Moss Polsters Research Articles

Surface pollen deposition on glacier forelands in southern Norway I: local patterns of representation and source area at Storbreen, Jotunheimen

The relationship between vegetation and surface pollen deposition is examined at Storbreen glacier foreland where a clear plant succession exists. The aim is to determine whether the distinct plant communities present produce characteristic pollen assemblages. The influence of environmental factors is also considered. Pollen assemblages from moss polsters, collected from 22 paired sampling sites across the foreland, are compared with local vegetation. Two-way indicator species analysis and detrended correspondence analysis are employed to identify clusters and sequences, initially in the vegetation data and subsequently in the pollen data sets. Vegetation and pollen data are compared simultaneously using canonical correspondence analysis. Three main plant communities are distinguished: pioneer, heath and snowbed. Broadly, each community produces characteristic pollen assemblages. Boundaries between groups are not clear-cut, reflecting the mosaic of plant communities present. Recognition of distinct plant communities is hampered by the prevalence of long-distance arboreal pollen and poor representation of entomophillously pollinated taxa. Late in the succession up to 78% of pollen could originate locally. Use of the non-arboreal pollen sum significantly improves correspondence with vegetation. The importance of indicator taxa is considered and both Salix and Empetrum are found to distinguish successfully early phases of succession from later phases. Strong correlations exist between the primary ordination axes of vegetation and pollen and with terrain age and altitude (for example, the correlation between altitude and total land pollen Axis 1 is r= -0.76). The surface data add new information to the interpretation of tree colonization in the area during the Holocene.

Modern pollen–vegetation relationships in subarctic southern Greenland and the interpretation of fossil pollen data from the Norselandnám

AbstractAim The objective of this paper is to explore the relationships that exist between vegetation and modern pollen rain in the open, largely treeless landscape of subarctic Greenland. The implications of these results for the interpretation of fossil pollen assemblages from the time of the Norselandnámare then examined.Location The study area is the sheep farming district of Qassiarsuk in the subarctic, subcontinental vegetational and climatic zone of southern Greenland (61° N, 45° W). Betweenc.ad1000–1500 this region was contained within the Norse Eastern Settlement.Methods Detrended Correspondence Analysis (DCA) of harmonized plant–pollen data sets is used to compare plant cover in 64 vegetation quadrats with pollen assemblages obtained from moss polsters at matching locations. Presence/absence data are also used to calculate indices of association, over‐ and under‐representation for pollen types.Results Good correspondence between paired vegetation–pollen samples occurs in many cases, particularly in locations whereSalix glauca–Betula glandulosadwarf shrub heath is dominant, and across herbaceous field boundaries and meadows. Pollen samples are found to be poor at reflecting actual ground cover where ericales orJuniperus communisare the locally dominant shrubs. Dominant or ubiquitous taxa within this landscape (Betula,Salixand Poaceae) are found to be over‐represented in pollen assemblages, as are several of the ‘weeds’ generally accepted as introduced by the Norse settlers.Main conclusions Due to their over‐representation in the pollen rain, many of the Norse apophytes and introductions (e.g.Rumex acetosaandR. acetosella) traditionally used to infer human activity in Greenland should be particularly sensitive indicators forlandnám, allowing early detection of Norse activity in fossil assemblages. Pteridophyte spores are found to be disassociated with the ground cover of ferns and clubmosses, but are over‐represented in pollen assemblages, indicating extra‐local or regional sources and long residence times in soil/sediment profiles for these microfossils. A pollen record forHordeum‐type registered in close proximity to a field containing barley suggests that summer temperatures under the current climatic regime are, at least on occasion, sufficient to allow flowering.

Modern pollen from the catchment and surficial sediments of a Scottish sea loch (fjord)

Modern pollen was extracted both from moss polster samples collected from a range of sites across the land surface of the Loch Sunart catchment, north‐west Scotland and from a number of marine sediment‐water interface sites in the sea loch. Comparisons between the pollen results and the major vegetation types growing in the catchment area revealed that in general the moss polsters contained a localised picture of the vegetation whilst the sea loch sediments varied much less from sample to sample and better represented a regional picture of the vegetation. It was anticipated that the pollen in the sea loch samples would be in a much poorer state of preservation because of the many pathways through which it travels in order to become incorporated into the sediments of the loch. However, this proved not to be the case and the study demonstrates that marine sediments of the kind found in Loch Sunart have the potential to provide a new source of Holocene vegetation data which is as good as the freshwater lake sediments that have been preferentially sampled in the past.

