High Temperature

Increased plant temperatures

Reduced net ^^tomata^closure during day

Less evaporative cooling

Increased plant temperatures

Less evaporative cooling

Less production of cell parts and substances

Possible changes in-" production of s, growth regulators photosynthesis

Less assimilation of food

Less food translocated and., stored throughout the tree

Increased respiration

Increased - consumption " of foods

LESS POTENTIAL ^ FOR RAPID CAMBIAL GROWTH (to A)

FIGURE 10.5 A schematic diagram showing how low precipitation and high temperature before the growing season may lead to a narrow tree ring in arid-site trees (Fritts, 1971).

lated with a number of different climatic factors in both the growth season (year tQ) and in the preceding months, as well as with the record of prior growth itself (generally in the preceding growth years, t x and t_2). Indeed in some dendroclimatic reconstructions, tree growth in subsequent years (i+1, t+2, etc.) may also be included as they also contain climatic information about year tQ.

Trees are sampled radially using an increment borer, which removes a core of wood (generally 4-5 mm in diameter) and leaves the tree unharmed. It is important to realize that dendroclimatic studies are unreliable unless an adequate number of samples are recovered; two or three cores should be taken from each tree and at least 20 trees should be sampled at an individual site, though this is not always possible. Eventually, as will be discussed, all of the cores are used to compile a master chronology of ring-width variation for the site and it is this that is used to derive climatic information.

10.2.2 Cross-Dating

For tree-ring data to be used for paleoclimatic studies, it is essential that the age of each ring be precisely known. This is necessary in constructing the master chronology from a site where ring widths from modern trees of similar age are being compared, and equally necessary when matching up sequences of overlapping records from modern and archeological specimens to extend the chronology back in time (Stokes and Smiley, 1968). Great care is needed because occasionally trees will produce false rings or intra-annual growth bands, which may be confused with the actual earlywood/latewood transition (Fig. 10.6). Furthermore, in extreme years some

Increment Borer Diagram

the growing season (earlywood, EW) are large, thin-walled, and less dense, while the cells formed at the end of the season (latewood, LW) are smaller, thick-walled, and more dense. An abrupt change in cell size between the last-formed cells of one ring (LW) and the first-formed cells of the next (EW) marks the boundary between annual rings. Sometimes growing conditions temporarily become severe before the end of the growing season and may lead to the production of thick-walled cells within an annual growth layer (arrows).This may make it difficult to distinguish where the actual growth increment ends, which could lead to errors in dating. Usually these intra-annual bands or false rings can be identified, but where they cannot the problem must be resolved by cross-dating (Fritts, 1976).

the growing season (earlywood, EW) are large, thin-walled, and less dense, while the cells formed at the end of the season (latewood, LW) are smaller, thick-walled, and more dense. An abrupt change in cell size between the last-formed cells of one ring (LW) and the first-formed cells of the next (EW) marks the boundary between annual rings. Sometimes growing conditions temporarily become severe before the end of the growing season and may lead to the production of thick-walled cells within an annual growth layer (arrows).This may make it difficult to distinguish where the actual growth increment ends, which could lead to errors in dating. Usually these intra-annual bands or false rings can be identified, but where they cannot the problem must be resolved by cross-dating (Fritts, 1976).

trees may not produce an annual growth layer at all, or it may be discontinuous around the tree or so thin as to be indistinguishable from adjacent latewood (i.e., a partial or missing ring) (Fig. 10.7). Clearly, such circumstances would create havoc with climatic data correlation and reconstruction, so careful cross-dating of tree ring series is necessary. This involves comparing ring width sequences from each core so that characteristic patterns of ring-width variation (ring-width "signatures") are correctly matched (Fig. 10.8). If a false ring is present, or if a ring is missing, it will thus be immediately apparent (Holmes, 1983). The same procedure can be used with archeological material; the earliest records from living trees are matched or cross-dated with archeological material of the same age, which may, in turn, be matched with older material. This procedure is repeated many times to establish a thoroughly reliable chronology. In the southwestern United States, the ubiquity of beams or logs of wood used in Indian pueblos has enabled chronologies of up to 2000 years to be constructed in this way. In fact, accomplished dendrochronologists can quickly pinpoint the age of a dwelling by comparing the tree-ring sequence in supporting timbers with master chronologies for the area (Robinson, 1976). Similarly, archeologically important chronologies have been established in western Eu-

FIGURE 10.7 Schematic diagram illustrating the potential difficulty presented by the formation of a partial ring (in 1847). In the lowest two sections the ring might not be sampled by an increment borer, which removes only a narrow wood sample. In the upper section the ring is thin, but present all around the tree circumference. Such missing or partially absent rings are identified by careful cross-dating of multiple samples (Glock, 1937).

FIGURE 10.7 Schematic diagram illustrating the potential difficulty presented by the formation of a partial ring (in 1847). In the lowest two sections the ring might not be sampled by an increment borer, which removes only a narrow wood sample. In the upper section the ring is thin, but present all around the tree circumference. Such missing or partially absent rings are identified by careful cross-dating of multiple samples (Glock, 1937).

rope. Hoffsummer (1996) for example, has used beams of wood from buildings in southeastern Belgium to establish an oak chronology extending back to A.D. 672, and in several regions of France chronologies of over 1000 yr in length have been assembled from construction timbers (Lambert et al., 1996). Dendrochronological studies have also been used in studies of important works of art, for example in dating wooden panels used for paintings, furniture, and even the coverboards of early books (Eckstein et al., 1986; Lavier and Lambert, 1996). Finally, tree stumps recovered from alluvial sediments and bogs have been cross-dated to form composite chronologies extending back through the entire Holocene. Long tree-ring series such as these are so accurate that they are used to calibrate the radiocarbon timescale (see Section 3.2.1.5). Tree rings are unique among paleoclimatic proxies in that, through cross-dating of multiple cores, the absolute age of a sample can be established. This attribute distinguishes tree rings from other high resolution proxies (ice cores, varved sediments, corals, banded speleothems, etc.) because in such proxies, comparable replication of records and cross-dating of many samples across multiple sites is rarely possible. Consequently, with the exception of specific marker

0 0

Post a comment