in Andisols cultivated for coffee is based on the geographical location of Andisols in the coffee region (i.e., 350,000 ha or about 40% of the total area of this region). The importance of this region, is exemplified by research on SWR characteristics financed by CENICAFE, the Colombian Research Center for Coffee (Hincapié and Tobón, 2010). Soil water retention (SWR) characteristics are categorized by three main components: (i) hygroscopic water or the soil water content at permanent wilting point (θPWP), (ii) plant available water storage (PAWS), and (iii) gravitational water (GW). θPWP is water that is held tightly by the colloidal fraction in soils, and occupies the smallest soil pores and thus it is not available to plants. PAWS is the water that is held by capillary forces, stored in medium size pores and is available to plants. GW corresponds to water in the macro-pores, and moves by gravitational force. Most GW drains from the saturated soil profile during the first few days after a rain event, but GW moves relatively slowly compared to overland 60 flow (i.e., water that does not infiltrate into the soil and moves as surface flow). Soils with high θPWP, such as soils with high clay content, generally have lower PAWS, which is often amended by irrigation for plant growth and productivity. In addition, soils with high θPWP, generally have lower GW, which implies less macro-pore volume to buffer the hydrological response to a rain event (O´Geen, 2013). Thus, GW contributes to hydrological buffer capacity as the soil water storage attenuates stream flow response to rainfall events (Herron, 2001). SWR studies of Andisols in Colombia have shown that in addition to having high θPWP, they also have high PAWS and GW, indicating high SWR and a wide pore size distribution (Diaz and Paz, 2002; Hincapié, 2011) which may be related to the presence of SRO minerals. Studies of Inceptisols are less easy to generalize because Inceptisols are in the early stages of development, and SWR will depend on soil mineralogy and the local ecosystem conditions. Despite the recognition of the importance of these soils in Colombia, relationships among mineralogical, chemical and physical soil properties, and SWR have been poorly studied. Given that Andisols in Colombia are mostly allophanic (Malagón et al., 1995; Jaramillo, 2002), studies on water retention would benefit from the determination of short-range order (SRO) minerals, dominant in allophanic soils. However, studies which have been conducted on the water retention characteristics of Colombian soils (Diaz and Paz, 2002; Hincapié, 2011; Henao, 2001) have not included the determination of SRO minerals. On the other hand, studies that evaluated the mineralogical, chemical and physical properties of Andisols in Colombia (Malagón et al., 1995; Jaramillo, 2002; Chinchilla et al., 2011) often lack a description of water retention characteristics. The study by Buytaert et al. (2006) on non-allophanic soils in Ecuadorian páramos found significant relations between soil organic carbon (SOC) content, bulk density (ρb), and soil water retention at 1,500 kPa. The study objectives addressed in this chapter were to compare the SWR characteristics of Andisols and Inceptisols, located at two mid-elevation sites in the Colombian Andes, and to assess the relationships between SWR characteristics and soil properties. Findings of this study will enhance the overall understanding of the differences between these two soil types and contribute to data on the characteristics of soils at mid-elevations in the Colombian Andes. 61 4.2 Experimental conditions and laboratory analyses Soils were sampled as described in Section 2.3. From each of the two soils (Andisols and Inceptisols), 18 soil pits (6 under natural forest and 12 in pasture) were excavated and composite soil samples were taken by horizon. These composite soil samples were analyzed for: short-range order (SRO) minerals, soil pH in H2O and in CaCl2, SOC, soil particle size distribution, soil bulk density (ρb), and soil particle density (ρs). Analysis of SRO minerals is described in Section 3.2.1.2, while analyses of soil chemical and physical properties are described in Section 3.2.2. Soil water retention cores (45 cm3) were also taken from each soil horizon to 1.20 m depth. All cores sampled from the 18 pits were analyzed for the following soil water retention characteristics: soil water content (θ) at saturation (Sat), field capacity (FC), 100 kPa, 500 kPa, and 1,500 kPa or permanent wilting point (PWP). Numbers of soil samples collected from each soil order by land use type and horizon are provided in Figure 4.1. 4.2.1 Soil water retention curves Soil water retention (SWR) curves were determined on undisturbed soil cores using pressure plate apparatus at tensions ranging from 10 to 1,500 kPa (Klute, 1986). The generally used value for FC (33 kPa) has been reported to be inappropriate for volcanic soils or Andisols from humid regions due to their aggregating properties, and 10 kPa has been suggested as a more appropriate value for estimating FC (Saigusa et al., 1987). In general, 1,500 kPa is considered to correspond to PWP for most soils. Therefore, in this study the following tensions were used: saturation at 0 kPa (Sat); 10 kPa or FC, 100 kPa, 500 kPa and 1,500 kPa or PWP. The pore radius associated with each pore size class was determined using the capillary rise equation (Eqtn. 3.1) and the associated soil water tension. r = 2τ cos ϕℎ𝑔𝜌𝑤 Eqtn. 3.1 Where: r = radius of capillary (or pore) 62 τ = surface tension of water against air, ~ 0.074N/m at 10°C ϕ = wetting angle (~0° for clean glass capillary or wettable soil, hence cos ϕ = 1) ρw = water density, 1000 kg/m3 g = gravitational acceleration, 9.81 m/s2 h = height of rise at equilibrium The components of soil water, their abbreviations, definitions, the corresponding pore size classes, the associated pore radius and the soil tensions used in this study are presented in Table 4.1. Given that 1,500 kPa was defined as the PWP, hydroscopic water was considered as soil water trapped by crypto-pores in pore radii < 0.1 µm. 63 Table 4.1 Soil water component and associated abbreviations, symbols and pore size class Soil water component Abbrevia-tion Symbol Pore size class1 Pore radius Soil water tension mm m kPa Total porosity f θSat Macro, meso, micro, ultramicro, and crypto-pores >0.015 – <0.0001 >15 - < 0.1 < 10 - > 1,500 Gravitational water GW θSat – θFC Macro-pores >0.015 >15 m < 10 Plant available water storage PAWS θFC – θPWP Meso-pores micro-pores and ultramicro-pores 0.0015 – 0.015 0.0003 – 0.0015 0.0001 - 0.0003 0.015 mm or 1.5 0.3 – 1.5 0.1 – 0.3 10 – 100 100 – 500 500 – 1,500 Hygroscopic water - θPWP Crypto-pores < 0.0001 < 0.1 > 1,500 1Based on characteristics and functions of pore size classes as provided by SSSA, 2001. 64 4.2.2 Statistical analyses Soil water retention characteristics, including the measured values of θSat, θFC, θ100kPa, θ500kPa, θPWP, and calculated values for plant available water storage (PAWS) and gravitational water (GW), were compared by horizons between Andisols and Inceptisols to evaluate significant differences using Mann Whitney U test and probability values (p-value) of 0.01, 0.05, and 0.1 (Fig. 4.1a). In addition, data were separated into natural forest and pasture to evaluate the differences between Andisols and Inceptisols within the same land use (Fig. 4.1a). Relationships between SWR characteristics and soil physical and chemical properties were assessed utilizing the non-parametric Spearman’s rank-order correlation. Spearman’s Rho correlation coefficients (r) with values > 0.4 and probability values (p) of 0.01 and 0.05 were used to indicate notable relationships (Fig. 4.1b). Correlations were conducted for: • Andisol and Inceptisol, all horizons ▪ Andisol and Inceptisol by horizon (A, B and C horizons) o Andisol only, all horizons ▪ Andisol by horizon (A and B horizons) o Inceptisol only, all horizons ▪ Inceptisol by horizon (A and B horizons) Correlations for C horizon data in individual soil orders were not assessed since there were only eight samples for Andisols and five samples for Inceptisols. Sample numbers for each analysis are provided in Figure 4.1. 65 Figure 4.1 Overview of samples collected to a) compare soil water retention (SWR) characteristics between Andisols and Inceptisols and b) determine relationships between SWR characteristics and soil properties 66 4.3 Results and discussion 4.3.1 Soil water retention characteristics Soil water retention (SWR) curves for Andisols and Inceptisols for A, B, and C horizons are presented in Figure 4.2. The SWR curves for Andisols are consistently above the curves of Inceptisols, indicating that Andisols hold more water than Inceptisols at every soil tension in all horizons. Soil texture of the Andisols and Inceptisols of this study were classified as loam and clay, respectively (Section 3.3.3.4), yet their SWR characteristics do not correspond with typical values for these textural classes as commonly cited in the literature (Rawls et al., 1982 and 2004.). The soils in this study have greater total porosity and higher θPWP than values often reported for clay textured soils (i.e., 55-60% total soil porosity and θPWP 20-24%) (Rawls et al., 1982 and 2004). The shape of the SWR curves of A horizons in Andisols and Inceptisols were similar (Fig. 4.2a) with both curves displaying a decrease in soil water content from FC to 100 kPa and from 500 to 1500 kPa. The steeper sections of the SWR curve indicate that in the A horizons of both soil orders, there were a higher proportion of meso-pores between 1.5 and 15 m, and ultramicro-pores with radii between 0.1 and 0.3 m. When it comes to the B horizon, in Inceptisols, the slope of the SWR curve was relatively uniform from saturation to PWP, indicating a uniform pore size distribution, while in Andisols, the slope was slightly steeper from 10 to 100 kPa, suggesting a larger volume of meso-pores between 1.5 and 15 m (Fig. 4.2b). Water retention curves for the C horizon, were similar in shape for both soil types with a slight change in soil water content between 100 and 500 kPa for Inceptisols (Fig. 4.2c). SWR characteristic curves under forest and pasture (Figs. 4.3 and 4.4) showed the same trends, but differences between the two soil orders were more pronounced under pasture than forest. 67 a) A horizon b) B horizon c) C horizon Figure 4.2 Median results by soil horizon for soil water retention (SWR) curves of Andisols and Inceptisols in: a) A horizon, b) B horizon and c) C horizon 20304050607080901000,1 1 10 100 1000Volumetric soil moisture (%)Log soil tension (kPa)Andisols Inceptisols.020304050607080901000,1 1 10 100 1000Volumetric soil moisture (%)Log soil tension (kPa)Andisols Inceptisols.020304050607080901000,1 1 10 100 1000Volumetric soil water (%)Log soil tension (kPa)Andisols Inceptisols0.0n= 17 n= 18 n= 19 n= 19 n= 8 n= 5 68 a) Forest A horizon b) Forest B horizon Figure 4.3 Median results and quartiles for soil water retention (SWR) curves in Andisols and Inceptisols for forest in: a) A horizon and b) B horizon 20304050607080901000,1 1 10 100 1000Volumetric soil water (%)Log soil tension (kPa)Andisols Inceptisols0.020304050607080901000,1 1 10 100 1000Volumetric soil moisture (%)Log soil tension (kPa)Andisols Inceptisols.0n= 5 n= 6 n= 6 n= 7 69 a) Pasture A horizon b) Pasture B horizon Figure 4.4 Median results and quartiles for soil water retention (SWR) curves in Andisols and Inceptisols for pasture in: a) A horizon and b) B horizon 20304050607080901000,1 1 10 100 1000Volumetric soil water (%)Log soil tension (kPa)Andisols Inceptisols.020304050607080901000,1 1 10 100 1000Volumetric soil moisture (%)Log soil tension (kPa)Andisols Inceptisols.0n= 12 n= 12 n= 13 n= 12 70 Median values for SWR characteristics (θSat, θFC, and θPWP) for each horizon were significantly different between Andisols and Inceptisols (Appendix F). However, when separating natural forests from pastures, statistically significant differences were observed only between Andisols and Inceptisols under pasture (Table 4.2). Differences under pasture were observed at every soil tension in the A horizon and at saturation (Sat), FC, and PWP in the B horizon. Despite these differences, PAWS and GW under pasture were similar between the two soil orders. In constrast, under forest, the two soils have similar SWR characteristics, but PAWS in the A horizon was statistically higher in Andisols than in Inceptisols (Table 4.2). The SWR characteristics of Andisols and Inceptisols in this study were compared to values reported by Rawls et al. (1982, 2004) who carried out studies to estimate water retention based on soil properties (Table 4.3). Both Andisols and Inceptisols have higher θPWP than those reported for clay textured soils, and moderate to high values for PAWS and GW. Both soils, but especially Andisols, due to their higher θPWP and location in a high precipitation region, may be more susceptible to compaction when physical degradation occurs (Toohey et al., 2018). Compaction in the Sonora watershed was observed during field work, particularly at sites such as water troughs where livestock gather, and horse trails used in forest harvesting. Soils at these sites lacked vegetation cover, and due to physical compaction, may be prone to water erosion (Kimble et al., 2000) and to the destruction of soil aggregates (Herrera et al., 2007). 71 Table 4.2 Median values for soil water retention (SWR) characteristics in A, B and C horizons for natural forest and pasture in Andisols and Inceptisols 1θSat, θFC, θ100 kPa, θ500 kPa, θPWP: soil water content (θ) at saturation (Sat), field capacity (FC), 100 kPa, 500 kPa and at permanent wilting point (PWP), respectively; 2Values in parenthesis are first and third quartile, 3PAWS: plant available water storage; 4GW: gravitational water; Number of samples (n) for natural forest, were 5 and 6 in Andisols, 6 and 7 in Inceptisols; for pasture were 12 and 13 in Andisols, 12 and 12 in Inceptisols, for A and B horizons, respectively; **, + Significant differences between Andisols and Inceptisols with Mann Whitney U test at p < 0.01, and p < 0.1, respectively Horizon Andisols Inceptisols Andisols InceptisolsA 79.2 (70.9-83.1)272.4 (66.8-80.8) 77.5 (76.1-83.7) 67.9 (65.4-70.4)**B 74.9 (70.3-77.7) 68.4 (67.4-69.5)+74.7 (66.3-75.6) 61.9 (60.8-64.1)**A 63.2 (55.9-67.5) 58.2 (53.2-61.9) 67.8 (64.9-71.0) 56.6 (51.7-58.9)**B 60.6 (50.9-65.9) 54.5 (52.3-59.1) 61.4 (55.5-64.7) 50.8 (49.3-53.5)**A 55.0 (45.0-58.2) 51.8 (46.5-55.4) 58.9 (55.1-63.2) 50.3 (46.0-53.1)**B 50.7 (44.9-54.5) 48.1 (46.2-55.9) 53.4 (46.2-56.8) 46.0 (44.1-49.0)A 50.8 (42.2-54.7) 49.8 (45.0-50.4) 55.8 (52.2-60.5) 47.7 (43.3-49.7)**B 48.6 (43.3-53.0) 45.6 (44.4-53.9) 50.4 (43.8-51.7) 43.3 (41.7-47.3)A 44.1 (36.4-50.5) 43.8 (40.9-47.5) 52.5 (49.5-57.5) 42.2 (40.2-43.5)**B 47.7 (42.1-49.7) 43.4 (40.9-44.2) 46.1 (43.4-51.7) 39.4 (36.8-43.2)**A 18.4 (15.9-20.9) 13.5 (12.1-16.9)+15.4 (11.8-17.3) 13.3 (11.9-15.7)B 14.3 (10.1-15.4) 12.5 (10.2-19.5) 12.6 (10.3-15.1) 11.0 (9.3-13.5)A 12.8 (10.9-21.2) 16.8 (12.9-18.9) 10.9 (9.5-12.4) 11.4 (8.9-18.1)B 11.4 (8.9-18.5) 10.4 (8.8-15.1) 11.5 (8.4-14.4) 11.7 (7.3-12.8)PAWS3 (%v/v)GW4 (%v/v)θSat1 (%v/v)θFC1 (%v/v)Natural forest Pastureθ100 kPa1 (%v/v)θ500 kPa1 (%v/v)θPWP1 (%v/v)72 Table 4.3 First and third quartile of volumetric water content for Andisols and Inceptisols in A horizon and literature values for three pure textural classes Soil water component or pore size class Volumetric water content (%) Soil textural class4 Natural forest Pasture Sand Loam Clay Andisols (n=5) Inceptisols (n=6) Andisols (n=12) Inceptisols (n=12) θPWP1 or crypto-pores 2-4 8-12 20-24 36-50 41-47 49-57 40-43** PAWS2 or meso-, micro- and ultramicro-pores 4-10 17-20 12-16 16-21 12-17+ 12-17 12-16 GW3 or macro-pores 16-18 10-13 5-8 11-21 13-19 9-12 9-18 Approximate total porosity 20-25 40-45 55-60 71-83 67-81 76-84 65-70 1θPWP: soil water content at permanent wilting point; 2PAWS: plant available water storage; 3GW: gravitational water; **, + Significant differences between Andisols and Inceptisols at p < 0.01 and p < 0.1, respectively. 4Source for soil textural class data: Rawls et al., 1982 and 2004. The high soil water content at PWP in both Andisols and Inceptisols, implies that a high portion of the pore volume contains water that is not available to plants. Despite this high hygroscopic water content, both soil types have considerable values for GW and PAWS. The value of PAWS in Andisols under forest was similar to a typical loam soil, while the PAWS values for Andisols under pasture and Inceptisols under both land uses were similar to that reported for a typical clay soil. This relatively high PAWS is important for forage and crops, particularly in the El Chocho watershed (Inceptisol site), given the lower annual precipitation at this site. The values of GW in Andisols and Inceptisols under both land uses were intermediate between sandy and loam soils (Table 4.3), implying a considerable hydrological buffering capacity. The high total soil porosity of both Andisols and Inceptisols relative to typical clay textured soils, may be due to high specific surface area (SSA) of the dominant colloids in the soils of the study sites. In particular allophane and imogolite (700-1500 m2/g), ferrihydrite (220-560 m2/g), Al and Fe oxides (90-140 m2/g), and SOC (800-900 m2/g) all have high SSA (Table 3.1) in comparison to clay minerals such as illite, chlorite and kaolinite, with SSA < 40 m2/g. 73 The higher values of SWR in Andisols relative to Inceptisols (Appendix F), may be related to the type of high surface area colloids and their abundance. Allophane for example, has a nanoparticle size and a hollow spherical structure of 3.5 to 5 nm in diameter (Van Wambeke, 1992) that stores water within its structure (Wada, 1985). Similarly, imogolite has a hollow tubular structure of 2 nm outer diameter (Nanzyo, 2002). Furthermore, these soil colloids, in addition to their nano-particle size and high specific surface area, have positive and negative charges (Table 3.1), forming aggregates with silt and clay particles, increasing pore volume (Kaiser et al., 2011; Regelink et al., 2013; Arias et al., 1996; Sei et al., 2002; Regelink et al., 2015). Comparing the Andisols of this study with other data from the Andes (Table 4.4) suggests that high organo-metallic compounds and high SOC (found in páramo ecosystems in non-allophanic soils) increase both macro and crypto-pore volume. The Andisol of this study, is predominantly allophanic with lower amounts of SOC and organo-metallic compounds relative to páramo ecosystems, but has considerable amount of SRO minerals, relatively lower gravitational and hygroscopic water content, and higher PAWS (Table 4.4). This suggests that Andisols, such as in this study, which have SOC < 17% by weight, but considerable amounts of SRO minerals (4-10% allophane), may have comparatively more PAWS. This comparison between Andisols in the Andes, highlights the importance of Andisols at mid-elevations, since they have both higher PAWS and considerable GW, which is important for food production and water supply. In Colombia, most Andisols are allophanic with a predominance of SRO minerals in the B horizon and low amounts of organo-metallic complexes in the A horizon. 74 Table 4.4 Water retention characteristics of Andisols of this study and other regional studies Country Ecosystem θPWP1 (cm3/cm3) PAWS2 (cm3/cm3) GW3 (cm3/cm3) SOC4 in A horizon (%) Allophane (%) Alp/Alo5 Reference Colombia Natural forest 36-506 16-21 11-21 10-15 4-7 0.07 - 0.16 This study Pasture 49-57 12-17 9-12 7-11 5-9 0.05 - 0.13 Ecuador Páramo 48-59 - - 32-37 - 0.9 – 1.11 Buytaert et al., 2006 Colombia Páramo 68-69 8.8-9.7 18-33 35-38 - - Diaz and Paz, 2002 Colombia Natural forest Pasture 45 48 8 15 29 19 16 8 - - - - Tobón et al., 2010 Colombia Coffee plantations 27-44 13-25 8-12 6-7 5-10 - Hincapié, 2011 1θPWP: soil water content (θ) at permanent wilting point (PWP); 2PAWS: plant available water storage; 3GW: gravitational water; 4SOC: soil organic carbon; 5Alp/Alo: Indicator for determing allophanic or non-allophanic soils; 6Results shown from this study are the first and third quartile in the A horizon 75 Inceptisols may develop on a variety of parent materials including volcanic ash. The Inceptisol in the study of Henao (2001), which has significant amounts of SRO minerals (Table 4.5) was developed on a volcanic ash, limiting the comparability to the Inceptisols of the El Chocho watershed sampled in my study. However, based on available research, it is possible that SWR characteristics of Inceptisols in the region, are related to SOC and the presence of SRO minerals such as allophane or ferrihydrite. Table 4.5 Water retention characteristics of Inceptisols of this study and other regional studies Country Ecosystem θPWP1 (cm3/cm3) PAWS2 (cm3/cm3) GW3 (cm3/cm3) SOC4 in A horizon (%) Allophane (%) Reference Colombia Natural forest 41-475 12-17 13-19 5-10 1.2-3.6 This study Pasture 40-43 12-16 9-18 4-6 1.2-2.1 Colombia Páramo 34-44 23-26 21-19 18 - Daza et al., 2014 Colombia Forest - 18-33 - - - Morales, 2008 Colombia Coffee plantations - - - 3.5 4-13 Henao, 2001 1θPWP: soil water content (θ) at permanent wilting point (PWP); 2PAWS: plant available water storage; 3GW: gravitational water; 4SOC: soil organic carbon; 5 Results shown are the first and third quartile in A horizon 4.3.2 Relationships between soil water retention characteristics and soil properties Correlations among SWR characteristics and soil properties for the two soils in this study are given in Appendix E, Table E2. There were few significant correlations using the complete data set. As the objective was to compare the two soil orders, correlations are presented for each soil separately comparing all horizons and each horizon (Table 4.6). Andisols showed positive correlations between SWR characteristics and the clay size fraction, SOC, and Al and Fe associated with SOM (pyrophosphate extractable Alp and Fep), when the complete dataset was used (Table 4.6a). In contrast, only SOC was positively correlated with SWR in Inceptisols. The only common factor in both soil orders was b, which was negatively correlated with SWR. In contrast to the complete data set where there were no correlations with SRO minerals and SWR, allophanes and ferrihyrite, presented significant positive correlations in A and/or B horizons in both soils (Table 4.6). 