Iron age to medieval entomogamous vegetation and Rhinolophus hipposideros roost in South-Eastern Wales (UK)

Karst cave systems are well developed in Wales (UK) and, in some instances, constitute important bat roosts. Ogof Draenen, near Blaenavon in south-east Wales, is the most recent major cave discovery (1994) with already > 70 km of passages explored spanning a vertical range of 148 m. With the exception of one small chamber (Siambre Ddu) located directly above the main Ogof Draenen system, very few bats have been noticed inside. Extensive accumulations of guano, attributable to Rhinolophus hipposideros, are however found in parts of the Ogof Draenen system. In places covering many square metres and sometimes building heaps > 0.5 m thick, these represent volumes not yet found in any other cave system in the British Isles. Although the date of the abandonment of the main Ogof Draenen system as a bat roost remains unknown, six radiocarbon dates on guano from Ogof Draenen place the occupation in the Iron Age to Medieval period at least. Palynological analysis was undertaken on ten samples distributed through the cave. Comparisons were made with a moss polster and a lake mud sample from the area to provide a first approximation of the regional modern pollen rain and with two modern guano samples, one from Siambre Ddu and one from Agen Allwedd cave (5 km to the north-west) to provide a temporal comparison with the fossil guano. Agen Allwedd cave currently is one of the largest active roosts for Lesser Horseshoe bats in Britain and lies close to the present northern limit of this endangered species in Europe. The main results are that (1) the cave appears to have been used both as a summer and a winter roost; (2) most of the Ogof Draenen guano is formed within c. 1600 14C years and, if the largest heap is continuous, it has accumulated within 750 14C years, i. e. 0.16 mm yr − 1 ; 3) the fossil guano samples reflect a relatively closed oak forest with more abundant ivy ( Hedera) and holly ( Ilex) than at present; (4) insect-pollinated plants such as Ilex, Acer, Hedera and Impatiens glandulifera are over-represented in the guano samples; (5) in addition to the usual causes of bat roost decline (pesticides, pollution), in the case of Ogof Draenen, we may add entrance blocked by rock collapse and decline of the local forest cover as well as change in its composition.

Modern pollen/vegetation/land-use relationships in mountain environments: an example from the Champsaur valley (French Alps)

This study aims at elucidating modern pollen spectra/environmental data relationships from both natural and human-induced vegetation types as an aid for palaeo-ecological reconstructions. A set of 51 surface moss polsters was sampled from different vegetation and land-use types in the Champsaur area (French Alps) and analysed to obtain modern pollen analogues of ancient cultural landscapes in mountain ecosystems. Samples were selected from grazed areas, mown meadows, cultivated fields, fallow land and deciduous and coniferous forests. Vegetation composition around the sampling points and seventeen types of environmental variables (e.g., management type, soil and topography) were collected all for these 51 sites. Patterns of modern local pollen variation in relation to the environmental variables were explored by means of canonical correspondence analysis (CCA) and associated statistical tests. This correlative model allows us to determine the major explanatory variables and to identify taxa indicative of particular anthropogenic activities, and thus may help to calibrate fossil pollen assemblages.

Modern pollen–vegetation relationships in the Champsaur valley (French Alps) and their potential in the interpretation of fossil pollen records of past cultural landscapes

This study is an attempt to evaluate the relationships between the vegetation and the modern pollen rain as a contribution to palaeoecological research. Pollen analysis of surface moss polsters and floristic records has been undertaken for 51 sampling points distributed all over the study area (Champsaur valley, Hautes-Alpes, France), within different vegetation and land-use types and along a west–east altitudinal transect, ranging from 870 to 2200 m a.s.l. The pollen and vegetation data were analysed independently using numerical methods (clustering and correspondence analysis) to investigate how vegetation is reflected in pollen assemblages. There was a good agreement between classifications and ordinations of the two data-sets and a pollen-analytical separation of different types of human activities was found despite the major gradients related to altitude and soil moisture. Both the AP/NAP ratio and the major pollen type percentage ranges given for each vegetation type were however very wide because of the typical fine-scale mosaic of the landscape in this region. Detailed comparisons of the two data-sets showed consistent differences between vegetation and pollen assemblages. Misclassifications of some modern pollen spectra were mainly attributed to differential pollen representation between species, but also to the effect of various land-use practices on flowering and pollination of herbaceous plant taxa. Moreover these differences, as well as discrepancies with earlier published data from lowlands of southern France and other parts of western Europe, are promoted by complex pollen dispersal which characterizes mountain environments. Although an overlap of characteristics thus exists between pollen assemblages, we were able to recognize specific features and indicator pollen taxa have been identified for each natural, semi-natural and human-induced vegetation type.

