Reconstruction of millennial forest dynamics from tree remains in a subarctic tree line peatland

Ecology, Sept, 1997 by Dominique Arseneault, Serge Payette

INTRODUCTION

Although long pollen and macrofossil records have allowed the reconstruction of vegetation development during the Holocene (Ritchie 1987), there is a lack of information on the ecological context of past tree growth. Deciphering the ecological processes by which vegetation responds to climate change at long time scales is severely impaired by the rarity of plant remains illustrating clear examples of ecosystem responses (Brubaker 1986). Filling this gap will improve our ability to predict vegetation dynamics in changing climates (Graumlich and Brubaker 1995).

Arctic tree line ecosystems are well-suited to analysis of climate effects, being climate-sensitive at different time scales (Ritchie 1987, MacDonald et al. 1993, Spear 1993, Payette and Lavoie 1994). Most well-drained sites in the northern Quebec forest-tundra are colonized by lichen-heath communities and scattered black spruce (Picea mariana (Mill.) BSP.), woodlands, and krummholz (stunted spruce). Charred plant remains, including spruce cones and needles, are found under the lichen mat in most sites, indicating that the landscape was forested sometime in the past (Payette and Gagnon 1985). Charcoal found in present-day wooded sites is usually older than 1000 yr BP (Payette and Morneau 1993), whereas charcoal from treeless sites is predominantly younger than 1000 yr BP (Payette and Gagnon 1985). Accordingly, it has been suggested that the shift from forest to tundra vegetation mainly resulted from reduced postfire regeneration due to climate cooling over the last millennium (Payette and Gagnon 1985).

Winter conditions above the snowpack constitute a critical environmental threshold influencing stem morphology of tree line conifers. Spruce growth forms vary from infranival (below snowpack) cushions on severely exposed sites, to skirted and whorled and flagged trees experiencing episodic stem growth and localized foliage losses above the snow cover on moderately exposed sites, to symmetrical trees in protected sites (Lavoie and Payette 1992). Seeds from damaged spruces are generally of low quality (Elliott 1979, Black and Bliss 1980, Sirois 1988). Two deforestation mechanisms have been proposed: (1) progressive reduction of forest density due to successive fires; and (2) abrupt postfire shift from krummholz to tundra vegetation (Payette and Gagnon 1985, Sirois and Payette 1991, Arseneault and Payette 1992). It is generally accepted that krummholz are often relict stands of former forests (Nichols 1976, Payette and Gagnon 1985), but the mechanisms driving changes from forest to krummholz have not been fully addressed from field evidence.

Wood remains at the surface of well-drained soils have been used to evaluate the responses of tree line woodlands and krummholz to fire and climate disturbances during the last 600 yr (Payette et al. 1985, 1989a, Arseneault and Payette 1992). However, such reconstruction further back in time is impaired by the lack of well-preserved wood. Northern peatlands are favorable for the long-term preservation of dead wood (Eronen 1979, Gagnon and Payette 1981, Eronen and Huttunen 1987, Kullman 1987, 1994, Pilcher et al. 1995, Lavoie and Payette 1996). Due to the ecological information they provide, spruce remains in peat can be used to analyze forest development over a longer time period. Black spruce wood in subarctic environments usually contains several diagnostic light rings (tree rings with few latewood cells; Filion et al. 1986), allowing the cross-dating of most undecomposed wood material. In addition, the cross-dating of charred stems buried in peat may be used to date past fires. Furthermore, because wood remains in peat are well-preserved, the original growth form profile of most buried trees can be adequately described.

Spruce growth forms, including the pattern of stem and branch development above the snowpack, yield direct evidence of climatic effects on tree growth (Begin 1991, Lavoie and Payette 1992). Growth forms develop in response to the death of needles and twigs exposed to frigid wind and drifting snow during winter Hadley and Smith 1983, 1986, 1987, 1989, Scott et al. 1993, Payette et al. 1996). The overwhelming influence of damaging winter conditions is exemplified by morphological anomalies above the snow-air interface, whereas the evergreen foliage below the snowpack is generally dense and healthy. Thus, temporal sequences of stem morphology can be used to identify climatic trends, because stem morphology integrates the influences of climatic conditions during both winter and the growing season. Stunted conifers usually display slower growth than arborescent trees (Payette et al. 1985, 1989a, Begin 1991, Lavoie and Payette 1992). A reduced photosynthetic mass is probably the main cause of the poor growth of shrubby spruces (Filion et al. 1985, Payette et al. 1989a, Lavoie and Payette 1992). For instance, it has been shown that black spruces experienced severe stem damage and reduced growth during the Little Ice Age ([approximately equal to]AD 1580-1880; Payette et al. 1985, 1989a).