Ivan Gams (Yugoslavia)
Many karst areas, especially in the Mediterranean where forests were natural vegetation are stony lands at present. Historical evidence for the deforestation age is scarce and it is urgent to develop new methods of dating. Our experiences with the use of rillenkarren for deforestation dating explained in the present paper.
Human-made smooth surface of tombstones, monuments, buildings etc. gradually become more and more rough and dotted with punctual or linear indents. They are developed due to euendolithic and endolithic blue-green algae (cyanobacteria), fungi and green algae. Their activity is mostly controlled by cumulative imbibitions time, which mainly depends on humidity and temperature. Rain water solution is also dependent on the exposition of the surface to rain bringing wind. Among the linear forms of solution rillenkarren (solution flutes) are the most known ones. Analysing them, LUNDBERG (1977) found no relation between their cross-section¡¢ depth¡¢ lenth and slope. DUNKLEY (1979) found little relationship between their length and slope, and GLOE and FORD (1980) established direct relationship between slope up to 60 degrees and length of the flutes, simulated in plaster of Paris (gypsum) in laboratorium.
Stone face in finegrained and homogene limestone is smooth also due to natural solution processes in the soil cover-bedrock interface. This is widely known. But the smooth stone face originating in subsoil is attributed, as for present karst literature, only to the cases of minor karst forms, as for example rundkarren. JENNINGS described the subsoil formation of cavernous subsoil weathering forms (kavernose karren), solution pits (geologische Orgeln), partially also solution notches, swamp slots and solution pans (kamenica). In the karst monographs of FORD-WILLIAMS (1989, 380-386) the subsoil formations are recognized as subsoil pit, subsoil clint-and-grike, cutters, rundkarren, soil pinnacles and as pinnacle karst and ruiniform karst.
Analyses of subsoil forms in the Dinaric Karst brought evidence on two differently orientated solution processes. The first one, in subsoil karst, with solution of humus and soil moisture makes the bedrock face smooth. The second one ruins the smooth face of the exposed stone due to mechanical weathering, algae, fungi and rain water solution. The subsoil origin of smooth surface of the outcropping stones is in many cases proved by the appearence of subsoil forms such as solution pipe, scallop, recession, bowl-like solution pan, rounded meanderkarren. On the generally smooth underground rock faces of compact homogene finegrained limestone and dolomitic limestone places can be found where soil is stripped off, e.g. behind quarries, new building, roads. Height of the outcropping stones with smooth faces can be used to estimate soil erosion. But there are some exceptions. Smooth stone face can also be generated below snow cover if lasting more than half a year. Dead leaves and litter in dense forests can cover the massive stones rising high above ground: below this material, especially if based on humus sheet, smooth stone face occurs as a natural phenomenon. In both cases the height of outcropping stone is no longer an �indicator of general soil level lowering but of the removing of the humus cap from the stone top.
The thesis on subsoil (subhumus) origin of smooth face of outcropping stones has led to the use of rillenkarren for dating the deforestation age. Nevertheless, the human or natural origin of smooth stone face, density and depth of its indents is a function of passed time after being exposed to all sorts of weathering in the open air. The idea was first proved valid on the northern side of the Hvar Island.
METHODOLOGY OF ANALYSES
As for morphometric elements, rillenkarren’s width, depth, length, cross-section, relation to slope degree were mostly analysed till now . We have established their depth as the most suitbable element of dating. But there is a question, in which point of their length it has to be measured. The depth of a channel is diminishing downwards until the rillenkarren face passes over to planar belt of non-channelled erosion. Locally steeper slope can modify this rule. Table 1 shows the histogramme (series of downward cross-sections) of one typical example of rillenkarren taken on a 1.3 m high outcropping stone built of gray Cretaceous finegrained homogene limestone. This example of rillenkarren begins in a ripple between two border channels. Measurements were performed along the road from Sezana to Lipica in the Kras Region in Slovenia, Yugoslavia. Depth and width of the channel are increasing fast up to 10 cm distance from the beginning of the channel, where they achieve their maximal values. Further on downward width is still progressing and depth is diminishing. There is an exception between the distance of 50 and 80 cm on a steeper part of the stone surface. At the beginning of the steeper part (at 50 cm) the depth is still diminishing but this is compensated by the next exceptionally deep cross-section. Presumably, this is the result of the more rapid flow above which only here begins the additional solution. After that, depth is diminishing again despite of still persisting steep slope.
