L. Dennis Gignac

Bryophytes as Indicators of Climate Change

Interest in climate change has increased tremendously in the past 10 to 15 yr, both within and outside the scientific community. The reason for this interest is directly related to the anticipated global warming that will result from increased concentrations of greenhouse gases in the atmosphere. As a result of this interest, several questions have been raised relative to climate warming. For example, how can we predict long term climatic change? How accurate are the predictions? What will be the severity and extent of the changes? How will biodiversity, ecosystems, and habitats be affected if climate change occurs as predicted? How long will it take for species and ecosystems to react to climate change? This essay will focus on the utilization of bryophytes to answer those questions. Bryophytes grow in almost all terrestrial and freshwater environments where plants can be found. These environments have a global distribution and are found in all climatic regimes with the exception of those on permanent ice. The success of bryophytes is largely due to their unique and very effective physiological water relation system that permits them to survive in the wide variety of climates in which they are found. This poikilohydric system permits them to grow during periods when water is available and to suspend their metabolism when water is lacking. Most genera are ectohydric and take up water through the whole surface of the plant and therefore do not need a root system to draw water from the soil. Also, nutrients are taken up through all surfaces from solutes in water that is in contact with the plants. As a result, bryophytes can grow on such very hard surfaces as rocks and tree trunks where higher plants cannot because their roots cannot penetrate the surface.

Habitat Limitations of Sphagnum along Climatic, Chemical, and Physical Gradients in Mires of Western Canada

Sphagnum distribution was studied on twenty-seven peatlands found along a transect extendingfrom the Queen Charlotte Islands, British Columbia, to central Alberta. Based on surface water chemistry, oceanic mires are either ombrotrophic bogs or poor fens, while in subcontinental areas, mires range from extreme-poor fens to moderate-rich fens. Species are grouped into fJive clusters; stands into ten. Species groups and stand dispersal are determined by climate and surface water chemistry, especially corrected conductivity and calcium, magnesium, and potassium con- centrations. Sphagnum species habitats are limited to mires having low cationic contents and corrected con- ductivities. Seven of eighteen species studied are limited by climatic factors to oceanic areas. Sphag- num fuscum is the most widespread of all the species studied, independent of climate and surface water chemistry. Only three Sphagnum species are present in moderate-rich fens. Most of the species height requirements along the hummock-hollow gradients are present on the mires studied, and the height relative to the water table does not limit the geographic distribution of these species. Height is limiting only for species found at either end of the topographic gradient. Tree-produced shade does not limit the habitat of any of the species studied.

Sphagnum-dominated Peatlands in North America Since the Last Glacial Maximum: Their Occurrence and Extent

Abstract Sphagnum-dominated peatlands occupy extensive tracts of land throughout the Boreal and Subarctic regions of North America, extending north onto the Low Arctic of the Canadian Shield and south along the west coast of Oregon, Rocky Mountains of Wyoming, and Appalachians of West Virginia. In addition, short pocosins found along the southeastern coast also can be considered as Sphagnum-dominated peatlands, even though they differ significantly from traditional concepts of boreal peatlands. Along the southern limit of Sphagnum-dominated peatlands, where climate is limiting, edaphic factors allow for the development of outliers. As the current distribution of Sphagnum-dominated peatlands is related to Sphagnum spore rain, past distributions of Sphagnum-dominated peatlands can be constructed from spore records preserved in lakes and peatlands. Here we present six time slices extending back to the Last Glacial Maximum to determine how Sphagnum-dominated peatlands have varied in both time and space. The spore record indicates that Sphagnum-dominated peatlands were present in North America during the Last Glacial Maximum although they were spatially limited to central Alaska, the Olympic Peninsula and Puget Trough of Washington, and to a narrow band in the eastern states of Kentucky, North Carolina, Tennessee, and Maryland. During the Late Wisconsinan Sphagnum-dominated peatlands shifted northwards in eastern North America and expanded farther into Alaska and the Midwest. The Late Wisconsinan/Holocene transition marks a time of overall increase in the area supporting Sphagnum-dominated peatlands, while extending farther in eastern Canada and western continental and coastal regions, they almost completely disappear in the Midwest where they were extensive earlier. Sphagnum-dominated peatlands generally reach their current extent about 2,000–3,000 years ago. Sphagnum-dominated peatlands have dramatically changed their distribution and abundance since the Last Glacial Maximum, and hence the carbon that is stored in these present-day important sinks has also changed dramatically. When compared to the estimated 220 Pg of carbon stored in North American peatlands today, less than 10% of this carbon was present in these peatland during the LGM.

