Dating in the Craven Dales

Abstract

The Craven Dales of North Yorkshire contain some of the finest examples of limestone geology and landscape in the UK. The extensive limestone pavements with their clints, grikes and other water-worn features, are a key attraction for both scientists and recreational visitors. Likewise the cave systems attract attention with their remarkable sediment accumulations, some of which are in excess of 500 000 years old and others contain the bones of mammals that are either extinct or no longer present in the British Isles. The glacial erratic boulders at Norber and the loessic sediments that, in places, mask the limestone have also provided stimulus for investigation. Summarized below are the findings of several recent studies that focussed on establishing the age of various features. Whilst the reports answer some long-standing questions, they also demonstrate that there is still much that can be learned about this seemingly familiar territory.

It is not difficult to understand why the rocks and karst landforms of the Craven Dales of North Yorkshire in northern England have generated considerable interest amongst scientists and non-scientists alike. The limestones of the area have enabled geologists to gather valuable information about the tropical Carboniferous seas and the creatures that lived and died there. They have provided quarrymen with a strong and pure rock much prized by industry and agriculture, and they have been fashioned into a profusion of caves and crags that challenge the imagination and skill of cavers and rock climbers.

The landforms have also stimulated scientific enquiry, particularly the extensive and flat limestone pavements with their abundance of clints, grikes and other water-worn features (together known as karren). A theme of recurrent interest has been the effect of glacial ice on pavement formation and the wider landscape. In recent years knowledge of the spatial and temporal aspects of ice cover in Craven during the Quaternary period and the landscape changes that followed decay of the last ice sheet has been advanced by the application of various dating techniques (see boxed text). Establishing the age of materials and events has enabled geologists, geomorphologists and archaeologists to answer some long-standing questions, and to better understand the processes that have shaped the landscape and the rates at which they operate.

Limestone pavements

Limestone pavements are among the most distinctive landforms in the British Isles and have fascinated generations of geologists and geomorphologists. It is probably the pavements, above all other things, which come to mind first when the Craven Dales of North Yorkshire are mentioned. The combination of hard, relatively pure Carboniferous limestones and recent glaciation has resulted in large tracts of bare rock surfaces that, in addition to their intrinsic landscape value, harbour unusual plant communities and several rare species.

Much of the research conducted on limestone pavements has focussed on their morphology and morphometry, with rather limited linkage being made to available geological information. It is now thought that the morphological diversity of pavements as illustrated in Figs 1A and B can be better understood through an appreciation of Lower Carboniferous (Asbian to Brigantian stages) geology and the cyclical nature of relative sea-level changes, and associated lime-rich sediment accumulation. The Craven limestones display a variety of characteristics related to this cyclicity and to contemporaneous Carboniferous sub-aerial weathering, calcrete development, and the formation of karstic surfaces.

Figure 1. Contrasting types of limestone pavement on the lower slopes of Ingleborough. a. Weak, thinly-bedded limestone with numerous small clints, deep and wide grikes, and considerable evidence of surface fragmentation (mechanical disintegration). b. Pavement developed on strong, massive limestone with extensive smooth clints, deep grikes and an absence of surface fragmentation.

In essence, some limestones are thinly-bedded and well-fractured, others are thicker-bedded and stronger. The weaker beds generally display small clints separated by deep and wide grikes, and an abundance of mechanically-fragmented limestone. Stronger more massive limestone beds, while also having deep grikes, possess smooth clint surfaces with little if any evidence of fragmentation.

Some of the limestone pavements undoubtedly result from severe sub-glacial erosion by abrasion and plucking (glaciokarstic pavements) others have been exhumed by glaciation, and they constitute a palaeokarstic surface—a pavement that was created during the Carboniferous period which has been revealed rather than created by glacial erosion.

Loessic sediments

Across much of the limestone outcrop there is a cover of very well-sorted brown silty sediment, up to 1.5 m thick, that supports a diverse flora. The sediment is not the residue of limestone solution, which usually yields less than 5 percent insoluble material, nor is it the result of in situ weathering of glacial tills; rather it is of wind-blown (aeolian) origin, having been winnowed from glacial and glaciofluvial deposits, and is known as loess (from the German word löss, meaning loose).