The interpretation of fen carr pollen diagrams: The representation of the dry land vegetation

Deposits of fen carr origin are frequently used in pollen-based studies, including those where the primary objective is reconstructing the Holocene history of dry land vegetation. However, there have been few studies of modern pollen–vegetation relationships in these circumstances. Here as an aid to the interpretation of fossil assemblages we determine the distribution of pollen derived from dry land sources within areas of fen carr. In particular, we examine the effect distance from the dry land/fen carr edge has on the pollen representation of the dry land vegetation components. Data on vegetation composition and modern pollen samples (from moss polsters) were collected from a series of sample points within two areas of fen carr and adjacent drier woodlands in Norfolk, UK. The spatial distribution of the pollen derived from dry arboreal elements is explored through trend surfaces and plots of pollen representation against distance away from the dry land/fen carr edge. The pollen assemblages from the fen carr sample points reflect and can be used to infer the composition of adjacent dry woodland areas where they are dominated by the relatively high pollen producers Quercus and Betula. The presence of dry taxa that occur away from the dry land/fen carr edge is difficult to detect even if these taxa are generally regarded as over-represented in the pollen record (e.g. Pinus sylvestris). Arboreal elements that are present in the vegetation close to the edge of the dry land show a general declining trend in pollen representation within the first 80 m of fen carr. Factors influencing this pattern include the variable input of pollen from wetland taxa and the spatial irregularity of the dry land/fen carr edge. Pollen recruitment into fen carr systems resembles other closed canopy sites. Palaeo-sites situated within c. 80 m of the dry land/fen carr edge afford the opportunity to reconstruct vegetation at the stand scale. However, adjustments need to be made for differential pollen dispersal, for example, the local presence of Fraxinus excelsior and Sorbus-type can only be detected within c. 20 m of the dry land edge. The interpretation of pollen sequences must also allow for changes in the position of this edge through time.

Estimating the spatial scale of pollen dispersal in the cultural landscape of southern Sweden

The primary aim of the study was to estimate the spatial scale of pollen dispersal and deposition for pollen assemblages from moss polsters in the cultivated landscape of southern Sweden, as a mean to improve future studies of the pollen/vegetation relationship in the region, and interpretation of fossil pollen data in terms of past cultural landscapes. This can be done by estimating the ‘relevant source area of pollen’ (RSAP) defined as the area around the pollen sampling point beyond which the pollen-vegetation relationship does not improve. Forty-two sites from nonfertilized grasslands in the traditional open agricultural (Open Region) and semi-open forested (Semi-Open Region) regions of southern Sweden were selected. The vegetation survey was performed within a 1500 m radius area around the moss polsters sampling area. The extended R-value (ERV) model was used to evaluate the pollen-plant abundance relationship. The RSAP for moss polsters in the Open Region was estimated to c. 400 m from empirical data. In the Semi-Open Region, however, the likelihood function score, an indicator of the goodness-of-fit of the data to the ERV model, showed an unexpected pattern of change, making it difficult to evaluate the RSAP. Simulations using hypothetical landscapes suggest that systematic selection of sampling sites could cause this pattern. Simulations also demonstrate that the size of vegetation patches affect the RSAP, i.e., the larger the vegetation patches are, the larger the RSAP becomes. Similar RSAP for the Open and Semi-Open Regions is obtained in simulations using the same patch size, and random selection. In the actual vegetation, patch size is comparable in the two regions, which would suggest that the RSAP for moss polsters in the Semi-Open Region is c. 400 m as well.

How open were European primeval forests? Hypothesis testing using palaeoecological data

Summary The Vera hypothesis that large herbivores maintained an open landscape in the primeval landscape of lowland Europe is tested using palaeoecological data. The hypothesis suggests that the high abundance of Quercus and Corylus apparent in European pollen diagrams could not have derived from a landscape dominated by closed canopy forests. If natural forest landscapes were indeed more open, current forest conservation management policy across Europe would need to be reconsidered. Relative proportions of Quercus and Corylus pollen are compared from regions which supported large herbivores with data from Ireland, where large herbivores were excluded. Similarity between the two data sets indicates that large herbivores were not required to maintain these taxa in the primeval landscape. Fine spatial resolution pollen data from small hollows in Europe and eastern USA were also reviewed. Data from moss polsters show that percentage arboreal pollen is a reliable indicator of canopy openness in theses sites. The palaeoecological data demonstrate that open canopy forest has only ever been maintained by human exploitation. Large herbivores in Europe do not therefore appear to have maintained an open landscape in primeval times although evidence suggests that they would have influenced the species composition of the forest canopy. It is proposed that data sources other than pollen should be explored as proxies for past forest structure to test this hypothesis more thoroughly.

