Paleoclimatology Global Warming

j_ Number of moraines

CD CD

IT 3

Years before A.D. 1972

FIGURE 4.16 Growth curve of Rhizocarpon geographicum and Rh. alpicola for the last four centuries based on measurements in the Sarek and Kebnekaise mountains, Swedish Lapland. Bar graph at right shows frequency of moraines as characterized by the appropriate maximum thallus diameters. Several discrete moraine groups are evident (Denton and Karlen, 1973b).

(b) Environmental Factors

Lichen growth is dependent on substrate type (particularly surface texture) and chemical composition (Porter, 1981c). Rocks that weather easily, or are friable, may not remain stable long enough for a slow-growing lichen to reach maturity. Conversely, extremely smooth rock surfaces may preclude lichen colonization for centuries and possibly many never support lichens. Extremely calcareous rocks may also inhibit growth of certain lichens. Measurements should thus be restricted to lichens growing on similar lithologies whenever possible.

Climate is a major factor affecting lichen growth rates; comparison of growth rates from different areas suggests that slower growth rates are found in areas of low temperature, short growing seasons, and low precipitation (Fig. 4.16). However, both macro- and microclimatic factors are of significance. In particular, lichens require moisture for growth and the frequency of small precipitation amounts, even from fog and dew, may be of more significance than annual precipitation totals. Radiation receipts are also important because they largely determine rock temperatures. Generally it is impossible to equate such factors on those rocks used to calibrate the lichen growth curve with rocks that are eventually to be dated. Commonly, calibration will be carried out on buildings or gravestones in a valley bottom, whereas the features to be dated are hundreds of meters higher than the calibration site. Similar problems may be encountered along extensive fjord systems where conditions at the fjord mouth are less continental than at the fjord head. Lichen growth is far slower in the more continental locations, even over distances as short as 50 km, probably due to the lower frequency of coastal fogs and generally drier climates inland. Increasing elevation also appears to be significant in reducing growth rates, even though moisture availability might be expected to increase (Miller, 1973; Porter, 1981c). Presumably, this is offset by longer-lying snow and a reduced growing season due to lower temperatures (Flock, 1978). All these factors may complicate the construction of a simple growth curve for a limited geographical area. Further problems arise due to the possible influence of long-term climatic fluctuations. Apart from the general effect of lower temperatures in the past, it is quite probable that in the high elevation and/or high latitude sites where lichenometry is most widely used, periods of cooler climate resulted in the persistence of snow banks, which would have reduced lichen growth rates (Koerner, 1980; Benedict, 1993). Growth curves may thus not be linear, but rather made up of periods of reduced growth separated by periods of more rapid growth (Curry, 1969). Lack of resolution in calibration curves may obliterate such variations, but this could account for apparent "scatter" in some attempts at calibration. There is evidence that such factors have been of significance in some regions; on upland areas of Baffin Island, for example, persistence of snow cover during the Little Ice Age is thought to have resulted in "lichen-free zones," where lichen growth was either prevented altogether or severely reduced (Locke and Locke, 1977). These zones can be recognized today, even on satellite photographs, by the reduced lichen cover of the rocky substrate compared to lower elevations where snow cover was only seasonal (Andrews et al., 1976). Similarly, attempts to date moraines that have periodically been covered by snow for long intervals would give erroneously young ages for the deposits (Karlen, 1979).

(c) Sampling Factors

It is of fundamental importance in lichenometric studies that the investigator locates the largest lichen on the substrate in question, but this is not always something one can be certain of doing (Locke et al., 1979). Furthermore, very large lichens are often not circular and may sometimes be mistaken for two individuals that have grown together into one seemingly large and old thallus. It is also possible that a newly formed moraine may incorporate debris from rockfalls or from older glacial deposits; if such debris already supports lichens, and if they survive the disturbance, the deposit would appear to be older than it actually is (Jochimsen, 1973). A number of innovative methods to improve the reliability of lichenometry have been proposed (McCarroll, 1994) and these generally provide a firmer statistical basis for the sampling procedure.