Deposition of the Craven loess was originally assumed to have occurred in the interval 18 000–12 000 years ago, following the decay of the last ice sheet but before the amelioration of climate that paved the way for woodland expansion and the effective ‘sealing’ of the ground surface from the erosive action of wind. However, a programme of optically stimulated luminescence (OSL) dating has revealed that the loess has a much more complex history than previously envisaged (see box for technical details of the technique).

Sampling was focussed on a number of dolines (solution depressions in the limestone), at elevations of 100–415 metres above sea level, where loess has accumulated and been preserved. In all the excavations the loess was friable and structureless, and, using the Munsell Soil Color system, generally yellowish brown (10YR5/4, 6/4) or brownish yellow (10YR6/6).

From the lower part of a pit at Dowkabottom (Grid Ref. SD 594 690; Figs 2A and B), near Arncliffe, OSL ages of 27 200 ± 2600 years and 27 800 ± 2600 years were obtained. These ages indicate that: (1) some of the Craven loess was deposited before the advance of the last ice sheet across the region; (2) the doline was in existence before the last ice sheet covered the area; and (3) the loess at this location was protected from the scouring action of the ice sheet. Prior to obtaining these OSL ages such statements would have been difficult to justify.

Figure 2. a. View across the large doline at Dowkabottom, above Littondale, with pile of loessic sediment from the excavated pit in centre of photograph. b. Excavated pit in loessic sediment at Dowkabottom. The survey pole is 1 m in length.

At New Close (SD 911 645), near Malham, an OSL age of 16 500 ± 1700 years provides a minimum age limit for the deglaciation of Craven, because the ice had to have vacated the area in order for the loess to accumulate. Thus, the New Close and Dowkabottom ages constrain the presence of the last ice sheet in this part of northern England to a period of about 10 000 years.

Elsewhere in Craven several loess samples have given OSL ages of around 8200 years. These cannot be primary loess, as the ice sheet had long gone, but they can be explained by the reworking of loess by overland flow, subsoil piping, and aeolian transport. The clustering of ages around 8200 years suggests a common cause for loess reworking and the most likely trigger mechanisms are human activities and/or climate change. However, the archaeological evidence for a significant human presence in the region and an associated major impact on landscape 8200 years ago is rather limited and this mechanism has been rejected.

In contrast, the evidence for marked deterioration of climate 8200 years ago in the North Atlantic region has been detected in the ice-core records from Greenland and sediments from the North Atlantic Ocean. Sediments from Hawes Water, north Lancashire, also indicate a climate downturn with mean annual air temperature reduced by approximately 2 °C. This probably resulted in greater amounts of winter snowfall, more meltwater flooding and soil saturation in the spring-summer period, and an increase in frost-related processes such as gelifluction and freeze-thaw. These conditions are likely to have had an adverse effect on vegetation and prepared the way for friable loess with little shear strength to be eroded and redeposited in topographic depressions such as dolines.

Use of the term ‘loessic sediments’ acknowledges the fact that not all silty sediments on the limestones of Craven are primary loess deposits; much of the material has a more complex history, being (re)deposited at different times by different processes. Although the loessic sediments of Craven are a shallow and patchy veneer in the landscape, their intimate linkages to the last ice sheet and subsequent environmental changes demonstrate an importance disproportionate to their thickness and extent.

Norber erratics

Norber Brow (SD 768 699) is the site of hundreds of greywacke erratic boulders of the Austwick Formation (Silurian). Many of the erratics are perched on pedestals or on the edge of steps of Malham Formation limestone (Carboniferous) that rise 30–50 cm above the level of the adjacent ground (Figs 3A,B). They have become the most widely known set of erratics in England because of their sizes, local topographic situation and accessibility.

Figure 3. a. Erratic boulder of Silurian greywacke on a limestone pedestal at Norber; scale bar is 30 cm in length. b. Norber erratic on a limestone step. The survey pole is 1 m in length. c. Solutional features and mechanical fragmentation characterize this area of limestone at Norber.