The interpretation of fen carr pollen diagrams: pollen–vegetation relationships within the fen carr

Sediments deposited in fen carr environments are frequently used in pollen-based reconstructions of vegetation history, although few studies of modern pollen–vegetation relationships in fen carr have been undertaken to aid the interpretation of fossil assemblages. It is often assumed that the wetland pollen signal will reflect the overall composition of the fen carr system although, given the closed canopy context, it may actually be dominated by local vegetation elements. This paper seeks to determine whether the vegetation heterogeneity within a modern fen carr is reflected in the pollen signal of the wetland taxa. Vegetation composition was recorded around a series of moss polster sample points at Calthorpe and Wheatfen, two areas of fen carr in the Norfolk Broadland, UK. The spatial interrelationship between the wetland components of the vegetation and their palynological equivalents in the moss polster pollen assemblages is explored through trend surfaces and a correlation based statistical test (the Mantel Test). The representation of the major vegetation components in the pollen record is highly variable. Values for Alnus glutinosa, a dominant canopy species, range from 5.3% to 73% of total land pollen (TLP) at Calthorpe and 2.5–61% TLP at Wheatfen. In part, this variation reflects the enhanced input of pollen from dry land sources close to the wetland edge although the variation recorded at points equidistant from the dry land margin suggests that other processes are also influential. The strongest positive associations in the Mantel Test are for under-storey and ground flora elements that are poorly represented in the pollen record. Taxa for which there is no association include those with few vegetation occurrences and those for which a large proportion of the pollen appears to have been derived from outside the study areas. We conclude that modern pollen samples collected from fen carr broadly reflect the vegetation composition of the study areas. However, the input of pollen from fen carr species is spatially inconsistent. Differences between the sample points can be attributed to the distribution and pollination biology of the under-storey and ground flora elements and the structure of the vegetation around sites. The results caution against assuming that changes in the pollen representation of wetland taxa, in fossil sequences constructed from fen carr deposits, necessarily reflect changes in the overall composition of the community.

Modelling pollen sedimentation in Danish lakes at c.ad 1800: an attempt to validate the POLLSCAPE model

AbstractAims To validate the POLLSCAPE simulation model of pollen dispersal and deposition, and evaluate the effect of factors such as pollen productivity, wind speed and regional plant abundance, using a data set of ad 1800 pollen assemblages and historical land cover data.Location Denmark.Methods ad 1800 land cover from historical maps is digitized for 2000 m radii around 30 Danish lakes (3.5–33 ha). The simulation model POLLSCAPE is used to predict sedimentary pollen assemblages in the lakes from the plant abundance data inferred from these maps, with different model parameter settings for wind speed, pollen productivity, regional pollen loading, etc. The model predictions are compared with observed ad 1800 pollen assemblages from the lake sediment records. Furthermore, pollen productivity is estimated from the ad 1800 pollen and vegetation data using the Extended R‐value model.Results Generally the model reproduces the patterns in the observed pollen assemblages, and for most pollen types there are significant correlations between observed and predicted pollen proportions. The pollen proportions predicted by the POLLSCAPE model are sensitive to the pollen productivity estimates used, the regional background pollen loading and average wind speed. There is a difference in background pollen loading between eastern and western Denmark, especially of Calluna pollen. The fit between predicted and observed pollen assemblages is best at wind speeds around 2.5 m s−1, and decreases rapidly at lower wind speeds. The pollen productivity estimates from the ad 1800 data set are comparable with estimates from moss polsters in modern analogues of traditional cultural landscapes in Sweden and Norway.Main conclusions The POLLSCAPE model reproduces the patterns in the observed pollen assemblages from the lakes well, considering the uncertainty in the historical plant abundance data. This study indicates that the simulation model can be a useful tool for investigating relationships between vegetation and pollen composition, but also that the simulated pollen proportions are sensitive to the pollen productivity estimates, the regional background and to wind speed.