Numerical data from the histogramme are listed in Table 1. The areas of cross-sections which are the channel volumes at the same time, are also cited in it. By dividing them with the width value of the profile, we get thickness of soluted sheet in mm. Relation between width and depth is augmenting up to 30 cm of distance and after that, it is diminishing (with the exception of the steeper sector).
The transformation of a series of rillenkarren on the 12-13 cm long profile is also shown at the end of Table 1. After the beginning the depth diminishes also here. Later, deeper and wider channels nearly supplant the middle channel and finally (distance 64 cm) absorb it near the transition to the belt of non-channelled erosion. Steeper slope locally modifies the normal transformation also here.
If we take the deepest channel at its beginning (5-10 cm in our cases). We approach the maximal thickness of solution sheet. At this distance the ripples are usually still evenly high and corresponding to the natural stone face.
But in line with the deepest channels in the middle. We usually encounter some minor rillenkarren on the border side and also in between them. An average depth along the line can be two times smaller than the deepest channels. We chose the deepest series for the sake of opinion that the minor channels were hindered in their development by lithological anisotropy, differences in water quantity, border channels supplanting, later removal of soil or humus cover etc.
The Carpenter’s Profile Gauge was used for measuring. South expositions (SW, SE) were favoured in the forest of the Kras Region, northern sides are often more covered with moss. Under the patches of moss the tops of ripples have already been lowered. According to these field observations, moss cover can not smooth out the stone face. The face under the moss is dotted with micropits and micropans.
Table 1 Histogramme of one single example of rillenkarren
With analyses of smooth stone faces we joined the international archaeological investigation in the northern part of the Hvar Island. The low karst plain Velo polje spreads among the bays of the settlements Stari Grad. Vrboska and Jelsa in the altitude 0-35 m. Stone walls bordering the parcels or being inside them are dense. According to the regular size of the parcels, some archaeologists support the idea of this being old Greek parcelling of the fields and erecting of stone walls. The Greek colony in Stari Grad (Pharos) lasted from 385 to 219 B.C. But the smooth faces of the stones in these walls have no signs of rain water solution that would have originated in open air. The opinion about the young age of the walls is also supported by the history of agriculture.Inland convenction of warm air in summer attracts cooler and humid air from the nearby sea. Therefore on Hvar, as elswhere on the Yugoslav plains near the Adriatic Sea., viniculture had not been possible until the beginning of this century, when vine diseases could be suppressed by chemicals. After that vineyards began to be moved from higher slopes to lowlands. At ploughing for cereals in the Middle Ages and planting of fruits (olives, fig-trees) later on stones were removed down to 15-30 cm of soil depth and for vineyards down to 50-80 cm deep. That’s why the stone walls became higher, wider and denser. In the stone walls along the 1830 m long profile through Velo polje we found so many stone that 236 kg of stones in average must have been removed from each square meter of land.
Rillenkarren are an exeptional phenomenon on the rare outcropping stones in Velo polje and their depth is in span from 0.9 to 2.5 mm. Their number being scare, we are not allowed to calculate the average depth. They are absent not only from dolomite and limestone breccia but also from homogene limestone.
On the slope bordering the north side of Velo polje and on the rocky terrace rising up to 90 m above the karst plain, the soil is far shallower and discontinuous, bushes and forest much denser. We found more rillenkarren there. Their average depth is 6.13 mm.
The second locality of our measurements is the already mentioned Kras between Sezana and Lipica. In the mixed forest there, there are several hundreds of outcropping stones of the height up to 2 m. There are thousands of rillenkarren on them, 41 measured samples gave their average depth of 6.2 mm.
The nearly equal depth of rillenkarren on the plain of Rudine and on the Kras is astonishing when we take into account the differences in precipitation and temperature. This can not be explained by facts, observed at simulation of rillenkarren in plastics which consists of unchanged lowering of channelled stone surface after their full development. The level tops of ripples between the channels testify in our cases against this thesis.
