G.M. MacDonald
Glen M. MacDonald is an associate professor in the Department of Geography. McMaster University, Hamilton. Canada.
How fossil pollen records from lakes and bogs provide a means for identifying long-term patterns of change in the occurrence and relative abundance of species in boreal forests.
Effective management for the conservation and sustainable use of boreal forests requires detailed forest monitoring to detect and manage the impact of natural and human - caused changes in the environment. Perhaps the most serious and widespread human-caused threat to the boreal forest is global climate change (Emanuel, Shugart and Stevenson, 1985; Kauppi and Posch, 1985; 1988). The response of vegetation to the widespread environmental modifications which might, in the future, be caused by climate change could include changes in plant community structure, shifts in the geographic ranges of plant species and alterations in the genetic structure of populations.
However, data on the present conditions of boreal forests are limited in both temporal and spatial coverage. The use of satellite imagery combined with intensive ground truthing and stand studies can enhance the spatial coverage of such data. However, the temporal extent of these records remains, at best, in the order of 50 to 100 years. This is less than the potential life span of most boreal tree species. The lack of long-term data raises two questions: are longer-term records required and is there a source of proxy data that can be used to provide such records retrospectively? The answer to the first question is clear. The current vegetation of the boreal zone has been developing since the close of the last glacial period approximately 10000 years ago. A monitoring record of 50 years represents a sampling of a two-hundredth of this time period. Given that the average life span of boreal trees ranges from 80 to 400 years and that the impact of natural environmental changes generally take at least one generation to become apparent, such observational records are likely to be insufficient to detect gradual, even if significant, changes in the forest.
In attempting to respond to the second question, this article first describes how fossil pollen records from lakes and bogs provide a means of identifying long-term patterns of change in the occurrence and relative abundance of species in boreal forests. It outlines some important conclusions that may be drawn from data regarding the development of boreal forest community structures, rates and patterns of change in the geographic distributions of boreal tree species, and the geographic distribution of genetic diversity.
The dominant boreal tree genera (Picea, Pinus, Larix, Abies, Betula and Populus) are wind-pollinated and produce copious amounts of readily transported pollen. Most of this pollen never fulfils its reproductive function and some of it is deposited in lakes and bogs where, under anaerobic conditions, the exterior portions of pollen grains can be preserved indefinitely. Palaeo-ecologists have well - established protocols for the analysis of pollen and spores from lakes and bogs. Various types of coring devices are used to take vertical sections of lake and bog sediments. Samples of sediment are treated with a series of chemical reagents that digest most mineral and organic matter while preserving the pollen and spores. The grains are then identified and counted using a microscope.
The pollen of most boreal trees can be identified to the level of genus although, in cases such as Picea glauca and P. mariana, the pollen is distinctive enough to allow identification of the species (Hansen and Engstrom, 1985). Chronological control for the pollen records are generally provided by radiocarbon dating.
The proportional mix of vegetation is difficult to determine from fossil pollen records. Genera such as Picea and Pinus, which are copious producers of pollen, are usually over represented in pollen records as opposed to the genera Abieis and Larix, which produce smaller amounts of pollen. The spatial sampling area represented in the fossil pollen records from lake sediments is inversely proportional to the settling velocity of the pollen grain and directly related to the radius of the lake or bog from which the sample is taken (Prentice, 1988).
FIGURE 1. Information provided by fossil pollen studies can provide insights that are unavailable from other sources on long-term changes in boreal forest composition. Fossil pollen from moderate-sizes lakes in the western Northwest Territories (A) show that the regional population size of Picea spp. Has been relatively constant over the past 4000 to 5000 years; fossil pollen data from the central Yukon (B) shows that the population of Pinus spp. have increased rapidly over the past 2000 years; the fossil pollen from a very small lake in northern Alberta (C) indicates that the abundance of Picea spp. in the local vegetation has varied as a consequence of fires.
A Lac Demain, Northwest Territoires, Canada
B Buggy Pond, Yukon Territory, Canada
C Rainbow Lake "A", Alberta, Canada
Pinus spp. pollen is generally light and readily transported by high-altitude winds. It has been calculated that the sampling area for 70 percent of the Pinus spp. pollen deposited in a small lake (300 m radius) may extend 59 km radially from the site (Prentice, 1988). By contrast, the sampling radius for 70 percent of the heavier pollen produced by Picea spp. may extend only 2 km from the same lake (Prentice, 1988).
