UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Reactive soil components and logging in Podzols of southwestern British Columbia Grand, Stephanie 2011

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
24-ubc_2011_fall_grand_stephanie.pdf [ 1.4MB ]
Metadata
JSON: 24-1.0105110.json
JSON-LD: 24-1.0105110-ld.json
RDF/XML (Pretty): 24-1.0105110-rdf.xml
RDF/JSON: 24-1.0105110-rdf.json
Turtle: 24-1.0105110-turtle.txt
N-Triples: 24-1.0105110-rdf-ntriples.txt
Original Record: 24-1.0105110-source.json
Full Text
24-1.0105110-fulltext.txt
Citation
24-1.0105110.ris

Full Text

REACTIVE SOIL COMPONENTS AND LOGGING IN PODZOLS OF SOUTHWESTERN BRITISH COLUMBIA by STEPHANIE GRAND B.Sc., Université Lyon I, 2001  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Resource Management and Environmental Studies)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  July 2011  © Stephanie Grand, 2011  ABSTRACT This study investigated soil characteristics and response to logging in a forested watershed of coastal British Columbia (Roberts Creek Study forest). We focused on soil organic carbon (SOC), labile and exchangeable elements, and short-range order (SRO) inorganic phases. These soil attributes were investigated in each pedogenic horizons (FH, Ae, Bf1, Bf2, BC, Cg) of 27 soil profiles. The profiles formed a disturbance chronosequence including undisturbed (control) forested sites, recently logged (cleared) sites harvested 1 to 5 years prior to sampling, and older logged (regenerating) sites harvested 8 to 15 years prior to sampling. Soils were coarse-textured humo-ferric Podzols developed on a glacial till underlain by granodioritic bedrock. We found that the average soil profile stored approximately 15.9 kg C / m2 over a depth of 1 m. Over 60% of the profile’s SOC was found at depth greater than 20cm. Predictors of SOC concentration included pyrophosphate and oxalate-extractable Al and Fe and the clay content. The forest floor of logged plots stored more C than undisturbed plots, most likely due to the addition of logging slash decay products. In the mineral subsoil, SOC was higher in cleared plots and similar to control levels in regenerating plots. This suggests that logging resulted in SOC gains to the mineral soil, but that these inputs were not stabilized. The subsoil played a key role in the overall response of SOC storage after logging and must therefore be taken into consideration in C dynamics studies. Short-range order inorganic phases such as imogolite-type material and ferrihydrite were a major component of the soil reactive fraction and were estimated to represent approximately 45% of clay-sized subsoil component. The major constituent of SRO phases was likely low-crystallinity proto-imogolite. Exchangeable ions and nutrients concentration were relatively constant in logged and control plots, with the exception of nitrate. Overall, the impacts of logging were not severe. We hypothesize that the good retention of SOC stores in logged plots and the presence of SRO material in lower mineral horizons contributes to the retention of nutrients in the profile and reduces the ecological effects of forest disturbance.  ii  PREFACE  I undertook this thesis research under the supervision of Dr. Les Lavkulich. Throughout this project, Dr. Lavkulich provided conceptual guidance and editorial advice. Laboratory manager Carol Dyck performed the analytical portion of chemical extractions (operation of the induction furnace, flow injection analysis spectrophotometer and inductively-coupled plasma atomic emission spectrometer). I conducted the rest of the laboratory work, including the infrared spectroscopy and transmission electron microscopy. A version of Chapter 2 has been accepted for publication: Grand, S. and L. M. Lavkulich. 2011. Depth distribution and predictors of soil organic carbon in Podzols of a forested watershed in southwestern Canada. Soil Science: In Press. I was responsible for the writing and revision of the manuscript as well as the rest of the thesis.  iii  TABLE OF CONTENTS ABSTRACT ...........................................................................................................................................................ii PREFACE ............................................................................................................................................................ iii TABLE OF CONTENTS .......................................................................................................................................... iv LIST OF TABLES................................................................................................................................................... vi LIST OF FIGURES ................................................................................................................................................ vii LIST OF ABBREVIATIONS................................................................................................................................... viii ACKNOWLEDGEMENTS ...................................................................................................................................... ix DEDICATION ....................................................................................................................................................... x CHAPTER ONE Introduction ................................................................................................................................ 1 Thesis objectives and organization ................................................................................................... 2 The field site ...................................................................................................................................... 3 Experimental design and sampling methodology ............................................................................. 6 Soil analyses ...................................................................................................................................... 8 Statistical analyses .......................................................................................................................... 14 Soils of the study area ..................................................................................................................... 16 CHAPTER TWO Soil organic carbon and related variables: Depth distribution and predictors ............................. 19 Synopsis ........................................................................................................................................... 20 Introduction..................................................................................................................................... 21 The soil profile ................................................................................................................................. 24 Predictors of SOC ............................................................................................................................. 37 Summary and conclusions ............................................................................................................... 42 CHAPTER THREE Soil organic carbon and related variables: Effects of logging..................................................... 45 Synopsis ........................................................................................................................................... 46 Introduction..................................................................................................................................... 47 C and N concentration..................................................................................................................... 54 C and N stocks ................................................................................................................................. 58 Organic matter size fractions .......................................................................................................... 61 Indicators of organic matter composition ....................................................................................... 62 Pyrophosphate extractable metals ................................................................................................. 69 Integrated effects of logging ........................................................................................................... 70 Conclusion ....................................................................................................................................... 72 CHAPTER FOUR Exchangeable ions and nutrients: Distribution and relations ..................................................... 75 Synopsis ........................................................................................................................................... 76 Introduction..................................................................................................................................... 76 pH and BS ........................................................................................................................................ 79  iv  Effective cation exchange capacity ................................................................................................. 83 N and P fractions ............................................................................................................................. 88 Conclusion ....................................................................................................................................... 92 CHAPTER FIVE Nutrients and labile ions: Effects of logging................................................................................. 94 Synopsis ........................................................................................................................................... 95 Introduction..................................................................................................................................... 95 Forest floor composition ............................................................................................................... 100 pH and exchangeable ions ............................................................................................................ 101 Overall effects of logging .............................................................................................................. 108 Conclusion ..................................................................................................................................... 109 CHAPTER SIX Poorly crystalline constituents: Characterization, distribution and relationships ......................... 111 Synopsis ......................................................................................................................................... 112 Introduction................................................................................................................................... 113 Selective dissolution results........................................................................................................... 116 TEM and FTIR results ..................................................................................................................... 126 Depth profiles ................................................................................................................................ 128 Importance of SRO phases ............................................................................................................ 131 Conclusion ..................................................................................................................................... 133 CHAPTER SEVEN Poorly crystalline constituents: Effects of logging .................................................................. 135 Synopsis ......................................................................................................................................... 136 Introduction................................................................................................................................... 136 Effects of logging........................................................................................................................... 139 A change in illuviation intensity? .................................................................................................. 141 Regenerating plots ........................................................................................................................ 142 Conclusion ..................................................................................................................................... 143 CHAPTER EIGHT General discussion and conclusions........................................................................................ 145 Overview ....................................................................................................................................... 146 Synthesis ....................................................................................................................................... 147 Significance ................................................................................................................................... 154 Limitations..................................................................................................................................... 155 Directions for future research ....................................................................................................... 157 BIBLIOGRAPHY ................................................................................................................................................ 158  v  LIST OF TABLES  Table 2.1  Sequence of horizons and average characteristics of control sites .......................25  Table 2.2  C and N stocks estimates in an average control profile ........................................26  Table 2.3  Soil texture and association of organic matter with different size fractions .........30  Table 2.4  C ratios in control plots .........................................................................................32  Table 2.5  Factor pattern for mineral horizons .......................................................................38  Table 2.6  Factor pattern for organic horizons .......................................................................38  Table 2.7  Regression results summary for the organic layer ................................................39  Table 2.8  Regression results summary for mineral horizons ................................................40  Table 3.1  Mean of selected soil variables by soil layer and treatment. ................................55  Table 3.2  Eigenvectors of the first 3 principal components .................................................71  Table 4.1  pH, base saturation and proportion of major ions on exchange complexes...........81  Table 4.2  Correlation coefficients between pH, base saturation and exchangeable Al. ........82  Table 4.3  Salt-extractable ions and cation exchange capacity of control plots .....................84  Table 4.4  Predictors of CECe as identified by multiple regression analysis .........................85  Table 4.5  Correlation matrix for CECe:C, pHCaCl2, SOC, Alp, Alsro+Fesro and Alp:Alo .........87  Table 4.6  P fractions in different horizons ............................................................................89  Table 4.7  Correlation coefficient between P fractions and selected variables ......................91  Table 4.8  Labile N in different horizons of control plots .....................................................92  Table 5.1  Effects of logging on forest floor composition ...................................................101  Table 5.2  Effects of logging on soil pH and CECe .............................................................103  Table 5.3  Effects of logging on salt-extractable cations and anions ...................................104  Table 6.1  Basic soil properties and selective dissolution results for control plots .............116  Table 6.2  Ratios and reactive fractions in control plots ......................................................118  Table 6.3  Correlation coefficients between the Alp:Alo ratio, pH and SOC .......................124  Table 7.1  Mean of reactive Al and Fe fractions ..................................................................140  Table 7.2  Mean of ratio variables .......................................................................................140  Table 8.1  Direction of significant differences observed between treatments .....................149  vi  LIST OF FIGURES  Figure 1.1  Location of Roberts Creek study forest ..................................................................3  Figure 1.2  Sampling locations ..................................................................................................7  Figure 1.3  Texture of different horizons ................................................................................18  Figure 2.1  Correlation of organic matter with mineral size fractions ....................................30  Figure 2.2  Soil organic carbon and C concentration of soil organic matter ...........................33  Figure 2.3  Soil C and pyrophosphate-extractable Al and Fe in the mineral soil ...................34  Figure 2.4  Soil C and pyrophosphate-extractable Al and Fe in the forest floor .....................35  Figure 3.1  Soil organic carbon concentration in mineral horizons .........................................57  Figure 3.2  Average C stocks in control, cleared and regenerating plots ...............................60  Figure 3.3  C:N variation in control, cleared and regenerating plots ......................................65  Figure 3.4  The C:N ratio as a function of soil organic carbon ...............................................67  Figure 3.5  Control, cleared and regenerating samples on principle components....................72  Figure 4.1  Relationship between base saturation and pH ......................................................83  Figure 4.2  Simple linear regression between soil carbon and cation exchange capacity .......86  Figure 5.1  N mineralization reactions .....................................................................................96  Figure 5.2  Effects of logging on salt-extractable nitrate ......................................................107  Figure 6.1  Values of the spodic index in different horizons of control plots .......................119  Figure 6.2  Relationship between Al and Si associated with short-range order material ......121  Figure 6.3  Imogolite-type material concentration as a function of soil pH .........................122  Figure 6.4  Pyrophosphate to oxalate-extractable Al ratio and ITM .....................................123  Figure 6.5  Evolution of the oxalate to dithionite-extractable Fe ratio with depth ...............125  Figure 6.6  Transmission electron micrographs ....................................................................126  Figure 6.7  Infrared spectra of the acid-dispersible clay fraction of Bf horizons ..................127  Figure 6.8  Depth profiles of Al and Fe in control plots ........................................................129  Figure 6.9  Depth profiles of C, N and P in control plots .....................................................130  Figure 6.10  Relative importance of clay-sized constituents in control plots .........................131  Figure 6.11  Relationship between non-crystalline phases and total clay concentration ........132  vii  LIST OF ABBREVIATIONS  BD: bulk density BS: base saturation CECe: effective cation exchange capacity FTIR: Fourier-transform infrared spectroscopy ITM: imogolite-type material Md: dithionite-extractable element Mexch: salt-extractable or exchangeable ion MH2O: water-extractable ion Mo: oxalate-extractable element Moxi: element associated with crystalline oxide Mp: pyrophosphate-extractable element Msro: element associated with short-range order inorganic material PC: principal component Posat: degree of P saturation of short-range order Al and Fe phases SEM: standard error of the mean SOC: soil organic carbon SOM: soil organic matter SRO: short-range order TEM: transmission electron microscopy  viii  ACKNOWLEDGEMENTS  I first wish to thank my supervisor Dr. Les Lavkulich for his guidance and unwavering support, and for attracting my attention to the wonderful world of poorly crystalline minerals. I thank my committee members Dr. Robert Hudson for his guidance in the field and helpful suggestions, and Dr. Hans Schreier for comments which led to the improvement of this thesis. I also thank two anonymous reviewers and Soil Science editor-in-chief, Dr. Robert Tate, for their helpful comments on the manuscript version of Chapter 2. Special thanks are owed to laboratory manager Carol Dyck for all her hard work and her attention to detail, as well as to field assistants Marina Romeo, Peter Shanahan and Bryan Forrest. I also thank the Coast Region Research Group for providing access to the Roberts Creek study forest, and research silviculturalist Brian D’Anjou for providing a field orientation. I wish to extend my gratitude to Dr. Maja Krzic for her inspiring dedication and for infecting me with her passion for all things ‘Soil’. Finally, my enduring gratitude goes to my parents for moral and financial support over the years. They never even complained.  ix  DE DI C AT I ON  To my parents, who first taught me to be curious about the world.  x  CHAPTER ONE  Introduction  1  THESIS OBJECTIVES AND ORGANIZATION This thesis documents soil processes and effects of logging disturbance on Podzols of coastal southern British Columbia. The major focus is on three aspects of soil biogeochemistry: (1) Soil organic matter (2) Plant nutrients and labile ions (3) Short-range order inorganic phases, such as imogolite-type material Soil organic matter (SOM) and soil organic carbon (SOC) dynamics is an issue of considerable current interest, since soils are the largest pool of terrestrial C. Soil management has the potential to significantly influence SOC sequestration (Lal, 2004). In order to study organic matter dynamics, we measured SOM concentration as well as total soil C and N. We also measured factors that have been shown to influence organic matter concentration and distribution in soils, such as pH, moisture, clay content, and organically-complexed Al and Fe. These results are presented in Chapters 2 and 3. The second part of the thesis investigates soil nutrients and labile ions. There is some evidence that logging increases soil acidity and negatively impacts the amount of plant-available nutrients, but results vary among ecosystems (Likens et al., 1970, Snyder and Harter, 1985, Johnson et al., 1991b, Olsson et al., 1996a, Johnson et al., 1997, Gravelle et al., 2009). We measured exchangeable / salt-extractable ions, exchangeable acidity, water-extractable ions, and plantavailable phosphorus. Chapters 4 and 5 document the relationships between these variables and the effects of logging. Short-range order (SRO) inorganic phases such as imogolite-type material (ITM) and ferrihydrite have received increasing attention over the last decade, largely due to analytical advances. Shortrange order materials contribute greatly to the reactivity of Podzolic horizons (Lundström et al., 2000a), but little is known about their response to disturbance. We investigated SRO phases by selective dissolution analysis, Fourier-transform infrared spectroscopy and transmission electron microscopy. The results of these analyses are presented in Chapters 6 and 7.  2  For each biogeochemical focus, the first chapter investigates the concentration, depth distribution of elements and relationships between variables. The objective is to gain insight into soil processes that control the concentration and distribution of elements in the profile. The second chapter focuses on the effects of logging, with the aim to understand and quantify the impact of disturbance on the ecosystem.  THE FIELD SITE Roberts Creek study forest This study was conducted at the Roberts Creek study forest (49° 27’ N, 123°41’ W) on the Sunshine Coast of southwestern British Columbia. The study forest is located between the towns of Roberts Creek and Sechelt approximately 40 km northwest of Vancouver (Fig. 1.1).  Figure 1.1: Location of Roberts Creek study forest. Reproduced with permission from D’Anjou (2002). The Roberts Creek forest was selected for this study because it contains a number of small logging plots that were harvested at various time within the last 20 years. It is also one of the 3  most productive forest environment in Canada. As a result, the characteristics of the area have been extensively studied by the BC Ministry of Forest, and studies detailing the effects of logging on biological communities and water quality have already been conducted (D'Anjou, 2002, Hudson and Tolland, 2002, Waterhouse and Harestad, 2002). This study complements existing research by focusing on soil processes and their response to logging disturbance. Biogeoclimatic setting The area lies within the Coastal Western Hemlock (CWH) biogeoclimatic zone and experiences a mean annual temperature of 10.2°C and mean annual precipitation of 1369 mm (Environment Canada, 2011). The area subzone is the Drier Maritime variant (CWHdm) (Pojar et al., 1991). The climate is oceanic and temperature variation between summer and winter is moderate, ranging from 3.7°C in January to 17.6°C in August (Environment Canada, 2011). Precipication averages 1370 mm annually and falls mostly in the form of rain. Two third of annual precipitation occurs between the months of October and March. Elevation ranges from 380 to 590 m above sea level and the forest is situated on a gentle (~ 15 %) southerly slope. The dominant species is Douglas-fir (Pseudotsuga menziesii) interspersed with smaller diameter western hemlock (Tsuga heterophylla) and western redcedar (Thuja plicata) (D'Anjou, 2001). Charcoal on standing and fallen snags indicates that the current forest regenerated following a wildfire that occurred around 140 years ago. Evidence for the fire is based on tree coring of the oldest standing trees along with remnants of burnt snags and fallen logs (D'Anjou, 2002). There are several older Douglas-fir veterans that survived the fire, as well as an old-growth western redcedar component that was selectively harvested in the early 20th century. Charcoal is also common in all layers of the soil profile, indicating that periodic wildfires occurred throughout the period of soil development (Zackrisson et al., 1996). Previous research conducted in the Roberts Creek study forest A brief review of previous research conducted by the BC Ministry of Forest provides an understanding of how this study complements previous works. The effects of forest harvesting on Roberts Creek forest structure, windthrows and regeneration were reported by D’Anjou (2001, 2002) and Hudson and D’Anjou (2001). Windthrow events were frequent during the first 4  fall and winter following harvesting, with trees growing on wetter soils being particularly susceptible to blowdown. Uprooting of trees growing in zero-order creeks caused large pulses of suspended sediment. Dispersed retention encouraged the natural regeneration of alder (Alnus spp.) followed by western hemlock, whereas natural regeneration in the clearcut blocks was similar to the composition of the surrounding forest and dominated by Douglas-fir, a relatively shade intolerant species. Logging effects on stream periphyton (benthic algal community) and invertebrates were reported by Hiebler (2003). Periphyton biomass was greater in harvested reaches than in forested headwaters. The increase in periphyton was attributed to the increase in incident solar radiation and in water temperature in logged reaches. Invertebrate density increased as periphyton biomass increased, suggesting bottom-up control of the aquatic food web. As far as terrestrial ecosystems are concerned, Waterhouse and Harestad (2002) found that forest harvesting negatively impacted nesting winter wrens in Roberts Creek because of changes to forest structure and habitat attributes. The effects of harvesting on hydrology and water quality were reported by Hudson (2001) and Hudson and Tolland (2002). The first report (Hudson, 2001) established that forest harvesting resulted in increased peak streamflow. The second report (Hudson and Tolland, 2002) focused on stream nitrate status following logging. One of the studied streams (F5) experienced a 4, 30 and 60 fold increase in nitrate levels 1, 2 and 3 years after logging, whereas in the other stream (F4), the nitrate levels were unchanged in year 1, then increased 12 and 18 fold in years 2 and 3. The difference in the level of response of the 2 streams could not be explained in terms of hydrology or harvesting intensity (Hudson and Tolland, 2002). Instead, Hudson suggested that differences in soil mineralogy and chemistry may be involved (personal communication). Prior to this study, the effects of logging on the soils of Roberts Creek had not been extensively examined. D’Anjou (2002) reported that physical soil disturbance was low due to the use of cable-yarding, with 3% or less of logged plots having exposed mineral soil; but with the exception of erosion, the response of Roberts Creek soils to disturbance had not been studied. By assessing the makeup and chemistry of the soils, and the evolution of these characteristics  5  following harvesting, this study fills a knowledge gap in our understanding of these coastal ecosystems and their response to logging disturbance.  EXPERIMENTAL DESIGN AND SAMPLING METHODOLOGY The chronosequence approach We chose to study a disturbance chronosequence to study the effects of logging on Roberts Creek soils. In the chronosequence approach, disturbed sites are compared to spatially distinct reference (control) sites. The other common method to assess the impacts of disturbance is referred to as the long-term experiment, or monitoring approach. In the monitoring approach, the control unit that establishes reference conditions consists in the pre-disturbance site (Dyck and Cole, 1994). The monitoring approach presents the advantage of greatly reducing problems caused by spatial variability. Forests in particular are heterogeneous systems that exhibit high spatial variability at the stand, watershed and regional scale (Fahey et al., 2005). However, in the absence of already existing baseline data, assessing medium and long-term impacts of ecosystem disturbance by monitoring is impractical due to the time needed to complete the study. In addition, in medium to long-term monitoring studies, the effects of of global change such as climate change or acid deposition may be both significant and indistinguishable from disturbance effects, thereby invalidating results. The chronosequence approach on the other hand does not require the existence of baseline data and involves a much smaller timeframe, but is generally only able to detect large effects (Yanai et al., 2003), due to the error term introduced in the experiment by spatial variability. Global change may still be a concern if it interacts with the effects of disturbance, but the interaction effect is likely to be small in short to medium-term studies. The most problematic assumption underlying the chronosequence approach is that the only difference between sites should be their disturbance regime, and that all other site properties should be similar (Dyck and Cole, 1994). 6  This assumption is unlikely to be met strictly under field conditions, since there are no identical natural systems. The chronosequence approach may still yield valid results however if researchers demonstrate reasonable similarity between sites (Pennock and van Kessel, 1997). In the Roberts Creek study forest, we observed no altitudinal, longitudinal or latitudinal gradients in vegetation, parent material, solum depth, or any of the soil properties measured. Since all sampling sites exhibited reasonable similarity in properties that are likely to affect the response variables, we expect the chronosequence to yield valid results. Experimental design We sampled 27 soil pits by morphological horizon. Nine soil pits were located on undisturbed forested plots (control), 11 were located on cleared stands (logged 2-5 years prior to sampling) and 7 in regenerating stands (logged 8-15 years prior to sampling). Samples from logged stands spanned 7 harvest clusters distributed throughout the experimental forest (Fig. 1.2). Control locations were interspersed in the areas between and around logged plots and at a distance of at least 30 m from the edge of the disturbance.  Figure 1-2: Approximate sampling locations in Roberts Creek Study Forest When sampling logged plots, our objective was to gain insight about the in-situ effects of vegetation removal over time, rather than the extent of mechanical disturbance caused by logging 7  equipment. Large differences can arise between logged plots due to changes over time in logging technology (Yanai et al., 2003) as well as the skill and commitment of logging crew to minimize soil disturbance. We sampled morphologically undisturbed soil profiles with no signs of mechanical disturbance or water erosion. We avoided old logging roads, equipment tracks and preferential flow channels with visible traces of erosion. Overall, the area of logged plots showing signs of disturbance or erosion was visually estimated to make up 10 to 25% of stand area. Sampling Sampling was conducted in August 2005. Each sample consisted of approximately 750g of soil. For each horizon or sub-horizon, soil was collected from all faces around a ~ 90 cm diameter soil pit. Samples were double bagged and transported to the laboratory in coolers. The forest floor was separated into two parts: (1) living moss and litter and (2) humic and fibric layer (FH). Only the FH layer was sampled. In a few instances, the C horizon occurred at depth greater than 120 cm and was not sampled.  SOIL ANALYSES Except for pH, all analyses were performed on the < 2 mm soil fraction. pH pH was measured on field moist samples within 48 h of sample collection. pH detemination was conducted potentiometrically in deionized water, in 0.01 M CaCl2 and in 1 M KCl. The soil:solution ratio was 1:2 for mineral horizons and 1:4 for organic horizons (Schofield and Taylor, 1955, Van Lierop, 1990). Suspensions were allowed to equilibrate for 30 min prior to measurement.  8  Soil moisture Gravimetric moisture content was determined by oven-drying field moist samples at 70°C (FH layer) or at 105°C (mineral horizons) to constant weight (Kalra and Maynard, 1991). All samples were collected during the dry season (August) and no rain occurred during the week preceding sampling. The gravimetric moisture content had no absolute meaning; rather, it allowed a comparison of moisture regime in plots of different age. Texture Texture was estimated after dispersion in Na hexamethaphosphate by a simple method combining wet sieving and sedimentation steps (Kettler et al., 2001). The sand fraction was separated by wet sieving, the silt fraction by sedimentation, and the remaining colloidal suspension corresponded to the clay fraction. Coarse fragments were an important component of the soil profile. The volume of cobble and stones was estimated visually. The gravel fraction was separated from the fine earths by sieving and weighed. The overall coarse fragment mass percentage was calculated assuming a specific gravity of 2.65 g/cm3. Organic matter Soil organic matter concentration was determined by loss-on-ignition in a muffle furnace. Samples were ovendried to constant weight then ashed at 375°C for 16 h and cooled in a dessicator (Kalra and Maynard, 1991). Weights were recorded to the nearest mg before and after combustion. Loss-on-ignition was performed on the < 2 mm soil fraction as a whole, as well as on the sand, silt and clay fractions separated for the purpose of textural determination (Kettler et al., 2001). This produced an estimate of the proportion of organic matter associated with each size fraction. The ashing temperature was set at 375 °C to prevent the evolution of significant amounts of hydration water from crystalline clay minerals (Guggenheim and Koster van Groos, 2001). At this temperature, loss of hydration water from poorly crystalline material such as ferrihydrite and ITM remains possible. Poorly crystalline oxyhydroxides may lose approximately 15% of their 9  mass between 100 and 400 °C (Šubrt et al., 1992). Assuming a poorly crystalline clay content of 2.5 % (Chapter 6), loss of dehydration water may amount to a 0.4% error in organic matter estimates. Total C and N Total C and N were measured on soil ground to pass through a 0.5 mm sieve by dry combustion using an induction furnace (LECO model CN-2000). All soils were free of carbonates and total C equals organic C (Girard and Klassen, 2001). Roots > 2 mm were not included in mineral soil analyses, possibly leading to under-estimates of C stocks. However, Homann et al. (2005) reported that 90% of the mineral soil SOC was in the < 2 mm fraction in coastal northwestern forests, so the error is not expected to be large. Forest floor samples were not sieved and all material was ground to pass through a 0.5 mm sieve. Total phosphorus and bases In the FH layer, total P and base cations were extracted by the Parkinson and Allen digestion procedure (Parkinson and Allen, 1975) followed by analysis on ICP-AES. This method was determined to be suitable for Podzols of coastal British Columbia (Cade-Menun and Lavkulich, 1997). Available phosphorus The Bray 1 extraction (Bray and Kurtz, 1945) was used to estimate plant-available soil P. This extraction procedure is suitable for acid soils in which the major portion of P exists as Al-P or Fe-P complexes. A dilute acid dissolves Al-P and to a lesser extent, Fe-P while fluoride ions prevent P readsorption (Benton Jones Jr., 2001). We followed the steps described in Kalra and Maynard (1991). Exchangeable acidity Exchangeable acidity (Al3+ and H+) was determined by extracting 20 g of soil with 60 mL of 2 M KCl followed by the titration procedure of Thomas (1982). This method provides measures of both total exchangeable acidity and exchangeable Al. 10  Exchangeable inorganic N Salt-extractable NO3-N and NH4-N were measured on the abovementioned KCl extracts using a Lachat QuikChem 8000 Flow Injection Analysis System (Hach company, 2009). Nitrate was determined using the cadmium reduction method followed by spectrophotometric detection of the nitrite formed (ISO standards, 1996). Ammonia was reacted with sodium salicylate and the resulting indophenol was measured with a spectrophotometer (Verdouw et al., 1978). Other salt-extractable ions Exchangeable and salt-extractable ions were leached from 10 g of soil with 160 mL of 1 N (NH4)2SO4 using a mechanical vacuum pump. Concentration of Ca2+, Mg2+, Na+, K+, Cl-, and PO43- in the extract was determined by inductively coupled plasma atomic emission spectroscopy (ICP-AES). Ammonium sulfate was chosen as a displacing salt over the more commonly used NaCl, KCl or NH4Cl salts because given the proximity to the ocean, Cl- was an ion of interest in the profile. In addition, sulfate is believed to be a more effective anion displacer than Cl- (Katou et al., 1996). A 2 M ammonium sulfate solution is weakly buffered at a pH of 5.2, which is close to the pH of the study soils (Chapter 4). Effective cation exchange capacity (CECe, cmolc / kg) was determined by the sum of charges pertaining to exchangeable cations (Ca, Mg, Na and K) and exchangeable acidity. Base saturation was calculated as the sum of charges pertaining to base cations (Ca, Mg, K, Na) divided by CECe (Hendershot et al., 1993). Water-soluble ions Water-soluble ions were extracted from soil suspensions with a 1:1.5 soil:deionized water ratio (Sheldrick, 1984). The suspensions were shaken on a reciprocal shaker at room temperature for 30 min, centrifuged at 400 g for 10 min and filtered through 2.5 μm filter paper. Al, Fe, Si, Ca, Mg, Na, K, Mn, and P elemental content in the extract was determined by ICP-AES.  11  Selective dissolution Reactive Al, Fe and Si fractions were evaluated by non-sequential extractions using sodium pyrophosphate (Bascomb, 1968), ammonium oxalate (McKeague and Day, 1966), and citratebicarbonate-dithionite (Mehra and Jackson, 1960). Extractions were conducted on field moist samples, since air drying can lead to irreversible changes in the surface properties of poorly crystalline materials and their chemical reactivity (Hernandez Moreno et al., 1985). For the pyrophosphate extractions, samples were shaken for 16 h in 0.1 M Na pyrophosphate at pH 10 and at a soil:solution ratio of 1:100. The oxalate extractant consisted of 0.2 M ammonium oxalate-oxalic acid buffer solution of pH 3.0 +/- 0.1. Samples were shaken for 4 h in the dark at a 1:100 soil:solution ratio. During the dithionite extraction, 5 g of soil was extracted with 2 g sodium dithionite and 40 mL of 0.3 M sodium citrate in a water bath heated to 75-80ºC. Five mL of sodium bicarbonate were added to buffer the solution pH near 7.3 and achieve maximum extraction effectiveness. All samples were fully discoloured (grey) after extraction was complete. Extracts were centrifuged at 450xg for 15 min. Pyrophosphate extracts were dispersed and had to be centrifuged at higher speed (7400xg) in order to obtain a clear supernatant. All extracts were filtered through a 2.5 μm cellulose filter (Whatman # 42). The concentration of Al, Fe and Si in the extracts were determined by inductively coupled plasma atomic emission spectroscopy (ICPAES Vista pro, Varian Inc.). Infrared spectroscopy and transmission electron microscopy Fourier-transfrom infrared spectroscopy and transmission electron microscopy are two effective ways to identify ITM in soils (Mossin et al., 2002, Parfitt, 2009), with the exception of phases with a low polymerization. Sample preparation for FTIR spectroscopy followed the general steps outlined by Farmer et al. (1980), including organic matter oxidation, salt removal, ultrasonic dispersion, acidification, and collection of the < 2 μm size fraction. Organic matter oxidation was done by one 8 h treatment with 1 M NaOCl at room temperature. NaOCl was preferred to H2O2 because it is less harmful to minerals (Siregar et al., 2004, Mikutta et al., 2005b). Heating was avoided to prevent conversion of weakly ordered minerals into more 12  crystalline ones (Kyle et al., 1975). While removal of some organic matter is necessary to maximize the dispersion of clays, the complete oxidation of all organic matter is not essential to IR analysis since the organic matter absorption band does not interfere with the ITM spectrum (Dahlgren, 1994). Samples were desalted by dialysis through a 1000 Da membrane, which retains particles larger than ~ 2 nm in diameter, and the samples were subjected to intense ultrasonic vibration at 100 W for 30s to aid dispersion. An acid dispersible clay suspension was subsequently prepared by adjusting the solution pH to 4 with dilute HCl. In acidic conditions, most crystalline phyllosilicates flocculate while ITM remains dispersed and in suspension. Samples were allowed to settle for 3 h. The supernatant was then collected in a beaker, frozen in liquid nitrogen, and freeze dried. The IR spectra were obtained on transmission mode on pellets containing 2 mg of the freezedried sample in 200 mg of KBr carrier. KBr was previously heated to 150ºC overnight to remove adsorbed water. Spectra were recorded between 4000 and 370 cm-1 on a Perkin Elmer 2000 operating at 0.5 cm-1 resolution. Each spectrum was obtained by averaging 50 scans to maximize the signal to noise ratio. The background signal was taken as the signal without a sample present (i.e. air signal) to avoid potential replication problems caused by KBr water bands (King et al., 2004). In a second run, the same pellets were heated at 350ºC, at which point ITM is destroyed, and another set of IR spectra was measured. The two sets of spectra were digitally subtracted to obtain the ITM contribution to absorbance. Transmission electron microscopy was performed on the acid dispersible clay fractions of illuvial horizons. The acid dispersible clay fraction was obtained in the way described above. One drop of diluted clay suspension was air dried on a C-coated Cu grid (Gilkes, 1994). Analyses were performed with a Hitachi H-800 TEM.  13  STATISTICAL ANALYSES All statistical analyses were carried out using SAS version 9.2 software (SAS Institute Inc., 2008) and all statistical tests were performed with an alpha level of 0.1. All means are given ± the standard error of the mean (SEM). Comparison of treatment and horizon means The effects of logging treatment and difference between horizons were studied using a mixed statistical model. Treatment effects (control / cleared / regenerating plots), horizon effects and treatment*horizon interaction were included as fixed effects. Observations were blocked by plot (logging operation) using a random group effect (G-side). To avoid pseudo-replication with respect to horizon effect, we included the horizon effect as a repeated measure (R-side random effect). This sets a common correlation among all observations of the same soil profile. We used the Toepliz covariance structure to model the correlation between horizons. The Toepliz structure is similar to the first order auto-regressive structure as pairs of observations separated by a common lag have a set correlation coefficient, but a different coefficient is calculated for each lag (Littell et al., 2006). The Toepliz structure was chosen because (1) it is a sensible model for our data, (2) the Bayesian information criteria was consistently better than for other covariance structures, and (3) there was no evidence of model lack of fit. Degrees of freedom were calculated using the Satterthwaite adjustment. Model diagnostics (normality, homoskedasticity, goodness of fit) were run on the conditional residuals (Haslett and Haslett, 2007). Normality of the residuals was tested by Shapiro-Wilk, Cramer-von Mises, and Anderson-Darling tests, which provide adequate power for limited sample sizes (Stephens, 1974). Homogeneity of variance and goodness of fit were assessed by observing conditional residual plots. Variables with non-normal residual distribution were transformed to achieve approximate normality. The Box-Cox procedure (Box and Cox, 1964) was used to identify appropriate transformations. The most commonly applied transformation was a log transformation. Other 14  transformations used included square root and reciprocal square root, and were successful in approximating residual normality. Treatment means were compared using a t-test with no provision for multiple inferences (Webster, 2007). If the interaction term was significant, treatment means were computed and compared separately for each horizon. In this last case, the analysis reduced to a single-factor experiment in which there are no repeated measures. Relationship between variables Relationships between variables of interest was investigated by examining correlation coefficients and performing simple or multiple regression analyses. If the treatment effect was found to interact with the relationships, meaning that the correlation or regression coefficients were significantly different between treatments, relationships were investigated separately for each treatment. When there was no evidence for an interaction of treatment with the relationships, correlation and regression results were given for the entire dataset. Correlation We used the Pearson product-moment correlation coefficient (r) to measure the amount of linear association between variables (Pearson, 1895). The significance of r was indicated by the associated p-value based on a t-test with n-2 degrees of freedom. Scatter plots were examined for evidence of non-linear association, heteroskedasticity, presence of outliers or non-homogeneous groups. Approximate bivariate normality was assessed on residual plots. Factor analysis and regression analysis The relationship between SOC and abiotic variables was investigated using a factor analysis followed by multiple regression analysis. The purpose of the factor analysis was to orthogonalize the independent variables and thus avoid the problems arising from multi-colinearity. Dimension reduction was not an objective of the analysis. Factor patterns and results of the regression analysis are given in Chapter 2.  