Figure 5 - uploaded by Vasile Ersek
Content may be subject to copyright.
Source publication
Flank margin caves are hypogenic
mixing chambers formed in the margin of the
fresh-water lens on carbonate islands and coasts.
They develop, by mixing dissolution, without
turbulent flow, in a very short amount of
geological time. The caves start as small
elliptical chambers with irregular protrusions.
As chambers increase in size, laterally adjace...
Citations
... The island is characterised by a sequence of Pleistocene shallow-water carbonate deposits covering the oceanic crust basement (Meyerhoff and Hatten, 1974;Supko, 1977;Carew and Mylroie, 1985). The dissolution of Bahamian carbonates produced karstic features such as karren, shallow depressions, blue holes, and the well-known flank margin caves (Roth et al., 2006;Labourdette et al., 2007). Flank margin caves (FMC) (Mylroie and Carew, 1990;Harris et al., 1995;Gulley et al., 2016) generally present subhorizontal branches that develop at the edge of the freshwater lens where the area of vadose/phreatic water mixing and fresh-seawater mixing zones are superimposed. ...
San Salvador (Bahamas) is a carbonate island with dozens of flank margin caves formed in the phreatic zone by fresh seawater mixing within the freshwater lens. These caves have no direct connection with the sea, and form at or close to the tidally influenced fluctuating water table. After sea-level fall, in their subaerial parts caves are enlarged mainly by rock dissolution and by erosion close to the water level, condensation-corrosion and breakdown processes. For understanding the geomorphological features observed in these caves and how they are related to light attenuation, we investigated three sampling sites in the tidally influenced zone of Lighthouse Cave, which has been re-invaded by seawater during the Holocene sea-level highstand. A freshwater lens no longer exists within or adjacent to the cave. Rock samples were collected above and below the internal lake shores close to the entrance, and in the twilight and dark zones of this cave. Light and electron microscopy examinations were conducted for detecting microbial cells, as well as bioconstruction and bioweathering features. In addition, a high precision laser scanner was used for characterising sample microtopography. Our data showed that the microtopography and geomorphology of the lake shore samples (cave entrance) are dominated by bioweathering, whereas the samples of the twilight and dark zones are controlled by a combination of both bioweathering and bioconstructive processes depending on light availability. Bioconstructive structures, such as semi-planar lamination, at the fluctuating water level of the Lighthouse Cave show that dissolution due to water mixing of sea and freshwater in the Holocene is no longer the most important speleogenetic process. We propose that the geomorphological evolution is strongly influenced by the degree of rock diagenesis more than the initial mechanism of speleogenesis.
... Dissolutional surfaces indicating phreatic development were also abundant (Figure 6). Morphological analysis of the cave maps indicated that banana holes plotted in the smaller end of the flank margin cave field (Figure 7), quite different from the field plot for stream caves as found in continental interiors (stream caves are turbulent-flow cave conduits; flank margin caves and banana holes form under laminar-flow conditions, see Roth et al., 2006). See Infante (2012) for all maps and data related to this study. ...
... flow flank margin caves (Roth et al., 2006) ...