Because the pollen deposited in lakes provides an integrated picture of the vegetation conditions within areas ranging from tens to thousands of km2, the spatial sampling qualities of pollen records have been likened to those of remote sensing data (Webb, Laseki and Bernabo, 1978). In the case of satellite imagery, each pixel provides an integrated sample of spectral reflectance from vegetation and other cover types over an area of 100 m2 to more than one million m2. However, pollen records can provide continuous monitoring of vegetation conditions within their sampling area for thousands of years while satellite coverage only goes back to the 1960s.
Fossil pollen records from sites in western and eastern Canada, Finland and Siberia (Fig. 2) give a perspective on the development of boreal plant communities following the last glacial period approximately 10000 years ago. These pollen records demonstrate that development of the boreal forest has followed very different patterns in each of these regions.
The first tree genus to appear in postglacial pollen records from the boreal region of western Canada is Populus (Fig. 2A - MacDonald, 1987). Picea glauca and P. mariana assumed their present dominance of the vegetation between 9000 and 8000 yrs BP (yr BP = 14C years before AD 1950). Betula papyrifera also became an important component of the forest at that time. The last tree genus to reach the region was Pinus (P. contorta ssp. latifolia and P. banksiana). Pines became important components of the vegetation between 7000 and 4000 yrs BP. In northeastern Canada, the first vegetation in the southern boreal region was characterized by herbs and shrubs (Fig. 2B - Engstrom and Hansen, 1985). The genus Picea did not come to dominate the vegetation until 8000 to 7000 yrs BP, with P. mariana replacing P. glauca as the dominant tree by 6000 yrs BP. Abies balsamea also became an important component of the vegetation at that time.
In eastern Finland (Fig. 2C - Huttunen and Stober, 1980) Betula spp. dominated the vegetation by 9000 yrs BP. Pinus sylvestris became the predominant tree species in the pollen record by 8500 yrs BP. Picea abies only became an important component of the vegetation at 5000 yrs BP.
The early postglacial vegetation in western Siberia (Fig. 2A - Khotinskiy 1984) included the genera Larix and arboreal species of Betula. Between 8 000 and 5 000 yrs BP the importance of the genera Abies and Pinus increased greatly at the expense of Larix and Picea.
From these data it is clear that the boreal forest in all of these regions evolved into its present form over the past 5000 years. It is noteworthy that all four of the sites sampled are in central portions of the boreal forest, in terms of postglacial vegetation change, and are relatively quiescent compared with sites at the southern and northern peripheries. In North America, Europe and Asia, the edges of the boreal forest have experienced pronounced changes in vegetation as a result of climatic variations over the past 10000 years (Hyvarinen, 1976; Khotinskiy, 1984; Ritchie, 1987). The factors that may have controlled forest development during the postglacial period remain the subject of debate; however, recent work suggests that, on time-scales of thousands of years and on a sub continental spatial scale, climate is likely to have been the primary factor determining rates and patterns of boreal forest development (Prentice, Bartlein and Webb, 1991). It is also likely, therefore, that the response of boreal vegetation to predicted future climate change will present a great deal of regional variability in patterns and rates of community change.
Lac Demain, Northwest Territoires, Canada - A
Moraine Lake, eastern Canada - B
Nizhnevartovskoye Bog, Siberia - D
Note: yr BP = 14C years before AD 1950.
Source: adapted from Mac Donald, 1987 (A); Engstrom and Hansen, 1985 (B); Huttenen and Stober, 1980 (C); Khotinskiy, 1984 (D).
FIGURE 3. The northward movement of the genera Picea and Pinus into the present boreal region of eastern North America (A,B); the northwestward movement of Picea abies and Pinus silvestris into Fennoscandia (C,D)
Northward movement of Picea spp. - A
Northward movement of Pinus banksiana and P. resinosa - B
Northward movement of Picea abies - C
Northward movement of Pinus silvestris - D
Source: adapted from Davis, 1981 (A, B); Huntley and Birks, 1983 (C, D)
Geographic networks of fossil pollen sites have been used to provide mapped reconstructions of past changes in the geographic ranges of boreal tree species in western Europe and North America (Fig. 3). The most significant aspect of these changes has been the northward spread of some tree species following the end of the last glacial period. These data are important for boreal forest management for two reasons.