76 Table 4.6 Spearman´s Rho correlation coefficients (r) between soil water retention (SWR) characteristics and soil physical and chemical soil properties in all horizons and in A and B horizons in: a) Andisols; and b) Inceptisols 1θSat, θFC, θ100 kPa, θ500 kPa, θPWP: soil water content (θ) at saturation (Sat), field capacity (FC), 100 kPa, 500 kPa and at permanent wilting point (PWP), respectively; 2PAWS: plant available water storage; 3GW: gravitational water; 4SOC: soil organic carbon; 5Pyrophosphate-extractable aluminum and iron (Alp and Fep); 6ρb: bulk density; correlations shown are the ones r > 0.4; Gray cells show correlations with p < 0.01 and white cells show correlations with p < 0.05 θSat1 (%v/v)θFC1 (%v/v)θ100 kPa1 (%v/v)θ500 kPa1 (%v/v)θPWP kPa1 (%v/v)PAWS2 (% v/v)GW3 (% v/v)θSat1 (%v/v)θFC1 (%v/v)θ100 kPa1 (%v/v)θ500 kPa1 (%v/v)θPWP kPa1 (%v/v)PAWS2 (% v/v)GW3 (% v/v)θSat1 (%v/v)θFC1 (%v/v)θ100 kPa1 (%v/v)θ500 kPa1 (%v/v)θPWP kPa1 (%v/v)PAWS2 (% v/v)GW3 (% v/v)a) Andisols All horizons (n= 44) A horizon (n= 17) B horizon (n= 19)Sand (%) -0.53 -0.52 -0.54 -0.58 -0.56 -0.56 -0.64Silt (%)Clay (%) 0.40 0.54 0.54 0.49 0.49 0.63 0.78 0.60 0.57SOC4 (%) 0.70 0.51 0.65 0.48 0.88Alp5 (g/kg) 0.49 0.51 0.51Fep5 (g/kg) 0.54 0.49 0.49Allophane (%) 0.60 0.54Ferrihydrite (%) 0.78ρb6 (kg/m3) -0.84 -0.53 -0.53 -0.80 -0.66 -0.81 -0.47b) Inceptisols All horizons (n= 42) A horizon (n= 18) B horizon (n= 19)Sand (%)Silt (%)Clay (%)SOC4 (%) 0.44Alp5 (g/kg) 0.49 0.67 0.50Fep5 (g/kg) 0.49Allophane (%) 0.46Ferrihydrite (%)ρb6 (kg/m3) -0.77 -0.50 -0.58 -0.84 -0.61 -0.51 -0.48 -0.65 -0.51 -0.63 -0.5877 4.3.2.1 Relationships between soil water retention (SWR) characteristics and texture The clay size fraction had an effect on SWR in the A horizons of Andisols. There were positive correlations between soil water content and the proportion of clay size fraction at tensions up to 500 kPa (Table 4.6a and Fig. 4.5), reflecting the high SWR characteristics of the clay fraction. Figure 4.5 Relationship between proportion of size fractions of clay size (clay) and sand size (sand) with soil water retention at field capacity (θFC) in A horizon of Andisols, n= 17 While the texture of Andisols of the Sonora watershed were dominantly sandy loam and loam, Inceptisols of the El Chocho watershed have a clay texture, with clay size fraction >40% in the majority of the samples (Section 3.3.3.4). Due to the high specific surface area, the % clay size fraction contributes to the relatively high SWR of Inceptisols. However, there were no statistically significant correlations between texture and water retention in Inceptisols (Table 4.6b). 4.3.2.2 Relationships between soil water retention characteristics (SWR), bulk density (ρb) and soil organic carbon (SOC) As expected, (ρb) was negatively correlated with soil water content in both soil orders. There were significant correlations in Andisols and Inceptisols between soil water content at saturation (θSat) and bulk density (ρb) using the complete data set and also in A and B r = 0.78 p < 0.01 r = -0.56 p = 0.02 78 horizons (Table 4.6). In Andisols, a higher r-value was found, when all data was compared (Figure 4.6a); in Inceptisols the highest correlation was found in the A horizon (Fig. 4.6b). a) b) Figure 4.6 Relationships between soil water content at saturation (θSat) and bulk density (ρb) in A, B and C horizons of: a) Andisols; and b) Inceptisols The importance of SOC in enhancing SWR in both soils was shown in the B horizon of Andisols where there was a significant correlation between SOC and θSat; and in Inceptisols with the complete data set (Fig. 4.7). In Andisols, although SOC was highest in the A horizon, there was no significant correlation between SOC and soil water content at any tension. This result leads to two observations: first, despite a lower concentration of SOC in the B horizon, SOC is an important property that increases the SWR; and second, the high correlation coefficient may be due the stabilizing effect of SRO on SOC in the B horizon (Section 3.3.3.2), which then suggests a synergistic effect of SRO and SOC that increases SWR in the B horizon. n = 19 r = -0.79 p = 0.02 n = 17 r = -0.81 p < 0.01 n = 19 r =-0.63 p < 0.01 n = 18 r =-0.84 p < 0.01 n = 44 r = -0.84 p = <0.01 n = 42 r =-0.77 p < 0.01 79 a) b) Figure 4.7 Relationships between soil water content at saturation (θSat) and soil organic carbon (SOC) in A, B and C horizons of: a) Andisols; and b) Inceptisols 4.3.2.3 Relationships between soil water retention (SWR) characteristics and short-range order (SRO) minerals Short-range order minerals, despite their low contribution to soil mass, had an influence on SWR characteristics. In the A horizon of Andisols there was a positive correlation between soil water content at 100 kPa (θ100kPa) and allophane (Fig. 4.8b), and in the B horizon between θSat and ferrihydrite (Fig. 4.8a). Ferrihydrite, even in lower amounts compared to allophane in Andisols, had the highest r-value with SWR. Ferrihydrite occurs mainly as coatings on soil separates, and the high number of functional groups increases SWR (Table 3.1). Additionally, the synergistic effect suggested previously between SOC and SRO minerals increasing SWR, is supported by these strong correlations found between ferrihydrite and θSat, and ferrihydrite and SOC in B horizon of Andisols (Fig. 3.11). This role of ferrihydrite in stabilizing SOC is also reported in the literature (Kaiser et al., 2011; Regelink et al., 2013), and may further enhance SWR characteristics. Furthermore, the influence of SRO minerals on SWR characteristics is indicated by the higher PAWS in the B horizon of Andisols (Fig. 4.2b). In contrast, in the B horizon of Inceptisols, there was a significant correlation between soil water content at permanent wilting point (θPWP) and allophane (Fig 4.8c). Although allophane content was low in Inceptisols (median values in A and B horizons were 1.4 and 1.2%, respectively), allophane with its high specific surface area and hollow spherical structure, may have contributed to greater SWR at higher tensions (Table 3.1). n = 17 n = 19 r = 0.88 p = < 0.01 n = 18 n = 19 n = 42 r =0.44 p < 0.01 n = 44 r = 0.70 p = < 0.01 80 a) b) c) Figure 4.8 Relationships between short range order (SRO) minerals in A and B horizons of Andisols or Inceptisols: a) ferrihydrite and soil water content at saturation (θSat) in Andisols; b) allophane and imogolite and soil water content at 100kPa (θ100kPa) in Andisols; and c) allophane and imogolite and soil water content at PWP (θPWP) in Inceptisols 4.3.2.4 Relationships between soil water retention (SWR) characteristics and organo-metallic complexes Although the influence of SOC was greater (Fig. 4.7), organo-metallic complexes had also an influence on SWR in Andisols. Positive correlations were found between Alp and Fep with θSat (r-values 0.49 and 0.54, respectively) (Table 4.6a), suggesting that organo-metalic complexes increase θSat. However, SOC had a stronger correlation coefficient (r-value 0.70). Alp found in the allophanic Andisols, from Sonora, was less than 3 g/kg, which is 204060801001201400,0 0,4 0,8 1,2θSat(%v/v)Ferrihydrite (%)A horizon B horizon. . 0.8 1.2n = 17 r = 0.60 p = 0.01 n = 19 r = 0.78 p < 0.01 n = 17 n = 19 n = 19 r = 0.46 p = 0.05 n = 18 81 very low compared to the concentrations found in non-allophanic soils where Alp may approach 30 g/kg (Shoji et al., 1993) and where Al-humus complexes are dominant over SRO minerals. Typically, pyrophosphate extractable minerals are strongly correlated to SOC since pyrophosphate extracts Al and Fe complexed with organic materials, which was the case in Andisols of this study (Section 3.3.2.1). This suggests that Fep and Alp are strongly associated with SOC and that poorly crystalline hydroxide phases are unlikely to have contributed significantly to the pyrophosphate extracts (Kaiser and Zech, 1996). However, this was not the case in Inceptisols (Table 3.5), indicating that Fep and Alp may be associated with poorly crystalline hydroxides. Crystalline hydroxide may be Al and Fe oxides such as goethite, hematite, gibbsite, boehmite or ferrihydrite (Kaiser and Zech, 1996; Parfitt and Childs, 1988). X-ray analyses in Inceptisols (Section 3.3.1), showed the presence of hematite and boehmite in A and B horizons, boehmite, hematite and goethite in the C horizon, and ferrihydrite (by chemical extraction) was significantly higher in Inceptisols than Andisols in the A horizon. Therefore, in Inceptisols, the presence of minerals with high specific surface area, such as iron and aluminum oxides, may play an important role in increasing SWR characteristics, as shown by the correlation in the B horizon of Inceptisols, between θFC and Alp (Fig. 4.9b). There was also a weak positive correlation between GW and Fep in the A horizon of Inceptisols (Fig. 4.9a). a) b) Figure 4.9 Relationships in Inceptisols between: a) pyrophosphate extractable Fe (Fep) and gravitational water (GW) in A horizon; and b) pyrophosphate extractable Al (Alp) and soil water content at saturation (θSat) in B horizon 0204060801001201400,00 0,05 0,10 0,15 0,20GW (%v/v)Fep (g/kg)0. . . . .0204060801001201400,00 0,10 0,20 0,30θFC(%v/v)Alp (g/kg)0.0 0. 0.2 .n = 18 r = 0.49 p = 0.04 n = 19 r = 0.49 p = 0.03 82 Texture, ρb, SOC, ferrihydrite, allophane and imogolite and organo-metallic complexes all were correlated with SWR characteristics in Andisols (Table 4.6). In contrast, in Inceptisols, correlations with ρb, Fe and Al oxides, allophane and imogolite were found, but coefficient values were lower (Table 4.6). These results show the importance of SRO minerals and SOC on SWR in both soils, and the influence of organo-metallic complexes in Andisols. Allophane and imogolite were found in both soils, although concentrations were higher in Andisols. In Inceptisols, Fe and Al oxides influence SWR. Interestengly, ferrihydrite was the SRO in Andisols with the highest correlation with SWR characteristics, even at concentrations < 1%. 4.4 Conclusions Both Andisols and Inceptisols have a high water retention capacity, particularly at high tensions (i.e., hygroscopic water). Despite the contrasting soil parent materials, climate and geographical conditions, Andisols and Inceptisols of this study share the presence of minerals with high specific surface area, that increase their ability to retain water. These minerals are allophane, imogolite, ferrihydrite and organo-metallic compounds in Andisols; and extractable ferrihydrite and other Al / Fe oxides in Inceptisols. The SWR characteristics of Andisols and Inceptisols were significantly different in pasture, but not under natural forest. PWP was highest under pasture in Andisols (52%) and similar for Inceptisols and Andisols under natural forest (~43%). Having a high soil water content at PWP, implies that a high portion of the pore volume in both Andisols and Inceptisols was neither PAWS nor GW. Yet, despite this high hygroscopic water content, both soil types have considerable values of PAWS and GW, with no significant differences between soil orders. In Andisols, positive correlations were found between soil water retention at different tensions and clay size fraction, SOC, ferrihydrite, allophane and imogolite, and organo-metallic compounds, which suggests that these soil properties increase soil water retention. There may also be a synergistic effect between ferrihydrite and SOC in the B horizon of Andisols that increases SWR characteristics. 83 In Inceptisols, positive correlations were found between soil water retention at different tensions with Alp and Fep, allophane and imogolite, and SOC. Alp and Fep in Inceptisols may be related to Al and Fe oxides, such as hematite, boehmite, goethite and ferrihydrite. These minerals, although present in low amounts may be increasing soil water retention in Inceptisols. In comparison to studies from páramo ecosystems (Diaz and Paz, 2002; Buytaert et al., 2006), the results of this study suggest that non-allophanic soils (Andisols from páramo ecosystems) had greater volumes of hygroscopic and gravitational water, while allophanic soils (Andisol of this study), had higher PAWS. Land management practices which maintain or increase soil organic matter are recommended for all soils, but particularly for soils containing SRO minerals, as SOC acts in conjunction with SRO minerals to enhance SWR characteristics. Thus, maintaining or increasing SOC in soils with SRO minerals is important for plant available water for both forest and rangeland productivity in the watersheds of this study. The relatively high PAWS in both soil orders is particularly important for forage and crop production in the El Choco watershed (Inceptisol site) due to the lower annual precipitation at this site. The high θPWP in both soil orders implies that soils retain water throughout the year. In the Sonora watershed (Andisol site), which has a wet climate, this implies that soils may be subject to compaction, especially under pasture grazed by cattle. 84 5. LAND USE IMPACTS ON SOIL WATER RETENTION CHARACTERISTICS OF ANDISOLS AND INCEPTISOLS AT TWO MID-ELEVATION SITES IN THE COLOMBIAN ANDES 5.1 Introduction People living in the Colombian Andes rely on services provided by mountain ecosystems such as water supply, agriculture, biodiversity conservation and carbon storage (De Groot et al., 2002; Labrière et al., 2015; Buytaert et al., 2011). Particularly important for these ecosystem services are natural land covers such as forest, wetlands and páramos (Forsyth, 1996; Calder, 1999, 2002; Roa-García, 2009). However, as discussed in Chapter 1, these natural land cover types have been converted to pasture for cattle grazing and to agricultural crops, land uses which may not be appropriate on steeper slopes. The Colombian National Geographical Institute (IGAC) determined that within the Andean region 54% of the area has inappropriate land uses, 41% of the area being classified as overused and 13% as underused. IGAC classifies land use in three soil categories: Group I for soils adequate for intensive or semi-intensive agriculture and cattle grazing; Group II for soils adequate for agriculture, cattle grazing, forestry and agro-forestry and Group III for preservation, conservation and eco-tourism. Soils under a land use different than recommended, that causes lower benefits than expected are classified as underused soils, while soils under a land use that causes damages- mainly soil erosion- to vulnerable soils (particularly those of Group III) are classified as overused soils (IGAC, 2012). Despite the relevance of the soil component in these ecosystems for water regulation, limited research has been conducted in the Andean region to assess the impacts of land use type on soil water retention (SWR) characteristics. The majority of studies that did evaluate the effects of land use on soil properties have been conducted in páramo ecosystems (elevation >3,500 m), which play an important role in regulating water for large cities in the Andes and in Colombia. For example, Diaz and Paz (2002) and Daza et al. (2014) reported that the conversion from natural vegetation to crops and pasture in the Colombian páramo reduced soil water content at field capacity (θFC), permanent wilting point (θPWP) and gravitational water (GW). Other studies in the Ecuadorian Andes, also found negative 85 effects on soil characteristics such as a reduction in SOC, increased bulk density (ρb), and lower θFC (Buytaert et al., 2002; Buytaert et al., 2005b; Podwojewski et al., 2002). However, soils at these study sites are non-allophanic (organo-metallic complexes are predominant) and have different mineralogy and soil chemical and physical properties; hence, are not directly comparable to the soils of this study. The objectives of this chapter were to (i) compare the effects of two common land uses (natural forest and pasture) on the SWR characteristics of Andisols and Inceptisols at two mid-elevation sites in the Colombian Andes, and (ii) determine the soil properties related to SRW characteristics. 5.2 Experimental conditions and laboratory analyses The same experimental conditions and laboratory analyses as outlined in the previous chapter (Section 4.2) are of relevance for this chapter. From each of the two soil orders (Andisols and Inceptisols), 18 soil pits (6 under natural forest and 12 in pasture) were excavated and composite soil samples were taken by horizon. These composite soil samples were analyzed for: pH in H2O and in CaCl2, SOC, soil particle distribution, short-range order (SRO) minerals, soil bulk density (ρb), and soil particle density (ρs). In addition, soil water retention cores (45 cm3) were taken from each soil horizon to 1.20 m depth. All cores sampled from the 18 pits were analyzed for the following soil water retention characteristics: soil water content (θ) at saturation (Sat), field capacity (FC), 100 kPa, 500 kPa, and 1,500 kPa or permanent wilting point (PWP). The number of soil samples collected on each soil order by land use and horizon are provided in Figure 5.1. Laboratory analyses are described in Sections 3.2.2 and 4.2.1. 5.2.1 Statistical analyses Soil water retention (SWR) characteristics, including the measured values of θSat, θFC, θ100kPa, θ500kPa, θPWP, and calculated values for plant available water storage (PAWS) and gravitational water (GW), were compared by horizon between natural forest and pasture to evaluate significant differences using Mann Whitney U test and probability values (p-value) of 0.01, 0.05, and 0.1 (Fig. 5.1a). In addition, data were separated into Andisols and 86 Incepisols to evaluate the differences between natural forest and pasture within the same soil order (Fig. 5.1a). Relationships between SWR characteristics and soil physical and chemical properties were assessed utilizing the non-parametric Spearman’s rank-order correlation. Spearman’s Rho correlation coefficients (r) with values > 0.4 and probability values (p) of 0.01 and 0.05 were used to indicate notable relationships (Fig. 5.1b). Correlations were conducted for: • Natural forest, all horizons ▪ Natural forest by horizon (A, B and C horizons) o Natural forest in Andisol, all horizons ▪ Natural forest in Andisol, by horizon (A and B horizons) o Natural forest in Inceptisol, all horizons ▪ Natural forest in Inceptisol, by horizon (A and B horizons) • Pasture, all horizons ▪ Pasture by horizon (A, B and C horizons) o Pasture in Andisol, all horizons ▪ Pasture in Andisol, by horizon (A and B horizons) o Pasture in Inceptisol, all horizons ▪ Pasture in Inceptisol, by horizon (A and B horizons) Correlations for C horizon data in individual soil orders were not assessed since there were less than three samples for the combination of land use and soil order with the exception of Andisols under pasture which had a total of seven samples. Soil properties including soil particle size, organo-metallic complexes (Alp and Fep), allophane, imogolite and ferrihydrite, and bulk density (ρb) were compared for each soil order by horizon for natural forest and pasture to evaluate any significant differences using Mann Whitney U test and probability values (p-value) of 0.01, 0.05, and 0.1 (Fig. 5.1c). Sample numbers for statistical analyses are provided in Figure 5.1. 