15  Multiple regression analysis was also used in Chapter 4 to identify the predictors of cation exchange capacity. As colinearity between variables of interest in Chapter 4 was not as pronounced, the orthogonalizing step of factor analysis was not required. Regression models were built by a stepwise selection method with a significance level of 0.1 for variable entry and retention. The stepwise selection method is a combination of the forward (adding significant variables to the model) and backward (deleting variables that become nonsignificant from the model) selection methods. It ensures that all significant variables, and only significant variables, are included in the final model (Hocking, 1976). All regression residuals were examined for normality, homoskedasticity, goodness of fit and independence. Normality was tested by Shapiro-Wilk, Cramer-von Mises, and Anderson-Darling tests. Homogeneity of variance and goodness of fit were assessed by observing a plot of residual versus predicted values. A plot of residuals versus profile number was used to check that the error terms for each horizon were uncorrelated. Principle component analysis We used a principle component analysis to show the combination of logging effects in an integrated manner. Results of the principle component analysis and eigenvectors are given in Chapter 3.  SOILS OF THE STUDY AREA Soils of Roberts Creek began developing following glacial retreat approximately 11,000 years ago (Clague and Luternauer, 1983, Blais-Stevens et al., 2001, Menounos et al., 2004). The solum ranged from 40 to over 120 cm in thickness and there was no recognizable gradient in soil depth across sampling sites. From top to bottom the following sequence of horizon was observed: LFH (forest floor), Ae, Bf1, Bf2, BCg and Cg. The Ae horizon was remarkably thick, often reaching  16  15 cm in thickness. One profile exhibited buried Aeb and Bfb horizons while other soil profiles showed no sign of mechanical disturbance. The disturbed profile was excluded from the analysis. The soil temperature class was Mesic and the moisture regime was Humid according to Canadian soil climate classification (Lavkulich and Valentine, 1978). The soil type was Humo-Ferric Podzol (Soil Classification Working Group, 1998), Aquentic Haplorthods (Soil survey staff, 2006) or Albic Gleyic Podzol (IUSS Working Group WRB, 2006) of sandy loam to loamy sand texture. The soil parent material was a basal till deposited on granodioritic bedrock. This compacted basal till prevented root penetration and restricted water infiltration as well as the downward extent of pedogenic processes. Roots ran parallel to the surface of this layer. A perched water table exists in the wet months of the year, as evidenced by mottling and gleying in the BC horizon. The forest floor was relatively thin, ranging from 3 to 8 cm in undisturbed sites. The humus type was mor in 26 of the 27 soil profiles investigated and mull-like in the remaining profile. In this profile earthworms were abundant despite the low pH (pHH2O = 5.0 – 5.2) and coarse, sandy loam texture. Earthworms were found down to the Bf1 horizon. The distribution of earthworms seemed localized as two other samples from the same plot did not yield earthworms. No study has systematically inventoried earthworm distribution in coastal forests of British Columbia, but invasive and native earthworm species have been reported at several locations (Panesar et al., 2000, Addison, 2009). A weakly expressed granular structure was observed in the Bf1 horizons. The BC horizon had occasional crumb aggregates that appeared cemented by a red, iron-rich phase. The Cg horizons had a platy structure, most likely developed due to the weight of the glacial overburden. Other horizons (Ae and Bf2) largely lacked structure development. The soils had a high proportion of coarse fragments, ranging from 25% of the mass of the Ae to 60% of the mass of the deeper horizons (BC – C). The texture was that of a sandy loam in the topsoil and of a loamy sand in the subsoil.  17  Particle size distribution of the fine earth fraction was largely homogeneous throughout the soil profile (Fig. 1.3). The Ae horizon had slightly less clay, suggesting eluviation, and slightly more silt, indicating possible eolian inputs following glacial retreat, than the rest of the profile. Overall there was no marked textural contrast between horizons, indicating that the soil profile is likely to have developed from a homogeneous parent material. Therefore, physical and chemical differences between horizons should be largely the result of pedogenic processes.  Figure 1.3: Distribution of the sand (2000 – 50 μm), silt (50 – 2 μm) and clay (< 2 μm) size fractions in different horizons  18  CHAPTER TWO  Soil organic carbon and related variables:  Depth distribution and predictors  19  SYNOPSIS Forest soils of the Canadian west coast may store significant amounts of organic matter due to the cool climate and high forest productivity of the area. The objectives of this study were to determine the distribution of soil organic carbon (SOC) in the profile and to identify the most important predictors of SOC in Podzols of Roberts Creek study forest located in southwestern British Columbia. We sampled 9 soil profiles in undisturbed forest plots by morphological horizon and measured SOC using a dry combustion method. We also determined soil pH, texture, moisture content, total nitrogen, loss on ignition, and pyrophosphate and oxalate extractable Fe and Al. The average soil profile stored approximately 15.9 kg C / m2 over a depth of 100 cm, which is higher than SOC stocks reported for well-drained inland Canadian forests. The organic layer (LFH) only accounted for one quarter of the C stock. Sixty percent of the profile SOC (including the forest floor) was found in the subsoil of depth greater than 20 cm. Studies of SOC dynamics that only sample the topsoil are therefore inappropriate. Even though the clay concentration was low (~5%), the clay fraction accounted for 1/3 of SOC. This suggests that organo-mineral interactions were an important factor for SOC accumulation. The major predictors of SOC in the mineral horizons were organically complexed Al and Fe and short-range order inorganic material. Crystalline clays also appeared to play a role in organic matter retention, but were not as important as poorly crystalline compounds. In the organic layer, organically complexed Fe correlated negatively with SOC, indicating that the amount of Fe available for adsorption to organic matter is limited. Organically complexed Al did not show the same negative association, suggesting the existence of a mechanism for preferential upward translocation of Al into the FH horizon.  20  INTRODUCTION Interest in soil organic matter (SOM) dynamics is ever increasing as the recognition of the potential role for soil carbon in climate change is added to more traditional concerns regarding the impacts of SOM on soil physical properties and nutrient content (Lal, 2004). Generally, information about soil C stocks is rather scarce (Nalder and Merriam, 1995, Kurz et al., 2002). Soil organic carbon (SOC) stocks have been measured in some inland Canadian forests (Bhatti et al., 2002, Bois et al., 2009) and on the coast of Washington and Oregon states (Smithwick et al., 2002, Homann et al., 2005), but to the best of our knowledge no information is available for the British Columbian coast. Considerable SOC stocks may exist in this region due to mild winters and the lack of pronounced moisture deficit during the summer (Environment Canada, 2011) favouring high primary productivity. Carbon depth distribution Information on soil C stocks is especially scarce below the top 10 to 20 cm of the soil profile. It is common in C dynamics studies to restrict sampling to the topmost layer(s) of the soil profile (e.g. Covington, 1981, Bock and Van Rees, 2002, Rey et al., 2008). However, in moist coastal forests of western Washington and Oregon, Homann et al. (2005) reported that more than half of the soil C stocks was found below 20 cm depth. This points to the need for field studies taking into account subsoil as well as topsoil C, especially since controls on C dynamics may be different (Salomé et al., 2010). Even though SOC turnover generally decreases with depth, both fast-cycling and more persistent SOC forms are present in significant amounts in the lower part of the soil profile (Trumbore, 2000, Diochon and Kellman, 2009). Carbon stabilization The prospect of using soils to sequester C requires a thorough understanding of mechanisms for SOC stabilization (Sollins et al., 2007). In the meanwhile, recent advances on the C saturation concept (Six et al., 2002) demonstrate that the intrinsic  21  characteristics of soils determine the stabilization capacity of SOM pools and limit accumulation of SOC. The association of SOM with mineral particles has been recognized as one of the main processes leading to the stabilization of organic matter matter (Mayer, 1994). Clay-sized particles are thought to be of particular importance for SOC retention. It has been established that clays may stabilize SOM through mechanical occlusion, sorption of the organic substrate to mineral surfaces, and / or inactivation of microbial extra-cellular enzymes (Paul, 1984, Oades, 1988, Sollins et al., 1996, Baldock and Skjemstad, 2000). Short-range order (SRO) and amorphous inorganic phases such as imogolite, allophane, ferrihydrite and other poorly defined Al and Fe phases have the potential to stabilize large amounts of SOC because of their reactivity and large specific surface area (Eusterhues et al., 2005). The contribution SRO to SOC retention is well established in volcanic soils (Zunino et al., 1982, Torn et al., 1997, Rasmussen et al., 2006). Podzols have received more recent interest. Riise et al. (2000) suggested that high-molecular weight SOM may precipitate on allophanic material. Kaiser et al. (2002) and Eusterheues et al. (2005) proposed that Fe oxides and SRO Al and Fe phases were the most important substrates for the formation of organo-mineral associations in Podzols. Mikutta et al. (2005a, 2006) and Kleber et al. (2005) suggested that the most important mechanism for SOM stabilization in acid subsoils is the interaction with poorly crystalline minerals and polymeric metal species. Spielvogel et al. (2008) demonstrated that a specific fraction of the SOC pool is preferentially associated SRO Al and Fe material and is then potentially stabilized for thousands of years (Eusterhues et al., 2005). Organic matter stabilization by polyvalent cations, especially Fe and Al, has also been extensively studied. In volcanic soils, the formation of organic-metallic complexes is commonly invoked to explain the inhibition of organic matter mineralization (Broadbent et al., 1964, Egli et al., 2008). In acid forest soils, the formation of organic-metallic complexes of Al and Fe has also been shown to be involved in the precipitation of dissolved organic matter and its subsequent stabilization (Skjemstad, 1992, Rasmussen et al., 2006). Scheel et al. (2007) showed that Al-SOM complexes isolated from acid forest 22  soils were up to 28 times more resistant to microbial decay than their metal-free counterparts. Al-humus complexation may be particularly effective at preventing microbial decomposition through the occurrence of Al toxicity (Tate and Theng, 1980) or inhibition of enzyme activity (Sollins et al., 1996). Al ions may also enhance the sorption of humic acids on clays through the formation of Al bridges between the clay and organic matter (Varadachari et al., 1991) and result in physical protection of sorbed organics. Relative importance of stabilization mechanisms While it is generally accepted that crystalline clays, SRO phases and polyvalent cations all have a stabilizing effect on SOM, the relative importance of each mechanism appears to be site-specific. Several authors have reported a strong correlation between SOC and the clay fraction (Spain, 1990, Arrouays et al., 1995, Alvarez and Lavado, 1998, Homann et al., 1998, Shen, 1999, Dexter et al., 2008). However, when present, reactive Al and Fe phases appear to be stronger determinants of SOM retention than crystalline phyllosilicates (Spielvogel et al., 2008). A number of authors have found that SOM stabilization was due primarily to SRO and non-crystalline inorganic phases (Zunino et al., 1982, Schnitzer and Kodama, 1992, Torn et al., 1997, Kaiser et al., 2002, Mikutta et al., 2006, Rasmussen et al., 2006), while others have emphasized the importance of organic-metallic complexes of Al and Fe (Broadbent et al., 1964, Boudot, 1992, Percival et al., 2000, Scheel et al., 2007, Egli et al., 2008). These apparently contradictory findings suggest that the determinants of SOC retention depend on each soil’s mineralogical assemblage, organic matter characteristics and characteristic pedogenic processes (Allison et al., 1949, Krull et al., 2003). Objectives The objective of this study is to determine whether subsoils are a major contributor to SOC content in Podzols of coastal British Columbia and to identify the most important predictors of SOC concentration. We measured SOC concentration in all pedogenic horizons and developped a regression equation relating SOC to clay content, SRO 23  inorganic material and organically complexed metals. Environmental variables such as pH and moisture were also considered, as well as a broad indicator of organic matter composition and maturity (C:N ratio). We ranked predictors according to their relative importance in determining SOC.  THE SOIL PROFILE Soils are coarse-textured and acidic, with pH measured in 0.01M CaCl2 ranging from 3.6 in surface horizons to 4.8 in the mineral subsoil (Table 2.1). The depth distribution of elements was characteristic of Podzolic soils. The leached Ae horizon has the lowest concentration of SOC, N, and extractable metals, which accumulate in the illuviated Bf horizons and to a lesser extent in the BC horizon.  24  Table 2.1: Sequence of soil horizons and properties (mean ± standard error or the mean, SEM) of control sites (n=9). H or izon T hickness1 pH C aC l2  T extur e  SOM 2  SOC 3  N3  C lay 4  M oistur e5  A l p6  F ep6  A l sr o7  F esr o7  FH  6±1  3.6 ± 0.1  organic  72.1 ± 4.0  38.1 ± 2.1 1.06 ± 0.07  n/a  0.87 ± 0.10  2.3 ± 1.3  0.7 ± 0.2  0.6 ± 0.2  0.6 ± 0.1  Ae  6±1  3.6 ± 0.1  sandy loam  2.5 ± 0.1  1.6 ± 0.2 0.06 ± 0.01 3.6 ± 0.4  0.10 ± 0.02  0.6 ± 0.1  0.5 ± 0.1  0.3 ± 0.0  0.2 ± 0.1  Bf1  13 ± 2  4.5 ± 0.1  sandy loam  4.0 ± 0.4  2.2 ± 0.3 0.09 ± 0.01 5.5 ± 0.5  0.13 ± 0.01  2.9 ± 0.3  1.6 ± 0.2  5.6 ± 0.8  2.5 ± 0.3  Bf2  30 ± 4  4.8 ± 0.1  loamy sand  3.5 ± 0.4  1.6 ± 0.2 0.08 ± 0.01 5.2 ± 0.6  0.13 ± 0.01  2.3 ± 0.5  1.2 ± 0.3  6.1 ± 1.2  2.0 ± 0.3  BCg  37 ± 6  4.7 ± 0.1  loamy sand  2.8 ± 0.3  1.4 ± 0.2 0.07 ± 0.01 4.9 ± 0.7  0.13 ± 0.02  2.3 ± 0.4  1.1 ± 0.4  5.3 ± 0.7  1.9 ± 0.2  Cg  n/a  4.8 ± 0.1  loamy sand  2.1 ± 0.3  0.9 ± 0.1 0.04 ± 0.01 3.7 ± 0.7  0.09 ± 0.01  1.8 ± 0.3  0.9 ± 0.3  5.3 ± 0.4  2.0 ± 0.1  1  Thickness of horizons in cm  2  Soil organic matter (SOM) concentration, expressed as mass %  3  Soil organic carbon and nitrogen concentration, expressed as mass %  4  Clay concentration (< 2μm), expressed as mass %  5  Gravimetric moisture content expressed on oven-dried soil basis  6  Organically-complexed (pyrophosphate extractable) iron and aluminum concentration (g/kg)  7  Iron and aluminum associated with short-range order inorganic material, calculated as the difference between oxalate and pyrophosphate extracts (g/kg)  25  C and N stocks distribution in the profile Table 2.2 shows estimates of the SOC and N stocks distribution in an average, 100 cm deep profile. Soil organic C and N contents were calculated using SOC and N concentrations, horizon thickness and estimated bulk density. In the mineral soil, the estimates were adjusted for coarse fragment content assuming a particle density of 2.65 g/cm3. Table 2.2: C and N stocks estimates (mean ± SEM) in an average control profile of 100 cm thickness (n=9) CARBON Profile portion  NITROGEN  Average stock  % of total  Average stock  % of total  (kg/m2)  stock  (kg/m2)  stock  Entire profile  15.9 ± 1.3  100  0.66 ± 0.06  100  Organic layer  3.6 ± 0.5  23  0.10 ± 0.01  16  Mineral soil  12.3 ± 1.3  77  0.55 ± 0.07  84  Ae  1.1 ± 0.2  7  0.04 ± 0.01  6  Bf1  2.6 ± 0.7  16  0.10 ± 0.03  15  Bf2  3.7 ± 0.6  24  0.18 ± 0.03  28  BC  4.8 ± 0.6  30  0.23 ± 0.03  35  Top 10 cm of profile  4.3 ± 0.3  27  0.13 ± 0.01  21  Top 20 cm of profile  6.1 ± 0.3  38  0.18 ± 0.01  28  Bulk density estimates Bulk density was not measured. The large proportion of coarse fragments and small scale variability makes accurate field determination of forest soil bulk density problematic (Curtis and Post, 1964). Fortunately, bulk density is strongly correlated to soil organic matter, and SOM concentration can be used to obtain estimates of bulk density (Curtis and Post, 1964, Federer et al., 1993, Prevost, 2004, Périé and Ouimet, 2008, Ruehlmann and Korschens, 2009, Baritz et al., 2010). Huntington (1989) observed that pedotransfer functions obtained in different studies are relatively consistent, which supports the estimation of bulk density from organic matter measurement to calculate estimates of SOC stocks. 26  To estimate bulk density we used the empirical regression equation (1) developed by Heuscher et al. (2005) using 687 samples from the Orthod soil suborder. The regression equation (1) is based on SOC concentration and depth of horizon, and takes into account the increase in bulk density with increasing soil depth due to overburden. (1)  BD = 1.780 – 0.379 SOC1/2 + 0.00123 depth  This equation was developed in mineral soil samples and is not appropriate for the organic layer. For the forest floor we used a pedotransfer function based on the organic density concept developed by Federer et al. (1993) (2): BD min × BD org  (2)  BD =  With  BD: soil bulk density  SOM × BD min + (1 - SOM) × BD org  BDmin: bulk density of pure organic fraction BDorg: bulk density of pure mineral fraction SOC: soil organic carbon concentration SOM: soil organic matter concentration For BDmin and BDorg we used the values reported by Périé and Ouimet (2008) of BDorg = 0.11 and BDmin = 1.77. These values are in good agreement with other studies conducted in medium to coarse textured acid forest soils (Federer et al., 1993, Prevost, 2004). Estimated bulk density averaged 0.15 g/cm3 in the forest floor. This value is within the range of expected values for LFH horizons in Canada of 0.11 – 0.15 g/cm3 (Shaw et al., 2005). Average estimated bulk density in mineral horizons ranged from 1.25 g/cm3 in the Bf1 horizon to 1.42 g/cm3 in the BC horizon. C and N distribution in the profile Estimated C stocks in the Roberts Creek soil profiles ranged from 11.7 to 25.1 kg C / m2 over a 100-cm depth, with an average of 15.9 kg C / m2 or 159 tons C / ha (Table 2.2). These values are generally higher than those reported for well-drained inland Canadian forests. Bhatti et al. (2002) reported an average of 10.9 kg C /m2 in boreal forest soils of central Canada, while Bois et al. 27  (2009) reported an average of 2.8 kg C / m2 in the forest floor and 8.8 kg C / m2 in the mineral soil of well-drained spruce – fir forests of central British Columbia. The humid climate and high primary productivity of coastal forests may contribute to the relatively high C accumulation on west coast forest soils. Previous studies have also reported relatively high C stocks on the coast of Washington and Oregon states compared to inland forests (Smithwick et al., 2002, Homann et al., 2005). Despite its high SOC concentration (38%, Table 2.1), the organic layer only accounted for about one quarter of the C stock in Roberts Creek soils (Table 2.2). The majority of SOC was found in the mineral soil (12.3 kg C / m2 on average). The first Podzolic horizon (Bf1) had the highest C concentration and stored twice as much SOC as the A horizon, but its contribution to the total SOC stock was eclipsed by the contribution of the deeper horizons, Bf2 and BC. Although the SOC concentration in these horizons was lower, their thickness resulted in a C store of more than half of the total SOC present in the profile. Poor drainage in the BC horizon may restrict the activity of decomposer for parts of the year, and may contribute to the retention of a sizable C pool in deeper horizons. Nitrogen stocks generally follow a depth distribution pattern similar to C stocks (Table 2.2). The main difference was that the forest floor stored relatively less N (16% of total N stocks), while deeper mineral horizons (Bf2 and BC) stored relatively more N (two third of total N stock). The N content of Roberts Creek soils was in good agreement with the observations conducted by Gessel et al. (1973) on soils of similar parent material in western Washington. It has previously been proposed that temperate to boreal forests store the majority of their SOC in the forest floor and / or the topmost mineral horizon (Jenny, 1950). When considering the humid temperate forest subset, Jobbagy and Jackson (2000) reported that 54% of SOC on average is stored in the upper 20 cm of the profile. The comparatively small proportion of SOC accounted for by the forest floor and top mineral horizons in Roberts Creek may be an indication of effective translocation of organic matter to the mineral subsoil through roots and illuviation (Sanborn and Lavkulich, 1989, Buurman and Jongmans, 2005). Mixing by soil animals was not an important process in these soils, as evidenced by the low organic C concentration in the A horizon and the mor humus type. 28  Studies of SOC commonly focus on the upper 10 to 20 cm of the soil profile (e.g. Covington, 1981, Olsson et al., 1996b, Tremblay et al., 2006, Kranabetter, 2009). In our soils, such approaches would account for less than a half of our SOC stocks and less than a third of N stocks. Since bulk density was estimated from depth and SOM concentration rather than being directly measured, the C and N stocks reported here are subject to some error. De Vos et al. (2005) reviewed the predictive quality of 12 models for bulk density estimation, and found that all models produced underestimates of field bulk density. Underestimation error was up to 9 - 36% (Boucneau et al., 1998, De Vos et al., 2005). It is thus likely that the estimates of C and N presented here are conservative estimates. SOM and size fractions Size fractionation is useful for separating SOM pools with different biogeochemical characteristics, functions and dynamics (Anderson et al., 1981, Kögel-Knabner et al., 2008, Xu et al., 2009). Table 2.3 shows the distribution of SOM in sand, silt and clay size fractions of the mineral soil. Organic matter concentration in sand, silt and clay was simply estimated by loss-onignition performed on the size fractions separated by a combination of wet sieving and sedimentation for the purpose of textural determination. Organic matter associated disproportionally with the clay fraction. The clay fraction only represented ~5% of the soil but accounted for ~ 1/3 of the organic matter pool (Table 2.3). This may be due to the fact that a major pathway for SOC to enter mineral horizons in Podzols is via dissolved organic matter precipitating on reactive mineral surfaces (Kaiser et al., 2002, Rumpel et al., 2004, Kalbitz and Kaiser, 2008). The correlation between SOC and the clay fraction (r2 = 0.20, p < .0001) was stronger than the correlation between SOC and the silt + clay fraction (r2 = 0.04, p = 0.02). There was a strong correlation between clay-sized SOM and % soil clay, whereas sand- and siltsized organic fractions show only weak correlation with soil clay and no correlation with sand and silt-size fractions (Fig. 2.1). This indicates that interactions between mineral and organic fractions were largely restricted to clay-sized material (Zinn et al., 2007). Clay-sized SOM is 29  likely retained by sorption to clay phases, whereas particulate organic matter accumulation mechanisms appear largely independent of the mineral phase (Kaiser et al., 2002). Table 2.3: Particle size distribution in control plots (n = 9) and association of soil organic matter (SOM) with different size fractions (mean ± SEM) Sand  Silt  C lay  SOM fr action  SOM fr action  SOM fr action  C lay-sized  associated  associated  associated  SOM : clay  with sand  with silt  with clay  r atio  ---- g / 100 g mineral soil ----  -------------------- g / 100g OM --------------------  g / 100g clay  Ae  69 ± 2  27 ± 2  3±0  57 ± 3  17 ± 1  26 ± 3  20 ± 1  B f1  75 ± 2  20 ± 2  5±1  54 ± 2  16 ± 1  30 ± 2  24 ± 3  B f2  76 ± 3  20 ± 3  5±1  48 ± 5  17 ± 1  35 ± 6  21 ± 2  BC  78 ± 3  17 ± 2  5±1  51 ± 4  17 ± 2  32 ± 4  22 ± 3  over all 75 ± 1  20 ± 1  5±0  51 ± 2  17 ± 1  31 ± 2  22 ± 1  Figure 2.1: Correlation of organic matter size fractions with corresponding mineral size fractions. The r2and p-value are associated with the linear regression between clay-sized organic matter and % clay. Regressions for sand and silt-size organic matter are not significant.  30  The last column of Table 2.3 gives the ratio of clay-sized SOM to clay (SOMclay:clay). The ratio was relatively constant, averaging 21.6 g SOM / 100g clay. This may suggest that a given amount of clay stabilizes a generally fixed amount of clay-sized organic matter, and that the clay fraction is saturated with organic matter (Six et al., 2002). A consequence of this hypothesis is that additional inputs of organic matter would only accumulate in labile soil C pools with a relatively fast turnover rate (Gulde et al., 2008). An exception to the constant SOMclay:clay ratio may occur in Bf1, where this ratio is slightly higher (albeit not significantly so) than in other horizons. The Bf1 horizon is rich in free Fe and Al and SRO material (Grand and Lavkulich, 2008) with a greater capacity for SOM stabilization (Torn et al., 1997, Basile-Doelsch et al., 2007). The maximum amount of organic matter that may be stabilized in Bf1 could thus increase (Six et al., 2002). These results are based on loss-on-ignition. As a result, clay-sized SOM fraction may be overestimated due to dehydration of clays, but the error cannot explain the overwhelming association of SOM with clay. Despite the coarse texture of Roberts Creek soils, organo-mineral interactions are believed to have a major influence on SOC accumulation. C:N and C:SOM ratios The C:N and C:SOM ratios are indicators of the composition of SOM. C:N ratio The total N content showed a strong linear association with SOC content. Pearson’s correlation coefficients were 0.86 (p = 0.003) and 0.90 (p < 0.0001) for the organic (FH) and mineral layers of control plots, respectively. The C:N ratio was generally wide (Table 2.4), reflecting the low N content of coniferous litter (McGroddy et al., 2004) and relatively slow transformation of organic matter. The C:N ratio was highest in the FH layer with an average of 36, and decreased with depth to reach a value of 22 in lower horizons (Table 2.4). This indicates that the organic matter at depth showed a higher degree of transformation (Norris et al., 2011).  31  Table 2.4: C ratios in control plots (n = 9). Means are given ± SEM and are not significantly different between horizons at the 0.1 level if followed by the same letter. Horizon C:SOM1  C:N2  C:Alp3  C:Fep3  C:(Al+Fe) p3  Alp: (Al+Fe)p4  (%)  (%)  FH  53.0 ± 1.2 b  36.2 ± 1.1 a 359.9 ± 63.8 a 948.6 ± 208.6 a 253.9 ± 44.7 a 72.0 ± 3.0 a  Ae  60.3 ± 4.4 a  29.3 ± 2.4 b 33.7 ± 4.1 b  Bf1  46.5 ± 12.0 b  18.8 ± 2.9 b  55.4 ± 2.5 c  55.2 ± 3.2 b 27.4 ± 2.1 b 7.7 ± 0.6 c  15.0 ± 2.0 c  5.0 ± 0.4 c  65.2 ± 2.4 b  Bf2  45.3 ± 3.7 c  22.2 ± 2.1 c 7.3 ± 0.5 c  17.8 ± 2.4 c  5.1 ± 0.5 c  69.1 ± 2.5 ab  BC  46.3 ± 2.2 c 22.5 ± 1.6 c 6.0 ± 0.4 c  18.7 ± 3.8 c  4.7 ± 0.4 c  71.5 ± 2.3 a  1  C concentration of soil organic matter (mass %)  2  C to N ratio (mass basis)  3  C to pyrophosphate-extractable Al and Fe ratios (mass basis)  4  proportion of Al in the pyrophosphate-extractable metals (mass %)  Even though the C:N ratio reached lower values in the subsoil, N availability need not necessarily improve. It is generally observed that undisturbed northern temperate forests are N limited (Keeney, 1980, Vitousek and Howarth, 1991) and N mineralization usually declines with increasing depth (Federer, 1983). A study of N availability by Garcia-Pausas et al. (2008) showed that although C:N ratios decreased with depth, N recalcitrance tended to also increase with depth. Even in the presence of near-optimal C:N ratios, N availability may be a limiting factor for plant growth and organic matter decomposition at depth. C:SOM ratio We estimated the C concentration of SOM by dividing %SOC obtained by an induction furnace by loss-on-ignition values. Organic matter contained 49% of organic C on average, yielding a conversion factor between values of SOC and SOM of 2.080. This value is significantly different from the Van Bemmelen factor of 1.724. The C concentration of organic matter generally decreased with depth, ranging from ~ 60% in the top mineral horizon (Ae) to ~ 45% in deep mineral horizons (Table 2.4). This precludes the use of any single factor to convert between SOM and SOC.  32  Organic matter of the Ae horizon had a significantly higher C concentration than SOM of adjacent horizons (Table 2.4). This indicates that organic matter dynamic in the Ae horizon is different from lower horizons and supports the proposition of Rumpel et al. (2004), who suggested that Podzols act as ‘chromatographic systems’. In a ‘chromatographic system’, the Ahorizon preserves C-rich, long-chain alkyl structures with a hydrophobic tendency. Small, highly oxidized, polar organic compounds are preferentially removed by leaching and are retained in illuvial horizons by adsorption onto Al and Fe soil minerals (Eusterhues et al., 2003). Accordingly, the illuviated horizons of Roberts Creek Podzols showed a lower C:SOM ratio than the Ae horizon.  Figure 2.2: Linear regression between soil organic carbon (SOC) and C concentration of soil organic matter (C:SOM) of control plots In mineral horizons, the C concentration of SOM increased as % SOC increased (Fig. 2.2). Samples rich in organic matter had a wide C:SOM ratio. This is consistent with the idea that organic-rich samples contained high amount of recalcitrant, high-C compounds. Comparatively organic-poor samples contained organic matter with less C, which may consist in a combination 33  of simple organic molecules such as amino acids and of more complex decomposition products such as microbial metabolites that have lost C through respiration processes. Pyrophosphate–extractable Fe and Al Correlation between SOC and Alp, Fep Pyrophosphate-extractable Al and Fe showed a strong correlation with SOC in the mineral soil (Fig. 2.3), indicating that Fep and Alp are strongly associated with SOM and that poorly crystalline hydroxide phases are unlikely to have contributed significantly to the pyrophosphate extracts (Kaiser and Zech, 1996).  Figure 2.3: Relationship between soil organic carbon (SOC) and pyrophosphate-extractable Al and Fe (Alp+Fep) in the mineral soil. Similar relationships are observed for Alp and Fep considered individually.  34  Figure 2.4: Relationship between soil organic carbon (SOC) and pyrophosphate-extractabe Al (Alp) and Fe (Fep) in the FH layer. As expected in the organic layer, the relationship between C and pyrophosphate extractable metals was weak. Contrary to what was observed in the mineral horizons, the correlation was negative (Fig. 2.4), indicating that in the FH layer the amount of metal available to form complexes with humus was limiting. Samples high in organics and with a low mineral content resulted in the formation of fewest organo-mineral complexes. Evidence for upward translocation of Al The negative correlation between pyrophosphate extractable metals and SOC in the organic soil was more pronounced for Fep. Although the slopes of the regression lines were similar for both metals (p = 0.80 in Wilk’s Lambda multivariate test), the regression line associated with Alp was not statistically significant due to the presence of high-Alp ‘outliers’ (Fig. 2.4). This suggests that there may be a mechanism for upward translocation of some Al into the FH horizon. Further indication for upward translocation of Al into FH was found by examining the proportion of Al in (Al+Fe)p (Table 2.4). In the mineral horizons, the contribution of Alp to (Al+Fe)p increased with depth. This is due to the increase in pH with depth. Jansen et al. (2003) showed that at pH 3.5, dissolved ‘free’ Al dominates, whereas at pH 4.5, Al-SOM complexes were more 35  common. On the other hand, complexation of Fe by SOM is similar at all pH (Nierop et al., 2002, Jansen et al., 2003), and as a result, at low pH SOM-Fe complexes should be more abundant than at high pH. Despite the low pH (3.6), the proportion of Alp averaged 72% in the forest floor, a value significantly higher than the top mineral horizons. This suggests the existence of a mechanism for Al enrichment of the FH horizon (Kaste et al., 2011). Lundström et al. (2000b) and Smits and Hoffland (2009) suggested that ectomycorrhizal fungi may play a role in upward transport of Al into organic horizons of Podzols. Whether a similar mechanism is at work in Roberts Creek soils deserves further investigation. The C:Al,Fep ratios Table 2.4 shows that both C:Alp and C:Fep tended to narrow with depth. This trend was more pronounced than the narrowing of C:N ratio, suggesting that metal enrichment is an active process. The C:metal ratios have been used to estimate the degree of Al and Fe saturation of humus (Matus et al., 2006). The C:Alp and C:Fep ratios were high in the FH layer, suggesting that organic matter complexation sites were not saturated with metals. These ratios decrease to intermediate values in the Ae horizon. In both the FH and Ae horizons, dissolved organic compounds percolating downward are not likely to precipitate from solution due to the relatively low metal saturation. Accordingly, the Ae horizon has a lower SOC concentration than the underlying Bf1 horizon. The C:Alp and C:Fep ratios were comparatively low in Bf-BC horizons. The C:Alp ratio was consistently below 10, suggesting that organic matter is very insoluble (Skjemstad, 1992). It is likely that in these horizons, formation of Al-SOM complexes played a role in the precipitation of dissolved organic matter and accumulation of stable SOM (Scheel et al., 2007). The C:(Al+Fe) p ratio was low (around 5) and comparatively constant in illuvial horizons (Table 2.4), suggesting that humus – metal complexation may have reached a maximum in the subsoil, and that most organic complexation sites are occupied.  36  PREDICTORS OF SOC We investigated the relationship between SOC and other soil variables using multiple regression analysis. Soil variables included pH, C:N ratio, % clay, moisture, Al and Fe associated with SRO inorganic material and organically-complexed Al and Fe. To avoid problems arising from multicollinearity in soil variables, we performed a factor analysis prior to multiple regression analysis (Kadono et al., 2009). The regression analysis was based on the orthogonal factors. Factors were obtained by factor analysis based on the correlation matrix followed by equamax orthogonal rotation. The equamax rotation aims to achieve a simple factor structure. The objectives of the factor analysis were to obtain a set of factors that (1) were orthogonal, eliminating interpretation problems linked to variable multicollinearity, and (2) showed a simple structure where each factor only loads on one or two variables, simplifying interpretation. Dimension reduction was not the objective of the factor analysis. Table 2.5 shows the correlation between original variables and orthogonal factors used in regression analysis of mineral horizons. Factors were named according to the variables that loaded on them. A separate set of factors was derived for the organic horizons, for which there was no oxalate and % clay data (Table 2.6). As shown in tables 2.5 and 2.6, the factor analysis successfully created orthogonal new variables or factors that correlated highly with only one, or in some instances two related initial variables. We then conducted a multiple regression analysis to determine which factors controlled SOC. We based the factor analysis on the correlation matrix so factors are standardized, meaning that the regression coefficients can be directly interpreted for sign and magnitude. The simple factor structure displayed in tables 2.5 and 2.6 allows us to confidently extend results obtained for factors to corresponding initial variables. In the following discussion the word ‘variable’ refers to the initial variables whereas ‘factor’ refers to the new set of variables constructed during factor analysis.  37  Table 2.5: Factor pattern for mineral horizons  1  pH moist2 Alp3 Fep3 CN4 Clay5 Alsro6 Fesro6  Fsro  Fp  Fmoist  Fclay  FpH  FCN  0.18 0.17 0.31 0.02 0.07 -0.03 0.91 0.90  -0.01 0.21 0.85 0.94 0.00 0.15 0.02 0.28  -0.07 0.89 0.32 0.19 -0.01 0.27 0.24 0.15  -0.03 0.34 0.15 0.22 0.10 0.94 -0.06 0.06  0.97 -0.10 0.05 -0.05 -0.10 -0.03 0.27 0.17  -0.11 -0.01 0.01 0.02 0.99 0.12 0.08 0.10  1.69  1.12  1.07  1.07  1.02  variance 1.86 explained  Table 2.6: Factor patterns for organic horizons Fmoist  FpH  FAlp  FCN  FFep  pH moist2 Alp3 Fep3 CN4  0.03 0.99 0.13 0.11 0.02  0.91 -0.02 0.26 0.34 -0.03  0.26 0.11 0.92 0.29 -0.05  -0.04 0.02 -0.06 -0.07 1.00  0.33 0.08 0.27 0.89 -0.05  variance explained  1.01  1.00  1.00  1.00  0.98  1  1  pH in 0.01M CaCl2  2  dry period gravimetric moisture content  3  pyrophosphate-extractable Al and Fe  4  carbon to nitrogen ratio  5  clay  6  Al and Fe associated with inorganic short range order material, calculated as oxalate-extractable Al and Fe minus  pyrophosphate-extractable Al and Fe  Organic layer results Regression results and factors significant at the α = 0.1 level in a t-test based on type III sum of squares are shown in table 2.7.  38  Table 2.7: Regression results summary for the organic layer showing significant independent variables and the associated coefficients. Dependent variable  % SOC  Independent variables  FFep, FAlp, Fmoist, FpH, FCN  Model R2 and significance  R2 = 0.42, p = 0.006  Significant factors (α = 0.1)  Coefficient  Intercept  34.39  FFep  -5.40  FCN  3.39  FFep and FCN were the only significant factors at the α = 0.1 level (Table 2.7), indicating that moisture, pH, and Alp were not important predictors of SOC in the organic layer. FFep was the most significant factor in the regression, but contrary to what was observed in the mineral horizons the coefficient was negative. As previously mentioned, this indicates that in the organic layer the amount of Fe available for adsorption by SOM was limited. This can be thought of as a ‘dilution’ effect, where larger amounts of SOM cause a dilution of the Fep pool. Alp did not show the same negative correlation with SOC as Fep, possibly due to the mechanism for upward translocation of Al into the FH layer previously discussed. The coefficient associated with FCN was positive, meaning that samples high in SOM had a wide C:N ratio. Wide C:N ratios are generally associated with low degrees of decomposition, and as mineralization progresses C:N ratios typically narrow as C is lost to the atmosphere. This suggests that FH samples with high SOC tended to have large amounts of fresh, undecomposed organic matter. The relatively low importance of pH and moisture as a determinant of SOC accumulation between profiles may be the result of the narrow range of environmental conditions covered in this study, where all soils are well-drained, highly leached and acidic.  