3.5 Conduit Drainage Patterns
Speleogenesis, the origin and development of caves and by extension karst conduits, is a dynamic hydrogeologic process and is the primary mechanism by which conduit flow systems evolve and acquire more “karstic” characteristics and complex organization (Klimchouk, 2015). Speleogenesis results in the most universally-recognized feature of karst aquifers—the networks of linked underground conduits that meander among bedding units, join as tributaries, and increase in size and order in the downstream direction (Palmer, 1991). Conduits act as master drains for a karst aquifer and do not form at random but rather form along groundwater paths of greatest discharge and solutional aggressiveness (Palmer, 1991). Thrailkill (1968) suggested that conduit development is most active in the mixing zone created at or near the water table (the unsaturated/saturated interface). Palmer states that “Solutional caves form where there is enough subsurface water flow to remove dissolved bedrock and keep undersaturated water in contact with the soluble walls. This is possible only where a pre–existing network of integrated openings connects the recharge and discharge areas.” Bakalowicz (2005) theorizes that conduit formation is often initiated by, and approximately follows, the pattern created by existing fractures. He envisions that hydraulic conductivity increases as a result of hydraulic and geochemical feedback between flow through the larger, more permeable fractures and the upper, near-surface, densely fractured, permeable part of the bedrock mass.
The geospatial pattern of conduit network formation may be extremely complex in plan-view and in three dimensions, influenced by stratigraphy and geological structure, as well as by locations of recharge. Common cave network patterns, shown in Figure 23 include branching-dendritic, which in plan-view resembles surface stream tributary systems; network or anastomotic mazes, often prevalent in karst aquifers with extensive fracture permeability and frequent episodic low-velocity flooding, respectively; and spongework, typical of eogenetic karst aquifers dominated by extensive matrix permeability. As described by Palmer (1991), “Their patterns depend on the mode of groundwater recharge. Sinkhole recharge forms branching caves with tributaries that join downstream as higher–order passages. Maze caves form where (1) steep gradients and great undersaturation allow many alternate paths to enlarge at similar rates or (2) discharge or renewal of undersaturation is uniform along many alternate routes. Flood water can form angular networks in fractured rock, anastomotic mazes along low–angle partings, or spongework where intergranular pores are dominant. Diffuse recharge also forms networks and spongework, often aided by mixing of chemically different waters. Ramiform caves, with sequential outward branches, are formed mainly by rising thermal or H2S-rich water.”
Stratigraphic contacts and the relative position and stability of the water table influence lateral distribution of conduit development. In carbonate rock sequences, individual stratigraphic units may contain considerable amounts of non-soluble, erosion-resistant siliceous mineral, particularly chert, or relatively less soluble dolomite and mudstones. These strata may act as bounding units that confine conduit and sinkhole development to one or more distinctive horizons (Figure 24).
Multilevel conduit networks are typical of karst terranes where the base-level of main surface streams is actively or episodically lowered due to stream downcutting in response to tectonic uplift, isostatic (glacial) rebound, or other causes. Multiple levels of formerly active higher-elevation conduits may be preserved, especially if protected from erosion under non-karstic caprock, as for example demonstrated in the internationally renowned Mammoth Cave-Flint Ridge Cave system in central Kentucky (USA), where several levels of horizontally extensive conduit networks are preserved under overlying sandstone conglomeratic caprock. In such multilevel conduit systems, the high-level conduits may be dry and part of the vadose-conduit zone as depicted in Figure 6, or, if they are located at elevations within the range of fluctuation of the water table, they may be reactivated as overflow routes.
The size and shape of individual conduits or conduit segments are likewise influenced by lithologic, stratigraphic, and structural conditions, as well as hydrologic conditions that occurred both when the conduit development was initiated and subsequently. Klimchouk (2015) states “Passages influenced by bedding–plane partings are sinuous and curvilinear … Closely spaced joints within favorable beds may produce a similar pattern (Powell, 1976). Solutionally enlarged joints and high–angle faults tend to produce fissure–like passages with lenticular cross sections and angular intersections. Where joints are prominent, they can determine the pattern of nearly every passage in a cave … Faults usually exert only local control of cave passages and determine the overall trend of relatively few caves (Kastning, 1977). Intergranular pores are significant to cave origin only in reef limestones and poorly lithified carbonates.”
Again, quoting Palmer (1991) and noting his use of the term phreatic refers to the saturated zone: “Phreatic passages originate along routes of greatest hydraulic efficiency (least expenditure of head per unit discharge). Such a passage enlarges solutionally over its entire perimeter and usually acquires a rounded or lenticular cross section. Most are tubular passages … although some phreatic caves are irregular and room–like … A passage along the water table may be water filled only during high flow and still meet the criteria for phreatic origin … Passages of vadose origin are formed by gravitational flow and trend continuously downward along the steepest available openings … Most vadose passages are canyon–like with floors entrenched below the initial route by free–surface streams … They may be tubular where entrenchment is limited by resistant beds or insufficient time. Water descending vertically along a fracture or a cluster of intersecting fractures may form a shaft, a well–like void with nearly vertical walls … A typical vadose passage consists of inclined canyons or tubes interrupted in places by shafts.” Examples of a tubular and a canyon-like conduit passage are provided in Figure 25 and Figure 26, respectively.
Depending on their diameters (that is, their hydraulic capacities) and organization (interconnection), conduit networks are capable of discharging large volumes of water and sediment rapidly through a karst aquifer (White, 1993). Flow velocities in well-developed and well-integrated conduit networks that range on the order of 100’s to 1000’s of feet per day (10’s to 100’s of meters per day) are not uncommon (White, 1988). “Sediment loads discharged by karst aquifers is a largely unrecognized and unappreciated process. Huge volumes of sediment are mobilized and transported in many karst conduits during and after storms when turbulent flow exceeds the critical shear stress of sediments. The mobilization and deposition of sediments in karst aquifers often affects the quality of karst groundwater resources and may have significant influence on transport and fate of subsurface contaminants.”
Flow discontinuities that occur within or between conduits, or between conduits of the unsaturated and the saturated zones, is an under-recognized characteristic of karst aquifers that differentiates them from other aquifer types. Horizontal discontinuities (for example, breaks in flow manifested as waterfalls or cascades) are common within and between horizontal conduit segments where flow is perched by a resistant or insoluble bed (chert for example). Vertical flow discontinuities are commonly observed as seeps and drips from fractures and permeable zones in conduit passage roofs, and as waterfalls or laminar sheet-flows on the walls in shafts and domes (Figure 27).