Effects of Fragment Size and Habitat Heterogeneity on Cryptogam Diversity in the Low-boreal Forest of Western Canada

Abstract The effects of fragmentation of the sub-humid low-boreal forest in agro environments on bryophyte and lichen diversity were analyzed on 44 woodlots in three regions of northern Alberta, Canada. Woodlots were selected to cover a wide variety of shapes and sizes in each area. Several microclimatic variables, which included temperature at three different heights above the forest floor, relative humidity, and the amount of light penetration were measured in five m diameter circular sample plots along transects that covered the length and the width of each fragment. The number of microhabitats that included living trees with creviced and smooth bark, standing dead, different decay classes of downed woody debris, soil and rock were tabulated at each sample plot. Bryophyte and lichen presence was noted for all microhabitats found within each plot. There was a trend of decreasing temperature and light intensity and increasing humidity up to 15 m from the edge of the fragments. A multiple regression analysis revealed that the distance from edge and the habitat heterogeneity were the most important variables affecting bryophyte and lichen species richness. Further analysis indicated that edge effects were significant for bryophytes, but not for lichens. The size and shape of fragments had a significant effect on habitat heterogeneity, and bryophyte and lichen diversity. Although there was a distinction between the floras in each region, the effects of fragment size and habitat heterogeneity were similar. An indicator species analysis selected several indicators of large, medium, and small fragments that were also indicators of habitat heterogeneity and species richness. The presence or absence of several indicator species is used to produce performance benchmarks for habitat heterogeneity and species richness. The use of indicator species offers an inexpensive and relatively easy method to evaluate edge effects and habitat heterogeneity on bryophyte and lichen diversity in woodlots.

Niche Structure, Resource Partitioning, and Species Interactions of Mire Bryophytes Relative to Climatic and Ecological Gradients in Western Canada

The bryophyte component of mire vegetation was analysed on 27 peatlands located on a transect from the Queen Charlotte Islands, British Columbia to central Alberta. The habitats studied range from hyperoceanic bogs and poor fens to subcontinental moderate-rich fens. Habitat niche dimensions are calculated for 11 Sphagnum and 6 non-Sphagnum species alongfive climatic gradients and such ecological variables as surface water chemistry, height relative to the water table, and overstory cover. Species niche overlaps were calculated along Jive gradients: aridity index, corrected conductivity andpH of mire surface waters, height, and overstory cover. Habitat partitioning by Sphagnum species indicates that climate is the most important factor affecting species niches and overshadows on a regional basis such locally important ecological gradients as surface water chemistry and height relative to the water table. Niches of Sphagnum austinii, S. papillosum, S. rubellum, and S. tenellum are restricted to oceanic areas by the climate. Among non-Sphagnum species, the overstory cover, pH, conductivity, and height gradients are more extensively partitioned than the aridity index gradient. Niche shifts among widespread bog and poorfen species, and species packing along the height gradients, suggest a competitive hierarchy where: 1) Sphagnum species are better competitors than most non-Sphagnum species; 2) lawn and carpet species are better com- petitors than hummock species; and 3) oceanic species are better competitors than widespread species.

The Utilization of Bryophytes in Bioclimatic Modeling: Predicted Northward Migration of Peatlands in the Mackenzie River Basin, Canada, as a Result of Global Warming