39  Mineral horizons results Regression results and factors significant at the α = 0.1 level in a t-test based on type III sum of squares are shown in Table 2.8. All factors entered in the model are significant at the α = 0.05 level, suggesting that all of their associated variables have an effect on SOC. Fp was the most important predictors of SOC (Table 2.8). Fp represents a linear combination of Alp and Fep with minor contributions from other variables (Table 2.5). This indicates that organically complexed Al and Fe were major determinants of SOC. The coefficient was positive, suggesting that Alp and Fep have a stabilizing influence on SOC, in agreement with the literature (Skjemstad, 1992, Rasmussen et al., 2006, Zanelli et al., 2006, Scheel et al., 2007, Egli et al., 2008). Table 2.8: Regression results summary for mineral horizons showing significant independent variables and the associated coefficients. Dependent variable  % SOC  Independent variables  Fsro, Fp, Fmoist, Fclay, FpH, FCN  Model R2 and significance  R2 = 0.76, p < 0.0001  Significant factors (α = 0.1)  Coefficient  Intercept  1.77  Fp  0.62  Fmoist  0.35  Fsro  0.32  Fclay  0.20  FpH  -0.19  FCN  0.10  The coefficient of Fsro was also positive (Table 2.8) and similarly suggests that Fe and Al associated with inorganic SRO material contributed to SOC retention (Eusterhues et al., 2005, Kleber et al., 2005, Mikutta et al., 2005a). Fclay was also a significant predictor of SOC, even in the presence of Fp and Fsro. This suggests that crystalline clays played a role in organic matter accumulation. 40  Other factors that correlated significantly with SOC include Fmoist, FCN and FpH (Table 2.8). Excess moisture can impede microbial activity and decrease SOM decomposition. However, Roberts Creek soils are coarse and well drained, and aeration is not believed to be a limiting factor of SOM decomposition in the solum. The positive correlation between SOC and moisture is better explained by the water retention capacity of SOM. The main function of FpH and FCN in the model was to differentiate between horizons. These factors were not significant when regression analysis was applied to any individual horizon. The coefficient associated with FpH was negative and the one associated with FCN was positive, indicating that horizons high in organic matter had a low pH and a high C:N ratio. Determinants of SOC retention in Roberts Creek Roberts Creek soils have a coarse texture and few aggregates. Stabilization of SOM due to physical protection by aggregates is likely to be limited. As a result, sorption of SOC to mineral surfaces and metals is likely to play an important role in SOM retention and stabilization (Baldock and Skjemstad, 2000, Eusterhues et al., 2005). Organically complexed metals, SRO inorganic phases and clay content were all identified as significant predictors of SOC by regression analysis, suggesting that these three matrix components contribute to SOC retention. Taken together they explained 58% of the total variance in SOC concentration. Both Fp and Fsro were much stronger predictors of SOC than Fclay. This agrees well with the weight of published evidence, which shows the prevailing role of reactive Al and Fe phases for controlling SOC storage (Torn et al., 1997, Kleber et al., 2005, Mikutta et al., 2005a, Kothawala et al., 2009). It can be concluded that poorly crystalline phases and polyvalent metals are largely responsible for mineral-mediated SOC retention in Roberts Creek. Crystalline clay minerals may be significant for SOC retention in the Ae horizon, where the amount of reactive Al and Fe phases is limited (Eusterhues et al., 2003). The coefficient associated with Fsro was roughly half of the coefficient associated with Fp, indicating that organically complexed Al and Fe were stronger predictors of SOC than SRO inorganic material. However, our data does not allow for an assessment of the relative role of organic-metallic interactions and SRO phases for SOC storage. We only measured total SOC, 41  which is not necessarily a good indicator of organic matter stability. Kadono et al. (2009) partitioned soil C into a potentially mineralizable and recalcitrant pool, and showed that oxalateextractable metals correlated only with recalcitrant soil C. Therefore, SRO inorganic material could be a major determinant of SOC recalcitrance and stability while showing a moderate correlation to total SOC. When regression analyses are conducted on individual horizons, the coefficient associated with Fp decreased with depth while the coefficient associated with Fsro increased, suggesting that the significance of inorganic SRO material for SOC stabilization increases with depth.  SUMMARY AND CONCLUSIONS C amount and distribution This chapter estimated the amount and distribution of SOC in a coastal forest of British Columbia. The average soil profile stored approximately 15.9 kg C / m2, which is generally higher than SOC stocks estimated for well-drained inland Canadian forests. The humid climate and high primary productivity of coastal forests may contribute to the relatively high SOC accumulation. The organic layer (LFH) only accounted for about one quarter of the C stock. The upper 10 cm of the profile stored less than one third of SOC and the upper 20 cm stored about 40%. The relatively small proportion of SOC accounted for by the forest floor and top mineral horizons may be an indication of effective translocation of organic matter to the mineral subsoil through roots and illuviation processes. Field studies restricted to the upper part of the soil profile would have much reduced interest and significance since they would overlook a major portion of SOC stocks.  42  Organic matter characteristics Organic matter characteristics evolved with depth. Organic matter found deeper in the profile had a lower C concentration and narrower C:N, C:Alp, C:Fep ratios. The C:Alp and C:Fep ratios declined more rapidly than C:N and C:SOM ratios, suggesting that organic matter was actively enriched with metals. The Ae horizon had a higher C:SOM ratio than adjacent horizons, suggesting a preferential retention of C-rich compounds such as long-chain, hydrophobic alkyl structures. Size fractions Despite the coarse texture, organic matter associated overwhelmingly with the clay fraction. The clay fraction represented less than 5% of the fine earth fraction but contained 32% of SOM, and is likely to have a disproportionate influence on SOM accumulation. The ratio of clay-sized SOM to clay was relatively constant, suggesting that a given amount of clay stabilizes a generally fixed amount of clay-sized organic matter, and that the clay fraction may be saturated with organic matter. Predictors of SOC This study identified the most important predictors of SOC in soils of Roberts Creek using a factor analysis followed by multiple regression analysis. This approach was successful as the orthogonalization step eliminated the problem of model instability due to variable multicollinearity. It produced easily interpretable regression equations that respected correlations present between SOC and original, untransformed variables. Processes determining SOC concentration were very different in the organic layer and mineral horizons. In the mineral horizons, clay content, organically complexed Al and Fe and SRO inorganic material were significant predictors of SOC, pointing to a mineral control of SOM retention. The most important determinants of SOC were reactive Al and Fe forms. Crystalline clays also appeared to play a role in organic matter accumulation, but were not as important as poorly crystalline compounds in determining % SOC.  43  In the organic layer, Fep correlated negatively with SOC, indicating that the amount of Fe available for adsorption by SOM was limited. Alp did not show the same negative association, probably due to a mechanism for upward translocation of Al into the FH horizon and possibly involving ectomyccorhizal fungi. Given the importance of Al in soil processes including organic matter retention, studies establishing the exact role of fungi in Al translocation would be of great interest.  44  CHAPTER THREE  Soil organic carbon and related variables:  Effects of logging  45  SYNOPSIS Knowledge about soil organic carbon (SOC) response to logging in conifer forests is limited. The objective of this study is to determine the short to medium term effects of clear-cut logging on SOC in a Douglas-fir-dominated forest of southern coastal British Columbia. We sampled a chronosequence comprising undisturbed (control) forested plots, recently logged (cleared) plots harvested 1 to 5 years prior to sampling, and older logged (regenerating) plots harvested 8 to 15 years prior to sampling. We collected soil samples from each pedogenic horizon (FH, Ae, Bf1, Bf2, BC) in 27 Podzolic profiles and measured SOC, pH, texture, moisture content, total nitrogen, loss-on-ignition, sum of exchangeable cations (CECe), and pyrophosphate extractable Fe and Al. We found that the mineral subsoil (Bf and BC horizons) played a key role in the overall response of SOC storage after logging. The entire profile must therefore be taken into consideration when studying the effects of disturbance. Carbon stocks in the forest floor were higher in logged plots than in control, probably due to gradual incorporation of decaying logging slash. In the mineral soil, SOC was higher in cleared plots and similar to control levels in regenerating plots. The increase in SOC in cleared plots occurred mainly in the sand and silt size fractions. This suggests that logging resulted in additional soil organic matter (SOM) inputs to the mineral soil, and that these inputs were not readily stabilized. Logging affected organic matter quality, as evidenced by decreases in the C:N and C:SOM ratios, and increases in the CECe:C ratio. Organic matter in regenerating plots showed indication of increased oxidation and maturity. The C:N ratio narrowed, indicating that logged plots have the potential to produce nitrate for several years. The CECe:C ratio is an indicator of environmental performance of the organic matter and increased in regenerating plots, suggesting SOM in regenerating plots has a high density of functional groups which could help retain nutrients on site. Overall SOC stocks did not vary significantly after logging. We propose that large inputs of logging slash coupled with relatively slow decomposition rates prevented C stocks from decreasing significantly for at least 15 years post harvest. However, our data suggest that steady 46  state is not re-established at 15 years after logging, and that older plots may show a decline in SOC and N stocks. In particular, the proportion of SOC found in the forest floor increased from 25% in control plots to 45% in regenerating plots. Forest floor SOC is not protected by interactions with minerals and may be more susceptible to decomposition than subsoil SOM.  INTRODUCTION Background Soil organic matter (SOM) influences soil fertility, long term productivity and soil development. In Podzols, the translocation of organic material and associated metals is one of the chief soilforming processes (Lundström et al., 2000b). Disturbances that alter the organic matter cycle have a wide range of effects because SOM influences so many biogeochemical processes. Understanding the dynamics of organic matter and associated variables during disturbance is critical to the preservation of key soil processes and ecosystem health. Recently, soil organic carbon (SOC) stocks and dynamics have received increased interest due to the possibility of using soils to sequester C. Under the Kyoto protocol, managed forest lands may be taken into account in national C budgets (Kurz and Apps, 2006). Therefore, local scale studies establishing baseline C stocks and their response to forest disturbance and regeneration are needed to allow accurate accounting and effective C management. Distribution of C stocks The forest industry provides information on forest stand volume. Remote sensing methods have been developed to measure biomass (Mickler et al., 2002), but cannot account for SOC. Generally, knowledge about C stocks in the forest floor and mineral soil, and the effects of forest management on each of these components is limited (Nalder and Merriam, 1995, Kurz et al., 2002). Information on soil C dynamics is especially scarce below the top 10 to 20 cm of the soil profile. The classic studies of logging effects on C and N content conducted by Covington 47  (1981) and Federer (1984) limited their scope to the forest floor. Other studies investigated the mineral soil response but restricted sampling to the topmost part of the mineral soil (e.g. Bock and Van Rees (2002) to 10 cm depth; Hendrickson et al. (1989), Borchers and Perry (1992) and Olsson et al. (1996b) to 20 cm depth; Parker et al. (2001) to 25 cm depth). Recently, several authors have emphasized the need to consider subsoil layers when evaluating C stocks and dynamics (Garcia-Pausas et al., 2008, Basile-Doelsch et al., 2009, Diochon et al., 2009, Knops and Bradley, 2009). Studies that investigate the entire soil profile generally report no effect of logging in the soil parent material or C horizon (Snyder and Harter, 1985), but any part of the solum may show treatment effects. Diochon and Kellman (2009) presented data consistent with a destabilization of the soil C pool and increased rates of decomposition in the deeper mineral soil (20+ cm) after logging, and proposed that concentrations of C below 20 cm may be driving the temporal response of soil C storage after logging. Indeed, even moderate changes in SOC distribution and dynamics in deeper horizons have the potential to strongly influence the overall SOC balance due to the large quantities of SOC at stake. Collecting data on C stocks in the entire soil profile is also essential to differentiate between net changes of soil C and simple translocation (Federer, 1984), as redistribution may be an important mechanism by which SOC is conserved in forest soil after disturbance (Hendrickson et al., 1989, Yanai et al., 2003). It is critical to establish whether SOC and N are simply translocated to the mineral soil or lost from the system, if any claim concerning greenhouse gas balance or longterm forest productivity is to be made (Yanai et al., 2003). Logging and C stocks Many models of the effects of forest harvesting predict a reduction of soil C stocks, reaching a minimum at about 10 – 20 years after logging, followed by a partial or complete recovery period during which soil C stocks increase (Aber et al., 1979, Covington, 1981, Jiang et al., 2002). In some models, harvesting is associated with a short-lived increase in soil C stocks as a result of increased inputs of above and below-ground biomass (Bengtsson and Wikstrom, 1993, Jiang et al., 2002).  48  Notwithstanding the usefulness of such models for the generation of present and future regional estimates, empirical data that confirm these models is relatively scarce (Yanai et al., 2003). Local effects such as logging type, harvest technology, site history, forest type, and climate greatly influence ecosystem response to disturbance. Soil properties also have the potential to exert a strong influence on SOC stocks and dynamics. Carbon retention is influenced by area or sitespecific characteristics such as pH (Nierop and Verstraten, 2003), moisture (Londo et al., 1999), N content (Moran et al., 2005), texture (Oades, 1988), and organo-metallic and organo-mineral interactions (Mikutta et al., 2005a, Rasmussen et al., 2005). Forest floor Logging can result in either an increase or decrease in forest floor C stocks, depending on how much logging slash is left behind (Johnson, 1992). In their recent meta-analysis, Nave et al. (2010) demonstrated that the impacts of logging on the forest floor are generally smaller in coniferous stands than in hardwoods. This could be due to slower decomposition due to litter recalcitrance and low temperatures (Johnson, 1995), and to the rapid resumption of carbon accumulation and limited nutrient loss after disturbance in conifer stands (Gholz and Fisher, 1982). Mineral soil In the mineral soil, short to medium term increases in C content are common. In northern hardwood forests, Johnson (1995) reported data consistent with moderate increases in C and N stocks at 3 years after logging in mineral horizons. In sub-boreal to boreal mixed woods, Hendrickson (1989) observed a mild increase in the top 20 cm of mineral soil in 3 years old stands. In coniferous forests, Gholz and Fisher (1982) found that Podzolic horizon Bh showed a maximum C content at 5 years after logging. Olsson et al. (1996b) observed increased C amounts in the mineral soil of both conventional stem and whole tree harvested plots. Increases in C and N stocks after logging are not generally emphasized in the literature, as many authors tend to focus their discussion on SOM losses rather than gains (Hendrickson et al., 1989, Johnson, 1995). The simplest explanation for the increase in mineral soil organic C content observed after logging is the mixing of the organic layer into the mineral horizons as a result of mechanical 49  disturbance (Federer, 1982, Gholz and Fisher, 1982, Ryan et al., 1992). Translocation of dissolved organic matter is another process that can increase mineral carbon stocks (Snyder and Harter, 1985, Kalbitz and Kaiser, 2008). Dissolved organic matter percolates from the forest floor down to subsoil horizons and plays an important role in the accumulation of organic matter in mineral soils of coniferous forests (Don and Kalbitz, 2005, Kalbitz and Kaiser, 2008). Increased concentrations and fluxes of dissolved organic matter are generally observed after clear-cutting (Dai et al., 2001, Diochon and Kellman, 2009, Morris, 2009). Kalbitz et al. (2004) established that logging leads to increased fluxes of dissolved organic matter from the forest floor to the mineral soil, where it has greater resistance to microbial decay. Abundant fresh organic matter inputs and increased microbial activity are a major determinant in this increase in DOC fluxes (Kalbitz et al., 2004). Additionally, the increase in effective precipitation that results from canopy removal may result in the intensification of organic matter eluviation – illuviation process (Borken et al., 2011). This mechanism has been invoked to explain differences in C concentration in mineral soils across precipitation gradients (de Wit et al., 2006). Hendrickson et al. (1989) showed that the ionic strength of precipitation reaching the forest floor decreases sharply after clear-cutting. This in turn intensifies the release of DOC from the forest floor and its migration into mineral soil (Evans et al., 1988). This mechanism has the potential to reduce post-harvest SOM losses (Hendrickson et al., 1989, Kalbitz et al., 2004). The increase in C content of the mineral soil observed after logging may be relatively shortlived. In hardwood stands, Johnson (1995) reported moderate increases in C and N stocks at 3 years after logging followed by decreases below baseline levels 8 years following logging. In mixedwoods, Pennock and Van Kessel (1997) observed increases in C levels in plots logged 1-3 years prior to sampling and decreases in plots logged 5-20 years prior to sampling. Long-term studies generally report a decline in mineral soil C content, followed by restoration of baseline stocks. Diochon et al. (2009) surveyed red spruce forest and found that storage of soil C reached a minimum at 32 years post-harvest, at which time C stores had declined by 50%. The data was consistent with a destabilization of the soil C pool and increased rates of decomposition in the deeper mineral soil (20+ cm) after logging (Diochon and Kellman, 2009).  50  The effects of logging are often difficult to detect. A number of studies failed to detect changes in soil C stocks after logging, despite dramatic changes in litter inputs (Lee et al., 2002). The lack of net logging effects may reflect a balance between competing processes such as fresh organic matter inputs, translocation, leaching and respiration losses (Hendrickson et al., 1989). Johnson et al. (1991a) found that the O-horizon mass and organic matter content increased after treatment, but that the total pool of organic matter in the solum did not change. They thus concluded that losses of organic matter via stream water and respiration were approximately balanced by inputs from decaying roots and litter. The ecosystem effect As previously mentioned, the best-studied ecosystem is that of the north eastern hardwood (Covington, 1981, Federer, 1984, Johnson, 1995, Dai et al., 2001). Relatively few studies have focused on coniferous forests, but there are indications that these forest soils are somewhat resilient to the effects of logging (Borchers and Perry, 1992, Parker et al., 2002, Chatterjee et al., 2009). Johnson (1995) included spruce-fir stands in their studies, and noted that the decrease in the forest floor after logging was much more moderate than in the hardwood stands, and that the C stocks in the mineral soil actually increased. They attributed this difference to slower decomposition processes due to litter recalcitrance and lower temperatures in the spruce-fir zone. In slash pine plantations, Gholz and Fisher (1982) noted that the profile C stocks remained relatively constant over time and that there was no apparent decline in total system carbon from 2 years onward, in contrast to what is usually observed in northern hardwoods. The author proposed that the rapid resumption of carbon accumulation in pine forests may limit nutrient loss after disturbance. Bois et al. (2009) surveyed sub-boreal spruce forests of central BC and found that forest floor and mineral soil C stocks did not correlate with stand age (ranging from 5 to 350+ years). The authors deducted that stands were resilient to partial and clear-cut harvesting systems in use in the area. Organic size fractions Beyond total SOC, monitoring the changes in specific SOM size fractions can provide insight into the dynamics of SOM after logging. Particle-size fractionation is a useful indicator of SOC  51  pools perturbations and of organic matter cycling (Borchers and Perry, 1992, Parker et al., 2002, Gartzia-Bengoetxea et al., 2009). Clay-sized SOM is considered to be the most stable fraction of SOM, with physical occlusion and the formation of complexes with mineral elements contributing to its stabilization (Paul, 1984, Oades, 1988, Sollins et al., 1996). In contrast, silt and sand-sized SOM fractions are considered to be more reactive than SOC associated with clay, due to weaker interactions with minerals (Tiessen and Stewart, 1983, Six et al., 2002). Because of the weak stabilization by minerals, most of the effects of logging are generally expected to be found in the sand and silt-sized SOM. However, Diochon and Kellman (2009) suggested that logging may lead to a destabilization of comparatively old organo-mineral complexes, which comprise most of the clay-sized SOM. Indicators of bulk SOM composition Bulk organic matter composition is another useful indicator of disruption in SOC cycling and SOM decomposition rates. The most commonly measured indicator of organic matter composition is the C:N ratio. The C:N ratio reflects differences in C and N net accumulation rates. A widening C:N ratio indicates that C net accumulation outpaces net N accumulation, while a narrowing C:N ratio is a sign that net C accumulation decreases relative to net N accumulation. Coniferous forests shed litter with a wide and relatively constant C:N ratio (McGroddy et al., 2004). The bulk of logging slash typically consists in coarse woody material and also tends to have a low N concentration. During the initial decomposition stage, the C:N ratio of fresh organic matter decreases as C is lost to the atmosphere (Johnson, 1995, Baldock and Skjemstad, 2000) and N immobilization dominates over mineralization (Keeney, 1980). Under the broad assumption that N inputs (atmospheric deposition and biological fixation) do not vary significantly, narrowing C:N ratios can be thought of as an indicator of SOM maturity and humification (John et al., 2005). A number of studies have reported that SOM undergoes intense decomposition and maturation after logging (Binkley, 1984, Bock and Van Rees, 2002, Hannam et al., 2005, Salmon et al., 2008). Londo et al. (1999) showed direct evidence that 52  harvesting significantly increased in situ respiration in a hardwood forest, while Diochon and Kellman (2009) presented isotopic evidence consistent with increased rates of decomposition after logging in coniferous forest. In contrast, Yanai et al. (2003) noted that decomposition rates of surface litter generally decreases after clear-cut logging due to inputs of highly recalcitrant material, but accelerated decomposition remain possible in the FH and mineral horizons. Prescott et al. (2000) reported that forest floor material lost mass at similar rates in forests and clearcuts, but pointed out that the response of decomposition to clear-cutting is highly variable and cannot be generalized. Another widely available indicator of SOM composition is the C concentration of organic matter (C:SOM ratio). A high C:SOM ratio suggests a predominance of C-rich, potentially hydrophobic compounds. Oxygen is the second most abundant element in SOM after C, such that a narrow C:SOM ratio is indicative of a higher degree of oxidation and a higher O content (Ussiri and Johnson, 2003). Oxygen-rich groups include functional groups such as carboxyl and phenolic groups (Johnson, 1995) and confer a general hydrophilic tendency to organic compounds. Objectives The aim of this study is to assess the effects of logging on SOC distribution and characteristics in the entire solum of our coastal British Columbia Podzolic soils. Using information on bulk SOM composition and size fractions, we make some inference about the dynamics of SOM after logging. We expect our conifer stands to be relatively resilient to SOC losses following harvest, and anticipate that the mineral subsoil plays a major role in C retention after logging. We further hypothesize that logging has a significant effect on bulk organic matter composition and that most of the changes take place in the silt and sand-sized organic matter pools, which are less stable than the clay-sized SOM.  53  C AND N CONCENTRATION SOC concentration Logging effects on SOC concentration were different in the forest floor and in the mineral soil. In the forest floor, SOC concentration was 20% lower in logged sites than in control stands, but this difference was not statistically significant (Table 3.1). In the illuvial (Bf-BC) horizons, SOC concentration was 40% higher in cleared than in control plots, but similar in regenerating and control plots (Fig. 3.1). Several authors (e.g. Tremblay et al., 2006, Kranabetter, 2009) have postulated that sampling the upper part of the soil profile is sufficient to study the effect of forest harvesting, but our results suggest that this approach may lead to under-estimates of the impacts of logging on C stocks by neglecting the response of the deeper mineral soil. The SOC concentration of illuvial horizons was similar in control and regenerating plots (Fig. 3.1), suggesting that the additional C present in cleared plots was not retained or that older C was metabolized. This is a surprising finding since Podzolic horizons are known to stabilize C by interaction with minerals and metals (Schmidt et al., 2000, Eusterhues et al., 2005, Kleber et al., 2005, Mikutta et al., 2006, Rasmussen et al., 2006, Scheel et al., 2007). A plausible explanation was put forward by Gholz and Fischer (1982), who proposed that the decrease in the C:N ratio following logging lead to rapid metabolism of new carbon sources. Diochon and Kellman (2009) presented data consistent with a destabilization of the soil C pool and increased rates of decomposition in the deeper mineral soil (20+ cm) after logging. In addition, organic matter generally interacts rather slowly with minerals (Parfitt, 2009). Buurman et al. (2007) suggested that mineral protection did not act on primary organic matter, which implies that fresh inputs of organic matter to the illuvial horizons are not necessarily stabilized.  54  Table 3.1: Mean ± standard error or the mean (SEM) of selected soil variables by soil layer and treatment. P-values at the end of each row document the statistical significance of treatment effect. Within each row, means followed by different letters are significantly different at the α = 0.1 level.  Soil property  Soil layer  ---------------------TREATMENT --------------------Control Cleared Regenerating n=9 n = 11 n=7  SOC1 (%)  FH Ae illuvial  38.1 ± 2.1 1.6 ± 0.2 1.7 ± 0.1 a  30.7 ± 4.7 1.7 ± 0.3 2.3 ± 0.2 b  29.2 ± 3.6 1.3 ± 0.2 1.6 ± 0.2 a  0.12 0.46 0.03  N1 (%)  FH Ae illuvial  1.06 ± 0.07 0.06 ± 0.01 0.08 ± 0.01  0.82 ± 0.12 0.08 ± 0.01 0.10 ± 0.01  1.03 ± 0.14 0.06 ± 0.01 0.07 ± 0.01  0.13 0.20 0.18  Clay-sized SOM2 (%)  Ae illuvial  0.80 ± 0.07 1.31 ± 0.214  0.81 ± 0.13 1.47 ± 0.15  0.60 ± 0.08 1.14 ± 0.14  0.16 0.65  Silt-sized SOM2 (%)  Ae illuvial  0.44 ± 0.03 0.65 ± 0.05 a  0.53 ± 0.05 0.86 ± 0.06 b  0.51 ± 0.09 0.59 ± 0.07 a  0.29 0.06  Sand-sized SOM2 (%)  Ae illuvial  1.47 ± 0.11 1.96 ± 0.13 a  1.34 ± 0.17 2.47 ± 0.18 b  1.11 ± 1.07 1.56 ± 0.17 c  0.18 0.004  C stock3 (kg/m2)  FH eluvial illuvial profile  3.6 ± 0.5 a 1.1 ± 0.2 11.1 ± 1.2 ab 15.9 ± 1.3  5.9 ± 1.2 ab 1.0 ± 0.2 13.1 ± 1.1 a 20.0 ± 1.5  7.6 ± 1.7 b 1.1 ± 0.3 8.5 ± 0.7 b 17.2 ± 1.8  0.05 0.89 0.01 0.21  N stock3 (kg/m2)  FH eluvial illuvial profile  0.10 ± 0.01 a 0.04 ± 0.01 0.51 ± 0.06 ab 0.66 ± 0.06  0.16 ± 0.03 a 0.05 ± 0.01 0.61 ± 0.07 a 0.83 ± 0.08  0.25 ± 0.04 b 0.05 ± 0.01 0.42 ± 0.04 b 0.73 ± 0.06  0.02 0.71 0.05 0.29  C:N4  FH Ae illuvial  36.2 ± 1.1 a 29.3 ± 2.4 a 24.2 ± 1.2  37.8 ± 4.5 a 21.3 ± 1.5 b 23.2 ± 1.0  29.0 ± 2.6 b 21.7 ± 2.3 b 21.2 ± 1.1  0.05 0.02 0.58  C : SOM 5 (% )  FH Ae illuvial  53.0 ± 1.2 a 60.3 ± 4.4 49.8 ± 2.0 a  54.1 ± 2.5 a 64.9 ± 4.8 50.6 ± 1.8 a  47.9 ± 1.6 b 59.1 ± 2.7 41.1 ± 1.8 b  0.05 0.60 0.02  N:SOM5 (%)  FH Ae illuvial  1.48 ± 0.05 a 2.33 ± 0.30 2.27 ± 0.15  1.53 ± 0.13 a 3.28 ± 0.40 2.25 ± 0.12  1.81 ± 0.10 b 2.87 ± 0.26 2.03 ± 0.12  0.05 0.19 0.67  p-value  55  Table 3.1 continued control  cleared  regenerating  p-value  FH Ae illuvial  1.10 ± 0.11 2.57 ± 0.38 0.92 ± 0.05 a  1.05 ± 0.17 3.12 ± 0.45 0.90 ± 0.09 a  1.42 ± 0.23 2.68 ± 0.17 1.30 ± 0.09 b  0.35 0.72 0.005  Moisture7  FH Ae illuvial  0.87 +/- 0.29 a 0.10 +/- 0.05 a 0.13 +/- 0.01 a  1.21 +/- 0.33 b 0.16 +/- 0.07 b 0.19 +/- 0.01 b  1.39 +/- 0.16 b 0.13 +/- 0.03 ab 0.13 +/- 0.01 a  0.01 0.05 0.09  (Al+Fe) p8  FH Ae illuvial  3.0 ± 1.5 1.1 ± 0.3 a 3.8 ± 0.4 a  3.4 ± 0.8 1.0 ± 0.2 a 5.0 ± 0.4 b  4.6 ± 2.2 0.5 ± 0.1 b 3.5 ± 0.5 a  0.54 0.03 0.02  253.9 ± 44.7 18.8 ± 2.9 a 5.0 ± 0.3  162.8 ± 50.1 18.5 ± 2.3 a 5.1 ± 0.3  142.0 ± 57.2 26.6 ± 1.9 b 4.6 ± 0.3  0.35 0.05 0.90  CECe: C  6  C:(Al+Fe) p9 FH Ae illuvial 1  Soil organic carbon and nitrogen concentration expressed as mass %  2  Clay, silt and sand-sized organic matter concentration, expressed as mass %  3  Carbon and nitrogen stocks in the humus (FH), eluvial (Ae), illuviated (Bf-BC) and entire profile (to 1m depth)  expressed in kg/m2 4  Carbon to nitrogen mass ratio  5  Carbon and nitrogen concentration of soil organic matter, expressed as mass %  6  Ratio of sum of exchangeable cations (molc) to soil organic carbon (kg)  7  Gravimetric moisture content expressed on oven-dried soil basis  8  Sum of organically-complexed (pyrophosphate extractable) iron and aluminum concentration (g/kg)  9  Mass ratio of carbon to organically-complexed iron and aluminum  56  Figure 3.1: Soil organic carbon (SOC) concentration in mineral horizons of control, cleared and regenerating plots. Points represent SOC means ± SEM.  N concentration Treatment effects on N concentration were similar to patterns observed for SOC, but were not statistically significant in any horizon (Table 3.1). In the forest floor, N concentration was 20% lower in cleared than in control plots. Contrary to SOC concentrations, N concentrations in control and regenerating plots were similar. In the mineral soil, N concentration was 1/3 higher in cleared than in control and regenerating stands.  57  C AND N STOCKS C stock Forest floor Logged sites typically have 20 cm of coarse woody debris on the surface consisting mostly in small logs and branches. The thickness of slash reached 40 cm in places. The forest floor (not including slash) increased in thickness from about 6 to 11 cm after logging, a significant increase of 75%. A simple explanation for the increase in organic layer thickness involves a combination of mineral matter mixing and inputs from the overlying slash layer, which may be considerable. Lee et al. (2002) calculated that harvesting operations in central Ontario mixedwoods may add up to 8.4 kg/m2 of organic matter to the forest floor. Fireweed (Epilobium angustifolium) became quickly established on logged plots and may have contributed some organic matter to the forest floor, but overall the contribution of primary succession species was likely small, since the thickness of slash left on site and low pH prevented the establishment of grasses and most other early seral species. Small amounts of mineral matter mix-in may have contributed additional material to the organic layer and could explain the slightly lower SOC concentration in logged plots. The increase in thickness of the organic layer caused the C and N stocks in the organic horizon to increase accordingly (Fig. 3.2), despite the small decrease in SOC concentration. Carbon stocks in the organic layer averaged 3.6 kg/m2 in control and 5.9 kg/m2 in cleared plots (Table 3.1). Carbon stocks were also high (7.6 kg/m2) in regenerating plots. The conversion of living biomass into detrital pools and subsequent incorporation into the forest floor is a likely cause for this increase. Mineral horizons The thickness of mineral horizons was not affected by logging. Consequently, C stocks followed the same pattern as SOC concentration (Table 3.1). Carbon stocks in the illuvial horizons were slightly higher in cleared than in control plots, and were lower in regenerating plots. 58  Mechanical disturbance was minimal in logged plots and mineral horizon boundaries generally showed not evidence of disturbance. The increase in C illuvial horizons was not likely to have been caused by incorporation of forest floor material, especially since the C content of the overlying Ae horizon was constant. A more likely explanation involves a temporary intensification of organic matter illuviation immediately after logging (Snyder and Harter, 1985, Kalbitz et al., 2004, Morris, 2009). Decomposing roots may also contribute SOM to the subsoil. In regenerating plots, the lower C stocks may be due to a decrease in illuviation rate and active decomposition of SOM. Overall profile Variations in total profile stocks were not found to be statistically significant (Table 3.1), due in part to the different behaviour of organic and mineral layers. Total profile stocks were 25% higher in cleared than in control plots (Fig. 3.2). Forest floors of regenerating plots had higher C stores than control or cleared plots, suggesting that the incorporation of logging residues still outweighed decomposition, leaching and translocation losses. This trend however did not compensate for losses of SOC from mineral horizons and total profile C stocks were 15% lower in regenerating than in cleared plots (Fig. 3.2). Changes were driven by variations in C stocks in the mineral sub-soil, in accordance with the proposition of Diochon and Kellman (2009). The longer term effect of logging was not investigated, so it is unknown whether the C stocks will keep declining below control levels after 15 years. The proportion of SOC present in the forest floor increased from 25% in control plots to 45% in regenerating plots. The increase was statistically significant (p-value = 0.01). The small proportion of SOC found in the forest floor in control plots suggests that nutrient cycling is tight in undisturbed soils, with rates of loss through decay and transfer approximately equal to those of gain from biomass (Simonson, 1959). In regenerating stands, a significantly larger proportion of soil C was present in the organic layer, even though the total SOC stocks remained similar to control stocks (Fig. 3.2). The variability of forest floor thickness also increased after logging, suggesting a disruption of previous balance between inputs and losses. Organic matter in the forest floor is more susceptible to degradation or mobilization due to the lack of protection afforded by metal complexation, phyllosilicates, free Al and Fe oxides, or short-range order  59  material such as allophane or imogolite. A long-term decline in profile C stocks is likely as unprotected organic matter from the forest floor decomposes (Nave et al., 2010).  Figure 3.2: Average C stocks in the soil profiles of control, cleared and regenerating plots The inability to detect statistically significant differences between treatments is common due to the high variability of forest soils. In addition, variability increased after logging. The variables SOC, SOM, C:SOM, N:SOM, and C:N all showed significant increase in variance in cleared plots as tested by Levene’s test for homogeneity of variance (test results not shown). Increased variability after logging is widespread (e.g. Huntington and Ryan, 1990, Xu et al., 2008, Bois et al., 2009) and impairs the ability to detect significant changes in C and N stocks and concentrations. N stock Nitrogen stocks follow the same trend as SOC stocks. In the forest floor, N stocks increased after logging, probably due to the accumulation of detrital organic matter. The increase was more pronounced than for C stocks. In regenerating stands, forest floor N stocks were 2.5 times higher than in control stands (Table 3.1).  60  Nitrogen stocks in Ae were relatively constant. In illuviated (Bf-BC) horizons, soil N stocks first showed a small increase, then decreased to values lower than control levels. The overall profile showed a good retention of pre-harvest N stocks. This contrasts with most studies of hardwood forests, which showed a significant decrease in soil N content for 5 – 15 years after clear-cutting (Covington, 1981, Federer, 1984, Hendrickson et al., 1989). This decline was generally attributed to leaching, uptake by regenerating vegetation, sequestration in logging slash through translocation by fungi, and denitrification (Aber et al., 1983). In Roberts Creek, the large inputs of detrital organic matter were likely to be the cause of forest floor N stocks increases. In the mineral soil, N may be gained by translocation of dissoved organic N, while losses may be reduced by complexation with reactive phases including ferrihydrite imogolite-type material, which are present in the subsoil (Grand and Lavkulich, 2008). Even though N stocks were largely conserved over the timeframe considered (15 years), it is possible that the decreasing trend observed for regenerating sites extend into the future, in which case young regenerating forests could experience N deficiency (Hendrickson et al., 1989). A longer term study is needed to assess this risk.  ORGANIC MATTER SIZE FRACTIONS The effects of logging were examined in more details by looking at the different size fractions of SOM. Logging did not affect clay-sized SOM significantly (Table 3.1). On the other hand, silt and sand-sized SOM in the illuvial horizons showed significant treatment effects following the general pattern observed for total SOC (Table 3.1). Sand-sized SOM recorded the strongest treatment effects. In illuvial horizons, sand-sized SOM was 26% higher than control in cleared plots and 20% lower in regenerating plots.  61  These results suggest that new inputs of SOM following logging tended to accumulate in sand and silt size fractions, while the clay-sized SOM fraction appeared comparatively supporting the hypothesis that the clay fraction was saturated with organic matter (Six et al., 2002, Gulde et al., 2008). This is common in soils with low clay content and few complexation sites (Borchers and Perry, 1992). Contrary to the findings of Diochon and Kellman (2009), we found no evidence that clear-cutting led to a destabilization of the most stable, clay-associated SOM. We found instead that the sand-sized particulate organic matter showed the largest amount of change in logged plots. Sand-sized SOM is not protected by interaction with minerals (Zinn et al., 2007) and may be susceptible to rapid decomposition. Accordingly, the increase we observed in cleared plot was followed by a decrease in regenerating plot. The sand-sized SOM pool do not appear to contribute to long-term C sequestration in the profiles.  INDICATORS OF ORGANIC MATTER COMPOSITION C:N ratio Effects of logging The overall effect of the logging disturbance was to decrease the C:N ratio. The decrease was only statistically significant in the organic and eluvial horizon but was observed in all horizons (Table 3.1). Soil C:N ratio generally shows an inverse relationship with net nitrification, with a C:N ratio of 25 – 30 in the topsoil generally considered to be a threshold below which net nitrification and nitrate leaching may take place (Gundersen and Rasmussen, 1990, Gundersen et al., 1998). In control sites, the C:N ratio in the uppermost layers (FH and Ae horizons) was high (Table 3.1), suggesting that these soils were not actively nitrifying. On the other hand, in logged plots the C:N ratio in Ae narrowed to an average of 21 (range 14 – 28). In regenerating plots, the C:N ratio also declined in the forest floor, narrowing to 29 on 62  average (range 24 – 42), while all mineral horizons displayed average C:N ratios of 25 or lower. This suggests that at least some of the profiles may act as a nitrate export source. Acid soils generally have high N release rates because the N-requirement of fungi tends to be lower than that of bacteria (Kooijman and Martinez-Hernandez, 2009). Indeed, Hudson and Tolland (2002) described a marked increase in nitrate in Roberts Creek streams draining plots harvested 2 to 3 years prior to sampling. The decrease in C:N ratio we observed suggests that soil organic matter undergoes intense decomposition and maturation after logging (Binkley, 1984, Bock and Van Rees, 2002, Hannam et al., 2005, Salmon et al., 2008). Several factors are likely to enhance organic matter decomposition after disturbance. After vegetation removal, dry season soil moisture content increased significantly (Table 3.1), due to the increase in effective precipitation and lack of vegetation uptake. In regenerating plots, soil moisture content remains elevated in the FH layer, perhaps reflecting incomplete canopy closure. The alleviation of summer drought may contribute to higher decomposition rates. Soil temperature is also likely to increase after logging as a result of increased solar irradiation, especially in the summer months (Johnson et al., 1985). These conditions stimulate microbial activity (Marks and Bormann, 1972, Meentemeyer, 1978, Hendrickson et al., 1989, Gabriel and Kellman, 2011). In addition, fresh needles and early successional litter may have higher N content and may be less recalcitrant than mature forest litter (Covington, 1981). In Roberts Creek, we observed active growth of fireweed (Epilobium angustifolium) in cleared plots, which may contribute easily degradable organic matter to the soil (Bradley et al., 2001b). A few alder trees were also observed in regenerating plots. Alders are known for their symbiotic association with the nitrogen-fixing actinomycete Frankia alni. The addition of alder litter to the forest floor may have contributed some nitrogen to soils of regenerating plots. Finally, fresh organic matter may produce dissolved organic compounds that can be translocated throughout the profile and exert a priming effect on existing soil organic matter (Bingeman et al., 1953, Fontaine et al., 2007, Crow et al., 2009, Zummo and Friedland, 2011). We suggest that these factors combine to stimulate decomposition and produce a more mature, N-rich SOM pool.  63  Timing of C:N decrease In the topmost mineral horizon (Ae), the C:N ratio decreased immediately after logging, suggesting that the organic matter in the top part of the profile started maturing rapidly following disturbance. In the FH layer, the variance of the C:N ratio increases significantly in cleared plots (p = 0.02 by Levene’s test for homogeneity of variances), whereas the mean value of the ratio remains roughly equal to control values (Table 3.1). We propose that the increased variance in FH reflects a combination of fresh organic inputs from logging slash, and increased maturation / decomposition processes of existing organic matter. Fig. 3.3 shows that logging disrupted the C:N depth profile. In control plots, the C:N ratio narrowed rather constantly with depth. The C:N ratio decreased dramatically in Ae in cleared plots while the lower part of the profile was generally unaffected, causing the Bf1 horizon to have a higher C:N ratio than the overlying A horizon. In regenerating plots, this trend was smoothed as C:N inturn decreased in illuvial horizons. The delay observed in illuvial horizons may be a sign that different processes are involved in C:N ratio decline and OM maturation. In the topsoil climatic effects may be more important, whereas in the subsoil priming effect from early successional vegetation dissolved organic matter and root exudates could play a greater role.  64  Figure 3.3: C:N variation in profiles of (a) control plots, (b) cleared plots, and (c) regenerating plots. Mean (thicker line) is shown surrounded by 90% confidence limits.  