Trees as Climatic Indicators

TREES AS CLIMATIC INDICATORS Harry P. Bailey University of California, Los Angeles The climatic record afforded by tree growth has been discussed in geographic literature in the United States over a period of 30 years. The discussion began with Ellsworth Huntington, who was interested in tree growth as a support for his thesis of climatic change (1). Later studies, published by the American Geographical Society and the National Geographic Society, have applied the data of tree growth to the dating of Indian ruins in the Southwest and to the formulation of rainfall indices for this and other areas. Because the names of several authors are associated with the literature of tree growth, the casual reader is not always aware of similarities in the techniques used in the studies it reports. The impression is easily gained that each study is an independent confirmation of the climatic significance of trees. Instead, it is the technique evolved by A. E. Douglass , of the University of Arizona, that has been followed in every instance ; indeed, nearly all the laboratory work, including the final preparation of graphs showing rainfall indices, has been done by Douglass and his staff. This statement applies not only to geographic writings, but also to nearly all tree-ring studies published in the United States. It is evident that additional perspective is required if the climatic significance of trees is to be defined accurately. Such perspective may be gained in part from pertinent studies in plant ecology; the relation between tree growth and environment is so poorly understood, however, that such studies do not provide a basis for conclusive judgment. Part of the difficulty of applying botanical evidence to the results of tree-ring analysis arises from the fact that botanists and foresters have usually studied the normal or representative habitat, whereas the trees found most useful in tree-ring analysis occupy distinctly non-representative sites. Additional checks may be supplied by other forms of historical data, such as météorologie observations, records of fluctations of lake levels, and descriptive literature. Such data are most abundant from the last century; they are almost completely lacking from earlier times. Archéologie stratigraphy provides some checks on the prehistoric record, but by defining sequences rather than by establishing precise dates. Most of the record of tree growth, therefore, lies beyond the range of a scientific check. It follows that only meager tools are available for evaluating it. This fact undoubtedly explains the controversial nature of the critiques of tree-ring analysis. I shall attempt here to discuss only those aspects of the subject that pertain to the final result of tree-ring study, namely the climatic record. Considerations of minor importance will be ignored. Since I approach the question from the field of climatology, my knowledge of tree-ring analysis is based essentially upon published materials rather than on personal experience. The publication of views thus arrived at seems justified by the fact that geographic literature contains little commentary on the climatologie value of tree-ring records. The following paragraphs give a summary of A. E. Douglass' accomplishments written by one of his most active disciples : 24Yearbook of the AssociationVol. 9 "When A. E. Douglass began in 1901 his classic studies in tree-ring analysis he was seeking to read and interpret what the trees had been able to record of solar and climatic changes. It was soon established that where there is a dominating climatic element which conditions growth, properly located trees respond faithfully to the changes in this element. Thus, in the arid southwestern United States, the relation between annual treegrowth and rainfall is very close, the correlation coefficient for the last century averaging 0.80. In the last two decades, studies have been chiefly directed along two paths: (1) the development of long climatic chronologies in tree growth, and (2) the analysis of these for long-period climatic fluctuations. "Two major chronologies have now been established: (1) a continuous annual record almost 3,250 years in length derived from the long-lived Sequoia gigantea of east-central California, and (2) a record extending back to 11 A.D. representative to an especially high degree of rainfall fluctuations during this...