Schrader, Barbara. 1991. Vegetation and conifer seedling response to clearcutting of alluvial spruce sites in southeast Alaska. M.S.
Regeneration of conifers was examined following clearcutting on 60 alluvial spruce sites in 12 drainages in northern southeast Alaska. Early successional vegetation was classified to community type.
Conifer seedling density in clearcut alluvial spruce sites was highly variable and concentrated primarily on well-decayed log microsites. Sitka spruce (Picea sitchensis) was the dominant species regenerating in the clearcuts, with a mean density of 6,000 seedlings/ha. However, average seedling density is not a good indicator of overall regeneration success because of the poor distribution of the seedlings, reflecting the patchy microsites. The highest numbers of seedlings per plot were found on well-decayed logs, which were very limited in availability. Competition from shrubs, in particular salmonberry (Rubus spectabilis), appears to limit seedling regeneration on the forest floor. While the tree reproduction appears to be limited by the availability of decayed log substrates, much unexplained variability in seedling density remains suggesting that additional unknown factors, other than simply the total amount of available substrate must also be important
Multiple linear regression was used to describe seedling densities. Basal area of the former stand and the percent of ground cover by slash were significant variables for spruce seedling densities. Percent cover of blueberry (Vaccinium spp.) and the percent cover of forest floor mineral soil were significant variables for hemlock seedling densities.
Eight early successional communities within the Sitka spruce zone were identified. These communities were: (1) Vaccinium alaskaense (2) Rubus spectabilis/Vaccinium alaskaense (3) Rubus spectabilis/Oplopanax horridum (4) Rubus spectabilis/Ribes bracteosum (5) Lysichitum americanum (6) Alnus sinuata (7) Alnus rubra and (8) Marsh/wetland species. In detrended correspondence analysis (DCA), vegetation communities separated out along moisture and disturbance gradients. Simple correlation revealed that the amount of forest floor soil present is positively associated, and the basal area of the former stand and total number of hemlock stems per plot are negatively associated with DCA Axis 1 (a complex moisture gradient).