Hydropedology and Surface/Subsurface Runoff Processes

HENRY LIN1, ERIN BROOKS2, PAUL McDANIEL3 AND JAN BOLL2 1Department of Crop and Soil Sciences, Pennsylvania State University, University Park, PA, US 2Department of Biological and Agricultural Engineering, University of Idaho, Moscow, ID, US 3Department of Plant, Soil, and Entomological Sciences, University of Idaho, Moscow, ID, US

The role of soils has long been recognized as critical to rainfall–runoff processes in watersheds. Hydropedology is an emerging interdisciplinary field that integrates pedology, hydrology, geomorphology, and other related bio- and geosciences to study interactive pedologic and hydrologic processes and the landscape–soil–hydrology relationships across space and time. This article presents an overview of hydropedology’s contributions to the understanding and modeling of surface/subsurface runoff processes, especially the diagnosis of soil features that can help answer “why-type” questions in watershed hydrology and the ubiquitous nature of preferential flow and its networks. We highlight two bottlenecks for advancing watershed hydrology and hydropedology: a conceptual bottleneck of modeling subsurface preferential flow networks and a technological bottleneck of nondestructively mapping or imaging subsurface architecture. Quantification of “soil architecture” at various scales and the identification of “hydropedologic functional units” in different landscapes offer promising potentials to advance hydrologic modeling. We present the information of linking surface/subsurface runoff processes to pedologic understanding at three scales of microscopic (macropores and aggregates), mesoscopic (horizons and pedons), and macroscopic (hillslopes and catchments) levels. Various examples from the literature are synthesized to illustrate the key points. Further research needs are suggested in the end.

INTRODUCTION in arid and semiarid regions, where rainfall intensities are high and soil infiltration capacity is reduced because of The role of soils (soil properties and their distribution surface sealing or pavement. The primary motivation for over the landscape) has long been recognized as criti- many of the early studies was to determine the magnitude cal to rainfall–runoff processes in a watershed. Hydrope- and return period of extreme events, as in many areas of the dology, as an intertwined branch of soil science and world these extreme events have been caused by infiltration- hydrology, embraces multiscale studies of interactive pedo- excess runoff. logic and hydrologic processes and their properties in the An alternative runoff generation paradigm, called vari- variably saturated and unsaturated surface and subsurface able source area (VSA) hydrology, has been suggested environments, and therefore possesses significant potential over four decades ago (Hewlett and Hibbert, 1967). The in enhancing the understanding and prediction of rain- VSA approach to watershed response suggests that runoff fall–runoff processes (Lin, 2003; Lin et al., 2006a). contributing to a storm hydrograph is generated from local- The traditional view of surface runoff generation has ized (typically near-stream) areas of the landscape. The been based on the infiltration excess,orHortonian, concept controlling mechanism is termed saturation-excess runoff of runoff production (Horton, 1940), where surface runoff is generation because of precipitation falling on soils that have the result of rainfall intensity exceeding infiltration capacity little or no available storage (typically high soil moisture at the soil surface. This surface runoff, sometimes also content or groundwater level), thereby precluding infiltra- called unsaturated overland flow, occurs more commonly tion. This “saturated overland flow” occurs more frequently

Encyclopedia of Hydrological Sciences. Edited by M G Anderson.  2008 John Wiley & Sons, Ltd. 2 RAINFALL–RUNOFF PROCESSES in riparian zones, valley bottoms, or local depressional areas hillslope and watershed scale so thresholdlike when the soil, in humid or semihumid regions. As a consequence, water- climate, vegetation, and water appear so tightly coupled? shed hydrologists must consider not only soil properties and This article attempts to shed light on some of these ques- their spatial distributions, but also the interactive roles of tions from a hydropedologic perspective. In particular, two topography, land use/land cover, geology, and other compo- bottlenecks are identified and discussed, where hydropedol- nents of the hydrologic cycle besides precipitation (Gburek ogy can make a unique contribution to advancing watershed et al., 2006). hydrology, including the quantification of soil architecture Another runoff, called subsurface return flow (or reflow), at various scales and the identification of hydropedologic occurs after water infiltrates the soil on an upslope por- functional units (HFUs) in different landscapes. tion of a hillslope and then flows laterally through the soil The article provides an overview of the current under- and exfiltrates (resurfaces) in the toeslope area or closer to standing of hydropedology’s role in rainfall–runoff pro- stream channel (Whipkey, 1965; Dunne et al., 1975). This cesses, especially the diagnosis of soils that can help answer overland flow is related to subsurface stormflow (also called “why-type” questions in watershed hydrology. We illustrate subsurface runoff, interflow, throughflow, lateral flow, tran- various aspects that show the diagnostic importance of soils sient groundwater, or some other names) (see Chapter 112, in rainfall–runoff processes at three scales – microscopic Subsurface Stormflow, Volume 3). Subsurface stormflow (macropores and aggregates), mesoscopic (horizons and is another runoff-producing mechanism operating in most pedons), and macroscopic (hillslopes and catchments) lev- upland terrains, especially in humid environment and steep els. Some research needs for the future are suggested in the terrain with conductive soils or soils with a water-restricting end. layer. While some studies have documented subsurface stormflow as unsaturated flow in the vadose zone, most studies have shown that subsurface stormflow is saturated HYDROPEDOLOGY AND ITS FUNDAMENTAL (or near-saturated) – due to either the rise of an existing QUESTIONS water table into more transmissive soil above (with ensuing Hydropedology, as an emerging interdisciplinary field, inte- lateral flow) or the transient saturation above an impeding grates pedology, hydrology, geomorphology, and other layer, soil–bedrock interface, or some zone of reduced per- related bio- and geosciences to study interactive pedo- meability (e.g. fragipan, argillic horizon, and other dense logic and hydrologic processes and the landscape–soil– layers in the soil profile). hydrology relationships across space and time (Lin, 2003; Field investigations conducted in various humid forested Lin et al., 2005a, 2006a). It aims to understand pedologic catchments throughout the world have also revealed the sig- controls on hydrologic processes and properties (e.g. soil nificance of subsurface preferential flow at multiple spatial as a central link in the hydrologic cycle and as a founda- and temporal scales. While overland flow is limited in these tion for rainfall–runoff processes) and hydrologic impacts forested catchments, evidence of subsurface preferential on soil formation, variability, and functions (e.g. hydrology runoff is ubiquitous (Sidle et al., 2001; McDonnell, 2003; as a critical driving force of soil formation and functional Lin, 2006). Growing evidence suggests that the subsur- dynamics). Essentially, hydropedology seeks to answer the face flow network is a key to understanding rainfall–runoff following two fundamental questions: processes including threshold behavior (Sidle et al., 2001; Weiler et al., 2005; McDonnell et al., 2007; Lin and Zhou, 1. How the structure and distribution of soils over the 2008), and that flow pathways often determine the hot spots landscape exert a first-order control on landscape water and moments of biogeochemical processes (Band et al., across spatiotemporal scales in the surface and the 2001; McClain et al., 2003; Boyer et al., 2006). This sug- subsurface? gests that a new thinking beyond the VSA paradigm is 2. How landscape water (and associated transport of needed (Sidle et al., 2000; McDonnell, 2003; McDonnell chemicals and energy by flowing water) impacts soil et al., 2007). The dynamic origin of network structures in evolution, variability, and functions? soils and hydrologic systems and recurrent patterns of self- organization are the subjects of recent research and model Landscape water here encompasses the source, storage, development (Rinaldo et al., 2006; James and Roulet, 2007; availability, flux, pathway, residence time, and spatiotem- Weiler and McDonnell, 2007; Lehmann et al., 2007). poral distribution of water (and the transport of chemicals There are significant questions remaining in hillslope andenergybyflowingwater)inthevariablysaturated hydrology and rainfall–runoff prediction. For example, and unsaturated surface and subsurface (Lin et al., 2006a). McDonnell et al. (2007) raised a number of such ques- While source, storage, availability, and flux of water have tions in their new vision of watershed hydrology: Why does been studied extensively in the past, attention to flow heterogeneity exist? Why is there preferential, networklike pathways (especially flow networks), residence time (age flow at all scales? Why are hydrological connections at the of water), and spatiotemporal pattern of flow dynamics HYDROPEDOLOGY AND SURFACE/SUBSURFACE RUNOFF PROCESSES 3

Hydrologic processes (water)

The The pedon The landscape Critical paradigm paradigm Zone

Pedologic Geomorphic processes processes (soil) (landscape)

Solar energy Basic structure 1 µm

Precipitation The Critical Zone Human use Transpiration Evaporation Runoff Ped Hydroecology 1 dm Soils Microscopic scale Soils (root zone) Ground water flow Subsurface flow Hydropedology Infiltration realm Pedosphere Deep vadose zone Deep vadose zone (pores partially 1 km filled with water) Water table Hydrogeology Macroscopic scale realm Ground water zone (pores entirely Ground water zone filled with water) 1 m Mesoscopic scale

Figure 1 Hydropedology as an interdisciplinary science that promotes integrated studies of hydrologic, pedologic, and geomorphic processes across spatial and temporal scales. It connects the pedon and landscape paradigms through linking phenomena occurring at the microscopic (e.g. pores and aggregates) to the mesoscopic (e.g. horizons and pedons) and the macroscopic (e.g. catenas and watersheds) levels, and eventually to the megascopic (e.g. regional and global) scales

(and its underlying organizing principles) has been limited sediment found in shallow permanently flooded envi- (McDonnell et al., 2007). ronments such as in an estuary). Hydropedology distinguishes itself from vadose zone • Hydropedology is not just about soils’ impacts on hydrology (or soil hydrology) in the following aspects: hydrology; equally important is hydrologic feedbacks to pedogenesis, why subsurface heterogeneity exists, and • Hydropedology emphasizes in situ soils in the land- how that impacts soil variability and soil functions. scape, where distinct pedogenic features (e.g. aggre- gation, horizonation, and redox features) and soil– As illustrated in Figure 1, hydropedology attempts to landscape relationships (e.g. catena, soil distribution connect the pedon and landscape paradigms through linking patterns, HFUs) are essential in understanding interac- phenomena occurring at microscopic (e.g. pores and aggre- tive pedologic and hydrologic processes, hence requir- gates) to mesoscopic (e.g. horizons and pedons) and macro- ing adequate attention to soil architecture, soil layering, scopic (e.g. catenae and watersheds) levels, and eventually soil mapping, soil morphology, soil geomorphology, to megascopic (e.g. regional and global) scales. Hydrope- and stratigraphy. dology, combined with hydrogeology, promotes an inte- • Hydropedology deals with the variably saturated and grated systems approach to study the interactions of water unsaturated zone in the surface and near-surface envi- with solid earth (i.e. soils, rocks, and anything in between) ronments, including shallow root zone, deep vadose beneath the earth’s surface. Above the ground, hydropedol- zone, temporally saturated zone, capillary fringe, wet- ogy interfaces with hydroecology and hydrometeorology lands, and even subaqueous soils (i.e. soils that form in to understand the feedback mechanisms between climate, 4 RAINFALL–RUNOFF PROCESSES

Soil process Degree of generalization Map scale Model scale of soil distribution and parameters + Globe i 4

Order 5 World soil map Upscaling Great (larger area) + + Region i 3 Small 1:100 000 000

Aggregation Order 5 NATSGO Macroscopic (larger area) Landscape (watershed) i + 2

1:7 500 000

Order 4 STATSGO Field (catena) i + 1

1:250 000

Order 3, Order 2 Mesoscopic Pedon i SSURGO

1:24 000 Profile horizon i − 1

Order 1 Components Ped (aggregate) Microscopic i − 2

1:12 000 Disaggregation <

(smaller area) Little Pedon Mixture Local Downscaling i − 3 (smaller area) 1:1 Large

Molecular i − 4 Point

(a) (b)

