Recovery Plan for the Northern Spotted Owl - DRAFT

III. Forest Protection in theWest Cascades SubregionA. The Natural History of DisturbanceFireIn the moist Douglas-fir forests of the Coast Range of Oregon, the WashingtonCascades, and the Olympics, fire return intervals are long (Fahnestock andAgee 1983) and most forests are first-generation post-fire forests less than 750years old. This would suggest a fire return interval somewhat less than 750years. The fire cycle model of Agee and Flewelling (1983), which is based onclimate, could not reproduce a natural fire rotation (essentially a fire cycle) lessthan 3,500 years using 20th Century climate patterns, and even with significantalteration in climate input to the model, fire return intervals could only bebrought down to about 900 years. They suggest that perhaps much largerthan average events may have occurred in the past (also suggested byHenderson and Peter 1981, for the southeastern Olympics) as a result of shorttermbut very extreme changes in two or more of the climate parameters thatdrive the model.Large fires have occurred in the historic past (Morris 1934) but our knowledgeof old-growth forest establishment dates is so weak as to preclude firm hypothesesabout disturbance pulses of the presettlement past. The forests aroundMt. Rainier appear to have had a major fire event about 750 years ago(Hemstrom and Franklin 1982) and similar aged stands have been identified inthe southern and western Olympics (Agee personal observation). A series ofabout 650-year-old and/or 450- to 500-year-old fires is apparent from the dataof Henderson and Peter (1981) in the southern Olympics; Franklin andHemstrom (1981) and Yamaguchi (1986) in the southern Washington Cascades;and Huff (1984) and legend (Quinault Natural Resources 1983) in thewestern Olympics. Although forest age-class data are sparse, these are alsotimes of sunspot minima identified by Stuiver and Quay (1980), using tree-ringanalysis of carbon- 14 activity. If large fire events are associated with theseperiods of general global cooling, they may represent periods where alteredsynoptic weather patterns, particularly during the growing season, containedhigher lightning frequency and foehn (east) wind patterns.444In moist Douglas-fir forests, long early seral tree recruitment (e.g., 75 to 100years for Douglas-fir) has been documented after disturbance by fire (Franklinand Hemstrom 1981). This pattern is not characteristic of all prehistoric fires.For example, Huff (1984) showed a 60-year recruitment interval for a ca. 1465fire in the western Olympics, while Yamaguchi (1986) shows that about 95percent of Douglas-fir was recruited within 40 years after a fire in ca. 1300near Mount St. Helens. However, even on these sites the regeneration period isdecades long and probably represents some regeneration from trees thatinitially colonized the burn and grew large enough to produce viable seed tohelp completely restock the stand. Lack of seed source, brush competition,and/or reburns have been identified as factors delaying regeneration on suchsites (Franklin and Hemstrom 1981). Patterns of reburns on the Tillamook fireof 1933 at 6-year intervals (1939, 1945, 1951) (Pvne 1982), at Mt. Rainier inthe late 19th Century, and at the southern Washington Yacolt burn of 1902(Gray 1990) are evidence these sites will rebum. High surface fire potentialduring early succession in Douglas-fir forest was identified by Isaac (1940) as a"vicious cycle" of positive feedback, encouraging rhizomatous bracken fern(Pteridium aqutihnum): this pattern was quantified by Agee and Huff (1987).Given sufficient sources for reignition (e.g., the original Yacolt and Tillamook