Ages of Some Large Giant Sequoias:
General Sherman Keeps Getting Younger
Nathan L. Stephenson, Ph.D
U.S. Geological Survey, Western Ecological Research Center
Sequoia and Kings Canyon Field Station, Three Rivers, CA 93271-9651
(Editors Note: This article first appeared in a different format in Madro˝o.)
General Sherman Tree
People have long been fascinated by the great size and longevity of giant sequoias (Sequoiadendron giganteum), which grow naturally only in isolated groves on the western slope of California's Sierra Nevada. Sequoias are the world's largest trees, reaching a maximum known bole volume of nearly 1500 m3 Precise cross-dating of tree rings on cut stumps has shown that sequoias can reach at least 3266 years in age, making sequoia the third longest-lived, non-clonal tree species known, exceeded only by the bristlecone pine of western North America's Great Basin (Pinus longaeva, the "Methuselah" tree at 4,767 years) and the alerce of Chile and Argentina (Fitzroya cupressoides at 3613 years).
Here I present age estimates for some large, well-known sequoias, thereby addressing one of the most frequently-asked questions about famous sequoias -- namely, "how old is this tree?" I additionally address two questions regarding sequoia sizes and ages. First, are the largest sequoias so massive because they are exceptionally old, as is often presumed, or because they have grown particularly rapidly? Second, are there likely to be any sequoias alive today that are older than the longest-lived sequoia yet known, which is known only from a cut stump?
These questions are difficult to answer because the only way to precisely determine a living sequoia's age is to crossdate tree rings on increment cores that intersect the tree's pith. However, the tremendous girth of large sequoias usually makes it impossible to reach their piths with hand-driven increment borers. Power increment borers with very long bits can sometimes be used to obtain cores that reach the pith, but have several disadvantages, which include unacceptably large holes left in the trees, poor quality of many of the cores extracted, and unacceptable use of noisy power tools on and around popular and frequently-visited sequoias.
I therefore estimated the ages of several large sequoias using a method that takes advantage of information from partial increment cores (cores that fall well short of a tree's pith). Unlike previous attempts to estimate the ages of large sequoias, this method has been tested on hundreds of sequoia stumps, does not systematically over- or underestimate tree ages, and offers confidence intervals on the final age estimates.
Pine: The oldest has been dated to 4,764 years.
long-lived Alerce Tree.
There has been a long-standing belief that the largest sequoias are the oldest. This is well illustrated by tracing the history of age estimates for the General Sherman tree, the world's largest tree. By the early 20th century, careful ring counts and crossdating had identified a handful of sequoia stumps more than 3000 years old, the oldest being about 3200 years old. (John Muir's reported count of 4000 rings on the "Muir Snag" in 1875 has not been repeated and was almost certainly in error, and other early claims of up to 11,000 rings counted on stump tops cannot be taken seriously.) Since none of these old stumps approached the great size of the General Sherman tree, most natural historians concluded that the General Sherman tree must be more than 3500 years old. One early researcher believed that the General Sherman tree was about 4000 years old, though he reported that an estimate based on "average number of rings counted ... in charred fragments from parts of the [General Sherman trees] burned trunk, in connection with the actual counts of rings of felled trees ... which have grown under conditions and situation similar to those of the Sherman tree" yielded an age of 5200 years. Popular publications, such as a 1931 program for a play performed among the sequoias not far from the General Sherman tree, tended to be more extravagant, proclaiming the tree to be 6000 years old. Ironically, the aforementioned play took place less than two months before the first quantitative estimate of the General Sherman tree's age based on increment cores, by A. E. Douglass.
Douglass, the founder of the modern science of dendrochronology, obtained six short cores from the General Sherman tree in 1931. He deemed two of the cores to be good enough to use for age estimation, finding that average ring width at 4.6 m above ground level was 0.81 mm. This ring width is less than that of Hartesveldts cores (Table 3) because it comes from a height where the General Sherman trees bole is narrower. Douglass stated that "[t]hese are ring sizes which, in relation to the total size of the tree and the probable rate at which rings increase in size toward the center, supplied an estimate of the age of the tree of 3500 years plus or minus 500 years" (Douglass 1946). I have found no quantitative description of how Douglass accounted for "the probable rate at which rings increase in size toward the [tree's] center."
To shed light on Douglass' age estimate, I applied the approach outlined in this paper to his data. Douglass' data yield an age of only 2380 years for the General Sherman tree in 1931, or 2450 years in 2000 (rounded to the nearest decade). This latter estimate is only 300 years older than the estimate based on Hartesveldt's cores (Fig. 1), and is well within that estimate's 95% confidence interval. However, I judge the estimate based on Hartesveldt's cores to be much more reliable than that based on Douglass' cores. Specifically, the estimate based on Hartesveldts cores required that fewer key parameters be estimated (such as the diameter of the General Sherman tree at 4.6 m above ground level in 1931, needed for using Douglass data), and was based on three cores widely spaced around the tree's bole, each of which was nearly twice as long as the longest of Douglass' two adjacent cores.