Modern pollen deposition in Lacandon forest, Chiapas, Mexico

The pollen rain from the Lacandon forest (Chiapas, Mexico) was analysed from moss polsters. The Lacandon forest is very important for its high diversity, being covered by a mix of tropical and temperate forests. One hundred sixty taxa of pollen were recorded; Moraceae, Leguminosae, Rubiaceae, Ulmaceae, Bursera and Protium were related with lowland tropical forests. Pinus, Quercus and Mimosoideae were related with montane rain forest/pine-oak forest (MRF/POF). Arboreal anemophylous pollen taxa were the most important elements in the pollen spectra, which indicates a regional signal of the vegetation.

The pollen spectra from moss polsters of Yururi Island in easternmost Hokkaido, Japan

The pollen spectra from moss polsters of Yururi Island in easternmost Hokkaido, Japan

Pollen–vegetation relationships in non-arboreal moorland taxa

Pollen assemblages from moss polsters from northwest Scotland are compared with the vegetation composition around the sampling points. Presence–absence data are used to calculate association, over- and under-representation indices for the relationship between plants and the pollen assemblages, and results are found to be broadly comparable with published results from Norway. Relevant pollen source areas for individual taxa are estimated using least-squares linear regression analysis of pollen percentage against distance-weighted plant abundance. The extended R-value model is applied to estimate the relevant pollen source area for the pollen assemblages. The results indicate relevant pollen source areas on the order of 2 m or less for Calluna vulgaris, Vaccinium type, Myrica gale, Cyperaceae, Poaceae, Potentilla type and Succisa pratensis. Changes in non-arboreal components of pollen assemblages from mire communities therefore appear to reflect predominantly local patch dynamics rather than wider landscape changes.

Detecting woodland remnants in cultural landscapes: modern pollen deposition around small woodlands in northwest Scotland

Modern pollen deposition in and around woodland fragments in a largely open cultural landscape in northwest Scotland is investigated using moss polsters and surface sediment samples from small lakes. The percentage of local arboreal (tree and shrub) pollen types declines rapidly with distance from the woodland edge, generally reaching background levels in less than 100 m. Even at the woodland edge or under the canopy, some samples contain less than 30% local arboreal pollen. The relationship between percentage of local arboreal pollen and distance from the woodland edge does not vary with changing topographic relationship between sample location and the woodland edge, but does differ according to the vegetation community outside the woodland. Comparison of pollen assemblages from 11 small lake basins (mean size 1.0 hectare) with assemblages from moss polsters collected near the lake edge suggests that the pollen source area for arboreal pollen types is in the order of 300 m for peat samples and over 400 m for lake sediments. Implications for interpretation of palaeorecords are considered.

Spatial variation of modern pollen in Oregon and southern Washington, USA

Surface sediments from 95 lakes provide information on the spatial variation of modern pollen spectra in Oregon and southern Washington. Percentages for 13 pollen types were compared within and between vegetation zones to characterize regional patterns of pollen spectra. The percentage data were also compared with climate variables to determine relationships between pollen percentages and regional climate gradients. The composition of modern pollen spectra corresponds well with the distribution of the pollen producers. Most pollen assemblages were generally dominated by Pinus, but those west of the Cascade Range were dominated by Alnus. Low percentages of Pseudotsuga/ Larix, Tsuga mertensiana, Abies, and Picea pollen coincided with local occurrence of the trees. The distributions of the pollen data were arranged along gradients of temperature and effective moisture. West of the Cascade Range, Alnus, Tsuga heterophylla, Pseudotsuga/ Larix, and Cupressaceae pollen were abundant and correlate well with moderate temperature and high effective moisture. In the shrub–steppe and woodlands east of the Cascade Range, where effective moisture is low, Artemisia, Cupressaceae, and Pinus pollen were dominant. At high elevations, Pinus, T. mertensiana, Abies, and Picea were common pollen types in areas with short growing seasons and high effective moisture. Pollen percentages collected from lake surface sediments, moss polsters, and soils were compared within a number of vegetation types to assess their similarity. The three types of sample yielded similar results for forested areas, but lake sediment samples from upper- and lower-treeline sites captured a more regional picture of the vegetation.

