Criticai, Study of the Cobalt Chloride Method of Measuring Transpiration1

In recent years increasing awareness of the importance of water in our national economy has focused attention on the significance of forests in this important phase of conservation of our natural resources. Much has been written in the past 30 years concerning various aspects of the water relations of woody plants, yet available data concerning total water loss from trees and forests are inadequate. This inadequacy is apparent when it is realized that many of our present-day estimates are based on experiments conducted by voN H6HNEL prior to 1900 in which potted seedlings were used as test plants (6, 7). Such estimates have little value for appraising seasonal use of water by forests. The present study was undertaken to supply needed information concerning a measurement procedure which could be used with facility under field conditions to give reliable estimates of transpiration by forests of the Tennessee Valley. Methods of measuring transpiration have been summarized from the literature (5, 10) and need not be described here. All of these methods are subject to criticism either on theoretical grounds or on the basis of operational difficulties. Most of them are impractical under field conditions, especially when applied to tree species. This is illustrated in the studies of MINCKLER (11) in which the water loss of 12 trees of five species was determined by an adaptation of the air-stream method. This procedure, one of the most acceptable under laboratory conditions, is quite cumbersome in application to field conditions. The cobalt chloride procedure, first proposed by Stahl, has particular appeal for field studies since it requires no elaborate equipment and determinations can be made simply and quickly (4). Numerous studies of plant transpiration have been made with this procedure (4, 8, 13). MEYER (9) and BLAYDES (1, 2, 3) have brought out a number of practical possibilities and limitations of the method and have emphasized that it does not measure absolute rates of transpiration that occur normally, but that the rates determined depend on certain abnormal conditions which may be imposed on the transpiring leaf surface by the cobalt chloride test paper. They considered that by careful standardization of the test paper the relative rates of