A bioclimatic model based on bryophyte species distribution and abundance relative to climate was coupled with climatic and geographical data obtained from Leemans and Cramer (1991) and the Canadian Climate Center (CCC) General Circulation model (GCM) at 1XCO2 in order to reconstruct the present geographical distribution of seven peatland types in the Mackenzie River Basin. The geographical distribution of 195 peatlands previously identified by type were used to test the validity of the reconstructions. The test revealed that the reconstruction using data from Leemans and Cramer was more accurate than the reconstruction using the CCC GCM data. For this reason, the CCC 1XCO2 data was subtracted from the CCC 2XC02 climatic data to produce an anomalies data set which was then added to the Leemans and Cramer data to project the distribution of the seven types of peatlands at 2XCO2. Results of this prediction were then compared to predictions using 2XC02 data obtained from the Geophysics Fluid Dynamics Laboratory (GFDL) GCM. The position of the southern limits of peatland distribution was compared to past distributions resulting from a warming period in the early to mid Holocene. Results of the predictions for the two climate change scenarios indicated a northward migration of the southern boundary of peatland ecosystems of approximately 780 km in the central portion of the Mackenzie River Basin. The model also predicted that Mid-Boreal peatlands would be located along a diagonal running from southeast to northwest from 600 longitude to an area just south of the Mackenzie Delta for both scenarios. High-Boreal and Subarctic peatlands were located to the north of the diagonal, while Low-Boreal peatlands were located to the south. However, the CCC anomalies + Leemans and Cramer predictions did not clearly define the Low-Boreal since Low-Boreal indicators were only located in the Cordilleran Ecological Province. Ecological diversity is anticipated to be maintained in the peatlands because all types were predicted to be found in the Basin at 2XC02 but at different locations. Comparisons between the predicted position of the southern limits of peatland distribution and that during the early to mid Holocene indicate that the models predictions were reasonable. Global warming resulting from increased concentrations of such greenhouse gases as CO2, CH4, and NOx could produce temperature increases as much as 1.5 to 4.60C by the middle of the 21st century (Mitchell et al. 1990). Northern and continental regions in particular will experience significant warming, especially during the spring and winter months. Increases in annual temperature for northern locations in North America are anticipated to be in the order of 3.7 to 4.60C (Cohen 1993). Elevated temperatures are expected to produce droughts that will affect site water balance and runoff, and cause major shifts in the distribution of ecosystems on a global scale (Houghton et al. 1990; Vitousek 1994). These effects will be quite noticeable in such ecosystems as wetlands, and in particular peatlands, which are especially sensitive to climate and fluctuations in water levels. Peatlands are wetlands that have a minimum accumulation of 0.5 m of peat, and are usually divided into two types: bogs and fens. This division is based on whether the water supply is ombrotrophic (bogs) or minerotrophic (Gore 1983; Sjors 1952). The largest concentration of peatlands is in the boreal and subarctic regions of the Northern Hemisphere (Gore 1983). The northern distribution indicates that peatlands are climatically sensitive ecosystems and that they are primarily found in areas that have wet and humid climates where there are no sustained dry periods (Moore & Bellamy 1974). 0007-2745/98/572-587

Peatland species preferences: An overview of our current knowledge base

1.75/0 This content downloaded from 207.46.13.128 on Tue, 06 Sep 2016 05:18:25 UTC All use subject to http://about.jstor.org/terms 1998] GIGNAC ET AL.: PREDICTED MIGRATION OF PEATLANDS 573 Peatland formation is a function of two climatic variables, precipitation and evaporation, and generally they occur where evaporation does not exceed precipitation i.e., where potential evapotranspiration ratios < 1 (Gignac 1993). Not only does the macroclimate affect the presence of peatlands, but it also affects the distribution of different forms of bogs, and the height that they attain above the surrounding landscape (Ivanov 1981; Moore & Bellamy 1974). Among climatic variables that have been correlated to the different types of bogs are: total precipitation (Eurola 1962; Ruuhijirvi 1960), number of precipitation days (Taylor 1983), and precipitation and temperature together (Ivanov 1981; Moore & Bellamy 1974). Fens can also be found in all climatic zones where peatlands occur, and some types of fens can be placed into different zones based on climatic factors (Gignac & Vitt 1990; Moore & Bellamy 1974). Peatland vegetation, particularly bryophyte species, are very sensitive to changes in the height of the water table (Gignac et al. 1991; Gorham & Janssens 1992; Nicholson et al. 1996). The height of the peat surface above the water table and thus the height of the vegetation are a function of climatic conditions, particularly temperature and precipitation (Gignac & Vitt 1990). Increases in temperature enhance evaporation and if precipitation remains constant or diminishes, it should cause a lowering of the water table. A decrease of the height of the water table will result in subtle changes in the presence, absence, and abundance of bryophyte species (Andrus et al. 1983; Gignac & Vitt 1990; Gignac et al. 1991; Vitt et al. 1975). Thus, peatland vegetation is very sensitive to climate. Changes in peatland vegetation should offer an early warning of large-scale climatic changes before they occur in more stable and larger ecosystems such as the boreal forest. Because peatlands are sinks and sources of the greenhouse gases CO2 and CH4, any modification of the climate could affect the quantities of these gases released to the atmosphere (Bubier et al. 1993; Roulet et al. 1993; Roulet et al. 1994). Accumulating deposits of peat retain water causing local water tables to rise (Ingram 1983) which has lead to large scale peatland formation through paludification in northern landscapes (Nicholson & Vitt 1990; Sjors 1963). Under a warmer drier climate methane emissions will initially increase with melting of permafrost and rising of water tables (Liblik et al. 1997). Once new drainage patterns establish, increased oxidation of the peat will release more CO2 to the atmosphere (Gorham 1991). Organic acids are also produced from decomposing mosses and as a result of increased decomposition caused by warmer climates, local streams may become acidified (Bayley et al. 1992; Hogg et al. 1992). The microclimate of the peatland will also change causing a reduction of evapotranspiration rates because of the shift from areas with free standing water to shaded areas as a result of increased tree or shrub cover (Ingram 1983). Subsurface soil temperatures will also increase with the resulting degradation of permafrost in many north-