65  Logging disrupts pre-existing C and N relationships Other evidence that logging disturbance results in the disruption of a pre-existing steady state between organic matter inputs and decomposition (Chaer et al., 2009) can be found by examining the relationship between SOC and C:N ratio. In control plots, no relationship was observed between SOC and C:N, in accordance with the observation of Waksman (1924), who noted that soils tend to achieve a relatively stable C:N ratio over time. In logged plots however, a positive relationship was observed (Fig. 3.4), indicating that the C:N ratio increased in organic-rich samples. This corresponds to a ‘nutrient dilution effect’(McGroddy et al., 2004). Organic C concentration increased more rapidly than N concentration in organic-rich samples and the C:N ratio increased. On the other hand, in samples low in SOM the C:N ratio is also low, consistently with the idea that decomposition reduces both the amount of organic matter and the C:N ratio. The relationship between SOC and C:N ratio was similar in both cleared and regenerating plots (grouped under ‘logging’ in fig. 3.4) and suggests that regenerating plots have not reverted to the steady state situation observed in undisturbed plots.  66  Figure 3.4: The C:N ratio as a function of soil organic carbon (SOC) in (a) the organic layer and (b) the mineral soil.  67  C concentration of SOM Effect of logging Like the C:N ratio, the C:SOM ratio generally decreased in regenerating plots (Table 3.1). Statistically significant trends were observed in the FH and illuvial horizons. This may be another indication of increased decomposition and maturation of organic matter after logging, with the C:N and C:SOM ratios decreasing as C is preferentially lost from SOM (Johnson, 1995). The N:SOM ratio was relatively constant and in mineral horizons no significant treatment trends were observed (Table 3.1). In the FH layer, an increase in N:SOM ratio was seen in regenerating plots, which may indicate intense decomposition. The Ae horizon did not show a decrease in the C:SOM ratio after logging. This suggests that different processes control SOM dynamics. In the Ae horizon, the preferential eluviation of oxidized, polar organic compounds may have a stronger influence on bulk organic matter composition than decomposition processes. The Ae horizon is likely to retain C-rich, long-chain alkyl structures with a hydrophobic tendency (Rumpel et al., 2004), regardless of any logging treatment. Bulk SOM composition and CECe One of the consequences of narrowing C:SOM ratios is the increase in oxygen content including oxygen-bearing functional groups such as carboxl and phenolic groups (Johnson, 1995). In acid soils with low clay content, a large proportion of the cation exchange capacity is provided by organic functional groups (Federer and Hornbeck, 1985). The decrease in C:SOM ratio we observed in regenerating stands should correspond to an increase in cation exchange sites. We tested this hypothesis by examining the ratio between the sum of exchangeable cations and soil C (CECe:C). We found that organic matter in the Ae horizon showed a comparatively constant CECe:C ratio, consistent with the constant C:SOM ratio, while the CECe:C ratio increased significantly in illuvial horizons of regenerating plots (Table 3.1). This confirms that in most of the profile, SOM of regenerating plots bore an increased number of exchange sites. The pH was constant or decreased after logging in Roberts Creek (Chapter 5), implying that much of the increase in 68  CECe:C ratio was due to actual changes in the character of SOM. Combined with the narrowing of C:N and C:SOM ratio, the increase in CECe:C ratio indicated that organic matter of regenerating plots has undergone increased decomposition and oxidation as a result of logging (Johnson et al., 1997). A high CECe:C ratio denotes SOM of high maturity and environmental performance (Miralles et al., 2009), which may help reduce nutrient loss after logging , minimize environmental impacts and improve forest regeneration (Johnson et al., 1997).  PYROPHOSPHATE EXTRACTABLE METALS Table 3.1 shows the effects of logging on the sum of Alp and Fep. Similar trends were observed for Alp and Fep individually (not shown). In the forest floor, the concentration of Alp and Fep was generally higher in logged than in control plots, likely because mixing of minerals into the forest floor enhanced metal availability. Metal availability is a limiting factor to the pool of organically complexed Al and Fe in the FH layer (Chapter 2). The C:(Al+Fe)p ratio narrowed accordingly. The variability displayed by the forest floor is high and these trends were not statistically significant. In the Ae horizons of regenerating plots, Alp and Fep decreased while the C:(Al+Fe)p ratio widened (Table 3.1). This indicates that SOM in the Ae horizons of regenerating plots complexed less Al and Fe, possibly because of a change in the average concentration of sorption sites. In the illuvial horizons, Alp and Fep concentrations were higher in cleared plots than in control or regenerating plots (Table 3.1). This trend matched the changes in SOC so that the C:(Al+Fe)p ratio remains essentially constant at ~ 5. This constant and relatively narrow C:metal ratio suggests that in the subsoil, metals are present in sufficient amounts for humus to reach its maximum metal sorptive capacity, and probably the maximum protection that can be afforded by metal complexation.  69  INTEGRATED EFFECTS OF LOGGING The effects of logging can be effectively recapitulated by plotting samples from different treatments on principal component axes. Input variables for the principal component analysis were SOC, C:N, C:SOM, Alp, Fep, moisture and CECe:C. Table 3.2 shows the coefficients of the first 3 principal components (PC) for mineral horizons. Eigenvector coefficients are the values used to lineary combine the original variables into orthogonal principal components. A high eigenvector coefficient signals that the associated original variable is an important part of the principal component considered, and that the original variable correlates highly with the principal component. The first 3 PC accounted for 77% of the total variance. The first PC (PC1) is an index of organic matter content, as shown by its correlation with SOC, moisture and organically-complexed Al and Fe. The third PC (PC3) is an index of organic matter “freshness” or “immaturity” and correlates positively with C:N and C:SOM ratios, and negatively with the CECe:C ratio. The interpretation of the second PC (PC2) is less obvious. PC2 correlated positively with the C:N ratio and the CECe:C ratio, but negatively with the C:SOM ratio. This contrasts with the other PCs, where the coefficients associated with C:N and C:SOM had the same sign. PC2 thus probably reflects the fact that some of the processes determining the C:N and C:SOM ratios are different. The C:SOM ratio is likely to be mostly influenced by the oxidation state of the organic matter, while the C:N ratio depends not only on the extent of organic matter decomposition, but also on the fate of the mineralized N. Inorganic N can be lost from the SOM pool by plant uptake, leaching or volatilization, or retained by microbial immobilization. This will result in variations in the C:N ratio of the bulk SOM pool, which do not always match the variations in the C:SOM ratio. PC2 captures this difference in behaviour. PC2 and PC3 represent similar amounts of variance (15 and 14%, respectively), indicating that either are equally valid representations of the range of the original variables. We chose to represent PC1 and PC3 as the x and y axes in the bivariate graph (Fig. 3.5). PC3 was preferred to PC2 on the second axis because PC3 is a better representation of the processes we are trying to illustrate. 70  Table 3.2: Eigenvectors of the first 3 principal components (PC1- PC3) for mineral horizon analysis. The last line shows the % variance explained by each factor. For PC1 and PC3, the highest coefficients are in bold font.  SOC1 C:N2 C:SOM2 Alp3 Fep3 Moisture4 CECe:C5 Variance explained  PC1 “SOM content” 0.50 0.09 0.25 0.49 0.47 0.41 -0.21 48%  1  Soil organic carbon concentration  2  Carbon to nitrogen and carbon to soil organic matter ratios  3  Pyrophosphate extractable Al and Fe  4  Gravimetric moisture content  5  Effective cation exchange capacity to carbon ratio  PC2 0.01 0.63 -0.48 0.17 0.14 0.08 0.57 15%  PC3 “SOM immaturity” 0.07 0.76 0.30 -0.18 -0.22 -0.19 -0.45 14%  Fig. 3.5 shows results for the first Podzolic horizon (Bf1). Similar graphs were obtained for the Bf2 and BC horizons (not shown). Fig. 3.5 shows clear differences between samples from control, cleared and regenerating plots. The major difference between samples from control and cleared plots was the score on PC1, while samples from regenerating plots were characterized by different scores on PC3. Samples from cleared plots tended to score higher than control on PC1, an index of SOM content. Likely causes of SOM increase include the accumulation of logging residues, decaying roots and translocation of dissolved organic carbon to mineral horizon. Samples from regenerating plots featured a difference in SOM quality, as demonstrated by a low score along PC3 (Fig. 3.5). A low score along the PC3 axis indicates low C:N and C:SOM ratios and a high CECe:C ratio. This suggests the presence of organic matter that is more oxidized and bears a larger number of functional groups.  71  The change in organic matter quality in regenerating plots is most likely the result of increased decomposition and oxidation. The SOM content was lower in regenerating than in cleared plots, supporting this proposition. Several studies also support this interpretation by providing molecular evidence of organic matter maturation following logging. Dai et al. (2001) reported an increase in the decomposition index of dissolved organic compounds of clear cut sites, while Kalbitz et al. (2004) reported an increase in aromaticity and complexity of organic molecules in forest floor leachate.  Figure 3.5: Distribution of control, cleared and regenerating samples along the first (PC1) and third principal components (PC3) for the first Podzolic horizon (Bf1).  CONCLUSION This study established the effects of clear-cut logging on SOC content, distribution and quality of SOM in Podzols of coastal British Columbia. 72  SOC content and distribution Forest floor C stock increased after logging, most likely because of the progressive incorporation of logging slash into the organic horizons. In mineral horizons, the SOC content was higher in cleared plots and lower in regenerating plots. The sand-sized fraction recorded the largest variations in SOC concentration after logging, while the clay fraction had a comparatively constant SOC concentration, suggesting that there was no extensive formation of new organomineral complexes. In cleared plots, the increase in subsoil SOM could be due to decaying roots and increased organic matter illuviation. New SOM inputs did not appear to be retained in regenerating plots, perhaps because of the lack of association with soil minerals. The overall soil profile showed a good retention of C stocks after logging. We propose that the large inputs of logging slash coupled with relatively slow decomposition rates due to cool temperatures and substrate recalcitrance prevented C stocks from declining significantly for at least 15 years after clear-cutting. The proportion of SOC found in the forest floor as opposed to mineral horizons increased significantly after logging. Forest floor SOC is generally more easily degraded or mobilized due to the lack of protection by the soil matrix and may be subject to longer-term losses. Keeping in mind that our study focused on areas that experienced only minor harvest-related physical disturbance, more significant effects are expected for high traffic areas of logged plots. In particular, C stocks may be greatly reduced in areas where the forest floor is lost. The majority of C stock was located in the mineral sub-soil. The overall response of soil C storage after logging was determined by changes in SOC in illuvial horizon (Bf and BC). Thus, we argue that C balance and C dynamics studies of Podzolic soils would be greatly improved by systematically taking into account the entire depth of illuvial horizons instead of focusing on the top 10 – 20 cm of the soil profile. Indicators of organic matter quality Samples from regenerating plots generally had a lower C:N and C:SOM ratios and higher CECe:C ratios than younger plots, suggesting that logging caused a change in organic matter quality. We propose that increased microbial decomposition rate due to higher temperature, moisture and higher substrate availability is responsible for the production of a more oxidized 73  SOM pool in regenerating plots. The narrower C:N ratio observed in logged plots suggest that they have the potential to produce nitrate for several years. On the other hand, the higher CECe:C ratio may help retain nutrients after logging, reduce environmental damage and promote rapid forest regeneration. The Ae horizon response to logging was distinct from the rest of the profile, suggesting that the controls of SOM dynamics in this horizon were different. The FH, Bf and BC horizons of logged plots displayed variations that could be explained by a combination of fresh organic inputs and increased decomposition. On the other hand, the Ae horizons of logged plots showed some indication of increased decomposition status of SOM (narrowing of C:N ratio), while other variables such as the C:SOM and CECe:C ratios remained unchanged. We hypothesize that the dominant influence on bulk SOM composition in Ae horizons of both control and logged plots is the preferential retention of C-rich, hydrophobic compounds. Overall effects of logging The main difference between control and cleared plots was the amount of SOC, while regenerating plots were distinguished by a change in bulk organic matter composition. This suggests that the response to logging comprised at least 2 stages. The first stage was characterized by SOM gains, probably resulting from the gradual assimilation of logging slash, root decay and increased translocation of SOC to mineral horizon. The second stage was characterized by SOM losses from the mineral soil and changes in bulk organic matter quality suggesting a higher degree of oxidation. Finally, our data suggest that the dynamic equilibrium characteristic of undisturbed forests was not re-established at 15 years after logging. Changes in the C:N depth profiles and the correlation between SOC and C:N provided an indication that the pre-existing steady state between organic matter inputs and decomposition had been disrupted by logging, and suggest that increased decomposition rates are likely to persist for some time after disturbance. A long term study focusing on the evolution of SOM amount and composition in plots logged 15+ years prior to sampling is needed.  74  CHAPTER FOUR  Exchangeable ions and nutrients:  Distribution and relations  75  SYNOPSIS Many classical concepts of soil chemistry were established in agricultural soils that are usually fine-textured with near-neutral pH. These concepts often need to be modified when considering acid forest soils. For example, studies conducted in northeastern US forests challenged the idea of a positive relationship between base saturation (BS), effective cation exchange capacity (CECe) and pH. The objective of this chapter is to establish the distribution of exchangeable cations and nutrients in Roberts Creek Podzols, and determine their relation to pH and to surface active soil components. We measured exchangeable cations and nutrients concentration in each pedogenic horizon, and investigated their relationship with pH, soil organic carbon, silt and clay, and pyrophosphate-extractable, oxalate-extractable and dithionite-extractable Al and Fe concentration. We found that the chemistry of the forest floor was different from the mineral soil due to the low degree of mixing between the two layers. In the forest floor, biological cycling maintained high nutrient concentrations, a high BS and high calcium saturation but a low pH. In the mineral soil, pH had a positive correlation to BS but a negative correlation to CECe. The formation of allophane-humus complexes may explain the anomalous relationship between pH and CECe. Nutrients concentrations were low in the mineral soil. A large portion of the soil phosphorus was strongly sorbed to short-range order mineral phases and did not contribute to labile P pools.  INTRODUCTION Soils of temperate and northern forests are receiving growing interest because of their role in carbon sequestration and biomass production, in addition to the more traditional function of timber harvesting. A good understanding of soil processes and the relationships between soil attributes is necessary to be able to protect soil functions and long-term productivity under new 76  demands. However, acid forest soils challenge some of the classical principles of soil chemistry as established largely in agricultural soils. In particular, the traditional relations between pH, base saturation (BS) and effective cation exchange capacity (CECe) may not apply (Ross et al., 2008). Base saturation Base saturation is generally calculated as the sum of charge of base cations (exchangeable Ca, Mg, K, Na) divided by CECe (Brady, 1990). In forest soils, BS is a measurement of the relative storage of nutrient cations and is thus an indicator of soil quality (Ross et al., 2008). It is generally accepted that pH and BS have a positive relationship. As the pH increases, the amount of exchangeable acidity is usually reduced, producing higher BS values (Brady, 1990). One of the first indications that the relationship between pH and BS may be more complex in forest soils came from studies of forest soil chemistry that found the highest BS values in the lowest pH horizon, which was the forest floor. Federer and Hornbeck (1985) reported a base saturation of 68 to 97% in organic layers of pH 3.7 to 5.5. Ross et al. (1991) established that the absence of a positive correlation between BS and pH was not unusual in acid forest soils. The presence of a negative or insignificant correlation between pH and BS was later ascribed to the non-acidic behaviour of Al at low pH (Skyllberg, 1994, Johnson, 2002, Ross et al., 2008). Ross et al. (2008) noted that at pH < 4.5, Al is largely nonhydrolyzing and thus acts as a ‘base’ (non-acid forming) cation. When adding Al to the sum of base cations, Skyllberg (1994) and Johnson (2002) were able to restore significant positive correlations between pH and “BS”. The studies discussed above were conducted in highly acidic soils that tended to be strongly affected by acid deposition. In soils of the US Pacific Northwest and western Canada, acid deposition is not considered to be significant and the acidity is mostly the result of natural pedogenic processes such as coniferous litter accumulation and leaching. The relationship between pH and BS may be different and deserves to be investigated.  77  CECe The relationship between pH and CECe is generally expected to be positive, due to the weak acidic behaviour of most exchange sites. At low pH, exchange sites may become protonated, thereby reducing CECe (Thomas, 1982, Brady, 1990). However, investigations in soils of the northeastern US (Johnson et al., 1991b, Johnson et al., 2000, Johnson, 2002) and Scandinavia (Skyllberg, 1994) have reported a negative correlation between pH and CECe. The negative relationship between CECe and pH could be due to the effect of Al. Aluminum is oxyphyllic and simple bonding bewteen Al and O is about 60% ionic and 40% covalent (Essington, 2004). Al is likely to bond to charged soil particles with a continuum of bonding strength (Ross et al., 2008). pH affects Al speciation and solubility, as well as organic matter configuration. It has the potential to affect the bonding strength of Al to organic compounds. Lower pH soils may have less Al in inner-sphere surface complexes and more of the surface charge could be measurable as CECe. Consistent with this hypothesis, Johnson (2002) and Ross et al. (2008) proposed that negative correlations between CECe and pH in mineral soils involved organically complexed (pyrophosphate – extractable) Al occluding exchange sites at high pH. Short-range order (SRO) aluminosilicates and oxyhydroxides may also influence CECe and its relation to pH. Seyboldl and Grossman (2006) found that an increasing pH tended to decrease the CECe of Andisols, which are known to contain SRO Al and Fe species, while the inverse occurred in the non-Andisols taxonomic groups. Bartoli et al. (2007) observed that in allophanic Andosols, the abundance of humus – allophane complexes was negatively correlated to the soil’s specific surface area. Bailey et al. (2008) proposed that the formation of humus-allophane complexes dramatically reduced the free negative charges of organic matter and thus the CECe of New Guinean soils. Johnson (2002) argued that the effect of interation between organic matter and SRO aluminosilicates and oxyhydroxides were likely to be small compared with metalorganic complexation at their sites, but recognized that in soils containing less OM and more Al and Fe oxides, this may not be the case. In addition to their influence on CECe, SRO material also influences the availability of anionic nutrients such as nitrate and phosphate.  78  Nitrogen and phosphorus One of the main factor limiting the productivity of northern forests is thought to be N availability (Gessel et al., 1973, Chappell et al., 1991, Weetman et al., 1997, Ingerslev et al., 2001), and there has been evidence that P is also an important limiting factor for the productivity of conifers (Bennett et al., 2003, Blevins et al., 2006). Forms, concentration and distribution of N and P in the soil profile are therefore of interest. In addition to their role in plant nutrition, N and P are also potential pollutants if they leach out of the soil profile and enter water bodies, where they can contribute to eutrophication (Ohle, 1953). The concentration of labile N and P and their relation to the soil solid phase are key determinants of N and P behaviour in the ecosystem. Objectives The objectives of this study are (1) to report the concentration and distribution of labile ions and nutrients in Roberts Creek soils, and (2) to establish their relation to master soil variables, such as pH, and to reactive components of the solid fraction, such as SRO Al and Fe phases. Special attention was given to the dependence of BS and CECe on pH. We identified predictors of CECe using multiple regression analysis and developed hypotheses about factors controlling CECe. We also used simple correlation analysis to investigate the interaction between BS and pH, and between inorganic P and N and reactive soil components.  PH AND BS  Roberts Creek soils are moderately to strongly acidic (Table 4.1). pH was lowest in the FH and Ae horizons (pHH2O < 5, pHCaCl2 < 4), most likely the result of the acidifying effect of leaching and vegetation. pHH2O, pHCaCl2 and pHKCl The difference between pH measured in water and in salt (ΔpH) was negative, indicating that all sites had a net negative surface charge (Burt, 2004). ΔpH values ranged from 0.6 to 1.3. 79  pH readings in 0.01 M CaCl2 solution and in 1 M KCl solution were very similar, despite the factor of 2 difference in molarity. The overall average for pHCaCl2 was 4.37, while the overall average for pHKCl was 4.34. A linear regression between pHCaCl2 and pHKCl returned a coefficient of determination r2 of 0.93 with an intercept not significantly different from 0 and a slope not significantly different from 1 (data not shown). Because Ca2+ is divalent, it is expected to displace acidic cations more effectively than K+, but this cannot fully compensate for the difference in mass ion effect between a 0.01 M and a 1 M solution. The most likely explanation for the similar pHCaCl2 and pHKCl values is that the exchangeable acidity pool is quite small and effectively displaced by low molarity solutions. Higher molarity solutions produce the same effective displacement. The ΔpH was positively correlated with SOC concentration (r = 0.36, p < 0.01). This suggests that a significant portion of the exchange sites, including exchangeable acidity, are provided by soil organic matter, as is generally the case in coarse textured forest soils developed from glacial till (Johnson, 2002, Ross et al., 2008, Ping et al., 2010). There was also a weak negative correlation (r = -0.22, p = 0.03) between ΔpH and Al and Si associated with SRO inorganic phases (Alsro and Sisro, calculated as oxalate-extractable minus pyrophosphate-extractable Al and Si). This could indicate that SRO minerals are contributing positively charged exchange sites, thus reducing the net total charge (Kleber et al., 2007). Base saturation Base saturation was highest in the forest floor (Table 4.1), despite the low pH. The Ae horizon had the lowest BS, while the lower parts of the B horizon had intermediate BS averaging about 50%. The forest floor also showed a significantly smaller spread of BS values than the mineral soil (see standard errors of the mean, Table 4.1). Base saturation in FH ranged from 73 to 90%, while in the mineral soil it ranged from 7 to 81%. The high BS saturation of the organic layer is likely to reflect the controlling influence of biological cycling on exchangeable cations. Biocycling bases (especially calcium) accumulate in the forest floor, while Al is effectively excluded (Table 4.1). This is because Al is not subject to  80  significant plant uptake and the availability of weatherable Al mineral is limited in mor humus layers (Ross et al., 2008). Table 4.1: pH, base saturation and proportion of major ions on exchange complexes of control plots (n = 9). Means are given ± standard error of the mean (SEM) horizon FH Ae Bf1 Bf2 BC 1  pHH2O 4.6 ± 0.1 4.6 ± 0.1 5.4 ± 0.2 5.6 ± 0.2 5.5 ± 0.2  pHCaCl2 3.6 ± 0.1 3.6 ± 0.1 4.5 ± 0.1 4.8 ± 0.1 4.7 ± 0.1  BS (%) 80 ± 1.4 30 ± 3.4 39 ± 3.2 51 ± 3.1 49 ± 3.3  Ca saturation1 (%) 68 ± 1.4 24 ± 2.9 31 ± 3.0 39 ± 2.8 36 ± 2.8  Al saturation1 (%) 8 ± 1.5 55 ± 3.9 55 ± 3.1 45 ± 3.1 47 ± 3.3  calculated as the charge pertaining to Ca or Al, divided by CECe  Base saturation correlated positively with pH in mineral horizons, but there was no correlation in the FH layer (Table 4.2). Including Al in the sum of ‘base’ cations, following the proposition of Skyllberg (1994) and Johnson (2002), failed to restore a significant positive relationships with pH (Table 4.2). Moreover, the nature of the relationship between pH and BS was not determined by a pH threshold. A positive correlation was observed between pH and BS in Ae but not in FH, even though FH and Ae horizons have a similar pH (pHH2O = 4.6, Table 4.1). While the nonhydrolyzing behaviour of Al below pH 4.5 undoubtedly influence the relationship between BS and pH in more acidic soils, this phenomenon cannot explain the different behaviours of FH and Ae horizons in Roberts Creek. Forest floors in Roberts Creek are largely isolated from the mineral soil due do the absence of mixing by soil animals. As a consequence, the FH layer is very low in weatherable minerals, including Al minerals (Ross et al., 2008). We propose that the relative proportion of exchangeable Al and bases in FH is primarily a function of weatherable mineral availability, which is in turn determined by the amount of disturbance in the profile and largely independent of pH.  81  Table 4.2: Pearson correlation coefficients between pHH2O, base saturation (BS), and exchangeable Al. The p-values indicate the probability of no correlation (null hypothesis).  FH (n = 20) Ae (n = 23) Bf1 (n = 23) Bf2 (n = 24) BC (n = 24) All (n = 114)  r p-value r p-value r p-value r p-value r p-value r p-value  pH pH pH pH pH pH  BS  Alexch  -0.11 0.64 0.76 0.00 0.73 0.00 0.67 0.00 0.71 0.00 0.07 0.47  0.23 0.33 -0.60 0.00 -0.64 0.00 -0.57 0.00 -0.65 0.00 -0.58 0.00  ‘BS’ including Al 0.12 0.62 -0.28 0.19 0.00 0.98 -0.27 0.20 -0.14 0.52 -0.66 0.00  Al saturation 0.18 0.46 -0.73 0.00 -0.76 0.00 -0.69 0.00 -0.74 0.00 0.09 0.29  Table 4.2 also illustrates the importance of considering correlations separately in pedogenic horizons dominated by different processes. The overall (all horizons pooled) correlation coefficient for BS and pHH2O was -0.15, suggesting a weak negative relationship, where in fact there was no correlation in the forest floor, and strong positive relationships in the mineral soil, with different intercepts in the Ae and Bf-BC horizons (Fig. 4.1).  82  Figure 4.1: Relationship between base saturation (BS) and pH in the FH, A and B horizons  EFFECTIVE CATION EXCHANGE CAPACITY Effective cation exchange capacity and exchangeable cations The effective cation exchange capacity was calculated as the sum of exchangeable base cations and exchangeable acidity. There was a large contrast between CECe in the forest floor (average 40 cmolc/kg, range 31-57 cmolc/kg) and the mineral soil (average 2.3 cmolc/kg , range 0.4 – 5.8 cmolc/kg). This is consistent with the idea that organic matter is responsible for the majority of the exchange sites in Roberts Creek Podzols. The low exchange capacity in the mineral soil is probably due to the coarse texture. When expressed on a silt+clay basis, CECe values ranged from 1.4 to 18.7 cmolc/kg, more in line with other acid soils of glaciated areas (Ross et al., 1991, Johnson et al., 2000).  83  Table 4.3: Salt-extractable ions and effective cation exchange capacity (cmolc/kg) of control plots (n = 9). Means are given ± SEM. CECe FH  40.3 ± 3.88 Ae 4.59 ± 0.29 Bf1 2.15 ± 0.21 Bf2 1.38 ± 0.31 BC 1.01 ± 0.18  Ca  Mg  K  Na  Al  Fe  H  27.0 ± 2.63 0.83 ± 0.13 0.69 ± 0.20 0.59 ± 0.24 0.32 ± 0.09  2.83 ± 0.29 0.12 ± 0.01 0.08 ± 0.02 0.06 ± 0.02 0.03 ± 0.01  1.95 ± 0.23 0.07 ± 0.01 0.07 ± 0.01 0.04 ± 0.01 0.03 ± 0.01  0.46 ± 0.04 0.05 ± 0.01 0.04 ± 0.00 0.04 ± 0.01 0.05 ± 0.00  3.48 ± 1.60 2.83 ± 0.32 1.13 ± 0.15 0.59 ± 0.16 0.53 ± 0.16  0.14 ± 0.03 0.14 ± 0.01 0.05 ± 0.00 0.02 ± 0.00 0.02 ± 0.00  1.72 ± 0.61 0.53 ± 0.00 0.08 ± 0.07 0.01 ± 0.01 0.01 ± 0.01  Exchangeable base cations decreased in the order Ca > Mg > K > Na in both FH and the mineral soil (Table 4.3). In the mineral soil, the strongest linear associations were observed between Ca and Mg (r = 0.94, p < 0.01), and between Al and Fe (r = 0.85, p < 0.01), suggesting that the chemical and biological behaviour of these elements in the ecosystem are similar. Ca was the dominant exchangeable cation in FH, while Al dominated in upper mineral horizons. In Bf2 and BC horizons, these two cations were present in approximately equal amounts. In FH and Ae horizons, exchangeable Al concentrations were above 2 cmolc/kg, which may indicate toxicity to organic decomposers (Dahlgren et al., 2004, Tonneijck et al., 2010). Exchangeable H was present in significant amounts in FH and Ae, reflecting the importance of organic acidity in these horizons (Sullivan et al., 2005). Predictors of CECe Multiple regression analysis was conducted to identify predictors of CECe in each horizon type (FH, A, B). Candidate variables comprised SOC, silt+clay, Alsro+Fesro, pHCaCl2, and the C:N and C:SOM ratios. Soil organic carbon and the silt and clay content were considered because these soil component provide exchange sites. Alsro+Fesro is an indicator of SRO inorganic phases, which may also contribute exchange sites. pH was included because of its influence on the protonation of surface groups. Finally, C:N and C:SOM were used as indicators of the composition and degree of maturity of soil organic matter, which may influence the functional 84  groups density. The regression analysis identified very different predictors in each horizon type (Table 4.4), indicating that stratification by horizon was necessary. Table 4.4: Predictors of CECe as identified by multiple regression analysis. For each layer, variables that explained significant amounts of CECe variations are shown with associated coefficient, partial r2 and results of the F-test based on the type III sum of squares. Horizons  Variable  Estimate  Partial R2  Forest floor R2 = 0.16 Eluvial R2 = 0.70  Intercept pH Intercept SOC C:SOM silt+clay Intercept pH SOC Alsro+Fesro C:SOM  -6.29 11.42 1.49 1.59 -0.05 0.12 9.27 -1.70 0.61 -0.06 -0.01  N/A 0.16 N/A 0.39 0.17 0.14 N/A 0.49 0.07 0.06 0.02  Illuvial R2 = 0.64  p>F (type III sum of squares) 0.79 0.09 0.22 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.10  Forest floor Table 4.4 shows that pH was the only significant predictor of CECe in the forest floor. pH was positively correlated with CECe, as expected given the weak acid behaviour of organic matter (Thomas, 1982, Brady, 1990). The r2 was only 0.16, indicating that a large portion of the variation in CECe remains unresolved. The SOC concentration was not a significant predictor of CECe. This suggests that in the FH layer, the charge and presumably the composition of organic matter is very variable, so that SOC concentration is not a good indicator of the amount of cation exchange sites.  Eluvial horizon Ae In the eluvial layer, SOC was the most important predictor of CECe (Table 4.4). Silt and clay concentration was also retained as significant predictors of CECe, but explained less variance. 85  The C:SOM ratio was negatively associated with CECe. A high C:SOM ratio may be indicative of the presence of charcoal or otherwise recalcitrant organic compounds with a low charge density, and thus a low CECe. Charcoal was common in the Ae horizon and may represent a significant fraction of the organic matter in some Ae horizons (Sanborn and Lavkulich, 1989). Figure 4.2 displays the relationship between CECe and SOC in the eluvial and illuvial layers. The intercept of simple linear regression was not significantly different from 0 (p = 0.20) in illuvial horizons. In the eluvial layer, the regression slope was similar to that of illuvial horizons, but the intercept was higher and significantly different from 0 (p = 0.001). This is another indication that in the eluvial horizon, a non-negligible portion of the exchange sites is provided by mineral components.  Figure 4.2: Simple linear regression between percent soil organic carbon (SOC) and effective cation exchange capacity (CECe, cmolc/kg) Illuvial horizons Bf and BC In the illuvial horizons, SOC also had a significant and positive relationship with CECe, while the C:SOM ratio showed a negative correlation. Silt and clay concentration was not an important 86  predictor of CECe , suggesting that inorganic functional groups are not a major contributor to CECe (Johnson, 2002, Ross et al., 2008, Ping et al., 2010). pH and Alsro+Fesro were also retained as significant predictors of CECe, but the coefficients were negative (Table 4.4). The simplest explanation for a negative correlation between CECe and pH involves SOC. Soil organic matter provides exchange sites while depressing pH. However, in the multiple regression model, pH had a depressing influence on CECe even in the presence of the SOC variable. pH also correlated negatively with the CECe:C ratio (Table 4.5), which can be thought as CECe ‘normalized’ for organic carbon content. This indicates that the negative relationship between pH and CECe cannot be considered to be entirely due to SOC. Table 4.5: Pearson’s coefficients correlation matrix for CECe:C, pHCaCl2, SOC, Alp, Alsro+Fesro and Alp:Alo (illuvial horizons only). Shaded cells highlight statistically significant correlations. n = 81 CECe CECe:C pHCaCl2 SOC Alp Alsro+Fesro Alp:C  r p-value r p-value r p-value r p-value r p-value r p-value r p-value  CECe 1.00 0.00  CECe:C 0.53 0.00 1.00 0.00  pHCaCl2 -0.69 0.00 -0.34 0.00 1.00 0.00  SOC 0.61 0.00 -0.19 0.11 -0.56 0.00 1.00 0.00  Alp 0.57 0.00 -0.11 0.35 -0.45 0.00 0.81 0.00 1.00 0.00  Alsro+Fesro -0.15 0.21 -0.44 0.00 0.19 0.10 0.35 0.00 0.32 0.00 1.00 0.00  Alp:C -0.06 0.64 0.11 0.38 0.07 0.57 -0.23 0.04 0.31 0.00 0.04 0.75 1.00 0.00  CECe: effective cation exchange capacity SOC: soil organic carbon Alp: pyrophosphate-extractable Al Alo: oxalate-extractable Al Alsro, Fesro: Al and Fe associated with short-range order minerals, calculated as the difference between oxalate and pyrophosphate extracts  87  Alternatively, the negative relationship between CECe and pH could be due to the effect of organically complexed Al (Alp). Johnson (2002) and Ross et al. (2008) proposed that Alp could occlude exchange sites. In Roberts Creek however, we observed a positive relationship between Alp and CECe (Table 4.5). In fact, Alp was nearly as strong a predictor (positive slope) of CECe as was SOC. There was no correlation between the Alp:C and CECe:C ratios, indicating that increased Al saturation of organic matter did not correlate with a decrease in exchange site density. These factors indicate that in Roberts Creek, organically complexed Al is not a good candidate to explain the negative correlation between pH and CECe. Short-range order minerals also have the potential to influence CECe and its relation to pH by forming organo-mineral complexes that reduce the negative charge of organic matter (Bartoli et al., 2007, Bailey et al., 2008). In Roberts Creek, Alsro+Fesro correlated negatively with CECe:C (Table 4.5), an indicator of the charge density of organic matter. This is consistent with the proposition that adsorption of SRO phases onto organic matter obstructs cation exchange sites. The formation of organo-mineral complexes involving SRO phases such as ITM and ferrihydrite therefore have the potential to explain the negative correlation between CECe and pH.  N AND P FRACTIONS Phosphorus Phosphorus concentrations were generally different in the FH, Ae and B horizons (Table 4.6). The FH layer had larger concentrations of P than the mineral soil for all P fractions. The Ae horizon had lower oxalate-extractable P (Po) and P extracted by the Bray procedure (Pbray) concentration than the B horizons, most likely due to leaching. The Ae horizon is depleted in colloidal Fe and Al forms known to retain P (Brady, 1990). Salt and water-extractable P (Pexch and PH2O) were both much higher in the forest floor than in the mineral soil (Table 4.6). Plant uptake and adsorption to soil minerals are probably responsible for the sharp decrease in labile P concentration with depth. Phosphorus leaching is still possible 88  during storm events when the soil becomes saturated and lateral transport is enhanced (Easthouse et al., 1992). Pbray is a widely used indicator of plant-available P (Benton Jones Jr., 2001). Table 4.6 shows that P availability was generally low in the mineral soil. In the FH and Ae layers, Pbray was lower than Pexch. Some P readsorption may have occurred during the Bray extraction. In mineral horizons, Pbray tended to increase with depth, particularly in BC and C. These horizons showed evidence of mottling. This suggests that periodic anaerobic conditions may have increased P extractability at depth. In acid soils, Po represents the sum of easily soluble reactive P (e.g. extractable by salt solutions or the Bray procedure) and non-soluble P adsorbed on SRO Al and Fe minerals (Garcia-Pausas et al., 2008). Table 6 shows that Po was the largest P fraction in all horizons, meaning that a large proportion of soil P is strongly adsorbed to SRO Al and Fe phases. Table 4.6: Operationally-defined P fractions in different horizons of control plots (n = 9). Means are given ± SEM. In each column, means followed by a different letter were significantly different at the α=0.1 level Horizon FH Ae Bf1 Bf2 BC  PH2O (mg/kg) 9.5 ± 5.2 a 0.2 ± 0.1 b 0.0 ± 0.0 b 0.0 ± 0.0 b 0.0 ± 0.0 b  Pexch (mg/kg) 116.4 ± 15.3 a 2.9 ± 0.3 b 1.2 ± 0.1 b 0.6 ± 0.0 b 0.5 ± 0.0 b  Pbray (mg/kg) 51.3 ± 15.5 a 1.6 ± 0.7 c 4.2 ± 1.2 bc 3.8 ± 0.7 bc 5.4 ± 0.8 b  Po (mg/kg) 308.3 ± 40.9 a 24.4 ± 2.2 c 151.4 ± 26.5 b 139.5 ± 22.5 b 114.0 ± 12.1 b  Posat (%) 4.4 ± 1.6 a 2.1 ± 0.3 b 1.1 ± 0.1 c 1.1 ± 0.1 c 1.1 ± 0.1 c  PH2O: water-extractable P Pexch: salt-extractable P Pbray: plant-available P extracted by the Bray procedure Po: oxalate-extractable P Posat: degree of P saturation of short-range order Al and Fe phases, calculated as the molar ratio of oxalateextractable P to the sum of oxalate-extractable Al and Fe  We calculated the degree of P saturation of poorly crystalline minerals (Posat) following Lookman et al. (1995), as the molar ratio of Po to (Alo+Feo). Table 4.6 shows that Posat was generally low, particularly in the illuvial horizon (Posat ≈ 1%). This indicates that SRO Al and Fe 89  minerals are far from saturated with P and are likely to sorb P from solution, rather than release it (Hartikainen et al., 2010). Table 4.7 shows correlation coefficients between P fractions and soil reactive phases. In the forest floor, Po was positively correlated to all reactive Al and Fe fractions, and Pbray was positively correlated with Alsro. This reflects the ability of the oxalate and Bray extraction to solubilize some portion of P associated with reactive Al and Fe forms. Pexch was positively correlated with SOC, which may provide weak sorption sites. Pexch and PH2O were negatively correlated with Alp, which may form insoluble complexes with P, thus depressing labile P concentrations (Giesler et al., 2005). Few significant relationships were observed in the Ae horizon (Table 4.7). Po was significantly and positively correlated to Alp and Alsro, and showed a stronger association with reactive Al fractions than with Fe fractions. The positive relationship between SOC and Po was likely driven by the strong linear association of Alp and SOC in the mineral soil (Chapter 2). Overall, P fractions were largely uncorrelated to solid phase reactive fractions, suggesting that other factors such as P additions from the overlying forest floor and eluvation losses to deeper horizons exert a large influence on P concentrations. In the B horizons, Po was strongly correlated to SRO inorganic Al and Fe phases but not to organically complexed Al and Fe (Alp and Fep). This is attributable to the fact that inorganic Al and Fe phases dominate over organically complexed metals in the subsoil (Chapter 6). Pexch correlated positively with SOC and silt and clay, which may provide weak sorption sites. Pexch was also found to be positively correlated with Alp and Fep, probably because of the strong relationship between SOC and pyrophosphate-extractable metals. PH2O correlated positively with SOC, Fep, and Feoxi, but negatively with SRO Al and Fe phases. This suggests that poorly crystalline minerals may depress P concentration in solution by forming insoluble Al-P and Fe-P complexes. The absence of correlation (r = -0.05, p = 0.82) between Po and PH2O support the proposition that P is strongly adsorbed on poorly crystalline mineral phases and contributes very little towards the labile P pools.  90  Table 4.7: Pearson’s correlation coefficient between P fractions and selected variables. Shaded cells highlight statistically significant correlations.  