In contrast, an age estimate based on linear extrapolation of Douglass' ring-width data, assuming no change in ring width toward the General Sherman tree's center (an unrealistic assumption), would yield an age of 3790 years in 1931. Thus, Douglass' estimate of 3500 (▒500) years apparently was little different from an estimate based on a simple linear extrapolation, and did not adequately consider the increase in ring widths toward the pith.
Douglass' age estimate was widely quoted (and sometimes exaggerated) from 1931 until the 1960s, when Hartesveldt et al. (1975) radically revised the estimate downward. Unlike Douglass, Hartesveldt and his colleagues explicitly stated their assumption as to how ring widths change within a tree: they assumed that basal area increment is constant (that is, trees add a constant amount of basal area each year). This is equivalent to substituting d = 2 into eq. 1 (Stephenson and Demetry 1995). Hartesveldt's notes (archived at Sequoia National Park) show that when he strictly adhered to this assumption, he estimated that in 1964 the General Sherman tree was only about 1600 years old. However, Hartesveldt's examination of growth patterns on sequoia stumps measured by Huntington (1914) indicated that strict adherence to this assumption sometimes underestimated the ages of sequoias (Hartesveldt et al. 1975). Thus, apparently based on a combination of assumed constant basal area increment and judicious comparisons with Huntington's data, Hartesveldt and his colleagues (1975) cautiously stated that the General Sherman tree "... is less than 2500 years old." According to my calculations using their original cores and data, their statement has a more than 75% probability of being true (Fig. 1).
As careful as Hartesveldt may have been in stating that the General Sherman tree was less than 2500 years old, the National Park Service, perhaps unable to bear such a precipitous decline in the tree's age, instead adopted 2500 years as the midpoint for a range encompassing the tree's estimated age. At the time of this writing, Park literature and the plaque at the General Sherman tree stated that the tree's estimated age was "2300 - 2700 years." Additionally, a popular book authored by Hartesveldt's colleagues dropped the qualifier "less than," stating instead that the tree "... is about 2,500 years old" (though a table on the same page gives the General Sherman tree's age as "2,500 - 3,000" years!). The most recent estimate of the General Sherman tree's age -- 2150 years (Fig. 1) -- is most closely aligned with Hartesveldts original statement that the tree is less than 2500 years old.
The relative youth of other famous sequoias may come as a disappointment to some. For example, the decline in the estimated age of the Grizzly Giant tree has been even more precipitous than that of the General Sherman tree. Clark (1910) reported that the Grizzly Giant had been growing so slowly over the last few centuries that its rings (presumably observed inside of a fire scar cavity) were "as thin as wrapping paper, too fine to be counted with the unaided eye." (On the contrary, measured ring widths [Table 3] and measured tree volume changes both indicate that the tree has been growing quite rapidly.) Comparing these purported ring widths with those of some fallen sequoias, Clark concluded that "the Grizzly Giant must be not less than six thousand years old," and that the tree was probably the oldest living thing on earth. Other early age estimates placed the Grizzly Giant at a more modest 3800 years old, while Hartesveldt later suggested that the tree "... is perhaps only 2500 years old." At the time of this writing, the National Park Service reported the age of the Grizzly Giant as 2700 years. However, I estimate the tree to be only about 1790 years old (Fig. 1), and that the probability of it being at least 2700 years old is less than 2%. Hartesveldt and his colleagues offered solace to those disappointed by the suggestion that certain large sequoias might be younger than expected: "... this [discovery] effects a change only in superlatives; the world's largest trees are the world's fastest-growing trees."PQ
Some readers may be disappointed by the broad confidence intervals associated with age estimates in Fig. 1. There is a great deal of uncertainty in estimating the ages of individual large sequoias, largely due to relatively abrupt and sustained changes in ring widths in the part of the bole not sampled by increment cores, and therefore invisible to us. Such changes in growth rates are due to unpredictable, site-specific events in the past, such as occasional, localized high-intensity fires. Thus, though Fig. 1 suggests that the General Sherman and Sentinel trees are the same age (2150 years), the broad confidence intervals additionally suggest that this correspondence is most likely a meaningless coincidence. However, most of the confidence intervals in Fig. 1 are based on relatively short cores. Confidence intervals could be tightened somewhat in the future by taking longer cores and, in the case of the Washington and Cleveland trees, more cores.