We can estimate the age of 6.13 and 6.2 mm deep channels when comparing them with the solution measurements on human-made vertical walls. KUPPER established the average increase of indents depths on the human-made vertical walls of the tombstones in Belgium to 2.5 mm in one century. According to this thesis our rillenkarren could be 24-25 centuries old. According to JONES_RACHAM WILLIAMS, 1984, the 13 mm deep indents developed on the vertical walls of dolomitic limestone after the Etruscan and the late Etruscan periods (4th cent. B.C.), Indents from the late Republican and the Imperial Roman periods (1st cent, B.C.-6th cent.A.D) are 5.5 mm and those from the Middle Ages (5-16th cent.A.D.) are 2mm deep. If compared with the depth of indents from the Roman period, our 6.13 to 6.2mm deep rillenkarren could be 21 centuries old. At that time the Hvar Island was a Roman colony. But the solution on the vertical walls is weaker than the solution on the inclined walls, not even to mention the horizontal surfaces (solution on the latter measured also with micro erosion meters, is not compared here as it is much faster). DANIN (1983) published a diagramme showing the relation between indent depths mostly produced by cyanobacteria on the man-made smooth stones in Israel. According to this diagramme, our 6.13 to 6.2 mm deep indents are from the 5th to 6th century B.C. Measurements of solution by means of standard limestone tablets hung in the air resulted in a sheet solution of 0.29 mm in one century in the Submediterranean climate. Our 6.13 to 6.2 mm deep rillenkarren on Hvar and on the Kras could be according to this 21 centuries old.
According to the measurements mentioned above, our 6.1-6.2 mm deep rillenkarren are 21-25 centuries old, most probably from the middle of the first millenium B.C. At that time the Dinaric Karst was inhabited by Illyrians. They fought for the land against the Greek colonists. They are known in the history as cattle-breeders who burnt forests to obtain pastures. Illyrians were first people who densely populated the karst area. In the distance of 3 km from the place of our measurements, between Sezana and Lipica, there are ruins of the Illyrian sconce on the Tabor hill.
DISCUSSION OF THE RESULTS
After the first Illyrian deforestation of the Dinaric Karst, many deforestation and forestation epochs have succeeded. The question arises as to which deforestation epoch we can attribute the beginning of our rillenkarren development. According to the age ascribed to them above, the first successive burnings of natural forests for maintaining the pastures were the most fatal for soil and humus (litter) cover of the bedrock. Dead leaves, litter and humus soil horizont beeing accumulated in the 7 millenia of Holocene also burnt togetherwith trees. After that, the rest of them have weathered in the direct sunshine or have been washed away by downpowers.
During the cattle-breeding period of agriculture in the inland of the Dinaric Karst which lasted till the recent century, the soil level was continuously lowered. This can be proved by the measurements of rillenkarren depths on the lower outcropping stones. We observed shallower channels there but we measured them less systematically on two localities only: (Tab.2)
On the low stones local conditions can substantially change the velocity of rillenkarren development.
Although most of the parts of the Dinaric Karst have gone through the siminar agricultural history, the rillenkarren appear only in some regions and even there, mostly on some stones only. They are nearly absent in the old fields with deeper soil which can be explained by progressive erosion of soil, originally covering nearly all bedrocks. The question about rare occurence of the rillenkarren can not be answered and explained only by the factors that we are familar with at present.
George A.Brook, David A. Burney, James B.Cowart, and Brooks B.Ellwood (USA)
Pollen spectra from speleothems in desert caves indicate that these deposits may be a valuable new source of paleovegetation data for the arid and semiarid regions of the world. In other dryland sediments pollen is often absent or poorly preserved because of the highly oxidizing conditions. Speleothem age and pollen data for Somalia, Zaire, and the U.S.A. indicate that East Africa was more arid than today during glacial maxima of the last 300,000 years and relatively moist during interglacials. By contrast, the American southwest was much wetter during glacials and relatively arid during interglacials. Only during interstadials was water more plentiful in both regions either due to lower temperatures or to increased precipitation.