SOC p-value Silt+clay p-value Alp p-value Fep p-value Alsro p-value Fesro p-value Aloxi p-value Feoxi p-value  Po  FH PBray Pexch PH2O Po  Ae PBray Pexch PH2O Po  B PBray Pexch PH2O  -0.56 0.25 n/a  0.57 0.18 n/a  0.39 0.09 n/a  0.56 0.19 n/a  0.85 0.03 0.95 0.00 0.82 0.04 0.73 0.10 0.97 0.01 0.71 0.18  -0.22 0.63 -0.41 0.36 0.74 0.05 0.55 0.20 -0.32 0.48 -0.57 0.18  -0.54 0.03 -0.37 0.15 0.36 0.42 -0.02 0.96 -0.51 0.38 -0.78 0.12  -0.70 0.08 -0.64 0.12 0.56 0.19 0.28 0.54 -0.26 0.57 -0.58 0.17  0.20 0.67 0.38 0.40 0.48 0.28 0.03 0.95 -0.21 0.66 -0.02 0.96 -0.31 0.50 0.44 0.32  0.03 0.89 -0.13 0.59 0.34 0.14 0.14 0.55 0.20 0.38 0.04 0.85 0.07 0.75 -0.37 0.10  0.87 0.01 0.03 0.94 0.81 0.03 0.48 0.27 0.67 0.10 0.27 0.55 0.58 0.42 0.31 0.69  0.18 0.42 -0.15 0.50 -0.05 0.81 0.14 0.52 -0.11 0.75 -0.15 0.66 0.75 0.25 -0.20 0.80  0.68 0.10 -0.33 0.48 -0.12 0.80 -0.62 0.14 -0.11 0.81 -0.47 0.28 -0.15 0.76 -0.53 0.22  0.18 0.15 -0.16 0.19 0.18 0.13 -0.08 0.53 0.71 0.00 0.75 0.00 0.47 0.12 -0.27 0.39  0.69 0.00 0.36 0.00 0.56 0.00 0.61 0.00 0.00 0.97 0.19 0.10 0.19 0.54 0.29 0.36  0.37 0.10 -0.08 0.74 0.20 0.38 0.38 0.09 -0.45 0.04 -0.38 0.09 -0.15 0.53 0.43 0.05  Po: oxalate-extractable P Pbray: plant-available P extracted by the Bray procedure Pexch: salt-extractable P PH2O: water-extractable P SOC: soil organic carbon Alp and Fep: organically complexed Al and Fe, extracted by sodium pyrophosphate Alsro and Fesro: Al and Fe associated with inorganic short range order phases, calculated as oxalate minus pyrophosphate extractable metals Aloxi and Feoxi: Al and Fe in oxides, calculated as citrate-bicarbonate-dithionite minus pyrophosphate for Al and citrate-bicarbonate-dithionite minus oxalate for Fe  Inorganic nitrogen With the exception of water-extractable nitrate (NO3H2O), whose concentration was almost nil throughout the profile, inorganic N was significantly higher in the forest floor than in the mineral horizons (Table 4.8). Salt-extractable N concentration was relatively high in the forest floor, but the organic layer was thin (3 to 8 cm) and forest floor inorganic N content was estimated to be less than 6 kg / ha. Inorganic N concentration was low throughout the profile in the mineral soil and the corresponding N content was estimated to be less than 15 kg / ha. 91  Table 4.8: Labile N (mg N /kg soil) in different horizons of control plots (n = 9). Means are given ± SEM. In each column, means followed by a different letter were significantly different at the α=0.1 level.  FH Ae Bf1 Bf2 BC  NO3exch 2.5 ± 1.1 a 0.3 ± 0.2 b 0.7 ± 0.1 b 0.7 ± 0.2 b 0.6 ± 0.2 b  NH4exch 63.7 ± 56.9 a 0.6 ± 0.2 b 0.4 ± 0.0 b 0.5 ± 0.4 b 0.6 ± 0.2 b  NO3H2O 0.2 ± 0.1 a 0.0 ± 0.0 a 0.0 ± 0.0 a 0.1 ± 0.0 a 0.1 ± 0.0 a  NH4H2O 20.3 ± 15.8 a 0.3 ± 0.0 b 0.3 ± 0.1 b 0.2 ± 0.1 b 0.4 ± 0.1 b  A significant proportion of inorganic N was water-soluble, suggesting that electrostatic interaction with soil colloids was weak. Unlike P, N showed no correlation to soil reactive fractions, suggesting that inorganic N concentrations were a function of the balance between decomposition and plant uptake, rather than interaction with soil colloids. Furthermore, sorption of competing ions such as Cl- originating from marine aerosols may reduce the potential for NO3interaction with SRO soil components (Katou et al., 1996). Finally, it should be noted that all samples were collected in the summer and during the active growing season. Root uptake is likely to have contributed significantly to the low levels of labile nutrients in the mineral soil. Nutrient concentration is expected to vary seasonally and would probably be significantly higher during winter.  CONCLUSION The forest floor and the mineral soil had contrasting distributions of exchangeable cations and nutrients. The organic layer had the largest organic acidity and lowest pH, yet it also had the highest BS and Ca saturation, and the lowest Al saturation of the profile. pH did not correlate with BS in the forest floor. Rather than being the result of non-acidic behaviour of Al, this is likely to reflect the importance of biological cycling, which concentrates bases while maintaining a low pH. pH and BS displayed the expected positive correlation in mineral horizons. 92  Cation exchange capacity was relatively high in the forest floor but very low in the mineral soil, most likely because of the coarse texture. In the mineral horizons, the most important predictor of CECe was SOC, indicating that the majority of exchange sites was provided by organic matter. pH and CECe had a negative relationship, in contradiction to the notion that more surface groups are de-protonated at higher pH, resulting in more cation exchange sites. We argue that the formation of humus – allophane complexes has the potential to explain the anomalous relationship between pH and CECe, provided the bonding strength of organo-mineral complexes increases with increasing pH. Further investigations of the behaviour of natural organo-mineral complexes and their surface charge under field conditions would be beneficial. The forest floor had significant amounts of labile N and P, suggesting that leaching losses are possible during rainfall events. In the mineral soil, labile N and P concentrations were low. A large proportion of soil P was strongly adsorbed to SRO mineral phases. The degree of P saturation of SRO Al and Fe phases was however low, indicating that SRO phases were unlikely to contribute significantly towards labile P pools. A negative correlation was observed between PH2O and SRO minerals, suggesting that they may even depress P concentration in solution.  93  CHAPTER FIVE  Nutrients and labile ions:  Effects of logging  94  SYNOPSIS Logging is known to disrupt forest biogeochemical cycles. This chapter investigates the effects of logging on soil chemistry in Roberts Creek, with special emphasis on the concentration and distribution of nutrients and labile ions. We collected data on the forest floor composition, and on pH and salt-extractable ions in all horizons of control, cleared and regenerating plots. We found that logged plots had higher soil NO3 concentration than control plots in all horizons. These elevated soil NO3 values were restricted to plots logged 1 to 3 years prior to sampling, suggesting that the increase in available NO3 after logging was short-lived. In the forest floor of logged plots, we observed lower concentrations of total and saltextractable P and K. Exchangeable K increased in the mineral soil, suggesting that some of the K leached from the forest floor was retained in Podzolic horizons. Logged plots had a significantly lower pH than control plots, but the difference was only of 0.2 pH units. Overall, the impacts of logging were less severe than reported in other studies, particularly those conducted in the northern hardwood biome. We hypothesize that the absence of significant acid deposition, the retention of a large pool of soil organic matter after logging and the presence of variable-charge colloids in the mineral soil combined to reduce the effects of logging in this coniferous ecosystem.  INTRODUCTION Forest harvesting has the potential to significantly alter site biogeochemical cycles. Biocycling elements are directly affected by the removal of biomass, while cycles predominantly under inorganic control are indirectly affected by chemical (e.g. pH) and physical (e.g. profile disturbance) changes in the soil. 95  Changes in soil labile or exchangeable element pools are of particular concern because these pools are immediate sources of nutrients for plants and affect forest regeneration and productivity. Exchangeable ions also equilibrate rapidly with soil solution composition, so that changes in ions prevalence on exchange complexes is an indicator of changes in soil solution composition, which in turn influences drainage water composition. Acidification and the N cycle The first studies of the effects of logging on soil and water chemistry were conducted in the northern hardwood forest biome. The results showed that in many watersheds, clearcutting results in soil acidification, increase in soluble Al, and a short-lived increase in NO3- concentration in soil and drainage water (Likens et al., 1970, Likens et al., 1978, Hendrickson et al., 1989, Johnson et al., 1991b, Dahlgren and Driscoll, 1994, McHale et al., 2007). The response was most pronounced when vegetation regrowth was inhibited by herbicides, but was still present in ‘normal’ forest harvesting experiments (Hornbeck et al., 1987). Likens et al. (1970) proposed that the increases in NO3- and acidity were linked. Increases in nitrification rates result in the production of H+ ions, according to the reactions below:  Figure 5.1: N mineralization reactions Because of its role in forest nutrition and its impact on water quality, N is one of the most critical nutrient in forest ecosystems. Following disturbance, changes in N concentration are often greater than changes in other elements (Vitousek et al., 1979, McHale et al., 96  2007). The potential for large N losses exists because large amounts of N are circulated annually in undisturbed forests (Vitousek et al., 1982). Disruption of one or more components of the N cycle can cause excess N availability. In cut forests, increases in mineralization rates, the lack of plant uptake and lack of canopy interception (Klopatek et al., 2006) combine to increase N availability, with a peak often reported at 3 to 5 years following harvest (Bradley et al., 2001a). One of the complicating factors to understanding and predicting N response to logging is that site-specific interactions between environmental conditions, biological variables and substrate quality, which greatly influence N dynamics (Grenon et al., 2004, Schimel and Bennett, 2004, Hazlett et al., 2007). The magnitude of N leaching from cut forests is very variable, ranging from non-existent to more than 50 folds in experiments controlling vegetation regrowth (Likens et al., 1970). The largest N exports after logging were reported in northern hardwoods ecosystems, while coniferous forests were generally less susceptible to N losses (Martin et al., 1984, Binkley and Brown, 1993, Lamontagne et al., 2000). Vitousek et al. (1982) proposed that N availability prior to disturbance was one of the main factors determining N leaching losses after disturbance. In their study, N-poor sites had lower N mineralization rates and higher N immobilization capacity than N-rich sites, resulting in better N retention after disturbance. Several N-poor sites also exhibited lags in nitrification, probably due to small initial nitrifier populations. Conifers in particular are known to produce allelophatic chemicals that inhibit nitrifier populations (White, 1994, Paavolainen et al., 1998). This nitrification lag may prevent N export to stream altogether if vegetation re-establishment is rapid enough (Vitousek et al., 1982). Coniferous forests of Southwestern Canada and the US Pacific Northwest generally have large amounts of C-rich organic matter on the forest floor and low N availability (Gessel et al., 1973), suggesting that net inorganic N production and losses after logging should be limited (Mann et al., 1988). Grenon et al. (2004) measured N dynamics in the forest floor and top mineral soil in 3 coniferous forests of British Columbia and found no increase in NO3- concentration and potential nitrification rate after clear-cutting at 2 of 97  the sites. At the 3rd site, clear-cutting resulted in large increases in both NH4 and NO3. This increase was attributed to poor colonization of cut blocks by early seral vegetation and by a decrease in organic matter quality, leading to lower N immobilization rates (Grenon et al., 2004). Strahm et al. (2005) also measured a large increase in soil solution N concentration at a highly productive Douglas fir site of Southwestern Washington. This increase was attributable to the fertilization history of the stands and the suppression of competing (early successional) vegetation with herbicides, confirming the importance of the type and abundance of early seral vegetation. Like in most conifer forests of western Canada, N availability in Roberts Creek soils is considered to be moderate to low. In the study plots, inorganic N concentrations were non-negligible in the forest floor (averaging 2.5 mg NO3-N / kg and 63.7 mg NH4-N / kg) but the organic layer is rather thin (6 cm on average). In the mineral soil, inorganic N concentration were low and averaged 0.5 mg/kg for both NO3-N and NH4-N (Chapter 3). Despite the low N availability, Hudson and Tolland (2002) presented evidence that Roberts Creek streams were relatively susceptible to NO3 increases after logging. The authors measured a 12 to 60 folds increase in stream and groundwater NO3 levels 1 to 3 years after logging, clearly exceeding responses measured in other BC watersheds (Hudson and Tolland, 2002). Other ions Together with NO3, K shows the most frequent change in soil and water after logging (Likens et al., 1970, Mroz et al., 1985, Mann et al., 1988, Johnson et al., 1991b). This is not surprising since the control on K concentration is mainly organic, as with NO3 (Vitousek, 1977). On the other hand, Ca, Mg and Na are thought to be predominantly influenced by mineral weathering (Vitousek, 1977). Although forest harvesting is thought to increase mineral weathering rates (Likens et al., 1970, Johnson, 1989), changes in the concentration of these elements is generally not as pronounced as the elements under organic control. Changes in base cations saturation on soil exchange complexes after logging are not consistent. Some studies have reported increases in exchangeable base cation 98  concentration (Belleau et al., 2006), while others described base cations losses (McHale et al., 2007). One of the factors influencing the response of base cations to logging is acidification (McHale et al., 2007). In watersheds experiencing strong acidification after logging, base cations tend to be displaced from exchange complexes to soil water by H+ and Al3+, and eventually flushed out of the system (Mroz et al., 1985, Lawrence et al., 1987, Neal et al., 1992, Dahlgren and Driscoll, 1994, Reuss et al., 1997). On the other hand, when acidification is mild or does not occur, base cation prevalence on exchange complexes either remains constant or may increase as decomposing organic matter releases significant amounts of Ca and K (Snyder and Harter, 1985, Hendrickson et al., 1989, Johnson et al., 1997). Effects of colloids The amount and type of colloids present in the soil influence the fate of nutrients and other elements that may be released by logging disturbance by controlling ion partitioning between the soil solid and solution phase. Negatively charged colloids are common, and as a result most soils exhibit a significant cation exchange capacity (Kimmins, 1997). Positively charged colloids are not as widespread, yet are critical to the retention of negatively charged essential plant nutrients such as nitrate and phosphate. In acidic forest soils, organic matter and Al and Fe fractions, particularly poorly crystalline ones, may provide anion sorption sites (Elliott and Sparks, 1981). Even ions with a weak affinity to form surface complexes such as NO3- (Parfitt, 1980) may be retained by charged colloids. Strahm and Harrison (2006) showed that the presence of variable charge soil components, mostly noncrystalline forms of Fe and Al, allowed for the concentration-dependent sorption of NO3- in coniferous soils of the Pacific Northwest. This mechanism may retain a significant portion of an otherwise highly leachable nutrient. Objectives The objectives of this study are to determine the effects of logging on soil acidity, nutrients and labile ions in Roberts Creek. We collected data on the forest floor 99  composition, pH and salt-extractable ions in all horizons of control, cleared and regenerating plots. We relate our findings to stream chemistry studies previously conducted in Roberts Creek. We hypothesize that changes in nutrients and labile ions concentration after logging will be small because (1) the sites are not nutrient-rich, (2) harvesting did not cause a dramatic loss of soil organic matter (SOM), a large source of effective cation exchange capacity (CECe) (Chapter 3), and (3) establishment of regenerating vegetation is rapid.  FOREST FLOOR COMPOSITION In temperate forests, the forest floor acts as a major reservoir of both rapidly and slowly available nutrients (Yanai, 1998). Monitoring changes in forest floor composition is one of the keys to understanding the evolution of other nutrient pools in the soil (Johnson et al., 1985). Table 5.1 shows that P, K and Zn forest floor concentrations were lower in logged plots than in control. The largest decrease was for P, which was reduced by 45% in logged plots. The C:P and N:P ratios were also wider in logged plots. This suggests that P was preferentially lost from the forest floor, perhaps as a result of translocation into decomposing coarse woody debris and stumps by fungi (Palviainen et al., 2010). The concentration of Fe, Al and Na was higher in cleared plots than in control or regenerating plots (Table 5.1). Correspondingly, the C:Al, C:Fe and C:Na ratios were narrower in cleared plots (data not shown). The Fe, Al and Na cycles are largely controlled by inorganic rather than biological processes. Their increase in logged plots is likely to reflect profile disturbance and mineral mixing into the forest floor after logging (Johnson et al., 1991b). Other elements (Ca, Mg, Mn, Cu) showed no significant change in concentration in different treatments. 100  Table 5.1: Effects of logging on forest floor composition measured by Parkinson and Allen digestion. Means are given ± standard error of the mean (SEM). P-values at the end of each row document the statistical significance of treatment effect. Within each row, means followed by a different letter are significantly different at the α = 0.1 level. Variable  C (%) N (%) P (%) C:P N:P Ca (g/kg) Mg (g/kg) K (g/kg) Na (g/kg) Al (g/kg) Fe (g/kg) Zn (g/kg)  ------------------------Treatment ------------------------Control Cleared Regenerating N=9 N = 11 N=7 38.1 ± 2.1 30.7 ± 4.7 29.2 ± 3.6 1.06 ± 0.07 0.82 ± 0.12 1.03 ± 0.14 0.10 ± 0.01 a 0.06 ± 0.01 b 0.06 ± 0.01 b 393 ± 18 558 ± 173 487 ± 63 11 ± 1 a 15 ± 3 b 17 ± 3 b 7.7 ± 0.5 6.6 ± 0.5 7.4 ± 0.6 0.3 ± 0.0 0.4 ± 0.1 0.3 ± 0.1 1.4 ± 0.1 a 1.1 ± 0.1 b 1.0 ± 0.1 b 0.8 ± 0.1 a 1.5 ± 0.3 b 1.2 ± 0.2 ab 0.4 ± 0.1 2.1 ± 1.0 0.6 ± 0.1 0.5 ± 0.2 a 2.7 ± 1.2 b 0.9 ± 0.2 ab 0.05 ± 0.01 a 0.02 ± 0.01 b 0.03 ± 0.00 b  p-value 0.12 0.13 0.00 0.15 0.03 0.30 0.41 0.02 0.06 0.13 0.10 0.02  PH AND EXCHANGEABLE IONS  CECe Effective cation exchange capacity was 20% and 36% higher in the Bf1 horizon of cleared and regenerating plots than in control plots, respectively (Table 5.2). Since most of the CECe in coarse forest soils is attributable to organic matter (Federer and Hornbeck, 1985, Johnson, 2002, Ross et al., 2008, Ping et al., 2010), the changes in CECe are likely due to changes in the quantity and/or quality of organic matter. In cleared plots we observed an increase in organic matter concentration corresponding to the increase in CECe (Chapter 3). However, organic matter decreased in regenerating plots, while CECe remained elevated. This may be due to accelerated humification of organic matter, as 101  proposed in Chapter 3. Intense decomposition increases the concentration of fulvic acids and the charge of organic matter, thereby increasing the number of exchange sites (Johnson et al., 1991b). pH and acid cations In Bf horizons, pH decreases slightly after logging (Table 5.2). There was no pH change in the forest floor or in the Ae and BCg horizon. This pattern is consistent with the hypothesis of a disruption of the N cycle. In the FH layer, hydrolysis of organic N consumes protons and counters the acidifying effects of nitrification (Qing-ru et al., 2006), resulting in a rather constant pH. Some of the ammonium ions released from the top part of the profile can then be leached to illuvial horizons where they are actively nitrified due to increased soil temperature and moisture, reduced plant uptake and possible increase in nitrifiers population after logging (Smith et al., 1968, Knoepp and Clinton, 2009). This could explain why the pH is constant in the topsoil and decreases in illuvial horizons. Overall, the pH decrease was small (0.1 to 0.2 pH units) and was not accompanied by an increase in Alexch, except perhaps in the Ae horizon (Table 5.3), or by a decrease in base saturation. This is likely due to the low rate of acid deposition in the area. Most studies that have reported strong acidification following logging were conducted on soils that had been depleted in base cations by acid deposition (Likens et al., 1970, Snyder and Harter, 1985, Johnson et al., 1991b, Dise and Gundersen, 2004, McHale et al., 2007). Roberts Creek soils are likely to have a larger buffering capacity than these acidified soils, thereby reducing the observed effects of logging.  102  Table 5.2: Effects of logging on soil pH and CECe. The treatment*horizon interaction was significant throughout the profile and treatment means were computed and compared separately for each horizon. Means are given ± SEM. P-values at the end of each row document the statistical significance of treatment effect. Within each row, means followed by a different letter are significantly different at the α = 0.1 level. Variable  pHH2O  pHCaCl2  CECe cmolc/kg  Horizon  FH Ae Bf1 Bf2 BCg FH Ae Bf1 Bf2 BCg FH Ae Bf1 Bf2 BCg  ------------------------Treatment ------------------------Control Cleared Regenerating N=9 N = 11 N=7 4.6 ± 0.1 4.7 ± 0.2 4.3 ± 0.2 4.6 ± 0.1 4.6 ± 0.1 4.6 ± 0.2 5.4 ± 0.1 a 5.3 ± 0.1 ab 5.2 ± 0.1 b 5.6 ± 0.1 5.5 ± 0.0 5.4 ± 0.0 5.5 ± 0.1 5.6 ± 0.1 5.5 ± 0.1 3.6 ± 0.1 3.7 ± 0.1 3.5 ± 0.2 3.6 ± 0.1 3.7 ± 0.1 3.7 ± 0.1 4.5 ± 0.1 a 4.3 ± 0.1 b 4.3 ± 0.0 b 4.8 ± 0.1 a 4.6 ± 0.0 b 4.5 ± 0.1 b 4.7 ± 0.1 4.7 ± 0.0 4.6 ± 0.0 40.3 ± 3.9 32.5 ± 3.9 37.3 ± 4.0 4.6 ± 0.3 5.3 ± 0.5 3.8 ± 0.4 2.1 ± 0.2 a 2.6 ± 0.3 ab 3.1 ± 0.4 b 1.4 ± 0.3 1.1 ± 0.1 1.3 ± 0.2 1.0 ± 0.2 1.0 ± 0.1 0.9 ± 0.2  p-value 0.27 0.89 0.06 0.12 0.54 0.49 0.88 0.05 0.05 0.55 0.37 0.51 0.10 0.77 0.91  103  Table 5.3: Effects of logging on salt-extractable cations and anions (Mexch). There was no treatment*horizon interaction in the Bf – BC horizons, which were grouped together under the name ‘illuvial’. Means are given ± SEM. P-values at the end of each row document the statistical significance of treatment effect. Within each row, means followed by a different letter are significantly different at the α = 0.1 level.  Alexch (mg/kg) Caexch (mg/kg) Mgexch (mg/kg) Kexch (mg/kg) Naexch (mg/kg) NH4exch (mg NH3N / kg) NO3exch (mg NO3N / kg) Pexch (mg tot P/kg) Clexch (mg/kg)  FH Ae illuvial FH Ae illuvial FH Ae illuvial FH Ae illuvial FH Ae illuvial FH Ae illuvial FH Ae illuvial FH Ae illuvial FH Ae illuvial  Control N=9 313.4 ± 143.8 254.7 ± 29.0 ab 66.0 ± 9.4 5410 ± 527 165 ± 27 104 ± 22 343.3 ± 34.8 15.1 ± 1.5 6.9 ± 1.1 761.0 ± 90.0 a 26.5 ± 5.6 19.1 ± 2.2 a 105.3 ± 9.9 a 11.4 ± 1.2 9.7 ± 0.7 63.7 ± 56.9 0.6 ± 0.2 0.5 ± 0.1 2.5 ± 1.1 0.3 ± 0.2 0.7 ± 0.1 ab 174.0 ± 33.9 a 3.6 ± 0.8 0.8 ± 0.1 250.8 ± 29.3 a 24.7 ± 2.1 a 21.0 ± 2.9  Cleared N = 11 285.6 ± 84.3 304.1 ± 49.0 a 72.1 ± 9.8 4446 ± 608 209 ± 27 112 ± 14 297.6 ± 47.3 15.8 ± 1.9 6.5 ± 0.7 383.5 ± 57.8 b 22.8 ± 3.2 28.7 ± 2.6 b 72.6 ± 12.4 b 9.9 ± 0.8 7.2 ± 0.4 30.1 ± 9.1 6.1 ± 2.7 2.1 ± 0.8 19.6 ± 16.0 1.7 ± 0.8 1.6 ± 0.5 b 74.2 ± 21.6 b 2.8 ± 0.4 0.8 ± 0.1 218.6 ± 53.0 ab 17.4 ± 2.9 b 21.7 ± 2.4  Regenerating N=7 221.1 ± 46.7 150.2 ± 45.1 b 88.3 ± 17.1 5323 ± 746 246 ± 62 111 ± 21 334.3 ± 47.1 20.9 ± 4.9 7.0 ± 1.2 392.6 ± 57.8 b 20.4 ± 2.0 15.7 ± 1.9 a 122.3 ± 12.9 a 16.0 ± 3.2 9.2 ± 0.5 13.2 ± 8.0 0.5 ± 0.2 0.7 ± 0.1 2.6 ± 0.7 0.4 ± 0.2 0.5 ± 0.1 a 98.4 ± 21.9 b 2.5 ± 0.4 0.7 ± 0.1 158.9 ± 15.9 b 20.6 ± 3.5 ab 18.2 ± 1.2  p-value 0.82 0.10 0.78 0.48 0.39 0.84 0.72 0.56 0.97 0.01 0.56 0.02 0.02 0.26 0.12 0.78 0.20 0.12 0.39 0.21 0.07 0.04 0.46 0.14 0.05 0.07 0.94  104  Base cations Caexch and Mgexch showed no treatment effects (Table 5.3). Concentrations in the illuvial horizons were constant and showed no indication of an increase in variability after logging. This contrasts with the data of Hiebler-Chariarse (2003), which showed a 25 to 50% increase in drainage water Ca and Mg in logged Roberts Creek watersheds. This suggests that rather than being unaffected by logging, soil Ca and Mg available pools reached a new steady state where leaching losses to drainage water were approximately balanced by inputs to the solum via accelerated weathering and organic matter decomposition (Johnson, 1989). In the forest floor, Kexch was lower in cleared and regenerating plots than in control. The proportion of K on exchange complexes was also lower in logged plots (p = 0.04, data not shown). Together with the decrease in total K concentration in the forest floor (Table 5.1), this suggests that K is lost from the forest floor following logging. This may be due to the fact that contrary to other nutrients, K readily leaches out of fresh organic material (Carlisle et al., 1967). In illuvial horizons, Kexch was higher in cleared plots but similar to control levels in regenerating plots (Table 5.3). This trend was also reflected in the proportion of K on exchange complexes (p=0.01, data now shown). Hiebler-Chariarse’s data (2003) showed that K+ increased 3-folds in Roberts Creek’s logged streams. This suggests that logging produces an increase in labile K in Roberts Creek soils, with some K retained on Bhorizons exchange complexes and some exported to drainage waters. Possible sources include forest floor leachates and accelerated mineral weathering immediately after logging (Bormann and Likens, 1979). In regenerating plots, decrease in effective precipitation and increase in vegetation uptake likely combine to bring K concentrations back to pre-harvest levels. Naexch tended to be lower in cleared plots and similar to control levels in regenerating plots in all horizons, but this trend was only found to be statistically significant in the forest floor (Table 5.3). The increase in effective precipitation reaching the soil after logging may promote leaching of weakly held Na ions out of the profile. Consistent with 105  this idea, Hiebler-Chariarse’s data (2003) showed a small increase in drainage water Na concentration after logging. Clexch concentration was also lower in cleared plots than in control or regenerating plots (Table 5.3). Phosphorus In the forest floor, Pexch was lower in cleared and regenerating plots than in control plots. This is in agreement with the results of the Parkinson and Allen digestion, which showed that total P concentration decreased after logging (Table 5.1). Possible causes include P leaching from sites with a low anion sorption capacity and P immobilization by decomposers. Losses of P from the forest floor may adversely affect regenerating vegetation nutrition, as the organic layer is known to supply a large portion of P requirements to humid conifer forests (Ballard, 1980). Nitrogen NO3exch was generally higher in cleared plots than in regenerating plots, which were similar to control plots (Table 5.3). In the forest floor, all salt-extractable ions were lower in cleared plots except for NO3exch, which was on average 8 times higher than in control. There was clear evidence of heteroskedasticity, with variance increasing in cleared plots and reducing the power of statistical tests. Figure 5.1 shows the dramatic increases in NO3exch in the mineral soil of some of the cleared plots. In the mineral soil, increases in NO3exch above 2 mg N/kg all occurred in plots with a C:N ratio of 25 or lower (Fig. 5.1). This is consistent with the idea that net nitrification is generally inversely correlated to the C:N ratio, with a C:N ratio of 25 – 30 generally considered to be a threshold below which net nitrification and nitrate leaching may take place (Gundersen and Rasmussen, 1990, Gundersen et al., 1998). Our data suggests that increased nitrification at favorable sites may have played a role in the increase of NO3exch in cleared plots.  106  Figure 5.2: Effects of logging on salt-extractable nitrate ( NO3exch) The variability in NO3exch concentration in cleared plots was very large. This is not entirely surprising, since cleared forests have been observed to leach nitrate at very variable rates (Vitousek et al., 1982, Grenon et al., 2004). The variability may be due to micro-climatic influences, differences in microbial population, patchiness of early successional vegetation establishment, and uneven distribution and quality of logging slash. Rosén and Lundmark-Thelin (1987) provided direct evidence that large quantities of soluble N, in both organic and mineral forms, are leached from thicker slash piles, increasing natural variability. Elevated values of NO3exch over 2mg N/kg all occurred in plots logged 1 to 3 years prior to sampling. In 5 years old plots and regenerating plots 8 to 15 years old, NO3exch was similar to control levels. This is likely due to the establishment of tree seedlings and competing vegetation, which restore N biological uptake. Early seral vegetation probably played a prevalent role in 5 year old plots, where the establishment of tree seedlings was not pervasive.  107  The timing of the observed increase in N availability (1-3 years) is in good agreement with many other studies (Likens et al., 1978, Johnson et al., 1985, Mann et al., 1988, Hendrickson et al., 1989, Hudson and Tolland, 2002, Kranabetter et al., 2006). Strahm et al. (2005) found that N availability was highest 3 to 6 years after logging, but their treatment included control of competing vegetation into the 4th growing season after disturbance. This shows that competing vegetation is likely to be an important factor in reducing N availability and N leaching losses after disturbance (Slesak et al., 2010).  OVERALL EFFECTS OF LOGGING The results from this study generally show that the effects of logging on Roberts Creek’s soil chemistry were not severe. The most notable changes were the increase in Kexch and NO3exch in cleared plots. Kexch was likely translocated from the forest floor to the mineral soil, while NO3exch concentrations were elevated throughout the profile. Average NO3exch concentrations remained modest (< 2 mg N/kg in the mineral soil, < 20 mg N/ kg in the forest floor), but some of the mineral samples from logged plots exceeded 10 mg N/kg and some of the organic samples exceeded 100 mg N/kg. Taken together with studies of drainage water chemistry showing that stream NO3 concentration increased after logging in Roberts Creek (Hudson and Tolland, 2002, Hiebler-Chariarse, 2003), these data suggest that some areas of cleared plots are actively producing and exporting NO3. Soils of Roberts Creek are relatively N-poor. We calculated that the mineral soil contained on average 0.55 kg N / m2 (Chapter 2), while the lower limit proposed by Gessel et al. (1973) for adequate Douglas fir forest N supply is 0.50 kg N / m2. Therefore even small N losses after logging could negatively impact forest productivity. Labile P and K concentrations decreased in the forest floor after logging. This could potentially affect the nutrition of regenerating vegetation, since fine roots responsible for nutrient acquisition are particularly abundant in the organic layer. Even in the case of K, 108  where increases in mineral horizons balance quite exactly losses in the organic layer, nutrient availability to shallow rooted plants including seedlings may be negatively impacted. The retention of organic matter after logging is critical to the retention of nutrients in Roberts Creek soils, since most of the cation exchange sites are provided by organic matter (Johnson et al., 1997). The increase in SOM in Podzolic horizons of cleared plots and the increase in SOM charge density in regenerating plots (Chapter 3) are particularly beneficial, since they help retain nutrients in the profile while regenerating vegetation gets established. More intensive forestry practices that do not retain as much organic matter on site, such as whole-tree harvest for the purpose of bioenergy generation, would most likely have very different consequences for nutrients dynamics in this ecosystem. Short-range order (SRO) Fe and Al minerals are known to bear some variable or permanent positive surface charge (Parfitt, 1980, Gustafsson, 2001), and have been shown to be involved in a concentration-dependent sorption of NO3- in soils of the Pacific Northwest (Strahm and Harrison, 2006). Short-range order Fe and Al minerals are abundant in Roberts Creek Podzolic horizons (see Chapter 6) and are likely to contribute to NO3 retention in the mineral subsoil, especially at high NO3 concentrations observed in some logged plots.  CONCLUSION Soil chemistry differences between logged and control plots of Roberts Creek were overall not dramatic. Acidification did occur in cleared plots but was mild, and in most horizons it was not accompanied by an increase in exchangeable Al or a decrease in base saturation. This was likely due to the absence of soil acidification prior to disturbance. The CECe of logged plots was either similar to, or higher than the CECe of control plots, most likely because of the good retention of organic matter in the profile. Podzolic 109  horizons in particular appeared to play an important role in retaining some portion of the most labile nutrients (K+ and NO3-) in the profile. It is likely that SRO Al and Fe phases contributed to NO3- retention in illuvial horizons. Despite the low N content of Roberts Creek undisturbed soils, we detected significant increases in available N in logged plots 1 to 3 years old. N also showed the largest increase in variability after logging. This suggests that local factors such as distribution of logging slash, micro-climate, in situ microbial populations, and early seral vegetation are likely to be important determinants of soil N concentration after logging. The increase in soil NO3 after logging is consistent with the results of local studies of drainage water composition, and suggests that Roberts Creek forest soils produce and export a measurable amount of NO3 following logging. Any N losses are a concern for long-term forest productivity, given the N-poor status of the soils. Other significant differences observed between logged and control plots included a lower Na, Cl, P and K concentration in the forest floor of harvested sites. Decreases in forest floor P and K are a concern for plant nutrition since in coarse acid soils, a large proportion of nutrient-acquiring fine roots are found in the forest floor. Exchangeable K was higher in the mineral soil of logged plots than in control, suggesting that some of the K lost from the forest floor is retained in mineral horizons. Changes in soil chemistry in Roberts Creek logged plots were attributable to a number of different proximal factors, in accordance with the idea that logging disrupts many compartments of element cycles. Losses of labile elements (K, Na, Cl) from the forest floor pointed to an intensification of leaching and increase in effective precipitation after logging, while the decrease in P could suggest translocation to coarse woody debris. The increase in available N throughout the profile suggested an increase in decomposition and mineralization rates, as mentioned in Chapter 3. Finally, the return of most nutrients to pre-harvest values in regenerating plots suggested a strong role for regenerating vegetation uptake.  110  CHAPTER SIX  Poorly crystalline constituents:  Characterization, distribution and relationships  111  SYNOPSIS Podzolic soils are known to contain significant concentrations of extractable Al, Si and Fe in illuvial horizons. The forms and amount of reactive Al, Si and Fe phases are major determinants of the reactivity and chemical properties of these soils, while their depth distribution sheds light on the dominant pedogenic processes. We investigated the amount, type and depth distribution of reactive Al, Si and Fe phases in Podzols of southwestern British Columbia using selective dissolution analysis (pyrophosphate, oxalate and dithionite extracts), transmission electron microscopy (TEM) and Fourier transform infrared spectroscopy (FTIR). In the Bf – Cg horizons, imogolite-type material (ITM) was the principal source of extractable Al and Si, while most of the extractable Fe was in the form of a poorly crystalline oxyhydroxide phase such as ferrihydrite. The Al:Si ratio of selective extraction procedure, together with TEM and FTIR results, indicated that the ITM probably consisted of low crystallinity proto-imogolite allophane. In the FH and Ae horizons, most of the extractable Al was found in Al-humus complexes. In the illuvial horizons, the predominance of oxalate-extractable metals indicated that SRO material dominated the mineral neoformations. The main predictor of Al partition between organic complexes and SRO inorganic phases was pH, suggesting that it is a key determinant of ITM distribution in the profile. All selective dissolution fractions showed a sharp break in concentration between the Ae and Bf horizon, suggesting that horizon differentiation was advanced and that Podzolization was well expressed in these young (~ 11,000 years old) soils. Imogolite-type material concentration averaged 2.3 % and ferrihydrite concentration averaged 0.3% in the Bf-BCg horizons. When expressed on a sand-free basis, ITM concentration averaged 11.5 % and ferrihydrite concentration averaged 1.8 %. Shortrange order inorganic phases were a major component of the soils’ reactive fraction, and may exert a strong influence on the chemical reactivity of these soils.  112  INTRODUCTION Poorly crystalline or short range order (SRO) phases are a very important component of soils reactive fraction, even at low concentrations (Jock Churchman, 2010). In tephric soils, SRO phases commonly make up the bulk of the clay fraction (Henmi and Wada, 1976). Many other soils contain non-negligible amounts of SRO phases (Ugolini and Dahlgren, 1991). Many Podzols in particular contain significant amounts of SRO material in their Podzolic horizons, where illuviation processes maintain high concentrations of reactive Al, Fe and Si (Wada, 1989). The most common SRO phases in Podzols are imogolite-type material (ITM) and ferrihydrite (Skjemstad et al., 1992, Karltun et al., 2000, Lundström et al., 2000b). Imogolite and allophane Imogolite-type material comprise imogolite, a tubular aluminosilicate of stochiometry (OH)3Al2O3Si OH (Cradwick et al., 1972, Farmer et al., 1983, Gustafsson, 2001), and proto-imogolite allophane, a mineral with the same chemical composition and local structure as imogolite (Harsh, 2000), but lesser crystalline order. Imogolite forms tubes about 2 nm in diameter and up to several micrometers in length (Farmer and Russell, 1990), whereas proto-imogolite allophane may form spheres of 3.5 to 5 nm in diameter, or fragments having an imogolite structure over a much shorter range (Parfitt and Henmi, 1980). Species with a very low polymerization state are generally undetectable with the Fourier transform infrared spectroscopy (FTIR) and transmission electron microscopy (TEM) techniques, but are readily extracted by chemical dissolution techniques. Other allophane species include halloysite-like and hydrous feldspathoid (stream deposited) allophane (Parfitt, 2009). These allophanes are rich in silica and have an Al:Si ratio close to 1. In contrast to ITM, Si-rich allophanes form at higher pH (pH 6 to 7) and are not generally observed in Podzols (Harsh, 2000). Imogolite-type material bears both negative and positive surface charges at most field conditions. It is well established that ITM bears pH-dependent charges (Parfitt, 1990). In addition, Gustafsson (2001) suggested that imogolite may have a structural charge. In his 113  model, a weak positive charge is developed on the outer tube walls, whereas a negative charge develops inside the tubes. The presence of ITM has been invoked to explain the very high chloride retention and phosphate sorption of many soils (Parfitt, 2009). Imogolite-type material is an important phase for sorption of nutrients and contaminants, especially negatively charged ones (Arai et al., 2005, Navia et al., 2005, Babel and Opiso, 2007, Isoyama and Wada, 2007, Müller and Duwig, 2007, Nishikiori et al., 2009, Opiso et al., 2009, Kaufhold et al., 2010, Nishikiori et al., 2010). The relationship between soil organic matter and SRO material such as ITM is complex. On one hand, SRO material is thought to promote soil organic carbon (SOC) accumulation and stabilization through the formation of organo-mineral complexes (Zunino et al., 1982, Mikutta et al., 2005a). On the other hand, organic molecules, particularly low molecular weight organic acid, may prevent ITM formation by chelating Al and disrupting the arrangement of atoms into the short-range crystalline structure (Huang and Violante, 1986, Mossin et al., 2002, Van Ranst et al., 2008). This process is termed the anti-allophanic effect (Shoji et al., 1993). As a result, either positive or negative correlations may be observed between ITM and organic matter. Prevalence Imogolite-type material occurs in a variety of pedogenic environments. It is particularly abundant in some soils formed from volcanic ash (Yoshinaga and Aomine, 1962, Shoji et al., 1982, Ugolini and Dahlgren, 1991, Jongmans et al., 2000, Parfitt, 2009). Interestingly, Ming et al. (2006) suggested that it may also occur on Mars. Imogolite-type material has been recognized in less acidic subsoils horizons of Podzols formed from a range of parent materials in various locations. Imogolite-type material has been identified in Podzols of Scotland (Tait et al., 1978, Farmer et al., 1980, Anderson et al., 1982, Farmer and Russell, 1990), Scandinavia (Gustafsson et al., 1995, Gustafsson et al., 1998, Karltun et al., 2000, Lundström et al., 2000b, Mossin et al., 2002), Germany (Gottlein and Stanjek, 1996), the European Alps (Zysset et al., 1999, Egli et al., 2004, Egli et al., 2007), New Zealand (Young et al., 1980, Parfitt and Henmi, 1982, Childs et al., 1983, Parfitt and Kimble, 1989, Parfitt, 1990), Russia (Byzova et al., 1990), USA 114  (Dahlgren and Ugolini, 1989, Johnson and Mcbride, 1989, Dahlgren and Ugolini, 1991) and eastern Canada (Brydon and Shimoda, 1972, Ross and Kodama, 1979, Kodama and Wang, 1989). In general, ITM content decreases from north to south in Podzols of Canada (Kodama and Wang, 1989) and Europe (Gustafsson et al., 1995, Gustafsson et al., 1998, Mossin et al., 2002). The ITM decrease in southern regions is attributed to (1) an increasing organic matter content, (2) increasing acidity and (3) increasing distance to the source of the soil parent material (Wang et al., 1986, Gustafsson et al., 1995, Mossin et al., 2002). Ferrihydrite Ferrihydrite (5Fe2O3•9 H2O) is a poorly crystalline Fe3+ oxide that forms very small spherical particles, 3 to 7 nm in diameter and bears variable charge (Schwertmann and Taylor, 1989). It is the most widespread poorly crystalline mineral found in soils (Schwertmann et al., 1986) and has been identified in Podzols of Scandinavia (Karltun et al., 2000), USA (Johnson and Mcbride, 1989), Australia (Skjemstad et al., 1992), New Zealand (Parfitt and Childs, 1988) and eastern Canada (Kodama and Wang, 1989). Objectives This paper reports the results of an investigation of SRO phases in Podzols of southwestern British Columbia, Canada. Poorly crystalline compounds were investigated using selective dissolution analysis, transmission electron microscopy and Fouriertransform infrared spectroscopy. Depth distribution and relations between selective dissolution Si, Al and Fe pools were examined in details to gain insight into pedogenic processes.  115  SELECTIVE DISSOLUTION RESULTS Table 6.1: Basic soil properties and selective dissolution results for control plots (n = 9). Means are given ± standard error of the mean (SEM). pHH2O  SOC1  clay  Alp2  Fep2  %  %  g/kg  g/kg  Sip2  Alo3  Feo3  Sio3  Ald4  Fed4  Sid4  FH  4.6 ± 0.1 38.1 ± 2.1 n/a  g/kg g/kg g/kg g/kg g/kg g/kg 2.3 ± 1.3 0.7 ± 0.2 2.2 ± 0.6 3.0 ± 1.4 1.3 ± 0.2 0.1 ± 0.0 3.3 ± 0.1 4.2 ± 0.1  g/kg 4.6 ± 0.4  Ae  4.6 ± 0.1 1.6 ± 0.2 3.6 ± 0.4 0.6 ± 0.1 0.5 ± 0.1 0.0 ± 0.0 0.8 ± 0.1 0.7 ± 0.1 0.0 ± 0.0 0.7 ± 0.1 2.9 ± 0.3  0.4 ± 0.0  Bf1  5.4 ± 0.1 2.2 ± 0.3 5.5 ± 0.5 2.9 ± 0.3 1.6 ± 0.2 0.1 ± 0.0 8.5 ± 0.8 4.1 ± 0.4 2.4 ± 0.3 7.4 ± 0.5 10.2 ± 0.7 1.1 ± 0.8  Bf2  5.6 ± 0.1 1.6 ± 0.2 5.2 ± 0.6 2.3 ± 0.5 1.2 ± 0.3 0.1 ± 0.0 8.5 ± 1.3 3.1 ± 0.5 2.7 ± 0.5 6.0 ± 0.4 7.8 ± 0.4  0.9 ± 0.1  BCg 5.5 ± 0.1 1.4 ± 0.2 4.9 ± 0.7 2.3 ± 0.4 1.1 ± 0.4 0.1 ± 0.0 7.5 ± 0.7 3.0 ± 0.5 2.3 ± 0.3 5.5 ± 0.5 5.7 ± 0.4  0.8 ± 0.1  5.5 ± 0.1 0.9 ± 0.1 3.7 ± 0.7 1.8 ± 0.3 0.9 ± 0.3 0.0 ± 0.0 7.2 ± 0.5 2.8 ± 0.2 2.2 ± 0.1 3.9 ± 0.3 4.3 ± 0.5  0.7 ± 0.0  Cg 1  Soil organic carbon  2  Pyrophosphate-extractable Al, Fe and Si  3  Oxalate-extractable Al, Fe and Si  4  Dithionite-extractable Al, Fe and Si  116  Pyrophosphate Pyrophosphate solutions extract Al and Fe (Alp and Fep) associated with organic matter (Dahlgren, 1994). Parfitt and Childs (1988) showed that pyrophosphate extractants may disperse Fe oxides such as ferrihydrite and goethite, leading to an over-estimate of Fep. Pyrophosphate may also remove some amorphous Al hydroxides and gibbsite (Kaiser and Zech, 1996). At Roberts Creek, there is a good correlation between SOC and pyrophosphateextractable metals in all mineral horizons (r2 = 0.62 in Ae and r2 = 0.69 in illuvial horizons, data not shown). The organic carbon to Fep ratios were above 10 and Alp was always greater than Fep, indicating that dispersion of inorganic Fe was not important (Parfitt et al., 1988, Zanelli et al., 2006). There was no indication that the pyrophosphate extraction extracted significant amounts of inorganic Fe and Al from Roberts Creek soils. Mean Alp concentration ranged from 0.6 g/kg in the Ae horizon to 2.9 g/kg in the Bf1 horizon. Mean Fep concentration was consistently lower than Alp, and ranged from 0.5 g/kg in Ae to 1.6 g/kg in Bf1. Sip concentrations were significant only in the FH layer and averaged 2.2 g/kg (Table 6.1). The pyrophosphate extractant probably dissolved some biogenic silica (Kodama and Wang, 1989). Alp and Alexch Exchange sites in Roberts Creek soils are believed to be derived primarily from organic matter (Federer and Hornbeck, 1985)(Federer, 1984). Both salt-extractable Al (Alexch) and Alp can then be thought of as forms of organically bound Al, with Alexch being the more easily extracted portion. The proportion of Alexch was low in illuvial horizons (B-C) (Table 6.2), indicating that most of the organically bound Al is relatively strongly complexed by SOM. In contrast, the proportion of Alexch was higher in FH, and neared 50% in the Ae horizon. This suggests that in the eluvial layer, most Al is held in weak association with the organic matter. 