The Utilization of Bryophytes in Bioclimatic Modeling: Present Distribution of Peatlands in the Mackenzie River Basin, Canada

The autecology of several peatland plant species has been studied on a wide variety of gradients that include elevation relative to the water table, surface water chemistry, mire margin to mire expanse, shade, and climate. The accuracy with which species’ realized niches are defined and the ability to predict community structure as conditions change along gradients vary considerably between studies. Many studies have quantified species niche dimensions along individual gradients, and although they have the ability to predict species abundance and distribution, they do not account for synergistic effects between gradients. Other methods analyze two gradients simultaneously producing rectangles and ellipsoids, but these methods have limited predictive ability. A variety of ordination techniques are often used to analyze species preferences along several gradients simultaneously. However, with the exception of Detrended Canonical Correspondence Analysis, these methods have limited predictive ability because gradients are defined as ordination axes and it is difficult to analyze each gradient individually. Species response surfaces calculated along several gradients simultaneously offer a more accurate definition of species’ realized niche dimensions. Response surfaces can be used to predict baseline community structure along several gradients, but they do not integrate such biotic factors as competition and succession, as well as such disturbances as fire, pollution, and peat harvesting.

Development of Sphagnum-dominated peatlands in boreal continental Canada

A model was developed that classified and projected the distribution of seven different types of peatlands in the Mackenzie River Basin. The model was based on the relationships between bryophyte indicator species, the types of peatlands they characterize, and regional climate. The model used the presence, absence, and abundance of 15 bryophyte indicator species to classify 81 peatlands in the study area into seven groups. Abundance values were calculated for each of the indicator species along three climatic gradients-Mean Annual Temperature (MAT), Mean Annual Total Precipitation (MATP), and Length of the Growing Season (LGS). The percent cover of all species were then ascribed to appropriate combinations of MAP, MATP, and LGS. The result produced a matrix consisting of 4,560 grid nodes where each node was identified by values for each of the three climatic variables and the types of peatlands that could be found at that climate. An independent data set consisting of climatic and ecological values and vegetation cover for 115 sites was used to test the ability and accuracy of the model to classify and project the climatic distribution of the seven peatland groups. The model correctly classified 106 of the 115 sites and of those, correctly projected the distribution of all but five of the test sites. The model accuracy was 70% for six of the seven groups, and > 90% for three of those. The accuracy for the remaining group was 50% and errors were mostly caused by the failure to project the distribution of three of the test sites. Other errors include: the inability to classify lichen dominated peatlands; the inclusion of wet lawns in bogs into one of the groups which caused a southward extension of that group by approximately 200 km. The overall model accuracy was 88%. Peatlands, primarily bogs and fens, cover extensive areas of the boreal and subarctic zones particularly in the northern hemisphere (Gore 1983). Several studies have demonstrated that peatland types and peatland vegetation are to a large extent controlled by macroclimate, physiology, and local gradients. The macroclimate plays an important role in the regional distribution of the different types of peatlands (Moore & Bellamy 1974). Among climatic variables that have been correlated to the different types are total precipitation (Eurola 1962; Ruuhijiirvi 1960), number of precipitation days (Taylor 1983), and precipitation and temperature together (Ivanov 1981; Moore & Bellamy 1974). The present distribution of peatlands in North America is a function of two climatic variables, precipitation and potential evaporation. Generally peatlands only occur in areas where evaporation does not exceed precipitation (Gignac 1993). Because bryophytes have a high fidelity to such environmental gradients as moisture and pH, they are particularly good indicators of different types of peatlands at the local level (Gignac & Vitt 1990; Horton et al. 1979; Nicholson et al. 1996; Vitt et al. 1975; Vitt & Slack 1975, 1984). The distributions of several of those species are also limited by climate and thus become indicators of climate. For example, in western North America the distribuion several such species as Sphagnum austinii, S. papillosum, S. rubellumn and S. tenellum are limited by climate to areas that have > 1,000 mm total precipitation and are thus indicative of oceanic peatlands (Gignac et al. 1991a). Others, such as Drepanocladus fluitans, Sphagnum lenense and S. riparium, are limited to areas having mean annual temperatres lower than O0C and are indicative of high boreal and subarctic peatlands (Nicholson & Gignac 1995). Also, several geographically widespread species can be indicators of peatlands based not on their presence and absence but on their abundance. The primary division of peatlands into bogs and 0007-2745/98/560-571

