Dendrochronology, or the study of tree ring growth patterns to date the age of archeological timbers, was initially developed in the 1920’s by Andrew E. Douglass using long-lived Ponderosa pines in the Southwest United States. An astronomer by training, Douglass was interested in historical sun spot activity and its relationship to earth’s climate. He surmised that by looking at yearly growth ring sequences in long-lived trees growing in an arid environment where moisture is key, he might be able to ascertain yearly variations in climate attributable to sunspot activity. (Baillie, 1982). To push the tree ring database back past the age of living trees, samples were taken from roof poles in Pueblo ruins which turned out to eventually overlap the living tree data. Besides fulfilling his research needs, this work revealed the feasibility of dating archeological structures.
In the 1980’s the advent of computer programs to collate the data and compile master chronologies enabled unknown samples to be compared to known masters with a high degree of accuracy. Recent work in Eastern Massachusetts focusing on Oak (Krusic and Cook 2001, Miles, Worthington and Grady 2002, 2003, 2005) and in the Connecticut River valley initially concentrating on Pitch pine (Flynt 2004) and expanding into oak, chestnut, and hemlock, has revealed the suitability of using dendrochronology as a mainstream research tool for analyzing and establishing construction timber felling dates in New England, a region heretofore considered too variable climatically to provide reliable results.
To aid with this study, a variety of dated master chronologies are available. The main Pitch pine master for the Deerfield-Montague area is composed of over 250 samples spanning the years 1526 to 1992. Several Connecticut River Valley Oak masters have also being compiled with 49 samples covering the years 1610-1846 included in one for Deerfield while a more general master for the Massachusetts portion of the valley consists of 78 samples also spanning the years 1610 to 1846. Also available is a Mt. Wachusett living oak master compiled by Harvard Forest composed of 35 samples spanning the years 1673-1997. A provisional Deerfield Area Chestnut master, so named due the data not being cross-verified with living trees, is composed of 19 samples spanning the years 1723-1846.
Procedures In procuring samples suitable for dendrochronology research, the analyst must be on the lookout for timbers, framing, and boards that exhibit several parameters. First, a bark, or waney, edge must be present if one wishes to establish with certainty the last year of growth. Second, there needs to be a sufficient number of rings in a sample to span several distinctive climactic variations that register as patterns of wide and narrow rings. Ideally, having 100 years of growth is best, but more often than not, samples will range from 60 to 100+ years. While it is feasible to get dates on young samples, spurious results are possible and thus must be reviewed carefully both with longer-lived samples from the same structure as well as with what documentary and stylistic research uncovers. This is especially relevant for this study as many of the samples turned out to be 60 years or less in age. Third, enough samples need to be obtained (10-12 per building episode is usually reasonable) to allow for comparison and the fact that often some will not date for one reason or another. It is also critical that an assessment be made of the building frame to ascertain that the members from which samples are extracted were not reused or inserted at a later date. Fourth, all samples must be labeled and entered into a log book that notes the position of each sampled timber within the structure, its species, whether or not it has wane, and any other information pertinent to the sample.
Samples are taken using a custom coring bit, chucked into a Bosch battery-powered drill that creates a 9/16” hole out of which is obtained a 3/8” core. Core samples are glued into custom wood mounts and sanded using successively finer grit paper (60-600 grit) both on a bench top belt sander and by hand sanding to create a mirror-smooth finish. All samples are then viewed under a Unitron ZST 7.5-45X binocular microscope fitted with cross hairs in one eyepiece to ascertain and mark the number of rings per sample. This is followed by a visual review of all samples from the structure to determine if site-specific growth patterns can be picked out. Each sample is then placed under the microscope on a Velmex Acu-Rite Encoder sliding stage calibrated to read to the nearest micron (.001mm). Measuring begins at the outer or last year of growth ring (LYOG), established as 1000, and proceeds to the center of the sample or first year of growth sampled (FYOG). At the junction of each growth ring, the analyst registers the interface by pushing a button sending the measurement to the computer via a Quick-Chek Digital Readout. In these studies, the measuring program Measure J2X iss used to compile each structure’s raw data files. The program transforms the ring widths into a series of indices that relate each ring’s growth to its neighbors, thus standardizing the climate-related influences on a year to year basis (Krusic 2001). Thus trees from a similar location but growing at different rates should exhibit similar indices. With the raw data in hand, using the program COFECHA, samples from each site can be compared with each other to determine if all were cut more or less at the same time or within the span of several years or more. The samples are also compared against dated regional master chronologies of the same species to determine the exact year or years when the samples in question were felled. As strong samples are uncovered, these are added to a fledgling site master and the raw data is again run against the site master to see if additional samples align.
Baillie, M.G.L. 1982 Tree-Ring Dating and Archeology. Croom Helm, London and Canberra.
Flynt, W. 2004. A Dendrochronological Study of a Select Group of Deerfield, Massachusetts Buildings. Deerfield, MA.
Krusic, P.J. and Cook E.R. 2001. The Development of Standard Tree-Ring Chronologies for Dating Historic Structures in Eastern Massachusetts, Phase I. Great Bay Tree-Ring Lab and The Society for the Preservation of New England Antiquities, Durham, NH and Boston.
Miles,D.W.H., Worthington, M.J. and Grady,A.A. 2002. Development of Standard Tree-Ring Chronologies for Dating Historic Structures in Eastern Massachusetts, Phase II. The Society for the Preservation of New England Antiquities and Oxford Dendrochronological Lab. Boston and Oxfordshire.
Miles,D.W.H., Worthington,M.J. and Grady, A.A. 2003 Development of Standard Tree-Ring Chronologies for Dating Historic Structures in Eastern Massachusetts, Phase III. The Society for the Preservation of New England Antiquities and Oxford Dendrochronological Lab, Boston and Oxfordshire.
Miles,D.W.H.,Worthington, M.J. and Grady,A.A. 2005 Development of Standard Tree-Ring Chronologies for Dating Historic Structures in Eastern Massachusetts, Phase IV. The Society for the Preservation of New England Antiquities and Oxford Dendrochronology Laboratory, Boston and Oxfordshire.