Corydalis aquae-gelidae is a large herbaceous perennial in the Fumariaceae that has high habitat specificity. It is endemic to the western Cascade Range of Oregon and Washington, almost entirely on the Mt. Hood and Gifford Pinchot National Forests. It is a federal C2 candidate under the Endangered Species Act and a Forest Service sensitive species. This study included a niche analysis, i.e., defining the levels of environmental factors where Corydalis aquaegelidae grew, and relating its abundance to environmental variation, using data from the southwestern Mt. Hood National Forest. Corydalis grew in relatively cool habitats (between late June and early September, diurnal means were 17.7°C for air temperature, 9.7°C for water temperature, and 10.6°C tor substrate temperature). Elevations were from 460 to 1300 m, on streams from headwaters to the fifth order Clackamas River. The stream reaches occupied by Corydalis had relatively small seasonal fluctuations in flow. An estimated 90% of the plants on the Oak Grove Fork were submerged yearly, and about 75% of the plants were between average summer low water level and winter high water level. The abundance of Corydalis aquae-gelidae was modeled using principal components analysis, resulting in a reproduction index (based on the number of seedlings and juveniles) and a biomass index (from the number of adults, their average leaf numbers and height, and percent cover). The abundance of Corydalis was greatest within 5 cm vertical distance below to 20 cm above the summer water levels, and within 150 cm horizontal distance to the water. Optimal water cover was within 10 to 30%. Abundance decreased with increasing organic material and fines (material <2mm) substrate cover. Optimal mineral and moss cover were 25 to 50% and 50 to 80% respectively. Abundance increased with increasing gravel and cobble cover. Abundance was suppressed where canopy cover was below 20%. GS substrates (gravel and coarse sand) were most favorable, followed by GX substrates (gravel and sand dominant, with other materials), then LX substrates (soil or mud dominant). Emergence from seeds and rootstocks occurred in late May and early June. Plants farther from the water flowered later than those closer to or in the water. Senescence occurred in late August to mid-September. Temperature differences and phenological state differences were generally greater within sites than across the elevational gradient. Corydalis aquae-gelidae seeds required 6 to 7 months of stratification before germination in the laboratory. Under lab conditions, cold substrates (less than 5.8°C AM low, 12.6°C PM high) reduced growth of Senecio triangularis seedlings, but not of Corydalis. Corydalis seedlings grew as rapidly in a greenhouse as in the field, but warmer, drier treatments produced less rapid growth than cool treatments with saturated substrates. Corydalis aquae-gelidae had relatively high allocation to roots, and slightly below average reproductive allocation, compared to data for other herbaceous species in the literature, including species from the same region. Development was slow; an estimated 8 years were required for reproductive maturity. Using a growth model and growth stage data from the field, a population age structure was hypothesized. Based on this age structure, Corydalis reproduction appeared to be adequate for population maintenance. Seed dispersal was mainly by explosive dehiscence and flowing water. Corydalis aquae-gelidae was pollinated by bumblebees (Apidae). VA mycorrhizae were present but infrequent in the juveniles and seedlings investigated. Slugs were the most frequent cause of herbivore damage. Downy mildew (Peronospora) occurred at relatively low elevations and under relatively low canopy cover. Both competition with and niche differentiation among streamside species were thought to occur. Corydalis aquae-gelidae appeared to be a relatively late-successional plant. The range of Corydalis aquae-gelidae may be restricted by dispersal problems, rather than by unfavorable environments, as apparently suitable habitat existed beyond its range. Timber harvest and road building have damaged Corydalis aquae-gelidae populations through direct physical disturbance and habitat degradation. Water diversion for the Stone Creek Hydropower Project should lead to loss of plants from the large Oak Grove Fork populations through drought and other habitat changes. Fisheries habitat improvement projects and grazing may also threaten Corydalis populations. / Graduation date: 1993