Two centuries of primary succession at Glacier Bay, Alaska: A test of a classic glacial retreat chronosequence

Abstract

The classic account of primary succession inferred from a 220-year glacial retreat chronosequence at Glacier Bay, Alaska was tested against site-specific dendroecological reconstructions of successional development and against measured 37-year changes in soil nitrogen pools. Differences between young and old portions of the chronosequence in the invasion and radial trunk growth of Sitka spruce (Picea sitchensis) indicate that the nitrogen-fixing shrub Sitka alder (Alnus sinuata) has been an important and long-lived species only at sites deglaciated since 1840. Confirming this result, the soil nitrogen pool at sites deglaciated since 1840 has accumulated rapidly (2.3 to 3.6 g N m$\sp{-2}$ yr$\sp{-1}$) and is currently larger at some sites than at older sites where alder has not been important. A late-successional decrease in soil nitrogen inferred from the chronosequence did not occur at sites where soil nitrogen was measured in 1952 and again in 1989. Consequently, uncritical use of the entire Glacier Bay chronosequence to infer successional trends in community composition, nutrient dynamics, or soil development is unwarranted. Although the nitrogen and organic matter added by alder thickets have an apparent facilitative effect on the growth of individual spruce trees, alder thickets are associated with substantially reduced stand density of spruce. Thus, at the level of the population, a long-term, net facilitation of spruce by alder has not been demonstrated at Glacier Bay. Differences in texture and lithology of soil parent material cannot explain the differences in successional development between young and old portions of the Glacier Bay chronosequence. However, distance from each study site to the closest seed source of Sitka spruce at the time of deglaciation explains up to 58% of the among-site variance in early spruce recruitment. Multiple successional pathways have apparently resulted from changes in seed rain to newly deglaciated surfaces caused by (1) linear ice retreat exposing land surfaces more distant from refugial old growth forests, and (2) shifts in the species composition of the advancing front of successional vegetation and in its seed output.