Bryophyte response surfaces along climatic, chemical, and physical gradients in peatlands of western Canada

1.35/0 This content downloaded from 157.55.39.225 on Tue, 13 Jun 2017 18:01:52 UTC All use subject to http://about.jstor.org/terms 1998] GIGNAC ET AL.: PRESENT DISTRIBUTION OF PEATLANDS 561 fens is based on the source of surface water. The supply of water in ombrotrophic peatlands (bogs) is entirely from precipitation and because of this, the pH of bog surface water is low (Sjirs 1952). In contrast, a portion of the surface water on minerotrophic peatlands (fens) has been in contact with mineral soil. Depending on the nature of the mineral substratum and the quantity of water in the peatland that has flowed over the substratum, the pH of surface water in fens can vary considerably. Bogs are usually classified according to their landform patterns into raised bogs, plateau bogs, eccentric-domed bogs, concentric-domed bogs, and blanket bogs (Glaser & Janssens 1986; Moore & Bellamy 1974). Fens are usually divided according to the species richness of the vegetation into poor and rich (Sjirs 1952). Associated with species richness are differences in pH of the surface water: poor fens have pH values usually below 5.6, while rich fen surface water pH values are usually circumneutral or higher. There are further subdivisions of the rich fen category into moderate and extreme, again depending on the species richness of the vegetation (Sjors 1952, 1963). Other types of fens are aapa and palsa fens, which are classified according to landform and can be either rich or poor. Aapa fens have patterns composed of strings (ridges) and flarks (pools), while palsa fens have mounds (palsas) that contain a core of permafrost (Moore & Bellamy 1974). The distribution of the different types of bogs is related to macroclimate (Moore & Bellamy 1974). Ivanov (1981) for example, determined that the maximum height of a bog is in part a function of the moisture surplus during the driest month. Moisture surplus is the excess of precipitation over evapotranspiration (Damman 1986). The convexity of the dome is also a function of the water balance for the driest month (Ivanov 1981). The surface of bogs in hyperoceanic and oceanic areas may be raised several meters above the water table because there is always a large monthly moisture surplus (Gignac & Vitt 1990). Conversely, continental bogs are only slightly raised above the water table because there are several months during the growing season when there is relatively little precipitation. Although fens can be found in all climatic zones where peatlands occur, some types of fens can be placed into different zones based on climatic factors. Aapa (string) fens are found in continental areas where precipitation is low and peatlands form in basins and are thus more susceptible to influence by mineral soil water. Palsa fens are located in high boreal and subarctic regions that have discontinuous permafrost (Moore & Bellamy 1974). Generally, rich fens, particularly extreme-rich fens, are located in subcontinental and continental areas and are rare or absent in hyperoceanic and oceanic areas (Gignac et al. 1991b; Gignac & Vitt 1990; Vitt et al. 1990). The Mackenzie River Basin provides an excellent geographical area in which to study the effects of cl mate on the distribution of peatlands. It is a large basin that is restricted to continental areas and has wide variations along several climatic gradients, particularly temperature. Also, and most importantly, its southern boundary generally corresponds to the southern limit of peatland development and, with a few exceptions, its northern boundary coincides with the northernmost development of extensive peatlands in western Canada. A previous analysis of the vegetation of peatlands in the Mackenzie River Basin (Nicholson et al. 1996) indicated that the distributions of several peatland types and their bryophyte indicator species were closely related to climate. Those results led to the development of a model that reconstructs the present distribution of peatlands in the Mackenzie River Basin based only on climatic variables. The purpose of this study is to delimit the climatic distribution of each type of peatland that was found in the study area. The following steps were used to model peatland distribution 1) relate bryophyte indicator species to different types of peatlands and the climate in the Mackenzie River Basin; 2) project the present climatic distribution of those types of peatlands; and 3) test the models validity and projections using an independent data