Figure 2 Hierarchical framework for bridging forms (soil distribution) and functions (soil processes) in hydropedology: (a) soil mapping hierarchy for depicting soil spatial distributions at different scales. The SSURGO, STATSGO, and NATSGO are county-, state-, and national-level soil maps, respectively, generated from different orders of soil surveys; (b) soil modeling hierarchy for understanding dynamic soil processes across scales. The pedon is the base level (ith scale), and other levels are defined with reference to it, with i+ levels for upscaling and i –levels for downscaling. These scale numbers (i’s), however, do not reflect quantitative relationships among scales; rather, they are used to aid in conceptual understanding vegetation, and soils. In the temporal dimension, hydrope- two hierarchies, however, have not yet been established. dology deals with both short- and long-term changes of the This is in part due to the mismatch of scales used in landscape–soil–hydrology relationships, including the use mapping and modeling. In a spatial sense, “aggregation” of pedogenesis and soil micro- and macromorphology as and “disaggregation” are used for mapping soil distribution, a signature of soil change and soil hydrology in the past which are defined irrespective of a process-based model and/or current conditions. (Figure 2a). In contrast, “upscaling” and “downscaling” are used in modeling soil processes and/or quantifying model Hierarchical Frameworks for Bridging Forms inputs/outputs, and thus are defined in the context of a and Functions in Hydropedology specific model (Figure 2b). The meaning of “scale” also Two hierarchical frameworks were suggested to bridge the carries different meanings between mapping and modeling: forms and functions in hydropedology (Lin and Rathbun, in cartography, map scale refers to a ratio of map to reality, 2003): one is soil mapping hierarchy that deals with and the scale becomes smaller as spatial information is soil spatial heterogeneity, and the other is soil modeling aggregated for a larger area; whereas in the modeling arena, hierarchy that addresses temporal dynamics (Figure 2). The scale is often used in a colloquial sense (without a specific soil mapping hierarchy relates to the “forms” of soils quantifier), so large scale loosely refers to a large area that depict the spatial pattern of soil types or specific extent. soil properties over the landscape of varying sizes, while Another constraint in connecting the forms and functions the soil modeling hierarchy relates to the “functions” of of soils for hydrologic modeling is the gap between soils that portray soil’s physical, chemical, and biological scales used in either mapping or modeling hierarchies. processes at different scales. Explicit links between these Although there are approximations that might be taken HYDROPEDOLOGY AND SURFACE/SUBSURFACE RUNOFF PROCESSES 5 to connect the scales of attribute-based mapping (such by planes of weakness (Figure 3). Three attributes – ped as spatial interpolation from point observations to areal grade, shape, and size – are used together to describe a soil coverages) or the scales of process-based modeling (such structure. For example, “strong fine granular soil structure” as extrapolation of smaller-area model parameters to larger- means a soil that separates almost entirely into discrete units area values), significant knowledge gaps remain, and direct of peds (i.e. strong grade) that are loosely packed, roughly translation from one scale to another is still lacking. Since spherical (i.e. granular), and mostly between 1 and 2 mm in different factors or processes may dominate at different diameter (i.e. fine size). Some soils have simple pedality, scales, and emerging new phenomena may occur as scale each unit being an entity without component smaller unit; changes, a hierarchical set of models – based on available others have compound pedality, in which large units are mapped attributes – may be constructed. As we focus on composed of smaller units separated by persistent planes of system patterns and filter out unimportant details, we might weakness (Figure 3c). also begin to see emergent features that could serve as Besides the shape, size, and grade of peds, the internal a natural skeleton to connect descriptions of soil and surface features of peds are also described in soil surveys, hydrologic responses across scales (Sivapalan, 2003). This consisting of (i) coats of a variety of substances unlike could then form the basis of models that are inherently the adjacent soil material and covering part or all of the simpler at the macroscales, but with sufficient links to surfaces, (ii) concentration of material on surfaces caused essential detailed heterogeneity and complexity observed by the removal of other materials, and (iii) stress formations at the microscales (McDonnell et al., 2007). Such an idea in which thin layers at the surfaces have undergone is somewhat similar to a hierarchy of identifiable dynamical reorientation or packing by stress or shear (Soil Survey patterns of processes, called successive dynamical levels,as Division Staff, 1993). The kinds of structural surface suggested by Baveye and Boast (1999). features include clay films, clay bridges, sand or silt coats, Jenny (1941) suggested the union of soil maps and soil other coats, stress surfaces, and slickensides (see examples functions to provide the most effective means of pedo- in Figure 3). All of them have significant impacts on flow logical research. At least two aspects hydropedology can andtransportprocessesinfieldsoils. contribute uniquely to such needed bridging of “forms” and Pores are considered separately in the U.S. Soil Survey’s “functions” and of different scales: One is through defining concept of soil structure; however, in Europe, Canada, and quantifying soil architecture at various scales, and the Australia, and some other countries, pore-related features other is through defining and identifying HFUs in different (e.g. pore-size distribution, connectivity, and tortuosity) are landscapes. Both efforts can advance hydrologic model- an integral part of soil structure (e.g. Brewer, 1976). Thus, ing via appropriate causal relationships and soil–landscape the US concept of soil structure is better termed pedality. datasets of comparable resolutions. Pedality and soil pore space are interrelated. But many soils have interpedal, intrapedal, and/or transpedal pores Defining a Broad Concept of ‘‘Soil Architecture’’ that are not necessarily represented by pedality. These and its Quantification at Different Scales pores, formed by biological activities (e.g. root channels Ped (a naturally formed unit of soil aggregate such as a and worm borrows), physical processes (e.g. desiccation block, column, granule, plate, or prism) is a unique term cracking and freezing–thawing), or chemical reactions (e.g. in soil science. Pedology (a branch of soil science that dissolution or binding of soluble chemicals and organic integrates and quantifies the morphology, formation, distri- matter) are critical in determining flow and transport in field bution, and classification of soils as natural or anthropogeni- soils. Therefore, in hydropedology, the term soil structure cally modified landscape entities) captures that uniqueness is used to encompass both pedality and pore space (Lin well. The natural soil architecture is of essence in under- et al., 2005a). standing soil physical, chemical, and biological processes Since soil structure generally refers to a specific soil as well as rainfall–runoff processes. It has been suggested horizon, a broader term of soil architecture is suggested that most that can be learned from sieved and repacked soil here to also include the overall organization of a soil profile samples has already been done, so whatever we do with soil (e.g. horizonation), a soil’s relationship with the landscape physics/hydrology should be done on intact, undisturbed (e.g. catena), and the overall hierarchical levels of soil soils, and preferably in situ in the field. A new era of soils structural complexity (Lin et al., 2005a). research will have to rely on “structure-focused” – passing Soil horizonation or layering is ubiquitous in nature, so the stage of “texture-focused” – to achieve better ways of it must be adequately addressed when measuring, model- quantifying flow pathways, residence times, and spatiotem- ing, and interpreting hydrologic processes in watersheds. poral patterns of landscape water. Various kinds and thicknesses of soil horizons and how The term soil structure hasbeenusedintheUSSoil they organize in soil profiles reflect long-time pedogene- Surveys, and elsewhere, to refer to the natural organization sis and the past and current landscape processes. The fact of soil particles into individual units (peds) separated that natural soils are layered has at least two significant 6 RAINFALL–RUNOFF PROCESSES

(a) Granular Prismatic Blocky Platy

Infiltration intake area

Root channel Slickenside Interpedal pore Crack Living roots (b)

Cutans (argillans) Secondary ped Primary Clay coating peds

Tertiary Cutans ped Soil matrix

Secondary ped

Primary 0.8 mm (c) peds

Figure 3 (a) Conceptual understanding of the effects of various shapes of soil structure on water flow (Courtesy of the USDA-NRCS). (b) Illustrations of preferential water flow paths (blue- or red-dyed areas) in structured clay soils. (c) A tertiary ped and a zoom-in illustration of clay coating (Cutan) on the face of a ped implications for hydrology: (i) interface between soil layers While the importance of soil architecture across scales of contrasting textures and/or structures would slowdown has long been recognized in soil science and hydrology, its water downward movement, which often leads to some kind quantification and incorporation into models have notori- of preferential flow (e.g. fingering flow, macropore flow, or ously lagged behind. This problem is due to many factors, funnel flow) and (ii) soil layering or discontinuity in soil including those highlighted below: hydraulic properties between layers would promote lateral flow or perched water table (PWT), especially in sloping • Inconsistent and fragmented concepts of soil structure, landscapes with a water-restricting layer underneath. as described above and further elaborated by Letey A catena (also called toposequence) is a chain of related (1991), Bronick and Lal (2005), Baveye et al. (2006), soil profiles along a hillslope with about the same age, and many others. One critical point is illustrated here: similar parent material, and similar climatic condition, “No art historian would attempt to characterize the but differs primarily in relief that leads to differences in architecture of buildings erected in France in the high drainage and soil thickness. Catenary soil development Middle Ages by cataloging the nature of the stones often occurs in response to the way water runs down the used, the size of the stones, the types of materials hillslope and recognizes the interrelationship between soil used to bind the stones together, or the geometry of and geomorphic processes (Hall and Olson, 1991; Moore the portions of walls into which the buildings break et al., 1993; Thompson et al., 1997; Lin et al., 2005a, when they are knocked down. With these parameters b). Catenae are thus also often called hydrosequences of alone, it might be somewhat difficult to distinguish Notre related soils, especially in depositional landscapes. Another Dame’s cathedral in Paris from stone houses in the Alps, important aspect of soil architecture along the hillslope for example. Clearly, a different perspective is needed, is related to preferential flow network, which is further emphasizing the openings (window frames, doorways, discussed later in this article. rooms, etc.) that the stones allow to create in one case HYDROPEDOLOGY AND SURFACE/SUBSURFACE RUNOFF PROCESSES 7

or another. In the same manner, a focus on voids in soils and integrate physical, chemical, biological, and anthro- might be a more fruitful way to look at soil architecture” pogenic impacts. Although a hierarchical organization (Letey, 1991; Young et al., 2001; Baveye et al., 2006). of soil aggregates has been well recognized (Tisdall and • Overemphasis on ground-sieved soil materials and soil Oades, 1982; Vogel and Roth, 2003) and fractal char- texture in the past, thus ignoring or downplaying the acterization of soil structure has been proposed (Perrier importance of the soil’s “natural architecture.” It has et al., 1999; Bartoli et al., 1998), there is still a lack been said that a crushed or pulverized sample of the of means of representing field soil structure in differ- soil is related to the soil formed by nature in the same ent scales in a manner that can be coupled into models way as a pile of debris is related to a demolished of flow, transport, and rate processes (Lin, 2003; Lin building (Kubiena, 1938). Lin (2007) also suggested et al., 2005a). that a crushed sample of the soil is as akin to a natural Quantifying soil architecture in the field across scales soil profile as a bulk of ground beef is to a live and at desirable spatial and temporal resolutions has been cow. The fundamental difference between in situ soils technologically limited. While landforms (e.g. digital ele- and disturbed soil materials lies in soil architecture. In vation models or DEMs) and vegetation (e.g. land use/land fact, soil is a living entity, with many dynamic forces cover) can now be mapped with high resolution (e.g. acting upon it, so its internal architecture forms and using Light dection and ranging or LIDAR and IKONOS evolves to serve various soil functions. That natural earth observation satellite, respectively), there is a “bot- soils are structured to various degrees at different scales tleneck” phenomenon for in situ high-resolution (e.g. sub- is the rule; whereas the existence of a macroscopic meter to centimeter) and spatiotemporally continuous and homogeneity is the exception (Vogel and Roth, 2003). noninvasive mapping or imaging of subsurface architecture • Lack of appropriate techniques and devices to quantify including flow networks. This technological bottleneck has soil architecture directly, especially in situ noninva- constrained our predictive capacity of many soil and hydro- sively. Traditionally, soil structure has been evaluated logic functions. by pedologists in the field using morphological descrip- tions or thin section observations, while soil physicists Defining ‘‘Hydropedologic Functional Unit (HFU)’’ have employed wet and dry sieving, elutriation, and sed- and its Quantification in Different Landscapes imentation to conduct aggregate analysis. In the absence At the hillslope scale, in the absence of detailed field inves- of direct quantification, soil structure has been fre- tigations, much of soil information needed for hydrologic quently evaluated by methods that correlate it to the modeling is obtained from soil survey reports. However, properties or processes of interest (such as water reten- the minimum size of soil map unit delineation in a standard soil survey is ∼1.6 ha (Buol et al., 2003), which is often tion, saturated hydraulic conductivity or Ksat, infiltration rate, and gas diffusion rate). In recent years, noninva- too coarse for most hillslope-scale studies. Consequently, sive methods that permit soils to be investigated without many soil map units are associations and complexes, where different soil types are mapped together as a matter of undue disturbance of their natural architecture have necessity dictated by the map scale (Soil Survey Division become increasingly attractive, which allow 3-D visu- Staff, 1993). Furthermore, polygon-based mapping in tra- alization of internal soil structure and its interactions ditional soil surveys has resulted in soil variation being with water. These methods include X-ray computing portrayed spatially as a step function that appears only tomography, soft X-ray, nuclear magnetic resonance, when crossing from one polygon to another (Zhu et al., γ -ray tomography, ground penetrating radar, and oth- 2001). This does not allow the continuum of variability in ers (e.g. Anderson and Hopmans, 1994; Perret et al., soil attributes on a hillslope or within a catchment to be 1999). Image analysis has brought new opportunities represented as it should. Future development of digital soil for analyzing soil structure, especially that of the pores, maps with greater spatial accuracy and resolution is needed their sizes, shapes, connectivity, and tortuosity (e.g. to provide a better interface with other digital databases and Vogel et al., 2002; Vervoort and Cattle, 2003). How- to facilitate spatially distributed hydrologic modeling. ever, although numerous attempts have been made to Traditional 3-D block diagrams used in soil surveys find either statistical relations or deterministic links to show soil–landscape relationships can be enhanced between soil structural data and hydraulic properties, to develop conceptual models of the landscape–soil– a significant gap remains between in situ soil struc- hydrology relationships (Figure 4). Enhanced 3-D block ture/architecture and field-measured soil hydraulic prop- diagrams with added information of water table dynam- erties at different scales. ics, water flow paths, hydric soils, restrictive layers, and • Lack of a comprehensive theory of soil structure/ other relevant information could significantly increase the architecture formation, evolution, quantification, and values of soil survey products. As illustrated in Figure 4, modeling that can bridge orders of magnitude in scale classical block diagrams of soil–landscape relationships 8 RAINFALL–RUNOFF PROCESSES

Dodge silt loam Lamartine silt loam Dodge silt loam Lamartine Elevation (n) Stream silt loam 920 Muscatine Pella silt loam 1–3% LL 900 Intermittent pond Pella Fayette Floodplain Garwin 880 2–30% Tama 0–1% Carlisle 2–15% CS muck muck 860 Downs 2–15% 840 Dinsdale Glacial till Tama 2–15% Catena 2–15% Loess Wabash 0 1/2 Mile 0–1% Flow lines of percolating water CS Calcareous silt Position of water table in spring LL Loached loess covering over calcareous glacial till Alluvium Glacial till (b) Position of water table in summer

Ground water

Clarion discharge area Vertic Typic Mollic Typic Typic Aquic Typic Ground water cumulic (calcareous) hapludalf hapludalf argiudoll argiudoll hapludoll endoaquoll endoaquoll recharge area Nicollet fayette downs tama dinsdale muscatine garwin wabash (leached) Storden Ochric Ochric Mollic Mollic Mollic Mollic Mollic Albic Clarion Albic Canisteo Clarion Bt Bt Bt Hydric soil Webster argillic argillic Bt argillic argillic Harpst ents Bg Bg Leached glacial till Bg cambic cambic Okobaji cambic Glacial Okobaji C C C till Cg Cg

Leached glacial Nicollet Alfisols Mollisols CalcareousCalcareous glacial till glacialsediments till or sediments Leached glacial till or sedim (a) "Edge effect" soluble salt accumulations

(c)