Holocene vegetation and land-use history in the environs of Lake Kahala, northern Estonia

Pollen profiles, based cores taken in Lake Kahala and from the adjoining mire, were used to establish general vegetation history and to reconstruct the extent and types of land-use over most of the Holocene. Modern pollen deposition was studied using moss polsters and the results were used in the interpretation of the fossil pollen data in terms of former land-use practices. The modern-day samples are from settlements, hay meadow and pasture, and overgrown pasture. Indications of human activity can be traced back to the Stone Age. At ca. 6400 cal. B.C., the first indications of possible woodland utilisation by humans are recorded. This may have involved grazing within the forests. From 4200 cal. B.C. onwards, animal husbandry with changing intensity was practised. Arable farming, involving cereals, was introduced to the area at ca. 1800 cal. B.C., but it was only at ca. 500 cal. B.C. that it assumed an important role in the farming economy.Secale cereale (rye) was introduced during the Roman Iron Age, intensive rye cultivation started at the end of the Iron Age, at ca. cal. A.D. 800. Ever increasing farming pressures triggered the formation of openalvars. Open landscape similar to that of today has persisted, with minor forest regeneration phases, since at least 500 cal. B.C.

Modern pollen assemblages from mown and grazed vegetation types in western Norway

The aim of this study was to investigate pollen–vegetation relationships on a local scale and to identify grazed and mown vegetation types by their pollen assemblages. Vegetation frequency analysis and pollen analysis of surface moss polsters from grazed and mown vegetation types in western Norway were carried out. Five one-square-metre plots within 10×10 m plots from 36 different sites were investigated using classification (TWINSPAN, DISCRIM) and ordination (correspondence analysis, canonical correspondence analysis) methods. The pollen assemblages reflected the local vegetation and there was good agreement between the classifications and ordinations of the two data sets. This agreement occurred despite both the inclusion of extralocal and regional pollen and the low representation of entomophilous species affecting the pollen assemblages. Gradients were found in the pollen assemblages relating to land-use practices as well as geographical regions. Pollen assemblages from heath were separated from the other investigated vegetation types and characterized by high percentages of Calluna. A pollen-analytical separation of mown meadows from grazed sites was identified, whereas a separation between mown meadows and grazed sites which had previously been mown was found to be unclear. Plantago lanceolata, Rumex acetosa type, Ranunculus acris type, and Asteraceae Cichorioideae were common in both mown and grazed vegetation types. Galium type, Cirsium type, and Succisa were more associated with grazing than mowing, whereas the opposite was true for e.g. Achillea type, Alchemilla, and Conopodium majus type and several pollen types were connected with the mown meadows. The modern pollen assemblages from mown and grazed vegetation types may aid in the identification of past hay meadows and pastures from local pollen diagrams.

Bryophyte Propagule Banks in a Feldmark on Subantarctic Macquarie Island

Subantarctic Macquarie Island has substantial areas of feldmark on its plateau above 200 m altitude. Samples of the substrate (5.5 cm in depth) from bare areas of feldmark contained viable propagules of bryophyte species found at adjacent and distant sites on the island. In laboratory conditions propagules of 15 bryophyte taxa germinated, allowing interpretation of reasons for bare patches in feldmark: bryophytes were successful at colonizing stable ground but when surface movement was present, burial and/or damage of propagules and young plants prevented colonization. Spherical moss polsters found in cryoturbatic areas in feldmark, however, represent a growth form that can tolerate surface movement. A conceptual model illustrating processes associated with colonization dynamics of bryophytes on feldmark terraces is presented. Ten of the 15 germinated taxa were nonlocal taxa which currently grow in plant communities at lower and hence warmer altitudes on Macquarie Island. The presence of viable propagules of these taxa provide an immediate and constant potential for dramatic vegetation change with climate change.

Bryophyte Propagule Banks in a Feldmark on Subantarctic Macquarie Island

Subantarctic Macquarie Island has substantial areas of feldmark on its plateau above 200 m altitude. Samples of the substrate (5.5 cm in depth) from bare areas of feldmark contained viable propagules of bryophyte species found at adjacent and distant sites on the island. In laboratory conditions propagules of 15 bryophyte taxa germinated, allowing interpretation of reasons for bare patches in feldmark: bryophytes were successful at colonizing stable ground but when surface movement was present, burial and/or damage of propagules and young plants prevented colonization. Spherical moss polsters found in cryoturbatic areas in feldmark, however, represent a growth form that can tolerate surface movement. A conceptual model illustrating processes associated with colonization dynamics of bryophytes on feldmark terraces is presented. Ten of the 15 germinated taxa were nonlocal taxa which currently grow in plant communities at lower and hence warmer altitudes on Macquarie Island. The presence of viable propagules of these taxa provide an immediate and constant potential for dramatic vegetation change with climate change.