Evidence of past environmental conditions in the world’s dry areas is often limited by the poor preservation of pollen grains and other organic matter. This severely restricts efforts to reconstruct past vegetation characteristics and to provide a 14C chronology for the data. As a result, desert paleoenvironments arein general poorly known and where there is information it rarely extends beyond about 30,000 B.P., the range of the 14C dating method. Lake-level studies are,for example, largely limited to this time range. These have shown that in East Africa the last glacial maximum was substantially drier than today and the Holocene, particularly the early Holocene, substantially wetter. The situation in the American southwest was apparently very different; lakes were highest here during the last glacial maximum and low or dry during the Holocene.
The recent discovery by Bastin and Brook et al. that cave speleothems may contain pollen raises the question as to whether these deposits–which can be 230Th/234U dated to ca 350,000 B.P. –could be a source of desert paleoenvironmental data to well beyond the range of 14C dating. In arid and semiarid areas speleothem ages alone can provide information on past wetter phases of climate as deposition is modest or absent today in extremely dry caves.
This paper reports on research conducted in Somalia, Zaire, and the U.S.A. between 1982 and 1987. The research attempts to determine 1) under what conditions desert speleothems (and cave speleothems in general) contain pollen, 2) to what extent the fossil pollen provides reliable data on past vegetation in deserts, and 3) whether pollen is preserved in very old formation–thus providing a window on the more distant past. In addition, speleothem ages are examined to establish if the pattern of glacial aridity and interglacial humidity in East Africa and the reverse in the southwestern U.S.A. (as revealed by lake-level studies) was typical of the last 300,000 years not just the last 30,000 years.
STUDY AREAS AND METHODS
There are numerous caves in and close to the North American Mojave, Sonoran, and Chihuahuan deserts. Carlsbad Cavern, in the northern Chihuahuan Desert at 1,325 m in the Guadalupe Mountains, was selected for study. Mean annual precipitation at the cave is 362 mm and the mean annual temperature is 17.2° C. Using an electric drilling rig three 4.5 cm diameter cores up to 3.2 m long were recovered from the Georgia Giant and Texas Toothpick stalagmites in the Green Lake Room and in the Lower Cave respectively.
There are extensive areas underlain by soluble rocks in the Somali-Chalbi Desert but the caves of the region are poorly known. Sediments from two caves in northern Somalia were examined in this study. Galweda Cave in the hyperarid coastal desert zone south of the Gulf of Aden is at 20 m elevation and presently receives less than 50 mm of precipitation annually, mean annual temperature is 30° C. Hayla Cave in the Golis Mountains immediately to the south is at 1,800 m with annual precipitation and temperature being 435 mm and 17¡æ respectively.
Work in the Ituri rainforest of northeastern Zaire was conducted principally at Matupi Cave in the Mount Hoyo block southwest of Lake Mobutu Sese Seko. The cave is at 1,100 m elevation on the western edge of the Western Rift Valley. Today, vegetation near the cave is species-rich equatorial rainforest. The mean annual temperature is 23° C and annual precipitation is about 1,687 mm.
The core drilled from the Georgia Giant stalagmite in Carlsbad Cavern was dated at regular intervals providing 46 ages. 230Th/234U ages were also obtained for the outside layers of the Texas Toothpick formation that were within the range of the dating method. 230Th/234U ages were also obtained for six speleothems from Galweda Cave, nine from Hayla Cave and five from Matupi Cave.
Speleothems from Carlsbad, Hayla, and Matupi Cave were also examined to determine if they contained pollen. Ten samples of more than 100 g were cut from selected locations along the Georgia Giant core and five samples from the Texas Toothpick core. Eight samples of 22-112 g were cut from seven Hayla Cave speleothems, and sixteen samples of 9-33 g were cut from seven Matupi Cave formations. Any pollen present was extracted using methods described in Brook et al. (1990).