Figure 4 (a) Classical block diagrams used in soil surveys to show soil mapping units and catenae in relation to the landscape (From Brady and Weil, 2004). Also illustrated are two contrasting landscape hydrology in relation to the soilscape: (b) A soil catena on a drumlin and adjacent lowlands in Dodge Co., WI, illustrating water flow direction and seasonal water table dynamics (From Hole, 1976); (c) Pattern of soil catena and parent materials in the Clarion–Nicollet–Canisteo association in Kossuth Co., IA, illustrating water and solutes in the uplands moving down into the Clarion soil profile where they either enter the groundwater or move laterally until they are discharged further downslope as return flow (Courtesy of L. Steffen, USDA-NRCS) could be used to indicate landscape hydrology, illustrating time of water in various soil–landscape units. These units water flow direction and water-table dynamics. These block can be identified and delineated using traditional soil survey diagrams could be further linked to watersheds or physio- methods and data in conjunction with various digital data graphic regions to provide valuable conceptual frameworks sources (e.g. DEM), geophysical investigations, and in situ of water movement over the landscape in different regions monitoring (Lin et al., 2008). Essentially, soil map units are (USDA-NRCS, 1997). better considered as landscape units rather than individual New ways of mapping soils in greater detail and soil types (Wysocki et al., 2000). It is also desirable to go with higher precision are desirable for precision research beyond the pure description of soil taxonomic classifica- and management. The concept of HFU, defined as a tions on a soil map (as traditional soil surveys did); rather, soil–landscape unit having similar pedologic and hydro- it would serve diverse applications much better by depict- logic functions, has been suggested as a means of ing soil functions, particularly those hydrologic functions cartographically representing important landscape–soil– critical to various local land uses. Hall and Olson (1991) hydrology functions (Lin et al., 2008). The goal of the have challenged traditional soil mapping: “Much effort has HFU is to subdivide the landscape into similarly functioning been expended on taxonomic classification of soils during hydropedologic units by grouping various geomorphic units the last few years but the importance of proper representa- that have similar storage, flux, pathway, and/or residence tion of landscape relations within and between soil mapping HYDROPEDOLOGY AND SURFACE/SUBSURFACE RUNOFF PROCESSES 9 units has been virtually ignored. The same mapping unit (2006), Beven (2002), and many others have noted, current is often delineated on convex, concave and linear slopes. models in watershed hydrology are based on well-known This mapping results in the inclusion of areas of moisture small-scale physics or theories such as the Darcy’s law accumulation, moisture depletion and uniform moisture flow and the Richards equation built into coupled mass balance within a given mapping unit.” equations. It has been observed that the dominant process Zhu et al. (2008) has proposed an approach toward governing unsaturated flow in soils may change from matrix precision soil mapping through integrating the existing flow to preferential flow under different conditions when second-order digital soil map with high-resolution terrain moving from the pore scale to the pedon scale (Bloschl¨ and indices and site-specific geophysical surveys. Their refined Sivapalan, 1995; Hendrickx and Flury, 2001). When mov- soil map obtained for a 19-ha agricultural landscape had ing from the pedon scale to the landscape/watershed scale, a considerably reduced variability within mapping units our knowledge for extrapolating the Darcy–Buckingham’s and an increased precision in soil boundaries. Such a law and the Richards equation to large heterogeneous area map provided a basis for specific soil property mapping is further constrained (Weiler and Naef, 2003). Further dis- through addition of soil coring, or monitoring, or other cussion on subsurface flow network is provided later in this means of soil sensing/mapping. On the basis of two article. soil properties important to hydrologic functions – surface soil texture and depth to clay layer, Lin et al. (2008) developed a map of HFUs for this agricultural landscape LINKING SURFACE/SUBSURFACE RUNOFF and tested this map for soil moisture dynamics and nitrogen PROCESSES TO PEDOLOGIC fertilizer management. Their results showed that the three UNDERSTANDING basic components making up an HFU (i.e. soil series In a new vision for watershed hydrology, McDonnell et al. or its variant, surface texture, and depth to clay layer) (2007) urged a paradigm shift from merely gauging the net displayed significant differences in soil moisture. They also hydrograph at the outlet of a catchment (or at the trench of showed that an optimal combination of hydrology (both a hillslope) and using small-scale theories to model water- soil moisture storage and drainage) and nutrient supply shed hydrology to a new one that emphasizes diagnosis (nitrogen fertilizer application rate) was required to obtain (i.e. diagnosis of soil–landscape patterns and watershed optimal growth of corn. functional traits), classification (i.e. watershed classifica- The HFU of different landscapes is a useful concept to tion based on similarity indices or dominant processes), be further developed. Such a mapping of HFUs can serve evolution (i.e. why watershed heterogeneity exists), and as a cartographic building block to increase the knowledge flow network (i.e. underlying optimality principles) across transfer and the bridging of “forms” and “functions” of soils scales. We believe pedology has significant potential to at different scales. This is consistent with the flexible box contribute to this new vision for watershed hydrology and models suggested by McDonnell (2003) and would facili- to help answer “why-type” questions, as mentioned in the tate the prediction of ungaged basins (Sivapalan, 2003). Introduction. In the following, we review various aspects There is a conceptual bottleneck that needs to be resolved of diagnostic soil–landscape features that are important for developing a new generation of hydrologic models. to understanding and modeling surface/subsurface runoff That is, should a continuous field or discrete objects be processes at the three hierarchical scales of microscopic, used to model surface and subsurface flow? Traditionally, mesoscopic, and macroscopic levels. hydrologic processes are generally conceptualized within the field domain (e.g. the Navier–Stokes equation and the Darcy’s law) (Goodchild et al., 2007). Classical hydrology Aggregate and Macropore Scale has applied findings from fluid mechanics, together with the necessary constitutive relations, to develop sets of gov- Pedality erning equations (much the same as atmospheric and ocean Through aggregation, pedality alters a soil’s pore-size dis- sciences have done). However, heterogeneities in land sur- tribution and pore continuity/connectivity. Notably, the face, hierarchical structures of soils, channel geometries, formation of peds creates interpedal pores that have size and preferential flow networks all make the land surface and shape different from intrapedal pores (Figure 3). Dye- and subsurface different from the continuous-field assump- tracing studies have clearly demonstrated that preferential tion (Consortium of Universities for the Advancement of flow often occurs along interpedal pores in structured soils Hydrologic Science, Inc. (CUAHSI), 2007). It is becoming (e.g. van Stiphout et al., 1987; Lin et al., 1996). Concep- more and more recognized that solid earth is not a con- tually, we would expect the following trend of increasing tinuous fluid; rather, it poses hierarchical heterogeneities capacity to transmit water vertically among different shapes with discrete flow networks embedded in both the surface of peds: platy < blocky, prismatic < granular (Figure 3a). and the subsurface. As McDonnell et al. (2007), Kirchner Higher Ksat has been reported for prismatic soil than its 10 RAINFALL–RUNOFF PROCESSES blocky counterpart in silty clay loam (Bouma and Ander- between peds, animal burrows, and channels formed by son, 1977a, b). Lin et al. (1999a) showed that, depending living and dead roots. From a hydrologic functional point on texture and initial moisture, soils with prismatic peds of view, these features often lead to preferential flow or had either higher or lower steady-state infiltration rates than bypass flow (Figure 3b). The definition of macropores, similar soils with blocky peds. For fine-textured soils (clay especially their size range, however, remains diverse. Some and clay loam), prismatic pedality had higher flow rates researchers have defined the size range of macropores on than blocky pedality because of common occurrence of the basis of flow function: that is, whether flow is subjected interprism macropores; whereas in medium-textured soils to capillary force or not. Others have used morphologi- (sandy clay loam and silt loam), interprism pores were cal approach to define macropore size. It appears that both less significant in soils with prismatic pedality, leading to functional and morphological approaches should be com- a lower water flow rate than in blocky soils (Lin et al., bined to define macropore so that meaningful interpretations 1999a). The hydraulic distinctions between blocky and pris- can be made in a comparable manner. As suggested by the matic ped shapes can therefore be addressed through the Soil Science Society of America (SSSA, 1997), the minimal associated difference in macroporosity and texture. size of macropores is 0.075 mm in diameter. Macropores Conceptual understanding of pedality’s impacts on soil are also commonly defined as the pores corresponding hydrology is intuitive, including flow pathway and flux to ≥−0.03-m water potential, that is, ≥0.5mm in radius (Figure 3). However, quantification of such impacts is for cylindrical pores or ≥ 0.5 mm in width for planar pores. less straightforward, in part because of the lack of a Most channels formed by the main roots of annual crops proper means for quantifying pedality. Pedality is usually and earthworms have a minimal radius of about 0.5 mm. expressed strongest at the surface and decreases with depth, However, soil pipes formed by tree roots or animal burrows resulting in a general trend of increasing ped size and are often much larger, with pore diameter often greater than decreasing ped grade with soil depth. We would also expect several centimeters. that the smaller the ped size, the higher the amount of Among various types of macropores, the general trend interpedal pores, and that the stronger the ped grade (i.e. the of increasing capacity to transmit water is (Lin et al., distinctness or strength of ped units), the more significant 1999a): vugh (isolated large void, usually irregular and not the interpedal pores. However, tillage and other human and normally interconnected with other voids of comparable biological activities could significantly alter this general size) < channel (tubular-shaped void) < fracture (planar trend. Changing soil moisture also changes the expression void between aggregates) < packing-void (void formed by of soil structure (especially in shrink–swell soils) and random packing of single skeletal grains or peds that do not the relative volumes of peds and pores, adding to the accommodate each other). Such a trend is supported by dye- complexity of finding mathematical solutions for modeling tracing studies of Bouma et al. (1977) and Lin et al. (1996); water movement in structured soils. both found the similar order as evidenced by dye-stained Spherical and blocklike pedality is most commonly seen areas in soils with these types of macropores. in surface soils, while subsoils generally display blocklike On the basis of steady infiltration rates measured on 96 or prismlike pedality. Platelike pedality is associated with soil horizons in Texas, Lin et al. (1999b) demonstrated that surface sealing, compacted soil layer (such as plowpan), soil structure was crucial in characterizing hydraulic prop- or sedimentation-inherited C horizon. Two structureless erties in macropore flow region, whereas texture had major cases – massive (commonly associated with heavy clayey impact on those hydraulic properties controlled by microp- soils) or single grains (associated with sands) – are also ores. Lin et al. (1997) showed that steady-state infiltration observed in soils. Often, a soil horizon’s pedality is a rates at zero tension (i0) were greatest for clay-textured soils mixture of several ped sizes (especially with compound with well-developed structure and macroporosity, while pedality), with one dominant ped shape and grade. How- clayey soils with poorly developed structure had low infil- ever, field qualitative descriptions of pedality often contain tration rates, leading to high variability in i0 for clayey soils. some degree of subjectiveness, leading to possible incon- At more negative supply water potential, where macropores sistency among individuals who describe it. were less active, infiltration rates were greatest for soils In soils with compound pedality, Lin et al. (1999a) found dominated by sand. that primary pedality had limited impact on infiltration at − water potential of > 0.03 m. Quisenberry et al. (1993) Soil Hydromorphology and Hydric Soil Field and Nortcliff et al. (1994) also reported that in a soil with Indicators both primary and secondary peds, preferential flow occurred mostly along secondary ped faces. Actions of water on soils often result in the formation of distinctive morphological features that are indicative of Macropores field water movement. Soil horizons, for instance, often Soil macropores encompass many types of structural open- develop in response to water movement (such as leach- ings in soils, such as desiccation cracks, pore space in ing or accumulation of certain materials). Even without the HYDROPEDOLOGY AND SURFACE/SUBSURFACE RUNOFF PROCESSES 11

Mnd/Fed---Bt 1992; Soil Survey Staff, 1999). Redox features, which are <0.06 usually considered hydric soil indicators, exclude those hydric soil indicators composed of C and S (USDA- 0.065–0.10 Elevation (m) NRCS, 1998). 10 >0.10 Hydric soils are defined as soils that formed under con- ditions of saturation, flooding, or ponding that lasted long enough during the growing season (repeated periods of more than a few days) to develop anaerobic conditions in

