Mechanisms of Primary Succession Following Deglaciation at Glacier Bay, Alaska

In primary succession following deglaciation at Glacier Bay, Alaska, we tested the hypothesis that the major effect of initial nitrogen—fixing colonizers is to facilitate establishment of late—successional dominants and that other possible causes of successional change (e.g., life history factors governing seed rain and competitive interactions among species) need not be invoked. Environment changed dramatically through the first 200 yr of succession. Soil organic matter increased 10—fold in the upper mineral soil with corresponding increases in soil moisture, total nitrogen (N), and capacity to support plant growth and declines in bulk density, pH, and total phosphorus (P). Plant growth in pioneer soils tended to be simultaneously limited by both N and P, as well as by unknown factors (perhaps lack of mycorrhizae), whereas only P limited growth in older soils. Light availability to seedlings declined through succession. Early—successional species (Epilobium latifolium, Dryas drummondii) had smaller seeds, younger age at first reproduction, shorter life—span, and shorter height at maturity than did mid—successional (alder, Alnus sinuata) and late—successional species (sitka spruce, Picea sitchensis). Seed rain of alder and spruce was negligible in the pioneer stage, increased prior to the stage in which a species was dominant, and was greatest in the stage in which a species dominated. Vegetation in each successional stage inhibited germination and initial establishment of sown alder and spruce seeds (except a tendency of the "black—crust") algal/microbial community in the pioneer stage to enhance survivorship). Removal of the surface litter layer generally enhanced germination and survivorship, particularly of alder. Comparisons of germination in the greenhouse and the field indicated that climatic or indirect vegetation effects (e.g., differential seed predation) and allelopathy also reduced germination and establishment in vegetated communities. Naturally occurring spruce seedlings grew most rapidly in the Dryas and alder stages and most slowly in the spruce stage. Similarly, growth of spruce seedlings transplanted into each successional stage was facilitated by the Dryas (nonsignificantly) and alder stages but inhibited by the spruce stage, relative to earlier successional stages. Facilitation of growth of natural and transplanted spruce seedlings by Dryas and alder stages was associated with higher N and P uptake and tissue nutrient concentrations, whereas nutrient uptake and concentration in spruce seedlings declined in the spruce stage. By contrast, transplanted alder seedlings grew rapidly and accumulated most nutrients in the pioneer stage and were strongly inhibited by subsequent stages. The facilitative effect of Dryas and alder comes primarily from inputs of organic matter and associated N. Addition of alder litter stimulated nutrient uptake and growth of transplanted spruce seedlings in the pioneer and Dryas stages, whereas shading had no effect on growth of spruce seedlings. Root trenching and planting of spruce near isolated alders indicated that, although the net effect of alder is facilitative, alder also inhibits growth of spruce seedlings through competition for soil resources. Strong root competition also occurs in the spruce stage. Alder competitively inhibits Dryas, primarily by shading but also through the physical and allelopathic effects of its litter. In general, both at Glacier Bay and elsewhere, life history traits determine the pattern of succession. Changes in competitive balance accompanying successional changes in environment provide the mechanism for changes in species dominance. Initial site conditions (and facilitation, where present) influence the rate of change and final state of community composition and productivity. We conclude that no single factor or mechanism fully accounts for primary succession at Glacier Bay.