The boulders have been the subject of scientific enquiry since the nineteenth century and it is now generally agreed that they were emplaced by ice that moved from Ribblesdale across the Sulber-Moughton ridge into Crummackdale from where they were quarried, from Silurian outcrops, carried for a distance of about 1 km, and then deposited on the limestone at Norber.

Norber has become the classic area in the British Isles for demonstrating the amount and rate of post-glacial surface lowering of limestone pavements by weathering. But for such estimates to be meaningful the timing of boulder deposition needs to be known with certainty. Previously, it was inferred that the erratics had been deposited between 12 000 and 15 000 years ago; now cosmogenic nuclide exposure dating has given a more reliable age estimate of 17 900 ± 1000 years for their emplacement (see box for technical details of the technique). Although this age is somewhat greater than had earlier been proposed it is compatible with the OSL age from loess at New Close (see above), and with other ages for ‘deglaciation’ from elsewhere in northern England.

The exposure age provides a more realistic timeframe for assessing the development processes that created the pedestals on which some of the erratics are perched. A long-held view has been that surface lowering of the limestone at Norber was achieved largely by dissolution. The newly established timeframe and the associated pattern of climate changes opens up the likelihood that a greater range of surface processes have contributed to limestone surface lowering, including mechanical disintegration as evidenced by small areas of fragmented limestone pavement (Fig. 3C).

Moughton limestone pavements

At Moughton (SD 778 728), overlooking the head of Crummackdale, there are extensive areas of exposed glaciokarstic limestone pavement (Fig. 4). Because these pavements are only 3 km north-northeast of Norber the exposure age of 17 900 ± 1000 years for erratic emplacement can also be taken as the best age estimate for deglaciation at Moughton. Since then, cosmogenic ³⁶Cl has been accumulating in the limestone and, assuming a constant rate of ³⁶Cl production, it is possible to determine the amount that should be present in the rock (see box for technical details of the technique). However, limestone is susceptible to both chemical and mechanical weathering which removes the outer layers of rock and with it some of the accumulated ³⁶Cl. Therefore on a weathered limestone surface the concentration of ³⁶Cl will be lower than expected. The difference between the expected and the actual concentrations of ³⁶Cl can be used to calculate a surface lowering rate since the limestone was first exposed. Analyses of four samples from Moughton indicate that the pavements have lost between 22 cm and 45 cm thickness (average of 33 cm) of limestone since deglaciation.

Figure 4. Part of the extensive glaciokarstic limestone pavement at Moughton at the head of Crummackdale.

Dating techniques—technical information

Optically stimulated luminescence (OSL) dating

When mineral grains (such as quartz and feldspar) are buried in sediment they receive ionising radiation from naturally-occurring isotopes of potassium, uranium and thorium. This radiation strips some electrons from the outer shells of the mineral atoms. The electrons then become trapped temporarily within defects in the mineral crystal lattice but can be released by directing a beam of light onto the sample; the amount of light emitted (luminescence signal) by the illuminated mineral grain is proportional to the total absorbed radiation dose. The amount of trapped electrons is related to both the duration and intensity of radiation exposure. When mineral grains are exposed to natural daylight, their trapped electrons are released thus ‘bleaching’ the signal—sometimes called ‘resetting the clock’. On subsequent reburial, the mineral grains once again start to accumulate trapped electrons. The luminescence signal is a measure of the time elapsed since burial and removal from sunlight. OSL dating has a range from a few hundred years to about 400 000 years.

Surface exposure dating with cosmogenic nuclides

Cosmogenic nuclides (or isotopes) are produced when a rock surface is bombarded by high-energy neutrons and other subatomic particles known as cosmic rays. Cosmic rays interact with all elements contained in the minerals of the rock and create new nuclei. The most commonly used nuclides, Beryllium-10 (¹⁰Be), Aluminium-26 (²⁶Al) and Chlorine-36 (³⁶Cl) accumulate in rock over time depending on the half-life of the isotope, the erosion rate, the composition of the rock and the intensity of the cosmic rays. By measuring the concentration of an isotope and knowing its rate of production, it is possible to establish how long the rock surface has been exposed to cosmic radiation. Exposure dating has a range from a few hundred years up to several million years.