0 the upper part (usually 0.15–0.3 m) of soil profiles (USDA- NRCS, 1998). Hydric soils are one of the three requirements (along with hydrophytic vegetation and wetland hydrology) for identifying jurisdictional wetlands in the United States. 210 m 140 m Hydric soil field indicators include a variety of features that are regional and texture-based, but all are formed predom- Figure 5 Landscape distribution of dithionite-extractable inantly by the accumulation or loss of Fe, Mn, C, or S Mn/Fe ratio (Mnd/Fed) in the Bt soil horizon in a 3-ha compounds (USDA-NRCS, 1998). While indicators related field in the Peidmont region of North Carolina. The map to Fe/Mn concentrations or depletions are the most com- was generated by kriging 60 sampling points occupying mon, other features (e.g. sulfide and various combinations a variety of geomorphic positions. Dash lines indicate the of carbon accumulations) have been used in specific kinds center of topographic low areas of soils that do not develop redox features. There are also so-called problem soils that are seemly hydric soils but formation of visible morphological features, soil chemistry whose morphologies are difficult to interpret or seem incon- and biology may provide clues regarding field-scale water sistent with the current landscape, vegetation, or hydrology movement. For example, difference in mobility between (Veneman et al., 1998). These include soils formed in gray- redox-sensitive Mn and Fe in acid soil systems allows ish or reddish colored parent materials, soils with high pH secondary Mn/Fe ratios to be used as pedochemical indi- or low organic matter, Mollisols with thick dark A horizon, cators of field-scale throughflow (McDaniel et al., 1992) Vertisols with shrink–swell, soils with relict redox features, (Figure 5). A subset of soil morphological characteristics, and disturbed soils such as cultivated soils and filled areas known as hydric soil indicators (USDA-NRCS, 1998), is (USDA-NRCS, 1998; Rabenhorst et al., 1998). Relict redox directly related to a specific set of hydrologic conditions. features do not reflect contemporary or recent hydrologic Soil morphology is also the testimony of long-term per- conditions of saturation and anaerobiosis; rather, they were sistent flow and transport processes that result in visible likely formed during past geological wetter climates. Typ- pedological features such as clay films, tonguing, plinthites, ically, contemporary and recent hydric soil morphologies and diverse soil structures that are of hydrologic sig- have diffuse boundaries, while relict redox features have nificance (Figure 3). Soil macro- and micromorphology abrupt boundaries (USDA-NRCS, 1998). thus have long been used to infer soil moisture regimes, Certain redox patterns occur as a function of the pat- hydraulic properties, and landscape hydrologic processes terns in which the ion-carrying water moves through the (e.g. Bouma, 1992; Lilly and Lin, 2004; Lin et al., 2005a). Soil hydromorphology deals with soil morphologi- soil and as a function of the location of aerated zones in cal features (especially redoximorphic features or redox) the soil. Characteristic color patterns are thus created by caused by water and their relations with soil hydrology. the reduced Fe and Mn ions removed from a soil if ver- Redox features (formerly called mottles and low chroma tical or lateral water flow occurs, or the oxidized Fe and colors) are formed by the processes of alternating reduc- Mn precipitated in a soil if lack of sufficient water flux. tion–oxidation due to saturation–desaturation and the sub- Consequently, the spatial relationships of redox depletions sequent translocation or precipitation of Fe and Mn com- and redox concentrations may be used to interpret water pounds in the soil (Soil Survey Staff, 1999). Types of redox and air movement in soils (Vepraskas, 1992). Interpreting features include (i) redox concentrations as accumulations directions of water movement from redox patterns is easi- of Fe/Mn oxides (e.g. nodules, concretions, masses, and est when the features have a consistent relationship with pore linings), (ii) redox depletions as low-chroma (≤2) soil structure including macropores. For example, when features formed by removal of Fe oxides (including Fe redox depletions occur around macropores and redox con- depletions and clay depletions), (iii) reduced matrix that centrations occur within matrix, it would suggest that water changes color upon exposure to air due to Fe(II) oxidation infiltration has occurred along macropores and reducing to Fe(III), and (iv) a reaction to an α,α-dipyridyl solu- condition has developed there because of perched saturated tion if the soil has no visible redox features (Vepraskas, layers (Vepraskas, 1992). As another example, when redox 12 RAINFALL–RUNOFF PROCESSES concentrations occur around macropores and redox deple- 12–25% (Huffman et al., 2001). A hydrophobic soil tions occur within matrix, it would indicate that soil matrix layer is often developed after intense wild fires. Burn- is wet for periods long enough for reducing condition to ing organic matter in a hot fire releases hydrocarbons be maintained while macropores become aerated because into the atmosphere as a gas, which penetrates into of faster drainage or plant roots (such as in rice plant) that soil aggregates at the mineral soil surface, condenses, transport air to macropores when the soil is still flooded and covers the aggregates with waxy substance after it (Vepraskas, 1992). cools (DeBano et al., 1970). This waxy coating repels water and greatly reduces the infiltration of the soil. Infiltration-excess Runoff Often, erosion from steep forested hillslopes after a A number of processes can cause the reduction in the soil’s hot-burning fire can be excessive. The thickness and infiltration capacity, leading to infiltration-excess runoff. persistence of hydrophobic layers varies primarily with These include surface sealing, soil freezing, hydrophobicity, burn severity; however, vegetation type, soil texture, and others. Infiltration-excess runoff events, however, can and soil moisture are also important. Thick hydropho- be minimized by proper management of the soil surface bic layers (i.e. >0.03 m) often persist for more than a such as mulching and crop residue cover. The following year (Huffman et al., 2001). elaborates each of these aspects. • Additional complicating factors may also reduce a soil’s infiltration capacity, including soil swelling, soil insta- • A surface seal or crust is caused by the breakdown bility, entrapped air, and soil layering. For example, of soil aggregates through raindrop impact or through infiltration into layered soil is always slowed down at a slaking process under saturated conditions (Ellison the layer interface because of the hydraulic discontinu- and Slater, 1945; Tackett and Pearson, 1965). Fine ity (either different hydraulic conductivity or moisture soil particles are dislodged from soil aggregates and holding capacity at a given soil water potential). In the are deposited as a thin layer over the soil surface. case of a fine-textured layer overlying a coarse-textured Through this process, large soil pores which would have been available for infiltration are “sealed” with the fine layer, “fingers” may develop after infiltrating water pore soil particles because the slaked clay particles are gel- pressure overcomes the hydraulic discontinuity, form- like. Surface sealing and crusting occurs in virtually all ing fingering flow that may enhance the subsequent arable soils during ponded infiltration or flood irrigation. infiltration. Given the right condition, infiltration into The abrupt contact of the soil surface with excess water structured, or macroporous, or cracking soil takes place causes weak aggregates to disintegrate and slake. The in only a fraction of total soil surface, called fractional migrating smaller particles can quickly form a seal on wetting (i.e. the fraction of total surface area that is wet- the soil surface within a short period of time. ted), thus vitiating the classical assumption of uniform • The infiltration capacity of a soil can also be greatly advancement of the wetting front. Such fractional wet- reduced when the soil freezes (seasonally or perma- ting may either reduce or accelerate infiltration depend- nently). Concrete frost is often nearly impermeable and ing on the initial and boundary conditions. can be found most often in fine-textured soils (Zuzel et al., 1982; Dunne and Black, 1971). In agricultural Horizon and Profile Scale areas, erosion from a partially thawed soil surface can be excessive. However, the infiltration capacity through Soil Profile Architecture and Water-restricting Soil soils that are initially dry or have large macropores is Layers often minimally affected by soil freezing (Harris, 1972). Soil profile is a vertical section of the soil through all its The thick litter layer and large macropores found in horizons and extending into the C horizon, while soil solum typical forest soils minimize the potential effects of soil refers to only the portion of a soil profile above the C freezing on infiltration (Lutz and Chandler, 1946). Often horizon. All materials above fresh, unweathered bedrock are in mountainous areas the deep snow cover insulates the also referred to as regolith, which is generally equivalent ground and prevents soil freezing. Crawford and Legget to a soil profile. (1957) observed that the frost depth was reduced about In the US Soil Taxonomy, a total of 8 diagnostic surface 0.3 m for every 0.3 m of undisturbed snow cover. horizons (called epipedons) and 23 diagnostic subsurface • Extensive field studies have shown that many soils horizons (called endopedons) have been identified, along are susceptible to hydrophobicity under dry condi- with 30 additional diagnostic features for mineral soils tions (Dekker and Ritsema, 2003). Many forest soils (Soil Survey Staff, 2006). The presence or absence of these can become seasonally hydrophobic through leaching horizons plays the major role in determining which class a of the hydrocarbons in the litter layer and subse- soil falls in Soil Taxonomy. quent drying on the mineral soil; however, these soils Numerous water-restricting subsurface soil horizons and become hydrophilic when soil moisture levels exceed features have been identified in Soil Taxonomy, including HYDROPEDOLOGY AND SURFACE/SUBSURFACE RUNOFF PROCESSES 13 the following, with their major distinguishing characteris- in seasonal PWT and water moving laterally within the soil tics, highlighted (genetic symbols are also shown in paren- as subsurface throughflow (Busacca, 1989; Kemp et al., thesis for diagnostic horizons) (Soil Survey Staff, 2006): 1998). These include the following diagnostic subsurface soil horizons (with their genetic symbols indicated in • Duripan (Bqm): hardpan, strongly cemented by silica. parenthesis): • Fragipan (Bx): dense pan, seemingly cemented when dry, brittle when moist, and slake when submerged in • Agric (A or B): silt, clay, and humus accumulation water. Typically has redox features and bleached prism under cultivation just below plow layer. ped faces. • Argillic (Bt): silicate clay accumulation. • Glacic layer (Bf or Cf): massive ice or ground ice in • Glossic (E): tongue resulting from the degradation of the form of ice lenses or wedges. an argillic, kandic, or natic horizon. • Ortstein (Bhm): cemented by humus and aluminum. • Kandic (Bt): argillic, kaolinite clays (low activity). • Permafrost: a perennially frozen soil horizon that • Natric (Btn): argillic, high in sodium, columnar, or ◦ ≥ remains below 0 Cfor 2 years in succession. prismatic structure. • Petrocalcic (Ckm): carbonate cemented horizon. • Oxic (Bo): highly weathered, primarily mixture of Fe, • Petrogypsic (Cym): gypsum cemented horizon. Al oxides and nonsticky-type silicate clays. • Placic (Csm): thin pan cemented by iron (or iron and • Spodic (Bh, Bs): organic matter, Fe and Al oxides manganese) and organic matter. accumulation. • Anhydrous conditions: a soil layer of cold deserts and other areas with permafrost (often dry permafrost) and In addition, stratified or dense geological materials (C or low precipitation. R horizons) also often develop a hydrologically restrictive • Aquic conditions: continuous or periodic saturation and layer that leads to PWT and lateral water movement. reduction, with redox features. The importance of soil–bedrock interface and bedrock • Cryoturbation: frost churning on permafrost table, ori- topography to subsurface stormflow has been recently ented rock fragments, and silt caps on rock fragments. recognized (e.g. Onda et al., 2001; Freer et al., 2002; • Densic contact: a contact between soil and densic McGlynn et al., 2002). materials, without cracks or with ≥10 cm crack spacing • Densic materials: relatively unaltered but noncemented Fragipan as an Example: Its Extent, Formation, materials that roots cannot enter except in cracks and Impacts on Perched Water Table (PWT) and (including till, volcanic mudflows, mine spoils, or other Hillslope/Catchment Runoff mechanically compacted materials). • Fragic soil properties: essential properties of a fragipan, A query into the US Department of Agriculture but lacks layer thickness and volume requirement for (USDA) national soils database indicated that there are ∼ a fragipan. 16.5 million ha fragipan soils in the United States, mostly • Gelic materials: materials that show cryoturbation distributed in the east and southeast (Figure 6). Other and/or ice segregation in the active layer (seasonal thaw countries, such as New Zealand, Scotland, and Italy, have layer) and/or upper part of the permafrost. also reported fragipan soils (Soil Survey Staff, 1999). • Lamellae: thin layer of oriented silicate clay accumula- This very slowly permeable pan is one of the many tion. water-restricting soil layers that frequently cause PWTs • Lithic contact: boundary between soil and a coherent and control VSA hydrology. Soils having PWTs occupy ∼ (often indurated) underlying material. 82 million ha in the United States (McDaniel et al., 2008). • Lithologic discontinuities: stone lines and others. The processes that produce fragipans are imperfectly • Paralithic contact: contact between soil and paralithic known, variously attributed to physical ripening, the weight materials, without cracks or with ≥10-cm crack spacing. of glaciers, permafrost processes, and other events during • Paralithic materials: partially weathered bedrock or the Pleistocene (Smeck and Ciolkosz, 1989; Soil Survey weakly consolidated bedrock (such as sandstone, silt- Staff, 1999). Some of the properties of fragipans are stone, or shale). inherited from buried paleosols. Fragipans most commonly • Petroferric contact: boundary between soil and a thin occur in soils that formed under forest vegetation. The ironstone sheet. upper boundary of most fragipans has a narrow depth • Plinthite: dark red redox concentrations that changes range of about 0.5–1 m below the surface, if the soil is irreversibly to ironstone on exposure to repeated wetting not eroded. It is interesting to note that fragipans form and drying. only in soils in which water moves downward through the profile and they are commonly at depths that rarely Other subsoil horizons might also act as an aquitard or freeze (Soil Survey Staff, 1999). The polygonal network aquiclude to downward moving water, ultimately resulting of bleached materials in fragipans is formed by reduction 14 RAINFALL–RUNOFF PROCESSES

Geographic extent of soils where: classification similar to '*fragi*'

WA

ME MT ND OR MN ID SD WI MI RL WY

IA NE NV ID UT IL IN CA CO KS

AZ OK NM

GA

TX Acres per soil survey area (total = 40 683 703) FL Fragipan Acres not Less than 5487 to 24 209 to 61 541 to reported 5472 24 162 61 411 598484

Figure 6 A general distribution map of soils with fragipans in the United States, including all soil series that have ‘‘fragi’’ anywhere in their taxonomic classification string (such as Fragiaqualfs, Fragiudalfs, Fragixeralfs, Fragiboralfs, Fragiudults, Fragiaquults, Fragiaquods, Fragiorthods, Fragiaquepts, Fragiochrepts, and others). The map was generated using the Series Extent Mapping tool (available at http://www.cei.psu.edu/soiltool/semtool phase3.html), which uses the SSURGO database. The inset shows a soil profile from east-central Pennsylvania, where the fragipan starts at ∼0.5m depth and extends to over 1.2 m

of free iron after water has saturated the cracks (Figure 6 relatively high lateral Ksat in surface soil layers create a inset). very “flashy” hydrologic system (McDaniel et al., 2008). Dense, slowly permeable fragipans are important control- ling factors in VSA hydrologic processes in heterogeneous Saturation-excess Runoff watersheds (Gburek et al., 2006). Needelman et al. (2004) Strictly speaking, the term saturation means 100% pore showed that hillslopes with fragipan-containing soils pro- space filled with water. In reality, field soil saturation is duced considerably more runoff than did those without more a “satiation,” meaning <100% pore space filled with fragipans. Even within a relatively homogeneous catch- water (because of air bubbles etc.), but with free water ment comprising one fragipan-containing soil series, the present. Three types of soil saturation/satiation are defined variable depth to the fragipan also gives rise to VSA in soil survey (Figure 8) (Soil Survey Staff, 2006): hydrology, which in turn influences the spatial pattern of • Endosaturation: The soil is saturated with water in all runoff generation (Figure 7). Many years’ monitoring in the layers from the upper boundary of saturation to a depth Palouse Region of northern Idaho and eastern Washington of 2 m or more from the mineral soil surface. by McDaniel et al. (2008) showed that the spatial distri- • Episaturation: The soil is saturated with water in one or bution of saturation-excess runoff is largely controlled by more layers within 2 m of the mineral soil surface and the distribution of fragipan and its varying depth from the also has one or more unsaturated layers with an upper surface (Figure 7). Areas of the catchment where fragipans boundary above a depth of 2 m, below the saturated are relatively close to the surface tend to become saturated layer(s). This is commonly associated with PWT on more quickly and generate more saturation-excess runoff. top of a relatively impermeable layer. The gleying by Subsurface lateral flow on top of the fragipan accounts for episaturation is sometimes called pseudo-gley or false redistribution of up to 90% of the precipitation received gley, in contrast to true gley caused by endosaturation. on site during late winter and early spring in the study • Anthric saturation: This is a special kind of saturation by McDaniel et al. (2008). The combination of low drain- that occurs in cultivated and irrigated (flood irrigation) able porosity, relatively shallow depth to the fragipan, and soils. HYDROPEDOLOGY AND SURFACE/SUBSURFACE RUNOFF PROCESSES 15