First, fossil pollen data provide evidence of the rates at which individual tree species, possessing certain levels of intraspecific variation, can shift their ranges in response to environmental change. Estimates for the maximum rate of northward spread of Picea spp. after the end of the last glacial period range from 376 m/yr in eastern North America (Delcourt and Delcourt, 1987) to 500 m/yr in northwestern Europe (Huntley and Birks, 1983). The maximum migration rate for northern pines (Pinus banksiana, P. strobus, P. resinosa) in eastern North America has been estimated to be 613 m/yr (Delcourt and Delcourt, 1987). In northwestern Europe the maximum migration rate for P. sylvestris has been estimated to be 1500 m/yr (Huntley and Birks, 1983).
Davis (1989) has pointed out that, although these rates of migration appear rapid, climate change from the greenhouse effect may require range adjustments at rates of 3000 m/yr.
Larch (Larix spp.)
Second, pollen data provide evidence of great interspecific differences in the timing of the northward spread of boreal trees. In western and eastern North America (Fig. 3A) Picea mariana and P. glauca were able to occupy much of their present range relatively rapidly following deglaciation (Davis, 1981; Delcourt and Delcourt, 1987; MacDonald, 1987). By contrast, the arrival of Pinus spp. at its northern limits in eastern North America occurred much later (Fig. 3B). It is likely that Pinus contorta ssp. latifolia is still migrating northward in northwestern Canada (MacDonald and Cwynar, 1985). In Fennoscandia (Denmark, Finland, Norway, Sweden) Pinus sylvestris expanded to its present limits relatively soon (Fig. 3D) after the end of the glacial period. By contrast, Picea abies reached its modern northwestern limits in Scandinavia only in the last 500 years (Fig. 3C). There is still uncertainty regarding the factors that controlled these intraspecific and interspecific rates of migration. One explanation for such continental patterns is that, owing to differences in genetic make-up and levels of adaptability, each species reacted to climate changes in a highly individualistic manner (Huntley and Webb, 1989). This would suggest that future climate change might lead to similar individualistic adjustments in the range limits of boreal tree species.
Fossil pollen data have long been scrutinized by forest geneticists in order to obtain information on the relationship between the Pleistocene history of boreal tree species and the present distribution of genetic variation in these species.
Patterns of genetic diversity in North American genera such as Picea, Pinus and Abies, and the evolutionary divergence of closely related species such as Pinus contorta and P. banksiana have been attributed to range disruptions caused by Quaternary glaciations (Critchfield, 1984) which split what were once contiguous populations into disjunct eastern and western or southern and northern populations. However, recent fossil pollen research in western Canada led to the conclusion that the low degree of genetic variability typical of Yukon populations of Pinus contorta ssp. latifolia could not be attributed to the isolation of pines in the Yukon during the last glacial maximum. Indeed, fossil pollen evidence indicates that Pinus contorta was completely absent from the Yukon until approximately 4000 years ago (MacDonald and Cwynar, 1985).
Pollen deposited in boreal forest lakes and bogs can provide a record of past species composition
It is possible that the low genetic variability of the Yukon populations was a result of the mode in which the pines migrated northward following the last deglaciation. The northward migration of the pines into western Canada was probably accomplished by the establishment of small outliers in advance of the main population. These small. isolated populations, in turn, may have served as the seed source for population growth at the northern edge of the species range limit and for the establishment of further outliers. Stochastic losses of genetic variability related to each of these founding events could have produced the northward gradient of decreasing genetic variability observed in Pinus contorta ssp. latifolia (Cwynar and MacDonald, 1987). This mode of migration may have also led to selection for the light, readily transported, seeds typical of Yukon populations of pines (MacDonald and Cwynar, 1987).
The research outlined above has two important implications for understanding changes in boreal forest genetic structure. First, the significant genetic difference between Yukon populations and more southerly populations have developed over a period of less than 4000 years. This suggests that significant and widespread differences in genetic structure can occur relatively rapidly without ice age disruptions to ranges. Second, if the loss of genetic diversity or the selection of light seeds resulted from the mode of northward migration, similar losses and/ or morphological changes are likely to be observed if plant population ranges alter rapidly in response to future global warming.
Fossil pollen data provide a means of identifying changes in the composition and relative abundance of species in boreal forest vegetation over periods of millennia. The spatial sensing properties of these data can be roughly compared with information from satellite remote sensing. Fossil pollen records provide an important source of baseline data for interpreting changes in boreal forests.