87 Figure 5.1 Overview samples collected to a) compare soil water retention (SWR) characteristics between natural forest and pasture in Andisols and Inceptisols; b) determine relationships between SWR characteristics and soil properties; and c) compare overall differences in soil properties between natural forest and pasture88 5.3 Results and discussion 5.3.1 Soil water retention characteristics in Andisols and Inceptisols: differences between natural forest and pasture Soil water retention characteristic curves under forest and pasture at the Andisol and Inceptisol study sites are shown in Figures 5.2 and 5.3. There was a marked difference in PWP in the A horizon under forest within the Andisols but no differences between land uses were seen at any tension in the Inceptisols (Fig. 5.2 and Table 5.2). In contrast, when comparing the SWR characteristics between natural forest and pasture in the B horizon, Inceptisols display the greater change (Fig. 5.3 and Table 5.2), with higher θSat and θFC under forest. a) Andisols b) Inceptisols Figure 5.2 Median results and quartiles for soil water retention (SWR) curves in A horizon of natural forest and pasture for a) Andisols, and b) Inceptisols 20304050607080901000,1 1 10 100 1000Volumetric soil moisture (%)Log soil tension (kPa)Forest Pastures0.020304050607080901000,1 1 10 100 1000Volumetric soil moisture (%)Log soil tension (kPa)Forest Pastures.0n= 5 n= 12 n= 6 n= 12 89 a) Andisols b) Inceptisols Figure 5.3 Median results and quartiles for soil water retention (SWR) curves in B horizon of natural forest and pasture for a) Andisols, and b) Inceptisols Analyzing the combined dataset from both soil orders (Table 5.1), GW was significantly higher under natural forest than pasture. Higher GW represents a larger hydrological buffering capacity, implying a larger capacity of the soil to retain rainwater for one to two days after a rainfall event (Herron, 2001). This water will flow relatively slowly by gravity compared to overland flow as it percolates through soil macropores. This relationship, however did not hold when comparisons were made within the soil orders. Even in the A horizon of Andisols, where the median SWR curve was steeper under forest, there were no statistically significant differences between natural forest and pasture in PAWS or GW (Table 5.2). 20304050607080901000,1 1 10 100 1000Volumetric soil moisture (%)Log soil tension (kPa)Forest Pastures.020304050607080901000,1 1 10 100 1000Volumetric soil moisture (%)Log soil tension (kPa)Forest Pastures.0n= 6 n= 13 n= 7 n= 12 90 Table 5.1 Median results of soil water retention (SWR) characteristics for forest and pasture in A and B horizons with all data for Andisols and Inceptisols 1θSat, θFC, θ100 kPa, θ500 kPa, θPWP: soil water content (θ) at saturation (Sat), field capacity (FC), 100 kPa, 500 kPa and at permanent wilting point (PWP), respectively; 2Values in parenthesis are first and third quartile, 3PAWS: plant available water storage; 4GW: gravitational water; Number of samples (n) for natural forest were 11 and 13, and for pasture were 24 and 25 for A and B horizons, respectively; + Significant difference between forest and pasture with Mann Whitney U test at p < 0.1. Horizon Forest PastureA 77.0 (67.3-79.8)273.6 (67.9-80.8)B 69.5 (67.6-74.9) 65.2 (61.9-74.9)A 59.9 (54.1-65.4) 62.6 (56.3-68.3)B 57.4 (51.8-62.6) 54.5 (50.5-61.7)A 52.5 (45.3-57.5) 53.5 (50.0-59.0)B 50.6 (45.9-54.3) 48.2 (44.7-55.3)A 49.8 (43.1-51.5) 51.0 (47.7-55.9)B 48.5 (44.1-52.6) 46.1 (42.2-52.1)A 44.1 (37.3-48.3) 47.8 (41.8-53.1)B 43.9 (41.8-47.7) 43.3 (37.5-47.8)A 16.4 (13.4-18.4) 14.0 (11.8-16.5)B 13.5 (10.4-15.8) 12.2 (9.8-14.5)A 16.7 (11.9-19.3) 10.9 (9.4-15.5)+B 10.4 (8.9-15.4) 11.6 (7.8-13.1)θPWP1 (%v/v)PAWS3 (%v/v)GW4 (%v/v)θSat1 (%v/v)θFC1 (%v/v)θ100 kPa1 (%v/v)θ500 kPa1 (%v/v)91 Table 5.2 Median results of soil water retention (SWR) characteristics for forest and pasture in A and B horizons of Andisols and Inceptisols 1θSat, θFC, θ100 kPa, θ500 kPa, θPWP: soil water content (θ) at saturation (Sat), field capacity (FC), 100 kPa, 500 kPa and at permanent wilting point (PWP), respectively; 2Values in parenthesis are first and third quartile, 3PAWS: plant available water storage; 4GW: gravitational water; Number of samples (n) in Andisols for natural forest were 5 and 6, and for pasture 12 and 13 for A and B horizons, respectively; in Inceptisols for natural forest were 6 and 7 and 12 and 12 for A and B horizons, respectively; +, *, ** Significant differences between forest and pasture with Mann Whitney U test at p < 0.1, p < 0.05 and p < 0.01, respectively. Horizon Forest Pasture Forest PastureA 79.2 (70.9-83.1)277.5 (76.1-83.7) 72.4 (66.8-80.8) 67.9 (65.4-70.4)B 74.9 (70.3-77.7) 74.7 (66.3-75.6) 68.4 (67.4-69.5) 61.9 (60.8-64.1)**A 63.2 (55.9-67.5) 67.8 (64.9-71.0) 58.2 (53.2-61.9) 56.6 (51.7-58.9)B 60.6 (50.9-65.9) 61.4 (55.5-64.7) 54.5 (52.3-59.1) 50.8 (49.3-53.5)*A 55.0 (45.0-58.2) 58.9 (55.1-63.2)+51.8 (46.5-55.4) 50.3 (46.0-53.1)B 50.7 (44.9-54.5) 53.4 (46.2-56.8) 48.1 (46.2-55.9) 46.0 (44.1-49.0)A 50.8 (42.2-54.7) 55.8 (52.2-60.5)* 49.8 (45.0-50.4) 47.7 (43.3-49.7)B 48.6 (43.3-53.0) 50.4 (43.8-51.7) 45.6 (44.4-53.9) 43.3 (41.7-47.3)+A 44.1 (36.4-50.5) 52.5 (49.5-57.5)** 43.8 (40.9-47.5) 42.2 (40.2-43.5)B 47.7 (42.1-49.7) 46.1 (43.4-51.7) 43.4 (40.9-44.2) 39.4 (36.8-43.2)A 18.4 (15.9-20.9) 15.4 (11.8-17.3) 13.5 (12.1-16.9) 13.3 (11.9-15.7)B 14.3 (10.1-15.4) 12.6 (10.3-15.1) 12.5 (10.2-19.5) 11.0 (9.3-13.5)A 12.8 (10.9-21.2) 10.9 (9.5-12.4) 16.8 (12.9-18.9) 11.4 (8.9-18.1)B 11.4 (8.9-18.5) 11.5 (8.4-14.4) 10.4 (8.8-15.1) 11.7 (7.3-12.8)θ500 kPa1 (%v/v)θPWP1 (%v/v)PAWS3 (%v/v)GW4 (%v/v)Andisols InceptisolsθSat1 (%v/v)θFC1 (%v/v)θ100 kPa1 (%v/v)92 5.3.2 Soil water retention characteristics under natural forest and pasture as affected by soil properties Correlations among SWR characteristics and soil properties are provided in Tables 5.3 through 5.5. Few significant correlations were found under natural forest using the complete data set; however, sand and clay size fractions, SOC, SRO and ρb were all correlated with SWR characteristics under pasture (Table 5.3). The lower number of samples under natural forest and the higher variability in SWR under forest (Fig. 4.3) relative to pasture (Fig. 4.4), may partially explain the lower number of significant correlations under natural forest. 5.3.2.1 Differences between land uses in Andisols Similar correlation patterns as those presented in Chapter 4 (Table 4.6a) are seen in the A and B horizons of Andisols, but with distinct differences between land uses (Table 5.4). In the A horizon, SWR was correlated with sand size fraction, clay size fraction and organo-metallic compounds under natural forest, and with SOC and SRO under pasture. Comparison between land uses, showed a significantly higher SOC under natural forest in relation to pasture (Table 5.6). Numerous studies in the tropics (e.g., Van Noordwijk et al., 1997; Navarrete et al., 2016) have found that land cover change from forests to pasture may result in either higher or lower SOC, depending on land management practices such as grazing intensity and soil C loss due to erosion. Comparison between land uses also showed a significantly higher sand size fraction under natural forest in relation to pasture (Table 5.6). The higher SOC, the higher measured sand size fraction, and the positive correlation between SOC and the apparent sand fraction (Fig. 5.4) suggest aggregation, which may contribute to the steeper SWR curve under natural forest. In addition, the pH under natural forest was significantly lower than under pasture (Table 5.6). Since both SOC and SRO have dominantly pH dependent charge (Table 3.1), SWR may be lower under natural forest. 93 Figure 5.4 Relationship between soil organic carbon (SOC) and sand size fraction in A horizon of natural forest and pasture of Andisols Regional studies which compared SWR characterstics between forest and pasture in Andisols are limited and suggest contrasting impacts on SWR in the A horizon (Table 5.7a). Tobón et al. (2010) found higher PAWS and lower GW under pasture than under natural forest, but similar θPWP values. In contrast, this study found no change in GW or PAWS but higher θPWP under pasture, while Roa et al. (2011) noted no significant differences in SWR between natural forest and pasture. All three of these sites were dominated by allophanic mineral soils. SOC was lower under pasture in this study consistent with Tobón et al. (2010). As noted earlier, forest conversion to pasture may result in lower SOC depending on management, particularly on sloping land were grazing and erosion may contribute to losses of soil C (Noordwijk et al., 1997; Navarrete et al., 2016). In the B horizon of the Andisol site, SWR characteristics were similar between land uses (Fig. 5.3a), and no significant differences were noted in SOC, SRO or bulk density (Table 5.6). Tobón et al. (2010) also found limited differences in SOC and SWR characteristics in the B horizon (Table 5.7). n = 5 r = 0.90 p < 0.04 n = 12 94 Table 5.3 Spearman´s Rho correlation coefficients (r) between soil water retention (SWR) characteristics and soil physical and chemical soil properties for both soil orders in all horizons and in A and B horizons in: a) natural forest; and b) pasture 1θSat, θFC, θ100 kPa, θ500 kPa, θPWP: soil water content (θ) at saturation (Sat), field capacity (FC), 100 kPa, 500 kPa and at permanent wilting point (PWP), respectively; 2PAWS: plant available water storage; 3GW: gravitational water; 4SOC: soil organic carbon; 5Pyrophosphate-extractable aluminum and iron (Alp and Fep); 6ρb: bulk density; correlations shown are the ones r > 0.4; Gray cells show correlations with p < 0.01 and white cells show correlations with p < 0.05 θSat1 (%v/v)θFC1 (%v/v)θ100 kPa1 (%v/v)θ500 kPa1 (%v/v)θPWP kPa1 (%v/v)PAWS2 (% v/v)GW3 (% v/v)θSat1 (%v/v)θFC1 (%v/v)θ100 kPa1 (%v/v)θ500 kPa1 (%v/v)θPWP kPa1 (%v/v)PAWS2 (% v/v)GW3 (% v/v)θSat1 (%v/v)θFC1 (%v/v)θ100 kPa1 (%v/v)θ500 kPa1 (%v/v)θPWP kPa1 (%v/v)PAWS2 (% v/v)GW3 (% v/v)a) Natural forest All horizons (n = 27) A horizon (n = 11) B horizon (n = 13)Sand (%) 0.61Silt (%)Clay (%) -0.58SOC4 (%) 0.53 0.41 0.50 0.71Alp5 (g/kg)Fep5 (g/kg) 0.68 0.63Allophane (%) 0.58 0.59Ferrihydrite (%) -0.43 -0.61ρb6 (kg/m3) -0.89 -0.54 -0.61 -0.58 -0.90 -0.74 -0.77 -0.61 -0.85b) Pasture All horizons (n = 59) A horizon (n = 24) B horizon ( n = 25)Sand (%) 0.52 0.64 0.66 0.63 0.66 0.67 0.46 0.41Silt (%)Clay (%) -0.51 -0.60 -0.60 -0.66 -0.69 -0.68 -0.47 -0.46SOC4 (%) 0.59 0.49 0.50 0.68 0.77 0.70 0.71 0.76 0.60Alp5 (g/kg) 0.69 0.66 0.53 0.52 0.59 0.63 0.67 0.61 0.62 0.69 0.81 0.76 0.55 0.52 0.71Fep5 (g/kg) 0.53 0.49 0.40 0.43 0.65 0.66 0.57 0.62 0.66Allophane (%) 0.56 0.49 0.51 0.77 0.80 0.79 0.75 0.82 0.80 0.61 0.58Ferrihydrite (%) -0.70 -0.77 -0.74 -0.80 -0.76ρb6 (kg/m3) -0.90 -0.75 -0.60 -0.60 -0.62 -0.45 -0.93 -0.89 -0.82 -0.76 -0.78 -0.50 -0.86 -0.4795 Table 5.4 Spearman´s Rho correlation coefficients (r) between soil water retention (SWR) characteristics and soil physical and chemical soil properties within Andisols in all horizons and in A and B horizons in: a) natural forest; and b) pasture 1θSat, θFC, θ100 kPa, θ500 kPa, θPWP: soil water content (θ) at saturation (Sat), field capacity (FC), 100 kPa, 500 kPa and at permanent wilting point (PWP), respectively; 2PAWS: plant available water storage; 3GW: gravitational water; 4SOC: soil organic carbon; 5Pyrophosphate-extractable aluminum and iron (Alp and Fep); 6ρb: bulk density; correlations shown are the ones r > 0.4; Gray cells show correlations with p <0.01 and white cells show correlations with p < 0.05 θSat1 (%v/v)θFC1 (%v/v)θ100 kPa1 (%v/v)θ500 kPa1 (%v/v)θPWP kPa1 (%v/v)PAWS2 (% v/v)GW3 (% v/v)θSat1 (%v/v)θFC1 (%v/v)θ100 kPa1 (%v/v)θ500 kPa1 (%v/v)θPWP kPa1 (%v/v)PAWS2 (% v/v)GW3 (% v/v)θSat1 (%v/v)θFC1 (%v/v)θ100 kPa1 (%v/v)θ500 kPa1 (%v/v)θPWP kPa1 (%v/v)PAWS2 (% v/v)GW3 (% v/v)a) Natural forest All horizons (n = 12) A horizon (n = 5) B horizon ( n = 6)Sand (%) -0.90 -0.90Silt (%)Clay (%) 0.68 1.00 0.90 1.00SOC4 (%) 0.69 0.89Alp5 (g/kg) 0.90 -0.90Fep5 (g/kg) 0.90 0.90 1.00 0.90Allophane (%) -0.90 -0.90Ferrihydrite (%) 0.68 0.83ρb6 (kg/m3) -0.74 -0.83b) Pasture All horizons (n = 32) A horizon (n = 12) B horizon (n = 13)Sand (%) -0.55 -0.62 -0.57 -0.59 -0.62Silt (%) 0.47 0.41 0.62Clay (%) 0.50 0.60 0.61 0.56 0.59 0.69 0.73SOC4 (%) 0.77 0.70 0.46 0.49 0.40 0.57 0.75 0.87 0.68 0.71 0.93Alp5 (g/kg) 0.58 0.62 0.55 0.56Fep5 (g/kg) 0.60 0.61 0.55Allophane (%) 0.60 0.59 0.90 0.89 0.73Ferrihydrite (%) 0.41 -0.60 -0.59 -0.66 0.86ρb6 (kg/m3) -0.85 -0.66 -0.55 -0.53 -0.44 -0.50 -0.81 -0.80 -0.66 -0.64 -0.65 -0.7796 Table 5.5 Spearman´s Rho correlation coefficients (r) between soil water retention (SWR) characteristics and soil physical and chemical soil properties within Inceptisols in all horizons and in A and B horizons in: a) natural forest; and b) pasture 1θSat, θFC, θ100 kPa, θ500 kPa, θPWP: soil water content (θ) at saturation (Sat), field capacity (FC), 100 kPa, 500 kPa and at permanent wilting point (PWP), respectively; 2PAWS: plant available water storage; 3GW: gravitational water; 4SOC: soil organic carbon; 5Pyrophosphate-extractable aluminum and iron (Alp and Fep); 6ρb: bulk density; correlations shown are the ones r > 0.4; Gray cells show correlations with p < 0.01 and white cells show correlations with p < 0.05 θSat1 (%v/v)θFC1 (%v/v)θ100 kPa1 (%v/v)θ500 kPa1 (%v/v)θPWP kPa1 (%v/v)PAWS2 (% v/v)GW3 (% v/v)θSat1 (%v/v)θFC1 (%v/v)θ100 kPa1 (%v/v)θ500 kPa1 (%v/v)θPWP kPa1 (%v/v)PAWS2 (% v/v)GW3 (% v/v)θSat1 (%v/v)θFC1 (%v/v)θ100 kPa1 (%v/v)θ500 kPa1 (%v/v)θPWP kPa1 (%v/v)PAWS2 (% v/v)GW3 (% v/v)a) Natural forest All horizons (n = 15) A horizon (n = 6) B horizon ( n = 7)Sand (%)Silt (%) -0.83Clay (%) 0.94 -0.79 -0.79SOC4 (%) 0.60 0.94 0.89Alp5 (g/kg) -0.51Fep5 (g/kg) 0.89Allophane (%)Ferrihydrite (%) -0.54 -0.94ρb6 (kg/m3) -0.77 -0.80 -1.00 -0.83 -0.83 -0.83b) Pasture All horizons ( n = 27) A horizon ( n = 12) B horizon (n = 12)Sand (%)Silt (%)Clay (%)SOC4 (%) 0.50 0.43Alp5 (g/kg) 0.65Fep5 (g/kg) 0.38Allophane (%)Ferrihydrite (%) 0.53ρb6 (kg/m3) -0.66 -0.52 -0.68 -0.64 -0.69 -0.5997 Table 5.6 Median results of soil properties for natural forest and pasture in A and B horizons of Andisols and Inceptisols 1Values in parenthesis are first and third quartile, 2SOC: soil organic carbon; 3Pyrophosphate-extractable aluminum and iron (Alp and Fep); 4ρb: bulk density; Number of samples (n) in Andisols were 5 and 6 in natural forest and 12 and 13 in pasture in A and B horizons respectively, in Inceptisols were 6 and 7 in natural forest and 12 and 12 in pasture in A and B horizons, respectively; +, *, ** Significant differences between forest and pasture with Mann Whitney U test at p < 0.1, p < 0.05 and p < 0.01, respectively Horizon Forest Pasture Forest PastureA 4.6 (4.5-4.7)1 5.1 (5.0-5.2)** 5.6 (5.4-5.8) 5.6 (5.4-5.7)B 5.5 (5.4-5.6) 5.7 (5.5-5.7)+ 5.1 (5.0-5.6) 5.5 (5.3-5.9)+A 52.6 (49.3-60.6) 42.2 (32.3-46.3)** 14.2 (10.2-36.4) 13.6 (11.6-19.7)B 60.0 (48.7-68.1) 55.1 (51.0-59.8) 20.0 (18.1-29.2) 18.9 (8.0-22.7)A 23.8 (21.7-29.0) 23.3 (22.3-31.1) 39.3 (27.0-52.6) 56.2 (50.2-60.1)*B 14.6 (12.1-15.3) 15.9 (14.7-17.6)+30.6 (22.5-45.6) 53.2 (42.8-58.3)*A 12.5 (10.0-14.6) 7.8 (7.1-11.5)* 7.1 (4.9-10.5) 5.0 (3.6-5.9)+B 3.8 (2.6-4.3) 3.5 (2.6-4.8) 1.7 (1.4-3.5) 1.6 (1.4-2.5)A 1.40 (1.09-2.14)1 1.53 (1.07-1.72) 0.06 (0.05-0.08) 0.08 (0.04-0.21)B 0.32 (0.22-0.57) 0.42 (0.27-0.55) 0.13 (0.08-0.23) 0.04 (0.03-0.12)*A 0.22 (0.14-0.39) 0.37 (0.15-0.50) 0.02 (0.01-0.03) 0.03 (0.01-0.06)B 0.01 (0.01-0.03) 0.01 (0.01-0.04) 0.03 (0.01-0.04) 0.01 (0.00-0.03)A 6.0 (4.4-7.3) 6.3 (4.6-9.2) 1.5 (1.2-3.6) 1.4 (1.2-2.1)B 15.5 (7.0-17.4) 14.8 (10.5-17.1) 1.6 (1.2-3.4) 1.2 (1.0-1.4)+A 0.45 (0.36-0.50) 0.49 (0.44-0.60) 0.83 (0.64-1.18) 1.2 (1.0-1.6)*B 0.44 (0.35-0.79) 0.7 (0.2-0.8) 0.6 (0.4-1.2) 0.7 (0.5-1.0)A 455 (326-740) 594 (461-665) 727 (539-847) 900 (840-935)+B 650 (544-894) 668 (578-829) 887 (834-1041) 1016 (958-1067)Andisols Inceptisols pH H2OSOC2 (%)Clay (%)Sand (%)b4 (kg/m3)Alp3(g/kg)Fep3 (g/kg)Allophane and imogolite (%)Ferrihydrite (%)98 Table 5.7 Soil water retention (SWR) characteristics and soil properties of Andisols of the Sonora watershed and other regional studies a) A horizon Land use GW PAWS2 θSat3 θFC3 θPWP3 SOC4 ρb5 Alp6 Alo6 Reference (cm3/cm3) (%) (kg/m3) (g/kg) (g/kg) Allophanic soils Forest Pasture 13 11 18 15 79 77 63 68 44 52 12 8 455 594 1.4 1.5 14 15 This study Forest Pasture 29 19 8 15 82 82 53 63 45 48 16 8 407 528 - - - - Tobón et al., 2010 Forest Pasture - - 7 8 - - 61 60 53 52 - - 600 700 - - - - Roa-García, et al., 2011 Non-allophanic soils Humid Páramo Pasture Dry Páramo Pasture - - - - - - - - - - - - 52 15 20 15 - - - - 10 5 7 4 680 760 740 990 4 2 2 1 6 3 4 2 Podwojewski et al., 2002 b) B horizon Allophanic soils Forest Pasture 11 11 14 13 75 75 61 61 48 46 3.8 3.5 650 668 0.3 0.4 39 35 This study Forest Pasture 19 19 16 16 74 80 55 61 39 45 7.8 9.2 612 627 - - - - Tobón et al., 2010 Non-allophanic soils HumidPáramo Pasture Dry Páramo Pasture - - - - - - - - - - - - 50 20 30 14 - - - - 7 3 5 2 - - 930 - 5 1 2.5 0.4 9 2.5 4 2 Podwojewski et al., 2002 1GW: gravitational water; 2PAWS: plant available water storage; 3θFC and θPWP: soil water content (θ) at field capacity (FC) and at permanent wilting point (PWP); 4SOC: soil organic carbon; 5ρb: bulk density; 6Alp and Alo: pyrophosphate and oxalate extractable aluminum (Al), respectively 99 5.3.2.2 Differences between land uses in Inceptisols In the case of Inceptisols, SWR characteristics in the A horizon were similar between land uses (Fig. 5.2b and Table 5.2). Although SOC was lower in pasture (similar to the Andisol site), the clay fraction and ferrihydrite were higher (Table 5.6). Ferrihydrite, due to its high specific surface area and large number of functional groups, may partially compensate in SWR for the lower SOC in the A horizon under pasture, in spite of its small contribution to soil mass (<1%) (Regelink et al., 2015; Goldberg et al., 2012). Bulk density was negatively correlated with SWR for both land uses (Table 5.5), similar to relationships found in Chapter 4 (Table 4.6b). Bulk density was slightly higher in the A horizon of Inceptisols under pasture (Table 5.6). As noted by Daza et al. (2014) compaction due to cattle grazing may increase ρb and reduce GW in Inceptisols (Table 5.8), although no significant change in GW was found in this study (Table 5.2). In the B horizon of Inceptisols, SWR was greater at low tensions under natural forest relative to pasture (Fig. 5.3b and Table 5.2). While SOC content was similar, Alp, allophane and imogolite were higher under natural forest (Table 5.6). Allophane and imogolite due to their hollow structures, high specific surface area and large number of functional groups, as discussed in Chapter 3, contribute to higher SWR. Alp may be an indicator of boehmite in Inceptisols, as explained in Section 3.3.3.1, and their relatively high specific surface area also contributes to SWR. Boehmite, allophane and imogolite may all contribute to aggregation and an increase in macro-pores, accounting for the increase in total porosity under natural forest. 100 Table 5.8 Soil water retention (SWR) characteristics and soil properties of Inceptisols of the El Chocho watershed and other regional studies a) A horizon Land use GW PAWS2 θSat3 θFC3 θPWP3 SOC4 ρb5 Alp6 Allophanes Ferrihydrite Reference (cm3/cm3) (%) (kg/m3) (g/kg) (%) (%) Allophanic soils Forest Pasture 17 11 13 13 72 68 58 57 44 42 7 5 727 900 0.06 0.08 14 15 0.83 1.2 This study Páramo6 Pasture Potato crops 21 15 10 23 17 18 78 57 49 57 42 39 34 25 21 18 10 12 700 830 900 - - - - - - - - - Daza et al., 2014 b) B horizon Forest Pasture 10 12 12 11 68 62 54 51 43 39 1.7 1.6 887 1016 0.13 0.04 1.6 1.2 0.6 0.7 This study Páramo6 Pasture Potato crops 19 15 13 37 29 23 100 64 63 81 50 50 44 21 24 13 11 1 1000 1200 1100 - - - - - - - - - Daza et al., 2014 1GW: gravitational water; 2θSat, θFC and θPWP: soil water content (θ) at saturation (Sat), field capacity (FC) and at permanent wilting point (PWP); 3SOC: soil organic carbon; 4ρb: bulk density; 5Alp and Alo: pyrophosphate and oxalate extractable aluminum (Al), respectively; 6Volumetric soil moisture was calculated with the published data of gravimetric soil moisture and bulk density 101 The effect of land use change on SWR has been shown to vary depending on site characteristics, management factors such as grazing intensity, and sampling design whether by horizon, as in this study, or by depth (Horel et al., 2015). Intensive land use commonly increases ρb and reduces SOC but research shows contrasting results in GW, PAWS and θPWP (Asghari et al., 2016; Pirastru et al., 2013), and may be a reflection of different inherent soil characteristics such as SRO, which are not appreciably influenced by land management. In mountainous regions, disturbance of the original land cover, may have a negative effect on soil structure and soil loss which are closely related with a reduction in SOC and SWR (Li et al., 2007). In this study, the conversion of natural forest to pasture, showed no changes in GW or PAWS in either soil. However, steeper SWR curves were noted under natural forest in the A horizon of Andisols and the B horizon of Inceptisols. 