117  Table 6.2: Ratios and reactive fractions calculated from selective dissolution data of control plots (n = 9). Means are given ± SEM. Horizon FH Ae Bf1 Bf2 BCg Cg  Alexch:Alp % 15.9 ± 2.4 52.8 ± 3.8 4.4 ± 0.8 2.2 ± 0.3 2.0 ± 0.2 1.8 ± 0.2  Alsro g/kg  Fesro g/kg  0.6 ± 0.2 0.3 ± 0.0 5.6 ± 0.8 6.1 ± 1.2 5.3 ± 0.7 5.3 ± 0.4  0.6 ± 0.1 0.2 ± 0.1 2.5 ± 0.3 2.0 ± 0.3 1.9 ± 0.2 2.0 ± 0.1  Alsro/Sio molar ratio 12.0 ± 3.6 18.7 ± 17.0 2.3 ± 0.1 2.3 ± 0.0 2.3 ± 0.0 2.4 ± 0.1  ITM % 0.0 ± 0.0 0.0 ± 0.0 2.2 ± 0.3 2.5 ± 0.5 2.2 ± 0.3 2.0 ± 0.1  FH % 0.1 ± 0.0 0.0 ± 0.0 0.4 ± 0.1 0.3 ± 0.1 0.3 ± 0.0 0.3 ± 0.0  Feoxi % 0.2 ± 0.1 0.2 ± 0.0 0.8 ± 0.1 0.6 ± 0.1 0.4 ± 0.0 0.2 ± 0.1  Oxalate Oxalate solutions extract Al, Fe and Si (Alo, Feo and Sio) from organic complexes, SRO Fe oxyhydroxides (ferrihydrite) and aluminosilicates (imogolite and allophane) (Dahlgren, 1994). In addition, the oxalate reagent removes some of the Al from the hydroxy-Al interlayer of 2:1 layer silicates (Dahlgren, 1994) and gibbsite (Parfitt, 2009), but their contribution is generally small (Johnson and Mcbride, 1989). Oxalate was the most efficient Al extractant in the illuvial horizons, as is usually the case when ITM is an important constituent of the poorly-ordered Al (Gustafsson et al., 1995). In accordance with the findings of Lavkulich et al. (1971), we measured high levels of oxalate-extractable Al in the B and C horizons. In the Bf1 horizon, Alo averaged 8.5 g/kg (Table 6.1) and ranged from 4.4 to 10.9 g/kg. These values are high considering the coarse texture of the horizon (75% sand). Oxalate was also and by far the most efficient extractant for Si in the B and C horizons. Sio concentration was highest in the Bf2 horizon, where it averaged 2.7 g/kg (Table 6.1) and ranged from 0.2 to 7.2 g/kg. Spodic index The spodic index, calculated as Alo + ½ Feo, exceeded the threshold for spodic horizon (5g/kg) (IUSS Working Group WRB, 2006, Soil survey staff, 2006) in all illuvial horizons (Fig. 6.1). A few profiles exhibited silandic-like properties in the subsoil, with Sio > 6 g/kg and Alo + ½ Feo > 20 g/kg. When expressed on a sand-free basis, over 80%  118  of the subsoil horizons exhibit silandic-like properties. This is an indication that SRO material dominates the mineral neoformations (IUSS Working Group WRB, 2006).  Figure 6.1: Values of the spodic index (Alo+ ½ Feo) in different horizons of control plots Alsro & Fesro The difference between oxalate and pyrophosphate values gives a measure of Al and Fe associated with SRO inorganic material (Alsro and Fesro) (Parfitt and Childs, 1988, Dahlgren, 1994). Alsro was highest in the Bf2 horizon, with an average of 5.6 g/kg and a range from 0.6 to 16.1 g/kg, while Fesro was highest in the Bf1 horizon, with an average of 2.5 g/kg (Table 6.2) and a range from 0.7 to 8.4 g/kg. Ferrihydrite Ferrihydrite concentration was estimated from Fesro, according to Childs et al. (1991): % ferrihydrite = 1.7 * % Fesro Other SRO Fe oxides such as nano-crystalline goethite or poorly crystalline lepidocrocite are also dissolved by the oxalate extraction (Schwertmann and Fitzpatrick, 1977, Thompson et al., 2006). Therefore, the poorly crystalline Fe phases may consist in a  119  mixture of ferrihydrite and other SRO Fe oxides. For simplicity, we use the term ferrihydrite since it is the most widespread poorly crystalline Fe phase in young Podzols. There was no ferrihydrite in the FH and Ae horizon, and low concentrations in the illuvial horizons. In the B horizon the ferrihydrite concentration averaged 0.3% (Table 6.2) and ranged from 0.0 to 1.4%. When expressed on a sand-free basis, ferrihydrite concentration averaged 1.8 % and ranged from 0.2 to 7.3%. (Lal, 2004) ITM We followed the common practice of using oxalate-extractable Al and Si to estimate ITM concentration (Farmer et al., 1983, Parfitt and Childs, 1988). We calculated ITM concentration from Sio and the (Alo-Alp)/Sio molar ratio, according to the formula of Mizota and Van Reeuwijk (1989):  % ITM = 100  % Si o 23.4 - 5.1x  with x = (Alo-Alp)/Sio representing the molar ratio of Al and Si in ITM. The Al:Si ratio was capped at 2.5 for this calculation, assuming that excess Al should be allocated to hydroxy-Al in interlayered minerals (Mizota and van Reeuwijk, 1989, Dahlgren, 1994). In the B and C horizons, concentrations of Sio were relatively high (2.2 to 2.7 g/kg) despite the coarse texture of the soils, and the Al:Si ratio was close to the ideal stochiometry of 2 in all samples (Fig. 6.2). This is indicative of the occurrence of ITM (Dahlgren, 1994, Zysset et al., 1999). The Al:Si ratio ranged from 1.86 to 2.72, indicating that the extracted material was rich in Al, and suggesting that ITM content was not overestimated by dissolution of Si from other soil minerals (Lilienfein et al., 2003).  120  Figure 6.2: Relationship between Al (Alsro) and Si (Sio) associated with short-range order material in illuvial horizons  As expected, ITM was not present in the FH and Ae horizons, as indicated by Sio concentrations below 0.05 g/kg and aberrant Al:Si ratios (Table 6.1 and 6.2). In these horizons, leaching conditions and low pH preclude ITM precipitation (Parfitt and Kimble, 1989). In the illuvial horizon, ITM concentration averaged 2.3% and ranged between 0.4 and 12.0%. When expressed on a sand-free basis, ITM concentration averaged 11.5 % and ranged from 0.5 to 58.0 %. Incidentally, samples with the highest ITM concentration were found in the cut block located around the F5 creek (12.0%, 6.5%, and 5.5% in the Bf1, Bf2, and BC horizons, respectively). This is the creek where Hudson and Tolland (2002) observed an anomalously high increase in drainage water NO3 concentration following partial harvest. Further investigation is needed to determine if the presence of high concentration of SRO phases is in some way linked to NO3 export.  121  Imogolite-type material was restricted to samples with pH ≥ 5.1 (Fig. 6.3). This is consistent with the proposition of Lindsay and Wathall (1996), according to which a pH below 5 hinders imogolite formation by preventing Al polymerization. In the low pH samples, Al-humus complexes were dominant.  Figure 6.3: Imogolite-type material (ITM) concentration as a function of soil pH Alp:Alo The Alp:Alo ratio is an indicator of the composition of the colloidal fraction, with high Alp:Alo ratios (≥ 0.5) indicating the prevalence of Al-humus complexes, and low ratios indicating that of inorganic SRO phases such as ITM (Shoji et al., 1988, Mizota and van Reeuwijk, 1989). In Roberts Creek Podzols, the Alp:Alo ratio was high in the FH and Ae horizons and decreased with depth (Fig. 6.4a). This indicates that in the upper horizons (FH and Ae), reactive Al was mainly bound to organic matter. The ratio was below 0.5 in illuvial horizons (Bf – BCg), indicating that in these horizons, the majority of reactive Al was bound to poorly crystalline inorganic phases (Zysset et al., 1999, Rasmussen, 2007). As  122  expected, the ITM content was negatively correlated with the Alp:Alo ratio (Fig. 6.4b). The ITM concentration was low (<2%) at Alp:Alo ratios greater than 0.5.  Figure 6.4: The pyrophosphate to oxalate-extractable Al ratio (Alp:Alo) and imogolitetype material ITM. (a) Values of the Alp:Alo ratio in different horizons of control plots. (b) ITM as a function of the Alp:Alo ratio  pH was a stronger predictor of the Al partition between the organic pool and the SRO inorganic pool than SOC (Table 6.3). When all horizons are considered together, pH explained 2/3 of the variance in Alp:Alo. Correlations coefficients between pH and Alp:Alo were also high in the illuvial horizons considered individually (Table 6.3). When the FH and Ae horizons were considered individually, there was no correlation between Alp:Alo and pH or SOC. This is likely due to the fact that both the pH and the SRO Al pool were low in all FH and Ae horizons, meaning that the spread of values was not large enough to observe correlations. The strong correlation between Alp:Alo and pH suggests that pH is the main determinant of Al partition between organic complexes and SRO inorganic phases, and thus regulates SRO material distribution in the profile. The anti-allophanic effect (Shoji et al., 1993), whereby ITM formation is prevented by organic chelation of Al, did not appear to play a 123  prevalent role. We observed a positive correlation between Alp and Alsro (r=0.51, p = 0.00), as well as a positive correlation between SRO material and SOC concentration (Chapter 2). This suggests that the stabilizing effect of SRO material on SOM, rather than the anti-allophanic effect, dominates the relationship between ITM and SOM in Roberts Creek. Table 6.3: Pearson’s correlation coefficients between the pyrophosphate to oxalateextractable Al ratio (Alp:Alo), pH and soil organic carbon concentration (SOC).  pH SOC  Horizon all Alp:Alo r -0.80 p-value 0.00 r 0.47 p-value 0.00  FH Alp:Alo -0.28 0.47 -0.30 0.43  Ae Alp:Alo 0.37 0.19 0.16 0.58  Bf1 Alp:Alo -0.49 0.01 0.18 0.36  Bf2 Alp:Alo -0.80 0.00 0.49 0.01  BC Alp:Alo -0.76 0.00 0.42 0.03  Dithionite Dithionite extracts Al, Fe and Si (Ald, Fed and Sid) from organic complexes and secondary oxyhydroxides, both crystalline and non-crystalline (Dahlgren, 1994). The presence of a significant amount of Sid indicates the presence of Si-bearing ferrihydrite, which can lead to an overestimate of ITM. Dithionite does not generally extract SRO aluminosilicates (Parfitt and Childs, 1988, Dahlgren and Saigusa, 1994), but some studies noted that the treatment may dissolve a fraction of allophane and imogolite, especially if these materials have very poor structural order, making the interpretation of Ald and Sid difficult (Sheldrick, 1984, Dahlgren, 1994).  In Roberts Creek, the concentration of Sid was only significant in the FH layer, and Alo exceeded Ald in all mineral horizons. If a Si-substituted oxide phase was present, it was therefore restricted to the forest floor and did not significantly interfere with ITM estimation in the illuvial horizons. Moreover, sources of Sid in the forest floor more likely consisted in silicic acid adsorbed to organic matter and some biogenic silica (Sauer et al., 2006, Saccone et al., 2007). 124  Dithionite was the most efficient extractant for Fe. Fed was highest in the Bf1 horizon, where it averaged 10.2 g/kg (Table 6.2) with a range from 6.9 to 17.4 g/kg. Feo / Fed The Feo/Fed ratio provides an indication of the proportion of free Fe present in poorly crystalline Fe oxides such as ferrihydrite (Skjemstad et al., 1992). Fig. 6.5 shows that the proportion of poorly crystalline oxides is lowest in the Ae horizon, and increases with depth. Intense leaching conditions in the Ae prevent the precipitation and persistence of SRO phases, while maintaining low Si concentrations that favour crystalline Fe oxides precipitation (Schwertmann, 1985). The Feo/Fed ratio is highest in BCg / Cg horizons, where approximately 50% of oxide phases are poorly crystalline material such as ferrihydrite. Seasonal anoxic conditions may favour the persistence of poorly crystalline phases in these deeper horizons.  Figure 6.5: Evolution of the oxalate to dithionite-extractable Fe (Feo:Fed) ratio with depth in control plots  125  TEM AND FTIR RESULTS The presence of ITM was further checked with transmission electron microscopy and Fourier-transform infrared spectroscopy. TEM The classic imogolite tubular morphology was not observed on the electron micrographs. Aggregates of spheroidal particles consistent with proto-imogolite allophane morphology were observed (Fig. 6.6a). The electron diffraction (EDX) data was consistent with an aluminosilicate composition. Diatom shells were common in the micrographs (Fig. 6.6b), confirming the presence of a biogenic Si pool.  (a)  (b)  Figure 6.6: Transmission electron micrographs of (a) possible proto-imogolite allophane aggregates and (b) a diatom skeleton. FTIR The infrared spectra are presented in Fig. 6.7. They were consistent with the presence of low crystallinity ITM.  126  Figure 6.7: Infrared spectra of the acid-dispersible clay fraction of 2 Bf horizons. (a) 4000 – 1200 cm-1 range (mid infrared) (b) 1200 – 400 cm-1 range (far infrared) (c) difference between spectra before and after dehydroxylation at 350°C (far infrared)  In the mid infrared, a broad vibration band was observed between 3800 and 2800 cm-1, with a maximum at 3450 cm-1 (Fig. 6.7a) . This is due to OH stretching vibration of structural OH or adsorbed water (Wada, 1989). The corresponding bending vibration band was seen at 1650 cm-1 (Fig. 6.7a). These bands point to considerable amounts of adsorbed and hydration water, and suggest the presence of allophanes or imogolite (Russell and Fraser, 1994, Woignier et al., 2005).  127  In the far infrared, a broad absorption region was observed between 1200 and 900 cm-1 (Fig. 6.7b). This region records Si-O-Al and Si-O-Si stretching and bending vibrations (Farmer et al., 1983, Wada, 1989), and can be used to differentiate between imogolite and allophane (Russell and Fraser, 1994). The absorption maximum was located around 1000 cm-1. The absence of a clear doublet near 1000 cm-1 indicated poor development in the imogolite c-axis direction. A narrow band was present at 940 cm-1. Figure 6.7c shows the result of subtraction between mid-infrared spectra before and after dehydroxylation at 350 °C, at which point allophane and imogolite are decomposed. After subtraction, the band at 940 cm-1 was better resolved and a clear, narrow band appeared at 970 cm-1. This band is characteristic of proto-imogolite allophane (Karltun et al., 2000). The 600 – 400 cm-1 region was complex (Fig. 6.7b). The bands at 534 and 472 cm-1 are indicative of dioctahedral phyllosilicates. The band at 431 cm-1 is indicative of ITM or kaolinite. As kaolinite bands were not observed in the far infrared, ITM was the most likely cause for this band. Due to instrumental limitation, the region between 400 and 250 cm-1 could not be investigated.  DEPTH PROFILES The depth distribution of soil reactive fractions is a reflection of the dominant pedogenic processes. Here we present the depth distribution of selective dissolution fractions and contrast them to the depth profiles of strongly biocycling elements (C, N and P). Podzolization depth profiles Overall, the selective dissolution fractions followed the classic eluviation – illuviation depth pattern of Podzols (Fig. 6.8). Such depth profiles are reminiscent of the depletion – enrichment profiles described in geochemical weathering studies (Brantley et al., 2007).  128  Figure 6.8: Depth profiles of Al and Fe associated with organic matter (Alp and Fep), short-range order inorganic phases (Alsro and Fesro) and crystalline oxides (Feoxi) in control plots  With the exception of crystalline iron oxides (Feoxi, calculated as 1.5*[Fed-Feo]), the concentration of reactive Al and Fe phases was negligible in the Ae horizon and increased dramatically in the Bf1 horizon. This indicates that Podzolization is well expressed. The eluviated Ae horizon was very thick, averaging 6.2 cm in thickness (range 2-15 cm), and indicating intense leaching and depletion of labile constituents. This is interesting considering that the soils are geologically young. Clague and Luternauer (1983) have reported that the region only became ice-free around 11,000 years before present. Podzolization therefore appears to be a rapid process in the area, perhaps due to the combination of the humid climate and free drainage of the upper soil horizons.  The concentration of Al and Fe reactive fractions remained high in the upper part of the Cg horizon. The limit between the BCg and Cg horizon is very sharp, with the BCg horizon consisting of single-grain material with extensive mottling, whereas the Cg horizon is very compact with a platy structure and gleying throughout. The fact that 129  concentration of Al and Fe reactive fractions are significant in the upper 5-10 cm of the Cg horizon indicates that pedogenic chemical processes operate ahead of the physical breakdown of parent material (Brantley, 2010).  Alsro and Fesro concentrations were highest in Bf1 and Bf2 (Fig. 6.8), indicating that the illuviation processes were most expressed in these horizons. Alp and Fep were highest in the Bf1 horizon and in the FH layer, reflecting the influence of organic matter. The concentration of crystalline iron oxides was highest in Bf1 and decreased sharply with depth. This is most likely the result of the interaction between the Podzolization process and seasonally reducing conditions in the lower horizons. Anoxic conditions favour iron oxides dissolution. The soluble Fe may then be leached out of the profile, or reprecipitate as poorly crystalline phases such as ferrihydrite during the dry season. Addition depth profiles The depth profiles of biocycling elements C, N and P are given in Fig. 6.9. These profiles stand in contrast with the depletion-enrichment profiles discussed above. The biocycling elements were distributed along an addition profile (Brantley et al., 2007), with a high concentration in the FH layer and a sharp decrease in the mineral soil.  Figure 6.9: Depth profiles of (a) C, (b) N and (c) P (%) in control plots While largely dominated by biological additions, the depth profiles of C, N and P still showed the effects of Podzolization. Concentrations were lower in the Ae than in the Bf horizons due to leaching and illuviation of organic compounds. The process of addition is 130  thus modulated by the process of Podzolization. Translocation and accumulation of organic matter in the illuviated horizons was moderate, with the SOC concentration averaging 2.2% in Bf1 (Table 6.1). Overall, the distribution of most reactive elements appears to be reducible to the effects of one dominant pedogenic process, modified by another, less prominent process.  IMPORTANCE OF SRO PHASES In this section we examine the average composition of the clay fraction in terms of organic matter, crystalline minerals and SRO phases, and investigate the relative contribution of SRO material to each horizon’s reactive components.  Fig. 6.10 shows the composition of the clay fraction of the Ae, Bf, and BCg-Cg horizons. The proportion of crystalline aluminosilicates was estimated indirectly by subtracting ITM, iron oxides and organic matter from the total clay concentration (Tonneijck et al., 2010).  Figure 6.10: Relative importance of clay-sized constituents in control plots, with claySOM = clay-sized organic matter, ITM = imogolite-type material, FH = ferrihydrite, clay = crystalline aluminosilicates and Feoxi = crystalline Fe oxides.  131  Eluvial horizon The Ae horizon had the lowest total clay content of all the mineral horizons (Table 6.1), but contained the largest concentration of crystalline aluminosilicates. Crystalline clays mineralogy was not determined. Based on previous studies of post-Wisconsin Podzolic soils developed under a cool humid climate, the dominant crystalline clay mineral is likely to be smectite (Ross, 1980). Clay-sized organic matter and crystalline Fe oxides made smaller contributions to the clay fraction (Fig. 6.10). Illuvial horizons On the other hand, in the Bf and BCg-Cg horizons, SRO inorganic constituents made up nearly half of the clay fraction. Crystalline phyllosilicates comprised less than a quarter of the clay fraction (Fig. 6.10), and most likely consisted in a mixture of 2:1 phyllosilicates such as vermiculite, chlorite and mica (Ross, 1980, Wilson, 1999). Most of the surface-active constituents consisted in organic or SRO inorganic compounds, as is often the case in Podzolic horizons. The low amount of crystalline aluminosilicates is attributable to the coarse texture of the parent material, as well as the young age of the soils, free drainage, and cool and humid climate. Low inputs of thermal energy and leaching of Si hinders the formation of crystalline aluminosilicates (Parfitt et al., 1983, Parfitt and Wilson, 1985). Formation of SRO material is kinetically favoured, and the cool and humid climate retards the ‘Ostwald ripening’ process, where SRO material is transformed into crystalline minerals (Schwertmann, 1985, Rasmussen, 2007). Relation between crystalline clays and SRO phases Interestingly, there was no correlation between total clay content and non-crystalline components (taken as the sum of clay-sized organic matter and SRO inorganic phases) in the Bf horizon, even though amorphous and SRO phases make up the bulk of the clay fraction (Fig. 6.10). On the other hand, there was an excellent correlation in the Ae horizon (Fig. 6.11).  132  Fig 6.11: Relationship between non-crystalline phases (sum of clay-sized organic matter and SRO inorganic phases) and total clay concentration  This is not a contradiction if we consider the possibility of a competing relationship between crystalline clays and SRO material in the illuvial horizons (Lindsay and Wathall, 1996). Poorly crystalline compounds of Fe, Al and organic matter may accumulate at the expense of well crystallized phyllosilicates, without a change in total clay content. In the Ae horizon, the major constituent of the non-crystalline phase is organic matter (Fig. 6.10). The good correlation between % clay and the non-crystalline phase reflects the association of clay and organic matter in this horizon.  CONCLUSION This study established the relative importance and depth distribution of reactive Al and Fe pools in Podzols of southwestern Canada. Selective extraction indicated that significant amounts of inorganic SRO components were present in the Bf, BCg and top 133  part of the Cg horizon. Short-range order components consisted of ITM and poorly crystalline oxyhydroxides such as ferrihydrite. Imogolite-type material was most likely present as low crystallinity proto-imogolite allophane. In the illuvial horizons (Bf – BCg), SRO inorganic components comprised about half of the clay fraction, while crystalline aluminosilicates made up less than a quarter. The remainder of the clay fraction consisted in organic matter and crystalline Fe oxides. Pedogenesis and Podzolization proceeded ahead of the physical breakdown of the parent material, with significant accumulation of SRO inorganic components and organic matter in the topmost part of the Cg horizon. The depth distribution of reactive Al and Fe components provided insight into dominant pedogenic processes. Reactive Al and Fe phases concentration reflected the paired eluviation – illuviation processes typical of Podzolic soils. Organically complexed Al and Fe dominated the reactive metal pool in the FH and Ae horizons. In the Bf – Cg horizons, the predominance of oxalate-extractable metals indicated that SRO material dominated the mineral neoformations. The accumulation of illuviated organic matter was relatively low (2.2% in the Bf1 horizon). The partition of Al and Fe between organic, SRO and crystalline oxide phases gave insight into pedogenic controls of Al and Fe forms. The main predictor of Al partition between organic complexes and SRO inorganic phases was pH, suggesting that it is a key determinant of ITM distribution in the profile. Organic matter concentration was a secondary determinant of Al partition. The proportion of poorly crystalline Fe oxyhydroxides depended on horizon position, with the highest values recorded in the BCg and Cg horizon. In these horizons, seasonal reducing conditions may favour the accumulation of poorly crystalline oxides. Overall, SRO inorganic phases are expected to be a large determinant of the chemical properties and reactivity of the sub-soil.  134  CHAPTER SEVEN  Poorly crystalline constituents:  Effects of logging  135  SYNOPSIS Short-range order (SRO) inorganic phases, such as imogolite-type material (ITM) and ferrihydrite, are an important component of the soil reactive fraction in Roberts Creek Podzols. This chapter investigates the effects of logging on SRO Al and Fe phases concentration, as calculated from selective dissolution data. We found that cleared plots had a higher concentration of SRO Al and Fe phases than undisturbed plots. A possible explanation for this observation is that logging disturbance caused a temporary increase in the intensity of illuviation processes, resulting in the precipitation of new SRO Al and Fe phases. Regenerating plots had concentrations of reactive Al and Fe phases similar to, or below control levels. In Roberts Creek Podzols, SRO Al and Fe phases thus appear very dynamic and responsive to changes in environmental soil conditions on a subdecadal scale.  INTRODUCTION Short-range order Al and Fe phases are important constituents of Podzols around the world (Parfitt, 2009). In Roberts Creek, selective dissolution data indicated that the illuvial horizons contain up to 12.0 % ITM and 1.4 % ferrihydrite. These SRO phases made up nearly half of the clay fraction (Chapter 6). Due to their large specific surface area and high reactivity, these materials are expected to exert a strong influence on soil biogeochemical processes (Parfitt, 2009). Significance of SRO phases One of the most important characteristic of SRO material is their large specific surface area (several hundred m2 / g). As a consequence, they are important phases for ion exchange and adsorption processes. At most soil pH, ITM and ferrihydrite bear positive charge, making them important phases for anion retention (Vance et al., 1996). 136  The dissolution and precipitation kinetics of SRO phases are many order of magnitude faster than kinetics of crystalline minerals. As a result, SRO phases constitute a reactive pool of Al, Fe and Si ions, which can be released into solution very rapidly (within days) following changes in the environmental conditions (Kleja et al., 2005, Yagasaki et al., 2006). It has been proposed that ITM and ferrihydrite control the immediate concentration of Al and Fe in the soil solution of Podzolic horizons (Farmer and Lumsdon, 2002). Effects of disturbance Short-range order phases are very reactive and they are the first phases to form after the release of ions from primary mineral weathering. They are also the first inorganic materials to register the effects of disturbance. Basile-Doelsch et al. (2009) noted that land use change such as cultivation or vegetation changes may modify the stability of mineral phases, particularly if the degree of crystallinity is low. Tanskanen and Ilvesniemi (2004) showed that ploughing induced dissolution of ITM in Podzols. Their study was conducted 17 years after ploughing occurred, but it is possible that ITM dissolution started very rapidly after disturbance. Logging is one of the most common disturbance of forested soils, and has the potential to affect SRO phases stability via effects on the biochemical characteristics of soils. Logging has been shown to result in increases in dissolved organic matter fluxes (Kalbitz et al., 2004), soil acidification (Johnson et al., 1991b), and increases in weathering rates (Likens et al., 1970). Short-range order material such as ITM has a narrow pH stability range. pH controls the amount of reactive Al3+ available for polymerisation and influences the type of aluminosilicate that forms. Imogolite-type material forms at acidic pH when the Al:Si ratio is high and octahedrally coordinated Al is favoured (Harsh, 2000), but it is believed that a pH below 5 hinders imogolite formation by preventing Al polymerization (Lindsay and Wathall, 1996). Other factors may modify this range in the field, since ITM has been observed in Podzolic B horizons where pHH2O was as low as 4.7 (Mossin et al., 2002). Nevertheless, a decrease in pH to values below 4 to 4.5 is likely to cause ITM 137  dissolution. McHale et al. (2007) proposed that clear-cutting of acidified northern hardwood forests resulted in the production of nitric acid which dissolved ITM and released its aluminum in monomeric form to soil and drainage waters. Organic matter is another major determinant of SRO phases stability. Organic substances interacts with soil minerals and influences the amount and forms of Al present in solution. A number of organic acids have been shown to enhance the formation of low crystallinity aluminosilicates by distorting the repeated structure normally found in crystalline minerals (Huang and Violante, 1986, Vance et al., 1996). On the other hand, high concentration of organic substances may complex a large proportion of metal ions and make them unavailable for crystalline minerals and SRO phases formation (Shoji et al., 1993, Mossin et al., 2002). In Podzols, dissolved organic matter is also thought to play an important role in the leaching of metals from the eluvial horizon and their translocation and precipitation in the illuvial horizons (Lundström et al., 2000a). As can be expected from changes in pH and dissolved organic matter, logging generally results in an increase in primary mineral weathering rates and an increase in dissolved Al and Si (Likens et al., 1970). This has important consequences for the formation of ITM, since sufficient Al and Si in solution is one of the key factor in the formation of ITM in soils (Harsh, 2000). There are thus many reasons to suspect that the amount of SRO phases such as ITM and ferrihydrite present in the soil may be affected by logging. Their large surface area, high reactivity and narrow stability range imply that short-term anthropogenic disturbances have the potential to affect their concentration. Because of the influence SRO phases have on biogeochemical processes, understanding the effects of disturbance on these materials may help shed light on other impacts of logging, such as organic matter (de)stabilization and nutrient retention. Johnson et al. (1991b) suggested that increased allophane and imogolite may have been responsible for the increase in cation exchange capacity of Podzolic horizons in logged plots of Hubbard Brook, while Hudson (personal communication) hypothesized that differences in SRO mineral pools may explain variations in N leaching losses in logged plots of Roberts Creek. The objective of this 138  chapter is to investigate the effects of logging on SRO Al and Fe phases in Roberts Creek.  EFFECTS OF LOGGING The effects of logging on Al and Fe reactive fractions, as well as on the ITM and ferrihydrite concentration as calculated from selective dissolution data, are presented in Table 7.1. Table 7.2 presents ratios illustrating the distribution of Al and Fe between the organic and inorganic phase. The treatment*horizon interaction was significant for the ratio variables. As a result, each horizon was analyzed separately. Only results for the Bf1 horizon are shown as deeper horizons showed no indication of a treatment effect. Table 7.1 shows that the concentration of reactive Al and Fe phases was generally higher in cleared than in control plots. In regenerating plots, SRO mineral fractions were similar or lower than in control. Ferrihydrite concentration was on average 34% higher in cleared plots and 23% lower in regenerating plots than in control. Imogolite-type material concentration was 25% higher in cleared plots and 31% lower in regenerating plots than in control. Table 7.2 shows that the proportion of Al and Fe associated with organic compounds was highest in regenerating plots. In control and cleared plots, the ratio between inorganic SRO Al and Fe and organically complexed Al and Fe was close to 2:1, while in regenerating plots, this ratio was close to 1:1.  139  Table 7.1: Mean ± standard error or the mean (SEM) of reactive Al and Fe fractions. Pvalues at the end of each row document the statistical significance of treatment effect. Within each row, means followed by a different letter are significantly different at the α = 0.1 level. Variable  Alo1 Alsro2 Feo1 Ferrihydrite3 ITM3  --------------------------Treatment-----------------------Control Cleared Regenerating N=9 N = 11 N=7 8.11 ± 0.55 ab 10.70 ± 1.14 a 6.46 ± 0.83 b 5.64 ± 0.53 7.29 ± 1.02 4.15 ± 0.71 3.43 ± 0.26 a 4.47 ± 0.35 b 2.81 ± 0.35 c 3.62 ± 0.27 a 4.85 ± 0.53 a 2.78 ± 0.50 b 23.2 ± 2.2 a 29.1 ± 4.15 a 15.9 ± 2.84 b  P-value 0.08 0.12 0.01 0.03 0.09  1  Oxalate-extractable Al and Fe (g/kg)  2  Al associated with short-range order inorganic phases, measured as oxalate minus pyrophosphate values  (g/kg) 3  Ferrihydrite and imogolite concentrations (g/kg) as calculated from selective dissolution data (Chapter 6)  Table 7.2: Mean ± SEM of ratio variables (Bf1 horizon only). Variable  Alp1 / Alo2 Fep1 / Feo2 Alsro3 / Alp1 Fesro3 / Fep1  ------------------------Treatment---------------------Control Cleared Regenerating N=9 N = 11 N=7 0.38 ± 0.06 a 0.42 ± 0.05 a 0.57 ± 0.08 b 0.39 ± 0.05 0.42 ± 0.05 0.56 ± 0.09 2.17 ± 0.44 a 1.79 ± 0.41 a 1.03 ± 0.34 b 1.9 ± 0.34 1.87 ± 0.41 1.12 ± 0.39  P-value 0.10 0.16 0.09 0.13  1  Pyrophosphate-extractable Al and Fe  2  Oxalate-extractable Al and Fe  3  Al and Fe associated with short-range order inorganic phases, measured as oxalate minus pyrophosphate  values  140  A CHANGE IN ILLUVIATION INTENSITY? One of the simplest explanations for the higher reactive Al and Fe fractions in the illuvial horizons of cleared plots (Table 7.1) involves a temporary increase in illuviation rate. The concentration of reactive Al and Fe in the illuvial horizons of Podzols is the result of a dynamic equilibrium between progressive illuviation gains, and regressive dissolution and transformation losses (Simonson, 1959). An increase in illuviation rate is likely to result in an increase in SRO Al and Fe phases. In Roberts Creek, factors that may favour an increase in illuviation intensity immediately following logging include (1) accelerated decomposition of organic matter and (2) increase in effective precipitation and soil moisture content (Chapter 3). Organic matter decomposition Decomposing organic matter is a source of soluble organic compounds, including fulvic acids and simple complexing acids such as oxalic and citric acids, which are believed to play a key role in the eluviation of metals from the organic and Ae horizons (Petersen, 1976, De Coninck, 1980, Buurman and Van Reeuwijk, 1984, Jansen et al., 2005). These soluble compounds are leached to the Bf horizon, where precipitation occurs due to adsorption, flocculation, polymerization and/or decomposition of the organic carrier (Skjemstad, 1992, Lundström et al., 2000a, Farmer and Lumsdon, 2001). Decomposition of organic matter and production of dissolved organic compounds has been shown to increase after logging (Qualls et al., 2000, Dai et al., 2001, Kalbitz et al., 2004). In Chapter 3, we presented evidence that organic matter in Roberts Creek logged plots was more mature and more decomposed than in undisturbed plots. In addition, SOC concentration followed the same pattern as reactive Al and Fe phases, with B-horizon SOC concentration being higher in cleared plots and returning to control levels in regenerating plots (Chapter 3). While part of the excess SOC in cleared plots undoubtedly originated from decaying roots, some fraction is likely to have accumulated as a result of increased translocation of dissolved organics from the upper parts of the profile (Strahm et al., 2009). Finally, we observed higher concentrations of organo-metal complexes in 141  the Bf1 horizon of logged plots (Chapter 3). This supports the idea that illuviation intensity was higher in cleared plots. It should be noted that translocation of organo-metal complexes is not the only possible pathway for accumulation of reactive Al and Fe phases to the Bf horizons. In at least some Podzols, illuviated metals may travel as inorganic colloids such as proto-imogolite sols (Farmer et al., 1980, Farmer and Lumsdon, 2001). Precipitation and soil moisture An increase in effective precipitation is usually observed after logging (Keppeler, 1998) due to lack of canopy interception. Correspondingly, we observed a significant increase in soil moisture content in Roberts Creek logged plots (Chapter 3). This may promote leaching processes in the upper part of the profile. In addition, rainwater in Roberts Creek is likely to contain sea salts that can generate hydrochloric and sulphuric acids through ion exchange in acid forest floors (Lagan, 1989). This can further increase metal eluviation (Farmer, 1987).  REGENERATING PLOTS In regenerating plots, the concentration of SRO Al and Fe phases was lower than in cleared plots, and similar to or lower than control plot (Table 1). This suggests that the accumulation of SRO material in illuvial horizons of cleared plots was short-lived. In regenerating plots, the soil moisture regime was similar to control plots (Chapter 3). Soil organic matter content in mineral horizon also returned to levels similar to control plots, as litter inputs from logging slash became reduced. It is therefore reasonable to assume that the illuviation rates returned to levels similar to control plots.  142  Ferrihydrite and ITM concentrations in regenerating plots were lower than in control (Table 7.1). This is rather unexpected, but might be attributed to changes in vegetation type. In regenerating plots, coniferous litter is more rare as alder and pioneer shrub species colonize the site. This may temporarily modify the Podzolization process, causing ITM and ferrihydrite to decrease below control levels. Table 7.2 shows that the fraction of Al bound to organic matter, as opposed to inorganic SRO phases, was significantly higher in regenerating plots. The organic and SRO inorganic pools are often thought to be in direct competition for Al (Shoji et al., 1993, Van Ranst et al., 2008). Our data suggests that in regenerating plots, formation of metalorganic complexes is favoured over the formation of SRO inorganic phases. A possible explanations involves a pH decrease. We showed in Chapter 5 that the pH of logged plots was lower than control plots, while Chapter 6 suggested that pH is the main determinant of the partition of Al between organic and SRO inorganic pools.  CONCLUSION We observed moderately higher concentration of reactive Al and Fe phases, including ITM and ferrihydrite, in the illuvial horizons of cleared plots. Regenerating plots had lower concentrations of SRO Al and Fe phases, suggesting that the increase in SRO material in cleared plots was short-lived. We hypothesize that an intensification of the eluviation – illuviation process afer logging may be responsible for these changes. In addition, the decrease in pH may have favoured the partition of Al towards organic pools in regenerating plots. These data suggest that in Roberts Creek Podzols, SRO inorganic phases are very dynamic and respond to changes in environmental soil conditions on a sub-decadal scale. Our data set is limited and further studies are needed to confirm our findings and investigate processes that may be responsible for field observations.  143  The rapid changes in concentrations of SRO material are likely to impact ecosystem response to disturbance. Short-range order material, such as ITM and ferrihydrite, have a large specific surface area and display both positive and negative charges. They have the potential to alleviate some of the impacts of logging, such as nutrient losses. In Roberts Creek, we observed an increased retention of exchangeable N in Bf horizons of cleared plots (Chapter 5). The respective role of organic matter and inorganic colloids in N retention should be investigated.  144  CHAPTER EIGHT  General discussion and conclusions  145  OVERVIEW This thesis was organized around 3 major foci: (1) soil organic matter (SOM), Chapters 2-3, (2) exchangeable and labile mineral elements, Chapters 4-5, and (3) short-range order (SRO) material, Chapters 6-7. For each of these foci, the profile characteristics, depth distribution of elements and pedogenic considerations were examined first (Chapters 2, 4 and 6). The effects of logging were subsequently investigated (Chapters 3, 5 and 7). One of the functions of Chapters 2, 4 and 6 was to describe conditions in undisturbed forest plots, thus establishing a baseline against which the effects of logging can be considered. Beyond establishing a baseline, these chapters also make standalone contributions to our understanding of Roberts Creek soils in particular and acid forest soils in general by describing key relationships between variables. They further our knowledge of SOM predictors (Chapter 2), acid forest soil chemistry (Chapter 4), and SRO material distribution and significance (Chapter 6). The effects of logging were studied in Chapters 3, 5 and 7. Because soil functions as an integrated biological, chemical and physical system, there was a large amount of interrelation between the information presented in each chapter. For instance, the variations in effective cation exchange capacity (CECe) were related to changes in SOM quantity and quality, since organic matter is believed to bear a large proportion of the soil’s exchange sites (Chapter 4). The dynamics of SRO inorganic materials and SOM also showed a large degree of interdependence. Short-range order material may contribute to stabilization of SOM (Chapter 2), while organic matter influenced the partition of metals between organic and inorganic pools (Chapter 7). The parallel examination of pedogenic considerations and effects of disturbance was useful, since many of the effects of logging could be understood as a change in the balance of pedogenic processes. Organic matter (Chapter 3) and Al and Fe reactive fractions (Chapter 7) were higher in the illuvial horizons of cleared plots and lower in  146  regenerating plots, suggesting that logging affected the intensity of the eluviation – illuviation process (Chapter 7).  SYNTHESIS Pedogenic considerations The horizon effect Chapters 2, 4 and 6 all began by presenting the concentration of elements and fractions in each horizon. As shown by the depth distribution of SOM, Al and Fe pools, the process of Podzolization was the dominant pedogenic influence on Roberts’ Creek soils. Horizon differentiation was pronounced. There was a sharp delineation between the FH layer and the mineral soil, due to the lack of mixing by soil animals. The Ae horizon composition also contrasted sharply with the underlying illuvial horizons, and there were strong indications that soil processes and regulators were distinct from B horizons. For instance, Chapter 2 showed that SOM quantity and quality were different in the Ae and B horizons. Organic matter concentration was lower than in the underlying Bf horizons, while the carbon concentration of SOM (C:SOM ratio) was the highest in the profile. This suggests that more oxidized, polar compounds with a low C concentration were preferentially leached from the Ae horizon. The Ae horizon also had the lowest amounts of extractable Al and Fe in the profile, as is expected in Podzols. The amount of extractable Al was very low, under 1 g/kg for all extractants (Chapter 6). In contrast with illuvial horizons, the largest pool of extractable Fe was dithionite-extractable (Fed), which indicates that Fe was present in crystalline oxides (Chapter 6). There was a positive correlation (r = 0.85, p = 0.01) between Fed and the clay content, suggesting that the Fe oxides occurred in association such as coatings with clay minerals. 