100 WA 20 75

10 50 1.2 0 OR ID 1.1 0 25 1.0 20 0.9 40 0 0.8 60 P1 80 20 0.7 100 X 0 10 WCR2 0.6 120 X 40 20 WCR1 0.5 140 60 3/13/01 - 12:00 80 0.4 160 100 0 140 120 0 180 160 0.3 180 20 100 40 P2 60 75 80 20 Flume S 100 10 50 W 120 0 20 E 140 40 0 60 0 25 160 100 80 (a) N 180 120 20 160 140 180 40 0 60 80 100 60 120 Soil 0 surface 140 40 20 3/13/01 - 18:00 80 60 160 120 100 180 140 180 160 40 100

20 75

PWT height (cm) 20 10 50 Top of 0 fragipan 0 25 20 0 40 0 (b) DEC JAN FEB MAR APR MAY JUN 60 80 100 0 120 40 20 140 80 60 100 160 120 (c) 3/14/01 - 12:00 160 140 180180

Figure 7 A case study showing the fragipan’s control on the spatial distribution of soil profile saturation in the Palouse Region of northern Idaho and eastern Washington: (a) The study site location (inset) and a block diagram showing depth to fragipan in meters WCR1 and WCR2 are locations of soil water content probes; P1 and P2 are locations of pedon sampling pits; dashed white line shows location of hydraulic conductivity sampling transect; black dotted rectangle shows location of the hydrologically isolated of the hydrologically isolated hillslope plot (see MCdaniel et al., 2008, for more details). Black circles indicate location of monitoring wells. Block distances are in meters; (b) An example of perched water table dynamics recorded by pressure transducer above the fragipan; (c) Shading indicates the percentage of the soil profile above the fragipan that was saturated. Sequence represents a 24 h period coinciding with peak runoff generation

Depending on the duration of saturation and whether or and saturated moisture content with soil depth, (iv) soil not reduction occurs, two conditions are associated with horizonation and effective vertical versus lateral Ksat,and soil saturation (Soil Survey Staff, 2006): (v) preferential flow in soil profiles. Each of these factors is further discussed in the following: • Aquic conditions: continuous or periodic saturation and reduction occur, leading to redox features (gleying). • The depth to a hydrologically restrictive layer is one • Oxyaquic conditions: saturated but not reduced and thus of the most critical parameters in understanding VSA no redox features. This may be related to transient water hydrology. This depth, which could be either the table or flashy soil hydrologic conditions. distance from the soil surface to a soil horizon that limits A combination of topography, soil architecture below the vertical water movement or a bedrock, is often referred surface, and vegetation largely determines the location and to as the soil depth in VSA hydrology. Obviously, frequency of saturation-excess runoff. Key soil attributes for the same amount of water input, a shallow soil critical to understanding the processes leading to saturation- will reach saturation leading to runoff sooner and more excess runoff include (i) depth to a hydraulically restrictive frequently than a deeper soil. The soil depth not only soil or bedrock layer, (ii) change in Ksat and drainable determines the maximum water volume that a soil porosity with soil depth, (iii) change in bulk density profile can hold before runoff is generated, but also 16 RAINFALL–RUNOFF PROCESSES

Well-drained soil

Soils with poor natural drainage

Epiaquic Endoaquic

Perched water table Spring Impermeable layer

(e.g., clay lense) Stream

Regional water table

Figure 8 An illustration of the difference between episaturation and endosaturation: episaturation is from the top and endosatuation is from the side or bottom. This cross section of a landscape also shows the regional versus perched water tables in relation to three soils, one well drained and two with poor internal drainage. The soil containing the perched water table is wet in the upper part, but unsaturated below the impermeable layer, and therefore is said to be epiaquic, while the soil saturated by the regional water table is said to be endoaquic. This cross section also illustrates the importance of subsurface architecture for understanding hillslope hydrology (From Brady and Weil, 2004)

directly relates to the water volume that a soil profile database presented by Holtan et al. (1968), (Beven, will hold before a PWT is developed. At this point, 1984). When deeper soil profiles are considered, that large macropores become active leading to downslope is, including C horizons and soil layers beneath water- lateral flow or subsurface stormflow (Whipkey, 1965). restrictive layers (Figure 8), then such an exponential Once this subsurface stormflow is generated, water decay function is misleading. As illustrated by West follows the topography of this impeding layer. Recent et al. (2008) for soils in the Southern Piedmont of Geor- studies have provided clear evidence that water often gia, five of ten sites had a lower soil horizon’s Ksat follows bedrock topography rather than the soil surface being higher than the overlying horizons. When all sites topography as assumed in many hydrologic models lumped together, mean Ksat in C horizons was 2.5 times (Burns et al., 1998; Freer et al., 2002). Although soil higher than the overlying BC and Bt2 horizons (West depth and the topography of water-impeding layer are et al., 2008). Lin (1995) also documented that field Ksat clearly important factors for understanding hillslope and (approximated by steady-state infiltration rates at zero catchment hydrology, there are few tools available to tension) in many of the 18 soil profiles measured in reliably map these features over large areas. Texas did not follow exponential decline trend from the • Soil Ksat and drainable porosity are commonly assumed soil surface to ∼2 m depth. Drainable porosity in deep, to have exponential decline with soil depth in hydro- freely draining soils is often defined as the difference logic models (Beven and Kirkby, 1979; McDonnell, between saturated moisture content and field capacity. 2003; Weiler et al., 2005). This holds in many solums In soils dominated by saturation-excess runoff due to a from organic-rich top layers to more compacted restric- restrictive layer, however, free drainage does not occur tive subsoil layers (Kendall et al., 1999; Brooks et al., as it would in deep soils. During the wet season a PWT 2004, 2007). For example, Beven (1982) demonstrated forms, which on sloping ground drains primarily by lat- using experimental data taken from hillslope hydrol- eral flow or has very little drainage on relatively flat ogy experiments on 24 soils in predominantly forested ground. At equilibrium soil water above the water table landscapes that both Ksat and soil porosity decreased is held under pressures equivalent to the elevation above exponentially with depth below the soil surface. These the water table. Thus, the drainable porosity for shallow exponential decay functions, however, were not as con- soils is the difference between saturated moisture con- sistent when later examining 38 soils from the broader tent and the moisture content at the soil surface, which HYDROPEDOLOGY AND SURFACE/SUBSURFACE RUNOFF PROCESSES 17

is higher than field capacity. PWT fluctuations in soils (Vache´ and McDonnell, 2006). A hillslope-scale exper- having a small drainable porosity can be very dynamic, iment conducted by Brooks et al. (2004) showed that and surface runoff will be more “flashy” relative to soils the hillslope-scale Ksat was 15, 5, and 2 times greater having a larger drainable porosity (Dunne et al., 1975). than the small-scale measurements in the A, Bw, and E • Saturated soil moisture (which can be derived from soil horizons, respectively. bulk density) has a similar effect on the generation of surface runoff as soil depth, because it determines Preferential Flow in Soil Profiles soil water storage capacity. Soils having low saturated Preferential flow is the process whereby water and dis- moisture content can store less overall water, causing solved chemicals move by preferred pathways at an accel- more frequent and greater amounts of saturation-excess erated speed through a fraction of a porous medium. runoff. Like Ksat, saturated moisture content typically Vervoort et al. (1999) suggested that preferential flow in decreases with solum depth from organic-rich top layers soil profiles is related to soil structural differences (e.g. to compacted restrictive layers, resulting in less overall macropore flow and fractional flow) or textural differ- water storage at greater soil depths (Beven, 1982, 1984; ences (e.g. fingering flow and funnel flow). Different Brooks et al., 2007). The bulk density of the soil is also pedological features are indicative of possible preferen- the most widely used parameter to estimate the amount tial flow in various field soils, especially if combined of open pore space. Pedotransfer functions have used with landscape observations. These include: (i) soil struc- bulk density along with soil texture to incorporate vari- tural features (such as pedality, coatings, macropores, ability in soil structure into effective Ksat measurements. and slickensides); (ii) redox features (indicating water- However, these parameters by themselves are often restricting layer and nonuniform water movement); (iii) more appropriate for explaining micropore flow rates water-restrictive soil layers (such as sloping lamellae indi- rather than macropore flow as macropores are often not cating the likely occurrence of funnel flow); and (iv) well represented by small soil cores used to determine lithologic discontinuities (indicating significant changes bulk density (Beven, 1984; Lin et al., 1999b). in particle-size distribution and thus possible fingering • Soil horizonation impacts the overall effective vertical flow or other types of preferential flow). A simple field technique was devised by Bouma (1997) to measure pref- versus lateral Ksat inasoilprofile.Thedepthdistri- erential flow in clay soils as a function of rain inten- bution of vertical Ksat is important in defining where in the soil profile will begin to saturate (i.e. defining sity/quantity and initial soil moisture content. Deriving the depth to a restrictive layer) and the relative magni- cracking patterns from theoretical soil swell–shrink char- acteristics turned out to be difficult, and very small pores tude of lateral K will determine runoff patterns in sat (such as slickenside fissures shown in Figure 3) with a catchment dominated by VSA hydrology. Surface a volume that cannot be measured with existing phys- runoff will be confined primarily to toe slopes or at ical methods can conduct large volumes of water (Lin sharp breaks in the slope from steep to flat in soils hav- et al., 1996). ing large lateral K . Surface runoff patterns will be sat Fluhler¨ et al. (1996) have suggested three regimes of more widespread and correlated with variability in soil flow and transport within a soil profile during a preferential depth in areas with low lateral Ksat (Dunne et al., 1975). flow process: (i) lateral distribution flow in the attractor The challenge for hydrologists trying to predict subsur- zone where preferential flow is initiated; (ii) downward face lateral flow or stormflow at the hillslope scale is to preferential flow in the transmission zone where water determine representative lateral Ksat. Using small soil moves along preferential flow pathways and bypasses a cores it is relatively easy to measure Ksat in the labora- considerable portion of the soil matrix; and (iii) lateral tory but these measurements may not represent the Ksat and downward dispersive flow in the dispersion zone where at larger scales. Without a method to physically charac- preferential flow pathways are interrupted and water flow terize soil architecture at the hillslope scale, especially becomes more or less uniform again. This model appears large macropores and their network, Ksat becomes a cal- to capture the general characteristics of vertical preferen- ibration parameter in most hydrologic models (Grayson tial flow in many soil profiles, and is also consistent with et al., 1992; Wigmosta et al., 1994). Most likely, the the depth function of soil hydraulic conductivity discussed calibrated Ksat values will not be able to adequately earlier. represent the flow of water through a hillslope without On the basis of a review of studies documenting accounting for differences in soil structure in soil pro- soil microbial biomass distributions with soil depth (e.g. files and how macropores are distributed within each Van Gestel et al., 1992; Dodds et al., 1996; Richter and soil layer. Tracer and isotope studies often reveal that Markewitz, 2001; Blume et al., 2002; Taylor et al., 2002; simplified representations of how water flows through a Fierer et al., 2003), the system of Fluhler¨ et al. (1996) can hillslope are inadequate to represent the true flow path be applied. Previous studies have demonstrated a consistent 18 RAINFALL–RUNOFF PROCESSES pattern of changing microbial biomass size and activity dis- development can then feedback to the variability in water tribution, characterized by three broad zones with distinct distribution and transport over the landscape. biogeochemical potentials that correspond to the three flow With the ever-increasing numeric processing power avail- zones of Fluhler¨ et al. (1996): (i) attractor zone, equiva- able today and the availability of spatially detailed eleva- lent to the A horizon and representing high-biomass surface tion maps, hydrologists have been able to better represent soils under the strong influence of plant roots, moisture, and landscape processes using various means from simple topo- temperature; (ii) transmission zone, equivalent to B and C graphic indices to detailed simulation models. Geographic horizons and representing lower more spatially heteroge- information system (GIS) models can use detailed elevation neous biomass than in the attractor zone (Konopka and maps to describe topographic influences on water fluxes Turco, 1991; Rodr´ıguez-Cruz et al., 2006); and (iii) dis- throughout a landscape. For example, Moore et al. (1991) persion zone, representing higher-biomass, water-capture summarized a wide range of topographic attributes that can zones at the soil–bedrock interface or right above a water- be used to map various hydrologic, geomorphic, and bio- restricted soil layer because of moisture and more nutrient- logical processes. The most well known static topographic rich conditions resulting from impeded or lateral flow (Buss index for representing the influence of topography on the et al., 2006). These three zones are distinguished from spatial distribution of saturation-excess runoff that utilizes “deep-subsurface” aquifers and geologic materials, many the upslope catchment area and local slope was incorporated of which have extremely low biomass (Kieft et al., 1995). into TOPMODEL (Beven and Kirkby, 1979). This simple wetness index utilizing only topography has been success- Catena and Catchment Scale ful in describing areas generating saturation-excess runoff patterns during wet seasons. However, growing evidence Topography: Linking Hydrology with Pedology has pointed out the critical importance of the topography of The topography of a landscape is responsible for providing underlying bedrock or water-restricting soil layer in defin- the overall direction of water flow and hydraulic gradient ing water flow paths and hence soil wetness distribution for downslope subsurface stormflow (Dunne et al., 1975; (e.g. Onda et al., 2001; Freer et al., 2002; McGlynn et al., see also Chapter 112, Subsurface Stormflow, Volume 3). 2002; McDonnell, 2003). In VSA hydrology, the spatial variability of soil moisture before an event and of subsurface stormflow during the Catena event dictates where saturation-excess runoff will occur on Spatial distribution of topographic attributes that charac- the landscape. Topographically convergent toe slopes often terize water flow also capture spatial variability of soil are the wettest locations leading to saturation-excess runoff. attributes at the hillslope scale. Because of the processes In addition to largely controlling the distribution and that occur along a hillslope, soils may be quite different magnitude of water flow through a landscape, topography in different portions of a landscape, but these processes is also considered one of the five major factors controlling and relationships may be similar across a larger area, par- soil formation. In the classic work by Russian geologist ticularly if geomorphology and stratigraphy are similar. V.V. Dokuchaev and later extended by Jenny (1941), the Numerous investigations of catenae have been completed five major soil-forming factors include climate, organisms, in order to address this phenomenon (e.g. Veneman and topography, parent material, and time. Topographic patterns Bodine, 1982; Evans and Franzmeier, 1986; Pennock and dictate the way in which mass and energy are distributed de Jong, 1990; Stolt et al., 1993; Khan and Fenton, 1994; within a landscape. The orientation of a hillslope also dic- Thompson et al., 1998). It is now believed that the greatest tates the amount of solar radiation received, and therefore variability in certain soil properties within a physiographic largely controls the rates of evaporation, transpiration, and region may occur along a hillslope rather than from one snow ablation. The topography controls wind fields leading side of the region to the other (e.g. Pennock and de Jong, to variability in evaporation, snow drift, and deposition. 1990; Lin et al., 2005b). This warrants a careful investi- The topography also greatly influences rates of soil ero- gation and understanding of hillslope soil variability and sion and deposition. The combination of these effects leads catenae before making broad generalizations about soil vari- to spatial patterns of vegetation diversity, biomass produc- ability within a physiographic region. tion, and soil variability (see Chapter 12, Co-evolution of Water table depth (drainage condition) and fluxes of Climate, Soil and Vegetation, Volume 1). water, solutes, and sediments typically differ in soils There is a close feedback between hydrology and soil along a catena (Schaetzl and Anderson, 2005). However, development within a landscape that is driving by differ- catenae in different climatic and physiographic regions may ences in topography. Topography drives the distribution exhibit markedly different relationships between soil and and transport of water and sediment, which in turn influ- hydrologic properties. In many low-relief landscapes of ences soil development (e.g. the development of restrictive humid regions, proximity of a water table to the soil surface layers through elluviation/illuviation processes). This soil increases with distance away from stream or drainage way. HYDROPEDOLOGY AND SURFACE/SUBSURFACE RUNOFF PROCESSES 19