Boreal forest plant communities have reached their present configuration in the past 5000 years or less. There are large regional differences in the patterns by which boreal communities developed. Future changes in boreal forests may be regionally variable and may produce community structures that lack modern counterparts.
The rates and patterns of change in the ranges of boreal trees since the end of the last glacial period have displayed high degrees of interspecific variability. This variability reflects highly individualistic responses of species to the environment. The response of tree species to future environmental change will likely display a similar high degree of interspecific and intraspecific variability.
Observable changes in the large-scale distribution of genetic diversity have occurred in some tree species since the last glaciation and it is possible that future e environmental changes could promote significant losses in the genetic diversity of boreal plant species.
Critchfield, W.B. 1984. Impact of the Pleistocene on the genetic structure of North American conifers. Proc. North American Forest Biology Workshop. 8: 70-118.
Cwynar, L.C. & MacDonald, G.M. 1987. Geographical variation of lodgepole pine in relation to population history. Am. Nat., 129: 463-469.
Davis, M.B. 1981. Quaternary history and stability of forest communities. In D.C. West H.H. Shugart & D.B. Botkin eds. Forest succession: concepts and application. New York, Springer-Verlag.
Davis, M.G. 1989. Lags in vegetation response to greenhouse warming. Climatic Change, 15: 75-82.
Delcourt, P.A. & Delcourt, H.R. 1987. Long term forest dynamics in the temperate zone. New York, Springer-Verlag.
Emanuel, W.R., Shugart, H.H. & Stevenson, M.P. 1985. Climate change and the broad - scale distribution of terrestrial ecosystem complexes. Climatic Change, 7: 29-43.
Engstrom, D.R. & Hansen, B.C.S. 1985. Postglacial vegetation change and soil development in southeastern Labrador as inferred from pollen and chemical stratigraphy. Can. J. Bot., 63: 543-561.
Hansen, B.C.S. & Engstrom, D.R. 1985. A comparison of numerical and qualitative methods of separating pollen of white and black spruce. Can. J. Bot., 63: 2159-2163.
Huntley, B. & Birks, H.J.B. 1983. An atlas of past and present pollen maps for Europe: 0-13000 years ago. Cambridge, Cambridge Univ. Press.
Huntley, B. & Webb III, T. 1989. Migration: species' response to climatic variations caused by changes in the Earth's orbit. J. Biogeog., 16: 5-19.
Huttunen, P. & Stober, J. 1980. Dating of palaeomagnetic records from Finnish lake sediment cores using pollen analysis. Boreas, 9: 193-202.
Hyvarinen, H. 1976. Flandrian pollen deposition rates and tree line history in northern Fennoscandia. Boreas, 5: 163- 175.
Kauppi, P. & Posch, M. 1985. Sensitivity of boreal forests to possible climatic warming. Climatic Change, 7: 45-54.
Kauppi, P. & Posch, M. 1988. A case study of the effect of CO2- induced climatic warming on forest growth and the forest sector: A. Productivity reactions of northern boreal forests. In M.L. Parry T.R. Carter & N.T. Konjin, eds. The impacts of climatic variations on agriculture, p. 183- 195. Dordrecht Kluwer Academic Publishers.
Khotinskiy, N.A. 1984. Holocene vegetation history. In A.A. Velichko, H.E. Wright, Jr. & C.W. Barnosky eds. Late Quaternary environments of the Soviet Union, p. 179-200. Minneapolis Univ. Minnesota Press.
MacDonald, G.M. 1987. Postglacial vegetation history of the Mackenzie River basin. Quaternary Res., 28: 245-262.
MacDonald, G.M. & Cwynar, L.C. 1985. A fossil pollen-based reconstruction of the late Quaternary history of lodgepole pine (Pinus contorta ssp. latifolia) in the western interior of Canada. Can. J. Forest. Res., 15: 1039-1044.
Prentice, I.C. 1988. Records of vegetation in time and space: the principles of pollen analysis. In B. Huntley & T. Webb III, eds. Vegetation history, p. 17-42. Dordrecht, Kluwer.
Prentice, I.C., Bartlein, PJ. & Webb III, T. 1991. Vegetation and climate in eastern North America since the last glacial maximum. Ecology, 72: 2038-2056.
Ritchie, J.C. 1987. Postglacial vegetation of Canada. Cambridge, Cambridge Univ. Press.
Webb III, T., Laseki, R.A. & Bernabo, J.C. 1978. Sensing vegetational patterns with pollen data: choosing the data. Ecology, 59: 1151-1163.