147  Together with the FH layer, the Ae was the most acidic horizon in the profile. It also had the lowest base saturation at only 30% (Chapter 4). Finally, a distinct feature of the Ae was that a significant portion of the exchange sites was provided by mineral components, as indicated by a plot of CECe versus organic carbon concentration (Chapter 4, Fig. 4.2). Overall, the Ae horizon was chemically, biochemically and mineralogically distinct from the underlying mineral horizons. On the other hand, the Bf1, Bf2, BCg and top part of the Cg horizons functioned as a pedogenic unit devoid of notable contrasts or boundaries. In the B-C unit, changes in extractable elements were gradational from one horizon to the next. The sharp break in physical properties between the BCg and Cg horizon was not accompanied by a break in chemical properties, indicating that pedochemical processes operated ahead of the physical breakdown of the parent material (Chapter 6). Implication for treatment effect The marked differences between the FH, Ae and illuvial horizons properties were reflected in the observed effects of logging. The interaction between treatment and horizon was generally significant, meaning that logging had different impacts on each of these layers. Table 8.1 presents a global summary of the changes observed between control, cleared and regenerating plots for the FH, Ae and illuvial horizons. Each horizon recorded changes in a distinct set of variables. For some variables, the timing of the change was different, as for instance for the decrease in carbon to nitrogen ratio (Chapter 3). Some variables even showed opposite directions of change in different horizons. For example, C and N stocks of regenerating plots were higher in the FH layer but lower in illuvial horizons (Chapter 3). In cleared plots, exchangeable potassium decreased in FH but increased in illuvial horizons (Chapter 5).  148  Table 8.1: Direction of significant differences observed between treatments for each soil layer (FH, A and B). The first set of arrows depicts the difference between control and cleared plots. The second set of arrows represents the difference between cleared and regenerating plots. Variable  control  SOC Silt-sized SOM Sand-sized SOM C stock N stock C:N C:SOM N:SOM CECe:C moisture (Al+Fe)p C: (Al+Fe)p Total P N:P Total K Total Na Total Fe Total Zn pHH2O pHCaCl2 CECe Alexch Kexch Naexch NO3exch Pexch Clexch Alo Feo Ferrihydrite ITM Alp:Alo Alsro:Alp Acronyms:  FH  A  ↘  ↗  ↗  ↘ ↗ ↘ ↗ ↗ ↘  ↘ ↘ ↘ ↘  ↗  B ↗ ↗ ↗ ↗ ↗  FH  ↗ ↗  ↘ ↗  B ↘ ↘ ↘ ↘ ↘  regenerating  ↘  ↘ ↘ ↗  ↘ ↘ ↗ ↗  A  ↗ ↗ ↘ ↘ ↗  ↗  ↗ ↘  cleared  ↘ ↘  ↗ ↘ ↘  ↘ ↘ ↘ ↘ ↗ ↘  SOC: soil organic carbon SOM: soil organic matter CECe: effective cation exchange capacity Mp: pyrophosphate-extractable element Mo: oxalate-extractable element Msro: element associated with SRO material, calculated as the difference between oxalate and pyrophosphate extract Mexch: exchangeable or salt-extractable element  149  The possible reason for these trends were discussed in details in previous chapters, but the overall interaction between horizon and treatment effect has implications for the interpretation of any work on the soil response to disturbance. As mentioned in Chapters 2 and 3, it is not uncommon for studies of logging impact to limit sampling to the topsoil. This is likely to produce biased results. Table 8.1 shows that in Roberts Creek, the majority of the differences between control and logged plots were recorded in illuvial horizons. A sampling program restricted to the forest floor and A horizon would therefore overlook most of the logging impacts. Furthermore, sampling is sometimes conducted by depth intervals (e.g. Borchers and Perry, 1992, Parker et al., 2001, Bock and Van Rees, 2002). Because the effects of logging are different in different horizons, the interpretation of depth-interval sampling programs is problematic. Large variations can arise depending on the percentage of each horizon present in a given depth interval, and mask treatment effects. Podzolization and the C stabilization potential Forests soils of the Pacific Coast are known to be potentially large sinks for carbon. We found indeed that soils of Roberts Creek stored large amounts of carbon compared to inland Canadian forests (Chapter 2). Roberts Creek C stocks were approximately 50% higher than reported averages for well-drained forest soils of central Canada or central British Columbia (Bhatti et al., 2002, Bois et al., 2009). About 60% of Roberts Creek soil’s carbon was found at depth greater than 20 cm (Chapter 2). After logging, changes in the subsoil C pools were the major determinants of overall changes in C pools (Chapter 3), suggesting that both stable and fast-cycling C forms were present at depth. This demonstrates that it is particularly important to consider the entire profile when conducting studies of C dynamics. Organically complexed Fe and Al, SRO inorganic material and clay minerals were major predictors of C concentration in the soil profile (Chapter 2), suggesting that organomineral and organo-metallic interactions play an important role in C stabilization in Roberts Creek. Organically complexed and SRO forms of Al and Fe were stronger predictors of SOC than crystalline clays and are likely to be a major determinant of C 150  stabilization. Because Al and Fe pools are highly dynamic in Podzols, the concept of SOM protective capacity or C saturation (Six et al., 2002, Gulde et al., 2008) may need to be revisited. The C saturation concept implies that the mineral phase has a maximum sorptive or protective capacity (Baldock and Skjemstad, 2000). Once that capacity is reached, additional dissolved organic carbon will not be retained. Since large amounts of Al and Fe are continually translocated from the upper part of the profile to the illuvial horizons, pedogenically active Podzols may not have a fixed C sorptive capacity (Kalbitz and Kaiser, 2008). Instead, the potential for organo-mineral associations may fluctuate with time depending on weathering rates and the strength of the eluviation / illuviation processes. In Roberts Creek, there are indications that the eluviation / illuviation balance fluctuates rapidly following changes in environmental conditions. We found that the SOM concentration, organically-complexed metals and SRO inorganic materials concentrations were all significantly higher in Bf horizons of harvested plots than in undisturbed plots (Chapters 3 and 7). Possible causes included an increase in effective precipitation and the effect of large additions of fresh organic matter to the litter layer as logging slash. Organic matter decomposition is likely very active after logging due to the increased temperature, moisture and substrate availability (Binkley, 1984, Londo et al., 1999, Diochon and Kellman, 2009). Decomposition generates mobile low-molecular weight organic compounds that are involved in metal translocation (Petersen, 1976, De Coninck, 1980, Buurman and Van Reeuwijk, 1984). While the effects of logging on the eluviation/ illuviation processes are likely to be transient, longer term changes in environmental conditions, such as global warming, have the potential to permanently alter the illuviation steady state condition. As a result, the potential for SOM protection by interaction with Al and Fe phases may be affected.  151  Impacts of logging Logging as ecosystem de-regulation Clear-cutting resulted in the disruption of biogeochemical processes that were fairly constant and predictable in the undisturbed mature forests. We observed changes in mean values of variables such as SOC, pH, exchangeable K, NO3 and SRO phases, among others (Table 8.1). More to the point, we observed an increase in the variability of soil attributes. This is the sign that the loss of biomass resulted in de-regulation of the ecosystem (Bormann and Likens, 1979). Even when the difference in means between control and harvested plots was small, harvested plots had a high likelihood of producing extreme values, as for example for soil NO3 (Chapter 5). Ecosystem resilience Even though we measured significant effects of logging, these effects were relatively mild when compared to other ecosystems, particularly the much studied northern hardwood (e.g. Likens et al., 1970, Covington, 1981, Hornbeck et al., 1987, Wang et al., 2006). We therefore suggest that the coniferous ecosystem could be more resilient to the impacts of logging (Chapters 3 and 5). In Roberts Creek, the overall resilience to the impacts of logging was largely attributable to the good retention of SOM in logged plots (Chapter 3). Organic matter is believed to be responsible for most of the soil’s cation exchange capacity (Chapter 4), and thus contribute to nutrient retention in the profile of cleared plots, where the absence of vegetation uptake makes nutrients most susceptible to leaching. Organic matter decomposition is also a source of nutrients that may be particularly important for the nutrition of regenerating vegetation. The retention of organic matter in soils depends on the dynamic equilibrium between SOM inputs and outputs. Chapter 3 presented evidence that in logged plots, both inputs and outputs of organic matter increased. The increase in organic matter inputs to the forest floor was believed to be due to the additions of logging slash, while inputs to the mineral soil likely included root decay and illuviation of soluble organic compounds. 152  Chapter 3 also presented indications that organic matter decomposition increased and resulted in the production of a more mature and oxidized SOM pool in regenerating plots. We proposed in Chapter 3 that key factors contributing to the maintenance of a sizable C stock in logged plots included (1) the abundance of logging slash inputs and (2) cool climate and substrate recalcitrance preventing runaway decomposition rates. For these reasons, climate change and modifications of logging practices may impair the homeostatic capacity of the soil system. Whole-tree harvesting is currently not a widespread forestry practice in British Columbia, but may receive growing consideration in the future as the demand for biomass to produce bioenergy products increases. By dramatically reducing the amounts of logging slash inputs, whole-tree harvesting is likely to decrease ecosystem resilience to the effects of logging. Logging and the Podzolization process In Roberts Creek soils, the process of eluviation – illuviation emerged as a key factor for understanding the effects of logging. Increases in SOM concentration in illuvial horizons of logged plots, leaching of labile elements such as K, Na and Cl from the forest floor, and changes in the concentration of SRO material (including imogolite-type material, ITM) all suggest that the eluviation – illuviation intensity was higher in cleared plots and decreased in regenerating plots (Chapters 3, 5 and 7). This is a plausible hypothesis, given the fact that logging is known to affect the effective precipitation amounts as well as dissolved organic matter dynamics (Chapter 7). We thus propose that even a shortterm disturbance such as logging can have a measurable impact on pedogenic processes. We further hypothesize that the Podzolization process also played a role in determining the degree of soil resilience to the effects of logging. The translocation of dissolved organic matter, Al and Fe species to the subsoil and their subsequent precipitation is likely to have made a significant contribution to the retention of organic C stocks in profiles of logged plots. This allowed nutrient cations such as K to be retained on organic exchange complexes, as suggested in Chapter 5. Anionic nutrients such as NO3 may also have been retained by electrostatic interactions with SRO Al and Fe illuviates bearing positive charges. 153  SIGNIFICANCE This study differentiates itself from classic studies of ecosystem response to logging because (1) the entire solum was investigated, rather than focusing only on the forest floor and topmost mineral soil, and (2) special attention was paid to SRO inorganic phases. By investigating the entire solum, we showed that the Podzolic B horizons play an important role in soil response to logging. Organic carbon and other elements, such as exchangeable potassium, increased in the illuvial horizons of harvested plots. This suggests that elements leached from the forest floor after logging can be at least temporarily retained in the illuvial horizons. The mineral sub-soil therefore shows the potential to mitigate logging impacts and should be taken into account in any study of the soil response to environmental disturbance. We found that SRO inorganic material constituted a major part of the soil’s reactive fraction (Chapter 6). In Chapter 7, we presented indications that the concentration of SRO phases was impacted by logging. To our knowledge, it is the first report of the impact of a temporary land cover change on SRO material dynamics. The changes in SRO phases concentration was attributed to changes in illuviation rates. The composition of Podzolic horizons results from the sum of two competing processes, addition of illuviated material and losses of material through leaching, decomposition or transformation. Logging disturbance has the potential to shift this balance, temporarily increasing the amount of illuviated material and SRO phases in the Podzolic horizons. Chapter 7 presented evidence consistent with a rapid (sub-decadal) response of SRO inorganic phases to disturbance. Because of the importance of SRO material for processes such as C stabilization and anion retention, this finding has significant environmental implications.  154  LIMITATIONS Statistical limitations The degree of replication in this study was low, which is inherent to large experiments designed to study the effects of disturbance on ecosystem processes. The difficulty of locating suitable sites and the time and funds required to sample and analyze soils to the depth of the parent material limits the level of replication. Sampling in this study was restricted to the Roberts Creek study forest. Within the forest, 9 soil pits were located on undisturbed forested plots, 11 were located in cleared stands and 7 in regenerating stands (Chapter 1). One of the consequences of a low degree of replication and limited sample size is a low power in statistical tests and a high probability for type II error (false negative) (Eberhardt and Thomas, 1991). Treatment effects were further confounded by the large magnitude of within-plot variability and the increase in variance in logged plots (heteroskedasticity). Only large logging effects could be detected, commonly in the range of 35 - 150% change. It is likely that we missed some moderate treatment effects. Because there was no replication at the forest or watershed scale, caution should be exercised if attempting to extrapolate these results to other forests. Published previous experiments can serve as a partial substitute for replication, so that Roberts Creek findings that are in agreement with the current literature are likely to be applicable to other forests in similar biogeoclimatic conditions (Coastal Western Hemlock, Chapter 1). On the other hand, there is a chance that results unsupported by the literature only reflect a local effect. Analytical limitations One of the main limitations of this study was that the quantitative determination of SRO phases was indirect. Quantitative SRO material measurements were based on results of selective dissolution analysis including pyrophosphate, oxalate and dithionite extractions (Chapter 1). These extractions are routinely used to determine SRO fractions and are one 155  of the most reliable mean to estimate SRO material concentration (Dahlgren, 1994), but as with any chemical extractant, the question of selectivity remains (Bascomb, 1968, Sheldrick, 1984, Kaiser and Zech, 1996). At this time however, there is no other method able to quantify low (i.e., < 10%) concentrations of SRO material in whole soil samples, especially if the material has a very low degree of crystallinity. Powder X-ray diffraction is able to identify relatively well-crystallized imogolite and allophane (Gustafsson et al., 1999), but fails to detect products with low crystalline order. Nuclear magnetic resonance is a useful method to qualitatively differentiate between types of allophanic material (Harsh, 2000). Transmission electron microscopy and infrared spectroscopy are well suited to the detection of ITM, but are only semi-quantitative at best in the absence of a rigorous calibration (Dahlgren, 1994). Calibration is problematic since SRO phases found in soil differ from synthetic species in properties such as the degree of crystallinity, presence of defects, association with low molecular weight organic molecules, etc. Another drawback of microscopic and spectroscopic methods is that they require separation or at least concentration of the SRO material. This is very difficult to achieve quantitatively, as a portion of SRO material may be lost during each phase of the purification process. To name a few of the challenges, the use of hydrogen peroxide to remove organic matter may dissolve some of the ITM (Lavkulich and Wiens, 1970, Siregar et al., 2004, Mikutta et al., 2005b). The use of sodium hypochlorite is less harmful to minerals, but residual Na interferes with flocculation of crystalline clays and must be removed by successive centrifugation steps or dialysis, causing more opportunities for losses of SRO material. The use of a dithionite treatment is recommended to remove iron oxides and aid in the dispersion of SRO material, but ITM with a very low degree of structural order may be dissolved by the treatment (Sheldrick, 1984, Dahlgren, 1994). On the other hand, incomplete dispersion of SRO material causes losses during the collection of the acid dispersible fraction (Chapter 1). Given all these difficulties, we argue that the use of selective dissolution is the most reliable quantitative measure of SRO material at this time.  156  DIRECTIONS FOR FUTURE RESEARCH The dynamics of Al emerged as an important determinant of soil processes. Chapter 6 established the significance of SRO Al phases to the soil’s reactive inorganic fraction. Chapter 2 pointed to the importance of Al to SOM retention, and suggested the existence of a mechanism for upward translocation of Al into the FH horizon. Investigations into the magnitude, mechanism, and specificity of metal transport into the FH horizon would be of great interest to the field of SOM dynamics. Chapter 2 established that the most important predictors of SOC concentration were reactive Al and Fe forms. As noted in this chapter, since Al and Fe are dynamic components being continually translocated to the illuvial horizons, the potential for SOM stabilization may not be a fixed soil property but should fluctuate as a result of short-term disturbance or long-term global change. The consequences for SOM retention should be investigated. Many of the logging impacts we observed were relatively short-lived. Most nutrients, Al and Fe fractions, and SOC pools returned to control levels in regenerating plots, at least partly due to the quick re-establishment of vegetative cover (Chapter 3 and 5). We saw some indirect evidence, however, that the bulk composition of soil organic matter had been altered and that organic compounds were more mature and more oxidized in regenerating plots than in both control and cleared plots. The proportion of Al and Fe associated with organic material was also greater in regenerating plots. A study including plots logged 15 to 30+ years prior to sampling is needed to establish whether differences observed in regenerating plots are part of longer term changes in soil attributes. Chapter 7 suggested that SRO material is significantly influenced by forest harvesting. Due to the reactive nature of SRO phases and their importance in processes such as organo-mineral interaction and nutrient sorption, a detailed knowledge of SRO material dynamics under changing environmental conditions is critical to our understanding of ecosystem response to the effects of logging. Further studies are needed to confirm our findings and investigate processes that may be responsible for field observations. 157  BIBLIOGRAPHY  Aber, J. D., D. B. Botkin, and J. M. Melillo. 1979. Predicting the effects of different harvesting regimes on productivity and yield in northern hardwoods. Can J For Res. 9:10-14. Aber, J. D., J. M. Melillo, C. A. McClaugherty, and K. N. Eshleman. 1983. Potential sinks for mineralized nitrogen following disturbance in forest ecosystems. Ecological Bulletins. 35:179-192. Addison, J. A. 2009. Distribution and impacts of invasive earthworms in Canadian forest ecosystems. Biol Invasions. 11:59-79. Allison, F. E., M. S. Sherman, and L. A. Pinck. 1949. Maintenance of soil organic matter.1. Inorganic soil colloid as a factor in retention of carbon during formation of humus. Soil Sci. 68:463-478. Alvarez, R., and R. S. Lavado. 1998. Climate, organic matter and clay content relationships in the Pampa and Chaco soils, Argentina. Geoderma. 83:127-141. Anderson, D. W., S. Saggar, J. R. Bettany, and J. W. B. Stewart. 1981. Particle size fractions and their use in studies of soil organic matter: I. The nature and distribution of forms of carbon, nitrogen, and sulfur. Soil Sci Soc Am J. 45:767772. Anderson, H. A., M. L. Berrow, V. C. Farmer, A. Hepburn, J. D. Russell, and A. D. Walker. 1982. A reassessment of Podzol formation processes. J Soil Sci. 33:125136. Arai, Y., D. L. Sparks, and J. A. Davis. 2005. Arsenate adsorption mechanisms at the allophane - water interface. Environ Sci Technol. 39:2537-2544. Arrouays, D., I. Vion, and J. L. Kicin. 1995. Spatial analysis and modeling of topsoil carbon storage in temperate forest humic loamy soils of France. Soil Sci. 159:191198. 158  Babel, S., and E. M. Opiso. 2007. Removal of Cr from synthetic wastewater by sorption into volcanic ash soil. Int J Environ Sci Technol. 4:99-107. Bailey, J. S., A. Ramakrishna, and G. Kirchhof. 2008. Relationships between important soil variables in moderately acidic soils (pH ≥5.5) in the highlands of Papua New Guinea and management implications for subsistence farmers. Soil Use Manage. 24:281-291. Baldock, J. A., and J. O. Skjemstad. 2000. Role of the soil matrix and minerals in protecting natural organic materials against biological attack. Org Geochem. 31:697-710. Ballard, R. 1980. Phosphorus nutrition and fertilization of forest trees, In: The role of phosphorus in agriculture. F. E. Khasawnek (ed.), ASA, CSSA and SSSA, Madison, WI, pp. 763-804. Baritz, R., G. Seufert, L. Montanarella, and E. Van Ranst. 2010. Carbon concentrations and stocks in forest soils of Europe. For Ecol Manage. 260:262-277. Bartoli, F., A. J. Poulenard, and B. E. Schouller. 2007. Influence of allophane and organic matter contents on surface properties of Andosols. Eur J Soil Sci. 58:450-464. Bascomb, C. L. 1968. Distribution of pyrophosphate-extractable iron and organic carbon in soils of various groups. Eur J Soil Sci. 19:251-268. Basile-Doelsch, I., R. Amundson, W. E. E. Stone, D. Borschneck, J. Y. Bottero, S. Moustier, F. Masin, and F. Colin. 2007. Mineral control of carbon pools in a volcanic soil horizon. Geoderma. 137:477-489. Basile-Doelsch, I., T. Brun, D. Borschneck, A. Masion, C. Marol, and J. Balesdent. 2009. Effect of land use on organic matter stabilized in organomineral complexes: A study combining density fractionation, mineralogy and δ13C. Geoderma. 151:77-86. Belleau, A., S. Brais, and D. Pare. 2006. Soil nutrient dynamics after harvesting and slash treatments in boreal aspen stands. Soil Sci Soc Am J. 70:1189-1199. Bengtsson, J., and F. Wikstrom. 1993. Effects of whole-tree harvesting on the amount of soil carbon: model results. New Zealand Journal of Forestry Science. 23:380-389. 159  Bennett, J. N., L. L. Blevins, J. E. Barker, D. P. Blevins, and C. E. Prescott. 2003. Increases in tree growth and nutrient supply still apparent 10 to 13 years following fertilization and vegetation control of salal-dominated cedar-hemlock stands on Vancouver Island. Can J For Res. 33:1516-1524. Benton Jones Jr., J. 2001. Laboratory guide for conducting soil tests and plant analysis. CRC press, Boca Raton. Bhatti, J. S., M. J. Apps, and C. Tarnocai. 2002. Estimates of soil organic carbon stocks in central Canada using three different approaches. Can J For Res. 32:805-812. Bingeman, C. W., J. E. Varner, and W. P. Martin. 1953. The effect of the addition of organic materials on the decomposition of an organic soil. Soil Sci Soc Am Proc. 17:34-38. Binkley, D. 1984. Does forest removal increase rates of decomposition and nitrogen release? For Ecol Manage. 8:229-233. Binkley, D., and T. C. Brown. 1993. Forest practices as nonpoint sources of pollution in North America. J Am Water Resour As. 29:729-740. Blais-Stevens, A., B. D. Bornhold, A. E. S. Kemp, J. M. Dean, and A. A. Vaan. 2001. Overview of Late Quaternary stratigraphy in Saanich Inlet, British Columbia: results of Ocean Drilling Program Leg 169S. Marine Geology. 174:3-20. Blevins, L. L., C. E. Prescott, and A. Van Niejenhuis. 2006. The roles of nitrogen and phosphorus in increasing productivity of western hemlock and western redcedar plantations on northern Vancouver Island. For Ecol Manage. 234:116-122. Bock, M. D., and K. C. J. Van Rees. 2002. Forest harvesting impacts on soil properties and vegetation communities in the Northwest Territories. Can J For Res. 32:713724. Bois, C. H., D. T. Janzen, P. T. Sanborn, and A. L. Fredeen. 2009. Contrasting total carbon stocks between ecological site series in a subboreal spruce research forest in central British Columbia. Can J For Res. 39:897-907.  160  Borchers, J. G., and D. A. Perry. 1992. The influence of soil texture and aggregation on carbon and nitrogen dynamics in Southwest Oregon forests and clearcuts. Can J For Res. 22:298-305. Borken, W., B. Ahrens, C. Schulz, and L. Zimmermann. 2011. Site-to-site variability and temporal trends of DOC concentrations and fluxes in temperate forest soils. Global Change Biol. 17:2428-2443. Bormann, F. H., and G. E. Likens. 1979. Reorganization: Loss of biotic regulation, In: Pattern and process in a forested ecosystem: disturbance, development and the steady state based on the Hubbard Brook ecosystem study. Springer-Verlag, New York, pp. 81-102. Boucneau, G., M. Van Meirvenne, and G. Hofman. 1998. Comparing pedotransfer functions to estimate soil bulk density in northern Belgium. Pedologie Themata. 5:67-70. Boudot, J. 1992. Relative efficiency of complexed aluminum noncrystalline Al hydroxide, allophane and imogolite in retarding the biodegradation of citric acid. Geoderma. 52:29-39. Box, G. E. P., and D. R. Cox. 1964. An analysis of transformations. Journal of the Royal Statistical Society Series B. 26:211-252. Bradley, R. L., J. P. Kimmins, and W. L. Martin. 2001a. Post-clearcutting chronosequence in the B.C. Coastal Western Hemlock Zone -- II. Tracking the assart flush. J Sustainable For. 14:23-43. Bradley, R. L., W. L. Martin, and J. P. Kimmins. 2001b. Post-clearcutting chronosequence in the B.C. Coastal Western Hemlock Zone -- IV. Modeling forest floor N dynamics and the possible role of denitrification. J Sustainable For. 14:6991. Brady, N. C. 1990. The nature and properties of soils. Macmillan Publishing Co., New York. Brantley,S. L. 2010. Bedrock to soil: where rocks meet life in the critical zone. Goldschmidt conference abstracts 2010. 74, A118. 161  Brantley, S. L., M. B. Goldhaber, and K. V. Ragnarsdottir. 2007. Crossing disciplines and scales to understand the Critical Zone. Elements. 3:307-314. Bray, R. H., and L. T. Kurtz. 1945. Determination of total, organic, and available forms of phosphorus in soils. Soil Sci. 59:39-45. Broadbent, F. E., R. H. Jackman, and J. McNicoll. 1964. Mineralization of C and N in some New Zealand allophanic soils. Soil Sci. 98:118-132. Brydon, J. E., and S. Shimoda. 1972. Allophane and other amorphous constituents in a Podzol from Nova-Scotia. Can J Soil Sci. 52:465-474. Burt, R. 2004. Soil survey laboratory methods manual version 4.0. Natural Resources Conservation Service, U.S. Government Printing Office , Washington, D.C. Buurman, P., and A. G. Jongmans. 2005. Podzolisation and soil organic matter dynamics. Geoderma. 125:71-83. Buurman, P., F. Peterse, and G. Almendros Martin. 2007. Soil organic matter chemistry in allophanic soils: a pyrolysis-GC/MS study of a Costa Rican Andosol catena. Eur J Soil Sci. 58:1330-1347. Buurman, P., and L. P. Van Reeuwijk. 1984. Proto-imogolite and the process of Podzol formation - a critical note. J Soil Sci. 35:447-452. Byzova, Y. V., T. A. Sokolova, and T. I. Sinani. 1990. Amorphous and weakly crystallized components of fine fractions in soils of the southern far-east. Soviet soil science. 22:98-111. Cade-Menun, B. J., and L. M. Lavkulich. 1997. A comparison of methods to determine total, organic, and available phosphorus in forest soils. Commun Soil Sci Plant Anal. 28:651-663. Carlisle, A., A. H. F. Brown, and E. White. 1967. The nutrient content of tree stem flow and ground flora litter and leachates in a sessile oak (Quercus petraea) woodland. J Ecol. 55:615-627.  162  Chaer, G. M., D. D. Myrold, and P. J. Bottomley. 2009. A soil quality index based on the equilibrium between soil organic matter and biochemical properties of undisturbed coniferous forest soils of the Pacific Northwest. Soil Biol Biochem. 41:822-830. Chappell, H. N., D. W. Cole, S. P. Gessel, and R. B. Walker. 1991. Forest fertilization research and practice in the Pacific Northwest. Nutr Cycling Agroecosyst. 27:129140. Chatterjee, A., G. F. Vance, and D. B. Tinker. 2009. Carbon pools of managed and unmanaged stands of ponderosa and lodgepole pine forests in Wyoming. Can J For Res. 39:1893-1900. Childs, C. W., R. L. Parfitt, and R. Lee. 1983. Movement of aluminium as an inorganic complex in some podzolised soils, New Zealand. Geoderma. 29:139-155. Childs, C. W., N. Masue, and N. Yoshinaga. 1991. Ferrihydrite in volcanic ash soils of Japan. Soil Sci Plant Nutr. 37:299-311. Clague, J. J., and J. L. Luternauer. 1983. Late quaternary geology of southwestern British Columbia. Field trip guidebook. GAC-MAC-CGU Joint Annual Meeting, Victoria, British Columbia. Covington, W. W. 1981. Changes in forest floor organic matter and nutrient content following clear cutting in northern hardwoods. Ecology. 62:41-48. Cradwick, P. D. G., V. C. Farmer, J. D. Russell, C. R. Masson, K. Wada, and N. Yoshinaga. 1972. Imogolite, a hydrated aluminium silicate of tubular structure. Nature Phys Sci. 240:187-189. Crow, S. E., K. Lajtha, T. R. Filley, C. W. Swanston, R. D. Bowden, and B. A. Caldwell. 2009. Sources of plant-derived carbon and stability of organic matter in soil: implications for global change. Global Change Biol. 15:2003-2019. Curtis, R. O., and B. W. Post. 1964. Estimating bulk density from organic matter content in some Vermont forest soils. Soil Sci Soc Am Proc. 28:285-286.  163  Dahlgren, R. A. 1994. Quantification of allophane and imogolite, In: Quantitative methods in soil mineralogy. J. E. Amonette and L. W. Zelazny (eds.), SSSA, Madison, WI, pp. 430-451. Dahlgren, R. A., and C. T. Driscoll. 1994. The effects of whole-tree clear-cutting on soil processes at the Hubbard Brook Experimental Forest, New Hampshire, USA. Plant Soil. 158:239-262. Dahlgren, R. A., and M. Saigusa. 1994. Aluminum release rates from allophanic and nonallophanic Andosols. Soil Sci Plant Nutr. 40:125-136. Dahlgren, R. A., M. Saigusa, and F. C. Ugolini. 2004. The nature, properties and management of volcanic soils, In: Advances in Agronomy. D. L. Sparks (ed.), Academic Press, pp. 113-182. Dahlgren, R. A., and F. C. Ugolini. 1989. Aluminum fractionation of soil solutions from unperturbed and tephra-treated Spodosols, Cascade Range, Washington, USA. Soil Sci Soc Am J. 53:559-566. Dahlgren, R. A., and F. C. Ugolini. 1991. Distribution and characterization of shortrange-order minerals in Spodosols from the Washington cascades. Geoderma. 48:391-413. Dai, K., C. Johnson, and C. Driscoll. 2001. Organic matter chemistry and dynamics in clear-cut and unmanaged hardwood forest ecosystems. Biogeochemistry. 54:51-83. D'Anjou, B. 2001. Effects of dispersed retention harvesting on stand structure and regeneration in a coastal mixed-conifer forest: Summary of year 6 results. Forest Research Technical Report TR-006. Research Section, Vancouver Forest Region, B.C. Ministry of Forestry, Nanaimo, B.C. D'Anjou, B. 2002. Roberts Creek Study Forest: harvesting, windthrow and conifer regeneration within alternative silvicultural systems in Douglas-fir dominated forests on the Sunshine Coast. Forest Research Technical Report TR-018. Research Section, Vancouver Forest Region, B.C. Ministry of Forestry, Nanaimo, B.C. De Coninck, F. 1980. Major mechanisms in formation of spodic horizons. Geoderma. 24:101-128. 164  De Vos, B., M. Van Meirvenne, P. Quataert, J. Deckers, and B. Muys. 2005. Predictive quality of pedotransfer functions for estimating bulk density of forest soils. Soil Sci Soc Am J. 69:500-510. de Wit, H. A., T. Palosuo, G. Hylen, and J. Liski. 2006. A carbon budget of forest biomass and soils in southeast Norway calculated using a widely applicable method. For Ecol Manage. 225:15-26. Dexter, A. R., G. Richard, D. Arrouays, E. A. Czyż, C. Jolivet, and O. Duval. 2008. Complexed organic matter controls soil physical properties. Geoderma. 144:620627. Diochon, A., L. Kellman, and H. Beltrami. 2009. Looking deeper: An investigation of soil carbon losses following harvesting from a managed northeastern red spruce (Picea rubens Sarg.) forest chronosequence. For Ecol Manage. 257:413-420. Diochon, A. C., and L. Kellman. 2009. Physical fractionation of soil organic matter: Destabilization of deep soil carbon following harvesting of a temperate coniferous forest. J Geophys Res -Biogeosci. 114:G01016-9. Dise, N. B., and P. Gundersen. 2004. Forest ecosystem responses to atmospheric pollution: linking comparative with experimental studies. Water Air Soil Poll. 4:207-220. Don, A., and K. Kalbitz. 2005. Amounts and degradability of dissolved organic carbon from foliar litter at different decomposition stages. Soil Biol Biochem. 37:21712179. Dyck, W. J., and D. W. Cole. 1994. Strategies for determining consequences of harvesting and associated practices on long-term productivity, In: Impacts of forest harvesting on long-term site productivity. W. J. Dyck, D. W. Cole and N. B. Comerford (eds.), Chapman & Hall, London, UK, pp. 13-40. Easthouse, K. B., J. Mulder, N. Christophersen, and H. M. Seip. 1992. Dissolved organic carbon fractions in soil and stream water during variable hydrological conditions at Birkenes, southern Norway. Water Resour Res. 28:1585-1596.  165  Eberhardt, L. L., and J. M. Thomas. 1991. Designing environmental field studies. Ecol Monogr. 61:pp. 53-73. Egli, M., A. Mirabella, G. Sartori, D. Giaccaid, R. Zanelli, and M. Plötze. 2007. Effect of slope aspect on transformation of clay minerals in Alpine soils. Clay Min. 42:373398. Egli, M., A. Mirabella, A. Mancabelli, and A. Sartori. 2004. Weathering of soils in alpine areas as influenced by climate and parent material. Clays Clay Miner. 52:287-303. Egli, M., M. Nater, A. Mirabella, S. Raimondi, M. Plötze, and L. Alioth. 2008. Clay minerals, oxyhydroxide formation, element leaching and humus development in volcanic soils. Geoderma. 143:101-114. Elliott, H. A., and D. L. Sparks. 1981. Electrokinetic behavior of a Paleudult profile in relation to mineralogical composition. Soil Sci. 132:402-409. Environment Canada. 2011. Canadian climate normals 1971 - 2000. Available from: http://climate.weatheroffice.gc.ca/climate_normals/index_e.html (verified 2011). Essington, M. E. 2004. Soil and water chemistry: an integrative approach. CRC press, Boca Raton, Florida. Eusterhues, K., C. Rumpel, and I. Kögel-Knabner. 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. Eusterhues, K., C. Rumpel, M. Kleber, and I. Kögel-Knabner. 2003. Stabilisation of soil organic matter by interactions with minerals as revealed by mineral dissolution and oxidative degradation. Org Geochem. 34:1591-1600. Evans, A., L. W. Zelazny, and C. E. Zipper. 1988. Solution parameters influencing dissolved organic-carbon levels in 3 forest soils. Soil Sci Soc Am J. 52:1789-1792. Fahey, T., T. Siccama, C. Driscoll, G. Likens, J. Campbell, C. Johnson, J. Battles, J. Aber, J. Cole, M. Fisk, P. Groffman, S. Hamburg, R. Holmes, P. Schwarz, and R. Yanai. 2005. The biogeochemistry of carbon at Hubbard Brook. Biogeochemistry. 75:109-176. 166  Farmer, V. C. 1987. The role of inorganic species in the transport of aluminium in podzols, In: Podzols and podzolization. D. Righi and A. Chauvel (eds.), INRA, Paris, pp. 187-194. Farmer, V. C., M. J. Adams, A. R. Fraser, and F. Palmieri. 1983. Synthetic imogolite properties, synthesis, and possible applications. Clay Min. 18:459-472. Farmer, V. C., and D. G. Lumsdon. 2001. Interactions of fulvic acid with aluminium and a proto-imogolite sol: the contribution of E-horizon eluates to podzolization. Eur J Soil Sci. 52:177-188. Farmer, V. C., and D. G. Lumsdon. 2002. A re-interpretation of 'Aluminium solubility mechanisms in moderately acid Bs horizons of podzolized soils' by Gustafsson et al. Eur J Soil Sci. 53:671-673. Farmer, V. C., and J. D. Russell. 1990. Structures and genesis of allophanes and imogolite and their distribution in non-volcanic soils, In: Soil colloids and their associations in aggregates. M. F. De Boodt, H. B. Hayes and A. Herbillon (eds.), Plenum Press, New York, pp. 165-177. Farmer, V. C., J. D. Russell, and M. L. Berrow. 1980. Imogolite and proto-imogolite allophane in spodic horizons: evidence for a mobile aluminium silicate complex in podzol formation. J Soil Sci. 31:673-684. Federer, C. A., and J. W. Hornbeck. 1985. The buffer capacity of forest soils in new England. Water Air Soil Pollut. 26:163-173. Federer, C. A. 1982. Subjectivity in the separation of organic horizons of the forest floor. Soil Sci Soc Am J. 46:1090-1093. Federer, C. A. 1983. Nitrogen mineralization and nitrification: depth variation in four New England forest soils. Soil Sci Soc Am J. 47:1008-1014. Federer, C. A. 1984. Organic-matter and nitrogen-content of the forest floor in even-aged northern hardwoods. Can. J. For. Res. 14:763-767.  167  Federer, C. A., D. E. Turcotte, and C. T. Smith. 1993. The organic fraction-bulk density relationship and the expression of nutrient content in forest soils. Can J For Res. 23:1026-1032. Fontaine, S., S. Barot, P. Barre, N. Bdioui, B. Mary, and C. Rumpel. 2007. Stability of organic carbon in deep soil layers controlled by fresh carbon supply. Nature. 450:277-280. Gabriel, C. E., and L. Kellman. 2011. Examining moisture and temperature sensitivity of soil organic matter decomposition in a temperate coniferous forest soil. Biogeosciences Discussions. 8:1369-1409. Garcia-Pausas, J., P. Casals, L. Camarero, C. Huguet, R. Thompson, M. Sebastià, and J. Romanyà. 2008. Factors regulating carbon mineralization in the surface and subsurface soils of Pyrenean mountain grasslands. Soil Biol Biochem. 40:28032810. Gartzia-Bengoetxea, N., A. González-Arias, A. Merino, and I. Martínez de Arano. 2009. Soil organic matter in soil physical fractions in adjacent semi-natural and cultivated stands in temperate Atlantic forests. Soil Biol Biochem. 41:1674-1683. Gessel, S. P., D. W. Cole, and E. C. Steinbrenner. 1973. Nitrogen balances in forest ecosystems of the Pacific Northwest. Soil Biol Biochem. 5:19-34. Gholz, H. L., and R. F. Fisher. 1982. Organic-matter production and distribution in slash pine (Pinus-Elliottii) plantations. Ecology. 63:1827-1839. Giesler, R., T. Andersson, L. Lovgren, and P. Persson. 2005. Phosphate sorption in aluminum- and iron-rich humus soils. Soil Sci Soc Am J. 69:77-86. Gilkes, R. J. 1994. Transmission electron microscope analysis of soil materials, In: Quantitative methods in soil mineralogy. J. E. Amonette and L. W. Zelazny (eds.), Soil Science Society of America, Madison, WI, pp. 430-451. Girard, I., and R. A. Klassen. 2001. A comparison of seven methods for analysis of carbon in soils. Current research E 11. Geological survey of Canada, Ottawa, ON.  168  Gottlein, A., and H. Stanjek. 1996. Micro-scale variation of solid-phase properties and soil solution chemistry in a forest podzol and its relation to soil horizons. Eur J Soil Sci. 47:627-636. Grand, S., and L. M. Lavkulich. 2008. Reactive soil components and pedogenesis of highly productive coastal Podzols. Geochim Cosmochim Acta. 72:A323. Gravelle, J. A., G. Ice, T. E. Link, and D. L. Cook. 2009. Nutrient concentration dynamics in an inland Pacific Northwest watershed before and after timber harvest. For Ecol Manage. 257:1663-1675. Grenon, F., R. L. Bradley, G. Joanisse, B. D. Titus, and C. E. Prescott. 2004. Mineral N availability for conifer growth following clearcutting: responsive versus nonresponsive ecosystems. For Ecol Manage. 188:305-316. Guggenheim, S., and A. F. Koster van Groos. 2001. Baseline studies of the Clay Mineral Society source clays: thermal analysis. Clays Clay Miner. 49:433-443. Gulde, S., H. Chung, W. Amelung, C. Chang, and J. Six. 2008. Soil carbon saturation controls labile and stable carbon pool dynamics. Soil Sci Soc Am J. 72:605-612. Gundersen, P., I. Callesen, and W. de Vries. 1998. Nitrate leaching in forest ecosystems is related to forest floor C/N ratios. Environ Pollut. 102:403-407. Gundersen, P., and L. Rasmussen. 1990. Nitrification in forest soils: effects from nitrogen deposition on soil acidification and aluminum release. Rev Environ Contam T. 113:1-45. Gustafsson, J. P. 2001. Modelling competitive anion adsorption on oxide minerals and an allophane-containing soil. Eur J Soil Sci. 52:639-653. Gustafsson, J. P., P. Bhattacharya, D. C. Bain, A. R. Fraser, and W. J. McHardy. 1995. Podzolisation mechanisms and the synthesis of imogolite in northern Scandinavia. Geoderma. 66:167-184. Gustafsson, J. P., E. Karltun, and P. Bhattacharya. 1998. Allophane and imogolite in Swedish soils. TRITA-AMI 3046. Royal Institute of Technology (KTH), Stockholm, Sweden. 169  Gustafsson, J. P. 2001. The surface chemistry of imogolite. Clays Clay Miner. 49:73-80. Gustafsson, J. P., P. Bhattacharya, and E. Karltun. 1999. Mineralogy of poorly crystalline aluminium phases in the B horizon of Podzols in southern Sweden. Appl Geochem. 14:707-718. Hach company. 2009. Lachat intruments. Available from: www.lachatinstruments.com (verified 2010). Hannam, K. D., S. A. Quideau, B. E. Kishchuk, S. W. Oh, and R. E. Wasylishen. 2005. Forest-floor chemical properties are altered by clear-cutting in boreal mixedwood forest stands dominated by trembling aspen and white spruce. Can J For Res. 35:2457-2468. Harsh, J. 2000. Poorly crystalline aluminosilicate clays, In: Handbook of Soil Science. M. E. Sumner (ed.), CRC press, pp. F-169-F-181. Hartikainen, H., K. Rasa, and P. J. A. Withers. 2010. Phosphorus exchange properties of European soils and sediments derived from them. Eur J Soil Sci. 61:1033-1042. Haslett, J., and S. J. Haslett. 2007. The three basic types of residuals for a linear model. Intern Statis Rev. 75:1-24. Hazlett, P. W., A. M. Gordon, R. P. Voroney, and P. K. Sibley. 2007. Impact of harvesting and logging slash on nitrogen and carbon dynamics in soils from upland spruce forests in northeastern Ontario. Soil Biol Biochem. 39:43-57. Hendershot, W. H., H. Lalande, and M. Duquette. 1993. Ion exchange and exchangeable cations, In: Soil sampling and methods of analysis. M. R. Carter (ed.), Lewis publishers, Boca Raton, FL, pp. 167-176. Hendrickson, O. Q., L. Chatarpaul, and D. Burgess. 1989. Nutrient cycling following whole-tree and conventional harvest in northern mixed forest. Can J For Res. 19:725-735. Henmi, T., and K. Wada. 1976. Morphology and composition of allophane. Am Miner. 61:379-390.  170  Hernandez Moreno, J. M., V. Cubas Garcia, A. Gonzalez Batista, and E. Fernandez Caldas. 1985. Study of ammonium oxalate reactivity at pH 6.3 (ro) in different types of soils with variable charge. I, In: Volcanic soils: weathering and landscape relationships of soils on tephra and basalt. E. Fernandez Caldas and D. H. Yaalon (eds.), Catena Verlag, pp. 9-16. Heuscher, S. A., C. C. Brandt, and P. M. Jardine. 2005. Using soil physical and chemical properties to estimate bulk density. Soil Sci Soc Am J. 69:51-56. Hiebler-Chariarse, J. A. 2003. Effect of forest harvesting on biological processes in headwater streams within the Flume creek experimental watershed, British Columbia. M.Sc. dissertation, UBC, Vancouver. Hocking, R. R. 1976. The analysis and selection of variables in linear regression. Biometrics. 