Tree Ring Width as an Index of Physiological Dryness in New England

Abstract

The use of tree ring width as a measure of relative secondary growth rates and growing conditions has received little attention in New England or any other humid section of the United States. Analyses covering short periods of time, dealing with a few trees only and often with several species in the same study, have led to contradictory and poorly supported conclusions, in sharp contrast with the extensive data and dependable conclusions from studies ill the southwestern states. Since the humid regions lack any such obvious factor for control of growth as the long droughts of relatively dry regions, it has remained an open question whether rainfall or any other factor, or simple set of factors, controls the growth rates of trees in wellwatered areas. This pl)aer l)resents the results of an intensive study of the relations of growth to climate shown by the eastern hemlock, Tsuga caneadonsis, in northern New England. The critical use of annual rings as measures of radial growth rates in response to environment requires (a) measurements sufficient in number and accuracy really to represent growth, and (b) use of the data for each tree with constant appreciation that they involve responses of a living organism to many factors internal and external, even though one or more factors may dominate. This means especially that more pronounced effects should be expected when the dominant factor is quantitatively much below normal than when it offers a relatively favorable opportunity for growth; for under these most favorable conditions the full and precise response of the tree may be prevented by a deficiency of even one other essential factor. Likewise, so much allowance must be male for interplay of factors when the dominant one is quantitatively not far from its average value, that the practical approach to the problem is to notice the responses to extreme values of the factor suspected of being outstandling in its control of growth rate. It is understood that such control may be either direct or indirect or 1)oth, in view of the many interrelated factors of temperature, evaporation, distributionn of rainfall from day to clay, soil texture and profile, root distribution, interception of rainfall by tree crowns, etc. Failure to allow for contributing factors like these, which all workers must admit as important, has led some writers to expect or require nearly perfect correlations under all conditions. Thus, Adams ('35, p. 22) concludes that Robbins ('21) 457