Banana holes are common karst features found within Pleistocene strandplains of San Salvador Island and throughout The Bahamas. Banana holes are typically meters to tens of meters wide, and one to six meters deep, ovoid in plan, and usually contain phreatic dissolution features. A study of banana holes was conducted to see if their origin could be better explained, and to determine if their locations could be predicted using remote sensing. A detailed banana hole investigation of the Line Hole, Jake Jones Road, Hard Bargain, and South Victoria Hill areas of San Salvador was conducted during field visits in December 2010, May 2011, and December 2011, locating 390 banana holes, 70 of which were mapped. Plotting banana hole locations in ArcGIS shows they follow linear trends and are associated with low inland ridges. Facies in banana hole wall rock show an upward progression from shallow subtidal (herringbone cross beds) to intertidal (laminar beds and or back-beach breccia blocks) to supratidal (eolian foresets) deposits, in agreement with a prograding beach environment. The flat topography of low, coast-parallel ridges mimics modern strandplains. Dissolution features found in banana holes include wall cusps, blind passages, bell holes, and thin wall partitions. These observations suggest that banana holes form at the lens margin within a prograding strandplain. As the strandplain advances the lens margin follows, creating syndepositional voids at the dissolutionally aggressive lens margin. This sequence of dissolution and abandonment creates rows of voids that represent paleo-shoreline positions. Comparison of banana hole and flank margin cave morphometry based on cave area versus enclosing rectangle area, shows an overlap, indicating a common dissolutional origin. This new model for banana hole formation requires that the term " banana hole " be used to describe flank margin voids formed within progradational facies, instead of a unique upper-lens dissolutional mechanism as previously proposed. In an attempt to predict banana hole locations, spatial models with unsupervised image classifications (Digital Elevation Model, ASTER 1a and Landsat 5 TM) were generated using different predictors: Normalized Different Vegetation Index (NDVI), Normalized Different Water Index (NDWI), Carbonate-Chlorite Index, Dolomite Index, and Slope. Most of the predictors had a strong capability to locate island karst features when combined with slope and elevation as the parameters; NDWI, NDVI, Carbonate-Chlorite Index, and elevation had the highest capability compared to other predictors. By using the fieldwork data from Dec 2010 and May 2011 as the validation dataset, the model showed a 95% percent correct, 0.4% false alarm ratio, 1% probability of false detection, and 93% critical success index. By using logistic regression to validate the model, it showed that Akaike's Information Criterion (AIC) is 10. By using geographic weight regression to validate the model, 91.5% of the validation data fell into standard deviation residual between-1 and 1, indicating that the model was accurate and had the ability to predict karst features such as flank margin caves and banana holes in the future.
... The differentiation of sea caves from flank margin caves is not especially useful in sealevel studies, as the caves both form at sea level. However, differentiating a sea cave from a flank Roth et al. (2006) almost exactly, and (b) revealed a smaller subset of cave size, under 20 m 2 , which cave mappers had ignored as too small (From Labourdette et al. 2007) margin cave is critical for calculation of denudation rates based on degree of cave destruction (Waterstrat et al. 2010). Sea caves form from the outside inward, and are always open to the surface. ...
Coastal karst cave development globally is biased towards the tropics and subtropics, where carbonate deposition is ongoing, and therefore carbonate coasts are common. The Carbonate Island Karst Model (CIKM) delineates the unique conditions that separate coastal karst from traditional karst areas of continental interiors. In these warm-water carbonate islands, diagenetically immature, or eogenetic carbonate rocks are host to a freshwater lens that creates flank margin caves in a diffuse flow environment. Diagenetically mature, or telogenetic carbonates, can also host flank margin caves. Flank margin caves can form rapidly, as carbonate sediment is deposited, to produce syndepositional caves called banana holes. Flank margin caves can survive as open voids for millions of years, and as infilled diagnostic features for tens of millions of years. Vadose fast flow routes called pit caves form as a result of surface micritization to provide point recharge to the freshwater lens. The presence of non-carbonate rocks can perch vadose flow, creating stream caves that terminate in the freshwater lens. When sea level falls to create large exposed carbonate platforms, phreatic conduit flow develops to carry recharge to the platform periphery. Collapse of these conduits, as well as bank margin fracture, account for the majority of blue holes in carbonate platforms. Closed depressions in eogenetic carbonate islands are commonly constructional, relicts from variable carbonate deposition. The most common sinkhole type is the cave-collapse sinkhole. Morphometric analysis of flank margin caves supports cave origin as the amalgamation of individual chambers, provides evidence of denudation rates, and can differentiate flank margin caves from some pseudokarst cave types.