Radiocarbon (¹⁴C) dating

Radiocarbon is produced in the upper atmosphere and enters the global carbon cycle. As a consequence of life processes all organisms accumulate and retain a small amount of radioactive carbon. When an organism dies, exchange ceases, and the ¹⁴C undergoes decay at a known rate. By comparing the amount of ¹⁴C remaining in fossil material with that present in modern material, an age of death can be determined for the organism. Radiocarbon dates can be obtained on wood, peat, charcoal, shell, coral, bone and soil organic matter. Because there have been long-term variations in the production rate of radioactive carbon, radiocarbon years are not the same as calendar years. Therefore a ‘calibration’ (conversion) must be applied and is usually done using tree-rings for the last 12 000 years. Beyond this limit, a combination of fossil corals, laminated marine sediments, and uranium/thorium dating has been used to extend the calibration to 24 000 years. Radiocarbon dating has a range from about 250 years to 60 000 years.

Uranium–thorium (U-Th) dating by thermal ionisation mass spectrometry (TIMS)

Uranium–thorium dating is routinely used for age determination of both marine and terrestrial carbonates. In caves, naturally occurring uranium isotopes transported in solution by groundwater become incorporated in calcite speleothems. The crystals within the speleothem act as a closed system so that as the uranium isotopes undergo radioactive decay to thorium and other daughter isotopes these products remain locked within the calcite. Unlike the parent uranium isotopes, the thorium isotopes are insoluble and in samples of pure calcite must originate from the decay of the parent uranium since crystallisation. Measurements of the ratios of the uranium isotopes to their daughter products enable the speleothem to be dated. U-Th/TIMS dating provides very precise radiometric age determinations for speleothems over a range from a few hundred years up to about 500 000 years.

This range of values indicates that there has been some spatial variation in the amount of limestone weathering. Earlier studies had proposed that limestone surface lowering (and pedestal development at Norber) was the result of dissolution beneath a cover of sediment but it is now becoming clear that surface lowering, like the loessic sediment described above, is more complex than previously thought.

Since deglaciation, climatic conditions in northern England have not been stable, rather there have been several marked changes in temperature and precipitation. These fluctuations from cold to temperate conditions are likely to have caused rates of surface lowering to vary, rather than remaining constant. In cold periods surface lowering is likely to have resulted from a combination of dissolution in association with snow melt, and freeze-thaw shattering; in temperate periods rainfall is likely to have caused dissolution.

The presence, thickness and composition of a former sediment cover on the Moughton pavements must also be considered with respect to surface lowering. The current absence of sediment cannot be taken to indicate that it was never present; it may have been eroded. A former sediment cover of glacial till and/or loess is likely to have been calcareous, and dissolution of the underlying limestone would only have become significant as the sediment was leached of its carbonate. Because sediment is lacking, we do not know for how long sub-sediment dissolution might have occurred. Furthermore, if sediment did previously exist, it might not have been of uniform thickness. All these factors could account for the range of surface lowering values.

Cave deposits

Many of the caves in Craven contain layers of laminated, clay-rich sediment, mammal bones and speleothems (calcite dripstones and flowstones), and since the late nineteenth century they have provided geologists and archaeologists with a wealth of information concerning environmental changes both in the caves and in their vicinity. Some of these materials are amenable to radiocarbon (¹⁴C) dating, others to uranium–thorium (U-Th) dating by means of thermal ionisation mass spectrometry (TIMS) (see box for technical details of the techniques), and both techniques have provided valuable age constraints on processes and events in cave history.