32:1-49. Homann, P. S., P. Sollins, M. Fiorella, T. Thorson, and J. S. Kern. 1998. Regional soil organic carbon storage estimates for Western Oregon by multiple approaches. Soil Sci Soc Am J. 62:789-796. Homann, P. S., M. Harmon, S. Remillard, and E. A. H. Smithwick. 2005. What the soil reveals: Potential total ecosystem C stores of the Pacific Northwest region, USA. For Ecol Manage. 220:270-283. Hornbeck, J. W., C. W. Martin, R. S. Pierce, F. H. Bormann, G. E. Likens, and J. S. Eaton. 1987. The Northern Hardwood forest ecosystem: ten years of recovery from clearcutting. USDA Forest Service Research Paper NE-RP-596. Northeastern Forest Experiment Station, Broomall, PA. Huang, P. M., and A. Violante. 1986. Influence of organic acids on crystallization and surface properties of precipitation products of aluminum, In: Interactions of soil minerals with natural organics and microbes. P. M. Huang and M. Schnitzer (eds.), Soil science society of america, Madison, Wisconsin, pp. 159-222. Hudson, R. 2001. Roberts Creek Study Forest: preliminary effects of partial harvesting on peak streamflow in two S6 creeks. Forest Research Technical Report EN-007.  171  Research Section, Vancouver Forest Region, B.C. Ministry of Forestry, Nanaimo, B.C. Hudson, R., and B. D'Anjou. 2001. Roberts Creek Study Forest: the effects of shelterwood harvesting and blowdown on sediment production in a small zero-order creek. Forest Research Technical Report EN-004. Research Section, Vancouver Forest Region, B.C. Ministry of Forestry, Nanaimo, B.C. Hudson, R., and L. Tolland. 2002. Roberts Creek Study Forest: effects of partial retention harvesting on nitrate concentrations in two S6 creeks three years after harvesting. Forest Research Technical Report TR-019. Research Section, Vancouver Forest Region, B.C. Ministry of Forestry, Nanaimo, B.C. Huntington, T. G. 1989. Carbon, organic-matter and bulk-density relationships in a forested spodosol. Soil Sci. 148:380-386. Huntington, T. G., and D. F. Ryan. 1990. Whole-tree-harvesting effects on soil nitrogen and carbon. For Ecol Manage. 31:193-204. Ingerslev, M., E. Malkonen, P. Nilsen, H. O. Nohrstedt, H. Oskarsson, and K. RaulundRasmussen. 2001. Main findings and future challenges in forest nutritional research and management in the Nordic countries. Scand J For Res. 16:488-501. ISO standards. 1996. Determination of nitrite nitrogen and nitrate nitrogen and the sum of both by flow analysis and spectrophotometric detection. Water quality ISO 13395. ISO central office, Geneva, Switzerland. Isoyama, M., and S. Wada. 2007. Remediation of Pb-contaminated soils by washing with hydrochloric acid and subsequent immobilization with calcite and allophanic soil. J Hazard Mater. 143:636-642. IUSS Working Group WRB. 2006. World reference base for soil resources 2006. FAO, Rome. Jansen, B., K. G. J. Nierop, and J. M. Verstraten. 2005. Mechanisms controlling the mobility of dissolved organic matter, aluminium and iron in podzol B horizons. Eur J Soil Sci. 56:537-550. 172  Jansen, B., K. G. J. Nierop, and J. M. Verstraten. 2003. Mobility of Fe(II), Fe(III) and Al in acidic forest soils mediated by dissolved organic matter: influence of solution pH and metal/organic carbon ratios. Geoderma. 113:323-340. Jenny, H. 1950. Causes of the high nitrogen and organic matter content of certain tropical forest soils. Soil Sci. 69:63-70. Jiang, H., M. J. Apps, C. Peng, Y. Zhang, and J. Liu. 2002. Modeling the influence of harvesting on Chinese boreal forest carbon dynamics. For Ecol Manage. 169:65-82. Jobbagy, E. G., and R. B. Jackson. 2000. The vertical distribution of soil organic carbon and its relation to climate and vegetation. Ecol Appl. 10:423-436. Jock Churchman, G. 2010. Is the geological concept of clay minerals appropriate for soil science? A literature-based and philosophical analysis. Phys Chem Earth. 35:927940. John, B., T. Yamashita, B. Ludwig, and H. Flessa. 2005. Storage of organic carbon in aggregate and density fractions of silty soils under different types of land use. Geoderma. 128:63-79. Johnson, C. E. 1989. Effects of whole-tree harvesting on chemical weathering rates in a small forested catchment. Trans Am Geophys Union. 70:1115. Johnson, C. E. 1995. Soil-nitrogen status 8 years after whole-tree clear-cutting. Can J For Res. 25:1346-1355. Johnson, C. E. 2002. Cation exchange properties of acid forest soils of the northeastern USA. Eur J Soil Sci. 53:271-282. Johnson, C. E., R. B. Romanowicz, and T. G. Siccama. 1997. Conservation of exchangeable cations after clear-cutting of a northern hardwood forest. Can J For Res. 27:859-868. Johnson, C. E., A. H. Johnson, T. G. Huntington, and T. G. Siccama. 1991a. Whole-tree clear-cutting effects on soil horizons and organic-matter pools. Soil Sci Soc Am J. 55:497-502.  173  Johnson, C. E., A. H. Johnson, and T. G. Siccama. 1991b. Whole-tree clear-cutting effects on exchangeable cations and soil acidity. Soil Sci Soc Am J. 55:502-508. Johnson, C. E., J. J. Ruiz-Mendez, and G. B. Lawrence. 2000. Forest soil chemistry and terrain attributes in a Catskills watershed. Soil Sci Soc Am J. 64:1804-1814. Johnson, D. W. 1992. Effects of forest management on soil carbon storage. Water Air Soil Pollut. 64:83-120. Johnson, J. E., D. W. Smith, and J. A. Burger. 1985. Effects on the forest floor of wholetree harvesting in an Appalachian oak forest. Am Midl Nat. 114:51-61. Johnson, M. G., and M. B. Mcbride. 1989. Mineralogical and chemical characteristics of Adirondack Spodosols - Evidence for paracrystalline and noncrystalline aluminosilicate minerals. Soil Sci Soc Am J. 53:482-490. Jongmans, A. G., L. Denaix, F. van Oort, and A. Nieuwenhuyse. 2000. Induration of C horizons by allophane and imogolite in Costa Rican volcanic soils. Soil Sci Soc Am J. 64:254-262. Kadono, A., S. Funakawa, and T. Kosaki. 2009. Factors controlling potentially mineralizable and recalcitrant soil organic matter in humid Asia. Soil Sci Plant Nutr. 55:243-251. Kaiser, K., and W. Zech. 1996. Defects in estimation of aluminum in humus complexes of podzolic soils by pyrophosphate extraction. Soil Sci. 161:452-458. Kaiser, K., K. Eusterhues, C. Rumpel, G. Guggenberger, and I. Kögel-Knabner. 2002. Stabilization of organic matter by soil minerals - investigations of density and particle-size fractions from two acid forest soils. J Plant Nutr Soil Sci. 165:451-459. Kalbitz, K., B. Glaser, and R. Bol. 2004. Clear-cutting of a Norway spruce stand: implications for controls on the dynamics of dissolved organic matter in the forest floor. Eur J Soil Sci. 55:401-413. Kalbitz, K., and K. Kaiser. 2008. Contribution of dissolved organic matter to carbon storage in forest mineral soils. J Plant Nutr Soil Sci. 171:52-60.  174  Kalra, Y. P., and D. G. Maynard. 1991. Methods manual for forest soil and plant analysis. Forestry Canada, Northwest Region, Northern Forestry Centre, Edmonton, Alberta. Karltun, E., D. C. Bain, J. P. Gustafsson, H. Mannerkoski, E. Murad, U. Wagner, A. R. Fraser, W. J. McHardy, and M. Starr. 2000. Surface reactivity of poorly-ordered minerals in podzol B horizons. Geoderma. 94:265-288. Kaste, J. M., B. C. Bostick, A. M. Heimsath, E. Steinnes, and A. J. Friedland. 2011. Using atmospheric fallout to date organic horizon layers and quantify metal dynamics during decomposition. Geochim Cosmochim Acta. 75:1642-1661. Katou, H., B. E. Clothier, and S. R. Green. 1996. Anion transport involving competitive adsorption during transient water flow in an andisol. Soil Sci Soc Am J. 60:13681375. Kaufhold, S., R. Dohrmann, Z. Abidin, T. Henmi, N. Matsue, L. Eichinger, A. Kaufhold, and R. Jahn. 2010. Allophane compared with other sorbent minerals for the removal of fluoride from water with particular focus on a mineable Ecuadorian allophane. Appl Clay Sci. 50:25-33. Keeney, D. R. 1980. Prediction of soil nitrogen availability in forest ecosystems: a literature review. For Sci. 26:159-171. Keppeler, E. T. 1998. The summer flow and water yield response to timber harvest, In: Proceedings of the conference on coastal watersheds: the Caspar Creek story. R. R. Ziemer (ed.), USDA, Forest Serv Pacific SW Forest & Range exptl stn, Berkeley, CA, pp. 35-43. Kettler, T. A., J. W. Doran, and T. L. Gilbert. 2001. Simplified method for soil particlesize determination to accompany soil-quality analyses. Soil Sci Soc Am J. 65:849852. Kimmins, J. P. 1997. Forest ecology: A foundation for sustainable management. PrenticeHall, Inc., Upper Saddle River, NJ. King, P. L., M. S. Ramsey, P. F. McMillan, and G. Swayze. 2004. Laboratory fourier transform infrared spectroscopy methods for geologic samples, In: Infrared spectroscopy in geochemistry, exploration geochemistry, and remote sensing. P. L. 175  King, M. S. Ramsey and G. Swayze (eds.), Mineralogical Association of Canada, London, Ontario, pp. 54-39. Kleber, M., R. Mikutta, M. S. Torn, and R. Jahn. 2005. Poorly crystalline mineral phases protect organic matter in acid subsoil horizons. Eur J Soil Sci. 56:717-725. Kleber, M., L. Schwendenmann, E. Veldkamp, J. Rößner, and R. Jahn. 2007. Halloysite versus gibbsite: Silicon cycling as a pedogenetic process in two lowland neotropical rain forest soils of La Selva, Costa Rica. Geoderma. 138:1-11. Kleja, D. B., W. Standring, D. H. Oughton, J. -. Gustafsson, K. Fifield, and A. R. Fraser. 2005. Assessment of isotopically exchangeable Al in soil materials using 26Al tracer. Geochim Cosmochim Acta. 69:5263-5277. Klopatek, J. M., M. J. Barry, and D. W. Johnson. 2006. Potential canopy interception of nitrogen in the Pacific Northwest, USA. For Ecol Manage. 234:344-354. Knoepp, J. D., and B. D. Clinton. 2009. Riparian zones in southern Appalachian headwater catchments: Carbon and nitrogen responses to forest cutting. For Ecol Manage. 258:2282-2293. Knops, J. M. H., and K. L. Bradley. 2009. Soil carbon and nitrogen accumulation and vertical distribution across a 74-year chronosequence. Soil Sci Soc Am J. 73:20962104. Kodama, H., and C. Wang. 1989. Distribution and characterization of noncrystalline inorganic components in Spodosols and Spodosol-like soils. Soil Sci Soc Am J. 53:526-534. Kögel-Knabner, I., G. Guggenberger, M. Kleber, E. Kandeler, K. Kalbitz, S. Scheu, K. Eusterhues, and P. Leinweber. 2008. Organo-mineral associations in temperate soils: Integrating biology, mineralogy, and organic matter chemistry. J Plant Nutr Soil Sci. 171:61-82. Kooijman, A. M., and G. B. Martinez-Hernandez. 2009. Effects of litter quality and parent material on organic matter characteristics and N-dynamics in Luxembourg beech and hornbeam forests. For Ecol Manage. 257:1732-1739. 176  Kothawala, D. N., T. R. Moore, and W. H. Hendershot. 2009. Soil properties controlling the adsorption of dissolved organic carbon to mineral soils. Soil Sci Soc Am J. 73:1831-1842. Kranabetter, J. M. 2009. Site carbon storage along productivity gradients of a late-seral southern boreal forest. Can J For Res. 39:1053-1060. Kranabetter, J. M., P. Sanborn, B. K. Chapman, and S. Dube. 2006. The contrasting response to soil disturbance between Lodgepole Pine and Hybrid White Spruce in subboreal forests. Soil Sci Soc Am J. 70:1591-1599. Krull, E. S., J. A. Baldock, and J. O. Skjemstad. 2003. Importance of mechanisms and processes of the stabilisation of soil organic matter for modelling carbon turnover. Functional Plant Biol. 30:207-222. Kurz, W. A., M. Apps, E. Banfield, and G. Stinson. 2002. Forest carbon accounting at the operational scale. For Chron. 78:672-679. Kurz, W., and M. Apps. 2006. Developing Canada's national forest carbon monitoring, accounting and reporting system to meet the reporting requirements of the Kyoto protocol. Mitigation Adapt Strat Global Change. 11:33-43. Kyle, J. H., A. M. Posner, and J. P. Quirk. 1975. Kinetics of isotopic-exchange of phosphate adsorbed on gibbsite. J Soil Sci. 26:32-43. Lagan, S. J. 1989. Sea-salt induced streamwater acidification. Hydrol Process. 3:25-41. Lal, R. 2004. Soil carbon sequestration to mitigate climate change. Geoderma. 123:1-22. Lamontagne, S., R. Carignan, P. D'Arcy, Y. T. Prairie, and D. Pare. 2000. Element export in runoff from eastern Canadian Boreal Shield drainage basins following forest harvesting and wildfires. Can J Fish Aquat Sci. 57:118-128. Lavkulich, L. M., S. Bhoojedhur, and C. A. Rowles. 1971. Soils with placic horizons on the west coast of Vancouver Island. Can J Soil Sci. 51:439-448. Lavkulich, L. M., and K. W. G. Valentine. 1978. The canadian system of soil and soil climate classification, In: The soil landscapes of British Columbia. K. W. G.  177  Valentine, P. N. Sprout, L. M. Lavkulich and T. E. Baker (eds.), B.C. ministry of environment, Victoria, B.C., pp. 59-65. Lavkulich, L. M., and J. H. Wiens. 1970. Comparison of organic matter destruction by hydrogen peroxide and sodium hypochlorite and its effects on selected mineral constituents. Soil Sci Soc Am Proc. 34:755-758. Lawrence, G. B., R. D. Fuller, and C. T. Driscoll. 1987. Release of aluminum following whole-tree harvesting at the Hubbard-Brook-Experimental-Forest, New-Hampshire. J Environ Qual. 16:383-390. Lee, J., I. K. Morrison, J. Leblanc, M. T. Dumas, and D. A. Cameron. 2002. Carbon sequestration in trees and regrowth vegetation as affected by clearcut and partial cut harvesting in a second-growth boreal mixedwood. For Ecol Manage. 169:83-101. Likens, G. E., F. H. Bormann, R. S. Pierce, and W. A. Reiners. 1978. Recovery of a deforested ecosystem. Science. 199:492-496. Likens, G. E., F. H. Bormann, N. M. Johnson, D. W. Fisher, and R. S. Pierce. 1970. Effects of forest cutting and herbicide treatment on nutrient budgets in the Hubbard Brook watershed-ecosystem. Ecol Monogr. 40:23-47. Lilienfein, J., R. G. Qualls, S. M. Uselman, and S. D. Bridgham. 2003. Soil formation and organic matter accretion in a young andesitic chronosequence at Mt. Shasta, California. Geoderma. 116:249-264. Lindsay, W. L., and P. M. Wathall. 1996. The solubility of aluminum in soils, In: The environmental chemistry of aluminum. G. Sposito (ed.), CRC Lewis publishers, Boca Raton, pp. 333-362. Littell, R. C., G. A. Milliken, W. W. Stroup, R. D. Wolfinger, and O. Schabenberger. 2006. SAS for mixed models. SAS Institute Inc., Cary, NC. Londo, A. J., M. G. Messina, and S. H. Schoenholtz. 1999. Forest harvesting effects on soil temperature, moisture, and respiration in a bottomland hardwood forest. Soil Sci Soc Am J. 63:637-644.  178  Lookman, R., N. Vandeweert, R. Merckx, and K. Vlassak. 1995. Geostatistical assessment of the regional distribution of phosphate sorption capacity parameters (FeOX and AlOX) in northern Belgium. Geoderma. 66:285-296. Lundström, U. S., N. van Breemen, and D. Bain. 2000a. The podzolization process. A review. Geoderma. 94:91-107. Lundström, U. S., N. van Breemen, D. C. Bain, P. A. W. van Hees, R. Giesler, J. P. Gustafsson, H. Ilvesniemi, E. Karltun, P. A. Melkerud, M. Olsson, G. Riise, O. Wahlberg, A. Bergelin, K. Bishop, R. Finlay, A. G. Jongmans, T. Magnusson, H. Mannerkoski, A. Nordgren, L. Nyberg, M. Starr, and L. Tau Strand. 2000b. Advances in understanding the podzolization process resulting from a multidisciplinary study of three coniferous forest soils in the Nordic Countries. Geoderma. 94:335-353. Mann, L. K., D. W. Johnson, D. C. West, D. W. Cole, J. W. Hornbeck, C. W. Martin, H. Riekerk, C. T. Smith, W. T. Swank, L. M. Tritton, and D. H. Van Lear. 1988. Effects of whole-tree and stem-only clearcutting on postharvest hydrologic losses, nutrient capital, and regrowth. For Sci. 34:412-428. Marks, P. L., and F. H. Bormann. 1972. Revegetation following forest cutting: mechanisms for return to steady-state nutrient cycling. Science. 176:914-915. Martin, C. W., D. S. Noel, and C. A. Federer. 1984. Effects of forest clearcutting in New England on stream chemistry. J Environ Qual. 13:204-210. Matus, F., X. Amigo, and S. M. Kristiansen. 2006. Aluminium stabilization controls organic carbon levels in Chilean volcanic soils. Geoderma. 132:158-168. Mayer, L. M. 1994. Relationships between mineral surfaces and organic carbon concentrations in soils and sediments. Chem Geol. 114:347-363. McGroddy, M. E., T. Daufresne, and L. O. Hedin. 2004. Scaling of C:N:P stochiometry in forests worldwide: implications of terrestrial Redfied-type ratios. Ecology. 85:2390-2401.  179  McHale, M., D. Burns, G. Lawrence, and P. Murdoch. 2007. Factors controlling soil water and stream water aluminum concentrations after a clearcut in a forested watershed with calcium-poor soils. Biogeochemistry. 84:311-331. McKeague, J. A., and D. H. Day. 1966. Dithionite- and oxalate-extractable Fe and Al as aids in differentiating various classes of soils. Can J Soil Sci. 46:13-22. Meentemeyer, V. 1978. Macroclimate the lignin control of litter decomposition rates. Ecology. 59:465-472. Mehra, O. P., and M. L. Jackson. 1960. Iron oxide removal from soils and clays by dithionite-citrate system buffered with sodium bicarbonate. Clays and Clay Minerals, Proceedings of the 7th National Conference.317. Menounos, B., J. Koch, G. Osborn, J. J. Clague, and D. Mazzucchi. 2004. Early Holocene glacier advance, southern Coast Mountains, British Columbia, Canada. Quat Sci Rev. 23:1543-1550. Mickler, R. A., T. S. Earnhardt, and J. A. Moore. 2002. Regional estimation of current and future forest biomass. Environ Pollut. 116:S7-S16. Mikutta, R., M. Kleber, and R. Jahn. 2005a. Poorly crystalline minerals protect organic carbon in clay subfractions from acid subsoil horizons. Geoderma. 128:106-115. Mikutta, R., M. Kleber, K. Kaiser, and R. Jahn. 2005b. Review: organic matter removal from soils using hydrogen peroxide, sodium hypochlorite, and disodium peroxodisulfate. Soil Sci Soc Am J. 69:120-135. Mikutta, R., M. Kleber, M. Torn, and R. Jahn. 2006. Stabilization of soil organic matter: Association with minerals or chemical recalcitrance? Biogeochemistry. 77:25-56. Ming, D. W., D. W. Mittlefehldt, R. V. Morris, D. C. Golden, R. Gellert, A. Yen, B. C. Clark, S. W. Squyres, W. H. Farrand, S. W. Ruff, R. E. Arvidson, G. Klingelhöfer, H. Y. McSween, D. S. Rodionov, C. Schröder, de Souza, P. A., Jr., and A. Wang. 2006. Geochemical and mineralogical indicators for aqueous processes in the Columbia Hills of Gusev crater, Mars. J Geophys Res. 111:E02S12.  180  Miralles, I., R. Ortega, G. Almendros, M. Sánchez-Marañón, and M. Soriano. 2009. Soil quality and organic carbon ratios in mountain agroecosystems of South-east Spain. Geoderma. 150:120-128. Mizota, C., and L. P. van Reeuwijk. 1989. Clay mineralogy and chemistry of soils formed in volcanic material in diverse climatic regions. ISRIC, Wageningen, The Netherlands. Moran, K. K., J. Six, W. R. Horwath, and C. van Kessel. 2005. Role of mineral-nitrogen in residue decomposition and stable soil organic matter formation. Soil Sci Soc Am J. 69:1730-1736. Morris, D. M. 2009. Changes in DOC and DON fluxes in response to harvest intensity of black-spruce dominated forest ecosystems in northwestern Ontario. Can J Soil Sci. 89:67-79. Mossin, L., M. Mortensen, and P. Nørnberg. 2002. Imogolite related to podzolization processes in Danish podzols. Geoderma. 109:103-116. Mroz, G. D., M. F. Jurgensen, and D. J. Frederick. 1985. Soil nutrient changes following whole tree harvesting on three Northern Hardwood sites. Soil Sci Soc Am J. 49:1552-1557. Müller, K., and C. Duwig. 2007. Transport and sorption of 2,4-dichlorophenoxyacetic acid in allophanic soils. Soil Sci. 172:333-348. Nalder, I. A., and H. G. Merriam. 1995. Simulating carbon dynamics of the boreal forest in Pukaskwa National Park. Water Air Soil Pollut. 82:283-298. Nave, L. E., E. D. Vance, C. W. Swanston, and P. S. Curtis. 2010. Harvest impacts on soil carbon storage in temperate forests. For Ecol Manage. 259:857-866. Navia, R., B. Fuentes, M. C. Diez, and K. E. Lorber. 2005. The use of volcanic soil as mineral landfill liner - III. Heavy metals retention capacity. Waste Manage Res. 23:260-269. Neal, C., B. Reynolds, C. J. Smith, S. Hill, M. Neal, T. Conway, G. P. Ryland, H. Jeffrey, A. J. Robson, and R. Fisher. 1992. The impact of conifer harvesting on stream water 181  pH, alkalinity and aluminium concentrations for the British uplands: an example for an acidic and acid sensitive catchment in mid-Wales. Sci Total Environ. 126:75-87. Nierop, K. G. J., and J. M. Verstraten. 2003. Organic matter formation in sandy subsurface horizons of Dutch coastal dunes in relation to soil acidification. Org Geochem. 34:499-513. Nierop, K. G. J. J., B. Jansen, and J. M. Verstraten. 2002. Dissolved organic matter, aluminium and iron interactions: precipitation induced by metal/carbon ratio, pH and competition. Sci Total Environ. 300:201-211. Nishikiori, H., K. Kobayashi, S. Kubota, N. Tanaka, and T. Fujii. 2010. Removal of detergents and fats from waste water using allophane. Appl Clay Sci. 47:325-329. Nishikiori, H., J. Shindoh, N. Takahashi, T. Takagi, N. Tanaka, and T. Fujii. 2009. Adsorption of benzene derivatives on allophane. Appl Clay Sci. 43:160-163. Norris, C. E., S. A. Quideau, J. S. Bhatti, and R. E. Wasylishen. 2011. Soil carbon stabilization in jack pine stands along the Boreal Forest Transect Case Study. Global Change Biol. 17:480-494. Oades, J. 1988. The retention of organic matter in soils. Biogeochemistry. 5:35-70. Ohle, W. 1953. Phosphor als initialfaktor der gewässereutrophierung. Wom Wasser. 25:127-149. Olsson, B. A., J. Bengtsson, and H. Lundkvist. 1996a. Effects of different forest harvest intensities on the pools of exchangeable cations in coniferous forest soils. For Ecol Manage. 84:135-147. Olsson, B. A., H. Staaf, H. Lundkvist, J. Bengtsson, and R. Kaj. 1996b. Carbon and nitrogen in coniferous forest soils after clear-felling and harvests of different intensity. For Ecol Manage. 82:19-32. Opiso, E., T. Sato, and T. Yoneda. 2009. Adsorption and co-precipitation behavior of arsenate, chromate, selenate and boric acid with synthetic allophane-like materials. J Hazard Mater. 70:79-86.  182  Paavolainen, L., V. Kitunen, and A. Smolander. 1998. Inhibition of nitrification in forest soil by monoterpenes. Plant Soil. 205:147-154. Palviainen, M., L. Finér, R. Laiho, E. Shorohova, E. Kapitsa, and I. Vanha-Majamaa. 2010. Phosphorus and base cation accumulation and release patterns in decomposing Scots pine, Norway spruce and silver birch stumps. For Ecol Manage. 260:1478-1489. Panesar, T. S., V. G. Marshall, and H. J. Barclay. 2000. The impact of clearcutting and partial harvesting systems on population dynamics of soil nematodes in coastal Douglas-fir forests. Pedobiologia. 44:641-665. Parfitt, R. L. 1980. Chemical properties of variable charge soils, In: Soils with variable charge. B. K. G. Theng (ed.), New Zealand society of soil science, Palmerston North, pp. 167-194. Parfitt, R. L. 2009. Allophane and imogolite: role in soil biogeochemical processes. Clay Min. 44:135-155. Parfitt, R. L., and C. W. Childs. 1988. Estimation of forms of Fe and Al: A review, and analyses of contrasting soils by dissolution and Moessbauer methods. Aust J Soil Res. 26:121-144. Parfitt, R. L., C. W. Childs, and D. N. Eden. 1988. Ferrihydrite and allophane in four Andepts from Hawaii and implications for their classification. Geoderma. 41:223241. Parfitt, R. L., and T. Henmi. 1982. Comparison of an oxalate-extraction method and an infrared spectroscopic method for determining allophane in soil clays. Soil Sci Plant Nutr. 28:183-190. Parfitt, R. L., and J. M. Kimble. 1989. Conditions for formation of allophane in soils. Soil Sci Soc Am J. 53:971-977. Parfitt, R. L., M. Russell, and G. E. Orbell. 1983. Weathering sequence of soils from volcanic ash involving allophane and halloysite, New Zealand. Geoderma. 29:4157. 183  Parfitt, R. L., and A. D. Wilson. 1985. Estimation of allophane and halloysite in three sequences of volcanic soils, new zealand, In: Volcanic soils: weathering and landscape relationships of soils on tephra and basalt. E. Fernandez Caldas and D. H. Yaalon (eds.), Catena Verlag, West Germany, pp. 1-8. Parfitt, R. 1990. Allophane in New Zealand - a review. Aust J Soil Res. 28:343-360. Parfitt, R. L., and T. Henmi. 1980. Structure of some allophanes from New Zealand. Clays Clay Miner. 28:285-294. Parker, J. L., I. J. Fernandez, L. E. Rustad, and S. A. Norton. 2002. Soil organic matter fractions in experimental forested watersheds. Water Air Soil Pollut. 138:101-121. Parker, J. L., I. J. Fernandez, L. E. Rustad, and S. A. Norton. 2001. Effects of nitrogen enrichment, wildfire, and harvesting on forest-soil carbon and nitrogen. Soil Sci Soc Am J. 65:1248-1255. Parkinson, J. A., and S. E. Allen. 1975. A wet oxidation procedure suitable for the determination of nitrogen and mineral nutrients in biological material. Commun Soil Sci Plant Anal. 6:1-11. Paul, E. 1984. Dynamics of organic matter in soils. Plant Soil. 76:275-285. Pearson, K. P. 1895. Note on regression and inheritance in the case of two parents. Proc R Soc London. 58:240-242. Pennock, D. J., and C. van Kessel. 1997. Clear-cut forest harvest impacts on soil quality indicators in the mixedwood forest of Saskatchewan, Canada. Geoderma. 75:13-32. Percival, H. J., R. L. Parfitt, and N. A. Scott. 2000. Factors controlling soil carbon levels in New Zealand grasslands: is clay content important? Soil Sci Soc Am J. 64:16231630. Périé, C., and R. Ouimet. 2008. Organic carbon, organic matter and bulk density relationships in boreal forest soils. Can J Soil Sci. 88:315-325. Petersen, L. 1976. Podzols and podzolization. DSR Forlag, Copenhagen.  184  Ping, C. L., G. J. Michaelson, E. S. Kane, E. C. Packee, C. A. Stiles, D. K. Swanson, and N. D. Zaman. 2010. Carbon stores and biogeochemical properties of soils under Black Spruce forest, Alaska. Soil Sci Soc Am J. 74:969-978. Pojar, J., K. Klinka, and D. A. Demarchi. 1991. Coastal western hemlock zone, In: Ecosystems of British Columbia. D. Meidinger and J. Pojar (eds.), B.C. Ministry of Forests, Victoria. Prescott, C. E., L. L. Blevins, and C. L. Staley. 2000. Effects of clear-cutting on decomposition rates of litter and forest floor in forests of British Columbia. Can J For Res. 30:1751-1757. Prevost, M. 2004. Predicting soil properties from organic matter content following mechanical site preparation of forest soils. Soil Sci Soc Am J. 68:943-949. Qing-ru, Z., L. Bo-han, Z. Li-tian, Z. Xi-hong, and T. Hong-xiao. 2006. Short-term alleviation of aluminum phytotoxicity by urea application in acid soils from south China. Chemosphere. 63:860-868. Qualls, R. G., B. L. Haines, W. T. Swank, and S. W. Tyler. 2000. Soluble organic and inorganic nutrient fluxes in clearcut and mature deciduous forests. Soil Sci Soc Am J. 64:1068-1077. Rasmussen, C., R. J. Southard, and W. R. Horwath. 2006. Mineral control of organic carbon mineralization in a range of temperate conifer forest soils. Global Change Biol. 12:834-847. Rasmussen, C. 2007. Soil genesis and mineral transformation across an environmental gradient on Andesitic Lahar. Soil Sci Soc Am J. 71:225-237. Rasmussen, C., M. S. Torn, and R. J. Southard. 2005. Mineral assemblage and aggregates control carbon dynamics in a California conifer forest. Soil Sci Soc Am J. 69:17111721. Reuss, J. O., R. Stottlemyer, and C. A. Troendle. 1997. Effect of clear cutting on nutrient fluxes in a subalpine forest at Fraser, Colorado. hydrol earth syst sc. 1:333-344.  185  Rey, A., E. Pegoraro, and P. G. Jarvis. 2008. Carbon mineralization rates at different soil depths across a network of European forest sites (FORCAST). Eur J Soil Sci. 59:1049-1062. Riise, G., P. Van Hees, U. Lundström, and L. Tau Strand. 2000. Mobility of different size fractions of organic carbon, Al, Fe, Mn and Si in podzols. Geoderma. 94:237-247. Rosén, K., and A. Lundmark-Thelin. 1987. Increased nitrogen leaching under piles of slash— a consequence of modern forest harvesting techniques. Scand J For Res. 2:21-29. Ross, D. S., R. J. Bartlett, and F. R. Magdoff. 1991. Exchangeable cations and the pH independent distribution of cation exchange capacities in spodosols of a forested watershed, In: Plant-soil interactions at low pH. R. J. Wright, V. C. Baligar and R. P. Murmann (eds.), Kluwer Academic Publishers, Dordrecht, the Netherlands, pp. 81-92. Ross, D. S., G. Matschonat, and U. Skyllberg. 2008. Cation exchange in forest soils: the need for a new perspective. Eur J Soil Sci. 59:1141-1159. Ross, G. J. 1980. The mineralogy of spodosols, In: Soils with variable charge. B. K. G. Theng (ed.), New Zealand Society of Soil Science, Palmerston North, pp. 127-143. Ross, G. J., and H. Kodama. 1979. Evidence for imogolite in Canadian soils. Clays Clay Miner. 27:297-300. Ruehlmann, J., and M. Korschens. 2009. Calculating the effect of soil organic matter concentration on soil bulk density. Soil Sci Soc Am J. 73:876-885. Rumpel, C., K. Eusterhues, and I. Kögel-Knabner. 2004. Location and chemical composition of stabilized organic carbon in topsoil and subsoil horizons of two acid forest soils. Soil Biol Biochem. 36:177-190. Russell, J. D., and A. R. Fraser. 1994. Infrared methods, In: Clay mineralogy: spectroscopic and chemical determinative methods. M. J. Wilson (ed.), Chapman & Hall, London, pp. 11-67.  186  Ryan, D. F., T. G. Huntington, and C. Wayne Martin. 1992. Redistribution of soil nitrogen, carbon and organic matter by mechanical disturbance during whole-tree harvesting in northern hardwoods. For Ecol Manage. 49:87-99. Saccone, L., D. J. Conley, E. Koning, D. Sauer, M. Sommer, D. Kaczorek, S. W. Blecker, and E. F. Kelly. 2007. Assessing the extraction and quantification of amorphous silica in soils of forest and grassland ecosystems. Eur J Soil Sci. 58:1446-1459. Salmon, S., L. Frizzera, and S. Camaret. 2008. Linking forest dynamics to richness and assemblage of soil zoological groups and to soil mineralization processes. For Ecol Manage. 256:1612-1623. Salomé, C., N. Nunan, V. Pouteau, T. Z. Lerch, and C. Chenu. 2010. Carbon dynamics in topsoil and in subsoil may be controlled by different regulatory mechanisms. Global Change Biol. 16:416-426. Sanborn, P., and L. M. Lavkulich. 1989. Ferro-humic podzols of coastal BritishColumbia. 2. Micromorphology and genesis. Soil Sci Soc Am J. 53:517-526. SAS Institute Inc. 2008. SAS 9.2. Cary, NC. Sauer, D., L. Saccone, D. J. Conley, L. Herrmann, and M. Sommer. 2006. Review of methodologies for extracting plant-available and amorphous Si from soils and aquatic sediments. Biogeochemistry. 80:89-108. Scheel, T., C. Dorfler, and K. Kalbitz. 2007. Precipitation of dissolved organic matter by aluminum stabilizes carbon in acidic forest soils. Soil Sci Soc Am J. 71:64-74. Schimel, J. P., and J. Bennett. 2004. Nitrogen mineralization: Challenges of a changing paradigm. Ecology. 85:591-602. Schmidt, M. W. I., H. Knicker, and I. Kögel-Knabner. 2000. Organic matter accumulating in Aeh and Bh horizons of a Podzol — chemical characterization in primary organo-mineral associations. Org Geochem. 31:727-734. Schnitzer, M., and H. Kodama. 1992. Interactions between organic and inorganic components in particle-size fractions separated from four soils. Soil Sci Soc Am J. 56:1099-1105. 187  Schofield, R. K., and A. W. Taylor. 1955. The measurement of soil pH. Soil Sci Soc Am Proc. 19:164-167. Schwertmann, U. 1985. The effect of pedogenic environments on iron oxide minerals. Adv Soil Sci. 1:171-200. Schwertmann, U., H. Kodama, and W. R. Fischer. 1986. Mutual interactions between organics and iron oxides. In: Interactions of soil minerals with natural organics and microbes. P. M. Huang and M. Schnitzer (eds.), Soil science society of America, Madison, WI. Schwertmann, U., and R. M. Taylor. 1989. Iron oxides, In: Minerals in soils environments. J. B. Dixon and S. B. Weed (eds.), Soil science society of America, Madison, WI, pp. 1-34. Schwertmann, U., and R. W. Fitzpatrick. 1977. Occurrence of lepidocrocite and its association with goethite in Natal soils. Soil Sci Soc Am J. 41:1013-1018. Seyboldl, C. A., and R. B. Grossman. 2006. Prediction of effective cation exchange capacity in low pH soils. Soil Sci. 171:3-12. Shaw, C. H., J. S. Bhatti, and K. J. Sabourin. 2005. An ecosystem carbon database for Canadian forests. Information Report NOR-X-403. Canadian Forest Service, Northern Forestry Centre, Edmonton, AB. Sheldrick, B. H. 1984. Analytical methods manual. Agriculture Canada, Ottawa, ON. Shen, Y. 1999. Sorption of natural dissolved organic matter on soil. Chemosphere. 38:1505-1515. Shoji, S., Y. Fujiwara, I. Yamada, and M. Saigusa. 1982. Chemistry and clay mineralogy of Ando soils, Brown Forest soils, and Podzolic soils formed from recent Towada ashes, northeastern Japan. Soil Sci. 133:69-86. Shoji, S., M. Nanzyo, and R. A. Dahlgren. 1993. Volcanic ash soils - genesis, properties and utilization. Elsevier, Amsterdam.  188  Shoji, S., M. Saigusa, I. Yamada, T. Takahashi, and F. C. Ugolini. 1988. Properties of Spodosols and Andisols showing climosequential and biosequential relations in Southern Hakkoda, Northeastern Japan. Soil Sci. 145:135-150. Simonson, R. W. 1959. Outline of a generalized theory of soil genesis. Soil Sci Soc Am Proc. 23:301-311. Siregar, A., M. Kleber, R. Mikutta, and R. Jahn. 2004. Sodium hypochlorite oxidation reduces soil organic matter concentrations without affecting inorganic soil constituents. Eur J Soil Science. 56:481-490. Six, J., R. T. Conant, E. A. Paul, and K. Paustian. 2002. Stabilization mechanisms of soil organic matter: Implications for C-saturation of soils. Plant Soil. 241:155-176. Skjemstad, J. O. 1992. Genesis of podzols on coastal dunes in southern Queensland .III. The role of aluminum organic complexes in profile development. Aust J Soil Res. 30:645-665. Skjemstad, J. O., R. W. Fitzpatrick, B. A. Zarcinas, and C. H. Thompson. 1992. Genesis of Podzols on coastal dunes in Southern Queensland .II. Geochemistry and forms of elements as deduced from various soil extraction procedures. Aust J Soil Res. 30:615-644. Skyllberg, U. 1994. Aluminum associated with a pH-increase in the humus layer of a boreal haplic podzol. Interciencia. 19:356-365. Slesak, R. A., T. B. Harrington, and S. H. Schoenholtz. 2010. Soil and Douglas-fir (Pseudotsuga menziesii) foliar nitrogen responses to variable logging-debris retention and competing vegetation control in the Pacific Northwest. Can J For Res. 40:254-264. Smith, W., F. H. Bormann, and G. E. Likens. 1968. Response of chemoautotrophic nitrifiers to forest cutting. Soil Sci. 106:471-473. Smithwick, E. A. H., M. E. Harmon, S. M. Remillard, S. A. Acker, and J. F. Franklin. 2002. Potential upper bounds of carbon stores in forests of the Pacific Northwest. Ecol Appl. 12:1303-1317. 189  Smits, M. M., and E. Hoffland. 2009. Possible role of ectomycorrhizal fungi in cycling of aluminium in podzols. Soil Biol Biochem. 41:491-497. Snyder, K. E., and R. D. Harter. 1985. Changes in solum chemistry following clearcutting of northern hardwood stands. Soil Sci Soc Am J. 49:223-228. Soil Classification Working Group. 1998. The Canadian system of soil classification. NRC Research Press, Ottawa. Soil survey staff. 2006. Keys to Soil Taxonomy. USDA-Natural Resources Conservation Service, Washington, DC. Sollins, P., P. Homann, and B. A. Caldwell. 1996. Stabilization and destabilization of soil organic matter: mechanisms and controls. Geoderma. 74:65-105. Sollins, P., C. Swanston, and M. Kramer. 2007. Stabilization and destabilization of soil organic matter: a new focus. Biogeochemistry. 85:1-7. Spain, A. V. 1990. Influence of environmental conditions and some soil chemical properties on the carbon and nitrogen contents of some tropical Australian rainforest soils. Aust J Soil Res. 28:825-839. Spielvogel, S., J. Prietzel, and I. Kögel-Knabner. 2008. Soil organic matter stabilization in acidic forest soils is preferential and soil type-specific. Eur J Soil Sci. 59:674692. Stephens, M. A. 1974. EDF statistics for goodness of fit and some comparisons. J Am Stat Assoc. 69:730-737. Strahm, B. D., and R. B. Harrison. 2006. Nitrate sorption in a variable-charge forest soil of the Pacific Northwest. Soil Sci. 171:313-321. Strahm, B. D., R. B. Harrison, T. A. Terry, B. L. Flaming, C. W. Licata, and K. S. Petersen. 2005. Soil solution nitrogen concentrations and leaching rates as influenced by organic matter retention on a highly productive Douglas-fir site. For Ecol Manage. 218:74-88. Strahm, B. D., R. B. Harrison, T. A. Terry, T. B. Harrington, A. B. Adams, and P. W. Footen. 2009. Changes in dissolved organic matter with depth suggest the potential 190  for postharvest organic matter retention to increase subsurface soil carbon pools. For Ecol Manage. 258:2347-2352. Šubrt, J., V. Štengl, and M. Skokánek. 1992. Decomposition of ferrihydrite prepared from Fe(NO3)3 aqueous solutions under varying pH. Thermochimica Acta. 211:107-119. Sullivan, T. J., I. J. Fernandez, A. T. Herlihy, C. T. Driscoll, T. C. McDonnell, N. A. Nowicki, K. U. Snyder, and J. W. Sutherland. 2005. Acid-base characteristics of soils in the Adirondack Mountains, New York. Soil Sci Soc Am J. 70:141-152. Tait, J. M., N. Yoshinaga, and B. D. Mitchell. 1978. Occurrence of imogolite in some Scottish soils. Soil Sci Plant Nutr. 24:145-151. Tanskanen, N., and H. Ilvesniemi. 2004. The amount of secondary Al in two ploughed podzolic forest soils. Geoderma. 119:249-260. Tate, K. R., and B. K. G. Theng. 1980. Organic matter and its interactions with inorganic soil constituents. In: Soils with variable charge. B. K. G. Theng (ed.), Soil Bureau, Lower Hutt, New Zealand, pp. 225-249. Thomas, G. W. 1982. Exchangeable cations, In: Methods of soil analysis. A. L. Page (ed.), American Society of Agronomy, Madison, WI, pp. 159-165. Thompson, A., O. A. Chadwick, D. G. Rancourt, and J. Chorover. 2006. Iron-oxide crystallinity increases during soil redox oscillations. Geochim Cosmochim Acta. 70:1710-1727. Tiessen, H., and J. W. B. Stewart. 1983. Particle-size fractions and their use in studies of soil organic-matter .2. Cultivation effects on organic-matter composition in size fractions. Soil Sci Soc Am J. 47:509-514. Tonneijck, F. H., B. Jansen, K. G. J. Nierop, J. M. Verstraten, J. Sevink, and L. D. Lange. 2010. Towards understanding of carbon stocks and stabilization in volcanic ash soils in natural Andean ecosystems of northern Ecuador. Eur J Soil Sci. 61:392-405.  191  Torn, M. S., S. E. Trumbore, O. A. Chadwick, P. M. Vitousek, and D. M. Hendricks. 1997. Mineral control of soil organic carbon storage and turnover. Nature. 389:170173. Tremblay, S., C. Périé, and R. Ouimet. 2006. Changes in organic carbon storage in a 50 year white spruce plantation chronosequence established on fallow land in Québec. Can J For Res. 36:2713-2723. Trumbore, S. 2000. Age of soil organic matter and soil respiration: radiocarbon constraints on belowground C dynamics. Ecol Appl. 10:399-411. Ugolini, F. C., and R. A. Dahlgren. 1991. Weathering environments and occurrence of imogolite / allophane in selected Andisols and Spodosols. Soil Sci Soc Am J. 55:1166-1171. Ussiri, D. A. N., and C. E. Johnson. 2003. Characterization of organic matter in a northern hardwood forest soil by 13C NMR spectroscopy and chemical methods. Geoderma. 111:123-149. Van Lierop, W. 1990. Soil pH and lime requirement determination, In: Soil testing and plant analysis. R. L. Westerman (ed.), Soil Science Society of America, Madison, Wisconsin, USA, pp. 73-92. Van Ranst, E., S. R. Utami, A. Verdoodt, and N. P. Qafoku. 2008. Mineralogy of a perudic Andosol in central Java, Indonesia. Geoderma. 144:379-386. Vance, G. F., F. J. Stevenson, and F. J. Sikora. 1996. Environmental chemistry of aluminum-organic complexes, In: The environmental chemistry of aluminum. G. Sposito (ed.), CRC Lewis publishers, Boca Raton, pp. 169-220. Varadachari, C., A. H. Mondal, and K. Ghosh. 1991. Some aspects of clay-humus complexation: effect of exchangeable cations and lattice charge. Soil Sci. 151:220227. Verdouw, H., C. J. A. Van Echteld, and E. M. J. Dekkers. 1978. Ammonia determination based on indophenol formation with sodium salicylate. Water Res. 12:399-402.  192  Vitousek, P. M. 1977. The regulation of element concentrations in mountain streams in the northeastern United States. Ecol Monogr. 47:65-87. Vitousek, P. M., J. R. Gosz, C. C. Grier, J. M. Melillo, and W. A. Reiners. 1982. A comparative analysis of potential nitrification and nitrate mobility in forest ecosystems. Ecol Monogr. 52:155-177. Vitousek, P. M., J. R. Gosz, C. C. Grier, J. M. Melillo, W. A. Reiners, and R. L. Todd. 1979. Nitrate losses from disturbed ecosystems. Science. 204:469-474. Vitousek, P. M., and R. W. Howarth. 1991. Nitrogen limitation on land and in the sea: How can it occur? Biogeochemistry. 13:87-115. Wada, K. 1989. Allophane and imogolite, In: Minerals in Soil Environments. J. B. Dixon and S. B. Weed (eds.), Soil Science Society of America, Madison, pp. 1051-1087. Waksman, S. A. 1924. Influence of microorganisms upon the carbon-nitrogen ratio in the soil. J Agric Sci. 14:555-562. Wang, C., J. A. McKeague, and H. Kodama. 1986. Pedogenic imogolite and soil environments - Case-study of Spodosols in Quebec, Canada. Soil Sci Soc Am J. 50:711-718. Wang, X., D. A. Burns, R. D. Yanai, R. D. Briggs, and R. H. Germain. 2006. Changes in stream chemistry and nutrient export following a partial harvest in the Catskill Mountains, New York, USA. For Ecol Manage. 223:103-112. Waterhouse, L. F., and A. S. Harestad. 2002. Habitat use of a dispersed retention area by breeding winter wrens. Forest Research Technical Report TR-020. Research Section, Vancouver Forest Region, B.C. Ministry of Forestry, Nanaimo, B.C. Webster, R. 2007. Analysis of variance, inference, multiple comparisons and sampling effects in soil research. Eur J Soil Sci. 58:74-82. Weetman, G. F., C. E. Prescott, F. L. Kohlberger, and R. M. Fournier. 1997. Ten-year growth response of coastal Douglas-fir on Vancouver Island to N and S fertilization in an optimum nutrition trial. Can J For Res. 27:1478-1482. 193  White, C. S. 1994. Monoterpenes: Their effects on ecosystem nutrient cycling. J Chem Ecol. 20:1381-1406. Wilson, M. J. 1999. The origin and formation of clay minerals in soils: past, present and future perspectives. Clay Min. 34:7-25. Woignier, T., E. Braudeau, H. Doumenc, and L. Rangon. 2005. Supercritical drying applied to natural “gels”: allophanic soils. J Sol Gel Sci Technol. 36:61-68. Xu, C., L. Guo, C. Ping, and D. M. White. 2009. Chemical and isotopic characterization of size-fractionated organic matter from cryoturbated tundra soils, northern Alaska. J Geophys Res. 114:G03002. Xu, Z., S. Ward, C. Chen, T. Blumfield, N. Prasolova, and J. Liu. 2008. Soil carbon and nutrient pools, microbial properties and gross nitrogen transformations in adjacent natural forest and hoop pine plantations of subtropical Australia. J Soils Sediments. 8:99-105. Yagasaki, Y., J. Mulder, and M. Okazaki. 2006. Comparing the activity of aluminum in two B horizons developed from volcanic ash deposits in Japan, dominated by shortrange ordered aluminosilicates and crystalline clay minerals, respectively. Geochim Cosmochim Acta. 70:147-163. Yanai, R. D. 1998. The effect of whole-tree harvest on phosphorus cycling in a northern hardwood forest. For Ecol Manage. 104:281-295. Yanai, R. D., W. S. Currie, and C. L. Goodale. 2003. Soil carbon dynamics after forest harvest: an ecosystem paradigm reconsidered. Ecosystems. 6:197-212. Yoshinaga, N., and S. Aomine. 1962. Imogolite in some Ando soils. Soil Sci Plant Nutr. 8:22-29. Young, A. W., A. S. Campbell, and T. W. Walker. 1980. Allophane isolated from a podsol developed on a non-vitric parent material. Nature. 284:46-48. Zackrisson, O., M. Nilsson, and D. A. Wardle. 1996. Key ecological function of charcoal from wildfire in the boreal forest. Oikos. 77:10-19.  194  Zanelli, R., M. Egli, A. Mirabella, M. Abdelmoula, M. Plötze, and M. Nötzli. 2006. 'Black' soils in the southern Alps: Clay mineral formation and transformation, X-ray amorphous Al phases and Fe forms. Clays Clay Miner. 54:703-720. Zinn, Y. L., R. Lal, J. M. Bigham, and D. V. S. Resck. 2007. Edaphic controls on soil organic carbon retention in the Brazilian Cerrado: texture and mineralogy. Soil Sci Soc Am J. 71:1204-1214. Zummo, L. M., and A. J. Friedland. 2011. Soil carbon release along a gradient of physical disturbance in a harvested northern hardwood forest. For Ecol Manage. 261:1016-1026. Zunino, H., F. Borie, S. Aguilera, J. P. Martin, and K. Haider. 1982. Decomposition of 14C-labeled glucose, plant and microbial products and phenols in volcanic ashderived soils of Chile. Soil Biol Biochem. 14:37-43. Zysset, M., P. Blaser, J. Luster, and A. U. Gehring. 1999. Aluminum solubility control in different horizons of a Podzol. Soil Sci Soc Am J. 63:1106-1115.  195  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0105110/manifest

Comment

Related Items