... One final interesting aspect of flank margin caves is that they occur on scales from small chambers up to immense caves without loss of their general morphology or position with respect to the land surface. Size-rank plots reveal that the caves self-select in three size groups based on growth of the initial voids, amalgamation of voids, and then amalgamation of clusters of voids (Roth et al., 2006). Computer modeling of void genesis creates the same plot, but adds a fourth group of tiny caves that were ignored by cave surveyors (Labourdette et al., 2007). ...
... In order to quantify the distinctions between flank margin caves and sea caves, a series of morphometric parameters were created (Roth, 2004;Roth et al., 2006;Waterstrat, 2007). From the cave maps, several measurements were taken: (1) perimeter (excluding the length of the entrance dripline); (2) area (enclosed by the walls and dripline); (3) entrance width; (4) maximum width (roughly parallel to entrance measurement); and (5) the length of the axes of the smallest possible rectangle to enclose the entire cave. ...
Coastal areas on carbonate islands commonly contain two types of caves: sea caves developed by wave erosion processes, and flank margin caves developed by dissolution at the edge of the fresh-water lens. Differentiating sea caves and flank margin caves in coastal settings is important, but can it be done reliably and quantitatively? Current methods use the degree of intricate wall-rock dissolution and the presence or absence of dense calcite speleothems to separate the two cave types. This study reports how analysis of cave maps creates three separate tools to differentiate coastal caves: area to perimeter ratio, entrance width to maximum width ratio, and rectangle short axis to long axis ratio. The study also presents some of the first sea cave data from eogenetic carbonate islands, specifically eolian calcarenites. The morphological and geometrical comparisons between Bahamian flank margin cave and sea cave maps using the three tools allows the two cave types to be statistically differentiated. The Bahamian sea cave data were also compared to sea cave data from California and Maine to demonstrate that Bahamian sea caves have a unique quantitative signature based on the youth and homogeneity of the host eolian calcarenite rock. The Bahamian sea cave data also indicate that sea cave formation may not be solely determined by differential rock weaknesses, as reported in the literature, but may also be a result of wave dynamics such as constructive interference.
... As a result, until recently the patterns that drive eogenetic caves such as flank margin caves were not easily interpreted. A rank order plot of flank margin caves based on areal footprint (Fig. 14), from the Bahamas, shows that the caves self-select into three size categories ( Roth et al., 2006). Areal footprint, or cave area, was selected as the size determiner because it is the best measure of how much dissolution has occurred. ...
The development of a comprehensive conceptual model for carbonate island karst began in the Bahamas in the 1970s. The use, initially, of cave and karst models created for the interior of continents, on rocks hundreds of millions of years old, was not successful. Models developed in the 1980s for the Bahamas, that recognized the youthfulness of the carbonate rock, the importance of fresh-water mixing with sea water, and the complications introduced by glacioeustatic sea-level change produced the first viable model, the flank margin cave model. This model explains the largest caves in carbonate islands as being the result of mixing zone dissolution in the distal margin of the fresh-water lens, under the flank of the enclosing land mass. The flank margin model, taken from the Bahamas to Isla de Mona, Puerto Rico, in the early 1990s, provided the first viable explanation for the very large caves there. Field work in the geologically-complex Mariana Islands in the late 1990s resulted in the development of the Carbonate Island Karst Model, or CIKM, which integrated the various components controlling cave and karst development on carbonate islands. These components are: 1) Mixing of fresh and salt water to create dissolutional aggressivity; 2) Movement of the fresh-water lens, and hence the mixing environments, by 100+ m as a result of Quaternary glacioeustasy; 3) The overprinting of glacioeustatic changes by local tectonic movements, where present; 4) The unique behavior of eogenetic (diagenetically immature) carbonate rocks; and 5) The classification of carbonate islands into simple, carbonate cover, composite, and complex categories. Current research involves the use of flank margin caves as predictors of past and present fresh-water lens configuration, the analysis of flank margin cave morphology as a measure of the processes that create them, and the CIKM as an indicator of paleokarst distribution.
