390 DENDROCHRONOLOGY Figure A comparison of the tree ring reconstructed and instrumentally recorded precipitation for the 1905 1987 calibration period (Reproduced with permission from Case RA and MacDonald GM (1995) A dendroclimatic reconstruction of annual precipitation on the western Canadian prairies since AD 1505 from Pinus flexilis James Quaternary Research 44: 267 275, copyright Elsevier 1995.) Figure A 487 year (AD 1505 1992) reconstruction of August July annual precipitation Drought years are indicated on the lower portion of the graph (ỵ) and are defined as years when precipitation is more than one standard deviation less than the average of the complete series The heavy line represents an 11 year low pass filter (Reproduced with permission from Case RA and MacDonald GM (1995) A dendroclimatic reconstruction of annual precipitation on the Western Canadian prairies since AD 1505 from Pinus flexilis James Quaternary Research 44: 267 275, copyright Elsevier 1995.) of course the latter approach is only possible where the data are available Many dendroclimatic studies have concerned themselves with reconstructing one particular factor in a limited geographical area, for example summer temperature in a particular country or state Another approach, which has been invaluable, is to produce a network of chronologies over national, continental or hemispherical ranges and to use these as the basis of more wide-scale reconstructions This approach enables an assessment of the reconstruction values at the intermediate points to be made, and can identify areas within the network where further sampling is desirable More ecologically based studies have shown changes in tree response that are elevation dependent (Figure 2) Micro-Anatomical Variations Whole ring, or earlywood/latewood width studies are the commonest form of dendroclimatological studies In parallel with these studies however, there is a long history of studies in the relationship of intraring features with external growing conditions Cell size, cell differentiation, cell distribution and other characteristics (e.g wall thickness) have all been studied Stable Isotope Studies This discussion has so far concentrated on the use of ring-width studies, but it is important to remember that dendrochronological studies (or tree-ring studies if one prefers) can and have used many other aspects