The most dramatic climatic events at the daily timescale are those concerning severe storms that are accompanied by floods and in some cases windthrow events. Daily precipitation and streamflow values vary by more than two orders of magnitude within each year. Fifteen-minute precipitation and streamflow values can vary by the same amount within a few days. Climatic/meteorologic events related to the ignition and spread of forest fire might also be considered in this category, although these events also incorporate effects of preceding droughts and associated drying of fuels. At the Andrews Forest, a small number of daily timescale events can have a large impact. Snyder (2000), for example, found that in the 50 years of records, most flood-related action and landslides occurred during only three major storm events. Similarly, most windthrow events in the northern Cascade Range, Oregon, since 1890 are associated with just three major individual storms (Sinton et al. 2000).
A very large 50-year-return period flood in 1996 led to a large direct response from the ecosystem. This February flood resulted from 290 mm of precipitation over 4 days that melted a large amount of already accumulated snow. Swanson et al. (1998) and Nakamura et al. (2000) list and document landslides and channel erosion and related disturbance of aquatic and riparian organisms and their habitats as responses to this flood. The hydrographic response varied with altitude because of the varying snowpack dynamics. At least 35 debris flows severely disturbed stream and riparian environments. There was a large amount of fluvial erosion. In some areas riparian vegetation was entirely removed in larger channels, and boulder and coarse woody debris movement was common (Johnson et al. 2000). Scouring in places uncovered objects that had long been buried. Wood samples exposed along the northeast side of Watershed 3 turned out to be over 46,000 years old (http:// www.fsl.orst.edu/lter/pubs/spclrpfr.htm). Many stream restoration project structures were washed away. Some biotic responses were very fast. Benthic algae recovered from the event within weeks. Again, this web page on the 1996 flood provides details on these effects: http://www.fsl.orst.edu/lter/pubs/spclrpfr.htm.
At the Andrews LTER site, windthrow events result mainly from southeasterly winds associated with storms arriving from the Pacific (Gratkowski 1956). Wind-throw events in winter in the northern Cascade Range of Oregon were found, in some cases, to highlight the importance of preexisting conditions. Sinton et al. (2000) found such events occurred particularly when winds were from the north or east with a preceding period of dry weather. High-pressure conditions in winter gave rise to icing on the branches of trees prior to some windthrow events (D. S. Sinton, pers. comm., 1996). As is well known, windthrow events cause forest gaps that subsequently undergo a cascade of successional events leading to the reestablishment of the forest. However, canopy gaps, especially those with fresh, clear-cut edges, are particularly prone to additional windthrow (Gratkowski 1956; Sinton et al. 2000).
In some cases ecosystems respond to the coincidence of two climatic events. One such example is the occurrence of Douglas-fir bark beetle outbreaks (Powers et al. 1999). In western Oregon and Washington, these insects are usually sapro-phytic, reproducing in freshly downed Douglas-fir trees. In rare instances, however, this species can kill live trees, and for a 2- to 3-year period can increase overall mortality in forests by a factor of 3 to 10. The coincidence of two climatic events is necessary for this to happen. First, a major windstorm or incidence of ice damage is necessary to create a large amount of breeding habitat. This allows the population to expand to sufficient numbers to attack living trees and kill them. Curiously, at least for this species of beetle (and spruce bark beetle as well), fire-killed timber is not a suitable enough habitat to increase the population. Second, the trees must be under stress during the growing season. This stress is usually caused by drought and reduces the trees' ability to respond to the beetle attack. Even in large numbers, the Douglas-fir bark beetle has little ability to overwhelm trees. Although the beetles can reproduce in live trees, they are unable to increase numbers in this habitat; therefore, outbreaks in live trees rarely last for more than 3 years despite the length of the drought. The rare coincidence of these two sets of climatic conditions means that Douglas-fir bark beetle outbreaks are rare events for western Oregon and Washington forests. Although the historical record of outbreaks is not long, outbreaks appear to occur at an average frequency of 50 years. These outbreaks do have important impacts: They alter forest composition (ironically by removing a more drought-resistant species), speed the rate shade-tolerant species dominate stands, temporarily increase the amount of detritus, and reduce the Net Primary Productivity (NPP) of the forest, with the end result of creating a temporary source of CO2 to the atmosphere.
Year-to-year oscillations in precipitation are responsible for variations in NPP, decomposition, and Net Ecosystem Productivity (NEP). By examining tree cores and litter fall records, Fraser (2001) found that tree growth and litter fall varied ±30% from year to year. Given lag of 4 to 5 years between leaf production and litter fall, the amount of combined variation is not clear; however, it is likely to be in a similar range. Year-to-year variation in decomposition rates has not been studied extensively, but fortuitous studies carried out during extremely dry and wet years indicate a range of ±30% (Valachovic 1998; Harmon, unpubl. data, 1992). This response is not likely to be mirrored in other forms of detritus, however, because their rates of drying, and response to moisture, differ substantially. Fine litter, for example, dries quickly and, because of its high ratio of surface area to volume, is rarely limited by excessive moisture. In contrast, large wood dries slowly (Harmon and Sexton 1995) and has a low enough surface-area-to-volume ratio that diffusion of oxygen can become limiting for decomposition when moisture content is high (Harmon et al. 1986). This means that summers with high precipitation can lead to fast decomposition of fine litter, but slow decomposition of large wood. Conversely, in summers that are dry, fine litter decomposition can be slow and that of large wood fast. As a result, the year-to-year variation in overall decomposition is likely to be dampened as the detritus pools are "decoupled" temporally from each other. By combining these sources of variation in NPP and decomposition, preliminary estimates are that NEP (the net exchange of carbon with the atmosphere) could vary as much as 2 Mg ha-1 year-1 in Douglas-fir/western hemlock old-growth forests (Harmon et al., in press). This is a substantial level of variation. Although these forests are thought to have an NEP close to zero over the long term, this variation means that in some years they are uptaking as much carbon as a forest in the peak carbon accumulation phase (Janisch and Harmon in press). Clearly, a more thorough examination of the cause of this year-to-year variation is necessary, but the key lessons that ecosystem processes are not responding to the same climatic variable in the same way and that the sign of the response could differ, even within a process, are likely to hold.
Was this article helpful?