5.4 Conclusions Land use effects on SWR characteristics in Andisols and Inceptisols appear limited, although SOC was lower under pasture than natural forest in both soils. Significant differences between pasture and natural forest were found only at PWP in the A horizon of Andisols, and θSat and FC in the B horizon of Inceptisols. In the Andisols of the Sonora watershed, as discussed earlier, aggregate formation gives rise to a pseudo-sand fraction that may be increasing the slope of the SWR curve in the A horizon under forest. In Inceptisols of the El Chocho watershed ferrihydrite appears to offset the effects of lower SOC under pasture in the A horizon resulting in no appreciable differences in SWR between the land uses. The limited differences in SWR between natural forest and pasture appear to reflect the effects of SRO minerals and organo-metallic compounds on SWR. The high water retention capacity of SRO and organo-metallic compounds may compensate the lower SOC under pasture, resulting in similar SWR characteristics between land uses. 102 6. FIELD SATURATED HYDRAULIC CONDUCTIVITY DURING DRY AND WET SEASONS UNDER PASTURE ON ANDISOLS AND INCEPTISOLS AT TWO MID-ELEVATION SITES IN THE COLOMBIAN ANDES 6.1 Introduction Field saturated hydraulic conductivity (Kfs) (quasi steady-state infiltration rate) is a key soil physical property that affects the partitioning of rainfall into infiltration and overland flow (OF) (Nimmo et al., 2009; Bonell, 1993). Kfs is dependent on soil properties such as texture, structure, soil organic matter, pore size distribution, and bulk density (ρb); it varies spatially due to the local geomorphology, topography and land cover, and temporally due to the antecedent soil water content (θ) (Bonell, 1993; Assouline, 2013). Consequently, a field-based measurement incorporating spatial and temporal variation is needed for an accurate determination of soil water movement (Diamond and Shanley, 2003). In the Andes, pasture has been recognized as a particularly important land use with a direct impact on runoff. Studies such as Tobón et al. (2010) and Zimmermann and Elsenbeer (2008) found higher runoff in pasture relative to forest, and this increase in runoff is generally attributed to livestock trampling and reduced infiltration (Leitinger et al., 2010; Chaves et al., 2008). Greater runoff under pasture in comparison to natural forests has been found in the Central Colombian Andes (Suescún et al., 2017; García-Leoz et al., 2018) and in Ecuador (Molina et al., 2007). These studies also suggested a decrease in soil infiltration associated with pasture, although Kfs was not directly measured. In spite of the relevance of Kfs and associated soil properties, limited research has focused on the role of soils in runoff generation in both the Andes and the tropics (Bonell, 1993; Ilsted et al., 2007). Runoff models, such as the Soil Water Assessment Tool (SWAT), require soil parameters including saturated hydraulic conductivity (Ksat) as input variables (Arnold et al., 2012). In Colombia, Ecuador and Perú, hydrological studies (Quintero et al., 2009; Uribe et al., 2013) have relied on soil texture for estimating Ksat values instead of the more accurate measured Kfs values. Modelling of runoff from watersheds could be improved with additional data on the main soil orders with consideration of spatial and temporal variability. 103 The objectives of this chapter were to compare the field saturated hydraulic conductivity (Kfs) under pasture in Andisols and Inceptisols in the Colombian Andes during the wet and dry seasons, and determine the factors affecting Kfs in these two soils. Kfs was then compared to rainfall intensity to provide a preliminary estimate of OF. 6.2 Experimental conditions and measurements The experiment was designed to compare Kfs on pasture between two soil orders and between two seasons. Kfs was measured using double ring infiltrometers, and as indicated in Chapter 2, provides an index of quasi steady-state infiltration rate as measured in the field. Measurements were taken on flat slope positions only (slope <20%) and near the locations of the sampled soil pits discussed in earlier chapters of this thesis in both Andisol and Inceptisol study areas. Measurements were repeated in the dry and the wet seasons. 6.2.1 Measurement of field saturated hydraulic conductivity Field saturated hydraulic conductivity (Kfs) was measured using double ring infiltrometers following the method developed by Bouwer (1986). Kfs was measured in duplicate at each site; a pair of double ring infiltrometers were used simultaneously approximately 10 m apart. Before every Kfs measurement, a soil sample was taken from the 0-15 cm depth, next to the site where the double ring was installed, for determination of the gravimetric soil water content (θgrav). Dates when Kfs was measured in the dry and wet seasons are presented in Table 6.1. Table 6.1 Dates when field saturated hydraulic conductivity (Kfs) were measured in the Sonora watershed (Andisol site) and in El Chocho watershed (Inceptisol site) Watershed - soil order Dry season Wet season Sonora - Andisol August 28-31, 2013 November 19-22, 2013 El Chocho - Inceptisol August 5-8, 2013 November 27-28, 2013 104 The double ring infiltrometer used for the measurements was 30 cm in height with sharpened bottom edges. The outer ring was 60 cm in diameter, while the inner ring was 30 cm in diameter. Rings were driven 15 cm into the soil. Inner and outer rings were filled with water (more than 7 cm above the ground level) prior to beginning the measurements. The infiltration rate was assessed by measuring the change in water level every 30 seconds for the first 5 minutes, every minute from 6 to 10 minutes, every 5 minutes from 15 to 45 minutes, and every 20 minutes until steady state. Outer and inner rings were periodically filled with water immediately after a water level measurement, to maintain the water level at a minimum of 7 cm above the ground level. The change of the water level was measured until steady state was reached; determined by an equal change in water level over at least two periods of 20 minutes. Steady state was reached in Andisols sites between 2 to 4 hours, while in Inceptisol sites, steady state was reached between 2 to 5 hours. θgrav was assessed gravimetrically by oven drying the soil sample at 105°C for 24 hours. Volumetric soil water content (θvol) was calculated using θgrav and the median ρb in the A horizon from each of the pasture sites on flat slope position. 6.2.2 Estimation of overland flow Overland flow will occur if the rainfall intensity exceeds the infiltration capacity of the soil; and Kfs defines a minimum absorption capacity, as the soil is capable of storing additional water, depending on the antecedent soil moisture conditions and micro-relief (Diamond and Shanley, 2003). Thus, estimating runoff as rainfall intensity > Kfs provides comparable results between sites but could result in an overestimation of the OF particularly in the dry season. To estimate OF, the rainfall intensity at 10-minute intervals (RI10) was compared to the median Kfs in the wet and the dry seasons (Kfswetseason and Kfsdryseason). Continuous precipitation data (every 0.2 mm), from within each watershed, was obtained from a research project (Roa and Brown, 2014); details are provided in Appendix G. Data was then organized for 10 minutes periods to obtain RI10. The interval of 10 minutes for rainfall intensity was defined as half of the time used for determing the quasi steady state infiltration rate (Kfs) at the end of the infiltration measurements, and captures the most 105 common rainfall events in the two watersheds (Fig. 6.3). For comparative purposes with other studies, the work of Pardo Gomez and Rodrígues Lopez (2014), was used to define low intensity events as the ones with RI10 < 10 mm/hr. Two years of precipitation data were utilized, allowing for the separation of wet and dry seasons. For each 10-minute interval within each season, RI10 was compared to Kfs and OF estimated as given in Table 6.2. Table 6.2 Rationale for overland flow (OF) estimation comparing rainfall intensity every 10 minutes (RI10) with field saturated hydraulic conductivity (Kfs) Dry season Wet season OF = 0 mm RI10 < Kfsdryseason RI10 < Kfswetseason OF = RI10 - Kfs RI10 > Kfsdryseason RI10 > Kfswetseason Organizing data by season during the two years for which data was available for both watersheds, allowed the estimation of the percentage of OF over total precipitation (OF/TP) by season, by year and over the two-year study period. 6.2.3 Statistical analyses Field saturated hydraulic conductivity (Kfs), θgrav and θvol were compared between Andisols and Inceptisols, over the entire study period and by season (Fig. 6.1a). In addition, data for Andisols and Inceptisols were evaluated separately to determine differences between dry and wet seasons within each soil order (Fig. 6.1a). The Mann Whitney U test and probability values (p-value) of 0.01, 0.05 and 0.1 were used for these comparisons. Relationships between Kfs and θgrav and θvol were assessed utilizing the non-parametric Spearman’s rank-order correlation. Spearman’s Rho correlation coefficients (r) with values > 0.4 and probability values (p) of 0.01 and 0.05 were used to indicate significant relationships (Fig. 6.1b). Correlations were conducted for: 106 • Andisol and Inceptisol, all seasons ▪ Andisol and Inceptisol by season (dry and wet seasons) o Andisol only, all seasons ▪ Andisol by season (dry and wet seasons) o Inceptisol only, all seasons ▪ Inceptisol by seasons (dry and wet seasons) Sample numbers for each analysis are provided in Figure 6.1 Figure 6.1 Overview of field measurements to a) compare field saturated hydraulic conductivity (Kfs) in pasture between Andisols and Inceptisols and b) determine relationships between Kfs and soil water content 6.3 Results and discussion Steady infiltration state or saturated hydraulic conductivity is a key parameter in hydrological modelling to partition rainfall into runoff and infiltration, and is used in irrigation calculations to determine the rate at which water should be applied to avoid a) Field hydraulic saturated conductivity in pasture: differences between Andisols and Inceptisols Kfs θgrav θvol Mann-Whitney U-test b) Relationships between field saturated hydraulic conductivity and soil water content Spearman´s rank order correlations n = 6x2 n = 6x2 n = 6x2 n = 6x2 Andisol Inceptisol Inceptisol Andisol Dry season Wet season Dry season Wet season Andisol and Inceptisol Andisol Inceptisol All seasons Wet season Dry season n = 24 n = 12 n = 12 n = 24 n = 12 n = 12 n = 6x2 n = 6x2 n = 6x2 n = 6x2 Wet season Dry season Andisol Inceptisol Wet season, n = 24 Dry season, n = 24 All seasons, n = 48 Wet season, n = 12 Dry season, n = 12 All seasons, n = 24 Wet season, n = 12 Dry season, n = 12 All seasons, n = 24 107 excess application and OF. Yet Kfs is challenging and time-consuming to measure in the field, and is often estimated from a more easily measured soil parameter such as textural class (NCSS, 1996). In the following sections Kfs values determined under pasture in Andisols and Inceptisols at two mid-elevation sites will be compared to each other, to infiltration categories based on soil texture, and to Kfs values reported in other studies conducted in the region. Seasonal variability and soil characteristics affecting Kfs will be discussed, followed by a first approximation of OF based on a comparison between Kfs and rainfall intensity. 6.3.1 Field saturated hydraulic conductivity (Kfs) in pasture: differences between Andisols and Inceptisols Infiltration categories classified by soil texture are given in Table 6.3, with clay textured soils having lower Kfs values than sandy soils. In contrast, results from this study (Table 6.4) show that Inceptisols in the El Chocho watershed, which have a clay texture, had significantly higher Kfs values than Andisols in the Sonora watershed with dominantly sandy loam to loam texture. Median Kfs values for Andisols fell within the “slow” infiltration category, while Inceptisols ranked as “moderately slow”. Table 6.3 Infiltration categories and hydraulic conductivity relative to texture class adapted from NCSS (1996) Infiltration category Texture class Hydraulic conductivity (mm/hr) Very slow Heavy clay < 1 Slow Clay, clay loam, silty clay, sandy clay loam 1 – 5 Moderately slow Sandy clay, silty clay loam, clay loam, silt loam, silt, sandy clay loam 5 – 20 Moderate Clay loam, silt, silt loam, very fine sandy loam, loam 20 – 60 Moderately rapid Fine sandy loam, sandy loam 60 -125 Rapid Loamy sand, fine sand 125 - 250 Very rapid Medium sand > 250 108 Table 6.4 Median field saturated hydraulic conductivity (Kfs) and soil water content (θ) in Andisols and Inceptisols 1Kfs: field saturated hydraulic conductivity; 2θgrav: field gravimetric soil water content; 3θvol: calculated volumetric soil water content; 4Values in parenthesis are first and third quartile; Number of samples (n) were 24 for Andisols and Inceptisols; *, ** Significant differences between Andisols and Inceptisols with Mann Whitney U test at p < 0.05 and p < 0.01, respectively Kfs values in Andisols in the Sonora watershed were lower than what is commonly reported in the literature (Nanzyo et al., 1993; Perrin et al., 2001; Neris et al., 2012; Tobón et al., 2010). The pseudo-sandy texture of Andisols in this study (Section 3.3.3.4) may suggest high Kfs values; however, this was not observed. The low Kfs values of Andisols in this study may be explained by a combination of soil properties and antecedent soil moisture regime. The high volume of micro-pores (Section 4.5.1 and Table 4.2), and the high soil water content throughout the year (Fig. 6.2), are related to both the high SWR characteristics of these soils (Section 4.3.1) and high rainfall. The precipitation (~3,000 mm/year) is high enough in the Sonora watershed (Andisol site), that the majority of θgrav was > 70% throughout the year (Fig. 6.2). The presence of volcanic ash layers and placic horizons (Sections 2.1.1; 2.1.3; and 2.1.4.1) may impede drainage and reduce measured Kfs values. Volcanic ash layers were found at 1 – 2 meters depth in the upper portion of the Sonora watershed, and a placic horizon was observed at approximately 2.8 meters in the lower watershed. No correlations were found in Andisols between Kfs and antecedent soil water content, even when high Kfs values are excluded (Tables 6.5 and 6.6). High SWR characteristics in Andisols in combination with the high precipitation in the Sonora watershed (~3,000 mm/year) maintain high θgrav which combined with imperfect drainage Andisols Inceptisols3 (0.1 - 7.1)49 (2 - 33)*102 (85-164) 57 (30 - 67)**0.51 (0.45-0.63) 0.47 (0.26 - 0.54)*θgrav2 (g/g)θvol3 (cm3/cm3)Kfs1 (mm/hr)109 impacts measured values of Kfs, reducing it even in the dry season to values close to 0 mm/hr (Fig. 6.2). In general, the Kfs values for the Inceptisols (Table 6.8) were found to be slow in the wet season consistent with the values reported in Table 6.3. However, during the dry season the Kfs values were moderate, even though the texture class is clay, which should place the Kfs into the slow category of Table 6.3. These dry season values are also high in comparison to infiltration data reported by studies conducted in similar soils in the region (Table 6.5) (Zimmermann and Elsenbeer, 2008; Toohey et al., 2018). The seasonal effect in measured infiltration in the Inceptisols of this study is related to the lower antecedent soil water content during the dry season. Recall that precipitation in the El Chocho watershed (Inceptisol site), ~1400 mm/year, is about 60% lower than in the Sonora watershed (Section 2.1.2) and is also lower than the precipitation at regional studies sites (Table 6.7). Thus, the negative correlation between Kfs and antecedent soil moisture in Inceptisols observed in Figure 6.2 with the majority θgrav < 70% (Tables 6.5 and 6.6) may in part explain the higher values of Kfs in comparison to regional studies and expected Kfs based on texture alone. Inconsistencies were found between measured Kfs values and Kfs predicted based on soil texture class for both Andisols and Inceptisols. Measured values of Kfs in Andisols were lower than predicted based on soil texture class, while Kfs measured in Inceptisols was higher than predicted, highlighting the importance of field data for use in hydrological modelling or agricultural practices. 110 a) Andisols b) Inceptisols Figure 6.2 Gravimetric soil water content (θgrav) vs field saturated hydraulic conductivity (Kfs) in the wet and dry seasons in: a) Andisols; and b) Inceptisols Table 6.5 Spearman´s Rho correlation coefficients (r) between field saturated hydraulic conductivity (Kfs) and soil water content 1Kfs: field hydraulic saturated conductivity; 2θgrav: gravimetric soil water content; 3θgrav: volumetric soil water content; gray cells show correlations with p < 0.01 Table 6.6 Spearman´s Rho correlation coefficients (r) between field saturated hydraulic conductivity (Kfs) and soil water content without high values (greater than 50mm/hr) 1Kfs: field hydraulic saturated conductivity; 2θgrav: gravimetric soil water content; 3θgrav: volumetric soil water content; gray cells show correlations with p < 0.01 n = 24 r = -0.65 p < 0.01 n = 24 111 Table 6.7 Field saturated hydraulic conductivity (Kfs) values reported in other studies carried out in the tropics and in the Andes Land cover Site/Country Elevation (m) Precipitation (mm/year) Texture Kfs1 (mm/hr) Method Reference Pasture (Andisols) Pasture (Inceptisols) Sonora watershed, Colombia El Chocho watershed, Colombia 2,088-2,331 1,768-2,111 2,955 1,393 Sandy loam, loam, sandy clay loam Clay, silty clay, silty clay loam Median dry – wet season 5 – 0.7 31 – 3 Double ring infiltrometer This study Pasture (Andisols) San Gerardo area, Costa Rica 1,520-1,620 4,400-6,000 Sandy loam, loam Average 14.7 8.7 under cow trails Guelph permeameter Tension infiltrometer Small-cores method Tobón et al., 2010 Pasture (Inceptisols) Reserva Biosfera de San Francisco, Ecuador 1,860 2,273 Silt loam Median 3 Ammozemeter (constant head permeameter) Zimmermann and Elsenbeer, 2008 Pasture (Inceptisols) Catie2 farm, near Turrialba, Costa Rica 650 2,500-3,000 Clay loam, loam Average 12 Double ring infiltrometer Toohey et al., 2018 1Kfs: field saturated hydraulic conductivity; 2 CATIE: Tropical Agricultural Research and Higher Education Center112 When separating Kfs data of Andisols and Inceptisols of this study by season, significant differences between soil orders only occurred in the dry season (Table 6.8). In contrast, wet versus dry season Kfs values were significantly different in both soil orders (Table 6.9). In Andisols, median Kfs values were <5 mm/hr falling into the slow and very slow infiltration categories. In contrast, in Inceptisols the difference in Kfs between seasons was more pronounced (Table 6.9). Moderate Kfs values in Inceptisols only occurred in the dry season, while Kfs values were classed as slow in the wet season. Significant differences in Kfs between seasons are also reported in the literature (Diamond and Shanley, 2003; Jejurkar and Rajukar, 2012) with antecedent soil moisture contributing to this seasonal effect. θgrav in the Andisol site was generally > 70% in contrast to θgrav < 70% in the Inceptisol site. Note that zero infiltration was measured in six of the sites in Andisols and three of the sites in Inceptisols during the wet season, indicating imperfect drainage particularly in the Sonora watershed. Conversely, true steady-state infiltration rate may not have been achieved at all Inceptisol sites during the dry season. These results reflect infiltration rates in the field and suggest that the use of seasonally measured Kfs values in hydrological models may increase our understanding of the factors and mechanisms affecting OF and water dynamics in Andean watersheds. Table 6.8 Median field saturated hydraulic conductivity (Kfs) and soil water content (θ) in Andisols and Inceptisols 1Kfs: field saturated hydraulic conductivity; 2θgrav: field gravimetric soil water content; θvol: calculated volumetric soil water content; 4Values in parenthesis are first and third quartile; Number of measurements were 12 for each soil order in each season; *, ** Significant differences between Andisols and Inceptisols with Mann Whitney U test at p < 0.05 and p < 0.01, respectively Andisols Inceptisols Andisols Inceptisols5 (3 - 24)431 (9 - 57)* 0.7 (0 - 6) 3 (0 - 11)84 (70 - 92) 24 (15 - 37)** 112 (94 - 139) 61 (57 - 73)**0.45 (0.38 - 0.49) 0.20 (0.13 - 0.32)** 0.60 (0.50 - 0.75) 0.53 (0.49 - 0.63)Dry season Wet seasonKfs1 (mm/hr)θgrav2 (g/g)θvol3 (cm3/cm3)113 Table 6.9 Median saturated hydraulic conductivity (Kfs), and soil water content (θ) by season in Andisols and Inceptisols 1 Kfs: field saturated hydraulic conductivity; 2 θgrav: field gravimetric soil water content; 3 θvol: calculated volumetric soil water content; 4 Values in parenthesis are first and third quartile; Number of measurements were 12 for each soil order in each season *, ** Significant differences between the dry and the wet season with Mann Whitney U test at p < 0.05 and p < 0.01, respectively High Kfs values (lines and crosses in Figure 6.2) were measured in both soil orders. These Kfs values > 50 mm/hr, may have been affected by site conditions. In the case of Andisols, as discussed in Chapter 2 (Section 2.1.3), high values may be the result of the random distribution of termites. Also, as discussed in Chapter 2 (Section 2.1.3), the high Kfs values of Inceptisols, may be the result of rocks associated with the deposition of materials from road construction and localized land subsidence. Regardless whether considering geographically isolated areas with high Kfs values, there were significantly higher Kfs values in Inceptisols relative to Andisols, and significantly higher Kfs values in the dry season relative to the wet season in both soil orders, highlighting the importance of measuring infiltration seasonally. See Appendix H, Tables H1 and H2 for the differences between soil orders and seasons with extreme values excluded. High Kfs values indicate the inherent soil variability encountered in the field, and thus should be considered when determing Kfs values for hydrological or other models. The variability of Kfs spatially and temporaly highlights the importance of field measurements which incorporate a larger surface area, to represent field conditions. The double ring infiltrometer, due to the large soil contact area, incorporates macro-pores and is Dry season Wet season Dry season Wet season5 (3 - 24)40.7 (0 - 6)* 31 (9 - 57) 3 (0 - 11)*84 (70 - 92) 112 (94 - 139)** 24 (15 - 37) 61 (57 - 73)**0.45 (0.38 - 0.49) 0.60 (0.50 - 0.75)** 0.20 (0.13 - 0.32) 0.53 (0.49 - 0.63)**Kfs1 (mm/hr)θgrav2 (g/g)θvol3 (cm3/cm3)Andisols Inceptisols114 thus representative of field conditions (Davis et al., 1999). This method; however, is labour intensive and requires time to obtain reliable measurements (Diamond and Shanley, 2003). 