Tree-ring dating is formally known as “dendrochronology” (literally, the study of tree time). It is the science of assigning calendar-year dates to the growth rings of trees, and Colorado figures prominently in its development and application in archaeology and other disciplines.
Tree-ring dating provides scientists with three types of information: temporal, environmental, and behavioral. The temporal aspect of tree-ring dating has the longest history and is the most commonly known—tree rings can be used to date archaeological sites, such as the Cliff Dwellings found at Mesa Verde National Park (MVNP) or historic cabins.
The environmental aspect of tree-ring dating today has the most worldwide application, as tree rings can be used to construct records of ancient temperature, precipitation, and forest fire frequency. They can also be used to build databases of stream flow, drought severity, insect infestation, and other environmental variables that trees record while they grow. The behavioral aspect of tree-ring dating, meanwhile, allows archaeologists to understand ancient wood-use practices, trade, and other activities.
Tree-ring dating may only be performed on tree species that produce one growth ring per year, and do so in response to annual variations in precipitation (and in some cases temperature). Everything else being equal, in a wet year trees will produce a larger growth ring. In a dry year, trees will produce a narrow growth ring. In particularly dry years, trees may fail to produce a growth ring at all. As a result, tree-ring dating requires use of a procedure called cross-dating.
Cross-dating is accomplished by documenting, analyzing, and matching repeated patterns of wide and narrow rings in tree-ring cores collected first from the same tree, then from trees in the same stand, and then from sites in the same region, all of which are responding to variations in the same climatic variable (e.g. precipitation). Then, by working backward from the current year, the dendrochronologist is able to determine the exact year in which each growth ring was formed, thus producing a master tree-ring chronology. Ring patterns from newly collected specimens, such as those from archaeological sites, are then compared to the master chronology in order to provide a tree-ring date for that specimen.
Three aspects of cross-dating warrant emphasis. First, tree-ring dating is about matching patterns, not counting rings. Second, sample sizes must be large in order to understand tree-growth variability in a given region. Third, one begins by studying living trees in a given area, cross-dating their ring series internally and working back in time to successively older specimens that are usually found as dead snags on the landscape or as construction beams in ancient dwellings.
In theory, tree-ring dating is a relatively straightforward process; in practice it can be astonishingly difficult. It requires rigorous sample collection and preparation, methodical attention to detail, and deep knowledge of tree-growth characteristics and wood attributes across vast regions.
Andrew Ellicott Douglass, an astronomer at the University of Arizona in Tucson, is considered the father of tree-ring dating. He introduced the American public to the technique in a December 1929 article in National Geographic entitled “Talkative Tree-Rings and the Tales They Tell.” In that article, Douglass published construction dates for six cliff dwellings at Mesa Verde National Park (MVNP) in southwestern Colorado, including Balcony House, Cliff Palace, Oak Tree House, Spring House, Spruce Tree House, and Square Tower House. Although the exact dates Douglass published have long since been refined, his general dating has not changed: the vast majority of cliff dwellings were built and occupied in the mid-1200s. These dates came as a shock to many archaeologists who, on the basis of little more than educated guesswork, thought that the cliff dwellings were much older.
In the 1930s archaeologist Earl Morris of the Carnegie Institution supplied Douglass with numerous wood specimens from Johnson Canyon, Colorado, south of MVNP, in an effort to extend his ability to date sites back to about 2,000 years ago. Also in the 1930s, Zeke Flora, an amateur archaeologist based in Durango, sent wood and charcoal specimens to Douglass. Subsequent examination of the Flora collection in the 1960s led to the discovery of the oldest tree-ring dated archaeological wood specimen in North America. From the Falls Creek rock shelters outside Durango, it dates to 272 BC.
A major expansion of visitor facilities at MVNP in the 1950s and 1960s led to huge improvements in archaeological tree-ring dating, particularly in the dating of charcoal samples from pithouse sites on top of Wetherill Mesa, in addition to dating cliff dwellings in the canyons between the mesas.
As we now know, the arid climate of southwestern Colorado makes it one of the best regions in the world for dating archaeological sites by tree-ring analysis. A recently completed database for the Four Corners region contains nearly 14,000 tree-ring dates from hundreds of archaeological sites; there are now over 4,300 tree-ring dates known from more than 140 archaeological sites in MVNP alone. The earliest date is AD 255 from a pithouse at the Soda Canyon Campground Site; the latest date is AD 1281 from a loose log found in a ceremonial room at Spring House. These large datasets allow archaeologists to understand the occupation and abandonment of southwestern Colorado with unprecedented precision.
In another major contribution to science, Douglass used tree rings to infer that a “Great Drought” had occurred across the American Southwest from AD 1276 until 1299. The Great Drought was at least partially responsible for the migration of people away from southwestern Colorado in the late thirteenth century; archaeologists are still examining the social, political, religious, and environmental implications of this important event in pre-Columbian history.