THE SPELEOTHEM AGE DATA
Under the present climatic conditions there is no speleothem deposition in Galweda Cave (rainfall < 50 mm) and minimal growth in Carlsbad (precipitation 363 mm) and Hayla Cave (precipitation 435 mm). With a precipitation close to 1,700 mm per year there is extensive deposition at Matupi Cave. The ages of speleothems from the northern Somali and southwestern U.S.A. caves can therefore provide information on past wetter climatic conditions when speleothem deposition was more rapid and more extensive than today. The ages of Matupi Cave speleothems indicate only that at the time of their deposition conditions at the cave were still relatively wet–as they are today (precipitation is 1,687 mm). Age data for the Georgia Giant and Texas Toothpick cores, together with speleothem ages for Ogle Cave and Carlsbad Cavern from Hill (1987) and Harmon and Curl (1978) are shown in Fig.1. Periods of speleothem deposition in northern Somalia are shown in the same figure for comparison.
It is clear from Fig.1 that maximum rates of speleothem accumulation were at very different times in the American southwest and East Africa. In North America deposition was most extensive during glacial and interstadial periods of the last 300,000 years, particularly during isotope stage 6 (the Illinoisian or Riss Glaciation). The warmer phases of interglacials were apparently much drier. This was particularly so during isotope stages 1 and 5e when there were hiatuses in the growth of the Georgia Giant. Isotope stage 5e appears to have been particularly dry as two other speleothems from Carlsbad and one from Ogle Cave also stopped growing at this time.
In contrast to the pattern in the southwestern U.S.A. there is no evidence of extensive speleothem growth in the Somali-Chalbi Desert during full glacial periods. In this region deposition was apparently most rapid during interglacials and to a lesser extent during interstadials. Particularly noticeable is that when deposition of the Georgia Giant speleothem ceased at the beginning of isotope stages 5 and 1, there was increased speleothem formation in northern Somalia.
Speleothem age and pollen data for the Chihuahuan and Somali-Chalbi deserts, and for the Ituri rainforest suggest that East Africa was much more arid than today during the last glacial maximum and probably also during other glacial maxima of the last 300,000 years. Interstadials appear to have been cooler and wetter than today, the increased availability of moisture being due either to reduced evapotranspiration or to a slight increase in precipitation. The wettest periods of the past were the times of greatest global warmth (peaks in the marine d 18O curve) when there is strong evidence of substantially increased monsoonal precipitation in the Horn of Africa. By contrast, the American southwest was wetter than today during glacial maxima and at least as dry as today during interglacial maxima. Only during interstadial times were the deserts in both North America and East Africa wetter than they are today and it is unresolved whether this was because of lower temperatures or increased precipitation.
Fig.1. Phases of speleothem deposition in East Africa and the American southwest during the last 300,000 years and arid phases indicated by pollen data.
The marine record is from Martinson et al. (1987)
Pollen spectra from actively growing speleothems in Hayla and Matupi Cave were found to reflect the present vegetation near these two caves implying that fossil pollen from speleothems can be used to reconstruct past vegetation characteristics. The recoverry of pollen from one Somali speleothem deposited 176,000 B.P. suggests that speleothems may be able to provide vegetation data for a considerable part of the Quaternary period. The absence of pollen in the two Carlsbad speleothems can probably be attributed to the great distance between them and the cave entrance and to their great depth beneath the surface. The Georgia Giant is 215 m below the cave entrance and the Texas Toothpick 245 m below it. The Hayla and Matupi Cave speleothems that contained pollen were growing fairly close to natural entrances and in passages close to the surface. In our view most of the pollen deposited on speleothems is airborne pollen and/or pollen introduced into the cave by animals, particularly bats. Modest numbers of pollen grains may be introduced by groundwaters under favorable conditions. Being so far from the natural cave entrance and at so great a depth, neither airbone nor waterborne pollen was apparently able to reach the Georgia Giant and Texas Toothpick stalagmites in Carlsbad Cavern.
In conclusion, cave speleothems in arid and semiarid regions appear to be a largely untapped yet extremely viable source of paleoenvironmental data. They may be able to provide important new information about the impact of man on his environment. In Africa the “Cradle of Mankind” they may also provide information on possible relationships between natural environmental changes and man’s evolution. In the last 200,000-300,000 years, a period within the range of 230Th/234U dating, man has evolved from Australopithecus through Homo erectus to Homo sapiens –modern humans.
(TO BE CONTINUED)