6.3.2 Estimation of overland flow Overland flow (OF) is largely regulated by the soil infiltration rate, rainfall intensity, antecedent soil water content and slope (Bonell, 1993). High OF may lead to floods in lowlands and to water scarcity in upper and mid-watersheds, as water is rapidly lost from watersheds. If OF is low, rain water can remain in the watershed, be stored in the soils, drain slowly to streams by gravity, be taken up by plants, or contribute to groundwater recharge. Thus, retaining water within soils is a key factor for communities facing water scarcity, climatic variability and population growth, especially in the Colombian Andes, where it is projected that 70% of urban municipalities will be affected by water scarcity by 2025 (IDEAM, 2000). The precipitation regime of the two watersheds is described in Table 6.10. In both watersheds, the majority of the rainfall events are of low intensity (RI10< 10 mm/hr) represented by 87% of the rainy days in the Andisol site and 93% of the rainy days in the Inceptisol site. These low intensity events accounted for 43% and 59% of total precipitation (TP) during the monitored years in Andisol and Inceptisol sites, respectively. Despite this similarity in the distribution of rainfall intensity, there is a large difference in the total precipitation between the sites. During the two years evaluated, precipitation in the Andisol site was 2.6 times greater than that of the Inceptisol site, and the number of days with precipitation was double at the Andisol site. 115 Table 6.10 Rainfall intensity classes for Sonora a) and El Chocho b) watersheds a) Sonora watershed (Andisol site) b) El Chocho watershed (Inceptisol site) The majority of precipitation in both watersheds has a low RI10 (<10 mm/hr). Figure 6.3 graphically compares Kfs to RI10 and indicates the amount of precipitation that may be partitioned to OF. Arrows represent the Kfs values in the dry and wet seasons for both soil orders, and precipitation is shown by the blue bars. Precipitation categories to the right of the green arrows (Fig. 6.3) are more likely to be partitioned to OF, as RI10 is higher than Kfs. Thus, more rainfall events in the wet season of both soils will be partitioned into OF. Sonora (Andisols site)Rain intensity class (mm/hr)No. Of days% days with rainPrecipitation (mm)% of total precipitation0-9.99 42.3 86.8 2698 4310-19.99 3.6 7.3 1224 1920-29.99 1.4 2.9 810 1330-29.99 0.7 1.5 609 1040-39.99 0.3 0.6 293 550-39.99 0.2 0.4 282 460-49.99 0.1 0.3 205 370-69.99 <0.1 0.1 86 180-79.99 <0.1 <0.1 41 190-99.99 <0.1 0.1 63 1100-109.99 <0.1 <0.1 18 <1110-119.99 - 0.0 - 0120-129.99 - 0.0 - 0Total 48.8 100 6331 100El Chocho (Inceptisols site)Rain intensity class (mm/hr)No. Of days% days with rainPrecipitation (mm)% of total precipitation0-9.99 24.7 93.3 1384 5910-19.99 1.0 3.8 337 1420-29.99 0.4 1.6 237 1030-29.99 0.2 0.7 139 640-39.99 <0.1 0.2 66 350-39.99 <0.1 0.2 53 260-49.99 <0.1 0.1 44 270-69.99 <0.1 0.1 37 280-79.99 <0.1 <0.1 14 190-99.99 <0.1 0.1 32 1100-109.99 - 0.0 - 0110-119.99 - 0.0 - 0120-129.99 <0.1 <0.1 20 <0.1Total 26.5 1 2365 100116 In the El Chocho watershed (Inceptisols), Figure 6.3b suggests that during the dry season when Kfs> 30 mm/hr, OF would be reduced. This graphical representation, shows a higher chance of occurrence of OF in Andisols in both seasons. In Inceptisols, OF may also occur in the wet season, but may be lower in the dry season since the infiltration rate in this season is faster. a) Sonora watershed (Andisol site) b) El Chocho watershed (Inceptisol site) Arrows represent Kfs values for pasture Figure 6.3 Number of days and total precipitation (TP) for each rain intensity (RI) class in: a) Andisol site; and b) Inceptisol site Wet season Wet and dry season Dry season 117 The ratio OF/TP was used to estimate the percentage of precipitation that results in OF. The results, given in Table 6.11, suggest that 75% of total precipitation in the Andisols site (̴ 2,400 mm/year) would result in OF, in comparison to only 30% in the Inceptisol site (̴ 350 mm/year). Note that in the wet season OF/TP was higher; around 89% and 36% in Andisol and Inceptisol sites, respectively, in comparison to 50% and 15% in the dry season. Suescún et al. (2017) measured OF in the Central region of the Colombian Andes and found values OF/TP < 3% in pasture on slopes of 18-22%, while Molina et al. (2007) in Ecuador found runoff coefficients (OF/simulated rainfall) between 36 and 98% under a land use (abandoned land) similar to pasture with slopes of 15-25%. These opposing results suggest that direct OF measurements are necessary to confirm the estimations of this study. However, during field measurements in the wet season in the Andisols site, runoff was observed during rainfall events which provides some confidence for the OF estimations at this site. In the Andisol site, high OF/TP is a concern for community water providers, as they are interested in conserving or increasing baseflows in the local streams, which are reliant on GW from the soil. In order to increase Kfs, reforestation and/or natural forest regeneration of pasture sites is recommended. The literature suggests that forest coverage has consistently higher Kfs than non-forest lands (Zimmerman et al., 2006, 2010; Germer et al., 2010). In addition, there is increasing evidence that soil infiltrability increases with time during natural forest regrowth (Deuchars et al., 1999; Ziegler et al., 2004; Zimmerman et al., 2006; Hassler et al., 2011) and with reforestation (Gilmour et al., 1987; Ilstedt et al., 2007). 118 Table 6.11 Estimation of overland flow (OF) over total precipitation in: a) the Andisol site; and b) Inceptisol site 1 Total rainfall was estimated based on the monitored rainfall as explained in Appendix GYear Season Months Monitored rainfall (mm)% monitored timeTotal rainfall TP1 (mm)Estimated OF(mm)OF/TP (%)Wet 1 Oct-Dec/12 1021 100 1021 912 89Year 1 Dry 1 Jan-Feb/13 569 100 569 302 532012-2013 Wet 2 Mar-May/13 1007 100 1007 893 89Dry 2 Jun-Sep/13 611 100 611 295 48Year 1 2012-2013 3208 3208 2403 75Wet 1 Oct-Dec/13 1209 99 1209 1077 89Year 2 Dry 1 Jan-Feb/14 511 100 511 290 572013-2014 Wet 2 Mar-May/14 867 100 867 751 87Dry 2 Jun-Sep/14 536 100 536 248 46Year 2 2013-2014 3123 3123 2365 76Sonora watershed - AndisolsYear Season Months Monitored rainfall (mm)% monitored timeTotal rainfall TP1 (mm)Estimated OF(mm)OF/TP (%)Dry 1 Aug-Sep/11 135 100 135 15 11Year 1 Wet 1 Oct-Dec/11 507 100 507 173 342011-2012 Dry 2 Jan-Feb/12 181 75 212 27 13Wet 2 Mar-May/12 422 84 464 169 36Dry 3 Jun-Jul/12 110 94 110 8 7Year 1 2011-2012 1355 1428 393 29Dry 1 Aug-Sep/12 156 100 156 31 20Year 2 Wet 1 Oct-Dec/12 226 100 226 88 392012-2013 Dry 2 Jan-Feb/13 173 60 201 25 13Wet 2 Mar-May/13 392 92 414 151 36Dry 3 Jun-Jul/13 64 100 64 7 12Year 2 2012-2013 1010 1061 302 30El Chocho watershed - Inceptisolsb) a) 119 6.4 Conclusions Significant differences were found in quasi steady-state infiltration rate in each soil type between the wet and the dry seasons, highlighting the importance of field measurements of Kfs and soil water content, in order to obtain data that represent temporal variations. Kfs under pasture in Andisols was classified in the very slow and slow categories, in the wet and in the dry seasons, respectively; while in Inceptisols, Kfs classified in slow and moderate categories, in the wet and in the dry seasons. The low Kfs in Andisols may be the result of soil properties and conditions within the Sonora watershed, specifically: high micro-pore volume, high soil water content even in the dry season, high precipitation in the watershed, and the presence of thick layers of volcanic ash and placic layers at depth that impede infiltration. Higher Kfs values in Inceptisols than in Andisols may be explained by lower precipitation in the El Chocho watershed that contributes to a range in θgrav from 20% to 85%, and a significant correlation between Kfs and θ. In Andisols of this study, spatial variability in Kfs was most likely related to termites, while in Inceptisols, land subsidence and rocks depositions associated with road construction likely contributed to variability. A map of Kfs values related to site conditions could be created, to identify regions with similar Kfs. Data separated by soil type, season and site conditions could then be used as input values in hydrological modelling to more accurately represent field conditions. Preliminary estimations of OF/TP during the two years of this study were found to be higher in the wet season in both Andisols and Inceptisols. Over 70% of rainfall was estimated to be partitioned into OF in the Sonora watershed (Andisols), indicating that management practices to increase infiltration should be implemented. 120 7. GENERAL DISCUSSION AND CONCLUSIONS 7.1 Overview Within a watershed context at mid-elevation (1,700-2,300 m) in the Colombian Andes, the goal of this research was to assess soil properties of the two most common soil types, as they affect soil water dynamics. The studied Andisols are located in the Central mountain range and Inceptisols are located in the Western mountain range. The colloidal fraction of the two soils are dominated by high surface area minerals and organic matter, and have similar soil water retention (SWR) characteristics even though the composition of the colloidal fraction is different. The specific objectives of the research were to: • Determine the mineralogy of the soils and relate these properties to water retention characteristics • Determine the soil water retention (SWR) characteristics of the A and B horizons • Compare the effects of natural forest and pasture land uses on SWR characteristics, and • Determine and compare field saturated hydraulic conductivity (Kfs) during dry and wet seasons under pasture, as it was not possible to compare to natural forest. The soils of this study, are located in different mountain ranges of the Colombian Andes, with different climate and parent materials, but are located in watersheds with similar elevation (1,700-2,300 m) and land uses (natural forest and pasture). Both watersheds drain to the Cauca River which flows north to the Atlantic Ocean, and both are the source of water for peri-urban communities. The two selected watersheds the Sonora and the El Chocho, are dominanted by Andisol and Inceptisol soils, respectively. 7.2 Summary Andisols and Inceptisols of this study have large total porosity, GW, θFC and θPWP in comparison to values reported for clay soils in the literature. Andisols in the Sonora watershed, even though they have a loamy texture, displayed high SWR. In contrast, the 121 Inceptisols of the El Chocho watershed were clay texture, yet SWR values in both A and B horizons of Andisols were greater than values for Inceptisols at every tension. Both soils however, have similar values of PAW and GW. Despite having different parent materials and climatic conditions, the soils of this study share the presence of colloids with high specific surface area, that increase their ability to retain water. These colloids contribute to the high hygroscopic water content of the two soils, aggregate formation and a wide range in pore size distribution. The inorganic-colloidal fraction is dominated by allophane, imogolite and organo-metallic compounds in Andisols, and ferrihydrite and Al/ Fe oxides in Inceptisols. Land use effects on SWR characteristics in Andisols and Inceptisols appear to be minimal. Although SOC was significantly lower under pasture relative to natural forest in the A horizon of both soils, significant differences in SWR were found only for PWP of Andisols. The similarity in SWR between natural forest and pasture may reflect the cumulative and integrative effects of SOC and SRO on SWR. Significant differences were found in Kfs in each soil between the wet and the dry seasons, highlighting the importance of seasonal, in situ field measurements of Kfs. Quasi steady-state infiltration or Kfs under pasture in Andisols was classified in the very slow (<1 mm/hr) and slow categories (5 mm/hr), in the wet and in the dry seasons, respectively. While in Inceptisols, Kfs was classified in the slow (3 mm/hr) and moderate categories (31 mm/hr), in the wet and in the dry seasons. The low Kfs in Andisols throughout the year was attributed to impermeable layers at depth, specifically volcanic ash and/or placic horizons. Higher values in Inceptisols were attributed to deep soils, and a pronounced seasonal effect on measured infiltration rates, reflected by a negative correlation between Kfs and antecedent soil moisture content. Preliminary estimations of overland flow (OF), determined from the ratio of OF to total precipitation, suggest that up to 88% of precipitation in the Sonora watershed (Andisol site) may be partitioned to OF in the wet season. In contrast, only 28% of precipitation is expected to form runoff in the El Chocho watershed (Inceptisol site). These results highlight the high OF in the Andisol site, and the need for a shift in management practices to increase infiltration in the watershed. 122 7.3 Significance This study makes an important contribution to our understanding of both soil types in the Colombian Andes, as it included mineralogical properties in the evaluation of factors affecting water retention characteristics, in addition to physical and chemical properties. Analyses of these properties by horizon facilitates a better understanding of the factors influencing water retention. For example, in the case of the dominant soil group of the Sonora watershed, it was found that this allophanic Andisol, had high amounts of SRO minerals in the B horizon, and low amounts of organo-metallic compounds in the A horizon, but a significant amount of organic matter in the surface and B horizons. In contrast, in the Inceptisols of the El Chocho watershed, the water retention is affected by minerals such as oxides and hydroxides of Fe (goethite and ferrihydrite) and Al (boehmite). These minerals with large specific surface areas, despite being present at small concentrations, as in the case of ferrihydrite, contribute to the high volume of micro-pores and moderate values of plant available water storage (PAWS) and gravitational water (GW). The presence of ferriyhidrite in Inceptisols, helped to explain the similarities between the two soils in terms of their water retention characteristics. Short-range order minerals contribute in two different ways in relation to SWR characteristics: they stabilize SOC and they increase the apparent sand size fraction. In Andisols, SRO minerals and SOC may have a synergistic effect increasing SWR characteristics. The increase in the apparent sand size fraction may be a reflection of aggregate formation, which increases macro-pores and GW. Hydrological modellers working in the Andean region commonly utilize texture to estimate soil water retention and hydraulic conductivity. This research has shown that soil texture is a poor surrogate for these estimations in soils with SRO minerals such as the Andisols and Inceptisols of this study. The apparent sand size fraction reflects the presence of aggregates rather than disaggregated soil particles. SWR of both soils was high. And soil water movement, in the Andisol site in particular, was impacted by a restrictive layer at depth resulting in imperfect drainage. Thus, the use of texture alone to predict soil water dynamics would not accurately represent watershed conditions. 123 This study contributes to the knowledge on SWR characteristics and the relationships with soil properties at mid-elevation in the Andes as most studies reported in the literature have focused on high elevation páramo ecosystems. This information is important as a large proportion of the Andean population live at these elevations and communities living in peri-urban and urban zones are dependent on these ecosystems for their water supply and livelihoods. 7.4 Implications for land use management Both Andisols and Inceptisols have unique SWR characteristics: although they retain significant water in their micro-pores, they have moderate PAWS and macro-pore volume (GW). Soils under pasture had significantly lower SOC in both watersheds. Thus, for communities interested in source water protection, a recommendation is to increase SOM in order to enhance both the hydrological buffering capacity of the soil and PAWS. Higher SOC will not only contribute to source water protection, but also provides carbon storage within the watersheds. The addition of organic matter to soils with SRO minerals, may foster carbon stabilization, aggregate formation and increase SWR characteristics. In Andisols, where OF was high, reforestation is recommended. With reforestation Kfs may increase and OF decrease, as suggested by the literature. Despite an increase in evapotranspiration (ET) with reforestation, the reduced OF, could compensate for the higher ET, and may result in higher dry season base flows in local streams. In the El Chocho watershed (Inceptisol site), precipitation is significantly lower, <1,400 mm/yr, and irrigation rates should be carefully controlled to avoid OF. 7.5 Suggestions for future research To improve our understanding of the role of soils in supporting crop production and regulating stream flow in the Sonora and El Chocho watersheds, the next step could be a hydrological study that includes a water balance, estimates of base flow in the dry seasons, water yield evaluation and modelling of stream flow under different climatic scenarios. The data of this study provides input variables for soil properties, which combined with 124 hydrological monitoring and modelling would support local communities in their efforts to adapt to climatic variability. Further analysis of the differences between land uses is needed to better understand the effect of land use change on SOC and changes in pore size distribution. While this study found limited differences in SWR between natural forests and pasture, and suggests that SRO minerals may play a role in mitigating impacts, additional detail on pasture management practices, erosion and characteristics of the forest floor would further our understanding about the relationships between land use and SWR. Since SOC found in Andisols and Inceptisols of this study was relatively high (~ 8% in Andisols and ~ 5% in Inceptisols) and SOC lability may be low in soils with SRO minerals as proposed by Bruun et al. (2010), the role of these soils as carbon sinks is important in the context of climate change (Lal, 2013). Therefore, a study of lability, turnover, and residence time of SOC in these soils is of interest. As aggregate dispersion in Andisols is problematic, the Bouyucous method conducted on samples at field water content (as opposed to air dried) is recommended. These data will provide more reliable texture information that may correlate with SWR characteristics and would be interesting for comparison purposes with the standard Bouyucous test results on air dried samples on both soil orders. To fully investigate the difference in SWR of soils containing SRO minerals, analysis of specific surface area, could be a more suitable approach, than texture analysis. Few studies of Inceptisols in Colombia have been conducted. Thus, additional characterization of these soils, including SWR, parent material and mineralogy will provide a better understanding of the range in water retention and related soil characteristics found in these soils. In particular, further research on the effect of ferrihydrite in moderating changes in SWR in surface horizons of Inceptisols, would be of interest. Future research on Andisols and Inceptisols in the Andes under a range of climate and ecosystems will contribute to an improved understanding of the factors affecting SWR characteristics and to developing regionally relevant data. In addition, exploration of 125 suitable methods for measuring infiltration in natural forest on slopes is recommended, in order to further study the differences of soil water dynamics between land uses. The contribution of this study was to give a complete analysis of the soil characteristics, including chemical, physical and mineralogical properties that affect SWR characteristics, of two great groups of the two most common soil orders in the Colombian Andes, Andisols and Inceptisols, at mid-elevations. 