From a part of Kinsey Cave (SD 804 657), north-west of Settle, a natural accumulation of Lateglacial large mammal bones dominated by the remains of brown bear (Ursus arctos) (Fig. 5) was excavated in the 1920s. Bones of two adult and one young adult brown bear have now been radiocarbon dated. The skull of one of the adult animals gave a ¹⁴C age, when calibrated, of 14 600 ± 400 years. This represents the earliest dated evidence for large mammals in northern England following the last period of glaciation. The age corresponds with the abrupt warming signal at 14 700 years ago in the ice core records from Greenland. In turn this suggests that between deglaciation at Norber, around 17 900 years ago, and 14 700 years ago the environment probably remained very cold, and coupled with the deposition of loess at New Close around 16 500 years ago it must have been both arid and windy as well. Such conditions are likely to have proved inhospitable for large mammals, and it was not until a marked warming around 14 700 years ago that they were able to colonise Craven.

Figure 5. Skull of brown bear (Ursus arctos), from Kilnsey Cave (photo courtesy of Terry O'Connor).

Lateglacial mammal bones and occasional human artefacts have been excavated at several other caves in Craven, the best known being Victoria Cave (SD 838 650, Fig. 6A), north-east of Settle. Radiocarbon dating on key specimens and the analysis of their contexts have allowed archaeologists and geologists to evaluate the nature of the ecosystem and the role of mammals in that system. The earliest dated evidence for humans comes from Victoria Cave where an antler rod-like artefact and a stone tool cut-marked wild horse vertebra excavated in 1870 show that people were hunting wild horses in the area by approximately 14 400 years ago. This suggests that humans reached Craven a couple of hundred years or so after brown bears and the onset of warming at 14 700 years ago.

Figure 6a. The entrance to Victoria Cave. b. Part of the laminated clay sequence in Victoria Cave. These clays are age constrained by U-Th/TIMS dating of speleothem to approximately 430 000 years.

Victoria Cave is also of great interest and significance to archaeologists and geologists because its sequence of deposits records evidence of clastic sediment accumulation and speleothem formation over multiple glacial–interglacial climate cycles. In the excavations of the 1870s a bone bed with species of an interglacial or ‘warm climate’ mammal fauna was discovered. The bones included those of hippopotamus (Hippopotamus amphibious), an extinct elephant (Palaeoloxodon antiquus), an extinct rhinoceros (Stephanorhinus hemitoechus), fallow deer (Dama dama) and spotted hyaena (Crocuta crocuta). In recent years the speleothem that covered the bone bed has been dated by U-Th/TIMS to 114 000 years ago, indicating that the mammal species are a representative of the fauna that inhabited Craven during the Last (Ipswichian) Interglacial, around 120 000–115 000 years ago.

Speleothems in Victoria Cave are now known to have formed during a number of interglacial periods, and samples from the different stratigraphical units have been dated using the U-Th/TIMS method. The results demonstrate that the earliest speleothem development was beyond the range of the dating technique (older than 600 000 years); other units correspond in age with Quaternary interglacial stages and postdate the warmest part of the respective stages. Some of the speleothem ages constrain deposition of the laminated clays. The laminated sediments have not yet been studied in detail but must have formed during glacial stages when the cave entrance was blocked by ice, and lakes existed inside the cave; the clays represent settlement from suspension (Fig. 6B).

The sedimentary sequence in Victoria Cave is one of the longest of any cave in Britain and is the most complete record of glacial–interglacial events in northern Britain. Although safeguarded by legislation, the remaining sediments in the cave are largely unprotected from visitor damage and rabbit burrowing, as a consequence they are eroding rapidly and a valuable environmental archive is being destroyed.

Tailpiece

The Craven Dales of North Yorkshire hold a unique record of both Carboniferous and Quaternary landscape changes. Although this area has long held a fascination for geologists, geomorphologists and archaeologists the results of recent studies have shed new light on the environmental changes that combined to give the area its distinctive character. A fuller appreciation of Craven (both above and below ground) has to recognise that: (1) limestone diversity and glacial processes account for the range of pavement types, (2) that the landscape has been shaped by processes that operated during several periods of severely cold climate, with glacial ice cover present, as well as in intervening temperate (interglacial) stages, and (3) that since the last ice sheet disappeared, significant landscape modification has occurred in response to climate changes of relatively small magnitude.