We reconstructed a 3500-yr history of fluctuations in treeline elevation and tree abundance in the southern Sierra Nevada. Treeline elevation was higher than at present throughout most of the last 3500 years. We studied foxtail pine (Pinus balfouriana) stands at the crest of the Sierra Nevada, where warm temperatures (ca. 100 BC to AD 900) allowed trees to migrate upslope above current treeline (the upper-most limit in elevation that upright trees exist). Subsequently, periods of severe droughts (AD 1000 to 1400) or cool temperatures (AD 1500 to 1900) caused treeline to move downslope to its current location.

How do we study treeline dynamics?
In the cool, dry climate of the Sierra Nevada, dead foxtail pine trees at high elevations are preserved in situ for millennia. Dead trees located above current treeline testify to the dynamic behavior of treeline in the past. In order to date each tree, we extracted increment cores from each tree and used tree-ring dating techniques to identify the calendar year of each tree's inner and outer ring. In order to document general trends within Sequoia National Park, we sampled five sites (Table) on the eastern crest of the Sierra Nevada (see Map). Our ability to interpret treeline movements in terms of past climatic variation was made possible by the results of SNGC studies of past climatic variation.

By comparing estimates of the numbers and positions of trees above and at treeline under different climatic regimes in the past, we can infer how climatic variation affects treeline dynamics. We see two distinctly different processes at work in our study areas. During some periods in the past, severe drought has reduced treeline populations by causing the death of adult trees. During other periods, low temperatures reduced the number of trees at treeline by inhibiting recruitment of seedlings. However, because foxtail pine can live for more than 1000 years, short-term failures in recruitment rarely causes large changes in the structure of these treeline forests.

What are the implications of this study for understanding the future of the Sierra Nevada?
Seedlings are currently being established above treeline. From our studies, we know that this incipient treeline movement reflects the relatively warm and moist conditions that have persisted for much of the 20th century. And, we know that these changes are still within the natural variability we have witnessed at these sites for the past 3500 years. Our results have three primary implications for the potential of anthrpogenically induced warming to cause future changes at treeline.

First, although large changes in forest structure and distribution occur on time scales of hundred of years, the rate at which these forests respond to rapid future climate change is likely to be slow, and treeline position may lag climate change by decades to centuries. Monitoring efforts at treeline might therefore emphasize change in density of existing stands as well as change in position of treeline.

Second, the period from AD 1050 to 1450 illustrates the extent to which water balance can reverse treeline response to temperature. Whereas climate appears to have been warm when treeline forests expanded, warmth does not necessarily lead to subalpine forest expansion. The observational record of the relatively wet 20th century leads to the prediction that treeline will rise if temperatures rise. The paleoecological record, however, indicates that future warming is unlikely to cause an expansion of subalpine forests if it is accompanied by a reduction in water supply.

Third, responses to past climate variation have involved more than simple upward or downward movements of the treeline boundary: they have involved relatively complex changes in distribution, accompanied by changes in forest structure. The impacts of even minor changes in the elevation of treeline are therefore likely to be substantial. A warming-induced rise in treeline elevation is likely to involve landscape-scale increases in biomass, productivity, and carbon pools as a result of increases in both forested area and density of current forests.