126 REFERENCES Algoe, I.C.; Stoops, G.; Vandenberghe, R.E.; Van Ranst, E., 2012. Selective dissolution of Fe-Ti oxides- Extractable iron as a criterion for andic properties revisited. Catena, 92: 49-54. Arias, M.; Barral, M.T.; Diaz-Fierros, F.; 1996. Effects of associations between humic acids and iron or aluminium on the flocculation and aggregation of kaolin and quartz. Eur. J. Soil Sci. 47: 335-343. Armenteras, D.; Rodríguez, N., 2007. Introducción. In: Armenteras, D., Rodríguez, N. (Eds.), Monitoreo de los ecosistemas andinos 1985–2005: síntesis. Instituto de Investigación de Recursos Biológicos Alexander von Humboldt, 15–17. Armenteras, D.; Rodríguez, N.; Retana, J.; Morales, M., 2011. Understanding deforestation in montane and lowland forests of the Colombian Andes. Regional Environmental Change 11:693-705. Arnold, J.G.; Kiniri, J.R.; Srinivasan, R.; Williams, J.R.; Haney, E.B.; Neitch, S.L., 2012. Soil water assessment tool, Input/Output documentation. Texas Water Resources Institute. Retrieved April 19, 2018, from https://swat.tamu.edu/media/69296/SWAT-IO-Documentation-2012.pdf Asghari, Sh.; Ahmadnejad, S.; Keivan Behjou, F., 2016. Deforestation effects on soil quality and water retention curve parameters in Eastern Ardabil Iran. Eurasian Soil Science, 49 (3): 338- 346. Doi: 10.1134/S1064229316030029 Assouline, S., 2013. Infiltration into soils: Conceptual approaches and solutions. Water Resources Research, 49: 1755-1772. Doi:10.1002/wrcr.20155 Bagarello, V.; Di Prima, S.; Iovino, M.; Provenzano, G., 2014. Estimating field-saturated soil hydraulic conductivity by a simplified Beerkan infiltration experiment. Hydrological processes, 28: 1095- 1103. Doi: 10.1002/hyp.9649 127 Bakshi, S.; He, Z.L.; Harris, W.G., 2015. Natural Nanoparticles: Implications for Environment and Human Health, Critical Reviews in Environmental Science and Technology, 45:8, 861-904. Doi: 10.1080/10643389.2014.921975 Barral, M.T.; Arias, M.; Guérif, J. 1998. Effects of iron and organic matter on the porosity and structural stability of soil aggregates. Soil and Tillage Research, 46: 261-272. Blake, G.R., 1965. Bulk density. In: American Society of Agronomy, Soil Science Society of America (Ed.), Methods of Soil Analysis. Part 1. Physical and Mineralogical Properties, Including Statistics of Measurement and Sampling. Blake, G.R., 2008. Particle density. In: Ward Chesworth (Ed.), Encyclopedia of soil science. University of Guelph, Canada. Boix-Fayos, C.; Calvo-Cases, A.; Imeson, A.C.; Soriano-Soto, M.D., 2001. Influence of soil properties in the aggregation of some Mediterranean soils and the use of aggregate size and satability as land degradation indicators. Catena, 44:44-67. Bompoti, N.; Crysochoou, M.; Machesky, M., 2017. Surface structure of ferrihydrite; Insights from modelling surface charge. Chemical Geology, 464: 34 – 45. Doi: 10.1016/j.chemgeo.2016.12.018 Bonell, M., 1993. Progress in the understanding of runoff generation dynamics in forest. Journal of Hydrology, 150: 217-275. Bouwer, H., 1986. Intake rate: Cylinder Infiltrometer, pages 825-844 of Methods of Soil Analysis, Part I. Physical and Mineralogical Methods, Agronomy Monograph no. 9, second edition, Madison: American Society of Agronomy-Soil Science Society of America. Bouyucous, G.J., 1962. Hydrometer method improved for making particle size analysis of soils. Agron. Journal, 54: 464-465. Bradley, R.S.; Vuille, M.; Diaz, H.F.; Vergara, W., 2006. Threats to water supplies in the tropical Andes. Science, 312: 1755–1756. Doi: 10.1126/science.1128087 128 Broadbent, F. E.; Jackman, R. H.; McNicoll, J., 1964. Mineralization of C and N in some New Zealand allophanic soils. Soil Sci. 98:118-132. Bronick, C.J.; Lal, R., 2005. Soil structure and management: a review. Geoderma, 124: 3-22. Bruijnzeel, L.A., 1989. Deforestation and dry season flow in the tropics: A closer look. Journal of Tropical Forest Science, 1: 229-243. Bruijnzeel, L.A., 2004. Hydrological functions of tropical forests: not seeing the soil for the trees? Agriculture, Ecosystems and Environment, 104: 185–228. Bruun, T.B.; Elberling, B.; Christensen, B.T., 2010. Lability of soil organic carbon in tropical soils with different clay minerals. Soil Biol. Biochem. 42: 888-895. Buytaert, W.; Deckers, J.; Dercon, G.; De Bievre, D.; Poesen, J.; Govers, G., 2002. Impact of land use changes on the hydrological properties of volcanic ash soils in South Ecuador. Soil Use and Management, 18: 94-100. Buytaert, W.; Sevink, J.; Leeuw, B.D.; Deckers, J., 2005a. Clay mineralogy of the soils in the south Ecuadorian páramo region. Geoderma, 127: 114-129. Doi:10.1016/j.geoderma.2004.11.021 Buytaert, W.; Wyseure, G.; De Bievre, D.; Deckers, J., 2005b. The effect of land-use changes on the hydrological behaviour of Histic Andosols in south Ecuador. Hydrological Processes, 19: 3985-3997. Doi:10.1002/hyp.5867 Buytaert, W.; Deckers, J.; Wyseure, G., 2006. Description and classification of nonallophanic Andosols in south Ecuadorian alpine grasslands (páramo). Geomorphology 73: 207-221. Buytaert, W.; Iñiguez, V.; De Bievre, B., 2007. The effects of afforestation and cultivation on water yield in the Andean páramo. Forest Ecology and Management, 251: 22-30. Doi:10.1016/j.foreco.2007.06.035 129 Buytaert, W.; Cuesta-Camacho, F.; Tobón, C., 2011. Potential impacts of climate change on the environmental services of humid tropical alpine regions. Global Ecol Biogeogr, 20: 19-33. Doi: 10.1111/j.1466-8238.2010.00585.x Calder, I.R., 1999. The Blue Revolution. Earthscan Publications, London, p. 192. Calder, I.R., 2002. Forests and hydrological services: reconciling public and science perceptions. Land Use and Water Resources Research, 2: 2.1–2.12. Cardona Mejía, J.F., 2012. Módulo pastos y especies forrajeras, núcleos municipales de extensión y mejoramiento para pequeños ganaderos, Asistegán. Federación colombiana de ganaderos- Fedegan, Bogotá. Cardona, P.J., 2005a. Caracterización botánica de la cuenca alta del río Barbas. Cali, 2015. Cardona, P.J, 2015b. Caracterización botánica de la cuenca alta del río Aguacatal. Cali, 2015. Chaves, J.; Neill, C.; Germer, S.; Neto, S.G.; Krusche A.; Eisenbeer, H., 2008. Land management impacts on runoff sources in small Amazon watersheds. Hydrological processes, 22: 1766-1755. Doi: 10.1002/hyp.6803 Childs, C. W.; Masue, N.; Yoshinaga, N., 1991. Ferrihydrite in volcanic ash soils of Japan. Soil Sci Plant Nutr, 37: 299-311. Childs, C.W., 2007. Ferrihydrite: A review of structure, properties and occurrence in relation to soils. Zeitschrift fur Planzanernahrung und Bodenkunde. Chinchilla, M.; Mata, R.; Alvarado, A., 2011. Andisoles, Inceptisoles y Entisoles de la subcuenca del río Pirrís, Región de los Santos, Talamanca, Costa Rica. Agronomia Costarricense, 35: 83-87. Cortes, A.; Franzmeier D.P., 1972. Weathering of primary minerals in volcanic ash-derived soils of the central Cordillera of Colombia. Geoderma, 8: 165-176. 130 CRQ- Corporación Autonoma Regional del Quindio, 2015. Precipitación diaria de la estación Bremen, desde 1971 to 2007. Armenia, Quindio. CVC- Corporación Autonoma Regional del Valle del Cauca, 2014. Precipitación diaria de la estación Villaracelly, Cali, desde 1981 a 2011. Grupo de Recursos Hidricos. Cali, Colombia. Dahlgren, R.A., 1994. Quantification of allophane and imogolite, In: J.E. Amonette and L.W. Zelazny (Eds.), Quantitative methods in soil minerology. SSSA, Madison, WI, pp. 430-451. Dahlgren, R.A.; Saigusa, M.; Ugolini, F.C., 2004. The nature, properties and management of volcanic soils. Advances in Agronomy, 82: 113-182. DANE- Departamento Administrativo Nacional de Estadística, 2005. Cuadros Población total censada según departamentos y municipios, Bogotá. Retrieved October 10, 2011, from http://www.dane.gov.co/censo/ Davis, S.H.; Vertessy, R.A.; Silverstein, R.P., 1999. The sensitivity of a catchment model to soil hydraulic properties obtained by using different measurement techniques. Hydrological Processes, 13 (5): 677-688. Daza-Torres, M.C.; Hernandez-Florez, F.; Triana, F.A., 2014. Efecto del uso del suelo en la capacidad de almacenamiento hídrico en el Páramo de Sumapaz-Colombia. Rev. FAc. Nal. Agr. Medellín 67 (1): 7189-7200. De Groot, R.S.; Wilson, M.A.; Boumans, R.M., 2002. A typology for the classification, description and valuation of ecosystem functions, goods and services. Ecol Econ, 41: 393–408. Doi:10.1016/S0921-8009(02)00089-7 Deuchars, S.A.; Townend, J.; Aitkenhead, M.J.; FitzPatrick, E.A., 1999. Changes in soil structure and hydraulic properties in regenerating rain forest. Soil Use and Management, 15 (3): 183-187. 131 Diamond, J; Shanley, T., 2003. Infiltration rate assessment of some major soils. Irish Geography, 36 (1): 32-46. Diaz, B.E.; Paz, B.L., 2002. Evaluación del régimen de humedad del suelo bajo diferentes usos, en los páramos Las Animas y Piedra de León, Tesis pregrado, Popayán, 89 pp. Dixit, S.; Hering, J.G., 2003. Comparison of arsenic(V) and arsenic(III) sorption onto iron oxide minerals: implications for arsenic mobility. Environ. Sci. Technol., 37: 4182–4189. Dixon, J.B.; Weed, S.B., 1989. Minerals in soil environments. 2nd. Ed. Soil Science Society of America, Madison, WI. DNP- Departamento Nacional de Planeación, 2010. Bases del Plan Nacional de Desarrollo 2010 – 2014, Prosperidad para todos, Bogotá. Duiker, S.W.; Rhoton, F.E.; Torrent, J.; Smeck, N.E.; Lal, R., 2003. Iron (hydr) oxide crystallinity effects on soil aggregation. Soil Sci.Soc. Am. J., 67 (2): 606-611. Duque-Escobar, G., 2007. Aspectos geofísicos y amenazas naturales en los Andes de Colombia. Primer Congreso Internacional de Desempeño Humano en Altura, desafío de la población de los Andes. Manizales, Colombia. Retrieved May 24, 2016, from http://www.galeon.com/geomecanica/andes-col.pdf Egli, M.; Nater M.; Mirabella, A.; Raimondi, S.; Plötze, M.; Alioth, L., 2008. Clay minerals, oxyhydroxide formation, element leaching and humus development in volcanic soils. Geoderma, 143:101-114. Etter, A.; Van Wyngaarden, W., 2000. Patterns of landscape transformation in Colombia, with emphasis in the Andean region. Ambio, 29 (7): 432-439. Eusterhues, K.; Rumpel, C.; Kögel-Knabner, I., 2005. Organo-mineral associations in sandy acid forest soils: importance of specific surface area, iron oxides and micropores. Eur J Soil Sci. 56: 753-763. Faivre, D., 2016. Iron oxides: From Nature to Applications. John Wiley & Sons. 132 Feller, C.; Beare, M.H., 1997. Physical control of soil organic matter dynamics in the tropics. Geoderma, 79: 69-116. Forsyth, T., 1996. Science, myth and knowledge: testing Himalayan environmental degradation in Thailand. Geoforum 27: 375–392. García-Leoz, V.; Villegas, J.C.; Suecún, D.; Flórez, C.P.; Merino-Martín, L.; Betancur, T.; León, J.D., 2018. Land cover effects on wáter balance partitioning in the Colombian Andes: improved water availability in early stages of natural vegetation recovery. Reg. Environ. Change, 18: 1117-1129. Doi.org/10.1007/s10113-017-1249-7 Gentry, A. H., 1982. Patterns of Neotropical plant diversity. Evolutionary Biology, 15: 1-84. Germer, S.; Neill, A.; Krusche, A.; Elsenbeer, H., 2010. Influence of land-use change on near-surface hydrological processes: undisturbed forest to pasture. J. Hydrol. 380: 473–480. Gilmour, D.A.; Bonell, M.; Cassels, D.S., 1987. The effects of forestation on soil hydraulic properties in the middle hills of Nepal: a preliminary assessment. Mountain Research and Development, 7 (3): 239-249. Goldberg, S.; Davis, J.A.; Hem, J.D., 1996. The surface chemistry of aluminum oxides and hydroxides. In: The environmental chemistry of auminum. CRC Press. 2nd Edition. United States of America.pp 271- 288. Goldberg, S.; Lebron, I.; Seaman, J.C.; Suarez, D.L., 2012. Soil colloidal behaviour. In: Huang, P.M.; Li, Y.; Summer, M.E. (Eds.), Handbook of soil sciences: properties and processes. Ed Second Edition, Florida. Griffiths, E.; Burns, R.G., 1972. Interaction between phenolic substances and microbial polysaccharides in soil aggregation. Plant and Soil, 36: 599-612. Grossl, P.R.; Sparks, D.L., 1995. Evaluation of contaminant ion adsorption/desorption on goethite using pressure-jump relaxation kinetics. Geoderma, 67: 87–101. 133 Grunwald, S., 2012. Andisols. Retrieved June 25, 2012, from http://soils.ifas.ufl.edu/faculty/grunwald/teaching/eSoilScience/andisols.shtml Guarín, F.; Gorin, G.; Espinosa, A., 2004. A Pleistocene stacked succession of volcaniclastic mass flow in Central Colombia: the Quindío-Risaralda fan. In: Allard et al. (Eds.), Debris avalanche and debris flows in volcanic terrains. Origins, behaviors, and mitigation. Acta Vulcanologica, 16 (1-2): 109-124 Handreck, K.; Black, N., 2005. Growing media for ornamental plants and turf. Third edition, University of New South Wales, Sidney, Australia. Hassler, S.K.; Zimmerman, B.; van Breugel, M.; Hall, J.S.; Elsenbeer, H., 2011. Recovery of saturated hydraulic conductivity under secondary succession on former pasture in the humid tropics. Forest Ecology and Management, 261 (10): 1634-1642. Henao, T., 2001. Caracterización de algunos suelos derivados de cenizas volcánicas de la zona cafetera central Colombiana. In: Proyecto UTP-GTZ, Suelos del eje cafetero. Universidad Tecnologica de Pereira, Pereira, Colombia. Hendershot, W. H.; Lalande, H.; Duquette, M., 1993. Chapter 16: Soil reaction and exchangeable acidity. In: Carter, M. R. (Ed.), Soil Sampling and Methods of Analysis. Lewis Publishers, Boca Raton. Herrera, M.C.; Lizcano, A.; Santamarina, J.C., 2007. Colombian volcanic ash soils. Characterisation and Engineering Properties of Natural Soils, 3: 2385-2409. Doi:10.1201/NOE0415426916.ch19. Herron, N., 2001. A water balance approach to assessing the hydrologic buffering potential of alluvial fan. Water Resources Research, 37 (2): 341-351. Hincapié-Gómez, E.; Tobón-Marín, C., 2010. Caracterización de las propiedades hidrofísicas de los Andisoles en condiciones de ladera. Suelos Ecuatoriales, 40 (2): 156-169. 134 Hincapié-Gómez, E., 2011. Estudio y modelación del movimiento del agua en suelos volcánicos de ladera. Tesis PhD, Universidad Nacional de Colombia, Palmira, 199 p. Hodnett, M.G.; Tomasella, J., 2002. Marked differences between van Genuchten soil water-retention parameters for temperate and tropical soils: a new water-retention pedo-transfer functions developed for tropical soils. Geoderma 108: 155-180. Horel, A.; Toth, E.; Gelybó, G.; Kása, I.; Bakacsi, Z.; Karkas, C., 2015. Effects of land use and management on soil hydraulic properties. Open Geosci (1): 742 – 754. Doi: 10.1515/geo-2015-0053 Huang, P.M., 1991. Ionic factors affecting the formation of short-range orders aluminosilicates. Soil Sci. SOc. Am. J., 55: 1172-1180. IBM-International Business Machines Corporation, 2016. Statistical Package for the Social Sciences SPSS software v.24. Retrieved May 12, 2017, from http://www.ibm.com/analytics/us/en/technology/spss/ IDEAM- Instituto de Hidrología, Meteorología y Estudios Ambientales, 2000. Estudio Nacional del Agua, Bogotá, Colombia. IDEAM, 2010. Estudio Nacional del Agua, Bogotá, Colombia. IDEAM, 2015. Estudio Nacional del Agua, Bogotá, Colombia. IGAC- Instituto Geográfico Agustín Codazzi, 1996. Suelos Departamento del Quindío. CRQ, Armenia. IGAC, 2012. Estudio de los conflictos de uso del territorio colombiano. Imprenta Nacional de Colombia, Bogotá, Colombia. IGAC; CVC, 2004. Levantamiento de suelos y zonificación de tierras del Departamento de Valle del Cauca. Imprenta Nacional de Colombia, Bogotá, Colombia. 135 Ilstedt, U.; Malmer, A.; Verbeeten, E.; Murdiyarso, D., 2007. The effect of afforestation on wáter infiltration in the tropics: a systematic review and meta-analysis. Forest Ecology and Management, 251:45-51. Doi:10.1016/j.foreco.2007.06.014 INATEC-Instituto Nacional Tecnológico, 2016. Manual del protagonista, pastos y forrajes. Nicaragua. INDERENA-Instituto Nacional de los Recursos Naturales Renovables y del Ambiente; IGAC; CONIF-Corporación Nacional de Investigación Forestal, 1984. Mapa de Bosques de Colombia. Bogotá, Colombia. IPCC-Interngovernmental Panel on Climate Change, 2007. Climate change 2007: impacts, adaptation and vulnerability. Working group II contribution to the Fourth Assessment Report. Cambridge Press. Jaramillo, 2002. Los suelos derivados de los piroclastos de la secuencia “El Cedral” en el altiplano de San Félix, Departamento de Caldas: aspectos taxonómicos. Retrieved June 20, 2014, from http://www.unalmed.edu.co/~geosuelo/Taxon_andisoles.pdf Jejurkar, C.L.; Rajukar, M.P., 2012. Infiltration studies for varying land cover conditions. International Journal of Computational Engineering Research, 2 (1): 72 – 76. Juo, A.S.R.; Franzluebbers, K., 2003. Tropical Soils, Properties and Management for Sustainable Agriculture. Oxford, University Press, New York. Kaiser, K.; W. Zech., 1996. Defects in estimation of aluminum in humus complexes of podzolic soils by pyrophosphate extraction. Soil Sci., 161:452-458. Kaiser, M.; Walter, K.; Ellerbrock, R.H.; Sommer, M., 2011. Effects of land use and mineral characteristics on the organic carbon content, and the amount and composition of Na-pyrophosphate-soluble organic matter, in subsurface soils. Eur. J. Soil Sci., 62: 226-236. 136 Kämpf, N.; Scheinost, A.C.; Schulze, D.G., 2012. Oxide Minerals. In: Huang, P.M.; Li, Y.; Summer, M.E. (Eds.), Handbook of soil sciences: properties and processes. Ed Second Edition, Florida. Kettler, T.A.; Doran, J.W.; Gilbert, T.L., 2001. Simplified method for soil particle-size determination to accompany soil-quality analyses. Soil Sci. Soc. Am. J., 65: 849-852. Kimble, J.; Ping, C.; Sumner, M.; Wilding, L., 2000. Andisols. In: Sumner, ME (Ed.), Handbook of soil science. CRC Press, Boca Raton, 209–224. Klute, A., 1986. Water retention: Laboratory methods. In: Klute, A. (Ed.), Methods of soil analysis, Part 1: Physical and mineralogical methods, 2nd edition, Madosin, WI: Soil Science Society of America, American Society of Agronomy, pp. 635-662. Krammer, M.G.; Sanderman, J.; Chadwick, O.A; Chorover, J.; Vitousek, P.M, 2012. Global Change Biology, 18: 2594 - 2605. Labrière, N.; Locatelli, B.; Laumonier, Y.; Freycon, V.; Bernoux, M., 2015. Soil erosion in the humid tropics: a systematic quantitative review. Agric Ecosyst Environ 203:127–139. Doi:10.1016/j.agee.2015.01.027 Lal, R., 2007. Carbon management in agricultural soils. Mitigation and Adaptation Strategies for Global Change, 12: 303-322. Lal, R., 2013. Soil carbon management and climate change. Carbon Management, 4 (4): 439-462. Doi: 10.4155/cmt.13.31. Leitinger, G.; Tasser, E.; Newesely, C.; Obojes, N.; Tappeiner, U., 2010. Seasonal dynamics of surface runoff in mountain grassland ecosystems differing in land use. J Hydrol 385: 95–104. Doi:10.1016/j.jhydrol.2010.02.006 Li, X.G.; Li, F.M.; Zed, R.; Zhan, Z.Y.; Singh, B., 2007. Soil physical properties and their relations to organic carbon pools as affected by land use in an alpine pastureland. Geoderma, 139: 98-105. Doi:10.1016/j.geoderma.2007.01.006 137 Malagón Castro, D.; Pulido Roa, C.; Llinas Rivera, R.D.; Chamorro Bello, C.; Fernández Lamus, J., 1995. Suelos de Colombia, origen, evolución, clasificación, distribución y uso. Instituto Geográfico Agustín Codazzi IGAC. Ministerio de Hacienda y Crédito Público, Bogotá. Malagón Castro, D., 2003. Ensayo sobre tipología de suelos colombianos- Enfásis en génesis y aspectos ambientales. Rev. Acad. Colomb. Cienc., 27 (104): 319-341. Manceau, A., 2011. Critical evaluation of the revised akdalaite model for ferrihydrite. American Mineralogist, 94 (4): 521-533. Manning, B.A.; Fendorf, S.E.; Goldberg, S., 1998. Surface structures and stability of arsenic(III) on goethite: spectroscopic evidence for inner-sphere complexes. Environ. Sci. Technol., 32: 2383–2388. Matis, K.A.; Zouboulis, A.I.; Malamas, F.B.; Afonso, M.D.R.; Hudson, M.J., 1997. Flotation removal of As(V) onto goethite. Environ. Pollut., 97: 239–245. Mejía, G.; Hohnke, H.; White, J.L., 1968. Clay mineralogy of certain soils of Colombia. Soil Sci. Soc. Amer. Proc., Vol 32. Mizota, C.; van Reeuwijk, L.P., 1989. Clay Mineralogy and Chemistry of Soils Formed in Volcanic Material in Diverse Climatic Regions. International Soil Reference and Information Centre, Wageningen. Molina, A.; Govers, G.; Vanacker, V.; Poesen, J.; Zeelmaekers, E.; Cisneros, F., 2007. Runoff generation in a degraded Andean ecosystem: interaction of vegetation cover and land use. Catena, 71: 357-370. Doi:10.1016/j.catena.2007.04.002 Morales Meneses, L.A., 2008. Evaluación de las propiedades físicas e hidráulicas del suelo bajo influencia de tres coberturas vegetales en Porce II, Antioquia, Colombia. Tesis de grado para optar por el título de Ingeniera Forestal. Universidad Nacional de Colombia, Facultad de Ciencias Agropecuarias, Medellin, Colombia. 