Finally, in establishing a calibrated growth curve for lichens, reference points at the "older" end of the scale are often obtained from a radiocarbon date on organic material overridden by a moraine. This date is then equated with the maximum-sized lichen growing on the moraine today. Such an approach can lead to considerable uncertainty in the growth curve. First, dates on organic material in soils overridden by ice may be very difficult to interpret (Matthews, 1980). Secondly, there may be a gap of several hundred years between the time organic material is overridden by a glacial advance and the time the morainic debris becomes sufficiently stable for lichen growth to take place. This would lead to overestimation of lichen age in a calibration curve. Thirdly, 14C dates must be converted to calendar years, which often results in very large margins of error, especially for the most recent Little Ice Age period, because of the nonlinearity between 14C and calendar ages in this interval. This only amplifies the uncertainties associated with lichen growth curves such as those shown in Fig. 4.14. In reality, each curve should probably have an error bar of around 15-20%, perhaps even more for growth curves that are extrapolated beyond a dated control point (Bickerton and Mathews, 1992; Beget, 1994).

A consideration of all these factors indicates that caution is needed in using lichenometry as a dating method, even for establishing relative age. Nevertheless, if consideration is given to the possible pitfalls, it can provide useful age estimates. Most problems would result in only minimum-age estimates on substrate stability, but in some cases, overestimation of age could result. Although it is worth being aware of the potential difficulties of the method, it is unreasonable to expect that all of the problems, discussed already, would subvert the basic assumptions of the method all of the time, and often the potential errors can be eliminated in various ways. Lichenometry is thus likely to continue to play a role in dating rocky deposits in arctic and alpine areas, and hence make an important contribution to paleocli-matic studies in these regions.

4.3.2 Dendrochronology

Although not a widely used dating technique in paleoclimatology, the use of tree rings for dating environmental changes has proved useful in some cases (Luck-man, 1994). The concepts and methods used in dendrochronological studies are discussed in more detail in Chapter 10 and need not be repeated here. Basically, dendrochronological studies are used in three ways: (a) to provide a minimum date for the substrate on which the tree is growing (e.g., an avalanche track or deglaciated surface); (b) to date an event that disrupted tree growth but did not terminate it; and (c) to date the time of tree growth, which was terminated by a glacier advance, or a climatic deterioration associated with a glacier advance (Luckman, 1988, 1995).

The first application is straightforward but to obtain a close minimum date assumes that the "new" surface is colonized very rapidly. This is highly probable in the case of avalanche zones (indeed, young saplings may survive the event) but in deglaciated areas surface instability due to subsurface ice melting and inadequate soil structure may delay colonization for several decades. Different approaches towards estimating the time delay before colonization of recently deglacierized terrain are described by Sigafoos and Hendricks (1961) and McCarthy and Luckman (1993). Unlike lichens, which may live for thousands of years, trees found on moraines are rarely more than a few hundred years old. Even when very old trees are located and dated, it cannot be assumed that they represent first generation growth. For example, Burbank (1981) found that a moraine on which the oldest tree was -750 yr old (dated by dendrochronological methods) was in fact older than 2500 yr B.P. according to the local tephrochronology.

A more widespread application of dendrochronology involves the study of growth disturbance in trees. When trees are tilted during their development they respond by producing compression or reaction wood on the lower side of the tree in order to restore their natural stance. This causes rings to form eccentrically after the event that tilted the tree; the event can be accurately dated by identifying the year when growth changes from concentric to eccentric (Burrows and Burrows, 1976). Such techniques have been used in dating the former occurrence of avalanches (Potter, 1969; Carrara, 1979) and hurricanes (Pillow, 1931) and the timing of glacier recession (Lawrence, 1950). They have also been successfully applied in more strictly geomorphological applications, in studying stream erosion rates and soil movements on permafrost (Shroder, 1980).

Valuable insights into the history of glacier advances in the Canadian Rockies have been obtained by Luckman (1996) using tree snags (partially eroded or damaged trees) or tree stumps, from areas exposed by receding glaciers. By cross-dating these samples, evidence of formerly more extensive forest in areas only recently deglacierized has been documented, and the timing of climatic deterioration and glacier advance has been clearly revealed (Fig. 4.17).

Last year of record

1000 1100 1200

Sample Glacier Erosion The Rockies
0 0

Post a comment