Even a cursory glance at the landscape reveals that vegetation is not distributed at random, but occurs in mosaics as an expression of several interacting variables. The whitebark pine (Pinus albicaulis, meaning white-stemmed pine) is a tree found at Crater Lake National Park generally above 6500 feet on exposed slopes in dry, rocky soils. This tree is easily identified by its whitish-gray bark and often twisted branches. Although Crater Lake National Park has no true timberline, whitebark pine forms the elfinwood or krummholz of timberline in many western mountain ranges.

Whitebark pine is a pioneer species colonizing subalpine habitats as the first tree. An amazing example of its pioneering ability can be seen at the Newberry caldera where whitebark pine is the only tree established upon the relatively recent obsidian surface; even the nearby lodgepole pines (P. contorta, subsp. murrayana) abruptly end near the toe of the flow. At the Crater Lake caldera, whitebark pine may have been the first tree to colonize the pumice slopes of old Mount Mazama within the first century following the climactic eruption. Whitebark pine is generally encountered as a pioneer tree, as there are several places around the caldera rim where old mother trees provided a favorable microclimate for the establishment beneath their canopy of subalpine fir (Abies lasiocarpa) or mountain hemlock (Tsuga mertensiana). Whitebark pine is arranged in ribbons or bands along the contours of Cloudcap and other habitats along the caldera's edge. These sites represent slightly higher, rocky substrate for the survival of whitebark seedlings since exposed areas devoid of snow earlier in the year have a significantly longer growing season.

Most pine seeds have wings for wind dispersal, but whitebark seeds have retained only a rudimentary wing. The dispersal agent has become the Clark's Nutcracker (Nucifraga columbiana). These birds have learned to retrieve whitebark seeds with their specialized beaks, storinga number in their sublingual pouch, and methodically storing seeds in soil caches. Only a fraction of the seed caches are retrieved, however, so some caches sprout seedlings in clumps which may grow into larger whitebark pine colonies.

Whitebark pine appears to be sensitive to a certain set of environmental conditions. Although it is often viewed only as an indicator of a short "rowing season and cold temperatures, this species occupies a niche in the subalpine forest that is far from simple. The tree can be found in relatively pure stands or in association with lodgepole pine and western white pine (P. monticola). The distribution of related species like limber pine (P.flexilis), bristlecone pine (P. longaeva), and foxtail pine (P. balfouriana) somewhat overlap that of the whitebark and can occupy what would often seem to be the latter's place forming the edge of timberline. Whitebark pine's distribution poses some nagging questions to dendrologists. For example, it provides the name for Nevada's Pine Forest Range but mysteriously remains absent in similar subalpine habitats on Steens Mountain in southeastern Oregon, only several air miles to the north. In southern Oregon, the whitebark pine may have disappeared on Mount Ashland in recent times and is presently almost gone from the top of Crater Lake's Wizard Island.

One of the reasons that whitebark and other pines are often so puzzling is that species of Pinus display much variation as well as many similarities. For example, whitebark pine and limber pine (the rarest native coniferous species in Oregon, but more common in the Great Basin and northern Rockies) mimic each other in many characters. Similarly, the ponderosa pine (P.) found along Annie and Sun creeks, for instance, display a strong Washoe pine (P. washoensis) element. This is thought to be a high elevation variant of the ponderosa's northwestern distribution and may account for its presence at higher elevations inside the caldera. Genetic variability in the park's whitebark pine may not be as great as in the ponderosa forests, but the loss of a population as small as the one on Wizard Island may imperil a distinct local seed source.

What is disturbing about the whitebark pine of Wizard Island is their seemingly rapid decline. Photographs taken at various times through the 1960s show living trees on top of the island. By July 1991, however, the authors could find only one living specimen. This small population's relatively sudden nosedive may be due to one or several causes. Might it be human activity, air pollution, drought, mountain pine beetle, or blister rust infection? The whitebark's decline is more likely tied to a combination of these factors, which makes the testing of single hypotheses (a key to the application of scientific method to the problem) very difficult or, at best, inconclusive.

Efforts aimed at monitoring environmental change in national parks like Crater Lake are generally handicapped by the lack of critical baseline information. Material available to the historian may help to reconstruct past conditions, but investigators should be aware of their possible shortcomings. The documentary record is limited to the historic period, whether it is in the form of photographs or writings.

Repeat photography is constrained by the scale and resolution of the original photo, as well as by the identifiable background features that allow a view to be replicated. Observations about the condition of flora throughout the park are usually fragmentary. Some describe what would seem to be unlikely events, even though the journalist may otherwise be credible. One example is a newspaper article of 1903 where Klamath Falls hotelier and photographer Maud Baldwin noted that Wizard Island was '"alive with grasshoppers. " Sufficient detail or locality data to verify an observation can be a problem, too. Much effort was expended by Crater Lake's chief park naturalist in the 1960s trying to track clown a colleague's discovery of the prostrate juniper (Juniperus communis) specimen probably living near the Watchman in 1929.

Other changes that might have occurred during the historic period lack any form of documentation. Just one of many examples in the park is the poor condition of Sun Meadow's vegetation when compared to the floral mosaic of Sun Notch. Simplistic explanations, such as sheep grazing prior to the park's establishment or a poor soil nutrient budget, are often offered by park staff when the limitations of available evidence or funding seem to frustrate efforts to study the situation further.

What the whitebark's disappearance on Wizard Island may illustrate, as have the attempts to understand fluctuations in Crater Lake's clarity, is that we really under stand very little about the park's ecosystems. Certainly more research is essential, but the limitations of available data have to be accepted since causation may be due to a number of factors not easily separable into testable hypotheses. Instead of certainty, all history and science can yield is a prediction of possibilities if the limitations affecting available evidence can be overcome through sound methodology.

Explanations based on models of complexity rather than simplicity will have to be complemented, however, by a willingness to admit that sometimes we do not have all the answers. Since whitebark pine ring the summit crater which provides the lake's name, what better symbol of uncertainty could there be for a phenomenon as complex as Crater Lake?