8097 - Laserfiche WebLink

Home Browse Search

the sites. Misclassified sites usually were classified as sites with similar soil moisture regimes. This was because the included variables either related directly or indirectly to the soil moisture regime. The dis- criminant analyses model for the herb-strata cluster-groups correctly classi- fied 82% of the sites. Eight variables (a = .01) were included in the latter model. In order of importance they include: (1) soil pH in the C sampling depth (28-32 cm); (2) electrical conductivity of soil in the A sampling depth (0-5 cm); (3) electrical conductivity of soils in the B sampling depth (13-17 cm); (4) mean basal area for all tree species at the site; (5) index of soil stoniness and/or compaction; (6) mean water potential of Rosa woodsii at the site; (7) soil moisture retention percentage at field capacity at the C sampling depth (28-32 cm); and (8) combined density for all shrub species. Better classification was achieved for the herb-strata cluster-groups than for those groups based on overs tory species. Appar- ently the herbaceous vegetation was more influenced by the top 30 cm of soil than were shrubs and trees that tended to be deeper rooted. Stepwise multiple regression analyses most frequently incorporated the variable water potential in equations explaining frequency of tree and shrub species. Environmental parameters selected for regressions to ex- plain aboveground standing crop phytomass of trees, forbs, and shrubs varied from species to species. However, they generally were associated with soil parameters of the A sampling depth (0-5 cm) or the C sampling depth (28-32 cm; see Hansen, 1977 for regression equations) . CONCLUSIONS One of the most characteristic attributes of riparian communities along Rock Creek is that they are constantly in a state of flux on a relatively short time scale. Physical perturbations appear to be the rule rather than the exception. Seasonal variations in streamflow arising from spring runoff and frequent summer droughts, for example, result in a variety of soil moisture conditions for the riparian communities. Alder stands, plant communities primarily affected by flooding, appear to be particularly resistant to flood damage. At higher and moister stream reaches, alders are not restricted to streamsides. However, at lower elevations alder are restricted to point bars of the stream where flood reduces competition. Alders fix atmospheric nitrogen making them ideally suited to establish on bare soils of low fertility. During periods of extremely high water, flooding was also observed in cottonwood communities. Conifer communities at higher elevations were flooded only during unusually high flood stages. While willow, wet meadow, and birch communities are rarely flooded by the river (except near beaver dams), they may be inundated by spring runoff. The drying patterns observed in willow communities were strongly correlated with the species of willow found at the site. River birch communities were found most commonly on deep soils (.7 to .9 m) along springs and seeps. Shallower soils along springs and seeps generally favored willow or alder communities. Seventeen charcoal layers occur in thick alluvial deposits (5.45 meter; 18 feet) along Rock Creek. This suggests that the occurrence of fire is also common to plant communi ties on most s'treams of the uintas. In contrast to cottonwoods and hawthorns, alders rarely were observed to resprout after fires. 64