An example of this can be seen on the broad, flat, low- been accumulated in the past decades (e.g. Kirby, 1978; relief Atlantic coastal plain in the southeastern United Beven and Germann, 1982; Gish and Shirmohammadi, States. The most poorly drained soils are found toward the 1991; Germann and Hensel, 2006). However, because of centers of the broad interstream divides, while the most the complex and dynamic nature of preferential flow in well-drained soils are restricted to the edges of the flats different geographic regions and at different space- and and slopes closest to the streams and estuaries (Daniels timescales, our ability to predict landscape variability in et al., 1971; Daniels et al., 1984). The opposite trend is water flow and solute transport has been limited (Jury, seen in the higher-relief landscapes associated with the 1999; Lin et al., 2005a, 2006a). This, in part, is due nearby Piedmont region. Water table proximity to the soil to the fragmented knowledge of quantitative relationships surface decreases with increasing distance away from the between preferential flow and soil structure, soil layering, streams or drainage ways. As a result, the well-drained landscape features, precipitation inputs, and antecedent soils occupy the uplands and the more poorly drained soils moisture conditions at different scales. It is interesting to occupy the lower slope positions near the streams (Daniels note that most preferential flow studies by the soil science et al., 1984). Similarly, fluxes associated with overland community has focused on the pore to pedon scales and flow, subsurface lateral flow, percolation, capillary rise, vertical flow, while the hydrology community seems to and return flow will also vary along a catena in different have focused more on the hillslope to catchment scales and climatic and physiographic regions (Schaetzl and Anderson, lateral flow. A more concerted and integrated effort will 2005; Sommer and Schlichting, 1997). These fluxes will advance our predictive capacity in truly 3-D flow system. vary spatially and temporally, resulting in VSA hydrologic There is an emerging recognition of a paradigm shift processes. from a “continuum”-based approach to a “network”-based Two opposite trends of soil depth and soil wetness view of hillslope and watershed hydrology (Rinaldo et al., distribution along steeply sloped catenae are illustrated 2006; Uhlenbrook, 2006; Sidle et al., 2001; McDonnell here. In the humid forested catchment developed from shale et al., 2007). A “3F” (form, function, and feedback) is on the ridge of central Pennsylvania (within the Ridge and needed to characterize various watershed networks, which Valley physiographic region), soil thickness and wetness generally increase in concave hillslopes from hilltop to would lead to a better understanding, modeling, and predic- valley floor, while that of the convex and planar hillslopes tion of hillslope hydrology and ungauged basins. It remains within the same catchment shows no clear difference from to be further tested, though, whether an internal network the hilltop to the hill bottom (Lin et al., 2006b). In contrast, structure exists in the subsurface of many hillslopes, which in the rugged Hill Country of central Texas with stair–step appears to govern vertical and lateral preferential flow topography developed from limestone (part of the Edwards dynamics and the threshold-like hydrologic response under Plateau physiographic region), soil thickness and infiltration varying precipitation, soil, and antecedent moisture condi- capacity decrease from the hilltop (upper riser with steeper tions (e.g. Weiler et al., 2005; Tromp-van Meerveld and slopes) to the hill bottom (tread with flatter slope) (Wilcox McDonnell, 2006; Lin and Zhou, 2008). These issues are et al., 2007). Wilcox et al. (2007) also found that the upper important because without a clear understanding of the ini- riser subsoils were saturated or very wet for extended tiation, connectivity, and persistence of preferential flow periods, and surface runoff was generated from lower part networks across space and time, we are unable to effec- of the hillslopes (lower risers and treads) during periods tively predict rainfall–runoff for the right reasons. of high rainfall. Therefore, runoff and erosion increased Analogous to stream branching networks, a hillslope/ from upper riser to tread. Despite the opposite trends, both landscape network is a set of linkages among differ- catenae described above demonstrate the controls of soil, ent reservoirs and pathways with varying conductance topography, and geology on runoff generation (Lin et al., of flow. During storms with wet soil conditions, such a 2006b; Wilcox et al., 2007). network often provides preferred pathways for water to flow downgradient with high velocities (Nimmo, 2007). Subsurface Preferential Flow Networks Threshold phenomena may occur at critical “nodes,” or in Hillslopes and Catchments junctions of a network (i.e. hot spots), where significant Uhlenbrook (2006) claimed that catchment hydrology is a changes may occur within a short period of time (i.e. science in which all processes are preferential. It has been hot moments). Threshold behavior of subsurface storm- well recognized that preferential flow can occur in just flow has been observed in many hillslopes and landscapes about any soil, leading to spatial concentrations of water (Whipkey, 1965; Mosley, 1979; Peters et al., 1995; Weiler and chemicals that are not well described by the classical et al., 2005; Tromp-van Meerveld and McDonnell, 2006; approach based on uniform flow assumption (Vervoort Lin and Zhou, 2008). Thus, the concept of subsurface et al., 1999; Hendrickx and Flury, 2001). A large body of preferential flow networks could provide a strong scien- research on preferential flow and on hillslope hydrology has tific advance in our understanding of seemingly complex 20 RAINFALL–RUNOFF PROCESSES

Highly permeable relative to the mineral soil and very high density of roots Formation of a perched water table temporary between organic- Hetrogeneity caused rich and mineral soil horizons by preferential flow pathways

Interaction of macropore with Impeding layer but the surrounding or adjacent soil some water flow into matrix resulting in expanded cracks and fractures preferential flow path Water movement through humus and loose soil zones Formation of a Water emerging perched water table from fractured above bedrock bedrock into soil Some macropores Preferential flow connecting to the bedrock through living and fractures decayed roots Preferential flow Macropore flow Bedrock Uneven bedrock Vertical direction topography Mineral soil Matrix flow Downslope direction Wet Resultant direction Organic-rich horizon Fractured flow Wet

Figure 9 Interconnected preferential flow pathways along a hillslope in a forested catchment in Japan, illustrating the importance of subsurface preferential network (From Noguchi et al., 1999) hydrologic (and biogeochemical) phenomena across land- modeling and prediction (Grayson et al., 1992; Kirchner, scapes. Sidle et al. (2001); Gish et al. (2005); Lin (2006), 2003; McDonnell et al., 2007) calls for contributions from and many others have reported evidence of preferential hydropedology and other related disciplines to enhance the flow self-organization in forested hillslopes and agricultural understanding of interactive pedologic and hydrologic pro- landscapes, where individual short preferential flow path- cesses and the complex landscape–soil–hydrology relation- ways are linked via a series of “nodes” in a network, which ships across space and time. This article has illustrated that may be switched on or off, or expand or shrink depending hydropedology can make unique contributions to advancing on local soil moisture conditions and landscape locations hillslope and watershed hydrology in many ways, among (Figure 9). Sidle et al. (2001) suggested different levels of which are the following key areas that require further nodes in a network to approximate the preferential move- research: ment of water and solutes over the landscape, including • Better understanding and quantification of soil archi- (i) those that are easily activated under various conditions tecture from the pore to the catchment scales, and its (including critical landscape locations such as topographic relationships to preferential flow dynamics across space depressions, contrasting soil layers and soil–bedrock inter- and time; faces, and direct physical linkages between mesopores and • Defining and mapping hydropedologic functional unit macropores), (ii) those that require very wet conditions to (HFU) in different landscapes and its incorporation into activate (such as disconnected large macropores, buried spatially distributed hydrologic modeling; pockets of organic matter, and discontinuous animal bur- • Breakthroughs in two bottlenecks – a new concep- rows and soil pipes), and (iii) those that do not promote tual/theoretical framework for modeling preferential preferential flow (such as soil matrix and dead-end macro- flow networks (which is different from the continuous- pores with no connecting nodes). This conceptual model field assumption) and a technological advance- could lead to a new way of modeling preferential flow net- ment in nondestructively mapping/imaging subsurface works across multiple scales. architecture; • Pattern recognition and prediction of the “forms” and “functions” of soils and watersheds across scales, SUMMARY AND FUTURE OUTLOOK which could offer comprehensive insights regarding heterogeneity and the underlying processes leading to It is clear that synergies can be generated by bridging classi- heterogeneity; cal pedology with soil physics, hydrology, and geomorphol- • Use of various tracers, including isotopes and microbes, ogy to advance the mechanistic understanding and process- in combination with advanced sensor networks, long- based modeling of rainfall–runoff processes. The challenge term environmental monitoring, and a new generation of getting the right answer for the right reason in hydrologic of computer models, to gain insights into optimality HYDROPEDOLOGY AND SURFACE/SUBSURFACE RUNOFF PROCESSES 21