138 Nanzyo, M.; Shoji, S.; Dahlgreb, R., 1993. Physical characteristics of volcanic ash soils. In: Shoji, S., Nanzyo, M., Dahlgren, R. (Eds.), Volcanic Ash Soils: Genesis, Properties and Utilization. Elsevier Science Publishers B.V., Amsterdam, p. 288. Nanzyo, M., 2002. Unique properties of volcanic ash soils. Global Environmental Research, 6: 99-112. Navarrete, D.; Sitch, S.; Aragao, L.E.O.C.; Pedroni, L., 2016. Conversion from forests to pastures in the Colombian Amazon leads to contrasting soil carbon dynamics depending on land management practices. Global Change Biology, 22: 3503-3517. Doi:10.1111/gbc.13266. NCSS- National Cooperative Soil Survey, 1996. National Cooperative Soil Characterization Database. National Cooperative Soil Survey, Lincoln, NE. Retrieved April, 18, 2018, from https://www.nrcs.usda.gov/wps/portal/nrcs/detail/soils/survey/?cid=nrcs142p2_053573 Neris, J.; Jiménez, C.; Fuentes, J.; Morillas, G.; Tejedor, M., 2012. Vegetation and land-use effects on soil properties and water infiltration of Andisols in Tenerife (Canary Islands, Spain). Catena, 98: 55-62. Nimmo, J.R.; Schmidt, K.M.; Perkis, K.S.; Stock, J.D., 2009. Rapid measurement of field-saturated hydraulic conductivity for areal characterization. Vadose zone, 8 (1): 142- 149. Doi:10.2136/vzj2007.0159 Nivia, A., 2014. Personal communication, 28 May 2014 in Servicio Geológico Colombiano, Cali-Colombia. Oades, J.M., 1984. Soil organic matter and structural stability: mechanisms and implications for management. Plant Soil, 76: 319-337. Doi:10.1007/BF02205590 O´Geen, A.T., 2013. Soil water dynamics. Nature Education, 4 (5): 9. Pardo Gomez, R.; Rodríguez Lopez, Y., 2014. Clasificación de tormentas tropicales según lluvias asociadas. Ingeniería Hidráulica y Ambiental, 35 (2), 35-51. 139 Parfitt, R.L.; Saigusa, M., 1985. Allophane and humus-aluminum in Spodosols and Andepts formed from the same volcanic ash beds in New Zeland. Soil Sci., 139:149-155. Parfitt, R.L.; Childs. C.W., 1988. Estimation of forms of Fe and Al: A review, and analyses of contrasting soils by dissolution and Moessbauer methods. Aust. J. Soil. Res., 121-144. Parfitt, R.L., 2008. Allophane and imogolite: role in soil biogeochemical processes. Clay Minerals, 44: 135-155. Doi: 10.1180/claymin.2009.044.1.135 Pasolac, 1999. Guía de Conservación de suelos y agua. Programa para la agricultura sostenible en laderas en America Central. Pasolac, San Salvador, El Salvador. Perrin, J.L.; Bouvier, C.; Janeau, J.L.; Menez, G.; Cruz, F., 2001. Rainfall/runoff processes in a small peri-urban catchment in the Andes mountains. The Rumihurcu Quebrada, Quito (Ecuador). Hydrological Processes 15: 843–854. Pirastru, M.; Castellini, M.; Giadrossich, F.; Niedda, M., 2013. Comparing the hydraulic properties of forested and grassed soils on an experimental hillslope in a Mediterranean environment. Procedia Enviromental Sciences, 19. 341 – 350. Doi: 10.1016/j.proenv.2013.06.039 PNNC- Parques Nacionales Naturales de Colombia, Dirección Territorial Suroccidente, 2005. Plan de Manejo 2005-2009. Parque Nacional Natural Farallones de Cali, Cali, Colombia. Podwojewski, P.; Poulenard, J.; Zambrana, T.; Hofstede, R., 2002. Overgrazing effects on vegetation cover and properties of volcanic ash soil in the paramo of Llangahua and La Esperanza (Tungurahua, Ecuador). Soil Use and Management, 18: 45-55. Quintero, M.; Wunder, S.; Estrada, R.D., 2009. For services rendered? Modelling hydrology and livelihoods in Andean payments for environmental services schemes. Forest Ecology and Management, 258: 1871-1880. Doi: 10.1016/j.foreco.2009.04.032. 140 Rahman, M.H.; Deurer, M.; Holmes, A.W.; Saunders, S.J.; Mowat, A.; Clothier, B.E., 2011. Comparison of three methods to estimate organic carbon in allophanic soils in New Zeland. 24th Annual FLRC Workshop, Massey University, Palmerston North, New Zeland. Rawls, W.J.; Brakensiek, D.L.; Saxton, K.E., 1982. Estimation of soil water properties. Transactions of the ASAE, 25: 1316-1320, 1328. Rawls, W.J.; Pachepsky, Y.A.; Ritchie, J.C.; Sobecki, T.M.; Bloodworth, H., 2003. Effect of soil organic carbon on soil water retention. Geoderma, 116: 61-76. Doi:10.1016/S0016-7061(03)00094-6. Rawls, W.J.; Nemes, A.; Pachepsky, Y., 2004. Effect of soil organic carbon on soil hydraulic properties. In: Pachepsky, Y; Rawls, W.J., Developments in soil science, Elsevier, 30: 95-114. New York. Regelink, I.C.; Weng, L.; Koopmans, G.F.; Van Riemsdijk, W.H., 2013. Asymetric flow field-flow fractionation as a new approach to analyse iron-(hydr) oxide nanoparticles in soil extracts. Geoderma, 202: 134-141. Regelink, I.C.; Stoof, C.R.; Rousseva, S.; Weng, L.; Lair, G.J.; Kram, P.; Nikolaidis, N.P.; Kercheva, M.; Banwart, S.; Comans R.N.J., 2015. Linkages between aggregate formation, porosity and soil chemical properties. Geoderma, 247-248: 24-37. Roa-García, M.C., 2009. Wetland and water dynamic in small tropical headwater catchments of the Andes. Doctoral dissertation, University of British Columbia. Roa-García, M.C.; Brown, S.; Schreier, H.; Lavkulich, L.M., 2011. The role of land use and soils in regulating water flow in small headwaters catchments of the Andes. Water Resources Research, 47: 1-12. Doi:10.1029/2010WR009582 Roa-García, M.C.; Brown, S., 2014. Final report: Adapting to climate change in rural Colombia: The role of water governance (IDRC grant number: 106344-001). Retrieved March 14, 2018, from https://idl-bnc-idrc.dspacedirect.org/bitstream/handle/10625/53381/IDL-53381.pdf?sequence=1&isAllowed=y 141 Rodríguez, N.; Armenteras, D.; Morales, M.; Romero, M., 2006. Ecosistemas de los Andes colombianos. Segunda edición. Instituto de Investigación de Recursos Biológicos Alexander von Humboldt. Bogotá, Colombia, 154 p. Saigusa, M.; S. Shoji; Nakaminami H., 1987. Measurement of water retention at 15 bar tension by pressure membrane method and available moisture of Andosols. Japanese Journal of Soil Science and Plant Nutrition, 58: 374–377. Sanchez, L.F., 2017. Personal communication, 6 June 2017 in Junta Administradora de Acueducto y Alcantarillado de la Vereda Alto Dapa, Acualtodapa, Yumbo-Colombia. Sei, J.; Jumas, J.C.; Oliver-Fourcade, J.; Quiquampoix, H.; Staunton, S., 2002. Role of iron oxides in the phosphate adsorption properties of kaolinites from the ivory coast. Clay Clay Miner., 50 (2): 217-222. SGC-Servicio Geológico Colombiano, 1984a. Mapa geológico regional de la plancha 224 Pereira. Regional Medellin. V1.0. Retrieved June 2, 2014, from http://aplicaciones1.sgc.gov.co/sicat/html/ConsultaBasica.aspx SGC, 1984b. Mapa geológico regional de la plancha 299 Ríos Jamundi, Claro, Timba. Retrieved June 2, 2014, from http://aplicaciones1.sgc.gov.co/sicat/html/ConsultaBasica.aspx Shannon, C.E., 1948. A mathematical theory of communication. Bell System Technical Journal, 27: 379-423. Shoji, S.; Fujiwara, Y., 1984. Active Al and Fe in the humus horizons of Andosols from northeastern Japan: their forms, properties, and significance in clay weathering. Soil Science, 137: 216-226. Shoji, S.; Nanzyo, M.; Dahlgren R.A., 1993. Volcanic ash soils - genesis, properties and utilization. Elsevier, Amsterdam. Singh, D.B.; Prasad, G.; Rupainwar, D.C., 1996. Adsorption technique for the treatment of As(V)-rich effluents. Colloid Surf. A, 111: 49–56. 142 Singh, V.P; Yadava, R.N., 2003. Water and Environment. Wastewater treatment and waste management. Allied Publishers Pvt. Lted. India. Soil Survey Staff, 2014. Keys to Soil Taxonomy, 12th ed, United Stated Department of Agriculture -USDA, Natural Resources Conservation Service-NRSC, Washington, DC. SSSA- Soil Science Society of America, 2001. Glossary of soil science terms. Soil Science Society of America, Madison, WI, United States. Suescún, D.; Villegas, J.C.; León, J.D.; Flórez, C.P.; García-Leoz, V.; Correa-Londoño, G.A., 2017. Reg Environ Change, 17: 827 – 839. Doi: 10.1007/s10113-016-1071-7 Tejedor, M.; Neris, J.; Jiménez, C., 2012. Soil properties controlling infiltration in volcanic soils (Tenerife, Spain). Soil Sci. Soc. Am. J., 77: 202-212. Doi:10.2136/sssaj2012.0132 Thorez, J., 1976. Practical identification of clay minerals; A handbook for Teachers and students in clay mineralogy. Institute of Mineralogy, Liege State University, Belgium. Tobón, C.; Bruijnzeel, L.A.; Frumau, K.F.A.; Clavo-Alvarado, J.C., 2010. Changes in soil physical properties after conversion of tropical montane cloud forest to pasture in northern Costa Rica. In: L.A. Bruijnzeel, Scatena, F.N., and Hamilton, L.S. (Eds.), Tropical Montane Cloud Forests: Science for Conservation and Management, Cambridge University Press. 502-515. Toohey, R.C.; Boll, J.; Brooks, E.S.; Jones, J., 2018. Effects of land use on soil properties and hydrological processes at the point, plot, and catchment scale in volcanic soils near Turrialba, Costa Rica. Geoderma, 315: 138-148. Ugolini, F.C.; Dahlgren, R.A., 1991. Weathering environments and occurrence of Imogolite /Allophane in selected Andisols and Spodosols. Soil Sci. Soc. Am. J., 55:1166-1171. Uribe, N.; Quintero, M.; Valencia, J., 2013. Aplicación del modelo hidrológico SWAT (Soil and water assessment tool) a la cuenca del río Cañete. Retrieved March 17, 2018, from 143 ftp://ftp.ciat.cgiar.org/DAPA/users/jvalencia/Modelo_SWAT_Ca%C3%B1ete/Segundo%20Informe_SWAT%202013.pdf Urrutia, R.; Vuille, M., 2009. Climate change projections for the tropical Andes using regional climate model: Temperature and precipitation simulations for the end of the 21st century. J. Geophys. Res., 114:1-15. Doi:10.1.1029/2008JD011021. Van Breemen, N.; Wielemaker W.G., 1974. Buffer intensities and equilibrium pH of minerals and soils: I. The contribution of minerals and aqueous carbonate to pH buffering. Soil Sci. Soc. Am. Proc., 38: 55-60. Van Noordwijk, M.; Cerri, C.; Woomer, P.L.; Nugroho, K.; Bernoux, M., 1997. Soil carbon dynamics in the humid tropical forest zone. Geoderma, 79: 187-225. Van Wambeke, A., 1992. Soils of the Tropics: Properties and Appraisal. Mc Graw Hill, Michigan, USA. Vuille, M.; Francou, B.; Wagnon, P.; Juen, I.; Kaser, G.; Mark, B.G.; Bradley, R.S., 2008. Climate change and tropical Andean glaciers: past, present and future. Earth-Science Reviews, 89: 79–96. Doi:10.1016/j.earscirev.2008.04.002 Vuille, M.; Franquist, E.; Garreaud, W.S.; Casimiro, L.; Cáceres, B., 2015. Impact of the global warming hiatus on Andean temperature. J. Geophys. Res. Atmos., 120: 3745-3757. Doi:10.1002/2015JD023126. Wada, K., 1985. The distinctive properties of Andosols. In: Advances in Soil Sciences Volume 2, Springer-Verlag, New York pp 174-223. Warkentin, B.P.; Maeda, T., 1980. Physical and mechanical characteristics of Andisols. In: Theng, B.K.G. (Ed.), Soils with Variable Charge. Offset Publications, Palmerston North, pp. 281-352. World Bank, 2012. Análisis de la gestión del riesgo de desastres en Colombia: un aporte para la construcción de políticas públicas. Bogotá. Retrieved April 15, 2018, from http://cedir.gestiondelriesgo.gov.co/dvd/archivospdf/5-GESTIONDELRIESGOWEB.pdf 144 Ziegler, A.D.; Giambelluca, T.W.; Tran, L.T.; Vana, T.T.; Nullet, M.A..; Fox, J.; Pinthong, J.; Maxwell, J.F.; Evett, S., 2004. Hydrological consequences of landscape fragmentation in mountainous northern Vietnam: evidence of accelerated overland flow generation. Journal of Hydrology, 287 (1-4): 124-146. Zimmerman, B.; Elsenbeer, H.; De Moraes, J.M., 2006. The influence of land-use changes on soil hydraulic properties: implications for runoff generation. Forest ecology and Management, 222 (1-3): 29-38. Zimmerman, B.; Elsenbeer, H., 2008. Spatial and temporal variability of soil saturated hydraulic conductivity in gradients of disturbance. Journal of Hydrology, 361: 78-95. Zimmerman, B.; Papritz, A.; Elsenbeer, H., 2010. Asymmetric response to disturbance and recovery: changes of soil permeability under forest-pasture-forest transitions. Geoderma 159: 209–215. 145 APPENDICES Appendix A. General characteristics of Andisols and Inceptisols The Andisol order is relatively a new soil order that was added in 1990 to the Soil Taxonomy (Soil Survey Staff, 2014). The radical “andi” comes from the Japanese word ando, meaning dark soil, which connotes the color being the typifying characteristic of volcanic ash soils (Van Wambeke, 1992). The vast majority of Andisols are formed from pyroclastic deposits (volcanic ejecta) such as ash, pumice, cinders and lava. Rapid cooling of the molten materials upon ejection prevents crystallization of minerals with long range order, and the resulting product is vitric material or volcanic glass, which are dominated by amorphous or short-range order (SRO) minerals (Grunwald, 2012). Term “short-range order (SRO) minerals” refers to a class of material that are noncrystalline or at best, poorly crystalline, although they consist of tightly bonded silicon, aluminum and oxygen atoms. The two principal clay minerals of this type are allophane and imogolite that are commonly formed from volcanic ash (often found in Andisols) and from organic matter and aluminum accumulation (found in Spodozols) (Huang, 1991). The molecular structure of allophane consists of incomplete 1:1 phyllosilicate layers that contain aluminum both in octahedral and tetrahedral positions with the formula Al2O3 (SiO2)1.3-2 (2.5-3) H2O. Defects produce hollow spherules of 3.5 to 5 nm in diameter (Van Wambeke, 1992). Imogolite Al2SiO3(OH4), is also a nanoparticle of tubular structure with an inner diameter of 1 nm and outer diameter of 2 nm (Nanzyo, 2002). The hollow spheres and tubes of allophane and imogolite of nanoparticle size (1-100 nm), result in microcoscopic pores that can store water within the structure (Wada, 1985). The form and size of these SRO minerals are related to their high water absorbing capacity and the high water content at high moisture tension (Buytaert et al., 2002). The high specific surface area of SRO minerals (Eusterhues et al., 2005) (Table A.1) and the presence of positive and negative charges (pH dependent charge), allow SRO minerals to bond organic compounds and participate in aggregate formation. Within the formed aggregates, soil organic carbon (SOC) is preserved from decomposition or microbial attack through physical protection (i.e., SOC inaccessibility and/or the existence of temporary 146 water saturated conditions) (Broadbent et al., 1964; Dahlgren et al., 2004; Egli et al., 2008; Parfitt, 2008). Therefore, it is not surprising that Andisols often have high organic matter content. The combination of SRO minerals, stable aggregates and high organic matter content, affects soil water characteristics in Andisols by enabling: • low soil bulk density (ρb), especially when higher SOC content is present in both non-allophanic soils (where organo-metalic compounds are predominant) and allophanic soils (where allophane and imogolite are predominant) (Nanzyo, 2002). Lower ρb indicate higher porosity, which in turn may increase soil water content. • wide range in pore size distribution. In mature Andisols, where SRO mineral concentrations are high, the proportion of macro-pores decreases and that of micro-pores increases. However, the proportion of macro-pores could remain around 10% (Nanzyo, 2002). • high permeability and rapid rainfall infiltration (Warkentin and Maeda, 1980; Shoji et al., 1993; Tejedor et al., 2012). Andisols may have the pseudo-sandy texture that is likely the product of aggregate formation (Shoji et al., 1993) when a complete dispersion of aggregates is virtually impossible because each of the inorganic colloids shows a different point of zero net charge (Shoji et al., 1993). Therefore, there is no available technique that accurately measures the texture of Andisols due to their strong soil aggregate structure (Shoji et al., 1993). Inceptisols are soils that are in the incipient (i.e., early) stages of formation toward mature soil orders but have not yet fully developed their diagnostic properties. Conceptually, Inceptisols are transitional soils with minimum or no appreciable development (Entisols) and to soils of various orders that have been accepted by pedologists as carrying the marks of well-defined soil-forming processes (Van Wambeke, 1992). In Inceptisols of the tropics it is common to find a cambic horizon, which is a noncemented subsurface horizon in which the marks of original rocks structures or thin bedding have been obliterated in at least one-half of their volume. Texture of Inceptisols is typically of very fine sand or finer (Van Wambeke, 1992). It is also common that one or more horizons are affected by 147 weathering of primary minerals, which releases free iron oxides and/or produces clays that give subsurface horizons stronger chromas or redder hues that the underlying horizons (Van Wambeke, 1992). Iron and aluminum oxides are produced from the weathering of primary and secondary silicate minerals. The iron and aluminum atoms are coordinated with oxygen atoms (the latter are associated with hydrogen ions to make hydroxyl groups). Some minerals such as gibbsite (an Al-oxide) and goethite (an Fe-oxide) consist of crystalline sheets, while ferrihydrite (an Fe-hydroxide 5Fe2O3 9H2O) is noncrystalline (Goldberg et al., 2012). Ferrihydrite structure is still been debated (Manceau, 2011; Faivre, 2016). Iron and aluminum oxides also play a significant role in soil water characteristics as they enhance aggregation. Films of gibbsite, goethite and hematite, and the amorphous hydroxide ferrihydrite coat and cement soil aggregates preventing their breakdown (Goldberg et al., 2012). Crystalline Fe-oxides and hydroxides are of less importance in enhancing aggregation than the amorphous oxides such as ferrihydrite (Duiker et al., 2003; Regelink et al., 2015) despite their small contribution to soil mass (<1%) (Regelink et al., 2015). Ferrihydrite dominates the surface area available for sorption of SOC, stabilizing it and forming organo-mineral assemblages (Kaiser et al., 2011; Regelink et al., 2013). In addition, ferrihydrite binds soil particles, silt and sand, which in association with soil organic matter creates secondary aggregates (Arias et al., 1996; Sei et al., 2002). This can be explained by smaller size of ferrihydrite particles (diameter <10 nm) and their high specific surface area (Table A.1). In terms of infiltration, Inceptisols tend to have intermediate values compared to other soil orders, influenced by a balanced set of soil properties such as a moderate level of aggregation and moderate structural stability, loam texture and moderate bulk density (Tejedor, et al., 2012). Soil organic matter content and composition affect soil structure and adsorption properties (Rawls et al., 2003). Organic matter helps stabilize soil structure and aggregates (Feller and Beare, 1997; Boix-Fayos et al., 2011), which in turn contribute to a significant total pore volume and a wide pore size distribution (Barral et al., 1998). Soil humic substances such 148 as humic acid, fulvic acid, and humin are nanoparticles (Bakshi et al., 2015) with high specific surface area (Table A.1). An increase in soil organic carbon in a fine textured soil results in an increase in water retention (Rawls et al., 2003). The stability of microaggregates is enhanced by multivalent cations, which act as bridges between organic colloids and clays, and by the binding action of polysaccharides, mainly mucilages produced by bacteria, but also by plant roots and fungal hyphae (Oades, 1984). 