Abstract: Tree rings record the growth histories of trees. The means by which researchers extract precise characteristics of tree rings are fundamental to the related fields. The objective of this study was to measure treering widths with statistical analysis software R. This method provides custom R functions to automatically or manually mark treering boundaries on digital images based on edge detection algorithm, and calculates treering widths using coordinates of successive boundary points. Width of inclined treering is corrected with R based on trigonometric functions. Custom R functions offer the ability to compare with multiple sample images to mark rings, handle broken samples and quickly reconstruct previous results. The results showed that this method allowed accurate measurement of treering widths. Tree ring widths calculated by this method was in good agreement with results measured with WinDENDRO. This method is simple and reliable. Digital images stored in the hard disk are convenient to maintain for long periods; R is open source software, and much cheaper than commercial tools like LinTab and WinDENDRO. In addition, R provides a wide variety of statistical and graphical functions, and can be easily extended with R packages. Thus, data analyses of treering widths could run in the same software without outputting.

Key words: land use/cover change, time series, the Three Gorges Reservoir Area, remote sensing, human activities