principles underlying the watershed functional traits Blume E., Bischoff M., Reichert J., Moorman T., Konopka A. (McDonnell et al., 2007; also see Chapter 12, Co- and Turco R. (2002) Surface and subsurface microbial biomass, evolution of Climate, Soil and Vegetation, Volume community structure and metabolic activity as a function of soil 1, Chapter 13, Pattern, Process and Function: Ele- depth and season. Applied Soil Ecology, 592, 1–11. ments of a Unified Theory of Hydrology at the Bouma J. (1992) Effects of soil structure, tillage, and aggregation Catchment Scale, Volume 1). upon soil hydraulic properties. In Interacting Processes in Soil Science, Wagenet R.J., Bavege P. and Stewart B.A. (Eds.), Lewis Publishers: pp. 1–36. Acknowledgments Bouma J. (1997) Long-term characterization: monitoring and modelling. In Methods for the Assessment of Soil Degradation, The fragipan map of the United States (Figure 6) was Lal R., Blum W.H., Valentin C. and Stewart B.A. (Eds.), generated with help from Sharon Waltman of the USDA- Advances in Soil Science, pp. 337–358. NRCS and Steven Crawford of Penn State University. The Bouma J. and Anderson J.L. (1977a) Water movement through paragraph on the summary of soil microbial distributions pedal soils. I. Saturated flow. Soil Science Society of America with soil depth and its possible relationships with the three Journal, 41, 413–418. zones of preferential flow was assisted by Mary Ann Bruns Bouma J. and Anderson J.L. (1977b) Water movement through of Penn State. pedal soils. II. Unsaturated flow. Soil Science Society of America Journal, 41, 419–423. Bouma J., Jongerius A., Boersma O., Jager A. and REFERENCES Schoonderbeek D. (1977) The function of different types of macropores during saturated flow through four swelling soil horizons. Soil Science Society of America Journal, 41, 945–950. Anderson S.H. and Hopmans J.W. (Eds.) (1994) Tomography Boyer E.W., Alexander R.B., Parton W.J., Li C., Butterbach- of Soil-Water-Root Processes. SSSA Special Pub. #36, Soil Bahl K., Donner S.D., Skaggs R.W. and Del Grosso S.J. (2006) Science Society of America, Inc.: Madison, WI. Modeling denitrification in terrestrial and aquatic ecosystems at Band L.E., Tague C.L., Groffman P. and Belt K. (2001) Forest regional scales. Ecological Applications, 16(6), 2123–2142. ecosystem processes at the watershed scale: hydrological and Brady N.C. and Weil R.R. (2004) The Nature and Properties of Hydrological Processes ecological controls of nitrogen export. , Soils, Fourth Edition, Prentice Hall: Upper Saddle River, NJ. 15, 2013–2028. Brewer R. (1976) Fabric and Mineral Analysis of Soils, Robert E., Bartoli F., Dutartre Ph, Gomendy V., Niquet S., Dubuit M. and Krieger Publishing Company: Huntington, NY. Vivier H. (1998) Fractals and soils structure. In Fractals in Bronick C.J. and Lal R. (2005) Soil structure and management: a Soil Science. Advances in Soil Science, Baveye P., Parlange J.- review. Geoderma, 124, 3–22. Y. and Stewart B.A. (Eds.), CRC Press: Boca Raton, FL, pp. Brooks E.S., Boll J. and McDaniel P.A. (2004) A hillslope- 203–232. scale experiment to measure lateral saturated hydraulic Baveye P. and Boast C.W. (1999) Physical scales and spatial conductivity. Water Resources Research, 40, W04208, predictability of transport processes in the environment. In DOI:10.1029/2003WR002858. Corwin D.L., Loague K. and Ellsworth T.R. (Eds.), Assessment of Non-Point Source Pollution in the Vadose Zone, Geophysical Brooks E.S., Boll J. and McDaniel P.A. (2007) Distributed and Monograph 108, American Geophysical Union: Washington, integrated response of a GIS-based hydrologic model in the DC, pp. 261–280. eastern Palouse region, Idaho. Hydrological Processes, 21, Baveye P., Bronick C.J. and Lal R. (2006) Comment on 110–122. “soil structure and management: a review”. Geoderma, 134, Buol S.W., Southard R.J., Graham R.C. and McDaniel P.A. (2003) 231–232. Soil Genesis and Classification, Fifth Edition, Blackwell/Iowa Beven K.J. (1982) On subsurface stormflow: an analysis of State Press: Ames, IA. response times. Hydrological Sciences Journal, 27, 505–521. Burns D., Hooper R., Kendall C., Freer J., McDonnell J. and Beven K.J. (1984) Infiltration into a class of vertically non- Beven K. (1998) Base cation concentrations in subsurface flow uniform soils. Hydrological Sciences Journal, 29, 425–434. from a forested hillslope: the role of flushing frequency. Water Beven K. (2002) Towards an alternative blueprint for a Resources Research, 34(12), 3535–3544. physically based digitally simulated hydrologic response Busacca A.J. (1989) Long quaternary record in eastern modelling system. Hydrological Processes, 16, 189–206, Washington, USA, interpreted from multiple buried paleosols DOI:10.1002/hyp.343. in loess. Geoderma, 45, 105–122. Beven K. and Germann P. (1982) Macropores and water flow in Buss H.L., Bruns M.A., Schultz M.J., Moore J., Mathur C.F. and soils. Water Resources Research, 18, 1311–1325. Brantley S.L. (2006) The coupling of biological iron cycling Beven K.J. and Kirkby M.J. (1979) A physically based variable and mineral weathering during saprolite formation, Luquillo contributing area model of basin hydrology. Hydrological Mountains, Puerto Rico. Geobiology, 3, 247–260. Sciences Bulletin, 24, 43–69. Consortium of Universities for the Advancement of Hydrologic Bloschl¨ G. and Sivapalan M. (1995) Scale issues in hydrological Science, Inc. (CUAHSI) (2007) Hydrology of a Dynamic modeling – a review. Hydrological Processes, 9(3–4), Earth – A decadal research plan for hydrologic science, 251–290. Washington, DC. 22 RAINFALL–RUNOFF PROCESSES

Crawford C.B. and Legget R.F. (1957) Ground temperature Grayson R.B., Moore I.D. and McMahon T.A. (1992) Physically investigation in Canada. Engineering Journal, 40, 1–8. based hydrologic modeling. 1. A terrain-based model for Daniels R.B., Gamble E.E. and Nelson L.A. (1971) Relations investigative purposes. Water Resources Research, 28(10), between soil morphology and water-table levels on a dissected 2639–2658. North Carolina coastal plain surface. Soil Science Society of Hall G.F. and Olson C.G. (1991) Predicting variability of soils America Proceedings, 35, 781–784. from landscape models. In Spatial Variabilities of Soils and Daniels R.B., Kleiss H.J., Buol S.W., Byrd H.J. and Phillips Landforms, Mausbach M.J. and Wilding L.P. (Eds.), Soil J.A. (1984) Soil systems in North Carolina, North Carolina Science Society of America: Madison, WI, pp. 9–24. Special Agricultural Research Service: Raleigh, North Carolina. Publication 28. Bulletin 467. Harris A.R. (1972) Infiltration rate as affected by soil freezing DeBano L.F., Mann L.D. and Hamilton D.A. (1970) Translocation under three cover types. Soil Science Society of America of hydrophobic substances into soil by burning organic litter. Proceedings, 36, 489–492. Soil Science Society of America Proceedings, 34, 130–133. Hendrickx J.M.H. and Flury M. (2001) Uniform and preferential Dekker L.W. and Ritsema C.J. (2003) Wetting patterns in water flow mechanisms in the vadose zone. In Conceptual Models repellent Dutch soils. In Soil Water Repellency, Ritsema C.J. of Flow and Transport in the Fractured Vadose Zone, National and Dekker L.W. (Eds.), Elsevier: Amsterdam, pp. 151–166. Research Council, National Academy Press: Washington, DC, Dodds W.K., Banks M.K., Clenan C.S., Rice C.W., Sotomayor D., pp. 149–187. Strauss E.A. and Yu W. (1996) Biological properties of soil and Hewlett J.D. and Hibbert A.R. (1967) Factors affecting the subsurface sediments under abandoned pasture and cropland. response of small watersheds to precipitation in humid regions. Soil Biology and Biochemistry, 28, 837–846. In Forest Hydrology, Sopper W.E. and Lull H.W. (Eds.), Dunne T. and Black R.D. (1971) Runoff processes during Pergamon Press: Oxford, pp. 275–290. snowmelt. Water Resources Research, 7, 1160–1172. Hole F.D. (1976) Soils of Wisconsin. Wisconsin Geological and Dunne T., Moore T.R. and Taylor C.H. (1975) Recognition Natural History Survey Bulletin 87, Soils Series 62. University and prediction of runoff-producing zones in humid regions. of Wisconsin Press: Madison, WI. Hydrological Sciences Bulletin, 20(3), 305–327. Holtan H.N., England C.B., Lawless G.P. and Schumaker Ellison W.D. and Slater C.S. (1945) Factors that affect surface G.A. (1968) Moisture-Tension Data for Selected Soils on sealing and infiltration of exposed soil surface. Agricultural Experimental Watersheds. USDA, ARS, Publication RS 41–144. Engineering, 26, 156–157. Horton R.E. (1940) An approach toward a physical interpretation Evans C.V. and Franzmeier D.P. (1986) Saturation, aeration, of infiltration capacity. Soil Science Society of America Journal, and color patterns in a toposequence of soils in North-central 5, 399–417. Indiana. Soil Science Society of America Journal, 50, 975–980. Huffman E.L., MacDonald L.H. and Stednick J.D. (2001) Fierer N., Schimel J.P. and Holden P.A. (2003) Variations in Strength and persistence of fire-induced soil hydrophobicity microbial community composition through two soil depth under ponderosa and lodgepole pin, Colorado Front Range. profiles. Soil Biology and Biochemistry, 35, 167–176. Hydrological Processes, 15, 2877–2892. Fluhler¨ H., Durner W. and Flury M. (1996) Lateral solute mixing Hurt G.W., Whited P.M. and Pringle R.F. (Eds.), USDA-NRCS processes – a key for understanding field-scale transport of (1998) Field Indicators of Hydric Soils in the United States. Ver. water and solutes. Geoderma, 70, 165–183. 4.0, USDA-NRCS, Ft. Worth, TX. Freer J., McDonnell J.J., Beven K.J., Peters N.E., Burns D.A., James A.L. and Roulet N.T. (2007) Investigating hydrologic Hooper R.P., Aulenbach B. and Kendall C. (2002) The role of connectivity and its association with threshold change in bedrock topography on subsurface storm flow. Water Resources runoff response in a temperate forested watershed. Hydrological Research, 38(12), 1269, DOI:10.1029/2001WR000872. Processes, 21, 3391–3408. Gburek W.J., Needelman B.A. and Srinivasan M.S. (2006) Jenny H. (1941) Factors of Soil Formation – A System of Fragipan controls on runoff generation: hydropedological Quantitative Pedology, McGraw-Hill: New York. implications at landscape and watershed scales. Geoderma, 131, Jury W.A. (1999) Present directions and future research in vadose 330–344. zone hydrology. In Vadose Zone Hydrology – Cutting across Germann P.F. and Hensel D. (2006) Poiseuille flow geometry disciplines, Parlange M.B. and Hopmans J.W. (Eds.), Oxford inferred from velocities of wetting fronts in soils. Vadose Zone University Press: New York, pp. 433–441. Journal, 5, 867–876. Kemp R.A., McDaniel P.A. and Busacca A.J. (1998) Genesis Gish T.J. and Shirmohammadi A. (Eds.) (1991) Preferential Flow: and relationship of macromorphology and micromorphology to Proceedings of the National Symposium. Dec. 16–17, 1991, contemporary hydrological conditions of a welded Argixeroll American Society of Agricultural Engineers: Chicago, IL. from the Palouse. Geoderma, 83, 309–329. Gish T.J., Walthall C.L., Daughtry C.S.T. and Kung K.-J.S. Kendall K.A., Shanley J.B. and McDonnell J.J. (1999) (2005) Using soil moisture and spatial yield patterns to identify A hydrometric and geochemical approach to test the subsurface flow pathways. Journal of Environmental Quality, transmissivity feedback hypothesis during snowmelt. Journal 34, 274–286. of Hydrology, 219(3–4), 188–205. Goodchild M.F., Yuan M. and Cova T.J. (2007) Towards a general Khan F.A. and Fenton T.E. (1994) Saturated zones and soil theory of geographic representation in GIS. International morphology in a Mollisol catena of Central Iowa. Soil Science Journal of Geographical Information Science, 21, 239–260. Society of America Journal, 58, 1457–1464. HYDROPEDOLOGY AND SURFACE/SUBSURFACE RUNOFF PROCESSES 23

Kieft T.L., Fredrickson J.K., McKinley J.P., Bjornstad B.N., Lin H.S., McInnes K.J., Wilding L.P. and Hallmark C.T. (1997) Rawson S.A., Phelps T.J., Brockman F.J. and Pfiffner Low tension water flow in structured soils. Canadian Journal S.M. (1995) Microbiological comparisons within and across of Soil Science, 77, 649–654. contiguous lacustrine, paleosol, and fluvial subsurface Lin H.S., McInnes K.J., Wilding L.P. and Hallmark C.T. sediments. Applied and Environmental Microbiology, 61, (1999a) Effects of soil morphology on hydraulic properties: 749–757. I. Quantification of soil morphology. Soil Science Society of Kirby M.J. (1978) Hillslope Hydrology, John Wiley & Sons Ltd: America Journal, 63, 948–954. Chichester. Lin H.S., McInnes K.J., Wilding L.P. and Hallmark C.T. Kirchner J.W. (2003) A double paradox in catchment hydrology (1999b) Effects of soil morphology on hydraulic properties: and geochemistry. Hydrological Processes, 17, 871–874. II. Hydraulic pedotransfer functions. Soil Science Society of Kirchner J.W. (2006) Getting the right answers for the right America Journal, 63, 955–961. reasons: linking measurements, analyses, and models to Lin H.S., Wheeler D., Bell J. and Wilding L. (2005b) Assessment advance the science of hydrology. Water Resources Research, of soil spatial variability at multiple scales. Ecological 42, W03S04, DOI:10.1029/2005WR004362. Modelling, 182, 271–290. Konopka A. and Turco R. (1991) Biodegradation of organic Lin H.S. and Zhou X. (2008) Evidence of subsurface preferential compounds in vadose zone and aquifer sediments. Applied and flow using soil hydrologic monitoring in the Shale Hills Environmental Microbiology, 57, 2260–2268. catchment. European Journal of Soil Science, 59, 34–49. Kubiena W.L. (1938) Micropedology, Collegiate Press: Ames, IA. Lin H.S., Zhu Q. and Schmidt J. (2008) Precision Soil Mapping Lehmann P., Hinz C., McGrath G., Tromp-van Meerveld H.J. and for Hydropedologic Functional Units and Its Application in Soil McDonnel J.J. (2007) Rainfall threshold for hillslope outflow: Moisture Monitoring and Nitrogen Management. Proceedings an emergent property of flow pathway connectivity. Hydrology of the 1st Global Workshop on High-resolution Soil Sensing and and Earth System Science, 11, 1047–1063. Mapping, University of Sydney: Australia. Letey J. (1991) The study of soil structure – science or art. Lutz H.J. and Chandler R.F. (1946) Forest Soils, John Wiley & Australian Journal of Soil Research, 29, 699–707. Sons, Inc.: New York. Lilly A. and Lin H.S. (2004) Using soil morphological attributes McClain M.E., Boyer E.W., Dent C.L., Gergel S.E., Grimm and soil structure in pedotransfer functions. In Development of N.B., Groffman P.M., Hart S.C., Harvey J.W., Johnston C.A. Pedotransfer Functions in Soil Hydrology, Pachepsky Y. and and Mayorga E. (2003) Biogeochemical hot spots and hot Rawls W. (Eds.), Elsevier, pp. 115–142. moments at the interface of terrestrial and aquatic ecosystems. Lin H.S. (1995) Hydraulic Properties and Macropore Flow of Ecosystems, 6, 301–312, DOI:10.1007/s10021-003-0161-9. Water in Relation to Soil Morphology. Ph.D. dissertation. McDaniel P.A., Bathke G.R., Buol S.W., Cassel D.K. and Falen Texas A&M University: College Station, TX. (Diss. Abstr. A.L. (1992) Secondary Mn/Fe ratios as pedochemical indicators 9534383). of field-scale throughflow water movement. Soil Science Society Lin H.S. (2003) Hydropedology: bridging disciplines, scales, and of America Journal, 56, 1211–1217. data. Vadose Zone Journal, 2, 1–11. McDaniel P.A., Regan M.P., Brooks E., Boll J., Barndt S., Lin H.S. (2006) Temporal stability of soil moisture spatial pattern Falen A., Young S.K. and Hammel J.E. (2008) Linking fragi- and subsurface preferential flow pathways in the Shale Hills pans, perched water tables, and catchment-scale hydrological Catchment. Vadose Zone Journal, 5, 317–340. processes. Catena, DOI:10.1016/j.catena.2007.05.011. Lin H.S. (2007) Cattle vs. ground beef: what is the difference? McDonnell J.J. (2003) Where does water go when it Soil Survey Horizons, 48, 9–10. rains? Moving beyond the variable source area concept Lin H.S., Bouma J., Pachepsky Y., Western A., Thomp- of rainfall–runoff response. Hydrological Processes, 17, son J., van Genuchten M.Th., Vogel H. and Lilly A. 1869–1875. (2006a) Hydropedology: synergistic integration of pedology McDonnell J.J., Sivapalan M., Vache’ K., Dunn S., Grant G., and hydrology. Water Resources Research, 42, W05301, Haggerty R., Hinz C., Hooper R., Kirchner J. and Roderick DOI:10.1029/(2005)WR004085. M.L. (2007) Moving beyond heterogeneity and process com- Lin H.S., Bouma J., Wilding L., Richardson J., Kutilek M. and plexity: a new vision for watershed hydrology. Water Resources Nielsen D. (2005a) Advances in hydropedology. Advances in Research, 43, W07301, DOI:10.1029/(2006)WR005467. Agronomy, 85, 1–89. McGlynn B.L., McDonnell J.J. and Brammer D.D. (2002) A Lin H.S., Kogelmann W., Walker C. and Bruns M.A. (2006b) Soil review of the evolving perceptual model of hillslope flowpaths moisture patterns in a forested catchment: a hydropedological at the Maimai catchments, New Zealand. Journal of Hydrology, perspective. Geoderma, 131, 345–368. 257, 1–26. Lin H.S. and Rathbun S. (2003) Hierarchical frameworks for Mosley M.P. (1979) Streamflow generation in a forested multiscale bridging in hydropedology. In Scaling Methods in watershed. Water Resources Research, 15, 795–806. Soil Physics, Pachepsky Y., Radcliffe D. and Selim H.M. (Eds.), Moore I.D., Gessler P.E., Nielsen G.A. and Peterson G.A. (1993) CRC Press: Boca Raton, pp. 353–371. Soil attribute prediction using terrain analysis. Soil Science Lin H.S., McInnes K.J., Wilding L.P. and Hallmark C.T. (1996) Society of America Journal, 57, 443–452. Effective porosity and flow rate with infiltration at low tensions Moore I.D., Grayson R.B. and Ladson A.R. (1991) Digital terrain into a well-structured subsoil. Transactions of the ASAE, 39, modeling: a review of hydrological, geomorphological, and 131–135. biological applications. Hydrological Processes, 5, 3–30. 24 RAINFALL–RUNOFF PROCESSES