149 Appendix B. Characteristics of soil profiles of Andisols (great group Hapludands) and Inceptisols (great group Dystrudepts) included in this study Table B.1 General characteristics of soil profiles in Andisols watershed (great group Hapludands) 1 pH determination in samples with SOC>12%, liquid volume was doubled; 2SOC: Soil organic carbon; 3 texture samples analyzed with Kettler method. Note: Based on filed observations there was 0% coarse fragments (2 – 75 mm in diameter) in A, B and C horizons. Site Horizon Elevation (m)SlopeDepth (m)Air dried colorpH1H2OpHCaCl2SOC2 (%)Bulk density (kg/m3)Sand (%)Silt (%)Clay (%)TextureP1 A 2148 Slope 33% 0.0-0.2 10YR 4/2 5.0 4.0 12.0 617 32 26 43 ClayB 0.2-0.6 10YR 6/3 5.7 4.9 2.4 1005 53 29 18 Sandy loamC 0.6-1.1+ 10YR 7/2 5.8 4.9 0.0 1270 62 20 18 Sandy loamP2 A 2141 Flat <20% 0.0-0.2 10YR 4/2 5.1 4.2 10.2 572 48 29 23 Loam3B 0.2-0.9 10YR 6/3 5.7 4.9 2.6 821 59 32 9 Sandy loam3C 0.9-1.2+ 10YR 7/2 5.7 4.9 0.3 1312 69 22 9 Sandy loam3P3 A 2228 Slope 59% 0.0-0.2 10YR 5/4 5.4 4.5 6.2 655 47 31 22 LoamB 0.6-1.1 10YR 7/6 5.8 5.4 2.9 984 50 33 17 LoamP4 A 2222 Flat <20% 0.0-0.5 10YR 5/3 5.2 4.1 8.6 455 19 54 26 Silt loamB1 0.5-2.7 10YR 5/4 5.6 4.7 6.3 557 52 33 15 LoamB2 2.7-2.9 10YR 7/2 5.5 5.1 0.9 837 57 26 17 Sandy loamC 3.1+ 10YR 8/3 5.4 4.5 0.5 550 41 29 30 Clay loamP5 A 2235 Slope 37% 0.0-0.3 10YR 5/3 5.0 4.1 7.3 656 44 34 22 LoamB 0.5-1.5+ 10YR 6/4 5.5 4.8 3.5 510 53 30 17 Sandy loamP6 A 2230 Flat <20% 0.0-0.2 10YR 5/3 5.2 4.2 6.1 733 40 40 20 LoamB 0.8+ 10YR 6/4 5.7 5.2 3.8 654 43 38 20 LoamP7 A 2205 Flat <20% 0.0-0.2 10YR 4/3 5.4 4.3 7.3 477 50 26 24 Sandy clay loamB 0.3-1.5+ 10YR 5/4 5.5 4.8 4.4 598 45 33 22 LoamP8 A 2233 Slope 24% 0.0-0.2 10YR 5/3 4.9 4.2 8.3 566 42 33 25 LoamB 0.6-1.1+ 10YR 6/6 5.6 5.3 4.7 668 60 28 12 Sandy loamP9 A 2216 Slope 44% 0.0-0.2 10YR 3/4 4.8 3.8 12.9 414 26 35 39 Clay loamB 1.0-1.7 10YR 7/4 5.7 5.2 2.6 693 63 21 16 Sandy loamC 1.7-2.1+ 7.5YR 8/1 4.9 3.6 0.6 709 11 24 65 ClayP10 A 2151 Flat <20% 0.0-0.15 10YR 4/2 5.3 5.1 13.8 293 34 33 33 Clay loamB 0.25-0.40 10YR 6/4 5.3 4.6 3.3 784 55 30 15 Sandy loamP11 A 2179 Slope 29% 0.0-0.3 10YR 4/2 5.0 4.0 7.2 677 42 35 22 LoamB 0.3-1.2+ 10YR 6/3 5.7 5.1 4.9 606 60 25 15 Sandy loamP12 A 2144 Flat <20% 0.0-0.2 10YR 4/2 5.2 4.1 7.0 669 45 33 22 LoamB 02-1.5+ 10YR 5/4 5.7 5.0 5.9 558 63 23 15 Sandy loamC - - - 10YR 8/1 4.9 4.0 0.8 768 53 23 25 Sandy clay loamC - - - 10YR 8/1 5.2 4.3 0.2 613 39 14 47 ClayC - - - 10YR 8/1 4.9 3.9 0.0 965 2 13 85 ClayB1 A 2104 Slope 59% 0.0-0.3 10YR 3/2 4.5 3.5 15.6 969 68 11 21 ClayB 0.3-1.2 10YR 6/3 5.3 4.6 2.2 1181 48 38 15 LoamB2 A 2131 Slope 53% 0.0-0.3 10YR 3/4 4.6 3.7 11.0 511 46 24 30 Sandy clay loamB 0.3-1.2+ 10YR 6/6 5.5 5.1 3.8 586 63 23 15 Sandy loamB3 A 2217 Slope 47% 0.0-0.3 10YR 4/2 4.6 3.7 9.0 455 52 20 28 Sandy clay loamB 0.3-1.2 10YR 6/3 5.4 4.5 4.1 798 49 34 17 LoamC 1.2-1.7+ 10YR 5/4 5.5 5.1 4.7 765 59 28 14 Sandy loamB4 A 2153 Slope 88% 0.0-0.2 10YR 4/3 4.9 3.7 12.5 329 53 25 22 Sandy clay loamB 0.2-0.6 10YR 5/4 5.5 5.0 4.8 418 68 20 12 Sandy loamB5 B 2077 Slope 73% 0.55-1.2+ 10YR 6/4 5.7 4.9 3.8 667 57 28 15 Sandy loamB6 A 2129 Slope 34% 0.0-0.15 10YR 4/2 4.6 3.6 13.5 323 53 23 24 Sandy clay loamB 0.25-0.9+ 10YR 6/4 5.6 5.1 2.7 633 69 19 12 Sandy loamSonora watershed - AndisolsPastureNatural forest150 Table B.2 General characteristics of soil profiles in Inceptisols watershed (great group Dystrudepts) 1 pH determination in samples with SOC>12%, liquid volume was doubled; 2SOC: Soil organic carbon; 3 texture samples analyzed with Kettler method. Note: Based on field observations, there were 5% coarse fragments (2 – 75 mm in diameter) in A horizon, 10% in B horizon and 15% in C horizon Site Horizon Elevation (m)SlopeDepth (m)Air dried colorpH1H2OpHCaCl2SOC2 (%)Bulk density (kg/m3)Sand (%)Silt (%)Clay (%)TextureP1 A 1850 Slope 45% 0.0-0.3 10YR 3/4 5.6 4.8 8.3 1053 13 32 55 ClayB 0.3-1.0+ YR 4/6 5.4 4.6 6.4 1020 6 27 66 ClayC 0.95+ 7.5YR 5/8 5.4 4.1 1.5 1122 44 43 13 ClayP2 A 1798 Flat < 20% 0.0-0.3 10YR 4/4 5.4 4.4 6.0 333 14 28 58 ClayB 0.5-0.8 10YR 6/6 5.3 4.0 1.9 832 21 20 58 LoamP3 A 1861 Slope 40% 0.0-0.3 10YR 4/4 5.7 4.5 3.6 906 19 17 64 ClayB 0.3-0.9+ 5YR 4/6 5.6 4.1 1.4 1011 8 36 56 ClayP4 A 1830 Flat < 20% 0.0-0.4 10YR 4/3 5.9 4.9 5.0 916 25 48 27 Clay loamB 0.6-1.2+ 10YR 6/6 5.7 4.8 1.6 984 25 32 43 ClayP5 A 1912 Slope 47% 0.0-0.1 10YR 3/2 5.7 4.8 5.5 839 13 26 61 ClayB1 0.1-0.6 10YR 5/6 6.0 5.2 1.6 1096 23 54 23 Silt loamB2 0.6-1.1+ 10.5YR 5/6 5.3 4.2 1.2 909 15 47 38 Silt clay loamP6 A 1863 Flat < 20% 0.0-0.4 10YR 5/4 5.5 4.6 3.6 1039 20 22 58 ClayB 0.4-0.9+ 7.5YR 5/6 5.7 4.7 2.7 1025 20 21 58 ClayP7 A 1876 Slope 40% 0.0-0.15 10YR 4/3 5.6 4.7 7.3 799 11 39 50 Clay3B 0.4-0.6 7.5YR 5/6 6.0 5.1 3.9 1041 9 35 56 ClayC 0.75-1.0+ 10R 5/6 6.1 5.0 2.0 972 14 21 65 ClayP8 A 1844 Flat < 20% 0.0-0.3 5YR 5/4 5.7 4.8 3.7 936 12 44 44 Silty clayB 0.7-1.0 7.5YR 6/6 5.9 4.9 1.3 949 19 30 51 ClayP9 A 1851 Slope 47% 0.0-0.2 10YR 5/4 5.3 4.3 5.1 842 11 31 58 ClayC 0.4-2.5+ 10R 5/6 4.9 3.6 1.3 959 9 35 55 ClayP10 A 1714 Flat < 20% 0.0-0.2 7.5YR 4/4 5.7 4.9 3.5 861 16 32 52 ClayB 0.5-0.6+ 7.5YR 5/6 5.4 4.4 1.6 1075 19 35 46 ClayP11 A 1891 Slope 47% 0-0.1 10R 3/2 5.4 4.1 4.4 893 12 13 76 ClayB 0.3-0.9+ 10R 5/8 5.2 3.7 1.4 1134 4 34 62 ClayP12 A 1844 Flat < 20% 0-0.2 10YR 4/2 5.5 4.5 5.0 931 26 21 53 ClayB 0.5-0.8+ 10YR 5/2 5.4 4.2 1.8 990 35 22 43 ClayB1 A 1932 Slope 34% 0.0-0.7 10R 3/1 5.5 4.6 6.5 807 47 23 31 Sandy clay loamB 0.75-0.9+ 10R 4/3 5.1 4.0 3.5 840 66 11 23 Sandy clay loamB2 A 1928 Slope 48% 0-0.05 10YR 3/1 5.7 4.9 14.9 388 33 15 52 ClayB 0.05-0.2 5YR 4/3 5.8 4.9 4.3 887 29 25 46 ClayC 0.2-0.55+ 5YR 5/4 5.3 4.1 2.9 805 64 13 23 Sandy clay loamB3 A 1959 Slope 48% 0-0.03 10R 3/1 6.0 5.1 9.0 590 9 43 48 Silty clay3B 0.06-0.80+ 10R 5/6 5.6 4.7 1.7 834 20 50 30 Clay loam3B4 A 1971 Slope 48% 0-0.1 5YR 3/3 5.8 5.1 7.7 721 12 32 56 Clay3B1 0.3-0.7 10R 6/8 4.8 3.4 1.4 947 23 46 31 Clay loam3B2 0.7-0.9+ 7.5YR 6/6 5.0 3.5 0.6 714 18 47 35 Silty clay loam3B5 A 1902 Slope 26% 0-0.35 10YR 5/2 5.5 4.4 3.8 964 16 56 28 Silty clay loamB 0.4-0.6+ 7.5YR 5/6 4.9 4.5 1.5 1041 17 62 22 Silt loamB6 A 1919 Slope 49% 0-0.17 10YR 4/2 5.4 4.3 5.2 733 11 65 24 Silt loamB 0.17-0.4 10YR 5/4 5.6 4.6 2.1 1117 19 28 53 ClayC 0.5-0.6+ 7.5YR 5/3 5.5 4.4 1.8 1137 59 15 25 Sandy clay loamEl Chocho watershed - InceptisolsPastureNatural forest151 Appendix C. Family, species and Importance Value Index (IVI) of vegetation found in natural forest of Sonora (Andisol site) and El Chocho (Inceptisol site) watersheds Table C.1 Family, species and Importance Value Index (IVI) of vegetation found in the Andisols watershed Family Species IndividualsRelative densityRelative frequencyRelative coverageIVICyatheaceae Cnemidaria horrida 37 0,14 0,06 0,17 0,37Arecaceae Wettinia kalbreyeri 27 0,10 0,07 0,11 0,29Cyatheaceae Alsophila cuspidata 13 0,05 0,06 0,07 0,18Melastomataceae Tibouchina lepidota 8 0,03 0,04 0,11 0,17Rubiaceae Guettarda crispiflora sabiceoides 16 0,06 0,04 0,07 0,17Arecaceae Geonoma undata 11 0,04 0,04 0,04 0,13Solanaceae Cuatresia riparia 19 0,07 0,02 0,03 0,13Rubiaceae Palicourea angustifolia 14 0,05 0,04 0,02 0,11Arecaceae Chamaedorea linearis 8 0,03 0,04 0,02 0,09Rubiaceae Guettarda hirsuta 8 0,03 0,02 0,02 0,07Melastomataceae Miconia sp.4 6 0,02 0,03 0,02 0,07Symplocaceae Symplocos quindiuensis 3 0,01 0,02 0,03 0,06Rubiaceae Elaeagia karstenii 5 0,02 0,03 0,01 0,06Rubiaceae Palicourea acetosoides 7 0,03 0,01 0,01 0,06Araliaceae Oreopanax floribundum 5 0,02 0,02 0,01 0,05Rubiaceae Ladenbergia macrocarpa 3 0,01 0,02 0,01 0,05Clusicaceae Chrysochlamys colombiana 4 0,02 0,02 0,01 0,04Fabaceae Inga sierrae 2 0,01 0,01 0,02 0,04Bignoniaceae Tecoma stans var. velutina 3 0,01 0,02 0,00 0,04Lauraceae Ocotea sp.1 1 0,00 0,01 0,03 0,04Piperaceae Piper aff. imperialis 4 0,02 0,01 0,01 0,04Hippocastanaceae Billia columbiana 2 0,01 0,01 0,01 0,03Myrtaceae Myrcia popayanensis 3 0,01 0,01 0,01 0,03Primulaceae Geissanthus francoae 3 0,01 0,01 0,00 0,03Sapindaceae Pouteria torta tuberculata 1 0,00 0,01 0,02 0,03Euphorbiaceae Alchornea glandulosa 1 0,00 0,01 0,02 0,03Melastomataceae Miconia curvipetiolata 2 0,01 0,01 0,01 0,03Verbenaceae Aegiphila grandis 2 0,01 0,01 0,00 0,03Sabiaceae Meliosma violacea 2 0,01 0,01 0,00 0,03Sapindaceae Allophyllus mollis 2 0,01 0,01 0,00 0,02Lauraceae Nectandra lineatifolia 2 0,01 0,01 0,00 0,02Lauraceae Ocotea insularis 2 0,01 0,01 0,01 0,02Melastomataceae Miconia smaragdina 3 0,01 0,01 0,00 0,02Melastomataceae Miconia theaezans 2 0,01 0,01 0,01 0,02Nyctaginaceae Guapira myrtiflora 2 0,01 0,01 0,01 0,02Lauraceae Ocotea lentii 1 0,00 0,01 0,01 0,02Melastomataceae Meriania speciosa 1 0,00 0,01 0,01 0,02Melastomataceae Miconia aff. coronata 2 0,01 0,01 0,00 0,02Rubiaceae Notopleura cf. capacifolia 2 0,01 0,01 0,00 0,02Melastomataceae Miconia sp.6 2 0,01 0,01 0,00 0,02Poaceae Chusquea latifolia 2 0,01 0,01 0,00 0,02Boraginaceae Cordia af. lucidula 1 0,00 0,01 0,01 0,02Clusicaeae Clusia crenata 1 0,00 0,01 0,00 0,01Cyatheaceae Trichipteris conjugata 1 0,00 0,01 0,00 0,01Solanaceae Solanaceae 1 0,00 0,01 0,00 0,01Staphylaceae Turpinia occidentalis 1 0,00 0,01 0,00 0,01Clusicaceae Chrysochlamys dependens 1 0,00 0,01 0,00 0,01Rubiaceae Ladenbergia oblongifolia 1 0,00 0,01 0,00 0,01Melastomataceae Miconia aff. acuminifera 1 0,00 0,01 0,00 0,01Rubiaceae Cinchona pusbences 1 0,00 0,01 0,00 0,01Melastomataceae Miconia sp.5 1 0,00 0,01 0,00 0,01Rubiaceae Palicourea calophlebia 1 0,00 0,01 0,00 0,01Rubiaceae Palicourea ovalis 1 0,00 0,01 0,00 0,01Solanaceae Solanum aphyodendron 1 0,00 0,01 0,00 0,01Meliaceae Trichilia martiana 1 0,00 0,01 0,00 0,01Ericaceae Cavendishia bracteata 1 0,00 0,01 0,00 0,01Pentaphylacaceae Freziera nervosa 1 0,00 0,01 0,00 0,01Euphorbiaceae Hyeronima macrocarpa 1 0,00 0,01 0,00 0,01Lauraceae Ocotea macrophylla 1 0,00 0,01 0,00 0,01Rubiaceae Palicourea cuatrecasasii (sp1) 1 0,00 0,01 0,00 0,01Rubiaceae Psychotria hazenii 1 0,00 0,01 0,00 0,01Annonaceae Guatteria crassipes 1 0,00 0,01 0,00 0,01264 1 1 1 3IVI: Importance Value Index152 Table C.2 Family, species and Importance Value Index (IVI) of vegetation found in the Inceptisols watershed Family Species IndividualsRelative densityRelative frequencyRelative coverageIVIHeliconiaceae Heliconia griggsiana 75 0,18 0,04 0,16 0,38Arecaceae Prestoea acuminata 30 0,07 0,03 0,06 0,16Piperaceae Piper sp. 3 30 0,07 0,01 0,04 0,13Piperaceae Piper sp. 2 26 0,06 0,03 0,03 0,12Moraceae Clarisia biflora 9 0,02 0,03 0,07 0,12Actinidiaceae Saurauia cuatrecasana 12 0,03 0,04 0,04 0,11Meliaceae Cedrela montana 3 0,01 0,02 0,07 0,09Anacardiaceae Toxicodendron striatum 9 0,02 0,03 0,04 0,09Tiliaceae Helicocarpus americanus 9 0,02 0,03 0,04 0,09Rubiaceae Palicourea thyrsiflora 14 0,03 0,03 0,02 0,09Siparunaceae Siparuna lepidota 14 0,03 0,03 0,03 0,08Myrsinaceae Myrsine guianensis 11 0,03 0,03 0,02 0,08Asteraceae Verbesina arborea 7 0,02 0,03 0,02 0,06Melastomataceae Miconia ochracea 9 0,02 0,02 0,02 0,06Poaceae Guadua angustifolia 13 0,03 0,01 0,02 0,06Nyctaginaceae Guapira costaricana 7 0,02 0,02 0,01 0,05Mimosaceae Inga coruscans 3 0,01 0,02 0,02 0,04Rubiaceae Palicourea sp. 1 6 0,01 0,02 0,01 0,04Heliconiaceae Heliconia huilensis 9 0,02 0,01 0,01 0,04Cyatheaceae Cyatheaceae 1 5 0,01 0,01 0,01 0,04Clusiaceae Vismia guianensis 2 0,00 0,01 0,02 0,04Boraginaceae Cordia cilindrosthachya 5 0,01 0,01 0,01 0,04Lauraceae Ocotea macrophylla 4 0,01 0,02 0,01 0,03Flacourtiaceae Hasseltia floribunda 4 0,01 0,01 0,01 0,03Meliaceae Ruagea glabra 2 0,00 0,01 0,01 0,03Sapindaceae Cupania americana 3 0,01 0,01 0,01 0,03Clusiaceae Chrysochlamys colombiana 4 0,01 0,01 0,01 0,03Flacourtiaceae Casearia megacarpa 2 0,00 0,01 0,01 0,03Urticaceae Myriocarpa stipitata 3 0,01 0,01 0,01 0,03Lauraceae Lauraceae 3 1 0,00 0,01 0,02 0,02Ulmaceae Trema micrantha 2 0,00 0,01 0,01 0,02Mimosaceae Zygia lehmannii 3 0,01 0,01 0,00 0,02Siparunaceae Siparuna laurifolia 3 0,01 0,01 0,00 0,02Piperaceae Piper sp. 1 3 0,01 0,01 0,00 0,02Melastomataceae Miconia lehmannii 3 0,01 0,01 0,00 0,02Lauraceae Ocotea lentii 1 0,00 0,01 0,01 0,02Euphorbiaceae Alchornea glandulosa 3 0,01 0,01 0,00 0,02Solanaceae Cuatresia riparia 4 0,01 0,01 0,01 0,02Lauraceae Beilschmiedia costaricensis 1 0,00 0,01 0,01 0,02Rubiaceae Guettarda crispiflora sabiceoides 2 0,00 0,01 0,00 0,02Siparunaceae Siparuna aspera 2 0,00 0,01 0,00 0,02Passifloraceae Passiflora arborea 2 0,00 0,01 0,01 0,02Boraginaceae Tournefortia scabrida 2 0,00 0,01 0,01 0,02Bignoniaceae Tecoma stans 2 0,00 0,01 0,00 0,02Euphorbiaceae Acalypha macrostachya 2 0,00 0,01 0,00 0,02Rubiaceae Palicourea sp. 2 3 0,01 0,01 0,01 0,02Melastomataceae Miconia sp. 3 2 0,00 0,01 0,00 0,02Euphorbiaceae Hyeronima scabrida 2 0,00 0,01 0,00 0,02Cyatheaceae Cyatheaceae 3 2 0,00 0,01 0,01 0,02Lauraceae Lauraceae 4 2 0,00 0,01 0,00 0,01Mimosaceae Inga sp. 2 1 0,00 0,01 0,01 0,01Rhamnaceae Rhamnus sphaerocarpa 2 0,00 0,01 0,00 0,01Euphorbiaceae Croton magdalenensis 1 0,00 0,01 0,00 0,01Melastomataceae Miconia minutiflora 2 0,00 0,01 0,00 0,01Boraginaceae Cordia sp. 1 0,00 0,01 0,00 0,01Lauraceae Ocotea sp. 2 2 0,00 0,01 0,00 0,01Bombacaceae Spirotheca rhodostyla 1 0,00 0,01 0,00 0,01Cyatheaceae Cyatheaceae 2 1 0,00 0,01 0,00 0,01Lauraceae Nectandra sp. 2 1 0,00 0,01 0,00 0,01Melastomataceae Miconia sp. 1 1 0,00 0,01 0,00 0,01Rosaceae Prunus carolinae 1 0,00 0,01 0,00 0,01Actinidiaceae Saurauia ursina 1 0,00 0,01 0,00 0,01Lauraceae Nectandra sp. 1 1 0,00 0,01 0,00 0,01Lauraceae Lauraceae 1 1 0,00 0,01 0,00 0,01Piperaceae Piper sp. 4 1 0,00 0,01 0,00 0,01Lauraceae Ocotea sp. 1 1 0,00 0,01 0,00 0,01Lauraceae Lauraceae 5 1 0,00 0,01 0,00 0,01Melastomataceae Miconia notabilis 1 0,00 0,01 0,00 0,01Rubiaceae Palicourea angustifolia 1 0,00 0,01 0,00 0,01Sabiaceae Meliosma sp. 1 0,00 0,01 0,00 0,01Annonaceae Raimondia sp. 1 0,00 0,01 0,00 0,01Euphorbiaceae Sapium stylare 1 0,00 0,01 0,00 0,01Lauraceae Lauraceae 2 1 0,00 0,01 0,00 0,01Melastomataceae Miconia sp. 2 1 0,00 0,01 0,00 0,01Mimosaceae Inga sp. 1 1 0,00 0,01 0,00 0,01Mimosaceae Inga densiflora 1 0,00 0,01 0,00 0,01Monimiaceae Mollinedia repanda 1 0,00 0,01 0,00 0,01Monimiaceae Mollinedia campanulacea 1 0,00 0,01 0,00 0,01Rubiaceae Ladenbergia oblongifolia 1 0,00 0,01 0,00 0,01Rubiaceae Coffea arabiga 1 0,00 0,01 0,00 0,01Zingiberaceae Renealmia ligulata 1 0,00 0,01 0,00 0,01411 1 1 1 3IVI: Importance Value Index153 Appendix D. Presence and intensities of crystalline minerals in Andisols and Inceptisols Table D.1 Presence and intensities of crystalline minerals in Andisols Abundance of soil minerals = 1: traces (intensities <10%), 2: present (<10% intensities <20%), 3: significant (<20% intensities <40%), 4: dominant (<40% intensities <100%) Hydro biotiteIllite or micasAmphiboles Chrysotile and Antigorite Chrysotile and Antigorite Cristobalite K feldspar Na feldspars QuartzChlorite or vermiculiteKaolinite, meta halloysiteCa and Mg carbo-natesMagnetite (Fe3O4)Hematite (αFe2O3)d=11.68, 3.87, 3.44d=9.93d=8.37, 3.12, 2.70d=7.52, 3.59d= 7.05, 3.63d=4.04 d=3.25d=3.74, 3.19d=4.24, 3.33d=14.06d=4.41, 3.52d=2.93, 2.84d=2.55 d=2.48A 1 - 1 - 1 4 1 2 4 1 - 1 - -B1 3 - 2 - 2 3 1 3 4 3 1 1 - -C - 1 1 2 - 4 1 - 1 - 4 1 1 1A 1 - 1 - - 4 - 3 3 1 - 1 - 1B 1 - 1 - - 4 - 3 3 1 - 1 - 1C 1 - 1 - - 4 - 2 2 1 - 1 - 1Secondary mineralsB3P4ProfileHorizonPrimary minerals154 Table D.2 Presence and intensities of crystalline minerals in Inceptisols Abundance of soil minerals = 1: traces (intensities <10%), 2: present (<10% intensities <20%), 3: significant (<20% intensities <40%), 4: dominant (<40% intensities <100%) Chlorite or vermiculiteMusco-viteNa feldspars Chrysotile and Antigorite Horn-blendeQuartzPartially dehydrated halloysiteKaolinite, meta halloysiteCa and Mg carbonatesCa and Mg carbonatesMagnetite (Fe3O4)Hematite (αFe2O3)Goethite (αFeO(OH))Boehmite (g-AlO(OH))d=14.17, 7.12, 3.54d=4.94d=4.05, 3.19d= 3.63 d= 3.13d=4.24, 3.33, 1.81d=7.28d=4.42, 3.56d=2.99, 2.93, 2.89, 2.23, 2.01d=2.28, 2.12, 1.97d=2.55d=2.66, 2.51d=4.15, 2.45d=2.34A - - 1 - - 4 2 4 1 2 1 - 1 -B - - 2 - - 4 2 3 - 1 1 - 1 -C - - 1 - - 4 2 3 - 1 1 1 1 -A - - - - - 4 2 3 1 1 2 1 3 1B1 - - - - - 4 2 3 1 1 2 3 3 1B2 1 - - - - 1 2 4 1 1 3 1 3 1Secondary mineralsB4ProfileHorizonP7Primary minerals155 Appendix E. Correlations between short-range order minerals and soil water retention characteristics with soil physical and chemical properties with all data of Andisols and Inceptisols Table E.1 Correlations between short-range order (SRO) minerals and indices and soil physical and chemical properties 1Pyrophosphate-extractable aluminum (Al) and iron (Fe) (Alp and Fep); 2SOC: soil organic carbon; 3ρs and ρb: particle and bulk density; correlations shown are the ones r > 0.4; Gray cells show correlations with p < 0.01 and white cells show correlations with p < 0.05 All horizons (n= 86) A horizon (n= 35)Alp1 (g/kg)Fep1 (g/kg)Allophane (%)Ferrihydrite (%) Alp1 (g/kg)Fep1 (g/kg)Allophane (%)Ferrihydrite (%)pH H2O -0.46 -0.79 -0.75 -0.59 0.69pH CaCl2 -0.70 -0.60 -0.46 0.58Sand (%) 0.49 0.75 0.71 0.60 0.75 -0.64Silt (%)Clay (%) -0.45 -0.72 -0.59 -0.55 -0.57 0.52SOC2 (%) 0.48 0.66 0.46 0.53 0.52 -0.75ρs3 (kg/m3) -0.42ρb3 (kg/m3) -0.55 -0.41 -0.52 -0.48 -0.57 -0.53 0.51B horizon (n= 38) C horizon (n= 13)pH H2OpH CaCl2 0.56Sand (%) 0.59 0.84Silt (%)Clay (%) -0.61 -0.82SOC2 (%) 0.43 0.59 0.88ρs3 (kg/m3) -0.55 -0.61ρb3 (kg/m3) -0.52 -0.71156 Table E.2 Correlations between soil water retention (SWR) characteristics and soil physical and chemical properties 1 θSat, θFC, θ100 kPa, θ500 kPa, θPWP: soil water content (θ) at saturation (Sat), field capacity (FC), 100 kPa, 500 kPa and at permanent wilting point (PWP), respectively; 2PAWS: plant available water storage; 3GW: gravitational water; 4SOC: soil organic carbon; 5Pyrophosphate-extractable aluminum (Al) and iron (Fe) (Alp and Fep); 6ρb: bulk density; correlations shown are the ones r > 0.4; Gray cells show correlations with p < 0.01 and white cells show correlations with p <0.05 All horizons (n= 86) A horizon (n= 35)θSat1 (%v/v)θFC1 (%v/v)θ100 kPa1 (%v/v)θ500 kPa1 (%v/v)θPWP kPa1 (%v/v)PAWS2 (% v/v)GW3 (% v/v)θSat1 (%v/v)θFC1 (%v/v)θ100 kPa1 (%v/v)θ500 kPa1 (%v/v)θPWP kPa1 (%v/v)PAWS2 (% v/v)GW3 (% v/v)Sand (%) 0.42Silt (%)Clay (%)SOC4 (%) 0.60 0.43 0.50 0.64 0.59 0.48 0.42 0.44Alp5 (g/kg) 0.58 0.58 0.43 0.43 0.48 0.46 0.59 0.48 0.51 0.55Fep5 (g/kg) 0.47 0.41 0.57 0.67 0.55 0.60 0.61Allophane (%) 0.49 0.43 0.44 0.52 0.56 0.47 0.45 0.49Ferrihydrite (%) -0.56 -0.65 -0.55 -0.58 -0.52ρb6 (kg/m3) -0.91 -0.69 -0.51 -0.47 -0.50 -0.51 -0.94 -0.79 -0.67 -0.58 -0.59 -0.59B horizon (n= 38) C horizon (n= 13)Sand (%) 0.61 0.40 0.41Silt (%)Clay (%) -0.63 -0.41 -0.48SOC4 (%) 0.57Alp5 (g/kg) 0.64 0.66 0.50 0.50 0.66Fep5 (g/kg)Allophane (%) 0.71 0.57 0.60Ferrihydrite (%)ρs2 (kg/m3) -0.48ρb6 (kg/m3) -0.87 -0.46 -0.40 -0.43 -0.89 -0.83 -0.85 -0.83 -0.80157 Appendix F. Comparison of median values of soil water retention characteristics (SWR) between Andisols and Inceptisols 1 θSat, θFC, θ100 kPa, θ500 kPa, θPWP: soil water content (θ) at saturation (Sat), field capacity (FC), 100 kPa, 500 kPa and at permanent wilting point (PWP), respectively; 2Values in parenthesis are first and third quartile, 3PAWS: plant available water storage; 4GW: gravitational water; Number of samples (n) for Andisols were 17, 19 and 8 and for Inceptisols 18, 19 and 5 for A, B and C horizons, respectively; **, *, + Significant differences between Andisols and Inceptisols with Mann Whitney U test at p < 0.01, p < 0.05 and p < 0.1, respectively Horizon Andisols InceptisolsA 78.0 (76.1-83.4)268.2 (66.7-71.8)**B 74.7 (66.3-75.9) 63.6 (60.9-67.8)**C 71.9 (61.2-77.3) 63.3 (60.4-68.5)A 66.8 (63.1-70.3) 57.4 (51.7-59.5)**B 61.4 (54.5-64.9) 52.3 (50.4-56.7)**C 62.1 (51.1-67.7) 53.0 (50.2-55.7)A 58.2 (54.1-60.4) 50.9 (46.5-53.2)**B 50.8 (45.2-56.8) 47.7 (45.1-51.8) +C 57.0 (45.0-61.3) 47.8 (45.8-50.8)A 55.1 (51.0-58.3) 48.2 (44.2-49.9)**B 48.6 (43.8-54.8) 44.7 (43.0-48.6)+C 54.0 (42.4-56.7) 46.2 (43.8-47.9)A 50.7 (48.7-54.0) 42.7 (40.6-45.6)**B 47.3 (43.8-51.2) 42.7 (37.4-43.5)**C 51.3 (41.9-53.5) 43.2 (40.0-45.8)A 16.4 (12.4-18.8) 13.5 (12.1-16.0)+B 13.9 (10.8-14.9) 11.6 (9.4-13.5)C 11.1 (9.2-14.0) 10.9 (8.4-11.2)A 11.9 (9.8-14.5) 13.9 (9.2-18.6)B 11.5 (9.1-15.1) 11.6 (8.6-13.2)C 10.1 (9.4-12.0) 10.7 (8.3-14.6)GW 4 (%v/v)θSat 1 (%v/v)θFC 1 (%v/v)θ100 kPa 1 (%v/v)θ500 kPa 1 (%v/v)θPWP kPa 1 (%v/v)PAWS 3 (%v/v)158 Appendix G. Precipitation for a) Sonora and b) El Chocho watersheds during the overland flow assessment period (mm) a) b) Source: Roa and Brown, 2014. To infill gaps in daily precipitation (February 15 to March 14, 2012; June 1 to 5, 2012 and February 6 to March 7, 2013) in the Inceptisol site, a regression through the origin was assessed with SPSS software (IBM, 2011) using the daily precipitation data of the nearby station Villa Aracelly, managed by the local environmental authority Corporación Autónoma Regional del Valle del Cauca (CVC). Season Year 2012 2013 2014January 238 273February 332 239March 216 338April 363 216May 428 313June 178 127July 54 58August 159 88September 220 263October 414 326November 305 490December 302 393Dry season 2Wet season 1Dry season 1Wet season 2Season Year 2011 2012 2013January 124 64February 57 109March 64 54April 287 124May 70 214June 80 37July 30 26August 77 84September 58 72October 144 109November 184 77December 179 39Incomplete dataWet season 1Dry season 1Wet season 2Dry season 2159 Appendix H. Results of field saturated hydraulic conductivity (Kfs) and soil water content in Andisols and Inceptisols without high values (>50mm/hr) Kfs data excluded from Andisols were two from the dry season; then, n= 12 for the wet season and 10 for the dry season. Data excluded from Inceptisols were three from the dry season and two from the wet season; then, n= 9 from the dry season and n =10 from the wet season. Table H.1 Field saturated hydraulic conductivity (Kfs) and soil water content comparison between soil orders 1 Kfs: field saturated hydraulic conductivity; 2θgrav: field gravimetric soil water content; 3 θvol: Calculated volumetric soil water content; 4Values in parenthesis are first and third quartile; ** Significant differences between Andisols and Inceptisols in the dry and wet season in pasture with Mann Whitney U test at p < 0.01 Table H.2 Field saturated hydraulic conductivity (Kfs) and soil water content comparison between seasons 1 Kfs: field saturated hydraulic conductivity; 2θgrav: field gravimetric soil moisture; 3 θvol: Calculated volumetric moisture; 4Values in parenthesis are first and third quartile; +, ** Significant differences between dry and wet season in pasture of Andisols and Inceptisols with Mann Whitney U test at p < 0.1 and p < 0.01, respectively "@en ; edm:hasType "Thesis/Dissertation"@en ; vivo:dateIssued "2018-09"@en ; edm:isShownAt "10.14288/1.0371261"@en ; dcterms:language "eng"@en ; ns0:degreeDiscipline "Soil Science"@en ; edm:provider "Vancouver : University of British Columbia Library"@en ; dcterms:publisher "University of British Columbia"@en ; dcterms:rights "Attribution-NonCommercial-NoDerivatives 4.0 International"@* ; ns0:rightsURI "http://creativecommons.org/licenses/by-nc-nd/4.0/"@* ; ns0:scholarLevel "Graduate"@en ; dcterms:title "Soil properties and land use affecting soil water dynamics in Andisols and Inceptisols at two mid-elevation sites in the Colombian Andes"@en ; dcterms:type "Text"@en ; ns0:identifierURI "http://hdl.handle.net/2429/66968"@en .