Needelman B.A., Gburek W.J., Petersen G.W., Sharpley A.N. and headwaters: a linked hydrogeomorphic paradigm. Hydrological Kleinman P.J.A. (2004) Surface runoff along two agricultural Processes, 14, 369–385. hillslopes with contrasting soils. Soil Science Society of Sivapalan M. (2003) Process complexity at hillslope scale, process American Journal, 68, 914–923. simplicity at the watershed scale: is there a connection? Nimmo J.R. (2007) Simple predictions of maximum transport rate Hydrological Processes, 17, 1037–1041. in unsaturated soil and rock. Water Resources Research, 43, Smeck N.E. and Ciolkosz E.J. (1989) Fragipans: Their W05426, DOI:10.1029/(2006)WR005372. occurrence, classification, and genesis, Soil Science Society of Noguchi S., Tsuboyama Y., Sidle R.C. and Hosoda I. America, Inc.: Madison, WI. SSSA Special Publication #24. (1999) Morphological characteristics of macropores and the Soil Science Society of America (SSSA) (1997) Glossary of Soil distribution of preferential flow pathways in a forested Science Terms, Soil Science Society of America, Inc.: Madison, slope segment. Soil Science Society of America Journal, 63, WI. 1413–1423. Soil Survey Division Staff (1993) Soil Survey Manual, U.S. Nortcliff S., Quisenberry V.L., Nelson P. and Phillips R.E. (1994) Department of Agriculture Handbook No. 18, U.S. Government The analysis of soil macropores and the flow of solutes. In Printing Office: Washington, DC. Soil Micromorphology: Studies in Management and Genesis, Soil Survey Staff (1999) Soil Taxonomy – A Basic System of Soil Ringrose-Voase A.J. and Humpherys G.S. (Eds.), Elsevier: Classification for Making and Interpreting Soil Surveys, Second Amsterdam, The Netherlands, pp. 601–612. Edition, USDA-NRCS Agricultural Handbook No. 436, U.S. Onda Y., Komatsu Y., Tsujimura M. and Fujihara J. (2001) Government Printing Office: Washington, DC. The role of subsurface runoff through bedrock on storm flow Soil Survey Staff (2006) Keys to Soil Taxonomy, Tenth Edition, generation. Hydrological Processes, 15, 1693–1706. United States Department of Agriculture – Natural Resources Pennock D.J. and de Jong E. (1990) Regional and caternary Conservation Service: Washington, DC. variations in properties of Borolls of southern Saskatchewan, Sommer M. and Schlichting E. (1997) Archetypes of catenas Canada. Soil Science Society of America Journal, 54, in respect to matter: a concept for structuring and grouping 1697–1701. catenas. Geoderma, 76, 1–33. Perret J., Prasher S.O., Kantzas A. and Langford C. (1999) van Stiphout T.P.J., van Lanen H.A.J., Boersma O.H. and Three-dimensional quantification of macropore networks in Bouma J. (1987) The effect of bypass flow and internal undisturbed soil cores. Soil Science Society of America Journal, catchment of rain on the water regime in a clay loam grassland 63, 1530–1543. soil. Journal of Hydrology, 95, 1–11. Perrier E., Bird N. and Rieu M. (1999) Generalizing the Stolt M.H., Baker J.C. and Simpson T.W. (1993) Soil-landscape fractal model of soil structure: the pore-solid fractal approach. relationships in Virginia: I. Soil variability and parent material Geoderma, 88, 137–164. uniformity. Soil Science Society of America Journal, 57, Peters D.L., Buttle J.M., Taylor C.H. and LaZerte B.D. (1995) 414–421. Runoff production in a forested, shallow soil. Canadian Shield Tackett J.L. and Pearson R.W. (1965) Some characteristics of soil basin. Water Resources Research, 31(5), 1291–1304. surface seals formed by simulated rainfall. Soil Science, 99, Quisenberry V.L., Smith B.R., Phillips R.E., Scott H.D. and 407–412. Nortcliff S. (1993) A soil classification system for describing Taylor J.P., Wilson B., Mills M.S. and Burns R.G. (2002) water and chemical transport. Soil Science, 156, 306–315. Comparison of microbial numbers and enzymatic activities in Rabenhorst M.C., Bell J.C. and McDaniel P.A. (1998) Quantifying surface soils and subsoils using various techniques. Soil Biology Soil Hydromorphology, Soil Science Society of America, Inc: and Biochemistry, 34, 387–401. Madison, WI. SSSA Special Pub. #54. Thompson J.A., Bell J.C. and Butler C.A. (1997) Quantitative Richter D.D. Jr and Markewitz D. (2001) Understanding Soil soil-landscape modeling for estimating the areal extent of Change, Cambridge University Press: Cambridge, UK, pp. hydromorphic soils. Soil Science Society of America Journal, 94–96. 61, 971–980. Rinaldo A., Banavar J.R. and Maritan A. (2006) Trees, networks, Thompson J.A., Bell J.C. and Zanner C.W. (1998) Hydrology and hydrology. Water Resources Research, 42, W06D07, and hydric soil extent within a Mollisol catena in Southeastern DOI:10.1029/(2005)WR004108. Minnesota. Soil Science Society of America Journal, 62, Rodr´ıguez-Cruz M.S., Jones J.E. and Bending G.D. (2006) Field- 1126–1133. scale study of the variability in pesticide biodegradation with Tisdall J.M. and Oades J.M. (1982) Organic matter and water- soil depth and its relationship with soil characteristics. Soil stable aggregates in soils. Journal of Soil Science, 33, Biology and Biochemistry, 38, 2910–2918. 141–163. Schaetzl R. and Anderson S. (2005) Soils: Genesis and Tromp-van Meerveld H.J. and McDonnell J.J. (2006) Thresh- geomorphology, Cambridge University Press: New York. old relations in subsurface stormflow 2: the fill and spill Sidle R., Noguchi S., Tsuboyama Y. and Laursen K. hypothesis: an explanation for observed threshold behav- (2001) A conceptual model of preferential flow systems in ior in subsurface stormflow. Water Resources Research, forested hillslopes: evidence of self-organization. Hydrological DOI:10.1029/2004WR003800. Processes, 15, 1675–1692. Uhlenbrook (2006) Catchment hydrology – a science in which all Sidle R., Tsuboyama Y., Noguchi S., Hosoda I., Fujieda M. processes are preferential. HP Today. Hydrological Processes, and Shimizu T. (2000) Stormflow generation in steep forested 20, 3581–3585. HYDROPEDOLOGY AND SURFACE/SUBSURFACE RUNOFF PROCESSES 25

USDA-NRCS (1997) Land Resource Regions and Major Land Weiler M. and Naef F. (2003) An experimental tracer study Resource Areas. National Soil Survey Handbook,U.S. of the role of macropores in infiltration in grassland soils. Government Printing Office: Washington, DC. Hydrological Processes, 17, 477–493. Van Gestel M., Ladd J.N. and Amato M. (1992) Microbial biomass Weiler M. and McDonnell J.J. (2007) Conceptualizing lateral responses to seasonal change and imposed drying regimes at preferential flow and flow networks and simulating the effects increasing depths of undisturbed topsoil profiles. Soil Biology on gauged and ungauged hillslopes. Water Resources Research, and Biochemistry, 24, 103–111. 43, W03403, DOI:10.1029/2006WR004867. Vache´ K.B. and McDonnell J.J. (2006) A process-based West L.T., Abreu M.A. and Bishop J.P. (2008) Saturated hydraulic rejectionist framework for evaluating catchment runoff conductivity of soils in the Southern Piedmont of Georgia, model structure. Water Resources Research, 42, W02409, USA: field evaluation and relation to horizon and landscape DOI:10.1029/(2005)WR004247. properties. Catena, 73, 174–179. Veneman P.L.M. and Bodine S.M. (1982) Chemical and Whipkey R.Z. (1965) Subsurface storm flow from forested slopes. morphological soil characteristics in a New England drainage- Bulletin of the International Association of Scientific Hydrology, toposequence. Soil Science Society of America Journal, 46, 2, 74–85. 359–363. Wigmosta M.S., Vail L.W. and Lettenmaier D.P. (1994) A Veneman P.L.M., Lindbo D.L. and Spokas L.A. (1998) Soil distributed hydrology-vegetation model for complex terrain. moisture and redoximorphic features: A historical perspective. Water Resources Research, 30, 1665–1642. In Quantifying Soil Hydromorphology, Rabenhorst M.C., Bell Wilcox B.P., Wilding L.P. and Woodruff C.M. Jr (2007) Soil and J.C. and McDaniel P.A. (Eds.), Soil Science Society of topographic controls on runoff generation from stepped land- America, Inc: Madison, WI, pp. 1–23. SSSA Special Pub. #54. forms in the Edwards Plateau of Central Texas. Geophysical Vepraskas M.J. (1992) Redoximorphic Features for Identifying Research Letters, 34, L24S24, DOI:10.1029/2007GL030860. Aquic Conditions, North Carolina Agricultural Research Wysocki D., Schoeneberger P. and LaGarry H. (2000) Service Technical Bulletin No. 301, North Carolina Agricultural Geomorphology of soil landscapes. In Handbook of Soil Research Service: North Carolina. Science, Sumner M.E. (Ed.), CRC Press: Boca Raton, FL, pp. Vervoort R.W. and Cattle S.R. (2003) Linking hydraulic E-5–E-39. conductivity and tortuosity parameters to pore space geometry Young I.M., Crawford J.W. and Rappoldt C. (2001) New methods and pore size distribution. Journal of Hydrology, 272, 36–49. and models for characterising structural heterogeneity of soil. Vervoort R.W., Radcliffe D.E. and West L.T. (1999) Soil structure Soil and Tillage Research, 61, 33–45. development and preferential solute flow. Water Resources Zhu A.X., Hudson B., Burt J., Lubich K. and Simonson D. (2001) Research, 35, 913–928. Soil mapping using GIS, expert knowledge, and fuzzy logic. Vogel H.J., Cousin I. and Roth K. (2002) Quantification of pore Soil Science Society of America Journal, 65, 1463–1472. structure and gas diffusion as a function of scale. European Zhu Q., Lin H.S. and Doolittle J. (2008) Functional soil mapping Journal of Soil Science, 53, 465–473. for soil moisture and crop management in an agricultural Vogel H.J. and Roth K. (2003) Moving through scales of flow and landscape. Agronomy Journal,Inreview. transport in soil. Journal of Hydrology, 272, 95–106. Zuzel J.F., Allmaras R.R. and Greewalt R. (1982) Runoff and soil Weiler M., McDonnell J.J., Tromp-van Meerveld I. and Uchida T. erosion on frozen soils in northeastern Oregon. Journal of Soil (2005) Subsurface stormflow runoff generation processes. In and Water Conservation, 37(6), 351–354. Encyclopedia of Hydrological Sciences, Anderson M.G. (Ed.), John Wiley& Sons Ltd, pp. 1719–1732.